An emerging perspective on persistent inequality suggests that the causal determinants of disparity may be found in the social interactions that an individual experiences over the course of his or her life. Three basic features characterize this perspective: 1. Individual preferences, beliefs, and opportunities are strongly influenced by one's membership in various groups. These preferences, beliefs, and opportunities may be formed by the social attitudes of the group members or of others, by market mechanisms, by social policy, or by some combination of geographic location and social attitudes. 2. Interaction effects occur among members of a given group, so that group level influences generate similar outcomes among group members. 3. Social stratification by income, race, education, or language can lead to divergence across group characteristics which results in greater inequality at a point in time, as well as decreased social mobility over time.
The Network on Economic Inequality and Social Interactions, initiated by the Foundation in December 1994, referred to this perspective as the "Memberships Theory of Inequality." With its focus on how different socioeconomic group memberships influence individual outcomes, memberships theory provided a complementary approach to the more traditional economic focus on the development of individual human capital. In addition to its implications for theoretical and empirical analysis, memberships theory also contained important policy implications. For example, the interactions-based perspective on inequality implies that altering group memberships may be an essential part of efforts to achieve greater equality of opportunity. Operationally, this perspective suggests that policies such as affirmative action and the deconcentration of public housing may advance access to opportunity in ways not previously recognized.
The network had 12 core members, including nine economists, one political scientist, and two sociologists. With Foundation support, network members focused their efforts in three areas. First, the network explored ways to identify and measure social interactions and their attendant effects on inequality. These studies mixed the development of econometric analyses for data such as the Panel Study of Income Dynamics with the exploration of natural "experiments" in group formation, such as the assignment of dormitory rooms to college students.
Second, the network pursued the development of theoretical models of interactions to examine several questions, including how social interactions in labor markets contribute to inequality and how government policies such as affirmative action influence the distribution and redistribution of resources. This theoretical work provides the context in which to interpret empirical analyses, especially with respect to implications for public policy.
Finally, the network explored computational simulations to study inequality, drawing on the various theoretical and empirical models pioneered by group members and other researchers. Computer simulations are a promising tool for studying the properties of large-scale systems and the direct and indirect effects of different policy interventions. Eventually, the group intended to develop a set of computational models which, when provided with detailed information about individual characteristics, such as education and race, and the multiple groupings in which human beings develop, can allow one to explore the consequences of different policies in a way which respects the richness and complexity of an individual's interactions over the life cycle.