Ex Post Moral Hazard in Private Health Insurance

This paper contributes to the discussion on possibilities to reveal ex post moral hazard in the Russian market for private health insurance. By 'ex post' we mean the period when a health insurance contract is valid. Moral hazard implies risky behavior of a respondent that increases health care utilization and/or decreases their incentives to prevent an insured event. In our empirical estimates, we explore the uniqueness of the Russian data that consists in the fact that many medical organizations provide services to respondents insured by enterprises. Adverse selection is hardly possible among such respondents. It gives us the possibility to observe ex post moral hazard, simply controlling for ex ante moral hazard by the use of individual fixed effects in panel data models. We use the RLMS-HSE data (2000-2015) for empirical estimates. We consider doctor visits, tobacco and alcohol consumption, physical exercises, and self-assessed health (SAH) as moral hazard indicators, estimating ordered choice regression models for each of the dependent variables mentioned above. To avoid inconsistency in estimates of parameters caused by the incidental parameter problem, we use the Blow-Up and Cluster (BUC) estimator. The results show a statistically significant increase in frequency of visits to the doctor and in alcohol consumption, as well as a decrease in SAH during the period of insurance. These results could be useful for insurance companies and could be accounted for in contracts for private health insurance.