Working Papers

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). Revision requested by the Review of Economic Studies.
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.

Reconciling Estimates of the Long-Term Earnings Effect of Fertility (with Simon Bensnes and Edwin Leuven)
Abstract: This paper provides methodological and empirical contributions to the child-penalty literature. We propose a new estimator that combines elements from standard event study and instrumental variable estimators, and show how these are related. We find that while all three approaches lead to similar conclusions about the impact of children on earnings differences between mothers and their partners – around 20 percent – they may lead to very different conclusions as to whether it is mothers or partners that drive this difference, the key policy issue. Where in the event study the complete impact comes from mothers, our findings suggest that only about one fourth of the impact is driven by maternal responses. The paper also has potential implications for event-study designs more generally. In particular, we validate event-study estimates using external information, and characterize bias. This provides new insights in the nature of selection into fertility, which shows that common intuitions regarding parallel trend assumptions may be misleading and that pre-trends may be uninformative about the sign of the selection bias in the treatment period.

Distributional Consequences of Patient Cost-sharing in a Single-payer Healthcare System  (with Simon Bensnes and Victoria Marone)
Abstract: Patient cost-sharing is used as a tool to limit over-utilization of insured healthcare services in almost all high-income countries. We study its (marginal) distributional consequences in the context of a publicly-funded universal health insurance system, where consumers (as tax-payers) are residual claimants on insurer spending. We highlight the critical distinction between consumers' level of demand versus their elasticity of demand for healthcare services, and estimate the relationship of each of these objects with socio-economic characteristics. Using detailed administrative data on the Norwegian national health insurance scheme, we study a 2010 policy change that raised the age threshold for copayment waivers, thereby increasing patient cost-sharing substantially for adolescents. We find that lower-income individuals have both higher average healthcare utilization as well as higher responsiveness to cost-sharing. Cost-sharing therefore places a larger burden on this group both in terms of the financial cost of out-of-pocket spending and in terms of reduced quantities of healthcare that these individuals choose to utilize.

Publications / Accepted Papers

Work in Progress