NBER Publications by Peter Arcidiacono

We argue that once we take into account the students' rational enrollment decisions, mismatch in the sense that the intended beneficiary of affirmative action admission policies are made worse off could occur only if selective universities possess private information about students' post-enrollment treatment effects. This necessary condition for mismatch provides the basis for a new test. We propose an empirical methodology to test for private information in such a setting. The test is implemented using data from Campus Life and Learning Project (CLL) at Duke. Evidence shows that Duke does possess private information that is a statistically significant predictor of the students' post-enrollment academic performance. We also propose strategies to evaluate more conclusively whether the evide...

In traditional signaling models, education provides a way for individuals to sort themselves by ability. Employers in turn use education to statistically discriminate, paying wages that reflect the average productivity of workers with the same given level of education. In this paper, we provide evidence that education (specifically, attending college) plays a much more direct role in revealing ability to the labor market. We use the NLSY79 to examine returns to ability early in careers; our results suggest that ability is observed nearly perfectly for college graduates but is revealed to the labor market much more gradually for high school graduates. As a result, from very beginning of the career, college graduates are paid in accordance with their own ability, while the wages of high ...

Using data on the universe of students who graduated from U.S. medical schools between 1996 and 1998, we examine whether the abilities and specialty preferences of a medical school class affect a student's academic achievement in medical school and his choice of specialty. We mitigate the selection problem by including school-specific fixed effects, and show that this method yields an upper bound on peer effects for our data. We estimate positive peer effects that disappear when school-specific fixed effects are added to control for the endogeneity of a peer group. We find no evidence that peer effects are stronger for blacks, that peer groups are formed along racial lines, or that students with relatively low ability benefit more from their peers than students with relatively high ability...

This study investigates whether models of forward-looking behavior explain the observed patterns of heavy drinking and smoking of men in late middle age in the Health and Retirement Study better than myopic models. We develop and estimate a sequence of nested models which differ by their degree of forward-looking behavior. We also study models which allow for heterogeneity in discounting, and thus test whether certain types of individuals are more likely to show forward-looking behavior than other types. Our empirical findings suggest that forward-looking models with an annual discount factor of approximately 0.78 fit the data the best. These models also dominate other behavioral models based on out-of-sample predictions using data of men aged 70 and over. Myopic models predict rates of sm...