Sawbones, cowboys and cheats

FIRST do no harm, doctors are wont to say. But some find it hard to sit on their hands. In a number of studies in the 1990s, Gianfranco Domenighetti, an economist at the Cantonal Health Office in Ticino, Switzerland, set out to discover whether surgeons performed more operations than were strictly necessary*. He and his colleagues found that the more sophisticated the patient, the less scalpel-happy the doctors. The best informed patients of all are, of course, other doctors. Sure enough, physicians went under the knife much less often than the average Ticino resident. (Lawyers' wives—whom doctors have good reason to fear—had the fewest hysterectomies of all.)

Surgeons belong to a class of experts—including computer engineers, car mechanics, taxi-drivers and others—who enjoy a fortunate position in relation to their customers. Not only do they provide a valued service (a cab ride, a repair, an operation), they also tell the customer what service she needs (a long trip, an engine overhaul, a hysterectomy). Their services are known as “credence goods”, because customers take it on faith that the supplier has given them what they need, and no more.

But as the Swiss studies show, it pays not to be too credulous. Customers can be overcharged—billed for something they did not get—or “overtreated”—given something they did not need. A mechanic might replace a car's gasket, but bill the customer for a new engine. Or he might replace the car's engine, when only a new gasket was needed.

How large do such dangers loom? For those people not married to lawyers, a new paper by Uwe Dulleck of Johannes Kepler University, Linz, and Rudolf Kerschbamer of the University of Innsbruck offers some consolation†. Customers may not know what the expert knows, but they know the incentives the experts face. If everyone acts on this knowledge, the market should, in theory, eliminate some of the incentives for expert dishonesty.

Suppose a customer can tell if his car has been fixed or not—it works, or it doesn't—but he cannot tell how it was fixed. In such cases, the mechanic has every reason to charge his customer for new brakes, even if he only replaced the brake pads. The customer should anticipate this danger. Indeed, he should resign himself to it: whatever the size of his car's problem, he can be sure his repair bill will be large.

Messrs Dulleck and Kerschbamer pursue this logic another step. If all customers share the same fatalism—as they should—what would the market for experts look like? When punters shop around for a mechanic or a plumber, they will ignore advertised prices for simple jobs. However attractive those rates may be, customers know they will never be lucky enough to pay them. They will instead prefer those experts who charge the least for elaborate procedures: new brakes, not new brake pads.

As a result, experts attract customers by shaving their prices for big jobs, and they do not lose any customers by raising their charges for small jobs. Consequently, the prices for all jobs, big and small, will tend to converge. In the extreme, Messrs Dulleck and Kerschbamer show, experts will charge a flat fee for all their services. In a competitive market, they will undercharge for expensive remedies, and overcharge for simple ones.

Is that extreme ever reached in real-life markets? Some estate agents now charge fixed fees for selling properties, shamed perhaps by the fact (demonstrated by Steven Levitt and Chad Syverson of the University of Chicago) that agents on commission sell their clients' homes more quickly and cheaply than their own.

At the start of their paper, Messrs Dulleck and Kerschbamer repeat some sage advice: if a car mechanic tells you he has replaced a part, ask him to put it in your boot. In many cases, customers can check that the expert really did what he said he did. Even Swiss doctors cannot pretend to remove someone's tonsils without really doing so. In such instances, customers cannot be overcharged. But they can still be overtreated. They know what procedure they received, but not what they needed.

Patient, heal thyself

Even if self-diagnosis is beyond them, however, customers can still diagnose the incentives experts face from the prices they post. If a surgeon enjoys fatter margins on bigger operations, he can be counted on to favour them. In principle, therefore, customers should flock to doctors who charge a uniform mark-up on all their procedures. In such surgeries, the price for complicated operations will be higher, but the margin will be the same. That way, the surgeon has no incentive to overtreat his customers. Do such surgeries exist in practice? That would be too much to hope. But many car garages now advertise standard job-completion times and then charge a uniform hourly rate. In other settings, the margins for quicker services are actually higher. New York taxis, for example, charge $2.50 the moment you sit in them, and another $2 for every mile covered.

Unfortunately, this pricing solution works only if taxi-drivers and mechanics are fully employed. When they have no trouble finding fares, taxi-drivers have no reason to take you the long way round. If they were not serving you they could be making as much money, or more, serving the next person. In quiet periods, however, the opportunity cost of “overtreating” clueless passengers falls, and the rewards rise. If the driver doubts he can find another fare, he would rather have you in his cab paying $2 a mile, than no one at all. In closing, Messrs Dulleck and Kerschbamer therefore offer advice that would otherwise seem counter-intuitive. If you are worried about being cheated by a taxi-driver or a mechanic, pick the busiest one you can find.

* For example, “Revisiting the Most Informed Consumer of Surgical Services”. International Journal of Technology Assessment in Health Care, No.4, 1993

† “On Doctors, Mechanics and Computer Specialists: The Economics of Credence Goods”. Journal of Economic Literature, March 2006