Preface

This book looks at the behavior of individuals at risk, insurance industry decision makers, and policy makers at the local, state, and federal levels involved in the selling, buying, and regulating of insurance. It compares their actions to those predicted by benchmark models of choice derived from classical economic theory. Where actual choices stray from predictions, the behavior is considered to be anomalous. We attempt to understand why these anomalies sometimes occur and sometimes do not, in many cases using insights from behavioral economics. We then consider if and how such behavioral anomalies could be modified.

This book is in no way a defense of the insurance industry nor an attack on it. Neither is it a consumer guide to purchasing insurance, although we believe that consumers will benefit from the insights it contains. Rather, we describe in this book situations in which public policy and the insurance industry’s collective posture need to change. This may require incentives, rules, and institutions that will help reduce inefficient and anomalous behaviors and encourage behavior that will improve individual and social welfare.

The benchmark model of supply assumes that competitive insurance firms know the loss probabilities and outcomes of the risks they are insuring against and base their premiums on this information. Furthermore, they are able to effortlessly change their premiums to reflect updates in their estimates of the risk. Insurance firms have access to the capital markets (at competitive interest rates) for any needed funds, even after experiencing a large loss. Investors who supply capital to insurers hold diversified portfolios. Losses are independent of one another, so that the law of large numbers minimizes the likelihood of an unexpectedly large total loss and makes it very unlikely an insurer will have to declare insolvency.

In that ideal world, insurance firms are also assumed to have accurate information on the risks of their customers and choose actions that maximize their expected profits. Under this model, firms should be willing to supply virtually any amount of insurance that buyers are likely to find attractive. The premiums they would charge are just high enough to cover their expected claims, including loading costs, which yield a rate of return on capital that investors could have earned elsewhere in the private market. The supply curve of insurance would be virtually horizontal because insurance uses only a tiny fraction of the global capital pool. In other words, the price of coverage should be largely unaffected by variations in the demand for insurance coverage or insurer demand for additional capital.

The classical approach to economics tends to elegance and simplicity, as we saw in the benchmark models of supply and demand. The outcome of a competitive insurance market generally means resources are allocated efficiently and outputs are produced at lowest cost; consumer welfare is maximized, given the resources available to the economy. The only reason for intervention by the public sector is to correct any inequities from the resulting premiums, such as providing some type of subsidy to low-income residents currently residing in hazard-prone areas who cannot afford homeowners’ insurance or to low-income households that might fail to buy health insurance without financial assistance.

But the real world is a considerably messier place. Individuals have difficulty understanding the purposes and concepts of insurance; firms often do not provide coverage at premiums that reflect risk. This chapter presents and analyzes anecdotal evidence of unusual insurance behavior reported by the media. Few of these journalistic examples use benchmark economic models to measure alleged mistakes by consumers considering buying insurance or insurers determining whether to offer coverage against a particular risk.

FOUR EXAMPLES OF INSURANCE BEHAVIOR IN PRACTICE

When Judy’s sister discovered she had cancer, Judy was frightened. Their mother had died of cancer and Judy became preoccupied with the possibility that a genetic link might endanger her, too. Not long after her sister’s diagnosis, Judy bought an expensive insurance policy that promised to pay her money on top of any medical benefits if she got cancer (but not if she contracted another disease).

Doug’s new car, an expensive Lexus sports car, was his dream come true. He maintained it meticulously and, thinking ahead, took out an insurance policy on the car with a very low $500 deductible, even though that increased his premium considerably. That way, he figured, repairs to his car would be nearly covered even if someone in a parking lot simply dented his fender. Three months later, Doug’s attention momentarily lapsed when his cell phone rang and he plowed into the car ahead of him at a stoplight. The body shop estimated damages to the Lexus at $1,500, well over Doug’s $500 deductible. Yet Doug chose to pay the entire cost out of pocket, fearful that making a claim would drive his insurance premiums even higher.

The benchmark model of demand developed in Chapter 2, based on expected utility theory, postulates a world in which the collection and processing of relevant information is costless to consumers, risk is perceived accurately, and the individual is assumed capable of choosing the amount of insurance that maximizes his or her expected utility. As long as people are risk averse, they are willing to pay a premium greater than the expected value of losses from a set of prespecified risky events. The maximum amount an individual is willing to pay for a given level of coverage depends on that individual’s degree of risk aversion. The optimal amount of coverage is determined by comparing the benefits of more financial protection should a disaster occur with the additional premiums for purchasing this additional coverage.

In the last chapter, we explored extensions to the benchmark model of demand that introduced imperfect information and search costs, but maintained the assumption that individuals choose options that maximize their expected utility. This chapter further relaxes some of the benchmark model assumptions and explores different theories of choice and behavior under risk. As we will show, commonly observed behaviors inconsistent with expected utility maximization can be explained by other theories supported by data from experiments, field studies, and consumers’ actual insurance-related decisions.

This chapter focuses on empirical examples of specific types of anomalies by insurers and those who supply capital to insurers. They may be caused partially by some of the concerns of insurance firm managers discussed in the previous chapter, such as fear of insolvency and the impact that this would have on their future job prospects. They may also be due to the effect rating agencies and regulators may have on the premiums insurers can or want to charge and the reserves they are forced to hold. Other deviations from benchmark supply behavior may be a result of the decision processes and heuristics used by insurance managers.

We first consider specific examples of insurer firm and manager behavior that do not adhere to the benchmark supply model and attempt to identify which, if any, of these behaviors can be classified as anomalies. We then consider anomalous behavior on the part of investors and other capital suppliers to insurance firms that create supply problems for insurers.

DECISION TO OFFER OR NOT OFFER TERRORISM COVERAGE

Insurers, like everyone else, have difficulty dealing with the uncertainty associated with terrorism. The likelihood of an attack is highly ambiguous and the actions taken by terrorists may change depending upon what protective measures are undertaken by those at risk. This latter feature distinguishes terrorism risk from other low-probability, high-consequence risks, such as hurricanes and other natural catastrophes where nature does not try to outwit the adoption of preventive measures and where catastrophe models using scientific and engineering data can help insurers determine premiums.