For each of the types, we summarize the empirical evidence and discuss the policy implications that follow from it. The evidence for ex ante moral hazard (i.e., insurance-induced increases in risk-taking) is rather weak but suggests that people engage less in preventive behaviors that are costly and hard to maintain when they obtain insurance. The evidence for ex post moral hazard (i.e., insurance-induced increases in usage of insured services) overwhelmingly indicates that it exists. However, the exact size of this effect in the Dutch health care system remains to be empirically determined. The numbers on insurance fraud indicate that it poses a significant problem. Furthermore, insurance fraud is deemed acceptable and common by many policyholders, and most seem unaware of the nature of insurance. Despite the large amount of data that has been gathered and analyzed, policyholders often lack the knowledge required for accurately predicting what effect policy changes will have. An important implication that follows from this is that insurance companies need to collect more data and run policy-experiments to gain the knowledge they need.
- file129317