How effective is the fight against insurance fraud?

Summary This thesis proposes a study of insurance fraud. In particular, the analysis focuses on the means available to insurers to combat fraud. The objective of the first chapter is to show how the failure of insurers to engage in credible control measures contributes to the deterioration of the insurance system. In the context of non-commitment, we show that the strategic behavior of each party alters the insurance contract in unexpected ways. Two major results are obtained. Unlike Mookherjee and Png (1989), we show that the optimal insurance contract provides the same payoff whether the insured is controlled or not. Moreover, if the cost of the audit is too high, then the optimal insurance contract becomes a contract with over-insurance. In the second chapter, the audit is assumed to be imperfect. On the one hand, the insurer is not always able to detect fraud. On the other hand, there is an imperfection related to a wrong assessment of the damage by the insurer and thus a positive probability of unfairly sanctioning an honest insurer. The analysis reveals that the isolated effect of the first type of error does not significantly affect the shape of the optimal insurance contract. However, since the insured is adverse to the risk, the possibility of being punished by error introduces an additional risk, which may induce him not to declare his damage. This leads us to reconsider the different possible equilibria resulting from the strategic behaviors of the two parties. We then show that the possibility of being unjustly sanctioned solves the insurer's credibility problem by making the threat of the audit credible. The last chapter raises the question of the effectiveness of controls in the presence of policyholders who differ in their perception of the probability of being audited. The insurance market is a competitive market in the presence of two types of insureds, a first group that overestimates the probability of an audit, and a second group composed of insureds who underestimate it. The objective of this chapter is then to analyze how this heterogeneity in beliefs can modify the relationship between an insurer and its insureds. It appears that it may be optimal to dissuade only the most pessimistic policyholders, and let the other part of the individuals commit fraud in a systemic way.