Academic journal article Journal of Risk and Insurance

An Empirical Analysis of the Effects of Increasing Deductibles on Moral Hazard

Academic journal article Journal of Risk and Insurance

An Empirical Analysis of the Effects of Increasing Deductibles on Moral Hazard

Article excerpt

ABSTRACT

Using information on timing and number of claims in a unique data set pertaining to comprehensive automobile insurance with the increasing deductible provision in Taiwan, the authors provide new evidence for moral hazard. Time-varying correlations between the choice of the insurance coverage and claim occurrence are significantly positive and exhibit a smirk pattern across policy months. This empirical finding supports the existence of asymmetric information. A subsample estimation depicts insured drivers' significant responses to increasing deductibles, which implies the existence of moral hazard. According to the probit regression results, the increasing deductible makes policyholders who have ever filed claims less likely to file additional claims later in the policy year. The empirical findings strongly support the notion that the increasing deductible provision helps control moral hazard.

INTRODUCTION

Rothschild and Stiglitz (1976) and Shavell (1979) pioneered work on adverse selection and moral hazard for insurance and inspired many theoretical studies (1) during the past three decades. The theoretical literature has identified many insightful mechanisms for understanding and controlling adverse selection and moral hazard in the insurance market, and more recent papers (2) have conducted empirical analyses to investigate whether adverse selection and moral hazard exist in the insurance market.

However, empirical evidence in the insurance literature has not yet provided a consensus regarding whether asymmetric information exists. Some research, such as that by Puelz and Snow (1994), Dionne and Gagne (2002), Fong (2002), and Finkelstein and Poterba (2004), finds evidence to support the existence of asymmetric information in the insurance market, whereas other research, such as Dionne, Gourieroux, and Vanasse (2001), Cawley and Philipson (1999), and Chiappori and Salanie (2000), uncovers no such evidence.

To determine whether insurance markets have asymmetric information problems, most researchers test the correlation between the choice of insurance coverage and claim occurrence. Correlation tests were simultaneously suggested by Chiappori and Salanie (with a nonparametric approach, 2000) as well as Dionne, Gourieroux, and Vanasse (with a parametric approach, 2001). The literature has been most interested in two types of asymmetric information: adverse selection and moral hazard. If the adverse selection problem exists in the insurance market as described by Rothschild and Stiglitz (1976), then the market could settle at a separating equilibrium in which the high risk purchase higher coverage whereas the low risk purchase lower coverage. Thus, the higher is the loss probability, the higher coverage the insured tend to buy. On the other hand, if the moral hazard problem does exist in the insurance market as shown by Shavell (1979), then the individual with higher coverage may have less incentive to avoid the occurrence of the risk. Therefore, the loss probability of the individual with higher coverage could be higher than that of the individual with lower coverage. Since both moral hazard and adverse selection predict the occurrence of the risk and the coverage of the insurance are positive correlated, a positive correlation implies adverse selection, moral hazard, or both.

Exploring the existence of asymmetric information in insurance markets is important, but it is at least as important to distinguish between adverse selection and moral hazard because the literature has developed alternative mechanisms to solve the adverse selection and moral hazard problems. In this regard, we note that Dionne and Gagne (2002) find that a replacement cost endorsement in an automobile insurance policy increases the probability of theft near the end of the protection. They interpret this result as an ex post moral hazard or opportunistic fraud after ruling out ex ante moral hazard and adverse selection. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.