Why insurance markets are a game of cat and mouse

Confusing, opaque and expensive. Buying and claiming on insurance policies can be a frustrating. It turns out, though, that insurance companies often feel the same way about their customers — a group of people whose inner secrets hold the key to their success or failure.

The reasons behind the mutual incomprehension lie at the heart of why many insurance markets either work badly, or do not work at all.

As the authors of Risky Business explain, insurance is what economists call a selection market — one where the identity of the customers is just as important as the price they are willing to pay. Some customers will be cheaper for the insurer to serve because they make few claims; others more expensive. The market only works if there is a mix of the two.

“The problem of selection introduces a two-sided game of cat and mouse in which insurers try to pick the right customers (and avoid the wrong ones), while the ‘wrong’ types of customers do all they can to get insurers to believe they’re actually the right ones,” the authors argue. The key problem, say Liran Einav, Amy Finkelstein and Ray Fisman, is that customers know far more about how risky they are than insurers do.

At its worst, selection can push some companies or entire segments of the industry out of business. If an insurance company finds customers are costlier to serve than it expected, it may raise prices. That could deter less risky customers from buying cover at all, leaving the insurer only with the more risky ones. And so the insurer’s costs rise again, and prices increase again, pushing yet more people out. Eventually, insurance either becomes too expensive for most people, or the insurance company goes bust. The authors furnish plenty of examples, from divorce insurance to unemployment cover.

The game of cat and mouse twists the insurance world in a thousand different ways as companies try to keep the market functioning well. For example, anyone with a company health insurance policy will be familiar with, and perhaps frustrated by, the rules that only allow changes to be made once a year. The authors point out that this is to stop people from buying cover as soon as they find out that they are ill. Similarly, throwing in free gym memberships is an attempt to weed out those people who do not like gyms — and so are perhaps unhealthier and more expensive for the insurer.

The authors, three US-based academics, keep the debate moving along with a chatty, breezy style, familiar to readers of Stephen Dubner and Steven Levitt’s Freakonomics books. It’s a book about insurance that doesn’t feel like a book about insurance. However, it is very much a book about US insurance. Examples from other parts of the world are lacking.

Where the book hits its stride is with some of the thornier problems insurance companies and their customers are beginning to face. Going back to health insurance, for example, the growing availability of genetic data leads to a new set of challenges as that information can be used to identify who is at higher risk of certain illnesses. Should governments allow insurers to use this information to price policies? If so, some people risk being excluded from the market because they lost the genetic lottery. But if insurers cannot use this information while their customers can, the market will be twisted in the other direction.

There are no easy answers here, and the authors do not try to offer any. There are, they say, only trade-offs. “Whatever balance between efficiency and fairness the government chooses, there will be winners and there will be losers,” they argue. “The losers will often have genuinely tragic stories to tell.”

These types of questions will become more common as insurers gather an ever wider range of data about their customers. The data may tell them, for example, that people with fair hair are more likely to drive too fast. Or that journalists who write book reviews are statistically more likely to have their house flooded. The information advantage could shift from the customer to the insurer, and the authors are a little too dismissive of the potential for big data to disrupt the market. But that potential is there. And it will not necessarily make insurance a less confusing, opaque or expensive place.

Risky Business: Why Insurance Markets Fail and What to Do About It by Amy Finkelstein, Liran Einav, and Ray Fisman, Yale University Press, $30, 280 pages