Working Papers
"The Impact and Incidence of Supplemental Health Insurance: Evidence from Sweden" (Email for current draft) with Mårten Palme, Johannes Spinnewijn and David Seim
Abstract This paper studies the prominent, yet under-researched role of supplemental health insurance (SHI) in universal healthcare systems committed to equal access to care. Linking novel data on SHI contracts to rich administrative data from Sweden, we show that SHI enrollment rises steeply with income, even though healthcare needs — as measured by mortality and health spending — are substantially higher among lower-income individuals. Exploiting variation in the timing of employer-sponsored SHI, we find that gaining coverage leads to large and persistent increases in healthcare utilization by about 25 percent. These effects reflect the relaxation of rationing in the public healthcare system, extend to downstream, high-value care, and are largest among lower-income individuals. Focusing on cancer patients, we further find that SHI is associated with earlier diagnosis and lower mortality. While offered by private insurers, SHI mostly increases publicly-funded healthcare and also shortens waiting times within the public system, generating both fiscal and congestion externalities. Overall, the results reveal a central tension in hybrid healthcare systems: supplemental health insurance improves access to care, yet uneven coverage distributes its benefits and costs unequally.
"Liquidity Effect in Non-Linear Health Insurance: Evidence from Sweden" (New draft coming soon )
Abstract This paper estimates liquidity effects in health care consumption. I use detailed administrative data on prescription drug purchases, monthly income, and liquid assets for the universe of Swedish Disability Insurance (DI) recipients to investigate if liquidity-constrained and unconstrained patients' demand for prescription drugs responds differently to changes in liquidity and health care prices. I first document that low-liquidity recipients time their prescription drug fills to paydays. Using price variation within the universal and uniform health insurance, I show that constrained households also reduced prescription purchases more in response to higher prices. I also provide evidence that low liquidity, but not low income in general, is associated with lower annual prescription drug purchases and higher hospitalization rates. These results have implications for health insurance design and welfare evaluation. If demand responses to subsidized health care prices are partly due to positive liquidity effects, previously documented welfare losses from providing generous insurance may be overstated.
"Does it matter who cares? Formal vs. informal care of the elderly" (New draft coming soon ) with Patrizia Massner
Abstract Rapidly aging populations in many countries present challenges not only for social welfare systems but also for younger generations. In this paper, we estimate the effects of subsidizing formal elderly care fees on seniors' utilization of formal care services, seniors' health outcomes, and the labor supply of their potential caregivers -- their children. Leveraging a Swedish reform in a difference-in-differences design, we find that decreasing the fee for formal elderly care is associated with both reduced healthcare utilization among the elderly and increased labor supply of their children. Estimating the welfare implications of the reform, we show that subsidizing elderly care is costly for the government in the short run but becomes self-financing within a decade after the implementation.
Presented at the 2023 NBER Summer Institute (Aging) and the 2023 Nordic Public Policy Symposium. Rewarded MIT Policy Impacts Early Career Scholar Grant 2023.
Abstract Can duration-dependent eligibility criteria - increasing the strictness for eligibility over time - in Short-Term Dis- ability Insurance (STDI) reduce STDI duration while improving long-run labor market outcomes? To answer this question, I leverage a reform in Sweden that introduced duration-dependent eligibility criteria for STDI. The results show that individuals who get screened out by the stricter eligibility criteria have lower labor force participation dur- ing the following four years after the STDI spell. Conversely, the probability of being dependent on unemployment insurance or being non-employed increases. Calculating the net effect on the government’s budget for the STDI recipients who are screened out because of the stricter eligibility criteria suggests large savings in the short run, but net losses after four years. The findings suggest that policymakers need to carefully consider the potential trade-offs between the direct short-run fiscal savings from policies aiming to shorten STDI duration and the potential long-run costs from unintended changes in labor supply and benefit substitution when designing eligibility criteria for STDI.
Work in progress
"Gatekeeping or Demand Side Incentive in Health Insurance: Evidence from Sweden" with Gustav Kjellsson
"Inequality in adherence to medication for chronic conditions", with Mårten Palme anc Gustav Kjellsson
"Mental Health and Employers: Evidence from Firm-to-Firm Movers"