Publications

There is a large gender wage gap among college graduates. This gender gap could be partially driven by differences in college major and prior skills. We use Swedish register data to study how much of the gender gap can be explained by differences in majors, skills, and skill prices. College majors explain 60 percent of the gender wage gap, but large gaps remain within majors. We find that within-major wage gaps are driven by neither differences in multidimensional skills nor returns to these skills. In fact, women are positively selected in terms of college preparation and skills in almost every major.

This paper develops a sufficient statistics approach for estimating the role of search frictions in wage dispersion and life-cycle wage growth. We show how the wage dynamics of displaced workers are directly informative of both for a large class of search models. Specifically, the correlation between pre- and post-displacement wages is informative of frictional wage dispersion. Furthermore, the fraction of displaced workers who suffer a wage loss is informative of frictional wage growth and job-to-job mobility, independent of the job-offer distribution and other labor-market parameters. Applying our methodology to US data, we find that search frictions account for less than 20% of wage dispersion. In addition, we estimate that between 40 to 80% of workers experience no frictional wage growth during an employment spell. Our approach allows us to estimate how frictions change over time. We find that frictional wage dispersion has declined substantially since 1980 and that frictional wage growth, while low, is more important toward the end of expansionary periods. We finish by estimating two versions of a random search model to show how at least two different mechanisms--involuntary job transitions or compensating differentials--can reconcile our results with the job-to-job mobility seen in the data. Regardless of the mechanism, the estimated models show that frictional wage growth accounts for about 15% of life-cycle wage growth.

We document gender differences in the booking of business air travel among similar workers within a firm. Women pay consistently less per ticket than men after accounting for a large set of covariates. A large proportion of the lower fares paid are explained by women booking earlier. We find that gender differences increase with age but find no deviation from this trend during the childbearing years. We also find that country-level gender differences in reciprocity are associated with the documented gender differences. The documented gender differences have important monetary implications for firms and suggest an important role for workers' morale.

Leveraging data from Sweden and Chicago, we study the educational pipeline for STEM and economics majors to better understand the determinants of the gender gap, and when these determinants arise. We present three findings. First, females are less likely to select STEM courses in high school, despite equal or better preparation. Second, there are important gender differences in preferences and beliefs, even conditional on ability. Third, early differences in preferences and beliefs explain more of the gaps in high school sorting than other candidate variables. High school sorting then explains a large portion of the gender difference in college majors.

This article uses networks to study price dispersion in seller-buyer markets where buyers with unit demand interact with multiple, but not all, sellers; and buyers and sellers compete on prices after they meet. The central finding of this article is that price dispersion is determined by the structure of the network. First, for any given network we characterize the pairwise stable matchings and the prices that support them. Second, we characterize the set of all graphs where price dispersion is precluded. Third, we use a theorem from Frieze (1985) to show that the graphs where price dispersion is precluded arise asymptotically with probability one in random Poisson networks, even as the probability of each individual link goes to zero. Finally, we calibrate our model to the documented price dispersion at the online trading platform eBay and show how counterfactual network structures at eBay would substantially decrease price dispersion.

This paper estimates returns to education using a dynamic model of educational choice that synthesizes approaches in the structural dynamic discrete choice literature with approaches used in the reduced form treatment effect literature. It is an empirically robust middle ground between the two approaches which estimates economically interpretable and policy-relevant dynamic treatment effects that account for heterogeneity in cognitive and non-cognitive skills and the continuation values of educational choices. Graduating college is not a wise choice for all. Ability bias is a major component of observed educational differentials. For some, there are substantial causal effects of education at all stages of schooling.

This paper analyzes the non-market benefits of education and ability. Using a dynamic model of educational choice we estimate returns to education that account for selection bias and sorting on gains. We investigate a range of non-market outcomes including incarceration, mental health, voter participation, trust, and participation in welfare. We find distinct patterns of returns that depend on the levels of schooling and ability. Unlike the monetary benefits of education, the benefits to education for many non-market outcomes are greater for low-ability persons. College graduation decreases welfare use, lowers depression, and raises self-esteem more for less-able individuals.

This paper develops robust models for estimating and interpreting treatment effects arising from both ordered and unordered multistage decision problems. Identification is secured through instrumental variables and/or conditional independence (matching) assumptions. We decompose treatment effects into direct effects and continuation values associated with moving to the next stage of a decision problem. Using our framework, we decompose the IV estimator, showing that IV generally does not estimate economically interpretable or policy relevant parameters in prototypical dynamic discrete choice models, unless policy variables are instruments. Continuation values are an empirically important component of estimated total treatment effects of education. We use our analysis to estimate the components of what LATE estimates in a dynamic discrete choice model.

Peters and Severinov (2006) (PS henceforth) characterize a perfect Bayesian equilibrium (PBE) in a competing auctions environment, where all buyers are linked to all the sellers. PS characterize a PBE using a simple bidding rule, whereby buyers select in which auction to bid. In this note we show that when buyers are linked with a subset of the sellers (i.e. when there are search frictions), the PS bidding rule is no longer guaranteed to be efficient nor a PBE of the competing auctions game of PS. Our results indicate that researchers should be cautious when using the PS bidding rule to make inference about the behavior of buyers and sellers in a market where frictions are present such as eBay.