Organizational demography and culture: insights from a formal model and simulation

Administrative Science Quarterly, Sept, 1998 by Glenn R. Carroll, J. Richard Harrison

An active line of contemporary research on organizations investigates the effects of demography on organizational outcomes. Inspired by Pfeffer's (1983) arguments, researchers have examined the consequences of demographic distributions in organizations in a variety of contexts and outcomes. Much of this research focuses on the demographic distribution that Pfeffer emphasized, the tenure or length of service (LOS) distribution in an organization or its top management team.(1) Outcome variables found to be associated with the LOS distribution include turnover (McCain, O'Reilly, and Pfeffer, 1983; Wiersema and Bird, 1993), innovation (O'Reilly and Flatt, 1986; Flatt, 1993), diversification (Wiersema and Bantel, 1992), and adaptiveness (O'Reilly, Snyder, and Boothe, 1993).

Researchers examining LOS distributions have put forward an impressive set of empirical findings over the last decade. An awkward number of different theoretical processes has been advanced to explain the observed associations, however, and many of these are based on questionable and unstudied assumptions. Several recent studies have also failed to find the expected relationship between the LOS distribution and organizational outcomes. For these reasons, we believe that the demographic research program could be improved by some analytical stock-taking and theoretical reformulation. More precisely, the demographic research program on organizations would be strengthened considerably at this point by the use of an explicit model, in particular, a formal mathematical model. According to Kreps (1990: 6-7), the main advantages of such a model are (1) clarity ("It gives a clear and precise language for communicating insights and contributions."); (2) ease of comparability ("It provides us with general categories of assumptions so that insights and intuitions can be transferred from one context to another and can be cross-checked between different contexts."); (3) logical power ("It allows us to subject particular insights and intuitions to the test of logical consistency."); and (4) analytical precision ("It helps us to trace back from 'observational' to underlying assumptions to see what assumptions are really at the heart of particular conclusions."). In a more casual vein, a model provides a different perspective on a research problem, and this fresh look often proves insightful in and of itself.

The features listed above promise to be useful to the research program on organizational demography. As we explain below, a synthetic but parsimonious recasting of various theoretical ideas used in this program implies a strong association between the demographic (LOS) heterogeneity of an organization and its cultural heterogeneity. Despite the role this association plays in designing and evaluating empirical research on the effects of LOS distributions, it (or any other theoretical link playing a similar role) has for the most part not been investigated directly. Instead, researchers typically assume - often implicitly and perhaps unwittingly - that the link between demographic heterogeneity and cultural heterogeneity is highly plausible under a wide variety of conditions and organizational contexts. A formal model of the underlying demographic and cultural processes would give us a new perspective on this research problem. It would potentially allow us to clarify the theoretical processes involved in the link, to examine the logical structure of the theory implicit in the usual research, to explore the general plausibility of the assumed link, to identify contexts (demographic and organizational) in which the link would be more or less strong, and to design new approaches to studying organizational demography.

THEORY AND RESEARCH ON ORGANIZATIONAL DEMOGRAPHY

Although several earlier minor contributions can be found, most contemporary research on organizational demography stems from Pfeffer's (1983) theoretical essay on the topic. In his essay, Pfeffer made a persuasive general case for the study of organizational demography. He also laid out a number of specific theoretical propositions about the causes and consequences of demographic phenomena in organizations. Most of these specific arguments concentrate on the properties of demographic distributions, especially the tenure or length of service (LOS) distribution of members of the organization or its top management team. The number of subsequent empirical studies examining this distribution shows that organizational researchers are especially intrigued with Pfeffer's arguments about the consequences of the LOS distribution on organizational outcome variables, including (1) innovation, adaptation, and performance; (2) administrative succession or turnover; (3) organizational control structures and processes; (4) interorganizational relations; (5) career trajectories; (6) cohort identity and conflict; and (7) the distribution of power within organizations. How LOS distributions are thought to affect these outcomes depends on the substantive details of the theorized process for a particular outcome, but a common general theoretical formulation holds that unevenness or heterogeneity in the LOS distribution is the demographic variable of primary interest. An example is the theoretical argument of Wagner, Pfeffer, and O'Reilly (1984: 77) about LOS distributions and turnover: