Research

Working Papers


Hospital Queues, Patient Health and Labor Supply (with Anna Godøy, Venke F Haaland and Mark Votruba).
Conditional Accept, American Economic Journal: Economic Policy.
Abstract: Long waits for health care raise concerns about the consequences of delayed treatment. We use variation in queue congestion to estimate effects of wait time for orthopedic surgery. We do not find that longer wait times lead to worse health outcomes. We do find persistent reductions in labor supply: long waits increase medium to long-term work absences and permanent disability receipt. The effect is driven by individuals who are already on sick leave at referral. Our results are consistent with patterns of state dependence, where extended periods of temporary disability while awaiting treatment create persistent barriers to returning to work.

Selection in Surveys: Using Randomized Incentives to Detect and Account for Nonresponse Bias (with Deniz Dutz, Santiago Lacouture, Magne Mogstad, Alex Torgovitsky, and Winnie van Dijk)
Abstract: We show how to use randomized participation incentives to test and account for nonresponse bias in surveys. We first use data from a survey about labor market conditions, linked to full-population administrative data, to provide evidence of large differences in labor market outcomes between participants and nonparticipants, differences which would not be observable to an analyst who only has access to the survey data. These differences persist even after correcting for observable characteristics, raising concerns about nonresponse bias in survey responses. We then use the randomized incentives in our survey to directly test for nonresponse bias, and find strong evidence that the bias is substantial. Next, we apply a range of existing methods that account for nonresponse bias and find they produce bounds (or point estimates) that are either wide or far from the ground truth. We investigate the failure of these methods by taking a closer look at the determinants of participation, finding that the composition of participants changes in opposite directions in response to incentives and reminder emails. We develop a model of participation that allows for two dimensions of unobserved heterogeneity in the participation decision. Applying the model to our data produces bounds (or point estimates) that are narrower and closer to the ground truth than the other methods. Our results highlight the benefits of including randomized participation incentives in surveys. Both the testing procedure and the methods for bias adjustment may be attractive tools for researchers who are able to embed randomized incentives into their survey.


Event Studies, Endogenous Timing and the Child Penalty (with Simon Bensnes and Edwin Leuven)
Estimates of the child penalty, the effect of children on mothers’ labor market outcomes, are large and persistent. They are typically based on event-study designs which assume that fertility timing is (conditionally) exogenous. We investigate the validity of this assumption in a population of women who undergo fertility treatment through in-vitro fertilization. Standard event study estimates for these women uncover long-run child penalties on earnings in the neighborhood of 20 percent, in line with the literature (f.e. Kleven et al., 2019b; Angelov et al., 2016). In contrast, estimates of fertility effects based on the instrumental variable approach of Lundborg et al., 2017 that exploits IVF, are zero two years after the IVF attempt. We reconcile these approaches using a 2SLS event-study specification that centers time on birth (as opposed to time of the IVF trial in Lundborg et al., 2017). We estimate long-run child penalties of around two percent. Most of the difference between the standard event-study and the 2SLS event-study estimates is explained by adjusting for the timing of the fertility attempt. We find that women have their first child when their wage profiles tend to flatten. Consequently, a standard event study overestimates the counterfactual wages as women who have children later are on wage profiles that continue to grow. This type of selection can invalidate event-studies even when pre-trends are the same. We finally disentangle extensive and intensive margin fertility effects by exploiting additional IVF attempts. These results show that the child penalty of the first child vanishes after just two years once we net out the effect of the second child.

Work in Progress


Publications / Accepted Papers