In recent years, testing for asymmetric information in insurance markets has gained much popularity. This leads to narrowing the gap between theory and empirical evidence. Empirical results also show directions for further theoretical developments. The theory of asymmetric information has been well understood for a long time (Akerlof ,1970; Rothschild and Stiglitz ,1976; Holmström ,1979; Shavell, 1979). The models for both phenomena, i.e. adverse selection and moral hazard, predict a positive correlation between risk and coverage. Although it is in general difficult to disentangle adverse selection from moral hazard, tests for asymmetric information as a whole are possible. While the theory has been highly developed, empirical studies have lagged behind. One reason is the scarcity of data sets in this field.
The aim of this project is both to test for asymmetric information in insurance markets and to develop new econometric methods. In Su and Spindler (2013, Journal of Business and Economic Statistics) a nonparametric test for asymmetric information is proposed and applied to both long-term care and automobile insurance. In a series of papers tests for asymmetric information are conducted in the German car insurance (Spindler, Winter and Hagmayer, 2013, Journal of Risk and Insurance), in the disability insurance (Spindler, 2013) and in the market for daily hospital benefits (Spindler, 2014, The Geneva Risk and Insurance Review).
