Dynamic Epidemiology: Spillovers from Limited Commitment and Private Information

Abstract

Can variations in immunization rates be accounted for by differences in information and commitment technologies? I document four salient facts regarding DPT vaccination rates: (i) lower immunization in countries with greater costs of contract enforcement, (ii) higher volatility in immunization in countries with larger informal sectors, (iii) less persistence in immunization relative to aggregate income, and (iv) negative skewness in the distribution of immunization over time. These patterns cannot be explained by efficient immunization in a frictionless economy. However, dynamic contracts subject to ex-post one-sided commitment and hidden income can rationalize these facts. This analysis shows that weak provision of public goods, such as the inefficacy of the judicial system and the degree of informality, can spillover to weak provision of preventive healthcare. A model estimated using U.S. data reveals that an income monitoring technology is welfare enhancing in the long-run, generating an increase of 4.27% in certainty equivalent consumption of the policyholder and 0.6% in the insurer’s surplus. Moreover, efficient risk outcomes estimated using output losses stemming from chronic diseases across U.S. states explain over 90% of ACA Prevention Fund allotments enacted by the CDC. For countries in which a large share of medical expenditure is incurred out-of-pocket, I propose an identification strategy that leverages national account data and labor statistics to construct a time-series for preventive healthcare expenditure. Using the consumption neutrality of the efficient risk choice, I also devise a test to show that the hypothesis of limited commitment cannot be rejected in the data.