Naked Economics (16 page)

The information problem at the heart of health care has not gone away: (1) The patient, who does not pay the bill, demands as much care as possible; (2) the doctor maximizes income and minimizes lawsuits by delivering as much care as possible; (3) the insurance company maximizes profits by paying for as little care as possible; (4) technology has introduced an array of massively expensive options, some of which are miracles and others of which are a waste of money; and (5) it is very costly for either the patient or the insurance company to prove the “right” course of treatment. In short, information makes health care different from the rest of the economy. When you walk into an electronics store to buy a big-screen TV, you can observe which picture looks clearest. You then compare price tags, knowing that the bill will arrive at your house eventually. In the end, you weigh the benefits of assorted televisions (whose quality you can observe) against the costs (that you will have to pay) and you pick one.
Brain surgery really is different.

The fundamental challenge of health care reform is paying for the “right” treatment—the “product” that makes the most sense relative to what it costs. This is an exercise that consumers perform on their own everywhere else in the economy. Bean counters should not automatically say no to super-expensive treatments; some may be wonderfully effective and worth every penny. They
should
say no to expensive treatments that are not demonstrably better than less expensive options. They should also say no to doing some tests “just to be sure,” both because these diagnostics are expensive, but also because when administered to healthy people they tend to generate “false positives,” which can breed expensive, unnecessary, and potentially dangerous follow-up care.

There is an old aphorism in advertising: “I know I’m wasting half my money; I just wish I knew which half.” Health care is similar, and if the goal of health care reform is to restrain rapidly rising costs, then any policy change will have to focus on quality and outcomes rather than just paying for inputs.
New York Times
financial columnist David Leonhardt describes the treatment for prostate cancer (where fabulously expensive technology does not appear to be delivering better health) as his own “personal litmus test” for health care reform. He writes, “The prostate cancer test will determine whether President Obama and Congress put together a bill that begins to fix the fundamental problem with our medical system: the combination of soaring costs and mediocre results. If they don’t, the medical system will remain deeply troubled, no matter what other improvements they make.”

But we’re not done with health care yet. The doctor may know more about your health than you do, but you know more about your long-term health than your insurance company does. You may not be able to diagnose rare diseases, but you know whether or not you lead a healthy lifestyle, if certain diseases run in your family, if you are engaging in risky sexual behavior, if you are likely to become pregnant, etc. This information advantage has the potential to wreak havoc on the insurance market.

Insurance is about getting the numbers right. Some individuals require virtually no health care. Others may have chronic diseases that require hundreds of thousands of dollars of treatment. The insurance company makes a profit by determining the average cost of treatment for all of its policyholders and then charging slightly more. When Aetna writes a group policy for 20,000 fifty-year-old men, and the average cost of health care for a fifty-year-old man is $1,250 a year, then presumably the company can set the annual premium at $1,300 and make $50—
on average
—for each policy underwritten. Aetna will make money on some policies and lose money on others, but overall the company will come out ahead—if the numbers are right.

Is this example starting to look like the Hope Scholarships or the used-car market? It should. The $1,300 policy is a bad deal for the healthiest fifty-year-old men and a very good deal for the overweight smokers with a family history of heart disease. So, the healthiest men are most likely to opt out of the program; the sickest guys are most likely to opt in. As that happens, the population of men on which the original premium was based begins to change; on average, the remaining men are less healthy. The insurance company studies its new pool of middle-aged men and reckons that the annual premium must be raised to $1,800 in order to make a profit. Do you see where this is going? At the new price, more men—the most healthy of the unhealthy—decide that the policy is a bad deal, so they opt out. The sickest guys cling to their policies as tightly as their disease-addled bodies will allow. Once again the pool changes and now even $1,800 does not cover the cost of insuring the men who sign up for the program. In theory, this adverse selection could go on until the market for health insurance fails entirely.

That does not actually happen. Insurance companies usually insure large groups whose individuals are not allowed to select in or out. If Aetna writes policies for all General Motors employees, for example, then there will be no adverse selection. The policy comes with the job, and all workers, healthy and unhealthy, are covered. They have no choice. Aetna can calculate the average cost of care for this large pool of men and women and then charge a premium sufficient to make a profit.

Writing policies for individuals, however, is a much scarier undertaking. Companies rightfully fear that the people who have the most demand for health coverage (or life insurance) are those who need it most.
This will be true no matter how much an insurance company charges for its policies.
At any given price—even $5,000 a month—the individuals who expect their medical costs to be higher than the cost of the policy will be the most likely to sign up. Of course, the insurance companies have some tricks of their own, such as refusing coverage to individuals who are sick or likely to become sick in the future. This is often viewed as some kind of cruel and unfair practice perpetrated on the public by the insurance industry. On a superficial level, it does seem perverse that sick people have the most trouble getting health insurance. But imagine if insurance companies did not have that legal privilege. A (highly contrived) conversation with your doctor might go something like this:

Insurance companies ask applicants questions about family history, health habits, smoking, dangerous hobbies, and all kinds of other personal things. When I applied for term life insurance, a representative from the company came to my house and drew blood to make sure that I was not HIV-positive. He asked whether my parents were alive, if I scuba dive, if I race cars. (Yes, yes, no.) I peed in a cup; I got on a scale; I answered questions about tobacco and illicit drug use—all of which seemed reasonable given that the company was making a commitment to pay my wife a large sum of money should I die in the near future.

Insurance companies have another subtle tool. They can design policies, or “screening” mechanisms, that elicit information from their potential customers. This insight, which is applicable to all kinds of other markets, earned Joseph Stiglitz, an economist at Columbia University and a former chief economist of the World Bank, a share of the 2001 Nobel Prize. How do firms screen customers in the insurance business? They use a deductible. Customers who consider themselves likely to stay healthy will sign up for policies that have a high deductible. In exchange, they are offered cheaper premiums. Customers who privately know that they are likely to have costly bills will avoid the deductible and pay a higher premium as a result. (The same thing is true when you are shopping for car insurance and you have a sneaking suspicion that your sixteen-year-old son is an even worse driver than most sixteen-year-olds.) In short, the deductible is a tool for teasing out private information; it forces customers to sort themselves.

Any insurance question ultimately begs one explosive question: How much information is too much? I guarantee that this will become one of the most nettlesome policy problems in coming years. Here is a simple exercise. Pluck one hair from your head. (If you are totally bald, take a swab of saliva from your cheek.) That sample contains your entire genetic code. In the right hands (or the wrong hands), it can be used to determine if you are predisposed to heart disease, certain kinds of cancer, depression, and—if the science continues at its current blistering pace—all kinds of other diseases. With one strand of your hair, a researcher (or insurance company) may soon be able to determine if you are at risk for Alzheimer’s disease—twenty-five years before the onset of the disease. This creates a dilemma. If genetic information is shared widely with insurance companies, then it will become difficult, if not impossible, for those most prone to illness to get any kind of coverage. In other words, the people who need health insurance most will be the least likely to get it—not just the night before surgery, but ever. Individuals with a family history of Huntington’s disease, a hereditary degenerative brain disorder that causes premature death, are already finding it hard or impossible to get life insurance. On the other hand, new laws are forbidding insurance companies from gathering such information, leaving them vulnerable to serious adverse selection. Individuals who know that they are at high risk of getting sick in the future will be the ones who load up on generous insurance policies.

An editorial in
The Economist
noted this looming quandary: “Governments thus face a choice between banning the use of test results and destroying the industry, or allowing their use and creating an underclass of people who are either uninsurable or cannot afford to insure themselves.”
The Economist,
which is hardly a bastion of left-wing thought, suggested that the private health insurance market may eventually find this problem intractable, leaving government with a much larger role to play. The editorial concluded: “Indeed, genetic testing may become the most potent argument for state-financed universal health care.”
⁴

Any health care reform that seeks to make health insurance both more accessible and more affordable, particularly for those who are sick or likely to get sick, will have devastating adverse selection problems. Think about it: If I promise that you can buy affordable insurance, regardless of whether or not you are already sick, then the optimal time to buy that insurance is in the ambulance on the way to the hospital. The only fix for this inherent problem is to combine guaranteed access to affordable insurance with a requirement that everyone buy insurance—healthy and sick, young and old—a so-called “personal mandate.” The insurance companies will still lose money on the policies that they are forced to sell to bad risks, but those losses can be offset by the profits earned from healthy people who are forced to buy insurance. (Any country with a national health care system effectively has a personal mandate; all citizens are forced to pay taxes, and in return they all get government-funded health care.)

This is the approach that Massachusetts took as part of a state plan to provide universal access to health insurance. State residents who can afford health insurance but don’t buy it are fined on their state tax return. Hillary Clinton supported a personal mandate in the 2008 Democratic presidential primaries; Barack Obama did not, though that arguably had more to do with distinguishing himself from his toughest Democratic opponent than it did with his analysis of adverse selection. Obviously, forcing healthy people to buy something that they would otherwise not buy is a heavy-handed use of government; it’s also the only way to pool risk (which is the purpose of insurance) when the distribution of risk is not random.

Here are the relevant economics: (1) We know who is sick; (2) increasingly we know who will become sick; (3) sick people can be extremely expensive; and (4) private insurance doesn’t work well under these circumstances. That’s all straightforward. The tough part is philosophical/ideological: To what extent do we want to share health care expenses anyway (if at all), and how should we do it? Those were the fundamental questions when Bill Clinton sought to overhaul health care in 1993, and again when the Obama administration took it up in 2009.

This chapter started with the most egregious information-related problems—cases in which missing information cripples markets and causes individuals to behave in ways that have serious social implications. Economists are also intrigued by more mundane examples of how markets react to missing information. We spend our lives shopping for products and services whose quality we cannot easily determine. (You had to pay for this book before you were able to read it.) In the vast majority of cases, consumers and firms create their own mechanisms to solve information problems. Indeed, therein lies the genius of McDonald’s that inspired the title of this chapter. The “golden arches” have as much to do with information as they do with hamburgers. Every McDonald’s hamburger tastes the same, whether it is sold in Moscow, Mexico City, or Cincinnati. That is not a mere curiosity; it is at the heart of the company’s success. Suppose you are driving along Interstate 80 outside of Omaha, having never been in the state of Nebraska, when you see a McDonald’s. Immediately you know all kinds of things about the restaurant. You know that it will be clean, safe, and inexpensive. You know that it will have a working bathroom. You know that it will be open seven days a week. You may even know how many pickles are on the double cheeseburger.
You know all of these things before you get out of your car in a state you’ve never been in.