Ex Ante Moral Hazard and Anticipatory Behaviour: Some Evidence

Controversy exists over whether health insurance reduces the individual incentives to invest in prevention activities; there is no consensus on the existence of ex ante moral hazard (EAMH). Past evidence shows that insurance seems to reduce investment in secondary prevention (e.g. check-ups), but have not supported the hypothesis of EAMH in the case of primary prevention (e.g. healthy lifestyles). In this paper, we first review the theoretical predictions in the case of primary and secondary preventions, and their main empirical evidence in the past literature. Then, we extend the general EAMH framework by assuming that healthy lifestyles reduce the probability of illness only in future periods. Assuming that there exists a payoff period before observing the lifestyle consequences, current preventive behaviour will be affected by anticipation of future insurance coverage. We call the effect “anticipatory behaviour” (AB). Therefore, it seems reasonable to test for the existence of EAMH some times before effectively receiving a health insurance. In the United States, Medicare is offered to almost all the population at age of 65. This exogenous variation in health insurance coverage is a natural experiment that allows us to test for EAMH and anticipatory changes in preventive activities. We use the nine waves of the US Health and Retirement Study (HRS) and test whether uninsured individuals change their lifestyle as they approach 65. Our estimates are based on double-robust approach that combine propensity score and regression methods and estimated for three different definitions of the uninsured group. Differences between the insured and uninsured in terms of physical activity and smoking behaviour are similar to the ones predicted by our model of AB. The effect on alcohol drinking behaviour is less clear but DR approach suggests rather the presence of pure EAMH.