Discrimination with endogenous group choices: A theory and its empirical application

Discrimination with endogenous group choices . We investigate the possibility of discrimination based on whether a worker undertakes some seemingly irrelevant activity. We study a model in which workers choose group identities and make human capital investments. Firms compete for workers based on their observable group identities and imperfect but informative signals of their qualifications. Assuming that the distribution of investment cost and the distribution of the utility derived from the group activity are independent, we show that equilibria where agents from different groups are treated differently can exist--even when the group independent equilibrium is unique. The engine of discrimination here is endogenous group choice. Group identity becomes a signaling instrument because a single crossing condition is endogenously generated as a property of the group dependent equilibrium. Disentangling college premium: Estimating a model with endogenous education choices . We extend the model in the first chapter into a structural general equilibrium model of education choices. It is then used as a framework to empirically disentangle the contributions of ability signaling and productivity enhancement to the college wage premium. The parameters of the model are estimated using 1990 U.S. Census 5% PUMS data. We find that college education enhances attendees' productivity by about 40%. In a counter-factual economy in which college education did not enhance attendees' productivity at all, the college wage premium would be at most 21% in contrast to a premium of approximately 59% in the data. Thus we conclude that productivity enhancement accounts for a minimum of nearly 64% of the college wage premium observed in 1989. Social culture and economic performance . We investigate the possible efficiency roles of differentially treating workers based on whether they undertake some seemingly irrelevant activity. The connection between higher paying jobs and the seemingly irrelevant activity is interpreted as "social culture". In the context of a technology adoption problem, we show that the society can partially overcome an informational free riding problem by allowing the firms to differentially treat workers based on some "cultural activity". Therefore social culture may affect the economic performance by altering the effective production technology of the economy.