Automobile insurance started up as soon as automobiles first appeared on the roads. It became rapidly clear that this insurance needed special consideration due to the nature of road activity. More specifically, driving activities generate externalities and individual risks are difficult to evaluate and monitor since they are not directly observable.
Two information problems have been extensively discussed in economic and financial literatures since the beginning of the 1960’s: 1) moral hazard referring to the effect insurance contracting has on the insured incentives whose actions are not perfectly observable by the insurer and, 2) adverse selection which refers to the effect private information about individual’s type has on risks exchanges. These two problems have been used intensively as stylized facts to explain the existence of partial insurance coverage (deductible and coinsurance), risk classification, and bonus-malus.
However, very few empirical studies have been able to verify adequately the real significance of the two problems. This lack of results is explained by two factors: the non-availability of micro data and the difficulty of setting up the adequate methodologies. The first paper, by Chiappori, is about the methodological difficulties of isolating the significance of the two information problems in insurance markets. The objective is to review the empirical models published in the recent literature.
Some of the main theoretical issues are also discussed. One important conclusion of this chapter is that multiperiod data seem necessary to separate the effects of moral hazard from those of adverse selection. Another conclusion is that the results can be affected by the methods used. In particular, some wrong conclusions on the presence of adverse selection may be due to specification problems in the econometric modeling.
Read More About: f95zone
