About me

I am interested in looking at old questions from new angles. For example, my job market paper studies the gender pay gap by asking how much of the difference in women’s pay relative to men’s is explained by differences in their productivity that may not be captured by easily observable characteristics. Using Danish data which links information about employees to their firm’s accounts, I find that productivity differences between men and women explain about 12 percentage points of the 16 percentage point gap in wages. The decoupling of pay and productivity happens only for non-mothers---mothers are 20 percent less productive than men and are paid twenty percent less than men. In contrast, non-mothers are as productive as men but are paid 12 percent less than men. Since only young non-mothers are underpaid, it is likely the case that employers are factoring the future probability that women will have children into their current salaries.

A question I ask in ongoing work is: what are the externalities of maternity leave on firms and co-workers? While a large literature is devoted to understanding the impact of maternity leave on children’s outcomes and the careers of women, less is known about the consequences of maternity leave at the workplace. I study spillover effects of maternity leave on co-workers and firms by examining a Danish reform that increased the length of parental leave in 2002. The policy change was unanticipated and retroactively applied, an ideal setting for a regression discontinuity study. While maternity leave length increased on average by one month for women giving birth in late 2001 compared to early 2002, there were no detectable effects of the reform on the wages or promotions of co-workers in any of the five years after the reform. In addition, there was no impact of additional maternity leave on firm value added or firm size. There is some evidence however that firm shutdown was more likely due to the extra maternity leave induced by the reform. Overall, the results suggest that moderate increases in the length of maternity leave should not affect co-worker wages or promotions and have limited (if any) affect on firms.

In earlier work I studied the well-examined question “what are the costs of job loss?” by looking at the behavior of workers who eventually become laid-off in the years before a layoff. Do these workers save before they are laid-off? Are they able to smooth the costs of job loss? Under a number of specifications, the data reveal higher savings in the years before a mass layoff among displaced workers compared to those in the same firm who keep their job, consistent with worker foresight about the idiosyncratic probability of losing their job. Knowledge of the idiosyncratic probability of future unemployment has important policy implications. For example, a recent paper by Nathan Hendren argues that private knowledge about layoff risk limits the amount of private insurance workers could buy and changes the nature unemployment insurance to insurance against the knowledge of future unemployment. Of course, knowledge of future unemployment also allows workers to better smooth consumption.

I have a number of other projects that study issues ranging from the demand for flexible jobs to household decision making under future preference uncertainty. You can find links to abstracts and papers in the left navigation panel.