a spatial interpretation of the

8/8/2019 A Spatial Interpretation of The

1/12

ABSTRACT. In this paper the density dependence model,which was developed in organizational ecology, is com-pared to the economic-geographical notion of agglomerationeconomies. There is a basic resemblance: both involve someform of positive feedback between size of the population andgrowth. The paper explores how the theoretical concepts

compare to each other, and if an interdisciplinary cross-fertilization between both is fruitful. It is found that there area number of important similarities in the underlying theories.These refer to the process of legitimation, which has someclose similarities to concepts derived from theories of newindustrial districts, such as social capital, institutional thick-ness, and innovative milieux. Differences remain importantas well. For instance, the sociological interpretation of com-petition is not transferable into notions of agglomerationeconomies. An important conclusion is that agglomerationeffects can and should be incorporated into the density depen-dence model.

1. Introduction

Industrial demography, or demography of the firm,is concerned with the analysis of demographicprocesses of entry, exit, and firm growth in indus-tries. Although the field is not new, recently it hasreceived renewed interest from disciplines such associology (Hannan and Freeman, 1977, 1989;Carroll and Hannan, 2000), geography (van Dijk and Pellenbarg, 2000), industrial organization(Geroski, 1995; Audretsch, 1997; Caves, 1998),

and demography (van Wissen, 2002). The mainreasons for this increased interest are twofold:first, it is related and runs parallel to the increasedawareness of the role of the SME sector in theeconomy, both in terms of its role as employment

generator, and as one of the agents of innovation.Second, the increased availability of (longitudinal)micro-information of firms allows the empiricaltesting of theories of processes of firm formation,firm growth and survival. It is probably fair to saythat organizational sociologists have been mostactive in this area in the last decade, as witnessedfor instance by the work of Carroll, Hannan, andFreeman (Hannan and Freeman, 1989; Hannan andCarroll, 1992; Carroll and Hannan, 2000). In thesociological approach to the demography of thefirm, which is called organizational ecology (inthis paper henceforth called OE) modelling andempirical testing of demographic processes of change in what is termed populations of firms arevery important. As a result, OE researchers havediscovered a number of illuminating empiricallaws of the demographic behaviour of populationsof firms over time. The model of density depen-dence , which states that the growth path of anindustry (in OE called organizational populations)over time is dependent on the number of firms(size) in that industry, is particularly interesting inthis respect. Understandably, their explanation of this density dependence is founded in sociological

theory, although there are a number of clearsimilarities with industrial organization and eco-nomics (Boone and van Witteloostuijn, 1995).

In spatial sciences such as economic geographyand regional science, the demography of firms isused to describe and explain interregional differ-ences in economic growth (Storey, 1994). Thisapproach has close links with a number of other

A Spatial Interpretation of the

Density Dependence Modelin Industrial Demography Leo van Wissen

Small Business Economics 22 : 253264, 2004.2004 Kluwer Academic Publishers. Printed in the Netherlands.

Final version accepted on May 28, 2003

Faculty of Spatial SciencesUniversity of GroningenP.O. Box 8009700 AV, GroningenThe Netherlandsand

Netherlands Interdisciplinary Demographic Institute NIDI The HagueThe Netherlands

E-mail: [email protected]


2/12


3/12

general model of long-term organizational evolu-tion, which is called the density dependence model(Hannan and Freeman, 1977; Hannan and Carroll,1992; Carroll and Hannan, 2000). According tothis model, vital rates of birth and death of firmsare dependent on the size of the population, thepopulation density. Generally it is found that thereis a non-linear effect of population size onfounding rates, with an increasing effect at lowlevels, and a decreasing effect at high levels.Similarly, there is a non-linear effect on failurerates, with a decreasing effect at low densities, andan increasing effect at high levels. This leads to aclockwise pattern of founding rates on density, anda U-shaped pattern of mortality on density. Thepopulation growth rate is the combined effect of

birth and death rates. Due to the large variationin the precise forms of these dependencies, popu-lation growth patterns vary considerably, butthey are variations of a basic growth pattern, asdepicted in Figure 1. When the size of the popu-lation is small, founding rates are small and mor-tality is relatively large. Overall, growth rates aretherefore small. For unsuccessful populations anegative growth rate may even result if mortalityexceeds founding from the start and the popula-tion will cease to exist without ever having grownto maturity. For more successful populations,initially growth rates are small, but as the densityincreases (the number of firms increases) foundingrates increase, mortality rates decrease, and theoverall growth rates increases. When foundingrates are at their maximum and mortality rates near

their minimum level, the growth rate of thepopulation is at its maximum. Beyond this level,the founding rate decreases, the mortality rateincreases, and consequently the growth ratedecreases. As a result, the size of the populationstabilizes at the level of the carrying capacity .This overall growth pattern of initially slowgrowth, followed by rapid growth, and proceededby stabilization or decline in the size of thepopulation is verified empirically in numerouspopulations of firms and organizations. See Carrolland Hannan (2000), for an overview of the liter-ature.

Two basic forces are responsible for the sizedependency of firm founding and failure: legiti-mation and competition . Both forces are linked to

the size of the population. Legitimation refers tothe extent that a new organizational form orindustry is known and accepted in society. Carrolland Hannan refer to this as the taken-for-granted-ness of an organizational form, or more formallyconstitutive legitimation . When a new industryemerges, customers are not familiar with theproduct, investors are reluctant, and there may belegal or institutional constraints that prevent freemarket introduction. Legitimation increases withthe number of firms: the product becomes morefamiliar for customers, knowledge increases, andinvestors become less reluctant. Founding and dis-banding are related to the level of legitimation of an organizational form. The founding rate is pro-portional, and mortality is inversely proportionalto the level of legitimation.

The second underlying force is competitionbetween firms. In OE the term competition haspredominantly a social interpretation. Here, com-petition means conflict, rivalry. It arises as a resultof the interactions within a social system. Inaddition to this social interpretation, there is alsoan economic interpretation to which we will comeback below. However defined, there is clearly a

positive relationship between the number of firmsand the level of competition. As the size of thepopulation increases linearly, it may be argued thatcompetition increases geometrically (Hannan andCarroll, 1992). This means that the addition of anew firm to a small population has less impactthan adding an extra firm to a large population.Founding and disbanding rates are related to com-petition. The founding rate is inversely propor-

Density Dependence Model in Industrial Demography 255

Figure 1. Density evolution over time in the density depen-dence model.


4/12

tional to the intensity of competition, whereasmortality is proportional to it.

The joint effects of legitimation and competi-tion explain to a large degree the specific S-shapedstructure of population growth rates over time,from emergence to the level of the carryingcapacity. Initially, when the population size isvery small, low legitimation intensity and lack of competition lead to low growth rates. As thepopulation grows, legitimation increases, andcompetition is still very low, so that the growthrate increases. At a certain level, the maximumlevel of legitimation is reached, and competitionstarts to increase fast. Consequently, the growthrate decreases fast to zero or even becomesnegative.

These trends have been observed in many anddiverse industries, such as the automobile industry,banking and insurance, beer breweries, computerindustry, but also non-market organizations suchas labour unions, or day-care centres. For example,there are a number of remarkable similarities inthe shape of the curve of the number of automo-bile manufacturers across countries (Carroll andHannan, 1995; Hannan et al., 1995, Carroll andHannan, 2002). The automobile industry startedaround 1895. In the first decade after the start of the industry growth in the number of firms wasslow, as a result of a very low level of legitima-tion: the product was initially largely unknown, orat best a curiosity. In the US and Germany thenumber of producers was less than 10, in Franceslightly more. After 1895 a period of rapid growthin the number of manufacturers set in, whichlasted in all countries until about 19151925. Bythat time the US counted about 350 producers, andFrance about 150, while the number in Germanywas around 80. These levels may be interpreted asthe carrying capacities of the respective markets.The rapid growth was caused by increased legiti-mation, until a level that the automobile was

taken for granted in society, or fully legitimized.With increasing numbers, competition startedto press hard and negatively on founding andsurvival rates, until net entry became zero. Afterthis period, failure rates exceeded founding rates,and the number of manufacturers dropped dra-matically in the next thirty or so years. After 1945the number of manufacturers in the US stabilizedat a number well below 50 until the late seven-

ties, and in France and Germany a comparabledevelopment took place, with a stabilizationaround 20 manufacturers. These are large corpo-rations and the automobile market is highly con-centrated. Although there are large variations inthe exact shape of these curves, the same clock-wise pattern has been observed in many differenttypes of organizations and industries. The densitydependence model tries to formalize this observedempirical regularity.

Since its introduction in 1989 the density modelhas gained popularity, especially among organi-zational sociologists. Despite its success, it hasalso received various criticisms. Broadly speaking,there are three major points of critique. First, noaccount is taken of firm size in the theory, whereas

clearly large and small firms have very differenteffects in a population (Winter, 1990; Baum andPowell, 1995). Second, legitimation and competi-tion explain the S-shaped form of populationgrowth, which leads to a stable population size atthe level of the carrying capacity. It fails to explainnegative growth rates and the negative slope of thedensity curve beyond the peak, since a decrease inthe population size would lead to less competi-tion and therefore a return of the growth rate tozero (Baum, 1995). Third, firms differ not onlywith respect to size and economic activity, but alsowith respect to geographical location, which maybe labeled spatial heterogeneity. The geographicaldimension of the population is especially impor-tant for its precise definition, and this also playsa dominant role in the definition of the legitima-tion and competition processes (Carroll and Wade,1991; Swaminathan and Wiedenmayer, 1991;Bigelow and Carroll, 1997). Unfortunately, thespatial dimension has only received limited atten-tion in OE.

We will concentrate in this article on the issueof spatial heterogeneity in the evolution of the size(density) of industries over time. More specifi-

cally, we will use the literature on spatial agglom-erations and compare it with the core concepts of the density dependence model. OE researchershave proposed a number of solutions to theproblem of spatial heterogeneity of populations(Carroll and Wade, 1991; Baum and Mezias, 1992;Lomi, 1995). More recently, Lomi and Larsen(1996, 2001) have introduced spatial proximityinto ecological models of populations. These

256 Leo van Wissen


5/12

approaches, while being highly valuable in itself,do not take into account agglomeration economiesexplicitly. As will be shown, agglomerationeconomies may be viewed as an extension of thedensity model at the local level. At the same time,the implications of agglomeration economies intothis framework will lead to additional require-ments for the density model. Before spelling outthese cross-linkages, in the next section we willvery briefly point out the key features of agglom-eration economies.

3. Agglomeration economies

Agglomeration economies may be regarded as costsavings that result from the spatial concentration

of production at a given location. Since Krugman(1995) the geographical dimension into main-stream economics is important, but it was Marshall(1882) who already introduced the concept of external economies . In Krugmans modelseconomies of scale, the size of the home marketand transportation costs generate positive returnsto scale. Marshall had a somewhat broader viewon external effects, which according to him arevarious types of benefits and cost savings obtainedoutside the market, but that may lead to increasedproductivity of a firm. These may be the avail-ability of skilled labour, the availability of spe-cialized suppliers of intermediary goods, butalso localized knowledge spillovers, or simply theatmosphere. These external economies areusually available in larger urban centres, wheresimilar economic activities are carried out in closeproximity. External economies are therefore oftenlocalized economies , and the regions where theyappear he called industrial districts . Industrialdistricts are clusters of manufacturing SMEfirms who benefit from external economies arisingfrom the availability of specialized labour, spe-cialized services and trade organizations, as well

as the availability of specialized machinery.Although there are different typologies of agglom-eration effects the most important distinction isbetween localization and urbanization economies.Localization economies result from the geograph-ical clustering of similar types of firms, whereasurbanization economies result from the geograph-ical clustering of different types of firms. Thus,localization economies involve specialization of

regions, whereas urbanization economies involveregional diversification. Localization economiesare external economies of scale and may arisewhen many firms in the same industry are locatedin the same region, and share the same specializedservices, specialized infrastructure, have possibil-ities for joint research and development activities,as well as region-wide marketing. Moreover, theymay benefit from the same specialized labourpool. Firms in this setting are substitutes.

Urbanization economies involve firms fromdifferent industries. Here, inter-firm interaction ishighly important, in terms of input-output rela-tionships of suppliers and deliverers. There is aunique mix of local industries that provides thepotential for cost advantages. Industries in this

setting are complementary: output from oneindustry is used as input by another industry.Porter (1990) emphasized the importance of anindustrial cluster of different industries that arelinked through intensive input-output relationshipsat the national level. In his view, these industryclusters may or may not imply a spatial clustering.Isard et al. (1959) developed the concept of industrial complexes , which are industry clusterscharacterized by intensive forward and backwardinput-output linkages and showing in addition alarge degree of geographical clustering as well(see also Czamanski, 1977). As it turns out,economic linkages are often associated with geo-graphical clustering (Ellison and Glaeser, 1997;Feser and Sweeney, 2000).

As a result of these external economies, firmshave lower production costs and have a benefitover firms in other regions. Due to this cost dif-ferential, more firms will move into the region,which leads to even larger cost advantages. As aresult, these locations become even more attrac-tive, which leads to additional economic growth,and so on. This cumulative causation process isvery important in explaining why economic

growth is highly unevenly distributed in space.In the geographical literature the economic

definition of agglomeration externalities has beenextended to include also many informal and/ornon-economic linkages between firms. In thisframework an extensive literature has developedaround a number of core concepts. The notion of the new industrial district is an extension of theMarshallian concept and includes also the social



6/12

and cultural background of the region, as well asthe global market in which the firms produce(Amstrong and Taylor, 2000). In this view,agglomeration effects are not only economic, butalso include social, cultural and institutionaldimensions. This involves notions of social capitaltheory of values, norms, and social networks withinwhich firms operate. Scott (1988) emphasized thatvertical specialization in a production column iseasier in urban districts due to lower transactioncosts. These arise because of larger opportunities,and less searching costs. Firm networks are builtaround a notion of trust and confidence among theparticipants, which reduces the need for formalcontracting and other formal rules and institutions.Other institutional factors, especially norms and

values, and other untraded interdependenciesmay also be important (Storper, 1997). Anothercentral notion is knowledge accumulation andspillover, which may lead to learning regions , andinnovative milieux (Lambooy, 1997). Especiallytacit knowledge is shared through informalnetworks of entrepreneurs and local workers.

Social structure and knowledge accumulationare important aspects of agglomeration, but are notsufficient for successful regional development. Aregion needs a certain institutional thickness aswell (Amin and Thrift, 1995): a local network of institutions and organizations supporting localfirms, e.g. banks, venture capitalists, chambers of commerce, supportive local government agencies,etc. (Malmberg and Maskell, 1996). Many of theseinformal and/or non-economic linkages are highlylocalized in character, and therefore emphasizethe importance of geographical clustering evenmore than the traditional economic linkages. Forinstance, Audretsch and Feldman (1996) showedthat technological spillovers typically work atthe very local level. Van Soest et al. (2002) andVan Oort (2002) reach similar conclusions usingdata on employment growth in the Netherlands.

Audretsch (1998) shows that economic activitiesthat are based on new knowledge have a tendencyto co-locate within a geographical region, the mainreason being that knowledge is communicatedmore easily at shorter distances. These non-marketexchanges between firms have been categorizedunder the name of embedding. Embeddedness isa frequently used concept in new approaches of economic geography (Amin, 1999). Economic

activities are framed within a particular socialcontext. It involves social relationships with otheragents, as well as with social institutions. Thissocial context has both positive and negativeconsequences for (potential) new firms. Positiveaspects are the provision of information andresources, facilitating, and activity channelling.Negative aspects are constraints on behaviour andinformation due to a (non-optimal) position in thenetwork.

4. Comparing density dependence and location4. theories

There are a number of important cross-linkagesbetween the concepts of agglomeration economies

and the OE theory of density dependence.Both involve some form of positive feedback between industry size and growth potential. Belowthis issue will be addressed in more detail, byanswering three related questions:

Are both theories consistent or conflicting witheach other?

Can the geographical dimension of the densitydependence model be made explicit, using theframework of agglomeration economies?

What can be gained by this comparison of soci-ological and economic-geographical concepts

for OE? For agglomeration theories?These questions will be addressed in turn below.

Complementary or conflicting theories?

We begin this discussion by elaborating upon thetwo underlying forces that according to OE shapethe development of the size of an industry overtime: legitimation and competition. Although com-petition is an economic concept, and legitimationnot, it is too simple to state that they representsociological versus economic arguments, since

legitimation also covers various economic mech-anisms.

Legitimation in OE may have two meanings.First, it can be interpreted as conforming to aset of rules or conventions. Second, and morerelevant, it may refer to constitutive legitimation,or taken-for-grantedness. New industries emergeas a result of an innovation. Initially, the new firmproducing the innovation (in OE the new organi-

258 Leo van Wissen


7/12

zational form) is unknown and lacks legitimation.This may imply for instance, that there is a lack of customers and information about the product,that there are no suppliers of goods, services, orcapital. Moreover, a pool of skilled labour withexperience in the production process does not yetexist. Further, an established network of similarproducers, who share knowledge and other infor-mation, is lacking, and there may be legal andinstitutional barriers to production. As a result, the(potential) new entrepreneur faces large startupproblems. Founding rates are low, and mortalityrates are high. Nevertheless, as time progresses,these limitations may be overcome. This is relatedto the build-up of a social structure of the industry(Carroll and Hannan, 2000), which includes a set

of roles and positions in a network of organiza-tions.These indicators of legitimation resemble in

many respects the emergence of agglomerationeconomies as presented above. First, they involveelements of localization economies: the size of thecustomer base, marketing, the size and quality of the labour pool, and a network of producers thatmay share common knowledge and experience.Second, it also contains some elements of urban-ization economies, especially the creation of input-output relationships with other industries. Third,the element of creating a social structure of anindustry is highly similar to the defining featuresof a new industrial district, based on a commonsocial and cultural background, as well as creatinginstitutional thickness and embedding.

New organizational forms, or entrepreneurs,when viewed in their role as innovator, are bydefinition stepping outside existing institutionalsettings. They create something new that didnot exist previously, or may even threaten theexistence of current firms and industries (cf.Schumpeters view of entrepreneurship as creativedestruction). In this way, innovation and creative

destruction can be defined as institutional disem-bedding. This feature of new firm populations isexactly the cause of lacking taken-for-grantedness.At the same time, an innovation creates a newmode of production that may be followed byothers. Most new firms are not pure innovatorsthemselves, but imitators of earlier innovators(Nelson and Winter, 1982). At the same time theydo conform to another important characteristic of

entrepreneurs, that of being high risk-bearers.They face higher intrinsic uncertainty because since the population is still very small and young they lack sufficient information to attach reliableprobabilities of success to their actions. The morefirms engage in the new routines of production,the more reliable the probabilities of success, andthe lower the fundamental uncertainty about entre-preneurial actions. In this way, entrepreneurialactions become established routines and in theprocess they loose their innovative character. Thisis an important aspect of legitimation as expressedby OE, and also very close to the concept of learning regions and regional knowledge accu-mulation, one of the important features of newindustrial districts.

Legitimation also involves the establishment of networks : of other firms in input-output marketrelations, but also in various other informalnetworks, with other entrepreneurs, and organiza-tions. Here, tacit and other knowledge isexchanged, and reciprocal relationships of trust areestablished. Thus, there are many similaritiesbetween the legitimation concept of OE and thegeographical concepts of new industrial districts ,and innovative milieux .

Despite these similarities there are strong dif-ferences as well. First, since in the density depen-dence model legitimation is dependent on the sizeof the own population, this implies that they areonly the result of localization economies. Thishowever ignores the inter-industry linkages andurbanization economies, which are crucial inindustrial districts. This problem arises becausethe density dependence model is a single industrymodel. A second difference is that agglomerationeconomies have in principle no upper limit, butaccording to OE there is a maximum to the levelof legitimation. This runs parallel to the assump-tion of a fixed (exogenous) carrying capacity fora population. While this may be the case in many

ecological applications, it does not hold for indus-tries. In urban regions under certain conditionspositive feedback between size of the agglomera-tion and growth may prevail for a long time,although beyond a certain limit negative feedback may set in, for instance as a result of congestion.Since this issue is also relevant when discussingthe issue of competition in OE we will return toit below.



8/12

A third difference is the lack of an explicit geo-graphical dimension in legitimation, whereas it isof central concern in agglomeration economies.Spatial proximity is highly important for the emer-gence of agglomeration economies. Transportationcosts between the industry and the home market,or between suppliers and deliverers are lower. Theavailability of a skilled labour market is a local-ized advantage as well. Moreover, knowledgespillovers, networks, local institutions, and theemergence of a social structure of industry all needspatial proximity. As discussed before, knowledgebased industries benefit more from spatial prox-imity. Therefore, sectoral differences exist in theneed for spatial clustering. Moreover, the industrylife cycle is also relevant in this respect. Industries

in the early stage of the life cycle are more depen-dent on knowledge and innovations, which maylead to a larger tendency for agglomeration (seealso Van Oort, 2002). In later stages of the lifecycle, firms are more likely to be engaged in pricecompetition, and there is less need for spatial clus-tering. However, the benefits of spatial proximityfor industry growth have not been recognized inOE until now. On the contrary, spatial proximityleads to negative competition effects on survivaland growth. Baum and Mezias (1992) for instance,in a study on the dynamics of the New York hotelindustry, find that hotels located in the districtswith the highest hotel density have the highestfailure risks. If this were the whole story inregional industry dynamics, firms would try tosettle as far as possible from similar firms in theindustry.

The second process that governs the densitydependence model is competition . Competition inOE refers to either or both of two types of inter-action. Structured competition refers to a limitednumber of actors that are in direct rivalry witheach other. Diffuse competition refers to a largeset of agents that compete for the same limited

resources. The latter form of competition is of direct relevance for industrial demography, sinceit is directly linked to the notion of density.Structured competition is studied extensively inindustrial organization. Competition is a muchmore straightforward notion than legitimation, andcan be analysed using economic concepts.Nevertheless, as already observed earlier, in OEcompetition is given a very simple interpretation:

it rises geometrically with the number of firms inthe population, given a fixed resource space(consumer market). Whereas legitimation is apositive feedback mechanism, competition leadsto negative feedback between industry size andgrowth. Thus, legitimation represents the cen-tripetal forces for spatial clustering, whereascompetition in OE represents the centrifugal ten-dencies. As a result of increased competitionpopulation growth will slow down until zero atmaximum population size (the carrying capacity).The OE notion of competition is therefore notcompatible with agglomeration economies. Muchcan be said about the lack of economic reasoningin this theory, but with reference to agglomerationeconomies three issues stand out. First, agglom-

eration economies lead to competitive advantagesfor firms. The more firms, the higher the positiveexternalities enjoyed, and the larger the competi-tive advantages. These may (partly) be viewed asexternal economies of scale, but they are disre-garded in OE. In OE the evolution of the industryis towards increasing market concentration in laterstages of the life cycle due to internal economiesof scale. Second, and already mentioned above,the assumption of a fixed resource space orexogenous carrying capacity is not justified. Afixed resource space presumes a closed marketwithout trade. Introducing trade would greatlycomplicate matters and bring us far beyond theboundaries of OE. It would introduce economicgeography into the model: transportation costs, thesize of the home market, relative prices and otherelements of trade theory; important concepts thatcannot be translated directly into a sociologicalmodel. Third, even without introducing trade, theassumption of a fixed resource space is nottenable. The resource space is not only made upof final consumers, but also by intermediatedemand. Many industries deliver primarily toother businesses. As a result of input-output rela-

tionships and positive feedback between differentindustries, the growth of one industry populationmay enhance the resource space of other indus-tries, and vice versa .

In summary, the process of legitimation is to acertain extent very close to the geographicalconcepts of new industrial district and innovativemilieux, especially in the non-economic exchangesbetween firms, and the build-up of a social struc-

260 Leo van Wissen


9/12

ture. Nevertheless, there are important differencesas well, especially the disregard of urbanizationlinkages between different organizational popula-tions, putting an upper limit on positive feedback between population size and growth potential, andthe lack of a geographical dimension. Moreover,the competition process as depicted in the densitydependence model lacks any agglomeration mech-anism of positive feedbacks, and is thereforenot compatible with any notion of agglomerationeconomies.

Introducing agglomeration effects in the densitydependence model?

Space matters in economic growth, and also in

legitimation and competition within and betweenindustries. The geographical boundaries of themarket define the population density as well as theresource space of the population. OE has its intel-lectual roots in human ecology, in which spatialcompetition processes play an important role(Park, 1925). Strangely, so far in OE, the spatialdimension has not been very important, and atten-tion was solely devoted to temporal variations inpopulation change. The theory of agglomerationeconomics makes it clear that there is a theoret-ical rationale for including geography into thedefinition of firm populations. At the sametime, according to the same theory, not all indus-tries have a similar need to cluster in space.Agglomeration effects are stronger in knowledgeintensive industries. As noted earlier, there is alsoa link with the industry life cycle, with spatialclustering being more relevant in the early stagesof the industry life cycle, where product innova-tions are important. As the industry develops, thegeographical pattern that emerged in the earlierstages may have a lasting impact upon the geo-graphical distribution of further growth. This maybe labelled spatial lock-in or path dependency.

There are a number of examples of spatialanalyses within OE. Broadly speaking, geograph-ical location in OE is viewed as one form of population diversification, which calls for differentmodels at different geographical scales (Barnettand Carroll, 1987; Carroll and Wade, 1991; Baumand Mezias, 1992; Bigelow and Carroll, 1997). Arelated development views resources as heteroge-neous, and deals with market partitioning in

segments (Carroll, 1984; Carroll and Hannan,2000). In the resource partitioning model, firmsoccupy niches in the market, which may bedefined as geographical niches. Until now, theresource partitioning model has only been appliedin a non-geographical setting. In the empiricalanalysis of geographically heterogeneous popula-tions two strands of research may be observed: astatistical approach where the optimal size of thegeographic region is determined as a result of ananalytical model, as in Lomi (1995), and anapproach where, based on a priori reasoning, thesize of the market area is fixed (Baum and Mezias,1992). The key issues in these papers are the geo-graphical definition of the population, dealing withheterogeneity within the population, and the effect

of local density on competition and founding anddisbanding rates. These are important issues, butfrom a geographical point of view the most impor-tant issue of agglomeration economies is lacking.

In order to introduce the notion of agglomera-tion economies into the density dependencemodel, a number of model extensions are neces-sary. First, the processes of legitimation and com-petition (i.e. the market or resource space) shouldbe defined geographically. These spatial scalesneed not be similar though (Zucker, 1989; Lomi,1995; Carroll and Hannan, 2000). Carroll andHannan for instance conjecture that legitimationand competition may work at different spatialscales. They argue that legitimation often worksat the national level, whereas competition is muchmore local. We concluded in this paper on thecontrary that in some situations this might wellbe the other way around. Many legitimationprocesses, similar to the creation of new industrialdistricts, are very localized in nature, whereas theresource space may be very large, for instancewhen dealing with firms operating on a globalmarket. Consumer awareness and marketing com-petition may be local whereas legitimation is a

process on a much larger geographical scale, butlegitimation entails much more than this. Wediscussed above that knowledge spillovers,institutions, social structure, and trust are verylocalized phenomena. A second model extensiondeals with between-industry linkages. Interactionsbetween firm populations should be incorporated,in order to capture urbanization economies. Thisis certainly an interesting but complicated topic.



10/12

In some respects the resulting models mightresemble elements of existing non-economictheories of agglomeration (Anas et al., 1998). VanWissen (2000) gives an example of such anapproach in the context of a microsimulationmodel of industrial demography. Third, local vari-ations in population density can also have apositive effect on founding and firm growth, espe-cially in the early life cycle stages of the industry,and in knowledge based industries.

Density effects in models of agglomerationeconomies?

In the previous subsection we addressed thequestion if introducing notions from agglomera-

tion economies in the density dependence modelwould be fruitful. We may also turn this questionaround, and ask what can be gained from this com-parative effort for our understanding of agglom-eration economies? In answering this questionthree issues should be emphasized. First, theconcept of legitimation offers a new sociologicalangle to a number of processes that are studied indepth by economic geographers, especially thenature of non-market exchanges and institutionalnetworks. Second, the density dependence modelis confirmed in numerous empirical studies, andhas attained the status of an empirical law, applic-able to many different organizational forms and inmany situations. There is also a substantial liter-ature on the exact specification of the relationshipbetween population density and founding anddisbanding processes of firms. This is quite dif-ferent in research focussing on agglomerationeconomies. The emphasis here is much more ontheoretical models and concepts. Empirical resultsare sometimes inconclusive, and despite the wide-spread conviction that it is of major importance,empirical confirmation often remains difficult(Van Oort, 2002). To start with, economies of

localization might be specified in terms of adensity dependence model, and this opens upinteresting new ways of empirical testing of agglomeration economies. Third, OE deals withfounding and disbanding of firms, as well asgrowth of incumbent firms. Looking at regionalgrowth from such a firm demographic perspec-tive may give a strong behavioural interpretationto agglomeration economies. Recently, this point

of view is used in studies on the emergence of industrial districts, such as Silicon Valley orDetroit, where new and successful firms in theregion start as spin-offs from successful incum-bents in the emerging period of the industry(Klepper, 2001, 2002). The density dependencemodel when applied in such a setting may there-fore help to open up the still largely mysteriousblack box of agglomeration economies.

5. Summary and conclusions

In this paper the density dependence model, whichwas developed in OE was compared to theeconomic-geographical notion of agglomerationeconomies. There is a basic resemblance: both

involve some form of positive feedback betweensize of the industry and growth. We explored froma theoretical perspective how the theoreticalconcepts compare to each other, and if an inter-disciplinary cross-fertilization between both isfruitful. Two driving forces are important for thetemporal evolution of industries in OE: legitima-tion and competition. When viewed from a spatialperspective, legitimation contains the centripetalforces, and competition the centrifugal forces inspatial cluster formation. Legitimation wasshown to have close links with the agglomerationeconomies literature, especially with various non-economic exchange linkages between firms,embeddedness, institutional thickness, and thecreation of social structure for new populationsof firms. In legitimation we find positive feedback effects similar to economies of localization.However, there are differences as well. Mostimportantly, urbanization economies are notcovered in the model, and the model lacks a clearspatial dimension. The notion of competition inthe density dependence model is only rudimentarydeveloped and the notion of external economies of scale is lacking here. Competition as viewed by

OE is a negative feedback mechanism, and there-fore not compatible with positive agglomerationeffects. The comparison with the literature onagglomeration economies shows that a number of important elements are missing here. Regional dif-ferences in founding and disbanding are at leastpartly the result of agglomeration economies andtherefore this should be taken into account in thedensity dependence model. The legitimation

262 Leo van Wissen


11/12

concept is a useful starting point here, since itlinks closely with existing agglomeration conceptsin economic geography. In addition, however, weneed inter-population linkages, in order to captureurbanisation economies. Such a step towardsmulti-population OE models is also foreseen byCarroll and Hannan (2002, p. 451), when dis-cussing future directions of the field:

[. . .] a scientifically sound multipopulation design likelyneeds to be developed. We think that an important nextstep in developing the demography of corporations andindustries will design and conduct research on the com-munities of interacting populations of corporations andindustries (Carroll and Hannan, 2002, p. 451).

If this direction is taken in OE, then some of thepotential cross-fertilizations may be realized in

this programme. Incorporating agglomerationeffects in a multi-industry density dependencemodel is an interesting but complex problem.Lessons could be learned here from existing (non-economic) dynamic models of agglomerationeconomies. These models often have complexdynamic properties resulting from non-lineardynamic relationships of interacting populations.

Another result of our comparison was thatagglomeration effects will vary between indus-tries, and are especially relevant in the formativeperiod of the industry. This life cycle aspect mayfit neatly in the framework of the density depen-dence model.

Finally, theories and models of agglomerationeconomies may also benefit from insights derivedfrom OE. The strong empirical roots of the densitydependence model might be relevant for alterna-tive tests of localization economies. Moreover, thedifferent conceptual angle, provided by the soci-ological concept of legitimation, but also the focuson demographic processes of founding and dis-banding, may prove to be fruitful for extendingour understanding of agglomeration economies.

References

Amin, A., 1999, An Institutionalist Perspective on RegionalEconomic Development, International Journal of Urbanand Regional Research 23(2), 365378.

Amin, A. and N. Thrift, 1995, Globalisation, InstitutionalThickness and the Local Economy, in Patsy Healey,Stuart Cameron, Simin Davoudi, Stephen Graham and AliMadani-Pour (eds.), Managing Cities: The New UrbanContext , Chichester: Wiley, pp. 92108.

Amstrong, H. and J. Taylor, 2000, Regional Economics and Policy , 3rd Edition, Oxford: Blackwell Publishers.

Anas, A., R. Arnott and K. Small, 1998, Urban EconomicStructure, Journal of Economic Literature 26 (3), 14261464.

Audretsch, B., 1997, Technological Regimes, IndustrialDemography and the Evolution of Industrial Structures,

Industrial and Corporate Change 6(1), 4982.Audretsch, B., 1998, Agglomeration and the Location of

Innovative Activity, Oxford Review of Economic Policy14(1), 1829.

Audretsch, D. B. and M. P. Feldman, 1996, R&D Spilloversand the Geography of Innovation and Production,

American Economic Review 86(3), 630640.Barnett, W. and G. Carroll, 1987, Competition and Mutualism

among Early Telephone Companies, AdministrativeScience Quarterly 32 (3), 400421.

Baum, J. A. C., 1995, The Changing Basis of Competitionin Organizational Populations: The Manhattan HotelIndustry, Social Forces 74 (1), 177205.

Baum, J. and S. Mezias, 1992, Localized Competition and theDynamics of the Manhattan Hotel Industry, AdministrativeScience Quarterly 37 (4), 580604.

Baum, J. A. C. and W. W. Powell, 1995, Cultivating anInstitutional Ecology of Organizations: Comment onHannan et al., American Sociological Review 60 (4),529.

Bigelow, L. S. and G. R. Carroll, 1997, Legitimation,Geographical Scale, and Organizational Density. RegionalPatterns of Foundings of American Automobile Producers,Social Science Research 26 (4), 377398.

Boone, C. A. J. J. and A. van Witteloostuijn, 1995, IndustrialOrganization and Organizational Ecology: The Potentialsfor Cross-Fertilization, Organizational Studies 16 (2),265298.

Carroll, G. R., 1984, Concentration and Specialization:Dynamics of Niche Width in Populations of Organizations, American Journal of Sociology 90 (1),12621283.

Carroll, G. R. and M. T. Hannan (eds.), 1995, Organizationsin Industry: Strategy, Structure, and Selection , New York:Oxford University Press.

Carroll, G. R. and M. T. Hannan, 2000, The Demographyof Corporations and Industries , Princeton: PrincetonUniversity Press.

Carroll, G. R. and J. Wade, 1991, Density Dependence inthe Organizational Evolution of the American BrewingIndustry across Different Levels of Analysis, SocialScience Research 20(3), 217302.

Caves, R. E., 1998, Industrial Organisation and New Findingson the Turnover and Mobility of Firms, Journal of Economic Literature 36 (4), 19471982.

Czamanski, S., 1977, Needless Complexity in theIdentification of Industrial Complexes: A Comment,

Journal of Regional Science 17 (3), 455457.Ellison, G. and E. L. Glaeser, 1997, Geographical

Concentration in US Manufacturing Industries: ADartboard Approach, Journal of Political Economy107 (5), 889927.

Feser, E. J. and S. H. Sweeney, 2000, A Test for Economic



12/12

a spatial interpretation of the

Documents