Nanyang Business School Forum on Risk Management and Insurance
Adverse Selection: Evidence form Commercial Surgery and Hospitalization Health Insurance
Asymmetric information has been a very crucial issue in the insurance literature in the last two decades. Many studies examine the problem of asymmetric information in various insurance markets such as automobile insurance (e.g., Puelz and Snow, 1994; Richaudeau, 1999; Chiappori and Salanie, 2000; Dionne et al., 2001; Cohen, 2005; Saito, 2006; Cohen and Einav, 2007; Kim et al., 2009; Gao and Wang, 2011; Zavadil, 2015 and Gao et al., 2016), medigap insurance (Fang and Keane, 2008), annuities insurance (Fong, 2002; Finkelstein and Poterba, 2004), life insurance (Cawley and Philipson, 1999; He, 2009; McCarthy and Mitchell 2010), long-term care insurance (Finkelstein and McGarry, 2006), and cancer insurance (Wang et al., 2011).
This paper examines adverse selection using a sample of commercial surgery and hospitalization health insurance policies from a Taiwan insurance company. Hofmann and Browne (2013) suggest that the topic of health insurance is more important than life insurance to examine asymmetric information because insureds accumulate funding capital to compensate for higher expected health expenditures to purchase health insurance rather than life insurance. Whether adverse selection exists has been a major research question for decades and there are no conclusive answers yet.
Our sample consists of 532,849 surgery and hospitalization insurance policies. The empirical results strongly support the existence of adverse selection. We find that a significant and positive relation between coverage and occurrence of claim and frequency of claim, indicating insureds with high-coverage are more likely to have high occurrence of claim and frequencies of claim. Our evidence also shows that insureds with high-coverage are more likely to have higher claim payments (number of days of hospitalization) than insureds with lower-coverage. In other words, adverse selection exists. The above result also coincides with the findings in literature, such as Wang et al. (2011) that evidence of adverse selection exists in the cancer insurance market through the survey data. We also find the requirement of physical examination can help to alleviate the severity of adverse selection. The evidence of interaction effects between physical examination and high coverage shows that insureds with high-coverage who go through physical examinations have less claim payments than those who do not go through physical examination. This evidence indicates that physical examination can mitigate adverse selection.
This study contributes to the literature in several ways. First, the study is the first to examine adverse selection in commercial surgery and hospitalization insurance. We believe using commercial surgical hospitalization policies has an important advantage. In general, policyholders needs doctors’ approval to go through surgery and hospitalization. It will reduce the possibility of moral hazard. Second, we have the information about whether policyholders go through physical examinations. With this data, it allows us to further test adverse selection and ways to control it. Third, we have complete underwriting and claim data such as those of insured with physical examination, surgical items, hospitalization, and benefits. Forth, our data contains gender, age, and city. Our study controls these variables. Finally, this paper offers significant and practical contributions for the life insurance industry. Life insurance companies need to charge enough premiums considering adverse selection.
The complete paper is available for download https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3236547