Adverse Selection, Statistical Discrimination and Antitrust Law: the Case of Italian Automobile Optional Insurance

Abstract

Risk classification is common practise in the insurance market: on one hand, it can help to refine the premium schemes and improve the economic situation of the insurance companies; on the other hand, it is a traditional remedy for situations of adverse selection, as the economic theory has demonstrated. But the application of classification could be seen as altering competition in the marketplace when the statistical discrimination implies the adoption of the same price policies by many firms in the market. The paper tries to verify this affirmation in relation with the Italian market of the optional insurance against car thefts: in this liberalised market the insurance companies had applied risk classifications, but in 1994 the antitrust authority identified a cartel with reference to the exchange of statistical data to define a common statistical discrimination between the main insurance companies. In the final part of the paper, a model estimated using data about Italian insurance companies confirms the presence of adverse selection problems. This result could support the defence of the insurance companies based on the use of statistical discrimination as a remedy against informational problems. But, giving the same result also for the years after the antitrust authority sentence and the prohibition to the application of statistical discrimination, the result of the model could also be seen as a demonstration of the fact that in this case statistical discrimination was not a good tool against adverse selection problem.