Social Insurance Design and International Differences in Health at Older Ages

Differences in health outcomes and longevity across similarly developed countries are long-standing puzzles in public health. Studying these differences and identifying their causes and consequences is crucial for determining what reforms or changes may have the highest potential to improve health outcomes in the US and abroad. In this project we ask whether international differences in costs that individuals face for prescription pharmaceuticals is a quantitatively important channel contributing to the disparities in health outcomes and longevity that have been documented between the US and other OECD countries. We focus on a comparison between the US and Sweden. Our project takes advantage of a series of natural experiments in the Swedish social insurance system, coupled with administrative data of unmatched quality, to explore the empirical link between prescription drug coverage and outcomes for several groups of common chronic conditions. First, using regression discontinuity designs that rely on discrete changes in coverage of some prescription drugs and non-linear insurance contract schedules, we estimate how changes in out of pocket costs for certain prescription drugs affect the health outcomes of conditions that they treat. Importantly, we will estimate both local average effects and heterogeneity in the link between out of pocket costs and health outcomes along the socio-economic gradient. We will then use US and Swedish administrative insurance claims data to document differences in out of pocket costs of pharmaceuticals for a select set of conditions. With these estimates in hand, we will conduct an exploratory decomposition analysis of whether differences in out of pocket costs of drugs may be an important channel accounting for differences in health outcomes between the US and Sweden. This project will advance our understanding of the role of health insurance system design for international variations in disease prevalence and mortality – this role has important policy implications, but has so far been underexplored.