Research

“Dying and Dissaving” (Job Market Paper)

Does a sudden drop in survival probability make people adjust saving? Standard life cycle theory suggests it would. Using Swedish administrative data, I link precisely measured negative health shocks to subsequent saving behavior. In this paper, I assess the quantitative impact of survival probability on the consumption-saving decision. By exploiting exogenous variation in the timing of health shocks, I also contribute to understanding the causal effect of health on wealth. I find that annual saving falls by 5 percent in response to a one standard deviation fall in survival probability. I find significant heterogeneity in the effect across the pre-diagnosis wealth distribution, but not across the pre-diagnosis income distribution. The rich dissave more whereas the poor increase saving in response to lower survival probability. People dissave less if they have a spouse whereas the presence of children has no significant effect.

“Cost of Loans and Moral Hazard” (with Y. Qi)

We document the effects of higher borrowing cost on private firms in the presence of financial frictions by exploiting a natural experiment and a unique and comprehensive dataset from Sweden. In June 2010, the central bank of Sweden increased the repo rate unexpectedly and exposed firms with long term loan maturing in the first and second half of the year to materially different cost of borrowing. After controlling for demand, we document the negative effects of debt overhang. We also find evidence of risk shifting among financially distressed firms.

“The impact of workplace gender composition on gender gaps in educational and labor market outcomes” (with E. Ranehill and A. Sandberg)

Are women’s performance and well-being adversely influenced by being in numerical minority in professional settings? We use Swedish register data on more than 100,000 PhD students, providing a unique opportunity to quantify the causal effect of peer group gender composition on subsequent outcomes. In particular, we use idiosyncratic variation in the gender composition of adjacent PhD cohorts to explore whether an increased share of male peers affects female academic achievement, health, and labor market outcomes negatively. Our results are relevant for both vertical and horizontal gender segregation and should be of interest to policy makers and employers aiming to increase female representation at top positions, and in male dominated occupations.