The population ecology of organizational founding: location dependence and unobserved heterogeneity

Administrative Science Quarterly,  March, 1995  by Alessandro Lomi

• E-mail
• Print
• Link

Organizational environments have spatial components that affect the evolutionary dynamics of organizational populations. First, geographical barriers of various kinds may allow enough isolation for different evolutionary paths to be explored in different regions (Eldredge and Gould, 1972; Mayr, 1976). Second, localized resource environments may pose complicated problems of adaptation for individual organizations (Carroll, 1985; Hannan and Freeman, 1989; Baum and Mezias, 1992). Finally, processes of legitimation and competition, which respond to organizational density. may vary depending on the geographical boundaries used to define organizational populations (Carroll and Wade, 1991; Hannan and Carroll, 1992). When taken together, the consequences of heterogeneity in the spatial distribution of resources have far-reaching implications for the ecological dynamics of organizational populations, because the level of spatial aggregation defines implicitly the population boundaries (Singh, 1993), and the intensity of competition among organizations is a function of the similarity in resource requirements (Hannan and Freeman, 1989; Baum and Mezias, 1992).

Based on the argument that markets and other institutional arenas eventually evolve at least to the national level, organizational ecologists usually specify population processes at that level (Hannan and Carroll, 1992). The propagation of organizational populations in space to the national and often supranational level, however, does not imply spatial homogeneity. Many organizations remain local and depend heavily on their immediate institutional and competitive environments for support, resources, and demand. But location dependence is not a phenomenon restricted to small or marginal organizations. The most striking feature of economic and organizational activity is their geographical concentration: Certain areas become so highly specialized that production in many industries is concentrated in space (Krugman, 1991a). It has frequently been mentioned, for example, that "Nighttime satellite photos of Europe reveal little of political boundaries but clearly suggest a center-periphery pattern whose hub is somewhere in Belgium" (Krugman, 1991b: 484). Similarly, the manufacturing belt in the U.S., the tile cluster and the textile district in northern Italy, the research triangle in North Carolina, the machine-tool district in Germany, the networks of suppliers surrounding the kaisha (lead manufacturers) in Japan, Silicon Valley, and Route 128 are all expressions that conjure up images of highly localized organizational activities (Piore and Sabel, 1984; Best, 1990; Porter, 1990; Preer, 1992).

The recurrence of patterns of organizational concentration in space across different industries and in a number of national contexts provides indirect evidence that location may be a general factor shaping the evolution of organizational populations. If forces exist that give evolutionary advantages to organizations located near other organizations or in specific geographical areas, then the internal structure of the organizational population can no longer be considered homogeneous, and organizational birth and death rates will vary systematically across locations. Also, if all organizations in a population do not compete for the same scarce resources or contribute to competition in the same way, then it is essential to select the appropriate level of analysis to examine the dynamics of the population, since different levels of spatial aggregation imply different assumptions about how general processes of legitimation and competition unfold.

This paper pursues these issues of location dependence and unobserved heterogeneity in the context of processes of organizational creation, for two reasons. First, in spite of the accumulation of empirical results in the population ecology of organizational founding, relatively little is known about the consequences of the level of spatial aggregation on founding rates in organizational populations. Lacking explicit theoretical indications, various researchers have tried different levels with different results (Barnett and Carroll, 1987; Carroll and Wade, 1991; Swaminathan and Wiedenmayer, 1991; Baum and Mezias, 1992; Baum and Singh, 1994; Hannan and Carroll, 1992; Freeman and Lomi, 1994). In their analysis of the U.S. brewing industry, for example, Carroll and Wade (1991) found support for ecological models of density dependence at the state and regional levels but not at the city level, while Carroll and Hannan (1989) found that founding rates of newspaper organizations depend on density in a way that is consistent with ecological theories only in small metropolitan areas. Second, while problems related to unobserved heterogeneity have been raised frequently in the context of processes of organizational mortality (Freeman, Carroll, and Hannan, 1983; Hannan, 1988; Petersen and Koput, 1991), virtually no research has been done on the implications of unobserved heterogeneity for organizational founding rates.