robin c. matthews - why has britain had full employment since the war?

16
THE ECONOMIC JOURNAL SEPTEMBER 1968 WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR?i IF one were asked to pick out the most important single respect in which the experience of the British economy in the post-war period has differed from that of earlier times, most people would agree in saying " full employ- ment of labour." There are other notable distinguishing features of the post-war period—chronically rising prices; an abnormally high ratio of domestic investment to income by historical standards; rapid growth of income per head—but full employment is the most striking and most pervasive in its effects. I am not here concerned with the question whether employment has been full in some absolute sense, or over-full, or rather less than full. The question I shall discuss is why employment has been so much more full, or more nearly full, than it was in the past. That it has been is not open to doubt, even when all due allowance is made for imperfections in the statistics. The unemployment percentage has averaged about 1-8% in the post-war period; in the inter-war period, if the figures are adjusted to a comparable basis, it averaged about 10^%; before 1914 the statistics show an average of about 4f %—not nearly so high as in the inter-war period, but still substantially above what we have had since the Second World War.^ When the question is asked, why have we had full employment since the war? most people tend to reply, without thinking very much, that it is because we have had a full-employment policy—we have had the Keynesian • A revised version of a University of London Special Lecture delivered at University College London on February 1, 1968. I am much indebted to Moses Abramovitz for helpful criticisms of the original text; to G. H. Feinstein and J. C. Odling-Smee, my collaborators on a forthcoming study on Growth in the British Economy, sponsored by the Social Science Research Gouncil (New York), for access to unpublished material; and to Mrs. P. Yudkin for research assistance. ^ The official unemployment percentages for the inter-war period need to be adjusted downwards to make them comparable with post-war figures on account of differences in coverage. The reli- ability and comparability of the pre-1914 figures are doubtful, as is well known, but when account is taken of offsetting biases it is unclear whether they need to be adjusted upwards or downwards. The evidence is surveyed by W. Galenson and A. Zellner, " International Gomparison of Unemploy- ment Rates," Appendix J, in Universities—^National Bureau, The Behaviour and Measurement of Unemptoyment (1957). No. 311—VOL. LXXVIII. NN

Upload: kmbence83

Post on 16-Nov-2014

317 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

THE ECONOMIC JOURNALSEPTEMBER 1968

WHY HAS BRITAIN HAD FULL EMPLOYMENTSINCE THE WAR?i

IF one were asked to pick out the most important single respect in whichthe experience of the British economy in the post-war period has differedfrom that of earlier times, most people would agree in saying " full employ-ment of labour." There are other notable distinguishing features of thepost-war period—chronically rising prices; an abnormally high ratio ofdomestic investment to income by historical standards; rapid growth ofincome per head—but full employment is the most striking and mostpervasive in its effects.

I am not here concerned with the question whether employment hasbeen full in some absolute sense, or over-full, or rather less than full. Thequestion I shall discuss is why employment has been so much more full, ormore nearly full, than it was in the past.

That it has been is not open to doubt, even when all due allowance ismade for imperfections in the statistics. The unemployment percentage hasaveraged about 1-8% in the post-war period; in the inter-war period, ifthe figures are adjusted to a comparable basis, it averaged about 10^%;before 1914 the statistics show an average of about 4f %—not nearly so highas in the inter-war period, but still substantially above what we have hadsince the Second World War.^

When the question is asked, why have we had full employment since thewar? most people tend to reply, without thinking very much, that it isbecause we have had a full-employment policy—we have had the Keynesian

• A revised version of a University of London Special Lecture delivered at University CollegeLondon on February 1, 1968. I am much indebted to Moses Abramovitz for helpful criticisms ofthe original text; to G. H. Feinstein and J. C. Odling-Smee, my collaborators on a forthcoming studyon Growth in the British Economy, sponsored by the Social Science Research Gouncil (New York),for access to unpublished material; and to Mrs. P. Yudkin for research assistance.

^ The official unemployment percentages for the inter-war period need to be adjusted downwardsto make them comparable with post-war figures on account of differences in coverage. The reli-ability and comparability of the pre-1914 figures are doubtful, as is well known, but when accountis taken of offsetting biases it is unclear whether they need to be adjusted upwards or downwards.The evidence is surveyed by W. Galenson and A. Zellner, " International Gomparison of Unemploy-ment Rates," Appendix J, in Universities—^National Bureau, The Behaviour and Measurement ofUnemptoyment (1957).

N o . 311—VOL. LXXVIII. NN

Page 2: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

556 THE ECONOMIC JOURNAL [SEPT.

revolution. Now supposing this were the right answer, it would be aremarkable thing. It would mean that the most important single featureof the post-war British economy has been due to an advance in economictheory. It would be a most striking vindication of Keynes' celebrateddictum about the ultimate primacy of abstract thought in the world ofaffairs.

However, this interpretation of events, at least in its simple form, isopen to serious objections.

The figures just quoted suggest that the inter-war unemployment problemwas of a different order from the pre-1914 unemployment problem. I shalltherefore deal separately with the comparison of the post-war period withthe inter-war period and the comparison of the post-war period with thelonger sweep of British economic development. I shall put forward twopropositions. The first is that as compared with the inter-war period therehas certainly been an increase in effective demand, in the Keynesian sense,relative to supply; but that it is non-proven that this has been due to govern-ment policy, at least in any simple sense. The second is that the declinein unemployment as compared with before 1914 is to a large extent not aKeynesian phenomenon at all.

First, then, the comparison with the inter-war period. Throughout thewhole ofthe inter-war period after 1920 there was clearly Keynesian demanddeficiency. Certain bottlenecks were encountered in the late 1930s, chieflyin the industries affected by rearmament, but not to such an extent as toinvalidate the proposition that a substantial increase in output could havebeen brought about by a general increase in demand.

Why did this deficiency of demand disappear in the post-war period ?In particular, was it due to government action ? The question at issue hereis not whether the Government was or was not pledged in the post-war periodto the policy of maintaining full employment or something near to it—ofcourse it was. The question is whether the high level of demand thatactually occurred was due to government action or whether it was due toother forces, as a result of which government action was not needed.

The hypothesis that the change was due to fiscal policy is open to a simplebasic objection. This is that throughout the post-war period the Govern-ment, so far from injecting demand into the system, has persistently had alarge current account surplus. This surplus has varied in amount; butgovernment saving has averaged about 3 % of the national income. This ismuch larger than at any previous time in the twentieth century. Fiscalpolicy as such therefore appears on the face of it to have been deflationaryin the post-war period, quite strongly deflationary in fact, rather than thereverse.

This is not to deny that in various post-war years fiscal policy has beenadjusted with the object of increasing demand above its existing level. Butin the years when this was done it was a matter of reducing the size of the

Page 3: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 557

surplus rather than turning it into a deficit. The overall effect has thereforebeen one of restraint.

There are a couple of ways in which the argument can be refined in anattempt to avoid the prima facie conclusion that fiscal policy was in netmore deflationary than in earlier periods.

In the first place, compared with before the war, there has been anincrease in the levels of both government expenditure and revenue relativeto national income, as well as a rise in revenue relative to expenditure.Therefore there may have been some inflationary effect of the type analysedin the theory of the balanced budget multiplier: that is to say, the extrataxation may have been partly at the expense of private saving. It isdifficult to estimate the magnitude of this effect, because our understandingof what determines private saving is very inadequate. As an example onecan assume that private saving varies proportionately with private disposableincome, other things equal. One can estimate the net effect of fiscal policy,including both the effect of the balanced budget multiplier and that of thebudget surplus, by asking what would have been the effect on the level ofdemand in 1937 (the last pre-war cyclical peak) if government expenditure,direct taxes and indirect taxes in 1937 had all borne the same ratio to nationalincome as they did in 1964 (the last post-war cyclical peak), on the assump-tion that exports, investment and the propensities to save and to importhad remained unchanged. It comes out from this calculation that incomein 1937 would in that case have been 1 % lower than it actually was. Thatis to say, the balanced-budget multiplier effect is less than sufficient to out-weigh the budget surplus effect. Alternative assumptions are possible, somepushing the result one way, some the other; but it does not seem possible,even on the most favourable assumptions, to get more than a mild overallnet stimulus—certainly not enough to explain the magnitude of the increasein activity that actually occurred. So while the balanced budget multiplierpoint has to be acknowledged as valid in principle, it does not appear toalter the argument qualitatively. More exact research on this must awaita better understanding of the saving function.

The second possible refinement ofthe argument is a point that has comeup in recent discussions of fiscal policy in the United States.^ Granted thatthere was a budget deficit in 1937 and a budget surplus in all post-war years,this is to some extent the result of the differences in the level of activity; fora high level of activity raises tax revenues. It could conceivably be arguedthat the level of government expenditure and tax rates prevailing in 1937would at a full-employment level of activity have produced an even largersurplus than there was in the post-war period, and that fiscal changes weretherefore in net inflationary. This is scarcely tenable statistically. Butsupposing it were true—or that it were true when taken in conjunction

* A very explicit earlier discussion is in N. Kaldor's Appendix G to W. H. Beveridge, FullEmployment in a Free Society (1944).

Page 4: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

558 THE ECONOMIC JOURNAL [SEPT.

vfith the effects of the balanced budget multiplier—it could not be takenas the main explanation of the increase in activity without assuming anunstable structure of demand. If a lowering of the government savingschedule, which is what is in effect implied, leads to a new equilibrium inwhich government saving is actually higher as a proportion of nationalincome, it follows that the rise in national income brought about by thechange in government policy must have raised investment by more than itraised private saving. This is not impossible. But it means that fiscalpolicy is of a pump priming character, setting in train a potentially unstableupward movement, rather than a direct and continuing support to demand.It is therefore rather different from the straightforward notion of the Govern-ment maintaining full employment through fiscal policy. The question ofpump priming will be reverted to presently.

It seems, therefore, that these refinements in the argument about fiscalpolicy, though valid in themselves, are not sufficient to upset the conclusiondrawn from the simple fact of the observed budget surplus, that fiscal policyhas not in any direct way been the explanation of the increase in activitycompared with before the war.

In order to analyse properly what the reasons have been, it would benecessary to have a full-scale econometric model of the various elements ofdemand, a model that could embrace both the inter-war period and the post-war period. We do not have such a model of the post-war period alone,far less one which covers the inter-war period as well. But inspection of thecomponents of demand in elementary Keynesian terms may help to identifythe elements in the situation.

If we include government expenditure as a form of consumption we canwrite the familiar identity

Y = I+C + X-MWriting as usual s for the ratio of saving (including government saving) toincome and m for the ratio of imports to income, we have

Ys -{- m

For completeness, account has also to be taken of income from abroad,which though not part of G.D.P. is a source of disposable income and there-fore affects the level of consumption demand. Taking 7 to mean G.D.P.,but measuring m and s as the ratios of imports and savings to G.N.P., wethen have (writing A for income from abroad)

s + m

We are interested for the present purpose not in 7 but in the ratio of

Page 5: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 559

Y to potential full-employment income, F. Dividing through by F, wehave

The relative importance of the proximate causes of the changes incompared with pre-war can then be ascertained by inspecting the move-ments of these five items. In this way we consider changes in consumptionand imports in terms of propensities, and we consider changes in the otherthree items, which are less directly determined by current income, in termsof their relation to full-employment income.

Components of Expenditure at Inter-war and Post-war Cyclical Peaks in theUnited Kingdom {Percentages)

(2)

fA{\ -s - m)

f(5)

I + X + A{\ - s -m)f

(6) (7) (8)

19291937

195519601964

9191

1009999

8810-3

21-314-1

15-6 21-817-2 20 0186 18-5

3-03-6

0-60608

33-127-9

38-037-8380

11-711-6

24-618-7

36-430-4

15-6 22-5 38-017-7 20-6 38-218-9 19-5 38-4

For definitions, see text and footnote below. Components do not necessarily add to totalsbecause of rounding.

Sources: For 1955, I960, 1964, National Imome and Expenditure 1966; for 1929, 1937, Londonand Cambridge Economic Service, The British Economy, Key Statistics 1900-66, Table A, adjustedto exclude property income from exports and imports of goods and services.

Columns (2), (3), (4), (6), and (7) in the table show the values of theseratios in two inter-war cyclical peak years and in three post-war cyclicalpeak years.^ In the post-war years all five of the ratios are higher than in1937, except that relating to income from abroad. The rise in investmentand exports relatively to full-employment G.D.P. serve to raise the level ofactivity; the increase in the saving and import propensities and the fall inincome from abroad tend in the opposite direction. The positive changesare larger than the negative ones, so F/F duly rises.

The investment item is the one which shows the biggest change comparedwith inter-war. This is true, even when the post-war period is compared

' All figures relate to values at current market prices. The figures given for YIV are roughapproximations; the exact values chosen do not affect the argument in the text. X is defined asexports of goods and services. / is measured as gross domestic capital formation plus half stock-building, and imports are measured as imports of goods and services minus half stockbuilding.Stockbuilding is allotted in this way because it is estimated that about half of stockbuilding in post-war years has consisted of stockbuilding of imports (cf. " Short-Term Economic Forecasting in theUnited Kingdom," Economic Trends, August 1964, p. 9). The simpler procedure of including allstockbuilding in investment would give exaggerated figures for m and I, because of the well-knowntendency of stockbuilding and imports to be abnormally high in cyclical peak years.

Page 6: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

560 THE ECONOMIC JOURNAL [SEPT.

with 1937, when investment was high by inter-war standards. It is evenmore marked if the post-war period is compared with 1929.

As compared with 1929, changes on the overseas side have not contributedsignificantly to the high level of activity in the post-war period. This holdswhether we look merely at the effect of the changes in the export item or atthe net effect of the changes in all three of the balance-of-payments items inthe table. As compared with 1937, however, changes on the foreign-tradeside have made a significant contribution, as might be expected. Onewould expect the recovery of exports from their depressed level in the 1930sto have been of some importance in raising the level of activity. A signifi-cant positive net contribution from the foreign-trade side to the change inthe level of activity, compared with 1937, remains even when account istaken of the deflationary effects of the fall in income from abroad and therise in the propensity to import (which, it will be remembered, is at currentprices and is defined to include services as well as goods). But this contribu-tion is smaller than the contribution arising from the investment-saving side.

It is thus clear that, comparing the post-war period with the inter-warperiod as a whole, the explanation of the rise in investment must lie at theheart ofthe explanation ofthe rise in the level of activity.

What have been the reasons for the high level of investment? Thissubject is a relatively unresearched one. Economists have been so pre-occupied with explaining why investment in this country has been lowcompared with other countries that they have devoted little attention toconsidering why investment has been so much higher relatively to nationalincome than it has ever been before in this country.

It is legitimate here to focus attention on private investment. Publicinvestment has certainly played its part. But investment in the publicsector has been on average a smaller proportion of total investment in thepost-war period than it was in the inter-war period, if we define the publicsector, as we obviously must do, to correct for the effects of nationalisation.To put it more exactly, investment in those industries that fall within thepublic sector now has been a smaller proportion of total investment thaninvestment in those industries was before the war. (This statement relatesto investment excluding dwellings, which cannot be assigned wholly toeither the public or the private sector, and which in any case has been amuch smaller proportion of total investment since the war than it was in the1930s.) We may therefore reasonably concentrate attention on investment inthe private sector, while not forgetting that public investment has also risen.

Two broad types of view can be put forward to explain the high levelof private investment in the post-war period. The first is that the marginalefficiency of investment schedule—the investment demand function—hasbeen abnormally high, because of such reasons as the pressure of demandon capacity and the scope for innovatory investment. The second is thatthe investment supply function—the willingness to do investment at any

Page 7: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 561

given expected rate of return—has been subject to some upward shift com-pared with before the war, because of a change in entrepreneurial attitudes,itself possibly due to government policy.

On the first hypothesis the high post-war investment has essentially thenature of a gigantic cyclical boom. By this I do not mean that it willnecessarily lead to a slump, but that the forces sustaining it have not beendifferent in essence from those of traditional cyclical booms. The economywas given a once-for-all hoist upwards during the war. When the warended income was therefore high. Moreover, the capital stock was low asa result of the low level of investment during the war itself and also duringthe inter-war period. Hence there was great scope for investment. Thefeed-back from this investment combined with the favourable effects of theforeign-trade situation combined to keep income up and keep up the demandfor more investment.

This view of what happened is related to the pump-priming idea men-tioned earlier. It says that government finance did play a role, in thatduring the Second World War itself it brought about a once-for-all increasein the level of activity, and that once this had been done cumulative forceswere set in train that kept investment high, so that the Government couldthen assume a passive or even restraining role without activity fallingseriously below the full-employment level.

The following question now naturally arises. Granted that the forcesencouraging investment in the post-war period did to some extent resemblethose of past cyclical booms, why did the resulting investment boom lastso much longer than cyclical investment booms have normally lasted?Could this have happened if there had not been some change as well inentrepreneurial attitudes towards investment?

It is not possible to give a simple or definitive answer to this. But it isarguable that the stage was set for an investment boom of unusual propor-tions, because of the amount of investment opportunities that had accumu-lated as a result of historical circumstances. There had not been a singlereal boom in domestic investment in the whole of the twentieth century upto the end of the Second World War. Investment was low during the SecondWorld War; it was low in the inter-war period, apart from house-building,because of the slump; it was low during the First World War; and even theperiod immediately before the First World War constituted a downwardphase of one of the long swings in home investment which at that timecharacterised the British economy. In the forty years preceding 1948 thedomestic stock of fixed capital is estimated to have risen at a rate of scarcelymore than 1 % per annum.^ In this way very substantial arrears of invest-ment opportunities were to be expected.

An interesting question concerns the role of technical progress in all this.It is probable that technical progress has been more rapid in the post-waj

1 Key Statistics, Table I.

Page 8: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

562 THE ECONOMIC JOURNAL [SEPT.

period than it used to be; this appears to be a world-wide phenomenon.This has helped maintain investment opportunities and so keep up the levelof demand. This acceleration of technical progress can be regarded on oneview as a consequence of the boom. On this reckoning it is part of the feed-back mechanism, part of the cumulative process. Alternatively, it can beregarded as independent in its origins, in which case it ranks as a separatefactor making for high investment and activity. To what extent it is due togovernment policy—British Government policy—is debatable. In so far asit is (probably only to a minor degree), it is due mainly to government policyin the technological and educational sphere rather than to the activities ofthe branches of government concerned with the management of demand.

Arguing along these lines, one cannot rule out of court the hypothesisthat the post-war investment boom was the result of a favourable conjunctionof objective circumstances affecting the marginal efficiency of investmentrather than the result of a shift in the investment supply function.

Let us now consider the hypothesis that there was an upward shift in theinvestment supply function. This provides more scope for attributing theresult directly to government. The possibilities here include, first, thatmonetary policy has operated so as to make finance relatively cheaper;secondly, that tax policy has encouraged investment; thirdly, that the beliefon the part of entrepreneurs that the Government would not permit a slumphas improved confidence—a safety net theory of the role of government,implying that although the acrobat has not actually fallen, his knowledgeof the presence of the safety net has given him confidence and improved hisperformance.

The general way in which such factors would operate if they did wouldbe to make businesses willing to do investment with the prospect of a lowerpre-tax rate of profit than they would otherwise have required or than theydid require before the war. Has this been the case? The realised rate ofprofit on all capital appears to have been lower than it was before the war,at least if measured net of depreciation.^ This movement of the profit ratemay be taken as a support ofthe hypothesis that the investment supply func-tion has shifted, for why otherwise would businesses have been willing to doso much investment in face of the relatively low rate of profit ? But theprofit rate on all capital is not necessarily a good guide to the expected profitrate required on new investment. This is the concept that we should inprinciple like to be able to identify. The impression one has from qualita-tive indicators of business attitudes is that it has not been particularly lowin the post-war period.^

* C. H. Feinstein, " Evolution ofthe Distribution ofthe National Income ofthe United King-dom since 1860," paper read to International Economic Association Conference on the Distribu-tion of National Income, Palermo, September 1964.

" Sir Robert Shone, " Planning for Economic Growth in a Mixed Economy," ECONOMIC JotmNAL,March 1965, p. 16, suggests that the rate of return on new investment usually aimed at, after depre-ciation but before tax, was around 20% in the early post-war years and around 15% more recently.

Page 9: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 563

This same conclusion, that required profit rates have not been particularlylow, can be derived from looking at earnings yields on ordinary shares, whichon certain assumptions can be taken to measure the real rate of return oninvestment which has to be achieved if the investment is not to prejudice theinterests of a company's existing shareholders. On the average, earningsyields have been higher in the post-war period than formerly, not lower. ̂They do not therefore suggest that the target rate of return on investmenthas iaeen particularly low.

At this point it is necessary for a moment to give up the simplificationadopted so far of treating the post-war period as a whole. It is fairly clearthat witkin the post-war period there have been changes tending to shift theinvestment supply function upwards. Otherwise we can hardly explainwhy the ratio of private investment to G.N.P. has not merely been high byhistorical standards but also has been rising quite steeply within the post-war period, at least until fairly recently. This is the opposite of what mighthave been expected in a period starting with once-for-all arrears of invest-ment opportunities. It has happened notwithstanding a downward trendof pre-tax profit rates (a trend which is what might have been expected toresult from gradual catching up with arrears). The most important causetending to raise the investment supply function within the post-war periodhas probably been the increasingly generous tax treatment of investment,culminating in the payment of actual investment grants.^ This has reducedthe level of the real rate of return that is needed in order to yield a givenprivate post-tax rate of return. This must have served to prolong the invest-ment boom. What is not so clear is whether taking the average of the post-war period as a whole, factors of this kind have lowered the required rateof return on investment compared with that prevailing before the war.It is in this sense that we cannot rule out the first of the two broad hypo-theses about investment, that its high post-war level has not been due to anyupward shift in the investment supply function.

To summarise the argument so far. The increase in the level of demandcompared with before the war has not been due to deficit financing. It hasbeen partly due to world developments, as reflected in our export markets.But the main cause has been the unprecedently high level of investment. Itis possible that there has been some upward shift in the investment supplyfunction due in one way or another to government policy, as compared withbefore the war. But it is unclear just how big this has been, if it has occurredat all. The alternative hypothesis is also perfectly tenable, that the maincause of the high investment has not been an upward shift in the investmentsupply function but a conjunction of circumstances similar to those thathave caused past booms of a cyclical character.

• Key Statistics, Table M.' This is convincingly argued byj . R. Sargent, " Recent Growth Experience in "me Economy

of the United Kingdom," ECONOMIC JOURNAL, March 1968, pp. 19-42, where the profit rate figuresare quoted.

Page 10: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

564 THE ECONOMIC JOURNAL [SEPT.

Before drawing any general conclusions, let me go on to the second topicof this lecture, the consideration of trends over a longer period going backto before 1914. The points involved here are rather different in kind.

Before 1914 unemployment was on average lower than in the inter-warperiod, but it was still substantially higher than we have become accustomedto since the Second World War. Unemployment was most severe in cyclicaldepressions, but it was not confined to them. " Distress from want ofemployment, though periodically aggravated by depressions of trade, is aconstant feature of industry and commerce as at present administered," saidthe Minority Report of the Poor Law Commission in 1909, summarising theevidence they had received.^ Does this mean that as compared with thattime there has been an increase in effective demand relative to supply in theKeynesian sense ?

Defective demand in the Keynesian sense is a situation in which anincrease in demand would lead to an increase in final output. It is not.sufficient that there should be an excess supply of labour; there must bean excess supply of all relevant factors if output is to be capable of respondingto a change in demand. At most cyclical peaks before 1914 it does not seemthat this was so. In general, the evidence from such sources as The Econo-mist's Annual Commercial Histories is that capital capacity at pre-1914cyclical peaks was pretty fully stretched in the major industries. It followsthat there cannot have been much demand deficiency for the final product.Naturally the situation was different in depressions, and then there norm-ally was excess capacity. However, the amplitude of cyclical fluctuationsin output relative to trend appears to have been only moderately greaterbefore 1914 than it has been since the Second World War—a fact whichis perhaps not generally appreciated. Since the Second World War theaverage shortfall of real G.D.P. in cyclical trough years below its peak-to-peak trend has been nearly 4%; between 1870 and 1914 it was 5%.^ So ifthere was not demand deficiency at cyclical peaks before 1914, it cannothave been all that severe relative to our post-war experience over the averageof the whole cycle.

This is the first reason for supposing that the unemployment of the pre-1914 period cannot be viewed in Keynesian terms as the result ofa deficiencyof demand for final product.

The second reason is the evidence we have about the nature of unemploy-ment before 1914. When one reads the accounts of unemployment in thosetimes one is irresistibly reminded of the characteristics of urban unemploy-ment in underdeveloped countries at the present time. There is the samecontrast between the more fortunate members of the working classes, who

^ The Public Organisation of the Labour Market: being Part Two of the Minority Report of the PoorLaw Commission (1909), edited by S. and B. Webb, p. 173. This conclusion was not a matter ofdisagreement between the Majority and Minority Reports.

^ R. C. O. Matthews, " Post-War Business Cycles in the United Kingdom," in Is the BusinessCyrle Obsolete?, edited by M. Bronfenbrenner (forthcoming).

Page 11: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 565

have more or less permanent jobs, and the body of unskilled casual labourers,who rarely succeeded in getting a full week's work even in periods of generalprosperity. The parallel to the situation in underdeveloped countries isfurther illustrated by the regional incidence of unemployment before 1914.Regional differences in unemployment were scarcely less marked then thanthey were in the inter-war period and have been since. But the geographicalpattern was almost exactly the reverse. Unemployment was low in theindustrial north, and about twice the national average in London, whereindustrial development of a modern kind had as yet made relatively littleimpact.^ There were similar regional disparities to the disadvantage of thesouth in wage levels.^ The only exception to this geographical transformationin the relative amount of unemployment between regions across the FirstWorld War is Ireland, the one major part of the United Kingdom that re-tains characteristics of an underdeveloped country to the present time. InIreland unemployment was above the United Kingdom average before1914, just as it has been ever since.

What is being suggested can be put into schematic theoretical terms asfollows. Unemployment of a factor can be caused either by defective de-mand for the final product or, if the scope for varying factor proportions islimited, by defective supply of a co-operating factor. The characteristic ofan underdeveloped country is that the supply of the co-operating factor capi-tal is inadequate for jobs to be available for all the labour. There is there-fore the phenomenon of a dual economy, with some workers engaged inrelatively capital-intensive industries such as manufacturing or railways, orelse in privileged occupations such as government service, while the rest ofthe labour force remains outside the circle in low-productivity employmentor else without any employment at all. This state of affairs had not entirelyceased to hold in Britain in 1914. The unskilled labour that jostled forjobs at the docks, on the building sites and in many other trades was theremnant of the chronic labour surplus associated with incomplete develop-ment. An increase in demand for final product at cyclical peaks would nothave permitted the absorption of this labour surplus unless it had been ac-companied by a switch in the structure of demand towards more labour-intensive goods and services.

Now it is to be expected that unemployment of this sort should diminishas economic development proceeds. Capital is accumulating, and, moreover,there is also an accumulation of human capital in the form of education whichwidens the range of occupations where people can be usefully employed with-out the assistance of much physical capital. Within the nineteenth centurythere was duly a reduction in this kind of unemployment in this country.

' W. H. Beveridge, Full Employment in a Free Society (1944), pp. 73-A.« A. L. Bowley, Prices and Wages in the United Kingdom 1914-20 {192\), pp. 170-1; E. H. Hunt,

" Labour Productivity in English Agriculture, 1850-1914," Economic History Review, 1967, pp. 280-92.

Page 12: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

566 THE ECONOMIC JOURNAL [SEPT.

This was noted by contemporaries, for example, Marshall.^ Marshall con-trasted the problem of industrial unemployment with the unemploymentand underemployment typical of pre-industrial societies. He quoted as anexample of the latter the conditions of artisans as he had observed themduring his celebrated stay in Palermo. " Scarcely anybody was ever thrownout of regular employment, because scarcely anybody ever was in it." ^ Hepointed out that the development of industry in this country had progres-sively lessened the proportion of the labour force in this situation.^

But although the underdeveloped country type of unemployment haddiminished by the end of the nineteenth century, it had not disappeared.One reason for this was that the outflow of labour from agriculture continued.The rural surplus labour moved to the towns, and part of it became urbansurplus labour. The urban surplus, especially in London, was in this waycontinuously replenished. This partly explains why, despite the favourableunderlying tendency, there is not any downward trend in the recordedunemployment percentage in the sixty years or so before 1914. Unemploy-ment and underemployment were diminishing, but the unemployment wasbeing increasingly transferred away from agriculture, where it was notmeasured, to the towns, where it was measured to a greater extent. Thishas, of course, been characteristic of many countries in the process ofdevelopment.

By the First World War the absorption of the rural surplus had largelyrun its course. Since then, the agricutturat labour force has continued todecline, but the quantitative importance of the flow relatively to the totallabour force has been much less than in the nineteenth century. Thestage might therefore seem to have been prepared for a more pronouncedmovement towards labour scarcity. But the long-run tendency towardsincreasingly full absorption of the labour force and increasing scarcity oflabour relatively to capital as development advances was entirely overborneand concealed in the inter-war period by Keynesian demand deflciency ona large scale. This was due to the peculiar circumstances of the inter-warperiod, in particular the structural shift to Britain's disadvantage in thepattern of world trade. The potentially beneficial effects on the demand forlabour of an increase in the supply of the co-operating factor capital wastherefore outweighed by a deficiency of demand for all the factors of produc-tion.

1 A. Marshall, Principles of Economics (8th edn.), pp. 687-8, and Official Papers, pp. 92-101." Official Papers, p . 93.° Writing as he was in the cyclical depression of the 1880s, Marshall was cagey about making

a specific assertion that unemployment had recently declined. He spoke rather of strong forcestending in that direction. But he does in the end commit himself to the effect that unemploymenthas declined in the following characteristically oblique sentence. " All these considerations sup-port the belief that if the average steadiness of employment throughout the country had remainedstationary, changing neither for better nor for worse, the unemployed benefit of the trade unionswould have increased faster than it has done " {Official Papers, p. 97).

Page 13: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 567

As a result, when overall demand deficiency was corrected after theSecond World War, the effect felt in the labour market was particularlysharp. The trend towards increasing labour scarcity had been concealedby the special circumstances of the inter-war period. After the war theeconomy experienced for the first time a situation in which there was neitherdemand deficiency ofthe inter-war type nor the characteristics of incompletedevelopment present still before 1914.

It may be noted here in parentheses that the model of the dual economywith infiexible coefficients of production, combined with the notion of em-bodied technical progress, implies that when the supply of capital catchesup with the supply of labour, so that the labour surplus is eliminated, scrap-ping of old equipment will have to proceed more rapidly than before, so asto prevent the capital stock from outrunning the labour supply. We haveno direct data on scrapping, but the indicators do seem to be consistent withthis idea and to suggest that the service life of capital is shorter now than itused to be in the nineteenth century.^

One particular and important way in which the long-run increase inlabour scarcity has manifested itself is in the response of employers to cyclicalfluctuations in demand. As stated earlier, the amplitude of cyclical fluctua-tions in output in the post-war period has not been much greater than it wasbefore 1914. But this is not true of fluctuations in employment. Fluctua-tions in employment have been of far smaller amplitude than they werebefore 1914. In the post-war period fluctuations in employment have beenonly about one-third of the amplitude of fluctuations in output. This is anew phenomenon. Before 1914, and also in the inter-war period, theamplitudes ofthe two were about the same.^ But although it is in this sensea new phenomenon, it can be regarded as a continuation of a longer-termtrend: a trend away from the situation where the employer hires labourcasually as he needs it towards a situation where an employer hires labouron a more or less permanent basis. The unemployment problem ofthe latenineteenth century manifested itself very largely in the excess supply ofcasual labour—the surplus labour crowded into those occupations where therewas some chance of work some times. However, the proportion of industriesin which regular employment was offered progressively increased. Thiswas recognised as an important phenomenon at the time. It was quoted byMarshall as a result of industrial development and as one of the chief causesmaking for more regular employment. " In many directions there is asteady increase in the proportion of employees who are practically hired bythe year." ^ The tendency exhibited in the post-war period for employers

* Thus the rates of depreciation on plant and equipment allowed by the Inland Revenue im-plied a longer life before the First World War than they did in the inter-war period and have donesubsequently. C. H. Feinstein, Domestic Capital Formation in the United Kingdom 1920-1938 (1965),p. 22.

" Matthews, " Post-War Business Cycles in the United Kingdom," cit.» Principles, p . 688.

Page 14: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

568 THE ECONOMIC JOURNAL [SEPT.

to keep labour on through cyclical recessions can be regarded as a furtherextension of the same process. From having been hired by the hour, labourpassed on to being hired by the week and then on to a situation in which itwas hired more or less permanently. When the labour supply is veryabundant, employers lack inducement to move away from a system of casualhiring. As labour becomes more scarce relative to capital, the costs of hir-ing and firing increase, and the employment contract becomes de facto alonger one. In this way the non-Keynesian force—the trend increase in thesupply of the co-operating factor capital—has lessened the sensitivity ofemployment of labour to Keynesian fluctuations in expenditure.^

I will now sum up. Part of the reason for low unemployment in the post-war period has been the trend increase in the scarcity of labour relative tocapital. This is non-Keynesian, it can confidently be expected to persist,and it has little or nothing to do with government policy. Not only has itbeen important in itself but also it has provided a measure of protection tolabour from the eiTects of fluctuations in the demand for final product. Inaddition, there has been a major change compared with the abnormallydepressed inter-war period in the level of effective demand relative to supply.It is unclear how much of this has been due to government policy. It wascertainly not directly due to fiscal policy except in so far as war-time fiscalpolicy started the post-war period off at a high level of activity. It is possiblethat government encouragement to investment has prolonged the boom, butthere were in any case powerful forces independent of government actionmaking for an investment boom of unusual proportions.

The argument presented in this lecture has, of course, been greatly over-simplified in a number of respects. For a proper analysis, much more would,in particular, need to be said about trends within the post-war period, andalso about fiuctuations in the post-war period. Moreover, there is one finalpoint. In playing down the Government's role, I do not at all mean todeny that a different policy on behalf of the Government could have pre-vented there being full employment. The underlying balance-of-paymentssituation in the post-war period was more difficult than in the inter-warperiod in a number of respects, notably because of the loss of propertyincome from abroad. If the Government had tried to deal with this ex-clusively by deflation we should not have had such a high level of activity.As it was, the Government protected the balance of payments by a numberof other measures, the most important of which has been the maintenancethroughout the post-war period of fairly strict control over capital exports.In the absence of such measures, and given the tendency of demand to behigh for other reasons, the budget surplus would have had to have exceeded

^ It is not suggested that this is the sole reason for the reduction in the short-period output-elasticity of employment. Other forces have probably also played a part, such as the increase inthe ratio of salary-earners to wage-earners and the increase generally in the number of workersengaged on overhead functions relatively to those engaged directly in production.

Page 15: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?

1968] WHY HAS BRITAIN HAD FULL EMPLOYMENT SINCE THE WAR? 569

what it was in the past by an even greater extent than it actually did, or elsemonetary policy would have had to be still more restrictive, or both. Inthis way the tendency for demand to be high for reasons independent ofgovernment action would have been checked by government action. Thatthis was not done is something for which economists and the Keynesianrevolution can take some credit.

R. C. O. MATTHEWS

All Souls College,Oxford.

Page 16: Robin C. Matthews - Why Has Britain Had Full Employment Since the War?