An economic and experimental analysis of insurance fraud and auditing.

Summary This work proposes an experimental approach to insurance fraud. We first present the two main paradigms that have been developed in the economic literature to analyze insurance fraud from a theoretical point of view: costly verification mechanisms and costly falsification mechanisms. According to the costly verification hypothesis, the insurer verifies claims by incurring an audit cost. This procedure can be deterministic or random. Under the falsification assumption, the insured incurs some cost to make the audit imperfect. The insurer is therefore unable to detect fraud with certainty. In parallel to this modeling, the experimentation carried out in these contexts confirms that the random audit dominates the deterministic audit and that the application of a severe penalty can deter fraud, mitigate the extent of falsification and compensate for the poor quality of the audit.