Ray Fisman’s New Book Examines Insurance Market—and its Flaws
Can insurance pay for college? Cover the expensively sick? How much do you need on a car? New book examines market—and its flaws

“Allstate cares not just how many auto policies it sells, but also who buys their coverage,” says BU economist Ray Fisman. “Careless drivers are much more expensive to have as customers than conscientious ones.” Photo by RobertCrum/iStock
BU Economist Ray Fisman Finds Insurance Fascinating. Really.
Can insurance pay for college? Cover the expensively sick? How much do you need on a car? New book examines market—and its flaws
Cruel and unusual punishment—at least in Woody Allen’s heist comedy Take the Money and Run—is confinement with an insurance salesman. But to Boston University economist Ray Fisman, such a tête-à-tête might be paradise. He’s cowritten a new book arguing that insurers and insurance “really are much more interesting than you think.”
He might even make you feel sorry for them.
Risky Business: Why Insurance Markets Fail and What to Do About It, coauthored with two economists from Stanford University and MIT, focuses on the ever-present, existential threat to insurers: the problem of selection. “Insurance sellers care not just about how much they sell but whom they are selling to. They care, in other words, about which customers they select,” the authors write. “And why do they care? Because some customers are much more expensive to insure than others.”

That reality creates real-life dilemmas. The book recalls how, in the 1980s, perfectly healthy men couldn’t buy health insurance in West Hollywood—though they could anywhere else in Los Angeles County—because that neighborhood was home to many gay men, who were at risk for expensive and then-untreatable AIDS. Nor is public insurance immune. Pre-Obamacare, when some states tried to insure high-risk residents, their efforts collapsed; healthy people skipped enrolling in the state plans, leaving a pool of sick residents that was too small and medically expensive to manage without charging exorbitant premiums. That’s why Obamacare required all people, healthy and sick, to have insurance.
Fisman, the BU College of Arts & Sciences Slater Family Professor in Behavioral Economics, and his coauthors examine other types of insurance, from automobile to dental to pet to insurance-like solutions for skyrocketing college costs. Policymakers can outlaw discrimination against the sick and others, the book notes, but they must be careful that such regulations don’t saddle insurers with selection-related costs that could bankrupt them. The Brink interviewed Fisman about his book and managing this “risky business.”
Q&A
with Ray Ray Fisman
The Brink: You offer different states’ experiences with mandating auto insurance as showing the perils of requiring too much coverage and too little. Can you summarize what problems ensue with each?
Fisman: Allstate cares not just how many auto policies it sells, but also who buys their coverage. Careless drivers are much more expensive to have as customers than conscientious ones. The problem, from Allstate’s perspective, is that the reckless motorists are willing to pay more for coverage than careful drivers for the simple and obvious reason that they’re more likely to crash their cars. This kind of selection can lead insurance markets to collapse completely.
One solution to this problem of “good” customers selecting out of the market is to mandate that everyone needs to participate. That way, selection problems can’t, by definition, wreck a market—everyone, whether a bad or good customer, has to participate. But that leads to the question of how much coverage you insist that people buy. Should we take a minimalist approach, like Florida, and require only $10,000 in personal injury protection? Well, for many drivers, that won’t be enough, so they’ll want more coverage. And who will be most eager to get more coverage? The bad drivers, of course—and we’re back to the problem of selection. And what if we take Maine’s approach and increase the personal injury insurance minimum to $50,000? Then we may be forcing lots of drivers to get more insurance than they want or need.
There isn’t a single right answer. As with most policies, it’s a question of tradeoffs, and clearly one for which Maine and Florida regulators have settled on very different solutions.
The Brink: The book says insurance-like solutions to expensive college costs—some organization fronting a student’s cost in exchange for a share of their income for a period after graduation—have never caught on large-scale. But hasn’t Australia successfully used a version of this to pay students’ tuition?
Fisman: Selection is a big problem for so-called income-share agreements (ISAs) as a way of paying for college. The basic idea is that if you make oodles of money when you graduate, you have to pay back more to the lender than if you get a lower-paying job. The selection issue is obvious; this is a more attractive proposition to students who know they will be starving artists or poets than it is for prospective investment bankers. It could help to, say, limit participants to economics and engineering majors. But there’s still going to be people who know they’re getting the degree to go to Wall Street and others who anticipate working for a nonprofit.
ISAs do exist. Purdue offers them as a financing option, the Australian government has an ISA program for all of its citizens, and a South American start-up, Lumni, offers ISAs in Colombia, Chile, Mexico, and Peru. But usually, as in the Australian case, they’re heavily subsidized, which is fine—it’s an alternative to traditional forms of financial aid that BU and other schools might use that are also subsidized. But attempts at turning income sharing into a profit-driven business have yet to succeed, and not for lack of trying.
The Brink: You note that public subsidies for insuring the young and healthy makes economic sense. Why?
Fisman: Because it’s the most efficient way of targeting the selection problem. We go through the argument in the book. In short, selection problems can lead to the following health insurance market “death spiral”: only less healthy people sign up for insurance, which makes the insurer’s costs really high. To cover the high costs, the insurer raises their prices. And that just makes the selection problem even worse, requiring yet-higher prices. In the end, prices are sky-high, if the market survives at all. This kind of thing actually happens.
Now, who is the first to select out of the market? Young and healthy people, who value coverage the least. A subsidy targeting this particular group can be an efficient way of short-circuiting the death spiral, and everyone can benefit from this. Since healthy people tend to be, on average, wealthy people, this ends up looking like a subsidy targeting the rich. Even if it’s good economics, it doesn’t play well—with good reason, perhaps—in the media, which makes it bad politics and can explain why it’s something we never observe in real-world policy.
Unfortunately, there is often this mismatch between what constitutes good economics and what makes it politically palatable, so we end up with policies that are popular, but end up destroying or distorting insurance markets.
The Brink: Don’t many nations with universal health insurance—public or private—insure the young and healthy successfully? Is the difficulty of balancing fairness and selection an argument for government health insurance—Medicare for All—to take this important choice off business’ back?
Fisman: I’m Canadian, and our universal health coverage is a point of national pride, as well as one of the ways we distinguish ourselves from the economic behemoth that lies to the south. So, I can’t help but be a fan of universal coverage, which does solve the selection problem: if everyone is put in the same healthcare pool, then by definition there are no selection problems.
However, the question of whether government should manage healthcare—rather than, say, a strict mandate that everyone get coverage—is one of those trillion-dollar policy questions that doesn’t lend itself to a glib response. Switzerland, for example, has what seems to be a well-functioning healthcare system based on private insurance coverage, which offers consumers choice of what insurance they buy. Canada, like many other countries, has at this point opted for a hybrid system—when I was growing up, there was only the public system—with some private clinics that require out-of-pocket payment or private insurance coverage.
All of these options have their strengths and weaknesses. One problem with policy discourse is there is this need to distill questions down to a sound bite with a clear conclusion. Unfortunately, most problems are more complicated than that.
This interview was edited for clarity and brevity.
Comments & Discussion
Boston University moderates comments to facilitate an informed, substantive, civil conversation. Abusive, profane, self-promotional, misleading, incoherent or off-topic comments will be rejected. Moderators are staffed during regular business hours (EST) and can only accept comments written in English. Statistics or facts must include a citation or a link to the citation.