Publications

Using data from over 115,000 university professors across public and private universities in the U.S. and Canada, we document significant and growing wage disparities among faculty. These disparities are largely driven by the cross-section of academic fields and their corresponding student future earnings. By employing a shift-share instrument based on persistent student placement patterns, we provide robust evidence of a causal relationship between student future earnings and faculty compensation, finding an elasticity of faculty compensation to student earnings after graduation in the [0.3 – 0.4] range. The elasticity of faculty wages to student earnings varies substantially across fields, being more than three times higher in Economics and Finance than in Humanities. We interpret these empirical facts as consistent with a mechanism where tuition and donation revenues per faculty, as well as faculty bargaining power, vary significantly across fields.

This paper empirically studies how college savings motives affect household stock market participation. More than 65% of households saving for college allocate at least a portion of their college savings to stocks, which are also the most popular class of risky assets for college savings investments. Employing the implementation of a financial aid program that eliminated college saving motives for some military households, I demonstrate that it reduced the probability of stock market participation outside of retirement accounts for these households by more than a quarter. I provide additional evidence that college savings motives explain a significant share of stock market participation by demonstrating that only one form of college savings, education savings accounts, accounts for 21% of stock market participation outside of retirement accounts for households with children.

This paper studies how menu design affects household participation and contributions in 529 college savings programs. Our empirical setting exploits two types of state-level staggered variation in program menu design: the inclusion of investment options tailored to different risk tolerance levels and the addition of an alternative menu through the creation of an additional 529 plan. We find that the inclusion of risk-based target-date options and the creation of an additional 529 plan both enhance participation and contributions in in-state college savings programs. The inclusion of static asset allocation options has no effect on participation but raises household contributions. Our results suggest that modifying the menu design of 529 college savings programs can increase investments in human capital.