migration and inequality in germany 1870-1913 (oxford historical monographs)
TRANSCRIPT
![Page 1: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/1.jpg)
![Page 2: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/2.jpg)
OXFORD HISTORICAL MONOGRAPHS
editors
j . maddicott r. j . w. evansj . harris b. ward-perkinsj . robertson r. service
p. a . slack
![Page 3: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/3.jpg)
This page intentionally left blank
![Page 4: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/4.jpg)
Migration and Inequalityin Germany1870–1913
OLIVER GRANT
CLARENDON PRESS � OXFORD
![Page 5: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/5.jpg)
3Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City NairobiNew Delhi Shanghai Taipei Toronto
With oYces in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal SingaporeSouth Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
� Oliver Grant 2005
The moral rights of the author have been assertedDatabase right Oxford University Press (maker)
First published 2005.
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriatereprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or coverand you must impose this same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britainon acid-free paper by
Biddles Ltd.King’s Lynn, Norfolk.
ISBN 0-19-927656-0
1 3 5 7 9 10 8 6 4 2
![Page 6: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/6.jpg)
PREFACE
This study has its origins in a D.Phil. thesis, ‘Internal Migration in
Germany 1870–1913’, which was motivated by the view that migration
is the central problem of an industrializing society. The initial stimulus
came from a visit to China in the spring of 1995, a visit which reawoke
my interest in the process of economic growth and its associated prob-
lems and opportunities. This combined with an older fascination with
the course of German history to produce the thesis which has now
evolved into this book.
Comparison is one of the most powerful tools at the disposal of a
historian, and one of the themes of this book is that historians should
not be timid or reticent about this. They should be prepared to make
broad comparisons across time and across continents. They should
insist that theirs is a discipline which can contribute to the study of
major issues of current concern: the causes of war and peace, economic
growth and the creation of modern industrial societies, the foundations
of democracy and the establishment of individual rights. If historians
turn their backs on these issues, then others will Wll this gap.
But, while historians can bring to these subjects the additional per-
spective which is gained by historical analysis, they should also be
aware of the researches on contemporary societies which social scientists
are making, and consider the ways that these can bring new perspec-
tives to bear on historical issues. This means confronting the methods
that social scientists bring to bear on these issues, in particular model-
building and the use of statistical analysis. Both will be used in the
chapters that follow.
The way that the thesis has evolved into this book has been
inXuenced by the experience of designing and teaching a postgraduate
course on comparative industrialization at Oxford University, entitled
‘Industrialization in Europe, North America and East Asia since 1700’.
I would like to thank Professor Avner OVer for having encouraged me
to do this. I also thank those students who have taken the course for
many interesting discussions on the problems of industrializing soci-
eties.
When the problems of Imperial Germany are set in this wider con-
text, it soon becomes apparent that, to a large extent, this was a success-
ful and progressive society, which dealt with the problems of rapid
![Page 7: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/7.jpg)
industrialization unusually well. It was in many respects the Wrst
‘Modell Deutschland’.
I owe thanks to many people who have helped directly and indirectly
over the past three years. I have beneWted from exemplary supervision
from Professors Charles Feinstein and Hartmut Pogge von Strand-
mann. I have much enjoyed our conversations. I have learnt much from
their observations and suggestions.
I had the additional good fortune to have had two excellent examin-
ers in Professors Niall Ferguson and Albrecht Ritschl. This work has
been considerably improved as a result of their comments.
Individual chapters have been read by Anthony Atkinson, Avner
OVer, Hans-Joachim Voth, Liam Brunt, James Foreman-Peck, Tom
Nicholas, and Eric Grimmer-Solem. I thank all these for their time and
attention.
I owe much to the library of the London School of Economics. This
has a large collection of German statistical materials on open access.
I am particularly grateful that the German statistical collection has
remained on open access during the rebuilding of the main library.
I would like to express my thanks to the staV of the Nuneham Cour-
tenay depository of the Bodleian Library, to Chris Newbould and John
Slater. The depository is not well suited to the arrival of visitors with
access permits, who sit in corridors and cause obstructions. Their toler-
ance was much appreciated.
I owe thanks to my wife and children, the arrival of the latter having
coincided with the beginning of this study. They have provided much
diversion and, at times, a necessary sense of proportion.
But my Wnal thanks are owed to my farm staV, to Paul Burns, Russell
Donovan, and Alan Roberts, who have taken on additional responsibil-
ities while I have been working on this study. It is beyond question
that, without their eVorts, this would not have been possible.
O.W.G.
18 June 2004
vi Preface
![Page 8: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/8.jpg)
CONTENTS
1. Imperial Germany as an Example of Industrialization
under Labour Surplus Conditions 1
2. Sources of Inequality in Rural Germany 22
3. The Pattern of Migration, 1870–1913 56
4. Migration in Germany 1870–1913: A Statistical
Analysis 97
5. Demography and Migration 124
6. Migration, Farm Size, and the Condition of the
Agricultural Labourer 181
7. Agricultural Productivity, Labour Surplus, and
Migration 215
8. Migration and Urban Labour Markets 253
9. Industrialization, Migration, and Inequality 293
10. Challenging the Kehrite View of Imperial Germany 330
References and Sources 361
Index 389
![Page 9: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/9.jpg)
This page intentionally left blank
![Page 10: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/10.jpg)
1
Imperial Germany as an Example
of Industrialization under
Labour Surplus Conditions
1. Introduction : ‘A house div ided ’—M igration
and Inequality in Imperial Germany , 1870–1913
‘A house divided cannot stand’ is Abraham Lincoln’s precis of the
issues facing the United States in the Civil War; it can also serve as a
summary of the views of a school of German historians, the Kehrite
school. The key concept of this school is the ‘primacy of internal polit-
ics’: the view that the internal weaknesses of pre-1914 Germany led to a
series of expedients which made German foreign policy a disruptive
force in pre-1914 European diplomacy, and which prevented the settle-
ment of outstanding disputes by any means short of a general European
conXict. The core of this policy was the building of a Xeet large enough
to challenge British naval hegemony.1
What were the sources of the internal weakness of Imperial Germany?
Historians have produced a list of suggested Xaws in the polity and social
structure of Imperial Germany: a series of ‘skirted decisions’ which
explain Germany’s failure to make political progress commensurate with
its economic advances.2 In this analysis, the fundamental assumption is
1 Kehr’s own work, originally written in the 1930s, is available in English, Kehr (1975)and (1977). Other important works in the Kehrite tradition include Wehler (1969) and(1985), Berghahn (1971), Stegemann (1977), and the collection in Sturmer (1976). Theexistence of a Kehrite school has been denied by Puhle (1978) and there is some disagree-ment over what the key Kehrite concept really is; for Eley (1986) it is either ‘the primacyof pre-industrial traditions’, p. 42, or the concept of Sammlungspolitik, p. 111. For thepurposes of this study, the key Kehrite concept is the ‘primacy of internal politics’:meaning by this, that the Kaiserreich suVered from internal problems, which had anadverse eVect on foreign policy to an extent which is unusual compared to other states.This last point is rarely made explicitly, but it is important if the concept is to have anygenuine force. For a simple statement of this view, see Rohl (1969), 10.
2 The phrase ‘skirted decisions’ is taken from the title of an essay in Mommsen (1995).Mommsen’s work stresses the political crisis produced by the inability of the Bismarckian
![Page 11: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/11.jpg)
that there were options which, if taken, would have alleviated the social
strains and enabled Germany to become a normal, modern, democratic
industrial state. And the implication of this is that such a state would
have been a more stable and predictable element of the pre-1914 diplo-
matic order.
These ‘skirted decisions’ were not just political and diplomatic: they
were derived from a fundamental unwillingness of the ‘pre-industrial’
elites of Imperial Germany to modernize structures and institutions
which had their roots in the traditions of an agrarian society.3 This
produced a bureaucracy recruited from a narrow and unrepresentative
social base, a political system which limited democratic control over the
executive, and a social order in which large numbers felt themselves
to be excluded or marginalized. In short, an outcome of ‘incomplete
modernization’, derived in part from the Bismarckian origins of the
Kaiserreich in a ‘revolution from above’.
Historians outside the group normally identiWed as Kehrite, or Kehr-
inspired, have accepted parts of this argument, or put it forward in
diVerent guises. The broader concept of a Sammlungspolitik, policies
designed to rally ‘loyal’ forces in support of the Imperial government,
has replaced Kehr’s narrower emphasis on the trade-oV of agricultural
protection for battleship construction. Modern studies place greater
stress on social and cultural divisions as the ultimate sources of the
weakness of the political structure: the rise of extremist ideas, the isol-
ation of important social strata, and the domination of political parties
by economic interest groups. Instead of a conscious ‘Flucht nach Vorn’,
an attempt to solve internal problems by foreign policy adventures,
more recent studies have set German foreign policy in the context of
a society where Social Darwinism had a ‘special resonance’, where
‘chauvinistic bile’ had a ‘broad appeal’, where paciWsm was ‘relatively
weak’ and ‘its opposite was correspondingly powerful’. Under these
circumstances, the ‘restless, querulous’ nature of German foreign
policy, the ‘fatalistic brinkmanship’ of the civilian leadership in the
constitution to adapt to changing economic and social conditions, and argues that ‘in July1914 the ‘‘latent crisis’’ of the German political system spilled over onto the internationalstage, setting in train a world war’, Mommsen (1995), 204. This can be regarded as avariation on the basic Kehrite theme.
3 According to Wehler (1985), 236, Bismarck, and the ‘Wilhelmine polycracy’ thatfollowed him, were determined to avoid the spread of democracy and moves towardsequal rights. The militancy of the property-owning elites, and their determination topreserve their privileges at all costs, is stressed in, for example, Puhle (1976), 341.
2 Industrialization under Labour Surplus Conditions
![Page 12: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/12.jpg)
crisis of July 1914, are seen, implicitly or explicitly, as expressions of
deeper Xaws.4
This study puts forward an alternative hypothesis: that many of the
internal stresses and strains of Imperial Germany were the inevitable
consequences of industrialization, and that there was little that German
political leaders could have done about this. Even if the political elites
had been prepared to make concessions this would not have produced
an immediate shift to a more harmonious society: the option of a ‘social
market economy’ was not available in pre-1914 circumstances. Instead,
the maturing of the industrial economy was a necessary precondition
for the attainment of social stability and political maturity.
The point that the transition from an agrarian to an industrial econ-
omy inevitably creates social and political strains is not a new one. Other
societies had also experienced these tensions. But the process of indus-
trialization in Germany was diVerent from that in the pioneer industrial-
izing countries because Germany had the advantages, and the
disadvantages, of ‘relative backwardness’, as originally pointed out by
Alexander Gerschenkron.5 Germany could make use of advanced indus-
trial technology developed elsewhere. This meant that the pace of
German industrialization was faster. It meant that the gap between
productivity in the mechanized industrial sectors and the traditional
sectors (agriculture and handcraft) was much greater. Transfers of
labour between these sectors was a major contribution to overall growth.
But the same transfers brought about social and institutional disruption.
Pre-1914 Germany was the Wrst economy to experience the tele-
scoped process of industrial development typical of those countries
which make their way to the front rank of industrial powers having
started from a position of relative backwardness. Many historians have
noted the rapidity of the process of social and economic transformation
in Imperial Germany. But the full implications of this have not been
examined.6 One important consequence is that the analysis of German
economy in the period needs to make use of concepts and tools taken
4 The quoted phrases are all from Blackbourn (1997), 426, 433, 446, 452, and 459. Thisis not a work in the Kehrite tradition as it is normally understood, but it is indicative ofthe way that Kehrite ideas have penetrated the historical mainstream. Works which dis-cuss foreign policy issues in isolation from the domestic context are rare.
5 Gerschenkron (1962); the main pioneer industrializing country was, of course, GreatBritain, but France also experienced a prolonged period of relatively slow industrialdevelopment, as emphasized by Keyder and O’Brien (1978).
6 One study which has made use of insights from development economics in theanalysis of inequality trends in nineteenth-century Germany is Dumke (1991).
Industrialization under Labour Surplus Conditions 3
![Page 13: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/13.jpg)
from the Weld of development economics. The concept of development
economics as a separate branch of economics rests on the idea that
‘relatively backward’ countries face diVerent problems and opportun-
ities to those which arise in more mature economies, or in societies
which have a more drawn out period of transition.
Much of the analysis which will be presented in this study makes use
of ideas whose origins lie in development economics: the ‘Kuznets
Curve’ relating income inequality to industrialization; the ‘Lewis Model’
of development under labour surplus conditions. Many societies have
found industrialization to be a disruptive process, and the path to matur-
ity has often been a rocky one even for those societies which have
successfully made the transition. The problems which Germany faced
have reoccurred in other contexts.
The historical debate over the pace of political progress relative to
economic development in Imperial Germany has a modern counterpart
in arguments over the role of authoritarian governments, or ‘Asian
values’, in the success of the East Asian ‘newly industrializing countries’
(or NICs). Developmental dictatorships such as the Park regime in
South Korea or the Kuomintang government in Taiwan have presided
over the initial stages of rapid industrialization, and political liberaliza-
tion has come later, when the most disruptive phase has been com-
pleted.7 It has been argued, by the World Bank amongst others, that an
‘insulated bureaucracy’ is an important component of a successful long-
term growth strategy.8 If so, then the alleged backwardness of the
German political system may have been advantageous in economic
terms: the state was able to focus on long-term goals rather than being
obliged, by the pressures of democratic politics, to resort to short-term
expedients which had damaging long-term consequences.9
There are now a number of examples of societies which have had
authoritarian governments during the early stages of industrialization,
and which have successfully managed the transition to democracy once
this phase is past. Besides the Korean and Taiwanese cases just cited,
there are other examples: Chile post-Pinochet, Spain after Franco, the
7 The role of authoritarian states in economic development in East Asia has beenexamined in a number of works; see for example Amsden (1989), Amsden (2001), andWade (1990).
8 World Bank (1994), 167.9 Many Third World countries have a problem of ‘urban bias’ as a result of policies
which were intended to improve urban conditions, but which then create privileged urbanelites.
4 Industrialization under Labour Surplus Conditions
![Page 14: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/14.jpg)
re-establishment of democracy in Brazil following military rule in the
1960s. Why were these societies able to make this transition when
Germany could not? It cannot, surely, be argued that social and polit-
ical tensions were any greater in Germany in the 1880s and 1890s than
in Chile at the time of the coup against Allende, or Spain just before the
outbreak of the civil war.
There were, undeniably, particular factors in the case of each of these
societies which helped to cause a reduction in the potential for social
and political conXict as industrialization proceeded. This study does not
propose a general theory of politics and industrial development. But
one particular factor is that the process of transition was not derailed by
any major external catastrophe. These countries did not face genuine
strategic dilemmas which in turn impinged on domestic politics. They
were not required to work out solutions to internal problems of transi-
tion at the same time as they dealt with foreign policy issues, which
were themselves moving towards a crisis for independent reasons.
This alternative view of pre-1914 Germany stresses the essential
autonomy of the foreign policy problems which beset the German gov-
ernment. In this ‘anti-Kehrite’ formulation, ‘revolutions from above’ are
commonplace; the processes of economic and political advance do not go
hand in hand but follow diVerent trajectories, so that ‘incomplete mod-
ernization’ is the normal state for an industrializing society; Imperial
Germany was not moving towards an internally generated catastrophe,
but was a society with as good a chance of achieving full economic
maturity, social modernization, and political democratization as any
other. The decisive factor in the equation, which shifted German history
onto a diVerent course, was the outbreak of war in 1914.10
2. M igration and Inequality in the
Development Process
Migration has a key role in industrialization, and it is a principal cause
of the social dislocation which accompanies the early stages of economic
growth in modern industrial societies. This will inevitably be a period
of high migration. It is the opportunity to shift labour out of low
productivity sectors such as agriculture which makes it possible to
10 Arguments about the outbreak of war are summarized in Pogge von Strandmann(1988), see also the collection edited by Schollgen (1990).
Industrialization under Labour Surplus Conditions 5
![Page 15: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/15.jpg)
achieve high rates of growth in underdeveloped economies. Indeed, one
of the deWning features which distinguishes ‘mature’ from ‘developing’
economies is that, typically, economic growth in the latter is accompan-
ied by large-scale migration from the countryside into cities, while in
the former it is not.
One view which provides a simple analysis of the eVects of migration
out of agriculture is the model of economic development Wrst set out byArthur Lewis in 1955: ‘a model of development with unlimited supplies
of labour’, generally known to development economists as the Lewis
Model.11 In his article, Lewis analysed the eVects of industrial develop-
ment in the context of a labour market where there is surplus labour, as
commonly found in countries in the early stages of industrialization.
The model predicts that, so long as capitalists can draw on these
supplies of surplus labour, then the fruits of industrial progress will go
largely to the holders of capital, in the form of rising proWts, and there
will be little increase in the level of real wages. Given that the owners of
capital are likely to be predominantly rich, and few in number, it can be
seen that this prediction will lead to rising inequality in cities and other
industrial areas, with a consequent increase in social divisions, and, also,
to a feeling amongst industrial workers that they are not gaining a fair
reward for their labours.
For students of German history the attraction of the Lewis Model is
that it provides a possible explanation both of the rapid growth of the
German economy 1870–1914 and of the tensions that accompanied this.
Moreover it also suggests a reason for one of these problems: the rise of
the socialist vote. Lewis named the two phases he described ‘classical’
(when there is a labour surplus and wages are held down) and ‘neo-
classical’ (when the labour surplus is exhausted and wages start to rise).
But one might note the similarity between his model and that of Karl
Marx and call the Wrst phase ‘Marxist’ or ‘quasi-Marxist’. According to
the model, this is a period when the capitalist economy behaves in a
way which seems to bear out some of Marx’s predictions: proWts rise as
a percentage of the national income; wages lag behind productivity
increases.12
11 Lewis (1954). For a modern discussion see Hayami (1997), 72–5 and 170–2.12 Lewis did not assume, as Marx did, that technology was always labour-saving. In a
purely Marxist model capital accumulation leads to increased competition, a fall in therate of proWt, and increasingly severe economic crises, so the ultimate level of the share ofproWts in national income is not clearly determined. However, Kolakowski has pointed outthat Marx’s views on whether capitalism led to the absolute or the relative impoverish-ment of the working class are also ‘by no means unequivocal’, Kolakowski (1978), i. 289.
6 Industrialization under Labour Surplus Conditions
![Page 16: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/16.jpg)
If the economy behaved as Lewis predicted, then the experience of
workers competing in urban labour markets with apparently unlimited
supplies of migrants from rural areas would be a dispiriting one. The
formation of trade unions could lead to an improvement in workers’
bargaining strength, but unions would have to be cautious about press-
ing claims for higher wages given the ability of employers to bring in
large numbers of migrants with no previous experience of trade union
activity. Trade unions would be more eVective where the workers they
represented possessed skills which could not easily be taught to newly
arrived migrants, as was the case for printers and other members of the
so-called ‘labour aristocracy’. These would be easier to organize.
In this context, the appeal of radical Marxist ideas would be under-
standable, together with a pessimistic view of the likely beneWts from
non-revolutionary trade union activity. But the diVerence between
Lewis’s model and Marx’s is that Lewis’s labour surplus phrase comes
to an end eventually. When this happens, urban labour markets are no
longer dominated by the Xow of surplus labour out of agriculture, and
employers who need to attract labour to new industries are obliged to
oVer superior wage levels to those obtained by fully employed workers
elsewhere. Trade unions can bargain from a position of strength and
union membership can oVer beneWts to unskilled as well as skilled
workers. The economy enters a new phase in which wages rise in line
with productivity and the proWt share stabilizes or may even decline.
The operation of the Lewis Model is illustrated in Figure 1.1. The
introduction of new technology causes a rise in productivity in
the advanced or industrial sectors. But with unlimited labour supplies,
the wage level is Wxed. Consequently, the gains from productivity in-
creases go exclusively to the owners of capital, producing a rise in the
proWts share. It is only when the labour surplus is exhausted—a
moment sometimes called the ‘turning point’ in the development eco-
nomics literature—that wages can rise above the Xoor set by the ready
availability of surplus labour in the traditional or rural sectors.13 Once
this has occurred the economy shifts to a new phase of development,
represented by the neo-classical model of mainstream economics, where
there are limitations on the supply of labour as well as capital. In this
phase, both capital and labour share in the fruits of economic develop-
ment. In mature capitalist economies real wages have tended to rise at
least as fast as the growth of productivity, if not somewhat faster. This
13 The term ‘turning point’ was coined by Fei and Ranis (1964).
Industrialization under Labour Surplus Conditions 7
![Page 17: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/17.jpg)
has meant that the share of wages in national output has been stable or
tending to increase.14
The implication of the Lewis Model is that this historical moment, the
‘turning point’ as the economy moves out of the labour surplus phase,
represents a new opportunity to achieve political and social stabilization.
There are now reasons for workers to favour ‘normal’ trade union activ-
ity, aiming to secure steady improvements in wages and other working
conditions, over more radical alternatives. ‘Revisionist’ views should gain
ground. Provided that the political system oVers a reasonable chance of
similar progress, leading to increased participation in political decision-
making, and that the whole system is not destabilized by some major
catastrophe, a society which has passed the ‘turning point’ of the Lewis
Model has a much better prospect of achieving full political, social, and
economic maturity—becoming a ‘normal’ modern, liberal, capitalist
democracy—than one which is still in the labour surplus phase.
The Lewis Model in this pure form has a clear and abrupt transition
from the labour surplus to the neo-classical model. In practice,
economic development does not divide quite so easily into phases.
14 The Wgure shows the eVect on factor shares if wages rise in line with productivity.This is not an automatic consequence of the transition to a ‘neo-classical’ economy,although it is broadly consistent with the experience of most OECD countries in thetwentieth century.
Turning point
Productivity
Profits
Labour surplus model Neo-classical model
Wages
Fig. 1.1. The pure Lewis Model: changing factor shares
during industrialization
8 Industrialization under Labour Surplus Conditions
![Page 18: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/18.jpg)
Moreover, there is generally some rise in wages during the early stages
of industrialization, even if the rise is not as fast as the increase in
productivity. Initially, industry can rely on nearby sources of labour.
But once this ‘local labour surplus’ is exhausted, wages have to be
raised to attract migrants from regions which are more remote from the
main industrial areas. This period of slow wage increases, which lag
behind rising productivity, can last until all sources of surplus labour
are used up. Then wages start to rise in line with productivity.
Figure 1.2 illustrates this ‘modiWed Lewis Model’. In the early stages
of industrialization the beneWts still go predominantly to the owners of
capital, even though there is some increase in wages. The transition to
the neo-classical model is a more gradual one. Instead of a sudden
‘turning point’ there is a slower transition to the neo-classical model.
What evidence is there that wages lagged behind productivity? The
actual record of German wage and productivity growth is set out in
Figure 1.3. There are some caveats to note. The Wgures used for these
calculations are taken from Walter HoVmann’s Das Wachstum der
deutschen Wirtschaft. The wage series used is derived from a combin-
ation of diVerent sources, and this means that it must be treated with
some caution.15
Productivity
Profits
Wages
Labour surplus model Neo-classical model
Fig. 1.2. The modified Lewis Model: changing factor
shares during industrialization
15 More reliable wage data becomes available from the 1870s and 1880s onwards, Bry(1960), Desai (1968), and Hohls (1995). However, the onset of German industrialization is
Industrialization under Labour Surplus Conditions 9
![Page 19: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/19.jpg)
The second caveat is that wages may lag behind productivity for
other reasons. If the terms of trade move against the industrial sector,
then factor returns will not keep pace with productivity.16 In the period
after 1895, the German export sector had to gain market share from
Britain, and this led to downwards pressure on industrial prices at a
time when world agricultural prices were recovering from the low point
generally dated to the 1850s, and Figure 1.3 shows that the gap between the two seriesopened mainly in the 1860s and early 1870s. It would also be interesting to compare realwages and the return on capital in the industrial sector, but HoVmann’s proWt Wgures havecome in for heavy criticism, Fremdling (1991).
16 The diVerence is that, with adverse movements of the terms of trade, factor returnsin the industrial sector are depressed in general, so both proWts and wages are held down,but under the Lewis Model only wages are held back.
1850
80
90
100
200
300
400
1856 1862 1868 1874 1880 1886 1892 1898 1904 1910
Real wages
NVA/head
Fig. 1.3. Indices of productivity and real wages in the advanced sectors of the
German economy (industry, mining, and transport), 1850–1913
Log scale; base is average for the years 1850–9 (¼100)
Source: Calculated from Hoffmann (1965). The series are averages for the three sectorscalculated from Hoffmann’s figures for employment, net value added, and wages. Thereal wage figures are produced using Hoffmann’s deflator for private consumption.
10 Industrialization under Labour Surplus Conditions
![Page 20: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/20.jpg)
of the mid-1890s. This shift in the terms of trade meant that the im-
pressive productivity performance of the German industrial sector in
the years just before the First World War was not fully reXected in real
wage gains.
Figure 1.3 shows that, according to HoVmann’s Wgures, real wages
did not keep pace with productivity in the modern sectors. Between
1850/9 and 1904/13 productivity rose by 202%; real wages rose by
119%. The rate of increase of productivity was 70% higher than the
rate of increase of real wages. The Wgure shows that the gap opened
substantially in the 1860s and early 1870s. It then closed in the 1880s
before widening again in the 1890s and 1900s.
The fact that so much of the widening of the gap occurred in the
1860s, which is a period for which there are few reliable wage statistics,
means that these results need to be interpreted with caution. What can
be said is that these are Wgures which are consistent with the modiWedversion of the Lewis Model as shown in Figure 1.2. Wages did not keep
up with productivity; most of the beneWts of industrialization went to
the owners of capital.
3. Towards a More General Theory of Inequality
and Industrialization :
The Weber–Lewis–Kuznets Hypothes is
The Lewis Model is not a complete model of inequality and industri-
alization: it lacks several important components. It does not explain how
the labour surplus is created, and how this phase comes to an end. It
provides a possible explanation of rising urban inequality, but to explain
the general movement of inequality in an industrializing society it is
necessary to widen the scope of the enquiry to include consideration of
rural inequality and the relationship between the two sectors.
Lewis’s view of the labour surplus is essentially a static one: less-
developed countries are overpopulated relative to economic opportun-
ities; there is underemployment both in agriculture and in an informal
urban sector.17 His views are, in some respects, very similar to those of
Max Weber, writing about conditions in east-Elbian agriculture in the
17 In societies which do not have comprehensive unemployment insurance there israrely much formal unemployment. Labour surplus conditions are more likely to Wndexpression in high levels of underemployment.
Industrialization under Labour Surplus Conditions 11
![Page 21: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/21.jpg)
late nineteenth century. Lewis wrote: ‘in over populated countries the
code of ethical behaviour so shapes itself that it becomes good form for
each person to oVer as much employment as he can’.18 Weber held that
‘the old economic order asks: ‘‘how can I give, on this piece of land,
work and sustenance to the greatest number’’ ’.19 Both these statements
describe a system of values which gives rise to behaviour which departs
from the proWt-maximizing norm.
Lewis did not try to explain why this changed, but Weber did. The
most important factor was the changing attitudes and values of the
land-owners. These were obliged by their desire to maintain their pos-
ition in society to match the consumption patterns of the rising urban
bourgeoisie, and this in turn led to the adoption of entrepreneurial,
proWt-maximizing behaviour.
Under capitalism, entrepreneurial land-owners would employ the
cheapest labour, use the best technology, and prefer market relation-
ships to paternalistic systems of mutual obligation. Thus, while the ‘old
economic order’ provided employment for surplus labour, this surplus
was released when capitalist values came to predominate.
In Weber’s view, therefore, it was the permeation of capitalist values
throughout society which led to the creation and release of the rural
labour surplus. This was an inevitable consequence of industrialization.
While Weber’s views on the exact mechanism which brings about
social change in rural societies during industrialization can be disputed,
his basic point that there are institutional arrangements in traditional
agriculture whose erosion during industrialization can create a period of
heavy migration into urban labour markets is more in tune with modern
analysis. This could arise for a number of reasons. Traditional marriage
patterns may have held back demographic increase. There may be insti-
tutionalized underemployment in agriculture, supported by systems
favouring revenue sharing or informal arrangements for mutual insur-
ance in hard times. Institutions which were appropriate to an isolated
agricultural sector with limited access to urban markets may hold back
the use of new technology: improved transportation may bring modern
inputs and a consequent increase in productivity, making it possible to
18 Lewis (1954), 142.19 The quotation is from a lecture Weber gave in the United States, Weber (1948), 367.
Weber’s major work on east-Elbian agriculture formed part of a survey by the Verein furSozialpolitik, Weber (1893). Other writings on this subject include an article whichappeared in 1894, Weber (1894, trans. 1979) and various speeches and addresses, whichcan be found in Weber (1993).
12 Industrialization under Labour Surplus Conditions
![Page 22: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/22.jpg)
shift large quantities of labour into the industrial sector while maintain-
ing or increasing food production.20
A more general view of the relationship between industrialization and
inequality in all sectors of society was provided by Simon Kuznets,
writing, like Lewis, in the 1950s.21 He was inXuenced by the work of
the Soviet economist Prokopovitch, who had used nineteenth-century
Prussian tax data to show a rise in inequality during industrialization.
As a result of Prokopovitch’s work, Kuznets was also aware of a gap
between rural and urban inequality in nineteenth-century Prussia. He
found a similar pattern in studies of inequality in the United States and
India.22 This led him to a general statement of the relationship between
inequality and economic development.
His hypothesis was that the shifting balance between a relatively
egalitarian, but poor, rural sector, and a relatively unequal, but richer,
urban sector would produce a curve, an ‘inverted U’, as inequality Wrstrises with industrialization, and then declines (Figure 1.4).
Migration has an important role in the mechanism which, according
to Kuznets, produced his curve. It was migration between the two
sectors which produced the shifting sectoral balance. Moreover, this
eVect would be intensiWed if the initial eVect of industrialization was to
raise the size of the rural–urban wage gap. Modern studies have shown
that the movement of the rural–urban gap is an important factor in the
overall income distribution: large gaps between rural and urban earn-
ings are associated with high levels of overall inequality.23 Poorer rural
workers and peasants account for a disproportionate number of the total
numbers on low incomes even in advanced countries, so a worsening of
the relative position of agriculture will tend to worsen the overall distri-
bution of income.
This factor in itself could explain part of the increase in inequality
predicted by Kuznets. If modern industrial centres are to grow rapidly
they will need to attract labour from rural areas. Initially this may come
from nearby regions, where there may be surplus labour. But, when this
surplus is exhausted, sources of labour must be found in further-oV
20 The view that many of the traditional institutions of agriculture in underdevelopedcountries are eVective solutions to the problems of remoteness, uncertain yields, andlimited access to markets is put forward in Dasgupta (1993).
21 Kuznets (1955).22 Ibid. 7; the German Wgures came from Prokopovitch (1926).23 Bourguignon and Morrison (1998), Bourgignon (1990); Basu (1997) discusses the
theory behind this.
Industrialization under Labour Surplus Conditions 13
![Page 23: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/23.jpg)
regions. This may require a substantial increase in the gap between
rural and urban earnings to create the necessary incentives for would-be
migrants.
Although Lewis and Kuznets had diVerent mechanisms in mind,
there are similarities between their views. The downward phase of the
Kuznets Curve is brought about by two factors: the declining import-
ance of the poorer rural sector in the overall distribution of income
(which reduces the importance of the gap between rural and urban
incomes as a contribution to general inequality) and a tendency for
inequality within the urban sector to start high and fall over time. This
last point can also be derived from the Lewis Model.
Putting the views of these three authors together produces a general
theory of the relationship between economic development and inequal-
ity: industrialization and the spread of modern values and relationships
unleashes a process of social change in agriculture, which in turn leads
to the transfer of surplus labour to the urban sector; labour surplus
conditions produce an unequal distribution of income in the urban
sector; the increased importance of the unequal urban sector leads to a
general rise in inequality; the ending of the labour surplus period
creates conditions which will eventually lead to a fall in inequality.
Inequality
Gross Domestic Product per capita
Fig. 1.4. The Kuznets Curve: an inverted U-shaped relationship between
inequality and economic growth during the early stages of industrialization
14 Industrialization under Labour Surplus Conditions
![Page 24: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/24.jpg)
4. Crit ic ism of the Lewis Model and Alternative V i ews
The Lewis Model has been widely used in development economics, and
it is still one of the mainstays of this branch of economic analysis, but it
has had its critics.
1. In the development economics literature the aspect of the Lewis
Model which has attracted most criticism is the view that agriculture
has only a passive role in economic development, and does not progress
technologically in the same way as the urban or industrial sector.24 One
of the most serious charges against the Lewis Model is that it has
contributed to the tendency of modern developing countries to under-
invest in agriculture.
2. Lewis makes a restrictive assumption about savings: ‘only capital-
ists save’. While this assumption may be justiWed for some developing
countries where banks and other savings institutions are not well estab-
lished, the experience of other industrializing countries shows that,
given the right institutional framework, even people who are not par-
ticularly well oV will save, and that these savings can be an important
component of total funds available for investment. A more realistic
assumption might be that ‘only capitalists invest in industry’: meaning
that, in the early stages of industrialization, only a minority have the
specialized knowledge of the new technology to be able to take advan-
tage of the opportunities it oVers. This, in turn, points to the import-
ance of institutional development, in making it possible for the savings
of those who lack technical knowledge to be channelled into industrial
investments: the German banking system performed such a function.
3. Lewis has been accused of overemphasizing the importance of
agriculture as a source of surplus labour. Surveys of modern developing
economies show that the main area of underemployment is in fact the
informal urban sector. Whether this also applies to nineteenth-century
Germany is another matter. The phenomenon of high rates of urbaniza-
tion, in countries which are still in the early stages of industrialization,
is one which is found in post-1945 developing countries and not in
nineteenth-century Europe. But an informal urban sector did exist:
Tagelohner, workers without permanent employment dependent on
occasional day-labouring jobs, were a signiWcant part of the urban
labour market.
24 Industrialization and urban growth produces change in agriculture in nearby regions,as was pointed out by Von Thunen (1826).
Industrialization under Labour Surplus Conditions 15
![Page 25: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/25.jpg)
4. In addition, there has been criticism of the concept of the labour
surplus itself. When, in 1979, Lewis was awarded the Nobel prize for
economics, it was shared with the Chicago agricultural economist Theo-
dore Schultz. Both had made a range of important contributions to
development economics, but on this subject they were in fundamental
disagreement: Schultz denied the existence of a labour surplus in trad-
itional agriculture, certainly in Lewis’s sense of a large body of labour
with zero marginal productivity.25
The essence of Schultz’s critique was, Wrst, to deny that resource allo-
cation was ineYcient in traditional agriculture, or that there was a prefer-
ence for idleness or a bias against thrift. From this it followed that labour
would be fully occupied to the point where marginal cost was equal to
marginal product. Secondly, he made use of empirical studies to show
that underemployment was not prevalent, and also demonstrated, using
the eVects of the 1918–19 inXuenza epidemic in India, that an exogenous
removal of agricultural labour did aVect food production.26
In Schultz’s view peasants were ‘poor but eYcient’.27 If investment
was low it was because the return on capital was also low, or the
discount rate was high. It was expensive for small subsistence farmers
to search for innovations; this, rather than resistance to change,
explained low technical progress. Peasant ineYciency was a myth im-
posed by Western agricultural experts who failed to understand the
seasonal nature of much work in traditional agriculture.
Although Lewis may well be the better known, or at least the more
cited, author, Schultz’s critique has been highly inXuential. The view
that peasants Wnd eYcient solutions to the problems they face is now a
commonplace. Moreover, Schultz’s stress on the importance of the
introduction of new factors—new productive knowledge, better educa-
tion, human capital in the broadest sense—appears prescient in the
light of subsequent developments in growth theory.
Nineteenth-century German agriculture oVers an excellent example
of the introduction of new, non-traditional factors exactly as described
by Schultz: new scientiWc knowledge, heavy investments in education,
and specialist training in agricultural sciences.
But before setting up Lewis and Schultz as polar opposites it should
be noted that in some respects they are not so far apart. Schultz thought
25 Possible deWnitions of the labour surplus are discussed in Reynolds (1975), 11.26 Major wars have also led to the sudden removal of large numbers from the labour
market; Prest (1948) gives some interesting examples of cases where this led to a fall ofproduction in less-developed economies, and some others where it did not.
27 Schultz (1964), 40.
16 Industrialization under Labour Surplus Conditions
![Page 26: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/26.jpg)
that the introduction of new factors could produce a relatively ‘cheap’
rise in productivity: the transformation of traditional agriculture could
be quite rapid. Now there is a clear distinction between Lewis’s view of
a static rural labour surplus, which pre-dates industrialization, and
Schultz’s more dynamic version, but the practical implications are simi-
lar: quite large amounts of labour can be released from agriculture while
food production is maintained or increased.28
Moreover, it is worth considering what the source of these new
inputs might be. Government investments in agricultural science and
education in nineteenth-century Germany were not the whole story.
Industrialization brought new agricultural inputs: basic slag from steel-
making; ammonium sulphate from gas production. Industrial progress
stimulated scientiWc research as well as beneWting from it, and this
created possibilities for agricultural advances in turn. Industrialization
raised the demand for agricultural products, some of which, such as
sugar beet, also required improved technical knowledge. Industrial
growth led to improved communications, bringing to rural society
better access to markets and to productive knowledge. Institutional
developments, such as improved banking systems, also aVected agricul-
ture. So, the transformation of traditional agriculture was, at least in
part, a consequence of industrialization.
Although Schultz and Weber came from diVerent traditions and
were writing in diVerent centuries, there is a similarity in their views in
that the creation of an agricultural labour surplus is a dynamic process,
which comes about as a result of industrialization and the eVect that
this has on agriculture. Weber’s primary stress was on the changes in
social attitudes but he also considered the eVect of changing technology,
in particular the introduction of sugar beet. This increased the seasonal-
ity of agricultural work, leading to the employment of seasonal migrant
labour, which in turn brought about a weakening of social ties.
5. The Lewis Model and the H i storiography
of Imperial Germany
Historical studies are mostly bounded in scope and in time, and this
one is no exception: July 1914 makes a convenient date for the ending
28 Ibid. 24–5. This eVect is greatly strengthened if, as is often the case, industrializationis accompanied by increased population growth, an issue which will be examined inChapter 5.
Industrialization under Labour Surplus Conditions 17
![Page 27: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/27.jpg)
of a study of conditions inside Imperial Germany. The year 1913 is the
last one for which a full set of data on peacetime conditions is available.
But the temptation to see the events of 1914–18, not just the war but
the revolutionary outbreaks which accompanied its ending, as the cul-
mination of forces and developments already under way within the
social structure of the Kaiserreich needs to be resisted. The sense that
the narrative structure needs a natural full-stop, not an accidental one,
pervades much that has been written about Imperial Germany. The
massive disruption of German political and social development which
followed is felt to need deep-seated explanations, not just the impact of
chance events. Hence the appeal of Sonderweg arguments which trace
the evolution of Xaws rooted in the character of the Bismarckian state
and the social or cultural background to its political travails.
The current consensus is not one to accept the Kehrite view of the
absolute primacy of internal politics in foreign policy without reserva-
tion. Instead the view which prevails is one which stresses the inter-
action of external and internal politics: the way that interest group
politics hemmed and hampered policy-making in both Welds; the con-
stricting eVect of unresolved constitutional problems on government
action; the eVect of a drift to radicalism and chauvinism on the tenor of
debate and the ability of political actors to see issues as resolvable, given
goodwill and a willingness to engage in rational analysis, rather than
unbridgeable divisions, calling for recourse to apocalyptical solutions.
The predominant view sees the intertwined problems of the Kaiserreich
as a Gordian knot which, by implication, only a radical discontinuity
could solve. Historians who take a positive and optimistic view, who
stress the real prospects for peaceful evolution without violent confron-
tation, have had a poor reception from their colleagues.29
The value of economic models to historians lies in their ability to
draw together diVerent pieces of evidence and present a coherent pic-
ture of social or economic evolution. The common features of the ex-
perience of many societies engaged in the process of industrialization
has naturally excited the interest of theorists and led to the construction
of theories which seek to explain these consistencies. While some his-
torians may distrust these attempts to impose a common structure on
historical events, the alternative approach, a history without theory,
leads to the particularization of every historical event, and an undue
29 See, for example, the reaction to Rauh (1973) and (1977).
18 Industrialization under Labour Surplus Conditions
![Page 28: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/28.jpg)
stress on the unique features of every society, ignoring many points of
similarity.
Migration is a particularly good example of this tendency to see
events without considering the broader context. It is a subject which
has rightly stimulated historical analysis. The movement of individuals
or families from one social and economic context to a diVerent one is a
dramatic historical event which brings about the disruption of trad-
itional relations, and leads to the creation of new ones. But it also forms
part of a larger picture, playing a role in the determination of wage
levels and other factor returns in the economy as a whole. This in turn
has an eVect on attitudes and motivations at the level of the individual.
For a society in the upwards phase of the Kuznets Curve, a drift to
radicalism is not inevitable, but the context is one which is, at the very
least, conducive to alienation and a sense of social exclusion. The diver-
gent paths of factor returns predicted by the Lewis Model, when trans-
lated into the actual experiences of workers exposed to the movements
of urban labour markets under labour surplus conditions, meant that a
generation of workers had poor living conditions, low life expectancy,
and insecure jobs and saw only slow improvements in these circum-
stances, despite rapid economic growth.
But these were not German peculiarities; they were the result of
general forces unleashed by the introduction of modern technology.
And as Germany began to climb out of the trap created by the labour
surplus phase of economic development, conditions in German cities
did improve, albeit slowly. This in turn created a more favourable
background for social reconciliation and political progress. The Lewis
Model suggests that the ending of the labour surplus phase may well
represent an opportunity for a wider process of political and social
modernization.
6. Standards of Comparison for the H i storiography
of the Kaiserreich
A major concern of historians has been the extent to which Imperial
Germany was an unusual state or society in the context of late nine-
teenth-century Europe. The view that there was something exceptional
about Germany takes diVerent forms: Wolfgang Mommsen has placed
the emphasis on Xaws in the political system; for Hans-Ulrich Wehler it
is the Janusgesicht of incomplete modernization, which raises broader
Industrialization under Labour Surplus Conditions 19
![Page 29: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/29.jpg)
questions of social and economic development. But one feature of these
studies is that the standard of comparison is taken from the experience
of the states of Western Europe: Wehler has written about choices
between ‘Modernisierung und Tradition . . . okzidentalen Gemeinsam-
keiten und deutschen Sonderkeiten’.30
The implication of the Weber–Lewis–Kuznets hypothesis is that this
comparison may not be appropriate. In the pioneer industrial states the
process of industrialization was more drawn out, the erosion of pre-
industrial, or pre-capitalist, values and institutions was slower.31 There
may have been a labour surplus period, when labour supplies were
abundant due to population growth or changes within agriculture, but
this was less acute and spread over a longer period. The growth of cities
was not so rapid. Patterns of inequality changed slowly, if at all. Polit-
ical and social institutions could evolve in situations where there was
less pressure for sudden change.32
In the British case, the existence of labour surplus conditions has
been detected in the operation of the labour market in the period up to
1850.33 This was the result of a relatively drawn-out process: compara-
tively slow industrial growth combined with demographic pressure and
the Wnal stages of the transition to large-scale agriculture which had
started in the sixteenth century. But the German experience was a
diVerent one. Although there had been pockets of proto-industrial ac-
tivity in the eighteenth century, the onset of rapid economic growth
was delayed until the ‘triple uniWcation’ of the mid-nineteenth century:
economic uniWcation with the creation of the Zollverein in 1834; polit-
ical and monetary uniWcation with the founding of the German Empire
in 1871; and, crucially for a country with limited access to sea transport,
the creation of a uniWed market as a result of the expansion of the
railway network in the 1850s and 1860s.
30 Modernisation or Tradition . . . Convergence to Western Norms or German Exceptional-ism, Wehler (1995), 1252; the use of the adjective okzidentalen obscures which Westernstate is being used as the standard of comparison. Since the strictures of Blackbourn andEley (1984) appeared, German historians of the Sonderweg persuasion have been lessspeciWc on this point, although it is often Britain that is the implied model.
31 In the British case, the switch to large-scale capitalist agriculture came well beforethe onset of industrialization.
32 The main focus of comparison is obviously Britain, but Belgium and the Nether-lands were also advanced industrial or commercial societies by 1830. Recent scholarshiphas also shown that the process of French industrial growth was a relatively drawn-outone.
33 As argued in Feinstein (1998), 651.
20 Industrialization under Labour Surplus Conditions
![Page 30: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/30.jpg)
German economic development was, like that other of Continental
countries, ‘jerked’ forwards by the application of British techology.34 In
the German case, the jerk was made even more abrupt by the overcom-
ing of barriers to internal market integration which had hitherto held
back economic development. The result was that German industrializa-
tion from 1850 onwards was unusually fast. The process was telescoped
in a way which was new in the experience of other European countries.
This created problems of adjustment and adaptation which were unpre-
cedented in their intensity.
If so, then the standard of comparison for Imperial Germany should
not be taken from Western Europe but, in the Wrst instance, from other
late-industrializing European countries—Russia, Italy, and Spain, for
example—and secondly, from other non-European countries which
have gone through the process of industrialization. There is a large
range of scholarship which has been devoted to the comparative study
of developing economies, and the diVerent social, economic, and polit-
ical problems associated with or resulting from the development pro-
cess. It is the basic contention of this book that the tools developed
within this Weld, the Weld of development studies, can be used to under-
stand and elucidate the problems of Imperial Germany. This will, it is
hoped, shed a new light on the diYculties, and the achievements, of the
Kaiserreich.
34 Pollard (1981), 84. The transforming nature of the new British industrial technologyis clear from the regional studies presented in chapter 3 of Pollard’s study.
Industrialization under Labour Surplus Conditions 21
![Page 31: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/31.jpg)
2
Sources of Inequality in Rural
Germany
1. Introduction
The Kuznets Curve—the hypothesis that inequality will initially rise
and then fall with industrialization—is predicated on the basis of a
series of assumptions made by Kuznets in his original article. One of
these is the assertion that, initially, inequality will be higher in the
urban than the rural sector.
However, consideration of the historical record, as well as the current
position in developing countries, shows that there are a wide range of
possible levels of inequality in the rural sector in underdeveloped or
newly industrializing countries. Factors such as climate, crop specializa-
tion, the eYcacy of land reform, the legacy of colonialism or feudalism
can produce very diVerent outcomes. The contemporary situation
shows wide divergences. On the one hand there are the ‘plantation
economies’ of central and southern America, which typically have large
inequalities in land ownership and rural incomes; at the other extreme,
the rural sectors of many newly industrializing East Asian countries are
characterized by small-scale intensive rice production, with a conse-
quent narrowing of disparities in income and wealth.1
Thus, a study of the evolution of inequality in an industrializing
society needs to start with a consideration of the state of rural society:
the allocation of property rights, the distribution of land holding and
land ownership, the position of rural labourers. Even if the predictions
of the Lewis Model concerning conditions in urban labour markets are
correct, and this leads to high inequality in the urban sector, there
are some societies in which the level of rural inequality is so high that
industrialization produces little or no increase in the overall level of
inequality.
1 Works which have stressed the importance of such geographical factors include Bray(1986) and Engerman and SokoloV (1997).
![Page 32: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/32.jpg)
2. The Agrarian Reform Movement in
N ineteenth-Century Germany :
General Considerations
Historical analysis of the nineteenth-century agrarian reforms has
tended to look for statistical indicators to judge the impact of the
reforms: Wgures for peasant (or smallholder) land losses, information on
the numbers of landless labourers and their incomes, the extent of
peasant indebtedness, the overall structure of land holding before and
after the reforms, the balance between rented land and land in owner-
occupation. These are important issues, and it is right that they should
be the main focus of research, but the use of statistical indicators is
meaningless without ‘benchmarks’, standards by which to judge the
diVerence between success and failure.
Much of the modern historical analysis of the agrarian reforms has
been inXuenced by the ‘incomplete’ or ‘absent’ revolution thesis of the
Sonderweg school of German historians: the view that the deWning char-
acteristic of German history has been a succession of ‘revolutions from
above’, which necessarily lacked the capacity to achieve thoroughgoing
social transformations of the type supposedly achieved in other coun-
tries which experienced ‘revolutions from below’. Implicitly, therefore,
the standards of comparison for the agrarian reforms are the revolution-
ary land settlements brought about by the French or Russian revolu-
tions (referring in the latter case to the pre-collectivization period).
These are possible points of comparison, but not the only ones.
There is now a large number of studies in the Weld of development
economics on land reform and its problems. This has a number of
lessons for historians. Perhaps the most important of these is the sheer
diYculty of land reform: it is quite rare to Wnd examples of land reform
programmes which have been even moderately successful. Agrarian
reforms have too often been either ineVective or else led to results
which were inimical to the development process.2
Land reform is a social and political process which places an unusual
strain on the capacity of the political system to deliver an unbiased
outcome: one which is not inXuenced either by the social composition
of the bureaucracy or by the ability of special interest groups to aVect
political decisions. It is a test of the ‘encompassing’ nature of the polit-
ical process: the extent to which decision-makers put the long-term
2 Modern works that make this point include Low (1996) and Ghimure (2001).
Sources of Inequality in Rural Germany 23
![Page 33: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/33.jpg)
interests of society, considered as a whole, ahead of considerations of
short-term expediency (which militate against oVending powerful social
groups) or the favouring of sectional interests.3
Moreover, it is a process which involves balancing a number of
diVerent objectives. The history of land reform shows that there are
many pitfalls. These include the following.
(i) The eVect of expropriation on property rights
Radical land reform without compensation can achieve a dramatic alter-
ation of rural property rights. However, expropriation on such a scale
will have adverse eVects on investment in other sectors: investors will
wonder if similar violations of property rights are possible in other
areas. This is one consideration which has led most governments to pay
compensation to landlords and others adversely aVected by land
reforms. The record of recent years shows that some compensation has
nearly always been paid: the main exceptions being land reforms
following the establishment of Communist governments which have
been coupled with more general restrictions on private property rights.
In the German case, compensation was paid by the peasants them-
selves, either in the form of monetary compensation or land given up to
landlords, in order to secure the elimination of labour services and
other onerous duties or restrictions. Arguably, the various German
states could have provided more of these payments themselves, but the
limited size of state budgets in the early nineteenth century probably
precluded this option.
(ii) The eVect on production
The objectives of land reform are Wrst to reduce rural inequality, and
secondly to promote increased agricultural productivity, the use of new
technology, and rising food production. Otherwise, inadequate agricul-
tural production may act as a brake on industrial development, produ-
cing problems such as the ‘scissors crisis’ of the Soviet economy in the
NEP period.4 Land reform should not create a rural sector which is an
obstacle to general economic development. It should not consist entirely
of subsistence farmers unable to produce a marketable surplus.
3 The concepts of ‘encompassing’ and ‘sectional’ interest groups are taken from Olson(1982).
4 Nove (1969), 93–6.
24 Sources of Inequality in Rural Germany
![Page 34: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/34.jpg)
The impressive productivity performance of the German agricultural
sector in the second half of the nineteenth century showed that the
reforms produced an eYcient market-orientated sector, which was well
able to take advantage of the opportunities presented by expanding
urban demand.
(iii) The importance of rural credit
A major theme which connects to this last point is the availability of
rural credit. In cases where land reform has had an adverse eVect on
production, the lack of credit has often been an important factor. More-
over, if peasants are obliged to pay compensation to former landlords,
then some means of Wnance will be required for this. In cases where
rural Wnance has been unobtainable or prohibitively expensive, there has
often been a return to landlord control, as indebted peasants have been
forced to sell land to local money-lenders or back to former landlords.5
This is probably the area where the Prussian reforms, in particular,
were most deWcient. In the early nineteenth century, state-sponsored
rural credit institutions, such as the Landschaften or rural mortgage
banks, lent mainly to larger estates. Smaller farmers had limited access
to credit, and this contributed to the extent of peasant land losses. By
contrast, in areas such as the Kingdom of Saxony, where more eVort
was made to give small and medium-sized farmers access to credit, the
post-reform land distribution was more egalitarian.
(iv) The importance of market access
As the objective is to create an agricultural sector which is both egalitar-
ian and also market-orientated, it follows that the costs of market access
will be an important factor. For peasant farmers, with relatively small
marketable surpluses, high transport costs and other obstacles can make
the use of the market mechanism an unattractive proposition. The con-
sequences are, Wrst, that peasant farmers revert to subsistence agricul-
ture, which makes it diYcult to pay for modern inputs or maintain
payments on debts, and secondly, that supplies of food and other
materials from the rural sector are limited.
5 In the Philippines there have been 11 consecutive land reform programmes, but thelack of rural credit and other follow-up services has meant that land has tended to passback into the hands of landlords and agricultural companies, Ghimure (2001), 13.
Sources of Inequality in Rural Germany 25
![Page 35: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/35.jpg)
Inevitably, diVerent areas of Germany had very diVerent experiences.
In the western regions the proximity of expanding urban markets made
life easier for peasant farmers who could intensify livestock production
or specialize in perishable high-value products for local markets. In the
east transport costs were high, which meant that the main marketable
crops would be cereals.6 This inevitably favoured the larger estates,
which tended to be more orientated towards cereal production. But the
elimination of internal tariVs and the creation of a single German
market, together with the expansion of the railway network, did im-
prove opportunities for farmers in some of the remoter areas.7
(v) The importance of contingency
But even the best-designed land reform is still subject to a range of
inXuences which are outside the control of the political authorities.
Two of the most important are demographic pressure, and the move-
ment of farm prices. Under circumstances where there is acute demo-
graphic pressure in rural areas, and the supply of fertile land is limited,
it will be extremely diYcult to avoid a rise in the landless population,
even with a highly egalitarian land distribution. The alternative is to
permit the subdivision of holdings to an extent which would substan-
tially reduce the marketable surplus, which would jeopardize other rural
policy objectives.
Another outside factor is the movement of farm prices. If land
reform leaves the peasant sector with an increased level of indebtedness
as a result both of compensation payments to landlords, and also invest-
ments consequent on the shift to a more market-orientated agricultural
system, then the real burden of these debts will be raised if prices fall,
and reduced if prices rise. Some notably successful land reform pro-
grammes, such as those of post-1945 Japan and eighteenth-century
Denmark, have been considerably assisted by subsequent bursts of
inXation which reduced the real value of peasant debts.
6 Which had a relatively high value-to-weight ratio; see Brinkmann (1935), 82–3 for thetransport costs of various commodities; this uses German data collected by H. Settegast inthe pre-1914 period.
7 East Prussia was an area which contained a large number of small or medium sizedholdings, and which was able to expand livestock production substantially in the latenineteenth century, moving away from wool production into more intensive livestocksystems.
26 Sources of Inequality in Rural Germany
![Page 36: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/36.jpg)
Figure 2.1 illustrates this point by comparing the movement of agri-
cultural prices in Prussia after 1816, and in Japan after 1946. The lower
line gives the movement of rye prices in Prussia (1815 boundaries);
superimposed on this is a second line giving the movement of rice
prices in Japan in the years following the land reform introduced by the
American occupation authorities in the years immediately after the
Second World War. The Japanese reforms were mainly carried out in
1946–7. In the years which followed there was a burst of inXation which
carried rice prices to levels two or three times those of the reform
period. A more gradual increase in subsequent years took prices up to
an index of 500 by the late 1960s.
By contrast, the years which followed the Prussian reforms saw a fall
in prices to levels which were around a third of the 1816–18 average by
1817
40
60
80
100
200
400
600
1820 1823 1826 1829 1832 1835 1838
Prussia
Japan
Fig. 2.1. A comparison of agricultural price movements in Prussia
(rye prices 1816–40) and Japan (rice prices 1946–69)
Three-year moving averages as indices (base for Prussia is 1816–18, for Japan 1946–8).Japanese index is superimposed (1946 corresponds to 1816).
Source: Japanese prices are from various issues of the Japanese statistical yearbook,1950–70; the index is a chained index, using whatever was reported, so that the 1946–7prices are wholesale prices (Tokyo), 1947–50 prices are black market prices (Tokyo), andthe 1950–70 prices are producer prices (all Japan). The German index uses the averagePrussian wholesale prices reported in ZdKPSB (1907), 84–92.
Sources of Inequality in Rural Germany 27
![Page 37: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/37.jpg)
the mid-1820s. The reforms were carried out under circumstances
which were considerably less favourable for peasant farmers than those
in post-1945 Japan.
The value to historians of the broader context provided by modern
writings on land reform is that it emphasizes the point that agrarian
policy needs to serve a range of objectives, and it must make a positive
contribution to the general objective of social and economic develop-
ment. Land reform cannot be judged according to a simple ‘black–
white’ dichotomy, where a resulting egalitarian land distribution is
‘white’ and an inegalitarian one is ‘black’. These other objectives also
need to be taken into account.
3. A Comparison of Agrarian Reforms in Germanyand Denmark
(i) The Danish experience
The consideration of the Danish land reforms makes an interesting
point of comparison for the later German reforms. Agrarian reform in
Denmark was essentially an eighteenth-century phenomenon, which
was largely completed by 1810. Most of the legislation was passed in
the period 1768–81 (although the abolition of serfdom came earlier in
1701). There are a number of points of interest which can be compared
and contrasted with the German experience.8
First, agrarian reform in Denmark was to a large extent a product of
‘enlightened’ state action. The state took the initiative and the reforms
were designed to serve state interests. These included the improvement
of state Wnances (which led the government to favour the peasantry who
were not exempt from taxation over the nobility who were). There was
also a conscious attempt to create a free independent peasantry; the
abolition of serfdom in 1701, for example, was connected with a deci-
sion to move to a conscript army. Neither the French revolution nor
industrialization had any part in the decision to introduce the reforms.
Secondly, despite this the reforms did not operate against aristocratic
interests in general. The nobility did give up their feudal rights but
8 The main sources used for this section are Baack (1977), Bjorn (1977), and Skrubbel-trang (1954); Tracey (1982) describes the evolution of Danish agriculture in the post-reform period.
28 Sources of Inequality in Rural Germany
![Page 38: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/38.jpg)
they were well compensated. Many estates chose to sell land to their
tenants, and it was this which increased the proportion of freehold land
held by peasants from the pre-reform level of under 10% to 40% in
1810. Between 1786 and 1807 226 out of 800 estates sold a majority of
their land. However, the nobility did achieve an increase in rents on the
land that remained in their possession, and the price of land rose by
93% between 1781–5 and 1796–1800. As land-owners, both nobility
and peasantry gained from this increase.
The main impetus to the reform came from the realization that im-
proved agricultural techniques, coupled with the enclosure and consoli-
dation of holdings, could bring about a large increase in output.
Experiments by progressive landlords in the 1760s showed that these
measures, which included the abolition of labour service and the
granting of copyhold tenancies, could lead to a doubling of production
(as occurred on the pioneering BernstorV estate between 1764 and
1783). The ‘increase in the cake’ was large enough to give everyone a
share (grain production rose 40% 1770–1803). What the government
did was to make sure that a substantial slice of the beneWts went to the
peasantry. In other states, notably Britain, the gains from agricultural
improvement went largely to landlords and large-scale tenant farmers.
Thirdly, and this emphasizes the voluntary, co-operative nature of the
reforms, some aspects were not the result of legislation. Thus the aboli-
tion of labour service or Hoveri was not decreed. Yet by 1807 this only
applied to a quarter of Danish farmland, having previously been fairly
universal. The reason was that it had become an ineYcient system which
obliged the peasantry to keep large numbers of horses instead of cattle.
So the reforms made it possible to raise cattle numbers, from 280,000 in
1774, to 560,000 in 1804 and 834,000 in 1837. At the same time the
number of horses fell (from 400,000 in 1804 to 325,000 in 1837).
Equally, the process of enclosure and consolidation was not regarded
by the peasantry with hostility. Indeed, where they were freeholders or
copyholders they often initiated such schemes themselves, and in areas
where peasant freehold was common, enclosure was earlier.9 This
points to the importance of the prior evolution of tenant rights. Where
these had become hereditary and Wrmly entrenched, with legal support,
the peasantry could bargain successfully with land-owners to secure
enclosure and the commutation of feudal duties even if the state stood
aside.
9 Bjorn (1977).
Sources of Inequality in Rural Germany 29
![Page 39: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/39.jpg)
A fourth point is the importance of Wnance. Here the state did estab-
lish a Royal Credit Bank which made loans to peasant farmers for the
purchase of freehold land or copyhold tenures. But the impact of this
bank was not large compared to other sources of Wnance. It lent 442,000
Rigsdalers for this purpose in 1786–98 (compared to 1,058,000 for land-
lord improvements). In the overall Wnancing of peasant land purchase
this was only a small share. In the 1790s it is thought that around 11⁄2million Rigsdalers were lent by banks for this purpose and another 1 mil-
lion by private investors. In 1797–1807 the Eukkase (or widow’s pension
fund) lent 2,245,000 million for the purchase of freeholds or copyholds.10
As far as the Wnancing of the cost of enclosure and other improve-
ments is concerned, this seems to have been mainly done by the land-
lords themselves. Tenants were required to repay the cost of enclosure
at an interest rate of 4%, which would have been a very reasonable rate
for agrarian credit in a pre-industrial society. However, landlords also
beneWted from the increase in rents, which may explain why they were
prepared to fund the costs of enclosure on favourable terms.
A Wfth point is that although the Danish reforms had an egalitarian
aspect, this did not mean that all beneWted equally. There were large
numbers of cottagers who did not gain freehold or copyhold land, either
through purchase or the enclosure of the common Welds. These bore the
burden of such labour service as remained. It is probably wrong to blame
the reforms for the creation of a ‘rural proletariat’ (which had probably
always existed in some form or another), but the fact that some groups
beneWted substantially while others did not was bound to increase social
divisions (at least between these groups). This was probably inevitable. No
eighteenth- or nineteenth-century government (and few since) had the
power to push through a completely egalitarian land redistribution. The
reforms had to take account of existing interests and work with these.
What the state could (and did) do was to alter the terms on which the
peasants bargained with the landlords. It forced the granting of longer
leases, regulated the conditions under which landlords could evict, and
obliged them to compensate tenants for improvements. The acquisition
of peasant holdings by large estates was prevented (as was the splitting
of holdings once consolidated).11 As one of the Danish ministers
involved, Count von Eggers, wrote in an open letter to Stein in 1807:
‘it is not enough . . . just to dissolve personal bondage . . . One must
secure them (the peasants) sustenance even as one gives them back their
10 Baack (1977), 37. 11 Tracey (1982).
30 Sources of Inequality in Rural Germany
![Page 40: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/40.jpg)
freedom; one must ensure that they remain proprietors and do not
become day-labourers.’12
Finally, the special conditions of Danish agriculture in the late eight-
eenth and early nineteenth century should be mentioned. In the Wrst place,
even before the reforms Danish agriculture had not been dominated by
‘demesne’ farming (large estates typically using labour services provided in
lieu of rent rather than wage labour). Large, directly farmed estates had
occupied only 13% of the agricultural area before the reforms; this was
reduced to 9% in 1835. Secondly, the reforms took place at a time of rising
agricultural prosperity. Between the 1730s and the 1790s land prices rose
sevenfold. Between 1771–80 and 1791–1800 rye prices rose 26%, oats
prices rose 32%, and barley 45%. This made it easier to provide adequate
landlord compensation without overburdening the peasantry. Thirdly,
Denmark was well-situated for agricultural exports. By 1803, 12% of grain
production was exported. With good access to foreign markets, in particu-
lar the expanding British market, there were favourable conditions for the
establishment of a market-orientated peasantry. Danish geography was
also well suited to an era when water transport was the only practical
means of moving bulk cargoes such as grain. Finally, the debts that the
peasants incurred as a result of the reforms were wiped out in a burst of
inXation in 1807–10. This gave the reforms a redistributional aspect which
was certainly unintended.13
(ii) Agrarian reform in central and north-west Germany
Comparison with the Danish example shows that the German reforms,
in particular those introduced in Prussia in 1807–11, diVered in several
respects. One of the principal objectives of the centrepiece of the
German reforms, the edict of 1811, was the abolition of labour services,
an issue which was not addressed by legislation in Denmark. Peasant
emancipation in the German case was, as Clapham put it, ‘two-
edged’:14 the peasant was freed from restrictions and gained much in
the way of rights, but at the same time the previous policy of the
Prussian government—Bauernschutz, the deliberate protection of
the peasantry against landlord demands—was also given up.15
12 Baack (1977), 39.13 A similar burst of inXation contributed to the success of the land reform imposed on
Japan by the US occupation authorities in 1945.14 Clapham (1921, repr. 1966), 43.15 Knapp (1891), 35–6.
Sources of Inequality in Rural Germany 31
![Page 41: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/41.jpg)
Labour services were not transformed into monetary obligations, but
rather redeemed by the cession of land to the landlord by the peasant (a
third of the total holding for tenants with hereditary rights, a half for
the others). Thus, the reforms had a built-in tendency to expand the
proportion of the total area farmed by large estates. Peasant land losses
to landlords were, however, balanced by gains as a result of enclosure.
The process still left many peasant holdings below the minimum
needed to maintain a self-suYcient rural enterprise (Spannfahige—
capable of maintaining a team of horses or oxen—is the term used in
the oYcial statistics).16
The enclosure movement in Germany started before the reforms and
continued long after. The result was that enclosure and emancipation
tended to be separate movements, with the consequence that the surge
in production which followed enclosure could not be used to provide
compensation for emancipation.
However, despite these diVerences, the fact is that in a large slice of
Germany, to the west of the Elbe and the north of the Main (and includ-
ing the Kingdom of Saxony), the reforms worked very much as they had
in Denmark: that is they produced a class of medium-sized peasant
holdings, mostly owner-occupied although the majority did rent some
land, free of labour service and other ‘feudal’ obligations, with the
common Welds largely enclosed and some consolidation of holdings.17
There were a number of reasons for this. In many areas there was, as
in Denmark, a deliberate policy bias in favour of the peasantry. The
Kingdom of Saxony is perhaps the best example of this.18 In the Rhine-
land and other western regions, French inXuence in 1806–13 may have
had some eVect, but was probably too brief to have made a major
impact.19 The availability of credit was an important factor. In Saxony
the Sachsische Landesrentenbank lent 86 million Thalers to peasants for
the purchase of labour service duties and other obligations in 1834–59.20
16 According to a Prussian survey of 1859, there were 344,701 Spannfahige holdingsand 604,501 Nichtspannfahige (holdings below the minimum size needed to maintain ateam of draught animals). The number of Spannfahige had fallen from 351,607 in 1816.However, the same survey also showed that in many areas the Spannfahige holdings hadmade net gains during the reform period; gains from the enclosure of common Weldsoutweighing losses through compensation for landlords, Meitzen (1868), 500.
17 In 1895 86% of agricultural land was owner-occupied, Rauchberg (1900), 580.18 Conze (1969).19 Dovring (1956), 144 emphasizes the impact of the French Revolution, but German
historians have tended not to support this point.20 Wehler (1995).
32 Sources of Inequality in Rural Germany
![Page 42: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/42.jpg)
In general the western peasant farmer was more likely to get access to
credit from Landeskreditkassen and Rentenbanken than his counterpart
east of the Elbe, where the dominant credit institutions, the Landschaf-
ten, lent mainly to the large estates.21
The legal framework in the large parts of north-western and central
Germany that came under Prussian administration was similar to that
in the east, where the reforms led to the transfer of substantial amounts
of land to the landlords, due to the fact that peasants were obliged to
oVer land in return for the abolition of labour services. The fact that
this did not happen suggests that labour services were either already
rare, or naturally declining, or that the peasants were able to oVer
monetary compensation instead.22 Local conditions could play a part.
Thus in the Magdeburg region the prosperity brought about by the
cultivation of sugar beet created a class of medium-sized peasant
farmers who often had shares in the sugar reWneries. Labour services
and payments in kind were ended by 1845, which was relatively early
for central Germany, and the resulting debts seem to have been paid oV
quite easily.23
In general the circumstances under which the reforms were intro-
duced seem to have been at least as important as the actual legislation
(and possibly more so). What these areas had in common with Denmark
was a low level of demesne farming before the reforms, well-entrenched
peasant rights which were often hereditary, and favourable economic
opportunities thanks to the proximity of growing urban markets which
also provided employment opportunities for surplus rural labour. It also
should be remembered that the reforms of 1807–11 applied only to
Prussia and to Spannfahige peasants (who could provide a team of
draught animals); only in the 1850s were the reforms extended to peas-
ants who were Nichtspannfahige and to important non-Prussian territor-
ies (the main reforms in Saxony date from this period). This coincided
with the period of economic growth brought about by the railway
boom. Almost certainly this was a favourable conjuncture. It meant that
labourers who did not receive land as a result of the reforms could Wnd
employment outside agriculture.24
21 Tcherkinsky (1921), 78. There were other sources of Wnance, but the Landschaftenwere the main sources of mortgage lending, ibid. 82.
22 This is the view of Conze (1969).23 Plaul (1986).24 Hempel (1957) describes how the Heuerlingen (the hired labourers of north-western
Germany) became increasingly divided between agricultural and non-agricultural employ-ment (having previously combined agricultural work and cottage industry).
Sources of Inequality in Rural Germany 33
![Page 43: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/43.jpg)
(iii) The reforms in Prussia east of the Elbe
This is the area which has received most attention from historians.
There is a widely accepted view that something went wrong with the
reforms in east-Elbian agriculture: ‘the beneWts of the agrarian reforms
went largely to the feudal aristocracy’, as one historian has written.25
The aristocracy gained substantially from the surrender of peasant land
as compensation for the loss of labour services, and also obtained 86%
of the land made available by the enclosure of the common Welds.
This view has been fairly well established since the writings of Theo-
dor von der Goltz and Adolph Meitzen in the nineteenth century. In
1868 Meitzen analysed the movement of the land distribution in the
original Prussian provinces over the 1816–59 period, and showed the
impact on peasant land-holdings of the loss of land given up to land-
lords. However, his analysis also showed that the peasantry had had
some compensation from gains made as a result of the overall increase
in the land under cultivation. Writing slightly later, in 1875, von der
Goltz examined the position of landless agricultural labourers, which
appeared to have worsened as a result of the reforms (although this
might also have been caused by demographic pressure).26
Subsequent writers, such as Max Weber and Gustav Knapp, en-
dorsed the views put forward by Meitzen and von der Goltz, and the
supposed failure of the land reforms in the Prussian east became a
generally accepted part of ‘received opinion’ on the broader question of
LandXucht: the depopulation of the rural east as a result of migration to
cities in central and western Germany.
There were Wve main reasons why the impact of the land reforms in
the east diVered from the eVect in the rest of Germany. The Wrst was
the remoteness of many eastern regions. In the period before the estab-
lishment of a railway network, marketing costs were extremely high for
any farm which did not have easy access to water transport. While the
east had its share of navigable rivers, such as the Oder and the Vistula,
there were large areas between these rivers which had poor communi-
cations, and little access to the expanding urban markets of western
Germany. In these circumstances, the establishment of a market-
orientated peasant agriculture faced almost insuperable obstacles.
25 Koselleck (1967), 498. The aristocratic estates made substantial gains from enclosure,adding over 100 hectares on average according to one calculation, Berthold et al. (1990),i. 234.
26 Meitzen (1868), von der Goltz (1875).
34 Sources of Inequality in Rural Germany
![Page 44: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/44.jpg)
The poor state of communications east of the Elbe is demonstrated
by the Wgures in Table 2.1, taken from a survey in 1867. The Wgures are
expressed in relation to land area. An average square mile of territory
east of the Elbe had only 44% of the total length of improved roads in a
typical square mile in the provinces west of the Elbe. It had 46% of the
railway mileage, and 78% of the total mileage of canals and navigable
waterways, for an area of comparable size in the rest of Prussia.
The chances of having reasonable access to improved roads or a railway
were correspondingly low.27
Secondly, access to credit was limited. Only about a tenth of the
rural population was able to obtain credit, and interest rates were in the
range 12–30%,28 much higher than the rate of 4% charged to Danish
peasants, and the similar rate which aristocratic borrowers paid on loans
from the Landschaften.29
27 The importance of having good communications is well demonstrated in contempor-ary Argentina, where the lack of good roads in rural areas is an obstacle to the develop-ment of small farms. Only large estates can aVord the high cost of getting produce tomarkets.
28 Koselleck (1967), 507.29 Tcherkinsky (1921), 43. Lutge also stresses the role of credit, Lutge (1963), 236.
Table 2.1. Communications in Prussian provinces east and west of
the Elbe, 1867
Miles of improved
road (Chausee)
Miles of
railway line
Miles of canals and
navigable rivers
East and West Prussia 53.4 9.0 14.2
Pomerania 63.0 10.8 13.4
Posen 67.2 10.7 13.0
Brandenburg 62.8 18.1 28.4
Silesia 82.9 24.3 9.0
Average east of the Elbe 64.6 14.4 15.7
Prussian Saxony 105.9 26.4 20.7
Westphalia 159.6 30.1 14.0
Rhineland 178.9 37.0 24.2
Average rest of Prussia 148.0 31.3 20.1
Note: All figures are relative to the total land area (excluding water) in square miles (unitis pre-1870 Prussian mile).Source: Calculated from figures in Meitzen (1863), i. 65–6, iii. 291.
Sources of Inequality in Rural Germany 35
![Page 45: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/45.jpg)
Thirdly, the main areas of industrial expansion were not in eastern
Germany but in the centre and the west. In the vicinity of these west-
ern cities, small-scale agricultural enterprises could prosper, supplying
perishable goods, such as fruit and vegetables, and also livestock pro-
duce. The economics of livestock production were much better suited
to peasant holdings than the extensive arable systems which character-
ized eastern Germany.30
Fourthly, the east was an area where the Prussian monarchy’s policy
of Bauernschutz—the protection and entrenchment of peasant rights—
had had less eVect. The position of the peasants before emancipation
was less favourable, and their bargaining position was consequently
weaker. This was an especially important factor in the parts of the east
which were not part of Prussia (such as Mecklenburg-Schwerin and
Mecklenburg-Strelitz) or where Prussian rule was more recently estab-
lished (such as Pomerania).31
Finally, the Gutsherrschaft administrative system, common in the
east, gave the aristocracy an important role in local government and
the administration of justice. It is probable that this had some eVect
on the agrarian reform process, either because the land-owners gained
control over the operation of the reforms, or because the government
was unwilling or unable to confront powerful local interests (a factor
which would have applied a fortiori in areas where a German aristocracy
ruled over a Polish peasantry).
In assessing the reforms, historians have tended to divide between
those who see this Wfth factor as the most important inXuence, and
those who recognize that the other factors meant that, even if the
government had been prepared to overrule aristocratic objections, the
conditions in eastern Germany in the early nineteenth century were not
favourable for small-scale, market-orientated, peasant farming.32 In
which case, a more radical reallocation of rural property rights would
have been damaging, holding back productivity gains, and producing a
subsistence sector unable to produce a marketable surplus.
For, considering the broader context of the contribution of the agrar-
ian reforms to German economic development, it is clear that there are
30 Von Thunen (1826) was the Wrst to analyse the eVect of urban proximity on agricul-tural systems; Levy (1911) explains the advantages that smallholdings specializing inlivestock produce had over larger units.
31 For a more detailed discussion of the importance of the pre-reform legacy see, interalia, Conze (1969), Lutge (1963), Harnish (1986), and Perkins (1986).
32 Schissler (1986) stresses the way the aristocracy controlled the reform process east ofthe Elbe.
36 Sources of Inequality in Rural Germany
![Page 46: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/46.jpg)
strong reasons to regard the eastern reforms as successful, despite the
failure to match the egalitarian outcome in western Germany. Free
labour, paid in cash or in kind, was, on some estimates, three times
more productive than labour contributed under the pre-reform
system.33 There was a large increase in investment: the new demesne
farms required three times as much capital as the pre-reform hold-
ings.34 And uncultivated land was reduced from 40.3% of the total
agricultural area in 1816 to 19.0% in 1864. This, combined with
increased yields (there was a 43% increase in the yields of the major
crops 1805–63), meant that in the Wrst half of the nineteenth century
the Prussian arable sector was sustaining growth rates of 1.0–1.3% per
annum.35 In the context of an economy which, in this period, was
largely pre-industrial (there were pockets of industrialization in Saxony
and the Ruhr but no general upwards shift in the rate of growth before
the 1850s), it appears that arable farming has a good claim to be
regarded as a ‘leading sector’ in this period.36
The new, post-reform, ‘demesne-farming’ of east-Elbian agriculture
was, in many respects, a highly Xexible contractual system. The land-
lord provided land to the peasant in return for labour services which
were employed in the production of marketable produce on the de-
mesne farm. In the pure form, the peasant received recompense only in
the form of land. But in practice the system showed more variation.
The Inste of eastern agriculture received some land and a share in the
harvest. The Deputate or Kossaten received a smallholding plus Wxed
payment (in kind or later in money).37 This Xexibility allowed the
system to adjust for the requirements of the production system. Thus
peasants were often given responsibility for the care of animals and the
upkeep of buildings, becoming in eVect small contractors, which
33 Tipton (1976), 24. 34 Koselleck (1967), 507.35 Finck von Finckenstein (1960), 99–100 provides Wgures for yields and the cultivated
area. The implied growth rate of production is easily calculated from these. The range of1.0 to 1.3% per annum reXects the eVect of choosing diVerent base years.
36 HoVmann (1963), 100–4, produced much higher Wgures for the growth of agricul-tural production in Germany as a whole (1.9% per annum in 1800–1860/5) and derivedfrom these an estimate of the increase in agricultural productivity of 1.3% per annum overthis period. These Wgures were, however, based on estimates of livestock productionwhich were much less well founded than those provided for the Prussian arable sector byFinck von Finckenstein. HoVmann also assumed that non-Prussian areas matched thePrussian increase in arable production, which is unlikely, given that the Prussian increasewas made possible by the incorporation of previously uncultivated areas in the east, anunder-utilized resource not available to other German states.
37 Finck von Finckenstein (1960), 197 gives payments to a Deputat around 1800.
Sources of Inequality in Rural Germany 37
![Page 47: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/47.jpg)
reduced monitoring costs for the landlord. Granting harvest shares gave
peasants an interest in the yield from the demesne Welds.38
The economics of the system shows that it was particularly well
suited to the settlement of areas where land was relatively abundant,
labour comparatively scarce, and markets either a long way away or
inherently risky. The peasant household had no need to market any of
its production beyond a small amount for local trade and barter.
Marketing costs were borne entirely by the landlord.
(iv) The extent of peasant land losses in Prussia
The extent of peasant land losses has remained one of the most serious
charges against the Prussian land reforms. Knapp claimed that the peas-
ants ‘lost’ a million hectares as a result of the reforms. This Wgure is
almost certainly an exaggeration. The situation was confused by the
point that ‘peasant’ holdings, which in the literature of the period are
generally deWned as holdings of at least 2 hectares, made both gains and
losses as a result of the reforms (losing land given up as compensation
for relief from labour services and gaining from the enclosure of the
previously uncultivated areas). The most careful examination of the net
gains and losses is probably that by Saalfeld, which is given in Table 2.2.
So, in total, peasant losses were not so large when it is considered that
labour services were removed from 16.95 million Morgen (the net loss of
5.5% of the aVected land is far less than the half to one third provided
for in the legislation), but this has to be set alongside a substantial
expansion of the total area under cultivation (the total arable area east of
the Elbe rose by 61% 1800–61).39 Thus, of the land made available by
the division of the commons only 14% went to the peasants; most of the
rest went to the large estates. It is estimated that there was an overall
increase in demesne land of 18%: 13.5% was due to the reforms and
4.5% due to purchases by the estates. Moreover, peasant losses were not
evenly distributed. In Westphalia peasant land-holdings rose by 9.9% in
area, and by 1.8% in Prussian Saxony, but falls of 13.2% and 12.8%
were recorded for Pomerania and Silesia (respectively).40
38 It has been suggested that labour service was highly ineYcient compared to ‘freelabour’ by a factor of 1 : 2 or 1 : 3 but this would have created a massive incentive to moveto wage labour; the long survival of demesne farming suggests that any ineYciencies werenot on this scale.
39 Dickler (1975), 273.40 All Wgures are from Saalfeld (1963).
38 Sources of Inequality in Rural Germany
![Page 48: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/48.jpg)
The losses have to be set against the removal of labour services. Of
the spannfahigen Bauern 84% had had their service obligations removed
(a total of 6 million days) by 1848 and of the 23.5 million Handdienst-
tage (days of labour service without animals) owed in 1816, 17 million
had also been relieved. To get some perspective on this, it implies
(assuming that the Wgure of 17 million Morgen for the aVected area is
correct) that a holding of 20 hectare (or 80 Morgen) might have owed
on average 80 days labour service without draught animals, and 28 with
draught animals: a total of 108 days.41 It would have been a substantial
gain to be free of this.
The process by which the larger estates gained at the expense of
peasant holdings was a complex one, in which free purchase played an
important role. There was a high rate of turnover in the land market in
the early part of the century, both for estates and peasant holdings.
Thus 24% of Prussian peasant land changed hands in 1816–59, but
most of these transfers (which totalled 8.2 million Morgen) were from
one peasant to another.42 Transfers to smaller or larger holdings were
only 21% of the total. Equally, the larger estates changed ownership
41 In practice it would have been more likely to be one or the other; so perhaps 50–60days with animals and no Handsdienst, or 120–50 days’ Handsdienst on its own; calculationsuse Wgures from Koselleck (1967), 498; note, the estimates are much lower than the 5–6days per week quoted by Harnisch (1986), but these levels could have been reached insome regions.
42 Saalfeld (1963).
Table 2.2. Peasant gains and losses, Prussia 1816–59, in million
Morgen
in Morgen (m.) as % of all peasant
land 1816
Losses: to smallholdings 1.29 3.8
to large estates 0.47 1.4
total 1.76 5.2
Gains: through enclosure 0.83 2.5
Net losses 0.93 2.7
Note: ‘Peasant’ holdings are defined as between 2 and 100 hectares; a ‘Morgen’ is,according to the dictionary (Langenscheidt) ‘a variable unit of between 0.6 and 0.9acres’ , but most commentators seem to assume that 4 Morgen ¼ 1 hectare.Source: Saalfeld (1963).
Sources of Inequality in Rural Germany 39
![Page 49: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/49.jpg)
quite frequently. Out of 888 estates in East Prussia in 1806 only 378
were in the same family in 1829.43
There were two main reasons for this. The Wrst was the cost of the
reforms and associated investments which led to build-up of debt. The
second was the fall in agricultural prices in the 1820s.
The reforms led to a considerable expansion of the area under culti-
vation, but this was expensive. New houses had to be built for the
labour needed to replace that previously provided by the tenants. The
cost of enclosing (and where necessary draining) the common Welds had
to be met. The peasants had to pay compensation where this had been
agreed, meet their share of the enclosure costs, and replace livestock
which were the landlord’s property.
Rather more is known about the debts owed by the large estates
which tended to be recorded (and periodically surveyed by the Prussian
government) but the position of the peasantry was also serious.44
Whereas the estates could borrow from the Landschaften at 4% the
peasantry had to pay considerably more, if they were oVered credit at
all. Certainly the nobility had problems: a third of estates in Pomerania
went into receivership during the period 1817–29.45 But the fact that
they, and those investors who subsequently bought up the bankrupt
estates, could borrow may have contributed to the gains made by the
large estates during this period.
If this is right then it was the economic success of the reforms which
contributed to their ‘social’ failure. For it was the increase in the area
under cultivation, and the consequent reduction in the uncultivated
area from 40% of the Prussian land area in 1816 to 19% in 1864, that
was the reform’s greatest achievement: ‘this increase of cultivated area
and total production was only made possible by the new energies re-
leased by the reform process’.46
(v) The regional pattern of land losses and transfers in Prussia
The impact of the reforms varied considerably in the diVerent Prussian
regions. In a sense this was an intriguing ‘natural experiment’: a similar
set of legal changes had diVerent eVects according to circumstances.
43 Koselleck (1967), 512.44 Meitzen (1885) is a rare attempt to survey debt on peasant holdings, which showed
that debts could be as high relative to property values as on larger estates.45 Koselleck (1967), 508.46 Lutge (1963), 228.
40 Sources of Inequality in Rural Germany
![Page 50: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/50.jpg)
Government surveys showed that previously existing diVerences in
agricultural structure, and in systems of agricultural tenure, aVected the
way the reforms operated in the diVerent regions. Table 2.3 gives some
Wgures for the most important reforms. The Wrst column gives the total
number of holdings which were completely released from all services.
The largest numbers of these were in Silesia and Prussian Saxony. The
next two columns give Wgures for the two of the most important cat-
egories of labour service: Spanndienst (work on the demesne farm with a
team of draught animals provided by the peasant household) and Hand-
dienst (work on the demesne farm without the obligation to provide
draught animals). The province where the amount of relief from Spann-
dienst was highest was Posen; the province with the highest amount of
relief from Handdienst was Silesia.
The Wnal column gives the area of the holdings whose land was
‘reformed’ (neu reguliert) through the enclosure of common Welds or theconsolidation of strips. These areas were largest in the four eastern
provinces: East and West Prussia, Posen, Pomerania, and Brandenburg.
The other provinces had little land in this category.
In general, the reforms are shown to have had little eVect in the
western provinces of the Rhineland and Westphalia. In these areas
feudal obligations were already in decline, as a result of developments
in the pre-reform era (including the period of Napoleonic reform).
Enclosure and the consolidation of holdings either took place in the
Table 2.3. The impact of the agrarian reforms in Prussia, 1816–48
Number of
holdings
released
from services
Given up
days of
Spanndienst
(000s)
Given up
days of
Handdienst
(000s)
Area of
reformed
holdings
(000 Morgen)
East and West Prussia 8490 375.1 568.2 1125.7
Pomerania 13,015 818.0 1465.1 1208.3
Posen 15,002 1985.9 4344.4 1388.0
Brandenburg 39,830 1145.1 2632.2 1231.3
Silesia 95,014 195.1 7547.5 205.3
Prussian Saxony 93,185 33.4 252.8 0.2
Westphalia 18,326 0.0 59.5 0.0
Rhineland 6789 0.0 0.0 0.0
Source: Figures from Meitzen (1868), iv. 302–3.
Sources of Inequality in Rural Germany 41
![Page 51: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/51.jpg)
pre-reform era or else were carried out without direct involvement by
the state or its agencies, which would exclude the reformed holdings
from these statistics. Another province where there was little eVect
from the reforms was Prussia (the province of Prussia was later divided
into East and West Prussia), apart from the area recorded in the fourth
column.
Turning to the other side of the reform process, the provision of
compensation to landlords for the loss of labour services and other
feudal duties, this shows a similar variation between diVerent regions
(Table 2.4). Large lump sum compensation payments were made in
three regions: Brandenburg, Saxony, and Silesia. In Posen, compensa-
tion was more likely to take the form of an annual monetary payment.
The heaviest payments in kind were made in Brandenburg.
The Wnal column gives the amount of land given up as compensation,
and this reXects both the extent of the reforms, and the amount of
compensation paid in other ways. Thus, the low level of land losses in
Saxony and Silesia can be connected to the relatively high level of
capital sum payments in these provinces. In Posen, land losses were not
so high, but there were substantial annual payments. The provinces
Table 2.4. The costs of agrarian reforms in Prussia, 1816–48:
compensation payments by peasants to landlords
Total compensation paid, in the form of:
Capital sums
(000 Thaler)
Annual
payments
in money
(0000
Thaler)
Annual
payments
in kind
(Scheffel
of rye)
Land
given up
(000
Morgen)
East and West Prussia 423.9 153.2 15,300 196.9
Pomerania 1403.4 140.2 36,900 590.6
Posen 104.7 492.6 10,400 206.9
Brandenburg 4172.2 266.4 108,500 405.9
Silesia 4192.5 338.9 51,900 118.4
Prussian Saxony 5932.1 134.0 23,700 14.0
Westphalia 1965.6 68.7 2100 0.3
Rhineland 350.3 5.9 600 0.1
Source: As for Table 2.3.
42 Sources of Inequality in Rural Germany
![Page 52: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/52.jpg)
with the largest losses were Pomerania and Brandenburg. In the former,
payments under the other categories were relatively low, and so it seems
that giving up land was the preferred method of compensation in this
province. In Brandenburg, the overall level of compensation appears to
have been relatively high, with high payments in all categories.
What this comparison shows is the variation in the reform process: a
typical reform in Posen might involve relief from an obligation to pro-
vide Spanndienst, labour service with draught animals, and the compen-
sation would be in the form of increased monetary payments; in Silesia,
it was more likely to be relief from Handdienst, with compensation in
the form of a lump sum payment. Concentrating on the extent of land
losses and gains by peasant households obscures these other dimensions
of the reform process.
Although the reforms did involve some loss of land by peasant
farmers, this has to be set against the gains that were made. The loss of
grazing land may have been outweighed by the ability to devote more
time to the holding (due to relief from labour services). The new hold-
ing was likely to be enclosed, which made it easier to introduce
improved husbandry techniques. Cattle could be kept instead of horses
or oxen to work the landlord’s Welds. For the landless labourers, how-
ever, the loss of common grazing was probably not outweighed by other
gains. The main social cost of the reforms was the rise in the number of
landless labourers, and the increase in the numbers of peasant holdings
which were not large enough to be self-suYcient.47
47 Berthold (1978).
Table 2.5. Consolidation of holdings in Prussian Saxony
Altengotten Großengotten
Number of farms 848 1241
Total area 6101 Morgen 6803 Morgen
Number of plots before consolidation 18920 16100
Average size 0.32 Morgen 0.42 Morgen
Number of plots after consolidation 913 1584
Average size 6.68 Morgen 4.29 Morgen
Source: Conrad (1891), 666.
Sources of Inequality in Rural Germany 43
![Page 53: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/53.jpg)
(vi) Consolidation and the reforms in southern Germany
In southern Germany the reforms started late and in one important
respect were unsuccessful: there was relatively little consolidation of
holdings. Fragmentation of holdings had been a fairly general conse-
quence of the open Welds system practised over much of Germany in
the eighteenth century. It was greatly increased by the system of part-
ible inheritance or Realteilung practised in the south (and other areas).
In those areas of central Germany where the reforms were fully imple-
mented there was considerable consolidation. Thus in two districts of
Prussian Saxony, Altengotten and großengotten, the average numbers
of separate plots were reduced as follows:
In the article from which these Wgures are drawn Conrad gives seven
examples of districts where consolidation had not taken place by the time
the article was written (it was published in 1891). In these the number of
separate plots per holding ranges from 21 to 59. Most are in the south,
though some are in central Germany (the Koblenz area, for example).
In general consolidation made slow progress in the Wrst half of the
century, and eVective legislation was not passed until the 1850s in
Prussia, Baden, and Hesse, and the 1860s in Bavaria and Wurttemberg
(further laws were passed in these states in 1886).48 Despite this there
had already been some consolidation in the north-west. Even after laws
had been passed there was much less consolidation in the south. In the
south-west in particular, where peasant rights were strongly protected,
there was opposition to the division of the common Welds.
The system which evolved in the south was characterized by a rela-
tively high population density, quite high reliance on non-agricultural
earnings to supplement the income from the holding, few large estates
but many smallholdings, and continuous division of holdings (not
necessarily on inheritance—children could be given a share on reaching
their majority). It was probably not very eYcient in economic terms.
Germany came to import livestock products from Denmark and Hol-
land which, arguably, could have been produced on southern farms.
But it provided a reserve of labour which was available for industrial
use when needed.49 And in downturns the rural economy could re-
absorb surplus labour, at least for a time.
Whether this system produced more or less discontent is hard to say.
Contemporary observers thought it gave the peasants a higher social
48 Haushofer (1957), 58. 49 Catt (1986).
44 Sources of Inequality in Rural Germany
![Page 54: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/54.jpg)
status.50 But, on the other hand, it was a vulnerable system, with a
tendency to overpopulation, and not one which was conducive to agri-
cultural progress. The dependence on non-agricultural earnings worked
well when there were local sources of employment, but less well when
craft earnings came under pressure due to competition from large-scale
industry. In practice only relatively few of those with smallholdings
found industrial employment: 27.5% of those with possession of hold-
ings of less than 2 hectares in Baden in 1907, and 23.4% in Wurttem-
berg, against 55.8% in Westphalia and 40.6% in Anhalt.51
(vii) Concluding comments
The comparison with Denmark, and the consideration of the reform
process in the diVerent parts of Germany, makes it plain that local
circumstances and the timing of the reforms could have important
eVects on the outcome. Past policies also had an eVect. The Wrm estab-
lishment of peasant rights before the reforms certainly aVected the
subsequent distribution of land. Thus in Mecklenburg and the former
Swedish Pomerania there had been little protection of peasant rights
and these areas became dominated by large estates: 75% of agricultural
land in the Stralsund region was in holdings of over 100 hectares in
1882. In Prussia east of the Elbe the proportion was 43%.52 So the
Prussian policy of peasant protection did have some eVect. But in view
of the cost of the reforms, the paucity of rural credit institutions, and
the fact that labour was becoming less scarce, it was probably inevitable
that the large estates were going to make gains, especially in the remote
areas east of the Elbe.
4. German Agriculture in Comparison
with Other European Countries
As already noted, the implicit assumption behind the ‘failed revolution’
hypothesis is that there was the possibility of a genuinely revolutionary
outcome, which would have resulted in a more complete modernization
of nineteenth-century German politics and society. For political
50 For example, Knapp (1887), 309. 51 Dillwitz (1975), 110–11.52 Perkin (1986); Dillwitz (1973); Conze (1969) refers to the annihilation of the peas-
antry in Mecklenburg and Pomerania in the eighteenth century.
Sources of Inequality in Rural Germany 45
![Page 55: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/55.jpg)
developments the implied point of comparison is usually Britain, whose
path of gradual progress to a widened franchise and a fully democratic
system of constitutional government contrasts with Germany’s halting
and uncertain moves in the same direction. For agriculture, the most
relevant comparison is the land settlement brought about by the French
revolution.
But it should not be assumed from this that all was well with rural
France in the nineteenth century. While the revolutionary settlement had
created a large number of small and medium-sized holdings, the level of
owner-occupation was lower than in Germany. In 1882, only 60% of
French agricultural land was in owner-occupation, 27% was rented, and
13% was in metayage, or share-cropping. By contrast, in Germany in
1895, over 86% of land was owner-occupied, just 12% was rented, and
only 1% was in share-cropping (this adds in Deputatland, which was a
form of share farming although applied to animal husbandry).53
A second problem was the failure of yields and productivity in
France to rise as much as in Germany. According to Bairoch, around
1800 French wheat yields were 85% of the German level; by 1910 they
were only 71% of the German Wgure. His estimates of agricultural
productivity show a rise for Germany of 371% 1800–1910, for France
the increase is 172%.54
There are various measures which can be used to compare the struc-
ture of German agriculture after the reforms with other countries. One
is the division of the agricultural labour force between employed and
self-employed workers. Labourers who were either landless, or who had
insuYcient land to provide full-time employment on their own holding,
would be obliged to take work on other farms or outside agriculture.
These would then be classiWed as employed, in contrast to those whose
own holding was large enough to support themselves and their family
without working oV the holding. The proportion of the agricultural
labour force classiWed as employed is thus an approximate measure of
53 German Wgures from Rauchberg (1900), 579; French from Auge-Laribe (1950), 93.The German Wgure was just below that of South Korea in 1980, 92% owner-occupied,which had increased from 35% in 1950, Wgures from Asian Productivity Organisation(1994), 37. The South Korean example is generally considered the best-managed landreform in recent years. By contrast, despite various eVorts at land reform, Pakistan stillhas 36% of land in tenancy and 18% in share-cropping, ibid. 70.
54 Figures from Bairoch (1989); Bairoch’s measure of agricultural productivity is cal-ories produced per male agricultural worker, which is not ideal but probably the onlymeasure which can be constructed for this period given the deWciencies of the datasources.
46 Sources of Inequality in Rural Germany
![Page 56: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/56.jpg)
the proportion that had access to holdings of a viable size. Table 2.6
provides a comparison drawn from national censuses carried out around
1900. Limitations in the data sources mean that the analysis has to be
limited to the breakdown of the male labour force.
This table shows that the German reforms produced an agrarian
sector which predominantly consisted of self-employed owner-
occupiers: the classic land reform objective. The percentage of the agri-
cultural labour force which was classiWed as being employed was low by
comparison with other European countries. The French level was
notably higher. The high level of employment in England and Wales is
not so surprising, given a historical process which had resulted in an
unusually inegalitarian allocation of rural property rights. However, a
number of other countries, not usually considered to be particularly
unequal, had lower levels of self-employment than Germany. Quite
evidently, the Danish land reforms did not produce a position in which
all had equal access to land.55
55 The extent of ‘landlessness’ is more diYcult to compare. Many employed workers inGermany, and elsewhere in Europe, had the use of a smallholding to grow potatoes andother vegetables. The Wgures given in Table 2.5 are not comparable with modern surveysof landlessness: see Jazairy et al. (1992), 406–7 for a summary of these.
Table 2.6. Percentages of the male agricultural labour force classified
as being employed (including servants), excluding family members,
various censuses circa 1900
Date of census % classified as employed
Austria-Hungary 1900 28.3
Norway 1900 29.3
Finland 1910 30.7
Germany 1907 36.7
Belgium 1901 36.8
Italy 1910 46.8
France 1901 47.2
Sweden 1900 47.6
Switzerland 1900 51.5
Denmark 1900 60.1
Netherlands 1899 67.6
England and Wales 1901 73.0
Source: Figures from Flora et al. (1983), adjusted to exclude family members andunclassified workers.
Sources of Inequality in Rural Germany 47
![Page 57: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/57.jpg)
Comparable Wgures on the distribution of agricultural land by size of
holding are only available for a few countries. Table 2.7 provides some
Wgures for England and Wales, Germany, and France. The Wgures have
been adjusted to allow for diVerences in the classiWcations by size class.
The British and German classiWcations use broadly comparable meth-
odology: German holdings were classiWed by total agricultural area, the
British by area under crops and grass. The French agricultural censuses
classiWed holdings by total area, which tended to increase the import-
ance of larger holdings, which would often include a disproportionate
amount of woodland, moor, or other uncultivated areas.
Comparison of the English and German Wgures shows that the distri-
bution of land by size of holding was much more unequal in England:
there was a smaller smallholding sector and the area in large holdings
was proportionately greater. Typically, English agricultural land was
concentrated in medium- or large-scale tenanted holdings of 100 acres
or more. Comparison with France is more diYcult, due to the diVerent
classiWcation methods used. An English statistician who examined this
issue in 1887, P. G. Craigie, thought that, whereas three-quarters of the
English agricultural area was in holdings of over 100 acres (roughly
Table 2.7. A comparison of the percentage distribution of land by
size of holding
England and
Wales 1885
Germany 1882 France 1882
Area in holdings of less
than 20 hectares
14.4 45.0 38.2
Area in holdings of
between 20 and 40 hectares
13.0 13.1 12.9
Area in holdings of
over 40 hectares
72.6 41.9 48.9
Area in holdings of
between 40 and 100 hectares
34.0 17.5 n.a.
Area in holdings of
over 100 hectares
38.6 24.4 n.a.
Note: England and Wales: classified by area under crops and grass; Germany: classifiedby total agricultural area; France: classified by total area.Source: British figures from Ministry of Agriculture, Fisheries and Food (1968), 20;French and German figures from Tracey (1982). To produce comparable figures, thedensity functions were interpolated using a log-log interpolation method.
48 Sources of Inequality in Rural Germany
![Page 58: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/58.jpg)
equivalent to 40 hectares), only a third of the equivalent French area
was in this class.56 This is a considerably lower proportion than the
48.9% share given in Table 2.7. If this is right, then the distribution of
agricultural area in France may have been rather more egalitarian than
in Germany, contrary to the impression given by the French distribu-
tion by total land area, shown in Table 2.2. Either way, the diVerence
was probably not that great.
However, the objective of agrarian reform is not just to produce a
relatively equal distribution of land-holding: it is also to produce a more
egalitarian pattern of land ownership. Comparison of the distribution of
land according to the size of the owned farm or estate is even more
diYcult, as there are no oYcial Wgures available for this period.
In England in the early 1870s there was a Local Government survey
of land ownership, sometimes referred to as the ‘New Doomsday
Book’.57 This survey showed that land ownership in England and
Wales was highly concentrated: 74% of privately owned land was in
estates of 300 acres or more. This is a distribution by size of total land
ownership, not ownership of agricultural land. Making some allowance
for this suggests that between 60% and 65% of agricultural land was in
estates of 300 acres or more (the lower Wgure assumes that all non-
agricultural land was in larger estates), and between 65% and 70% in
estates of 100 hectares or more. This contrasts with the farm census
Wgures which show that 28% of agricultural land was in holdings of 300acres or more in 1875 (or 38% in holdings of over 100 hectares). Thus,
the distribution of land by ownership was much more inegalitarian
than the distribution by holding.
The high level of owner-occupation in Germany meant that the
gap between the distribution by ownership and by holding was nothing
like as large as in Britain. The distribution by ownership may well
have been more egalitarian than the distribution by holding size. This
could arise if larger holdings rented from more than one landlord, or
if they included some rented land as well as some which was owner-
occupied.
The best source of information on the structure of ownership in
German agriculture is a survey carried out by the Reich statistical oYce
in 1938. This produced a careful cross-tabulation of the distribution of
56 Craigie (1887), 110; Levy (1911) is also a useful source on the structure of Englishagriculture in comparison with other countries.
57 The original returns were summarized in Bateman (1879), but this study makes useof the adjusted results published in Brodrick (1881).
Sources of Inequality in Rural Germany 49
![Page 59: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/59.jpg)
land by size of estate and by size of holding (something which is
not available for British agriculture). The results are summarized in
Table 2.8.
This table shows a fairly close correspondence between the size dis-
tribution by holding size and the pattern of ownership. There are some
regions where the ownership distribution was signiWcantly more equal
than the holding distribution: Brunswick, Silesia, Westphalia, and
Hanover are the most notable. There are two regions where the owner-
ship distribution was signiWcantly less egalitarian: Mecklenburg and
Anhalt. But in general the two series do not diVer by much.
Table 2.8. The distribution of land by size of holding and by size of
estate, by province, 1938
% of land in
individual
ownership
% of agricultural
land in holdings of
over 100 hectares
% of all privately
owned land in estates
of over 100 hectares
East Prussia 72.0 32.2 32.0
Brandenburg 71.1 35.7 31.1
Pomerania 71.8 46.5 40.0
Silesia 71.7 38.1 26.7
Prussian Saxony 72.1 25.9 25.7
Schleswig-Holstein 83.8 16.8 14.4
Hanover 77.1 19.7 10.1
Westphalia 81.1 19.1 5.7
Hesse-Nassau 48.2 6.4 6.7
Rhineland 59.9 10.5 4.7
Bavaria 72.1 5.7 4.5
Saxony 72.4 17.4 13.8
Wurttemberg 63.9 1.1 3.9
Baden 61.5 10.9 5.9
Thuringia 62.3 9.8 11.3
Hesse 62.7 9.9 5.1
Mecklenburg 82.8 35.0 48.2
Oldenburg 84.4 11.8 19.3
Brunswick 56.2 13.2 2.9
Anhalt 55.5 26.1 35.2
All Germany 70.0 22.3 18.7
Source: SDR (1938).
50 Sources of Inequality in Rural Germany
![Page 60: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/60.jpg)
The drawing of conclusions for the pre-1914 era from a survey
undertaken in 1938 is obviously something which needs to be done
with some care. The crisis which aVected German agriculture in the
1920s would have altered the relationship between the two series to
some extent. One consequence could have been an increase in the own-
ership of land by institutions, both public sector and private (as a
result of the repossession of mortgaged estates, for example). The Wrstcolumn of Table 2.8 shows that, in some regions, the percentage of land
in private ownership had fallen to under 60% (though this does
not appear to have much eVect on the relationship between the two
distributions).
With these caveats in mind, a hedged statement is advisable: on the
evidence currently available, there is no reason to anticipate that the
ownership distribution in the German regions before 1914 would have
been markedly less egalitarian than the holding distribution, with the
exception of regions such as Mecklenburg where the 1938 survey
showed this to be the case. In other words the holding distribution is a
good guide to inequality in land ownership in pre-1914 Germany
(which follows from the high level of owner-occupation in German
agriculture).
Looking at the map of German agriculture in 1895 (Figure 2.2), the
only areas with a level of inequality in land-holding which came up to
the English level (38.6% of land in holdings of over 100 hectares) were
those east of the Elbe. Even in this area there were several regions
where larger holdings were relatively rare: in the East Prussian region
of Gumbinnen holdings with over 100 hectares accounted for 33.1% of
the total agricultural area, in the Silesian Regierungsbezirke of Oppeln
and Liegnitz they occupied 31.2 and 28.4% respectively.
But when allowance is made for the substantial gap in Britain be-
tween inequality in land ownership and in land-holding, the diVerence
between the two countries becomes a massive one. The English level of
65–70% of land in the ownership of estates with at least 100 hectares
was way above the most probable level in most regions east of the Elbe.
It was equalled only in parts of Pomerania (the Stralsund Regierungsbe-
zirk and possibly Koslin as well) and, most probably, in Mecklenburg.
In Posen and Bromberg, the English level could have been reached if
the ownership distribution had been substantially less equal than the
holding distribution.
The image of east-Elbian agriculture as one dominated by large
estates, on the English pattern, is to a large degree a false one. The
Sources of Inequality in Rural Germany 51
![Page 61: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/61.jpg)
typical farm in Brandenburg, Silesia, East Prussia, and the Danzig
region of West Prussia was more likely to be an owner-occupied hold-
ing of around 30–50 hectares. Even where larger estates predominated,
they were very diVerent from the English model: an average Junker
estate might consist of around 250 hectares farmed ‘in hand’; the
equivalent English aristocratic estate in the 1890s would be almost en-
tirely let out to tenants, and considerably larger.
In Germany 14% of land was in some form of tenancy in 1895; the
Wgure for England and Wales in 1891 was 85%.58 Tenanted percentages
in German regions came nowhere near the English Wgures. Figure 2.3
gives the percentage of the total agricultural area which was rented, by
region. The map shows that the highest percentages were in Alsace-
Lorraine and in Mecklenburg. The lowest was in Bavaria. In general,
the rented percentage was not particularly high east of the Elbe, apart
58 Rauchberg (1900), 579; Ministry of Agriculture, Fisheries and Food (1968), 25.
<5%
5−10%
10−15%
15−20%
20−25%
25−30%
30−35%
35−40%
40−45%
45−50%
50−55%
55−60%
60−65%
>65%
Fig. 2.2. Percentage of land in holdings of over 100 hectares, 1895
52 Sources of Inequality in Rural Germany
![Page 62: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/62.jpg)
from Mecklenburg. Levels in the other eastern provinces were below
those in central and western regions such as Hanover, Prussian Saxony,
and the Rhineland.
In most of Germany, especially in the west and the south, large
estates were a rarity. The typical farm was small, 10–20 hectares, and
owner-occupied. There was little employed labour. The rural sector
was therefore, by the standards of contemporary European countries, a
relatively egalitarian one. The low proportion of landless labourers in
the rural population as a whole, and the high level of owner-occupation,
mean that the structure of nineteenth-century German agriculture
compared well with the situation in many less-developed economies
today.59
5. The Kuznets Curve and German Agriculture
The Kuznets Curve is a hypothesis about the development of inequality
in industrializing economies based on a series of assumptions about
inequality in diVerent sectors. However, the analysis of particular coun-
tries shows that historical circumstances and geographical conditions
matter a lot and can lead to large diVerences: Do these favour large
estates or plantations? Has there been an attempt at land reform? Was
this successful or unsuccessful?
In the German case, the agrarian reforms produced a relatively egali-
tarian rural sector, which meant that the starting point for the Kuznets
Curve was a low one.60 If inequality in the industrial sector turned out
to be high, then a shift into this sector from the more egalitarian rural
sector would tend to push up overall inequality. The structure of
German agriculture also meant that the experience of rural–urban
migration produced a much more dramatic change in social circum-
stances than in Britain. In Britain the tripartite division of industrial
society—owner or shareholder at the top, salaried manager or foreman
in the middle, wage-earning worker at the bottom—was not dissimilar
to the tripartite division of rural society—large land-owner who rented
59 Prosterman and Riedinger (1987), 41 gives the current position in developing econ-omies.
60 The Prussian rural distribution was an egalitarian one by modern standards. Thisissue is further discussed in Chapter 9, which includes estimates of income inequalityderived from tax statistics.
Sources of Inequality in Rural Germany 53
![Page 63: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/63.jpg)
out land at the top, tenant farmer in the middle, landless rural labourer
at the bottom.
The position in rural Germany was a very diVerent one. Migrants
from the rural areas in the south and the west came from areas where
peasant proprietorship was the dominant system. If they were the sons
and daughters of peasant farmers they would have had little experience
of wage labour. If they came from the families of the relatively small
number of landless labourers, the social context would still have been a
dissimilar one (working alongside neighbours who did own some land)
to that provided by an urban factory. Even in the east, the position of
the agricultural labourer was not that of a wage-earning labourer paid
by the hour. There were a variety of contractual arrangements, most of
which gave a degree of autonomy to the labourer and his family, at least
as far as the organization of the day’s work was concerned.61
61 The system was one which required the worker to supply speciWed hours of labourat harvest time and other peak periods, but which left the organization of duties such ascow-keeping up to the labourer. Outside peak periods, the worker and his family were
>5%
5−10%
10−15%
20−25%
15−20%
25−30%
30−35%
Fig. 2.3. Rented land as a percentage of the total agricultural area, 1882
54 Sources of Inequality in Rural Germany
![Page 64: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/64.jpg)
For these reasons, the structure of German agriculture had an eVect
on the impact of industrialization and migration: the relative change
was a greater one. Industrialization brought about a transformation of
society which was more rapid than in early-industrializing economies
such as Britain. It was also a more profound change than in Britain.
expected to occupy themselves on a smallholding provided by the land-owner. See Weber(1893) and Chapter 6 below for further details.
Sources of Inequality in Rural Germany 55
![Page 65: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/65.jpg)
3
The Pattern of Migration,
1870–1913
1. Introduction
The discussion in the Wrst two chapters has made it clear that there are
particular reasons to be concerned with the process of migration during
the early stages of industrialization. First, migration is the conduit
whereby a rural labour surplus holds down wages in urban labour
markets during the labour surplus phase of the Lewis Model. Secondly,
the transfer of population from a relatively egalitarian rural sector to a
more unequal urban sector is an integral part of the Kuznets process,
pushing up overall inequality. Both these factors make it important to
consider the factors which aVected the pace of migration, and which
eventually led to a change in the pattern of migration as the labour
surplus phase came to an end.1
The important feature of the labour surplus phase is the movement
of population from predominantly agricultural regions to the expanding
urban centres, resulting in a high level of net losses in the former areas,
and large gains in the latter. But this feature was not the only type of
migration in industrializing societies. It was superimposed on a pattern
which already existed in pre-industrial societies: of seasonal migration
connected with the harvest demands and other types of work inXuencedby weather conditions. It also merged into a more modern pattern,
characteristic of mature industrial societies, of migration between urban
centres, responding to shifts in labour demand by industry. But neither
of these other streams has the implications predicted by the Lewis
Model.
Other migration streams which have diVerent implications are those
which cross international boundaries. Where there is a large outward
1 Internal migration is not the only source of downward pressure on labour markets.Kindleberger (1967), 20 argues that in late nineteenth-century America there were laboursurplus conditions created by heavy immigration. In post-1945 West Germany there was asimilar eVect as a result of the inXux of refugees, ibid. 34.
![Page 66: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/66.jpg)
Xow of emigrants this will transfer a portion of the rural labour surplus
to foreign labour markets, thus acting as a ‘safety valve’, relieving pres-
sure on over-supplied urban markets in the country of origin. But
where there is large-scale immigration, then surplus conditions else-
where can intensify the ‘labour surplus’ phase in the host country.
EVectively, the labour market in the receiving country is integrated
with that of the country of origin.
Those parts of the migration process which have tended to attract
most attention are moves which result in a permanent gain or loss of
population. In the aggregate this is net migration: a surplus or deWcit is
created when moves in are deducted from moves out. But in most soci-
eties net internal migration is relatively small compared to gross internal
migration. There are many moves which are temporary or reversible.
There are life-cycle moves: young men (less often women) who move to
cities intending to return, perhaps hoping to have accumulated enough
savings to buy a smallholding or some other business in their region of
birth. There are seasonal moves: farmers who take winter jobs in cities
but return for the harvest; other agricultural workers who move with the
harvest, starting in the warmer southern regions, and moving north and
east as the harvest season progresses. There are certain skilled occupa-
tions which have always been associated with high mobility: stone-
masons and other skilled building workers moved with major
construction projects in medieval Europe, such as bridges or cathedrals.
Railway construction had a similar eVect in the nineteenth century,
attracting large numbers of workers, skilled and unskilled.2
Most of the available statistical sources tend to measure net migration
better than gross migration. Census returns generally record place of
residence at the time of the census and place of birth: it is rare to Wnd
any evidence on places of residence between these two points. This
means that many temporary moves between these dates are unrecorded.
Germany is relatively unusual in that, thanks to the system of
MeldepXicht, compulsory registration after a change of residence, there
is some evidence on gross migration.3 This makes it possible to examine
the relationship between the labour surplus transfers and the general
pattern of mobility.
2 Von der Goltz (1875) includes the results of a survey of rural migration in the early1870s, which shows both the traditional pattern of seasonal migration, and the impact ofrailway construction.
3 The MeldepXicht system and its value as a migration source is described in Meyer(1938).
The Pattern of Migration, 1870–1913 57
![Page 67: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/67.jpg)
2. German M igration and Urbanization
in International Comparison
Most of this study is concerned with the eVect of internal migration in
Germany. But it is worth considering whether the German experience
diVered from that of other European countries, and if so, in what way?
It is not easy to achieve comparability of migration statistics between
countries. The drawing of the regional boundaries aVects migration
records. A country with a large number of regional units will tend to
have more recorded migration than one with a relatively small number
of larger units. An accurate comparison needs to be based on a similar
set of regional units for all countries.
In the late nineteenth century, the American statistician, Adna
Weber, attempted such a comparison. His results are given in Table 3.1.
These show that, compared to all other European countries, the German
states of Prussia and Saxony had the highest rates of internal migration.
In many other countries there were relatively few internal moves. This is
particularly striking for the Scandinavian countries, which had high
rates of external migration, but much less internal migration.4
4 Norway and Sweden had the highest rates of emigration after Ireland, Hatton andWilliamson (1998), 33.
Table 3.1. Percentage of the population born outside
the country where enumerated
Date of census % born elsewhere
Saxony 1885 30.9
Prussia 1890 30.3
England and Wales 1891 28.4
Denmark 1890 22.1
Austria 1890 19.8
France 1891 18.7
Switzerland 1888 17.9
Netherlands 1889 13.0
Norway 1875 12.8
Sweden 1880 11.5
Hungary 1890 10.8
Source: Weber (1899), 249.
58 The Pattern of Migration, 1870–1913
![Page 68: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/68.jpg)
Another approach is to compare rates of urbanization. If cities are
growing fast, so that the proportion of the population living in cities is
rising, then this can come about either through higher rates of natural
increase or through migration. In practice, in most European countries,
the Wrst factor was not important: the rate of natural increase in cities
was generally below that in rural areas. So, rising urbanization was a
consequence of migration.
Table 3.2 presents some Wgures for the proportion of the population
living in cities in various European countries. These are countries for
which results are available both for the period around 1870 and again
for the period just before the First World War. The results for the
Netherlands in 1909 are given in brackets as there are no comparable
Wgures for the earlier period. The table is in two parts: Part A gives the
percentage of the population living in cities of over 100,000 inhabitants;
Part B gives the proportion living in towns and cities with over 20,000inhabitants.
The table shows that by 1910 Germany was the third most urbanized
country in Europe after Britain and the Netherlands. There is a sign-
iWcant gap between Germany and the fourth-placed country on both
measures. In 1870, by contrast, Germany was behind France and Bel-
gium on both measures, and around the same level as Italy, Denmark,
and Ireland (the relative position depending on the measure used).
The increase in urbanization between the two dates (measured as the
annual gain in percentage points) is larger in Germany than in any
other country. The German gain during the period 1871–1910 was 16.5
percentage points using the ‘over 100,000’ measure and 22.2 using the
‘over 20,000’ measure. The next-placed country was England and
Wales with gains of 12.1 and 18.6 respectively (1871–1911). The gain
for France was just 5.7 percentage points on the ‘over 100,000’ meas-
ure, and 9.3 percentage points for the ‘over 20,000’ measure.
Urbanization in nineteenth-century Germany proceeded at a much
faster pace than in other European countries. The pressure on urban
services and amenities was correspondingly high, as was the pressure on
urban labour markets in the Lewis Model framework. Where high urban
growth resulted from high inward migration, this also meant that a
higher portion of the population was exposed to the experience of social
dislocation resulting from moves from a rural society still dominated by
traditional attitudes and values to a very diVerent urban environment.
Table 3.3 provides Wgures for a more detailed comparison of
Germany with England and Wales over the whole of the nineteenth
The Pattern of Migration, 1870–1913 59
![Page 69: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/69.jpg)
Table 3.2. Urbanization in various European countries, c. 1870
and c. 1910
A: Living in cities >100,000
Date of
census
% Date of
census
%
England and Wales 1871 25.8 England and Wales 1911 37.9
Scotland 1871 22.7 Scotland 1911 30.1
Denmark 1870 10.2 (Netherlands) (1909) (29.7)
France 1872 9.1 Germany 1910 21.3
Belgium 1866 8.1 Denmark 1911 16.8
Ireland 1871 7.8 Ireland 1911 15.8
Italy 1861 5.4 France 1910 14.8
Germany 1871 4.8 Belgium 1910 11.0
Austria 1869 3.7 Italy 1911 10.9
Sweden 1870 3.3 Austria 1910 10.6
Norway 1875 0 Norway 1910 10.3
Switzerland 1870 0 Sweden 1910 9.2
Finland 1870 0 Switzerland 1910 8.6
Netherlands not available Finland 1910 4.7
B: Living in towns and cities >20,000
Date of
census
% Date of
census
%
England and Wales 1871 42.0 England and Wales 1911 60.6
Scotland 1871 31.8 Scotland 1911 49.5
Belgium 1866 15.4 Germany 1910 34.7
France 1872 14.8 Belgium 1910 28.8
Germany 1871 12.5 Italy 1911 28.2
Ireland 1871 11.6 Denmark 1911 27.0
Italy 1861 11.1 France 1910 24.1
Denmark 1870 10.2 Ireland 1911 21.5
Norway 1875 8.5 Switzerland 1910 21.1
Austria 1869 6.5 Norway 1910 18.1
Switzerland 1870 6.4 Austria 1910 16.9
Sweden 1870 5.8 Sweden 1910 16.1
Finland 1870 1.6 Finland 1910 9.3
(Netherlands) (1909) (40.4)
Source: Data from Flora et al. (1996), 249–80; the French and Belgian results useddiVerent-size classes and were adjusted to make them comparable by extrapolating thedensity function.
60 The Pattern of Migration, 1870–1913
![Page 70: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/70.jpg)
century. This shows that in 1801 England was much more urbanized
than Germany in 1871. Interpolating between the census years, it
appears that Germany reached the English 1801 Wgure of 23.0% in
1893. Germany then moved to 34.7% in 1910. Using the same ap-
proach on the results for England and Wales shows that this Wgure was
attained around 1853. So, while it took Germany 17 years to get from
23.0% to 34.7%, it took England and Wales 52 years to make the same
move. Urbanization was three times as fast in Germany than in England
during the equivalent period.5
Looking at results for individual cities, it is clear that one reason why
urbanization was so fast in Germany was that all the main cities grew
fast. The annual rates of growth of the Wve largest cities in 1871–1900
were:
Table 3.3. Percentage of the population living in towns and
cities >20,000
Germany England and Wales
Date of census % Date of census %
1801 23.0
1811 24.4
1821 25.9
1831 28.5
1841 30.7
1851 33.6
1861 38.2
1871 12.5 1871 42.0
1880 16.1 1881 47.9
1890 21.0 1891 53.7
1900 28.8 1901 58.2
1910 34.7 1911 60.6
Note: A comparison with France shows that the move from 4.8% of the population livingin cities of over 100,000 to 14.8% took 59 years in France (1852–1911) but 26 years inGermany (1871–97).Source: Data from Flora et al. (1996), 262, 278.
5 There is no simple relationship between urbanization and economic growth, but thiscomparison provides another piece of evidence (if more were needed) indicating that therate of growth in Britain during the industrial revolution was relatively slow.
The Pattern of Migration, 1870–1913 61
![Page 71: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/71.jpg)
Berlin 3.70%
Hamburg 2.85%
Munich 3.52%
Leipzig 3.66%
Dresden 3.22%6
In Britain there were cities which matched these rates of growth:
Liverpool grew at 3.30% per annum in 1801–31; Glasgow at 3.03% in
1801–41. But there were other cities which grew less fast: Birmingham
at 2.4% in 1801–31, and London at 1.95% in 1801–41. The slow growth
of London, which was already a large city in the eighteenth century, is
particularly important in holding back the overall rate of increase.7
Comparison with cities outside Western Europe shows that there
were American cities which matched or surpassed German rates of
growth: New York grew at 3.9% per annum in 1825–71 and Chicago
went from a population of 80,000 in 1855 to one of 420,000 in 1876, an
annual growth rate of 8.2%. Massive immigration made these rates of
growth possible.
But the German rate of urbanization was unusually fast by the stand-
ards of nineteenth-century Europe: faster than any other European
country had experienced. This meant that internal migration was also
relatively high. The natural rate of population increase for Berlin was
only 0.95% per annum in 1841–1900; it required high rates of inward
migration to raise the annual growth rate to 3.7%.
3. Measuring Gross and Net M igration
As mentioned in the Introduction, there was an obligation on the
German population to register a change of residence. This created a
source of data on migration which is particularly valuable in that it
gives information about total moves. The obligation did not apply to
temporary changes and there may also have been some evasion. In 1902
the Berlin statistical oYce published adjusted Wgures which took
account of under-recording, the extent of this being estimated from a
comparison with the census results. No adjustment to the amount of
6 Figures from Schrott (1903), who provides a valuable adjustment to the populationWgures which takes account of boundary changes.
7 British Wgures taken from the 1879 edition of the Encyclopaedia Britannica, articles onvarious cities.
62 The Pattern of Migration, 1870–1913
![Page 72: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/72.jpg)
inward migration was considered necessary, but the numbers of out-
ward moves were adjusted upwards by 17%. However, it is possible
that the Berlin police were unusually vigilant. Others may have been
more lax. It is noticeable, for example, that recorded gross migration is
signiWcantly lower for Bavarian cities than in other parts of Germany.8
Keeping these reservations in mind, these Wgures still show a society
which had high levels of mobility. From 1898 onwards the available city
statistics were collected and published in the Statistisches Jahrbuch
Deutscher Stadte. By 1907 this included Wgures for 48 cities, with a total
population of 11.3 million. These cities recorded a total number of
moves in of 2.06 million and moves out of 1.81 million. Total moves
amounted to 346 per 1,000 head of population.9 There will, however,
be some double counting in this (moves from one city to another will be
counted twice). Moreover, some individuals will have left and come
back in the same year, thus counting as two moves.
The adjusted series produced by the Berlin statistical oYce runs
from 1841 to 1902. For this study the series has been extended to 1912
using the same adjustment procedure, and the results of the 1905 and
1910 censuses. Figure 3.1 shows the results.
There are four periods which can be distinguished. The Wrst runs
from 1841 to the late 1850s. In this period overall mobility is relatively
low. On average total moves are 141 per 1,000 head of population in
1841–59. Net gains are 1.2% per annum on average, but these Xuctuateconsiderably. In the second period, from 1860 to the early 1870s, overall
mobility is much higher and rises to a peak of over 270 moves per 1,000
in 1871–3. Net gains are also high, averaging 3.4% in 1860–73,
although these do still Xuctuate.10
It is worth taking note of the timing of this phase: it coincides with
the onset of faster industrial growth brought about by the boom in
railway construction. It is a period in which agricultural prices were
tending to rise. The increase in mobility was not connected to any
agricultural crisis.
8 Berlin Wgures from SJB (1903), 158; Wgures for all cities were published annually inSJDS.
9 SJDS (1908), 39; all the main cities are included with the exception of Munich.Total moves includes all moves to and from each city; moves within the city boundariesare not included.
10 This is partly due to the inXuence of the war years 1866 and 1870; in both theseyears net migration is low, presumably as a result of the conscription of many potentialmigrants.
The Pattern of Migration, 1870–1913 63
![Page 73: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/73.jpg)
In the third period, from 1874 to 1890, mobility is below the 1871–3
peak although there is an upward trend. Net gains are substantial,
although, at 2.2% per annum, these are lower than in the previous
period. From 1890 onwards net migration falls, while gross migration
rises, eventually rising above the level of the 1871–3 peak. By 1910–12
Berlin was no longer gaining population through migration.11 So the
Wnal period is one of high mobility, a Xexible pre-1914 labour market,
but not one of high net migration. Whether this is typical of Germany
as a whole is a question which will be considered later.
What was the reason for this high mobility? There was a seasonal
element, which leads one to conclude that some migrants did return
home to help with farmwork in the summer. Figure 3.2 gives a monthly
breakdown of migration in and out of Berlin in 1901–2.12 There was a
11 The chart shows a net loss; however, this is partly due to a shift in population toareas which were not included within the pre-1914 city boundaries: Charlottenburg,Neukolln, Schoneberg, Spandau, Lichtenberg, and Wilmersdorf. Including these in a‘Greater Berlin’ unit (and allowing for some under-recording of moves out) alters theoverall result to one of balance: moves out equalled moves in.
12 Data from SJB (1903), 159; the moves out have been adjusted upwards by thestatistical oYce.
1841−100
0
100
200
300
400
1849 1857 1865 1873 1881 1889 1897 1905
Net moves
Total moves
Fig. 3.1. Net and gross migration, Berlin 1841–1912: net moves and total
moves per 1,000 head of population
Source: The original series was in SJB (1903), 158; the extension uses annual data fromSJDS, adjusted for the census results for 1905 and 1910.
64 The Pattern of Migration, 1870–1913
![Page 74: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/74.jpg)
substantial move out in March, at the start of the sowing season, and a
smaller outwards movement in the summer months. In all Berlin lost
0.76% of the total population in the period February–July and gained
0.70% in the remaining months. However, this is probably a serious
underestimate of the true movement; many of those who returned for
just a few weeks in the summer may not have needed to register a
change of address. In general, however, this is not a major contribution
to overall mobility. As the chart shows, there were a large number of
moves in both directions in most months. Monthly net migration is
small relative to gross migration, as is annual net migration.
Results from other cities also show a seasonal pattern. Figure 3.3
provides a similar analysis of moves to and from Cologne in 1912–13.13
Here the seasonal pattern is a rather diVerent one: there is net outwards
migration in the June–August period, equivalent to 0.5% of the total
population, and gains in the remaining months amounting to 2.9% of
the total population. Again, monthly net migration is small relative to
gross migration.
13 SJK (1914), 30; the moves out are not adjusted, so these may be underestimated.
Jan.
−2
−1
0
1
2
3
Mar. May July Sept. Nov.
Net moves
Moves in
Moves out
Fig. 3.2. Net moves, moves out and moves in, as percentages of the total
population, Berlin 1901–2 (averages of the two years)
Source: Data from SJB (1903), 159.
The Pattern of Migration, 1870–1913 65
![Page 75: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/75.jpg)
It is also worth considering what aspects of migration are obscured or
downgraded by the tendency to concentrate on net migration. Two
studies are of interest. The Wrst is an analysis by the Berlin statistical
oYce of the occupations of migrants in 1910. In this year there were
263,991 registered moves into Berlin and 206,552 moves out, an appar-
ent gain of 56,839. Table 3.4a compares the breakdown by occupation
of this Wgure with the distribution of total moves (in and out).
In this year Berlin took in numbers of industrial workers, and these
accounted for a third of net moves, but only a quarter of all moves. By
contrast, the professional categories, oYcials, students, and workers in
trade and commerce (a highly mobile group according to census
returns) all occupy a larger share of total moves than net moves.
The Cologne statistical oYce classiWed migrants in 1913 according to
their last place of residence or intended destination. This again can be
used to compare net and gross migration. The results (in Table 3.4b)
show that net migration Wgures tend to be biased towards moves over
longer distances, while the importance of moves within the locality, in
this case the Regierungsbezirk Koln, is underestimated.
Net moves
Moves in
Moves out
Jan.−2
−1
0
1
2
Mar. May July Sept. Nov.
Fig. 3.3. Net moves, moves out and moves in, as percentages of the total
population, Cologne 1911–12 (averages of the two years)
66 The Pattern of Migration, 1870–1913
![Page 76: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/76.jpg)
To summarize, using Wgures for net migration concentrates attention
on the movement of migrants over substantial distances into urban
industrial jobs, but there are other short-distance moves which are also
an important part of internal migration in the period: a student going to
a large city to study; a doctor or oYcial moving to a diVerent post; a
small trader trying his or her luck in a series of diVerent localities. Most
of the rest of this study will be concerned with net migration, or moves
revealed by the census comparison of place of residence with place of
Table 3.4a. A comparison of the distribution by occupation of net
moves and total moves, Berlin 1910
% share in
net moves
% share in
total moves
Occupied in agriculture and Wshery 1.15 0.78
Occupied in industry 33.62 26.54
Occupied in trade and commerce 15.74 20.01
Domestic servants 25.60 23.09
General labourers 16.63 11.98
Professional, state oVicials, and military personnel 1.67 7.40
Pensioners and rentiers 1.14 1.31
Students 0.65 4.20
UnclassiWed 3.80 4.71
Note: Unoccupied family members are included, classiWed according to the occupation ofthe head of the household.Source: Calculated from SJB (1908–11), 215, which classiWed migrants into 43occupational categories. These have been amalgamated for use in Table 3.4a.
Table 3.4b. A comparison of the distribution of net moves and total
moves, by last place of residence or intended destination, Cologne 1913
% share in
net moves
% share in
total moves
Rest of Regierungsbezirk Koln 18.4 29.4
Rest of Rhineland province 43.5 38.8
Hesse-Nassau and Westphalia 12.7 12.1
Rest of Germany 25.5 19.7
Source: Calculated from SJK (1913), 27.
The Pattern of Migration, 1870–1913 67
![Page 77: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/77.jpg)
birth. But it should be borne in mind that these moves represent just
the visible part of a much larger iceberg.
4. Sources of M igrants
Where did migrants come from? Did this change over time? The analy-
sis of census data provides one way to answer these questions. By
looking at cohorts, census returns sorted by the date of arrival in the
city, it is possible to compare the origins of migrants in diVerent
periods. Two points should be noted in advance. The Wrst, following
from the discussion in the previous section, is that this will only capture
permanent moves, or at least moves which had not been reversed by the
time of the census. The second is that the analysis will be aVected if
there are variations in death rates according to migrant origins. If
migrants from the rural east, for example, live in poorer quality hous-
ing, and have lower life expectancy in consequence, then there will be a
tendency for these to be under-represented in earlier cohorts.
It is possible to make such an analysis by cohort of the 1910 Berlin
census. This is shown in Table 3.5. In 1910 the Berlin census asked for
the last place of residence prior to migration. This makes it possible to
compare those migrants who moved directly from their place of birth to
those who made an intermediate move.
The Wrst point to note is the decline in the numbers of moves from
the areas nearest to Berlin, especially the provinces of Brandenburg,
Prussian Saxony, and Anhalt. These provided over half the migrants in
the pre-1870 period, but only a third after 1900. There is a shift in
favour of the eastern provinces in the period to 1890, but thereafter the
east’s share declines. By 1906–10 migrants were drawn much more
evenly. There were rises in the share contributed by the west and
north-west, and by the other central provinces, and there was also a rise
in the foreign contingent.
There is a problem here: return migration could have the eVect of
boosting the east’s share over time, as eastern migrants were less likely
to return or make further moves. Thus, the eastern share will tend to
rise for cohorts furthest from the census date. Fortunately, there are
other ways to establish the trend of net migration and these will be
used in a later section.
Looking at the balance between migrants who moved from their
place of birth, and those who had already made at least one move, it is
68 The Pattern of Migration, 1870–1913
![Page 78: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/78.jpg)
clear that there is a considerable change. Over 70% of the pre-1870
cohort were on their Wrst move. By 1906–10 this had dropped to below
30%. There is also a shift in the source of migrants, from smaller towns
and rural areas to cities (the census data makes it possible to group
migrants from 87 major towns and cities, the Berlin suburbs not
Table 3.5. Analysis of the 1910 Berlin population by arrival cohort:
percentage distribution of those born outside the city by last place of
residence prior to migration (both sexes)
Date of arrival in the city
1906–
10
1901–
5
1896–
1900
1891–
5
1886–
90
1881–
5
1876–
80
1871–
5
pre-
1870
All moves:
Regbez. Potsdam 16.5 16.0 16.4 15.2 14.2 16.4 17.3 15.8 22.4
Other Brandenburg/
Pr. Saxony/Anhalt
16.7 17.9 19.5 19.3 19.5 22.3 23.5 23.4 28.1
Other Centre 8.3 7.0 6.2 5.4 4.8 5.1 4.6 3.7 3.4
total centre 41.5 40.8 42.0 39.9 38.5 43.8 45.5 42.9 53.8
East 37.5 45.5 46.7 49.2 53.1 48.4 47.5 51.3 41.4
West and
north-west
10.6 6.9 5.8 5.9 4.4 4.5 4.0 3.0 2.9
South and
south-west
3.1 2.1 1.7 1.8 1.4 1.2 1.2 0.9 0.5
Foreign 7.3 4.8 3.7 3.2 2.6 2.1 1.8 1.9 1.4
Moved from place of birth:
total 40.6 44.4 48.2 49.7 53.0 55.8 60.7 63.7 71.1
From towns and
rural areas
29.2 34.4 39.4 41.1 44.8 47.7 52.8 56.5 61.9
From cities 8.7 8.1 7.3 7.3 7.2 7.3 7.1 6.4 8.6
Moved from somewhere other than place of birth:
total 59.4 55.6 51.8 50.3 47.0 44.2 39.3 36.3 28.9
From towns and
rural areas
36.9 35.0 33.8 32.3 30.5 28.8 25.7 24.2 19.3
From cities 18.0 17.7 15.8 16.1 15.0 14.1 12.6 10.9 8.9
Source: These calculations were made using material from microWlm tapes kindly madeavailable by Professor Dudley Baines, which are now in the library of the London Schoolof Economics. These contain the full published record of the Berlin censuses 1861–1910.The results were summarized using the regional classiWcation set out in Appendix 2C.The data is drawn from table V.3 of the 1910 census, Berlin Statistisches Amt (1913–16),i. 32–9.
The Pattern of Migration, 1870–1913 69
![Page 79: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/79.jpg)
incorporated in the city, and the industrial regions of Zabrze and Bor-
beck). But it should be noted that these urban areas only contributed
27% of migrants in 1906–10; the majority still came from rural areas
and small towns.
Comparison with the results of earlier censuses is made diYcult by
the fact that these only recorded place of birth, not last place of resi-
dence. It is possible to take encouragement from the Wnding that most
of those who arrived in the earlier years had come from their place of
birth, but this is still an obvious source of discrepancy between the two
data sources. Table 3.6 provides an analysis of the population resident
in 1890, by place of birth and date of arrival. This shows a steady rise
in the proportion coming from the eastern provinces. In the 1870s and
1880s about half of all migrants came from this area. At the same time
there was decline in the proportion coming from the area around
Berlin, which is particularly marked in the case of the nearby Kreise:Teltow, Niederbarmin, and Osthavelland.
At this stage it would be premature to attempt to draw Wrm conclu-
sions, but the pattern which is emerging suggests that the German
Table 3.6. Cohort analysis of the Berlin population 1890: percentage
distribution of those born outside the city by place of birth
(both sexes)
Date of arrival in the city
1881–
90
1871–
80
1861–
70
1851–
60
1841–
50
1831–
40
1821–
30
pre-
1820
Nearby Kreise 6.3 6.4 7.4 10.2 12.3 16.4 20.0 22.6
Rest of Regbez.
Potsdam
9.4 11.2 13.5 16.1 17.8 20.7 21.5 20.9
Other Brandenburg/
Pr. Saxony/Anhalt
20.7 23.1 26.6 30.8 32.0 30.3 28.1 24.8
Other Centre 3.5 2.6 2.3 2.5 3.2 3.0 4.2 5.7
TOTAL CENTRE 39.9 43.3 49.8 59.6 65.3 70.4 73.8 74.0
East 49.3 50.3 45.0 34.9 28.6 24.2 20.4 15.2
West and north-west 6.2 3.9 3.5 4.1 4.3 3.8 3.3 5.1
South and south-west 1.6 0.8 0.5 0.5 0.8 0.7 0.9 2.4
Foreign 3.1 1.7 1.2 0.9 1.0 0.9 1.3 4.9
Source: Calculated from data in Berliner Statistisches Amt (1893), table V.2; the regionalclassiWcation is set out in Appendix 2C.
70 The Pattern of Migration, 1870–1913
![Page 80: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/80.jpg)
experience of migration during the period of industrialization can be
divided into three periods:
1. Early (up to the 1860s): low mobility, most migrants drawn from
the regions nearest to the cities.
2. Middle (from the 1860s to the 1890s): higher mobility, and a
major shift of population out of the rural east.
3. Late (after the mid-1890s): a highly mobile, Xexible labour
market, but one less dominated by the absorption of a labour
surplus from the rural east.
The cohort analysis made by the Berlin statistical oYce was also used
to examine the extent to which migrants from diVerent regions made
further moves after arrival. The basic requirement was an assessment of
the probable death rate of a given cohort. After this had been deducted,
a comparison of cohort numbers at diVerent census dates gave the
percentages of those who remained and those who moved again. This
was not an estimate of return migration as such, but return migration
formed a part of the total of those moving. Ideally it would have been
desirable to adjust for diVerences in likely death rates, but this infor-
mation was not available, and the statistical oYce applied the same
death rate to all groups (though diVerent rates were applied to each
sex). Some error would have resulted from this. If migrants from the
east, having lower wages and poorer living conditions than others, had
higher death rates then there will be a tendency to overestimate the
numbers moving: some of those assumed to have moved having, in fact,
died.14
Table 3.7 provides a summary of this analysis, using the 1875, 1880,and 1890 censuses. The table shows that those who moved over short
distances, coming from the adjacent areas of Brandenburg and Prussian
Saxony, were less likely to make further moves. Of those who moved
from more distant areas, those coming from the eastern provinces have
much lower rates of further migration when compared to the rest of
Germany. Two other points which emerge are the high rates of further
migration for foreigners and the lower rates for women compared to
men.
14 The view that the poor had higher death rates is well supported: see Spree (1988) forexample. There is no evidence that eastern migrants had lower wages; nevertheless, givenlower wage levels in the region of origin it can be assumed that reservation wage levelswould be lower for eastern migrants facing similar costs of migration, and it is thereforeprobable that wages would have tended to be lower.
The Pattern of Migration, 1870–1913 71
![Page 81: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/81.jpg)
This may appear a surprising result. The likelihood of a temporary
move could be expected to be inversely related to distance: individuals
might make short moves intending to return, to take advantage of un-
usual short-term employment opportunities, but over long distances the
cost of the move would mean that most would be intended to be per-
manent. Instead the reverse seems to be the case.
It might be that short-distance migrants are better informed, and
therefore less likely to make mistakes, but examination of Figure 3.4,
which gives the rate of further moves in 1880–90 by region, shows that
there is no simple relationship with distance: the more distant provinces
such as East Prussia, Alsace-Lorraine, and the Rhineland show very
diVerent results.
Figure 3.4 suggests that there are two factors which aVect the like-
lihood of further moves. The Wrst is what might be termed ‘cultural
aYnity’: migrants from Alsace-Lorraine and from the Catholic areas of
southern Germany were less likely to make permanent moves to Berlin.
The same point applies to foreigners. The second is the income level of
the region of origin. Migrants from the richer western and north-western
provinces (Schleswig-Holstein, Hanover, Rhineland, and Westphalia)
were more likely to make further moves. So this analysis shows that there
was a fundamental diVerence between migration from the east and
Table 3.7. Estimates of the numbers making further moves after
arrival, Berlin 1875–80 and 1880–90
% of each migrant cohort who move away from Berlin
Arrived pre-1875,
moved in 1875–80
Arrived pre-1880,
moved in 1880–90
men women men women
Arrived from:
Brandenburg/
Pr. Saxony
25.3 15.4 26.0 21.8
East 27.6 16.8 25.4 19.0
Rest of Germany 43.6 23.4 46.4 32.1
Foreigners 51.0 35.6 53.3 44.7
Source: Calculated from data in Berlin Statistisches Amt (1893), 44.
72 The Pattern of Migration, 1870–1913
![Page 82: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/82.jpg)
migration from other areas: the Wrst was a permanent shift of a popula-
tion surplus in the sense that it represents the rectiWcation of an imbal-
ance between the available labour force and perceived relative
opportunities in the region of origin; the second was part of a general
pattern of rising mobility.
The east was the principal source of surplus labour for the Berlin
labour market in the period. But there were other streams of long-
distance migration, seasonal as well as permanent. Figures 3.5 and 3.6
illustrate this by making use of a survey of local authorities in the 1860s
(at the Kreis level or equivalent). The administrators were asked about
migration to or from their district. Many replied in general terms
(‘much movement at harvest time’ or such like), but some identiWed
particular destinations or sources, and sometimes gave reasons as well.
>700
600−700
550−600
500−550
450−500
400−450
350−400
300−350
250−300
200−250
Berlin
Fig. 3.4. Estimated numbers making further moves in 1880–90 by region:
numbers moving again per 1,000 of those arrived in Berlin pre-1880 from each
region (men only)
Source: As for Table 3.7.
The Pattern of Migration, 1870–1913 73
![Page 83: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/83.jpg)
The survey was not a comprehensive one, but it does provide a useful
picture of movements in Germany at the time.15
Figure 3.5 gives some examples of long-distance migration connected
with agricultural activities. It shows that, even in the 1860s, Germany
was attracting foreign seasonal labour at harvest time, although in
this period the main sources were Bohemia, Sweden, and Italy. There
were also harvest-related movements inside Germany though these were
mainly over relatively short distances. Mostly the movements were from
15 The authorities were not asked about the means of transportation used by thesemigrants. Some of these movements were suYciently short to have been possible on footor using horse-drawn transport. Others were so long as to imply the use of the railwaynetwork.
Italy - Harvest and brick-making
Bohemia- Harvest
Eisenach
Holland- Harvest
Sweden - Harvest
Peat-cutting
Sugar beet
Peat-cutting
Fig. 3.5. Examples of long-distance migration connected with agricultural
activities
Source: Examples taken from survey replies given in von der Goltz (1875). There wassubstantial harvest-related migration within Germany over shorter distances; Bertold et al.(1990), 268 has a map of these movements.
74 The Pattern of Migration, 1870–1913
![Page 84: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/84.jpg)
east to west, though there were some exceptions, such as the seasonal
migration from the Eisenach region to eastern Silesia at harvest time.
The sugar beet areas of central Germany were attracting seasonal mi-
grants, although the authorities indicated that peat-cutting in these
regions was also an important reason for migration to these areas.
Figure 3.6 gives some examples of migration connected with industrial
or urban activities. It shows, Wrst, those regions which speciWcally men-
tioned outward migration to Berlin or the Ruhr. The latter area was
mainly drawing migrants from western or central Germany in this period
(large-scale migration from the east came later). Berlin was attracting
migrants from eastern Brandenburg, and parts of Silesia and Posen.
These were, it appears, relatively concentrated streams, from a limited
number of areas, compared to the more widespread eVect of railway
construction. The Wgure shows the districts reporting major movements
Areas reportingmigration to railwayconstruction
Fig. 3.6. Examples of long-distance migration connected with industrial or
urban activities: areas reporting moves to Berlin or the Ruhr; areas reporting
migration to take part in railway construction
Source: As for Figure 3.5.
The Pattern of Migration, 1870–1913 75
![Page 85: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/85.jpg)
to railway construction. These were in predominantly rural areas in the
south, the centre, and in the east. Comparison with the previous Wgure
suggests that railway construction may well have Wtted in with the pre-
industrial pattern of seasonal migration at harvest time: young people
from regions accustomed to provide harvest labour to other areas went to
the railways instead. By contrast, moves to cities represented a more
fundamental change in lifestyle and social surroundings.
The transition from a pre-industrial to an industrial pattern of migra-
tion was not a sudden change for the supplying regions: there were
several stages. Regions might start with a tradition of harvest-time out-
migration, then move on to temporary migration connected with
railway building or other major construction projects. It was then a
possibility that a portion of those who made temporary moves for these
reasons might decide to make permanent moves. This in turn would
establish migrant cores, which could then expand as those making the
initial moves were joined by friends or other family members.
5. D i fferences in M igration by Sex
Many of the statistics on migration are available for both sexes. In this
study these are amalgamated unless there is some evident diVerence
which might be of interest, such as that shown in Table 3.6. Otherwise
every table would be twice as long.
There are indeed diVerences in the migration record between the
sexes. In general men moved more. In 1901–11 in Berlin there were
annually, on average, 301 moves per 1,000 for the male population, but
only 222 moves per 1,000 for the female population (total moves to and
from Berlin, Wgures adjusted by the statistical oYce). But this higher
rate of gross migration did not mean that net migration was necessarily
less for women. In fact in 1891–1912 the female population of Berlin
gained proportionately more through migration: an average annual gain
of 0.46% against one of 0.39% for the male population. Looking at the
country as a whole, and dividing it into 17 fairly large regional group-
ings, the 1907 census found that 14.7% of men were living outside their
region of birth, and 12.4% of women. This is a signiWcant diVerence,
but one which shows that women were by no means immobile, even if
they moved less often than men.16
16 Calculations from the analysis of the 1907 Occupational Census published in SJDS(1909).
76 The Pattern of Migration, 1870–1913
![Page 86: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/86.jpg)
Male migration was more volatile. Figure 3.7 compares annual
net migration for each sex for Berlin during the period 1891–1912.
The magnitude of the swings in male migration is clearly greater,
and evidently related to economic conditions. For example, the so-called
Elektroboom of 1898–1900, which was particularly marked in Berlin, pro-
duced a larger male inXow, and the collapse of this boom in 1901–2 led to
a net outXow of men, but not to a net outXow of women.
Looking at the historical record, there is evidence to show that
female mobility rose more slowly than male. Figures 3.8 and 3.9 make
use of the cohort analysis of the Berlin population in 1890 and 1910
given in Tables 3.5 and 3.6. In the Wrst of these, the proportion of
migrants who came from the central provinces is compared for both
sexes; in the second the analysis is repeated for the east. Two sets of
lines are shown. As mentioned in the discussion in section 4 above, the
census results are not comparable: the 1910 census classiWes by last
place of residence prior to migration; the 1890 census classiWes by place
of birth. So there are separate lines (A and B) for each census.
The Wgures show that in the Wrst half of the nineteenth century there
was a marked diVerence between the two sexes: female migrants were
much more likely to be drawn from the central provinces, and less
likely to come from the east. Female migration was predominantly
191119071903189918951891
30
20
10
0
−10
−20
−30
Men
Women
Fig. 3.7. Net migration of men and women, Berlin 1891–1912, per 1,000
head of population of each sex (moves out adjusted by the statistical office)
Source: Calculated from data in SJB (1913), 158.
The Pattern of Migration, 1870–1913 77
![Page 87: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/87.jpg)
short-distance. As migration from the east rose, female migration lagged
behind male migration, reaching the 30% mark about 20 years later, for
example. But by the 1860s the gap had narrowed considerably. And
from 1870 onwards there is no tendency for more female migrants to
come from the central regions, and the eastern proportion is rather
higher for female migration than for male.
In general, the diVerences between male and female migration are not
so large after 1870, at least when net migration is the main focus of
interest. The main diVerences are in the pre-1870 period, and in the
fact that men moved more often, and thus contributed a higher propor-
tion of total moves. Young men made more frequent moves than young
women, but women were nearly as likely to make permanent changes of
residence.
1905189518851875186518551845183518251815
80
70
60
50
40
30
Men (A)
Women (A)
Women (B)
Men (B)
Fig. 3.8. Percentage of migrants drawn from the centre, by arrival cohort,
Berlin 1815–1910
A 1890 census results classified by place of birth
B 1910 census results classified by last place of residence
Source: Figures given in Tables 3.5 and 3.6.
78 The Pattern of Migration, 1870–1913
![Page 88: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/88.jpg)
6. German Emigration
The Lewis Model deals with the operation of an economy seen in
isolation, without considering the geographical context: How easy is
migration across international frontiers? What are the conditions in
neighbouring countries? Are wages in these countries lower or higher?
In the model, the urban wage is determined by conditions in domestic
agriculture: by the existence, or non-existence, of a labour surplus in
this sector. But if there is migration from neighbouring countries with
lower wages, then the whole wage structure will be inXuenced by condi-
tions in these countries. And if there is substantial out-migration then
1905189518851875186518551845183518251815
60
50
40
30
20
10
Men (A)
Men (B)
Women (B)
Women (A)
Fig. 3.9. Percentage of migrants drawn from the east, by arrival cohort, Berlin
1815–1910
A 1890 census results classified by place of birth
B 1910 census results classified by last place of residence
Source: Figures given in Tables 3.5 and 3.6.
The Pattern of Migration, 1870–1913 79
![Page 89: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/89.jpg)
the internal labour market will be aVected by developments in the
recipient countries.
German external migration underwent a major structural change
around 1895. Comparison of census results with recorded births and
deaths for the territories included in the post-1871 German empire
show a total net loss of 2.34 million in 1841–71 (an annual average of
78,100) and 2.46 million in 1871–95 (an average of 102,500). But after1895 this Xow came to an end. Between 1895 and 1910 the net loss was
just 14,000, an annual average of less than a thousand.
Calculations based on census returns may underestimate the extent of
the change. By 1910 Germany was admitting large numbers of foreign-
ers on temporary permits. Holders of these permits were required to
leave the country by December, and then re-apply if they wished to
continue to reside in Germany. As German censuses were carried out
in December, those who complied with this regulation would have been
omitted from the census total.17 A census taken in some other month
would have shown a larger population, and the calculated rate of migra-
tion into Germany would have been greater. Germany may well have
already been a net gainer through migration.18
If so then this was an unusual development. Of the other European
countries, only Belgium was a net gainer before 1914. Britain did not
start to gain population through migration until the 1970s. This change
had the eVect of reinforcing the links between Germany and the other
labour markets of central and eastern Europe, and decoupling Germany
from the converging labour markets of the ‘Atlantic economy’.
Discussion of German emigration in the nineteenth century is dom-
inated by German emigration to the United States. Partly this is due to
the nature of the statistical sources. There are records of German over-
seas migration, both from the ports of departure and from the recipient
countries, but there is very little on movements across land frontiers. In
1912 the German statistical oYce made a survey of foreign censuses, to
establish the numbers of Germans living outside the Reich. The Wgures
are given in Table 3.8. The statistics are not fully comparable: some
countries recorded all those born in Germany; others only included
those who still retained German citizenship. Some countries which lack
suitable census data are excluded: Brazil is the most important of these.
17 Those who chose not to comply with the regulation would have had a good reason toevade the census.
18 Return migration was also important: see Schniedewind (1994) for a study of this.Other studies of German migration include Bade (1980 and 1991) and Marschalck (1973).
80 The Pattern of Migration, 1870–1913
![Page 90: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/90.jpg)
The pre-eminence of the United States as an emigrant destination is
evident. But quite large numbers went to neighbouring states, although
these in return provided migrants to Germany, reducing net migration.
The most substantial Xow of net migration out of Germany to a
European destination was to Switzerland (there were 68,000 Swiss
living in Germany in 1910).
The United States accounted for 91% of recorded overseas migration
in 1871–1913. Figure 3.10 shows that there were years when this fell to
80% or below. Typically these were years when migration to the
United States was low, but there is no evidence that migrants chose
other destinations when conditions were unfavourable for migration to
the United States (migration to the United States was not negatively
correlated with Xows to other destinations).
Towards the end of the period there is some evidence that German
migrants were looking for alternative destinations. There was a rise in
migration to Canada and Argentina. However, the Xows were not large
compared to the movements to the United States in the period up to
1895. By contrast, after 1900 British emigration switched to the
Empire, which took 65% of those leaving in 1911–13 (up from 28% in
1891–1900).19
Comparison of migrant destinations with the regional origins of mi-
grants is possible from German Wgures from 1899 onwards. Figure 3.11
gives migration to the United States as a percentage of all recorded
migration to non-European destinations, 1900–9. The main areas from
which there was migration to alternative destinations were a group of
central regions—Saxony, Prussian Saxony, and Berlin-Brandenburg.
19 British Wgures from Thomas (1973), 57; the German Wgures are calculations usingWgures for migrant destinations from SJDR.
Table 3.8. Germans living in foreign countries 1910–11
United Statesy 2,501,000 Britain and
Ireland y56,000 Argentina 27,000
Switzerland 220,000 Netherlands 38,000 Luxembourg 22,000
Austria-Hungary 126,000 Denmark 35,000 Chile 11,000
France 90,000 Australiay 33,000 Italy 11,000
Belgium 57,000
yCountries who recorded all those born in Germany, all others recorded only Germancitizens.Source: Kaiserlich Statistisches Amt (1916), 3*.
The Pattern of Migration, 1870–1913 81
![Page 91: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/91.jpg)
This group includes some important industrial centres. At the other
end of the scale, in the agricultural regions of eastern and southern
Germany there was little migration to destinations other than the
United States. The result is that there is a positive correlation between
the percentage occupied in agriculture in each region, and the percent-
age going to the United States (a coeYcient of 0.56), indicating that the
Xow to the United States was more ‘agricultural’ than the Xow to other
destinations.20
100
80
60
40
20
01871 1876 1881 1886 1891 1896 1901 1906 1911
Asia
Africa
Australia
Other North andSouth America
Brazil
United States
Fig. 3.10. Percentage distribution of recorded migration to non-European
destinations (from German sources) 1871–1913
Note: The German figures initially covered only migration from German ports andAntwerp. Later this was expanded to include migration from French ports. Migrationfrom British ports was not included, which affected the figures for migration to Australia.Source: Data from SJDR (1885), table 5, p. 23 and subsequent issues.
20 The negative correlation between the percentage going to Brazil (the next most im-portant destination) and the percentage occupied in agriculture is even stronger (a coeY-cient of �0.72). This Wnding that the ‘more advanced’ regions sent migrants to the ‘lessadvanced’ destination is intriguing although perhaps not of any great signiWcance. It has aparallel in the better-known division between northern and southern Italy (the south wasmore likely to send migrants to the United States), Hatton and Williamson (1998), ch. 6.
82 The Pattern of Migration, 1870–1913
![Page 92: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/92.jpg)
However, the basic conclusion has to be that the story of German
emigration is, in aggregate, the story of German migration to the
United States. There was no other Xow which was large enough to have
a signiWcant eVect on labour market conditions inside Germany. The
record of migration to the United States is presented in Figure 3.12.
Only moves from Germany to the United States are given; there are no
Wgures for return migration before 1909.
The chart includes two series for German migration to the United
States, one from German sources and the other from US immigration
records. The two sources agree relatively well. Between 1890 and 1900there is little diVerence. Before 1890 the discrepancy is probably due to
gaps in the German records. After 1900, the diVerence is less easy to
explain. The US authorities may have classiWed a number of German-
speaking immigrants from Russia as Germans. This could explain why
<86%
86−88%
88−90%
90−92%
92−94%
94−96%
96−98%
>98%
Fig. 3.11. Migration to the United States as a percentage of all migration to
non-European destinations 1900–9, by region
Source: Calculated from data in SJDR (1901–10).
The Pattern of Migration, 1870–1913 83
![Page 93: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/93.jpg)
the discrepancy rose after 1900. Some travellers, intending to stay for a
limited period, may also be classiWed as immigrants in the US Wgures.21
The migration Wgures show three main peaks: in the early 1850s, in
the late 1860s and early 1870s, and in the period 1881–93. After this
German emigration to the United States remained at a low level. These
changes can be linked to the relative movement of the German and US
economies: emigration was high when conditions were favourable in the
United States or unfavourable in Germany. But after 1895 this relation-
ship becomes much weaker: the collapse of the Elektroboom in 1900–1
produced only a modest rise in emigration, for example.22
1910190018901880187018601850184018301820
300
200
100
0
Emigration:US data
US−Germanwage ratio
Emigration:German data
Fig. 3.12. Annual German emigration to the United States (in 1,000s) and the
US-German wage ratio
Source: Migration figures from USDepartment of Commerce (1975), and SJDR, annualissues; wage ratio calculated fromWilliamson (1995), 164–7. The source for the German wagedata is a series for unskilled building wages in Rostock, Nuremburg, and Berlin, originallycollected by Robert Kuczynski, also published in Kuczynski (1945) and Bry (1960).
21 There are large discrepancies in statistics of German migration 1900–13 for all themajor non-European destinations. These are discussed in Burgdorfer (1926). One problemwas German emigration through British ports, but there is no obvious reason why thisshould only have caused diYculty after 1900.
22 Grant (2002d ) contains a detailed analysis of the decline of German emigration tothe United States.
84 The Pattern of Migration, 1870–1913
![Page 94: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/94.jpg)
Figure 3.12 also gives the US–German wage ratio, calculated from
Williamson’s series in purchasing power parities. There was a major
disturbance to this series in the 1860s, as a result of the American Civil
War. Once allowance is made for this, it is clear that there is no evident
tendency for wages in the two countries to converge. Germany was an
exception to the general tendency for wages in the Atlantic economies
to move closer. One reason for this was that declining emigration to the
United States, combined with increasing immigration from eastern and
central Europe, had the eVect of strengthening links with other over-
supplied labour markets in these regions, and weakening those with the
converging Atlantic economies.23
German emigration to the United States declined for a number of
reasons, due to changes within Germany and within the United States.
One factor was increased competition within the United States from
sources of ‘new immigration’: countries in southern and eastern Europe
which supplied an increased proportion of US immigrants from the
1890s onwards.24 Another was a reduction in the availability of land in
the United States which suited German farming skills as the frontier of
settlement moved into more arid western regions. But a further possible
reason is that Germany was moving out of the labour surplus phase
when migration, internal and external, was dominated by the release of
surplus agricultural labour.
German emigration had a strong agricultural character. Figures 3.13and 3.14 show the variation in the intensity of migration by region, for
two periods: 1871–9 and 1890–9. The regional rates of emigration (per
head of resident population) are expressed as percentages of the national
average. The variation between regions is striking: in 1871–9 the aver-
age annual rate of emigration was 0.44 per 1,000 in East Prussia, and
4.53 in West Prussia, ten times as large. In the 1870s the highest rates
of migration were in the rural east, in West Prussia, Mecklenburg, and
Pomerania. There were also high rates in the north-west, and in
the south-west, which had been an area of high emigration earlier in the
23 Important studies on convergence in the international economy of the late nine-teenth century include O’Rourke et al. (1994 and 1996) and Taylor and Williamson(1997). Although Germany contributed large absolute numbers of migrants in the yearsbefore 1895, the numbers were not as high relative to the size of the German populationas for Ireland and some Scandinavian countries. This meant that the convergence processwas not a strong one even in the years when emigration was high.
24 In the literature on US immigration these are the ‘new immigration’ sources, todistinguish them from the traditional sources: Britain, Ireland, Germany, and the Scandi-navian countries.
The Pattern of Migration, 1870–1913 85
![Page 95: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/95.jpg)
century. The pattern was not greatly altered by the 1890s: the highest
rates were then in Posen, West Prussia, and Pomerania, the
Mecklenburg rate having dropped.
There is a relationship between migration intensity and the percent-
age of the population engaged in agriculture. Rates of migration tended
to be below average in the more industrial regions, as can be seen from
the maps: Westphalia, Saxony, Silesia, the Rhineland, and Berlin-
Brandenburg all had low rates of emigration. Table 3.9 gives correlation
coeYcients for various periods, which show that this association
between high emigration and high rates of agricultural employment was
at its highest in 1885–94, and thereafter tended to fall.
Although German migrants came largely from agricultural regions
the statistical sources show that they were not predominantly occupied
as farmers or farmworkers immediately prior to departure. Table 3.10
250−300
200−250
150−200
125−150
100−125
80−100
60−80
40−60
20−40
<20
>300
Fig. 3.13. Regional emigration rates as percentages of the national average,
1871–9
Source: Calculated from data in SJDR (1881), table 5, pp. 14–15 and subsequent issues.These include all recorded migration to non-European destinations. The earlier figuresomit migration through French ports, which may affect the rates from some westernprovinces.
86 The Pattern of Migration, 1870–1913
![Page 96: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/96.jpg)
compares the previous occupations of German immigrants with those of
other groups. This shows that relatively few German immigrants had
previously been occupied in agriculture.25 The proportions were, how-
ever, rather higher than for the other ‘old migrant’ groups.
25 Marschalck (1973), 79–80 gives the Wgures from German sources; see alsoMarschalck and Kollmann (1973), 537.
Table 3.9. Correlation between rates of emigration by region
and the percentage occupied in agriculture in each region, for
diVerent periods, 1871–1913
1871–4 0.54 1885–9 0.59 1900–4 0.53
1875–9 0.47 1890–4 0.61 1905–9 0.45
1880–4 0.51 1895–9 0.48 1910–13 0.32
Source: Annual data from SJDR for recorded immigration by province.
250−300
200−250
150−200
125−150
100−125
80−100
60−80
40−60
20−40
<20
>300
Fig. 3.14. Regional emigration rates as percentages of the national average,
1890–9
Source: Calculated from annual data in SJDR.
The Pattern of Migration, 1870–1913 87
![Page 97: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/97.jpg)
However, this may be misleading. If there is surplus labour in agricul-
ture, then the Wrst to move will be the younger sons of farmers, and the
children of landless labourers. These will be unable to Wnd work in
agriculture, and may well start by taking non-agricultural jobs in their
region of birth. But if they are determined to work in agriculture
somewhere, then the realization that only emigration to a country where
land is readily available oVers this possibility will eventually make them
decide to move. However, the records will show that these are people
whose occupations prior to migration were outside agriculture, in a rural
handcraft sector or as unskilled non-agricultural labourers.
The recorded occupations of migrants after arrival in the United States
also show that relatively few found work in agriculture immediately. But it
does appear from the US census returns that German migrants had a
stronger preference for farming than some other groups, notably the
British. Table 3.11 gives Wgures for the percentage of the male labour
force who were classiWed as farmers by the United States. As this shows,
by 1910 German migrants were twice as likely as British migrants to be
farmers, having started from a position of equality in 1870.
The table also shows that only a relatively small proportion of the
German-born population were farmers, a rather lower Wgure than the
Table 3.10. Occupations of immigrants to the United States prior
to departure (as percentages of migrants 1899–1910 excluding those
without an occupation)
Irish Scandinavian German English
Professional 1.5 1.3 3.9 11.3
Farmers 0.3 2.5 3.0 2.4
Skilled workers 13.4 21.5 35.0 53.9
Unskilled: farm workers 4.8 7.6 18.0 2.5
Unskilled: non-agricultural 80.1 67.0 40.2 29.8
Note: The table gives results for the four largest ‘old migrant’ groups. Amongst the‘new migrant’ groups the Poles, Slovaks, and Croats had higher proportionspreviously occupied in agriculture than the Germans (‘nationalities’ are as classiWedby the US immigration authorities, which means that Polish-speaking migrants aretreated as a separate group).Source: Calculated from US Immigration Commission (1910–11), 100–76; thedistinction between skilled and unskilled is largely that used by the immigrationauthorities; the subsequent division into categories (industrial, handicrafts, etc.)follows the author’s classiWcation.
88 The Pattern of Migration, 1870–1913
![Page 98: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/98.jpg)
one for the whole US labour force. Even if the 4.9% of the 1910
German-born workforce who were farm-workers is added to the
numbers for farmers, the total is only 22.3%, which is well below the
equivalent Wgure for the United States as a whole (35.1%). Most
German migrants were in non-agricultural jobs.
However, once again, this comparison may well lead to an under-
estimate of the importance of agricultural opportunities to German
migrants. A higher proportion of immigrants may have become farmers
eventually. A typical immigrant life-cycle might include an initial
period of non-agricultural work to pay oV debts incurred during the
passage and to accumulate capital before applying for a Homestead land
grant. Some were obliged, by the terms under which they had travelled,
to take non-agricultural jobs to pay oV debts to steamship companies
and their agents.
The sudden decline in German emigration around 1895 Wts with the
picture given earlier, of a ‘labour surplus’ phase of migration out of the
rural east which came to an end in the late 1890s. Migrants who wanted
to remain in Germany went to the major cities. Those who were more
adventurous, or determined to go to a country where land was available,
or had contacts with Germans already in the United States, went across
the Atlantic. This trans-Atlantic ‘safety valve’ reduced the pressure on
urban labour markets quite considerably in 1880–5. In 1880 around
55% of those leaving the eastern provinces of Germany went abroad.
By the early 1900s this had changed: in 1900–5 less than 10% of those
leaving these regions left the country.26
Table 3.11. Farmers as a percentage of the US male labour force
1870–1910
1870 1880 1890 1900 1910
As a % of the total US labour force 27.8 28.3 27.1 22.1 19.5
As a % of those born in Germany 21.5 24.3 21.2 20.6 18.4
As a % of those born in Britain 20.2 19.8 14.7 13.1 9.0
German–British ratio 1.06 1.23 1.44 1.57 2.04
Source: Census results from US Department of Commerce (1975) and Hutchinson (1956);the 1870 Wgures are estimates using the movement of the total German-born populationengaged in agriculture. (Changing census deWnitions may account for some of the jumpsin the Wgures, but the relative movements should be unaVected.)
26 Calculated indirectly from a comparison of natural population movements, recordedpopulation, and registered emigration by region using Wgures published annually in the
The Pattern of Migration, 1870–1913 89
![Page 99: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/99.jpg)
7. Immigration into Germany 1870–1913
At the same time as German emigration was declining there was an
increase in immigration into Germany. The German population census
showed a substantial increase in the numbers of foreigners living in
Germany, from 433,000 in 1890, to 779,000 in 1900 and 1,260,000 in
1910. This indicates that the net annual rate of immigration was around
41,000–46,000 in the 1890s and 57,000–64,000 in the 1900s.27 But the
census was taken in the month of December, and this would tend to
aVect the numbers since migrants on short-term permits were expected
to return home in this month. The occupational census of 1907, taken
in June, gave the numbers of foreigners living in Germany as 1,342,000,
which was greater than the Wgures from the December censuses for
1905 (1,030,000) and 1910 (1,260,000). But even this may have been
below the peak, which is likely to have come in the harvest months of
August and September.
The total numbers who came to Germany on temporary permits
were considerable. Russian Wgures show that, on average, 755,000 tem-
porary eight-month passports were issued each year between 1910 and
1913 to those wishing to work in Germany and their families. Over the
same period, the German Wgures show that 484,000 work permits were
issued, on average, each year to Russians (including Poles and Germans
of Russian nationality), and 267,000 to other nationalities. Italian emi-
gration Wgures show that, on average, 69,000 Italians went to Germany
each year between 1910 and 1913, and the Germans issued 54,000 work
permits to Italians.28 So the numbers of workers, and their dependants,
coming to Germany on a temporary basis each year could have
exceeded one million.29 How long they stayed is a more diYcult ques-
tion. Some could have come for a few weeks, for a harvest job just over
Statistisches Jahrbuch des Deutschen Reichs. The resulting Wgures are approximate. Therewas some tendency to under-record external migration, but this has to be set against thefact that return migration was not recorded at all.
27 To get this estimate a mortality rate had to be applied, which should be lower thanthe German average to allow for the diVerent age structure of the migrant population.The range was produced by using, Wrst, a mortality rate half the overall German rate, andthen one 85% the overall rate. Data from SJDR (1913).
28 These were mostly seasonal workers; comparison of the Italian emigration Wgureswith the German census Wgures indicates that 90–95% of Italian migrants in 1900–10went back each year. Ferenczi and Willcox (1929), 810 and 828–31; see also DelFabbro (1996).
29 There might well have been a signiWcant amount of unrecorded migration fromother areas, from Bohemia for example.
90 The Pattern of Migration, 1870–1913
![Page 100: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/100.jpg)
the frontier; others may have stayed for the full eight months allowed
by the Russian passport.30
What is also uncertain is the extent to which this was a new develop-
ment, rather than the imposition of a system of passports and permits
on a Xow which was already in existence. Russian Wgures show a steep
increase in the numbers of passports, from an annual Wgure of around
100,000 in 1900. German Wgures show a more moderate rise in the
number of permits: 733,000 of these were issued in Prussia alone in
1907 (all nationalities); 768,000 in the whole of Germany in 1913.31
Surveys in the 1890s suggested that the use of seasonal migrant labour
in German agriculture was already substantial.32 It may well be that it
was the extension of this to German industry which was the novelty.
Table 3.12 gives the numbers of foreigners living in Germany
according to the December censuses. The Wgures show that the largest
contribution came from Austria-Hungary. The numbers of Russians
was relatively small (more came in the summer on temporary permits).
The Xow of migrants from Austria-Hungary has attracted less attention,
but it was also an important inXuence on the German labour market.
Table 3.12. Numbers of foreigners, by country of birth, resident in
Germany 1900, 1905, and 1910
1900 1905 1910
Russia 47.0 106.6 137.7
Austria-Hungary 391.0 525.8 667.2
Switzerland 55.5 62.9 68.3
Italy 69.7 98.2 104.2
France 20.5 20.6 19.1
Spain and Portugal 0.9 1.3 2.0
Belgium, Netherlands, and Luxembourg 113.5 127.6 172.0
Scandinavia 38.9 41.1 39.2
Britain and Ireland 16.1 17.3 18.3
Balkans including Greece and Turkey 4.1 5.8 8.1
Others 21.8 23.2 24.0
totals 779.0 1030.4 1260.0
Source: Kaiserlich Statistisches Amt (1916).
30 For a description of the work permit system, see Bade (1984), also Nichtweiss (1959).31 Burgdorfer (1929), 379; there are no Wgures for before 1905.32 Weber (1893).
The Pattern of Migration, 1870–1913 91
![Page 101: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/101.jpg)
Attention has also tended to focus on the Xow of labour into agricul-
ture. The 1907 occupational census showed that there were 288,300
foreign-born agricultural workers, 3.3% of the labour force. There were
rather more foreign industrial workers (501,000) and these were 4.6%
of the industrial labour force. The industries most dependent on for-
eign labour were mining and metals (7.5% foreign), quarrying (8.8%),
and construction (7.4%).33
Looking at the regional distribution of workers in these three indus-
tries, it is also noticeable that the highest concentrations of foreign
workers were not in the east but more in the south-west (Figure 3.15).
The Wgure for Alsace-Lorraine (27%) is perhaps something of an
anomaly, but there are also high Wgures for Baden (14.3%) and Wurt-
temberg (12.0%). By contrast the distribution of foreign agricultural
>1
1−2
2−4
4−6
6−8
8−10
10−12
12−14
14−16
>20
Fig. 3.15. Foreign-born workers as a percentage of all those occupied in mining
and metals, quarrying, and construction, 1907
Source: 1907 occupational census, SDR (1908), n.f. 210.
33 SDR n.f. 210 (1908).
92 The Pattern of Migration, 1870–1913
![Page 102: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/102.jpg)
workers is more or less as expected (Figure 3.16). The highest Wgure
was the 11.3% of the Mecklenburg agricultural labour force which was
foreign-born. Next came Prussian Saxony-Anhalt with 8.1%.
Although the conventional view may be that the typical pre-1914
migrant was a Russian Pole working in eastern agriculture, this shows that
a typical migrant was just as likely to be someone from Austria-Hungary
working in a quarry or on a building site in south-western Germany.
8. M igration and Labour Surplus
The previous section has stressed the growing links between Germany
and the labour markets of eastern Europe, particularly Russia and
Austria-Hungary. In combination with the decline in emigration to the
United States, which weakened the connection between Germany and
<1
1−2
2−4
4−6
6−8
8−10
10−12
12−14
14−16
>20
Fig. 3.16. Foreign-born workers as a percentage of all those occupied in
agriculture, June 1907
Source: 1907 occupational census, SDR (1908), n.f. 210.
The Pattern of Migration, 1870–1913 93
![Page 103: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/103.jpg)
the converging Atlantic labour markets, this development reinforced the
‘Continental’ nature of the German labour market, insulating it from
the forces driving the process of convergence in the Atlantic economies.
It also meant that one source of elastic labour supplies—the labour
surplus in German agriculture—was supplemented by another—migra-
tion from neighbours to the east. This in turn held back the increase in
German wages.
This was an inevitable consequence of the eastern movement of the
process of industrialization in Europe. Both inside Germany and out-
side Germany industrialization led to increased mobility, traditional
institutions which promoted immobility were eroded, and improved
communications made migration attractive and aVordable. In the last
quarter of the nineteenth century these same forces were unleashed in
the countries to the east and the south of Germany. This in turn set oV
a new wave of migration to the United States, displacing potential
German emigrants, and also increased migration across Germany’s land
frontiers to the south and the east.
A theme which runs through the chapter is the identiWcation of
periods within the 1870–1913 timespan covered by this study, and the
characterization of these periods. Evidence presented points to the im-
portance of the years around 1895 as a dividing point. After this exter-
nal migration is greatly reduced, and the rate of migration out of the
east falls. There is also a shift in the characteristics of the typical
migrant, from being someone with no previous experience of urban life,
to a person who had already moved at least once, and who was more
likely to have come from another city.
The importance of the 1895 caesura is demonstrated by Wgures which
will be presented in Chapter 9. These show that the Kuznets Curve,
the relationship between income inequality and economic development,
peaks around these years. The years up to 1895 are characterized by
rising inequality and high net migration: the labour surplus phase of the
Lewis Model. After 1895 there was a highly mobile labour market, but
static or falling inequality; Germany was moving out of the labour
surplus phase.
Appendix 3A. A Note on Regional Units
The regions used in pre-1914 German statistics are a source of some
diYculty in that they vary considerably in size, and the boundaries
94 The Pattern of Migration, 1870–1913
![Page 104: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/104.jpg)
Table 3A.1. Regions and regional groupings
Regional
grouping
Regional unit used
in this study
Includes the following
regional units used
in German statistical
publications
East East Prussia East Prussia
West Prussia West Prussia
Pomerania Pomerania
Posen Posen
Silesia Silesia
Mecklenburg Mecklenburg-Schwerin
and -Strelitz
Centre Berlin/Brandenburg Berlin and Brandenburg
Pr. Saxony/Anhalt Provinz Saxony and Anhalt
Thuringia Saxony-Weimar,
-Altenburg, -Meiningen,
and -Coburg-Gotha,
both Reub and both
Schwarzburgs
Saxony Saxony (Kingdom of)
West and north-west Schleswig-Holstein Schleswig-Holstein,
Hamburg, and Lubeck
Hanover/Oldenburg/
Brunswick
Hanover, Oldenburg,
Brunswick, and Bremen
Westphalia Westphalia, Waldeck,
and both Lippes
Hesse-Nassau Hesse-Nassau
Rhineland Rhineland
Hesse Hesse
South and
south-west
Bavaria excl. Pfalz
Pfalz
Baden
Wurttemberg/
Hohenzollern
Alsace-Lorraine
Bavaria excl. Pfalz
Pfalz
Baden
Wurttemberg and
Hohenzollern
Alsace-Lorraine
The Pattern of Migration, 1870–1913 95
![Page 105: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/105.jpg)
reXect a historical process: Prussian expansion leading to eventual
German uniWcation, not any Napoleonic striving for geographical logic.
Therefore, all the statistics in this study are presented in tables or
Wgures which use regional units designed to eliminate the smaller
regions. On occasions, these are then grouped into four main regional
categories: east, centre, west and north-west, and south and south-west.
The basic system is shown in Table 3A.1. In addition, wherever pos-
sible, Oldenburg is split: the separate enclaves of Lubeck and Birken-
feld being included in Schleswig-Holstein and Rhineland respectively.
For some purposes, chieXy agricultural, the term ‘east-Elbian’ is used:
here Berlin/Brandenburg is added to the other eastern provinces.
For the tables and maps presented in this study a certain amount of
additional work was required to produce results which conformed to
these boundaries. The Wgures could not be taken directly from oYcial
publications.
96 The Pattern of Migration, 1870–1913
![Page 106: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/106.jpg)
4
Migration in Germany 1870–1913:
A Statistical Analysis
1. Introduction : A Note on Methodology
and Source Materials
The study of migration has been one of the most active areas of
German historical research in recent years. This has produced a
number of excellent local studies, making use of regional archives.
These have described the eVects of large-scale migration: the process of
assimilation, the problems faced by rapidly expanding cities, the process
whereby an industrial workforce was created from a stream of rural
migrants.1 The response of the German state and the upper echelons of
German society to these problems has been examined, and, at times,
found wanting. Studies have repeatedly shown pre-1914 Germany to be
a society of high inequality, and limited social mobility.2
The aim of this book is not to produce another local study. It is
instead to try to grasp the process in its entirety: to put migration in a
theoretical context which explains the origins of the problems described
by the regional studies, and to answer the question ‘Why is migration in
this period so important?’ The Lewis Model, the concept of a ‘labour
surplus’ phase in economic development, provides such a context, and
the application of this model is therefore a central theme.
The remit of the study is therefore a wide one. The aim is to test
certain hypotheses about the causes and consequences of migration.
The level is the national one: not what was happening in certain cities
or regions, but the eVect on Germany as a whole. And, wherever pos-
sible, hypotheses have been expressed in quantiWable terms.
1 Examples include Barfuss (1985), Bolsker-Schlicht (1997), Del Fabbro (1996),Jackson (1997), Kollmann (1960), Kosters-Kraft (1998), Lourens and Lucassen (1999),Oberpennig (1998), Reich (1998), Wenneman (1997), Crew (1979), Catt (1986), and thevarious studies in Hoerder and Nagler (1995), Robler (1995), and Bade (1987). Studies ofmigration with a wider remit include Matzerath (1985), Caumanns (1994), Kollmann(1959), Tipton (1976), and Hochstadt (1999).
2 Kaelble (1986).
![Page 107: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/107.jpg)
This is not because this is regarded as the only way to approach the
subject, or as necessarily superior to other techniques of historical en-
quiry; it is because this is a type of study not currently available. The
idea is to produce a work which is complementary to the regional and
local studies: for example the collection published as People in Transit.3
Moreover, the main emphasis has been on the use of statistical sources
available in nineteenth-century government publications. This period
was one in which statistical enquiry was developing, both in scope, in the
areas covered by published statistical series, and in the use that was made
of the results. There was a belief that the proper use of statistics could
lead to a better understanding of social problems and, at times, produce
solutions. The classic case was the use of mortality tables to prove the
connection between breast-feeding and low infant mortality.4
The new German empire had an Imperial Statistical OYce, which
oversaw the collection of national statistics and the administration of
censuses. There were also state statistical oYces, which collected a lot
of the data and conducted enquiries of their own. The most important
was the Prussian Statistical Bureau, headed by Ernst Engel in the 1860s
and 1870s, but others also made interesting contributions: the Saxon
Statistical OYce made a series of studies of income distribution in the
1900s which were unusually comprehensive for the time. In addition,
all the major cities had statistical oYces, and typically published a
statistical yearbook. The most important of these was the Berlin Statis-
tical Bureau, whose director in the years before the First World War
was Robert Kuczynski.
The analysis of these materials used the techniques available at the
time: where causation was suspected a series was ranked by one vari-
able, and results given for a second variable. This could show a rough
correlation: if the eye running down the page could pick up a tendency
for the second variable to rise or fall. But regression analysis was not
used, and any relationship involving more than one explanatory variable
could not be studied.
The result is a vast amount of statistical material presented in a form
which is not easy to understand or digest. Taking one example: the
1907 Zeitschrift of the Prussian Statistical Bureau contained an article
by Dr M. Broesike on internal migration in Prussia which included
sixteen of these ‘visual correlation’ tables, some of which extended over
two or more pages, and contained as many as 489 observations (an
3 Hoerder and Nagler (1995). 4 Berlin, Statistisches Amt (1893), 82.
98 Migration 1870–1913: A Statistical Analysis
![Page 108: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/108.jpg)
example is shown as Figure 4.1).5 There were also a number of other
tables containing basic statistical material. This was a barrier to compre-
hension greater, in most respects, than the obstacle presented by the
terminology of modern econometrics.
So the approach used in this study has been to seek, in the Wrst place,
to achieve a thorough knowledge of the materials collected and made
available by the various statistical oYces, and secondly to subject these
to examination using modern econometric analysis.
For the general reader the statistical results have been summarized in
a series of tables which are included in the main text of this chapter,
5 Broesike (1907).
Fig. 4.1. Part of a ‘visual correlation’ table from Broesike (1907)
Note: This is part of a table covering 489 Prussian Kreise. Broesike has classified them bythe dominant farm size; the first column gives the percentage of the total agricultural areain holdings of this size. He is looking for a relationship between this and rates of migrationfor 1895–1900 and 1900–5 (the two columns at the right). He also includes a classification(Kb or Gb) according to the next most important farm size class, and follows this with thepercentage of the total agricultural area in holdings of this size. So, the reader is asked tolook down the two right-hand columns, and consider if migration tends to rise or fall withthe dominant farm size, keeping in mind the possible effect that might flow from thefarm size sector ranked second.
Migration 1870–1913: A Statistical Analysis 99
![Page 109: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/109.jpg)
and in subsequent chapters which also make use of econometric analy-
sis. These provide a ‘verbal exposition’ of the econometric results. A
more detailed analysis is then presented in appendices, which may be of
interest to those readers who are well versed in statistical matters. They
can, however, be skipped without losing any essential part of the argu-
ment presented in each chapter.
2. The Changing Nature of M igration , 1871–1913
The basic hypothesis considered in this study is that Germany went
through a ‘labour surplus’ phase of economic development, and that
this phase was coming to an end by the late 1890s. The shift to a more
modern pattern of migration post-1900 created conditions which were
favourable to political stabilization and the transition to a mature demo-
cratic society. The Lewis Model provides the theoretical basis for these
claims.
One of the best sources of Wgures for migration is the indirect calcu-
lation of net migration by region based on the comparison of census
returns. For each census year from 1871 to 1910, the Imperial Statis-
tical OYce compared the actual increase in population by region with
the recorded numbers of births and deaths since the last census (gener-
ally Wve years before). Deducting the recorded natural increase from the
actual increase produced a Wgure for net migration. This was done for
eighty-one regions and eight periods, thus producing a panel data set of
648 observations.
The reliability of these calculations will depend on the accurate
recording of births and deaths. By and large, this was not unsatisfactory
in Germany in the period. The main question raised by demographers
concerns the recording of stillbirths in the Wrst half of the century, but
this is not a problem as it does not aVect the overall rate of natural
increase (the issue is whether these were recorded as a birth followed
by a death or left unrecorded).6
One obvious problem is that the destination of migrants is not
recorded by this procedure. Other data sources will be examined in the
next section which do provide this information. It is, however, possible
to make the same calculation for Germany as a whole, which provides a
6 See the discussion of this in Knodel (1974).
100 Migration 1870–1913: A Statistical Analysis
![Page 110: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/110.jpg)
Wgure for net external migration. This gives an idea of the balance
between external and internal migration.
In 1900 the Imperial Statistical OYce extended the analysis back to
1841. In this pre-uniWcation period censuses were organized by the
diVerent states, and did not always take place in the same year. In
addition, boundary changes imply that a calculation of net external
migration, for the territories included in the post-1871 state, will in-
clude a certain amount of internal migration within French and Danish
territories (pre-1871 and pre-1864 respectively).
For these reasons the pre-1871 Wgures should be treated with some
care, although they are probably reliable for Prussian territories. The
overall record 1841–1910 is given in Table 4.1.
The table uses four regional groupings. Of these two are consistent
sources of net outwards migration: east, and south and south-west. But
the trajectories are very diVerent. The east was not a major source of
net outwards migration prior to 1861. Indeed, parts of the east were
gaining through migration: East Prussia and West Prussia both made
gains in 1851–61. Thereafter, outwards migration rises until, in the
decade 1880–90, the region is losing 1% of its total population annually
Table 4.1. Net migration 1841–1910, in average annual rates per
100,000 of the resident population
East Centre West and
north-west
South and
south-west
All Germany
1841–51 �55 þ37 �273 �301 �167
1851–61 �65 �98 �331 �433 �243
1861–71 �342 þ22 �260 �284 �229
1871–5 �738 þ323 þ119 �365 �192
1875–80 �431 þ48 �99 �203 �174
1880–5 �959 þ35 �252 �559 �427
1885–90 �1038 þ490 þ167 �268 �137
1890–5 �661 þ99 þ50 �259 �177
1895–1900 �815 þ351 þ449 �60 þ35
1900–5 �662 þ316 þ285 �113 þ18
1905–10 �663 þ225 þ217 �192 �51
Source: Pre-1871 data from SDR n.f.150, pp. 190*–193*; post-1871 data from SDR a.f.25(2) Ubersicht 1B, pp. 50–2 (1871–5 Wgures) and subsequent census publications,combined in the regional units given in Appendix Table 3A.1.
Migration 1870–1913: A Statistical Analysis 101
![Page 111: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/111.jpg)
through migration. This rate of out-migration then falls back in the
decade after 1900. In the south and south-west, by contrast, out-
migration was already substantial in the pre-1871 period. There is a
continuous outXow, with Xuctuations, until 1895, although the rates are
well below those in the east. After 1895, the rate of out-migration is
much reduced.
The experience of the two gaining regions is also diVerent. In the
pre-1871 period the west and north-west was also a major source of
out-migration. This was a period in which this region contributed heav-
ily to German migration to the United States. But after 1871 this
changed. There are periods of in-migration as well as out-migration.
After 1885 the region consistently absorbs migrants, post-1895 this is
on a larger scale. By contrast there is consistent migration into the
central region; the only period in which outwards net migration is
recorded is 1851–61. The rate of absorption Xuctuates, with peaks in
1871–5 and 1885–90, but it is more consistent after 1895.
The story in summary is that, in the early part of the century, there
was a substantial labour surplus in western Germany, north and south.
This was chieXy exported, although some was transferred to the centre
and east. In the rural east the amount of land under cultivation was still
rising so any local surpluses could be mainly absorbed within the
region. From the 1860s onwards industrial growth in the west and the
centre meant that more migrants could be found employment within
these regions, but at the same time the east emerged as a major source
of out-migration. Before 1895 the rate of progress of the industrial
regions was not suYcient to take in all the migrants out of the east and
the south, resulting in a continuing high rate of emigration for
Germany as a whole (Wnal column of Table 4.1).
Statistical analysis of the migration Wgures can be used to extract
additional information from the census results.7 Although the published
census returns do not distinguish between agricultural and non-
agricultural areas, the use of modern econometric techniques can pro-
duce estimated rates of net gain and loss for each period. These are
given in Table 4.2.
The table shows that there were four periods in which the net rate of
out-migration from agricultural areas exceeded 2.0% per annum, and
7 Results of the regressions used to make these calculations are given in Appendix 4B,which also contains a discussion of the statistical issues involved, and further commentson the pattern of migration revealed by this analysis.
102 Migration 1870–1913: A Statistical Analysis
![Page 112: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/112.jpg)
these also produced annual rates of gain for the non-agricultural areas
of at least 1.0%.8 After 1900, the rate of migration drops to one which
produces a signiWcantly lower rate of gain for the non-agricultural
regions: the rate of gain for 1905–10 is the lowest of all.
This does not mean that the overall level of migration had necessarily
dropped; rather it had changed its character from one dominated by the
transfer of surplus labour out of agriculture, to a pattern where this was
one stream amongst others, although still important. One way of dem-
onstrating this change is to consider what percentage of the pattern of
net gains and losses can be explained purely by the extent to which the
region specializes in agriculture. In a ‘labour surplus’ period a large
proportion of migrant moves are from agricultural to non-agricultural
areas, and most of the variation in net gains and losses by region is due
to this factor. In other periods transfers between industrial regions due
to the relative performance of diVerent industries will be an important
inXuence and the role of rural–urban migration will be reduced. Table
4.3 gives Wgures produced by the statistical analysis of the census
results, which show how German internal migration changed over the
period covered by this study.
8 The rate of gain is lower relative to the size of the non-agricultural population for tworeasons. First, the total non-agricultural population was larger. Secondly, some of thoseleaving agriculture emigrated and so did not transfer to non-agricultural regions insideGermany.
Table 4.2. Results of statistical analysis of net migration by region,
1871–1910: rates of net gain or loss through migration for agricultural
and non-agricultural areas (annual rates of gain or loss as percentages
of the total population)
Estimated rate of net loss for
an entirely agricultural region
Estimated rate of net gain for an
entirely non-agricultural region
1871–5 �2.33 þ1.60
1875–80 �1.25 þ0.73
1880–5 �2.50 þ1.01
1885–90 �2.89 þ1.65
1890–5 �1.64 þ0.68
1895–1900 �2.48 þ1.34
1900–5 �1.58 þ0.77
1905–10 �1.55 þ0.61
Migration 1870–1913: A Statistical Analysis 103
![Page 113: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/113.jpg)
The table shows that, before 1900, labour surplus migration was the
most important factor in internal German migration. Even in periods
when the rates of net gain and loss were reduced, such as 1875–80, the
importance of the ‘labour surplus’ component remained high. The peak
for labour surplus migration came in the 1880s: in this period over 50%
of the variation can be explained by this one factor: was the region
agricultural or not? The 1890s are a transitional period, with a less
pronounced agricultural character. By 1900–5 only a third of the pat-
tern of net gains and losses can be explained by the proportion occupied
in agriculture, and by 1905–10 this has dropped to just over a quarter.
The labour surplus period was over.
So, what happened in the 1890s that brought about this change? Was
there a static lump of surplus labour in agriculture which had simply
moved out by then? Or were there other changes which improved the
prospects of the agricultural sector?
Some Wgures on these issues are brought together in Figure 4.2. All
are taken from HoVmann. The two indices of arable producer prices and
imported cereal prices show the eVect that world market conditions had
on German producer prices. World market prices fell from the 1870s
onwards and this brought down German producer prices despite tariV
protection. But from the mid-1890s onwards world market prices began
to recover, and this in turn brought about a recovery of German arable
producer prices. The consequences of this can be seen from the move-
ment of the index of land prices. The index fell slowly from the late
Table 4.3. ‘Rural labour surplus’ migration as a
percentage of all migration (the percentage of the
variation in net gains and losses which can be
explained by the proportion occupied in agriculture)
1871–5 44.2
1875–80 43.5
1880–5 57.1
1885–90 63.9
1890–5 39.8
1895–1900 48.9
1900–5 33.1
1905–10 26.9
Note: These are the R2 statistics from the regressions given inAppendix Table 4B1.
104 Migration 1870–1913: A Statistical Analysis
![Page 114: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/114.jpg)
1870s, when it was over 67, to a low point of 59.6 in 1895. From then
onwards it was rising, the level reached in 1913 representing a rise of
68% in eighteen years.
The uppermost line in Figure 4.2 gives industrial wages as a percent-
age of agricultural. In the early 1870s industrial wages were 40–50%
above agricultural wages, and dipped below this in the late 1870s.
Thereafter the ratio rose and remained in the range 170–180% from
the late 1880s onwards. The improved agricultural conditions post-
1895 did not lead to a fall in this ratio, but agricultural wages did keep
pace with industrial wages during this period.
187040
60
80
100
120
140
160
180
200
1875 1880 1885 1890 1895 1900 1905 1910
Wage ratio
Import prices
Land pricesArable prices
Fig. 4.2. Indicators of agricultural conditions, 1870–1913 (all indices have
1913 ¼100)
‘Wage ratio’ is the ratio of industrial to agricultural wages; ‘arable prices’ is an index ofarable producer prices; ‘import prices’ is an index of the price of imported cereals.
Source: Calculated from data in Hoffmann (1965), 492–5, 561–2, 569–70, and 610–11.Hoffman’s index of agricultural workers’ earnings is based on figures for cash payments,but figures given in Pohl (1908) also show gains for workers paid in kind or by harvestshares. Earnings of Insteleute with threshing shares in Oletzko (East Prussia) rose 73%between 1880 and 1903 for example.
Migration 1870–1913: A Statistical Analysis 105
![Page 115: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/115.jpg)
The movement of this ratio can also be interpreted in the light of
the migration patterns considered in the previous chapter. In the mid-
century, the industrial regions were drawing migrants predominantly
from nearby rural areas. Consequently, the premium required to per-
suade migrants to come from these areas needed to be only a moderate
one. But by the late nineteenth century the premium had to be suY-
cient to persuade migrants to move over much longer distances. It is
noticeable that the premium did not fall back after 1900, even though
agriculture was then doing better. Industry still had need of long-
distance migrants.
3. Analysing Net M igration by Region , 1871–1910
Scrutiny of the pattern of net migration shown in Table 4.1 in the light
of the evidence on changing agricultural conditions given in Figure 4.2
shows that there is a possible connection between migration and agri-
cultural conditions: the improvement in land prices after 1895 appears
to have had some eVect, reducing the rate of migration out of the east
for example. This raises the possibility that a more precise modelling of
the eVect of changing agricultural conditions on migration might pro-
duce an equation which can be estimated econometrically.
The value of this does not lie only in giving precise values to rela-
tionships which can be guessed at from visual inspection—it also makes
it possible to compare diVerent causes of migration. There are two
issues which are of particular interest. The Wrst is the eVect of agricul-
tural structure, the extent to which the prevalence of large estates had
an eVect on migration. The second is the relationship between migra-
tion and demographic variables: how far was the regional pattern of
migration caused by diVerences in rates of population increase? Both
these issues will be examined in greater depth in subsequent chapters;
the aim here is to show the part they played in the overall pattern of
migration.
The full set of econometric results is given in Table 4B.2 in the
Appendix; Table 4.4 provides a summary of the results.
The analysis shows that demographic factors had an important inXu-
ence on the pattern of migration: regions with high birth rates tended to
have high rates of out-migration; regions with high death rates took in
migrants to replace these losses. The birth rate which matters most is
the one which aVects current levels of entry onto the regional labour
106 Migration 1870–1913: A Statistical Analysis
![Page 116: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/116.jpg)
market. It is the lagged birth rate—the birth rate for 20 years before—
which is used in the estimated equations.
The estimates conWrm that the level of agricultural occupation was
an important inXuence on migration: agricultural regions tended to
shed population and non-agricultural regions made gains. But they also
show that agricultural structure mattered: the presence of large farms
and estates tended to increase out-migration. This question will be one
of the main themes of Chapter 6, but the result appears to conWrm a
point made by a number of contemporary authors, including Max
Weber.9
The eVect of changing agricultural conditions on migration can be
estimated as part of the overall procedure: eVectively this asks whether
migration out of agricultural areas was higher in bad periods for the
agricultural sector—when land prices were low or the rural–urban
wage gap was high. The analysis shows that times of low land prices
were indeed periods in which migration out of rural areas tended to be
9 It was one of the major themes of Weber’s study of the conditions of the agriculturalworkers east of the Elbe, Weber (1893).
Table 4.4. Summary of results of statistical analysis of net migration
for 81 regions, Wve-year time periods, 1871–1910
I. Secure statistical results
Factors which increase net outwards migration:
High birth rates in the past
Presence of large farms and estates
Increases in the rural–urban wage gap
Factors which decrease net outwards migration or lead to net gains
from migration:
High death rates in the current period
High rates of non-agricultural employment
II. Less secure statistical results
Factors which increase net outwards migration:
Low land prices
Note: The classiWcation of results as ‘secure’ or ‘less secure’ is based on the analysis anddiscussion given in the Appendix. The assessment is based on the usual statistical tests ofsigniWcance, plus the ‘robustness’ of the results—the extent to which a coeYcient isaVected by changes in the form of the estimated equation or the inclusion of additionalvariables.
Migration 1870–1913: A Statistical Analysis 107
![Page 117: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/117.jpg)
higher, but this result is not a particularly secure one in statistical terms.
A stronger result is obtained when the eVect of changes in the rural–
urban wage gap is included in the estimated equations: upwards move-
ments in this gap led to increased out-migration; falls held back the Xow
out of rural areas.
How important are these variables in the explanation of the overall
pattern of migration? One way to approach this is to try to explain
migration between the four main regional groupings. An important
caveat should be noted: there is a lot of migration within these units
which is not covered by this analysis.
Table 4.5 compares the eVect of predicted migration using the coeY-
cients calculated for the two demographic variables with the actual
Wgures for net migration previously given in Table 4.1.10 The results
show that demographic factors contributed to out-migration from the
east, though the eVect is not large in relation to the total Xow. But the
10 The calculation is a simulation which allows for the eVect through the NETMIGkt�1
term.
Table 4.5. Demography and migration: the estimated eVect of
demographic variables on net migration by regional grouping (columns
headed a) compared with actual net migration, annual rates per 100,000
(columns headed b)
East Centre West and
north-west
South and
south-west
a b a b a b a b
1871–5 �24 �738 �15 þ323 �14 þ119 0 �365
1875–80 �27 �431 �17 þ48 �19 �99 �1 �203
1880–5 �29 �959 �21 þ35 �27 �252 �16 �559
1885–90 �34 �1038 �28 þ490 �33 þ167 �26 �268
1890–5 �41 �661 �40 þ99 �43 þ50 �36 �259
1895–1900 �46 �815 �52 þ351 �52 þ449 �44 �60
1900–5 �48 �662 �58 þ316 �55 þ285 �43 �113
1905–10 �58 �663 �66 þ225 �60 þ217 �44 �192
Note: The equation used in these calculations was NETMIGjt ¼ 0:45NETMIGjt�1þ0:26 DEATHRTjt �0:23 BIRTHRTjt�4 (the j subscripts denoting the use of regionalgroupings); these coeYcients are from column a of Appendix Table 4B.2.
108 Migration 1870–1913: A Statistical Analysis
![Page 118: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/118.jpg)
same eVect should have led to a similar level of out-migration from the
centre and the west, which in fact gained through migration. In the case
of the south and south-west, there was a very small eVect from the
demographic variables in the period of highest out-migration (1871–85)
and a larger eVect in the period of lower out-migration after 1895.
Some of the reasons for this can be seen from the comparison of the
various regional groupings given in Table 4.6. The east did have a high
birth rate, but it also had a high death rate (and more of the variation in
the death rate fed into migration); the overall rate of natural population
increase was not unusually high. The south had the lowest rates of
natural population increase, yet initially a high rate of out-migration.
The other factors show quite large variations which contribute rather
more to the pattern of migration. Table 4.7 provides standardized esti-
mates of the eVect of each factor on net migration over the whole period
1871–1910 (the standardization adjusts the calculated eVect so that the
national average is zero). The impact of the demographic variables is
small compared to the others. The most important eVect is the percent-
age of the population occupied in agriculture: the south and the east
lost population through migration because they were more agricultural
and because they had less industry. The south and the east diVered in
the importance of larger farms and large estates, and this apparently
caused a further rise in out-migration from the east. For the south, it
caused a small reduction. This also had diVerent eVects on the two
regions which gained through migration.
The eVect of demographic factors on migration will be considered
further in the next section and in the next chapter. In aggregate, it does
Table 4.6. Average values by regional grouping, 1871–1910
East Centre West and
north-west
South and
south-west
Death rate, current period
(annual %)
2.63 2.45 2.22 2.62
Birth rate, 20 years previous
(annual %)
4.17 3.95 3.69 3.87
Percentage of land in holdings
of over 100 hectares (1895)
45.6 26.4 7.9 3.1
Percentage of the population
occupied in agriculture (1895)
49.3 25.4 31.3 46.6
Migration 1870–1913: A Statistical Analysis 109
![Page 119: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/119.jpg)
not appear that out-migration from the eastern provinces to the rest of
Germany was driven to any signiWcant extent by higher rates of popula-
tion growth than those in other regions. But population growth was
high relative to the limited employment opportunities available in east-
ern agriculture. This drove the demographic surplus to seek employ-
ment in cities both inside the east and in other regions. It was the
combination of demographic increase with the relative decline of agri-
culture which made the east a major source of migrants in the ‘labour
surplus’ period.
But this uses a broad classiWcation. Results in subsequent chapters
show a diVerent picture for individual Kreise and regions. High rates of
population growth at this level were an important factor in migration.
4. Rural–Urban M igration from the 1890 Census
The analysis in the previous section used net migration Wgures calcu-
lated from census returns. This data gives no indication of the destin-
ation of the migrant Xows. But a comparison of place of residence with
place of birth from census returns can provide information of this
type.11
An individual census is a snapshot, describing a moment in time,
which tells us quite a lot about migration but which will leave out many
Table 4.7. Estimates of the eVect of diVerent factors on the pattern
of net migration by regional grouping (annual average net migration
per 100,000), 1871–1910
East Centre West and
north-west
South and
south-west
Both demographic variables �3 �2 �3 þ9
Percentage of land in holdings of
over 100 hectares
�135 �31 þ70 þ96
Percentage of the population occupied
in agriculture
�290 þ331 þ178 �220
Note: A similar equation to that given in the previous footnote is used for each variable;the coeYcients used are �0:32 for LAND>100j and �0:168 for % AGOCCPjt.
11 This type of data has been used by Kollmann (1959) amongst others.
110 Migration 1870–1913: A Statistical Analysis
![Page 120: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/120.jpg)
moves. Those recorded as living away from their place of birth will
include many who made other moves before reaching their current
place of residence. Some of those living in the region of their birth may
have left and then returned. In addition, there is no way of knowing
which moves are temporary and which are intended to be permanent.
Migration calculated from census records in this way is more than just
net migration (since a calculation of the Xow from one region to another
is possible in both directions) but it will also tend to understate the
overall level of mobility.
One advantage of this type of data is that it is possible to separate
diVerent types of migration. The 1890 census is particularly useful. The
published results provided an analysis of the population resident in
major cities according to place of birth. This can be used to examine
moves to the twenty-seven largest cities from areas outside these cities,
thus excluding moves between major cities, and moves between rural
areas. The analysis can consider whether the likelihood of a move was
aVected by conditions in the region of origin or the recipient city, identi-
fying factors which promoted or retarded migration between urban and
rural areas. The full set of econometric results is given in Appendix 4B,
Section (iii); Table 4.8 provides a summary of the Wndings.
There are a number of general Wndings which provide an overview of
the migration process in the period up to 1890. Migration was strongly
aVected by the distance between the recipient cities and the regions of
origin. If the recipient city lay within the region of origin this increased
the likelihood of a move. This indicates a strong preference for short
moves over long-distance ones. It also means that the proximity of
major cities would have an eVect on overall rates of migration out of
rural areas. In remote regions, the large distances needed to be travelled
to reach the various major cities would have a depressing eVect on
migration in general.
High wages in the recipient cities, and low levels of unemployment
or casual employment as revealed by the 1895 occupational census,
tended to increase migration: moves were more likely to cities with
stable or buoyant labour markets. There were certain industries which
attracted migrants: clothing and construction oVered easy entry into the
labour market for migrants who lacked previous experience of industrial
work. By contrast, cities with high levels of employment in textiles did
not attract migrants: the main period of rapid expansion of these cities
had been in the 1860s and 1870s, and they had less need of migrants
once this period had passed.
Migration 1870–1913: A Statistical Analysis 111
![Page 121: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/121.jpg)
In the regions of origin, poor economic prospects, as indicated by a
low rate of formation of Aktiengesellschaften (joint stock companies), had
a positive eVect on outwards migration. This is interesting as it indi-
cates a ‘forward-looking’ element to migration: migrants responded to
the direction of economic change as well as current conditions.
When these Wndings are combined with the result of those given in
Table 4.4, they show how important economic factors could be in
aVecting the course of migration: migrants were more likely to move if
the urban–rural wage gap increased; they chose destinations with high
wages and good chances of employment; they left regions which were
lagging in terms of economic development.
Demography was also important, although the strongest and most
secure results are obtained for the recipient cities: high rates of natural
population increase in these had a negative eVect on inwards migration,
since such cities could Wll a greater proportion of available vacancies
from their own natural increase. By contrast, while it appears that high
rates of population increase in the regions of origin may have promoted
out-migration from these regions, this Wnding is much less secure.
Table 4.8. Summary of results of statistical analysis of
rural–urban moves from 1890 census
I. Secure statistical results
Characteristics of the regions of origin which increase outwards
migration:
Proximity to the recipient cities
Presence of large farms and estates
Poor economic prospects
High rates of productivity growth in agriculture
Characteristics of the recipient cities which increase inwards migration:
High wages
Low rates of natural population increase
Low numbers of unemployed workers or workers without permanent jobs
Presence of the construction and clothing industries
Absence of the textile industry
II. Less secure statistical results
Characteristics of the regions of origin which increase migration:
High rates of natural population increase
High levels of agricultural employment
112 Migration 1870–1913: A Statistical Analysis
![Page 122: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/122.jpg)
Agricultural conditions are shown to have had an important eVect on
out-migration. The presence of large farms and estates increased the
likelihood of outwards migration in the regions of origin. In addition,
agricultural change, as measured by the rate of productivity increase
between 1880/4 and 1893/7, is associated with high rates of out-migra-
tion. This could be an indication that labour was shed as a direct conse-
quence of changes in agricultural technology—the use of mechanized
threshing, for example—or it could show that there was a general pro-
cess of agricultural transformation, involving the replacement of trad-
itional social and economic relations with modern capitalist ones, which
had the eVect of both raising productivity and increasing the likelihood
of migration. Alternatively, improvements in transportation may have
brought new inputs to remote areas, raising productivity, as well as
making it easier for potential migrants to leave.
5. Conclusions
In general the statistical results presented in this chapter and the appen-
dices conWrm the analysis presented in the previous chapter. There was
a period of intense rural–urban migration in the 1870s and 1880s, and
this was aVected by conditions within agriculture relative to those in
industry, and by changes within agriculture. Agricultural transform-
ation brought about the release of surplus labour onto urban labour
markets. After a period of transition in the 1890s, the pattern of migra-
tion changed to one which was no longer dominated by the movement
of surplus rural labour. The ending of this ‘labour surplus’ phase
created conditions which were favourable to the creation of a modern
mature industrial society, and to a process of social and economic rec-
onciliation which could have led to the emergence of democratic insti-
tutions and a reduction in social conXict.
Migration 1870–1913: A Statistical Analysis 113
![Page 123: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/123.jpg)
Appendix 4A. Description of Regress ion Variables
and Sources
(i) Appendix Tables 4B.1 and 4B.2; Tables 4.2, 4.3, 4.4, 4.5, 4.6, and 4.7
in text
NETMIGkt
BIRTHRTkt
DEATHRTkt
Net migration, births, and deaths per 10,000 head
of population from census results, from SDR a.f.
20 and 52, n.f. 32, 68 and 150–1, VSDR 1897 iv,
1906 iv, and 1911 iv.
%AGOCCPkt The population occupied in agriculture as a per-
centage of the total occupied population, from the
occupational censuses, from SDR a.f. 4, n.f. 104–6
and 204–6: 1895 and 1907 results are also in Kaelble
and Hohls (1995).
LAND > 100k The percentage of the total agricultural area in
holdings of over 100 hectares, from 1895 Agricul-
tural Census, SDR n.f. 212, pp. 489–500.
LANDPRICEt HoVmann’s index of land prices, HoVmann
(1965), 569–70.
DAGINDWGAPt Change in the ratio of industrial and agricultural
wages, using HoVmann’s Wgures, and comparing
Wve-year averages, HoVmann (1965), 492–5.
(ii) Appendix Table 4B.3, Table 4.8 in text
P(MIG)kj(dependent
variable)
Numbers born in rural part of region k who are
recorded as living in city j, divided by total
numbers born in rural part of k, calculated from
SDR n.f. 68, pp. 136–53.
NATPOPINCk Natural rate of population increase in the region of
origin, 1880–90, from 1885 and 1895 census
results, per 1,000 head of population, SDR n.f. 52
and VSDR 1897 iv.
AGVALUEk The value of Aktiengesellschaften founded 1871–90from VSDR 1907 iv, p. 373; the Wgures are then
divided by the regional population 1895 (1895
census).
%AGRICk The population occupied in agriculture as a per-
centage of the total occupied population, from the
114 Migration 1870–1913: A Statistical Analysis
![Page 124: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/124.jpg)
1895 occupational census, Kaelble and Hohls
(1995), 167–88.
DAGPRODk Percentage increase in agricultural value added per
full-time labour unit, 1880/4 to 1893/7, from
regional accounts prepared for this study (see
Chapter 7).
NATPOPINCj Natural rate of population increase for the recipi-
ent city, 1890–5, from SJDS (1892), table viii,
p. 39, and subsequent issues.
AVEWAGESj Average of Ortsubliche Tagelohne for men and
women, 1892, from SJDS (1913), 827–8.
%CONSTRCTj
%CLOTHj
%TEXTILESj
%UNEMPj
Calculated from the results of the 1895 occupa-
tional census, given in SJDS (1898), 350–1; also in
SDR n.f. 107–8.
Appendix 4B. Econometric Analysis
(i) Analysis of net migration 1871–1910, using census data
The post-1871 census results can be used to look at the rates of transfer
between the eighty-one regions (the pre-1871 Wgures use diVerentregional units and so are not comparable). Analysing each time
period separately can show if the course of migration changed—if
there were some periods which were dominated by the transfer of
surplus labour out of agriculture and others which show a diVerent
pattern. Table 4B.1 presents the results of a series of regressions run
for each time period in succession, with a single explanatory variable:
the percentage of the total occupied population which is occupied in
agriculture (%AGOCCP).
What this table shows is the extent to which the pattern of net
migration by region can be interpreted solely as a transfer of labour out
of the more agricultural regions, either to the industrial regions or to
foreign countries. The t-ratios are all signiWcant at the 95% conWdence
level, which shows that the extent to which a region specializes in
agriculture is always a factor in migration, but it is the R2 statistics
in the Wnal column that are the most eloquent. These show the extent
to which the incidence of this form of migration, the transfer of surplus
labour out of agriculture, can explain all the observed variance in net
Migration 1870–1913: A Statistical Analysis 115
![Page 125: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/125.jpg)
migration by region. At the peak in 1885–90, with an R2 of 0.639, this
is evidently the dominant form of migration: 64% of the migration
pattern can be explained by this factor alone. But by 1905–10 it only
accounts for around a quarter of observed variance. By this time migra-
tion was more heavily inXuenced by other considerations.12
The slope coeYcients give an indication of the pace of the transfer.
A high coeYcient indicates that a large proportion of the agricultural
population migrated in that period. This peaks in 1885–90, although it is
also high in 1871–5 and 1895–1900. By 1900–10 it is considerably lower.
This table provides clear statistical support for the view that the
pattern of migration shifted from a ‘labour surplus’ one in the 1870s
and 1880s, to a more ‘normal’ one post-1900, where the movement of
population out of agriculture (which still continues even in advanced
industrial economies) is just one stream amongst other migrant move-
ments, not a dominant inXuence on the labour market as a whole.
(ii) Panel data analysis of migration 1871–1910
The separate analysis of each time period can produce interesting con-
trasts in the slope coeYcients, and in the Wt of the estimated equations,
12 The question of non-linearity in the functional form, which might aVect the R2
calculations, was considered. F-tests generally led to the rejection of additional X2 vari-ables. No signiWcant improvement in the R2s was found to be possible.
Table 4B.1. Regressions of net migration by period on % occupied in
agriculture: dependent variable is net migration per 100,000 (annual
rate) for 81 regions
Period constant coeYcient on
%AGOCCP
t-ratio R2
1871–5 1595 �39.22 7.91 0.442
1875–80 734 �19.87 7.80 0.435
1880–5 1006 �35.06 10.36 0.571
1885–90 1653 �45.44 11.84 0.639
1890–5 677 �23.12 7.23 0.398
1895–1900 1335 �38.15 8.70 0.489
1900–5 769 �23.49 6.25 0.331
1905–10 610 �21.63 5.39 0.269
Note: Regression variables and sources given in Appendix 4A.
116 Migration 1870–1913: A Statistical Analysis
![Page 126: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/126.jpg)
but combining the data in a single panel makes it possible to test
hypotheses about the reasons for the changes in the slope coeYcients.
The analysis in this section uses the same sources as the previous
section but analyses the data as a single panel.
There are some restrictions on the analysis which follow from the
nature of the data. Only eight time periods are available for the 1871–
1910 period, and the pre-1871 data is not comparable either in terms of
the regional units used or the lengths of the time periods. This means
that any lag structure has to be a simple one: eVectively the only possi-
bility is the inclusion of a NETMIGkt�1 term (net migration for the kth
region lagged one period) amongst the explanatory variables, implying,
by way of the Koyck transformation, a lag structure of the form
E(Xt) ¼ aXt þ a2Xt�1 þ a3Xt�2 þ a4Xt�3 . . . : for the expected values
of the explanatory variables.
An additional problem is that the sum total of regional net migration
is the external balance, which Xuctuates considerably, as shown in
Table 4.2. Ideally, this would be modelled separately, and in a way
which took into account the eVect of alterations in economic conditions
in the main destination for German emigrants, the United States.13
But, given the limited number of time periods, it is not possible to
estimate a complex model of this type for the regional data set.
So, the data set was analysed as a Wxed eVects panel data set, includ-
ing dummies for all time periods, which means that the external balance
is treated as an exogenous factor. The estimated model only deals with
the internal transfer of population between the diVerent regions.
Results are given in Table 4B.2. The Wrst column gives the basic
model: net migration as a function of two demographic variables (the
current death rate and the birth rate lagged 20 years), the percentage
occupied in agriculture and the amount of land in holdings of over 100
hectares. As shown in the previous section, the percentage occupied in
agriculture has an important eVect on migration, but this table shows
that there is a further eVect from the prevalence of larger farms and large
estates: there is a signiWcant negative coeYcient on the LAND > 100kterm, suggesting that this causes additional out-migration.
This point merited further investigation, and it is one of the themes
examined in the next two chapters. In brief, these results show that,
13 A model of this type, using annual data, was included in chapter 6 of the originalD.Phil. thesis, Grant (2000). It is intended to publish a separate article based on thischapter.
Migration 1870–1913: A Statistical Analysis 117
![Page 127: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/127.jpg)
once the issue is examined at a lower level of aggregation, using a Kreis-
level data set (with data for around 400 Prussian Kreise), then it is
discovered that what mattered was the geographical position of these
estates, the fact that they were mainly located in remote eastern areas,
not the fact that they were large as such.
The demographic terms are both signiWcant. Their importance is
somewhat obscured by the lag structure adopted; part of the eVectcomes through the NETMIGkt�1 term. The estimated coeYcients
imply that, after Wve periods (or twenty-Wve years), 41% of an increase
in the birth rate lagged four periods would be exported through in-
creased net migration, and 47% of a decrease in the current death
rate.14
The next two columns in Table 4B.2 introduce variables which are
inter-acted with the %AGOCCPkt variable. These are variables which
might aVect the rate of labour transfer between agriculture and indus-
try, and thus explain the variation in the coeYcients on %AGOCCP
shown in Table 4B.1. A negative coeYcient therefore shows that a rise
in this variable increases the rate of transfer from agriculture to indus-
try; a positive coeYcient shows that it holds it back.
The second column uses the change in the ratio between agricultural
and industrial wages to explain the shifts in the pace of migration. This
model produces an improved Wt. Part of the increase in migration be-
tween the 1870s and the 1890s can therefore be attributed to the eVectof the rise in this ratio. Given the lag structure used, this would still be
aVecting migration for some time after this, even though the ratio
ceased to rise.
The third column uses a diVerent variable: the level of land prices as
an indicator of agricultural prosperity. The estimated coeYcient is posi-
tive, indicating that higher land prices did reduce the rate of migration,
but it is not signiWcant at the 5% level (the hypothesis that there is in
fact no eVect on migration from this variable cannot be rejected with
95% conWdence). So, migration is best explained by the equation used
for column b, which uses the change in the agriculture–industry wage
gap. It was the widening of this gap which was responsible for the high
rate of rural urban migration in the 1880s.
What is also notable about the results in Table 4B.2 is the stability of
the coeYcients on the demographic variables, and on the structural
14 The reason for lagging the birth rate 20 years is that the Berlin census results showthat most migrants arrived in their early twenties.
118 Migration 1870–1913: A Statistical Analysis
![Page 128: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/128.jpg)
variables LAND > 100k and %AGOCCPkt. The importance of these
variables in the overall pattern of migration is conWrmed. The labour
surplus was created by a combination of demographic factors and the
relative decline of agriculture.
(iii) Analysis of rural–urban moves from the 1890 census
The 1890 census provides Wgures for moves to the twenty-seven largest
cities from those living outside these cities, divided between twenty-
nine provinces. This excludes city–city moves and rural–rural moves.
This data has been used to create a variable Pkj, which is the proportion
Table 4B.2. Fixed eVect regressions (OLS with time period dummies)
of net migration, annual rates per 1,000 population, Wve-year time
periods 1871–1910
a b c
NETMIGkt�1 þ0.454 þ0.469 þ0.450
(5.86) (5.98) (5.73)
DEATHRTkt þ0.260 þ0.231 þ0.247
(3.43) (3.11) (3.27)
BIRTHRTkt�4 �0.230 �0.214 �0.223
(4.26) (4.13) (4.16)
%AGOCCPkt �0.168 �0.163 �0.170
(6.60) (6.42) (6.65)
LAND > 100k �0.0320 �0.0320 �0.0328
(2.78) (2.84) (2.85)
%AGOCCPkt* �0.0119
DAGINDWGAPt (5.50)
%AGOCCPkt* þ0.00380
LANDPRICEt (1.81)
N 648 648 648
F-test of regression 89.10 89.02 83.18
[12,635] [13,634] [13,634]
R-sqd 0.627 0.646 0.630
adj R-sqd 0.620 0.639 0.623
RSS 112213990 106588139 111304403
Note: t-statistics (in brackets) calculated using White’s heterocedasticity-correctedstandard errors. Appendix 4A gives a full description of variables and sources; alternativerandom eVects estimates of the equations are given in Appendix 2B of Grant (2000).These do not alter the basic results although some of the coeYcients did change.
Migration 1870–1913: A Statistical Analysis 119
![Page 129: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/129.jpg)
of the total numbers born in province K, excluding those born in major
cities, who move to city J.
The model is a simple linear log-odds ratio model: log (Pkj=(1� Pkj)
¼ f (Xk1 . . . :Xkn, Xj1 . . . :Xjm), where f(.) is a linear function, and
Xk1 . . . :Xkn, Xj1 . . . :Xjm are characteristics of the region of origin K
and the recipient city J used as explanatory variables. The only non-
linear variable is the distance travelled which is entered in a log form.
The functional form produces a logistic curve, thus preventing pre-
dicted values from falling below zero.
The LOGDISTANCEkj term gives the log of the distance in kilo-
metres between the recipient city and the nearest border of the region
of origin. Where the recipient city was within the boundaries of the
region a notional distance of 2 kilometres was used (this could not be
zero as the term was to be in log form). To avoid distorting the results
by this procedure a dummy variable DUM1CONTkj was introduced
for moves to cities within the regional boundaries.
Table 4B.3 gives the results of four diVerent regressions run using
this data. In the Wrst three columns the only variables included for the
Table 4B.3. Regression results for linear log-odds migration
model, 1890 census
a b c d
Constant �5.982 �5.620 �5.367 �5.082
LOGDISTANCEkj �1.331 �1.334 �1.328 �1.281
(26.3) (26.4) (27.2) (30.4)
DUM1CONTkj �0.655 �0.676 �0.668 �0.499
(2.45) (2.53) (2.55) (2.21)
Characteristics of the region of origin:
NATPOPINCk þ0.01088 þ0.01118
(1.26) (1.30)
%AGRICk þ0.01108 þ0.06684
(2.64) (1.43)
LAND > 100k þ0.0154 þ0.0141 þ0.0156 þ0.0155
(6.90) (6.13) (7.54) (8.73)
AGVALUEk �0.00196 �0.00392 �0.00381
(2.13) (4.63) (5.24)
DAGPRODk 0.0140 0.0140
(5.49) (6.37)
120 Migration 1870–1913: A Statistical Analysis
![Page 130: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/130.jpg)
recipient city are the level of wages (AVEWAGESj) and the rate of
natural population increase (NATPOPINCj). A number of explanatory
variables which were found to inXuence migration were introduced for
the region of origin. The results in the Wrst column show that the
prevalence of larger farms and estates (LAND > 100k) has a signiWcant
positive eVect on the likelihood of a move from the region of origin.
This conWrms the Wnding reported in the previous section.
The rate of natural population increase has a signiWcant negative
eVect on the likelihood of migration into the recipient city (a high rate
of population increase reduces migration into the city) but the coeY-
cient on the NATPOPINCk term (population growth in the region of
origin) was not signiWcant. This is an intriguing result: variations in
population growth between cities matter; those in rural areas are less
Characteristics of the recipient city:
NATPOPINCj �0.0120 �0.0120 �0.0120 �0.0101
(�8.95) (�8.98) (�9.13) (�5.13)
AVEWAGESj þ0.0195 þ0.0195 þ0.0195 þ0.0123
(11.55) (11.57) (11.79) (7.896)
%CONSTRCTj þ0.162
(6.156)
%CLOTHj þ0.105
(5.283)
%TEXTILESj �0.040
(9.74)
%UNEMPj �0.140
(7.29)
N 783 783 783 783
S 1.031 1.028 1.010 0.869
RSS 823.3 818.5 790.8 582.5
F-Test 201.6 177.8 214.5 209.4
[8,774] [8,774] [7,775] [11,771]
R-sqd 0.645 0.647 0.659 0.749
adj.R-sqd 0.642 0.644 0.656 0.746
Note: RONATINC and RCNATINC are natural rates of population growth for theregion of origin and the recipient city; %AGRIC is the % of the population occupied inagriculture; %LAND > 100HA and %LAND20/100 refer to the distribution of theusable agricultural area by size of holding; AVEWAGES is the average wages of day-labourers in 1892 for each city; AGPROD82 and DAGPROD82=95 are Wgures for thelevels of agricultural productivity 1880/5 and the increase in 1880/5–1893/7. Thesemake use of regional agricultural accounts prepared by the author. Further details on thesources are in Appendix 4A.
Migration 1870–1913: A Statistical Analysis 121
![Page 131: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/131.jpg)
important. This theme will be examined in more detail in the next
chapter.
The percentage of the population occupied in agriculture in the
region of origin (%AGRICk) is a variable whose importance is consid-
erably reduced when compared to the results given in the previous
section. But this is not so surprising. The eVect of using rural–urban
moves as the dependent variable is to eliminate much of the previous
eVect of this variable. More urban regions had lower overall net migra-
tion for two main reasons: Wrst, there were proportionately less farmers
and farm-workers, and secondly, there were more moves to cities within
the region. Moves to cities within the region are now recorded; and by
using moves relative to the rural population as the dependent variable
the importance of the Wrst point is also reduced.15
The second column shows the eVect of including the term
AGVALUEk. This gives the value per head of population of Aktienge-sellschaften (joint stock companies) founded 1871–1910 by region. This
was included in an attempt to examine the extent to which migration
was aVected by the prospects of a region as well as its current position.
The %AGOCCPkt term is a backward-looking measure: it reXects what
has happened in the past, not what is likely to happen in the future.
The formation of Aktiengesellschaften is a forward-looking decision by
investors and should be related to the dynamism of the region, to the
extent to which it has specialized in fast-growing industries for
example.
The introduction of the AGVALUEk variable reduces the signiW-
cance of the %AGRICk term (column b). For the last two columns this
term and the NATPOPINCk term are both dropped. The coeYcient on
AGVALUEk is raised as a result. Some of the eVect of the %AGRICk
variable is captured by AGVALUEk (Aktiengesellschaften were more
likely to be formed in areas which were already industrial). But the
greater signiWcance of AGVALUEk is due to its ‘forward-looking’ or
dynamic quality.
What exactly were these dynamic factors which caused AGVALUEk
to have this inXuence? There were three forces at work. The Wrst was
an increase in the importance of large Wrms, which were more likely to
be Aktiengesellschaften. The second was the impact of foreign trade on
the German economy, which led to greater specialization in those areas
15 The fact that some eVect is still exerted by %AGRICk is explained by the fact thatthere are some regions which were industrial but which did not have large cities.
122 Migration 1870–1913: A Statistical Analysis
![Page 132: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/132.jpg)
where Germany had a comparative advantage, and in turn led to faster
growth in regions specializing in these Welds. The third was a tendency
for regional specialization to rise, due to economies of scale internal and
external to the Wrm. Increased vertical integration, leading to the com-
bining of processing and primary production in a single region, was one
aspect of this.
In the third column the increase in agricultural productivity by
region between the years 1880–4 and 1893–7 is introduced (DAGPRODk).
16 This variable is found to have a signiWcant positive eVect on
the likelihood of migration: migration is higher from regions with high
productivity increases. This result can be interpreted as supporting the
view that there was a substantial agricultural labour surplus, and the
release of this surplus both increased productivity and led to out-migra-
tion. But an alternative view is that these are both consequences of a
change in institutions and social relations. The impact of industrializa-
tion led to a transformation of agrarian society which both raised prod-
uctivity and led to increased out-migration.
In the last column (d) the breakdown of the employed population in
the recipient city by occupation is used to produce a series of additional
explanatory variables. The results show that certain industries attracted
migrants, leading to an increase in the likelihood of inward migration
from rural areas. Two which had this eVect were construction and
clothing (%CONSTRCTj and %CLOTHj). Other industries had a
negative eVect. One of these was textiles (%TEXTILESj). The main
period of rapid expansion for this industry had passed by 1890, and the
main textile areas had less need of migrants. The results also show that
there is a negative eVect from the percentage of workers who were
without regular employment according to the census (the variable
is called %UNEMPj although it is not the same as registered
unemployment).
16 These Wgures are derived from regional accounts. The methods used are describedin chapter 5 of Grant (2000), and also in Grant (2002a).
Migration 1870–1913: A Statistical Analysis 123
![Page 133: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/133.jpg)
5
Demography and Migration
1. Introduction
The last chapter presented some evidence on the inXuence of demo-
graphic variables on migration. The results were equivocal. Regressions
showed an eVect from demographic variables on net migration for
eighty-one regions 1871–1910, but, when migration between four major
regional groupings was considered, the net eVect of demographic factors
was shown to be small when compared with other explanatory variables.
The analysis of the 1890 census showed that variations in the natural
rate of population increase in the recipient cities had an important eVect
on migration, but that the eVect from population growth in the regions
of origin was less certain.
This chapter will attempt to elucidate these issues, in the context of a
general discussion of the relationship between migration and demog-
raphy in Germany 1870–1913. In this period Germany was going
through the ‘demographic transition’, usually deWned as a movement
from a state in which both fertility and mortality are high to one in
which both are considerably lower. This transition has been much stud-
ied by demographers: one conclusion which has emerged is that there is
no single model of the demographic transition—much will depend on
the circumstances of individual countries or regions. The German
experience conWrms this: there were important diVerences between the
regions and these have implications for migration. SpeciWcally, the
demographic experiences of two regions which were major sources of
potential migrants in the period, the south and the east, were very
diVerent. This had a considerable eVect. It is one reason why there was
heavy out-migration from certain heavily agricultural regions in the
east, while migration from similar southern regions was lower.
So, diVerences between the south and the east will be one theme of
the chapter. A second theme will be a comparison of the demographic
background to two periods of heavy out-migration: that which aVectedsouthern and western Germany in the earlier part of the nineteenth
century and the movement out of eastern Germany in the period after
![Page 134: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/134.jpg)
1870. The intention here is to look for signs of demographic stress,
direct evidence of severe population pressure. To what extent was
migration caused by overpopulation? Are there diVerences between the
two periods?
Comparison between the south and the east leads to an examination
of the role played by agricultural factors. The agricultural systems of
the two regions were very distinct: diVerent sized-farms, diVerent typesof labour contract, diVerent inheritance systems. Did these have demo-
graphic eVects?
As regards the demographic experience of the recipient towns and
cities, this has already been studied in some depth. The high fertility of
the coal-mining regions in the west is a well-known fact.1 This has clear
implications for migration. It is a fairly simple calculation to show that
high fertility reduced the need for migration into the coal-mining areas.
A higher proportion of the required increase in the total labour force
was supplied through the natural increase of the resident population
than was the case for Berlin, for example. This was an important factor
in migration.2
A more general theme is the contribution of demographic movements
to the ‘labour surplus’ phase of the Lewis Model. The period of the
‘demographic transition’, when mortality rates fall ahead of a reduction
in fertility, is potentially one of high population growth, contributing to
the overall labour surplus.
2. The German Demographic Experience in the
N ineteenth Century
An overview of German population movements in the nineteenth cen-
tury is provided in Figure 5.1. This uses calculations made by the
Prussian statistical oYce for all the territories included in Prussia after
the incorporation of Hanover and Schleswig-Holstein in 1866. This
series extends back to 1816. A second series is superimposed, which is
for all the post-1871 German territories, and makes use of a series
1 Wrigley (1961); see also Haines (1977) and (1979).2 Valuable studies of demographic change in Germany in the period include Ballod
(1914), Hohorst (1977), Kollmann (1974), Knodel (1988), Marschalck (1987), Lee(1979a), and Pollard (1980). This chapter also draws on the theoretical analysis of Birdsell(1988) and Habbakuk (1958). The regional classiWcation follows the deWnitions set out inAppendix Table 3A.1.
Demography and Migration 125
![Page 135: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/135.jpg)
produced by the Imperial Statistical OYce. This only goes back to
1862.3 There is a fairly close correspondence between the two series,
which is as expected, given that Prussia accounted for around 61% of
the total German population.
The reliability of the earlier Wgures is open to question. The main
area of concern is the Wgures for stillbirths prior to 1841. These had to
be estimated as they were not then recorded. If these are underesti-
mated then the birth rate and the death rate are aVected, but not the
overall rate of population growth.
The natural rate of population increase peaked in the 1816–22
period, possibly as a reaction to the ending of the disruptions caused by
3 Calculated from SJDR (1915), 22, SJPS (1910), 14.
18160
100
200
300
400
500
Prussian births
Prussian deaths
Germanbirths
Germandeaths
Prussian natural
1826 1836 1846 1856 1866 1876 1886 1896 1906
increase
German naturalincrease
Fig. 5.1. Births, deaths, and the natural rate of population increase, per 10,000
head of population, Prussia and Germany 1816–1913 (post-1871 territorial
units)
Note: Figures are three-year rolling averages.Source: The source used is SJPS (1915), 55; these estimates were endorsed by Kollmann(1980), who also gives figures for the individual provinces.
126 Demography and Migration
![Page 136: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/136.jpg)
the Napoleonic wars, but then fell back sharply in the agricultural crisis
which followed, from 1.9% in 1822 to just 0.2% in 1831. This was
brought about by a combination of a rise in the death rate and a fall in
the birth rate. Thereafter there was an irregular recovery, which peaked
in the mid-1870s, when the natural rate of increase reached 1.5% per
annum. This period saw considerable Xuctuations in both fertility and
mortality, which are apparent even when a three-year moving average is
used. After the 1870s both series are much more stable, and tend to
decline, with the decline in the death rate being more pronounced and
starting much earlier. The birth rate does not show a substantial decline
until after 1900, so that the rate of natural increase remained high,
peaking at 1.56% per annum (reached in both 1898 and 1902).4
The decline of fertility after 1900 was too late to aVect the pre-1913
labour market. The combined eVects of high fertility in previous
periods and declining mortality meant that the rate of increase of the
potential labour force remained high right up to 1913. Comparison of
the employment censuses of 1895 and 1907 shows the male labour force
increasing at a rate of 1.5% per annum and an even faster rate for
female employment.5
The pattern is a classic demographic transition curve. But the peak
was reached at diVerent times in diVerent regions. Figure 5.2 illustrates
this point, classifying each region by the timing of the peak rate of
natural increase between census years (normally at Wve-year intervals).Most regions peaked in either 1895–1900 or 1900–5, as did the
national average. But there was a group which peaked much earlier, in
the 1870s. This included Berlin, Pomerania, and some other central
regions. There is another group in which population growth was still
rising by 1905–10, which had not, therefore, clearly peaked before the
First World War. These are the regions shaded black. This is quite a
scattered group of regions, with all areas represented apart from the
centre. There is no clear pattern; it includes agricultural as well as
industrial regions.
Figure 5.2 helps to explain why there was relatively little eVect from
demographic factors when the pattern of migration between the major
regional groupings was considered in the last chapter. There were
4 German Wgures, SJDR (1912), 22.5 Data from Kaelble and Hohls (1989). There are some changes in the relative positions
of the various regions as a result of diVerent rates of industrial growth between censusdates, but the overall ranking remains remarkably constant. The classiWcation would beessentially the same if the 1882 or 1907 censuses were used in place of the 1895 census.
Demography and Migration 127
![Page 137: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/137.jpg)
diVerences within these groupings which had a tendency to cancel each
other out. Population growth falling in Pomerania was balanced by
rising population growth in other eastern regions, for example.
Figures 5.3 and 5.4 demonstrate the eVect of these movements on the
overall rate of population growth by region, comparing the 1870s with
the 1900–10 period. The Wrst map has a ‘mottled’ appearance: most
parts of the country contained regions both of high and of low growth.
Again, this helps to explain why demography had a limited eVect on the
major groupings.
By 1900–10 there is a clearer pattern: the areas of high growth are
predominantly in the west and in the east; the centre and the south
contain the low-growth regions. In most areas the rate of increase was
faster in the 1900–10 period. The main exceptions are in the centre,
conWrming that this was the region which Wrst went into the downward
phase of the demographic transition.
1895−1900
1890−5
1900−5
1905−10
1875−80
1871−5
1880−5
1885−90
Fig. 5.2. The demographic transition in Germany 1871–1910: German
regions classified by the period of highest natural population increase
Source: Data for this figure, and the next two, taken from census publications, SDR(1876), a.f. 20, SDR (1881), a.f. 57, SDR (1886), n.f. 32, SDR (1891), n.f. 68, SDR (1901),n.f. 150–1; VSDR 1897 iv, 1906 iv, and 1911 iv.
128 Demography and Migration
![Page 138: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/138.jpg)
From the viewpoint of a study of migration, the most important
map is Figure 5.3. This is because diVerences in population growth in
1900–10, in so far as these reXect diVerences in birth rates, are not
particularly relevant to migration Xows before 1914. A rise in the
number of children will not increase migration immediately; in fact it
could reduce migration if families with children are less likely to mi-
grate. It is only when these children enter the labour market, and take
decisions based on an assessment of their own prospects, that there is
likely to be an increase in out-migration from a region of high popula-
tion growth.
Analysis of the Berlin Wgures from the 1885 census supports these
points. The median age of migrants arrived in the previous year was
22.8; 52.2% of migrants were aged between 17 and 26.6 These Wgures
also show that those aged between 20 and 21 had the highest probability
of migration.
6 Calculated from Berlin, Statistisches Amt (1887), table Vb.
1.6−1.8
1.4−1.6
>1.81.2−1.4
0.8−1.0
<0.8
1.0−1.2
Fig. 5.3. The natural rate of population increase 1871–80 (annual averages)
Source: As for Table 5.2.
Demography and Migration 129
![Page 139: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/139.jpg)
So, it is the earlier, ‘mottled’ map of 1871–80 which has more sign-
iWcance. The more pronounced regional pattern of 1900–10 was of little
consequence for migration before 1914.
3. A Comparison of Demographic Movements
in the East and the South
The discussion in the previous section implicitly treated demographic
factors as exogenous: as a pattern of behaviour which was unaVected by
economic circumstances and prospects. But there are ways in which
economic conditions can have eVects on population growth, and, once
this is accepted, the main point about the interaction between demo-
graphic factors and migration becomes a rather diVerent one. It is the
extent to which the natural rate of increase responds, or fails to respond,
to diVerences in relative economic opportunities between regions that
will aVect migration.
1.6−1.8
1.4−1.6
>1.81.2−1.4
0.8−1.0
<0.8
1.0−1.2
Fig. 5.4. The natural rate of population increase 1900–10 (annual averages)
Source: As for Table 5.2.
130 Demography and Migration
![Page 140: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/140.jpg)
Contemporary observers were well aware of mechanisms whereby
economic conditions could aVect both mortality and fertility. In 1879
von Fircks pointed to the coincidence of periods of rising mortality and
high cereal prices.7 German demographers also recognized the import-
ance of delayed marriage—the ‘Western European marriage pattern’—
for the control of fertility. Indeed, they were particularly conscious of
this, as Germany was very much on the frontier of this division: the
area of delayed marriage in central and eastern Europe followed the line
of German inXuence.8 In the German view, delayed marriage was a
consequence of cultural progress, and this was the reason why Ger-
many, and other parts of Western Europe, diVered in this respect from
lands further east: ‘In primitive tribes or in countries with a low level of
cultural development, such as Russia, Serbia or Bulgaria, the average
age at marriage is very low; marriage takes place shortly after puberty
for both sexes.’9
The same study argued that in Germany marriage was linked to
economic conditions. The custom which delayed marriage until the
couple were able to set up as an independent family unit meant that in
favourable economic circumstances this would be attained earlier, while
less good conditions would hold it back.10 Statistical analysis of mar-
riage rates in Prussian provinces, 1816–75, supports the view that eco-
nomic circumstances did inXuence marriage rates. High prices for foods
consumed by the rural labourers and their families, principally rye and
potatoes, were associated with lower marriage rates. However, high
prices for goods which were produced in rural areas but mainly con-
sumed by the better-oV or by city-dwellers, such as wheat and beef,
were linked to higher rates. Marriage was put oV when times were bad
and brought forward when times were good. The full analysis is given
in Appendix 5E.11
With such a relationship, when there was out-migration caused by a
rate of population growth which was excessive relative to available eco-
nomic opportunities, this would set in motion demographic changes
which would eventually reduce population growth in the regions of
7 Von Fircks (1879), 48.8 Including, for example, the future Baltic states, as can be seen from the map at the
end of Coale and Watkins (1986).9 Nadobnik (1908), 69.
10 Ibid. 80.11 Hohorst (1977) also found a relationship between marriage rates and economic con-
ditions for linen workers.
Demography and Migration 131
![Page 141: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/141.jpg)
out-migration. A crisis of overpopulation would cause a deterioration of
employment prospects. This would hold back marriage rates and act as
a stabilizing mechanism. The question then becomes: why did this
mechanism work in some regions and not in others?
The importance of this point is illustrated by the results of a classiW-
cation of German regions given in Appendix 5A and summarized in
Table 5.1. The data set is the one used in the previous chapter: net
migration is calculated using the indirect method of census comparison,
and the period of analysis is 1871–1910. The regions were divided into
four equal groups according to the percentage of the population occu-
pied in agriculture (1895 occupational census). These groups were then
subdivided into two according to an estimate of predicted net migration
resulting from population growth. This made use of the statistical an-
alysis reported in Appendix 4B (Table 4B.2).12 As before, the birth rate
is lagged four periods, or twenty years.
The classiWcation shows that there were considerable diVerences be-
tween the demographic records of the diVerent regions. In the high
population growth sub-group of the regions classiWed as highly indus-
trial, the diVerence between the lagged birth rate and the current death
rate was 1.66% per annum; for the low population growth regions it
was 1.29%. The predicted eVect from these demographic diVerences
amounts to a net diVerence of 137 persons per 10,000 resident popula-
tion (108 plus 29). The actual diVerence was 227 (565 minus 238). So,these demographic factors accounted for 42% of the observed diVerence
in net migration rates between the two sub-groups. At the other end, in
the highly agricultural group, the demographic eVect accounts for 48%
of the observed diVerence in net migration between the high and low
population growth groups.13 Demographic factors played a large part in
the determination of migration patterns.
Looking at individual regions, there are some major diVerences
between regions which are not so far apart geographically. In the highly
industrial group there are two comparisons of this type which stand
out: in the west, the Aachen region has one of the lowest rates of
population growth, the Arnsberg region one of the highest; in Saxony
12 The equation used in these calculations was the same as that used forTable 4.5: NETMIGjt ¼ 0:44 NETMIGjt�1 þ 0:26 DEATHRTjt � 0:22 BIRTHRTjt�4.The regions were divided into eight sub-groups of ten; the extra region is in the ‘moder-ately agricultural–low population pressure’ sub-group.
13 For the ‘moderately industrial’ group, however, the predicted eVect is much greaterthan the observed diVerence.
132 Demography and Migration
![Page 142: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/142.jpg)
the Zwickau region has a high rate of natural increase, while the Baut-
zen region is much lower.
In the ‘highly agricultural’ group there is a clear geographical div-
ision. Six of the ten ‘high population growth’ regions are in the east,
including the Wve highest; eight of the ten ‘low population growth’
Table 5.1. A classiWcation of 81 German regions according to the
availability of non-agricultural employment and population growth,
1871–1910, unweighted averages (migration and demographic variables
are per 10,000 head of population)
Regions classiWed as
Highly
industrial
Moderately
industrial
Moderately
agricultural
Highly
agricultural
All
regions
A. Regions with low population growth
% occupied in
agriculture
25.2 37.8 45.6 55.3 41.0
Average birth
rate (T-4)
3591 3588 3683 3496 3590
Average death
rate (T)
2305 2433 2555 2447 2435
Predicted
demographic
eVect
þ29 þ82 þ104 þ122 þ84
Actual net
migration
þ238 �206 �344 �469 �195
B. Regions with high population growth
% occupied in
agriculture
16.7 35.6 45.8 58.1 39.1
Average birth
rate (T-4)
4079 3725 3888 4050 3936
Average death
rate (T)
2414 2217 2386 2412 2357
Predicted
demographic
eVect
�108 �72 �63 �94 �84
Actual net
migration
þ565 �246 �567 �921 �292
Source: ClassiWcation of regions uses 1895 occupational census, Kaelble and Hohls (1989).
Demography and Migration 133
![Page 143: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/143.jpg)
regions are in the south, including the Wve lowest. The basic point is
clear: in the south population growth was adjusted downwards in
regions which had limited opportunities for non-agricultural employ-
ment; in the east this did not occur. Consequently there was more
migration out of the eastern regions.
The mechanism which achieved this was marriage restriction.
Figures 5.5 and 5.6 make this point. The Wrst shows the overall recordof population growth for the south and the east. It combines Wgures
1818
0
100
200
300
400
500
600
1832 1847 1862 1877 1892 1907
Deaths
Births
South : births
Natural increase
South : deaths
East : births
East : deaths
South : naturalincrease
East : naturalincrease
Fig. 5.5. Births, deaths, and the rate of natural increase, five-year average
annual rates, 1816–1910 (per 1,000 head of population)
Source: Calculated using data from Kollmann (1980) for the pre-1871 figures; the post-1871 figures come from the census publications. For this comparison Mecklenburg-Strelitz and Alsace-Lorraine were omitted from the regional groupings, thus ‘south’ isBavaria, Wurttemberg, and Baden.
134 Demography and Migration
![Page 144: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/144.jpg)
given by Kollmann for the pre-1871 period with Wgures produced by
the Imperial Statistical OYce for the period after uniWcation. Kollmann
has estimated stillbirths for periods when these are not available, so
there are reasons to treat the pre-1870 Wgures with some caution. How-
ever, the basic conclusions should not be seriously aVected. These are,
Wrst, that the natural rate of increase was fairly consistently higher in
the east, and secondly that this owed more to fertility than to mortality.
Mortality Xuctuated up to 1880 (rather more in the east than in the
1818
0.0
0.1
0.2
0.3
0.4
0.5
0.6
1832 1847 1862 1877 1892 1907
East: Im index
East: marriages
South: marriages
South: Im index
Fig. 5.6. Crude marriage rates and the Princeton Im index, 1816–1901, five-
year average annual rates
Note: Marriage rate is number of marriages per 100 head of population.Source: Data from Knodel (1974), Appendix table 2.1, and Kollmann (1980).
Demography and Migration 135
![Page 145: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/145.jpg)
south), and thereafter fell in both regions. Fertility was generally higher
in the east, apart from the 1871–85 period, when it was the same in
both areas.14
Figure 5.6 compares the marriage Wgures for the two regions, and
includes average values of the Im index used by the Princeton study of
population growth.15 This is a weighted index of the proportion of
women of diVerent ages who are married, the weights reXecting changes
in the natural fertility of women in diVerent age groups. It should
therefore represent an unbiased estimate of the potential eVect of mar-
riage on birth rates. However, the limitations of the available census
data meant that it could not be calculated before 1867. The crude
marriage rate, which is available, is a much less valuable measure. As it
includes remarriages, it is aVected by the death rate: higher death rates
lead to more remarriages of those who have lost partners. The same
problem applies to the average age of those getting married, if this does
not distinguish remarriages (which is the case with the Prussian
Wgures).
Figure 5.6 shows that there is a relationship between the crude mar-
riage rate and the Im index. The low marriage rate in southern Ger-
many before the 1860s clearly aVected the Wrst Im index. Even though
the southern marriage rate was rising in the 1860s this still left the
south well below the east. The southern Im index was 79% of the
eastern Wgure. A higher southern marriage rate in the years 1860–75meant that the gap narrowed, and this is the period in which the south-
ern birth rate rose to the eastern level. But in the late 1870s the south-
ern marriage rate fell again, and both the Im indices and the birth rates
moved apart.
The low birth rate of southern Germany owed something to legal
restrictions on marriage in Bavaria, but this was not the only factor.
While these contributed to the extremely low Bavarian Im index of
0.380 in 1867, the indices for Wurttemberg and Baden, where restric-
tions were lifted earlier, were also low (0.403 and 0.400 respectively). In
addition, while the lifting of the Bavarian restrictions in 1868 certainly
contributed to the high marriage rates in the following years, in the less
favourable economic circumstances of the late 1870s there was a return
to low marriage rates, which contributed to the fall in the southern Im
14 The reasons for high eastern fertility are examined in Section 3.15 All the Princeton indices for Germany are taken from the appendices in Knodel
(1974).
136 Demography and Migration
![Page 146: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/146.jpg)
index relative to the east. It seems therefore that the legal restrictions
reinforced a tendency to low marriage rates which had other causes as
well.16
Turning from the incidence of marriage to the level of fertility within
marriage (the Princeton Ig index), there is no sizeable diVerence be-
tween the east and the south on this measure for the period 1867–1910.
The southern average was slightly lower, 0.741 against 0.745, but thiswas balanced by a larger number of births outside marriage.
Statistical analysis of the percentage unmarried did not produce a
clear relationship between this variable and economic circumstances.
The extent to which marriage was sensitive to economic conditions
varied considerably between regions, especially once the non-Prussian
regions were included in the analysis. One reason for this may be that
some areas were restricting fertility within marriage. This was the con-
clusion reached by Knodel in his study of village fertility patterns. If so,
then these areas would have had less reason to put oV marriage in times
of economic hardship.17
Another statistical analysis which did not produce results was an
attempt to Wnd causation running from migration to marriage patterns.
If migration aVected the sex-ratio in regions of out-migration then it
might also aVect the likelihood of marriage for women in these regions.
The lack of any clear relationship between migration and the marriage
prospects of women is demonstrated by Figure 5.7 which gives the
proportion of women aged 49–54 who had never been married, by
region, from the 1880 census.
Figure 5.7 shows that there was no consistent relationship between
migration and the likelihood of marriage. Marienwerder, a region of
high out-migration, had one of the lowest proportions of never-married
women. In general the regions of out-migration in the east had no
tendency to have more never-married women than the recipient regions
of the centre. Some of the reasons for this can be understood from the
discussion in the previous chapter. Women did migrate, even if they
16 Works dealing with the eVect of the marriage restrictions in Southern Germany andAustria include Matz (1980), Ehmer (1991), and Mantl (1997). The eVect of these restric-tions may have been to increase illegitimacy rather than reducing the overall birth rate. Astudy of 5,263 applications in Wurttemberg during the restricted period showed that 6%were refused, but these couples may have applied successfully later, Matz (1980), 267.
17 Knodel (1988). This study of demographic behaviour in 14 villages shows consider-able divergence between the diVerent villages, but the sample is too small, and theregional coverage is too restricted, to draw general conclusions about the reasons for thesediVerences.
Demography and Migration 137
![Page 147: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/147.jpg)
made fewer moves than men.18 So women were quite capable of leaving
regions where prospects were poor. Moreover, some of the men who
made temporary moves out of their region of birth in their early twen-
ties returned later to be married in that region. Thus out-migration did
not inevitably lead to a decrease in the likelihood of marriage.19
But in the south and the south-west the position was a diVerent one.
Here there are extremely high proportions of never-married women.
The Bavarian Wgures are almost certainly aVected by the legal restric-
tions in force up to 1868, but high Wgures are also found in Baden,
18 Women also made temporary moves, Wnding employment as domestic servants, Leeand Marschalck (2000), 379, but men were more mobile in general, Hochstadt (1999),93–4.
19 Analysis of Kreis-level results for 1895 did indicate that there was an eVect frommigration on marriage rates inside Prussia by that date: see the note to Table 5D.2 inAppendix 5D.
12−14%
6−8%
<6%
8−10%
10−12%
14−16%
16−18%
18−20%
>20%
Fig. 5.7. Percentage of women aged 49–54 who had never been married,
by region, 1880 census
Source: Calculated from SDR (1881), a.f. 52, pp. 144–87.
138 Demography and Migration
![Page 148: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/148.jpg)
Wurttemberg, and Alsace-Lorraine.20 Figures in some western regions
are also above average. These include many of the areas where there
was high out-migration in the Wrst half of the nineteenth century.
This leads to a more general point. Looking at the history of migra-
tion in Germany over the whole nineteenth century it is possible to
identify two distinct movements of large-scale out-migration. The Wrst
was centred on south-western Germany, began in the 1830s, and
peaked in the 1850s. The second was centred on eastern Germany,
began in the 1860s, and peaked in the 1890s.
The Wrst was accompanied by clear signs of demographic distress.
The death rate was high and the marriage rate fell. Partly in conse-
quence, there was a rise in the number of illegitimate births. Kollmann
has referred to this as a period of ‘Ubervolkerungskatastrophe’. 21 In a
process not untypical of pre-industrial Western Europe, an excess
supply of labour relative to economic opportunities was eliminated,
partly by out-migration and partly by reduced marriages. The excess
supply of labour was most probably caused by falling mortality from the
late eighteenth century onwards.22
The second period has none of these characteristics. The death rate
was falling in all areas from the 1870s onwards, including the areas of
out-migration. Marriage rates were high and tending to rise. Illegitim-
ate births fell, and were below the national average in the east.
This is not to argue that all was well in eastern Germany. There is
evidence of poor conditions in an earlier period from a diVerent source,
which is the height of recruits, generally taken as a good indicator of the
nutritional standards of children and young adults.23 Figure 5.8 gives
the proportion of recruits rejected for being less than 5 feet tall, in the
20 Marriage restrictions may also account for the relatively high Wgure for Mecklen-burg, Lubinski (1995); for Bavaria see Lee (1977). The period after the ending ofthe marriage restrictions saw a gradual convergence of marriage rates in the south andsouth-west with those in other parts of the country, but this could also be explained bythe improvement in economic circumstances in these regions after 1870.
21 Kollmann (1976), 15. The reasons for this ‘overpopulation catastrophe’ havebeen debated: see inter alia Lee (1979 and 1980), Marschalck (1987), Kollmann (1974).One important question is the impact of agrarian reforms introduced in the earlynineteenth century. But it is quite possible to see it as a consequence of a period ofrelatively high population growth, set oV by declining mortality in the eighteenth century.
22 SchoWeld et al. (1991), Lee (1980), and Marschalck (1982) discuss this point.23 See, inter alia, Komlos (1995) and Baten (1999). There are other possible reasons for
the low stature of eastern recruits, such as the eVect of diseases, Leunig and Voth (1996),but in the absence of clear evidence of regional diVerences in disease epidemiology, it isreasonable to assume that the main cause of the variation in stature was nutritional.
Demography and Migration 139
![Page 149: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/149.jpg)
old Prussian provinces (post-1815 boundaries), taken from a survey in
the 1860s. This would obviously reXect conditions in the 1840s when
these recruits were born.
The range is a large one: fewer than 2% of possible recruits were less
than 5 feet tall in Berlin and Munster; in Marienberg 16.8% were, and
in Posen the Wgure was 22.8%.24 This is a period when the east was
attracting migrants from other parts of Germany. The map shows that
conditions in the eastern Prussian provinces in these years were harsh
and unappealing. It conWrms the view that only acute demographic
pressure in other parts of Germany could have led migrants to move to
these remote areas where living conditions were precarious and food
24 The Wgures are much higher for the Polish areas, which might reXect a disposition toreject Polish recruits, although putting down incorrect heights would have been an oddway to do this when there were other, less easily veriWable, reasons which could have beengiven (such as ‘bodily weakness’).
120−40
100−20
140−60
160−80
20−40
<20
40−60
60−80
>18080−100
Fig. 5.8. Numbers of possible recruits in the 1860s rejected for being less
than 5 feet tall, per 1,000, by Regierungsbezirke
Note: The unit of measurement is the old Prussian Fu�.Source: Meitzen (1868), 322–3.
140 Demography and Migration
![Page 150: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/150.jpg)
supplies often inadequate. What is striking is that, in the early part of
the century, migrants moved from relatively rich provinces to the
poorer ones in the east, but after 1860 they moved from the poorer east
to the richer parts in the centre and the west.25
There are, therefore, reasons to see the two migration movements as
representing diVerent types of labour surplus: the Wrst a pre-industrial
labour surplus, driven by demographic conditions, the second, a labour
surplus of the industrial period, not primarily caused by population
growth, but rather by changes in the relative demand and supply of
labour between sectors. Rising industrial demand for labour was sup-
plied by the release of surplus labour from agriculture.
4. Population Growth in the Regions of Origin
The argument has been made that the agrarian reforms of the early
nineteenth century created the conditions for a large increase in popula-
tion, and that this provided a stimulus to industrial growth. This is
sometimes referred to as the ‘Ipsen–Kollmann thesis’ after the two
authors who are most closely associated with this view. The eVect of
the land reforms has, however, been disputed. Some demographers,
including Marschalck and Lee, have argued that the intensiWcation of
work which, they argue, was brought about by the reforms contributed
to a worsening of child care and a consequent rise in infant mortality.26
There is a problem with the chronology of the Ipsen–Kollmann story
which is apparent from Figures 5.1 and 5.5 in the previous section: the
peak of German population growth in the years around 1900 is far
removed from the period of the land reforms. Rising population growth
after 1870 looks like a response to industrialization, not a cause. It can
be argued that the surge in marriages and in births in 1816–22 owed
something to emancipation (though the ending of the Napoleonic wars
is also a possible explanation). This, in turn, meant that there was a
ready supply of labour for aspirant industrialists, as well as providing
25 Kollmann’s Wgures (calculated on a three-year basis using census returns) show thatthere were substantial gains in East and West Prussia, Silesia, and Posen through migra-tion between 1816 and 1846. It is not possible to say exactly where these migrants camefrom, but there was out-migration from Brandenburg in 1816–31, Prussian Saxony in1820–31, and Westfalia in 1831–7 and again in 1840–6. Kollmann (1980).
26 Kollmann (1976), Lee (1979a), Marschalck (1976); see also Harnisch (1977).
Demography and Migration 141
![Page 151: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/151.jpg)
the labour required for the expansion of the area under cultivation in
eastern Germany in the Wrst half of the century.
The emphasis placed on emancipation has behind it a ‘top–down’
view of population growth: peasants and rural labourers will marry
early, and have more children, unless laws or other restrictions imposed
from above prevent them from doing so. But this underestimates the
ability of rural society in Western Europe to regulate population growth
in its own interests. Knodel’s 1988 study of fourteen German villages
showed that some were restricting fertility within marriage as early as
1800, while the average age of women at Wrst marriage was raised from
25.5 in 1700–49 to 26.9 in 1825–49.27
This was not exactly a benign process. The burden fell heavily on
women, particularly if reduced marriage opportunities led to an increase
in illegitimate births. The high mortality of illegitimate children (84%
higher than for those born within marriages in Prussia 1875–1900) iseloquent testimony to the social pressures placed on unmarried
mothers.28
If rural society was able to exercise a crude but eVective control over
population increase, then it is possible to interpret the increased birth
rate which followed the agrarian reforms as a response to a perceived
improvement in economic opportunities. But for some these hopes were
disappointed. In the 1830s bad harvests and high prices brought about a
worsening of the condition of the landless agricultural labourer. How-
ever, from 1860 onwards rising non-agricultural demand for labour
brought about a recovery, both in real wages and in births.29
Agricultural institutions have consequences for the fertility decisions
of households. One example is the contractual labour relationship typ-
ical on large estates in the east. The Inste or Deputate received a small-
holding from the landlord (from which they derived much of their
income) in return for an obligation to provide labour at harvest time.
This obligation was not just to provide the labour of a single person,
but of additional Scharwerker, who could be family members or Knechte
or Magde (other labourers living in the Inste household). For a house-
hold living in this system it was valuable to have children who could
27 Knodel (1988), 125.28 Calculated from Wgures in SJPS (1915), 55. The Wgures show that on average 55.5%
of illegitimate children died in their Wrst year, while 19.5% of those born within marriagesdied. This was not just a feature of rural society; the diVerence was just as large for thoseborn in cities.
29 The role of economic factors in population growth is discussed in Birdsell (1988).
142 Demography and Migration
![Page 152: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/152.jpg)
provide Scharwerker labour; it was also necessary that, if the family
were to retain the smallholding when the parents were too old or inWrm
to work during the harvest, there should be children willing and able to
perform these duties.30
A second institution which may aVect fertility decisions is partible
inheritance. There are reasons to expect this to aVect marriage decisions
in particular. Studies show that landless labourers married later, so any
institution which spread land ownership more widely should lead to
earlier marriage and consequently higher fertility. But there are factors
which might work against this. If the main concern of parents when
making fertility decisions was the prospect of support from their chil-
dren in old age, then there was a risk that, with partible inheritance,
large family sizes could lead to a position where none of the children had
holdings large enough to provide eVective support to their parents. Ex-
cessive subdivision of holdings would not be in anyone’s best interest.31
To examine the relationship between these institutional factors and
fertility a data set was prepared which was based on the 1895 census
results for the Prussian Kreise. Limiting the analysis to Prussian terri-
tories had the advantage that the legal framework would be the same for
all areas. It also meant that the results of the 1892 and 1901 surveys of
Ortsubliche Tagelohne could be used for rural wages. Finally, Wgures for
30 Weber (1895), 185; see also Harnisch (1984) for a diVerent perspective on tenurialrelations in eastern Germany, but which also stresses the importance of the post-reformsystem as a source of labour for industrialization.
31 See Berkner and Mendels (1980) for a survey of the literature on this point.
Table 5.2. Summary of Wndings from statistical analysis of census
results for 424 rural Prussian Kreise 1895: analysis of total fertility
Factors which increase total fertility (numbers of children under 1 year relative
to the total number of women of child-bearing age)
I. Secure statistical results
A large Polish-speaking proportion of the population
Presence of large farms and estates
High levels of rural wages
Low levels of partible inheritance
II. Less secure statistical results
A high level of Catholicism
A large Lithuanian-speaking proportion of the population
Demography and Migration 143
![Page 153: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/153.jpg)
partible inheritance were available from an 1896–1902 Prussian survey
which recorded what was actually happening to property which
changed hands as a result of inheritance. This is preferable to the use of
a classiWcation of regions according to an assessment of the predominant
system which does not record actual practice, since practice may diVer
from custom. It is the actual splitting of holdings which is likely to have
the strongest eVect on demographic behaviour.32 However, it does
mean that the main areas of partible inheritance in the south and south-
west are excluded from the analysis.
The econometric analysis of this data set is reported in Appendix 5D.
The results are summarized in Tables 5.2, 5.3, and 5.4. The Wrst of
these shows the results of the analysis of total fertility: the numbers of
children aged under 1 year relative to the total number of women aged
15–50. The main Wndings are that total fertility is highest in areas with
a high Polish-speaking population, in areas where wages are relatively
high and in areas where there were large farms and estates. In the rural
east, the presence of large numbers of Poles, and of larger estates,
would tend to raise fertility, but lower wages would have the opposite
eVect. These results support the view that the institutions of east-
Elbian estates, in particular the Scharwerker system, did raise fertility,
and thus lead to out-migration.
Partible inheritance, on the other hand, does not appear to have led
to an increase in fertility; on the contrary, fertility was lower in the
Table 5.3. Summary of Wndings from statistical analysis of census
results for 424 rural Prussian Kreise 1895: analysis of marriage rates
Factors which increase the proportion of women aged 15–50 who are married
I. Secure statistical results
A low level of Catholicism
A large Polish-speaking proportion of the population
High levels of rural wages
The proximity of cities (over 50,000 inhabitants)
II. Less secure statistical results
High levels of partible inheritance
32 The survey was published in ZdKPSB (1905); studies of inheritance systems includeMiaskowski (1882) and Sering (1899–1910). In the partial inheritance areas of southernGermany the problem of excessive subdivision aVected the attitudes of parents to lateroVspring, and may have contributed to high infant mortality. Lee (1981), 33 and 96.
144 Demography and Migration
![Page 154: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/154.jpg)
Prussian areas which practised divided inheritance. If partible inherit-
ance led to earlier marriage, then this did not feed through into higher
birth rates. The analysis of marriage rates (summarized in Table 5.3)shows that partible inheritance did tend to be associated with a higher
marriage rate, though the result is not a particularly secure one, and the
size of the eVect is not large. The analysis also shows that Catholic areas
tended to have low marriage rates, outside the Polish-speaking areas.
This indicates that marriage restriction may not have been the only
reason for low marriage rates in Catholic areas in the south and south-
west: Catholics inside Prussia also had lower marriage rates.
Marriage rates were higher in areas with good economic prospects: in
Kreise with high wages and in areas close to major cities which would have
beneWted from better access to urban markets and improved employment
opportunities. However, the proximity of major cities did not lead to a
rise in overall fertility. If the measure of fertility is altered to reXect the
impact of variations in the marriage rate—dividing the number of chil-
dren under 1 year by the number of women aged 15–50 who are
married—then the statistical analysis shows that fertility on this measure
was lower in areas close to major cities and higher in remoter areas (Table
5.4). This Wgure cannot be regarded purely as a measure of fertility within
marriage since it will include children born outside marriage as well (these
cannot be distinguished from the census results as published). However, it
Table 5.4. Summary of Wndings from statistical analysis of census
results for 424 rural Prussian Kreise 1895: analysis of fertility relative to
the number of married women
Factors which increase fertility measured by the number of children under
1 year relative to the number of married women of child-bearing age
I. Secure statistical results
A large Polish-speaking proportion of the population
A high level of Catholicism
Presence of large farms and estates
High levels of rural wages
Low levels of partible inheritance
Large distances from cities (over 50,000 inhabitants)
II. Less secure statistical results
A large Lithuanian-speaking proportion of the population
Demography and Migration 145
![Page 155: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/155.jpg)
will reXect changes in marital behaviour. The Wnding suggests that
married couples in areas close to cities were restricting fertility and this
was cancelling the eVect of higher marriage rates
Similarly, in areas of partible inheritance, the tendency for marriage
rates to be higher was cancelled by low rates of fertility relative to
married women. The fear of excessive subdivision may have led married
couples in these areas to restrict family size. Outside the Polish area,
Catholic families followed the opposite strategy: high rates of fertility
relative to married women were countered by low marriage rates.
There were therefore three factors which had strong eVects on demo-
graphic behaviour, but which had less eVect on overall fertility, and
hence on overall population growth, largely because of cancelling eVects
from marriage rates on the one hand, and reproductive behaviour
within marriage on the other. These factors were Catholicism, the prox-
imity of large cities, and partible inheritance. This leaves three other
factors which did have substantial eVects on overall fertility: high
wages, the presence of large estates, and the Polish-speaking percentage.
In these cases, the eVect on marriage rates and the eVect on fertility
relative to married women both ran in the same direction.
The impact of this on overall population growth is illustrated in
Table 5.5, which gives predicted values for diVerent areas, distinguish-
ing between the implied rate of fertility in areas dominated by larger
estates and that for areas where smaller holdings predominated.33
33 These are illustrative Wgures, which assume that the behaviour of the non-agricultural population in Kreise dominated by larger estates followed the example set by
Table 5.5. Predicted values for total fertility (numbers of
children aged under 1 year per 1,000 women of child-bearing age)
On large farms
and estates
(over 100 hectares)
On other holdings
(under 100 hectares)
Average values 66.0 57.9
Polish areas 79.6 71.5
Non-Polish areas 64.3 56.2
Daily wages of 1.51 Marks 68.4 60.3
Daily wages of 1.08 Marks 63.3 55.2
Note: The predictions use the coeYcients given in column a of Table 5D.1.
146 Demography and Migration
![Page 156: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/156.jpg)
Outside the Polish areas, fertility in the rural east was held back by low
wages: the predicted level for a larger estate paying the average wage
east of the Elbe (1.08 Marks) is only slightly higher than that for a
smaller holding in the western area where wages reached the average
for Kreise west of the Elbe (1.51 Marks). However, in the Polish areas
population growth was much higher, despite low wages.
The previous section raised the question why population growth was
high in the east even in regions which had few opportunities for non-
agricultural employment, creating a labour surplus which had to
migrate to Wnd work. A contrast was drawn between the east and simi-
lar regions in the south, where population growth was restricted.
The analysis presented in this section provides answers to this ques-
tion. It shows that there was a tendency for total fertility to fall in
regions where wages were low, but the operation of this adjustment
mechanism in the east was limited by other structural factors:
1. A higher proportion of women were married in the east; the east
does not appear to have practised marriage restriction as a means
of population control to the same extent as the German-speaking
Catholic areas of southern and western Germany, or the Catholic
areas inside Prussia.
2. The Catholic population of the east was predominantly Polish-
speaking; this minority had high fertility.
3. There was a tendency for fertility to be higher in rural areas
where the farm structure was dominated by large farms and
estates; this was much more common in the east.
4. Partible inheritance was rare in the east; this was a factor which
tended to reduce overall fertility.34
But the demographic experience of the east can be seen in a wider
context. Both partible inheritance and the presence of large farms and
estates are related to the overall distribution of property and income. In
practice, inequality in the distribution of income in rural areas is closely
the estate labourers. This may, however, have been the case. If the presence of largeestates led to a change in demographic behaviour only on these estates, then an interactionterm, combining the percentage of land in larger holdings and the percentage of the totalpopulation occupied in agriculture should have a signiWcant and positive eVect on overallfertility. A variable of this type was tried, but it had no signiWcant eVect.
34 The evidence suggests that partible inheritance areas were quicker to move to someform of fertility restriction within marriage in the late nineteenth century. In the earlierpart of the century, this pattern of behaviour was not widespread in rural areas, so that thedemographic problems of the south and south-west led instead to a policy of fertilitycontrol by marriage limitation (which was encouraged by legislation).
Demography and Migration 147
![Page 157: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/157.jpg)
linked to inequality in the distribution of land. And, over time, partible
inheritance should tend to reduce inequality in property distribution.
Modern demographic studies of developing countries have shown a link
between high inequality and population growth. The Wndings reported
in this section point in the same direction.35
5. M igration and Population Growth in Rural Areas
The eVect of population growth on migration can be studied using a
data set for the Prussian Kreise prepared by the Prussian Statistical
Bureau in the interwar years and used in an article by Peter Quante,
published in 1933.36 This made estimates of the rate of migration out of
the population occupied in agriculture by applying to the agricultural
population rates of fertility and mortality recorded for the population
residing in rural areas. These rates of natural increase were then used to
derive estimates of net migration, using the indirect method (subtract-
ing the estimated rate of natural increase from the actual increase as
recorded by the occupational censuses).37
There are some problems with this approach. The population occu-
pied in agriculture was a large component of the total population resi-
dent in rural areas but there were others, occupied outside agriculture,
who also lived in rural areas. If rates of mortality or fertility for this
rural, non-agricultural, population were to diVer markedly from those
for the agricultural population, there would be a bias in the estimates.38
The second problem is that, although the data set is described as one
which gives migration out of agriculture (landwirtschaftliche Abwander-
ung) it is not in fact one which necessarily records migration. Someone
who switched from agricultural employment to a non-agricultural occu-
pation without leaving the Kreis would be recorded as a move. Given
the nature of agricultural employment in the late nineteenth century,
35 The empirical evidence, and the theoretical reasons for expecting such a link, aresurveyed in Birdsell (1988).
36 Quante (1955); the data was also used in Quante (1958).37 These estimates are derived from the occupational censuses, which were conducted
in June. These Wgures would therefore include some movements which were purelyseasonal (such as migration to sugar beet areas to do hoeing and other cultivations)although the main period of seasonal migration came later in the year, at harvest time.
38 In 1907 55.2% of the German population was occupied in agriculture, while 41.5%lived in rural areas (taking an average of the 1905 and 1910 censuses). So around 15% ofthe rural population had an occupation outside agriculture.
148 Demography and Migration
![Page 158: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/158.jpg)
which typically involved living on the farm where the worker was
employed, a switch to non-agricultural work would usually involve a
move of some description, but this could be no more than a shift to a
nearby town or village. The data set is best described as one which
gives the net agricultural ‘quitting rate’: the rate at which persons occu-
pied in agriculture switched to other occupations. The results of this
analysis are summarized in Table 5.6; the full econometric results are
given in Table 5D.3 in Appendix 5D.
The most striking result from this analysis is the close relationship
between natural population increase and migration. The estimated rela-
tionship indicates that at least 100% of any population increase in rural
areas was exported to other sectors. There was no capacity to absorb
additional labout within agriculture. Any population growth fed directly
into migration (at least within the Kreis).
The quit rate was heavily inXuenced by the position of the Kreis; itwas much higher in areas close to the border with Russian Poland and
fell progressively for Kreise further and further to the west. This would
have led to a high rate of migration out of the east but for the tendency
for the quit rate to be higher on smaller holdings than on larger estates.
Both the presence of large estates (over 100 hectares), and, to a lesser
extent, the presence of medium-sized holdings (20–100 hectares) are
associated with lower quit rates than those on smaller holdings of below
20 hectares.
Table 5.6. Summary of results of statistical analysis of moves out of
agriculture for 390 rural Prussian Kreise, 1882–95
I. Secure statistical results
Factors which increase net moves out of agriculture:
A high rate of natural population increase
Being closer to the border with Russian Poland
Factors which decrease net moves out of agriculture or lead to net gains:
Presence of large farms and estates (over 100 hectares)
Presence of medium-sized farms (20–100 hectares)
A large Polish-speaking proportion of the population
A large Lithuanian-speaking proportion of the population
II. Less secure statistical results
Factors which increase net moves out of agriculture:
High levels of non-agricultural employment in the Kreis
Demography and Migration 149
![Page 159: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/159.jpg)
This Wnding contrasts strongly with the view expressed by many
contemporary writers that large estates led to a higher rate of out-
migration. As the previous section has shown, one explanation of this is
that large estates were associated with higher birth rates and it was this,
not any tendency to shed labour as such, that produced an association
between large estates and out-migration. This issue will be considered
in more detail in the next chapter.
Quit rates were also lower in the Polish and Lithuanian areas. In
these areas population pressure was high, particularly so in Polish-
speaking areas, and this was the main factor driving out-migration.
Table 5.7 gives the predicted quit rates for the 1882–95 period for
Kreise at diVerent distances from the border with Russian Poland, con-
trasting the predicted rates for smallholdings and larger estates. In areas
close to the border with Russian Poland the predicted rates are consid-
erably higher than in the centre of Germany (approximately 500 kmfrom the border). The predicted quit rates are also much lower for the
Polish areas, but the calculations are based on the assumption that the
rates of natural population increase were similar, and they were in fact
much higher in these areas. If the Polish-speaking population had had a
similar rate of population increase to the Germans living in the rural
east, then Polish migrants would have been a relatively unimportant
factor in the overall pattern of German migration.
Table 5.7. Predicted quit rates (per 1,000 of the total agricultural
population) for rural areas with similar rates of natural population
increase, based on analysis of Wgures for Prussian Kreise 1882–95
On large farms and estates
(over 100 hectares)
On small farms
(under 20 hectares)
Average values 166.5 313.4
Distances from the
border with Russian
Poland:
50 km 225.8 372.7
500 km 148.0 294.9
Polish areas 69.2 214.9
Non-Polish areas 182.6 328.3
Note: The calculations use the coeYcients from column c of Table 5D.3, and averagevalues for all other variables.
150 Demography and Migration
![Page 160: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/160.jpg)
On average the quit rate for Kreise east of the Elbe was 21% higher
than for Kreise west of the Elbe (284.5 per 1,000 versus 235.5 per
1,000). About half of this can be attributed to higher rates of population
increase feeding directly into migration. The remainder was due to
other factors. Chief amongst these was the proximity of the border with
Russian Poland. Cheap seasonal migrant labour from Russian
Poland could be used to displace permanently employed workers. The
availability of this labour source lowered wage rates in the areas close to
the border, providing an additional reason for farm labourers in these
regions to look for employment elsewhere in Germany.
6. Population Growth in the Recip ient C it ies
The regressions in Chapter 4 showed that the rate of natural population
increase of recipient cities was an important inXuence on migration
rates into those cities. Table 5.8 compares lists of the 20 cities with the
highest rates of natural population increase and those with the lowest
rates (selected from a list of 96 for which Wgures were available).
The signiWcance of these diVerences is evident when compared to the
average rate of growth of German cities in this period, which was 2.6%
per annum (total increase of population from all sources). The cities in
the Wrst list could have supplied much of this increase from the natural
increase of the resident population; those in the second list would have
needed to take in migrants.
A few points emerge from this comparison. The cities in the Wrst list
are predominantly industrial cities, with a heavy concentration in the
Ruhr. Those in the second list are well scattered around the country,
and tend to be longer-established cities, often the capitals of their
regions, or cities which for other reasons might be expected to have a
strong middle-class presence.39
There are two views of the demographic relationship between the
cities and their surrounding regions. The Wrst stresses the leadership
role of the cities, and of the middle class within these cities. Middle-
class attitudes and behaviour would Wrst aVect the working class living
within the cities, and then transfer to the surrounding rural areas. The
39 Laux (1989) shows that natural population growth rates were lower in cities whichwere predominantly commercial or administrative, and these cities tended to have higherrates of in-migration in consequence.
Demography and Migration 151
![Page 161: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/161.jpg)
second view is the one expressed by Kollmann: migrants from rural
areas bring their behaviour patterns with them to the cities.
The results reported in Tables 5.2, 5.3, and 5.4 showed that there
was a relationship between demographic behaviour and the proximity of
major cities. In areas nearer to cities the marriage rate was higher, but
fertility relative to the number of married women was lower compared
to more remote regions. This could be interpreted as supporting the
view that changes in demographic behaviour were initiated in the cities
and then moved outwards to contiguous areas. However, the analysis
also indicated that fertility in rural areas responded to factors which
were speciWc to those areas: the presence of large estates and the
Table 5.8. German cities with highest and lowest rates of natural
population increase 1908–12
20 cities with highest rates 20 cities with lowest rates
Annual rate of
increase (%)
Annual rate of
increase (%)
Hamborn 3.16 Potsdam 0.28
Borbeck 3.09 Hildesheim 0.34
Herne 2.93 Regensburg 0.47
Gelsenkirchen 2.93 Ulm 0.51
Recklinghausen 2.63 Gorlitz 0.51
Oberhausen 2.61 Mulhausen 0.62
Bochum 2.51 Frankfurt a. Oder 0.66
Konigshutte 2.31 Metz 0.70
Essen 2.21 Braunschweig 0.72
Gleiwitz 2.18 Liegnitz 0.73
Dortmund 2.12 Dessau 0.75
Dusseldorf 2.10 Rostock 0.76
Rixdorf 2.04 Koblenz 0.76
Mulheim a.d. Ruhr 1.99 Bromberg 0.77
Mannheim 1.92 Wiesbaden 0.79
Plauen 1.88 Berlin 0.82
Ludwigshafen 1.86 Solingen 0.83
Duisburg 1.86 Augsburg 0.83
Saarbrucken 1.75 Wurzburg 0.84
Osnabruck 1.74 Brandenburg 0.84
Source: Calculated from SJDS (1911), table xii, pp. 56–7, and subsequent issues.
152 Demography and Migration
![Page 162: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/162.jpg)
institutions associated with these in eastern agriculture, and the preva-
lent inheritance system. Demographic behaviour outside cities was not
just a lagged response to the changes taking place in urban areas.
If migrants from rural areas did transfer demographic behaviour
from their regions of origin then it should be possible to Wnd this eVect
by statistical analysis. The birth rate in rural areas was higher than in
cities, and the diVerence was increasing. The crude birth rate in Prus-
sian cities was 97.6% of the rural level in 1876–80 but only 82.5% in
1906–10.40 Despite this it was not possible to Wnd a statistical relation-
ship between migration and urban fertility. Two data sets were tried.
The Wrst made use of the 1895 census results for the major German
cities (showing the proportion of the population born in rural areas);
the second used Wgures for movements into German cities in the period
1905–10.41
One reason for this may be that the demographic behaviour of mi-
grant families diVered in ways which tended to cancel each other out.
Migrants tended to marry later, which should have reduced overall
fertility.42 But they may have been more reluctant to adopt ‘modern’
patterns of reproductive behaviour, which would have pushed fertility
up. Studies based on household Wgures show relatively small diVerences
in the size of completed families.43
Demographic behaviour was inXuenced by conditions within the cities.
Table 5D.4 in Appendix 5D gives the results of an econometric analysis of
the birth and death rates in 79 German cities in 1910–12. These are
summarized in Table 5.9. The main inXuences came from the occupa-
tional structure of the cities. In cities which had high levels of industrial
employment—which also meant relatively low levels of employment in the
professions and other middle-class activities, and in domestic service—the
40 SJPS (1915), 55–7.41 Using the 1905–10 data set, no signiWcant coeYcients were found either for net
migration or total moves per head of population. The dependent variable was the crudebirth rate. For 1895, a relationship was found between the average marital fertility of theregions from which migrants were drawn (using the Princeton Ig index) and maritalfertility in the recipient cities. The problem here is that most migrants were drawn fromthe provinces within which the cities were situated, and so this result could also beinterpreted as demonstrating the inXuence that demographic behaviour in cities had onthe surrounding countryside. An important drawback with this approach is that it doesnot consider the actual behaviour of migrant families, only the aggregate Wgures for citieswith high levels of migration. A family-level study may well show diVerent results.
42 Lee and Marschalck (2000), 383–4.43 Lee and Marschalck (2000), 384 report that completed family size was 4.3 children
for in-migrants in the city of Bremen 1820–75, but 4.6 for native-born households.
Demography and Migration 153
![Page 163: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/163.jpg)
birth rate tended to be higher. It was higher still for the cities dominated
by ‘heavy industry’, mining and metal production.
There was an association between low birth rates and a high vote for
the Social Democratic party in the 1912 election. These were both
indicators of a change in working-class attitudes. Arguably, both were
signs of a greater self-conWdence: an attempt on the part of working-
class households to control or at least inXuence their own destiny. They
were no longer prepared to accept that their lives should continue to be
dominated by forces outside their control. This turned out to be a more
important and consistent eVect than Catholicism, which had little eVect
on demographic behaviour once industrial structure had been included
in the statistical analysis.
Death rates were also inXuenced by industrial structure and by wage
levels: low death rates were found in cities with relatively high levels of
industrial occupation compared to other forms of employment, and in
Table 5.9. Summary of results of statistical analysis of birth rates
and death rates in 79 German cities, 1910–12
Secure statistical results
I. Factors which are associated with high birth rates:
A high level of industrial employment relative to other occupations
A high level of employment in mining and metal production relative to
other industrial occupations
A low Socialist vote in the 1912 election
II. Factors which are associated with low death rates:
High wages
A high level of industrial employment relative to other occupations
A low level of employment in mining and metal production relative to
other industrial occupations (less secure result)
III. Factors which are associated with high rates of natural increase:
A high level of industrial employment relative to other occupations
A high level of employment in mining and metal production relative to
other industrial occupations
High wages
A low Socialist vote in the 1912 election
Factors which have no consistent eVect on the rate of natural increase:
Catholicism
A high level of migration
154 Demography and Migration
![Page 164: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/164.jpg)
cities with high wages. On the other hand, the heavy industry sector
was associated with higher death rates when compared to other indus-
tries (although the result is not a particularly secure one in statistical
terms). This was most probably due to the nature of the work, and the
illnesses and accidents which resulted from this.
The rates of natural increase which resulted from these birth rates and
death rates show eVects from the same factors: population growth was
highest in industrial cities, with high wages, and relatively low socialist
votes. The positive eVect of high birth rates in the ‘heavy industry’ cities
outweighed the tendency for these to have higher death rates, so the
overall eVect from this factor on population growth was a positive one.
Again the analysis shows no consistent eVect from Catholicism, and
migration, as measured by moves in and out relative to population, was
not found to have any eVect on demographic behaviour.
Table 5.10 illustrates these points for ten selected cities, those with
the highest and lowest rates of population increase. The list of cities
with low rates of increase includes several with small Catholic percent-
ages, but alongside them is Muhlhausen, which was 75.8% Catholic.
Table 5.10. Predicted rates of population increase (per 100,000
resident population) for selected cities, 1910–12
Catholic
%
Wages
1912
Industry
%
Mining/
metal
%
SDP
vote
1912
Rate of
increase
A. 5 cities with lowest rates of increase
Potsdam 8.7 3.25 32.8 0.15 47.4 23.3
Gorlitz 14.0 2.50 49.7 0.13 48.8 38.7
Mulhausen 75.8 2.40 63.3 0.01 56.1 40.3
Wiesbaden 31.8 3.20 37.4 0.13 35.8 50.7
Brunswick 6.5 3.20 50.2 0.13 55.6 53.7
average 27.3 2.91 46.7 0.11 48.7 41.3
B. 5 cities with highest rates of increase
Beuthen 55.1 3.40 66.4 42.8 35.7 189.7
Duisburg 52.8 3.25 65.8 33.0 30.9 192.3
Bochum 50.4 3.30 68.2 45.8 36.8 229.0
Gelsenkirchen 49.7 3.40 74.3 56.5 36.8 252.0
Recklinghausen 75.2 3.30 73.5 64.2 27.3 262.5
average 56.6 3.33 69.6 48.5 33.5 225.1
Demography and Migration 155
![Page 165: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/165.jpg)
Other Catholic cities with low rates of population increase were
Wurzburg, Augsburg, and Koblenz. It is this inconsistency that makes
the relationship between Catholicism and population growth such a
weak one.
On the other hand the low growth cities are distinguished by gener-
ally low levels of industrial employment, relatively low wages and
high SDP votes. But above all, they are not cities which have large
mining or metal production sectors. By contrast the cities with high
rates of population growth are all ones where the heavy industry sector
was important. They tended to have lower socialist votes, but wages
were high.
The classiWcation of regions given in Appendix 5A shows that, over
the whole 1871–1910 period, comparison of the death rate and the
lagged birth rate identiWes the Ruhr regions, Arnsberg and Dusseldorf,
as two of the regions which were both highly industrial and which had
high population growth. But a comparison with other industrial regions
shows that the Ruhr was not so unusual when values for the whole
1871–1910 period are considered. Population pressure was as high in
parts of Saxony, Leipzig, and Zwickau, for example. What made the
natural rate of increase in the Ruhr so distinctive is that it remained
high while population growth in other urban areas was tending to fall.
The more isolated working-class culture of the Ruhr was a factor in this
delayed demographic transition.
7. Population Growth and the Lewis Model
So far the analysis in this chapter has concentrated on relative diVer-
ences in population growth between regions. For migration this is the
most important factor. Is the pattern of demographic change in line
with the distribution of economic opportunities? If it is not, then there
will be an adjustment role for migration.
But, in the context of the Lewis Model, the aggregate level of popu-
lation increase also matters. Lewis argued that demographic response
would be a component of the ‘unlimited supplies of labour’ assumed by
his model. He placed most emphasis on the decline in the death rate:
Because the eVect of development on the death rate is so swift and certain,
while its eVect on the birth rate is unsure and retarded, we can say for certain
that the immediate eVect of economic development is to cause the population to
grow; after some decades it begins to grow (we hope) less rapidly. Hence in any
156 Demography and Migration
![Page 166: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/166.jpg)
society where the death rate is around 40 per thousand, the eVect of economic
development will be to generate an increase in the supply of labour.44
The reasons Lewis gave for the ‘swift and certain’ decline in the death
rate were, Wrst, that improved communications would reduce the inci-
dence of local famines, and secondly, that there would be improvements
in public health and medicine.
The relationship between development and the death rate does not
appear quite so certain now as it did to Lewis: rather more is known
about high urban mortality in industrializing cities. Life expectancy in
German cities remained low into the twentieth century. Table 5.11
gives life expectancy at birth for the Prussian population, as calculated
by the Prussian Statistical OYce using mortality tables for 1901–5. This
44 Lewis (1954), 144.
Table 5.11. Life expectancy at birth, Prussia 1901–5
Men Women
Towns
and cities
Rural
areas
Towns
and cities
Rural
areas
East Prussia 37.6 42.0 44.8 45.2
West Prussia 37.5 43.4 42.9 45.9
Berlin 41.8 - 47.3 -
Brandenburg 41.7 43.2 47.0 47.5
Pomerania 37.9 46.5 43.6 48.6
Posen 38.7 45.5 43.9 48.5
Silesia 36.5 40.0 41.7 43.5
Pr. Saxony 41.1 44.9 45.2 47.7
Schleswig-Holstein 44.9 52.9 49.7 54.8
Hanover 44.7 50.3 49.4 51.4
Westphalia 41.7 46.6 46.6 48.6
Hesse-Nassau 45.2 49.6 49.5 50.9
Rhineland 42.9 45.8 47.6 47.7
Note: The reasons for high mortality in the east are discussed by Spree (1988), 58. Seealso Hennock (2000) for a discussion of attitudes to urban sanitation. These Wgures mayunderstate the problems of eastern cities. Appendix 5B gives life-expectancy Wgures forindividual cities. The results show that male life expectancy in the major eastern citieswas 29–35 years. There had been an improvement in urban life expectancy since the1870s, but this was still below rural life expectancy.Source: Zdp/SB (1908), i. 26–7.
Demography and Migration 157
![Page 167: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/167.jpg)
table shows that in all regions, and for both sexes, life expectancy in
urban areas was below that in the countryside, and the gap could be a
substantial one: 7 years for men in Posen, 81⁄2 years in Pomerania.
National averages tend to obscure this diVerence, as the urban popula-
tion was concentrated in the west and centre, regions where life expect-
ancy was high, while the east had proportionately more of the rural
population. The ‘centre of gravity’ of the rural population was further
to the east compared to that of the urban population.45
One evident conclusion is that, in the absence of balancing variations
in fertility, cities would have to draw in migrants from rural areas to
make up for their higher mortality rates. Lower rural mortality should
have made it easier to Wnd migrants from the countryside nearby.
Migration over longer distances had a mixed eVect on mortality. On
the one hand, it transferred population from rural areas to cities, which
would tend to increase mortality. On the other, it moved population
from the east to the west, which should have led to a reduction. But, as
most moves were over shorter distances, and did result in net gains by
cities, it can be seen that the overall eVect would have tended to raise
mortality.
If urbanization tended to raise mortality, then the decline in the
German death rate from 1870 onwards becomes even more impressive.
However, by European standards German life expectancy in the period
just before the First World War was not high. At 49.0 years in 1910–11(both sexes) it was below France (50.4), England and Wales (53.5), and
the Netherlands (56.1). In Sweden, Norway, and Denmark life expect-
ancy was over 57 years. Comparative studies show that the decline of
mortality in Germany was relatively late; in other Western European
countries it had started earlier.46
45 Life-expectancy tables for the early 1900s show a substantial diVerence betweenurban and rural areas for the male population at all ages, while there was little diVerencefor women past the Wrst year of life (urban infant mortality was high for both sexes). Formen the gap was 3.8 years at birth and 3.6 years at age 10, Vogele (1998), 50–1. Thiscalculation, for the population as a whole, underestimates the gap between urban areas andcontiguous rural areas.
46 SchoWeld et al. (1991), 47. There had been considerable progress in reducing mortal-ity in German cities since the 1870s: see Vogele (1998) and Vogele and Woelk (2000) forstudies of this. The fact that German life expectancy was still not high by Europeanstandards is an indicator of the severity of the problems faced in the 1870s and 1880s.The high rates of migration in this period had contributed to a major crisis of mortality,which was only gradually overcome. By 1905, however, migrant life expectancy in themore advanced cites was below that of the native-born population, Lee and Marschalck(2000), 381.
158 Demography and Migration
![Page 168: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/168.jpg)
On the other hand, Lewis’s ‘uncertain and retarded’ decline in fertil-
ity set in fairly quickly. The Princeton study showed that fertility
within marriage started to fall from the late 1870s onwards. This was
initially oVset by a rise in the proportion married. If there had not been
this decline in marital fertility, then population growth in the 1880s and
1890s would have been much higher.
The Princeton study showed that the decline in marital fertility started
earlier in urban areas, but village studies, and the evidence of the regres-
sions presented in this chapter, make the point that this was not a com-
pletely new phenomenon. Some rural areas had already reduced fertility
within marriage. If rural areas tended to place more emphasis on mar-
riage restrictions, it may well be that the reason for this was that the
degree of social control required to maintain a high average age at Wrst
marriage was easier to achieve in village communities.
It seems therefore that the decline in marital fertility which began in
the 1870s had its roots in attitudes and objectives which were also
found in rural society, and which had already found expression in the
Western European pattern of delayed marriage. EVectively, population
growth was linked to economic objectives: the household unit sought to
regulate population increase so as to achieve a desired standard of living
for itself and its children. Late marriage was particularly linked to the
institution of the nuclear family, and to the custom that children should
set up a separate household on marriage. Parental approval of marriage
would depend on the couple having suYcient means to do this, and to
achieve a respectable standard of living in the new household.47
This implies that there was a gradual upwards adjustment of the
objective. Even in pre-industrial Europe there was some increase in
living standards, although this was intermittent and uncertain. If the
household made no adjustment of the desired standard of living, then
in times of prosperity there would have been a surge of population
growth suYcient to wipe out any gains achieved. By and large this did
not happen.
If this argument is correct then the German rural population which
moved into industrial cities carried with it attitudes which predisposed
it to control fertility within marriage. The incentive for a married
couple to move out of a family home consisting of a one- or two-room
47 This meant that eVective population control could be exercised simply by refusingpermission to set up a new household, as happened in Mecklenburg for example, Lubinski(1995).
Demography and Migration 159
![Page 169: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/169.jpg)
apartment in a Berlin Mietskaserne (or tenement block) was at least as
great as the incentive to move in a rural village. The strategy may have
changed. Increased employment opportunities for married women with-
out children would have made it more attractive to marry early but put
oV having children. But the target was the same.48
Moreover, the ability to adjust objectives would have also aVected
behaviour. Cities oVered new opportunities: white-collar jobs, appren-
ticeships, accessible secondary schools. Even if migrants themselves
were not able to take advantage of these possibilities, their children
might, if there were not too many of them. In cities with a strong
middle-class presence these opportunities may well have been greater,
or at least more apparent.49
The demographic aspect of the Lewis Model was therefore curtailed.
Germany had to deal with quite a high rate of population increase in
the 1880s and 1890s, but this would have been higher still if cities in
central Germany had not started to reduce fertility from the 1870s
onwards. This in turn made these cities take in migrants from other
areas where the shift in demographic behaviour was less advanced.
The intention in this chapter has been to concentrate on those
aspects of demographic behaviour which aVect migration. This meant
that particular attention was given to mechanisms which might adjust
population growth to available economic opportunities within a region.
The main conclusion is that this did not happen in the rural east, and
that this was an important reason why the region was a major source of
migrants in the 1871–1913 period. The evidence of the regressions at
the Kreis level shows that demographic factors had major eVects on
migration. But this does not mean that the east was in a crisis of
overpopulation in 1870–1913 similar to that of the early nineteenth
century. There was less evidence of demographic distress. It was not
just population pressure which led migrants to leave the east.
48 Neumann (1978) provides evidence of fertility control in urban working-class house-holds.
49 The idea that marriage in Western Europe was linked to economic circumstances isfairly widely accepted. This may have been a long-established pattern of behaviour, asargued by Habbakuk (1958). But this has the implication that European households wereaccustomed to adjust demographic behaviour in general to economic circumstances, andmay therefore have been unusually predisposed to reduce fertility in response to improvedopportunities to invest in human capital, leading to the downward phase of the demo-graphic transition. The extent of the responsiveness of demographic variables to economicconditions may be a European peculiarity.
160 Demography and Migration
![Page 170: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/170.jpg)
The next two chapters will look more closely at conditions in the
rural east. The position of the rural labourer will be considered. Evi-
dence on wage movements, and on changing social relationship will be
presented. The role of changing technology will also be examined. But
it is important to remember that the east was not the only agricultural
region. Changes were taking place in agriculture in other regions. It was
the combination of the fact that the east was predominantly agricultural,
with a rate of population increase which was high relative to opportun-
ities in local agriculture or industry, that made migrants leave the east.
Demography and Migration 161
![Page 171: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/171.jpg)
Appendix 5A. A Class if ication of German Regions
A. Highly industrial regions
1. Regions with high population growth: 2. Regions with low population growth:
Arnsberg (43.0/23.3) Starkenburg (37.4/22.4) Wiesbaden (33.7/20.5) Aachen (36.1/23.2)
Zwickau (48.6/28.9) Brunswick (35.6/20.7) Koln (38.7/24.9) Bremen (35.5/22.9)
Dusseldorf (41.0/22.8) Dresden (40.3/24.7) Saxony-Coburg-Gotha (35.4/22.2)Hamburg (35.1/22.8)
Leipzig (42.3/25.1) Berlin (38.7/23.7) Rheinhessen (35.3/22.4) Lubeck (32.0/20.3)
Reub -a.l. (42.4/25.5) Potsdam (38.1/23.9) Reub -n.l. (40.3/26.2) Bautzen (36.6/24.6)
B. Moderately industrial regions
1. Regions with high population growth: 2. Regions with low population growth:
Neckarkreis (41.6/24.5) Sch.-Lippe (33.1/18.7) Magdeburg (38.8/24.7) Hildesheim (34.0/22.3)
Erfurt (39.2/22.9) Saxony-Meiningen (35.9/21.4)Mannheim (38.0/24.8) Breslau (40.6/28.4)
Anhalt (38.3/21.8) Schleswig-Holstein (33.6/20.0)Schwarzburg-Rudolfs. (35.1/21.9) Upper Alsace (34.6/25.1)
Merseburg (40.7/24.6) Saxony-Weimar (35.0/22.0) Munster (32.0/23.3)
Schwarzburg-Sond. (35.7/21.1)Hanover (35.8/21.7) Saxony-Altenburg (40.9/27.9) Lorraine (29.4/22.0)
Karlsruhe (38.0/24.5)
C. Moderately agricultural regions
1. Regions with high population growth: 2. Regions with low population growth:
Danzig (44.2/27.0) Trier (37.5/22.5) Kassel (36.0/22.4) Upper Alsace (34.7/24.0)
Stettin (40.1/24.3) Frankfurt a. Oder (38.3/24.1) Oldenburg (33.0/20.7) Mittelfranken (38.4/27.6)
Oppeln (43.7/27.3) Schwarzwaldkr. (41.7/26.7) Koblenz (35.3/22.7) Liegnitz (37.4/28.2)
![Page 172: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/172.jpg)
Pfalz (38.1/22.8) Minden (35.8/21.8) Donaukreis (41.5/27.9) Oberpfalz (40.6/30.9)
Lippe (36.0/21.1) Mecklenburg-Schwerin
(32.9/20.4)
Stralsund (36.1/24.0)
Mecklenburg-Strelitz (32.1/21.7)
Oberbayern (39.6/30.3)
D. Highly agricultural regions
1. Regions with high population growth: 2. Regions with low population growth:
Konigsberg (40.7/22.0) Aurich (32.8/18.9) Waldeck (35.2/21.5) Konstanz (34.7/24.2)
Marienwerder (46.7/26.7) Gumbinnen (42.4/27.1) Sigmaringen (37.8/26.1) Unterfranken (34.2/24.1)
Potsdam (44.2/25.1) Stade (34.7/21.0) Oberfranken (34.6/23.5) Freiburg (31.5/22.5)
Bromberg (45.8/27.1) Jagstkeis (40.3/25.8) Oberhessen (31.2/20.7) Schwaben (39.6/29.8)
Koslin (42.6/26.5) Osnabruck (34.3/20.8) Luneburg (31.0/20.8) Niederbayern (39.5/31.0)
Note: Regions classiWed according to percentage occupied in agriculture 1871–1910. Figures in brackets are average birth rate per 1,000lagged 20 years followed by the average death rate.
![Page 173: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/173.jpg)
Appendix 5B. L ife Expectancy at B irth in Major C it ies ,
in Years , Calculated from 1893–4 Mortality Tables
Men
East Prussia 36.2 Konigsberg 29.5 Rest of Province 36.6
West Prussia 38.6 Danzig 33.6 Rest of Province 38.9
Brandenburg 37.9 Berlin 35.8
Pomerania 40.5 Stettin 29.2 Rest of Province 41.0
Posen 39.5
Silesia 34.9 Breslau 30.0 Rest of Province 35.2
Prussian Saxony 39.6 Magdeburg 36.2 Rest of Province 39.8
Halle 36.0
Schleswig-Holstein 43.9 Altona 37.0 Rest of Province 44.4
Hanover 43.2 Hanover 41.0 Rest of Province 43.3
Westphalia 40.9
Hesse-Nassau 42.6 Frankfurt am Main 38.5 Rest of Province 42.8
Rhineland 39.6 Koln 33.6 Rest of Province 40.1
Dusseldorf 37.5
Elberfeld 39.7
Barmen 42.0
Crefeld 36.6
Aachen 35.0
All Prussia 39.0
Women
East Prussia 40.1 Konigsberg 35.8 Rest of Province 40.3
West Prussia 42.2 Danzig 39.5 Rest of Province 42.4
Brandenburg 41.9 Berlin 41.1
Pomerania 43.5 Stettin 35.2 Rest of Province 43.8
Posen 42.9
Silesia 38.5 Breslau 35.2 Rest of Province 38.6
Prussian Saxony 42.8 Magdeburg 39.6 Rest of Province 42.9
Halle 41.9
Schleswig-Holstein 46.4 Altona 42.1 Rest of Province 46.7
Hanover 44.8 Hanover 43.5 Rest of Province 44.8
Westphalia 42.9
Hesse-Nassau 44.7 Frankfurt am Main 44.4 Rest of Province 44.7
Rhineland 42.5 Koln 38.0 Rest of Province 42.6
Dusseldorf 43.0
Elberfeld 46.6
Barmen 45.8
Crefeld 41.4
Aachen 41.1
All Prussia 42.1
Source: Kucyzinski (1897).
164 Demography and Migration
![Page 174: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/174.jpg)
Appendix 5C. Descr iption of Regress ion Variables and
Sources
(i) Tables 5.2 and 5.5
TOTFERTILITY Numbers of children aged under 1 year per
1,000 women aged 15–50, 1895 census, PS
(1897–8) vol. 148, pp. 60–176.MARRFERTILITY As above, children per 1,000 married women
aged 15–50.
%MARRIED As above, percentage of women aged 15–50
who were married.
LAND>100ha The percentage of the total agricultural area
in holdings of over 100 hectares, from 1895
Agricultural Census, SDR n.f. 212.
%CATHOLIC The percentage of the population who were
Catholic, 1890 census, PS, vol. 121, pp. 152–
98.
%POLISH As above, the percentage of the population
who spoke Polish.
%LITHUANIAN As above, the percentage of the population
who spoke Lithuanian.
RURALWAGES Average wages (Ortsubliche Tagelohne) for
rural areas; the Wgures are averages of male
and female daily rates, 1892 and 1901,
ZdKPSB (1904), 320–8.
%PARTIBLE The percentage of land transferred by inher-
itance in 1896–1902 which was divided,
ZdKPSB (1905), 262–7.
CITYDISTANCE(b) Distance in kilometres from the Kreis mid-
point to the nearest city with at least 50,000
inhabitants in 1900 (measured from contem-
porary maps).
(ii) Tables 5.4 and 5.6
Net quit rate
(dependent variable)
The estimated number of net moves out of
agriculture, by Kreis, 1882–95 and 1895–
1907, per 1,000 of the average population
occupied in agriculture, from Quante (1933),38–58.
Demography and Migration 165
![Page 175: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/175.jpg)
NATINCREASE As above, estimated numbers of total births
less total deaths, per 1,000 of the average
population occupied in agriculture.
RURALWAGES92 As for RURALWAGES, using only 1892
Wgures.
RUSSDISTANCE Shortest distance from the Kreis mid-point
to the border with Russian Poland.
CITYDISTANCE(c) Distance to the nearest city with at least
20,000 inhabitants in 1900.
(iii) Appendix Tables 5D.4 and 5D.5
Birth rates, death rates,
rates of natural increase
(dependent variables)
All Wgures are rates per 10,000 head of
population, taken from SJDS (1911), 67–8;
(1912), 70–1; and (1913), 76–7.CATHOLIC% The percentage of the population who were
Catholic, 1910 census, SJDS (1911), 678–9.
WAGES1912 Average wages (Ortsubliche Tagelohne) for
male workers, aged 16 and over, from SJDS
(1913), 826.
INDUSTRY% The share of the total occupied population
(both sexes) occupied in industry, 1907
Occupational Census, SDR (1909), n.f.
volumes 207 and 209.
MINING/METALS% As above, the share of the industrial work-
force occupied in the mining and metal pro-
duction sector.
SDPVOTE1912 Social Democrat share of the vote from Die
Reichstagswahlen von 1912, SDR, vol. 250
(1913).
MOBILITY Data from SJDS 1911–13: total moves for
each city in 1910–12 divided by total popu-
lation.
(iv) Appendix Table 5E.1
MARRIAGEt
(dependent variable)
The number of marriages in each province
per 1,000 head of population in each period
(annual data), from Kollmann (1980).
166 Demography and Migration
![Page 176: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/176.jpg)
WARDUMMY Dummy for war years (1866, 1870, and
1871).
RYEPRt Average rye prices for each province, from
ZdKPSB (1907), 84–92.
POTATOPRt,
WHEATPRt, BEEFPRt,
PORKPRt and
BUTTERPRt
As above, for potatoes, wheat, beef, pork,
and butter.
WAGESt Wages paid to male workers in Prussian state
forests, daily rate in Marks, from Eggert
(1883), decade averages.
Appendix 5D. Econometric Results
(i) The analysis of fertility in the rural Prussian Kreise
The aim of this analysis was to examine factors which aVected fertility
decisions in rural areas, so as to consider the eVect of agricultural
conditions on migration when these factors operated indirectly—
through their eVect on the demographic balance in the rural area. Mi-
gration would tend to occur when the supply of labour exceeded the
ability of the local economy to supply employment. Information on
fertility was derived from the 1895 census, by comparing the numbers
of children to the numbers of women of child-bearing age.
Two measures of fertility were used as dependent variables: TOT-
FERTILITY is the number of children aged under 1 year per 1,000women aged 15–50; MARRFERTILITY is the number of children
aged under 1 year per 1,000 married women aged 15–50. This second
variable has the disadvantage that it includes all children, not just those
born within marriage (these are not distinguished by the published
census results). A third dependent variable is the percentage of women
aged 15–50 who were married.50
The explanatory variables include two which measure the ethnic
composition of the population, speciWcally the presence of Polish- or
50 In societies with high infant mortality these fertility variables will be biased due todeaths in the Wrst year. On average around 8% of children would have died before theywere recorded by the census. However, being rather brutal about this, from a migrationviewpoint, what matters is not the number of births but the number of children whosurvived. So this bias is not a serious problem.
Demography and Migration 167
![Page 177: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/177.jpg)
Lithuanian-speaking minorities. The inclusion of these variables is im-
portant if the eVect of Catholicism is to be accurately estimated. The
coeYcients on %LITHUANIAN are not signiWcant, which is not so
surprising as there were only six Kreise where there was a substantial
minority (or majority—Lithuanian-speakers were 58% of the popula-
tion of Kreis Heydekrug in East Prussia). The variable was left in the
equations to avoid creating a bias in the other estimates.
The behaviour of the Polish-speaking and German-speaking Cath-
olics is shown to have been very diVerent. Overall fertility was 3%
higher, compared to the Prussian average, for German Catholics, but
Table 5D.1. Regression results for the analysis of fertility in rural
Prussian Kreise, 1895, estimation by OLS, t-ratios in brackets
Dependent variable is:
a
TOTFERTILITY
b
MARRFERTILITY
c
%MARRIED
Constant 89.5 183.3 23.0
LAND>100ha þ0.214 þ0.342 �0.0007
(4.90) (4.56) (0.13)
%CATHOLIC þ0.0406 þ0.494 �0.0399
(2.01) (14.25) (15.5)
%POLISH þ0.324 þ0.234 þ0.0350
(9.44) (3.96) (8.00)
%LITHUANIAN þ0.216 þ0.508 �0.0244
(1.32) (1.81) (1.17)
RURALWAGES þ21.0 þ24.2 þ2.10
(6.92) (4.64) (5.42)
N 477 477 477
S 13.90 23.89 1.77
R2 0.319 0.442 0.385
adj R2 0.311 0.436 0.378
F-test of regression 44.13 74.84 58.99
RSS 91148 269470 1485
Note: DeWnitions and sources are given in Appendix 5C. The analysis is a crude one bythe standards of modern demographic studies: there is no attempt to measure fertility bycohort or to make allowance for diVerences in the age structure of the female populationin the diVerent Kreise. The main reason for this was that the data needed to do this werenot available from the sources used. The analysis serves the needs of this study, which isto examine the eVect of demographic factors on migration.
168 Demography and Migration
![Page 178: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/178.jpg)
27% higher for Polish Catholics. The reasons for this are of interest.
Although fertility within marriage was high for German Catholics, they
were much less likely to be married (16% less likely than the rest of the
Prussian population).Thus the practice of late marriage found in south-
ern Germany was also found in the Catholic areas of Prussia (something
which can be seen from Figure 5.7 as well).
The presence of large farms and estates is shown to raise fertility,
though it does not aVect the likelihood of marriage. The coeYcient
implies that the rural population in areas where farms were over 100
hectares had fertility rates 16% higher than the rest of the population.
But the eVect of this on the east was cancelled by the Wnding that low
wages depressed fertility.
In Table 5.D2 the eVect of adding additional variables is considered.
The addition of %PARTIBLE (the percentage of land transferred by
inheritance in 1896–1902 which was divided) and CITYDISTANCE(a)
(the distance to major cities) did not have a large eVect on the other
estimated coeYcients, although the overall eVect of Catholicism was
reduced.51 The eVect of partible inheritance was to reduce overall fertil-
ity, which may surprise some. The likelihood of marriage was increased,
however, which leads to the conclusion that, barring major changes in
the percentage of children born outside marriage, there must have been
a major diVerence in fertility within marriage in the regions where
partible inheritance was common. The implied reduction in marital
fertility is about 15%. What this suggests is that, in these regions,
families were making eVorts to reduce fertility in order to avoid further
subdivision of the family property.52
Remoteness, as measured by the distance from major cities, had the
eVect of raising fertility within marriage, and reducing the percentage
married. Rural areas closest to cities had a more ‘modern’ demographic
pattern: earlier marriage, and lower fertility within marriage. The over-
all eVect on total fertility was not signiWcant.The same data set was also used for the analysis of mortality rates.
The results were of no particular interest. Neither LAND>100ha nor
51 One reason could be that there were 55 Kreise observations which had to be droppedbecause %PARTIBLE was not available, and these were predominantly in the Catholicareas of the Rhineland.
52 According to a contemporary observer, Wydgozinski in Sering (1899–1910), i. 116,there was indeed a tendency in the partible inheritance areas to restrict fertility in order toavoid the excessive splitting of holdings: a ‘two-child’ system was practised (Zweikinder-system).
Demography and Migration 169
![Page 179: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/179.jpg)
%PARTIBLE had any signiWcant eVect on mortality. The only clear
result was a tendency for mortality to rise where wages were low.
(ii) Analysis of migration by Kreis, 1882–1895
The connection between migration and population growth can be exam-
ined by making use of an analysis conducted by the Prussian Statistical
Table 5D.2. Regression results for the analysis of fertility in rural
Prussian Kreise, 1895, estimation by OLS, t-ratios in brackets
a b c
TOTFERTILITY MARRFERTILITY %MARRIED
Constant 88.4 166.7 24.8
LAND>100ha þ0.195 þ0.346 �0.0040
(4.34) (4.52) (0.72)
%CATHOLIC þ0.0393 þ0.0435 �0.0350
(1.61) (10.45) (11.50)
%POLISH þ0.336 þ0.267 þ0.0337
(9.17) (4.33) (7.43)
%LITHUANIAN þ0.216 þ0.411 �0.0147
(1.31) (1.47) (0.72)
RURALWAGES þ23.0 þ33.0 þ1.33
(6.58) (5.58) (3.05)
%PARTIBLE �0.149 �0.330 þ0.0112
(2.94) (3.81) (1.77)
CITYDISTANCE(b) �0.00013 þ0.00601 �0.000658
(0.11) (3.14) (4.74)
N 424 424 424
S 13.9 23.7 1.72
R2 0.345 0.436 0.333
adj R2 0.334 0.426 0.321
F-test 31.4 45.9 29.6
RSS 80443 233943 1238
Note: This data set was also used to examine the eVect of out-migration on marriagerates. The inclusion of a term giving rates of out-migration between 1895 and 1905improved the Wt, and the variable had a t-statistic of 11.00, indicating a highly signiWcantresult. As the migration Wgures refer to a period after the census used for the dependentvariable, the result cannot be regarded as an entirely satisfactory one, and there mustbe a question about the direction of causation. For this reason, the result has not beenreported, although it certainly suggests that marriage rates were depressed in regions ofout-migration.
170 Demography and Migration
![Page 180: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/180.jpg)
Bureau and used in an article in 1933.53 This made indirect estimates of
the rate of migration out of agriculture by applying rates of fertility and
mortality recorded for the rural population. These estimated rates of
natural increase for the population occupied in agriculture were then
used to derive estimates of net migration, using the indirect method
(comparing census returns with the predicted natural increase).
The dependent variable is best described as the net agricultural ‘quit-
ting rate’: the rate at which persons occupied in agriculture switched to
other occupations, since it is not clear whether or not the people who
left agriculture actually migrated. It may be surmised that most did
make some move: although some may have found non-agricultural em-
ployment in their village, the majority probably had to leave for a local
town at the very least. But many may not have left the Kreis, particu-
larly if this was in an area where non-agricultural employment was
rising rapidly.
The data set makes it possible to examine the eVect of variations in
population growth on the net quitting rate, along with the inXuence of
other variables, such as agricultural structure. Bivariate correlation an-
alysis shows a positive relation between the presence of large farms and
estates (where LAND>100HA is high) and the net quitting rate: out-
ward migration was apparently higher on the larger estates. However, in
multiple regression analysis, this eVect is reversed once the rate of
natural population increase is introduced into the estimated equations.
In other words, the apparent relationship between large estates and out-
migration is entirely due to demographic factors.
The Wrst column of Table 5.D3 gives the results of Ordinary Least
Squares estimation of the basic equation. The estimated coeYcient on
LAND>100HA is negative, as is the coeYcient on LAND20–100HA,
though smaller. Large estates were the ones with the lowest quit rates;
quit rates were highest on holdings of less than 20 hectares.
The quit rate was reduced by the presence of Poles or Lithuanians.
The fact that Poles, in particular, have Wgured quite heavily in accounts
of migration, has much to do with high birth rates: these non-German
groups had no great inclination to migrate, and were rather less likely to
leave agricultural employment than Germans in similar circumstances.
It was population pressure which drove up migration from the Polish-
speaking areas.54
53 Quante (1955); the data were also used in Quante (1958).54 If NATINCREASE is omitted from the estimated equation, the coeYcient on
%POLISH becomes positive and signiWcant.
Demography and Migration 171
![Page 181: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/181.jpg)
The negative coeYcient on %AGOCCP1882, the percentage occu-
pied in agriculture, shows that the availability of opportunities to Wnd
non-agricultural employment within the Kreis increased the quit rate.
This, however, would not lead to increased net out-migration. In the
Table 5D.3. Regression results for moves out of agriculture by
Kreis, 1882–95
Estimation method
OLS OLS IV IV
a b c d
Constant 184.4 182.2 155.9 190.4
LAND>100HA �1.245 �1.230 �1.469 �1.550
(6.42) (6.30) (4.10) (3.84)
LAND20-100HA �0.585 �0.565 �0.643 �0.624
(3.76) (3.58) (3.65) (3.10)
%AGOCCP1882 �0.507 �0.540 �0.361 �0.431
(3.19) (3.27) (1.42) (2.15)
RUSSDISTANCE �0.161 �0.159 �0.173 �0.154
(10.30) (10.00) (7.95) (5.21)
%POLISH �0.826 �0.843 �1.109 �1.183
(5.33) (5.38) (2.70) (2.25)
%LITHUANIAN �1.687 �1.698 �1.749 �1.711
(2.99) (3.00) (3.01) (2.94)
NATINCREASE þ1.051 þ1.043 þ1.218 þ1.271
(21.39) (20.70) (5.32) (3.66)
CITYDISTANCE(a) þ0.110
(0.76)
%PARTIBLE �0.17
(0.02)
RURALWAGES92 �0.39
(0.84)
N 390 390 390 339
S 46.45 46.47 47.15 46.41
R2 0.672 0.673 0.474 0.483
adj R2 0.666 0.666 0.464 0.469
F-test 111.84 97.82 49.16 34.12
RSS 824158 822914 849148 708756
Note: Dependent variable is the estimated net quitting rate per 1,000 for the totalagricultural population.
172 Demography and Migration
![Page 182: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/182.jpg)
more industrial areas the quit rate was high, but many of those leaving
found non-agricultural work within the locality.
But the most important result to emerge from this analysis is the
importance of the rate of natural population increase. The eVect is a
large one. The estimated coeYcient in column a is just over 1.0, indi-
cating that, in Kreise with above-average rates of population increase,
over 100% of this surplus left agriculture. There was no capacity to
absorb surplus labour within agriculture.
In regions where non-agricultural employment was increasing, sur-
plus labour could Wnd work without migrating. In other regions where
there were limited non-agricultural opportunities the surplus had to
migrate. Demographic pressure fed directly into migration in these
regions.
There are problems of possible endogeneity bias in this equation.
High migration rates aVect the population balance and this will have an
inXuence on the rates of natural increase. In areas with high inward
migration, the population will tend to be younger, and this will reduce
the crude death rate, pushing up the rate of natural increase. One way
to deal with this is to use instrumental variables, if a suitable instrument
can be found. The distance to large cities turned out to be a possible
instrument. Its inclusion had little eVect on the estimated equation
(column b) but it was well correlated with the rate of natural increase.55
Using this as an instrument produced some changes in the estimated
values (comparing column a with column c), and reduced the overall Wt
of the equation, but pushed up the coeYcient on NATINCREASE to
over 1. The new coeYcient suggests that 120% of the natural increase
in the population left agriculture. This implies that the overpopulated
Kreise were releasing a surplus built up in previous periods of high
population growth as well as the surplus created by high current rates
of population increase.
The Wnal column introduces other variables. There is no eVect frompartible inheritance on the quit rate. High rural wage levels have a small
negative eVect on the quit rate, but the coeYcient is not signiWcant.
(iii) Analysis of demographic behaviour in urban areas
Econometric analysis of demographic variables for German cities can be
used to show the eVect of industrial structure and other factors on rates
55 A correlation coeYcient of 0.274 (p-value of 0.000).
Demography and Migration 173
![Page 183: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/183.jpg)
of natural increase. Cities with low rates of natural increase would need
higher rates of inward migration to provide the increased labour force
needed for industrial expansion.
The basic model uses three variables: the Catholic percentage of the
total population, the level of wages, and the percentage of the occupied
population who are occupied in industry. In column a of Table 5D.4
the dependent variable is the birth rate for the years 1910–12. Theresults show that the birth rate was highest in Catholic industrial cities.
The Wnding that the industrial percentage has an eVect on the birth rate
is consistent with the view that the working class in these cities was
relatively isolated and therefore less likely to emulate middle-class
reproductive behaviour, although other explanations cannot be ruled out.
In column b the eVect of industrial structure is examined by includ-
ing MINING/METALS% (the percentage of the industrial workforce
occupied in mining and metal production) amongst the explanatory
variables.56 The inclusion of this variable has an eVect on the coeY-
cients estimated for the other variables. The eVect of Catholicism is
greatly reduced, as is the eVect of industrial employment in general.
The mining and metal production sector was characterized by large-
scale enterprises and a strong ‘isolated’ working-class culture.
The eVect of Catholicism is further reduced, eVectively eliminated,
by the inclusion of SDPVOTE1912, the Social Democrat vote in the
1912 election as a percentage of all votes. This should not be inter-
preted as a direct causation; rather, both changed demographic behav-
iour, and a willingness to vote for the SPD, should be taken as
indicators of changes in working-class mentality. The more conWdent,
and independent, working class which was emerging in many German
cities was prepared to express this in votes for the more radical pro-
gramme of political and social change espoused by the SPD. Similar
changes in social attitudes led to an alteration in reproductive behav-
iour, and a reduction in family size, which led to a fall in the birth rate.
The Wnal column of Table 5D.4 gives the results of a similar analysis
for the death rate. Here, the importance of wage levels is much greater,
having had an insigniWcant eVect on birth rates. Mortality was consider-
ably lower in better-oV cities. Mortality was also lower in more indus-
56 Other variables of this type were also considered (for other industrial sectors and forthe service sectors). This was the one which the analysis showed to have had a consistentand signiWcant eVect on demographic behaviour.
174 Demography and Migration
![Page 184: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/184.jpg)
trial cities, as long as the industry involved was not mining or metal
production: these two variables have opposite signs.
Combining the birth rates and the death rates produces the overall
rate of natural increase, which is the most signiWcant variable for the
explanation of migration patterns. This is analysed in the regressions
reported in Table 5D.5. These show that the eVect of Catholicism is
not secure or consistent. High wage levels have a positive eVect on rates
of population increase, and this coeYcient is signiWcant at the 5% level
once MINING/METALS% is included in the explanatory variables
(columns b to d). The industrial structure variables (INDUSTRY%
and MINING/METALS%) have a consistent and signiWcant eVect on
population growth. The eVect of heavy industry is somewhat reduced
by the tendency of this sector to produce high death rates as well as
high birth rates, but the overall eVect is a positive one.
Table 5D.4. Regression results for the analysis of birth and death
rates in 79 major cities, 1910–12
Dependent variable is:
Birth rate Birth rate Birth rate Death rate
a b c d
Constant 180.3 232.0 241.2 259.4
CATHOLIC% þ0.641 þ0.293 þ0.038 þ0.162
(3.67) (2.03) (0.22) (1.88)
WAGES1912 �6.8 �8.6 þ0.8 �25.1
(0.57) (0.93) (0.08) (4.51)
INDUSTRY% þ1.609 þ0.738 þ0.932 �0.680
(4.14) (2.16) (2.76) (3.33)
MINING/METALS% þ2.220 þ2.057 þ0.352
(6.88) (6.48) (1.82)
SDPVOTE1912 �0.916
(2.57)
N 79 77 77 77
S 42.78 33.11 31.89 19.79
R2 0.297 0.584 0.620 0.361
adjusted R2 0.269 0.561 0.593 0.325
F-test of regression 10.56 25.29 23.13 10.16
RSS 137246 78929 72210 28194
Note: Dependent variables are annual rates per 10,000 inhabitants.
Demography and Migration 175
![Page 185: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/185.jpg)
The socialist vote continues to be associated with a reduced rate of
population increase (column c). The Wnal column includes MOBILITY
(the number of total moves in and out of each city relative to the total
resident population) amongst the explanatory variables. This has no
signiWcant eVect on demographic behaviour. This conWrms the Wnding,
reported in the chapter, that it is not possible to discover any relation-
ship between migration and demographic behaviour in urban areas,
using the sort of aggregated analysis reported in this chapter. A micro-
level study, using individual household data, might produce a diVerent
result.
Statistical analysis of results for an earlier period, the years 1899–1901, broadly conWrmed the Wndings of this section. The more limited
data availability meant that the analysis had to be conWned to 27 major
cities, with a resulting loss of degrees of freedom. But the econometric
results also showed a strong eVect on the rate of natural increase from
Table 5D.5. Regression results for the analysis of the population
growth in 79 major cities, 1910–12
a b c d
Constant �69.2 �27.4 �18.6 �11.5
CATHOLIC% þ0.432 þ0.131 �0.112 �0.176
(3.01) (1.13) (0.83) (1.05)
WAGES1912 þ16.9 þ16.4 þ25.5 þ28.9
(1.74) (2.21) (3.34) (3.42)
INDUSTRY% þ2.186 þ1.418 þ1.603 þ1.620
(6.84) (5.18) (6.04) (5.34)
MINING/METALS% þ1.868 þ1.713 þ1.726
(7.22) (6.87) (5.67)
SDPVOTE1912 �0.873 �0.971
(3.12) (2.97)
MOBILITY �0.3568
(0.9)
N 79 77 77 69
S 35.19 26.54 25.07 26.15
R2 0.460 0.697 0.733 0.700
adj R2 0.439 0.680 0.714 0.671
F-test 21.34 41.36 39.03 24.09
RSS 92860 50721 44620 42410
Note: Dependent variable is the annual rate of natural increase per 10,000 inhabitants.
176 Demography and Migration
![Page 186: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/186.jpg)
the percentage occupied in industry, and from heavy industry, and also
conWrmed that population growth was highest where wages were high.
Catholicism was found to be relatively unimportant.
Appendix 5E. Marriage and Economic C ircumstances
The argument that late marriage was a distinctive Western European
institution, which linked marriage rates to economic conditions, and
Table 5E.1. Panel data analysis of marriage rates in Prussian
provinces 1816–75
Random eVects model Fixed eVects model
a b c d
Constant 3.83 3.98 � �MARRIAGEt�1 þ0.515 þ0.526 þ0.503 þ0.52143
(15.48) (16.09) (9.44) (10.50)
WARDUMMY �1.176 �1.151 �1.209 �1.161
(8.09) (7.93) (7.07) (6.78)
RYEPRt �0.0149 �0.0155 �0.0140 �0.0151
(5.90) (6.19) (5.16) (5.93)
POTATOPRt �0.0012 �0.0011 �0.0014 �0.0011
(2.30) (2.01) (2.76) (2.29)
WHEATPRt þ0.0120 þ0.0123 þ0.0116 þ0.0121
(6.55) (6.72) (5.91) (6.30)
BEEFPRt þ0.0266 þ0.0301 þ0.0261 þ0.0306
(3.52) (4.11) (3.14) (3.73)
PORKPRt �0.0104 �0.0141 �0.0098 �0.0146
(1.69) (2.44) (1.45) (2.05)
BUTTERPRt þ0.0014 þ0.0011 þ0.0014 þ0.0011
(0.52) (0.39) (0.48) (0.38)
WAGESt �0.0001 �0.0050 þ0.0027 �0.0048
(0.02) (1.37) (0.47) (1.42)
TIME �0.0072 �0.0101
(1.67) (1.79)
N 711 711 711 711
R2 0.668 0.667 0.687 0.685
RSS 0.256 0.246 0.226 0.228
Note: t-statistics in brackets, calculated using White’s heterocedasticity-corrected standarderrors.
Demography and Migration 177
![Page 187: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/187.jpg)
thus formed a ‘positive check’ on population growth, to use Malthus’s
terminology, is associated with the work of Hajnal, although it is cer-
tainly of much older origin.57 The idea can be found in the work of
German demographers in the late nineteenth century, and indeed
delayed marriage is Malthus’s moral check on population growth.
There has, however, been criticism of the view that non-Western soci-
eties inevitably tended to push population growth to the point where
wages were reduced to the subsistence level. Some form of population
control seems to have operated in Tokugawa Japan and in Imperial
China, although the mechanism did not involve delayed marriage.58
If marriage was related to economic circumstances in Western
Europe, then it should be possible to Wnd some connection between
economic variables and marriage rates. In the period before the trans-
formation to an industrial society these variables will be primarily re-
lated to agricultural conditions. Good harvests should produce gains in
real wages for agricultural workers, as a result both of lower living costs
and of higher demand for rural labour. High urban demand for agricul-
tural produce should also raise rural living standards, and lead to higher
marriage rates.
To test this hypothesis, a data set was prepared which included
annual marriage rates for the original Prussian provinces (1816 bound-
aries), and average prices for each province from the oYcial price
surveys. The data set ran from 1816 to 1875. This made it possible to
include wage data from a survey of day-labourer rates in the Prussian
state forests for each region, which should be a good guide to rural
wage rates. The wage Wgures refer to decades. It would have been
desirable to have had annual wage data, but a source of this type is not
available for the period. The analysis runs for a period when Prussia
was still a predominantly rural society.
The data set was analysed as a panel data set; the results are given in
Table 5E.1. The Wrst two columns give the results when a random
eVects model is used; the second two when the analysis uses a Wxed
eVects model (with dummies for each province). The two sets of results
are very close, indicating that the random eVects results can be regarded
as secure.
57 Hajnal (1965).58 Infanticide was one means employed: Lee and Wang (1999), Hanley and Yamamura
(1977).
178 Demography and Migration
![Page 188: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/188.jpg)
The eVect from wage levels is weak and not secure in statistical
terms. The introduction of a time trend (columns a and c) eliminates
the eVect of this variable. Butter prices were also found to have little
eVect on marriage rates. The other price variables were associated with
signiWcant changes in the marriage rate.
There are three price series which have negative coeYcients: rye,
potatoes, and pork (although the result for pork is not a robust one).
These were the basic items of consumption for the poorer sections of
the agricultural population, and the less well-oV urban dwellers as well.
High prices for these commodities hit the living standards of the agri-
cultural workers and led to a postponement of marriage plans. The
eVect of rye prices is a particularly strong one.
There are two price series which are positively linked to marriage
rates: those for beef and wheat. These are items which did not form a
major part of the diet of the agricultural labourers and their families,
but which were mainly consumed by the better-oV in the countryside,
or sent to urban markets. High prices for these commodities were an
indicator of high urban demand, and a general rise in prosperity. As the
agricultural sector exported these products to the cities, while consum-
ing relatively little itself, high prices raised rural incomes. So, the posi-
tive link with marriage rates shows that rising rural prosperity was a
factor which brought forward marriages. It can also be argued that high
prices for these ‘luxury goods’ were indicators of rising urban wages,
which may have led to increased marriage rates in urban areas.
In general these results support the view that marriage in Prussia in
the pre-industrial phase and the early stages of industrialization was
inXuenced by economic circumstances, particularly in the rural sector.
When conditions were favourable, leading to gains by the rural sector,
then the marriage rate rose. More couples were able to aVord the cost of
setting up a household on their own. But when times were hard, then
this was put oV.This analysis leaves open the question whether regions which had
poor economic conditions had permanently low marriage rates. The
fact that no satisfactory result is obtained from the wage data argues for
the view that there was no permanent eVect; marriages were brought
forward or put oV according to circumstances, but the rate was not kept
down over the long term in regions where wages were low. The price
data might obscure this eVect if poorer regions had relatively high
prices for rye and potatoes but relatively low prices for beef and wheat,
though there is no evidence that this was in fact the case. Against this,
Demography and Migration 179
![Page 189: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/189.jpg)
the econometric results for the Kreise given in Section 5D(i) do show a
positive eVect from rural wages on marriage rates in 1895, so it may be
that the analysis reported in this section is too aggregated to Wnd this
eVect. It might be found at the Kreis level but not at the provincial
level.
180 Demography and Migration
![Page 190: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/190.jpg)
6
Migration, Farm Size, and the
Condition of the Agricultural Labourer
1. Introduction
One theme which ran very strongly through contemporary writings on
migration, particularly migration out of eastern agriculture, was that
there were institutional problems east of the Elbe which caused the
high rates of migration. These were particularly related to the preva-
lence of large estates, and the contractual labour system practised on
these estates. The property-owning peasant farmers typical of other
regions were, it was thought, less likely to leave. The landless east-
Elbian agricultural workers were drawn to the cities, leaving empty
dwellings and uncultivated Welds behind them. Or, even more seriously,
they were replaced by Polish workers, which threatened the population
balance in the disputed eastern provinces which Prussia had gained as a
result of the partitions of Poland in the eighteenth century.
Many of these studies expressed the view that there was something
wrong with agriculture east of the Elbe, and this feeling has tended to be
reproduced in modern historical studies: the east was backward, bound by
tradition, hostile to modernity, and so on.1 This chapter and the next one
question some of these stereotypes. The theme of this chapter is the insti-
tutional structure of east-Elbian agriculture, and the part this played in
migration out of the rural east. The next chapter considers the techno-
logical transformation of eastern agriculture in the late nineteenth century.
2. A Modern Perspective on the Institutions
of Eastern Agriculture
Adam Smith and the classical economists of the eighteenth and
early nineteenth centuries were hostile to share-cropping and other
contractual systems which deviated from the employer–wage labourer
1 A work published in 1898 described large eastern estates as the principal obstacles tothe establishment of a ‘just society’ and a ‘modern exchange economy’, Oppenheimer(1898), 9. See Frauendorfer (1957), i. 407–8 for a summary of Oppenheimer’s work.
![Page 191: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/191.jpg)
or landlord–tenant norms. The reason for this was the eVect that such
institutions had on the incentives faced by the share-cropper or tenant
farmer: a share-cropper would only receive a proportion (typically 50%
or less) of the return to additional eVort, or to investments in new
productive knowledge. By contrast, a tenant farmer who paid a Wxed
money rent would receive the entire reward, which should lead to the
adoption of more progressive techniques, greater eVort on the part of
the farmer, and higher output.
The classical economists also favoured the use of cash payments: the
certainty provided by wages paid in a unit of known and stable value
would be preferable to the uncertainty of payments in kind, or payment
in the form of access to land or other beneWts. It would take extreme
conditions, where the value of a currency was uncertain or changing
rapidly (such as a period of hyper-inXation), to persuade rational and
well-informed economic agents to give up the beneWts of money and
return to barter and other forms of non-monetary exchange.
From this perspective, the traditional institutions of eastern agricul-
ture were not conducive to the rational pursuit of economic gain. An
example of payments and beneWts provided to a Deputat household on
an eastern estate in the 1800s included:
free accommodation and heating;1⁄2 Morgen fruit or vegetable garden;
1 Morgen potato land (fertilized and cultivated);
housing and fodder for one cow, plus a calf every other year;
the possibility of having hens or geese, or 1–2 sheep;
permission to keep a pig;
68 ScheVel of rye and other cereals for each adult male worker and
30 ScheVel for each female worker.2
As the nineteenth century progressed, there was a gradual shift to
cash payments, but, even in the years just before the First World War,
payments in kind remained an important source of income for eastern
agricultural workers. Table 6.1 gives income sources for households in
Wve eastern regions (East Prussia, Pomerania, Brandenburg, Silesia, and
Mecklenburg-Schwerin) compared to the average for all regions.3 This
shows that eastern workers received 60% of total household income in
2 Finck von Finckenstein (1960), 198.3 Averaging all the eastern regions obscures the fact that there were considerable diVer-
ences between the eastern provinces—cash payments were much more common in Silesia,for example. See Wygodzinski (1917), 26 for a more detailed analysis of payment systems.
182 Migration, Farm Size, Condition of the Labourer
![Page 192: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/192.jpg)
the form of harvest shares and other payments in kind, considerably
more than the national average.
EVectively, the eastern worker and his family had two roles: the
household practised smallholder production on the land supplied by the
landlord; but the household was also a small contractor, who cultivated
and harvested the estate Welds, and looked after estate livestock in
return for a share in production, as well as the land allocated to the
household (and some cash payments).
The critical view of share-cropping which can be found in the work
of the classical economists was widely accepted by economists and other
social scientists until the 1950s and 1960s, and formed a part of a
generally dismissive attitude to the institutions and values of traditional
rural society. But from the 1970s onwards a series of studies have
appeared which have taken a more positive stance. Although the main
focus has been on the advantages and disadvantages of share-cropping,
the analysis has broader implications for the study of agricultural insti-
tutions. The authors responsible for this new approach have been
working mainly in the Welds of development economics and principal-
agent theory; but their insights can be applied to the historical study of
east-Elbian agriculture.
The essence of this approach lies in the concepts of ‘monitoring
costs’ and ‘missing markets’, particularly for credit and insurance. The
monitoring of a worker’s eVort and diligence is costly for employers,
and this is particularly the case where the workplace is a diVuse one
Table 6.1. Sources of income of workers in eastern agriculture,
1910–14 (in Marks), compared with the German average
East of the Elbe All Germany
Value of house plus Wrewood 127 103
Value of household production using
land supplied by landlord
138 94
Value of harvest shares 210 124
Other payments in kind 29 34
Other beneWts 18 13
Cash payments 338 501
Total household income 860 870
Cash payments as % of total 40.5 57.6
Source: Averages calculated from budgets given in Mittler (1929), 210–15.
Migration, Farm Size, Condition of the Labourer 183
![Page 193: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/193.jpg)
(like an eastern estate) rather than concentrated (as in a modern fac-
tory). But the obvious answer to this would be to make the workers into
independent tenant farmers paying a Wxed rent, who would then have
every incentive to monitor their own activities as they would gain the
full rewards from additional eVort or increased diligence.
What is the reason why this option is not taken? One possibility lies
in the economies of scale available to large holdings, which may have
access to advanced machinery and management skills which are not
available to smaller farmers. In this case the ‘missing market’ is the one
which might allow other farmers to use such factors. Another possible
reason is that credit markets are closed to smaller farmers, or that credit
is prohibitively expensive. A third reason may be that smaller farmers
wish to have some insurance against crop failures and other disasters,
and, in the absence of formal insurance markets, they may be obliged to
enter into contractual arrangements with landlords which provide some
element of support in bad years. Finally, the costs of getting goods to
market may be so high that only larger estates can run the risk of
producing marketable crops.
Recent theory predicts that, in circumstances where there are missing
markets and high monitoring costs, then crop-sharing contracts will be
eYcient solutions to these problems. They combine an incentive elem-
ent to promote self-monitoring and higher eVort, and also provide solu-
tions to the problems which hinder smaller farmers’ access to product
markets, credit, insurance, and modern technology.4
The situation in east-Elbian agriculture in the early part of the nine-
teenth century was one which combined many features of the various
‘missing markets’ hypotheses: credit was not easy to obtain, crop yields
and prices were highly variable but insurance against these risks was not
available, new technologies such as sugar beet cultivation were being
introduced which favoured larger estates, urban markets were distant
and diYcult to predict, and transport was costly.5
4 Articles which set out the modern approach to agricultural contracts include Stiglitz(1974) and Newbery (1977). Bardhan (1989) provides an overview of the various ap-proaches. Two articles which were particularly useful in the preparation of this sectionwere Newbery (1989) and LaVont and Matoussi (1995). Newbery deals with the eVect thatmarketing costs have on the optimal choice of contract, and argues that a mixture of arental contract and a wage labour contract might provide an eYcient solution. The insti-tutional arrangements found in east-Elbian agriculture provided such a combination.LaVont and Matoussi analyse the eVect that missing credit markets have on the optimalchoice of contract.
5 The analysis of the importance of marketing costs follows Dasgupta (1993), 232: ‘TheWxed rental system is particularly ineYcient . . . where information about market conditions
184 Migration, Farm Size, Condition of the Labourer
![Page 194: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/194.jpg)
The contractual system of eastern agriculture was well suited to these
circumstances. It provided a guaranteed amount of work for the farm-
worker and his family at harvest time, which protected them against
Xuctuations in labour demand. Although payments by harvest share
meant that they were exposed to these Xuctuations, the crops grown on
the household smallholding provided an alternative source of food.
Access to buildings and fodder for livestock reduced the capital cost of
keeping a cow and other animals.
The main disadvantage from the viewpoint of the farmworker was
that it only provided full employment at harvest time (when the system
often required the labourer to provide additional Scharwerker). It could
be described as a form of institutionalized underemployment: in oV-peak
periods the family occupied itself as best as possible on the household
smallholding or in other tasks. This was equally one of the principal
advantages from the landlord’s point of view: there was no requirement
to provide tasks to keep the estate workers busy during the winter.
The second advantage for the land-owner was that it economized on
cash outgoings. Although eastern estate owners had advantages over
small farmers in that they had better access to credit, and that they
were also better placed to face the costs and the risks of bringing crops
to market, they were also credit-constrained even if at a rather higher
level, and marketing costs were still high. There could be years when
market prices would not cover the costs of bringing in the crops from
remote estates. The normal practice in such years was to feed grain not
required for human consumption to livestock. This would not produce
any income until the animals were sold. EVectively, by making pay-
ments in kind, part of the risk associated with price Xuctuations was
transferred to the estate workers, the value of whose ‘wages’ would
Xuctuate with price changes. But as most of their income was spent on
food anyway (or would have been if they had been paid money wages),
this was not such a serious disadvantage.
Industrialization promotes institutional change in rural society partly
because it provides solutions to many of these ‘missing market’
dilemmas. The spread of banks and other credit institutions eventually
makes it possible for small farmers to borrow, and also to save (as a form
of self-insurance). Improved employment opportunities can also reduce
the need for insurance: if work on construction sites, for example, can be
is of importance in generating farm proWts, and where the tenant does not possess thisinformation to any great extent.’
Migration, Farm Size, Condition of the Labourer 185
![Page 195: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/195.jpg)
found in times when agricultural incomes are low then it would be less
necessary to accept disadvantageous labour contracts which provide an
element of insurance. As urban centres grow, and transport links are
improved, market opportunities improve, integrating markets which
were previously separated, and thus reducing price variability.
Institutional arrangements such as the contractual relationships of
eastern agriculture form a part of a wider system of values and arrange-
ments, behavioural norms and beliefs, which characterize pre-industrial
rural societies. Revenue-sharing norms can support high levels of under-
employment, causing families to share resources with other members of
an extended kinship network, and with other villagers. In traditional
societies revenue-sharing and other forms of reciprocal behaviour can be
substitutes for missing markets in credit and insurance. Industrialization
will tend to alter these attitudes, mainly because wider employment
opportunities will reduce the need for insurance mechanisms. The move
to increased mobility will damage support mechanisms which rely on a
close-knit network of mutual obligations between diVerent kinship
groups. Partha Dasgupta has written about this:
Costs of mobility involve those associated with transportation, but they involve
many other things besides. Markets being thin, the outside world presents
considerable uncertainties to the individual. Life in the village may be harsh,
but there is an element of security provided by the family and the community.
This is sustained by norms of behaviour by which people abide. You can’t
simply pack up and leave for better prospects in a neighbouring village. For one
thing, you may not be welcome; for another, your departure would be seen as
an unsocial act, and this could inXict costs on your kin through a reduction of
the support system your village provides for them.6
The implications of this are, Wrst, that the cost of migration includes
not only the cost to the individual of giving up the support available in
the home village, but also the consequences for those family members
that remain of this ‘unsocial’ act. But, secondly, the system of reci-
procity itself will be at risk once there is signiWcant out-migration. If
families with members who migrate are perceived as having withdrawn
from a system of mutual obligation, then the system will cease to be one
which involves the whole community.7 While it might survive a few
6 Dasgupta (1993), 234.7 Dasgupta does not draw this conclusion, but it is a logical consequence of the system
he describes. The system can, however, survive considerable seasonal or short-termmigration.
186 Migration, Farm Size, Condition of the Labourer
![Page 196: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/196.jpg)
defections, there will be a critical level below which it will no longer
function. Once this happens, the cost of migration will fall dramatically,
since migrants will no longer have to set the loss of this insurance
beneWt against the gains from higher urban wages.
In this analysis, as in Max Weber’s, one consequence of industrializa-
tion is the erosion of traditional relationships in rural society. Initially,
this may be a slow process, but the system could collapse quite quickly
once migration is established. These traditional relationships confer a
beneWt on those living in rural areas which is additional to the income
received from farming (or other activities). The consequence of the
removal of this beneWt will be the release of surplus labour onto urban
labour markets.
Dasgupta also provides an analysis of the eVect of eYciency wages on
rural unemployment. The model is a simple one: workers are either
permanently employed or employed only at peak periods or perman-
ently unemployed. Those without work are obliged to feed themselves
using common-property resources. Labour eVort is related to nutri-
tional status, therefore land-owners pay eYciency wage rates which
exceed the market-clearing rate, so there is involuntary unemployment.
If there is a reduction in the rural population through out-migration,
then the Wrst consequence is an improvement in the average amount of
nutrition available from the common Welds and other common
resources. This in turn provides an improved nutritional carryover for
those who go on to work during the peak period. In consequence, land-
owners are able to cut peak period eYciency wages, resulting in
increased employment and a general rise in production. As Dasgupta
puts it: ‘agricultural output would be expected to increase were popula-
tion to decline in size. This is surplus labour with a vengeance.’8
More generally, the existence of eYciency wage rates either in indus-
try or agriculture will tend to create unemployment if wages are thereby
raised above market-clearing levels. In modern economies most of this
will be open and recorded, due to the availability of beneWts. In the
nineteenth century, when the availability of beneWts was restricted or
non-existent, underemployment was hidden; people scraped by on the
occasional casual job, joined the informal urban sector, went into
domestic service, or returned to family farms. The size of this pool of
underemployed labour is not easy to estimate, and much will depend on
8 Dasgupta (1993), 509 and 518–23.
Migration, Farm Size, Condition of the Labourer 187
![Page 197: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/197.jpg)
the deWnition of underemployment. Nevertheless, the reabsorption of
discouraged or temporarily unemployed urban workers was one way in
which agriculture helped to stabilize urban labour markets. In this case,
the source of the surplus may lie as much in the operation of the urban
labour market as in the state of agriculture.
3. Contemporary works on M igration
out of Eastern Agriculture
There are a number of distinctive features of Germany’s position which
aVected contemporary attitudes to migration.9 These included:
(a) Germany’s geographical position: Germany was not an island, but
linked to the labour markets of central and eastern Europe. Migra-
tion could Xow across Germany’s eastern borders relatively easily.
(b) Political considerations: the size of the rural vote was an import-
ant factor in the political structure of imperial Germany, and
those who sought to maintain it had positions of inXuence or
authority within the system.
(c) The strength of German socialism: from an early stage Germany
had a well-organized social democratic party, which polled over
50% of the vote in three German cities in the Reichstag elections
of 1877, and nearly 40% in Berlin in that year.
(d) The British example: Germans had the opportunity to study
conditions in British industrial cities. Many found this an
alarming prospect for Germany: something to be avoided, or
improved upon, if possible.
(e) Ethnic composition: parts of the German east contained substantial
numbers of Poles and Lithuanians, and the departure of German-
speaking migrants from such areas had implications for the govern-
ment’s policy of strengthening the German identity of these regions.
Attitudes to migration were certainly inXuenced by the social prob-
lems associated with urbanization. Adolf Weber, writing in 1908, wasable to use oYcial statistics to demonstrate that crime was higher in
cities, as were divorce, suicide, adultery, the incidence of venereal dis-
ease, and the number of children on welfare, while church attendance
was lower. He added a number of other charges: city art was ostenta-
9 Haberle (1938) surveys German studies before 1938.
188 Migration, Farm Size, Condition of the Labourer
![Page 198: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/198.jpg)
tious, there was less patriotism, and city life hindered the moral and
spiritual progress of the nation. Despite this, Weber did not condemn
all aspects of city life: he thought that productivity was higher and
educational opportunities were better.10 Most serious commentators
recognized that urbanization had its beneWts, and that it was an inevit-
able consequence of economic growth.
So, although the debate over the question ‘Agrar- oder Industriestaat’attracted attention in the 1890s, in practice no serious eVort was made
to hold back urbanization or the growth of industry. The most that it
was hoped to achieve was that Germany should remain both ‘Agrar-
und Industriestaat’, to use the title of a work by one participant in the
debate, the historical economist Adolf Wagner.11 It was also considered
to be important to preserve the Mittelstand, and peasant proprietors
constituted a large part of this class. This led to a concern with the
interests of small and medium-sized farmers: as Gustav Schmoller put
it, ‘the issue of the preservation of the position of the peasantry involves
our entire future as a society’.12
This piece of rhetoric requires some elucidation. Schmoller feared
that industrialization would lead to the concentration of property, and
the polarization of society between a propertyless working class and a
rich capitalist minority, just as Marx had predicted. Membership of the
Mittelstand was deWned not just by income level but also by the owner-
ship of at least a small amount of property. Gustav Knapp, speaking
after Schmoller in the same debate, praised labour relations in agricul-
ture in western Germany: the Heuerlingen of this region had some
animals of their own, long-term contracts with their employers (them-
selves not aristocratic land-owners but larger peasant farmers), and
worked ‘shoulder to shoulder with the employer’.13 The consequence
was that social democracy had made no progress in this area.14
Migration transferred agricultural workers into a very diVerent soci-
ety. In large-scale industry there was little personal contact between
employers and employees; employment was insecure and contracts were
10 Weber (1908). Adolf Weber (1876–1963), economist, posts included chairs at theuniversities of Cologne, Breslau, Frankfurt, and Munich.
11 Wagner (1902); the debate is described in Lebovics (1967) and Barkin (1970). AdolfWagner (1835–1917), economist, professor at Berlin university, leading member of thehistorical school, and one of the founders of the Verein fur Sozialpolitik. He representedthe Christian Social Party in the Prussian State Assembly 1882–5.
12 Verein fur Sozialpolitik (1893), 2.13 Ibid. 12.14 Ibid. 29.
Migration, Farm Size, Condition of the Labourer 189
![Page 199: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/199.jpg)
short-term; workers had little property of their own. This was fertile
ground for socialism.
The concern of the historical economists with the polarization of
society and the decline of the Mittelstand was intensiWed by a series of
studies of Prussian tax returns which appeared to show that inequality
was widening. The Wrst of these, by Engel, appeared in 1868. Some of
the subsequent studies were produced by members of the Verein furSozialpolitik; Schmoller and Wagner were amongst those who worked
on these estimates.15
These studies led to a debate about the reasons why inequality was
widening. Adolf Wagner thought that the cause was technological pro-
gress favourable to large enterprises. But the essence of the historical
school’s analysis of this, and other, economic problems was to stress the
importance of the historical context: society did not move along prede-
termined lines. Decisions were made which aVected the outcome. They
held the view that the state could bring about the reconciliation of
capital and labour, both by direct action and by encouraging favourable
institutional developments.16
Some made the same connection between rural conditions, migration,
and urban labour markets which lies behind the Lewis Model. Max
Sering, in a report presented to the Verein fur Sozialpolitik in 1893,
blamed excessive migration for depopulating rural areas in the east, as
well as holding down the wages of industrial workers:
while the cities and industrial districts are accumulating an ever larger reserve
army, which rarely Wnds full employment, even in times of maximum economic
activity, and which is forcing down the living standards of other workers, the
wide Welds of the eastern arable districts are short of labour . . . 17
Interesting points in this passage include the use of the term ‘reserve
army’, familiar from Marxist terminology.
15 Engel (1868 and 1875); Schmoller (1895) and Wagner (1904); there had been atradition of concern about the distribution of income in German political economy, goingback at least as far as von Thunen (1826), whose views are given in Pogge von Strand-mann (1999). Ernst Engel (1821–96), statistician, director of the Prussian Statistical OYce1860–82, represented the National Liberal Party in the Prussian State Assembly, 1867–70,Neue Deutsche Biographie, iv. 500–1.
16 Wagner (1871) and Schmoller (1900); also HelVerich (1915); the views of the histor-ical economists on income distribution are discussed in Dumke (1991) and Grimmer(1999); other writers considered ways to reconcile the interests of capital and labour,amongst them Walter Rathenau, Pogge von Strandmann (1988a). Karl HelVerich was adirector of the Deutsche Bank and secretary of the Treasury during the First World War.
17 Sering (1893), 7.
190 Migration, Farm Size, Condition of the Labourer
![Page 200: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/200.jpg)
For a group of writers on agricultural matters, the extent of migra-
tion out of agriculture was far greater that could be justiWed by the
diVerence between urban and rural wages. This group, which included
Knapp, Max Weber, and Sering, blamed developments within agricul-
ture.18 Sering noted that in 1885–90 the eastern provinces had lost
640,000 people through migration, 75% of the natural increase in the
population. This was an indication of a profound social malaise: ‘it
suggests that all social classes had a feeling of profound dissatisfaction,
that there was a sickness of the social organism itself’.19 The view that
migration was a symptom of some major problem within rural society
was an inXuential one, and led to schemes of ‘internal colonization’: the
settlement of landless rural workers, or disaVected urban workers, in
colonies of land-owning peasant farmers.20
Migration from rural areas into German cities was a matter of con-
cern to a range of opinion in Germany. A number of authors took the
view that it was caused by a deterioration in agricultural conditions, in
particular by changes in the position of landless agricultural workers.
One strand of this analysis concerned the eVect of the agrarian
reforms of the early nineteenth century. In 1873 Theodor von der
Goltz pointed out that migration was highest from areas of Großgrund-
besitz—the large east-Elbian estates—and unknown in areas dominated
by smaller peasant holdings. In 1887 Gustav Knapp argued that the
reforms had led to a loss of traditional rights and increased the number
of landless labourers:21 In Knapp’s view it was the poverty and lack of
freedom of the eastern agricultural worker which was responsible for
out-migration.
In the western provinces, where there is often serious poverty amongst small
farmers as a result of the splitting of holdings, there are certainly poor agricul-
tural workers, but nobody who feels to the same extent the lack of freedom in
every aspect of their lives. . . . The poorest forestry worker in the Black Forest,
the most down-trodden hay-maker in Kanton Uri, have a higher station in
society than the Insten in our wide eastern provinces.
18 Knapp (1887), Weber (1893 and 1894): also von der Goltz (1875). G. F. Knapp(1842–1926), economist and statistician, professor of statistics at Leipzig and Strasburg,NDB xii. 152. T. A. von der Goltz (1836–1905), farmer and agricultural historian, heldchairs of agriculture at Konigsberg and Jena, NDB vi. 635–6. See Frauendorfer (1957),i. 392–4, 394–6 and 436–7 for the background to this debate.
19 Sering (1893), 7.20 These schemes also served nationalist aims, Grimmer (1999), 293–4.21 von der Goltz (1875), Knapp (1887 and 1891).
Migration, Farm Size, Condition of the Labourer 191
![Page 201: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/201.jpg)
Knapp blamed the operation of the Stein-Hardenburg laws for the
original loss of land and status by the agricultural workers east of
the Elbe. But he also thought that their position was deteriorating as
the land-owners intensiWed their demands for labour to work the estate
Welds, thus reducing the time the labourers had for their own small-
holdings:
The aristocratic landlords want to have all the labourers’ time for their own
enterprises . . .Where this development has been completed . . . the worker on an
estate will be turned into a state of utter dependency . . . he is no longer a true
Inst: he has no house, he has no arable acres of his own, he no longer has a cow.
These early studies had an eVect on opinion. The view that large estates
were predominantly to blame for migration out of agriculture became
widely accepted. A study published in 1930 commented: ‘It is a com-
monly held view that the large landholdings considerably increase mi-
gration out of agriculture.’22
The debate produced three major surveys: one of the conditions of
the peasantry in 1883, another of the conditions of agricultural labour-
ers in 1890–1, and a survey of Prussian inheritance systems in 1899–
1910. These are valuable surveys of conditions in German agriculture in
the period.23 There are, however, some diYculties in using the infor-
mation they provide. They are, in eVect, collections of articles, which
reXect the concerns and interests of individual authors. Data is not
generally provided on a comparable basis.24
The 1890–1 survey was in three parts. The third of these, covering
Germany east of the Elbe, was written by Max Weber. It was his
second major publication. His contribution is useful for a number of
reasons. It was more systematic than the others. It includes useful data
on wage rates and payments in kind, and attempts to link these Wgures
to those produced by earlier surveys. And Weber attempted to provide
an explanation of the process of social change revealed by the survey.
22 Golding (1930), 214. See Oppenheimer (1898) for a particularly forceful presentationof this view.
23 These were published in the Schriften des Verein fur Sozialpolitik (1883), vols. 22–4and (1892–3), vols. 53–5, and in Sering (1899–1910).
24 A comment which reXects the frustrations of the modern statistician, faced withwhat could have been immensely valuable sources of data. There are some very goodarticles in these surveys: one example is the article by Paasche on Mecklenburg-Schwerin,Paasche (1883).
192 Migration, Farm Size, Condition of the Labourer
![Page 202: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/202.jpg)
4. Max weber and the Condition of the
Agricultural Labourer in the East
Of all Weber’s major works this is probably the least known. It remains
untranslated. Yet it is a thorough survey of conditions in eastern agri-
culture, a valuable source of material, and one of the earliest works to
consider the question of a ‘labour surplus’ in traditional agriculture:
how is it created? under what conditions is it released?
The questionnaire results were used as a basis. Weber reported some
of the more reactionary views of the respondents without comment, as
for example the opinion that it was not desirable to educate the rural
population.25 But he also used the survey to express his own views on
the social relationships found in eastern agriculture. He developed this
analysis in an article in the 1894 Archiv fur Soziale Gesetzgebung und
Statistik, which also responded to some of the criticism made of the
original study at the 1892 Colloquium of the Verein fur Sozialpolitik. In
1908 he returned to the theme in a lecture in the United States.26
Weber followed Knapp in that he thought that the larger estates were
primarily responsible for out-migration. He saw the creation of the
labour surplus as a result of an alteration in social relationships on these
estates. Paternalist attitudes were replaced by a commercial, capitalist
approach. This in turn led to the introduction of more intensive arable
production, associated with sugar beet cultivation, and increased use of
mechanization. Contractual German workers were replaced by seasonal
Polish workers.
Weber’s attitude was inXuenced by political concerns, a desire to
maintain and, if possible, increase the German population in the dis-
puted lands of the Prussian east. He was a member of the nationalist
organization, the Ostmarkverein (also known as Hakatisten from the
initials of the organization’s founders), which argued for the compul-
sory purchase of Polish estates and the settlement of German farmers
on lands previously occupied by Poles.27 Weber supported the organiza-
tion’s aims. A newspaper report of a speech he gave in March 1893
went: ‘above all else, Herr Weber argued for the pursuit of a policy
25 Weber (1893), 197–8.26 Weber (1894 and 1948). See Honigsheim (1949) for an assessment of Weber’s work
in this area.27 See Hagen (1980) and also Galos et al. (1966) for accounts of the conXict between
Germans and Poles in the east. Schmoller and Adolf Weber were also members of theOstmarkverein, Hagen (1980), 269.
Migration, Farm Size, Condition of the Labourer 193
![Page 203: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/203.jpg)
which served German interests, he was against the equal treatment of
Poles and Germans’.28 His criticism of the Junker estates should be
seen in this context. In the same month, in another speech, he said: ‘the
presence of large estates is the factor which is, currently, most respon-
sible for the increase in the Polish population in the East’.29
However, in the 1890s the concern over foreign agricultural workers
was premature. In 1907 there were 195,000 foreign-born agricultural
workers in Prussia, constituting 1.8% of the agricultural labour force:
1.0% in East Prussia, 1.6% in West Prussia, 1.3% in Posen, and 2.7%
in Pomerania. The largest concentration was in Prussian Saxony, where
they amounted to 4.8% of the population occupied in agriculture.30
This suggests that, up to this point, the main cause of migration from
outside Germany had been the high seasonal demand for labour in the
sugar beet Welds of central Germany.
It has been estimated that the number of foreign workers rose to over
a million by 1913 (including 356,000 employed in industry). This is
diYcult to check given the lack of an occupational census after 1907.
The Preubische Feldarbeiterzentrale issued a total of 412,000 permits in
1907–14, but, as Russian-born Poles, in particular, were obliged to
return home every year, it is possible that these were only used for a
few years on average.31
In general, permanently settled foreigners do not seem to have
replaced Germans to any considerable extent before 1907. There was
some displacement by temporary seasonal migrants, although this could
be interpreted as the replacement of German labourers who left volun-
tarily to seek better wages in the cities. This question is related to the
claim that the position of the agricultural labourers in the east was
deteriorating, socially or economically. In this respect there is a diVer-
ence between Knapp on the one hand, who saw in the persistence of
payments in kind and contractual labour services evidence of the
dependent status of the eastern workers, and explained migration as an
attempt to escape this, and Weber on the other, who took a rather more
favourable view of the traditional labour relationships.
Weber’s view was that the Inste and Deputate of eastern agriculture
had had the status of Kleinwirte or small contractors, bound individually
to the landlords in a system which was certainly one of subordination,
but which also emphasized common interests (through harvest shares
28 Weber (1993), 825. 29 Ibid. 179 30 Ibid. 489.31 Figures are from Wehler (1995), 546.
194 Migration, Farm Size, Condition of the Labourer
![Page 204: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/204.jpg)
and access to communal grazing). The keeping of livestock by the
labourers was a crucial sign not only of material well-being, but also of
status and of ‘cultural-level’. The point here was that, in Weber’s view,
a reduced level of cow-keeping (such as Weber found in Silesia) was
generally associated with a higher level of female participation, in work
both oV and on the holding, and this was a backward step culturally:
‘the ability to keep, preferably, a milking cow, or at least some cattle, is
the most important precondition for a well-adjusted family life’.32
Weber held that this system of patriarchal labour relations had come
under threat from two directions. The Wrst was the shift to a ‘capitalist’
agriculture: more intensive use of the land and greater livestock produc-
tion. This had led to the reduction of the smallholdings given to
workers, and the division of the common grazing Welds: ‘the loss of
access to grazing on the village commons at the time of enclosure was
the most grievous blow that could have been struck at the property
owning labourer’.33 He described the reduction of the Masurian labour-
ers to the status of Milchdeputate (looking after the landlord’s cows
instead of their own) as ‘something which embittered social relations
and which destroyed the sense of having interests in common’.34
The second problem was the unwillingness of the children of agricul-
tural labourers to remain part of the system. The Instleute households
were required to provide two or three workers at harvest time, so the
participation of children was vital, especially as it became more diYcult
to Wnd substitutes (‘cripples, old men, women of loose morals, have
been oVered as such’).35 Weber did not believe that there was a connec-
tion between the migration of the children of agricultural labourers to
the cities and the level of wages. Indeed, he thought that migrants
experienced a decline in wages. In his view it was a somewhat perverse
desire for an illusion of freedom which led the children of eastern
agricultural labourers to prefer proletarianization and lower wages in
cities to the paternal system and higher ‘cultural level’ on oVer in
eastern agriculture: ‘according to the survey returns, the motive was a
desire for independence even if it could only be achieved at the cost of a
reduction in living standards’.36
It should be borne in mind that Weber was making these remarks
while reporting on the results of a survey of the condition of agricul-
tural labourers, conducted by means of a questionnaire sent out to
32 Weber (1893), 184. 33 Ibid. 34 Ibid.35 Ibid. 185. 36 Ibid. 193.
Migration, Farm Size, Condition of the Labourer 195
![Page 205: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/205.jpg)
landlords. Weber had no other evidence to support the statement that
migrants out of agriculture did not improve their economic position
apart from the opinions expressed by these respondents.
The eVect of out-migration was damaging to the institutional struc-
ture of eastern agriculture as a result of its impact on the Scharwerker
system. If the Inste were unable to provide additional workers, then
they were not fulWlling their side of the bargain which led to the land-
owner oVering a smallholding in return:
The farmer’s own children move away, at the latest, after completing military
service, though many leave earlier, so that they can devote themselves to life in
the cities as soon as possible, without considering whether this will bring about
the collapse of the family business. As a result, the Inste are no longer able to
fulWl a contractual obligation to provide three workers to the estate, as used to
be commonly required in East Prussia, many are unable to even provide two, if
their wives don’t work with them.37
The allegation that, as a result of the shift to capitalist values, labour-
ers had less opportunity to keep animals themselves is not supported by
the evidence, which shows continued cow-keeping on smallholdings,
although it was probably true in certain districts.38 However, the fact
that cow-keeping was lower in industrial districts like Silesia is not
necessarily an indicator that the position of the labourers was worse in
these areas. It could just as well show that better opportunities to supple-
ment agricultural earnings through industrial work made the keeping of
cows less attractive. Whether or not the increase in the keeping of live-
stock on smallholdings was suYcient to compensate for the undoubted
reduction in grazing rights on uncultivated lands, as these were enclosed
and brought into arable production, is a more open question.
The argument that changed values had brought about a deterioration
in the living standards of the agricultural workers is also not supported
by the evidence. On the contrary, there is good evidence (some pro-
vided by Weber himself) to show that agricultural wages were rising.
Weber supplied a mass of evidence on the conditions of employment of
37 Weber (1893), 185. Another factor which might have aVected the Scharwerkersystem was the enforcement of laws against child labour during school hours. Weber didnot mention this as a factor, but, according to other authors involved in the survey, thishad a signiWcant eVect on the use of child labour in some regions, Schriften des Vereins furSozialpolitik, 52, p. 332.
38 Between 1881 and 1895, the value of livestock kept by holdings of less than 2hectares rose by 16.2% and it rose by 17.4% on holdings of between 2 and 5 hectares; therise on holdings of over 100 hectares was only 9.5%, Rauchberg (1900), 590.
196 Migration, Farm Size, Condition of the Labourer
![Page 206: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/206.jpg)
diVerent classes of agricultural labourers, and comparisons with the
position in 1849 and 1873 (the latter date uses von der Goltz’s Wgures).
Where a high proportion of the payments are in kind or through har-
vest shares or the allocation of smallholdings the comparisons are hard
to make. There seems to have been an increase in cash payments to
groups traditionally paid in kind. But the value of these other beneWts
may have altered, and there is some evidence that the labour contribu-
tions demanded by landlords had risen.
However, comparisons are possible for those workers paid in cash
ohne Kost (without provision of board and lodging). There are a number
of localities where Wgures are available for 1849 and 1892 and these have
been brought together to form Table 6.2. What this table shows is a
very considerable improvement in the wages paid to agricultural
workers in the east over the period. The last line gives indices which
show wages rising by a factor of 2¼–21⁄2 1849–92.This is important for a number of reasons. First, the suggestion that
eastern agricultural workers experienced declining wage rates is encoun-
tered in some publications.39 This does not appear to be the case.
Secondly, it puts the problems of eastern grain producers in a new
perspective. Between 1870/4 and 1890/4 arable producer prices fell by
10.3%, while this table shows wages rising by 34–35%. Producers were
being squeezed from both directions.
5. M igrat ion , Farm S i ze and Agricultural Techniques
The regressions in the previous chapter showed that the apparent rela-
tionship between farm size and migration was reversed once allowance
was made for demographic factors. Large farms contributed to out-
migration indirectly, through the positive relationship between farm
size and the natural rate of population increase. There was no tendency
for the quit rate to be higher on larger holdings once the eVect of high
population increase was removed. Indeed, the quit rate was shown to be
lower on larger holdings.
The data set used in this analysis made use of a 1933 study of net
migration in the Prussian Kreise from 1882 to 1895. In this, an indirect
estimate of the quit rate for the agricultural population was produced
39 Dickler (1975), for example.
Migration, Farm Size, Condition of the Labourer 197
![Page 207: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/207.jpg)
Table 6.2. Daily earnings in Marks of ‘free’ agricultural day-labourers (ohne
Kost) in full-time employment where possible (earnings of those who are only
temporarily employed are in bold type)
Summer wages Winter wages
1849 1873 1892 1849 1873 1892
East Prussia: Gumbinnen
Ragnit 0.5/0.6 0.85 1.0/1.8 3/0.5 0.62 0.8/1.2
Gumbinnen 0.5/0.6 0.85 1.2/1.4 0.3/0.4 0.62 0.8/1.0
Insterburg 0.6/0.7 1.55 1.5/2.0 0.4 0.66 1.0/1.2
Sensburg 0.5/0.6 1.1 1.0/1.6 0.4 0.57 0.8/1.2
East Prussia: Konigsberg
Mohrungen 0.5/0.6 1.23 1.4/2.0 0.4/0.5 0.73 0.9/1.2
Rossel; Allenstein;
Ortelsburg
0.5/0.6 1.25/1.3 1.0/1.5 – 0.7/0.81 0.7/1.0
West Prussia: Marienwerder
Stuhm 0.8/1.0 2.03 1.25/2.5 – 1.41 –
Rosenberg 0.8 – 1.25/1.75 0.4 – 1.0/1.2
West Prussia: Danzig
Stargard 1.0/1.5 1.62 2.75/4.0 – 0.92 1.0/1.5
Putzig 0.75 – 1.0/1.75 0.5 – 0.75/1.0
Pommern: Koslin
Schlawe;
Rummelsberg; Stolp
0.8 0.6 – – 1.5/3.0 0.8/1.25
Pommern: Stettin
Anklam 1.0 1.68 1.5/2.0 0.62 1.0 1.25/1.75
Greisenberg 0.75 1.53 1.5/2.0 0.6 1.31 1.2
Pommern: Stralsund
Greiswald 1.0/1.2 2.47 3.0 – 1.18 –
Silesia: Oppeln
Neustadt 0.5 1.38 2.0 0.4 0.9 1.5
Silesia: Liegnitz
Liegnitz 0.5 0.96 1.5 0.4 0.73 1.0
Brandenburg: Frankfurt
Lebus 0.6 1.5 1.5/2.0 – 0.87 1.0/1.25
Brandenburg: Potsdam
Barnim 1.0 1.56 2.1 0.75 1.06 1.2
Unweighted
averagesa0.74 1.37 1.83 0.52 0.87 1.17
as an index
(1849¼100)
100 185 247 100 167 225
aGaps in the data have been Wlled by extrapolation (such as using the average relationship betweensummer and winter wages for the year in question); ranges were entered as the mid-point values.
198 Migration, Farm Size, Condition of the Labourer
![Page 208: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/208.jpg)
using a combination of results from the occupational censuses of these
years, and records of births and deaths by Kreis.
An alternative approach is to examine the movement of the agricul-
tural population directly using the results of the occupational censuses.
This eVectively excludes the inXuence of demographic factors by
looking only at the trend of agricultural employment. Implicitly, it
assumes that any demographic surplus will be exported, either to non-
agricultural sectors within the region, or to other regions. The analysis
presented in the previous chapter showed this to be a relatively well-
founded assumption.
The data set prepared for this analysis covered a rather longer period,
1882 to 1907, and included 797 Kreis-level units, covering the whole of
Germany not just Prussia. The results reinforced the analysis in the
previous chapter. Although there is a simple negative relationship be-
tween the movement of the agricultural population and the presence of
large holdings, this is a spurious result. More sophisticated analysis
showed that the reason why the permanent agricultural population de-
clined in areas where there were larger farms and estates (or rose less
rapidly) is that these estates were situated in areas where the agricul-
tural population was declining for diVerent reasons: these were remote
areas, with poor access to urban markets, and they were situated in
regions close to the border with Russian Poland, which meant that
seasonal Polish workers could be used.40
The results of the statistical analysis are summarized in Table 6.3.
The most important inXuences on the movement of the agricultural
labour force were the level of non-agricultural employment within the
Kreis in 1882—there tended to be a decline in the more agricultural
regions—and geographical location. In purely agricultural areas the pre-
dicted level of agricultural employment was static, meaning that any
demographic surplus would be exported, while in the areas which had
high levels of non-agricultural employment, agricultural employment
rose by 35–40% (37% for a region which was 80% non-agricultural in
1882). In these regions, quite substantial rates of population increase
could be accommodated within the agricultural sector.
Remoter regions tended to maintain a higher level of agricultural
employment. The predicted level of employment in regions 100 kilo-
metres from major cities (over 100,000 inhabitants) is 7% higher in
1907 compared to a region with a similar starting point in 1882, but
40 The full details of this analysis were originally given in chapter 4 of Grant (2000)and have been published separately in Grant (2002c).
Migration, Farm Size, Condition of the Labourer 199
![Page 209: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/209.jpg)
which was 20 kilometres from a major city. This suggests that regions
closer to major cities were quicker to shed surplus labour.
There was also a major east–west division: the agricultural population
declined fastest, or rose less rapidly, in the rural east. Table 6.4 illus-
trates this for two hypothetical regions: Region A with 60% of land in
holdings of over 100 hectares, Region B with only 10%. Both have 65%
of the population initially occupied in agriculture, and 75 kilometres
between their mid-point and the nearest large city.
This table shows the importance of location. In areas close to the
border there is little predicted increase in the agricultural population.
The average German rate of natural population increase of 1.3%
(1871–1913) implies an increase of 38% over 25 years. The rates in
Table 6.3. Summary of results of statistical analysis of the change
in the agricultural population 1882–1907 in 793 German Kreise
Secure statistical results
Factors which are associated with a less rapid rise in the population occupied in
agriculture or with a decrease in the agricultural population:
A high rate of agricultural employment in 1882
A situation close to the border with Russian Poland
Factors which are associated with a less rapid decrease in the population occu-
pied in agriculture or with an increase in the agricultural population:
The presence of large farms and estates
Being a long distance from large cities (over 100,000 inhabitants)
Having a high level of sugar beet production in 1907
Having high use of mechanized threshing in 1907
Table 6.4. Predicted change in the agricultural population over
25 years for two hypothetical regions
Distance from the border with Russian Poland in km
50 100 200 500
Predicted change in the agricultural population over 25 years
A 60% in
large farms
þ3.5 þ6.5 þ12.4 þ30.3
B 10% in
large farms
þ3.3 þ4.8 þ7.7 þ16.6
200 Migration, Farm Size, Condition of the Labourer
![Page 210: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/210.jpg)
some eastern provinces were substantially higher (1.6% per annum in
West Prussia, 1.7% in Posen). In consequence, the anticipated rate of
out-migration would be high from all the agricultural regions close to
the border. But at a distance of 500 kilometres (approximately the
German average) the position changes. Here the larger estates increased
employment opportunities more than small or medium-sized farms did,
and so Region A is predicted to increase the agricultural population by
more than Region B. The implied rate of out-migration from Region
A is substantially lower than from Region B.
The same analysis can also be used to examine the eVect of more
advanced agricultural techniques, such as the introduction of sugar beet
cultivation and also the use of labour-saving machinery such as steam-
threshing. The results given in Table 6.3 show that the intensiWcation
of arable production associated with the cultivation of sugar beet led to
an increase in employment opportunities: when controlling for other
factors the eVect of sugar beet cultivation is a positive one. The more
intensive agricultural crops such as sugar beet made increased demands
on the farm labour force, for weeding and harvesting, and also were
associated with increased livestock numbers (who could be fed on the
pulp left after the sugar had been extracted), which also raised labour
requirements.
Similarly, the use of steam-threshing was not associated with a
decline in employment opportunities, when this factor was considered
separately from other factors inXuencing agricultural employment.
Rather, the use of steam-threshing is associated with rising agricultural
employment. The larger estates did often use steam-threshing, but, as
already noted, these holdings were tending to increase employment for
other reasons. If steam-threshing did replace labour previously used to
thresh by hand (and this was the main advantage of the technology)
then it appears that these workers were redirected to other tasks. The
positive connection between the use of something which was quite
clearly a labour-saving technology, and the change in agricultural em-
ployment, suggests that this technology tended to be used mainly in
areas which were already short of labour.
6. Influences on Rural Wage Levels
It was a commonly held view that large holdings caused out-migration
partly because they paid low wages. A study published in 1907
Migration, Farm Size, Condition of the Labourer 201
![Page 211: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/211.jpg)
commented: ‘the accusation has often been made against large farms
and estates that by their nature, as well as by paying poor wages, they
have created migration out of agriculture, or at least caused it to in-
crease’.41
If so, it should be possible to Wnd this eVect by examining variations
in rural wage levels. These can be studied making use of the surveys
conducted to provide information for the Reichsversicherungsamt. Theaccident insurance legislation required that the Ortsublicher Tagelohn
should be recorded for each locality. The Prussian surveys are particu-
larly useful in that wage rates in rural areas are distinguished from
those in towns. The data was used to construct estimates of average
rural wages for each Kreis, combining the 1892 and 1901 surveys. The
full statistical results are given in Appendix 6B, Table 6B.1; Table 6.5
provides a summary of the main Wndings.
The main inXuences on wage levels came from geographical location
and from the costs of living in the locality: wages were higher in areas
where food prices were high. The impact of the border with Russian
Poland was also an important factor. Wage rates were substantially
lower in Russian Poland: male day-labourer rates were 0.52 Marks in
Russian Poland in 1890, compared to 1.13 Marks in a German border
region such as Gumbinnen (in 1892), and 1.66 Marks in a more pros-
Table 6.5. Summary of results of statistical analysis of rural
wages levels in 392 Prussian Kreise 1892/1901
I. Secure statistical results
Factors which are associated with a higher level of rural wages:
High food prices
A situation further from the border with Russian Poland
Factors which are associated with lower wage levels:
A high rate of agricultural employment in 1882
A high level of household employment in 1882
Having a high level of sugar beet production in 1897
II. Less secure statistical results
Factors which are associated with a higher level of rural wages:
The presence of large farms and estates
The proximity of large towns and cities (over 50,000 inhabitants)
41 Broesike (1907), 22.
202 Migration, Farm Size, Condition of the Labourer
![Page 212: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/212.jpg)
perous central region such as Magdeburg.42 The equivalent Wgures for
female day-labourers were 0.37, 0.71, and 1.01 Marks respectively. In
the areas closer to the border, the availability of cheap migrant labour
had a depressing eVect on German wage rates.
Table 6.6 gives the predicted levels of wages for rural areas at diVer-
ent distances from the border with Russian Poland.43 The strong eVect
of the border on rural wage levels is shown. The importance of non-
agricultural employment is also demonstrated. Wages were low in the
east mainly because of the proximity of the border and the relative
shortage of non-agricultural employment.
Once allowance is made for these factors, it can be shown that the
presence of large farms and estates did not lead to lower wages. On the
contrary there is a slight tendency for wages to be higher in areas where
there were signiWcant numbers of holdings of 100 hectares or more.
There is no justiWcation for the view that larger estates held down
wages.44 Wages were low in the rural east for other reasons.
Other factors which were associated with low wages include a high
level of household employment in 1882. Cottage industry tended to be
found in areas of high population pressure, and could therefore be an
Table 6.6. Predicted average daily wages in Pfennigs, in rural
areas, 1892–1901
Distance from the border with
Russian Poland (km)
50 100 200 500
Predicted level for a region with 20%
of the population occupied in agriculture
121.6 124.6 130.8 149.2
Predicted level for a region with 80%
of the population occupied in agriculture
98.2 101.2 107.4 125.8
42 Figures for Russian Poland are from Knoke (1911), 32, who obtained them from theWarsaw Statistical OYce and converted these into Marks; German data from Neuhaus(1904), 320–28.
43 The calculations use the coeYcients given in column a of Appendix Table 6B.1(other variables take average values).
44 The statistical result showing that large holdings tended to push up wages is not astrong one. However, the alternative hypothesis, that large estates paid signiWcantly lowerwages, can be rejected with a high degree of conWdence.
Migration, Farm Size, Condition of the Labourer 203
![Page 213: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/213.jpg)
indicator of surplus labour conditions within the Kreis.45 The Wnding
that wage levels within agriculture were lower in regions with high
levels of household employment supports this view.
The statistical analysis also found an association between low wages
and sugar beet cultivation: wages were lower in Kreise with high areas
under sugar beet relative to other arable crops in 1897. This is a sur-
prising result when considered alongside the Wnding reported in the
previous section, that sugar beet cultivation led to an increase in agri-
cultural employment. It appears to support Weber’s contention that
there was a change in the behaviour of land-owners in market-
orientated agriculture, such as the areas practising intensive arable
farming associated with sugar beet cultivation. Sugar beet cultivation
raised the demand for labour but led to a fall in the price of labour, in
apparent contradiction of the basic laws of supply and demand.
The reason seems to lie in the relationship revealed by the import-
ance of the proximity of the border with Russian Poland: the availability
of migrant labour from Russian Poland was a major factor in wage
determination in German agriculture. Much of the eVect of migrant
labour is captured by the distance variable, as the most important factor
was the proximity of the frontier. However, in the sugar beet areas the
use of foreign migrant labour was established early, and this led to the
development of networks of agents who supplied labour on a regular
basis. It was therefore easier for farmers in these regions to obtain
migrants for sugar beet cultivation and other tasks. This had a depress-
ing eVect on wages.
7. D irect Evidence on Underemployment in Rural Areas
This section looks for evidence that underemployment in German agri-
culture was a factor in the relative decline of the rural population. Was
there a shift away from areas where there was underemployment? If so,
then this would support the concept of a pre-industrial rural labour
surplus, which is then released during industrialization.
The analysis reported in Section 5 made use of a data set covering all
the German rural Kreise. This section restricts the analysis to the Prus-
45 See for example the description of conditions in the Taunus uplands, in BauerlicherZustand in Deutschland (1883), i. 156–7, where the main cottage industry had been nail-making.
204 Migration, Farm Size, Condition of the Labourer
![Page 214: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/214.jpg)
sian Kreise, which makes it possible to use other data sources, in par-
ticular the more detailed publication of the occupational census results
found in Prussian statistical sources. Table 6.7 provides a summary of
the Wndings; the full econometric results are given in Appendix 6B,
Table 6B.2.
The main results repeat those shown earlier when the full data set
was used: the agricultural population declined faster or rose least in
areas close to the border with Russian Poland, and in areas where there
was a high level of agricultural employment in 1882. There was a
tendency for the decline to be held back in remote areas (measured by
the distance from major cities).
A number of authors thought that the excessive division of land in
areas of partible inheritance had led to surplus population and under-
employment.46 If so, then this might have produced a negative relation-
ship between partible inheritance and the movement of the labour
force, as migration reduced the labour surplus in these regions. How-
ever, the analysis shows that the relationship is a positive one not a
negative one: partible inheritance is associated with a slightly larger
46 Sering (1899–1910), Conrad (1891).
Table 6.7. Summary of results of statistical analysis of the change
in the agricultural population 1882–1907 in 397 Prussian Kreise
I. Secure statistical results
Factors which are associated with a less rapid rise in the population occupied in
agriculture or with a decrease in the agricultural population:
A high rate of agricultural employment in 1882
A situation close to the border with Russian Poland
A high level of household employment in 1882
A high level of self-employment in 1882
Factors which are associated with a less rapid decrease in the population occu-
pied in agriculture or with an increase in the agricultural population:
Being a long distance from large cities (over 100,000 inhabitants)
II. Less secure statistical results
Factors which are associated with a less rapid decrease in the population occu-
pied in agriculture or with an increase in the agricultural population:
A high level of partible inheritance
Migration, Farm Size, Condition of the Labourer 205
![Page 215: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/215.jpg)
increase in the agricultural labour force in the Kreise where this was
practised, although the result is not a secure one in statistical terms.
The previous chapter has provided a possible reason for this. In areas
of partible inheritance demographic behaviour had adjusted so as to
avoid excessive subdivision. There was, therefore, no reason for there
to be any more underemployment, or surplus labour, in these regions
than there was in other rural areas.47
The analysis also considered some factors which might be indicative
of a high level of rural underemployment. These are the percentage of
the non-agricultural labour force which is self-employed and the per-
centage occupied in cottage industry. The justiWcation for including
these variables is that underemployment within agriculture in a particu-
lar Kreis may be related to underemployment outside. If there is surplus
agricultural labour, some of this will take non-agricultural work, even if
there is not full employment in this sector either. Expected returns in
both sectors will tend to equalize. And total annual returns will be a
multiple of expected daily earnings and the anticipated number of fully
employed days. So, unless there is a compensating diVerence in
expected earnings, underemployment in one sector will spill over into
the other.48
The results show that high levels of household employment and self-
employment are associated with slower increases in agricultural employ-
ment. In Kreise with this ‘labour surplus’ pattern of employment the rate
of natural population increase was considerably greater than the pre-
dicted increase in agricultural employment, leading to out-migration.
Table 6.8 illustrates this point, by providing predicted rates of increase
in agricultural employment for diVerent Kreise, ranked by the estimated
labour surplus (which is a weighted combination of the percentage in
self-employment and the percentage in household employment).
The estimates show that, in a Kreis in the top decile for ‘estimated
labour surplus’, with high levels of household employment and self-
employment, the predicted increase is well below the average rate of
natural increase (38% over the whole period). The implied rate of out-
migration from rural areas for such a Kreis is 25% of the population
in 1882. By contrast, a Kreis in the lowest decile would have had a
47 A number of regressions, using diVerent data sets and diVerent periods, were runlooking for a relationship between migration and partible inheritance. All these producedresults which were insigniWcant.
48 This implies that there is some Xoor to wages in both sectors, so eYciency wages arepaid.
206 Migration, Farm Size, Condition of the Labourer
![Page 216: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/216.jpg)
predicted increase in agricultural employment nearly equivalent to the
rate of population increase, leaving a migration Xow equivalent to just
7% of the population in 1882. This measure provides strong evidence
for the view that labour surplus conditions in rural areas contributed
heavily to migration in the period.
8. The Lewis Model and East-Elbian Agriculture
The focus in this chapter has been on the eVect that large farms and
estates had on migration out of rural areas. A central theme has been to
examine the view of Max Weber that there was a change in the values
and attitudes of land-owners which led to deterioration in the condition
of the agricultural worker and the release of surplus labour.
The evidence does not support the view that there was a general fall
in real wages after 1870, nor was there any overall tendency for larger
farms to shed labour. On the other hand, it does appear that the shift to
more intensive arable production systems, characterized by the intro-
duction of sugar beet, did lead to a change in labour relations, involving
the increased use of migrant labour, and this had a depressing eVect on
Table 6.8. Analysis of the change in the agricultural population,
1882–1907: predicted rates of change for 397 Prussian Kreise ranked by
estimated surplus labour
Ranking of Kreis by
estimated labour
surplus
Household
employment
1882 (%)
Self-employment
1882 (%)
Predicted change
in agricultural
population
1882–1907 (%)
Top decile average 18.74 39.36 þ12.95
Top quartile average 9.10 38.05 þ16.84
Average all Kreise 3.68 28.89 þ23.34
Lower quartile average 1.60 19.51 þ28.85
Lower decile average 0.87 15.59 þ31.09
Note: The ranking is obtained by using the coeVicients given in column c of Table 6B.2for %HOUSEIND and %SELFEMPLOYED, and applying them to the Kreise data toproduce an estimate of the labour surplus. Note also that the increase in the agriculturallabour force shown between 1882 and 1895 was at least in part due to the under-recording of female labour in the earlier census.
Migration, Farm Size, Condition of the Labourer 207
![Page 217: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/217.jpg)
wages in the areas most aVected. This eVect was strongest in the east,
due to the proximity of the border with Russian Poland.
One constant theme running through the results presented in this
chapter is the inXuence of Russian Poland, both on wage rates and on the
numbers occupied in agriculture. This inXuence seems to have increased.
In Weber’s view this was due to a change in the behaviour of land-owners.
There are, however, other factors which might account for this.
In the Wrst place, Russia was starting to industrialize. From the 1880s
onwards growth was rapid.49 This would have set oV a similar process
of increased mobility and migration to the one this study has described
for Germany. This could have made it easier for German farmers, and
their agents, to obtain migrant Polish workers. Another consequence of
this rise in mobility was a dramatic increase in the emigration of Rus-
sian Poles to the United States.
Secondly, there may have been a change in the attitude of the
workers themselves. The bargain represented by the traditional con-
tractual relationship of eastern agriculture was one-sided. Although
Weber interpreted the contractual relationships of eastern agriculture as
reXecting a pre-capitalist desire on the part of the land-owners to
employ as many as possible, the evidence he provides is not consistent
with this. The essence of the Inste system was that it did not oVer year-
long full-time employment (as some landlord who sought only to maxi-
mize employment might have done). Rather it ensured that the peak
seasonal labour requirements of the land-owner were met, even if this
meant considerable underemployment for the rest of the year. The
provision of smallholdings and grazing rights meant that this ‘surplus
labour’ could be absorbed in the Inste household economy during the
slack periods. The system he described was one motivated by self-inter-
est not paternalism.
Industrialization made it less necessary for agricultural workers to
accept this. There is evidence that some eastern workers, and in par-
ticular their children, gave up their contractual status voluntarily, in
response to a general rise in the demand for labour and increased wages
outside eastern agriculture.50 So land-owners were obliged to look else-
where for labour to meet peak seasonal requirements. Russian Poland
was the nearest alternative.
49 Gregory (1997).50 There is a report on this in the Jahrbuch fur die Amtlich Statistik des Preußischen
Staats II (1867), 284.
208 Migration, Farm Size, Condition of the Labourer
![Page 218: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/218.jpg)
Taking the long view, the movement of German farmers into the
eastern provinces, and into other regions of eastern Europe, was a
product of high population pressure in other parts of Germany, and a
lack of suitable land for settlement nearby. In the eighteenth century
the Prussian state was able to attract settlers from Wurttemberg and
Switzerland.51 But around 1860, this eastwards shift of the German
population, which had been going on for centuries, came to an end.
Instead there is a movement of the German population to the centre
and west, and this is followed by an inXow of migrants from lands
further east. The reason for this was the rise of industry in the west
and the centre.
Weber, and other nationalists, saw this as a problem. But for those
who left eastern agriculture it was an opportunity. Conditions in indus-
trializing cities were harsh, but for many they represented an improve-
ment on what they had left behind. It was no longer necessary to accept
institutional underemployment.
The broad conclusion of this chapter is favourable to the Lewis
Model and to the concept of the rural labour surplus. The evidence
points to the existence of a reserve of underemployed rural labour
available for use in industry. But it also makes the point that geograph-
ical factors are important. Germany was not an island. It was connected
to conditions in the labour markets of eastern and central Europe.
Surplus labour in these markets was available for use in Germany. The
operation of the Lewis Model in one country can be aVected by condi-
tions in neighbouring economies.
51 Keyser (1959), Hauser (1965).
Migration, Farm Size, Condition of the Labourer 209
![Page 219: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/219.jpg)
Appendix 6A. Description of Regress ion Variables
and Sources
%DAGPOP Percentage change in the agricultural population
(includes dependants of occupied population)
1882–1907, from occupational censuses, SDR
n.f. 2 (for 1882) and SDR n.f. 209 (for 1907),
adjusted to 1882 Kreis boundaries where these
had changed.
FOODPRICES A weighted index of food prices for 1899–1901,
for 165 Prussian towns, weights are 1900 shares
in private consumption from HoVmann (1965),applied to the nearby Kreise, prices from Neu-
haus (1904).
RURALWAGES Average wages in Pfennigs (Ortsubliche Tage-
lohne) for rural areas; the Wgures are averages of
male and female daily rates, 1892 and 1901,
ZdKPSB (1904), 320–8.
%LAND>100HA The percentage of the total agricultural area in
holdings of over 100 hectares, from 1895 Agri-
cultural Census, SDR n.f. 112, pp. 489–500.
%AGOCCP1882 The population occupied in agriculture as a
percentage of the total occupied population,
from the 1882 occupational census, SDR n.f. 2.
RUSSDISTANCE Shortest distance in kilometres from the Kreis
mid-point to the Russian border (measured
from contemporary maps).
CITYDISTANCE(a) Distance in kilometres from the Kreis mid-
point to the nearest city with at least 200,000
inhabitants in 1900.
CITYDISTANCE(b) As above, for cities with at least 100,000 inhab-
itants.
%SUGAR1897 Sugar beet area as % of the total arable area, by
Kreise, for 1897, from PS vol. 154 (1898), 8–163.
%PARTIBLE The percentage of land transferred by inherit-
ance in 1896–1902 which was divided, ZdKPSB(1905), 262–7.
%HOUSEIND The percentage of the non-agricultural occu-
pied population occupied in Hausindustrie, 1882
occupational census, from PS vol. 83 (1884),
part 2, pp. 3–466.
%SELFEMPLOYED As above, for those recorded as self-employed.
210 Migration, Farm Size, Condition of the Labourer
![Page 220: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/220.jpg)
Appendix 6B. Results of Econometric Analysis
(i) Statistical analysis of wage levels in rural Prussian Kreise 1892/1901
The dependent variable is taken from the Wgures provided by the Reichs-
versicherungsamt. These were used to calculate average rural wages for
each Kreis combining the 1892 and 1901 surveys.52 In the surveys
diVerent wage rates were given for male and female workers, and for
those aged below 16. The Wgures used are weighted averages of these,
the weights being the share of these categories in the total agricultural
52 Figures are from Neuhaus (1904).
Table 6B.1. InXuences on rural wages in the Prussian Kreise
a b c d
Constant 69.0 84.6 77.8 87.8
FOODPRICES þ0.581 þ0.486 þ0.516 þ0.491
(4.40) (3.72) (3.86) (3.78)
RUSSDISTANCE þ0.0615 þ0.0649 þ0.0681 þ0.0626
(15.08) (16.05) (14.35) (14.96)
%AGOCCP1882 �0.390 �0.458 �0.459 �0.442
(8.38) (9.45) (9.31) (9.04)
%HOUSEIND �0.308 �0.288 �0.271 �0.282
(3.94) (3.77) (3.48) (3.70)
%SUGAR1897 �4.310 �4.142 �4.953
(4.39) (4.13) (4.82)
%LAND>100HA þ0.060
(1.15)
CITYDISTANCE(a) þ0.011
(0.66)
CITYDISTANCE(b) �0.0480
(2.03)
N 392 392 392 392
S 13.76 13.42 13.43 13.37
R2 0.738 0.752 0.753 0.754
adj. R2 0.735 0.748 0.748 0.750
F-test of regression 272.9 233.6 167.0 196.9
RSS 73440 69564 69253 68830
Note: Dependent variable is average daily wages in Pfennigs, 1892 and 1901 surveyscombined (RURALWAGES), estimated using OLS, t-ratios in brackets.
Migration, Farm Size, Condition of the Labourer 211
![Page 221: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/221.jpg)
workforce for the whole of Germany (from the 1907 occupational
census). They are thus standardized estimates, removing the eVect of
variations in the age or sex distribution of the labour force by Kreis.
These were then used as the dependent variable in the regressions
shown in Table 6B.1.
The Wrst column gives the basic results. One important inXuence on
wages was the cost of living in the locality. Ideally, it would be possible
to construct an index of real wages, making allowance for housing costs,
and the prices of all consumer goods, but the data sources are not
available for such a calculation. However, there was a survey of food
prices in 165 Prussian market towns, published in the Zeitschrift des
Koniglich Preußischen Statistischen Bureaus. The data is not given at the
level of the individual Kreise, and only Wve items are covered (rye,
wheat, potatoes, beef, and pork).
This survey was used to construct a weighted index of food prices
(FOODPRICE) for nearby Kreise. The Wrst line of Table 6B.1 shows
that this does indeed have an important eVect on wages. Wages are
higher in areas with high food prices.
Other inXuences include the level of employment in non-agricultural
sectors in the Kreis: the coeYcient on %AGOCCP1882, the percentage
occupied in agriculture according to the 1882 occupational census, is
strongly negative. Wages were lower in regions where there was little
alternative employment. However, the coeYcient on employment in
‘cottage industry’ (%HOUSEIND—non-agricultural household
employment) is a negative one. As the presence of other forms of
non-agricultural employment has a positive eVect, this implies that the
presence of cottage industry may well be an indicator of underemploy-
ment. Cottage industry was found in areas which had a long-term
problem of overpopulation and underemployment.
However, the most important eVect on the structure of wages in
German rural areas was the distance in kilometres from the frontier
with Russian Poland (RUSSDISTANCE). This is an important eVect:
wages rise by 6–7 Pfennigs per 100 km. A regression equation with just
RUSSDISTANCE as the sole explanatory variable has an R2 of 0.673,
indicating that this factor alone explains over two-thirds of the observed
variance in money wages.
Adding sugar beet (column b) improves the overall Wt. The coeY-
cient is negative, indicating that wages in sugar beet regions were lower
than would be expected. This coeYcient is not greatly aVected by the
inclusion of other variables in columns c and d.
212 Migration, Farm Size, Condition of the Labourer
![Page 222: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/222.jpg)
The third column introduces %LAND>100HA, the percentage of
land in large holdings. If large estates did pay low wages then this
variable should have a negative coeYcient. In fact the estimated coeY-
cient is positive although not signiWcant at the 5% level. There is no
evidence to support the view that larger farms and estates paid low
wages; the evidence suggests that they may have raised wages.
The CITYDISTANCE variables introduced in columns c and d are
used to test the robustness of the estimated coeYcients on the other
variables. The strongest eVect on wages comes from CITYDISTAN-
CE(b) (the distance to cities with over 100,000 inhabitants). The co-
eYcient is negative—wages are lower further away from the major
cities—which is as expected. The other results are not seriously aVected.
Table 6B.2. Underemployment and the change in the agricultural
labour force
a b c
Constant 21.6 30.3 35.2
%AGOCCP1882 �0.500 �0.458 �0.555
(7.63) (5.90) (6.86)
RUSSDISTANCE þ0.0468 þ0.0485 þ0.0477
(10.52) (11.09) (11.86)
CITYDISTANCE(a) þ0.123 þ0.123 þ0.181
(5.14) (5.24) (7.39)
%PARTIBLE þ0.107 þ0.099
(1.30) (1.23)
%HOUSEIND �0.330 �0.334
(3.25) (3.00)
%SELFEMPLOYED �0.359 �0.512
(2.58) (3.50)
N 347 347 397
S 17.47 17.05 19.47
R2 0.409 0.442 0.432
adj. R2 0.402 0.432 0.424
F-test of regression 59.1 44.7 59.4
RSS 104361 98505 148182
Note: Dependent variable is change in total agricultural population 1882–1907 as apercentage of the 1882 Wgure (%DAGPOP), estimated using OLS, t-ratios inbrackets.
Migration, Farm Size, Condition of the Labourer 213
![Page 223: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/223.jpg)
(ii) Underemployment and migration
The analysis in this section looks at the connection between under-
employment and migration, making use of data sources which are only
available for the Prussian Kreise. The dependent variable in Table 6B.2
is the change in the total agricultural population in the Prussian Kreise
between the two occupational censuses of 1882 and 1907. The Wrst
three variables are those found to have the largest eVect on the move-
ment of the agricultural population. The decline was steepest in areas
which were closest to the border with Russian Poland: by contrast,
farms in central and western Germany tended to raise employment.
The decline was also steeper in areas close to major cities, as is shown
by the positive coeYcient on CITYDISTANCE(a).
There is a strong negative eVect from the percentage occupied in
agriculture in 1882 (%AGOCCP1882). This can be interpreted as
showing that there was underemployment in many predominantly agri-
cultural areas at that time, so that the decline in the agricultural labour
force was higher in Kreise which had little non-agricultural employ-
ment. The data set was used to study the eVect of partible inheritance,
using %PARTIBLE (the percentage of land transferred by inheritance
1896–1902 which was partitioned). Columns a and b show that the
coeYcient on this variable is positive and not signiWcant. There was no
tendency for partible inheritance to produce out-migration; there was a
slight eVect in the opposite direction.
Column b adds two other variables. These are employment in cottage
industry and self-employment (%HOUSEIND and %SELFEM-
PLOYED). Both are expressed as percentages of total non-agricultural
employment, so they are possible measures of underemployment out-
side agriculture.
The results in columns b and c show that coeYcients on both these
variables are negative and signiWcant. So, summarizing these results,
there is good evidence that there was underemployment in rural areas
in 1882, and that, in 1882–1907, this was reduced through migration, as
the rural labour surplus was transferred to urban labour markets.
214 Migration, Farm Size, Condition of the Labourer
![Page 224: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/224.jpg)
7
Agricultural Productivity,
Labour Surplus, and Migration
1. Introduction
In the previous chapters a number of diVerent explanations of the
agricultural ‘labour surplus’ have been considered. Some have been
rejected, some have found at least partial support: the agrarian reforms
did create a class of landless labourers, there was considerable demo-
graphic pressure in parts of rural Germany, there was underemploy-
ment in the agrarian sector and its ‘cottage industry’ oVshoot, there
were institutional and social changes in agriculture which contributed
to out-migration.
In Chapter 1 there was a discussion of Theodore Schultz’s critique of
the concept of the labour surplus. The essence of Schultz’s view is that,
although productivity may be low in traditional agriculture, this is not
due to ineYciency or underemployment, but rather to a lack of
resources, in particular to a lack of ‘non-traditional resources’: educa-
tion, science, new productive knowledge. Schultz also stressed the im-
portance of access to produce markets and to capital.
Industrialization brings to agriculture access to improved inputs. The
creation of a railway network and other transport improvements make it
possible to transport bulky inputs like fertilizers and animal feedstuVsover long distances. This also makes it easier to get agricultural produce
to market, and, by integrating markets, helps to reduce price volatility.
Education is another productive factor which is likely to be aVected
by industrialization. Industrializing societies, particularly ones like
Germany which develop a specialization in advanced technology, Wnd it
desirable to invest in educational improvements so as to raise the skill
level of the workforce. If the system is a universal one, then the rural
sector will also have access to a higher level of general education. In
specialist Welds such as scientiWc education, advances in basic science
often have implications for agriculture. Nineteenth-century Germany
was a classic case of this: advances in the understanding of organic
![Page 225: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/225.jpg)
chemistry fed into an improved understanding of plant nutrient needs
and how to supply these.
Industry also supplies improved inputs directly. Steam power could
be applied to agricultural tasks, such as ploughing and threshing. In-
dustrial by-products such as ‘basic slag’ turned out to be valuable fertil-
izers. Improvements in mechanical engineering made it possible to
design and build machines such as reaper-binders or seed drills, which
were suYciently robust to stand up to being pulled over Welds while in
operation.
These ‘non-traditional inputs’ can, therefore, lead to productivity
increases and improved yields as farmers react to the stimulation pro-
vided by rising urban demand and better transport links, by investing
in new technology. This does not lead directly to out-migration, but it
does have indirect eVects which might produce this. In the Wrst place, it
will lead to social change within agriculture: as this becomes more
market-orientated, traditional social attitudes and institutions will tend
to be eroded. Secondly, it creates the possibility that the nutritional
needs of the country could be supplied by a much smaller agricultural
sector (measured by the relative size of the agricultural population).
Whether or not there is a relative decline in the agricultural sector as a
result depends on the evolution of the foreign trade position. It is quite
possible for countries to industrialize, to develop modern industrial
sectors, and at the same time remain major exporters of primary prod-
ucts, including agricultural goods. Examples include Brazil and Austra-
lia in the twentieth century and the United States in the nineteenth. In
these examples, the industrial sector concentrated on supplying the
home market. Under these circumstances, the relative importance of the
primary-producing sectors can remain high despite industrialization.
But consideration of these examples shows quite clearly that they
were all ‘resource rich’ countries, with comparatively small populations
relative to the availability of land and other resources. Late nineteenth-
century Germany was not ‘resource rich’ in these terms, especially as
the expansion of European settlement and falling transport costs were
bringing new areas into the world trading system which had far more
favourable land–labour ratios. Germany’s comparative advantage in
world trade lay in her scientiWc base, in the education system, and in
the skills and knowledge possessed by urban workers. Germany was
‘skill rich’ not ‘resource rich’.
The logic of Germany’s position was, therefore, that the German
economy should export skill-intensive industrial goods and import food
216 Agricultural Productivity, Labour Surplus, Migration
![Page 226: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/226.jpg)
and raw materials. This meant that demand for home-produced agricul-
tural goods was limited to a declining share of the home market. Under
these circumstances, improvements in agricultural productivity meant
that the available demand could be supplied by a smaller labour force.
As Germany became more integrated in the world trading system, re-
sources, including labour, would shift to the exporting industrial sector,
and leave the sector exposed to rising imports, which was agriculture.
The context for the discussion of these issues is the more general
question of the relationship between agriculture and industry in the
circumstances of the late nineteenth century, when cheap US grain
exports reduced prices on world agricultural markets. European coun-
tries reacted in diVerent ways, inXuenced by internal political and social
pressures, and by economic conditions. Many tried to mitigate the
eVect on their agricultural sectors by increasing tariVs on agricultural
goods. Germany was not exceptional in this.
An important point about Germany is that it was already industrial-
izing, and that it had a comparative advantage in industrial goods. This
meant that it was able to pay for increased agricultural imports by
exporting industrial goods. This was not an option available to all
European countries. Perhaps only Britain and Belgium were in a similar
position.1
A second point about the German position is that the arrival of cheap
US grain exports in the 1870s came at a time when Germany was still
predominantly rural, and when German cities were already under pres-
sure from rapid growth. By contrast, the agricultural population in
Britain had become relatively small by that time, and British cities had
been growing steadily for a century or more. The consequences to
British society of a free-trade policy were absorbed much more easily.
A simple calculation illustrates this point. Suppose that a country
with the rural–urban balance of Britain in 1871 (42% of the population
in towns with at least 20,000 inhabitants) maintained a free-trade policy
in face of rising agricultural imports, and that this caused 10% of the
rural population to move to the cities over the next ten years. The
additional migrants would amount to 14% of the initial urban popula-
tion: a burden for the urban areas but not an insurmountable one. The
same calculation for a country with a similar balance to Germany in
1 For a general discussion of these issues, see O’Rourke (1997). There is an importantdistinction to be made between countries which had the capacity to raise industrial exportsand other less industrial countries, whose best option was probably to reorientate agricul-tural exports to the livestock sector, as Denmark did.
Agricultural Productivity, Labour Surplus, Migration 217
![Page 227: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/227.jpg)
1871 (12.5% in towns with at least 20,000 inhabitants) would produce a
movement equivalent to 70% of the initial urban population. This
would have raised the growth rate of German cities from 3.0–3.5% per
annum to over 8%. Such a rate would have created intolerable strains.
If there is a relationship between the social costs of migration and the
rate of urban growth, there may be good reasons for a government to
seek to slow down the decline of the agricultural sector. Agricultural
protection is probably undesirable on economic grounds, but it may be
a reasonable response for a society which fears the consequences of a
collapse of the agricultural sector.2
2. Agriculture in the German Economy
The relative importance of the German agricultural sector fell in 1870–1913, despite protection. Table 7.1 presents some calculations, from
Wgures provided by HoVmann, which show how the changing structure
of the German economy aVected the position of agriculture.
HoVmann made a reasonable adjustment to the agricultural labour
force Wgures to allow for under-recording of family labour in the earlier
censuses.3 The resulting estimates show the agricultural labour force
rising at a slow rate, 25% in 42 years (1871 to 1913), while the non-
agricultural labour force more than doubled (a rise of 125%). The
estimated rise in agricultural productivity 1870/2 to 1911/13 is 1.0%
per annum, which is relatively high compared to other European coun-
tries (such as Britain), but still below the rise in the non-agricultural
sector (1.4% per annum). Consequently, the fall in the agricultural
share of total net value added is rather greater than the fall in the share
of the total labour force.4
It is possible to decompose the fall in the agricultural share of net
domestic product into two parts: that which is due to changes in the
structure of domestic demand and that which is due to changes in
foreign trade. As the table shows, there was a small net deWcit on
2 There are other reasons to support agriculture, including the strategic one. A countrydependent on food imports is in a vulnerable position, as was demonstrated by theGerman experience in the First World War, a point discussed in OVer (1989).
3 On the under-recording of female labour in the 1882 census, see the discussion inTipton (1976) and also comments in the introduction to the 1895 survey, SDR n.f. 112.
4 The increase in labour productivity in British agriculture averaged 0.5% per annum1873–1913, Matthews et al. (1982), 228–9.
218 Agricultural Productivity, Labour Surplus, Migration
![Page 228: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/228.jpg)
agricultural trade in 1870–2, equivalent to 0.7% of NDP (1913 prices).
By 1911–13 this had risen to 4.5% of NDP. The change in the external
balance contributed 20% of the total fall in the agricultural share of net
value added.5 But this shows that there would have been a substantial
fall in the agricultural share even without the contribution of the exter-
nal sector. The shift in the structure of domestic demand had rather
more eVect than the change in the foreign trade position.
Comparison of the net import Wgures with the Wgures for total agri-
cultural production (not just value added) shows that agricultural self-
suYciency in all sectors, deWned as total domestic production as a per-
centage of total consumption, fell from 98.7% in 1870–2 to 88.1% in
1911–13. This exaggerates Germany’s ability to feed itself without
using imports, since it takes no account of imported fertilizers and
other non-agricultural inputs. Moreover, the importance of fertilizers,
in particular, was far greater than that implied by the Wgure of their
cost relative to total production.6
Figure 7.1 concentrates attention on trade in cereals. In the period
before 1870 Germany exported cereals in most years. Germany sup-
plied 23.5% of British cereal imports in 1856–60, but this share was
declining, in 1871–5 it was only 8.2% (the share of the United States
5 This Wgure is based on an estimate of the size of the agricultural share in NDPassuming that the deWcit on agricultural trade had remained at the 1870–2 Wgure. It takesaccount of the point that imports were themselves Wnal products; their importance tends,therefore, to be exaggerated by a comparison with domestic value added.
6 A calculation of self-suYciency including the eVect of imported fertilizers is noteasily done. Some fertilizers also had non-agricultural uses (nitrates were used in themanufacture of explosives). Germany exported potash and imported phosphate and nitro-gen. Simply adding up tons imported and exported will give a misleading impression.
Table 7.1. Agriculture in the German economy 1870–1913 (all
calculations use 1913 prices)
Agricultural share in: Exports as % of NDP Imports as % of NDP
NDP Labour
force
Non-
agricultural
Agricultural Non-
agricultural
Agricultural
1870–2 38.7 49.2 10.1 2.7 13.0 3.4
1911–13 23.0 34.9 18.0 1.8 16.4 6.3
Source: Data from Hoffmann (1965); Hoffmann builds up his accounts from estimates of netvalue added by sector, the final figure is therefore net domestic product (NDP).
Agricultural Productivity, Labour Surplus, Migration 219
![Page 229: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/229.jpg)
rose from 18.8% to 40.9% over the same period).7 But before conclud-
ing that it was US competition that was the cause of this loss of share,
it should be noted that German producer prices actually rose in
1856/60 to 1871/5: by 13% for wheat, and 15% for rye. The eVects of
increased US competition in third markets was outweighed by buoyant
domestic demand.
A careful examination of the timing of the fall in cereal producer
prices in the 1870s (given in Figure 7.2) shows that this came after the
shift to a net import position. The 1870s were a period of relatively
high producer prices, certainly better than the early 1860s. There is no
evident downward trend in cereal prices until after the peak in 1873–4.
7 Lambi (1963), also Hobson (1997).
1910190018901880187018601850
30
20
10
0
−10
Fig. 7.1. Net imports of cereals as a percentage of total domestic consumption
(including use as animal feed), Germany (post-1871 boundaries) 1850–1913
Note: It is difficult to find any clear evidence in this diagram of effects from changes intariff policy, Bismarck’s duties of 1879–80 produced, at the most, a brief interruption inthe upwards trend in imports. The diagram also prompts the question: was it conceivable,in the context of pre-1914 power politics, for any country to have been as dependent onoverseas trade for vital supplies, and yet not build a navy to protect this trade? See Offer(1989) for a discussion of the consequences. The issue is also considered in Chapter 10 ofthis study.Source: Calculated from Hoffmann (1966), 292–3, 530–5, and 537–43.
220 Agricultural Productivity, Labour Surplus, Migration
![Page 230: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/230.jpg)
It was the low prices of the 1878–9 period that changed the attitude of
the German agricultural lobby, which had previously supported free
trade.
At the same time domestic demand for food was buoyant, increasing
the viability of a strategy which relied more on protected domestic
markets, and less on exports. Table 7.2 gives some Wgures for per capita
consumption. Between 1850–4 and 1870–4 there were substantial rises
in most categories, apart from pulses and other cereals. The increases in
the consumption of sugar and pork are particularly notable. As the
German population rose by 16% over this period, the overall increase
in demand was high.
Rising demand for livestock produce increased the amount of cereals
required as livestock feed: principally oats, but also some barley and
rye; potatoes were also fed to livestock. This reduced the area that
could be allocated to exports. One consequence of this period of relative
agricultural prosperity was a rise in average rents, from 14.10 Marks
189018851880187518701865186018551850
300
250
200
150
100
50
0
Wheat
Rye
Barley
Oats
Fig. 7.2. Cereal producer prices 1850–95 (in Marks per ton)
Note: The ranking of the crops by production in 1870–2 is, first, rye (40% of the total byweight), then oats (28%), wheat and spelt (17%), and finally barley (15%).Source: From Hoffmann (1966), 552–4.
Agricultural Productivity, Labour Surplus, Migration 221
![Page 231: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/231.jpg)
per hectare in 1850 to 38.16 in 1880.8 This is a clear indication that
problems in third markets had not been the principal cause of the
reduction in exports.
After 1880 the connection between rising imports and falling produ-
cer prices becomes clearer. There is a rise in net imports from 7% of
total consumption in the period around 1880 to 15% in the mid-1890s,and there is an evident downward trend in producer prices. TariVs did
have some eVect in holding back the fall in prices, as Table 7.3 shows.
The fall in producer prices was 53 Marks per ton (1871/80 to 1891/
1900), but the Hamburg price fell by 93 Marks per ton over the same
period. Comparison of this price with the prices in Nebraska and New
York shows the eVect of falling transport costs: a total reduction of
55%, 70% of this coming from a reduction of costs within the United
States.
After 1895 net imports of cereals remained at around 15% of total
consumption, but net imports of all agricultural produce continued to
rise. In 1911–13 net imports were 15.6% of total food consumption; net
imports of foodstuVs other than cereals and sugar having risen from
8 Henning (1977), 223.
Table 7.2. Per capita food consumption: five-year averages
Wheat
and rye
Potatoes Pulses and
other cereals
Sugar Beef &
veal
Pork Milk Eggs
Annual consumption per capita in kg: In numbers:
1850–4 90 120 26 2 10 7 272 47
1870–4 114 169 18 5 12 13 296 54
1890–4 127 242 12 9 15 18 313 72
1909–13 128 200 11 18 17 25 381 102
Note: There are reasons to question the 1850–4 figures. The implied increase in percapita calorie consumption of these foods between the first two periods is high (20.8% in1850/4 to 1870/4 against an increase of 7.8% in 1890/4 to 1909/13). Some of this maybe explained by the increased use of cereals and potatoes for domestic alcohol production.But, in the earlier years, Hoffmann may also have underestimated personal consumptionof pulses and other cereals (as opposed to their use as livestock feed). High potatoconsumption in 1890–4 may also be spurious, if a high proportion was fed to livestock inthese years. But in general, the implied calorie consumption figures for the period after1870 are not unreasonable.Source: Calculated from Hoffmann (1966), 622–33.
222 Agricultural Productivity, Labour Surplus, Migration
![Page 232: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/232.jpg)
2.7% of total consumption in 1870–2 to 12.0% in 1894–6 and 15.5% in
1911–13. The rise after 1894–6 was not due to a increase in meat
imports, which fell slightly in the period to 1911–13 (including trade
in live animals), but to rises in other categories such as fruit (up
168%) and eggs and dairy products (up 180%) over the same period.
By 1911–13 Germany was importing nearly 11% of total consumption
of eggs and dairy produce.9
The impressive performance of German agriculture in raising pro-
duction has, therefore, to be set against a background of buoyant
demand. Domestic production was not suYcient to supply all the food
required by the expanding urban population. But it was important that
productivity rose by as much as it did. A fairly simple calculation shows
that, if agricultural productivity had risen by 0.8% per annum instead
of 1.0%, domestic food production in 1911–13 would have been 26%
9 Figures from HoVmann. The question of the relative rates of protection given todiVerent agricultural sectors has been a matter of controversy since Gerschenkron’s 1943study, Gerschenkron (1943), Hardach (1967), and Webb (1982).
Table 7.3. Wheat prices, tariffs and transport costs
(a) Average wheat prices, by decade
(Marks per Ton)
1871–80 1881–90 1891–1900
Nebraska 106 99 81
New York 195 155 123
Hamburg (f.o.b.) 227 164 135
Average German producer prices 224 187 171
(b) German duties on imported wheat (minimum duty payable)
1879–85 1885–7 1887–91 1891–1902 1902–6 1906–10
Tariff in Marks/
Ton
10 30 50 35 55–75
(sliding
scale)
55
Source: Prices in Nebraska, New York, and Hamburg taken from Henning (1977), 219,German producer prices calculated from Hoffmann (1966), 552–4, tariffs from Tracey(1982). These are the minimum rates, so the lower rates for 1891–1902 reflect the tradetreaties with Austria-Hungary, Romania, and Russia. American grain was still paying the1887–91 rate.
Agricultural Productivity, Labour Surplus, Migration 223
![Page 233: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/233.jpg)
lower. Under these circumstances, the achievement of a similar level of
self-suYciency to that actually attained in 1911–13 would have required
additional capital and labour. Using a joint marginal output coeYcient
of 0.85 for capital and labour (and assuming that the capital input rose
in proportion with labour), it can be calculated that the agricultural
labour force would have had to be 40% higher, so that agriculture
would have continued to account for 49% of the total labour force.
This would have had a major eVect on German industrial growth. It
also implies a much lower rate of rural–urban migration.10
A similar calculation shows that the cost of achieving full agricultural
self-suYciency would also have been high. To achieve the necessary
increase in production 16% more labour would have been needed in
agriculture. This would have meant that 40% of the labour force would
have still been occupied in agriculture in 1911–13 instead of 35%.
Again, this would have held back industrial growth.11
There was a cost to protection. Food prices were raised, though by
less than the price diVerences given in Table 7.3. Calculations using
price ratios from the 1905 survey by the British Board of Trade, shown
in Table 7.4, show that food prices were 14.5% above the British level
(weighted average of the Wgures in the table). According to the house-
hold budget surveys, these items accounted for 40% of total household
expenditure. So there was certainly an eVect on real wages from protec-
tion. But this was oVset by lower prices for milk and potatoes, presum-
ably reXecting lower German production costs in these sectors.
Germany faced some diYcult choices about the position of agricul-
ture within the whole economy, and it is not surprising that this was an
issue which was much debated. The outcome which was chosen was
one that involved a slow decline in self-suYciency, a rate of industrial
growth which was fast by the standards of the time, large-scale migra-
tion to the cities, and a level of agricultural protection which imposed
some costs on urban consumers.12 It was in many ways a sensible
10 Van Zanden (1991) suggests coeYcients of 0.35 for land, 0.5 for labour, and 0.15 forcapital (livestock). This gives a joint coeYcient on mobile factors (capital and labour) of0.65. This is probably too low. Additional investment could have raised the value of land(by drainage for example) even though there was little suitable land left uncultivated. Ajoint coeYcient of 0.85 is more reasonable.
11 Full self-suYciency is deWned as an increase of production suYcient to reduce netimports to zero. Germany would still have imported some products and exported others.There would also have been imports of fertilizers.
12 There was, however, a political price. The crisis in agriculture, even if it was a crisis ofadjustment, contributed to the radicalization of right-wing politics, Rettalack (1988), 101.
224 Agricultural Productivity, Labour Surplus, Migration
![Page 234: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/234.jpg)
solution, which avoided the social costs of a complete agricultural col-
lapse, but still allowed urban living standards to increase.
3. ‘Few better farmers in Europe ’ : Technological
Progress in German Agriculture 1870–1913
In 1921 J. H. Clapham published The Economic Development of France
and Germany 1815–1914. Writing about German agriculture in com-
parison with France, and, by implication, in comparison with Britain as
well, Clapham was expressing opinions widely held by contemporary
observers of pre-1914 German agriculture. Germany was perceived as
the leader of ‘scientiWc agriculture’:
there were few better farmers in Europe than the best of the Junkers in the later
nineteenth century . . . after the decade 1840–50 Germany had little to learn
from any country in agriculture and forestry. In agricultural chemistry she was
undoubtedly the leader.13
In his writings Clapham placed great emphasis on the role of the
larger farms and estates of eastern Germany: ‘these were the Welds on
13 Clapham (1921), 206.
Table 7.4. German food prices as percentages of British prices, 1905
(percentage of total German household expenditure spent on each
category shown in brackets)
Wheat and rye flour 140 (2.3) Beef 122 (4.4)
Wheat and rye bread 131 (10.0) Mutton 137 (0.3)
Sugar 119 (1.4) Pork and lard 123 (8.4)
Potatoes 88 (2.6) Butter 105 (4.6)
Milk 75 (5.8)
Source: Price ratios from Board of Trade (1908); household expenditure proportions takenfrom Desai (1968), 122, who used the results of a household budget survey published inthe 1909 Reichsarbeitsblatt. Additional information on the consumption of different meatswas taken from Hoffmann (1966), 632. Price ratios for pork were also applied to lard;those for wheat flour were used for rye as well. The price ratio for bread was derivedusing the baker’s premium found by Peterson (1995), 271: 24% of the British price forwheat flour. This was used as an estimate of the absolute cost of producing bread fromflour in both countries. The Board of Trade survey provides a better estimate of theeffect of German tariffs than calculations based on the movement of prices insideGermany. World prices were falling in the late nineteenth century; German tariffs heldback this decline.
Agricultural Productivity, Labour Surplus, Migration 225
![Page 235: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/235.jpg)
which the Wght for scientiWc agriculture was won’. But he also pointed
out that these diVered from the typical English estate of the time, which
was rented out to tenant farmers:
the German Empire was . . . a land of free land-owning peasants and powerful
cultivating squires . . . In a broad survey of modern German agriculture tenant
farming might almost be omitted . . . East of the Elbe where big estates were
common they were nearly all in hand. The typical eastern Junker lived on and
directed the management of an estate of perhaps 2000 acres of cultivated land . . . 14
ScientiWc agriculture involved, in particular, the application of the
discoveries of organic chemistry to agricultural practice. But it was also
recognized that knowledge of these new practices had to be diVused.
This involved a major expansion of agricultural education. By 1908
Germany had 26 Agricultural Mittelschulen, 45 Ackerbauschulen, 43
Agricultural Fachschulen, and, for those without access to a full-time
secondary school (or unable to aVord the time), there were Winterschu-len (204 of these in Prussia alone).15 Agricultural faculties had been
established at many universities. There was a network of research sta-
tions and experimental farms run by the provincial authorities and the
local Landwirtschaftsgesellchaften.16
The new agricultural science was based on an improved knowledge of
the nutrients required by plants, and the extent to which these would be
removed when the crop was harvested. Soil testing was used to establish
the extent to which nutrients could be supplied from resources already
present in the soil. DeWciencies could then be made up using a combin-
ation of animal manure, artiWcial fertilizers, and ‘green manure’: nitrogen-
ous plants which were ploughed in to provide food for subsequent crops.17
The story of one estate, Kittnau in Kreis Graudenz in West Prussia,
illustrates the extent of the changes:
14 The entrepreneurial abilities of the Junkers have been recognized by some authorswho are critical of their political inXuence, for example Rosenberg (1978), 94. On theother hand, Puhle (1986), 86, claims that the Junkers refused to invest or modernize theirestates, and had a ‘general disinterest in the nature of agricultural production’. Accordingto Schissler (1986), 36, ‘historians agree that the Junkers represent a great political,economic and social burden for Prusso-German history’.
15 Figures from Muller (1985).16 The German agricultural education system was surveyed in 1881–3 by the Royal
Commission on technical education. This part of the report was published in the Journalof the Royal Agricultural Society of England and Wales (1885), 126–64 and 518–46.
17 Valuable studies of late nineteenth-century agricultural technology and its eVect onproductivity include Evershed (1884), Dyer (1896), Rauchberg (1900), Sering (1932),Schremmer (1988), Haushofer (1957), Frauendorfer (1956), Rolfes (1976), Henning(1977), Perkins (1986), Reif (1994), and Fussell (1962).
226 Agricultural Productivity, Labour Surplus, Migration
![Page 236: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/236.jpg)
Previously there had been little cultivation of root crops. In 1884 a sugar beet
factory was constructed at the railway station at Melno, 5 km from the estate,
and the manager, Oberamtmann Muller, immediately switched to intensive
sugar beet cultivation. Clover was replaced with a rotation using lucerne, and a
third of the farmed land was put into sugar beet. The arable land was drained, a
Fowler steam plough was bought and sugar beet was cultivated according to the
practices developed in Prussian Saxony. In order to move the sugar beet
promptly to the factory, around 5 km of roads were built at a total cost of
40,000 Marks. At the same time cereal yields also increased as a result of the
improved husbandry system.18
What evidence is there of a major change in the performance of
German agriculture? The Wrst place to look is movement of yields rela-
tive to other countries. Figure 7.3 gives average wheat yields in four
countries 1848–1913. In the 1850s German yields were at a similar level
1908189818881878186818581848
4
3
2
1
0.9
Germany
log scale
FranceDenmarkBritainGermany, eastGermany, rest
Fig. 7.3. Wheat yields 1848–1913, 1,000 kg per hectare, rolling five-year
averages
Source: Calculated from data in Mitchell (1981), from national sources, but then adjustedusing the 1905–9 Institut International d’Agriculture comparison of cereal yields: InstitutInternational d’Agriculture (1915), 28. The British figures were converted to metric tonsusing estimates of bushel weights from the 1879 Encyclopaedia Britannica.
18 Dade (1913), i.19. This is a collection of farm histories, providing a valuable descrip-tion of German agriculture in the period.
Agricultural Productivity, Labour Surplus, Migration 227
![Page 237: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/237.jpg)
to France, some way below those in the leading countries: Britain and
Denmark (yields were also high in Belgium and the Netherlands). From
the late 1850s onwards, German yields increased faster than in other
countries, so that they caught up with the leading group, and overtook
Britain. The German performance was even more impressive when the
fact that Britain and Denmark were reducing their wheat acreage, while
it was expanding in Germany, is taken into account.
The Wgure also shows average yields for east-Elbian Germany, and
for the rest of Germany, for the years for which this is available. The
convergence of yields in eastern Germany with those in the rest of the
country is evident. Part of the improvement in German yields was due
to a more even performance within Germany itself.
Wheat provides an interesting comparison because it was a widely
grown crop. But rye was the more important crop in Germany. Figure 7.4
repeats the analysis for rye. Britain has to be omitted as the rye crop was
insigniWcant. This again shows that German yields caught up with those in
the leading country, Denmark in this case. German yields showed no
upwards trend until 1860. After that they started to rise, but slowly. There
1908189818881878186818581848
2log scale
1
0.9
0.8
Germany
France
Denmark
Germany, east
Germany, rest
Fig. 7.4. Rye yields 1848–1913, 1,000 kg per hectare, rolling five-year averages
Source: As Figure 7.3.
228 Agricultural Productivity, Labour Surplus, Migration
![Page 238: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/238.jpg)
was a marked acceleration after 1890. Again, this was accompanied by a
strong process of convergence within Germany, although a gap between
the east and the rest remained at the end of the period.
Another basic indicator of agricultural conditions in the nineteenth
century was livestock intensity. Even though artiWcial fertilizers were
coming into use, animal manure was still the main source of plant
nutrients—nitrogenous fertilizers in particular were not much used. So,
high livestock intensity would not only be associated with a high level
of production of livestock produce, but it would also have a beneWcial
eVect on arable yields.
International comparison of livestock numbers (Table 7.5) shows that
overall livestock intensity rose by 39% in Germany between 1873 and
1913, which brought the German Wgure up to the British level. German
livestock intensity was some way above France, but well below the
Danish Wgure. Danish agriculture had reorientated away from cereal
exports and was now concentrating on the export of livestock produce.
The composition of the German livestock Wgures changed. Sheep
numbers fell sharply, and there was a large increase in pig numbers.
Table 7.5. Livestock numbers (in thousands) and livestock intensity
Horses Cattle Pigs Sheep LU/100ha
Germany
1873 3352 15777 7124 24999 45.0
1913 4558 20994 25659 5521 62.5
% change þ36.0 þ33.1 þ260.2 �77.9 þ38.9
France
1882 2838 12997 7147 23809 37.3
1913 3222 14788 7036 16131 39.9
% change þ13.5 þ13.8 �1.6 �32.2 þ6.9
Denmark
1871 325 1239 442 1842 54.6
1914 487 2463 2467 515 105.1
% change þ49.8 þ98.8 þ458.1 �72.0 þ92.4
Britain
1872 1258 5625 2772 27922 57.8
1912 1441 7026 2656 25058 64.9
% change þ14.5 þ24.9 �4.2 �10.3 þ12.3
Source: Data from Mitchell (1981), conversion to Livestock Units (LU) uses weights fromWagner (1896).
Agricultural Productivity, Labour Surplus, Migration 229
![Page 239: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/239.jpg)
The livestock structure in 1913 resembled the Danish pattern, rather
than that of France or Britain, where sheep numbers remained high.
The more intensive livestock systems of Germany and Denmark had an
eVect on the value of manure as fertilizer. The manure of in-housed
cattle and pigs would be stored and spread on the Welds in the spring,
which reduced nutrient loss through leaching; sheep by contrast would
be left out on the Welds all year, which meant that part of the nutrients
would have leached away by the time the crops were Wnally planted.19
Going from these comparisons of livestock numbers and yields to
comparisons of productivity involves much greater problems of com-
parability. It is not a job for the faint-hearted. One of the more ambi-
tious exercises is Bairoch’s series of estimates of calorie production per
male agricultural worker reaching back to 1800. There are two problems
here. Female labour is ignored, so genuine variations in the male–
female ratio by country will aVect the results, and the aggregation of
production by caloriWc content can be criticized on the grounds that
calorie content is a poor guide to economic value. The calorie content
of a potato is greater than that of a bunch of grapes but the price of the
latter is much higher.
These objections apply most strongly to Mediterranean agriculture,
whose products—fruit, olives, wines—often have a high value relative
to their calorie content. Productivity in these countries will be under-
estimated. Conversely, it will be overestimated in countries which spe-
cialize in root crops. The calorie output per hectare of potatoes is much
higher than the equivalent yield of wheat, yet the value of output is not
so diVerent.
However, for all these problems Bairoch’s estimates are the only ones
available for a substantial group of countries in the early nineteenth
century, and some relatively crude aggregation is probably the only
option for this period. The results are given in Figure 7.5.
According to Bairoch, agricultural productivity in the United King-
dom was substantially greater than in any other European country
around 1800. It was a third higher than the Dutch level, the nearest at
that time. In the 1860s Britain was overtaken by Denmark, and around
1890 by Germany. Germany rose from Wfth position, level with France,
to be second only to Denmark in 1910. Of the other countries, relative
losers included Austria-Hungary, France, and (perhaps more surpris-
ingly) the Netherlands; the other relative gainer was Sweden.
19 Nutrient loss through leaching is now a major agricultural concern, on environmen-tal grounds, but it was recognized in the nineteenth century: see Voelcker (1883) on this.
230 Agricultural Productivity, Labour Surplus, Migration
![Page 240: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/240.jpg)
Some of the German improvement is due to the ‘root-crops’ bias.
Germany had a proportionately larger acreage of potatoes and sugar
beet, and this would produce an overestimate of the value of German
agricultural output.20 The value of French ‘Mediterranean’ produce
would be similarly underestimated. The use of male labour as the de-
nominator also favours Germany. In the early twentieth century around
half the German agricultural labour force was female; only 10% of the
British was (as recorded). Thus, dividing by the male labour force raises
estimates of German productivity in 1910 by 73% relative to Britain. It
also raises it against other countries: by 70% against the Netherlands,
by 26% against Denmark, and 16% against France, for example.21
20 In 1913 19.7% of the German arable area was down to root crops: three times theFrench proportion, and eight times the British, Perkins (1981), 80; this had risen from14.8% in 1883 and 2.3% in 1800, Schremmer (1988), 50.
21 Calculated from estimates in O’Brien and Prados de la Escosura (1992).
18004
5
6
7
89
10
20
30
50
40
1810 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910
Sweden
Switzerland
Netherlands
Britain
Denmark Germany
Belgium
France
Austria-Hungary
Fig. 7.5. Bairoch’s estimates of agricultural productivity 1800–1910 (calories
produced per male occupied in agriculture)
Source: From Bairoch (1989).
Agricultural Productivity, Labour Surplus, Migration 231
![Page 241: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/241.jpg)
So, if Bairoch’s estimates produce a distorted picture how can they be
improved? The Wrst step is to express output in some unit of value, to deal
with the problems created by the calorie measure. The output of diVerent
commodities in the various countries should be valued at uniform prices,
and combined as an index of production measured in some common unit
of account. Generally the unit of account is some monetary measure of
given purchasing power (a purchasing power parity or PPP). However,
one author (Van Zanden) has chosen to express all prices in terms of
wheat equivalents (thus a cow might be worth the equivalent of four tons
of wheat, a ton of sugar beet half a ton of wheat, etc.).
Table 7.6 gives a range of estimates from diVerent sources, indicating
the unit of account, and the productivity measure, which are used by
each study. The table concentrates on the comparison between Britain
and Germany. This table shows once again that the use of male labour
as the denominator aVects all these results. The use of value added as
the numerator is also important. This is because German agriculture
was more ‘self-suYcient’, it relied less on inputs from other sectors. So
‘value added’ is a higher share of total output.
The fact that British agriculture made much greater use of employed
labour, which was predominantly male, is one explanation of the diVer-
ent rates of female participation between the two countries. However,
Table 7.6. German agricultural productivity compared to the United
Kingdom, 1910–13
Gross output per head Value added per head
Both sexes Men only Both sexes Men only
O’Brien/Prados de
la Escosura (PPP)
56/58 97/100 77/79 134/136
Bairoch (calories) 127
Van Zanden
(wheat equivalent)
77 133
Fremdling (PPP) 42 51
Broadberry (PPP) 67
Source: This is a reworking of a similar table in O’Brien and Prados de la Escosura(1992): their estimates are combined in a range (German weights and British weights);Broadberry’s estimate is added, from Broadberry (1997); and Bairoch’s estimate is revisedfrom 106 (which came from an earlier study) to 127, from Bairoch (1989). Quite whyBairoch made this revision is not clear.
232 Agricultural Productivity, Labour Surplus, Migration
![Page 242: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/242.jpg)
the diVerence between Germany and the Netherlands was almost as
great, and this is more diYcult to explain. Institutional diVerences were
less important. So it may be that the German occupational censuses
were rather more diligent in the recording of the labour contributed by
wives and other female family members than surveys in other countries.
If this is right, then comparisons which use both sexes may understate
productivity in German agriculture.
Looking at the diVerent estimates, those of Fremdling use a small
number of prices to establish the PPP: just the prices for wheat and beef
for the whole agricultural sector. These are probably unfavourable to the
German side (potatoes, pork, and rye might have produced a diVerent
Wgure). The estimates using purchasing power parity calculations and
value added per head are the ones which should be preferred on a priori
grounds. Discounting the Fremdling estimates, this leaves those by
Broadberry and O’Brien/Prados de la Escosura. There should probably
be some allowance for the superior coverage of female labour
in Germany. This suggests that the most likely level of productivity in
German agriculture was around 75–80% of the British level in 1910–13.
As far as other countries are concerned, estimates by O’Brien and
Prados de la Escosura (Table 7.7) show Germany to be ahead of France
(on most measures) but behind Denmark. The comparison with the
Netherlands is heavily aVected by the inclusion or exclusion of female
labour. Allowing for some under-recording in the Netherlands would
put productivity in Germany ahead of the Dutch level in 1910.
Between 1890 and 1910 Germany seems to have made gains relative to
Britain, France, and the Netherlands, but not relative to Denmark (where
the increase was exceptionally high). Van Zanden has constructed an
Table 7.7. Production and productivity in selected countries, as
compared to the British level, 1890 and 1910 (UK ¼ 100)
Output per worker Output per male
worker
Output per hectare
1890 1910 1890 1910 1890 1910
Netherlands 82 89 89 92 192 237
Denmark 44 82 82 147 140 202
France 52 72 72 82 128 136
Germany 63 89 89 118 148 205
Agricultural Productivity, Labour Surplus, Migration 233
![Page 243: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/243.jpg)
index of total productivity (combining land, labour, and livestock). This
shows that total productivity rose by 1.53% per annum in Germany in
1870–1910, which is just ahead of Denmark (1.31%) and substantially
more than in the Netherlands (0.82%), France (0.46%), and Britain
(0.19%).22 Broadberry’s estimates also show German productivity rising
from 55.7% of the British level in 1871 to 67.3% in 1911.23
So a tentative conclusion is that productivity growth in German
agriculture in 1870–1913 was higher than in most other countries with
the possible exception of Denmark. This was partly due to the use of
techniques already established in other countries. One study has oVered
the following view of the increase in German production per hectare
1870–1913: it was 50% due to increased fertilizer use (including both
artiWcial fertilizers and livestock manure), 15% due to improved breed-
ing, and 35% due to improved husbandry and reduced losses, including
harvest losses.24 Breeding was mainly new technology, although some
crops and improved breeds came from foreign countries. Husbandry
improvements, on the other hand, often involved the use of machinery,
and livestock systems, which had been largely developed outside
Germany: drills, reapers, the stall-feeding of cattle. The major German
contribution was the development of a husbandry system for lighter
soils: the Lupitz system.25
This section will conclude with an attempt to compare crop nutrient
supply from artiWcial fertilizers to that available from livestock manure.
The use of artiWcial fertilizers was clearly new technology, and an area
where Germany was an acknowledged leader. The Wgures presented in
Table 7.8 make use of Wgures produced by Schremmer for the nutrients
supplied by artiWcial fertilizers. He also provided estimates of the total
tonnage of animal manure, but not the nutrient content. The table gives
the estimates of nutrient value, using Wgures for the analysis of animal
manure given by the German scientist, Augustus Voelcker, who became
22 Van Zanden (1991).23 Broadberry (1997), 18; Broadberry makes use of the 1937 benchmark comparisons
made by Rostas (1948); these incorporate a crude, but possibly justiWable, adjustment ofthe German agricultural labour force (removing three out of four wives); Broadberry alsouses HoVmann’s estimates of the agricultural labour force 1870–1913, thus incorporating(as does everyone else who uses HoVmann in this respect) HoVmann’s upwards adjust-ment of the German agricultural labour force in the 1870s and 1880s to allow for asupposed under-recording of female labour in the census of 1882. This has the eVect ofreducing estimated productivity growth in German agriculture.
24 Estimates given in Muller (1895).25 Brouwer (1957) and Dyer (1896).
234 Agricultural Productivity, Labour Surplus, Migration
![Page 244: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/244.jpg)
consulting chemist to the Royal Agricultural Society of England and
Wales.
These Wgures should be regarded as approximations. Schremmer’s
Wgures are themselves estimates. The nutrient value of manure depends
on factors such as storage practice, time of application, losses through
leaching, and the way that animals are fed. Margins of error are likely
to be high.
Keeping these reservations in mind, the estimates show that, by the
years just before the First World War, artiWcial fertilizers were supply-
ing around 40% of the potash and phosphate applied, but only 16% of
the nitrogen (an additional source of nitrogen came from the use of
clover and other nitrogen-Wxing legumes in the rotation). But they did
represent a large share of the increase since 1878–80: over 50% for
potash and phosphate, and 23% for nitrogen. Total nutrient applica-
tions rose twice as fast after 1878/80 as in the period 1800–1878/80,
and this acceleration was entirely due to the use of artiWcial fertilizers.So, the contribution of new technology was, in this area, a major factor.
4. Regional Patterns of Productiv ity Growth
A problem which has restricted discussion of agricultural developments
in Germany has been the lack of estimates of regional productivity
levels. HoVmann’s accounts provide national Wgures, but no attempt
was made to provide a regional breakdown. Thus, it was not possible to
Table 7.8. Plant nutrient supply, 1800–1914, average annual
applications divided by the total agricultural area
1800 1878–80 1891–3 1898–1900 1911–14
A. Estimates of nutrients available from animal manure
Nitrogen (kg/ha) 5 14 19 33
Phosphate (kg/ha) 4 12 16 28
Potash (kg/ha) 4 11 15 26
B. Nutrients available from artificial fertilizers
Nitrogen (kg/ha) 0.7 2.2 6.4
Phosphate (kg/ha) 1.6 10.3 18.9
Potash (kg/ha) 0.8 3.1 16.7
Source: Schremmer (1988) and Voelcker (1883).
Agricultural Productivity, Labour Surplus, Migration 235
![Page 245: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/245.jpg)
consider issues such as whether the problems of eastern agriculture
were due to poor productivity rather than low prices. The best way to
create Wgures for regional productivity levels is to construct agricultural
accounts for each region: dividing estimates of total net value added by
the labour force (in full-time labour units to allow for the eVect of part-
time employment) produces Wgures for labour productivity.
It was considered that the provision of regional accounts would be of
value, not only to this study, but also to others working on German
agriculture, so three sets were prepared, for 1880–4, 1893–7, and
1905–9. The details of the calculations have been published separately.
The results are given in Table 7.9, which shows estimates of labour
productivity in German agriculture for 21 regional units for three Wve-
year periods: 1880–4, 1893–7, and 1905–9. The table also gives the
Table 7.9. Productivity in German agriculture: estimates from regional
agricultural accounts in 1913 prices, 95% confidence intervals in brackets
Net Value Added per FLU (Full-time Labour Unit) Annual % rates
of growth
1880–4
(Marks)
1893–7
(Marks)
1905–9
(Marks)
1880/4–1905/9
East Prussia 480 (454–506) 745 (703–786) 934 (877–992) 2.71 (2.87–2.54)
West Prussia 648 (615–680) 923 (879–967) 1078 (1022–1134) 2.07 (2.18–1.95)
Berlin/Brandenburg 731 (697–764) 1001 (957–1045) 1222 (1163–1280) 2.08 (2.19–1.97)
Pomerania 852 (811–893) 1175 (1121–1229) 1433 (1363–1503) 2.11 (2.22–1.99)
Posen 636 (607–664) 891 (850–932) 1179 (1124–1234) 2.51 (2.63–2.38)
Silesia 550 (521–579) 765 (728–802) 960 (911–1009) 2.26 (2.38–2.13)
Pr. Saxony/Anhalt 1089 (1046–1131) 1356 (1304–1407) 1388 (1330–1445) 0.98 (1.03–0.93)
Schleswig-Holstein 1145 (1072–1218) 1323 (1243–1403) 1709 (1600–1819) 1.62 (1.74–1.50)
Hannover/Old./Bruns. 779 (737–821) 1075 (1023–1126) 1136 (1075–1197) 1.53 (1.62–1.43)
Westphalia 579 (544–614) 865 (822–908) 834 (787–881) 1.48 (1.58–1.37)
Hesse-Nassau 524 (497–552) 798 (760–837) 769 (727–811) 1.55 (1.65–1.45)
Rhineland 535 (504–565) 761 (723–799) 757 (713–800) 1.41 (1.50–1.31)
Bavaria excl. Pfalz 510 (484–536) 691 (654–729) 667 (623–712) 1.08 (1.16–1.00)
Pfalz 480 (458–501) 791 (756–826) 699 (664–735) 1.53 (1.61–1.44)
Saxony 789 (744–834) 1052 (999–1105) 1395 (1316–1474) 2.32 (2.46–2.17)
Wurttemberg/Hoh. 594 (565–622) 641 (608–674) 661 (619–704) 0.44 (0.47–0.41)
Baden 517 (492–543) 631 (601–661) 635 (599–671) 0.83 (0.88–0.77)
Hesse 685 (654–716) 949 (908–990) 1016 (965–1067) 1.60 (1.68–1.5)
Mecklenburg 1267 (1211–1323) 1442 (1380–1505) 1716 (1632–1801) 1.23 (1.3–1.16)
Thuringia 750 (717–783) 1029 (984–1075) 1088 (1034–1142) 1.50 (1.59–1.42)
Alsace-Lorraine 596 (571–622) 609 (580–638) 676 (639–713) 0.51 (0.54–0.47)
Germany 672 855 982 1.53
236 Agricultural Productivity, Labour Surplus, Migration
![Page 246: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/246.jpg)
annual rate of increase 1880/4 to 1905/9. As the calculations involve a
degree of uncertainty, conWdence intervals have been estimated both for
the levels of productivity and for the rates of growth. The regional
units are those deWned in Appendix 3A.26
The Wrst point which emerges clearly from Table 7.9 is the high rates
of productivity growth in eastern Germany. Productivity growth rates
of over 2.0% per annum are impressive by any standards. While this
does not prove that there had been underutilized resources in these
provinces, this is certainly consistent with this explanation.
In Chapter 4 regressions were presented which showed that agricul-
tural productivity growth had a positive eVect on migration. The table
shows that the reason for this is the high rate of increase in the east, in
regions which also had high out-migration. There is a contrast with
southern and south-western Germany, which was also agricultural, but
where productivity growth was slower. Out-migration was also less.
This is the eVect which is picked up by the regression results.
If there had been a labour surplus in eastern agriculture then it could
have caused both high out-migration, as some of those who were under-
employed left, and high productivity growth, as others became more
fully employed. A relationship between high productivity growth and
high out-migration would be one consequence of this.
Looking at the maps of regional productivity (Figure 7.6), it is evi-
dent that, in the Wrst period, there was a small group of central regions
with high productivity: Schleswig-Holstein, Mecklenburg, and Prussian
Saxony. By 1905–9 these regions were still amongst the leaders, but the
gap with regions to the west and east had closed. The south and south-
west, on the other hand, were lagging.
Table 7.10 provides a summary of the changes in the relative position
of the east-Elbian provinces. Their share in the total labour force was
relatively static, as was the share in livestock production. The factor
which increased the share of total value added, and thus produced faster
than average productivity growth, was the gain in the share of total
production of major roots and cereals (sugar beet, potatoes, wheat,
barley, spelt, oats, and rye). It was the gains made by the arable sector
which drove the process of convergence.
Table 7.11 provides a decomposition of the increase in the production
of major roots and cereals, comparing the east-Elbian provinces with the
rest of Germany. The major part of the shift in the balance of production
26 Details of the methods used to construct these accounts are given in Grant (2002a).
Agricultural Productivity, Labour Surplus, Migration 237
![Page 247: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/247.jpg)
1,300−1,400
1,200−1,300
1,400−1,500
1,600−1,700 1,000−1,100
1,100−1,200
800−900
700-800
>1,700
600−700
1,500−1,600 900−1,000
Productivity in 1880-4Net valve added per full-time labour unit in 1913 Marks
Productivity in 1905-9Net valve added per full-time labour unit in 1913 Marks
1,000−1,100
900−1,000
1,100−1,200
1,200−1,300 700−800
800−900
600−700
500−600
>1,300
<500
Fig. 7.6. Agricultural productivity by region
238 Agricultural Productivity, Labour Surplus, Migration
![Page 248: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/248.jpg)
in favour of the east came about as a result of a faster increase in yields.
There was a contribution from an increase in areas, and another from a
composition eVect (the substitution of higher yielding, or higher value,
crops for others which were less productive or less valuable). But the
main factor was the strong tendency for yields in eastern Germany to
converge on the higher levels of central and western Germany.
What was the cause of this convergence? Clapham suggests that sugar
beet cultivation played a crucial role in making producers aware of the
importance of artiWcial fertilizers. One source which conWrms this is a
survey of fertilizer production and use in Prussia published in the 1887
Landwirtschaftliche Jahrbucher.27 This brought together reports from six-
teen regional research stations. It showed that, typically, fertilizer use
was two or three times as high for sugar beet as for other crops. It also
27 Thiel (1887).
Table 7.10 East-Elbian share in total production and value added
1880–4 1893–7 1905–9
Production of:
Major roots and cereals 33.5 37.7 43.3
Other crops 31.6 27.7 27.8
Livestock produce 33.0 32.2 33.3
Total net value added 32.8 33.7 37.6
Labour force in FLU 32.7 32.7 32.2
Source: Calculated from accounts given in Grant (2002a).
Table 7.11. Decomposition of change in production of major roots
and cereals 1880/4–1905/9
East-Elbian
Germany
Rest of
Germany
Total % increase in production 118.5 48.0
Increase in production per hectare 95.5 39.7
Increase in areas 11.8 5.9
Decomposition of increase in production per hectare:
Composition effect 2.9 0.9
Weighted increase in yields 89.8 38.1
Note: Calculations use 1905–9 prices.
Agricultural Productivity, Labour Surplus, Migration 239
![Page 249: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/249.jpg)
showed that most fertilizer was supplied from Wrms within the region.
This suggests that sugar beet may have stimulated the use of fertilizers
in other ways besides the ‘demonstration eVect’ mentioned by Clapham.
The introduction of sugar beet into a region would raise demand, and
this might then lead to the construction of additional fertilizer factories.
Having a fertilizer factory nearby would then reduce transport costs and
so lead to increased use on other crops.
Sugar beet cultivation expanded faster in the east. Sugar beet rose
from 0.6% of the eastern arable area in 1880–4 to 2.2% in 1905–9,
which was a larger increase than in the rest of Germany (1.7% to 2.5%).
So the sugar beet eVect is one which promoted convergence. But the
pattern of convergence was much wider than this. Sugar beet cultivation
was limited to areas with suitable soil and climate (Figure 7.7). Initially,
the centre of cultivation was Prussian Saxony, especially the Magde-
burg region. From there it spread west and east. But certain regions
8−10
6−8
4−6
3−4
2−3
1.5−2
1−1.5
0.5−1
<0.5
>10
Fig. 7.7. The sugar beet regions: sugar beet as a percentage of the total
arable area, 1907
Source: Data from agricultural part of the 1907 occupational census, SDR n.f. 212.
240 Agricultural Productivity, Labour Surplus, Migration
![Page 250: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/250.jpg)
never had much. These included the regions of southern Germany,
which had relatively low productivity, but also some more successful
regions: Westphalia, Schleswig-Holstein, and the Kingdom of Saxony.
East Prussia, which recorded the fastest productivity increase, had little
sugar beet cultivation.
Arable production in Eastern Germany caught up from a position
well below the national average in 1880–4. Table 7.12 gives the results
of some calculations based on the statistical analysis of yield variations,
using statistical techniques similar to those used elsewhere in this study.
These show that the spread of sugar beet eVect did indeed promote
convergence. But only 20% of predicted convergence is attributable to
this eVect. There was a more general process of convergence which was
not driven by the spread of sugar beet.28 Sugar beet cultivation was
only part of the catch-up story. Improved techniques were coming into
use in regions outside the sugar beet areas as well.
Statistical analysis can be used to examine whether the presence of
larger estates contributed to increases in yields. Large estates might
have been better placed to take advantage of the new technologies:
more aware of scientiWc advances, having easier access to sources of
Wnance. There is documentary evidence which suggests that larger
farms were more likely to use artiWcial fertilizers. For example, a report
on conditions in a Gutsbezirk in Kreis Stolp in 1890–1 observed: ‘In this
area, as in the whole region, it is only the large farms and estates that
make use of artiWcial fertilizers and concentrated livestock feed.’29
Table 7.12. Explaining the convergence of arable yields, east-Elbian
Germany compared to the rest of Germany, 1880–4 to 1905–9
Value of arable production
per hectare
(average 1880–4 ¼ 100):
Convergence
effect
Sugar beet
effect
Predicted
gain
1880–4 1905–9
East 78.7 172.0 þ27.6% þ24.0% þ103.9
Rest 116.7 172.8 �21.7% þ12.5% þ43.0
28 This calculation uses the results of regressions given in Grant (2002a).29 Landwirtschaftliche Jahrbucher, xix, erg. 4 (1891). Every year this publication in-
cluded a series of reports on conditions in various parts of Prussia, which are a usefulsource of information on the state of agriculture.
Agricultural Productivity, Labour Surplus, Migration 241
![Page 251: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/251.jpg)
However, the analysis of yields at the Kreis level provides a more
nuanced view. The larger farms and estates maintained an advantage in
wheat yields, which was a crop grown largely for sale in urban markets,
and thus better suited to the more market-orientated enterprises. But
when rye yields were examined it was found that the advantage which
the larger holdings had had in the 1870s had largely disappeared by
1900; the smaller and medium-sized farms had caught up.30 The new
techniques had become more widely diVused.
Even if the larger holdings had an advantage when the use of artiWcial
fertilizers is considered, this was compensated by the greater livestock
intensity on smaller holdings, which meant that these holdings were
deriving more nutrients from animal sources and so had less need of
artiWcial fertilizers. The livestock censuses show that the value of live-
stock per hectare was higher on smaller farms. They also show that the
total value of livestock kept was rising faster on smaller holdings. Be-
tween 1882 and 1895 the total value of livestock on all German holdings
of less than 20 hectares rose by 17.7%, but it only rose by 9.5% on
holdings of over 100 hectares.31
So, the emphasis on the leading role of larger farms and estates may be
misplaced. Even in the east there were many small and medium-sized
farms. There were three east-Elbian provinces where holdings of under
100 hectares accounted for over 60% of the total agricultural area: East
Prussia (60.5% in 1895), Brandenburg (64.8%), and Silesia (66.1%).
Technological change, leading to increased yields, on these farms was an
important component in the success of eastern agriculture in the period.
The main reason for the improved performance of the arable sector
east of the Elbe was that the new technology, in particular the use of
artiWcial fertilizers, was particularly valuable in areas where the natural
level of soil fertility was low, or areas with free-draining sandy soils. On
the sandy soils of eastern Germany increased nutrient supply was par-
ticularly beneWcial. These were free-draining, ‘hungry’ soils, where the
rapid leaching of nutrients meant that continuous fresh applications
were needed to maintain soil fertility. Table 7.13 gives the results of a
survey of soil quality in the original Prussian provinces in the 1860s.
The diVerence between Prussia east and west of the Elbe is very notice-
able. In the west the predominant soil type was a heavier loam, with a
higher clay content, and, in consequence, much less free-draining. This
30 Details of the statistical analysis are given in Grant (2002a); there is a further discus-sion of this issue in Grant (2002e).
31 Rauchberg (1900), 590.
242 Agricultural Productivity, Labour Surplus, Migration
![Page 252: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/252.jpg)
Table 7.13. Soil quality in the Prussian Regierungsbezirke, from a
survey in the 1860s
Loam Sandy loam Sand Moor Grsrt
I. East of the Elbe
Konigsberg 22.4 57.8 16.8 2.9 29.6
Allenstein 6.2 49.1 39.0 5.7 15.2
Gumbinnen 22.3 50.6 19.0 8.1 24.1
Danzig 27.6 38.5 25.5 8.3 24.7
Marienwerder 14.5 33.6 48.4 3.5 23.3
Potsdam 11.6 44.8 33.3 10.3 34.6
Frankfurt 12.7 27.7 50.6 9.1 33.9
Stettin 4.9 52.0 30.2 12.9 35.7
Koslin 8.9 36.2 47.1 7.8 21.7
Stralsund 1.4 70.3 15.4 12.9 58.9
Posen 11.6 49.9 32.8 5.7 25.2
Bromberg 7.7 42.5 39.2 10.6 26.0
Breslau 48.3 24.6 25.1 2.0 53.8
Liegnitz 32.0 28.8 36.3 2.9 42.6
Oppeln 31.2 33.5 33.5 1.8 36.6
Average 17.9 40.9 34.7 6.5 31.3
II. Other Prussian provinces
Magdeburg 37.0 20.1 35.7 7.2 61.4
Merseburg 66.2 11.4 22.1 0.4 89.7
Erfurt 69.8 25.4 4.7 0.1 60.7
Munster 29.2 16.5 44.5 9.8 53.3
Minden 55.0 12.3 29.2 3.5 59.1
Arnsberg 93.3 3.6 3.1 0.0 52.9
Koblenz 90.1 9.9 0.0 0.0 53.5
Dusseldorf 41.4 38.4 16.0 4.2 96.2
Koln 68.4 23.3 7.5 0.8 88.7
Trier 65.6 22.3 12.0 0.1 38.0
Aachen 71.5 18.2 5.4 4.9 78.4
Average 60.8 17.3 18.8 3.1 65.3
Note: Percentage distribution of the land area (excluding areas under water) by soil type,together with the average level of Grundsteuerreinertrag (Grsrt) for cultivated ground, inMarks per Hectare.Source: Data from Meitzen (1868).
accounted for over 90% of the area in Koblenz and Arnsberg. In the east
the predominant types were the lighter ‘sand’ and ‘sandy loam’. Outside
Silesia, heavier soils were comparatively rare.
Agricultural Productivity, Labour Surplus, Migration 243
![Page 253: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/253.jpg)
The land tax assessment (Grundsteuerreinertrag) was closely linked to
soil quality. The Wnal column gives the average assessment for each
Regierungsbezirk. The level east of the Elbe was less than half the aver-
age for the remaining Prussian provinces, conWrming the poorer quality
of the eastern soils.
Writers on agricultural matters, inside Germany and outside, con-
sidered that the major achievement of German agriculture in the period
was the raising of yields on the lighter, sandy soils. The ‘Lupitz
system’, developed by the Mecklenburg farmer A. Schultz-Lupitz, at-
tracted the attention of English agriculturalists.32 There was govern-
ment support for research into yields on lighter soils. Work at
experimental farms, and Weld trials conducted by the local Land-
wirtschaftsgesellchaften, demonstrated the value of fertilizers to local
farmers. The results were some spectacular gains. Average rye yields in
Kreis Uelzen in the Luneburg Heath rose 205% between 1882–4 and
1910–14, from a position 29% below the national average to one 15%
above.33 This meant that moorland areas, and areas which had ex-
tremely light sandy soils, which had previously been left uncultivated,
could be brought into production.
ScientiWc research in the early part of the century, in particular the
discoveries of Justus von Liebig, had led to a greater understanding of
plant nutrient requirements and the possibilities of providing these
through artiWcial fertilizers. There was, however, no simple relationship
between the discoveries of Liebig and their application to the poorer soils
of northern Germany. To begin with, practical, and cost-eVective, means
of supplying soluble nutrients to plants had to be found. The Wrst suc-
cessful ‘artiWcial fertilizer’, superphosphate produced by the application
of sulphuric acid to mined phosphate, was developed by Gilbert and
Lawes in England, making use of Liebig’s suggestion that the value of
bonemeal could be improved by treatment with sulphuric acid.34
32 Brouwer (1957) describes the work of Schultz-Lupitz; his system was introduced toEnglish readers in an article in the 1896 Journal of the Royal Agricultural Society ofEngland and Wales, Dyer (1896).
33 Figures from Schneppe (1960), who also gives details of government-sponsoredresearch on the light soils of the Luneburg Heath. It was a general complaint of Germanagriculturalists that the quality of German soil was poor. See, for example, the remark byDr H. Gerlich at a meeting of the Royal Statistical Society in 1899, Journal of the RoyalStatistical Society, 62/4 (1899). Sandy soils have advantages and disadvantages, highfertilizer requirements being undeniably one of the latter. The author of this study farmson soils of this type in central Sussex. See also Thiel (1887) for details of fertilizer use inthe Prussian regions.
34 Frauendorfer (1956). There is a tendency in writings on nineteenth-century agricul-tural technology to focus on developments within the boundaries of the state to which the
244 Agricultural Productivity, Labour Surplus, Migration
![Page 254: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/254.jpg)
But then these ideas and new products had to be incorporated in a
practical farming system, and this took years of experimentation.
Schultz-Lupitz, the son of a pharmacist whose university education
made him aware of the importance of Liebig’s discoveries, was inspired
by this to buy a 250-hectare farm in 1855. But it was only in 1879–80
that his experiments with diVerent fertilizers and rotations produced a
system which consistently raised yields. The value of Liebig’s discover-
ies lay in the conWdence they gave to innovative farmers, and to re-
searchers at experimental stations, that ways could be found to raise
fertility on poorer soils. Without this conWdence, Schultz-Lupitz would
never have engaged in twenty-Wve years of initially fruitless work.35
A second major advance in the period was improvements in yields
through breeding. Again, German successes in this area attracted atten-
tion in England.36 New strains of wheat and rye were developed. But
perhaps the greatest success was the raising of the yield and sugar con-
tent of sugar beet. This last was increased from 8.5% in 1870 to 16% in
1914, an important contribution to the success of this industry.37
New productive knowledge was an important part of the success of
German agriculture in 1870–1913. Traditional agriculture was ‘trans-
formed’: new techniques were introduced, as were new crops and new
inputs. But, equally, there was a contribution from old technology as
well. Much of the increase in total plant nutrients came from increased
livestock numbers, deWnitely an old technology. The convergence of
yields within Germany may have been due to new technology, but the
rise in average German yields relative to other European countries
owed a lot to rising livestock intensity.
One question which arises is how this relatively favourable view of
German agriculture can be reconciled with the evidence of high debt
levels and Zwangsversteigerungen (forced sales of agricultural property).38
writer belongs, and which were therefore published in the writer’s own language: oneexample is Fussell (1962). Frauendorfer’s article is an exception. It gives full weight todevelopments in Britain, France, and Germany, and is good on interactions betweenresearchers in these countries. The conclusion which emerges is that, while France andGermany led in basic scientiWc research, British research on the practical application ofthese ideas was of considerable importance. Britain did not lag in the development offertilizer technology.
35 Brouwer (1957).36 Evershed (1884).37 Muller (1985).38 ZdKSB (1905), p. xlii; Zwangsversteigerungen were reported annually in the ZdKSB
in the 1880s and 1890s.
Agricultural Productivity, Labour Surplus, Migration 245
![Page 255: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/255.jpg)
But a distinction can be drawn between crises of agriculture, which aVect
the whole sector badly enough to bring about a fall in the total area
under cultivation, and crises which aVect individual farmers who have
over-borrowed or paid too much for assets whose value has subsequently
fallen. In crises of the latter type, land does not go out of production,
as the assets of bankrupt farmers are bought up by other farmers, who
are less indebted or more eYcient. The ‘crisis’ of German agriculture
in the 1890s was of this latter type. There was a genuine agricultural
crisis in Britain in the 1880s and 1890s, and again in the 1930s. In
these periods land did go out of production. This is the acid test. Un-
doubtedly, some estate owners did conform to the stereotype found in
many works on east-Elbian agriculture, with an addiction to ‘ruinous,
late-feudal, conspicuous consumption’ which led to increased indebted-
ness.39 But east-Elbian estates considered as a whole did not deserve this
criticism.
Germany 1870–1913 is an excellent example of Schultz’s view of
agricultural transformation. Germany is also a good example, one might
almost say a model, of how this can be done: improved agricultural
education, the setting up of research stations, investments in scientiWc
research, trial plots and other demonstrations organized through local
farming societies, good interaction between researchers and innovative
farmers. In all these areas German agriculture was a leader in European
agriculture in 1870–1913.But the consequence of this was that the declining share of the home
market available to German farmers, after imports had been deducted,
could be supplied by a smaller labour force. So, this also contributed to
out-migration. And the more modern farming techniques reduced the
variability in crop yields, which meant that institutions which provided
insurance against crop failure were less necessary. So, improved tech-
nology led to institutional change as well.40
5. Rural Underemployment , Labour-Saving
Technology , and M igration
There is no evident labour-saving bias to the technological changes dis-
cussed in the last section. Indeed, many of the changes in German
39 Puhle (1972), 16.40 The decline in the variability of crop yields is discussed in Grant (2002c).
246 Agricultural Productivity, Labour Surplus, Migration
![Page 256: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/256.jpg)
agricultural practice should have raised labour requirements: the area
under root crops was rising and these required more labour, livestock
numbers were rising, and there was a shift from sheep (less labour inten-
sive) to stall-fed cattle and pigs.41 What was impressive was the ability of
German agriculture to deliver a large increase in food production,
1870–1913, with a relatively small increase in the labour force. This is
consistent with the view that there was initial underemployment.
There was one area in which labour-saving technology was intro-
duced, and that was equipment for harvesting and threshing. Some
authors, including Max Weber, have argued that the introduction of
steam-powered threshing machines reduced out-of-season labour re-
quirements and led to out-migration. However, the statistical analysis
showed that steam-powered threshing machines were not associated
with reduced agricultural employment. This section will look brieXy at
the reasons for this Wnding.When considering the eVect of machinery it is important to exclude
the eVect of farm size. Larger farms do use more machinery, but, in this
period, they were also intensifying production and introducing more
labour-intensive crops. The question of the eVect of farm size on labour
requirements was the main issue examined in the previous chapter.
To exclude the farm size eVect, a standardized index of access to
steam-threshing was constructed, which used results from the 1907
census for the use of this machinery by size of farm. These results were
then weighted by the average size distribution for the whole country,
not the regional distribution. The index therefore gives the answer to
the hypothetical question, ‘What percentage of farms would have used
steam-threshing machines if the size distribution had been the same as
the national average?’ The results are shown in Figure 7.8.
There are some considerable diVerences. The range is from an index of
12 to one of 68 (100 would indicate that all farms used steam-threshing).
There was a clear regional pattern. The use of steam-threshing was
highest in central Germany, in Hesse, Brunswick, and Franconia, and in
parts of Westphalia and Hanover. It was not high in the east, using this
standardized measure.
For a study primarily concerned with migration the important point
to note from Figure 7.8 is that few of the areas identiWed as having high
41 Figures given in Muller (1985) give the labour requirement of a sugar beet rotationas 97% higher than a Norfolk four-course rotation, and 346% higher than a three-Weldsystem.
Agricultural Productivity, Labour Surplus, Migration 247
![Page 257: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/257.jpg)
use of steam-threshing are areas of high out-migration. So it is not
surprising that migration regressions, or regressions with the change in
the agricultural labour force as the dependent variable, do not Wnd that
steam-threshing displaced labour and led to migration.
By and large, there was not a major labour-saving eVect from changes
in agricultural technology in the period. The livestock sector remained
almost entirely unmechanized apart from mechanical milk separators.
All feeding was still done by hand. Most arable improvements increased
yields rather than displacing labour. Labour would only have been
displaced if the area under cultivation had fallen in consequence.
The rises in stock numbers shown in Table 7.7 should have required
a substantial amount of additional labour. The fact that this was not
required suggests that there had been underemployment. It is hard to
Wnd any improvement in technology which might have enabled the
number of pigs to increase by 260% without using more labour if
labour had been fully utilized in 1873.
10−20
20−30
30−40
40−50
50−60
60−70
Fig. 7.8. Standardized index of the use of steam-powered threshing machines,
by Regierungsbezirke, 1907 (100% use ¼ 100)
Source: Calculated from data in SDR n.f. 212.
248 Agricultural Productivity, Labour Surplus, Migration
![Page 258: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/258.jpg)
Contemporary reports speak of underemployment, particularly
amongst family members. One report, on conditions in a Gemeinde in
Silesia in 1886, mentions that farmers are seeking to rent more land to
make more use of family labour ‘which otherwise would not be fully
utilized’.42 The same report also recorded that the cost of poor relief in
the Gemeinde had risen because of the decline in cottage industry.43
This is supported by the regression results given in Chapter 6, whichshowed that the decline of the agricultural population was higher in
areas where there had been cottage industry. The decline in this source
of employment created underemployment in some areas.
A direct indicator of underemployment might be the ratio of popula-
tion to agricultural land in rural areas. If underemployment leads to
migration there could be a positive relationship between this variable and
migration. This relationship cannot be found at the Kreis level, but there
are considerable variations between Kreise in land quality and in
the availability of non-agricultural employment. Data on migration at the
level of the Gemeinde, or village, are not generally available from
the oYcial statistical sources, but a study of migration in the 1880s,
published in 1890 in the Prussian Landwirtschaftliche Jahrbucher, pro-
vided data for 720 villages in seven rural Kreise. The analysis made use of
the indirect method for estimating rates of migration by decade (compar-
ing the natural rate of increase from recorded births and deaths with the
actual increase as shown by the census returns). The econometric analy-
sis is given in Appendix 7A. It shows that population density, the rela-
tionship between the total population and the agricultural area, had a
strong eVect on migration: 30% of the recorded variation in rates of out-
migration in these rural areas could be explained by this factor alone.
Table 7.14 provides a summary of the results by showing the pre-
dicted rates of out-migration for various Gemeinden. The predicted rate
for a village in the top decile ranked by population density is 7.8 times
that of a village in the bottom decile; the predicted rate for a village in
the top quartile is 4.6 times that for one in the lowest quartile.
This simple analysis is suYciently strong and clear to make it evident
that there was overpopulation and underemployment in parts of rural
Germany in the 1870s and 1880s, and that this was a major inXuence on
migration. There was surplus labour. The validity of this aspect of the
Lewis Model is conWrmed.
42 Landwirtschaftliche Jahrbucher, xvi, erg. 2 (1887), 54. 43 Ibid. 76.
Agricultural Productivity, Labour Surplus, Migration 249
![Page 259: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/259.jpg)
It is extremely diYcult to produce an estimate of the scale of the
rural labour surplus in Germany in the period. Much depends on the
deWnitions used. Are underemployed family members treated as part of
the surplus? Is low productivity employment, in cottage industry or on
family smallholdings, regarded as a form of underemployment? Modern
development economists, using surveys of work patterns, have come up
with a wide range of results. Estimates of underemployment in develop-
ing countries have been as high as 70% (for Peru). But a more typical
result is around 20–37%. Much depends on the methods used.44
Rural underemployment is a diVerent concept from formal, measur-
able unemployment in societies which pay unemployment beneWt. It is
not static. It is aVected by changing social and technological conditions.
So it is inevitably diYcult to deWne and measure.
In nineteenth-century Germany the amount of surplus labour was
aVected by changes in the terms of trade, in particular by the fall in
imported grain prices to 1897 and the subsequent recovery, and by
changing conditions within agriculture, by institutional developments, and
by new agricultural technology. The introduction of new technology
meant that a rising urban population could be fed by a rural workforce
which was static or falling, thus making it possible for the demographic
surplus in rural areas to move out of agriculture into industry. Migration
out of agriculture was, therefore, part of a dynamic process, as agriculture
adjusted to the new conditions created by industrialization, and was in
turn transformed by the eVects of industrialization.
Table 7.14. Summary of results from the analysis of migration from
720 rural Gemeinden, 1880–9: predicted rates of out-migration by
decade per 100 head of resident population
Ranking of Gemeinde by
population density
Population density
(population relative to
total agricultural area)
Predicted rate of
out-migration
Top decile average 4.06 16.99
Top quartile average 2.68 12.49
Average all Gemeinden 1.27 6.61
Lower quartile average 0.47 2.72
Lower decile average 0.37 2.18
Source: All data from Kaerger (1890).
44 The 70% Wgure is from a CIA study, Todaro (1997), 244.
250 Agricultural Productivity, Labour Surplus, Migration
![Page 260: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/260.jpg)
Appendix 7A. Econometric Analysis of M igration in
Rural Gemeinden
The data set was derived from an article in the 1890 Landwirtschaftliche
Jahrbucher and analysed in the regressions shown in Table 7A.1. There
were seven Kreise represented, mainly in the rural east. The article
looked at Die Sachsengangerei, the pattern of migration from rural east-
ern Kreise to work on the sugar beet Welds of Provinz Sachsen (Prussian
Saxony).
The results show a consistent eVect from population density. The
relationship is not a linear one, as is shown by the signiWcance of the
quadratic term in column b. The predicted eVect of rising population
density is large. A village with population density 70% greater than the
average is predicted to have a rate of migration 42% above the average;
one with twice the average population density is predicted to have a rate
of out-migration 81% above the average (using the coeYcients from
column b).
As a check on the results the Wnal column gives results obtained
when the regression equation includes a full set of Kreise dummies.
Table 7A.1. Analysis of migration from rural Gemeinden, 1880–9:
dependent variable is rate of out-migration per 100 head of resident
population
No Kreise dummies used Kreise dummies used
a b c
Constant 2.20 0.20 �0.78
POPDENSITY þ3.10 þ5.48 þ4.20
(16.27) (12.23) (10.11)
POPDENSITYSQD �0.332 �0.241
(7.83) (4.70)
N 720 720 720
S 6.20 6.06 7.31
R2 0.269 0.302 0.470
adj. R2 0.268 0.300 0.462
F-test of regression 264.6 177.4 62.8
RSS 27760 26311 20002
Source: All data from Kaerger (1890).
Agricultural Productivity, Labour Surplus, Migration 251
![Page 261: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/261.jpg)
This removes any eVect from the location of the Kreis, and other factors
operating at this level. There is a reduction in the size of the estimated
coeYcients, as would be expected, but the results remain signiWcant.
Variations in population density at the village level were important
factors in migration.
252 Agricultural Productivity, Labour Surplus, Migration
![Page 262: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/262.jpg)
8
Migration and Urban Labour Markets
1. Introduction
Lewis’s description of his model as a model of economic development
with unlimited supplies of labour has, naturally, tended to focus atten-
tion on labour supply. Yet, as pointed out in Chapter 1, the assumptions
he made about the supply of capital to industry are also important. His
model could be described as one in which there is a shortage of capital
instead of one in which labour is abundant.
The assumption that ‘only capitalists save’ is an extremely powerful
one in the context of a developing economy where an increase in the
overall rate of investment is required. EVectively, if capitalist savings
are a function of the level of capitalist income and the anticipated rate
of return on new investments, then one of these (or both) must rise if
there is to be an increase in the rate of investment. If borrowing from
the rest of society (through the banking system) is precluded, then it is
likely that there will be a shift in the distribution of income in favour of
capitalists under this assumption.1
A more realistic form of the assumption is that ‘only capitalists
invest’ and that they use savings out of their own incomes, or those of
their family and friends. This has two implications: Wrst, that know-
ledge of potentially proWtable investment opportunities is not widely
diVused—other groups may save, but they are unable to achieve the
high rates of return available on new industrial investments. Secondly,
the cost of Wnancial intermediation is high. Bank Wnance for industry is
either unobtainable or prohibitively expensive, and industrialists have
1 This is not a watertight theoretical proposition. It is possible that the prospect ofhigher returns on new investment might induce capitalists to raise savings out of anunchanged share of national income, and that the fact that the rate of return on newinvestment had risen (which would tend to raise the proWts share) might be balanced by afall in the rate of return on existing investments as a result of the introduction of newtechnology: a Schumpterian ‘creative destruction’ process. However, the assumption ‘onlycapitalists save’ does make it considerably easier to construct models in which there is ashift in the distribution of national income in favour of capitalists during the early stagesof industrialization. Kaldor (1956) and Pasinetti (1962) model this eVect.
![Page 263: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/263.jpg)
little access to pools of savings that may be available elsewhere in the
economy.
However, Wnancial institutions and corporate structures are not
static, and these points, which may be a reasonable description of con-
ditions in the early stages of industrialization, should have less relevance
in a mature economy. The modern corporation, or joint-stock company,
brings together technically proWcient managers, who have knowledge of
investment opportunities but lack substantial savings of their own, and
investors who have savings but not specialized knowledge. A developed
banking sector, with access to integrated capital markets, will bring
down the cost of intermediation and draw upon additional sources of
savings.
In Germany in 1870–1913 there were important changes in corporate
structure and in the Wnancial system which enabled industry to draw
upon savings from outside the limited group of industrial capitalists,
and their families and friends, which had often been the main source of
Wnance for new industrial projects in the early part of the century. The
‘labour surplus’ phase of industrial development was brought to an end
as much by this as by changes within agriculture or the diVerent phases
of the demographic transition.2
This chapter will consider the pace of change, paying particular at-
tention to the role of the Wnancial markets. There were diVerent phases
in the development of the Wnancial sector which had important conse-
quences for the availability of funds for industrial investment. The
analysis is set in the context of a general discussion of the relationship
between migration and conditions in urban labour markets. For there
were other changes in the structure of the urban, or industrial, sector
which aVected migration. The most importance of these was the eVect
of Germany’s external trade position. As Germany became more inte-
grated in the world economy, the importance of the external sector rose
and the German economy became more specialized.3 Changes in the
international terms of trade, and in the structure of German exports,
2 Lindenlaub and Kohne-Lindenlaub (1991) provide an analysis of the early Wnancingof the Krupp enterprise. The view that there was a capital shortage in Germany has beenattacked by Borchardt (1982). The evidence he presents is, however, consistent with theview that there was no shortage of savings, but that there was a requirement for insti-tutional developments to turn these into industrial investments. Walter Rathenau alsocriticized the view that there was a capital shortage in Germany, Pogge von Strandmann(1988a).
3 Germany had always been inXuenced by foreign trade, but the transport improve-ments of the nineteenth century greatly increased the importance of the trade sector.
254 Migration and Urban Labour Markets
![Page 264: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/264.jpg)
had consequences for urban labour markets, and in turn aVected the
Xow of migration.
This chapter also considers some of the consequences of migration
for urban labour conditions: the extent to which it stabilized urban
markets (reducing registered unemployment), the eVect on wage vari-
ations, the eVect of skill transfers, and the impact on the relative
position of women. The question of the eVect on overall income distri-
bution is considered in Chapter 9.
2. Structural Changes in the German
Economy 1870–1913
Table 8.1 gives Wgures for aspects of the German economy, comparing
Wve-year averages, and bisecting the whole period. The early 1890s is animportant dividing point; after this the growth rate of per capita GDP
was 0.5% higher than in the earlier period.
In this second period the rate of investment was also signiWcantly
higher, and this results in a faster rate of increase in the capital stock
per industrial worker (including mining). The acceleration was a
marked one: the rate of increase is 74% higher in the second period.
The importance of foreign trade was also rising. In the Wrst period the
increase in export volumes was only slightly faster than the increase in
GDP; in the second period it was considerably faster. In both periods
Table 8.1. Structural change in the German economy (all Wguresuse 1913 prices)
1870/4–1890/4 1890/4–1909/13
Annual rate of increase of GDP 2.2% 3.2%
Annual rate of increase of per capita GDP 1.3% 1.8%
Annual rate of increase of total capital
per industrial worker
1.5% 2.6%
Annual rate of increase of exports 2.4% 5.1%
Annual rate of increase of exports of
manufactures
3.6% 5.8%
Note: The export-orientation of the pre-1914 economy was high even by the standardsof the post-1945 period: Abelshauser (1983), 148 and 193.Source: All Wgures are calculations from HoVmann (1965).
Migration and Urban Labour Markets 255
![Page 265: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/265.jpg)
the increase in the export of manufactured goods (Wnished and semi-
Wnished) was faster than the increase of total exports.
The division into two periods is a simpliWcation of a rather more
complex pattern, in which Xuctuations in the international economy
also played an important part. Figure 8.1 gives Wgures for total net
investment, and net industrial investment, as percentages of net domes-
tic product. There were Wve main periods. The Wrst was a strong boom
in the early 1870s, dominated by railway investment. This was followed
by a sharp recession which depressed investment into the 1880s. The
third period, which lasted to the mid-1890s, saw a moderate recovery;
boom conditions were not restored until the so-called Elektroboom of
the late 1890s.4 The Wnal period, after 1900, was one of considerable
Xuctuations, but less violent than either the boom of the 1870s or the
4 Hentschel (1978) uses the term ‘Elektro-boom’.
18700
10
20
Rolling five-year averages
1874 1878 1882 1886 1890 1894 1898 1902 1906 1910
Industrialinvestment
All investment
Fig. 8.1. Net investment (all sectors) and net industrial investment as
percentages of net domestic product, 1870–1913 (all Wgures use 1913 prices)
256 Migration and Urban Labour Markets
![Page 266: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/266.jpg)
Elektroboom. They were the normal trade cycle Xuctuations of an indus-
trial economy. The level of total investment remained substantially
higher than in the period before 1895.5 Even in recessions it did not
sink back to the level of the 1870s.
One important source of structural change was the development of
the Wnancial sector. The collapse of the boom of the 1870s did consider-
able damage to the German Wnancial sector. Share prices fell 64%between 1872 and 1877, which greatly reduced the ability of companies
to raise money from the stock market. In 1871–3 a total of 2,781 million
Marks was raised from company Xotations; in 1874–9 just 284 millions.6
This increased the importance of bank Wnance, but the banking sector
had its own problems, stemming from the large number of bank Xota-
tions and formations during the boom. In 1870–4, 103 banks were
founded; 29 of these failed in 1873–4, accounting for a third of all
liquidations (by capital size) during the crash.7 There were further
liquidations and consolidations during the rest of the decade. Of the 36
banks Xoated in 1871 only 15 survived to 1880, of the 35 Xoated in 1872
just 10, and not one of the Wve Xoated in 1873.8
The consolidation of the banking system which followed gave the
German Wnancial system a distinctive character. Between 1870 and
1888 the Deutsche Bank, the largest bank, absorbed 71 other banks.9
The other major joint stock banks, or Kreditbanken, also made acquisi-
tions. By 1890 there were two other banks of similar size (with issued
capital of 75–80 million Marks): the Disconto-Gesellschaft and the
Darmstadter Bank. The Dresdner was slightly smaller (issued capital of
60 million Marks). Together with the Berliner Handelsgesellschaft and
the Schaaf’hauschen Bankverein (50 million and 36 million Marks re-
spectively) these six joint stock banks dominated industrial banking, as
far as the larger companies were concerned.10 Their inXuence was par-
ticularly strong in the new issue market.
HoVmann has produced Wgures for the Wnancing of investment
which, given the gaps in the available Wgures, are best regarded as
illustrative rather than deWnitive. In particular, little is known about the
5 Many of these phases reXected changes in international conditions. Internationaldevelopments are described in Milward and Saul (1977).
6 Wehler (1995), 554–5.7 Tilly (1986), 124.8 Pohl (1976), 26; these Wgures refer to stock market Xotations, Tilly’s to foundations.9 Wehler (1995), 628; Bohme (1966), 347.
10 Jeidels (1905), 66.
Migration and Urban Labour Markets 257
![Page 267: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/267.jpg)
contribution made by companies which were not Aktiengesellschaften
(joint stock companies whose accounts are available). Table 8.2 uses the
available Wgures, and expresses them as percentages of net domestic
product.11
The table shows that the rise in industrial investment accounted for
most of the increase in total investment after 1895. The increase of 3
percentage points was also matched by a similar rise in recorded sources
of Wnance, and this, in turn, can be shown to have resulted from the
increased contribution of two sectors: the Aktiengesellschaften and the
banking sector.
The Wnal section of the table breaks down the contribution of the
banking sector. This shows that the rise in the position of the Kredit-
banken, from 19% of the total banking sector to 29%, was an important
factor in the overall increase. But smaller banks and Sparkassen were
also important sources of Wnance, particularly to smaller and medium-
sized companies. They raised funds from smaller investors, which were
11 Calculated using data from HoVman (1966), 812–13; HoVmann’s title for his table is‘The Wnancing of domestic net investment’, though there are enough uncertainties tomake this a rather optimistic title. HoVmann separates the activities of railway companiesfrom other joint-stock companies, so these are included in ‘other sources’. One justiWca-tion for this is that a number of railway companies passed into state ownership in thisperiod, resulting in large payments to shareholders, and making it diVicult to disentanglethe private and the public sector in this industry.
Table 8.2. Investment and sources of Wnance, 1870–94 and
1895–1913; all Wgures are expressed as percentages of net domestic
product (current prices)
1870–1894 1895–1913
All net investment 11.51 14.66
of which: net industrial investment 3.63 6.49
All recorded sources of Wnance of which: 7.12 10.5
Banks 3.04 5.56
Aktiengesellschaften 0.77 1.61
Other sources 3.35 3.34
Contribution of banking sector
Kreditbanken 0.58 1.61
Other banks 2.46 3.95
Source: All Wgures from HoVmann (1965).
258 Migration and Urban Labour Markets
![Page 268: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/268.jpg)
then placed in the Wnancial markets, thus contributing to the overall
pool of available funds.
As before, the division into two periods is a simpliWcation which ob-
scures more complex trends. Figure 8.2 gives the available Wgures for
sources of Wnance. The ‘all sectors’ series shows a gradual recovery from
the trough of the mid-1870s, reaching a plateau of around 10% of NDP
after 1900. But the Wgures also show the process of structural change: the
rise in the importance of the banking sector, and the slow revival of the
contribution of the Aktiengesellschaften, from a negative position in the
late 1870s. These two sectors dominate the period of the Elektroboom.
These Wnancial statistics show how damaging the consequences of
the 1870s boom and recession were to the structure of the German
capital markets. Recovery was a long, slow process. They also show the
importance of the institutional changes which eventually overcame this
legacy. Somewhere around 1895 the process of recovery and rebuilding
1870-2
0
2
4
6
8
10
12
14
1875 1880 1885 1890 1895 1900
Aktien gesellschaften
sectors
Other financial
Banks
All sectors
1905 1910
Fig. 8.2. Sources of Wnance 1870–1913
Note: All Wgures are rolling Wve-year averages expressed as percentages of net domesticproduct, current prices.
Migration and Urban Labour Markets 259
![Page 269: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/269.jpg)
was complete. The German economy moved out of a ‘capital shortage’
phase and into a phase in which buoyant investment played a leading
role in economic growth.12
A second major area of structural change lay in the external sector.
This was inXuenced by similar technological developments to those
which lay behind the Elektroboom: the growth of technologically ad-
vanced industries in which Germany had a comparative advantage.
This contributed to the acceleration of economic growth after 1895, and
thus to the ending of the ‘labour surplus’ period.
There was a shift in the composition of German exports (Figure 8.3):
exports of manufactured goods (Wnished and semi-Wnished) rose relative
12 The development of the German banking sector has been a subject of debate sincethe work of Riesser (1911) and Jeidels (1905). The views of Gerschenkron (1962) providedfurther stimulation. More recent contributions include Pohl (1976, 1984, and 1986), Neu-berger and Stokes (1974), Tilly (1986), Edwardes and Ogilvie (1995), Fohlin (2002),Feldenkirchen (1982), and Wellhoner (1989).
19101906190218981894189018861882187818741870
6000
5000
4000
3000
2000
1000
0
Finished goods
Agricultural goods
Semi-finished goods
Raw materials
Fig. 8.3. German exports by category, in million Marks (1913 prices)
Source: All Wgures calculated from HoVmann (1965), 530–5.
260 Migration and Urban Labour Markets
![Page 270: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/270.jpg)
to other exports. In 1870–4 manufactured goods accounted for 52.5%
of exports; by 1890–4 this was 65.3%, and 74.2% in 1909–13.
The composition of the most important sector, Wnished goods, also
changed (Figure 8.4). Exports of Wnished goods in 1913 can be divided
into Wve roughly equal categories: textiles (1,150 million Marks), metal
goods excluding machinery (1,135 million Marks), chemicals and related
products (1,042 million Marks), machinery (1,025 million Marks), and
other Wnished goods (1,007 million Marks). Of these, other Wnished
goods showed the slowest rate of growth, but was the largest category in
the 1870s. Textile exports rose rapidly between 1870 and 1887, but then
remained relatively static until 1900. The fastest rates of growth were
recorded by the other three categories. Between 1894/6 and 1911/13 the
annual rates of increase were: 8.5% for chemicals and related goods,
8.7% for metal goods excluding machinery, and 15.3% for machinery.
The rapid growth of exports was made necessary by an adverse shift
in the German terms of trade after 1895. This is shown in Figure 8.5.
Other industries
Textiles
MachineryMetal goods
Chemicals
19101906190218981894189018861882187818741870
1200
1000
800
600
400
200
0
Fig. 8.4. German exports of Wnished goods, by category, in million Marks
(1913 prices)
Migration and Urban Labour Markets 261
![Page 271: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/271.jpg)
HoVmann’s index of export prices divided by import prices fell by 22%
between 1894/6 and 1911/13. This was partly due to the recovery of
world food prices, but another factor was a fall in the prices of German
manufacturing exports relative to the cost of raw material imports,
which was steeper (32% over the same period).13
Although there was a general rise in raw material costs in this
period, German industry was aVected more than others. HoVmann’s
index of German manufactured export prices divided by the prices of
imported manufactures fell by 30% during the period 1895–1913.
Foreign-produced manufactures rose in price while the prices of
13 The adverse shift in the terms of trade was one reason why German wages did notconverge, or did not converge strongly, with those in Britain and the United States. Thereare a range of wage indices for Germany in the period, and some show rather strongerwage growth. The main studies are Kuczynski (1945), Grumbach and Konig (1957),Gommel (1979), Desai (1968), Bry (1960), Orsagh (1969), and Hohls (1995).
Manufactures - raw materials
All exports - all imports
19121910
19081906
19041902
19001898
18961894
18921890
18881886
18841882
1880
160
150
140
130
120
110
100
90
80
Fig. 8.5. The terms of trade of the German economy 1880–1913
Note: 1913 ¼ 100; export prices divided by import prices and manufactured export pricesdivided by raw material import prices.Source: From HoVmann (1965), 548.
262 Migration and Urban Labour Markets
![Page 272: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/272.jpg)
German-produced goods were static. The average cost of raw material
imports rose by a half between 1895 and 1913. The cost of imported
food rose by a third.
From the German point of view this was a serious development.
Calculated over the whole 1870–1913 period, the income elasticities of
food imports and imports of raw materials were 1.31 and 1.58 respect-
ively.14 The combined volumes of these two items would be expected to
rise 45% faster than the rate of growth of national income. As these
categories accounted for 73.5% of Germany’s imports in 1894–6 it is a
fairly simple calculation (assuming that other categories of imports rose
in line with GDP) that, faced with the adverse price movements experi-
enced in 1895–1913, German exports would have to rise 80% faster
than the growth of overall national income to maintain balance of pay-
ments equilibrium. The required rate of increase for 1895–1913 was
5.7% per annum.15
Maddison’s Wgures for the period show that world trade grew at an
annual rate of 3.9% in 1870–1913.16 So the implication of this calcula-
tion is that Germany would have to make substantial gains from other
exporters to maintain a rate of growth of per capita GDP of around
1.8% per annum. Given the resources available to the German econ-
omy, it is clear that the main target would have to be Britain.
This provides a rather diVerent perspective on the Anglo-German
trade rivalry of the period before 1913. From the German point of
view, Germany’s economic success in 1870–1913 consisted of a rate of
growth which still left per capita GDP 24% below the British level in
1913 (having been 41% behind in 1870).17 German wages in 1905 were
83% of the British level at current exchange rates, and between 71%
and 75% of the British level in real terms, according to the Board of
Trade survey of that year. Phelps-Brown and Browne have pointed out
that the original Board of Trade wage comparison was too heavily
weighted towards craftsmen’s wages, which were higher relative to the
unskilled in Germany. They used Desai’s indices, which are more
broadly based, and this brought the German level down to 75% of
14 Calculated by dividing the rates of increase of these import categories, 1870/4 to1909/13, by the overall growth of German GDP, Wgures from HoVmann (1965).
15 The actual annual rate of increase of exports 1894/6 to 1911/13, 5.4%, did not quitematch this, which was one reason why Germany had a deWcit on visible trade in 1911–13.
16 Maddison (1994).17 Maddison (1995); this is an upwards revision of the Wgures for Germany in Maddi-
son (1991 and 1994). According to Prados de la Escosura (2000), 26–7, the gap was only12%.
Migration and Urban Labour Markets 263
![Page 273: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/273.jpg)
the British at current exchange rates, and 63–67% at equivalent pur-
chasing power.18
It is also notable that there was little tendency for German wages to
converge on the British level, despite Germany’s faster rate of growth.
Williamson’s comparison shows real wages rising 47.2% in Germany
1870/4 to 1909/13, and 46.3% in Britain.19 Throughout this period
German exporters had a substantial competitive advantage over their
British rivals. Per unit labour costs were between 72% and 79% of the
British level in 1905.20
Yet German exporters may well have needed an advantage of this
order to make the required gains at Britain’s expense. Taking market
share away from an already established rival is not an easy business.
German exporters had to keep prices low. Between 1895 and 1913
prices of German manufacturing exports were reduced by around 20%
relative to the domestic price level. The need to gain market share by
oVering lower prices meant that German wages had to be held down.
As a result the impressive performance of the German economy was not
fully reXected in real wage increases.
There is a wider implication to this point. The argument that
developing economies are likely to encounter deteriorating terms of
trade is a well-established (though controversial) view in development
economics. It is a possible alternative to the Lewis Model. In the Lewis
Model, urban wages are held down by the availability of under-
employed rural labour. If the terms of trade are deteriorating, then
wages will be held down in the export-orientated sectors by the eVects
of falling export prices.21
There is a way out of this dilemma, although it is one which rela-
tively few developing economies have managed to take, and that is to
‘leapfrog’ up the technological ladder: to establish a strong position in
new industries rather than cutting prices to gain an increased share in
existing ones. This is what Germany achieved, and Figure 8.4 showed
18 Phelps-Brown and Browne (1968). Their index covers 18 manufacturing sectors,together with transport and mining, so it is more than just an index of manufacturingwages.
19 Williamson (1995), 164–7; this is an index of unskilled wages (mainly buildinglabourers), not those paid in export-orientated industries.
20 This combines Broadberry’s Wgure for relative manufacturing labour productivity in1905 with the wage Wgures produced by the Board of Trade survey, and by Browne andPhelps Brown: Broadberry (1997a), Phelps-Brown and Browne (1968), and Board ofTrade (1908).
21 The arguments are outlined in Todaro (1997), 478–80.
264 Migration and Urban Labour Markets
![Page 274: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/274.jpg)
that much of German export growth was due to successes in new indus-
tries: chemicals, steel, electrical equipment, machine tools.22
But, although the weight of these newer industries was rising (chem-
icals, machinery, and metal goods accounted for 30.1% of exports in
1911–13, against 5.6% in 1870–2), Germany also pushed up exports in
categories where price advantages were probably more important:
exports of textiles, other Wnished goods, and semi-Wnished goods all
rose in 1895–1913. The increase in German exports was due to a com-
bination of successes in new technology and cost advantages in older
industries.
There are two ways of looking at the economy of Wilhelmine Ger-
many. One is a static analysis, looking at indices of real GDP per capita
or wages at purchasing power. These comparisons show that Germany
in 1913 was not a particularly rich country by the standards of the time,
although progress had been made since 1870. The other is to look at
potential: the structure of industrial output and, in particular, the pat-
tern of exports. Here, Germany had established a strong position in a
number of industries with excellent growth prospects in the twentieth
century. The potential for strong export-led growth, eventually realized
by the Federal Republic after 1945, was already in existence by 1913.23
Moreover, the move up the technological ladder made it possible for
Germany to escape from the terms of trade dilemma shown in Figure
8.5. Instead of competing on price in markets where British Wrms had
established positions, Germany was now able to oVer new products,
which Britain did not produce, and also compete on quality, by incorp-
orating new technology in goods such as electrically-driven machine
tools. This meant that export growth could be sustained without cutting
22 There is an argument, put forward by Paul Krugman, in Brezis et al. (1993), thatleapfrogging is an automatic process. This argument does not survive close scrutiny of therelatively small number of cases where leapfrogging has occurred. There were, forexample, strong reasons why the dyestuVs industry should have located in Britain notGermany (as it initially did in the 1860s). Britain had an ample supply of the mainfeedstock (coal tar) and a dominant position in the industry which was the main customer(textiles). Such strong positive externalities, running from an existing technology to a newone, generally mean that a more advanced economy has a considerable advantage in a newtechnology compared to a less advanced economy. Germany’s success in newer industriesowed much to Germany’s strong position in science, education, and training.
23 This is the converse of the British position. Static analysis of the British economy in1913 shows that it was still doing well, but the composition of exports was ill-suited torapid growth in the twentieth century: something that no amount of apologetics for thelate Victorian economy can obscure. Britain was ‘over-committed’ to industries with poorgrowth prospects, Richardson (1965).
Migration and Urban Labour Markets 265
![Page 275: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/275.jpg)
prices, and holding down wages. The prospects for a sustained increase
in real wages were much better as a result.
3. M igration and Industrial Structure
The rapid growth of exports in certain sectors had eVects on migration.
These industries were not evenly distributed over the whole country but
relatively concentrated. Table 8.3 gives Wgures for the four main regional
groupings, and also for the three advanced industrial regions: Berlin-
Brandenburg, the Kingdom of Saxony, and Rhineland-Westphalia.
Table 8.3. Percentage distribution of the non-agricultural occupied
population, and the population occupied in three export-orientated
industries, 1895 and 1907
Metal-
working
Machinery Chemicals All three
industries
Total
non-
agricultural
occupied
population
1895
East 13.8 14.9 9.7 13.8 17.7
Centre 26.1 34.6 29.1 28.9 29.0
Berlin/Brandenburg 9.6 12.4 12.3 10.7 10.9
Saxony 8.7 12.9 6.6 9.8 9.9
West and north-west 39.3 28.2 36.8 35.8 32.6
Rhineland-
Westphalia
24.9 11.9 15.7 20.2 16.1
South and south-west 20.8 22.2 24.4 21.6 20.7
1907
East 11.9 10.9 8.3 11.2 16.0
Centre 26.2 37.6 28.9 31.2 29.8
Berlin/Brandenburg 11.1 15.4 12.3 13.0 12.2
Saxony 8.0 13.5 6.8 10.2 9.8
West and north-west 42.1 32.9 41.0 38.2 35.0
Rhineland-
Westphalia
28.2 16.6 17.6 22.5 18.4
South and south-west 19.8 18.6 21.8 19.4 19.1
Source: Figures are calculations from data in Kaelble and Hohls (1989).
266 Migration and Urban Labour Markets
![Page 276: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/276.jpg)
These three advanced regions accounted for 40.7% of the total occupied
population in the three export-orientated industries in 1895, and this
had risen to 45 percent by 1907. Metal-working and machinery were
more concentrated than chemicals (possibly for geographical reasons).
The east lost ground between 1895 and 1907. Its share of these three
industries fell, as did its share of the total non-agricultural occupied
population. The failure of eastern industry to keep pace with the rest of
the country was a major factor in the high rate of migration out of this
region. The south did rather better. It had a high share of employment
in chemicals so its share of the export-orientated industries was rela-
tively high, though this fell between 1895 and 1907.24
Figure 8.6 gives the change in the regional pattern of non-
agricultural employment, 1895–1907. There is an evident shift towards
25−30
30−5
20−5
10−15
40−5
35−40
60−70
70−80
<10
>80
15−20
50−60
45−50
Fig. 8.6. Percentage increase in non-agricultural employment by region,
1895–1907
Source: Calculated from data in Kaelble and Hohls (1989).
24 Hohorst, in Pollard (1980), 231, gives estimates of regional income levels whichconWrm that the east was falling behind.
Migration and Urban Labour Markets 267
![Page 277: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/277.jpg)
the centre and the west. In part this reXects the growth of cities: these
areas had more large cities and these were the areas where employment
in industry and services was increasing most rapidly. But it is also
noticeable that there were a number of regions in the west (Stade,
Luneberg, and Oldenburg, for example) which did not contain large
cities, where non-agricultural employment rose faster than in eastern
regions, such as Konigsberg, Breslau, or Stettin, which did.
Again, the south and south-west has a diVerent pattern, containing
both regions where there was little increase in non-agricultural employ-
ment (Upper Alsace, Lower Bavaria, and Jagstkreis in Wurttemberg)
and others where the increase was over 40%—Mittelfranken, Mann-
heim, and Lorraine (there were none in the east). This was one reason
why most migration in the south was short or medium-distance; the
east was the main source of long-distance migration because there were
less opportunities to Wnd non-agricultural work within the region as a
whole.
There were two trends which did not help the east. The Wrst was the
tendency for regions to become more specialized and for industries to
become more concentrated. External economies, internal and external
to the Wrm, meant that industrial growth was highest in regions which
were already relatively industrial.25 The second was vertical integration:
the tendency for processing industries to locate in areas close to sources
of primary production. Again, this could be both internal and external
to the Wrm.
Both these eVects can be demonstrated using the occupational cen-
suses. Table 8.4 provides a summary of the results of a comparison of
the censuses for 1895 and 1907 (the full results are given in Table 8B.1
in Appendix 8B). The analysis shows that there was a tendency towards
concentration both in the industrial and in the service sectors. Regions
which had above-average levels of non-agricultural employment had
higher rates of increase for employment in these sectors compared to
regions with below-average rates of employment in industry and ser-
vices. The estimates suggest that a region with 70% of the occupied
population employed outside agriculture would experience a 46% rise
in non-agricultural employment; one with only 40% would have a rise
of 34%.26
25 Steller (1903) gives Wgures for the formation of machinery Wrms which show that theeast was losing ground in this area.
26 Using the coeVicients given in the Wrst column of Appendix Table 8B.1 (OLSestimates).
268 Migration and Urban Labour Markets
![Page 278: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/278.jpg)
Analysis of results for individual industries identiWed three which
were becoming more concentrated: metal-working, machinery, and
clothing. This suggests that an ‘agglomeration eVect’ was in operation
in these industries. The miscellaneous, ‘other industries’, sector was
also moving in this direction. Only one sector showed a tendency to
move towards dispersion—utilities (though the result is not a secure
one). In this sector a tendency towards dispersion might be expected, as
a result of the spread of electriWcation and gas production, and wide-
spread investments in improved water supplies.
Taking the machinery sector as an example, the results imply that a
region with 0.5% of the occupied population in this industry in 1895
would experience a 78% increase in employment; one with 2.0% would
have an increase of 120%. There would be a pull of migrants towards
the towns and cities associated with these ‘concentrating’ industries.
A similar approach can be used to examine the evidence for increased
vertical integration. There was strong association between linked indus-
tries, such as mining and metal production: employment in one was
Table 8.4. Summary of results of statistical analysis of trends in
employment shares for 81 regions, 1895–1907: trends towards
concentration and dispersion
I. Secure statistical results
Sectors or industries showing a trend towards concentration:
The industrial sector as a whole
The service sector as a whole
Metal working
Machinery
Clothing
Miscellaneous industries
II. Less secure statistical results
Sectors or industries showing a trend towards dispersion:
Utilities
III. Other industries showing no clear trend to concentration or dispersion
Mining and quarrying
Metal production
Chemicals
Textiles
Food and drink
Construction
Migration and Urban Labour Markets 269
![Page 279: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/279.jpg)
high in regions where the other was also present. The question con-
sidered was whether these links were getting stronger. If so, then em-
ployment in the ‘downstream’ industry would be rising faster in areas
where the ‘upstream’ industry was already established. The strongest
evidence for this was found in the machinery sector: this industry was
locating in areas where metal-working was already established, and also
in areas where high employment in utilities indicated good supplies of
power and other services. There was some evidence that metal-working
Table 8.5. Summary of results of statistical analysis of trends in
employment shares for 81 regions, 1895–1907: trends towards vertical
integration (linked pairs)
I. Groups of industries showing a trend towards increased vertical integration
Secure results:
Machinery—metal-working
Machinery—metal-working—utilities
Less secure results:
Metal-working—metal production
II. Groups of industries showing no trend towards increased vertical integration
Mining—metal production
Clothing—textiles
Table 8.6. Summary of results of statistical analysis of movement of
insured workers 1909–1913
I. Secure statistical results
Characteristics of the recipient regions which increased inwards migration:
High employment in industry and services relative to agriculture
High employment in services relative to industry
II. Less secure statistical results
Characteristics of the recipient regions which increased inwards migration:
High employment in chemicals
High employment in metal-working
High employment in construction
Characteristics of the recipient regions which decreased inwards migration:
High employment in utilities
270 Migration and Urban Labour Markets
![Page 280: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/280.jpg)
was expanding more in areas where employment in metal production
was already high.
The implications of industrial structure for migration are considered
in table 8.6, which summarises the statistical analysis given in Appensix
8B section ii. This is based on the recorded movements of insured
workers in 1909–1913. The analysis shows that the likelihood of migra-
tion into a given region was influenced by the availability of non-agri-
cultural employment, particularly in the service sector, which seems to
have attracted migrants more than industry.
Industries which were found to have a particular appeal to migrants
in the period included 2 export-oriented industries – chemicals and
metal-working – which were expanding rapidly in the period just before
the First World War, and construction, which had always employed
large numbers of transient workers. On the other hand, it seems that
jobs in utilities were not easily obtained by migrants, as this sector is
associated with lower migration.
4. M igrat ion and Economic Cycles
The argument put forward earlier was that there was a change in the
nature of migration, from a destabilizing, labour-surplus pattern up to
1895 to a stabilizing pattern after this date, more characteristic of
modern mature economies. Migration then acted as a buVer, absorbing
part of the eVects of short-term economic Xuctuations on employment
in urban labour markets. But there was still substantial rural–urban
migration, which helped to reduce the strain on urban labour markets
in recessions. This may help to shed some light on a diVerent question
which has been much debated: why was recorded unemployment rela-
tively low before 1914?
Town migration Wgures provide a source for annual movements. The
discussion in Chapter 4 showed that these are not without their prob-
lems: there were gaps in the coverage, diVerences in deWnitions and in
legal requirements to register a change of residence. From 1898 onwards,
comparability was improved, and annual Wgures were published in the
Statistisches Jahrbuch Deutscher Stadte. For the years 1904–12 it is pos-
sible to produce a series covering 25 cities with a total population of 7.5
million. All the major cities are included with the exception of Munich.
These Wgures are given in Table 8.7, together with labour exchange
Wgures for 23 of these cities, and two series for economic conditions in
Migration and Urban Labour Markets 271
![Page 281: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/281.jpg)
the whole economy. The Xuctuations of the economy can be followed in
the second column, giving the annual change in industrial production.
There was a period of expansion in 1904–7, followed by a brief reces-
sion in 1908, and a recovery in 1909–12. The Wnal two columns show
how labour market conditions responded to these Xuctuations. In the
recession there was both a rise in applicants per vacancy, and in un-
employment.
The way that migration was aVected by changes in economic condi-
tions is quite clear. There was a sharp fall in net gains in 1908–9 from
the much higher rates in 1904–7. It is noticeable that the response of
migration was at least as great as the change in recorded unemploy-
ment. Statistical analysis (reported in Table 8B.4 in Appendix 8B,
Section (iii)) shows that a change of one percentage point in unemploy-
ment produced an equivalent change in migration. The eVect of migra-
tion was to halve the change in unemployment (assuming that all those
deterred from migrating would otherwise have become unemployed).
However, this may underestimate the importance of migration. The
unemployment Wgures are derived from trade union sources, and so
Table 8.7. Summary of gains and losses through migration 1904–1912:
25 cities with full sets of migration Wgures, together with the change
in total industrial production, a comparison of labour exchange
applications to the number of open places, and recorded
unemployment
Net gains
as % of
population
% change in
industrial
production
Ratio of
applicants
to vacancies
Recorded
unemployment
1904 2.50 þ4.2 1.26 2.1
1905 2.73 þ3.7 1.03 1.6
1906 2.96 þ4.3 0.95 1.2
1907 2.14 þ7.8 1.26 1.6
1908 0.26 �0.9 1.79 3.0
1909 0.78 þ4.4 1.65 2.9
1910 1.24 þ5.0 1.30 1.9
1911 1.28 þ6.1 1.15 1.9
1912 1.11 þ7.2 1.21 2.0
Source: Calculated using data from SJDS (1906), 280 and 482–3, and subsequent issues;unemployment from Bry (1960), 325–6; industrial production from HoVmann (1965),451–2.
272 Migration and Urban Labour Markets
![Page 282: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/282.jpg)
record numbers unemployed as a percentage of union members. This
may be a reasonable approximation for unemployment amongst the
working population. But the migration variable used in the regressions
is net moves relative to the total population. If participation rates
amongst migrants are above average, particularly in the ‘volatile com-
ponent’ of migration, those who move in response to changes in eco-
nomic circumstances, then the ‘unemployment-relieving’ eVect will begreater. As we know that migrants are predominantly young and of
working age, it is indeed probable that participation rates will be above
average.
A few calculations illustrate the importance of this point. If the par-
ticipation rate amongst ‘volatile’ migrants had been 60%, against a
German average in 1907 of 46%, then the ratio between the change in
unemployment and the change in migration (of those available for em-
ployment) would be raised from 1 : 1 to 1.37 : 1. If the participation rate
had been 80%, the ratio is increased to 1.83 : 1.
Figure 8.7 shows the eVect of adding the change in migration to the
change in unemployment. Adjustment 1 assumes the participation rate
of migrants to be 60%; adjustment 2 makes it 80%. The combined
eVect is to greatly increase the volatility of the whole series.
It seems clear, from these Wgures, that migration in and out of the
urban sector made an important contribution to the reduction of
recorded unemployment in the period before 1914. There are two add-
itional points which follow from this. The Wrst is that this implies the
existence of a pool of underemployed labour in agriculture. At times
when the economy was expanding rapidly, this might be reduced to
quite a low level, but in recessions it would rise to substantial propor-
tions if the reduction in migration recorded for these 25 cities applied
to all moves into the non-agricultural sector.
The second is that a decision by an unemployed migrant to return to
his or her district of origin is not without its costs. There is the cost of
the journey. In addition, there may be advantages to remaining in a city
even if unemployed. It would be easier to look for job opportunities. It
is likely that, when conditions improve and new vacancies appear, the
Wrst to obtain these jobs will be the unemployed still resident in the city
or industrial district, not those who have returned to a parent’s farm or
smallholding.
If this is correct, then the implication is that, if unemployment ben-
eWts were to be made available to workers who had previously been
unable to claim (as in fact happened in the 1920s and 1930s), then, even
Migration and Urban Labour Markets 273
![Page 283: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/283.jpg)
if these were kept quite low, a substantial number of the unemployed
would prefer to remain in the cities rather than migrate back to rural
areas. The payment of beneWt would change the position from one in
which there was a substantial, though Xuctuating, pool of hidden rural
underemployment, to one in which there was overt unemployment in
cities.
5. Some Consequences of M igration
(i) Migration and wage convergence
The consequences of migration are, on the whole, likely to be favour-
able for the migrants themselves, even if they may be not so good for
those already resident in the recipient cities who face increased compe-
1905−4
−2
0
2
4
6
1906 1907 1908 1909 1910 1911 1912
Change in unemployment
Plus migration
Plus migrationadjusted (1)
Plus migrationadjusted (2)
Fig. 8.7. Percentage changes in unemployment plus migration, 1904–12
Note: Annual changes in the unemployment rate in % points, relative to the workingpopulation, showing the eVect of allowing for changes in the rate of migration underdiVerent assumptions about the participation rate of migrants.
274 Migration and Urban Labour Markets
![Page 284: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/284.jpg)
tition for jobs and housing. The ‘congestion eVect’ on urban amenities
and infrastructure may also be a cost not included in the calculations
made by individual migrants. But the transfer of labour from relatively
unproductive activities, or from regions with high underemployment,
should beneWt the economy as a whole.
The available wage and migration data are not suViciently good to
make it possible to measure the eVect of migration on urban wages by
econometric analysis. It is possible to look at some general trends in
wages, in particular the question of wage convergence. Table 8.8 pro-
vides an analysis based on the surveys of ‘typical’ day-labourer rates
(Ortsubliche Tagelohne) made for insurance purposes. The Wrst part of
the table gives regional averages (weighted by the city population).
These show that wages in eastern cities were well below those in other
Table 8.8 (a) Day-labourer rates (Ortsubliche Tagelohne) in 85 cities,
various years; weighted averages (by population); German average ¼ 100
1884 1892 1902 1912
Men
East 77.9 86.6 86.5 85.5
Centre 102.2 107.8 106.7 109.7
West and north-west 110.8 110.3 106.4 104.8
South and South-west 101.0 100.3 106.2 109.7
Women
East 73.6 73.8 79.1 78.8
Centre 100.8 101.7 99.9 108.0
West and north-west 113.4 113.5 109.9 106.7
South and south-west 105.2 102.3 109.3 110.6
Source: Calculated using data from SJDS (1913), 826.
Table 8.8 (b) CoeVicients of variation for day-labourer rates in
85 cities
1884 1892 1902 1912
Men 0.172 0.149 0.133 0.138
Women 0.198 0.169 0.170 0.166
Note: The coeVicient of variation is the standard deviation divided by the mean: it istherefore a standardized measure of dispersion.
Migration and Urban Labour Markets 275
![Page 285: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/285.jpg)
parts of the country, although the gap did fall. The period of ‘catch-up’
was 1884–92 for men and 1892–1902 for women.
In 1884 the highest wages were in cities in the west and north-west,
but this was gradually altered, as this region tended to lose ground to
the centre and the south. Tying this in to migration is not so easy. The
main migrant Xow was from the east to the centre and the west, so it
probably contributed to the convergence of eastern wage rates, but there
was no obvious connection with the diVerent trends of wage rates in the
centre and in the west.
Wage dispersion (as measured by the coeVicient of variation—part
(b) of the table) tended to fall, both for men and women, although most
of the fall occurred in 1884–92. This indicates that the German labour
market was becoming more integrated: regional diVerences were eroded
by the migration of labour from low wage areas. This occurred despite
the relatively uneven pattern of economic development shown in Figure
8.6, which should have caused wage dispersion to rise. The migration
process was suViciently powerful to overcome this eVect.
The wage Wgures produced by these surveys have the advantages of
being contemporary estimates, and produced using a common method-
ology, but they are not based on actual data: the local insurance oVices
were asked to provide Wgures for what they considered to be the ‘typ-
ical’ day-labourer rate for the locality. It would be desirable to supple-
ment this evidence of convergence with data from some other source,
recording actual payments to some category of workers. There are vari-
ous possible series, but relatively few that cover a substantial number of
regions or cities over a decade or more. One possible series gives the
earnings of miners (hard coal, lignite, and iron ore) in nine categories
covering eight diVerent regions, 1889–1913 (Figure 8.8). This series
shows a consistent fall in the coeVicient of variation for above-ground
workers, but the position with regard to below-ground workers was less
clear. Of the two, it is probably the above-ground series which gives the
more representative picture of general labour market conditions. Below-
ground workers were often on piece rates which meant that geological
problems would aVect earnings. The diViculties experienced by some
Silesian coal mines (an area where earnings were low), for example, may
have held back convergence.
Another series gives the weekly earnings of unskilled building
workers in eight cities (Figure 8.9). This also shows a strong process of
convergence. However, convergence was much weaker for the more
skilled categories, carpenters and masons (although data is only available
for four cities).
276 Migration and Urban Labour Markets
![Page 286: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/286.jpg)
18890.08
0.10
0.12
0.14
0.16
0.18
0.20
0.22
1892 1895 1898 1901 1904 1907 1910 1913
Below ground
Above ground
Fig. 8.8. CoeVicients of variation for the daily earnings of miners,
1889–1913, nine categories
Source: Calculated using data from Bry (1960), 342–5.
189910
20
30
40Marks
1901 1903 1905 1907 1909 1911 1913
BerlinDresdenElberfeldFrankfurt am MainKielLubeckNurembergRostock
Fig. 8.9. Weekly earnings of unskilled building workers, eight cities, 1899–1913
Source: Calculated using data from Kuczynski (1945).
Migration and Urban Labour Markets 277
![Page 287: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/287.jpg)
The evidence of convergence is particularly striking when contrasted
with the results given in Section 3 of this chapter, which showed a
process of divergence with regard to non-agricultural employment: the
east was not keeping up with the more advanced parts of the country. If
wages were converging while industrial structure was not, then the
most likely explanation is that this was caused by migration, if no other
force promoting convergence can be identiWed.However, at the end of this period there were still major disparities
in wage rates between diVerent parts of Germany. Figure 8.10 gives the
25−30
30−35
20−25
10−15
40−45
35−40
>10
15−20
60−70
70−80
>80
50−60
45−50
Fig. 8.10. Day-labourer rates in urban areas in German regions, 1912
(weighted averages of male and female rates)
Note: Regional averages were calculated by weighting the rates for each city by its share ofthe total urban population in each region (where the number of cities in the region isgreater than one). Regional units which did not have a major city were amalgamated withnearby regions (Koslin with Stettin for example).Source: Data for 94 cities from SJDS (1913), 826.
278 Migration and Urban Labour Markets
![Page 288: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/288.jpg)
Ortsubliche Tagelohne (day-labourer rates) for urban areas throughout
Germany in 1912. The Wgures are weighted averages of male and female
rates, using the share of each sex in the total occupied population for
the whole of Germany (the male share was 66%). Figure 8.10 conWrms
the point that urban wages in the east were below levels in the rest
of the country, and it shows that this applied just as much to the indus-
trial areas, such as Silesia, as to the other, more agricultural, eastern
regions. It also shows that wages in many southern cities were high.
Day-labourer rates in Freiburg, Mannheim, Stuttgart, and Munich
were as high as anywhere in Germany.
As in the earlier analysis of rural wage rates, the inXuence of the
Russian–Polish border is strong. There is a clear east–west gradient,
although high rates were also found in the centre, in the Berlin area,
and in Saxony. Relating this to the discussion in earlier chapters, it
provides a reminder of the point that German labour market conditions
were inXuenced by Germany’s position in central Europe, and the prox-
imity of Russian Poland in particular.
(ii) Migration and skill endowments
There is one possible consequence of migration which is less favourable
for the region of origin: if migration depletes the skills available in a
region signiWcantly faster than the overall rate of out-migration then the
region will become an unattractive location for advanced industries
with high requirements for skilled labour. A vicious circle could de-
velop: a lagging region loses skilled workers and then lags even more as
a result.
The German occupational census of 1907 is well suited to the analy-
sis of this question, as it recorded the place of birth of respondents, as
well as classifying them by the skill requirements of their occupations.
Consequently, it is possible to compare the numbers born in a given
region, of a particular skill level, with the numbers resident at the time
of the census. Regions which experience out-migration will have a ratio
of numbers born to numbers resident of less than 1. The question is: is
this ratio lower for the higher skill categories?
Table 8.9 provides an analysis of this type for the main regional group-
ings. The results show that the east had a net loss of around a quarter of
the men born in the unskilled category, but a lower Wgure of just 10% of
women. However, female participation rates fell with age, so this may
just reXect the fact that many unskilled female migrants withdrew from
the labour market after migration. Net losses for the east in the skilled
Migration and Urban Labour Markets 279
![Page 289: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/289.jpg)
category are about the same for men and women, which means that they
are no higher than the unskilled for men, but rather higher for women.
In general the table shows that the east did not lose disproportionate
numbers of skilled or entrepreneurial male migrants, with the exception
of the ‘managerial and technical’ category, where there were high losses
both of men and women. So the question revolves around two issues:
how important were the high losses in this category? and can the high
losses of skilled and entrepreneurial women relative to unskilled be
disregarded?
Looking at the absolute numbers, it is clear that the most important
ratio in the more skilled categories is the ratio for skilled male workers:
there were more of these than any other category. The fact that the east
did not lose disproportionate numbers of these leads to the conclusion
that the eastern economy did not experience a major loss of economic
potential through migration. On the other hand the Wgures do suggest
that for white-collar workers of both sexes, and for entrepreneurs and
the self-employed in the service sector, the east was a relatively un-
attractive place to work. This would have contributed to migration. It
Table 8.9. Ratios of numbers born to numbers resident, by skill
category and region, 1907 occupational census
East Centre West and
north-west
South and
south-west
Entrepreneurs and
self-employed: industry
Men
Women
0.803
0.845
1.077
1.148
1.062
1.044
0.995
0.993
Entrepreneurs and
self-employed: services
Men
Women
0.713
0.706
1.134
1.200
1.075
1.059
0.996
0.980
Managerial and
technical: industry
Men
Women
0.587
0.622
1.129
1.153
1.070
1.021
0.950
0.960
Skilled workers:
all sectors
Men
Women
0.740
0.744
1.069
1.185
1.090
1.024
1.015
0.964
Unskilled workers:
all sectors
Men
Women
0.746
0.903
1.273
1.202
1.152
1.035
0.973
0.996
Source: Calculated from 1907 occupational census, SDR n.f. 210.
280 Migration and Urban Labour Markets
![Page 290: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/290.jpg)
would also have had some detrimental consequences for the growth of
non-agricultural employment.
Turning to the other regions: the centre and west made net gains in
all categories. It is noticeable that, in both regions, the highest net gains
were in the ‘male, unskilled category’; the educational and training
system within these regions was supplying most of the skills needed
from those born within the region. The south had net losses in all
categories except skilled male workers. So the migration balance was
slightly favourable to the ratio of skilled to unskilled in the south. The
eVect was not so large: migration raised the ratio of skilled to unskilled
male workers from 0.0835 to 0.0871.
In general, therefore, the migration pattern of the more skilled was
not massively diVerent from that of the less skilled, and skill transfers
were not a major contribution to the economic success or failure of the
diVerent regions.
(iii) Migration and the position of women
The analysis in Sections (i) and (ii) has shown up some diVerences
between the position of men and women. According to the Wgures in
Table 8.9, the gap between male and female wages was larger in the
east: female pay was 78.8% of the national average in eastern cities in
1912, while the Wgure for men was 85.5%. The analysis of skill trans-
fers has indicated that opportunities for white-collar employment, an
important area for women, were better outside the east. So, it would
appear that migration out of the east should have led to an improve-
ment in women’s prospects.
There are two important problems with this view. The Wrst is the
eVect of migration on women’s place in the household economy. Analy-
sis of the budgets of rural households in 1910–14 shows that the house-
hold economy contributed as much as 44% of total household income
in some rural areas.27 This cannot be exactly equated with the female
contribution, but it does give an idea of the potential size of the contri-
bution which could be made by a married woman who did not have
formal employment. There is good evidence that migrant households
preferred, if possible, to replicate this position in industrial areas, by
obtaining smallholdings. This was one factor which contributed to the
27 Mittler (1929), 210–15.
Migration and Urban Labour Markets 281
![Page 291: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/291.jpg)
large rise in small agricultural holdings in industrial regions.28 There
was an increase of 72,975 in the numbers of holdings of less than 2
hectares in Westphalia in 1882–1907.29 However, this was only possible
in industrial areas where agricultural land could be found in close prox-
imity. This was often the case in mining areas but it was rare in the
larger cities.
The second point is that the gap between male and female earnings
was higher in urban areas than in the countryside, and it was higher still
in the major cities. Table 8.10 gives Wgures calculated from the 1892 and
1902 surveys of Ortsubliche Tagelohne. The second part also shows that
the gap between rates for men and women in cities was tending to rise up
to 1902, although there was some reduction in 1902–12.
In summary, migration had, on the whole, negative consequences for
the economic position of women. Although some may have beneWted
Table 8.10. The relative position of women: ratios of male to
female wages (Ortsubliche Tagelohne)
(a) Average ratios for diVerent localities in Prussia (1892 and 1902 surveys
combined)
Cities Towns Rural
Male to female ratio 1.646 1.532 1.519
Source: Data from Neuhaus (1904).Note: Neuhaus (1904) gives, for all the Prussian Kreise, Wgures for rural and urban rates.These have been divided between ‘cities’ (urban areas with a population of at least 50,000in 1912) and ‘towns’ (other urban areas). The averages given in the table are unweightedaverages following this classiWcation.
(b) Average ratios for 86 German cities
1884 1892 1902 1912
Unweighted 1.591 1.618 1.679 1.657
Weighted by population 1.585 1.639 1.696 1.665
Source: Calculated using SJDS (1913), 826.
28 The smallholding could support the family in times of unemployment or sickness.‘A small piece of land is eVectively a savings bank for the worker’, Wygodzinski in Sering(1899–1910), i. 140.
29 Dillwitz (1973), 98; see Tenfelde (1977), 114–18 for the importance of smallholdingsto the Ruhr miners.
282 Migration and Urban Labour Markets
![Page 292: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/292.jpg)
from the improved opportunities to achieve white-collar employment,
or to become self-employed in the service sector, these were a minority.
Wages for women in cities were lower relative to men than in the
countryside, and there was a loss of economic status as a result of
reduced opportunities within the household economy.30
However, life expectancy Wgures show a rather diVerent picture
(Table 8.11). Here, it seems that men were the main losers. At a typical
migrant age of around 20, men experienced a loss of nearly 4 years in
life expectancy by moving to a city. This was 3.5 years more than the
loss experienced by women. Possible reasons for this include the more
dangerous nature of the work done by men in urban areas, and the fact
that more women were employed as domestic servants, which may have
involved better housing or residence in areas with better sanitation. On
this point, therefore, it does not appear that women were the main
losers from migration.
6. Conclusions
The main theme of this study is the movement of labour out of
agriculture into urban labour markets, and the conditions and the
30 Kocka (1990), 465 gives Wgures for the gap between male and female earnings: Ritterand Tenfelde (1992), 215 also show that there was a gap in status at the workplace as well.On the other hand, mortality Wgures, as given in Appendix 5B, show that male lifeexpectancy fell by more when rural and urban rates are compared.
Table 8.11. Comparison of life expectancy in major Prussian cities to
expectancy in the rest of the province, calculated from mortality tables
for 1893–4
At birth At 1 year At 20 years
Men: Major cities 34.7 47.7 56.9
Rest of province 40.0 51.8 60.8
DiVerence �5.3 �4.1 �3.9
Women: Major cities 39.9 52.2 62.4
Rest of province 42.9 53.5 62.8
DiVerence �3.0 �1.3 �0.4
Source: Calculated from tables in Kucyzinski (1897); Wgures for individual cities givenin Appendix 5B.
Migration and Urban Labour Markets 283
![Page 293: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/293.jpg)
circumstances under which a ‘labour surplus’ is Wrst created and then
released. The preceding chapters have, therefore, concentrated on the
rural sector—on technological, demographic, and institutional develop-
ments within this part of German society. But the urban context is also
important. If the urban sector is under pressure—from capital shortage
or from adverse movements in the international terms of trade—then
the ability of the urban economy to absorb the rural surplus, without
putting downwards pressure on wage rates, is impaired.
From this point of view, there were two main phases in the develop-
ment of the industrial sector in this period. The Wrst lasted from the
1870s to the 1890s, and was characterized by favourable movements in
the terms of trade for the export sector. However, it was also a period
when the German capital markets were suVering from the after-eVects
of the Grunderzeit boom of the 1870s. This restricted access to capital at
a crucial time in German economic development.
The second phase began in the 1890s and lasted until the First
World War. In this period the capital markets still experienced periodic
crises, but these were relatively quickly overcome. Investment levels
were high, and only temporarily aVected by cyclical downswings. On
the other hand, the export sectors were adversely aVected by the move-
ment of the terms of trade. The recovery of world food and raw mater-
ial prices from the trough of the 1890s had a part in this, but German
industry was also having to cut prices to gain a share on world markets:
the price of German manufactures was falling relative to those of other
countries.
In the long run, Germany overcame this problem by moving up the
technological ladder, by developing new products rather than compet-
ing in established markets. This process was already under way before
1914. The Wgures for ‘high technology’ exports make this clear. This
created the prospect for rapid export growth without adverse move-
ments in the terms of trade, as would be the case for the Federal
German Republic after 1945. If world trade had not been disrupted by
the First World War and its aftermath, then the Kaiserreich might well
have moved into this third phase, which would have made it much
easier to achieve real wage improvements.
284 Migration and Urban Labour Markets
![Page 294: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/294.jpg)
Appendix 8A. Description of Regress ion Variables
and Sources
(i) Appendix Table 8B.1
Dependent variables For each industry: % change in numbers oc-
cupied, calculated from occupational census
data given in Kaelble and Hohls (1989).
Independent variables As above, % of the total occupied popula-
tion occupied in each industry.
(ii) Appendix Table 8B.2
As Table 8B.1, industry variables as indicated.
(iii) Appendix Table 8B.3
Dependent variable Logistic transformation, Yjk ¼ log (r=1� rjk), where rjk is the average annual
numbers of moves from region j to region k,
1909–13 divided by the occupied population
in region k in 1907. Movement data from
Reichsarbeitsblatt (annual issues) 1910–14;
occupied population from Kaelble and
Hohls (1989).
LOGDISTANCEkj The distance in kilometres between each
region measured from maps, being the
smallest distance between the frontiers. The
distance between contiguous regions was
entered as 10 kilometres, this being the value
which was found to reduce the coeVicient on
a contiguity dummy to zero.
POPREGORIGINj The total population of the region of origin
1907, data from Kaelble and Hohls (1989).
INDUSTRY%k As above, the percentage of the total occu-
pied population occupied in industry, 1907,
in the recipient region.
SERVICES%k As above, for services.
CHEMICALS%k As above, for chemicals.
MACHINERY%k As above, for Maschinenbau.
Migration and Urban Labour Markets 285
![Page 295: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/295.jpg)
METALWORK%k As above, for Metallverarbeitung.
CONSTRUCTION%k As above, for construction.
UTILITIES%k As above, for Versorgungsleistung (gas, water,
and electricity).
REGORIGIN%AGj As above, for agriculture in the region of
origin.
(iv) Appendix Table 8B.4
Dependent
variable (MIGt)
Data from SJDS (1907), 69, and subsequent
issues, combining annual results for 25
cities: total net moves in was divided by total
population. The selected cities were those
with complete sets of data for the period.
UNEMPLOYMENTt The annual average rate of unemployment
amongst trade union members, data from
Bry (1960), 325–6.
APPL=VACANCIESt Data from SJDS 1904–13, combining results
for 23 cities: total numbers of registered ap-
plicants at labour exchanges divided by the
number of vacancies, annual Wgures.
INDPRODUCTIONt Industrial production, data from Hofmann
(1965), 451–2.
Appendix 8B. Econometric Results
(i) Analysis of employment change by industry, 1895–1907
The changing structure of German industry had a direct eVect on
patterns of migration. As industry became more concentrated and more
export-orientated, migrants were drawn towards these areas. The ten-
dency to concentration can be demonstrated using the occupational
censuses. Table 8B.1 gives the results of a series of equations estimated
using the form Yki ¼ aþ bXki, where Yki is the percentage increase of
employment in industry k in region i in 1895–1907 and Xki is the
percentage of the total occupied population in region i occupied in
industry k in 1895. Any tendency towards concentration should pro-
duce a positive estimate for b.
286 Migration and Urban Labour Markets
![Page 296: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/296.jpg)
There was a problem with heteroscedastic errors in some of the
estimated equations. So alternative estimates were produced using a
two-stage weighted least squares procedure. There are few major diVer-
ences between the two sets of results, but the 2S WLS results are
somewhat clearer, due to the elimination of some nearly signiWcant, but
in fact rather spurious, results.31
31 The 2S WLS procedure follows the procedure set out in Greene (2000), 514–15.A weighting variable was calculated, which was wi ¼ 1=(Xi)
z for each industry, the value zbeing obtained by an interactive maximum likelihood procedure. This was then used inWLS estimation of the equation. The reason why this problem arose was that there weresome industries where employment was low and highly variable. In metal production, forexample, employment was often in single Wgures: 1 person was employed in Bromberg in1895, but 9 in 1907, an increase of 800%. Using OLS produced absolute errors stronglyrelated to the explanatory variable.
Table 8B.1 Regressions of change in employment 1895–1907 on the
percentage of the total occupied population employed in each industry
in 1895, estimated using ordinary least squares and a two-stage
weighted least squares procedure (81 regions)
Equations estimated using OLS Equations
estimated
using 2S WLS
constant b t-ratio R2 adj. R2 b t-ratio
All industry 18.04 0.406 (2.74) 0.087 0.075 0.574 (3.91)
All services 2.40 0.797 (4.26) 0.187 0.177 0.895 (4.00)
Mining and
quarrying
93.33 �2.74 (0.42) 0.002 0.000 �0.04 (0.00)
Metal
production
274.57 �70.4 (1.65) 0.035 0.022 �4.57 (0.92)
Metal-working 16.02 4.12 (3.50) 0.134 0.124 4.12 (3.50)
Machinery 64.16 28.1 (3.18) 0.113 0.102 40.2 (4.18)
Chemicals 82.69 �28.2 (1.14) 0.016 0.004 �1.24 (0.10)
Textiles 2.15 0.71 (1.17) 0.017 0.000 0.18 (0.54)
Clothing �9.65 4.11 (3.68) 0.146 0.135 5.44 (3.33)
Food and drink 20.46 1.29 (1.20) 0.018 0.005 1.16 (1.01)
Construction 52.00 �1.79 (1.22) 0.018 0.006 �0.47 (0.35)
Utilities 170.32 �164.3 (1.41) 0.025 0.012 �87.3 (1.73)
Other industry 10.26 1.47 (4.00) 0.169 0.158 1.34 (4.04)
Migration and Urban Labour Markets 287
![Page 297: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/297.jpg)
There is good evidence of a move towards concentration both in
industry and services. Analysis of the service sectors (not reported)
showed that this result was heavily aVected by the domestic service
sector.32 Within the industrial sector there are four categories which
show clear trends towards concentration: metal-working, machinery,
clothing, and ‘other industries’. The sector with the best evidence of a
move towards dispersion is utilities, although this is not a result sign-
iWcant at the 5% level.
The occupational censuses can also be used to look for evidence of
vertical integration. This analysis is reported in Table 8B.2. Correlation
analysis revealed a number of industries which were linked geographic-
ally. The highest value in 1895 was 0.83 for the correlation of employ-
ment in mining and metal production. But these industries were already
so well integrated that there was little scope for further movement in
this direction. Column a of Table 8B.2 shows that there was no sign-
iWcant eVect on the change in employment in metal production from
the percentage employed in mining in 1895.
Column e shows that there was little eVect from textile employment
on the change in the numbers occupied in the clothing industry: the
coeVicient on textile employment is positive but not signiWcant. Simi-
larly, the results in column b indicate that there was some eVect from
employment in metal production on the increase in employment in
metal-working, but the coeVicient was not signiWcant.The machinery industry was the one which showed the strongest
inXuence from other industries. As column c shows, the increase in
employment in this industry was inXuenced by the presence of metal-
working in 1895. In column d other variables are added. These results
indicate that the strongest eVect on the rate of growth of machinery
employment came not so much from the percentage employed in ma-
chinery in 1895, but rather from the presence of metal-working and
from employment in utilities. This last eVect may well reXect the in-
creased importance of electrical power, as the electrical industry was
included in this category. The machinery industry was moving to the
areas most advanced in the use of electricity.
32 Which suggests a ‘middle-class agglomeration’ eVect. However, this result is notunimportant: domestic service was an important source of employment opportunities forfemale migrants, and the availability of such opportunities was therefore a factor in migra-tion decisions by women.
288 Migration and Urban Labour Markets
![Page 298: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/298.jpg)
(ii) Industrial development and migration in 1909–13
The eVects of industrial developments on migration can be studied
using data provided by the various Landesversicherungsanstalten—the
oVices which administered accident and sickness insurance. When an
insured worker left one region for another, a Quittungskarte was ex-
changed between the regions. Total numbers of Quittungskartenaustauch
were published annually in the Reichsarbeitsblatt from 1910 onwards.
Table 8B.2. Regressions of change in employment 1895–1907 on the
percentage of the total occupied population employed in related
industries in 1895, estimated using OLS (81 regions)
Metal
production
Metal-working Machinery Clothing
a b c d e
Constant 281.12 17.01 30.84 31.12 �10.08
MINING95 �17.55
(0.63)
METALPROD95 �30.52 þ2.63 þ9.38
(0.40) (1.10) (1.45)
METALWORK95 þ3.47 þ14.96 þ9.96
(2.65) (4.36) (2.71)
MACHINERY95 þ18.27 þ8.76
(2.20) (1.01)
UTILITIES95 þ295.72
(3.96)
CLOTHING95 þ4.02
(3.57)
TEXTILES95 þ0.20
(0.78)
N 78 81 81 81 81
S 398.80 20.09 56.48 51.45 14.09
R2 0.040 0.148 0.287 0.424 0.153
adj R2 0.014 0.126 0.269 0.393 0.131
F-test of regression 1.56 6.76 15.73 13.97 7.04
RSS 11926363 31491.7 248817 201211 15480.2
Note: t-ratios in brackets. As before there were heteroscedasticity problems with columna; however, no corrective mechanism was found which produced signiWcant results forthis equation, so the results are not reported.
Migration and Urban Labour Markets 289
![Page 299: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/299.jpg)
This provides a more complete picture of annual movements than any
other published source. Even though data are only available for Wve
years, 1909–13, it still gives a valuable insight into the operation of the
German labour market in the years just before the First World War.
One reservation is that the data covers only workers who were already
insured. Younger migrants, who had not been insured, would not be
included.
In Table 8B.3 the linear log-odds model, used in Chapter 5, is ap-
plied to the Quittungskartenaustauch data. Here, the focus is on the
characteristics of the recipient region (k), and so the dependent variable
is a logistic transformation of the number of recorded moves into region
k from region j, divided by the total occupied population in that region
(from the 1907 occupational census).
The Wrst column gives the basic model. The probability of a move
into a given region is inXuenced by distance, by the size of the popula-
tion of the region of origin, and by the amount of non-agricultural
employment in the recipient region. The much higher coeVicient on
service employment conWrms that this sector was an important source
of opportunities for migrants. The percentage occupied in agriculture
in the region of origin was also entered, but this variable was not found
to be signiWcant and so was dropped.33
Columns b, c, and d show that, entered separately, the export-orien-
tated industries have a positive eVect on the probability of inward mi-
gration. Due to collinearity, when entered together only chemicals and
metal-working have signiWcant coeVicients. The Wnal column shows the
results of a step-by-step procedure, in which all eleven industries and
seven service sectors were entered. The procedure identiWed four in-
dustries as having a signiWcant eVect on migration (no service sectors
survived the procedure). These are chemicals, metal-working, and con-
struction, with positive coeVicients, and utilities, with a negative coeVi-
cient. The attraction of work in construction to migrants is not
unexpected: it oVered easy entry into a relatively unskilled urban occu-
pation. Moreover, the transitory nature of the work meant that in any
given period a number of workers would be on the move, as projects in
one area came to an end and others started up elsewhere. The negative
coeVicient on utilities suggests that few jobs in this sector were open to
migrants.
33 This might show that agricultural regions were no longer major sources of migrants,but as many migrants out of agricultural regions would not have had insurance cards, thisconclusion may not be justiWed.
290 Migration and Urban Labour Markets
![Page 300: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/300.jpg)
(iii) Migration and trade cycles 1904–12
Annual data for movements into and away from major cities can be
obtained from the published series of town migration Wgures produced
as a result of the system of MeldepXicht, compulsory registration after a
change of residence. For 1904–12 a series can be produced covering 25
cities, which includes all the major cities with the exception of Munich.
Table 8B.3. Regression results for linear log-odds migration model:
1909–13
a b c d e
Constant �5.685 �6.104 �5.830 �5.822 �7.723
LOGDISTANCEkj �0.454 �0.455 �0.451 �0.453 �0.451
(38.4) (38.9) (38.1) (38.4) (38.8)
POPREGORIGINj þ0.378 þ0.380 þ0.379 þ0.379 þ0.380
(16.29) (18.09) (17.97) (17.94) (18.32)
INDUSTRY%k þ0.0170 þ0.0138 þ0.0162 þ0.0164 þ0.0267
(5.62) (4.49) (5.37) (5.44) (5.21)
SERVICES%k þ0.0637 þ0.0633 þ0.0630 þ0.0571 þ0.0799
(15.94) (16.03) (15.83) (11.91) (6.35)
REGORIGIN%AGj �0.0003
(0.14)
CHEMICALS%k þ0.0762 þ0.0699
(3.11) (2.74)
MACHINERY%k þ0.0356
(2.46)
METALWORK%k þ0.0553 þ0.0510
(4.56) (4.00)
CONSTRUCTION%k þ0.0495
(4.38)
UTILITIES%k �0.7230
(2.34)
N 930 930 930 930 930
S 0.9393 0.9289 0.9344 0.9362 0.9169
R2 0.713 0.720 0.716 0.715 0.728
adj R2 0.712 0.718 0.715 0.714 0.725
F-test of regression 460.1 474.6 466.9 464.3 307.9
RSS 815.2 797.2 806.7 809.9 774.2
Note: t-ratios in brackets.
Migration and Urban Labour Markets 291
![Page 301: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/301.jpg)
Although there are only Wgures for nine years, it was still possible to
produce satisfactory econometric results (Table 8B.4) once Wrst diVer-
ences have been used (to deal with the high autocorrelation of residuals
shown in column a).
Various diVerent dependent variables related to economic conditions
were tried. Column b uses the change in unemployment; column d uses
the change in industrial production. The coeVicients on both are sign-
iWcant. The best results are those given in column c, possibly because
the applicants/vacancy ratio was obtained from the cities themselves,
and thus recorded conditions in these cities rather than conditions in
the economy as a whole. The results in column d are the weakest, and
the Durbin Watson statistic is within the inconclusive range
(d�L ¼ :773, d�U ¼ 1:332) for a sample of this size.
Table. 8B.4. Regression analysis of migration into major cities
1904–12
Dependent variable is:
MIGt DMIGt DMIGt DMIGt
a b c d
Constant 4.335 �0.187 �0.190 0.790
UNEMPLOYMENTt �1.320
(�3.88)
D UNEMPLOYMENTt �1.046
(�5.78)
DAPPL=VACANCIESt �2.616
(�7.44)
DINDPRODUCTIONt�1 �0.223
(�2.31)
N 9 8 8 8
S 0.568 0.341 0.274 0.636
R2 0.683 0.848 0.902 0.471
adj R2 0.638 0.822 0.886 0.383
F-test 15.08 33.4 55.29 5.34
RSS 2.259 0.698 0.449 2.424
DW 0.55 3.08 2.91 1.01
Note: t-ratios in brackets.
292 Migration and Urban Labour Markets
![Page 302: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/302.jpg)
9
Industrialization, Migration,
and Inequality
1. Introduction
The analysis contained in the preceding chapters has provided ample
evidence of the existence of labour surplus conditions in Germany in
the late nineteenth century. The sources of this surplus were various:
population pressure as a result of the demographic transition; the
release of underemployed labour in agriculture and cottage industry;
the increased use of seasonal migrant labour in agriculture; techno-
logical change which raised agricultural productivity; altered social rela-
tions inside agriculture as a result of changed economic circumstances.
The combined impact of these changes was to create a period when
labour supplies were highly elastic. High levels of rural–urban migra-
tion provided the motor for rapid economic growth. But this also meant
that Germany faced social and economic problems which were unpre-
cedented in the experience of European industrialization.
This amounts to a qualiWed endorsement of Max Weber’s views
about the origins of the agricultural labour surplus. The main qualiWca-
tion is that the release of surplus labour was brought about by other
factors besides Weber’s mechanism involving the decline of the patri-
archialische Arbeitsverfassung due to the spread of capitalist values. But
the basic point, that there is surplus labour within agriculture and that
this surplus is released as a result of changes associated with industrial-
ization, is endorsed.
Comparison with other countries shows that the German experience
was unique. No other European country had such a rapid transition to
an urban industrial society. The rate of growth of the major cities was
unusually high; the numbers migrating internally was also above the
levels in other countries. Only the United States went through a similar
experience, but this was in a very diVerent context, where most
migrants were immigrants.
![Page 303: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/303.jpg)
Intertwined with these problems were other sources of diYculty. As
a late-industrializing country, with a massive need to import food and
raw materials, Germany had to expand exports rapidly, gaining share at
the expense of other countries. This meant that wages had to be kept
down. Germany’s position in central Europe made it a natural destin-
ation for migrants from other countries in central and eastern Europe.
This connected Germany to other ‘labour surplus’ countries and pro-
longed the labour surplus period in German industrialization. The re-
quirements of industrial growth also placed strains on the German
capital markets, resulting in the Wnancial crisis of the 1870s. A ‘capital
shortage’ phase coincided with the period of labour surplus.
The implications of the labour surplus period are set out in the
propositions of the Lewis Model: growth will be high as under-
employed labour is transferred to productive occupations; wages will
tend to lag behind productivity in the advanced or industrial sector; the
proWts share will rise; the beneWts of industrialization will go dispropor-
tionately to the owners of capital. Inequality in the urban sector will be
high and remain so until the labour surplus phase is over, when the
‘turning point’ is reached.
As noted in Chapter 1, the Lewis Model is one possible explanation
of the Kuznets Curve: the prediction that inequality will rise during the
initial stages of industrialization, and then fall as the economy reaches
maturity. Bringing together the views of Weber, Lewis, and Kuznets
produces a coherent model of economic development: the impact of
industrialization and the spread of capitalist values brings about the
release of surplus labour from agriculture; this holds back wage
increases in industry and raises the proWt share; which in turn leads to a
rise in inequality in the economy as a whole.
This chapter examines the evidence of a rise in inequality, making
particular use of Prussian tax statistics. This is an unusually valuable
source, which makes it possible to study the evolution of income
inequality in an industrializing society over a long period, something
which is rarely possible even with modern statistical materials.1
1 An important point is that there were no conXicts long enough to aVect the distribu-tion of income in the 1815–1914 period, something which makes it hard to draw generalconclusions from the movement of inequality in the major industrial countries in thetwentieth century.
294 Industrialization, Migration, Inequality
![Page 304: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/304.jpg)
2. The Kuznets Curve and German H i story
The idea that, during industrialization, the relationship between growth
and inequality follows the path of an inverted ‘U’, with inequality Wrst
rising and then falling, has long been associated with Simon Kuznets,
whose 1955 article gave it wide currency. However, the idea that the
distributional consequences of industrialization might not be favourable
was certainly not a new one. Apart from its role in the socialist critique
of capitalism, it was also a concern of the German historical school, and
discussion of income distribution was an important component in the
German debate over Manchestertum: the pros and cons of laissez-faire
and free trade economics. Evidence that the income distribution was
worsening fuelled fears that the lower and middle classes were not
beneWting from industrialization to the same extent as the rich.
Kuznets was inXuenced by the same evidence, drawn from Prussian
tax records. The Soviet economist Prokopovitch had used these in a
study of income distribution, and, at the time of the 1955 article, this
was the only empirical support that Kuznets had for the initial rise in
inequality.2 As a result of Prokopovitch’s work, Kuznets was aware of
the diVerence between rural and urban inequality in nineteenth-century
Prussia. He found a similar pattern in studies of inequality in the
United States and India.3 This led him to a more general statement of
the relationship between inequality and economic development.
Migration had a central role in Kuznets’s own explanation of his
‘inverted U’ relationship. Kuznets’s model had three parts:
1. There are two sectors A (rural) and B (urban); incomes are higher
in B; there is migration from A to B;
2. Income inequality is higher in B;
3. Income inequality in B will tend to fall.4
Kuznets showed, using mathematical examples, that under certain
assumptions the Wrst part, on its own, could produce a ‘Kuznets
Curve’. This eVect was intensiWed by the second part. But Kuznets also
thought that income inequality in the three countries he was consider-
ing (Britain, Germany, and the United States) had in fact stabilized
while the Wrst two components were tending to cause it to rise, and he
2 Prokopovitch’s work was published by Keynes in the Economic Journal, Prokopovitch(1926).
3 Kuznets (1955), 7; the German Wgures came from Prokopovitch (1926).4 Kuznets (1955), 7–8 for parts 1 and 2, p. 17 for part 3.
Industrialization, Migration, Inequality 295
![Page 305: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/305.jpg)
attributed this to the third component (for which he had no direct
evidence).
The Lewis Model provides an explanation for parts 2 and 3 of Kuz-
nets’s model: it explains high inequality in sector B, and it also predicts
that this will fall once the labour surplus phase is over.
Later Kuznets made a cross-section study of inequality in developing
countries which appeared to support his hypothesis. However, this
result has been disputed, and more recent studies of contemporary
evidence from developing countries have tended not to show an
inverted U-shaped curve, so that one participant at a conference was
led to declare: ‘the Kuznets curve is a Wction’.5
If the Kuznets Curve is a contemporary Wction, then this leaves some
questions. Does the historical evidence for the curve stand up? And, if it
does, how could it be that the relationship held for some industrializing
countries in the nineteenth century and not for their twentieth-century
successors? Is the Kuznets Curve ‘historical fact, but contemporary
Wction’? Is it just a ‘peculiarity of German history’?
Since Kuznets wrote his article some additional evidence of high
inequality during the early stages of industrialization has emerged. Lin-
dert and Williamson concluded, with reference to the United States in
the mid-nineteenth century, that ‘there apparently was a Kuznetsian
fall from grace as America embarked on the path of modern economic
growth’.6 Charles Feinstein has reinforced the pessimistic view of
working-class living standards during the British industrial revolution:
‘the majority of the working class . . . had to endure almost a century of
hard toil before they really began to share in any of the beneWts that
they had helped to create’. In the British context this was a continuation
of a pattern of high inequality already established in agriculture.7
In the next section of this chapter, new estimates of inequality,
derived from Prussian tax statistics, are presented which show that
Kuznets’s ‘inverted U’ hypothesis is relatively well supported for the
1820–1914 period. The peak of the curve was reached around 1900.
Thereafter inequality tended to decline. So, the Kuznets Curve is not a
Wction as far as nineteenth-century Germany is concerned.
5 Dani Rodrik, the discussion was published in Bruno et al. (1996), the remark is onp.159. For examples of the studies which led to this remark, see Anand and Kanbur(1993), Ravaillion and Chen (1997), and Dollar and Kraay (2000).
6 Lindert and Williamson (1980), 284.7 Feinstein (1998).
296 Industrialization, Migration, Inequality
![Page 306: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/306.jpg)
If the Kuznets Curve has some basis in historical fact, then this
suggests that it is one possible outcome of the process of industrializa-
tion, but not a relationship set in stone. Given favourable circum-
stances, some countries can avoid the initial rise in inequality; others,
facing less favourable conditions, may not experience the subsequent
fall postulated by Kuznets. Which leaves the question, what are these
favourable or unfavourable factors? In Germany’s case, what were the
reasons for the existence of a Kuznets relationship in the 1820–1914
period?
3. The German Debate
Germany, as one of the second wave of industrializing powers, could
observe the eVects of industrialization on British society and draw con-
clusions. There were horriWed reactions from German visitors ranging
from Friedrich Engels to Richard Wagner, who saw London as ‘the
fulWlment of Alberich’s dream—Nibelheim, world dominion, activity,
work, everywhere the pressure of steam and fog’.8
The German debate began once it was clear that Germany was
industrializing, and facing similar problems. The eVect of industrializa-
tion on income distribution was one aspect of this debate. Gustav
Schmoller, writing in 1895, summarized the position as follows:
Optimistic free-traders and defenders of the current social order have certainly
asserted that economic freedom and the current organisation of society will, on
their own, guarantee an increasingly equal distribution of income, while the
historical school, the more pessimistic critics of our time, and above all the
socialists have, in the main, come to the opposite conclusion.9
As this quote indicates, concern over the movement of the income
distribution was not limited to the left. A wide range of opinion feared
that industrialization would lead to widening inequality unless govern-
ment action was taken to prevent this. Considerable impetus was given
to the pessimistic view by an article by Ernst Engel, the director of the
Prussian Statistical OYce, published in the Zeitschrift des Koniglich
Preußischen Statistischen Bureaus in 1875. This contained an analysis
of the Prussian tax statistics which appeared to show that the beneWts of
8 Following a boat trip down the Thames to Greenwich in 1877, C. Wagner (1978–80).9 Schmoller (1895), 1067–8.
Industrialization, Migration, Inequality 297
![Page 307: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/307.jpg)
industrialization were going, in the main, to the higher-income groups.
Engel speciWcally referred to the way that the data seemed to support
Schmoller’s concerns: ‘There is no doubt that the concentration of
wealth has increased.’10
The historical school placed much more emphasis on the role of insti-
tutions, particularly the state, in the determination of income inequality.
Schmoller argued that there was a tendency for income inequality to
rise but that state action could counteract or even reverse this. This
was a small, but still signiWcant, breach with free market economics.
Although the historical school was not a large group, they were very
active, founding journals and producing surveys (including many
used as sources for this study), so they had a considerable eVect on
opinion.11
The view that the income distribution was worsening was challenged
by Soetbeer, who argued that the lower classes were in fact doing better
than Engel had thought. Engel was using income tax Wgures at a time
when the Prussian tax system was a hybrid, with the lower-income
groups paying a Klassensteuer, a tax which was intended to reXect ability
to pay, but which may have been less representative at times of rising
money wages. Soetbeer made a more systematic analysis of the income
distribution, which was less aVected by changes in the tax system. Adolf
Wagner, writing in 1904 but using tax data from 1822 onwards, took
the view that both the upper- and the lower-income groups were doing
well, but at the expense of the middle classes.12
In 1913 HelVerich made an analysis of developments since 1896
which came to the conclusion that, ‘Contrary to the much asserted view
that society is developing in a plutocratic direction, the lower-income
classes are gaining their full share of the increase in the national
income’.13 HelVerich had made a full analysis of the national income
and its distribution. His work was one of a series which made use of the
data produced by the revised Prussian tax system, which had been in
operation since 1892. This was showing a declining trend in inequality,
at least amongst taxpayers. It had also stimulated a discussion of the
10 Engel (1875), 142.11 For discussion of Schmoller’s views, and the role of the historical school, Dumke
(1991) and Grimmer (1999).12 Soetbeer (1882) and Wagner (1904). Soetbeer was a statistician who published on a
number of issues, including monetary reform.13 HelVerich (1915), 128. Karl HelVerich was a director of the Deutsche Bank and
secretary of the Treasury during the First World War.
298 Industrialization, Migration, Inequality
![Page 308: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/308.jpg)
statistical problems involved in the construction and analysis of income
distributions.14
Despite this progress, there are serious problems with the work of
these authors. One is the lack of an agreed standard for the measure-
ment of inequality. A second is a tendency to underestimate gains by
those below the income tax threshold. Both Soetbeer and HelVerich put
in unchanged Wgures for average incomes in this group. Yet, if an
income distribution is shifted upwards, so that the numbers falling
below some Wxed threshold falls, the average income of this group will
tend to rise unless there is some balancing shift within the exempt
group.15 A third problem is that comparisons are often between single
years, and over short time periods. A much longer view, less dependent
on estimates for single years, is needed.
4. The Pruss ian Tax Data : Problems and Opportunities
There are three main problems with the use of tax data to examine
income inequality. First, part of the population is normally exempt
from the tax, due to income levels below the tax threshold, or some
other reason for exemption. So, coverage is inevitably incomplete. Sec-
ondly, as the income assessment is linked to the tax payment, there is
an obvious incentive to minimize declared income and conceal sources
of income wherever possible. Some downwards bias must be antici-
pated. Thirdly, the tax system tends to change over time, which has
eVects on coverage, on incentives to evade, on deWnitions of income, on
procedures for checking returns, and so on.
These are serious problems, but not insuperable. There are two ways
of dealing with the Wrst problem. One is to divide the income received
by diVerent groups of taxpayers by total personal income, obtained
from national income accounts. This is an approach used in modern
studies by Atkinson and Piketty.16 It has been applied to the Prussian
income tax data by Geisenberger and Muller.17 There are, however, a
14 See Bresciani (1907) for a discussion of the statistical issues; this article anticipatesmany of the techniques used in modern studies, including this one. Other studies ofincome (and property) inequality include Kuhnert (1906 and 1916) and Biederman (1918).Constantine Bresciani (or Bresciani-Turroni) was an Italian economist who studied inBerlin in 1902–7.
15 A point which can be established by simulation using a log-normal distribution.16 Piketty and Saey (2001), Atkinson (2002).17 Geisenberger and Muller (1972).
Industrialization, Migration, Inequality 299
![Page 309: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/309.jpg)
number of problems with this approach. The main one is that it can
only be used for periods for which national income accounts are avail-
able, so the Geisenberger and Muller study starts in 1874. The second
is that the income side of the national income accounts is generally the
least reliable. Data on physical production is normally available from an
earlier stage than information on wages, rents, and proWts. HoVmann’s
estimates of income from capital and land in the nineteenth century
were derived using methods which have come in for criticism.18 More-
over, for a series which is derived from Prussian tax statistics, the
divisor should be based on income accounts for Prussia, and these are
problematic, as there is no information on inter-regional trade or trans-
fers of property income (post-1871 Prussia contained 61–62% of the
total German population).19
An alternative approach (which was the one used to prepare the
estimates given in this study) is to Wnd other sources of information on
the incomes of those excluded from the tax system. In nineteenth-
century Prussia, the most important low-income group was that con-
taining the landless agricultural labourers. Information on these, and on
incomes in low-paid non-agricultural occupations, can be used to con-
struct an overall distribution. This in turn can be used to derive a Gini-
coeYcient, or other measure of general inequality, whereas the Wrst
approach only yields a measure of the share of top taxpayers in total
income.
The second problem, the problem of evasion, is inherent in the use
of income tax data.20 It is a serious source of possible error, particularly
in cross-section comparison with other countries (where enforcement
procedures may be diVerent). It is a problem in a time series for a single
country if there are changes in the proportion of income which is not
declared, as a result of improved methods of oversight, for example. It
is clear from contemporary comment on the tax system that there were
18 Problems with HoVmann’s Wgures are discussed in Fremdling (1988 and 1995).19 The best income distribution Wgures are those derived from the Saxon tax statistics,
Jeck (1970); this source has also been used by Hentschel (1979) and Kiesewetter (1988).However, Saxony was a small and not very typical part of Germany, which industrializedearly and had a relatively equal land distribution. Given the importance of Prussia, with62% of the population in 1907, it is clear that any study which attempts to draw conclu-sions for Germany as a whole must make use of the Prussian data. More recently Germanstudies have tended to look at other aspects of inequality, such as social mobility—seeKaeble and Federspiel (1990) and Kocka (1983)—and there has been little work onincome distribution. Kaelble (1986) surveys work up to that date.
20 For a more detailed examination of the pros and cons of tax data as a source ofinformation on income distribution, see Atkinson (2002).
300 Industrialization, Migration, Inequality
![Page 310: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/310.jpg)
changes of this type.21 The reform of the Prussian tax system in 1891–2
led to a more rigorous examination of tax returns. It is important to
make some allowance for these changes.
The tax system was not constant in Prussia in the period, but evolved
in stages from a Klassensteuer, under which households paid a Wxed
charge, reviewable every three years, based on an assessment of ability
to pay (which took account of property owed, and indebtedness as well
as expected income), to a modern income tax system. Comparability
between the diVerent systems is not to be expected.
These last two problems were addressed by producing estimates of
income inequality for periods when the tax system was not subject to
major changes (1822–46, 1853–73, 1874–91, and 1892–1914), and then
looking for contemporary estimates or other data sources to chain the
series together. This proved to be possible for 1873–4 and 1891–2, but
not for 1846–53. During this period the Klassensteuer went through a
series of changes: urban areas which previously paid a diVerent tax were
brought into the system and the top bands were changed to a system
more explicitly linked to income. It was not possible to Wnd evidence
from other sources which could cover this period.
In addition to selecting an approach which should minimize potential
sources of error, an attempt has been made to quantify the eVect of
those errors which remain. By constructing diVerent variants for each
period, based on alternative sets of assumptions, it was possible to
examine the sensitivity of the Wgures to the assumptions used in the
central estimates, and Wnd extreme values produced by a range of pos-
sible assumptions. These variants were then used to calculate variance
estimates which were, in turn, employed to construct conWdence inter-
vals for the central inequality measures.22 Appendix A contains the
resulting estimates and conWdence intervals.23
The approach is one which is appropriate for a data source which has
serious deWciencies by modern standards. It generates a range of prob-
abilities as well as a single set of point estimates. Taking as an example
the weakest period, 1822–46, the tax system in these years was only
21 Meisel (1911) gives data on the number of returns rejected or challenged under thepost-1892 system. On average, a challenge to an assessment led to an increase of 25–30%in the amount of taxable income. See also Wagner (1880), 612, on the pre-1892 system.
22 This follows the approach set out in Feinstein and Thomas (2001).23 A full discussion of the issues involved in these calculations, together with details of
the calculations and references to the sources used, has been published separately in Grant(2002b).
Industrialization, Migration, Inequality 301
![Page 311: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/311.jpg)
indirectly related to income and coverage was not universal. Neverthe-
less, the data does provide certain pieces of evidence. The numbers in
Stufe 1, the highest band, rose by 60% between 1822 and 1846, while
the total population in the areas subject to the tax rose by only 39%.
The numbers exempt from the tax on the grounds of having insuYcient
property or income rose steadily, from 36.6% of the population in 1822
(including dependants) to 41.5% in 1846.24 These two factors strongly
suggest a hollowing out of the distribution, with numbers rising at each
extreme.
Moreover, there is evidence from other sources which suggests that
real wages for those at the bottom of the distribution were falling in this
period. Table 9.1 gives an index derived from Wgures for earnings in the
Prussian state forests, deXated by local food prices.
Taking these factors into account, the balance of evidence strongly
suggests a worsening of the income distribution in the period. But by
how much? This is less easy to say. The variance calculations produce a
95% conWdence range for the increase in the Gini-coeYcient of 12.2% to
16.2%, assuming that the central estimates are subject to uncorrelated
24 The Wgures of those exempt have been adjusted to allow for the eVects of anextension of exemptions to those aged 14–16 and over 60 in 1831.
Table 9.1. Rural wages in Prussia 1820/9–1875/9
(post-1815 boundaries)
1820–9 104.6
1830–9 101.1
1840–9 88.7
1850–9 81.9
1860–9 91.4
1870–4 91.3
1875–9 100.0
Note: Index based on earnings of male day-labourers in thePrussian state forests (1875–9 ¼ 100), deXated by local food priceindices.Source: Earnings in Prussian state forests come from Eggert (1883);this gives the results produced by a circular sent to all Oberforsterin November 1879 asking for the movement of wages since 1800.The survey was not repeated. Prices were taken from ZdKPSB(1907), 84–92.
302 Industrialization, Migration, Inequality
![Page 312: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/312.jpg)
errors. But examination of the highest and lowest variants (considering
the possibility that there are correlated errors produced by incorrect
assumptions), and applying variance calculations to these estimates,
produces a much wider range: an increase of between 4.3% and 21.4%.
This demonstrates that the statement that inequality worsened in the
period can be made with a high degree of certainty, even though the
exact scale of the increase is less easy to determine.
Thus, for all these caveats and notes of caution, the Prussian tax data
does present a major opportunity, provided that the data is handled
with caution. For 91 years, 1822–1913, the Prussian population paid
some form of income-related tax. Coverage was, at times, remarkably
high (74% in 1891). The period covers a society going through a pro-
cess of transformation from an essentially rural, pre-industrial economy
in the 1820s to a position as one of the world’s most advanced industrial
economies in 1913. Moreover, and in contrast to the twentieth century,
there were fewer unexpected shocks to the income distribution: world
wars with concomitant rises in taxation and prices, major world depres-
sions on the scale of the 1930s, unanticipated periods of high peacetime
inXation as in the 1970s. It is a unique opportunity to study the evolu-
tion of inequality in an industrializing society.
5. The Kuznets Curve in N ineteenth-Century Pruss ia
(i) Estimates for Prussia
Figure 9.1 gives the base estimates chained together to produce an
index. Because of the gap between 1846 and 1853 there are two base
years, 1846 and 1853. The estimated Gini-coeYcients are converted to
indices with these years set to 100.25
The shape is approximately that of the Wrst part of Kuznets’s
‘inverted U’. There was a downwards movement in inequality after a
peak in the early 1900s, though the extent of the fall was not large (a fall
of just 5% in the average Gini-coeYcient estimated for 1910–14 com-
pared to 1903–7). The main downwards phase of the Kuznets Curve for
25 The Gini-coeYcient is a measure of the extent to which the income distributiondiVers from perfect equality, all divergences being given equal weight. It is, in eVect, anaverage of the absolute values of all deviations from the average income level, expressed ona scale of 0 to 1, where 0 is perfect equality (all incomes are the same) and 1 is perfectinequality (one person has all the income).
Industrialization, Migration, Inequality 303
![Page 313: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/313.jpg)
Germany, as for other industrialized countries, came in the twentieth
century. Modern estimates give a Gini-coeYcient for Germany in the
1990s of 0.25, which is approximately the same level as the chained
estimate (using 1892–1914 as the base estimates and making no allow-
ance for changes in 1846–53) for Prussia in the 1820s.26 So the Prussian
or German Gini-coeYcient appears to have risen 35–40% in the nine-
teenth century, and fallen by a similar amount in the twentieth. This is
in line with Kuznets’s hypothesis.
In the Wrst period, 1822–46, industrialization was just beginning, and
Prussia was still a predominantly rural society. In these years, the main
inXuences on the level of inequality came from developments within the
rural sector. These include the eVects of the Prussian land reforms, the
consequences of a relatively high level of population growth, leading to
a period of quite severe demographic pressure, and the eVect of ruralexpansion in the east, bringing into cultivation land in remote areas
which were not well suited to market-orientated peasant agriculture. In
addition, agricultural prices fell in the 1820s and 1840s, which would
have disadvantaged the sector as a whole, without necessarily leading to
a worsening of inequality within agriculture.
26 Atkinson et al. (1995), 40.
1912190218921882187218621852184218321822
130
120
110
100
90
80
Fig. 9.1. Estimates of inequality from Prussian tax data 1822–1914: indices
of estimated Gini-coefficients (1846/1853¼100)
304 Industrialization, Migration, Inequality
![Page 314: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/314.jpg)
The second period saw the onset of a period of more rapid growth in
Germany. The start of more widespread industrialization is generally
dated to the 1850s. The tax statistics analysed by Engel seemed to show
that the fruits of industrialization were not being equally distributed.
Engel’s analysis was probably too pessimistic. The tax system in this
period was a hybrid. The richer taxpayers were subject to an income
tax, which was responsive to increases in earned and unearned incomes;
the less well-oV paid a revised version of the Klassensteuer. This was a
period in which, according to available evidence, wages were rising in
industry and agriculture. This may have led to a lag in the adjustment
of Klassensteuer rates to improved earnings. More importantly, it means
that any estimate of the overall income distribution should allow for
some improvement in the incomes of those below the tax threshold.
The base estimates for this period show an increase in the Gini-
coeYcient of between 4.3% and 6.0%, but the systematic error variants
show a much wider range, from 20.2% to 3.5%. This uncertainty is
largely caused by the question of the extent to which the incomes of the
exempt population moved up with other incomes. However, this period
is also identiWed as one when inequality most probably moved upwards.
The fruits of the early stages of German industrialization were not
evenly distributed.
In the next period, 1873–82, there was a change in the terms of trade
which was unfavourable to agriculture. The widening gap between agri-
cultural and non-agricultural earnings had the eVect of pushing up
inequality. As the exempt population contained large numbers of land-
less agricultural labourers, the assumptions made with regard to this
group have a signiWcant eVect on the level of overall inequality. The
base estimates show an increase in the Gini-coeYcient of between
11.6% and 13.3%. The systematic error variants show a range from
13.6% to 5.7%. The lower Wgure comes from a variant which is based
on a more optimistic view of the movement of real earnings in agricul-
ture in the period.
So the Wrst three periods studied, covering 1822 to 1892, less the
omitted six years in 1846–52, all show strong evidence of rising inequal-
ity, even when allowance is made for possible sources of error or bias.
The upwards phase of the Kuznets Curve is relatively established for
nineteenth-century Prussia. There were a number of diVerent factors at
work, but the general trend was towards greater inequality.
The Wnal period, 1892–1914, can be divided into two: a Wrst part inwhich inequality may have continued to rise, and a second one in which
Industrialization, Migration, Inequality 305
![Page 315: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/315.jpg)
inequality began to fall, albeit slowly and uncertainly. The break came
around 1906. Given the nature of the data, there are considerable diY-
culties with annual movements in the inequality measures. The main
problem is the estimation of the movement of the income levels of the
exempt. There are a number of possible wage series, but these show
diVerent patterns of cyclical movements.27 Thus, while the base esti-
mates for the period 1892–1906 show an increase in the Gini-coeYcient
of 5.5–6.0%, the range of possible variants runs from þ8.3% to
�1.3%. The period is best regarded as one of relative stabilization,
with Xuctuations around a core level which may have still been moving
upwards. The exact timing of the peak is uncertain, but it probably lay
in the early 1900s.
In this period, the recovery of the German capital markets, combined
with the ending of the main period of labour surplus migration out of
agriculture, produced an improvement in conditions in urban labour
markets. The increase in real wages still lagged behind productivity,
but this owed more to the adverse movement of the terms of trade. As
German industry forced its way onto world markets, it had to keep
prices low, and this restricted wage rises.
From 1906 onwards there was a slight decline. The base estimates for
1906–14 show a fall of 3.1–3.4%, with a range for the other variants of
between �5.4% and �2.7%. There was a cyclical rise in inequality in
1913–14 which reduced the overall decline to 1914.28 But, making some
allowance for these Xuctuations, the evidence for this period indicates
that the downward phase of the Kuznets Curve had begun.
Although the timing of some of these movements can be uncertain,
the overall trends are clearer. Throughout most of the nineteenth cen-
tury inequality was rising. In the 1890s this upwards movement came to
an end, and, after a period of Xuctuation, with no strong trend upwards
or downwards, inequality began to fall, although the fall before 1914
was a slight one.
(ii) Comparison with other countries
To get some idea of the scale of inequality in Prussia in the period, it is
necessary to make comparisons with other countries. This is the only
27 The main wage series are those given in HoVmann (1965), Desai (1968), Bry (1960),and Hohls (1995).
28 The tax system worked in arrears, so the assessments for 1914 were not aVected bythe outbreak of war in that year (as would have been the case with a PAYE system).
306 Industrialization, Migration, Inequality
![Page 316: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/316.jpg)
way to create a standard of measurement, to decide whether a given
Gini-coeYcient is ‘high’ or ‘low’.
The peak Gini-coeYcient for Prussia in 1906 is estimated to have
been 0.34 (Table 9A.5 in Appendix 9A). The World Bank data set
measuring inequality in modern economies shows that only seven out of
the 70 developing economies had lower Gini-coeYcients than this. So,
even at the 1906 peak, Prussian incomes were more equally distributed
than in most developing countries today.29 On this basis the overall
level of inequality in Prussia in the period appears to have been rela-
tively low for a country going through the disruptions of industrializa-
tion. However, the increase over levels in the earlier part of the century
may well have been a factor creating resentment and pessimism over the
future distribution of the rewards of industrial progress.
Another point of comparison is the experience of other European
countries in the late nineteenth century. Figure 9.2 compares the Gini-
coeYcients estimated for Prussia with estimates for Britain given by
Charles Feinstein. The Prussian estimates are chained backwards from
the estimates for 1892–1913 (which are the most reliable). ConWdence
intervals are also shown (from Appendix 9A). It is not possible to apply
the same procedure to the British Wgures, but, given the nature of the
data, it can be assumed that the conWdence intervals would be larger
than the Prussian ones, even those given for the 1822–46 period.
It appears that the movement of inequality in Britain and Prussia
followed diVerent paths. Inequality in Britain was consistently high,
and not much aVected by industrialization. Estimates produced by
Peter Lindert show a rise in inequality in the late eighteenth century,
but Britain was already a highly unequal society before industrializa-
tion.30 The estimated Gini-coeYcient for 1913 (0.48) would put Britain
in the top half of the World Bank data set, with a higher level of
inequality than 45 of the 70 developing economies.
This is not surprising given the much less equal distribution of land
ownership in Britain. Industrialization brought about a transfer of
population out of an agricultural sector which was itself highly unequal.
The fact that inequality did not rise as a result of industrialization was
due to the fact that inequality was just as high inside agriculture as
outside.
29 Deininger and Squire (1996), table 1; the calculation excludes Eastern Europe andthe former Soviet economies. There are massive problems of comparability in an exerciseof this type. For a discussion of some of these see Atkinson and Brandolini (1999).
30 Lindert (1994).
Industrialization, Migration, Inequality 307
![Page 317: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/317.jpg)
The Prussian experience was very diVerent. Inequality in the early
part of the century was relatively low, reXecting the more equal distri-
bution of land ownership. But industrialization was accompanied by an
increase in overall inequality. There are a number of possible reasons
for this. There was an increase in the number of landless labourers in
the early part of the century, at a time when real wages in agriculture
were falling. This could have been responsible for part of the increase
in inequality in the 1822–46 period. There was a rise in the urban–rural
wage gap in the 1870s, which would have increased overall inequality in
this period. Both these factors were, however, relatively short-term
eVects. The longer trend needs a diVerent explanation.
19071892187718621847183218171802
0.6
0.5
0.4
0.3
0.2
0.1
Prussia
Upper 95% C I
Lower 95% C I
Britain
Gini-coefficients
Fig. 9.2. Estimates of inequality from Prussian tax data, 1822–1914 (including
calculations of upper and lower 95% confidence intervals), compared with
figures for Britain
Note: The 1846–53 gap is covered by assuming that the Gini-coefficient for 1846 was thesame as that for 1853. An important point to note is that the probability of a movementfrom the upper limit of the 95% range to the bottom is 1/1600, which means that theconfidence intervals for rates of change are tighter than might be assumed from thisgraph: see the calculations presented in Appendix 9A.
Source: British figures from Feinstein (1988), 723.
308 Industrialization, Migration, Inequality
![Page 318: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/318.jpg)
The most likely cause was a shift of population from a relatively
equal rural sector to a less equal urban or industrial sector. There are
two sources of information on inequality in the urban sector. One is the
tax data for the period after 1896. A breakdown into two sectors is
available for a number of years. These have been used to produce
estimates of inequality which are given in Table 9.2. Some caveats
should be noted. These are based on a much cruder breakdown than
that used for the estimates given in Appendix 9A, just six tax-paying
classes plus the exempt. Consequently margins of error are likely to be
rather larger than those given in the appendix. In addition, the same
Wgures for household size by income class are used for both sectors,
which is a possible source of bias.31
The results show that urban Gini-coeYcients were 28% higher than
rural on average. So, a shift of population from one sector to another
would have an eVect on overall inequality.32 The table also shows that
inequality in both sectors fell substantially in 1900–13. This, however,
would have been partially balanced by the eVect of the continuing shift
of population into the less equal urban sector.
A second source of evidence comes from budget surveys in urban
areas. Results from these are given in Table 9.3. Again there are caveats
to note. These are based on household frequency distributions with no
information about household size. This has the eVect of raising the
Table 9.2. Gini-coeYcients estimated for rural and
urban sectors from Prussian tax data, 1896–1913
Urban Rural
1896 0.429 0.351
1900 0.438 0.349
1905 0.396 0.321
1910 0.362 0.261
1913 0.355 0.265
Source: Income tax Wgures from Biedermann (1918); income of the exemptcalculated using HoVmann’s Wgures for wages in industry and agriculture.
31 Household size was deWnitely larger in agriculture, but this would only be a sourceof bias if the diVerence was not the same for all income classes.
32 Overall inequality is not, however, just a weighted average of the two sectors. Thesize of the gap between the sectors is also important.
Industrialization, Migration, Inequality 309
![Page 319: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/319.jpg)
estimated Gini-coeYcient by around 0.08 (many low-income house-
holds had few dependants). The surveys were undertaken at diVerent
times, using methods of recording and selection which could have
varied considerably. This may well explain some of the surprising
jumps in inequality shown for Hamburg, for example.
Allowing for these reservations, the results still show a pattern of
inequality substantially higher than the estimates given for the rural
sector. They indicate that rural–urban migration would have been a
powerful force driving up overall inequality. And, more broadly, they
support the view that Germany went through the ‘labour surplus’ phase
of the Lewis Model in this period, and that this was driving up inequal-
ity in urban areas. These were the circumstances under which the
urban voters started to turn to the Social Democratic Party.
6. The Kuznets Curve : Contemporary F ict ion?
H i storical Fact? German Peculiarity?
If the Kuznets Curve was a product of historical circumstances, which
may be speciWc to Germany, this leaves a question: what was it that
produced the relationship in this period and not in others?
Two sector models, connected by migration, have a certain value in
the elucidation of points such as the importance of migration costs and
the signiWcance of expectations. But they also obscure other points,
such as the role of technological change within sectors. They focus
attention on changes in the relative importance of the diVerent sectors,
not on changes within sectors.
Looking at the process of development in the more advanced econ-
omies over the past 200 years, this can be summarized as an alteration
Table 9.3. Gini-coeYcients calculated from household budget
surveys
Hamburg 1867–8 0.651 Leipzig 1875 0.643
Hamburg 1873–4 0.467 Leipzig 1885 0.613
Hamburg 1881–2 0.720 Leipzig 1900 0.605
Hamburg 1890–1 0.534 Dresden 1880 0.559
Hamburg 1900–1 0.496 Breslau 1880 0.542
Breslau 1900 0.595
Source: Calculated from survey results given in Kuczynski (1908).
310 Industrialization, Migration, Inequality
![Page 320: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/320.jpg)
of the production function for the economy as a whole. The pre-indus-
trial production function was one in which, by deWnition, the level of
productive knowledge was low, and the stock of tangible capital was
also low. Physical eVort played an important part in the production
process, and the other major factor was the quantity of suitable agricul-
tural land. It might also be noted that the productive knowledge which
was available was rarely covered by patent rights, and so there were few
‘owners’ of such knowledge able to charge for its use: medieval har-
nesses for draught animals were major technological improvements but
nobody held a patent.33
The modern industrial or post-industrial economy is one in which
agricultural land has little importance in the overall production func-
tion. Physical eVort is also relatively unimportant. The stock of tangible
capital is much higher. There has also been a massive increase in
the level of productive knowledge, and a substantial part of national
income now goes to those who possess this knowledge, as skills, patent
rights, experience gained through learning-by-doing, or unique capabil-
ities possessed by Wrms and other institutions.34
The modern debate about income distribution mainly revolves
around the distribution of skills and the relative rewards received by
skilled and unskilled labour. Ways to improve the overall income distri-
bution include the provision of training for those at the bottom of the
income distribution, and the achievement of high and uniform stand-
ards in education. The modern German vocational training system,
which had its roots in the pre-1914 period, is generally seen as a par-
ticularly good example of these measures.35 German vocational training
was already admired and studied by other countries in this period. So,
if the wider diVusion of skills is indeed a major factor tending to bring
about a more equal distribution of income, then this factor was already
operating in pre-1914 Germany.
There are some reasons to expect that skill acquisition and diVusionmay lag during the earlier stages of industrialization and migration has a
part in this. Vocational training typically takes place immediately after
the end of formal education, which was around the age of 14 in
Germany in the period. Relatively few migrants had arrived at their
destination by this age. Only 13.9% of those arriving in Berlin in
33 Mokyr (1990), 36–8.34 The role of human capital in the movement of the income distribution is discussed
in Atkinson (1996a).35 Nickell (1996).
Industrialization, Migration, Inequality 311
![Page 321: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/321.jpg)
1881–5 were aged less than 14 at the time of arrival.36 Migrants arriving
later would Wnd it more diYcult to obtain formal vocational training.
They would be too old to be accepted as apprentices. Employers might
be more likely to oVer training to the children of resident workers than
to ‘footloose’ migrants. Moreover, from the point of view both of em-
ployer and of employee, the prospective rate of return on investment in
vocational training would fall with age, as the anticipated period of
future employment was reduced.
Migrants could receive training prior to migration, but this would be
less valuable. Most vocational training involves a practical element,
training ‘on the job’. This is only possible in the vicinity of the industry
concerned: shipyard apprenticeships will only be available in towns
where there are shipyards. Even more general courses are less likely to
be available in sparsely populated rural areas. The establishment of a
course on electrical engineering is only possible when there are a suY-cient number of prospective students in the catchment area to make the
course a viable proposition. Such agglomeration eVects will tend to raise
skill levels in cities and the areas around them; rural migrants will not
have the same opportunities.
Kuznets made a similar point, though with less emphasis on skill
acquisition:
The very fact that after a while, an increasing proportion of the urban popula-
tion was ‘native’, i.e. born in cities rather than in the rural areas, and hence
more able to take advantage of the opportunities of city life in preparation for
the economic struggle, meant a better chance for organisation and adaptation, a
better basis for securing greater income shares than was possible for the newly
‘immigrant’ population coming from the countryside or abroad.37
The consequence of this is that there will be a delay after the onset of
industrialization before skill diVusion starts to aVect the distribution
of income. As for the distribution of rewards to holders of other forms
of productive knowledge, such as patent rights, in the modern economy
this knowledge is more likely to be held by corporations than individ-
uals, and so this is bound up with the broader question of the distribu-
tion of capital ownership. However, it can be said that the initial
establishment of property rights over certain forms of productive know-
ledge, through patent laws and laws on copyright, would tend to increase
36 Calculated from Berlin, Statistisches Amt (1887), table V.37 Kuznets (1955), 17; skill acquisition is also an important component in Williamson’s
model, Williamson (1985).
312 Industrialization, Migration, Inequality
![Page 322: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/322.jpg)
inequality unless there are oVsetting factors (the increase in the import-
ance of this source of income might be at the expense of another form
of income which was even less evenly distributed, such as agricultural
rents).
In modern societies the ultimate ownership of capital is relatively
widely spread, with institutional investors such as pension funds, with
large numbers of beneWciaries, holding a high percentage of total equity.
Institutional improvements which widen the range of savings options
available to the less wealthy, and which, by bringing down the cost of
intermediation, reduce the gap between the rate received by small
savers and the rate of return on industrial investments, will tend to
decrease inequality. The arguments presented in the last chapter sug-
gested that only a small group of relatively wealthy investors were able
to share in the high proWts available from industrial investments in the
early stages of industrialization. As the Wnancial system developed, a
greater range of investors came to share in these beneWts.
If the mature industrial economy is one in which a wider diVusion of
skills and improved savings opportunities are tending to reduce in-
equality, it does not necessarily follow that, in the absence of these
factors, a country in the early stages of industrialization will experience
a rise in inequality. The rising importance of capital may give a labour-
saving bias to the development of technology in the early stages of
industrialization, as argued by JeVrey Williamson, but the eVect of thisfactor could be negated by a decline in the share of rents, if land
ownership is highly concentrated, as agriculture becomes less import-
ant.38 Much will depend on the starting position: the distribution of
income in the pre-industrial economy. This in turn will be heavily
inXuenced by inequalities in land ownership.
Kuznets’s view was that ‘inequality in the percentage shares within
the distribution for the rural population is somewhat lower than in that
for the urban population’.39 His evidence came from a comparison of
income distributions in India, from US data for the period before the
Second World War, and from Prokopovitch’s study of Prussian tax
data. But these studies may demonstrate the importance of institutional
factors in patterns of land ownership. The United States, for example,
had a relatively equal rural income distribution partly as a result of
38 Williamson (1985), 86–90; Williamson’s model is discussed in Feinstein (1988).39 Kuznets (1955), 7–8.
Industrialization, Migration, Inequality 313
![Page 323: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/323.jpg)
successive Homestead Acts which had favoured the establishment of
medium-sized family farms.
Other factors which aVect land ownership include climate, infrastruc-
ture, and geographical isolation. There are certain tropical crops which
are associated with ‘plantation’ production systems, due to economies of
scale favouring larger units.40 Where public sector provision of infra-
structure is poor, so that farmers have to provide this themselves, the
cost of this will be a problem for all except the largest estates. This
eVect is intensiWed by geographical isolation. The current expansion of
large-scale soya-bean production in the Brazilian Mato Grosso is a good
example of the eVect produced by a lack of infrastructure when com-
bined with geographical remoteness. The example of US agriculture
shows that soya beans can be successfully produced on small and
medium-sized units, given favourable circumstances.41
The state has an important role in this, both because it sets the
legislative framework and because it is responsible for infrastructure
investments. The United States in the nineteenth century was suY-
ciently democratic to be able to carry through a land settlement policy
broadly favourable to the interests of small and medium-sized units; the
modern Brazilian state does not appear to have either the desire or the
capacity to do this.42
Even in conditions favourable to large-scale production the state can
alter the playing Weld so as to prevent the establishment of larger units.
British colonial policy in Fiji was hostile to the cultivation of sugar cane
on plantations, with the result than many Indian immigrants brought in
as indentured labourers were eventually able to set up as small farmers
(on leased land). The much-disputed current EU banana import policy
is an attempt to protect small-scale producers on Caribbean islands
from competition from plantations in central America.
Societies make choices which aVect land ownership, and these are
often choices which reXect the balance of political power. Two eight-
eenth-century constitutional monarchies, Britain and Denmark, made
diVerent choices, which resulted in very diVerent patterns of land
ownership. The range of crops grown was similar (the Danish shift to
40 Engerman and SokoloV (1997) make this point.41 Article on the Mato Grosso in the Financial Times, Survey of Brazil (3 Oct. 1994), 16.42 In the Brazilian case the problem may well be that the cost of dealing with the social
problems created by one of the world’s most unequal income distributions leaves littlemoney for infrastructure investments in remote regions. In which case, this is an exampleof the way that high inequality can breed further inequality.
314 Industrialization, Migration, Inequality
![Page 324: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/324.jpg)
livestock production came later); neither country was geographically
remote. In the British case, the political system produced an outcome in
which the interests of land-owners were placed above all other consider-
ations.43 The Danish reforms were more beneWcial to the peasantry.
It has also been argued that a more equal distribution of land, and a
more equal starting point, will promote skill acquisition and thus pre-
vent a rise in inequality in the urban or industrial sector.44 Putting a
child into secondary education or some form of vocational training is a
considerable investment for families in poorer countries: there is the
cost of feeding and clothing, there are often fees to be paid, but most
important of all there is the loss of potential earnings, a valuable add-
ition to the family budget. With high initial inequality, and borrowing
constraints on poorer families, rather fewer will be able to aVord this
investment.
Statistical results which support these views are inXuenced by the
fact that, since 1945 there have been a number of ‘high growth–high
equality’ cases, countries with relatively equal land distributions which
have achieved high growth rates: Japan, South Korea, Taiwan, and the
Irish Republic are some examples.45 There have been others with high
levels of inequality in land ownership, which have had diYculty sus-
taining high growth rates even when these have been achieved for short
periods: Brazil is a good example.
Consideration of the four ‘high growth–high equality’ cases reveals
that in all these countries there was a reallocation of property rights in
agriculture at a time when some exceptional event made it possible to
achieve this without it being seen as a precursor of further measures. If
there was an expectation of repeated expropriations, this would aVect
investment in agriculture and in other sectors where there might be
political pressure for a reallocation of property rights.
These events include defeat, or near-defeat, in war, or in a civil war,
and the end of colonial rule. Prussia is a good historical example. Defeat
by Napoleon at Jena made it possible to carry out a land reform without
creating the expectation that similar measures detrimental to the free
exercise of property rights would be introduced at some future date.
These ‘Jena-like’ windows of opportunity appear rarely, if at all. In
their absence, property rights tend to evolve slowly. Sometimes,
43 The importance of starting points was also emphasized by Williamson (1991).44 Perrson and Tabellini (1994), Galor and Tsiddon (1994), and Clarke (1995); see also
Bernabou (1996) for a survey of the arguments.45 There is an interesting study of the Taiwan case in Fei et al. (1979).
Industrialization, Migration, Inequality 315
![Page 325: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/325.jpg)
technological changes, such as those which (arguably) made the enclos-
ure of common Welds advantageous, can create conditions where suY-
cient gains can be made to compensate some owners for the loss of their
property rights, while reallocating these rights to others. In general, the
declining relevance of feudal rights and obligations made it necessary
for these to be recast in a more modern form, but this did not deter-
mine the eventual outcome.
These factors produce a more complex mechanism than the one
outlined by Kuznets, and one which is heavily inXuenced by historical
circumstances. There are now many possible paths, not just one. Alter-
native paths include some which are radically diVerent. A country with
a extremely high level of inequality in land ownership might experience
little or no increase in inequality during the early stages of industrializa-
tion simply because the starting position was so high.46 If political
problems, or low skill acquisition, then holds back economic growth,
this country may never break through to economic maturity and the
downward phase of the Kuznets Curve. It could become stuck in a high
inequality–low growth trap: an initial pattern of high inequality being
perpetuated through low skill acquisition.
A second country with a highly equal land distribution may avoid the
Kuznets Curve for completely diVerent reasons. The transition to the
downward phase could be so fast that the initial rise in inequality would
be insigniWcant. If high skill diVusion is established right at the onset of
industrialization then this factor could outweigh other forces tending to
increase inequality. In which case there would be no Kuznets Curve:
inequality would start low and remain low.
Kuznets explained high inequality in his sector B, the urban or
industrial sector, using an argument which contains elements of the
Lewis Model:
It seems most plausible to assume that in earlier periods of industrialisa-
tion . . . [sector B’s] income distribution was more unequal than that of the
agricultural population. That would be particularly so during the periods when
industrialisation and urbanisation were proceeding apace and the urban popula-
tion was being swelled, and fairly rapidly, by immigrants—either from the
country’s agricultural areas or from abroad.47
46 Lindert’s Wnding that inequality in Britain started high and remained high mayindicate that Britain was in this category, Lindert (1994).
47 Kuznets (1955), 16.
316 Industrialization, Migration, Inequality
![Page 326: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/326.jpg)
He also saw the concentration of savings as an important factor tending
to raise inequality in sector B.48 Lewis himself endorsed the use of his
model as an explanation of high inequality: ‘the elasticity of supply of
unskilled labour is so high (whether in the dual or the classical models)
that it dampens real wages . . . The fruits of growth are concentrated in
few hands.’49
This study has looked at the labour surplus assumption of the Lewis
Model in some depth. The conclusions are broadly favourable to the
view that there is a labour surplus phase in industrialization, though
much depends on circumstances: in the German case the inXuence of
geographical location, in particular the proximity of Russian Poland,
has been shown to be considerable.
However, there are two points which have emerged which have
important implications for the Kuznets Curve. The Wrst is demo-
graphic. In Germany the onset of the fall in birth rates characteristic of
the ‘demographic transition’ came early: there was a downward trend in
the Princeton Im index (the index of births within marriage allowing
for the age distribution of married women) from the late 1870s
onwards. This was led by the urban areas. In 1880–1 marital fertility in
cities of over 20,000 inhabitants was 12.9% below the Wgure for rural
areas; by 1905–6 it was 31.4% below, having fallen by 25.6% compared
to the fall of just 5.5% in rural areas.
But this fall was itself a continuation of a pattern of behaviour
already present in rural society in Western Europe: the restriction of
population growth principally through delayed marriage. The main
diVerence was that the population in the urban areas began to marry
earlier and to restrict fertility within marriage. This, however, was not
unusual, as Protestant rural areas followed a similar pattern of behav-
iour; in Catholic areas, by contrast, people married later. The situation
in many contemporary developing economies is very diVerent. As
Todaro put it: ‘LDC (less-developed country) birth-rates today are
substantially higher than they were in pre-industrial Europe. This is
largely because of early and almost universal marriage in contemporary
Third World countries . . . since the initial level of birth-rates in West-
ern Europe was generally low as a result of either late marriage or
48 Ibid. 6–8.49 In Ranis and Schultz (1988), 14; Williamson commented (at the same conference):
‘the upswing of the Kuznets curve is not an unambiguous fact of history’, ibid. 23. Taylor(1979) and Taylor and Anda (1988) explore the distributional implications of the LewisModel.
Industrialization, Migration, Inequality 317
![Page 327: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/327.jpg)
celibacy, overall rates of population growth seldom exceeded the 1%
level even at their peak.’50
The result has been that many development economists take a rela-
tively pessimistic view of the likely pace of population increase: Hayami
writes (in 1997): ‘the probability of low-income countries escaping
quickly from the strong pressures of exogenously given population
growth appears to be relatively slim over the next couple of decades’.51
If the Kuznets Curve is indeed ‘historical fact but a contemporary
Wction’ then this very diVerent pattern of demographic behaviour is
probably part of the explanation. The labour surplus phase in the in-
dustrialization of Western Europe was curtailed, partly by the rapid
decline of the urban birth rate, and partly by the fact that birth rates
were already controlled in rural areas. Hence, the more rapid transition
to the downward phase of the Kuznets Curve.
The point is also of interest for the light it sheds on household
motivation and behaviour. Take, for example, the Wnding reported in
Chapter 5, that birth rates were lower in areas of partible inheritance.
This suggests a concern for the well-being of subsequent generations
and a willingness to adapt behaviour to secure an improvement. Too
many children threatened the viability of the family holding. So
numbers were restricted. This may not be entirely altruistic; parents
may have considered that their own interests, in having children able to
provide support in their old age, were better served by having relatively
few children who were better-oV. But it also suggests that, if a move to
a city brought improved educational and training opportunities, then
similar considerations would lead to a decision to restrict family size so
that available resources were suYcient for children to make use of these
opportunities.
The second point concerns the capital side of the Lewis Model.
DiVerential saving rates, especially in the form used by Lewis, ‘only
capitalists save’, create a powerful mechanism which may well drive up
inequality in an economy which does not draw on international sources
of capital. However, once the possibility of large-scale international
capital Xows is introduced into the equation, then the domestic savings
rate becomes less relevant. With perfect capital mobility, and zero
transactions costs (for turning international savings into domestic in-
vestment), the whole Lewis Model disappears, as does any concern
50 Todaro (1997), 196 and 200. 51 Hayami (1997), 60.
318 Industrialization, Migration, Inequality
![Page 328: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/328.jpg)
with the rate of population growth and the agricultural labour
surplus.52
There was certainly a degree of capital mobility in Western Europe
in the mid-nineteenth century, but the largest Xows were those associ-
ated with the international bond market: there was little direct invest-
ment in industry. In the last quarter this began to change, and there
was much more direct investment in those countries which began to
industrialize in this period: Italy, Russia, and Austria-Hungary. The
electrical industry, which also emerged in this period, was dominated
by a small group of German and American Wrms which had subsidiaries
in diVerent countries. But this was a relatively new development.
The position facing contemporary developing countries is not the
same. There is a large pool of international capital available for invest-
ment in all sectors of the economy. Provided it is able and willing to
oVer conditions favourable to the interests of foreign investors, a suc-
cessful developing economy should be able to attract suYcient invest-
ment to avoid Wnding itself with surplus labour, even with high
population growth and large-scale migration out of agriculture.53
Given these factors, it is not surprising that modern studies do not
Wnd a Kuznets relationship between economic growth and inequality.
The more successful developing economies have managed to avoid the
rise in inequality that Kuznets predicted. The less successful have failed
to make the breakthrough to economic maturity and the downward
phase of the curve. The Kuznets Curve is one path among many
possibilities.
7. Sources of Inequality in Germany 1870–1913
(i) General comments
From the discussion in the previous section it appears that there are
ways whereby a society going through the early stages of industrializa-
tion can avoid a rise in inequality. How well does nineteenth-century
52 The consequences of diVerential savings were examined by Kaldor (1956) and Pasi-netti (1962).
53 This is not, however, something that is either easy to achieve, or costless to thedeveloping country. Dependence on external Wnance can lead to instability as a result ofchanging sentiments in international Wnancial markets, which is one reason why successfuldeveloping economies such as Singapore or South Korea have taken steps to encouragedomestic savings.
Industrialization, Migration, Inequality 319
![Page 329: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/329.jpg)
Germany measure up to these standards? Was the rise shown in Figure
9.1 inevitable?
Germany started with a relatively equal distribution of income, due
to a more equal distribution of land ownership, which was generally
more egalitarian outside Prussia, with some exceptions such as Meck-
lenburg. Historians have criticized the land reforms of the early nine-
teenth century on the grounds that they left a relatively unequal
distribution of land ownership east of the Elbe and also that there was a
rise in the numbers of landless labourers. But a broader comparison
shows that the reforms, coupled with the previous policy of
Bauernschutz which protected and entrenched peasant rights, were at
least partially successful.54
Unlike the most successful of the modern developing countries,
Germany was not able to avoid a rise in inequality during industrializa-
tion. In comparison with these countries, Germany suVered from two
main disadvantages: there was little foreign direct investment (by
modern standards), and Wnancial institutions were relatively under-
developed. In particular, the system of universal banks closely tied to
industry, which Germany pioneered and which has reappeared in other
countries such as Japan or South Korea, was then relatively unknown.
In consequence, there was a period when industrial investment was
held back, more by institutional problems than by a general shortage of
capital, which ran from the 1870s to the mid-1890s.The shift of the terms of trade against agriculture in the 1870s would
not, under reasonable assumptions about factor immobility, have led to
a permanent rise in inequality. However, it had a powerful short-term
eVect, and it was particularly unfortunate that, just as the movement of
the international terms of trade required an increased rate of migration
from rural areas into cities, the rate of industrial investment should be
held back by the institutional problems mentioned in the last paragraph.
It also coincided with a period when surplus labour was being released
from agriculture as a result of institutional changes, such as the decline
of the contractual labour system studied by Weber.
German conditions were also inXuenced by geography, in particular
the proximity of Russian Poland, which had a considerable inXuence on
the structure of rural wages. By restricting wage growth this factor
54 Historians who have analysed the agrarian reforms include Conze (1969), Dickler(1975), Harnisch (1986), Koselleck (1967), Lutge (1963), Plaul (1986), Perkins (1986), andSaalfeld (1963).
320 Industrialization, Migration, Inequality
![Page 330: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/330.jpg)
contributed to overall inequality even in areas where the numbers of
migrants were small. Competition from migrants from eastern and cen-
tral Europe also aVected German migration to the United States, but it
was fortunate that this eVect was mainly felt after 1895, by which time
the ‘labour surplus’ period was over. During the period up to 1895, the
safety valve of emigration to the United States was open. It reduced the
pressure on urban labour markets inside Germany, and thus helped to
prevent an even steeper rise in inequality.
Given these factors, it is probable that high levels of inequality were
inevitable in the urban, or industrial, sector in 1870–95. Nothing much
could have been done to avoid this. Higher levels of agricultural protec-
tion could have held back the rate of decline of rural employment, but
at the cost of a signiWcant reduction in industrial growth, and there
would also have been a detrimental eVect on urban wages. Examples
from modern developing countries show that, under similar circum-
stances, the introduction of measures to raise urban wages creates
‘urban bias’, and a division of the urban labour market between a rela-
tively well-paid formal sector, and a low-paid, or underemployed, infor-
mal sector. Some measures might record this as an increase in
inequality.
The same point applies to the activities of trade unions. In this
period, trade unions were better organized in the more skilled sectors,
so their inXuence, in so far as they did aVect wage levels, would have
tended to widen divisions within the urban labour market.
The deliberate restriction of internal migration through government
action, a practice followed in a number of developing countries (notably
China), would have had the eVect of widening the overall income distri-
bution by holding a number of potential migrants in low-paid rural
employment. It is a policy which is only likely to appeal to governments
which give little weight to the interests of those living in rural areas.
The factors that tend to bring the initial rise in inequality to an end
have been identiWed as higher investment and increased skill diVusion.
In both these areas the German record can be described as exemplary:
it was one which other countries have sought to emulate. Both the
vocational training system, and the Wnancial system as it developed in
the 1890s, were well suited to the requirements of the German economy
in this period, and these were the forces which brought about the
decline in inequality, in combination with the recovery of German
agriculture after 1895, which reduced the urban–rural gap and also
decreased the pressure on urban labour markets.
Industrialization, Migration, Inequality 321
![Page 331: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/331.jpg)
(ii) Analysis by simulation
The Wrst year for which separate distributions for both the rural and
urban sectors are available is 1896. These can be used to simulate the
eVect on overall inequality of diVerent scenarios. Between the 1820s
and the 1890s there was a total increase in inequality of around 40%;
this is the change that has to be explained.
Table 9.4 presents simulations which show that, using Kuznets’s
assumptions, migration will push up inequality (scenarios C and E) but
that this eVect can be reversed if more realistic assumptions are made,
so that migrants come from the lower part of the rural distribution and
move to the lower part of the urban distribution (scenarios D and F).
So migration, in itself, is not a very convincing explanation of the
Kuznets Curve.
Table 9.4. Simulations using 1896 rural and urban income
distributions
I. Using 1896 sector weights: 40% urban, 60% rural
Scenario EVect on overall inequality
A All rural incomes fall 10% Gini-coeYcient rises 2.6%
B Incomes of poorest 50% in rural
sector fall 10%
Gini-coeYcient rises 3.8%
C 10% of rural population moves to
urban sector—no eVect on sector
distributions (Kuznets process)
Gini-coeYcient rises 1.8%
D 10% of the poorest half of the rural
population move to bottom half of
the urban sector
Gini-coeYcient falls 2.1%
II. Using circa 1850 sector weights: 10% urban, 90% rural
E 10% of rural population moves to
urban sector—no eVect on sector
distributions (Kuznets process)
Gini-coeYcient rises 7.1%
F 10% of the poorest half of the rural
population move to bottom half of the
urban sector
Gini-coeYcient falls 3.8%
G Overall eVect of move from
10% urban/90% rural to 40%
urban/60% rural (Kuznets process)
Gini-coeYcient rises 14.5%
322 Industrialization, Migration, Inequality
![Page 332: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/332.jpg)
However, the Kuznets process as a whole, the shift from a predomin-
antly rural society to one in which the urban sector is substantially
increased, will increase overall inequality, as scenario G shows. But the
eVect only accounts for around a third of the increase in inequality
which actually occurred between the 1820s and the 1890s. Other factors
which might have contributed are simulated in scenarios A and B: a
shift in the terms of trade which has an adverse eVect on agriculture
raises inequality, but the eVect is greater if it is strongest on those at the
bottom of the rural distribution (scenario B).
In general, the strength of the rise in inequality shown by the analysis
of the Prussian income tax data is such that it was almost certainly
brought about by a combination of adverse factors. These were:
1. a shift of population from a more equal rural sector to a less equal
urban sector (Kuznets process);
2. increased inequality in the rural sector (due to a rise in the
numbers of landless labourers);
3. increased inequality in the urban sector (due to the operation of
the Lewis Model);
4. a shift of the terms of trade against agriculture.
(iii) Econometric analysis
The Prussian tax statistics can also be used to examine some of the
causes of inequality in the rural Kreise. In 1916 an analysis of tax
returns in 1914 at the Kreis level was published in the Zeitschrift of the
Prussian Statistical Bureau. The assessments were based on income in
the previous year, so this is an analysis of the Prussian income distribu-
tion in the Wnal year before the war.55
The data gives the numbers of taxpayers, and their dependants in
four categories: those with annual incomes below 900 Marks, those with
between 900 and 3,000 Marks, those with between 3,000 and 9,500
Marks, and those with over 9,500 Marks. This made it possible to
produce estimates of inequality at the Kreis level, and these could then
be subjected to statistical analysis. The results of the overall analysis are
summarized in Table 9.5.
The analysis showed that diVerent-sized cities were associated with
diVerent levels of inequality in nearby regions. Larger cities tended to
55 There is a full report on the sources used and the results of the econometric analysisin Grant (2002b).
Industrialization, Migration, Inequality 323
![Page 333: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/333.jpg)
have high levels of inequality, and this appears to spill over into nearby
regions. But the eVect of smaller towns was more beneWcial: inequality
was lower in regions which were close to these. Kreise which were
remote from these smaller towns tended to have high inequality.
Inequality was lower in the more agricultural Kreise, conWrming the
sectoral diVerences shown in Table 9.2. There was also a pronounced
east–west division: inequality tended to fall as the distance from Rus-
sian Poland increased. However, in the rural areas the land distribution
was a major inXuence on overall inequality, so it may have been the
unequal land distribution which was causing inequality to be high in
the rural east. This can be demonstrated by considering the rural Kreise
separately; an analysis of this type is summarized in Table 9.6.
This conWrmed some of the points made in Table 9.5: inequality was
high in areas close to big cities, and in those which were remote from all
urban areas. But the inXuence of Russian Poland was considerably
weakened, and may well be insigniWcant in statistical terms. While the
earlier table might be interpreted as showing that inequality was
‘imported’—the availability of cheap labour from Russian Poland was
pushing up German inequality—it now appears that inequality in the
east was to a large extent ‘home-grown’. It was the unequal land distri-
bution in the east which caused inequality to be high in these areas.
Turning to the major cities, Table 9.7 reports the results of the analy-
sis of inequality in these urban areas. There was no tendency for in-
equality to be higher in cities further to the east, which conWrms that the
Table 9.5. Summary of results of statistical analysis of inequality in
583 Prussian Kreise, 1913/14
I. Secure statistical results
Factors which tend to increase inequality:
Proximity of the border with Russian Poland
Remoteness (as measured by the distance from towns and cities with at least
20,000 inhabitants)
Factors which tend to decrease inequality:
High rates of agricultural employment
II. Less secure statistical results
Factors which tend to increase inequality:
Proximity of large cities (at least 100,000 inhabitants)
Source: Summary of the results given in table 13 of Grant (2002b).
324 Industrialization, Migration, Inequality
![Page 334: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/334.jpg)
east–west division was mainly due to diVerences in the land distribution.
However, inequality was higher in cities with high levels of migration,
measured by the total number of moves, in and out, relative to the
resident population. In these cities, with highly Xexible labour markets
and elastic labour supplies, there was a larger gap between rich and poor.
This provides support for the view that the migration of the rural labour
surplus into urban areas was a factor driving up inequality. However,
the main ‘labour surplus’ period was over by 1914. If an eVect can be
Table 9.6. Summary of results of statistical analysis of inequality
in 478 rural Prussian Kreise, 1913/14
I. Secure statistical results
Factors which tend to increase inequality:
Remoteness (as measured by the distance from towns and cities with at least
20,000 inhabitants)
Presence of large farms and estates (over 100 hectares)
Factors which tend to decrease inequality:
High rates of agricultural employment
Proximity of large cities (at least 100,000 inhabitants)
II. Less secure statistical results
Factors which tend to increase inequality:
Proximity of the border with Russian Poland
Presence of medium-sized farms (20 to 100 hectares)
Table 9.7. Summary of results of statistical analysis of inequality
in 44 Prussian cities, 1913/14
I. Secure statistical results
Factors which increase inequality:
High levels of migration (moves both in and out)
II. Less secure statistical results
Factors which tend to increase inequality:
Presence of textile, mining and printing industries
Presence of utilities
Factors which tend to decrease inequality:
Presence of leather industry
Industrialization, Migration, Inequality 325
![Page 335: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/335.jpg)
found for this date, it suggests that the impact would have been much
greater in the 1870–95 period, when the bulk of the transfer took place.
The industrial structure of the cities was found to have an eVect on
inequality, although the statistical analysis was not entirely satisfactory.
Industries which were associated with high inequality included textiles
and mining, which may reXect the larger scale of operation of these
industries, but also printing, which was not organized on a large scale.
The leather industry, which was relatively unmechanized and organized
in small workshops, was associated with low levels of inequality.
8. Could the R i se of Inequality have been Avoided?
The results produced so far are in line with the conclusion reached by
some modern cross-section studies: there is a conditional Kuznets
Curve, which can emerge when the historical circumstances are right
but which, under diVerent conditions, might be obscured by other
factors. The question then is, what were the forces that produced the
Wrst stage of the Kuznets Curve in nineteenth-century Prussia? Could
the increase of inequality have been prevented?
The Wrst point is that, even at the peak, inequality was not particularly
high in Prussia by modern standards. Although income inequality in the
urban sector was quite high, the overall Wgure was held down by a relatively
egalitarian rural sector. What created the Kuznets Curve in Prussia was the
shift from a rural society with a low Gini-coeYcient to an urban or indus-
trial sector which had a level of inequality which was high by comparison
with the rural sector though not so unusual when compared to other coun-
tries going through the process of industrialization.
The second point is that, from the 1870s onwards, there was a shift
in the terms of trade against the agricultural sector as a result of devel-
opments on world markets. This had the eVect of widening the gap
between rural and urban incomes, and thus worsening the overall
income distribution. Although this was not the only cause of the rise in
inequality in the period, it contributed to the rise in the Gini-
coeYcient. This was due to the changing terms of trade on world
markets. Only a massive rise in agricultural protection could have kept
this from aVecting German agriculture.
A third point is that, in the early part of the century, inequality was
rising in the rural sector, due to a range of factors: demographic pres-
sure, expansion into remote eastern areas, the eVects of the agrarian
326 Industrialization, Migration, Inequality
![Page 336: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/336.jpg)
reforms, and changing social relations. This was only indirectly con-
nected to industrialization.
These factors provided a considerable reinforcement to the mechan-
ism speciWed by Kuznets himself. It would be quite possible for a
society which industrialized under diVerent circumstances to escape the
Kuznets Curve. The Prussian example points to some factors which
might help to achieve this: such as a relatively low level of demographic
pressure, a prosperous agricultural sector in which incomes kept pace
with those in urban areas, agrarian policies which promoted an equal
distribution of income in this sector, and better credit provision for
small farmers. These are the ‘good’ ways to escape a Kuznets Curve.
But the comparison with Britain also reveals that there is a ‘bad’ way to
escape it, and that is to start oV with such a high level of inequality in
the rural sector that the transition to an industrial society does not push
up overall inequality even though incomes in the urban sector are also
unequally distributed.
In short, the Kuznets Curve, while not just a peculiarity of German
history, owes its origins in Prussia to some features of nineteenth-
century Prussia which may or may not be replicated in modern
developing countries. It is the product of circumstances rather than an
automatic consequence of industrialization. But there was not much
that government action in nineteenth-century circumstances could have
done about this. A period of rising inequality was probably something
that just had to be endured.
Appendix 9A. Est imates of income inequality from
Pruss ian tax statist ics
Table 9A.1. Estimates of income inequality in Prussia 1822–46
1822 1826 1831 1836 1841 1846
Gini-coeYcient
as indices (1846¼100)
0.260 0.266 0.274 0.278 0.286 0.297
Gini-coeYcient 87.6 89.4 92.3 93.5 96.4 100.0
Upper 95% CI of
Gini-coeYcient
96.5 98.5 101.7 103.1 106.2 110.2
Lower 95% CI of
Gini-coeYcient
78.7 80.3 82.9 84.0 86.6 89.8
Industrialization, Migration, Inequality 327
![Page 337: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/337.jpg)
Table 9A.2. Estimates of income inequality in Prussia 1853–73
1853 1855 1858 1861 1864 1867 1870 1873
Gini-coeYcient
as indices (1853¼100)
0.345 0.348 0.352 0.353 0.354 0.355 0.357 0.363
Gini-coeYcient 100.0 100.9 102.1 102.3 102.6 102.8 103.4 105.2
Upper 95% CI of Gini-coeYcient 111.7 112.7 114.1 114.3 114.6 114.9 115.5 117.5
Lower 95% CI of Gini-coeYcient 88.3 89.0 90.1 90.3 90.5 90.7 91.2 92.8
Table 9A.3. Estimates of income inequality in Prussia 1873–5
1873 1874 1875
U10/L40 1.44 1.51 1.47
Gini-coeYcient as indices with 1853¼100 0.315 0.325 0.318
Gini-coeYcient 105.2 108.5 106.2
Table 9A.4. Estimates of income inequality in Prussia 1875–91
1875 1876 1878 1880 1882 1883 1885 1887 1889 1891
Gini-coeYcient
as indices (1853¼100)
0.340 0.341 0.357 0.371 0.373 0.370 0.374 0.374 0.379 0.382
Gini-coeYcient 107.1 107.6 112.5 116.8 117.7 116.7 117.9 118.0 119.5 120.5
Upper 95% CI of Gini-coeYcient 112.1 112.7 117.8 122.3 123.3 122.2 123.5 123.6 125.1 126.1
Lower 95% CI of Gini-coeYcient 102.1 102.5 107.2 111.3 112.2 111.2 112.4 112.5 113.8 114.8
328Industrialization
,Migration,
Inequality
![Page 338: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/338.jpg)
Table 9A.5. Estimates of income inequality in Prussia 1892–1914
1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902
Gini-coeYcient
as indices (1853¼100)
0.321 0.324 0.324 0.323 0.315 0.318 0.324 0.327 0.331 0.333 0.334
Gini-coeYcient 119.2 120.1 120.1 119.8 117.0 118.1 120.0 121.4 122.9 123.6 124.1
Upper 95% CI 123.9 124.8 124.8 124.6 121.6 122.8 124.8 126.2 127.8 128.6 129.0
Lower 95% CI 114.5 115.3 115.3 115.1 112.4 113.4 115.3 116.5 118.0 118.7 119.1
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914
Gini-coeYcient
as indices (1853¼100)
0.331 0.332 0.335 0.340 0.338 0.334 0.333 0.319 0.316 0.315 0.318 0.329
Gini-coeYcient 122.6 123.1 124.3 126.0 125.5 124.0 123.5 118.2 117.3 117.0 117.9 122.0
Upper 95% CI 127.5 128.0 129.3 131.1 130.4 128.9 128.4 122.9 122.0 121.6 122.6 126.8
Lower 95% CI 117.7 118.2 119.4 121.0 120.5 119.1 118.6 113.5 112.6 112.3 113.3 117.1
Table 9A.6. Analysis of the estimated changes in each period
1822–46 1853–73 1875–91 1892–1906 1906–14
Change in Gini-coeYcient (%) þ14.2 þ5.2 þ12.5 þ5.7 �3.2
Upper 95% CI of change þ16.2 þ6.0 þ13.3 þ6.0 �3.1
Lower 95% CI of change þ12.2 þ4.3 þ11.6 þ5.5 �3.4
Upper 95% CI of highest variant þ21.4 þ20.2 þ13.6 þ8.3 �5.4
Lower 95% CI of lowest variant þ4.3 þ3.5 þ5.7 �1.3 �2.7
Source: ***Take in JJ From fo. 386***
Industrialization,
Migration,
Inequality329
![Page 339: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/339.jpg)
10
Challenging the Kehrite
View of Imperial Germany
1. Introduction
The view that the Kaiserreich suVered from internal Xaws is so deeply
entrenched in German historiography that one doubts if it will ever be
dispelled, however much evidence is produced against it.1 The list of
Xaws has changed over time: concepts such as the lack of a middle-class
identity, ‘organized capitalism’ as a distinctive economic system, ‘social
imperialism’, the importance of the ‘failed revolution’ of 1848 have
been given up or downgraded. The term Sonderweg is used less. But
new lists are produced, which stress diVerent aspects of the social order,
internal structure, political system, and cultural outlook of Imperial
Germany: the importance of organized economic pressure groups, the
impotence of parliamentary control, the appeal of Social Darwinism,
the drift to extremist nationalism, and so on.
This study has not attempted to confront these arguments directly;
instead it has put forward an alternative view, taking many of its cues
from the comparative study of developing countries in general, and
from development economics in particular. The argument in its essence
is that the Kaiserreich suVered from no more than the normal problems
of a rapidly industrializing society.
This is not to argue that there were no special features and character-
istics of German society in the period. There were indeed unusual and
unique attributes, as there are in all societies. But as Richard Evans has
1 Even historians specializing in foreign policy have accepted much of the ‘Xawednation’ thesis: see Hildebrand (1995), 5–6 and 9. Hildebrand also accepts the interlockingof domestic and foreign policy, Hildebrand (1989), 77. The view that a more realisticforeign policy, involving an alliance with either Britain or Russia, was blocked for internalreasons is set out in Geiss (1976), 116, and Kehr is cited in support of this. Calleo (1978)provides a more sympathetic view of the ‘German problem’, and argues that GermanuniWcation should not be seen as ‘a malevolent accident which befell an otherwise harmo-nious European state system’, ibid. 3, but then goes on to accept the now discreditedargument about the connection between ‘organized capitalism’ and German colonialexpansion, put forward in Wehler (1969).
![Page 340: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/340.jpg)
pointed out, commenting on the evolution of the Sonderweg thesis in
the works of Hans-Ulrich Wehler, the list of Xaws which Wehler has
put forward in recent work as an explanation of the particular course of
German history contains nothing that cannot be found in some degree
in other European states in the late nineteenth and early twentieth
centuries.2
And, it might be asked, why is the comparison limited to Europe in
this period? Industrialization is no longer a purely European phenom-
enon. It is an experience through which a number of non-European
societies have passed or are passing. And in virtually every case it has
been a disruptive and dislocating process.
Evans also commented that the Sonderweg argument has become one
which is increasingly concentrated on the failure to make political pro-
gress commensurate with the pace of social and economic transform-
ation. This is a point which is widely accepted even by historians who
are not associated with the Sonderweg view. It is one of the common-
places of Wilhelmine historiography.3
But a broader comparison of societies, inside and outside Europe,
going through the initial stages of industrialization reveals that few if
any have made political progress during this period. Even in Britain,
the earliest phase of industrial growth, usually dated 1780 to 1830, was
not a period which saw much in the way of progressive democratic
politics. In many other countries, there have been declines into civil
war, authoritarianism, or military government.
When this broader context is considered, the insistence that Imperial
Germany should have made political progress commensurate with its
economic advance can appear rather bizarre.4
2. Social , Economic , and Political
Trajectories in Imperial Germany
In the history of the Kaiserreich the results of the 1912 election have a
particular signiWcance: the Social Democratic Party gained 110 seats,
2 In a review of Wehler’s Deutsche Gesellschaftsgeschichte, reprinted in Evans (1997).3 Theodor Schieder was one example of a historian of a diVerent persuasion who made
this point, Schieder (1960), 85.4 The view that industrialization and progress towards democracy are linked is a com-
ponent of ‘modernization’ theory, and this is often regarded as a typically Europeanphenomenon; see Eley (1996) for comments on this from a German perspective.
Challenging the Kehrite View 331
![Page 341: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/341.jpg)
reversing the losses made in the ‘Hottentot’ election of 1907, and
exceeding the previous high-water mark of 81 seats achieved in 1903. It
was now the largest party in the Reichstag; if it could gain the support
of the left liberals and the parties of the national minorities (Poles,
Danes, and Alsatians) it would form a ‘disaVected’ block of 185 out of a
Reichstag of 397 deputies.
This had a number of important consequences. On the right it pro-
duced a sense of marginalization, which was intensiWed when the gov-
ernment made concessions to the SPD, including the introduction of
direct taxation at the Reich level, in order to secure the joint passage of
the 1913 Finance Bill and the Army Bill of that year. Kehrite historians
such as Dirk Stegmann have argued that the radicalization of the right,
and the consequent abandonment of Sammlungspolitik, were important
factors in the intensiWed foreign policy crisis which culminated in the
events of July 1914.5
More broadly, the results of the election called into question the
quasi-democratic system of constitutional checks and balances which
maintained the Kaiserreich’s uneasy compromise between diVerent
social and regional groups. A Reichstag which possessed legislative
powers but which did not control the executive had been a successful
solution to many of the problems faced in 1871: how to provide repre-
sentation for others in an empire dominated by Prussia; how to secure
local autonomy from encroachments by a central executive; how to
reconcile previously dominant groups (the landed aristocracy, the urban
Burgertum) to the experiment of universal male suVrage.6 But the con-
tinuation of this compromise required that all participants accepted the
implied limitations, and this was particularly diYcult for the Social
Democrats, because they were largely excluded from participation in
the political process at the level of local and regional government.
Any compromise which would have a reasonable prospect of securing
Social Democratic support over the long term, rather than as a short-
term tactical expedient (as in 1913), would have had to have included
reform at the local and regional level as well as the strengthening of
5 Stegmann (1970), 257–93; according to this, the election led to a renewed imperialistagitation, ibid. 277. The ‘shameless, chronic persecution’ of Bethmann-Hollweg by thesegroups had, it is argued, an eVect on his policy in July 1914, Chickering (1984), 289.
6 The Bismarckian constitution has attracted much criticism from historians: ‘mon-strous’ and ‘irregular’ are two adjectives that have been used, Sauer (1968), 413. But thisunderestimates the problems it had to resolve. For an alternative view, see Rauh (1973),47 V.
332 Challenging the Kehrite View
![Page 342: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/342.jpg)
Reichstag authority.7 There are many historians who would argue that
such a compromise was not acceptable to powerful interest groups, and
that a military coup, or Staatsstreich, would have occurred had not the
outbreak of war intervened to make this unnecessary.8 The war was
welcomed by such groups as a means of halting the progress towards
democracy.
Thus, in the Kehrite scheme, the election has a special signiWcance: itis the point where sharpening internal divisions directly impinged on
Germany’s strategic position, making the need for a solution to the
problems of Germany’s diplomatic isolation into a pressing imperative.
It also made it clear that any internal solution would have to involve
major concessions to a party with a revolutionary tradition and regarded
by its conservative opponents as disloyal.9
Against this, there is the more favourable picture presented by the
economic trends discussed in the previous chapters. Combining these
trends it is possible to construct, as a mental picture, a curve of the
intensity of the socio-economic problems facing Imperial Germany,
meaning by this both purely economic problems, and social problems
having economic roots. The trajectory of this curve shows a rising trend
to 1895–1900 and then declines. But this was not the only set of prob-
lems which beset the Kaiserreich. There were diVerent sets involving
socio-cultural issues (dissatisfaction with or alienation from social and
cultural structures and trends) and socio-political questions, deWningthese to include purely political problems and social divisions which
manifest themselves within the operation of the political system.
Voting behaviour is one symptom of the latter, though the movement
of opinion within parties is another aspect which is also important
although less easy to measure. If the shares of the diVerent parties are
plotted (Figure 10.1) this shows the steady gains made by the Social
Democrats from the late 1880s onwards and the way that this squeezed
7 For a discussion of the tactical nature of the SPD’s support of the government in1913, see Groh (1973), 434–6. The background to the vote is analysed in Stargardt (1994),108–24.
8 The danger of a Staatsstreich was probably not that great, Pogge von Strandmann andGeist (1965).
9 The extent to which the SDP was ever a revolutionary party is debatable. Marxistideas certainly inXuenced the leadership, but had less appeal at lower levels, Lidtke (1985).Its ‘disloyalty’ would seem to have been at least partly disproved by its vote for the 1913Army Bill. Despite this, Epkenhans (1991) has argued that there was a fear that a politicalcompromise would lead to a ‘storm of disarmament’, ibid. 413, and that this, together withthe internal political situation, led to the Flucht nach Vorn of July 1914.
Challenging the Kehrite View 333
![Page 343: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/343.jpg)
other parties. However, it would be wrong to conclude from this that
there was a radicalization of internal politics in the Wilhelmine period.
The internal politics of the Social Democratic Party have also to be
considered. These were moving in a more centrist direction, and the
party was making a determined eVort to broaden its support and move
outside its natural constituency of the urban working class. This was
rewarded by a gain in electoral support.
The area of German politics which was shifting in a more radical
direction was the conservative right. But this was losing ground. The
best result for the right-wing parties came in the 1893 election, when
the combined right-wing vote was 16.9%, but this had fallen to 9.6%
by 1912. There was a fall in seats gained from 88 to 46. The shift to
more radical policies was not beneWcial in electoral terms.
In short, the pattern of electoral politics in Imperial Germany was
not one of increasing polarization. Parties which adopted more moder-
ate policies gained votes; parties which shifted towards the extremes
lost votes. This was exactly as would be expected in a normal, stable,
Minoritiesand others
Conservativesand other right
Nationalliberals
Freeconservatives
Lift liberals
Centre
SPD18710
20
40
60
80
100
1874 1877 1878 1881 1884 1887 1890 1893 1898 1903 19121907
Fig. 10.1. Percentage shares of different parties in Reichstag elections,
1871–1912
334 Challenging the Kehrite View
![Page 344: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/344.jpg)
democratic system. But the gains made by the SPD meant that the
party had to be admitted into the Wilhelmine political establishment:
the option of continuing to ignore it no longer existed.
3. The Dog that D idn ’t Bark :
Economic Issues and the 1912 Election
The SDP’s ability to break out from its core constituency depended on
its success in appealing to artisans, tradesmen, workers in small work-
shops in the urban areas, and in attracting votes in smaller towns which
were not major industrial centres.10 For these voters, the importance of
the economic issues stressed in earlier party policy, such as public
ownership of major enterprises, was at best peripheral.
It is worth considering quite how small the Social Democrats’ core
constituency was. In 1910 the urban population, those living in major
towns and cities (91 cities with at least 50,000 inhabitants), was 17million out of a total German population of 64 million. A survey of
employment in these cities showed a total of 1.6 million men (and 0.5
million women), aged 16 and above, employed in enterprises with at
least 10 employees. Bearing in mind that some of these would have
been too young to vote, and also that 29% of the urban population were
Catholics, this suggests that the Social Democratic core, urban Protest-
ant working-class men, was between 1.2 and 1.3 million voters. But the
SDP vote was well above this level. It totalled 4.25 million in 1912. Itwas 1.78 million in 1893 (when the size of the core was considerably
smaller).
Figure 10.2 contains a comparison of the SDP share of the vote with
the percentage of the German population living in urban areas. This
illustrates the way that structural change in German society was
favouring the party’s electoral advance. It also suggests that the party
broke out of its core constituency relatively early, in the 1890s. In the
early period the party’s share of the vote was roughly in line with
the share of the German population living in cities of at least 100,000:
the major industrial centres where large-scale enterprise was common.
But from 1890 onwards the party’s share moves with the percentage of
the population living in towns and cities with at least 20,000 inhabit-
ants, many of which were not industrial centres at all.
10 Sperber (1997), 64, estimates that the party had 45% of the Protestant middle-classvote, and 13% of the Catholic middle-class vote in 1912.
Challenging the Kehrite View 335
![Page 345: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/345.jpg)
The graph shows the deviations from trend caused by the anti-socialist
laws of the 1880s, and it also shows the exceptional nature of the 1907
election: the 1912 election represented a recovery of the relationship
previously established. It is also worth noting that the proportion of the
German population living in towns with at least 20,000 inhabitants was
rising by 6 percentage points each decade. If the SDP could hold onto
the electoral share implied by this trend, its vote would rise proportion-
ately—by over a million votes every ten years.
The ability of the party to address broader social and political issues
made it attractive to a swath of opinion outside the larger cities. It was
also helped by the ineVectiveness of other opposition parties: voters
who wanted signiWcant change in the political structure of Imperial
Germany had no other party to turn to.
The statistical analysis of election results can elucidate some of these
points, though it is necessary to stress that statistical tests can only be
run where there are questions involving measurable issues. It is possible
to look at questions such as ‘Did high unemployment in particular
constituencies cause a rise in the socialist vote?’; others, such as ‘Did
the desire for constitutional change lead to an increase in the SDP
191119061901189618911886188118761871
40
30
20
10
0
SPD vote
% in cities >20,000
% in cities >100,000
Fig. 10.2. The share of the Social Democratic Party in Reichstag elections
compared with the proportions of the German population living in urban areas,
1871–1912
336 Challenging the Kehrite View
![Page 346: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/346.jpg)
vote?’, are not so easily answered. Moreover, statistical analysis at the
aggregate level does not provide direct evidence on voter motivation;
this would require surveys of individual voter behaviour, attitudes, and
social background. It shows which factors are associated with higher or
lower probabilities of a vote for a given party. It tests the inXuence of
the general social and economic environment on voting behaviour.11
The data set prepared for this analysis was split in two. The Wrstincluded the major cities and the second looked at results in Prussia
outside these urban areas. This enabled statistical sources giving infor-
mation on economic conditions in the urban areas to be used in the Wrst
data set, and sources on agricultural structure to be included in the
second.
Table 10.1 gives a summary of the statistical analysis of voting be-
haviour in 70 large towns and cities (the full results are given in Table
10B.1 in Appendix 10B). The analysis Wrst demonstrated the import-
ance of city size: the SDP share was higher in the larger cities. The
second important factor was religious aYliation: Catholics were much
less likely to vote for the SDP (the proportion of non-Catholics voting
for the SDP was 2–21⁄2 times that for Catholics).12
Table 10.2 shows the implications of these results for cities of diVer-
ent sizes.13 The importance of city size is clearly demonstrated. The
growth of cities tended to favour the SPD. Migration patterns which
caused cities to expand were therefore a factor in the party’s advance.
Catholicism remained an obstacle even in the larger cities. However,
when Prussian cities are separated from cities outside Prussia an im-
portant diVerence is revealed: the eVect of Catholicism was considerably
stronger inside Prussia. In Prussian cities the Catholic population was
much less likely to vote SPD. In a city of 500,000, non-Catholics would
be three times as likely to vote SPD as Catholics. But outside Prussia
11 To give an example of this: only a survey of individual voters could show whetherunemployed workers were more likely to vote for the SPD. A statistical association be-tween high unemployment in a given constituency and a high SPD vote might show this,but it might also show that employed workers in a constituency with high unemploymentwere more likely to vote SPD. The two eVects cannot be distinguished. All that can beshown is a general result that unemployment inXuences voting behaviour.
12 The SPD was gaining ground with the Catholic working class, though its share wasonly 22% against 54% for Protestant workers, according to estimates by Sperber (1997),37–8. Contrary to the view put forward by Schmadeke (1995), 183–91, Catholicism wasnot an insuperable obstacle to the SPD. Schmadeke’s conclusion, that the SPD’s rise wascoming to an end, does not appear to be justiWed by the evidence.
13 The coeYcients used are those given in column a of Appendix Table 10B.1.
Challenging the Kehrite View 337
![Page 347: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/347.jpg)
this was less important. In a similar-sized non-Prussian city the ratio
between non-Catholics and Catholics would only be 1.3–1.14
It might be thought that this was a reXection of the importance of the
Polish-speaking Catholic population in Prussian cities. But inspection
14 This conWrms a point made by historians who have looked at regional electionresults. The SPD was a strong contender for the Catholic working-class vote in non-Prussian areas. Blackbourn (1980), 193, comments on the strength of the party inworking-class Catholic areas in Baden and Wurttemberg. The SDP had 28.3% of the votein Baden in 1912 (Baden was 59.3% Catholic) but only 9.8% in Regierungsbezirk Koblenz(65.5% Catholic) and 4.6% in Regierungsbezirk Trier (79.3% Catholic). Figures fromRitter (1980).
Table 10.1. Summary of results of statistical analysis of the 1912
SPD vote in 70 major German cities
I. Secure statistical results
Factors which were associated with a higher SPD vote:
Size of the city (larger cities had higher SPD shares)
A large SDP vote in 1878
Factors which were associated with a lower SPD vote:
Catholicism
Catholicism within Prussia (additional eVect)
Factors which had little or no inXuence on voting behaviour:
Levels of inequality (measured by the U10/L40 ratio)
Local labour market conditions (the ratio of applicants to vacancies)
II. Less secure statistical results
Factors which were associated with a higher SPD vote:
High wages
High mobility
Large-scale industry (large factory size)
Table 10.2. Predicted SPD share of the vote for cities of different
sizes and different religious affiliations
Number of inhabitants in the city
50,000 100,000 500,000 1,000,000
City population is:
80% Catholic 20.5 24.6 35.7 41.1
20% Catholic 37.8 43.4 56.7 62.2
338 Challenging the Kehrite View
![Page 348: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/348.jpg)
of the data reveals that this was not the case. The socialist vote was also
unusually low in Rhineland cities like Aachen and Trier which had few
Poles. Rather it seems to reXect the strength of Rhineland Catholicism,
the social cohesion of a minority in a predominantly Protestant soci-
ety.15 The after-eVects of the Kulturkampf almost certainly played a
part as well.
There was a strong element of persistence in the socialist vote: the
cities which had a relatively large vote in 1877 tended also to have high
votes in 1912. Table 10.3 gives the Wgures for the two elections for the
ten cities with the highest votes in 1912. The SPD share was already
above 40% in six out of the ten cities. Much of the growth in the
party’s overall share came from the spread of socialist ideas from these
core cities, initially to other large cities, and then to smaller towns and
other areas.16
15 Another possible reason, suggested by Rohe (1982), 236, is that in Prussian areassuch as the Ruhr there was a confessional aspect to employment relations: typically,Catholic workers were employed by Protestant entrepreneurs. This meant that, in theseareas, Catholicism developed a ‘social’ dimension which was less evident in non-PrussianCatholic regions, where workers and employers were both likely to be Catholics. See alsoEvans (1997) and Sperber (1983) for the consequences of the Kulturkampf.
16 Hamburg and Saxony were strongly represented at the original meeting of theAllgemeiner Deutscher Arbeiterverein in 1863, Berger (2000), 49. See also the maps of SDPsupport in Schmadeke (1995) which show the strength of the party in areas aroundSaxony, particularly Thuringia and Franconia. There was a geographical element to thediVusion of SPD support.
Table 10.3. Cities with the highest SPD vote in 1912; comparison
with 1877 percentage share of votes cast (first-round votes)
SPD vote 1912 SPD vote 1877
Berlin 75.0 39.2
Chemnitz 64.3 54.9
Altona 63.4 51.0
Hamburg 61.2 40.0
Nuremberg 60.7 48.3
Zwickau 60.6 61.4
Charlottenburg 58.7 16.2
Mulhausen 56.1 �Brunswick 55.6 46.1
Solingen 55.6 35.8
Challenging the Kehrite View 339
![Page 349: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/349.jpg)
The analysis also showed that there was only a limited eVect on the
socialist vote from economic conditions within the cities. There was
some evidence that the socialist vote was higher in the more prosperous
cities (where wages were higher), and in cities with large-scale industry,
measured by factory size, though the connections were not strong ones.
This in turn reXected the success of the party in widening its appeal
beyond the conWnes of its core constituency: workers in large-scale
industry.17 The presence of this core constituency is just discernible in
the results, but the eVect is only a slight one. Another weak Wnding (in
statistical terms) is the existence of a relationship between high mobil-
ity, the total number of moves in and out of each city in 1910–12
relative to population, and the size of the SPD vote. The socialist vote
rose in cities with more Xuid labour markets, but the result is not a
strong or a secure one.
Two factors which were shown to have virtually no eVect on voting
behaviour were labour market conditions and inequality. The Wrst was
measured by the ratio of applicants to vacancies at town labour
exchanges. Where economic conditions were adverse, the ratio of appli-
cants to vacancies would be high. This was not a factor which had any
discernible impact on the SPD vote. Inequality was measured by the
ratio between the estimated share of the top 10% of households in total
income and the estimated share of the bottom 40%. In the more
unequal towns, the top 10% had a share which was Wve or six times as
large as that of the bottom 40%; in the more egalitarian towns the
shares were about equal. But these diVerences had no signiWcant eVect
on votes cast in the 1912 election.18
The analysis of the vote in rural or small-town Prussian constituen-
cies is summarized in Table 10.4 (full results are in Table 10B.2). The
SPD share in these constituencies averaged 20.6%, which was rather
less than in the towns analysed earlier (42.2% on average). However,
the SPD advance in these small town and rural constituencies was an
important component of the party’s overall gains. The average share
had risen from 16.3% in 1907.
17 The SPD no longer represented a ‘homogeneous group of proletarians’, Fairbairn(1997), 216.
18 The methods used to produce these estimates of the income distributions in thePrussian Kreise are described in Grant (2002b). Eley (2002), 56–7, comments on the lackof a relationship between economic factors and the SPD vote in a comparison ofRemscheid and Hamborn: Hamborn workers were the ‘very epitome of the brutalised andexploited factory proletariat’ but the SPD was weaker there than in Remscheid.
340 Challenging the Kehrite View
![Page 350: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/350.jpg)
The analysis shows that there were two factors which explained
much of the variation in the SPD vote: the percentage of the population
occupied outside agriculture and the Catholic percentage. The agricul-
tural population was less likely to vote for the SPD, and the relative
decline of agriculture was a factor working in the SPD’s favour in these
rural constituencies.19 Catholicism was an important factor working
against the SPD especially in these Prussian areas. In the Polish-speak-
ing regions the SPD vote was held down even further, although it was
higher in constituencies where Lithuanians were present.
The SDP vote was highest in areas near to the larger cities (with over
100,000 inhabitants) but it was also high in areas near to towns with a
population of at least 20,000. But the party’s share was reduced in the
19 This result is also found in the analysis by Fairbairn (1997), 228. Fairbairn alsoshows that the negative eVect from agricultural employment became stronger between1890 and 1912, ibid. 230. As pointed out by Steinbach, electoral research in Germanhistory is ‘relatively immature’ and the type of multi-variable statistical analysis of elect-oral results presented in this chapter is rare for pre-1914 Germany, Steinbach (1992), 121.Fairbairn’s analysis is an exception. Suval (1985), 122, produces correlations which areinteresting but does not use multivariate analysis. Nor does Schmadeke (1995). Winkler(1993), 251 does not give t-statistics, so there is no way of assessing the quality of theresults shown.
Table 10.4. Summary of results of statistical analysis of the 1912 SPD
vote in 204 constituencies outside the major cities (Prussian
constituencies only)
Secure statistical results
I. Factors which were associated with a higher SPD vote:
A high level of non-agricultural employment
Large numbers of Gutsbezirke relative to other Gemeinden
A high Lithuanian-speaking population
II. Factors which were associated with a lower SPD vote:
Catholicism
A high Polish-speaking population
Presence of large farms and estates (over 100 hectares)
Remoteness from major cities (over 100,000 inhabitants) and from towns
(over 20,000 inhabitants)
III. Factors which had little or no influence on voting behaviour:
Levels of inequality (measured by the U10/L40 ratio)
Wage rates
Challenging the Kehrite View 341
![Page 351: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/351.jpg)
more remote constituencies, and this was an eVect which was independ-
ent of the occupational structure of these constituencies. Towns and
large cities were the centres of socialist activity, and this had an eVect
on nearby areas.
Economic variables had little eVect apart from the occupational struc-
ture. Neither wage levels nor inequality (as measured by income in-
equality) had any eVect on voting behaviour. This repeats the Wndingreported earlier from the analysis of votes in urban constituencies: eco-
nomic issues do not appear to have been important factors in the 1912
election. Inequality in land ownership, however, as measured by the
percentage of the agricultural area which was in large farms and estates
(over 100 hectares), was a factor working against the SPD. Traditions of
deference may have kept the socialist vote down, or fears of retaliation if
the secrecy of the ballot was not thought to be absolutely guaranteed.20
However, a factor which operated in the opposite direction was the
presence of Gutsbezirke. The SPD vote was higher in constituencies
where a large percentage of the local Gemeinden were classiWed as Guts-
bezirke: aristocratic estates which had in the past had an important
administrative function. The other categories were Landgemeinden and
Stadtgemeinden. Around 25% of the administrative units in these con-
stituencies were Gutsbezirke, though this exaggerates their importance
as they were usually much smaller than the Landgemeinden.
In the past the aristocratic owners of the Guter had had considerable
control over the local police and administration, and judicial functions
of their own. These were largely abolished by the Kreisordnung of
1872.21 What remained was an important position in local society, and
some Wscal privileges, notably an exemption from the supplement
charged on the Prussian income tax assessment to Wnance the expend-
iture of the Landgemeinden and Stadtgemeinden. Against this, the owners
of the Guter had to provide some services which would otherwise have
been paid out of local government funds.22
20 Anderson (2000), 157–9, argues that the secrecy of the ballot was not assured insome rural areas.
21 The judicial role of the Ritterguter was removed in 1849–51; the police functionswere abolished in 1872, von Below (1891), 675 V. However, the Ritterguter remainedseparate administrative districts until 1927, Puhle (1986), 85, and some historians haveargued that much of their authority and inXuence survived the 1872 Kreisordnung, Carsten(1989), 120.
22 The supplement could be a substantial one, an increase of 50% or more in 663Gemeinden according to Meisel (1911), 298. The aristocratic owners of the Guter alsobeneWted from favourable tax treatment from the Landrate (district administrators),according to Witt (1985). This would have been a further cause of resentment.
342 Challenging the Kehrite View
![Page 352: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/352.jpg)
Not all large farms and estates were Guter: there were many estates
that had been sold by their original owners or that had never the func-
tions of a Gut. And not all Guter were large agricultural holdings: some
estates had been broken up or rented out while still retaining the title
and privileges of a Gut. In places the system was a relic with little
practical relevance. What these results indicate is that this system still
caused resentment.
In general, economic factors (income inequality, wage levels, un-
employment, and enterprise size) did not have much eVect on the SPD
vote. The party was appealing to a broader constituency, concerned
with other issues. But political inequality, the sense of being excluded
from decision-making, whether at the local or the national level, did
matter. And this is the main signiWcance of the result with the Gutsbe-
zirke variable.23
It is not the case that the Gutsbezirke were a particularly onerous part
of the political and administrative system of Prussia in 1912. There
were other more important concerns for electors willing to vote for
radical change. Their presence happens to be something which can be
measured, and which varied considerably between diVerent constituen-
cies. Other issues, such as the three-class electoral system, aVected all
constituencies, so there is no variation in the importance of the variable
which can be picked up by statistical analysis. These results indicate
that there was a desire for change in the political and administrative
system, and that this desire for change was an important factor in the
SDP vote. Voters who wanted progress towards democracy turned to
the SDP as the party most likely to achieve this.24
The SPD responded to this new role. Its policy statements and leaX-
ets emphasized constitutional issues, and it signalled an increased will-
ingness to play a constructive role in the political process.25 Moreover,
and more dramatically, there was a shift in the social make-up of the
23 The question of political ‘fairness’, the sense of being excluded by the suVragesystem, was one of the central issues of Wilhelmine politics, Fairbairn (1992).
24 Another way to protest against the three-class electoral system was to abstain in localelections. The turnout in Landtag elections could fall as low as 8% for electors in the thirdclass, Anderson (2000), 196. In some rural areas the left-liberals also beneWted fromdissatisfaction with the electoral system, Thompson (2000), 301. The electoral system alsokept many German city administrations in liberal hands, despite SPD majorities, and itmay be surmised that resentment of this was a factor in the strength of the SPD vote inurban areas, Pogge von Strandmann (1992).
25 Bertram (1964), 169; this trend towards an emphasis on political objectives was inpart a reaction to the SDP’s setback in the 1907 election: Crothers (1941), 213 and 229;Schorske (1955), 225–6.
Challenging the Kehrite View 343
![Page 353: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/353.jpg)
SDP’s Reichstag candidates. In 1907 the party had Welded 132 working-
class candidates (Arbeiter, Handwerker, and Angestellte); by 1912 this
had fallen to 57. By contrast, the numbers of private sector managers
and oYcials (Privatbeamte) had risen from 50 to 175.26 Table 10.5 gives
the social breakdown of the SPD candidates. Private sector managers
and businessmen constituted a majority of the candidates oVered in
1912. The other striking feature of the table is the under-representation
of anyone in public sector employment: oYcials (OVentliche Beamte,
Justizbeamte), army oYcers, and teachers. These were just 1.5% of
26 The size of the shift raises the question whether it represented a genuine change, orwhether candidates previously identiWed as Angestellte or even Arbeiter were now classify-ing themselves as Privatbeamte. If so, it still represents a shift in the party’s approach:candidates were stressing their middle-class aYliations instead of seeking to appeal to anarrow working-class constituency.
Table 10.5. Social origins of the SDP candidates, 1907 and 1912
elections, compared to those of other parties (as percentages)
SPD
candidates
Candidates of
other parties
1907 1912 1907 1912
Entrepreneurs,
businessmen:
Agriculture
Commerce and industry
0.8
16.6
1.5
10.0
21.4
11.1
19.5
10.4
Managers, officials: Public sector 3.0 1.5 10.9 9.1
Legal 0.0 0.0 12.3 8.0
Private sector 12.6 43.9 1.6 4.9
Religious: Protestants 0.3 0.0 1.4 3.3
Catholics 0.0 0.0 4.7 3.6
Professional: Army officers 0.0 0.0 1.2 1.8
Lawyers 2.8 3.5 7.1 10.0
Doctors 0.5 0.5 2.3 2.7
Writers 23.1 23.3 9.5 9.5
Teachers 0.3 0.0 6.3 9.0
Employees: Workers 33.2 14.3 5.3 5.6
Others: 7.0 1.5 4.8 2.6
Total number of candidates 398 399 1285 1153
Source: Calculated from figures in Bertram (1964), 158.
344 Challenging the Kehrite View
![Page 354: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/354.jpg)
SPD candidates in 1912, but 28.6% of those put forward by other
parties.27
The composition of the SPD’s candidates reXected the shift in its
policy orientations: it represented the politically excluded middle classes
as well as the working class. The statistical analysis presented in this
section conWrms that economic issues, which would have been major
concerns of working-class electors, played a relatively unimportant role
in the explanation of the size of the SDP vote. The party was appealing
to a broader constituency: those who wanted political change as well as
those who wanted radical alterations in the structure of society.
4. Economics and the T irp itz Plan :
The Forces Driv ing Germany towards War
The weakness of the political structures of the Bismarckian Reich had
left voters who wanted political change with little alternative but to vote
for a party which, in its original policy statements, had stood for a far-
reaching programme of revolutionary social and economic change. The
rapprochement between the aims of the SDP leadership and those of its
newer and wider base was a gradual one, as the party shifted its em-
phasis from economic issues to the pursuit of political reform, and
began to indicate its willingness to work within the system. However,
the growing strength of the party and the desire for change which it
represented presented a serious problem for the Bismarckian system.
Could the system evolve in a democratic direction without a violent
upheaval?
It is an open question whether this could have been accomplished if
the only issues facing pre-1914 Germany had been internal ones. Could
concessions have been made which would have strengthened parliamen-
tary control without conceding full majority rule? How would the
strengthened Reichstag authority have been regarded by the non-
Prussian states? How far were the conservative forces opposed to
further democratization prepared to go to obstruct such a settlement?
Would the radicals within the SPD have accepted a compromise?
But this is all speculation. For the issues of democratic control and
accountability which were raised by the trajectory of electoral results
27 It is likely that public sector employees would have been reluctant candidates,fearing discrimination against anyone who put themselves forward as candidates for an‘unpatriotic’ party.
Challenging the Kehrite View 345
![Page 355: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/355.jpg)
meant that the Kaiserreich confronted the problem of constitutional
change at just the same time as the trajectory of Germany’s diplomatic
isolation and weakening strategic position caused a series of crises in
external politics, one of which ended in the outbreak of the First World
War. If there was ever a chance of achieving a normal transition to a
democratic settlement, this disappeared in July 1914.
To consider the background to the deteriorating external position it
is worth returning to one of the arguments put forward in Chapter 8,
when explaining why economic developments had created the condi-
tions for political stabilization post-1900: the favourable prospects for
German trade. Germany was able to use its comparative advantage in
the production of industrial goods, particularly in the ‘high technology’
sectors, to gain market share in world trade, and it could then use these
revenues to import an increasing volume of food and raw materials.28
This was a development which had profound strategic implications, and
it meant that relations with Britain, previously a peripheral issue in
German diplomacy, were now of central importance.
In the 1860s Bismarck was able to largely ignore British positions on
issues such as the annexation of Schleswig-Holstein and the progress
towards German uniWcation: little more than lip service was paid to
British concerns, British oVers of mediation were treated with barely
disguised condescension, and substantive British proposals were mostly
ignored. But this reXected the realities of the position: Britain was in no
position to intervene eVectively in a land war in continental Europe;
Prussia was a largely self-suYcient economy, whose ability to wage war
would not be aVected by a British blockade and which had few non-
European interests vulnerable to British sea power.29
To have maintained a position of self-suYciency and strategic inde-
pendence would have required policies which supported a higher level
of agricultural production than that achieved under the Kaiserreich.
TariVs would have had to be high enough to enable the grain sector to
match production costs in the American Mid-West. Resources would
have had to have been maintained in agricultural production, not
switched to industry, requiring a substantial reduction in the rate of
rural–urban migration. To achieve this, agricultural prices and incomes
28 The shift in the German position with regard to cereals was given in Table 7.1.There were substantial increases in the importation of other foodstuVs and raw materials.
29 Bismarck’s remark that Britain was a quantite negligeable in European politics wasquite justiWed from a Prussian point of view as long as Prussia remained solely a Contin-ental power, Kennedy (1980), 191.
346 Challenging the Kehrite View
![Page 356: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/356.jpg)
would have had to be increased relative to urban levels. German
farmers would have beneWted at the expense of the urban consumer.
By choosing a moderate level of protection, which did not prevent a
substantial rise in imports, Imperial Germany adopted a policy which
provided some relief to agriculture but which did not prevent an in-
crease in living standards for the urban working class. Improved
incomes for workers contributed to the post-1896 fall in the urban
Gini-coeYcient. This in turn created economic conditions favourable to
the normalization of internal politics. It raised the likelihood that the
Social Democratic Party could be persuaded to turn away from revolu-
tionary politics and become an accepted part of the political system.
Germany faced a genuine socio-strategic dilemma. Economic self-
suYciency could be maintained only by the pursuit of policies which
were clearly hostile to the interests of the urban working class and
which would have perpetuated the alienation of that class. By contrast,
rising urban living standards could be secured only by recourse to
imported foodstuVs and raw materials. To a considerable extent these
came from British possessions, or from areas dominated by British
capital, and the overwhelming proportion crossed seas controlled by the
Royal Navy.30
Even in the event of a war with France and Russia, from which
Britain stood apart, the state of the German navy around 1900 was a
serious strategic deWciency. In 1903 Germany was outnumbered by
France alone. Germany had 9–12 ships of the line; France had 21. It
would only be possible to have a local superiority over the French
Atlantic Xeet if the French Mediterranean Xeet were prevented from
moving north and if the Russian Baltic Fleet did not join with the
French.31
To see Germany’s Flottenpolitik solely in terms of its eVect on in-
ternal politics, as historians in the Kehrite school tend to do, to see it as
a part of a Sammlungspolitik, a cause for the rallying of forces loyal to
the Empire, is to ignore or obscure this genuine dilemma.32 The eco-
nomic facts are clear: Germany in the late 1890s was not in the same
30 That there were particular problems associated with Germany’s geographical pos-ition is a point made in Sturmer (1990), but this study does not deal with the interactionof geography and the evolution of foreign trade.
31 Buchel memorandum, 27 Mar. 1903, Lambi (1984), 203.32 There is a way in which some Kehrite historians have linked the Flottenpolitik to
trade issues and that is to argue that German industry needed colonial expansion to gainnew markets, which in turn required a powerful Xeet, Wehler (1969), 112–26.
Challenging the Kehrite View 347
![Page 357: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/357.jpg)
position as Prussia in the 1860s. Ignoring Britain was no longer an
option. Bismarckian condescension was not an appropriate policy. The
trade routes which brought imported food and colonial products to
Germany crossed those which brought the same goods to Britain. Only
one power could control these seas. Germany would have to confront or
conciliate.
In the Kehrite literature, the existence of a genuine strategic dilemma
is either downplayed or ignored altogether. Economic factors are not
considered as a part of the geo-political equation. Yet the German
government was certainly aware of the importance of foreign trade to
the German economy. Hohenlohe wrote in a letter in 1897: ‘Germany’s
trade has grown to such an extent that the government is obliged to
protect it. This can only be done by a Xeet which is more than just a
coastal protection force, but one capable of ensuring the safe passage of
our imports.’33 And the Kaiser wrote ‘richtig’ in the margin of a gov-
ernment report against the following passage: ‘without her overseas
trade Germany could not exist, and this trade would not prosper for
long without the protection of a Xeet capable of commanding respect’.34
Tirpitz himself was well aware of the economic implications of the
plan. He saw an intimate connection between the expansion of the Xeet
and the growth of German trade. In his public statements he tended to
play down the trade implications in order to avoid antagonizing the
agrarian lobby, who would have preferred Germany to remain self-
suYcient in food, rather than exporting industrial products to pay for
food imports. But there was an awareness of the strategic consequences
of the changing structure of German imports.35
Critics of the Tirpitz naval programme need to explain how
Germany could have resolved this dilemma. What was the alternative
to building a navy? A British alliance was one possibility, and this was
considered at various times by Germany’s leaders. Yet this had the clear
implication that Germany could be drawn into conXict with other Euro-
pean powers (most probably Russia) as a result of disputes arising from
Britain’s imperial interests and possessions. The British Empire would
33 Letter from Hohenlohe to SoldendorV, 7 Nov. 1897; quoted in Berghahn (1971),133.
34 The document was a report from the Prussian representative in Bavaria, Berghahn(1971a), 134. In 1899 Tirpitz predicted that, without a Xeet, Germany would sink back tothe status of a poor agricultural country, Herwig (1987), 35.
35 Deist (1976), 110–12. Trade issues were, however, used to gain the support ofinXuential lobby groups and industrial organizations, Pogge von Strandmann (1972).
348 Challenging the Kehrite View
![Page 358: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/358.jpg)
be defended ‘on the banks of the Elbe’, at immense cost to Germany.36
To have eschewed naval expansion in the absence of a formal British
alliance would have meant that Germany would have little choice but to
accept British mediation in the event of a prolonged European war.
Britain would be the arbitrator of Europe, with unpredictable conse-
quences. In addition, German trade would have been unprotected from
attack by the French or Russian navies should such a war involve these
powers.37
Eventually, after 1945, West Germany was able to reap the rewards,
in terms of rising living standards and political stabilization, of a trade-
orientated economic policy. But this was done in the context of an
American alliance which did not just provide security to West German
trade routes but also oVered an eVective guarantee of land frontiers as
well. A pre-1914 British alliance oVered only the Wrst part of this: trade
security at the expense of a more dangerous position on land.
To recognize this dilemma is not to deny that internal weaknesses
had an adverse eVect on German foreign policy. A more far-sighted
foreign policy, more Xexible and less inXuenced by domestic political
considerations, might have kept potential opponents isolated, might
have prevented the formation of a power bloc hostile to Germany,
might have pursued a policy of naval expansion less obviously driven
by hostility to British interests, and might, in short, have avoided
the drift towards war on terms which were unfavourable to Germany.
The contrast with British policy is instructive. Once German policy had
shifted to one of evident confrontation, the British response was indeed
far-sighted and Xexible. Outstanding disputes with France, Russia,
and the United States were settled, even at the cost of giving up long-
held positions (conceding a Russian sphere of inXuence in Persia, recog-
nizing the Monroe doctrine). An empire built on an implicit sense of
racial superiority concluded an alliance with Japan. German policy, by
contrast, was hesitant and needlessly antagonistic. Internal weaknesses
had a part in this. But Germany was not the only country pre-1914
whose foreign policy was adversely aVected by internal weakness.
36 As the elder Pitt had claimed to have won Canada in 1759, by distracting France inGermany.
37 The Kehrite historians’ concentration on the Tirpitz plan has obscured the fact thatmuch of German naval planning was concerned with France and Russia—for example, the1899 plan, Lambi (1984), 193. Any navy large enough to match the combined strength ofFrance and Russia, and therefore give Germany the option of planning for a prolongedwar against these two countries, was also large enough to be a threat to Britain.
Challenging the Kehrite View 349
![Page 359: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/359.jpg)
Russia and Austria-Hungary are two obvious examples. It was the qual-
ity of the British policy response to Germany’s challenge which was
truly exceptional.38
Kehrite historians have ignored the economic dimension of the
Tirpitz plan, and concentrated on the internal political aspects. The
result has been to underestimate the genuine dilemma faced by German
policy-makers, and to present the plan as a cynical attempt by an out-
moded elite to retain power and obstruct political reform. Even if such
motives did play a part, the success of the naval programme in at-
tracting widespread support owed a lot to the economic imperatives
which lay behind it.
5. Growth in Mature and in Industrializ ing Societies :
The Economic Preconditions for Polit ical Progress
Economic growth in mature industrial societies diVers from that in
industrializing societies in that the main engine of growth is improve-
ments in productivity within each sector not the transfer of labour and
other resources from sectors where productivity is low into more
modern sectors. In mature economies the proportion of the labour force
in agriculture and other lagging sectors (if agriculture is still lagging) is
typically a low one, below 10%. Further transfers from this sector can
make only a modest contribution to overall growth.
In the early stages of industrialization, structural shifts are an import-
ant component of overall growth. Indeed, productivity growth within
each sector is often quite modest, so that the contribution of structural
change is a vital one. Thus a high rate of migration out of agriculture
into the urban or industrial sector is intrinsic to the growth process, as
well as a major force of social transformation.
The Lewis Model is valuable in that it provides an explanation for
some features of the operation of urban labour markets in industrial-
izing countries, and may help to explain the high level of urban inequal-
ity found in Germany and other countries. But this is only part of the
broader process of social change brought about by rural–urban
38 The contempt with which German foreign policy is treated by German historianscan come as a surprise: was German policy really that of a ‘deluded upstart’, Hildebrand(1995), 10.
350 Challenging the Kehrite View
![Page 360: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/360.jpg)
migration: a rural society bound by traditional ties of obligation, defer-
ence, and reciprocity is cast aside and a new urban structure, built on
very diVerent principles, has to be constructed in its place. Inevitably,
this takes time. Initially, the sense of dislocation produced by migration
is intensiWed by the contrast between the stability of rural institutions
and values compared to those of an urban society which is in a state of
Xux.As the growth process continues, the contribution made by structural
change will tend to fall, unless exceptional factors intrude (such as the
check to structural change experienced by Germany and many other
European countries in 1913–50, which left room for a rapid catch-up in
1950–73). As the rural labour surplus is used up, the opportunities for
rapid growth as a result of the redeployment of labour which was
previously underemployed will be reduced. Migration in and out of
urban labour markets will become a stabilizing rather than a destabil-
izing process.
The analysis of the previous chapters has concentrated on the social
and economic consequences of the ‘labour surplus’ phase of industrial-
ization. It has looked at the eVects of migration out of agriculture into
urban labour markets, it has considered the movement of inequality in
the rural and urban sectors, and it has examined the eVects of change
and transformation in agriculture. Standing back from these trends, it is
possible to discern the broad movements: a pattern of sharpening ten-
sions and divisions up to the 1890s followed by a period of stabilization
and improvement. The elements of this better position post-1895 are
the following:
1. An end to the ‘labour surplus’ pattern of migration dominated by
the transfer of surplus population out of agriculture. The sharp
decline in emigration to the United States after 1895 is a further
indication of the changing nature of migrant behaviour.
2. The stabilization and slight decline in the overall Prussian Gini-
coeYcient after 1900, combined with the fact that inequality within
the rural and urban sectors was declining from 1896 onwards.
3. The improved position of German agriculture after 1895, as indi-
cated by rising land prices, the reduction in forced sales, and
increases in rural wage levels. This in turn reduced the pressure
on urban labour markets.
4. The recovery of the German capital markets from the crisis of the
1870s was complete by 1900; the robustness of the new system
Challenging the Kehrite View 351
![Page 361: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/361.jpg)
was demonstrated by the fact that investment levels remained
high despite short-term Wnancial crises in the post-1900 period.
The urban sector was better able to Wnance the creation of jobs for
rural migrants.
5. The strong growth of German exports, particularly in ‘high tech’
sectors, promised well for the prospects of the German economy in
the twentieth century. So long as world trade remained open, and
volumes increased, Germany could expect a steady improvement in
its share of world exports, making it possible to Wnance an increased
bill for imported raw materials and foodstuVs, and to deliver con-
tinued improvements in living standards for German citizens.
From the mid-1890s onwards the German economy developed in
directions which created the preconditions for social and political stabil-
ization.39 These favourable preconditions were not enough, in them-
selves, to guarantee the successful transformation of the Kaiserreich into
a modern democracy. Even if there had been no outbreak of war in
1914 there was still much that could have gone wrong. But circum-
stances were more favourable in the early twentieth century than they
had been in the 1870s and 1880s.
In the end, though, the weakness of the Kehrite argument lies in the
normality of the pre-1914 German experience. Germany went through
a process of rapid industrialization, and this brought with it economic
and social problems common to many countries going through this
phase. The tools needed to understand this are available in the com-
parative study of industrializing societies. There was no Sonderweg
before 1914. The outbreak of war in 1914 had other roots, other causes.
6. July 1914: A Derailment or the End of a Sackgasse?
The idea of this study was conceived during a visit to China in 1995.
There was an evident labour surplus in the countryside; the migration
of this surplus to already crowded cities was clearly going to have
serious social, economic, and political consequences. This observation
39 This view contrasts with the one put forward by some historians that industrializa-tion inevitably led to new social conXicts, Puhle (1976), 340. For GeoV Eley, the ‘centralcontradiction of Wilhelmine politics’ was failure of the state to adapt to the new task ofsocial reconciliation caused by the basic ‘antagonism of capital and labour’, Eley (1980),349. The eVect of economic and social change has also been blamed for the readiness ofthe middle classes to embrace radical nationalist ideas, Chickering (1984), 302.
352 Challenging the Kehrite View
![Page 362: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/362.jpg)
prompted a general reXection. If there is a labour surplus period in
economic development—a ‘hump’ which many countries have to sur-
mount—does this aVect the stability of countries going through this
process? Are they potentially dangerous factors in world politics? Is this
an explanation of the political problems of pre-1914 Germany? Did
these in turn aVect German foreign policy? Will a similar process make
China a disruptive force in world politics in the twenty-Wrst century?The conclusions are in many ways more optimistic than the fears
which were prompted by this visit. The autonomy of the foreign policy
problems and the internal problems of the German empire is the main
conclusion to emerge. While this is not a conclusion which necessarily
applies to every state in every era, it is at least more comforting than the
implication of the Kehrite view: that a state which lacks the ability to
tackle its internal problems will turn to an aggressive foreign policy in
an attempt to defuse domestic tensions.
The picture given in this study is one of a society which, by 1914,
was having considerable success in dealing with the problems of rapid
industrialization and the ‘labour surplus’ period. These were problems
which were unprecedented in their scale and intensity. Germany 1870–
1913 was the Wrst ‘developing economy’, in the sense that it was the
Wrst to go through a process of telescoped industrialization and fast
economic growth characteristic of countries ‘catching up’ with a tech-
nological leader. The speed of German industrialization and urbaniza-
tion was rapid, perhaps three times as fast as Britain at a similar stage.
If there is a connection between the economic and political problems
of Imperial Germany, it probably lies in an earlier period, before 1895.
Evidence presented in this study shows that the time of maximum
rural–urban migration was also one of intense urban problems: high
inequality and low life expectancy. It was during this period that Marx-
ist ideas became inXuential in working-class districts. This develop-
ment, and the violence of the reaction by the Bismarckian state which it
provoked, had a profound eVect on the political system of Imperial
Germany. It is in this period that the ‘Weber–Lewis–Kuznets’ hypoth-
esis has most value for those seeking to explain the course of German
history.
But by 1914 the chances of democratic transformation were much
improved. There remained massive obstacles. For many historians,
these were so great as to be insuperable.40 Consequently, it can be
40 It was a situation of ‘intolerable complexity’, according to Eley (1980), 350.
Challenging the Kehrite View 353
![Page 363: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/363.jpg)
argued that the rising demand for change inevitably created tensions
which, even if these were not directly linked to German policy during
July 1914, produced a background which aVected the decision-makers’
ability to resolve the crisis. The atmosphere of pessimistic fatalism, the
general weariness with the political process and the compromises it
required to keep operating, a willingness to take risks and consider
radical solutions—these were dangerous sentiments in a context which
required clear thought and an ability to appreciate the consequences of
a mistaken policy.41
The view that the Kaiserreich was in a political cul-de-sac, a Sack-
gasse, where the options open to the executive were narrowing, is one
which is held by many historians outside the Kehrite school.42 The
arguments of Manfred Rauh, that a process of ‘quiet Parliamentariza-
tion’ was already under way, have been widely derided.43 However, if
the option of peaceful progress was ruled out, then the solution would
have had to have been a violent one of some sort.
But, the implications of the arguments about economic preconditions
put forward in this study are that the path to democracy was getting
easier, not more diYcult. In which case, even if Rauh overstated the
extent of the progress made by 1914, the prospects for further advances
were improving. The events of 1914 represented a derailment, not the
hitting of the buVers at the end of a Sackgasse.
Appendix 10A. Description of Regress ion Variables
and Sources
(i) Appendix Table 10B.1
SDP1912
(dependent variable)
Social Democrat share of the vote in urban
constituencies, from Die Reichstagswahlen von
1912, SDR vol. 250 (1913).
41 The view that the policy which led to the outbreak of war was based on an inabilityto appreciate the consequences of German policy, leading to the mistaken view that theconXict could be limited, was put forward in Jarausch (1969). But the consensus now isthat Germany pushed for war, Pogge von Strandmann (1988).
42 The term Sackgasse is taken from the title of a paper by Volker Berghahn, Berghahn(1971).
43 The phrase is found in Rauh (1973), 347. The point is made more forcefully in Rauh(1977).
354 Challenging the Kehrite View
![Page 364: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/364.jpg)
POPULATION(log) Log of total population in 1910, from SJDS
(1911), 67–8.
CATHOLIC% The percentage of the population who were
Catholic, 1910 census, SJDS (1911), 678–9.
PRUSSIA-CATH% As above, Prussian cities only.
NONPR-CATH% As above, non-Prussian cities only.
MOBILITY Average moves in and out, divided by total
population. Data from SJDS (1911), 67–8, and
subsequent issues.
TOWNWAGES1912 Average wages (Ortsubliche Tagelohne) for male
workers, aged 16 and over, from SJDS (1913),
826.
FACTORYSIZE Total number of employed workers divided by
the number of enterprises, 1912, from SJDS
(1916), 386–7.SDP1878 Social Democrat share of the vote in urban
constituencies, from SDR a.f. vol. 37 (1879).
INEQUALITY Estimated levels of inequality, as ratios of the
income shares of the top 10% to the bottom
40% (U10/L40 ratios). Data from Kuhnert
(1916); estimation procedure is given in
Appendix B of Grant (2002b).
(ii) Appendix Table 10B.2
SDP1912
(dependent variable)
Social Democrat share of the vote in non-
urban constituencies, from Die Reichstagswah-
len von 1912, SDR vol. 250 (1913).
%AGOCCUP07 The percentage of the population occupied in
agriculture, from the 1907 occupational census,
SDR vol. 209 (1909).%CATHOLIC The percentage of the population who were
Catholic, 1890 census, PS vol. 121, pp. 152–98.
%SPEAKPOLISH As above, the percentage of the population
who spoke Polish.
%SPEAKLITHUAN As above, for Lithuanian speakers.
CITYDISTANCE(a) Distance in kilometres from the Kreis mid-
point to the nearest city with at least 200,000
inhabitants in 1900 (measured from contem-
porary maps).
Challenging the Kehrite View 355
![Page 365: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/365.jpg)
CITYDISTANCE(c) As above, for cities with at least 50,000 inhab-
itants.
LAND>100HA The percentage of the total agricultural area in
holdings of over 100 hectares, from 1895 Agri-
cultural Census, SDR n.f. 212.
LAND20-100HA As above, for holdings of between 20 and 100
hectares.
%GUTSBEZIRKE The numbers of Gutsbezirke as percentages of
all administrative units, from PS vol. 81
(1884), part 1, pp. 20–65.
INEQUALITY As Table 10.1 (averages of Kreis results for
each constituency).
RURALWAGES Average wages (Ortsubliche Tagelohne) for rural
areas; the Wgures are averages of male and
female daily rates, 1892 and 1901, ZdKPSB(1904), 520–8.
Appendix 10B. Econometric Analysis
The dependent variable in these regressions was the share of the Social
Democrats in the 1912 election. The analysis showed that the SDP vote
was highest in the larger cities (the log of the resident population has a
strong eVect on voting behaviour), and also that the presence of a large
Catholic population held it down (column a of Table 10B.1). In the
second column a distinction is drawn between non-Prussian cities and
Prussian cities. The inXuence of Catholicism is shown to have been
signiWcantly greater inside Prussia.
In the other columns a variety of economic variables are introduced.
These turned out to have, at the most, a slight eVect on voting behav-
iour. Two labour market indicators were used. MOBILITY is a meas-
ure of migration (total moves per head of population for 1910–12); this
was weakly linked with an increase in the SPD vote, but the result was
neither robust nor statistically signiWcant. A second indicator, the ratio
of applicants to jobs reported by the labour exchanges, was completely
insigniWcant (a t-statistic of 0.16 when added to the equation estimated
in column c).44 Short-term movements in job market conditions did not
44 The variable gave the ratio of applicants to vacancies for 1910–12, derived fromtown labour exchange statistics (SJDS 1911–13).
356 Challenging the Kehrite View
![Page 366: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/366.jpg)
Table 10B.1. Regression results for linear log-odds voting model, 1912 election: dependent variable is logistic
transformation of the SDP vote in major cities; estimated by OLS
a b c d e f
Constant �1.51 �1.55 �1.94 �2.30 �1.22 �2.91
POPULATION(log) þ0.332 (3.79) þ0.341 (4.28) þ0.350 (4.10) þ0.314 (3.48) þ0.167 (2.26) þ0.271 (2.35)
CATHOLIC% �0.0143 (5.72) �1.985 (6.90)
PRUSSIA-CATH% �0.0183 (7.34) �0.0183 (6.98) �0.0188 (7.72) �0.0086 (2.87)
NONPR-CATH% �0.0068 (2.29) �0.0026 (0.67) �0.0079 (2.65) �0.0039 (1.46)
MOBILITY þ0.0098 (1.22)
TOWNWAGES1912 þ0.263 (1.51) þ0.228 (1.75) þ0.617 (2.75)
FACTORYSIZE þ0.0022 (1.55) þ0.0016 (1.27) þ0.0010 (0.57)
SDP1878 þ0.232 (4.61)
INEQUALITY �0.063 (0.72)
N 70 70 65 70 55 46
S 0.567 0.516 0.512 0.503 0.332 0.520
R2 0.435 0.540 0.539 0.576 0.706 0.637
adj R2 0.418 0.519 0.519 0.543 0.669 0.592
F-test of regression 25.82 25.84 19.70 17.40 19.19 14.03
RSS 21.563 17.560 15.699 16.185 5.302 10.822
![Page 367: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/367.jpg)
appear to matter at all. Wage levels were positively linked with the
socialist vote (column c), as was factory size (the average number of
employees per enterprise with at least 10 workers), but both results are
below the 10% signiWcance level.
In the fourth column the socialist share of the vote in 1878 was
included (SPD1878). As might be expected, this has an eVect on the
other coeYcients. It shows the persistence of the socialist vote and the
importance of the early establishment of a separate working-class
culture and organization.
Finally, measures of income inequality derived from Prussian tax
statistics were included: the ratio of the share of the top 10% of the
distribution to that of the bottom 40% (INEQUALITY). This meant
that the analysis could only be run for the 46 Prussian cities. The
results showed no connection between inequality and the socialist vote.
Inequality does not appear to have been a factor in the electoral per-
formance of the SPD in urban areas.
Table 10B.2 gives the results of a similar analysis for the remaining
Prussian constituencies. This is a data set covering 204 rural or small
town constituencies. Two variables have a major eVect on voting behav-
iour: the percentage occupied in agriculture and the Catholic propor-
tion. Both coeYcients are strongly negative, and the two variables
together explain much of the variance in the dependent variable
(column a). Again, the slow relative decline of the agricultural popula-
tion was a factor working in favour of the SDP. Adding the percentage
of the population who spoke Polish (column b) conWrms the point made
earlier, that there remains a strong negative eVect from Catholicism
even when allowance is made for this factor. However, Lithuanian-
speakers showed a preference in favour of the SPD, so the presence of
this group tends to raise the SPD vote (these were found in Memel and
a few other constituencies in the Gumbinnen region of East Prussia).
Cities also had an eVect on the surrounding countryside. The two
city distance variables, measuring the distance from major cities, both
have coeYcients that are negative and signiWcant: the SPD vote tended
to fall as the distance from major cities rose. The eVect of the larger
cities (CITYDISTANCE(a) over 200,000 inhabitants) is larger than
that for smaller cities (CITYDISTANCE(c) over 50,000 inhabitants),
as would be expected given the strength of the SPD vote in the major
cities.
The eVect of large farms and estates was examined using two vari-
ables. The Wrst was the percentage of agricultural land in holdings of
358 Challenging the Kehrite View
![Page 368: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/368.jpg)
Table 10B.2. Analysis of the SDP vote in Prussian constituencies (excluding urban constituencies), 1912.Dependent variable is logistic transformation of vote
a b c d e
Constant 1.417 1.262 1.588 0.816 1.307
%AGOCCUP07 �0.0473 (14.9) �0.0371 (12.1) �0.0383 (11.9) �0.0360 (10.8) �0.0370 (12.0)
%CATHOLIC �0.0240 (15.4) �0.0181 (11.2) �0.0182 (11.2) �0.0179 (10.9) �0.0183 (10.8)
CITYDISTANCE(a) �0.0127 (4.37) �0.0128 (4.42) �0.0126 (4.30) �0.0128 (4.37)
CITYDISTANCE(c) �0.0110 (2.37) �0.0110 (2.37) �0.0102 (2.15) �0.0107 (2.27)
%SPEAKPOLISH �0.0122 (4.25) �0.0127 (4.41) �0.0122 (4.27) �0.0121 (4.24)
%SPEAKLITHUAN þ0.0281 (2.37) þ0.0294 (2.48) þ0.0298 (2.47) þ0.0281 (2.36)
LAND>100HA �0.0199 (3.80) �0.0182 (3.37) �0.0206 (3.88) �0.0200 (3.80)
%GUTSBEZIRKE þ0.0296 (5.67) þ0.0295 (5.67) þ0.0321 (5.31) þ0.0294 (5.61)
INEQUALITY �0.225 (1.28)
RURALWAGES þ0.150 (0.83)
LAND20-100HA �0.00124 (0.36)
N 204 204 204 204 204
S 0.810 0.661 0.660 0.662 0.663
R2 0.680 0.793 0.795 0.794 0.793
adj R2 0.677 0.785 0.785 0.784 0.784
F-test 214.00 93.44 83.52 83.00 82.70
RSS 131.75 85.30 84.59 85.00 85.25
![Page 369: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/369.jpg)
over 100 hectares (LAND>100HA). This variable had a negative eVect
on the SDP vote: workers on large farms and estates tended to be less
likely to vote for the SPD, perhaps for reasons of deference or loyalty to
the estate-owner, or possibly out of fear that such a vote would be
discovered and lead to retribution. However, the second variable,
%GUTSBEZIRKE, has a positive sign. This had a persistent eVect on
voting behaviour, and this result is robust to the inclusion or exclusion
of other variables. The estimated coeYcients for this variable are unex-
pectedly large given the relative unimportance of the Guter in this per-
iod. They are rather higher than the coeYcients on LAND>100HA,
indicating that resentment of the Guter was apparently a rather stronger
eVect on voting behaviour than deference on larger holdings.
The remaining columns introduce some other variables: the results in
column d show that the level of wages did not aVect voting behaviour,
in contrast to the results for the larger cites. Those in column e show
that the presence of medium-sized holdings did not aVect the results
either (these holdings typically employed some labour). But perhaps the
most interesting result from this table is in column c, which shows that
inequality at the Kreis level, measured by income inequality, had no
signiWcant eVect on the SPD vote. This conWrms the result given in
Table 10.B1, column f, that the level of inequality was not a factor in
the SDP vote in the cities. Political inequality, as represented by the
%GUTSBEZIRKE variable, had much more eVect.
360 Challenging the Kehrite View
![Page 370: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/370.jpg)
REFERENCES AND SOURCES
a. abbreviations used for main statist ical sources
1. Publications of the Imperial Statistical OYce
ErgVSDR Erganzungsheft zur Vierteljahrsheft zur Statistik des Deutschen Reichs
MSDR Monatsheft zur Statistik des Deutschen Reichs
SDR Statistik des Deutschen Reichs
SJDR Statistisches Jahrbuch fur das Deutsche Reich
VSDR Vierteljahrsheft zur Statistik des Deutschen Reichs
2. Publications of the Prussian Statistical OYce
PS Preußische Statistik
SJPS Statistisches Jahrbuch fur das Preußischen Staat
ZdKPSB Zeitschrift des Koniglich Preußischen Statistischen Bureaus
3. Other publications
SJB Statistisches Jahrbuch der Stadt Berlin
SJDS Statistisches Jahrbuch deutscher Stadte
SJK Statistisches Jahrbuch der Stadt Koln
b. references
Abel , W. (1962), Geschichte der deutschen Landwirtschaft vom fruhen Mittelalter
bis zum 19. Jahrhundert (Stuttgart: E. Ulmer).
Abelshauser , W. (1983), Wirtschaftsgeschichte der Bundesrepublik Deutschland
1945–1980 (Frankfurt am Main: Suhrkamp).
Albrecht , K. (1927), ‘Die Preußischen Gutsbezirke’, Zeitschrift des Preuß-
ischen Statistischen Landesamts, 3: 344–477.
Amsden , A. (1989), Asia’s Next Giant: South Korea and Late Industrialization
(Oxford: Oxford University Press).
——(2001), The Rise of ‘the Rest’: Challenges to the West from Late-Industrial-
izing Economies (Oxford: Oxford University Press).
![Page 371: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/371.jpg)
Anand , S. and Kanbur , S. M. (1993), ‘The Kuznets process and the
inequality–development relationship’, Journal of Development Economics, 40/
1: 25–52.
Anderson , M. L. (2000), Practicing Democracy: Elections and Political Culture
in Imperial Germany (Princeton: Princeton University Press).
Asian Productivity Organisation (1994), Agrarian Structure and Reform Measures
(Tokyo: APO).
Atkinson , A.B. (1996), ‘Income distribution in Europe and the United
States’, Oxford Review of Economic Policy, 12: 15–27.
——(1996a), ‘Seeking to explain the distribution of income’, in J Hills (ed.),
The New Inequalities: The Changing Distribution of Income and Wealth in the
United Kingdom (Cambridge: Cambridge University Press).
——and Brandolini , A. (1999), ‘Promises and Pitfalls in the Use of ‘‘Sec-
ondary’’ Data-Sets: Income Inequality in OECD Countries’, manuscript,
NuYeld College, Oxford and Banca D’Italia, Research Department.
——Rainwater , L. and Smeeding , T. M. (1995), Income Distribution in
OECD Countries, OECD Social Policy Studies no. 18.
Aug E-Larib E, M. (1950), La Politique agricole de la France de 1800 a 1940
(Paris: Presse Universitaire).
Baack , L. J. (1977), Agrarian Reform in Eighteenth Century Denmark, Univer-
sity of Nebraska Studies 56.
Bade , K. J. (1980), ‘German emigration to the United States and continental
emigration to Germany in the late nineteenth and early twentieth centuries’,
Central European History, 12/4: 348–77.
——(1984), ‘Vom Auswanderungsland zum ‘‘Arbeitseinfuhrland’’: kontinentale
Zuwanderung und Auslanderbeschaftigung in Deutschland im spaten 19. und
fruhen 20. Jahrhundert’, in K. J. Bade (ed.), Auswanderer, Wanderarbeiter,
Gastarbeiter (OstWldern: Scripta Mercaturae Verlag).
——ed. (1987), Population, Labour and Migration in 19th and 20th Century
Germany (Leamington Spa: Berg).
——(1991), ‘German transatlantic emigration in the nineteenth and twentieth
centuries’, in P. C. Emmer and M. Morner (eds.), European Expansion and
Migration (New York: Berg).
Baines , D. (1994), ‘Population, migration and regional development,
1870–1939’, in R. Floud and D. McCloskey (eds.), The Economic History of
Britain since 1700, vol. 2 (Cambridge: Cambridge University Press).
Bairoch , P. (1989), ‘Les trois revolutions agricoles’, Annales, 44/2: 317–54.
Ballod , C. (1914), ‘Die Bevolkerungsbewegung der letzten Jahrzehnte in
Preußen und in einigen anderen wichtigen Staaten Europas’, Zeitschrift des
Koniglich Preußischen Statistischen Bureaus, 3: 237–90.
Bardhan , P. (1989), ‘Alternative approaches to the theory of institutions in
economic development’, in P. Bardhan (ed.), The Economic Theory of Agrar-
ian Institutions (Oxford: Oxford University Press).
362 References and Sources
![Page 372: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/372.jpg)
Barfuss , K. M. (1985), Gastarbeiter in Nordwestdeutschland 1884–1918
(Bremen: Staatsarchiv).
Barkin , K. D. (1970), The Controversy over German Industrialisation
1890–1902 (Chicago: University of Chicago).
Basu , K. (1997), Analytical Development Economics: The Less Developed Econ-
omy Revisited (Cambridge, Mass.: MIT Press).
Bateman , J. (1879), The Great Landowners of Great Britain and Ireland
(London: Harrison).
Baten , J. (1999), Ernahrung und wirtschaftliche Entwicklung in Bayern
(1730–1880), Beitrage zur Wirtschafts- und Sozialgeschichte 82 (Stuttgart:
F. Steiner).
Bauerlicher Zustand in Deutschland (1883), Schriften des Verein fur Sozialpolitik,
vols. 21–4.
Below , G. von (1891), ‘Rittergut’, in Handworterbuch der Staatswissenschaft,
suppl. vol. I: 672–5.
Benabou , R. (1996), ‘Inequality and growth’, NBER Macroeconomics Annual,
xi: 11–74.
Berger , S. (2000), Social Democracy and the Working Class in Nineteenth and
Twentieth Century Germany (Harlow: Longman).
Berghahn , V. (1971), ‘Das Kaiserreich in der Sackgasse’, Neue Politische
Literatur, 16: 494–506.
——(1971a), Der Tirpitzplan: Genesis und Verfall einer innenpolitischen Krisen-
strategie unter Wilhelm II (Dusseldorf: Droste Verlag).
——(1994), Imperial Germany,1871–1914: Economy, Society, Culture and Polit-
ics (Providence and Oxford: Berghahn).
Berkner , L. K. and Mendels , F. (1980), ‘Inheritance systems, family struc-
ture and demographic patterns in Western Europe, 1700–1900’, in C. Tilly
(ed.), Historical Studies of Changing Fertility (Princeton: Princeton University
Press).
Berlin, Statistisches Amt (1887), Die Berliner Volks-Zahlung vom 2. Dezember
1885 (Berlin).
——(1893), Die Bevolkerungs- und Wohnungs-Aufnahme vom 2. Dezember 1890
in der Stadt Berlin (Berlin).
——(1913–16), Die Grundstucks-Aufnahme vom 15. Oktober 1910, sowie die
Wohnungs- und Bevolkerungs-Aufnahme vom 1. Dezember 1910 in der Stadt
Berlin und 44 Nachbar-Gemeinden (Berlin).
Berthold , R. (1978), ‘Die Veranderungen im Bodeneigentum und in der
Zahl der Bauernstellen, der Kleinstellen und der Ritterguter in den
preußischen Provinzen Sachsen, Brandenburg und Pommern wahrend der
Durchfuhrung der Agrarreformen des 19. Jahrhundert’, Jahrbuch fur
Wirtschaftsgeschichte, Sonderband, 10–116.
References and Sources 363
![Page 373: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/373.jpg)
Berthold , R. et al., eds. (1990), Geschichte der Produktivkrafte in Deutschland
von 1800 bis 1945. Band 1: Produktivkrafte in Deutschland 1800 bis 1870
(Berlin: Akademie Verlag).
Bertram , J. (1964), Die Wahlen zum Deutschen Reichstag vom Jahre 1912.
Parteien und Verbande in der Innenpolitik des Wilhelminischen Reichs (Dussel-
dorf: Droste).
B i edermann , E. (1918), ‘Die Einkommens- und Vermogensverhaltnisse der
preußischen Bevolkerung im Zeitraum 1895–1914, Zeitschrift des Koniglich
Preußischen Statistischen Bureaus, 60–86.
B irdsell , N. (1988), ‘Economic approaches to population growth’, in H.
Chenery and T. Srinivasan (eds.), Handbook of Development Economics, vol.
I (Amsterdam: North-Holland).
B jorn , C. (1977), ‘The peasantry and agrarian reform in Denmark’, Scandi-
navian Journal of Economic History, 25/1: 117–37.
Blackbourn , D. (1980), Class, Religion and Local Politics in Wilhelmine Ger-
many: The Centre Party in Wurttemberg before 1914 (New Haven: Yale Uni-
versity Press).
——(1997), The Long Nineteenth Century: The Fontana History of Germany
1780–1918 (London: HarperCollins).
——and Eley , G. (1984), The Peculiarities of German History (Oxford).
Board of Trade (1908), ‘Report of an Enquiry by the Board of Trade into
Working Class Rents, Housing and Retail Prices Together with Rates of
Wages in Certain Occupations in the Principal Industrial Towns of the
German Empire’, Parliamentary Papers, Accounts and Papers (47), cd 4032.
B ohme , H. (1966), Deutschlands Weg zur Gro�macht: Studien zum Verhaltnis
von Wirtschaft und Staat wahrend der Reichgrundungszeit 1848–1881 (Berlin
and Cologne: Oxford University Press).
Bolsker-Schlicht , F. (1997), ‘Deutsche Saisonarbeiter in den Niederlan-
den bis zum 19. Jahrhundert’, in H.-H. Nolte (ed.), Deutsche Migrationen
(Munster: LIT Verlag).
Borchardt , K. (1982), ‘Was there a capital shortage in the Wrst half of the
nineteenth century in Germany?’, in K. Borchardt, Perspectives on Modern
German Economic History and Society (Cambridge: Cambridge University
Press).
Borjas , G. J. (1994), ‘The economics of immigration’, Journal of Economic
Literature, 23 (Dec.), 1667–1717.
Bourgignon , F. (1990), ‘Growth and inequality in the dual model of devel-
opment: the role of demand factors’, Review of Economic Studies, 57: 215–28.
——and Morrison , C. (1998), ‘Inequality and development’, Journal of De-
velopment Economics, 57: 233–57.
Bowley , A. L. (1946), Elements of Statistics, 6th edn. (New York: Staples).
Boyer , G. R. (1990), An Economic History of the English Poor Law, 1750–1850
(Cambridge: Cambridge University Press).
364 References and Sources
![Page 374: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/374.jpg)
Bray , F. (1986), The Rice Economies: Technology and Development in Asian
Societies (Oxford: Blackwells).
Bresciani . C. (1907), ‘Uber die Methoden der Einkommenverteilungstatistik’,
Jahrbucher fur Nationalokonomie und Statistik, 33: 577–600.
Brezis , E., Krugman , P. and Ts iddon , D. (1993), ‘Leapfrogging in inter-
national competition: a theory of cycles in national technological leadership’,
American Economic Review, 83/5: 1211–19.
Brinkmann , T. (1935), Theodore Brinkmann’s Economics of the Farm Business
(Berkeley: Social Science Research Council).
Broadberry , S. N. (1997), ‘Anglo-German Productivity DiVerences: A Sectoral
Analysis’, unpubl.
——(1997a), The Productivity Race: British Manufacturing in International Per-
spective, 1850–1990 (Cambridge: Cambridge University Press).
Brodrick , G. (1881), English Land and Landlords (London: Cassell, Petter,
Galpin).
Broes ike , M. (1907), ‘Die Binnenwanderung im Preußischen Staat’, Zeits-
chrift des Koniglich Preußischen Statistischen Bureaus, 1: 1–62.
Brouwer , W. (1957), ‘A. Schultz-Lupitz zum 125. Geburtstag’, Zeitschrift fur
Agrargeschichte und Agrarsoziologie, 5/1: 69–75.
Bruno , M., Ravaillon , M. and Squire , L. (1996), Equity and Growth in
Developing Countries: Old and New Perspectives on the Policy Issues, Policy
Research Working Paper 1563 (Washington: World Bank).
Bry , G. (1960), Wages in Germany 1871–1945 (Princeton: Princeton University
Press).
Buchsteiner , I. (1994), ‘Besitzkontinuitat, Besitzwechsel und Besitzverlust in
der Gutswirtschaft Pommerns 1879–1910’, in H. Reif (ed.), Ostelbische Agrar-
gesellschaft im Kaiserreich und in der Weimarer Republik (Berlin: Akademie
Verlag).
Burgd orfer , F. (1931), ‘Migration across the frontiers of Germany’, in I
Ferenczi and W. F. Willcox, International Migrations, vol. 2 (New York:
NBER).
Calleo , D. (1978), The German Problem Reconsidered: Germany and the World
Order, 1870 to the Present (Cambridge: Cambridge University Press).
Carsten , F. (1989), A History of the Prussian Junkers (Aldershot: Scolar
Press).
Catt , C. (1986), ‘Farmer and factory worker: rural society in Imperial Ger-
many, the example of Maudach’, in R. J. Evans and W. R. Lee (eds.), The
German Peasantry (London: Croom Helm).
Caumanns , U. (1994), Technischer Fortschritt und Sozialer Wandel in deutschen
Ostprovinzen (Bonn: Kulturstiftung der deutschen Vertriebenen).
Chickering , R. (1984), ‘We men who feel most German’: a cultural study of
the Pan-German League (London: George Allen & Unwin).
References and Sources 365
![Page 375: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/375.jpg)
Clapham , J. H. (1921, repr. 1966), The Economic Development of France and
Germany 1815–1914 (Cambridge: Cambridge University Press).
Clarke , G. (1995), ‘More evidence on income distribution and growth’, Jour-
nal of Development Economics, 47: 403–27.
Coale , A. J. and Watkins , S. C., eds. (1986), The Decline of Fertility in
Europe (Princeton: Princeton University Press).
Conrad , J. (1891), ‘Bodenzersplitterung’, in Handworterbuch der Staatwis-
senschaft, vol. 3 (Jena: G. Fischer).
Conze , W. (1969), ‘The eVects of nineteenth century liberal agrarian reform
on social structure in central Europe’, in F. Crouzet (ed.), Essays in European
Economic History (London: E. Arnold).
Craigie , P. (1887), ‘The size and distribution of agricultural holdings in
England and abroad’, Journal of the Royal Statistical Society, 100: 86–142.
Crew , D. (1979), Town in the Ruhr: A Social History of Bochum 1860–1914
(New York: Columbia University Press).
Crothers , G. D. (1941), The German Elections of 1907 (New York: AMS).
Dade , H. (1913), Die deutsche Landwirtsaft unter Kaiser Wilhelm II (Halle:
C. Marhold).
Dasgupta , P. (1993), An Inquiry into Well-Being and Destitution (Oxford:
Oxford University Press).
Deininger , K. and Squire , L. (1996), ‘A new data set measuring income
inequality’, World Bank Economic Review, 10/3: 565–91.
——and——(1997), ‘Economic growth and income inequality: re-examining
the links’, Finance and Development (Mar.), 38–41.
——and ——(1998), ‘New ways of looking at old issues: inequality and
growth’, Journal of Development Economics, 57: 259–87.
Deist , W. (1976), Flottenpolitik und Flottenpropaganda: das Nachrichtenbureau
des Reichsmarineamtes 1897–1914 (Stuttgart: Deutsche Verlagsanstalt).
Del Fabbro , R. (1996), Transalpini: Italienische Arbeitwanderung nach Sud-
deutschland im Kaiserreich 1870–1918, Studien zur historischen Migrations-
forschung 2 (Osnabruck: Rausch).
Desai , A. V. (1968), Real Wages in Germany, 1871–1914 (Oxford: Oxford
University Press).
D ickler , R. A. (1975), ‘Organisation and change in productivity in Eastern
Prussia’, in W. N. Parker and E. L. Jones (eds.), European Peasants and their
Markets (Princeton: Princeton University Press).
D illwitz , S. (1973), ‘Die Struktur der Bauernwirtschaft von 1871 bis 1914’,
Jahrbuch fur Geschichte, 9: 47–127.
Dodd , D. (1993), Historical Statistics of the United States: Two Centuries of the
Census, 1790–1990 (Westport: Greenwood).
Dollar , D. and Kraay , A. (2000), ‘Growth is Good for the Poor’, manu-
script, www.worldbank.org/research
366 References and Sources
![Page 376: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/376.jpg)
Dovring , F. (1956), Land and Labour in Europe, 1900–1950 (The Hague:
Martinus NijhoV).
Dumke , R. (1991), ‘Income inequality and industrialisation: the Kuznets curve
re-examined’, in Y. Brenner, H. Kaelble, and M. Thomas (eds.), Income
Distribution in Historical Perspective (Cambridge: Cambridge University
Press).
Dyer , B. (1896), ‘Green manuring’, Journal of the Royal Agricultural Society of
England and Wales, 773–9.
Edwardes , J. and Ogilvie , S. (1995), Universal Banks and German Industri-
alisation: A Reappraisal, CEPR Discussion Paper 1171.
Eggert , U. (1883), ‘Die Bewegung der Holzpreise und Tagelohn-Satze in
den preußischen Staatsforsten von 1800 bis 1879’, Zeitschrift des Koniglich
Preußischen Statistischen Bureaus, 23/1: 1–44.
Ehmer , J. (1991), Heiratsverhalten, Sozialstruktur, Okonomisches Wandel, Kri-
tische Studien zur Geschichtswissenschaft 92 (Gottingen: Vandenhoeck &
Rupprecht).
Eley , G. (1980), Reshaping the German Right (New Haven: Yale University
Press).
——(1986), From UniWcation to Nazism: Re-Interpreting the German Past
(Boston and London: Allen & Unwin).
—— , ‘German history and the contradictions of modernity: the bourgeoisie,
the state and the mastery of reform’, in G. Eley (ed.), Society, Culture, and the
State in Germany, 1870–1930 (Ann Arbor: University of Michigan Press),
43–103.
——(2002), Forging Democracy: The History of the Left in Europe 1850–2000
(Oxford: Oxford University Press).
Engel , E. (1868), ‘Die Ergebnisse der Classensteuer, der classiWcierten Ein-
kommensteuer und der Mahl- und Sachlachtsteuer im preußischen Staate’,
Zeitschrift des Koniglich Preußischen Statistischen Bureaus, 1: 25–84.
——(1875), ‘Die Ergebnisse der Classensteuer und der classiWcierten Einkom-
mensteuer im preußischen Staate’, Zeitschrift des Koniglich Preußischen Statis-
tischen Bureaus, i. 108–48.
Engerman , S. L. and Sokoloff , K. L. (1997), ‘Factor endowments, insti-
tutions, and diVerential paths of growth among new world economies: a view
from economic historians of the United States’, in S. Haber (ed.), How Latin
America Fell Behind: Essays on the Economic Histories of Brazil and Mexico,
1800–1914 (Stanford: Stanford University Press).
Epkenhans , M. (1991), Die wilhelminischen Flottenrustung 1908–1914
(Munich: Oldenbourg).
Erickson , C. (1957), American Industry and the European Immigrant (Cam-
bridge, Mass.: Harvard University Press).
Evans , E. L. (1997), The German Centre Party 1870–1933: A Study in Political
Catholicism (Carbondale: Southern Illinois University Press).
References and Sources 367
![Page 377: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/377.jpg)
Evans , R. (1997), Rereading German History: From UniWcation to ReuniWcation
(London: Routledge).
Evershed , H. (1884), ‘Improvements of the plants of the Weld’, Journal of the
Royal Agricultural Society of England and Wales, 77–113.
Fairbairn , B. (1992), ‘Interpreting Wilhelmine elections: national issues, fair-
ness issues and electoral mobilisation’, in E. L. Jones and J. Rettalack (eds.),
Elections, Mass Politics and Social Change in Modern Germany: New Perspec-
tives (Cambridge: Cambridge University Press).
——(1997), Democracy in the Undemocratic State: The German Reichstag Elec-
tions of 1898 and 1903 (Toronto: University of Toronto Press).
Fe i , J. E., Ranis , G. and Kuo , S. Y. (1979), Growth with Equity: The Taiwan
Case (World Bank and Oxford University Press).
Fe instein , C. H. (1972), National Income, Expenditure and Output of the
United Kingdom 1855–1965 (Cambridge: Cambridge University Press).
——(1988), ‘The rise and fall of the Williamson Curve’, Journal of Economic
History, 47/3: 699–729.
——(1998), ‘Pessimism perpetuated: real wages and the standard of living in
Britain during and after the Industrial Revolution’, Journal of Economic His-
tory, 58/3: 625–58.
——and Thomas , M. (2000), ‘A Plea for Errors’, manuscript, All Souls Col-
lege, Oxford.
Feldenkirchen , W. (1982), ‘Zur Kapitalbeschaftigung und Kapitalverwen-
dung bei Aktiengesellschaften des deutschen Maschinenbaus im 19. und
beginnenden 20. Jahrhundert’, Vierteljahresschrift fur Wirtschaftsgeschicht, 49:
38–74.
Ferenczi , I. and Willcox , W. F. (1929–31), International Migrations, 2 vols.
(New York: NBER).
Ferguson , N. (1994), ‘Public Wnance and national security: the domestic
origins of the First World War revisited’, Past and Present, 142: 141–68.
——(1998), The Pity of War (London: Allen Lane).
F inck von Finckenstein , H. W. Graf (1960), Die Entwicklung der Land-
wirtschaft in Preußen und Deutschland 1800–1930 (Wurzburg: Holzner).
F ircks , A. von (1879), ‘Ruckblick auf die Bewegung der Bevolkerung im
Preußischen Staat 1816–74’, Preu�ische Statistik, 48/1.
Flora , P., Kraus , F. and Pfennig , W. (1983), State, Economy and Society in
Western Europe 1815–1973: A Data Handbook, 2 vols. (Frankfurt am Main:
Campus).
Fohlin , C. (2002), ‘Corporate capital structure and the inXuence of universal
banks in pre-World War I Germany’, Jahrbuch fur Wirtschaftsgeschichte,
113–34.
Frankenstein , K. (1893), Die Arbeiterfrage in der deutschen Landwirtschaft
(Berlin: R. Oppenheim).
368 References and Sources
![Page 378: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/378.jpg)
Frauendorfer , S. (1956), ‘Naturwissenschaften und Technik als dyna-
misches Moment’, Zeitschrift fur Agrargeschichte und Agrarsoziologie, 4/2:
113–27.
——(1957), Ideengeschichte der Agrarwirtschaft und Agrarpolitik im deutschen
Sprachgebiet, Band 1 (Munich: Bayerischer Landwirtschaftsverlag).
Fremdling , R. (1991), ‘Productivity comparisons between Great Britain and
Germany, 1855–1913’, Scandinavian Journal of Economic History, 39/1:
28–42.
Fussell , G. E. (1962), ‘The early days of chemical fertilisers’, Nature, 195:
750–4.
Galor , O. and Ts iddon , D. (1994), Human Capital Distribution, Techno-
logical Progress and Economic Growth, CEPR Discussion Paper no. 971
(London: CEPR).
Galos , A., Gentzen , F. and Jak okczyck , W. (1966), Die Hakatisten: Der
Deutsche Ostmarkverein 1894–1934 (Berlin: VEB).
Ge isenberger , S. and MUller , J. H. (1972), Die Einkommensstruktur in
verschiedenen deutschen Landern 1874–1914, Schriften zu Regional- und Ver-
kehrsproblemen 10 (Berlin).
Ge iss , I. (1990), ‘The German version of imperialism, 1898–1914: Weltpoli-
tik’, in G. Schollgen (ed.), Escape into War? The Foreign Policy of Imperial
Germany (Oxford: Berg).
Gerschenkron , A. (1943), Bread and Democracy in Germany (Berkeley: Uni-
versity of California Press).
——(1962), Economic Backwardness in Historical Perspective (Cambridge, Mass:
Harvard University Press).
Ghimure , K., ed. (2001), Land Reform and Peasant Livelihoods: The Social
Dynamics of Rural Poverty and Agrarian Reform in Developing Countries, UN
Institute for Social Development (New York: UN).
Goldin , C. (1994), ‘The political economy of immigration restriction in the
United States, 1890–1921’, in C. Goldin and G. Libecap (eds.), The Regu-
lated Economy: A Historical Approach (Chicago: University of Chicago Press).
Golding , F. (1930), ‘Die Wanderbewegung in Ostpreu�en seit der Jahrhun-
dertwende mit besonderer Berucksichtigung der Abwanderung vom Lande’,
Zeitschrift des Preußischen Statistischen Landesamts, 69/2: 200–34.
Goltz , T. Frhr . von der (1875), Die Lage der landlichen Arbeiter im
deutschen Reich (Berlin: P. Parey).
——(1893), Die landliche Arbeiterklasse und der preu�ische Staat (Jena: G.
Fischer).
Gommel , R. (1979), Realeinkommen in Deutschland: ein internationaler Vergleich
(Nurnberg: Im Selbstverlag).
Grant , O. (2000), ‘Internal Migration in Germany, 1870–1913’, Oxford
D.Phil. thesis.
References and Sources 369
![Page 379: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/379.jpg)
Grant , D. (2002a), Productivity in German Agriculture: Estimates of Agricul-
tural Productivity from Regional Accounts for 21 German Regions: 1880/4,
1893/7 and 1905/9, Oxford University Discussion Paper in Economic and
Social History, available at http://www.nuV.ox.ac.uk/ Economics/History/
——(2002b), Does Industrialisation Push Up Inequality? New Evidence on the
Kuznets Curve from Nineteenth Century Prussian Tax Statistics, Oxford
University Discussion Paper in Economic and Social History, available at
http://www.nuV.ox.ac.uk/Economics/History/
——(2002c), ‘Max Weber and Die Lage der Landarbeiter im ostelbischen Deutsch-
land: a statistical examination’, forthcoming in the Jahrbuch fur Wirtschafts-
geschichte, 2: 61–84.
——(2002d ), ‘Explaining the Decline of German Emigration to the United
States after 1895’, manuscript, NuYeld College, Oxford.
——(2002e), ‘ ‘‘Few Better Farmers in Europe?’’ Productivity, Change and
Modernisation in East-Elbian Agriculture 1870–1913’, manuscript, NuYeld
College, Oxford.
Greene , W. H. (2000), Econometric Analysis, 4th edn. (New Jersey: Prentice-
Hall).
Gregory , P. R. (1997), ‘Searching for consistency in historical data: alterna-
tive estimates of Russia’s industrial production, 1887–1913’, Journal of Eco-
nomic History, 57/1: 196–202.
Grimmer-Solem , E. (1999), ‘The Science of Progress: The Rise of Historical
Economics and Social Reform in Germany, 1864–1894’, Oxford D.Phil.
thesis.
Groh , D. (1973), Negative Integration und revolutionarer Attentismus: Die
deutsche Sozialdemokratie am Vorabend des Ersten Weltkrieges (Frankfurt am
Main: Propylaen).
Grumbach , F. and Konig , H. (1957), ‘Beschaftigung und Lohne der
deutschen Industriewirtschaft 1888–1954’, Weltwirtschaftsarchiv, 79: 125–54.
Guttsman , W. (1981), The German Social Democratic Party 1875–1933
(London: Allen & Unwin).
Habakkuk , J. (1958), ‘Population growth and economic development’, in
J. Habakkuk et al., Lectures on Economic Development (Istanbul: Sermet Mat-
baash).
Haberle , R. (1938), ‘German approaches to internal migration’, in D. Thomas
(ed.), Research Memorandum on Migration DiVerentials (New York: Social
Science Research Council).
Hagen , W. (1980), Germans, Poles and Jews: The Nationality ConXict in the
Prussian East, 1772–1914 (Chicago: University of Chicago Press).
Haines , M. (1977), Economic–Demographic Interrelations in Developing Agricul-
tural Regions: A Case Study of Prussian Upper Silesia, 1840–1914 (New York:
Arno).
370 References and Sources
![Page 380: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/380.jpg)
——(1979), Fertility and Occupation: Population Patterns in Industrialisation
(New York: Academic).
Hajnal , J. (1965), ‘European marriage patterns in perspective’, in D. Glass
and D. Eversley (eds.), Population in History (Chicago: Aldine).
Hanley , S. B. and Yamamura , K. (1977), Economic and Demographic Change
in Pre-Industrial Japan, 1600–1868 (Princeton: Princeton University Press).
Hardach , K. W. (1967), Die Bedeutung wirtschaftlicher Faktoren bei der Wie-
dereinfuhrung der Eisen- und Getreidezolle in Deutschland 1879 (Berlin:
Duncker & Humblot).
Harnisch , H. (1977), ‘Die Bedeutung der kapitalistischen Agrarreform fur
die Herausbildung des Inneren Markte und die industrielle Revolution in der
ostlichen Provinzen Preußens in der ersten Halfte des 19. Jahrhunderts’,
Jahrbuch fur Wirtschaftgeschichte, 63–82.
——(1984), Kapitalistische Agrarreform und Industrielle Revolution (Weimar: H.
Bohlhaus).
——(1986), ‘The background to the agricultural reforms in feudal Prussia east
of the Elbe 1760–1807’, in R. J. Evans and W. R. Lee (eds.), The German
Peasantry (London: Croom Helm).
Hatton , T. J. (1993), A Model of UK Emigration 1870–1913, CEPR Discus-
sion Paper no. 771.
——(1997), ‘The immigrant assimilation puzzle in late nineteenth century
America’, Journal of Economic History 57: 34–62.
——and Will iamson , J. G., eds. (1994), Migration and the International
Labour Market, 1850–1939 (London: Routledge).
——and ——(1996), ‘What drove the mass migrations from Europe in the
nineteenth century?’ Population and Development Review, 20: 533–59.
——and ——(1998), The Age of Mass Migration: Causes and Economic Impact
(Oxford: Oxford University Press).
Hatzius , J. (1994), Regional Migration, Unemployment and Vacancies: Evidence
from West German Microdata, Oxford, Institute of Economics and Statistics,
Applied Economics Discussion Paper no. 164.
Hauser , A. (1965), ‘Schweizer Bauern als Kolonisten in Preußen und
Litauen’, Zeitschrift fur Agrargeschichte und Agrarsoziologie, 13: 212–19.
Haushofer , H. (1957), Die deutsche Landwirtschaft im technische Zeitalter,
vol. 5 of G. Franz (ed.), Deutsche Agrargeschichte (Stuttgart: Ulmer).
Hayami , Y. (1997) Development Economics: From the Poverty to the Wealth of
Nations (Oxford: Oxford University Press).
Helferrich , K. (1915), Deutschlands Volkswohlstand (Berlin: G. Stilbe).
Helling , G. (1966), ‘Zur Entwicklung der Produktivitat in der deutschen
Landwirtschaft im 19. Jahrhundert’, Jahrbuch fur Wirtschaftsgeschichte,
129–41.
Hempel , L. (1957), ‘Heuerlingswesen und Croftersystem’, Zeitschrift fur
Agrargeschichte und Agrarsoziologie, 5/2: 169–79.
References and Sources 371
![Page 381: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/381.jpg)
Henning , F. W. (1977), ‘Produktionskosten und Preisbildung fur Getreide in
den letzten Jahrzehnten des 19. Jahrhunderts’, Zeitschrift fur Agrargeschichte
und Agrarsoziologie, 25: 214–36.
Hennock , E. P. (2000), ‘The urban sanitary movement in England and Ger-
many, 1838–1914: a comparison’, Continuity and Change, 15/2: 269–96.
Hentschel , V. (1978), Wirtschaft und Wirtschaftspolitik im wilhelminischen
Deutschland (Stuttgart: Klett-Cotta).
——(1979), ‘Erwerbs- und Einkommensverhaltnisse in Sachsen, Baden und
Wurttemberg vor dem Ersten Weltkrieg’, Vierteljahrschrift fur Sozial- und
Wirtschaftsgeschichte, 66/1: 26–73.
Herbert , U. (1990), A History of Foreign Labour in Germany 1880–1980 (Ann
Arbor: University of Michigan Press).
Herwig , H. (1987), ‘Luxury Fleet’: The Imperial German Navy 1888–1918
(London: AshWeld).
Hess , K. (1994), ‘Zur wirtschaftlichen Lage der Gro�agrarier im ostelbischen
Preu�en 1867/71 bis 1914’, in H. Reif (ed.), Ostelbische Agrargesellschaft im
Kaiserreich und in der Weimarer Republik (Berlin: Akademie Verlag).
H iggins , M. and Will iamson , J. G. (1999), ‘Explaining Inequality the
World Round: Cohort Size, Kuznets Curves and Openness’, manuscript,
Federal Reserve Bank of New York and Harvard University.
H iggs , R. (1971), ‘Race, skill and earnings: American immigrants in 1909’,
Journal of Economic History, 31: 420–9.
H ildebrand , K. (1989), Deutsche Au�enpolitik, 1871–1918 (Munich: Olden-
bourg).
——(1995), Reich–Nation-State–Great Power: ReXections on German Foreign
Policy 1871–1945 (London: German Historical Institute).
Hobson , J. M. (1997) The Wealth of States: A Comparative Sociology of Inter-
national Economic and Political Change (Cambridge: Cambridge University
Press).
Hochstadt , S. (1996), ‘The socio-economic determinants of increasing mo-
bility in nineteenth century Germany’, in D. Hoerder and L. P. Moch (eds.),
European Migrants (Boston: Northeastern Press).
——(1999), Mobility and Modernity: Migration in Germany 1820–1989 (Ann
Arbor: University of Michigan Press).
Hoerder , D. and Nagler , J., eds. (1995), People in Transit (Cambridge:
Cambridge University Press).
Hoffmann , W. (1963), ‘The take-oV in Germany’, in W. Rostow (ed.), The
Economics of Take-OV into Sustained Growth (London: Macmillan).
——(1965), Das Wachstum der Deutschen Wirtschaft seit der Mittel des 19. Jahr-
hunderts (Berlin: Springer Verlag).
Hohls , R. (1995), ‘Germany’, in P. Scholliers and V. Zamagni (eds.), Labour’s
Reward (Aldershot: E. Elgar).
372 References and Sources
![Page 382: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/382.jpg)
Hohorst , G. (1977), Wirtschaftswachstum und Bevolkerungsentwicklung in
Preußen 1816 bis 1914 (Munster: University of Munster).
Honigsheim , P. (1949), ‘Max Weber as historian of agriculture and rural life’,
Agricultural History, 23/3: 179–212.
Hourwich , I. (1922), Immigration and Labour: The Economic Aspects of Euro-
pean Immigration to the United States (New York: Huebsch).
Hunt , J. C. (1974), ‘Peasants, grain tariVs and meat quotas’, Central European
History, 7: 311–31.
Hutchinson , E. P. (1956), Immigrants and their Children 1850–1950 (New
York: Wiley).
Institut International d’Agriculture (1915), Annuaire International de Statistique
Agricole 1913/14.
Jackman , R. and Savouri , S. (1992), ‘Regional migration in Britain: an
analysis of gross Xows using NHS central register data’, Economic Journal,
102: 1433–50.
Jackson , J. J. (1997), Migration and Urbanisation in the Ruhr Valley 1821–1914
(New Jersey: Humanities Press).
Jarausch , K. (1969), ‘The illusion of limited war: Chancellor Bethmann Holl-
weg’s calculated risk’, Central European History, 48–76.
Jazairy , I., Alamgir , M. and Panuccio , T. (1992), The State of World
Rural Poverty: An Enquiry into its Causes and Consequences, International
Fund for Agricultural Development.
Jeck , A. (1970), Wachstum und Verteilung des Volkseinkommens (Tubingen:
Mohr).
Jeidels , O. (1905), Das Verhaltnis der deutschen Gro�banken zur Industrie,
Staats- und Sozialwissenschaftliche Forschungen, 24.
Jenks , J. W. and Lauck , W. J. (1926), The Immigration Problem: A Study of
American Immigration Conditions and Needs (New York: Huebsch).
Kaelble , H. (1976), ‘Social stratiWcation in Germany in the 19th and 20th
centuries: a survey of research since 1945’, Journal of Social History, 10:
144–65.
——(1986), Industrialisation and Social Inequality in Nineteenth Century Europe
(Leamington Spa: Berg).
——and Federspiel , R. (1990), Soziale Mobilitat Berlin 1825–1957, Histor-
ische Statistik von Deutschland, vol. X.
——and Hohls , R. (1989), Die Regionale Erwerbsstruktur im Deutschen Reich
und in der Bundesrepublik 1895–1970, Quellen und Forschungen zur histor-
ischen Statistik von Deutschland.
Kaerger , K. (1890), ‘Die Sachsengangerei’, Landwirtschaftliche Jahrbucher, 19:
237–522.
Kaiserlich Statistisches Amt (1916), Die Deutschen im Ausland und die Auslander
im Deutschen Reich, Erganzungsheft zur Vierteljahrsheft zur Statistik des
Deutschen Reichs 4.
References and Sources 373
![Page 383: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/383.jpg)
Kaldor , N. (1956), ‘Alternative theories of distribution’, Review of Economic
Studies, 23/2: 94–100.
Kamphoefner , W. D. (1986), ‘At the crossroads of economic development:
background factors aVecting emigration from nineteenth-century Germany’,
in I. A. Glazier and L. De Rosa (eds.), Migration across Time and Nations
(New York: Holmes and Meier).
——(1987), The Westfalians: From Germany to Missouri (Princeton: Princeton
University Press).
——(1999), ‘Deutschen-Amerikaner: Musterknaben der Einwanderung?’,
IMIS-Beitrage, 12.
Kehr , E. (1975), Battleship Building and Party Politics in Germany (Chicago:
University of Chicago Press).
——(1977), Economic Interest, Militarism and Foreign Policy (Berkeley: Univer-
sity of California Press).
Kennedy , P. (1980), The Rise of the Anglo-German Antagonism, 1860–1914
(London: Allen & Unwin).
Keyder , C. and O’Brien , P. (1978), Economic growth in Britain and France,
1780–1914: Two Paths to the Twentieth Century (London: Allen & Unwin).
Keyser , E. (1959), ‘Die Ansiedlung von Schwaben in Westpreu�en
1776–1805’, Zeitschrift fur Agrargeschichte und Agrarsoziologie, 7/2: 169–76.
K i esewetter , H. (1988), Industrialisierung und Landwirtschaft: Sachsens Stel-
lung im regionalen Industrialisierungsprozess Deutschlands im 19. Jahrhundert
(Cologne: Bohlau).
K indleberger , C. (1967), Europe’s Postwar Growth: The Role of Labour
Supply (Cambridge, Mass.: Harvard University Press).
Klemm , V. (1994), ‘Die Agrarwissenschaften und die Modernisierung der
Gutsbriebe in Ost-und Mitteldeutschland’, in H. Reif (ed.), Ostelbische Agrar-
gesellschaft im Kaiserreich und in der Weimarer Republik (Berlin: Akademie
Verlag).
Knapp , G. F. (1887), Die Bauernbefreiung und der Ursprung der Landarbeiter in
den alteren Theilen Preussens (Leipzig: Duncker & Humblot).
——(1891), Die Landarbeiter bei der Stein-Hardenbergischen Gesetzgebung
(Strassburg: Heitz and Mundel).
Knodel , J. (1974), The Decline of Fertility in Germany 1871–1939 (Princeton:
Princeton University Press).
——(1988), Demographic Behaviour in the Past (Cambridge: Cambridge Univer-
sity Press).
Knoke , A. (1911), Auslandische Wanderarbeiter in Deutschland, Diss., Leipzig.
Kocka , J., ed. (1975), ‘Soziale Schichtung und Mobilitat in Deutschland im
19. und 20. Jahrhundert’, Geschichte und Gesellschaft 1/1 (Gottingen: Van-
denhoeck und Ruprecht).
——(1990), Arbeitsverhaltnisse und Arbeiterexistenzen: Grundlagen der Klassenbil-
dung im 19. Jahrhundert (Bonn: Dietz).
374 References and Sources
![Page 384: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/384.jpg)
Kolakowski , L. (1978), Main Currents of Marxism, 3 vols. (Oxford: Oxford
University Press).
Kollmann , W. (1959), ‘Industrialisierung, Binnenwanderung und ‘‘Soziale
Frage’’ ’, Vierteljahrschrift fur Sozial- und Wirtschaftsgeschichte.
——(1960), Sozialgeschichte der Stadt Barmen im 19. Jahrhundert (Tubingen:
J. C. B. Mohr).
——(1964), ‘The population of Germany in the age of industrialisation’, in
H. Moller (ed.), Population Movements in Modern European History (London:
Collier-Macmillan).
——(1974), ‘Demographische ‘‘Konsequenzen’’ der Industrialisierung in
Preu�en’, in W. Kollmann (ed.), Bevolkerung in der Industriellen Revolution
(Gottingen: Vandenhoeck und Ruprecht).
——(1976), ‘Bevolkerungsgeschichte 1800–1870’, in H. Aubin and W. Zorn
(eds.), Handbuch der deutschen Wirtschaft- und Sozialgeschichte, vol. 2 (Stutt-
gart: Klett-Cotta).
——(1980), Quellen zur Bevolkerungsgeschichte Deutschlands, vol. 1 of Quellen
zur Bevolkerungs-, Sozial- und Wirtschaftsstatistik Deutschlands (Boppard:
Boldt).
——and Marschalck , P. (1973), ‘German emigration’, Perspectives in Ameri-
can History, 7: 499–554.
Komlos , J. (1995), The Biological Standard of Living in Europe and America
1700–1900: Studies in Anthropometric History (Aldershot: Variorum).
Konitz , C. von (1895), ‘Uber das Verhaltnisse zwischen Rohertrag und Rein-
ertrag in landwirtschaftlichen Betrieben verschiedener Wirtschaftssysteme’,
Landwirtschaftliche Jahrbucher 24: 309–444.
Koselleck , R. (1967), Preu�en zwischen Reform und Revolution (Stuttgart:
Klett-Cotta).
Kosters-Kraft , M. (1998), Wanderarbeit und Gro�baustelle: Niederlander
beim Bau des Dortmund-Ems-Kanals 1892–1900, Studien zur historischen
Migrationsforschung 8 (Osnabruck: Rausch).
Kuczynski , J. (1945), A Short History of Labour Conditions in Germany, 1800
to the present day (London: F. Muller).
Kuczynski , R. (1897), Der Zug nach der Stadt (Stuttgart: J. G. Cotta).
——(1908), Einkommen und Miete (Berlin: Berlin Statistische Amt).
——(1913), Arbeitslohn und Arbeitszeit in Europa und Amerika (Berlin:
J. Springer).
Kuhnert , F. (1911), ‘Die Ergebnisse der preu�ischen Einkommens- und
Erganzungssteuer: Veranlagung nach Stadtegruppen fur das Jahr 1908’, Zeits-
chrift des Koniglich Preu�ischen Statistischen Bureaus, 1–54.
——(1916), ‘Einkommensgliederung der preu�ischen Bevolkerung im Zei-
traum 1902–1911’, Zeitschrift des Koniglich Preu�ischen Statistischen Bureaus,
269–309.
References and Sources 375
![Page 385: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/385.jpg)
Kuznets , S. (1955), ‘Economic growth and income inequality’, American Eco-
nomic Review, 45: 1–28.
Laffont , J.-J. and Matouss i , M. (1995), ‘Moral hazard, Wnancial con-
straints and share-cropping in El Oulja’, Review of Economic Studies, 62/3:
381–99.
Lambi , I. N. (1984), The Navy and German Power Politics 1862–1914 (Boston:
Allen & Unwin).
——(1963), ‘Free trade and protection in Germany 1868–1879’, Vierteljahrs-
chrift fur Sozial- und Wirtschaftsgeschichte, Part 44.
Laux , H.-D. (1989), ‘The components of population growth in Prussian cities,
1875–1905, and their inXuence on urban population structure’, in R. Lawton
and W. R. Lee (eds.), Urban Population Development in Western Europe from
the Late Eighteenth to the Early Twentieth Century (Liverpool: Liverpool Uni-
versity Press).
Lebovics , H. (1967), ‘ ‘‘Agrarians’’ versus ‘‘industrializers’’: Social Conserva-
tive resistance to industrialism and capitalism in late nineteenth century Ger-
many’, International Review of Social History, 12: 31–65.
Lee , J. Z. and Wang , F. (1999), One Quarter of Humanity: Malthusian Myth-
ology and Chinese Realities, 1700–2000 (Cambridge, Mass.: Harvard Univer-
sity Press).
Lee , W. and Marschalck , P. (2000), ‘Demographic change and industrial-
isation in Germany, 1815–1914: Bremen in comparative perspective’, History
of the Family, 5/4: 373–90.
Lee , W. R. (1977), Population Growth, Economic Development and Social Change
in Bavaria 1750–1850 (New York: Arno).
——ed. (1979), European Demography and Economic Growth (London: Croom
Helm).
——(1979a), ‘Regionale DiVerenzierung im Bevolkerungswachstum Deutsch-
lands im fruhen 19. Jahrhundert’, in R. Fremdling and R. H. Tilly (eds.),
Industrialisierung und Raum (Stuttgart: E. Klett).
——(1980), ‘The mechanism of mortality change in Germany 1750–1850’,
Medizinhistorisches Journal, 15.
——(1981), ‘Family and ‘‘modernisation’’: the peasant family and social change
in nineteenth century bavaria’, in R. Evans and W. R. Lee (eds.), The German
Family: Essays on the Social History of the Family in Nineteenth and Twentieth
Century Germany (London: Croom Helm).
Lees , A. (1979), ‘Critics of urban society in Germany, 1854–1914’, Journal of
the History of Ideas, 40: 61–83.
Leps ius , R. (1966), ‘Parteinsystem und Sozialstruktur. Zum Problem der
Demokratisierung der deutschen Gesellschaft’, in W. Abel (ed.), Wirtschaft,
Geschichte und Wirtschaftsgeschichte: Festschrift zum 65. Geburtstag von Fried-
rich Lutge (Stuttgart: G. Fischer).
376 References and Sources
![Page 386: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/386.jpg)
Leunig , T. and Voth , H.-J. (1996), ‘Did smallpox reduce height? Stature
and the standard of living in London, 1770–1873’, Economic History Review,
49/3: 541–60.
Levy , H. (1911), Large and Small Holdings: A Study in English Agricultural
Economics (Cambridge: Cambridge University Press).
Lewis , W. A. (1954), ‘Development with unlimited supplies of labour’, Man-
chester School of Economics and Social Studies, 20: 139–92.
L idtke , V. (1985), The Alternative Culture: Socialist Labour in Imperial Ger-
many (New York: Oxford University Press).
L indenlaub , J. and Kohne-Lindenlaub , R. (1991), ‘Business Wnancing by
Krupp 1811–1848’, German Yearbook on Business History, 9–86.
L indert , P. (1994) ‘Unequal living standards’, in R. Floud and D. McCloskey
(eds.), The Economic History of Britain: Volume One: 1700–1860 (Cambridge:
Cambridge University Press).
——and Will iamson , J. G. (1980) American Inequality (New York: Aca-
demic Press).
L ipton , M. (1980), ‘Migration from rural areas of poor countries: their impact
on rural productivity and income distribution’, World Development, 8: 1–24.
Lourens , P. and Lucassen , J. (1999), Arbeitswanderung und beruXiche Spe-
zialisierung: die lippischen Ziegler im 18. und 19. Jahrhunderts, Studien zur
historischen Migrationsforschung 6 (Osnabruck: Rausch).
Low , D. (1996), The Egalitarian Moment: Asia and Africa 1950–1980 (Cam-
bridge: Cambridge University Press).
Lubinski , A. (1995), Entlassen aus dem Untertanenverband: Die Amerikaaus-
wanderung aus Mecklenburg-Strelitz (Osnabruck: Rausch).
LU tge , F. (1963), Geschichte der deutschen Agrarfassung von fruhen Mittelalter
bis zum 19. Jahrhundert (Stuttgart: E. Ulmer).
McGouldrick , P. F. and Tannen , M. B. (1977), ‘Did American manufac-
turers discriminate against immigrants before 1914?’, Journal of Economic
History, 37: 723–46.
Maddison , A. (1991), Dynamic Forces in Capitalist Development (Oxford:
Oxford University Press).
——(1994), ‘Explaining the economic performance of nations, 1820–1989’, in
W. Baumol et al. (eds.), Convergence of Productivity (Oxford: Oxford Univer-
sity Press).
——(1995), Monitoring the World Economy, 1820–1992 (Paris: OECD).
Mantl , E. (1997), Heirat als Privileg: obrigkeitliche Heiratsbeschrankungen in
Tirol und Vorarlberg 1820 bis 1920 (Vienna: Verlag fur Geschichte und Poli-
tik).
Marschalck , P. (1973), Deutsche Uberseewanderung im 19. Jahrhundert (Stutt-
gart: Klett).
References and Sources 377
![Page 387: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/387.jpg)
Marschalck , P. (1987), ‘The age of demographic transition: mortality and
fertility’, in K. Bade (ed.), Population, Labour and Migration in 19th and 20th
Century Germany (Leamington Spa: Berg).
Massey , D. S. (1988), ‘Economic development and international migration in
comparative perspective’, Population and Development Review, 14: 383–413.
Matthews , R., Fe instein , C. and Odling-Smee , J. (1982), British Eco-
nomic Growth 1856–1973 (Oxford: Oxford University Press).
Matz , K.-J. (1980), Pauperismus und Bevolkerung: Die gesetzlichen Ehebeschran-
kerungen in der suddeutschen Staaten wahrend des 19. Jahrhunderts (Stuttgart:
Klett-Cotta).
Matzerath , H. (1985), Urbanisierung in Preu�en 1815–1914 (Stuttgart: W.
Kohlhammer).
Me isel , F. (1911), ‘Moral und Technik bei der Veranlagung der preußischen
Einkommensteuer’, Jahrbuch fur Gesetzgebung, Verwaltung und Volkswirtschaft
im Deutschen Reich, 35/1: 285–374.
Me itzen , A. (1868), Der Boden und der landwirtschaftlichen Verhaltnisse des
Preußischen Staates (Berlin: Finanzministerium).
——(1885), ‘Ermittelungen uber die durchschnittliche Hohe der Grund-
buchschulden der bauerlichen Besitzungen in 52 Amtsgerichtsbezirken der
Preußischen Staat nach dem Stand des Jahres 1883’, Landwirtschaftliche Jahr-
bucher, 14/2: 1–32.
Meyer , F. (1938), ‘German internal migration statistics: methods, sources and
data’, in D. Thomas (ed.), Research Memorandum on Migration DiVerentials
(New York: Social Science Research Council).
M iaskowski , A. von (1882), ‘Das Erbrecht und die Grundeigentumvertei-
lung im deutschen Reich’, Schriften des Vereins fur Sozialpolitik, 20.
M ilward , A. and Saul , S. B. (1977), The Development of the Economies of
Continental Europe, 1850–1914 (London: Allen & Unwin).
Ministry of Agricul ture , Fisher ies and Food (1968 ) , A Century of
Agricultura l Stat is t ics : Great Bri ta in 1866–1966 (London:
HMSO).
M itchell , B. R. (1981), European Historical Statistics 1750–1975, 2nd edn.
(London: Macmillan).
M ittler , E. S. (1929), Untersuchungen uber Landarbeiterverhaltnisse, Berichte
uber Landwirtschaft (Berlin: P. Parey).
Mokyr , J. (1990), The Lever of Riches: Technological Creativity and Economic
Progress (Oxford: Oxford University Press).
Mommsen , W. (1995), Imperial Germany: Politics, Culture and Society in an
Authoritarian State (London: Arnold).
Nadobnik , M. (1908), ‘Die Abnahme des durchschnittlichen Heiratsalters in
Deutschland’, Zeitschrift des Koniglichen Preussischen Landesamts, 48: 69–84.
Neuburger , H. and Stokes , H. H. (1974), ‘German banks and German
growth, 1880–1913’, Journal of Economic History, 34: 710–31.
378 References and Sources
![Page 388: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/388.jpg)
Neuhaus , G. (1904), ‘Die ortsublichen Tagelohne gewohnlicher Tagearbeiter
in Preussen 1892 und 1901’, Zeitschrift des Koniglich Preu�ischen Statistischen
Bureaus, 310–46.
Neumann , R. P. (1978), ‘Working class birth control in Wilhelmine Ger-
many’, Comparative Studies in Society and History, 20: 408–28.
Newbery , D. (1977), ‘Risk sharing, share-cropping and uncertain labour
markets’, Review of Economic Studies, 23: 187–207.
——(1989), ‘Agricultural institutions for insurance and stabilization’, in
P. Bardhan (ed.), The Economic Theory of Agrarian Institutions (Oxford:
Oxford University Press).
N ichtweiss , J. (1959), Die auslandischen Saisonarbeiter in der Landwirtschaft
der ostlichen und mittleren Gebiete des Deutschen Reichs: Ein Beitrag zur
Geschichte des preußisch-deutsch Politik von 1890 bis 1914 (Berlin: Rutten und
Loening).
N ickell , S. (1996), ‘The low-skill low-pay problem: lessons from Germany
for Britain and the US’, Policy Studies, 17: 17–21.
Nove , A. (1969), An Economic History of the U.S.S.R. (London: Allen Lane).
Nugent , W. (1992), Crossings: The Great Transatlantic Migrations, 1870–1914
(Bloomington: Indiana University Press).
Nurkse , R. (1953), Problems of Capital Formation in Underdeveloped Countries
(Oxford: Basil Blackwell).
O’Brien , P. and Prados de la Escosura , L. (1992), ‘Agricultural prod-
uctivity and European industrialisation’, Economic History Review, 45:
514–36.
O’Rourke , K. (1997) ‘The European grain invasions, 1870–1913’, Journal of
Economic History, 57/4: 775–801.
——Taylor , A. M. and Will iamson , J. G. (1996), ‘Factor price conver-
gence in the late nineteenth century’, International Economic Review, 37/3:
499–530.
——Will iamson , J. G. and Hatton , T. (1994), ‘Mass migration, commod-
ity market integration and real wage convergence’, in T. J. Hatton and J. G.
Williamson (eds.), Migration and the International Labour Market, 1850–1939
(London: Routledge).
Oberpennig , H. (1998), Migration und Fernhandel im ‘Todden-System’: Wan-
derhandler aus dem nordlichen Munsterland im mittleren und nordlichen Europa
des 18. und 19. Jahrhunderts (Osnabruck: Rausch).
Offer , A. (1989), The First World War: An Agrarian Interpretation (Oxford:
Oxford University Press).
Olson , M. (1982), The Rise and Decline of Nations: Economic Growth, StagXa-
tion and Social Rigidities (New Haven: Yale University Press).
Oppenheimer , F. (1898), Gro�grundeigentum und soziale Frage (Berlin:
VITA).
References and Sources 379
![Page 389: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/389.jpg)
Orsagh , T. (1969), ‘Lohne in Deutschland 1871–1913: neuere Literatur’,
Zeitschrift fur die gesamte Staatswissenschaft, 125: 476–83.
Paasche , H. (1883), ‘Die rechtliche und wirtschaftliche Lage des Bauern-
standes in Mecklenburg-Schwerin’, in Bauerliche Zustand in Deutschland,
vol. 3, Schriften des Verein fur Sozialpolitik, 24.
Pas inetti , L. (1962), ‘Rate of proWt and income distribution in relation to the
rate of economic growth’, Review of Economic Studies, 29: 267–79.
Perkins , W. A. (1981), ‘The agricultural revolution in Germany’, Journal of
Economic History, 10/1: 71–118.
——(1984), ‘The German agricultural worker, 1815–1914’, Journal of Peasant
Studies, 11/3: 3–27.
——(1986), ‘Dualism in German agrarian historiography’, Comparative Studies
in Society and History, 28/2: 287–306.
Perrson , T. and Tabellini , G. (1994), ‘Is inequality harmful for growth?’,
American Economic Review, 84/3: 600–21.
Peterson , C. (1995), Bread and the British Economy: c.1770–1870 (Aldershot:
Scholar).
Phelps Brown , E. H. and Browne , M. (1968), A Century of Pay (London:
Macmillan).
Plaul , R. (1986), ‘The rural proletariat: the everyday life of rural labourers in
the Magdeburg region, 1830–80’, in R. J. Evans and W. R. Lee (eds.), The
German Peasantry (London: Croom Helm).
Pogge von Strandmann , H. (1972), ‘Nationale Verbande zwischen Welt-
politik und Kontinentalpoltik’, in H. Schottelius and W. Deist (eds.), Marine
und Marinepolitik im kaiserlichen Deutschland 1871–1914 (Dusseldorf: Droste).
——(1988), ‘Germany and the coming of war’, in R. Evans and H. Pogge von
Strandmann (eds.), The Coming of the First World War (Oxford: Oxford
University Press).
——ed. (1988a), Walter Rathenau: Industrialist, Banker, Intellectual and Polit-
ician: Notes and Diaries 1907–1922 (Oxford: Clarendon).
——(1992), ‘The Liberal power monopoly in the cities of Imperial Germany’,
in E. L. Jones and J. Rettalack (eds.), Elections, Mass Politics and Social
Change in Modern Germany: New Perspectives (Cambridge: Cambridge Uni-
versity Press).
——(1999), ‘Revolutionsfurcht, Sozialreformen und wirtschaftliches Gleichge-
wicht: Anmerkungen zur Thunenrezeption’, in O. N. Haberl and T. Korenke
(eds.), Politische Deutungskulturen: Festschrift fur Karl Rohe (Baden: Nomos).
——and Geiss , I. (1965), Die Erforderlichkeit des Unmoglichen: Deutschland am
Vorabend des Ersten Weltkrieges (Frankfurt am Main: Europaische Verlagsan-
stalt).
Pohl , E. (1908), Die Lohn und Wirtschaftsverhaltnisse der Landarbeiter in Masu-
ren in den letzten Jahrzehnten, Diss., Konigsberg.
380 References and Sources
![Page 390: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/390.jpg)
——(1984), ‘Forms and phases of industrial Wnance up to the Second World
War’, German Yearbook on Business History, 75–94.
Pohl , M. (1976), Einfuhrung in die deutsche Bankengeschichte (Frankfurt am
Main: Knapp).
——(1986), Entstehung und Entwicklung des Universalbankensystems (Frankfurt
am Main: Knapp).
Pollard , S., ed. (1980), Region und Industrialisierung (Gottingen: Vanden-
hoeck und Ruprecht).
——(1981), Peaceful Conquest: The Industrialization of Europe 1760–1970
(Oxford: Oxford University Press).
Prados de la Escosura , L. (2000), ‘International comparisons of real prod-
uct: an alternative data set’, Explorations in Economic History, 37/1: 1–41.
Prest , A. R. (1948), War Economies of Primary Producing Countries (Cam-
bridge: Cambridge University Press).
Prokopovitch , S. (1926), ‘The distribution of national income’, Economic
Journal, 36: 69–82.
Prosterman , R. and Ri edinger , J. (1987), Land Reform and Democratic
Development (Baltimore: Johns Hopkins University Press).
Puhle , H.-J. (1972), Von der Agrarkrise zum Prafaschismus (Wiesbaden:
F. Steiner).
——(1976), ‘Parlament, Parteien und Interessenverbande 1890–1914’, in M.
Sturmer (ed.), Das Kaiserliche Deutschland (Dusseldorf: Droste Verlag).
——(1986) ‘Peasants and Lords in the Kaiserreich’, in R. Moeller (ed.), Peas-
ants and Land in Modern Germany: Recent Studies in Agricultural History
(London: Allen & Unwin).
Quante , P. (1933), ‘Die Flucht aus der Landwirtschaft’, Zeitschrift des Koni-
glich Preu�ischen Statistischen Bureaus, 3: 277–380.
——(1958), Die Abwanderung aus der Landwirtschaft, Kieler Studien 48.
——(1959), ‘Die Bevolkerungsentwicklung der preu�ischen Ostprovinzen im
19. und 20. Jahrhundert’, Zeitschrift fur Ostforschung, 8/4: 481–99.
Radetzki , W. (1930), Die inlandischen landwirtschaftlichen Wanderarbeiter
Deutschlands, Arbeiten aus dem Institut fur Wirtschaftslehre der Universitat
Breslau.
Ranis , G. and Schultz , T. P. (1988), The State of Development Economics:
Progress and Perspectives (Oxford: Basil Blackwell).
Rauchberg , H. (1900), ‘Die Landwirtschaft im deutschen Reich’, Archiv fur
soziale Gesetzgebung und Statistik, 15: 554–97.
Rauh , M. (1973), Foderalismus und Parlamentarismus im Wilhelminischen Reich
(Dusseldorf: Droste).
——(1977), Die Parlamentarisierung des Deutschen Reiches (Dusseldorf: Droste
Verlag).
References and Sources 381
![Page 391: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/391.jpg)
Ravallion , M. and Chen , S. (1997), ‘What can new survey data tell us
about recent changes in distribution and poverty?’, World Bank Economic
Review, 11: 357–82.
Rehbein , F. (1911), Das Leben eines Landarbeiters (Jena: E. Diederichs).
Re ich , U. (1998), Aus Cottbus und Arnswalde in die Neue Welt: Amerikaauswan-
derung aus Ostelbien im 19. Jahrhundert (Osnabruck: Rausch).
Re if , H., ed. (1994), Ostelbische Agrargesellschaft im Kaiserreich und in der Wei-
marer Republik (Berlin: Akademie Verlag).
Retallack , J. (1988), Notables of the Right (Boston: Unwin Hyman).
Reynolds , L. G., ed. (1975) Agriculture in Development Theory (New Haven:
Yale University Press).
R ichardson , H. W. (1965), ‘Over-commitment in Britain before 1930’,
Oxford Economic Papers, 17: 237–62.
R iehl , W. H. (1854), Die Naturgeschichte des Volkes als Grundlage einer
deutschen Sozial-Politik, vol. 1: Land und Leute (Stuttgart: J. G. Cotta).
R i esser , J. (1911), The German Great Banks (Washington: National Monetary
Commission).
R itter , G. (1980), Wahlgeschichtliches Arbeitsbuch: Materialien zur Statistik des
Kaiserreichs 1871–1918 (Munich: C. A. Beck).
——and Tenfelde , K. (1992), Arbeiter im Deutschen Kaiserreich 1871 bis 1914
(Bonn: Dietz).
Rohe , K. (1982), ‘Die ‘‘verspatete’’ Region’ in P. Steinbach (ed.), Probleme der
politischen Partizipation im Modernisierungsprozess (Stuttgart: Klett-Cotta).
Rohl , J. (1969), Germany without Bismarck: the Crisis of Government in the
Second Reich, 1890–1900 (Tubingen: Rainer Wunderlich Verlag).
Rolfes , M. (1976), ‘Landwirtschaft 1850–1914’, in H. Aubin and W. Zorn
(eds.), Handbuch der deutschen Wirtschaft- und Sozialgeschichte, vol. 2 (Stutt-
gart: Klett-Cotta).
Romer , C. (1986), ‘Spurious volatility in historical unemployment data’, Jour-
nal of Political Economy, 94: 1–37.
——(1989), ‘The pre-war business cycle reconsidered: new estimates of gross
national product, 1869–1908’, Journal of Political Economy, 97: 1–37.
Rosenberg , H. (1985), ‘The pseudo-democratisation of the Junker class’, in
G. Iggers (ed.), The Social History of Politics (Leamington Spa: Berg).
Roßler , H., ed. (1995), Es zieht eben einer den Anderen nach: Wanderungen und
ihre Wirkungen auf ausgewahlte Gebiete Schleswig-Holsteins und Ostelbiens
(Scripta Mercaturae Verlag).
Rostas , L. (1948), Comparative Productivity in British and American Industry,
NIESR Occasional Paper no. 13.
Saalfeld , D. (1963), ‘Zur Frage des bauerlichen Landverlustes im Zusam-
menhang mit den preu�ischen Agrarreformen’, Zeitschrift fur Agrargeschichte
und Agrarsoziologie, 11: 163–71.
382 References and Sources
![Page 392: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/392.jpg)
Sauer , W. (1968), ‘Das Problem des Nationalstaats’, in H. Bohme (ed.), Pro-
bleme der Reichsgrundungszeit 1848–1879 (Cologne: Kiepenheuer und Witsch).
Schama , S. (1995), Landscape and Memory (London: HarperCollins).
Schieder , T. (1961), Das Deutsche Kaiserreich von 1871 als Nationalstaat (Col-
ogne: Westdeutscher Verlag).
Schissler , H. (1986), ‘The Junkers: notes on the social and historical sign-
iWcance of the agrarian elite in Prussia’, in R. Moeller (ed.), Peasants and
Land in Modern Germany: Recent Studies in Agricultural History (London:
Allen & Unwin).
Schm Adeke , J. (1995), Wahlerbewegung im Wilhelminischen Deutschland. Vol. 1:
Die Reichstagswahlen von 1890 bis 1912: eine historisch-statistische Untersu-
chung. Vol. 2: Die Reichstagswahlen von 1890 bis 1912: Wahlergebnisse und
Strukturen im Kartenbild (Berlin: Akademie Verlag).
Schmoller , G. (1895), ‘Die Einkommensverteilung in alter und neuer Zeit’,
Jahrbuch fur Gesetzgebung, Verwaltung und Volkswirtschaft im deutschen Reich,
19: 1067–94.
——(1895a), ‘Uber die Einkommensverteilung’, Jahrbuch fur Gesetzgebung und
Statistik.
——(1900), Grundriss der allgemeinen Volkswirtschaftlehre (Munich: Duncker
und Humblot).
Schneppe , F. (1960), ‘Die Ertragssteigerung auf den leichten Boden der
Luneburger Heide’, Zeitschrift fur Agrargeschichte und Agrarsoziologie, 8/1:
14–22.
Schniedewind , K. (1994), Begrenzter Aufenthalt im Land der unbegrenzten
Moglichkeiten: Bremer Ruckwanderer aus Amerika 1850–1914 (Stuttgart: Franz
Steiner).
Schofield , R. S., Reher , D. and Bideau , A., eds. (1991), The Decline of
Mortality in Europe (Oxford: Oxford University Press).
Sch ollgen , G., ed. (1990), Escape into War? The Foreign Policy of Imperial
Germany (Oxford: Berg).
Schorske , C. (1955), German Social Democracy 1905–1917: The Development
of the Great Schism (Cambridge, Mass.: Harvard University Press).
Schott , S. (1903), ‘Das Wachstum der deutschen Grosstadte seit 1871’, Sta-
tistisches Jahrbuch deutscher Stadte, 11: 129–49.
Schremmer , E. (1988), ‘Faktoren, die den Fortschritt in der deutschen Land-
wirtschaft im 19. Jahrhundert bestimmen’, Zeitschrift fur Agrargeschichte und
Agrarsoziologie, 36/1: 37–77.
Schultz , T. W. (1964), Transforming Traditional Agriculture (New Haven:
Yale University Press).
Sering , M. (1893), ‘Die Innere Kolonisation im ostlichen Deutschland’,
Schriften des Verein fur Sozialpolitik, 56.
References and Sources 383
![Page 393: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/393.jpg)
Sering , M. ed. (1899–1910), Die Vererbung des landlichen Grundbesitzes im
Konigreich Preu�en, 4 vols., Landwirtschaftliche Jahrbucher, vol. 28, supp. vol.
1; vol. 29, supp. vol. 3; vol. 37, supp. vol. 5; vol. 34, supp. vol. 2.
——ed. (1932), Die Deutsche Landwirtschaft, Berichte uber Landwirtschaft
Sonderheft 50 (Berlin: P. Parey).
S i emens , W. von (1893, trans. 1966), Inventor and Entrepreneur: Recollections
of Werner von Siemens (London: Lund Humphries).
S i evers , K., ed. (1981), Die deutsche und skandinavische Amerikaauswanderung
im 19. und 20. Jahrhunderts (Neumunster: Karl Wachholtz).
Skrubbeltrang , F. (1954), Agricultural Development and Rural Reform in
Denmark, FAO Agricultural Studies no. 22.
Soetbeer , A. (1882), Umfang und Verteilung des Volkseinkommens im Preu�-
ischen Staate, 1872–1878 (Leipzig: Duncker & Humblot).
Sperber , J. (1983), ‘The shaping of political Catholicism in the Ruhr Basin,
1848–1880’, Central European History, 16/4: 347–67.
——(1997), The Kaiser’s Voters: Electors and Elections in Imperial Germany
(Cambridge: Cambridge University Press).
Spree , R. (1988), Health and Social Class in Imperial Germany (Oxford: Berg).
Squire , L. (1981), Employment Policy in Developing Countries. (Washington:
World Bank).
Stargardt , N. (1994), The German Idea of Militarism: Radical and Socialist
Critics 1866–1914 (Cambridge: Cambridge University Press).
Stark , O. (1991), The Migration of Labour (Oxford: Basic Blackwell).
Statistisches Reichsamt (1932), Das deutsche Volkseinkommen vor und nach dem
Krieg, Einzelschriften zur Statistik des Deutschen Reichs 24 (Berlin: R. Hob-
bing).
Stegmann , D. (1970), Die Erben Bismarcks: Parteien und Verbande in der
Spatphase des wilhelminischen Deutschlands—Sammlungspolitik, 1897–1918
(Cologne: Kiepenheuer & Witsch).
Steinbach , P. (1992), ‘Reichstag elections in the Kaiserreich: the prospects
for electoral research in the interdisciplinary context’, in E. L. Jones and
J. Rettalack (eds.), Elections, Mass Politics and Social Change in Modern
Germany: New Perspectives (Cambridge: Cambridge University Press).
Steller , P. (1903), ‘Die Maschinenindustrie Deutschlands’, Schriften des Ver-
eins fur Sozialpolitik, 106/3.
Stern , F. (1961), The Politics of Cultural Despair (Berkeley: University of
California Press).
Stiglitz , J. (1974), ‘Incentives and risk sharing in share-cropping’, Review of
Economic Studies, 41: 219–55.
St Ubel , H. (1935), Staat und Banken im Preu�ischen Anleihewesen von
1871–1913 (Berlin: C. Heymann).
St Urmer , M., ed. (1976), Das Kaiserliche Deutschland (Dusseldorf: Droste
Verlag).
384 References and Sources
![Page 394: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/394.jpg)
——(1990), ‘A nation state against history and geography: the German di-
lemma’, in G. Schollgen (ed.), Escape into War? The Foreign Policy of Imperial
Germany (Oxford: Berg).
Suval , S. (1985), Electoral Politics in Wilhelmine Germany (Chapel Hill: Uni-
versity of North Carolina Press).
Taylor , L. and Anda , P. (1988), ‘Long run income distribution and growth’,
in H. Chenery and T. Srinivasan (eds.), Handbook of Development Economics,
vol. 1 (Amsterdam: North-Holland).
Taylor , A. M. and Will iamson , J. G. (1997), ‘Convergence in the age of
mass migration’, European Review of Economic History, 1: 27–63.
Taylor , L. (1979), Macro Models for Developing Countries (New York:
McGraw-Hill).
Tcherkinsky , M. (1922), The Landschaften and their Mortgage Credit Oper-
ations in Germany 1770–1920 (Rome: International Institute of Agriculture).
Temple , J. (1999), ‘Income Distribution in the Harris–Todaro Model’, manu-
script, NuYeld College, Oxford.
Tenfelde , K. (1977), Sozialgeschichte der Bergarbeiterschaft an der Ruhr im 19.
Jahrhundert (Bonn: Verlag Neue Gesellschaft).
Thiel , H. (1887), ‘Der Verbrauch von Kunstdunger im Preussischen Staate’,
Landwirtschaftliche Jahrbucher, 16: 447–79.
Thomas , B. (1973), Migration and Economic Growth: A Study of Great Britain
and the Atlantic Economy, rev. edn. (Cambridge: Cambridge University
Press).
Thompson , A. P. (2000), Left Liberals, the State and Popular Culture in Wil-
helmine Germany (Oxford: Oxford University Press).
Th Unen , J. H. von (1826), Der Isolierte Staat in Beziehung auf Landwirtschaft
und Nationalokonomie (Berlin: Wiegandt, Hempel und Parey).
T illy , R. (1986), ‘German banking 1850–1914: development assistance for the
strong’, Journal of European Economic History, 15: 113–52.
T ipton , F. R. (1976), Regional Variations in the Economic Development of
Germany in the Nineteenth Century (Middletown, Conn.: Wesleyan University
Press).
Todaro , M. P. (1969), ‘A model of labour migration and urban unemploy-
ment in less developed countries’, American Economic Review, 59: 138–48.
——(1997), Economic Development (New York: Longman).
Tracey , M. (1982), Agriculture in Western Europe: Challenge and Response,
1880–1980 (London: Granada).
Twose , A. (2001), Statistics and the German State, 1900–1945: The Making of
Modern Economic Knowledge (Cambridge: Cambridge University Press).
US Department of Commerce (1975), Historical Statistics of the United States:
Colonial Times to 1970.
US Immigration Commission (1909–10), Immigrants in Industries, 61st Con-
gress, 2nd Session, Senate Documents, no. 633, vols. 68–85.
References and Sources 385
![Page 395: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/395.jpg)
US Immigration Commission (1910–11), Statistical Review of Immigration
1820–1910, 61st Congress, 3rd Session, Senate Documents, no. 756, vol. 20.
Van Zanden , T. (1991), ‘The Wrst green revolution’, Economic History
Review, 44: 643–64.
Verein fur Sozialpolitik (1893), ‘Verhandlungen des 20. und 21. Marz 1893’,
Schriften des Vereins fur Sozialpolitik.
Voelcker , A. (1883), ‘Manure’, Encyclopaedia Britannica, 15: 505–12.
Vogele , J. (1998), Urban Mortality Change in England and Germany,
1870–1913 (Liverpool: Liverpool University Press).
——and Woelk , W., eds. (2000), Stadt, Krankheit und Tod: Geschichte der
stadtischen Gesundheitsverhaltnisse wahrend der Epidemiologischen Transition
(Berlin: Duncker & Humblot).
Wade , R. (1990), Governing the Market: Economic Theory and the Role of
Government in East Asian Industrialisation (Princeton: Princeton University
Press).
Wagner , A. (1871), Rede uber die Soziale Frage. Gehalten am 12 Oktober 1871
(Berlin: Wiegandt und Grieben).
——(1902), Agrar- und Industriestaat (Jena: G. Fischer).
——(1904), ‘Zur Methodik der Statistik des Volkeinkommens und Volksver-
mogens’, Zeitschrift des Koniglich Preu�ischen Statistischen Bureaus, 2: 41–122.
——(1904a), ‘Weitere statistische Untersuchungen uber die Verteilung des
Volkseinkommen in Preu�en’, Zeitschrift des Koniglich Preu�ischen Statis-
tischen Bureaus, 4: 229–67.
Wagner , C., trans. G. Skelton (1978–80), Cosima Wagner’s Diaries (London:
Collins).
Wagner , P. (1896), Die Steigerung der Rohertrage der Landwirtschaft im Laufe
des 19. Jahrhunderts, Diss., Jena.
Walker , M. (1964), Germany and the Emigration, 1816–1885 (Cambridge,
Mass.: Harvard University Press).
Webb , S. (1982), ‘Agricultural protection in Wilhelmian Germany: forging an
empire with pork and rye’, Journal of Economic History, 42: 309–26.
Weber , A. (1908, trans. 1976), ‘From ‘‘Die Grossstadt und ihre soziale Pro-
bleme’’ ’, in A. Lees and L. Lees (eds.), The Urbanisation of European Society
in the Nineteenth Century (Lexington: Heath).
Weber , A. F. (1899), The Growth of Cities in the Nineteenth Century: A Study
in Statistics (New York: Macmillan).
Weber , M. (1893), ‘Die Lage der Landarbeiter im ostelbischen Deutschland’,
Schriften des Verein fur Sozialpolitik, 55.
——(1894, trans. 1979), ‘Developmental tendencies in the situation of East
Elbian rural workers’, Archiv fur soziale Gesetzgebung und Statistik, 7: 1–41,
translated in Economy and Society (Berkeley: University of California Press).
386 References and Sources
![Page 396: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/396.jpg)
——(1948), ‘Capitalism and rural society in Germany’, in H. H. Gerth and
C. Wright Mills (eds.), From Max Weber: Essays in Sociology (London:
Kegan Paul).
——(1993), Landarbeiterfrage, Nationalstaat und Volkswirtschaftspolitik, ed.
W. Mommsen, Vol. 4 of M. Weber, Gesamtausgabe Abteilung I: Schriften und
Reden, 1892–1899.
Wegge , S. (1997), Migration Decisions in Mid-Nineteenth Century Germany,
Diss., Northwestern University, Evanston, Illinois.
——(1998), ‘Chain migration and information networks: evidence from nine-
teenth century Hesse-Cassel’, Journal of Economic History, 58/4: 957–86.
Wehler , H.-U. (1969), Bismarck und der Imperialismus (Cologne: Kiepenheuer
& Witsch).
——(1985), The German Empire 1871–1918 (Leamington Spa: Berg).
——(1995), Deutsche Gesellschaftsgeschichte. Vol. 3: 1849–1913 (Munich:
C. Beck).
Wellh oner , V. (1989), Gro�banken und Gro�industrie im Kaiserreich (Gottin-
gen:Vandenhoeck und Ruprecht).
Wenckstern , H. von (1911), Existenz-Bedingungen se�hafter Landarbeiter,
Diss., Rostock.
Wenneman , A. (1997), Arbeit im Norden: die Italiener im Rheinland und West-
falen des spaten 19. und fruhen Jahrhunderts, Schriften des IMIS 2 (Osnabruck:
Rausch).
W ill iamson , J. (1988), ‘Migration and urbanisation’, in H. Chenery and
T. Srinivasan (eds.), Handbook of Development Economics, vol. 1 (Amsterdam:
North-Holland).
W ill iamson , J. G. (1985), Did British Capitalism Breed Inequality? (Boston:
Allen & Unwin).
——(1991), Inequality, Poverty and History (Cambridge, Mass.: Basil Black-
well).
——(1995), ‘The evolution of global labour markets since 1830: background
evidence and hypotheses’, Explorations in Economic History, 32: 141–95.
——(1997), Growth, Distribution and Demography: Some Lessons from History,
NBER Working Paper no. 6244.
W inkler , J. (1993), Sozialstruktur, politische Traditionen und Liberalismus: eine
empirische Langsschnittstudie zur Wahlentwicklung in Deutschland, 1871–1933
(Opladen: Westdeutscher Verlag).
W itt , P.-C. (1970), Die Finanzpolitik des Deutschen Reiches von 1903 bis 1913
(Hamburg: Matthiesen).
——(1973), ‘Der preußische Landrat als Steuerbeamter 1891–1918: Bemerkun-
gen zur politischen und sozialen Funktion des deutschen Beamtentums’, in
I. Geiss and B. Wendt (eds.), Deutschland in der Weltpolitik des 19. und 20.
Jahrhunderts (Dusseldorf: Bertelsmann Universitatsverlag).
References and Sources 387
![Page 397: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/397.jpg)
Witt , P.-C. (1985), ‘The Prussian Landrat as tax oYcial, 1891–1918’, in G.
Iggers (ed.), The Social History of Politics (Leamington Spa: Berg).
World Bank (1994), The East Asian Miracle: Economic Growth and Public Policy,
World Bank Policy Research Report.
Wrigley , E. A. (1961), Industrial Growth and Population Change: A Regional
Study of the CoalWeld Areas of North-West Europe in the Later Nineteenth
Century (Cambridge: Cambridge University Press).
Wygodzinski , W. (1917), Die Landarbeiterfrage in Deutschland (Tubingen:
J. C. B. Mohr).
388 References and Sources
![Page 398: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/398.jpg)
INDEX
absent revolution thesis 23
agrarian reform 45
and aristocracy 34, 36
and banking system 32–3
in central and north-west
Germany 31–3
and child care 141
and compensation 24, 41–2
and consolidation of holdings 44
and credit 25, 32–3, 35
and demographic pressures 26
and Denmark 26, 28–31
and difficulty of 23
east of the Elbe 34–8
aristocracy’s role 36
distance from industrial areas 36
limited credit 35
peasants’ weak position 36
poor communications 34–5
success of 36–8
supposed failure of 34
and effects of 191
egalitarianism 53
and farm prices 26–8
and infant mortality 141
and labour service 31–2, 39, 41
and land ownership
distribution 320
and landlords’ land gains 32, 33
and market access 25–6
and objectives of 24, 28
and population growth 141
and productivity 24–5, 29
in Prussia 31
cultivated land area 40
peasant land losses 38–40
regional variation in impact
of 40–3
and regional variations in 43
and social costs of 43
and Sonderweg school 23
in southern Germany 44–5
and standards of comparison 23
and statistical analysis of 23
as test of political process 23–4
agriculture:
and contemporary views of 181
and cultivated land area 40
and debt levels 245–6
and demesne-farming 37–8
and economic development 15
and the economy:
decline of importance in 218–25
domestic demand 219, 220, 221,
223
international trade 216–23
protection 217–18, 221,
222, 224
and education 226
and emigration 81–2, 85–8
and enclosure 32, 34, 41–2
and farm size 48–9, 51–2
and immigration 92, 93
and impact of capitalist
values 12
and industrialization 17, 185–6
consequences of 187
impact of 215–18
and institutional problems
with 181
and Kuznets Curve 53–4
and labour contract system 183,
185, 208
and labour force 218
employed 46–7
productivity increase
246–7
self-employed 47
![Page 399: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/399.jpg)
agriculture: (Continued)
and labour surplus 193, 209
and land ownership
distribution 49–53
tenancies 52–3
and Lewis Model 207–9
and migration 5, 6, 12–13, 102–4,
106, 109
concerns over 190–2
consequences of 187
costs of 186–7
large estates 192, 193, 194
market conditions 104–5, 107–8,
113
population density 249–50,
251–2
seasonal labour 74–5
as symptom of social
problems 191
and missing market hypothesis 184
and mutual obligations 186–7
and owner-occupation 46, 47
and payments-in-kind 182–3
and population changes 199–201
and prices 104–5, 220–1, 222
fall in 197
and productivity 16–17, 24–5, 29,
215–16
artificial fertilizers 234–5,
239–40, 242, 244
breeding improvements 245
compared with France 46
impact of increase 216
improvements in 218, 223–4
international comparisons 230–4
labour force 246–7
labour-saving technology 247–8
large estates 241–2
livestock intensity 229–30, 242,
245
migration 113, 123, 237, 246
regional patterns of 235–45
soil quality 242–4
sugar beet cultivation 239–41
yield improvement 227–9, 239,
244
and quit rate from 148–51, 171–3,
197–9
farm size 149–50, 171, 197
geographic variations 150–1
population growth 173
and structure of:
fertility decisions 142–3, 169
migration 55, 107, 113
and technological
improvements 201, 216,
225–6, 234–5, 242–5
labour-saving 247–8
and transport costs 25–6
and underemployment 12, 16,
185–8, 204–7, 214,
247–9, 273
partible inheritance 205–6
and wages:
influences on 201–4, 207–8,
211–13
rising level of 196–7
and Weber (Max) on 193–7
see also agrarian reform
Argentina, and emigration to 81
aristocracy, and agrarian reform 34,
36
Army Bill (1913) 332
artificial fertilizers, and
productivity 234–5, 239–40,
242, 244
‘Asian values’, and industrialization 4
Atkinson, A B 299
Australia 216
Austria-Hungary, and migration to
Germany 91
authoritarianism, and
industrialization 4–5
Baden, and agrarian reform 44
Bairoch, P 46, 230, 232
390 Index
![Page 400: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/400.jpg)
banking system:
and agrarian reform 32–3
Denmark 30
and consolidation of 257
and expansion of 185
and investment 15, 254
financing of 257–9
and rural credit 25
and underdevelopment of 320
Bauernschutz 31, 36
Bavaria, and agrarian reform 44
Belgium, and immigration 80
Berlin, migration to 63–5
and further moves 71–3
and migrant occupations
66–7
and sources of migrants
68–70, 75
Berlin Statistical Bureau 98
Berliner Handelgesellschaft 257
birth rate 127
and cities 153–4
and decline in 317
and economic conditions 142
and less-developed
countries 317
and migration 106–7
and rural-urban differences 153
Bismarck, Otto von 346
Blackbourn, D 3 n4
Brandenburg, and agrarian
reform 41, 42, 43
Brazil 216
and industrialization 5
and land ownership
distribution 315
and soya-bean production 314
Broadberry, S N 233, 234
Broesike, Dr M 98
Browne, M 263
bureaucracy, and unrepresentativeness
of 2
Canada, and emigration to 81
capital:
and mobility of, and Lewis
Model 318–19
and ownership distribution 312,
313
Catholicism:
and demographic behaviour 154,
168–9
and population growth 146, 155–6,
174
and Social Democratic Party
support 337–9
child care, and agrarian reform 141
Chile, and industrialization 4
China, and labour surplus 352, 353
cities:
and birth rates 153–4
and contemporary views of 188–9
and death rates 154–5, 157–8,
174–5
and employment 268
and fertility rates 145–6,
152, 169
and impact of agricultural
decline 217–18
and inequality 323–4, 325–6
and life expectancy 157–8, 283
and migration to 63–6, 111–12
congestion effect 275
further moves 71–3
migrant occupations
66–7
sources of 68–70
and opportunities in 160
and population growth 151–6,
173–7
and rate of growth of 59, 61–2,
151, 293
and skill levels 312
see also urbanization
Clapham, J H 31, 225–6, 239
Index 391
![Page 401: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/401.jpg)
Cologne:
and migrant occupations 66
and migration to 65–6
communications:
and agrarian reform 34–5
and improvements in 157
and migration 113
compensation, and agrarian
reform 24, 41–2
Conrad, J 44
consolidation, and agrarian reform 44
constitution, and election
of 1912 332
corporations, and investment 254
Craigie, P G 48–9
credit:
and agrarian reform 25, 32–3, 35
and Danish agrarian reform 30
and expansion of 185
Darmstadter Bank 257
Dasgupta, Partha 184 n5, 186, 187
death rates 127
and cities 154–5, 157–8, 174–5
and economic conditions 131
and economic development 156–7
and migration 158
and migration analysis 71, 106
and wages 154–5, 170
debt, and agriculture 40 n44, 245–6
demography:
and birth rates 127
cities 153–4
decline in 317
economic conditions 142
less-developed countries 317
rural-urban differences 153
and comparison of east and
south 147
fertility rates 137
marriage and migration 137–9
marriage rates 134–7
migration movements 139–41
non-agricultural
employment 132–4
and death rates 127, 131
cities 154–5, 157–8, 174–5
economic development 156–7
wages 154–5, 170
and demographic transition 124
regional variations 127–8
and economic factors:
influence of 130–1
non-agricultural
employment 132–4, 172–3
and fertility rates 127, 131, 137
agricultural institutions 142–3
decline in 159
factors affecting 144, 146, 147,
152–3
large estates 146, 169
partible inheritance 143, 144–5,
169, 318
proximity to cities 145–6, 152,
169
in rural Prussian Kreise 142–5,
167–70
and infant mortality 141
and life expectancy 283
international comparisons 158
regional variations 157–8
and marriage rates 131
economic conditions 131, 137,
177–80
migration 137–9
proximity to cities 152
regional variations 134–7
and migration 112, 121–2, 124–5,
160
birth rate 106–7
death rates 106
population growth 129–30
regional variations 108–10
and population growth 126–7
Catholicism 146, 155–6, 168–9,
174
392 Index
![Page 402: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/402.jpg)
in cities 151–6, 173–7
delayed marriage 177–8
industrial structure 156, 174,
175
industrialization 141
inequality 147–8
Lewis Model 156–60
regional variations 128–30,
132–4
rural migration 148–51, 170–3
rural regulation of 142, 159, 317
wages 175
and population movements 125–7
Denmark, and agrarian reform 26,
28–31
and agricultural productivity 29
and context of 31
and financing of 30
and nobility 28–9
and non-legislative aspects 29
and peasants 30–1
and state initiative 28
and unequal benefits 30
Deutsche Bank 257
development economics:
and relative backwardness 3–4
see also Lewis Model
Disconto-Gesellschaft 257
Dresdner bank 257
economic development:
and agriculture 15
as alteration of production
function 310–11
and death rates 156–7
and financial sector 254
and inequality 14, 295
and investment 253–4
and land ownership
distribution 315
and ‘leapfrogging’ 264–5
and migration 6, 112, 122–3, 350
and model of 294
economy:
and agriculture:
decline in importance of 218–25
domestic demand 219, 220, 221,
223
industrialization 17, 185–6, 187,
215–18
international trade 218–21,
222–3
prices 220–1, 222
protection 217–18, 221, 222, 224
and consequences of migration:
skill endowments 279–81
wage convergence 274–9
and economic stabilization 351–2
and migration and economic
cycles 271–4, 291–2
and structural changes in 255–66
Anglo-German rivalry 263–4
export growth 255–6, 260–1,
265–6
financial sector 257
growth potential 265
investment 256–7
investment finance 257–9
new industries 264–5
raw material costs 262–3
recovery from recession 259–60
terms of trade 261–2, 264
wages 263–4
education:
and agriculture 226
and income distribution 311–12
and industrialization 215–16
Eggers, Count von 30–1
election of 1912:
and foreign policy 333
and political consequences
of 332–3
and Social Democratic Party
support:
agricultural population 341
aristocratic estates 342–3
Index 393
![Page 403: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/403.jpg)
election of 1912: (Continued)
candidate composition 343–5
city size 337, 339
desire for change 343
econometric analysis 356–60
economic conditions 340, 342,
343
feelings of political
inequality 343
increase in 331–2
inequality 340, 342
population mobility 340
proximity to towns/cities 341–2
religious affiliation 337–9, 341
rural/small town areas 340–3
and statistical analysis, limitations
of 336–7
electoral politics:
and radical right 334
and urbanization 335–6
and voting trends 333–4
see also election of 1912
Elektroboom 256, 259
Eley, G 1 n1
elites, and unwillingness to
modernize 2
emancipation, and population
growth 142
emigration:
and agricultural character of 85–8
and decline in 89
and destination countries 81
and Germans living abroad 80–1
and labour surplus 56–7, 89
and regional variations in 85–6
and structural change in 80
to United States 81, 83–4, 321
agricultural employment 88–9
decline in 85
influence of agriculture 81–2
previous occupations 86–8
wage ratio 85
see also immigration; migration
employment:
and agriculture 46–7, 218, 246–7
and export-oriented
industries 266–7
and household 206, 212, 214, 249
and migrant participation rates 273
and non-agricultural 132–4, 172–3,
199
industrial concentration 268–9,
286–8
regional pattern of 267–8, 271
vertical integration 268, 269–71,
288
and self-employment 47, 206, 214
and sugar beet cultivation 201
see also underemployment
enclosure 32, 41–2
and aristocracy 34
and Danish agrarian reform 29, 30
Engel, Ernst 98, 190, 297–8, 305
Engels, Friedrich 297
Evans, Richard 330–1
exports:
and finished goods 261
and growth in 255–6
and terms of trade 10, 254
expropriation, and agrarian reform 24
famines 157
farm size:
and agriculture 48–9, 51–2
and productivity 241–2
and quit rate from
agriculture 149–50, 171, 197
and wages 201–2, 203, 213
see also large estates
Feinstein, Charles 296, 307
fertility rates 127
and agricultural institutions 142–3
and decline in 159
and delayed marriage 131, 159,
177–8
and economic conditions 131
394 Index
![Page 404: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/404.jpg)
and factors affecting 144, 146, 147,
152–3
and large estates 146, 169
and partible inheritance 143,
144–5, 169, 318
and Polish-speaking areas 145, 146,
147, 167–8
and proximity to cities 145–6, 152,
169
and regional variations in 137
in rural Prussian Kreise 142–5,
167–70
and wages 146, 147
see also farm size
Finance Bill (1913) 332
financial sector:
and development of 254, 257
and financing of investment 257–9
and underdevelopment of 320
Finckenstein, Finck von 37 n35
Fircks, A von 131
foreign direct investment 320
foreign policy:
and election of 1912 333
and industrialization 5
and internal politics 1
and naval expansion:
economic motives for 347–8
lack of alternatives 348–9
and social context of 2–3
France:
and agriculture:
farm size 48–9
owner-occupation 46
productivity 46
and urbanization 59
Fremdling, R 233
Geisenberger, S 299, 300
gender, and migration 76–9, 137–8
see also women
Gerschenkron, Alexander 3
Gilbert and Lawes 244
Goltz, Theodor von der 57 n2,
191
and Prussian land reform 34
Gutsherrschaft administrative system,
and agrarian reform east of the
Elbe 36
Hajnal, J 178
Handdienst, and agrarian reform 41
Hayami, Y 318
Helfferich, K 298, 299
Hesse, and agrarian reform 44
historical school, and income
distribution 297–8
historiography:
and economic models 18–19
and German exceptionalism
19–20
of Imperial Germany 1–2, 18,
331–2
and Tirpitz naval plan 350
see also Kehrite school
Hoffmann, Walter 9, 37 n36, 104,
218, 235, 257, 300
household behaviour, and fertility
decisions 318
household employment:
and underemployment 206, 214,
249
and wage levels 212
illegitimacy 139
and death rates 142
immigration 90–3
and agriculture 92, 93
from Austria-Hungary 91
and countries of origin 91
and increase in 90
and industrial workers 92, 93
and labour surplus 57
and net gains from 80
and regional variations in 92–3
from Russia 90, 91
Index 395
![Page 405: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/405.jpg)
Imperial Germany:
and Anglo-German
relations 263–4, 346
and democratic transition 345–6
economic stabilization 351–2
potential for 353–4
and electoral politics of 331–5
growth of Social Democratic
Party 335–45
and German exceptionalism 19–20
and industrialization 3, 331–2
standards of comparison 20–1,
332
success in coping with 353
and internal weakness of 1–3, 330,
349
and naval expansion:
economic motives for 347–8
lack of alternatives 348–9
naval weakness 347
and problems faced by 333
and socio-strategic dilemma
of 346–7
Imperial Statistical Office 98
income distribution, see inequality
incomplete revolution thesis 23
industrial structure:
and birth rates 155
and death rates 154–5
and inequality 326
and migration 266–71
export-oriented industries 266–7
industrial concentration 268–9,
286–8
patterns of non-agricultural
employment 267–8, 286–90
vertical integration 268, 269–71,
288
and population growth 156, 174,
175
industrialization:
and adoption of capitalist values 12
and agriculture 17
eastern 185–7
impact on 215–18
structure of 55
and authoritarianism 4–5
and contemporary concerns
with 189–90
and education 215–16
and foreign policy 5
and German debate on 295
and inequality 6, 13–14, 297–8,
305, 307–8, 316, 320
and internal weakness of Imperial
Germany 3, 331–2
and Kuznets Curve 13–14, 53, 295
and Lewis Model 6–11
and migration 5–6, 7, 56, 94
periods of 71
sources of 75–6
and nature of German 3, 20–1
in pioneer countries 20
and population growth 141
and productivity growth 350
and relative backwardness 3–4
and rural society 185–6
consequences for 187
and speed of 59, 61, 62, 151, 293,
353
and standards of comparison 20–1,
332
and trade unions 7
and urbanization 15
inequality:
and avoidance of 326–7
and cities 323–6
and contemporary concerns
with 190
and debate over 297–9
and economic development 14, 295
and electoral politics 340, 342, 343
and historical school 297–8
and industrial structure 326
and industrialization 6, 13–14,
297–8, 305, 307–8, 316, 320
396 Index
![Page 406: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/406.jpg)
and international
comparisons 306–8
and Kuznets Curve 13–14, 53, 295
and labour surplus 310
and land ownership
distribution 49–53, 313–16,
324
and Lewis Model 6–11, 317
and migration 53, 56, 309–10, 316,
322, 325
and population growth 147–8
and productive knowledge 311–13
skills 311, 316, 321
vocational training 311–12, 321
and proximity to Russian
Poland 324
in Prussia, nineteenth
century 303–6
and rural society:
divergences in 22
rural-urban gap 13–14, 54–5,
108
and savings’ concentration 317
and skill endowments 311–12,
315–16, 321
and sources of 319–21
analysis by simulation 322–3
econometric analysis 323–6
and tax data:
opportunity provided by 303
problems with use of 299–302
and terms of trade 320, 323, 326
and urbanization 316, 323, 326
and worsening of 302–3
infant mortality:
and agrarian reform 141
and illegitimate children 142
inheritance, see partible inheritance
insurance 289–90
internal politics:
and foreign policy 1
and industrialization 3, 331
and primacy of 1
international trade:
and agriculture 216–17, 218–21,
222–3
and Anglo-German rivalry 263–4,
346
and export growth 255–6, 260–1,
265–6
employment 266–7
and German naval policy 347–8
and impact of 254–5
and new industries 264–5
and raw material costs 262–3
and terms of trade 10, 254, 261–2,
264
investment:
and banking system 254
and capital mobility 319
and corporations 254
and economic development
253–4
and financial sector 254
and financing of 257–9
and improved opportunities
for 313
and inequality 321
and institutional constraints
on 320
and pattern of 256–7
and savings 253–4
Ipsen-Kollmann thesis 141
Ireland, and land ownership
distribution 315
Italy, and migration to Germany 90
Japan:
and agrarian reform 26
and agricultural prices 27
and land ownership
distribution 315
joint-stock companies, and
investment 254
Junkers, and opinions of
226 n14
Index 397
![Page 407: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/407.jpg)
Kehrite school:
and election of 1912 333
and internal weakness of Imperial
Germany 1–2
and key concept of 1
and radicalization of the right 332
and Tirpitz naval plan 350
and weakness of argument 353
Kindleberger, C 56 n1
Knapp, Gustav 189
and migration from rural
east 191–2
and peasant land losses 38
and Prussian land reform 34
Knodel, J 137, 142
knowledge, productive 311–13
and skills 311, 316, 321
and vocational training 311–12,
321
Kollmann, W 135, 139, 152
and Ipsen-Kollmann thesis 141
Koselleck, R 34
Kuczynski, Robert 98
Kuznets, Simon 13, 295, 312,
316–17
Kuznets Curve 4
and agriculture 53–4
and assumptions of 22
and conditionality of 326, 327
as contemporary fiction 296
and disputing of 296
and escaping from 327
as German peculiarity 296, 297
as historical fact 296, 297
and inequality and
industrialization 13–14, 53,
295
and Lewis Model 294, 296
and migration 295, 322
in Prussia, nineteenth
century 303–6
international comparisons 306–8
labour market:
and ‘Continental’ nature of 80, 94
and impact of capitalist values 12
and industrial development 6–8
and integration of 276
and non-agricultural employment,
and migration 132–4, 172–3,
199
and population increase 127
see also employment
labour service:
and abolition in Denmark 29
and agrarian reform 41
abolition of 31–2, 39
and fertility decisions 142–3
and inefficiency of 38 n38
labour surplus:
and criticism of concept of 16, 215
and eastern agriculture 193, 209
and emigration 56–7, 89
and immigration 57
and impact of capitalist values 12
and implications of 294
and inequality 310
and Kuznets Curve 13–14
and Lewis Model 6–9, 294
and migration 56, 93–4, 97, 100,
102, 103–4, 141
and political stability 353
and sources of 293
and underemployment 204–7,
249–50
land ownership:
and agrarian reform 320
and Danish agrarian reform 29
and distribution of 49–53
and income distribution 313–16,
324
and marriage rates 143
and political power 314–15
and re-allocation of 315–16
and skill acquisition 315
398 Index
![Page 408: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/408.jpg)
and the state 314
and turnover of 39–40
land prices 104–5, 106
and migration 107–8
land reform, see agrarian reform
land tax assessment 244
Landesrentenbank 32, 33
Landflucht 34
land-owners, and adoption of
capitalist values 12
Landschaften 25, 33
large estates:
and economies of scale 184
and fertility rates 146, 169
and migration 192, 193, 194
and productivity 241–2
and quit rate from
agriculture 149–50, 171
and wages 201–2, 203, 213
Laux, H-D 151 n39
‘leapfrogging’, and economic
development 264–5
Lee, W 141
less-developed countries, and birth
rates 317
Levy, H 36 n30
Lewis, Arthur 156–7, 317
and employment 12
and the Lewis Model 6
Lewis Model 4
and capital mobility 318–19
and criticisms of 15–16
and eastern agriculture 207–9
and explanatory value of 350
and inequality 6–11, 317
and Kuznets Curve 294, 296
and labour supply 6–8
and labour surplus 6–9, 294
and migration 79–80, 97, 100
and modification to 9
and population growth 156–60
and propositions of 294
and shortcomings of 11
and ‘turning point’ 7, 8, 294
and wages 9–11
Liebig, Justus von 244, 245
life expectancy:
and international comparisons 158
and migration 283
and regional variations in 157–8
Lindert, Peter 296, 307
living standards:
in eastern Germany 139–41
and regulation of population
growth 159
and wage levels 212–13
local government, and Social
Democratic Party 332
Lupitz system 244, 245
Magdeburg, and peasant farmers
33
Malthus, Thomas 178
market access, and agrarian
reform 25–6
marriage:
and decline in fertility rates 159
and delayed 131, 159, 177–8
and economic conditions 131, 137,
177–80
and land ownership 143
and migration 137–9
and partible inheritance 145
and proximity to cities 152
and regional variations in 134–7
and wages 178–80
Marschalck, P 141
Marx, Karl 6
Mecklenburg, and agrarian
reform 45
medicine 157
Meitzen, Adolph:
and peasant debt 40 n44
and Prussian land reform 34
Meldepflicht system 57
middle class, and cities 151, 160
Index 399
![Page 409: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/409.jpg)
migration:
and agriculture 5, 6, 12–13, 102–4,
106, 109
concerns over 190–2
large estates 192, 193, 194
market conditions 104–5, 107–8,
113
productivity 113, 123, 237, 246
quit rate from 148–51, 171–3,
197–9
structure of 55, 107, 113
and demographic factors 112,
121–2, 124–5, 160
birth rate 106–7
death rates 106, 158
population growth 129–30
regional variations 108–10
and economic cycles 271–4, 291–2
and economic development 350
and economic factors 112, 122–3
non-agricultural
employment 132–4, 172–3,
199
and factors affecting attitude
towards 188
and gross migration 57
and historical research 97
and impact of 19
and industrial structure 266–71
export-oriented industries 266–7
industrial concentration 268–9,
286–8
patterns of non-agricultural
employment 267–8, 286–90
vertical integration 268, 269–71,
288
and industrialization 5–6, 7, 56, 94
and inequality 53, 56, 309–10, 316,
322, 325
rural-urban gap 13–14, 54–5, 108
and international 56–7
and international comparisons 58
and intra-urban 56
and Kuznets Curve 295, 322
and labour surplus 56, 93–4, 97,
100, 102, 103–4, 141
and Lewis Model 79–80, 97, 100
and marriage patterns 137–9
and measurement of 62–8
and net migration 57
and periods of 71, 94, 139
and population density 249–50,
251–2
and population mobility 63–6
migrant occupations 66–7
and regional variations in 101–2
and rural population
growth 148–51, 170–3
and rural-urban 56, 102–4, 110–13
concerns over 190–2
consequences of 187
costs of 186–7
as symptom of social
problems 191
and seasonal 56, 64–5, 74–5
by sex 76–9, 137–8
and skill endowments 279–81
vocational training 311–12
and social change 350–1
and sources of migrants 68–76
agriculture-related 74–5
further moves 71–3
industrial-related 75–6
and statistical analysis of:
changing nature of
(1871–1913) 100–6
net migration by region
(1871–1910) 106–10, 115–16
panel data analysis
(1871–1910) 116–19
regression variables and
sources 114–15
rural-urban (1890
census) 110–13, 119–23
and statistical sources for 98
and underemployment 214, 249
400 Index
![Page 410: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/410.jpg)
and unemployment 271–4
and urbanization 59–62, 111–12
and wages 56, 94, 111
wage convergence 274–9
wage dispersion 276
and women 76–9, 138–9
life expectancy 283
wages 281–3
see also emigration; immigration;
Lewis Model
missing market:
and industrialization 185–6
and share-cropping 184
Mittelstand, and preservation of 189
mobility 63
and cities 63–6
and migrant occupations 66–7
and sex of migrants 76–9
and sources of migrants 68–76
modernization, and incompleteness
of 2
Mommsen, Wolfgang:
and German exceptionalism 19
and ‘skirted decisions’ 1 n2
monitoring costs, and share-
cropping 183–4
Muller, J H 299, 300
Newbery, D 184 n4
newly industrializing countries
(NICs) 4
nobility, and Danish agrarian
reform 28–9
nutrition, and indicators of 139–40
O’Brien, P 233
occupations, and migration 66–7
Ostmarkverein 193
Pakistan, and land reform 46 n53
partible inheritance:
and fertility decisions 143, 144–5,
169, 318
and marriage rates 145
and underemployment 205–6,
214
patents, and productive
knowledge 311, 312
peasants:
and agrarian reform:
bias in favour of 32, 320
credit availability 32–3
Denmark 28, 30–1
east of the Elbe 36
labour service 31–2, 39
land losses 38–40
and demesne-farming 37–8
and efficiency of 16
and non-agricultural earnings
44–5
and preservation of 189
Phelps-Brown, E H 263
Philippines, and agrarian reform 25
n5
Poland, see Russian Poland
Polish-speakers:
and fertility rates 145, 146, 147,
167–8
see also Russian Poland
Pomerania:
and agrarian reform 41,
43, 45
and large estate failures 40
and peasant landholdings 38
Posen, and agrarian reform
41, 42
Prados de la Escosura, L 233
prices:
and agrarian reform 26–8
and agriculture 104–5, 197, 220–1,
222
and death rates 131
and marriage rates 178–9
and wages 212
production, and economic
development 310–11
Index 401
![Page 411: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/411.jpg)
productivity:
and agrarian reform 24–5, 29
and agriculture 16–17, 24–5, 29,
215–16
artificial fertilizers 234–5,
239–40, 242, 244
breeding improvements 245
compared with France 46
impact of increase 216
improvements in 218, 223–4
international comparisons 230–4
labour force 246–7
labour-saving technology 247–8
large estates 241–2
livestock intensity 229–30, 242,
245
migration 113, 123, 237, 246
regional patterns of 235–45
soil quality 242–4
sugar beet cultivation 239–41
yield improvement 227–9, 239,
244
and industrialization 350
and wages 7, 9–11
see also knowledge, productive
profits, and industrial development 6
Prokopovitch, S 13, 295
property rights:
and agrarian reform 24
and evolution of 315–16
and exceptional re-allocation
of 315
and productive knowledge 312–13
protection, and agriculture 217–18,
221, 222, 224
Prussia:
and agrarian reform 31, 42
consolidation 44
peasant land losses 38–40
regional variation in impact
of 40–3
and agrarian reform east of the
Elbe 34–8
aristocracy’s role 36
distance from industrial areas 36
limited credit 35
peasants’ weak position 36
poor communications 34–5
success of 36–8
supposed failure of 34
and agricultural prices 27–8
and rural credit 25
see also taxation data, Prussian
Prussian Statistical Bureau 98
public health 157
Quante, Peter 148
railway construction, and
migration 75–6
Rauh, Manfred 354
recruits, and height of 139–40
regional government, and Social
Democratic Party 332
regions 94–6
Reichstag, and constitutional position
of 332–3
relative backwardness:
and development economics 3–4
and industrialization 3
religious affiliation, and voting
behaviour 337–9, 341
Rentenbanken 33
‘revolutions from above’ 23
Rhineland, and agrarian reform
41
Royal Credit Bank (Denmark) 30
Ruhr, and sources of migrants 75
rural society:
and efficiency wages and
unemployment 187
and fertility rates 142–5,
167–70
and impact of capitalist values 12
and industrialization 185–6
consequences of 187
402 Index
![Page 412: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/412.jpg)
and inequality 22
rural-urban gap 13–14, 54–5,
108
and migration:
concerns over 190–2
consequences of 187
costs of 186–7
large estates 192, 193, 194
population density 249–50,
251–2
as symptom of social
problems 191
and mutual obligations 186–7
and population changes 199–201
and population growth:
migration 148–51
regulation of 142, 159, 317
and wages, influences on 201–4,
207–8, 211–13
see also agrarian reform
Russia:
and industrialization 208
and migration to Germany 90, 91
Russian Poland 201
and inequality and proximity
to 324
and influence on wage
levels 202–3, 204, 208, 213,
279, 320–1
as source of labour 194, 199, 208
Saalfeld, D 38
Sammlungspolitik 2
savings:
and economic development 15
and improved opportunities
for 313
and inequality 317
and investment 253–4
Saxony, Kingdom of:
and agrarian reform 32, 41, 42
and peasant landholdings 38
and rural credit 25
Saxony Statistical Office 98
Schaaf’hausen Bankverein 257
Schissler, H 36 n31
Schmoller, Gustav 189, 190, 297
Schremmer, E 234, 235
Schrott, S 62 n6
Schultz, Theodore 16–17, 215,
246
Schultz-Lupitz, A 244, 245
science, and agriculture 225–6,
244–5
seasonal migration 56, 64–5
self-employment:
and agriculture 47
and underemployment 206, 214
serfdom, and abolition in
Denmark 28
Sering, Max 190, 191
share-cropping:
and criticisms of 181–2
as efficient solution 184
and missing market 184
and monitoring costs 183–4
and positive views of 183
Silesia:
and agrarian reform 41, 42
and peasant landholdings 38
skill endowments:
and income distribution 311, 316,
321
and land ownership
distribution 315
and migration 279–81
and vocational training 311–12
‘skirted decisions’, and internal
weakness 1–2
smallholdings, and migrants 281–2
Smith, Adam 181
social change, and rural-urban
migration 350–1
Social Darwinism 2, 330
Social Democratic party 310
and broadening appeal of 335
Index 403
![Page 413: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/413.jpg)
Social Democratic party (Continued)
and electoral support for:
agricultural population 341
aristocratic estates 342–3
candidate composition 343–5
city size 337, 339
desire for change 343
econometric analysis 356–60
economic conditions 340, 342,
343
feelings of political
inequality 343
gains in 1912 election 331–2
growth of 333–6
inequality 340, 342
population growth 154, 156,
174, 175
population mobility 340
proximity to towns/cities 341–2
religious affiliation 337–9, 341
rural/small town areas 340–3
and internal politics of 334
and local/regional government 332
Soetbeer, A 298, 299
Sonderweg school 23, 330, 331
South Korea:
and industrialization 4
and land ownership
distribution 315
and land reform 46 n53
soya-bean production 314
Spain, and industrialization 4
Spanndienst, and agrarian reform 41
Spannfahige, and agrarian reform 32,
33
state, the, and land ownership
distribution 314
statistics, and contemporary use
of 98–9
steam-threshing 247–8
Stegmann, Dirk 332
Stein-Hardenburg laws 192
stillbirths 126, 135
sugar beet cultivation:
and artificial fertilizers 239–40
and employment 201
and productivity 239–41
and wages 204, 212
Switzerland, and emigration to 81
Tagelohner 15
Taiwan:
and industrialization 4
and land ownership
distribution 315
tariffs, see protection
taxation data, Prussian:
and income distribution 303–6
econometric analysis 323–6
measurement of 297–9
worsening of 302–3
and opportunity provided by 303
and problems with use of 299–302
technology:
and agriculture 216, 225–6, 234–5,
242–5
impact on employment 201
labour-saving 247–8
tenancies, and agricultural land
52–3
tenant rights, and Danish agrarian
reform 29, 30–1
terms of trade:
and inequality 320, 323, 326
and movement in 10–11, 254
Thunen, J H von 36 n30
Tirpitz, Alfred von 348
Tirpitz naval plan 348, 349
and Kehrite school 350
Todaro, M P 317
trade unions:
and industrialization 7
and wage levels 321
transport costs, and agriculture
25–6
‘turning point’, and Lewis Model 7,
8, 294
404 Index
![Page 414: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/414.jpg)
underemployment 11 n17
and agriculture 12, 16, 185–8,
204–7, 214, 247–9, 273
partible inheritance 205–6
and definition problems 250
and informal urban sector 15
and labour surplus 204–7, 249–50
and migration 214, 249
unemployment:
and efficiency wage rates 187
and migration 271–4
United Kingdom:
and Anglo-German rivalry 263–4,
346
response to German foreign
policy 348, 349
and farm size 48
and immigration 80
and land ownership
distribution 49, 51
and political development of 46,
331
and urbanization 59–62
United States:
and agriculture 216
exports 217, 219–20
and city growth rates 62
and emigration to 81, 83–4, 321
agricultural employment 88–9
decline in 85
influence of agriculture 81–2
previous occupations 86–8
wage ratio 85
and income distribution
313–14
urbanization:
and contemporary views of 188–9
and death rates 157–8
and electoral politics 335–6
and impact of agricultural
decline 217–18
and industrialization 15
and inequality 316, 323, 326
and international comparison of
rates of 59–62
and life expectancy 157–8, 283
and migration 59–62, 111–12
and population growth 151–6,
173–7
and Social Democratic Party
vote 335–6
and social dislocation 59, 351
and speed of 59, 61, 62, 151, 293,
353
Van Zanden, T 233–4
vocational training:
and income distribution 311, 321
and migration 311–12
Voelcker, Augustus 234–5
voting behaviour, see electoral politics
wages:
and advantages of cash
payments 182
and agricultural:
agricultural/industrial
ratio 105–6
influences on 201–4, 207–8,
211–13
rising level of 196–7
and Anglo-German
comparison 263–4
and death rates 154–5, 170
and decline in rural 302
and efficiency wage rates 187
and fertility rates 146, 147
and German/American ratio 85
and labour supply 6, 7
and marriage rates 178–80
and migration 56, 94, 111
impact of 190
wage convergence 274–9
wage dispersion 276
women 281–3
and payments-in-kind 182–3
Index 405
![Page 415: Migration and Inequality in Germany 1870-1913 (Oxford Historical Monographs)](https://reader033.vdocuments.us/reader033/viewer/2022051006/57ade2de1a28abbe3a9aec03/html5/thumbnails/415.jpg)
wages: (Continued)
and population growth 175
and productivity 7, 9–11
and proximity of Russian
Poland 202–3, 204, 208, 212,
279, 320–1
and regional variations 275–6,
278–9
and rural-urban gap 108
and trade unions 321
Wagner, Adolf 189, 190, 298
Wagner, Richard 297
Weber, Adna 58
Weber, Adolf 188–9
Weber, Max 11–12, 17, 107, 187,
191, 192, 207
and eastern agriculture 193–7
capitalist organization 195
children of labourers 195
impact of migration 196
labour relationships 194–5
large estates 193, 194
political concerns 193–4
and labour surplus 293
and labour-saving technology 247
and Prussian land reform 34
Wehler, Hans-Ulrich 2 n3, 331
and German exceptionalism 19–20
Westphalia:
and agrarian reform 41
and peasant landholdings 38
Williamson, J G 264, 296, 313
women:
and illegitimate children 142
and migration 76–9, 138–9
life expectancy 283
wages 281–3
working class 154, 156, 174
World Bank 4, 307
Wurttemberg, and agrarian reform 44
Zollverein, and creation of 20
406 Index