Parties with better information may be tempted to exploit their information advantage at the expense of their trading partners; this possibility is called moral hazard. The taxi driver who takes the long route to jack up the fare, or the auto mechanic or the dentist who recommends unnecessary services—these are all examples of moral hazard.

No one wants to be ripped off but the problem of moral hazard runs deeper than a transfer of wealth from buyer to seller. Let’s focus on the example of the mechanic. Suppose the mechanic tells you that you need an engine overhaul when all your car really needs is a minor new part. Overhauling the engine isn’t just a ripoff, it’s a waste of time and resources—the economy is producing a good that no one actually wants or needs.

447

A second problem occurs when buyers, knowing that the mechanic might rip them off, decide not to buy any service at all. Frankly, we don’t always listen to our dentist, especially when our dentist recommends expensive treatments. Sometimes that is the right thing to do but some services really are needed even when they are expensive. Refusing service means there’s no possibility of a ripoff but also no possibility of mutual gain through trade.

The bottom line is that if we lived in a world of symmetric information, sometimes we would get less treatment and sometimes we would get more. Both of these deviations represent a cost of asymmetric information.

For most goods, asymmetric information isn’t a big deal. You may not know how the new item on the restaurant menu tastes but you buy it once, and if you like it, you buy it again, and if not, well, the cost of trying and sampling was low. In the case of auto mechanics, dentists, and surgeons—not to mention marriage partners (see later in this chapter!)—the decisions we must take are more expensive, longer-lasting, and more difficult to reverse, so we need to think carefully about asymmetric information.

There are two solutions to asymmetric information and moral hazard problems: either provide more information, thereby reducing the asymmetry, or reduce the incentive for the knowledgeable party to exploit their information advantage. Let’s examine each of these solutions.

The internet has played a big role in helping overcome asymmetric information problems by making it easier for buyers to pool their knowledge. On Angie’s List, for example, buyers can find reviews of local auto mechanics and plumbers. Yelp provides similar reviews for restaurants. Reviews of sellers on Amazon Marketplace help potential buyers judge which sellers are most likely to deliver on their promises. Reviews have two advantages. First, they make it easier to avoid shady mechanics. Second, they raise the cost to mechanics of exploiting their information advantage.

The shady mechanic always faces a risk: If buyers think that they have been ripped off, they won’t return with repeat business. Before the Internet, an upset buyer might have warned away a few friends and family. But with the Internet, an upset buyer can warn the world and that decreases the incentive for sellers to attempt a moral hazard ripoff.

More generally, the Internet has increased the value of having a good reputation. Two economists, Daniel Houser and John Wooders, found that sellers with good reputations on eBay were able to sell their goods for higher prices.2 The internet has also increased the cost of having a bad reputation, as students who post compromising pictures on Facebook may discover to their chagrin.

For all of their virtues, however, consumer review sites like Yelp overcome asymmetric information problems only imperfectly. We know some authors who ask their friends to post friendly and presumably non-objective reviews on Amazon. Some stores have hired fake reviewers to praise their product, or condemn the product of a rival. In response to these problems, Yelp even set up a sting operation to catch fakers. In other words, a new kind of asymmetric information problem is introduced into the market, namely assessing the honesty and accuracy of the reviews themselves.

448

The general problem is that the more important a rating becomes, the greater the incentive to fake that rating or manipulate it. U.S. News & World Report reviews and rates U.S. colleges and universities, and these ratings have a big influence on where students want to study. But where does U.S. News and World Report get the data to rate universities? Often from the universities themselves! We won’t mention names (don’t worry, we’re sure your institution wouldn’t do anything wrong), but colleges have engaged in such practices as lying about the class rank and SAT scores of admitted freshmen, lying about student retention rates, lying about fund-raising success, and misrepresenting student–faculty ratios. Other times, the college does not lie but rather manipulates admission practices to achieve higher scores. One university offered financial incentives for admitted students to retake the SAT, so that the university could report a higher average SAT score among its students.3 Other colleges have delayed the admission of students with lower SAT scores so that these lower scores would not be counted when average SAT data were collected.

When we can’t rely on seller or buyer reviews, third-party organizations can sometimes be trusted to provide independent advice. The magazine and Web site Consumer Reports tests and rates a wide variety of products, from washing machines to cars to computers, not to mention bassinets and blood glucose meters—all to help their subscribers and paid-up users of their Web site make better informed decisions. To demonstrate its credibility, Consumer Reports does all of its own tests (it doesn’t rely on what the sellers claim) and it will not accept advertisements or other forms of payment from the companies whose products are being rated.

Organizations like Consumer Reports, however, are probably underprovided relative to their social usefulness. Many forms of information are a public good, which means that it is difficult to exclude nonpayers (and also that consumption is nonrival; see Chapter 19). Nonexcludability makes it hard to sell information even when the information is valuable. Let’s say, for instance, that an individual can pay for Consumer Reports or that she can go to Google and try to get that same information for free. Most people opt for the cheaper option, and indeed if you Google “best new cars” you will come across plenty of free information, some of it even drawn from Consumer Reports. Once the underlying research has been produced, it tends to make its way into the world, whether in someone else’s article, blog post, or tweet, and Google will bring us much of that information, again for free. The result is that many people free ride on all the testing and research that CR does to produce its product ratings. Since not everyone who benefits from CR’s research pays for it, the market for such information is smaller than ideal.

By the way, did you know that Consumer Reports forbids its ratings from being used in any product advertisements? CR argues that this prevents even the hint of impropriety. Notice that this policy also makes it harder to find CR ratings, except by buying Consumer Reports magazine. Thus, the no-advertising policy also helps CR to prevent free riders.

Since buyers are reluctant to pay for information, independent third-party reviewers often aren’t as independent as we would like. The bond-rating companies Standard and Poor’s and Moody’s evaluate the creditworthiness of companies and also of financial securities, including mortgage securities. The rating companies are supposed to help buyers pick safe securities on the basis of objective data. The recent track record of these companies, however, isn’t good. Right up until the financial crash of 2007–2008, for example, the credit rating companies were rating many mortgage-backed securities as very safe. A short time later, the largest wave of real estate defaults in the United States since the Great Depression made quite a few of these securities very bad investments. If you are wondering who paid for these ratings, it was the firms whose securities were being rated, not potential buyers. If credit rating companies are too critical, they may lose business from their true customers, the finance firms who want favorable reports about the securities that they are selling. Arguably, the rating companies were biased because they were trying to please those who paid them.

449

The ratings companies would have better incentives if they were paid directly by buyers, but again the buyers won’t pay because of free rider problems, as there are many buyers for any potential security but typically only one main issuer. In general, information problems are difficult to solve completely because damping down one problem often inflates another.

Better informed buyers are one solution to asymmetric information problems. A second solution is to give sellers less of an incentive to exploit their information advantage. We have already seen how reputation can make sellers less willing to exploit their information advantage. Splitting the selling of a service from the diagnosis of how much service is needed can also help to reduce moral hazard. Before signing the deed, for example, most house buyers will hire a house inspector, and indeed often a mortgage lender will require this. Since the house inspector is paid by the buyer, the inspector has no incentive to underreport bad news. In addition, it’s illegal for house inspectors to profit from any repairs that they recommend, so inspectors also do not have an incentive to overreport bad news.

Similarly, for major medical decisions, it makes sense to seek a second opinion. Not only will you get more information, but if you tell the second doctor that she will not be performing the service regardless, you are likely to get an opinion stripped of moral hazard.

Part of the original Hippocratic Oath, which physicians swear, reads, “Whatever houses I may visit, I will come for the benefit of the sick, remaining free of all intentional injustice [and] of all mischief.” The oath and the professional ethics and training that come with it probably do limit moral hazard. But as always, economics is about thinking on the margin, and even physicians appear to make choices that on the margin are shaded to their benefit.

Obstetricians are typically paid more when a baby is delivered via caesarean section, a surgical procedure, than by vaginal delivery. Are you surprised that when the fee for c-sections increases, so does the number of c-sections? Incentives matter. Interestingly, when the patient is herself a physician, and thus better informed than the average patient, we do not see this relationship, and that further indicates the importance of more symmetric information. A further test of the theory of moral hazard as applied to obstetricians comes from HMOs, or health maintenance organizations, which are one form of health care provider. HMOs pay their obstetricians a salary that is not dependent on the delivery method. As expected, patients in HMOs have lower c-section rates (relative to similar non-HMO patients). Don’t assume, however, that we can remove incentives completely. C-sections are also more common on Fridays, probably because physicians (and some patients!) would like to free up their weekends.4

450

There are even ways to overcome some of the problems with taxi drivers. If you don’t trust the standard cabs, try booking a vehicle through Uber or some of the other smart phone apps. You can specify the destination and arrange the fare in advance.

Instead of trying to eliminate a seller’s incentives, another method of reducing moral hazard is to better align the buyer’s and seller’s incentives. How do you know your lawyer is working hard on your case and not being lazy? Without a lot of expert knowledge and monitoring, it’s hard to know. To overcome the potential moral hazard problem, plaintiffs often pay their lawyers with a contingent fee—the lawyer is paid, typically a third of any judgment, only if the case is won. Real estate agents are also paid only if the house sells but in this case the contingent fee, typically 3% of the price of the house, is much smaller than a lawyer’s fee. As a result, real estate agents are more eager to sell houses quickly than are homeowners. If waiting an additional week raises the price by $1,000, that’s an extra $970 for the homeowner and only $30 more for the agent. Interestingly, when agents sell their own homes, they use their information advantage and keep their houses on the market about 10 days longer than similar houses sold by owners with agents.5

Finally, don’t think that moral hazard is limited to markets. Politicians typically have more information than voters and bureaucrats often have more information than politicians. A deep-seated problem of political science is how to incentivize politicians and bureaucrats to act in the public interest and not to use their information advantage for their own interest. Competitive elections, checks and balances, a free and independent press, and other institutions all can be understood as ways to limit moral hazard in politics (we take up more of these issues in Chapter 20).

Let’s be honest: Moral hazard is everywhere because self-interest is everywhere (even if you don’t think that self-interest is the only motivation). Recall Big Idea Two from Chapter 1: Good Institutions Align Self-Interest with the Social Interest. When it comes to moral hazard, reputation, ratings, and contract design, such as contingent fees, are some of the institutions that align self-interest with the social interest. Most importantly, note that these institutions are continually evolving in response to new challenges.

Let’s turn now to another problem of asymmetric information.