Asymmetric Information with multiple risks: the case of the Chilean Private Health Insurance Market

Abstract Since Rothschild and Stiglitz (1976) that asymmetric information is considered a potential setback in insurance markets, particularly so in health insurance, due to adverse selection. The recent literature establishes that failure to find evidence of adverse selection based on traditional tests may be due to its multidimensionality. Differently from the received literature, the source of the multidimensionality that we analyze here is based on the presence of more than one source of risk. We propose a test of asymmetric information, in the vein of Chiappori and Salanié (2000), that accomodates multi-risk and multi-coverage and apply it to a rich dataset from the Chilean private health insurance market. We start by extending the competitive separating equilibrium of RS to allow for two sources of risk, namely the risks of inpatient and outpatient care. This extension is important for at least three reasons: a) individuals may have different risk profiles along these dimensions such that risks may not be ranked; b) most insurance contracts have different coverages along these two dimensions of care; c) there may be asymmetric information regarding one risk dimension and not the other. In our application to the Chilean private health insurance market, we find a surprising result, that while there is evidence of Adverse Selection regarding inpatient risk, there is no such evidence for outpatient risk.