Fixing American Health Care

Of all major sectors of the economy, health care might be the most complicated from a policy perspective. Already roughly 15 percent of the U.S. Gross Domestic Product, the health care sector’s role in the economy and its cost to most Americans is rapidly increasing.

Government already plays a large role in health care. In 2008, the federal government’s two largest health care programs, Medicare and Medicaid, are expected to cost $591 billion — about 35 percent of the nation’s entire health care budget. Federal and state expenditures are estimated to make up 45 percent of all direct spending on health care. In addition, federal tax preferences for employer-paid health insurance will amount to about $247 billion in forgone income and payroll taxes.

Private markets, if properly structured, can satisfy private needs with surprising diversity and ingenuity. Many of the structural problems in markets are introduced by government itself, often in an attempt to make markets “fairer” or more affordable. These interventions often exacerbate the very problems they seek to address because the discretion they give government is often used to further the interests of those with power and wealth. A more direct solution would be to transfer wealth directly to the poor and not distort markets.

But even such a direct transfer of wealth would not solve the problems in U.S. health care, for a couple of reasons. First, normal health care is different from most types of goods. A great deal of health care consists of normal consumption such as regular eye checkups, visits to the dentist, and annual exams. These expenses are as predictable and routine as regular car maintenance. In an unregulated market there is no inherent reason why they should not be paid out of pocket by the patient when they are incurred, the same way that clothing, food, and housing are. The fact that some individuals may not be able to afford these basic services is not a failure of the health care markets any more than hunger represents a failure of the private food markets.

Medical care has another aspect, however. Occasionally individuals suffer from accidents or illnesses that present them with large unplanned costs they cannot pay. Like other catastrophic events, such as house fires and auto accidents, individuals need insurance so that they can spread the risk of occurrence among the general population. Individuals therefore need adequate health insurance to ensure that they are able to purchase treatment even when they are seriously ill.

It is important to note, however, that insurance does not reduce the overall risk but merely spreads it, ideally to others who can bear it better. If one in ten individuals get cancer, then each individual will have to pay roughly one-tenth the cost of obtaining treatment for cancer over his life in the form of insurance premiums. Alternatively, if everyone can expect to need one major operation sometime during his life, then each individual will eventually end up paying for the cost himself, even if he is insured, since the premium will reflect this expectation. From this perspective, insurance merely becomes a way of spreading the cost out over one’s own lifetime.

Although these two separate aspects of health care are distinct in theory, they are often confused in fact. Many individuals purchase insurance even for predictable, routine visits, such as an annual check-up. The reason stems not from any inherent market inefficiency or popular preference. As with many other aspects of health care in America, the conflation of routine consumption and insurance is driven by government policies.

The strong tilt in the market toward employer-paid health care results from decisions that were made more than half a century ago with no real thought to how they would impact health care. World War II imposed severe manpower shortages on the U.S. economy. In order to prevent companies from bidding employees away from each other and to keep the costs of production down, the federal government imposed wage controls on the private sector. These controls did nothing to eliminate the pressure behind the shortages, although they may have prevented people from doing things to address it. Companies quickly started to get around the controls by offering non-monetary benefits, including health care. Later, the Internal Revenue Service ruled that these benefits were not taxable as income. Even though they had a clear monetary benefit to the worker, he did not have to pay income or payroll tax on their value.

Because the employer generally makes the relevant choices, employee benefits usually cost the employer more than they are worth to the employee. Put another way, if employees were given the choice, they would generally choose for the employer to give them a pay raise rather than spend money on additional benefits. For this reason, many employer-provided benefits might have faded out after price controls were lifted were it not for the fact that the IRS had ruled that they were tax exempt.

A simple example illustrates the economics of tax exemption. Let’s assume a worker making $90,000 faces a marginal tax of 25 percent. In addition, he and his employer pay 15.6 percent in payroll taxes. What happens if the employer is willing to spend $1,000 more per year in order to keep the worker? If the employer raises the worker’s salary, the latter will receive only $633 after all taxes. But if the employer spends the money on extra health care for the worker, no taxes are paid. Thus, even if the employee only values the increased benefits at $800, he will prefer to receive them over an increase in pay. The employer of course is indifferent. The result, however, is that social welfare is decreased by $200.

There are two other problems with the result. First, the employee has an artificial preference for receiving health benefits through his work, even though in this example the benefits are worth less than their cost to the employer. If benefits were treated as taxable income, this preference would disappear. Second, as with tax deductions, the tax exclusion benefits the rich much more than it benefits the poor. The amount of subsidy is directly proportional to the marginal tax bracket, which increases with income. It is true that, unlike a deduction, the exclusion from income benefits all workers who pay income tax, even those who do not itemize deductions. But the middle and upper classes are much more likely to work for employers who offer health benefits, further tilting the advantage toward them. In 2008, these subsidies are expected to equal $247 billion in lost tax revenue.

Linking health benefits to employment has several disadvantages. First, it starves the market for individual health insurance. In light of the tremendous tax advantage given to employer-provided benefits, it should be no surprise that the market for individuals purchasing health insurance on their own is small and therefore expensive. As a result, those workers whose employers do not offer health insurance often have a difficult time finding and affording it. They also do not benefit at all from the tax break. Because the market is so thin, individuals with preexisting health conditions and the elderly may find it impossible to get affordable insurance. Yet there is no innate reason why an individual market could not work for most people. Although every economic sector has important differences, it is not at all clear that an individual health insurance market could not function as well as fire or auto insurance markets do. In the latter case there is a public mandate to purchase the insurance, but individuals arrange coverage on their own and few would think of looking to their employer for help.

A second negative effect is that the tax exclusion involves companies in an activity that they are not necessarily very good at. Much of the management efficiency since the 1990s has resulted from companies returning to their core competencies — those activities that they do best. Choosing and administering health benefits is seldom one of them. From the company’s standpoint, health benefits often work as a competitive disadvantage because established companies with older workers and retirees tend to have much higher health care costs. They are always vulnerable to new companies that can enter the market without these liabilities associated with past operations. General Motors, for example, paid $5.2 billion in health care in 2004 — or $15,000 for every car it produced; $4 billion of this cost went to retirees. Toyota, which has a much younger work force and fewer retirees, would have a built-in cost advantage even if it were to offer the exact same benefits.

These costs allocate health care inefficiently, but it is not clear that they place U.S. companies at a disadvantage. Some businessmen have worried that high health care costs hinder their international competitiveness, a point made recently by the president of the Business Roundtable: “Our soaring healthcare costs put American goods and services at a significant competitive disadvantage, and they slow economic growth.” Certainly companies in other parts of the world normally do not pay health care coverage for their workers, but they do pay much higher payroll taxes in order to fund government coverage. But even when compared to countries such as China, where government coverage is rudimentary, American employers are not harmed if they can pass these costs on to workers. There is good evidence that on the whole, they do. Increases in overall compensation, including benefits, roughly track producer productivity over the medium term, indicating that employers pay workers what they are worth — no more, no less. As health care costs rise, it is employees who bear most of the cost through lower salary increases than they otherwise would get. Since the total cost of compensation to employers is the same, they may not suffer any competitive disadvantage.

The one case in which this is clearly not true is in heavily unionized companies that have already committed to paying expensive benefits and which therefore have heavy fixed costs unrelated to current production or productivity. In these industries, employee strikes can be prohibitively expensive even if the short-term costs of labor far exceed its marginal benefit. The union has no incentive to agree to benefit cuts because the employer cannot afford to stop production or relocate without abandoning a large amount of sunken investments. In these situations, a more rational pay system that preserves the employer’s long-term competitiveness is not likely to occur until the company actually goes into bankruptcy and the owners demonstrate a willingness to write off their investment. Such was the fate of the steel and airline industries, and the American automobile industry seems likely to follow course.

It is clear, however, that the linkage between health benefits and work restricts the job mobility of workers, especially those with health concerns. This might not have mattered very much when most workers expected to stay with one employer throughout most of their careers. But it does matter in a world where company mergers and dissolutions are common events and where workers can expect to change companies and even careers a number of times during their lives. It adds yet another factor to the consideration of whether to take a particular job. The handicapped, seriously ill, or elderly are especially limited in their choices. They must accept a job with an employer that offers good benefits, and can only leave for another that also offers comprehensive benefits. The advantage that large companies have in either purchasing insurance or self-insuring translates into an advantage in attracting workers, not necessarily because workers value the insurance more but because the benefits are much cheaper than those they could get on their own.

The dominance of employer-provided health care also makes workers insensitive to the cost of the services that they receive. In a private market, prices are normally determined through a delicate interplay of the supply and demand for all goods. Suppliers always want higher prices and larger profit margins. Consumers want prices as low as possible. For most consumers the list of things that they want is much larger than the amount of things that they can afford. They therefore trade off goods and services against each other, buying those that give them the most value. The fact that consumers seldom pay the full cost of their health care distorts this balancing act and tilts consumption far more toward “free” health care and away from other goods. This in turn draws more resources into the health care sector and weakens the pressure on suppliers to provide true value.

To some extent, employers have stepped in to exert this market discipline. Because health care costs are rising so rapidly and have already become a large cost item, companies have started to participate directly in efforts to hold down costs. In Minnesota, a coalition of private and public employers has banded together into the Buyers Health Care Action Group in order to get the health care system to focus on increasing the value being delivered to their workers. Working with labor unions, they formed the Smart Buy Alliance to push for better measurement of health quality, wider adoption of best practices, more efficient administration, and better public education. Companies have also subjected workers to higher deductibles and co-payments, thereby making them bear more of the direct cost. Evidence shows that these changes do have an impact on usage but, since employers are still choosing the basic plans and paying most of the costs, and since the individual market for health insurance is so small, consumers still have limited incentives and ability to trade off consumption decisions in the same way that they do for housing or food. Because demand is not accompanied by the normal desire to hold down prices, supply is relatively unconstrained and prices are disconnected from quality.

Demand and prices are further inflated by the fact that, as mentioned, the federal government already accounts for more than a third of all spending on health care, mainly through Medicare and Medicaid and the purchase of health care for federal employees. Since this spending also is not tightly linked to quality or the relative benefit of non-health goods and services, it draws additional resources into the medical sector without imposing on them any real requirement that they increase marginal welfare to the same degree that spending on other social priorities might. These programs have also had the deleterious effect of introducing political considerations into decisions about who should get or offer care, at what prices, and with what degree of quality. Because the government, unlike a market, is inherently poor at equating costs with benefits, a large percentage of medical spending is wasted, either in the sense that the services associated with $100 of government spending are worth much less to those who receive them, or through outright fraud.

The latter is known to be large. In 2003 the Government Accountability Office (GAO) put Medicaid on its list of high-risk programs due to its size, growth, and management weaknesses. Although GAO has conducted dozens of studies on the program, it believes that the total amount of fraud cannot be precisely quantified. Meanwhile, a study of the Medicare program estimated payments made in error to be $19.9 billion in 2004. Another estimate puts the figure at $35 billion. But even these estimates are likely to understate the level of government waste because they accept all payments made under the rules as legitimate. Yet many of these payments are almost certain to confer far fewer benefits than they cost the taxpayers. In a normal market, consumers will seldom spend $100 unless they receive benefits worth at least that amount. But when government is paying for a service, beneficiaries may come to insist that the government provide services even if it costs the government $100 and even if they would not pay $70 if they had to bear the cost themselves. In this case we can say that, in some sense at least, $30 has been wasted.

Unfortunately, this type of waste is almost certainly large but immeasurable. Among some patients and doctors, government care has become its own culture, with patients having multiple visits with multiple doctors and doctors including numerous procedures for each visit in order to increase the reimbursement the government will pay. In order to control this, the government has developed an elaborate system of controls that add a great deal of administrative cost to the system. Advocates of a single-payer system — a system in which a single entity, generally the government, pays all health care providers — often point to its low administrative costs, but they seldom include in these costs either the lost value outlined above or the administrative costs imposed upon private parties to learn and comply with the rules. The results of these controls can be perverse. For example, Intermountain Health Care, a network of hospitals in Utah and Idaho, claims that medical practices that it put in place to save lives significantly reduce its revenue because Medicare pays for procedures to cure disease over precautions to prevent them. This de-linking of value from payment further inflates both the demand and the price of health care. Inflation is then often pointed to as an example of why a private market will not work.

The broken link between expenditures and quality in American health care has several negative consequences. First, health care costs rise more than they would if consumers were insisting on value for their money. Second, there are few incentives to increase either quality or efficiency over time. Over the last few decades, manufacturing industries have undergone a series of revolutions involving restructuring, reengineering, supply chain management, and quality control. These changes have brought about dramatic improvements in both quality and price, pushing back the tradeoffs that normally dominate the two. This type of pressure does not drive continuous improvements in the medical field. Researchers have identified many areas where better management could save large amounts of money — yet the industry faces little pressure to adopt them.

Instead, the tradeoff between cost and quality barely exists. Studies in the Dartmouth Atlas of Health Care show that medical practices and costs differ widely across the country, with some areas routinely performing procedures that are known not to be effective. A New England Journal of Medicine study showed that even experienced doctors vary widely in the number of precancerous polyps they discover in routine colonoscopies, with the main factor seeming to be the amount of time they spend on the procedure. The research on variations in medical care produces three main results. First, variations in resource use are very large: Medicare spending per patient in Miami is about two and a half times what it is in Minneapolis. Second, resource use is sensitive to supply, indicating that hospitals and doctors feel free to bill out equipment whether it is cost-justified or not. Normally, more MRI machines in an area would drive down the price of each procedure, but in health care, it seemingly gives doctors an incentive to schedule more MRI exams. Third, more aggressive treatment and higher spending does not result in better patient outcomes. In part this is because sicker patients often require more intervention just to get the same result as healthier ones. But a large part of the discrepancy stems from the unique payment structure and incentives associated with health care.

Because patients are used to passing the bills on to someone else, they have developed a sense of entitlement to health care that they do not apply to the quality of their housing, transportation, or college educations. They expect the best, regardless of cost, and resent any decision to deny them care, even if there is no medical evidence that the care they want is justified. It very well might be that a seventy-five-year-old man who had to pay the full cost of a hip replacement himself would balk at the cost and decide to manage on his own, even if money was no object. But let his HMO or the government tell him that it is not cost-effective and he will truly believe that a gross injustice has been done to him. The path of least resistance is often to allow the procedure and then build it in into future costs. This of course raises premiums and has the effect of pricing many people out of health care.

At the same time, doctors usually have a financial incentive to schedule as many procedures as possible, and they are generally able to do this because they know the patient will not bear the cost. These pressures on the doctor’s side are increased by the desire to avoid being second-guessed in a malpractice suit.

Although government introduces several inefficiencies into health care, there are two major objections to a totally free market that do point to a need for some boundaries on how the private markets should work. This need for boundaries is not unique; all markets need them, if only to ensure that companies fulfill their contracts once they enter into them. In banking, for instance, few would question the wisdom of some level of federal deposit insurance or the requirement that banks maintain a certain level of capitalization in order to absorb financial shocks. The need for these controls does not automatically justify other constraints such as the old prohibitions on banks selling insurance or paying interest on checking accounts, however. Similarly, in health care, the fact that some level of government regulation is likely to be beneficial does not justify a single-payer system in which the government runs virtually all health care.

The first major problem with a free market in health care stems from society’s decision that it will not deny vital care to a person who needs it simply because he cannot pay. A homeowner who loses his house to fire or flood usually has little reason to think that the government will cover his loss if he does not have insurance. This provides a strong incentive to purchase it. But a person who does not purchase health insurance still has good reason to think that he will receive a minimal level of care if he suffers a severe injury or illness, whether or not he can pay. As a result, he may be less inclined to purchase insurance. These costs are then borne by either taxpayers or other users of health care. The Federation of American Hospitals claims that hospitals have to absorb $40 billion in unpaid bills each year. These costs must either be absorbed by the hospitals, in which case fewer of them are likely to offer emergency room service, or passed on to other patients. If the latter, then the costs of health care will rise further, possibly causing other people to drop insurance or forgo care.

In theory, this problem could be solved relatively easily by requiring everyone to purchase an insurance policy as is done with car insurance (although in practice, one in seven drivers still does not carry insurance). Any requirement would also force the government to define what the minimum level of insurance is. This would inevitably lead to pressure from lobbying groups to cover their specific diseases, procedures, and drugs. Drug manufacturers will press for mandated coverage of drugs for impotence and hair loss. Psychologists and psychiatrists recently succeeded in getting a law that makes insurance plans cover mental health care. Each of these mandates adds to the cost of insurance without necessarily delivering an equal or greater benefit. Normally individuals would purchase such plans only if the benefits outweighed the costs. With a mandate, the decision on whether this condition is met is taken away from the individual and given to the government. Adding too many requirements prices many workers out of the market.

The second problem with a free market in which everyone purchases individual insurance is that future medical costs can often be predicted by past ones. Some medical events, such as automobile accidents and breast cancer, hit suddenly and unpredictably (although to the extent that cancer is genetic, science may someday be able to predict the probability of cancer). Many, however, are associated with past and existing health. As a result, individuals with chronic health problems have very high expected future costs. These people can only find affordable health care if the pool of individuals participating in an insurance plan contains a large number of relatively healthy people. If this is the case, then people with good health subsidize those with poor health. The problem is that the former have no incentive to participate in such an arrangement. If given other options, such as Medical Savings Accounts, that permit them to pay lower insurance costs, they are likely to do so. This of course raises the premiums of those still in the insurance pool, exacerbating the problem of cross subsidization and making health care unaffordable to the sick.

But forcing everyone to participate in the pool is not necessarily the best approach, either. For one thing, it is not very clear why a young worker on the bottom of the career ladder should subsidize an older worker who might make two or three times his salary. Certain inequalities are inherent in life. We know we cannot restore eyesight to the blind by making everyone else’s eyesight just a little bit worse, but some people think that we have an obligation to protect people who happen to have high health care costs by reducing the income of everyone else.

Is there really any inherent injustice in expecting someone with $20,000 in annual medical costs to pay for them? Shouldn’t the real problem be whether he has the ability to make these payments himself without suffering the type of extreme economic hardship that we would want to protect any deserving person from, whatever the cause? If the answer is yes, then we are faced not so much with a health care problem as with an income problem that in this particular case is linked to health. Since the costs of preexisting conditions are largely predictable, they cannot be spread out among a broader population through insurance unless we force everyone into the same insurance pool and maintain the practice of including most health costs, except those that are relatively minor or predictable, in the insurance policies that everyone buys. We have already seen the complications that this creates.

It is true that purchasers of health insurance might very well wish to purchase policies that not only cover the costs of an unexpected, expensive event, but also protect them against the large annual costs associated with a chronic disease. The policy might therefore cover all annual expenses over a set figure such as $10,000. But this option does not exist for people with preexisting conditions, many of whom are in the lower and middle class and will therefore have difficulty bearing the cost themselves. Protecting these individuals by having someone else pay the cost raises several problems, however, the most important of which is that any protection is likely to remove their incentive to minimize costs or to forgo treatment that costs society more than it benefits them. Someone or something else therefore has to impose this discipline.

Nevertheless, some of these individuals will need a subsidy in order to afford health care. It is difficult for private markets to provide this subsidy, however. Insurers may be willing to extend coverage, but only if they can charge the proper risk-adjusted premium, which is likely to be too high for many to afford. If insurers are forced to offer the same premium to everyone, then consumers with few health concerns are likely to find that their benefits are not worth the premiums. They will look for cheaper insurance or, if none is available, possibly do without. Another alternative is to have the government subsidize insurance for individuals with preexisting conditions. But this sometimes creates perverse incentives as individuals seek to avoid losing their eligibility as their income rises.

Recent problems in Maine illustrate some of these difficulties. Because the state’s insurance plan, DirigoChoice, offers very comprehensive coverage, many individuals cannot afford it. Others who can afford it decide to go without because they do not see the need to pay for coverage that they are unlikely to use. Those who do sign up often need significant care, frequently for preexisting problems, which increases the cost of the program. Because the program is means-tested, different people are required to pay different amounts. At least one person interviewed stated in the New York Times that he limited the amount of hours he worked in order to remain eligible for the low premium rate. (He still dropped the insurance after rates rose by 13.4 percent, however.)

For this reason, many people believe that some form of compulsion is needed in order to get younger and healthier individuals to participate in the insurance pool. But forcing the healthy to purchase insurance will not automatically solve the problem of high-cost individuals. Left to themselves, people who expect to use very few health services will form a separate pool with lower premiums. The rules must force them to participate in the same insurance pool as the sick, thereby subsidizing them.

Some observers call for a single-payer program in which the government provides health care to all individuals, similar to what other major industrialized countries such as Japan, Germany, the United Kingdom, and Canada offer their citizens. In theory, a centralized system would reduce administrative costs, achieve the greatest savings through standardization and volume purchasing, and rationalize medical care by deciding on the best treatment in each case, thereby eliminating unnecessary procedures. In practice, government administrators can never display the drive toward efficiency, continual innovation, and rapid evolution that a properly designed private sector can. As Eric Cohen and Yuval Levin recently observed in Commentary:

Everywhere it has been tried, the single-payer model has yielded inefficient service and lower-quality care. In Britain today, more than 700,000 patients are waiting for hospital treatment. In Canada, it takes, on average, seventeen weeks to see a specialist after a referral. In Germany and France, roughly half of the men diagnosed with prostate cancer will die from the disease, while in the United States only one in five will. According to one study, 40 percent of British cancer patients in the mid-1990s never got to see an oncologist at all.

These single-payer systems were started decades ago when medicine was still relatively simple and the national population was young and healthy. As long as health care costs were a minor part of the total economy, a relatively large amount of inefficiency could be tolerated, especially if it was possible to maintain current services by avoiding investments in future quality improvements. Now each of these systems is increasingly being forced to deal with the same tradeoffs facing American medicine, except that the single-payer system has far less flexibility to make the necessary changes. Their problems would be even worse if the United States did not continue to originate and subsidize most of the medical advancement and innovation in the rest of the world.

The primary task of medical reform cannot be to extend the best quality health care coverage to all Americans. Such a goal, like affordable housing, is likely to be self-defeating, both because of its indeterminate nature and because the vast resources that it will require are likely to have the net effect of pricing more Americans out of health care. The definition of health coverage is likely to expand gradually to cover much more than the necessities of adequate physical health. Already it has been expanded to include hip replacements for eighty-year-olds, birth control, remedies for baldness, and drugs for improving one’s sex life. This adds to the cost of giving the poor coverage for broken bones, heart attacks, and diabetes.

Even if a narrow definition of health care is maintained, government funding is usually not well linked to quality. When this is the case, costs increase up to whatever threshold is set by the government. After that, quality deteriorates. And — as in housing and education — increases in funding are unlikely to produce automatic increases in quality. They are likely to create a situation where the real cost of a given level of quality rises, pricing the poor out of the market. Since there will never be enough government funds to provide a sufficient level of service to all who qualify, many are likely to be left without. Unlike in housing, rationing in health care is unlikely to take the form of people being denied any form of government assistance. Instead, everyone will have coverage for basic care but cost-control measures will gradually limit the scope of that care by imposing waiting lines and denying certain medications and procedures. The existence of a private care option will exacerbate this trend as those who can afford it opt out for a better level of service even if they have to pay for it themselves. This trend is already happening in public education even though local control of school districts protects wealthier neighborhoods from the effects of deteriorating districts. The same pattern in health care is also evident in the fact that more and more doctors are refusing to see Medicare and Medicaid patients because government reimbursement rates are so low. At the same time, richer individuals are paying more in order to get a higher standard of care than most people receive. Although some people have criticized this practice as unethical, it is likely to continue as the quality of government and private programs deteriorates.

There is no magic bullet that will solve the many inefficiencies of health care. In fact, tradeoffs are inherent in any market. But there are strong reasons to think that giving individual patients more control over their own health care and subjecting the sector to greater competition will create an environment in which quality is gradually improved and prices are brought under control.

Before we begin to explore what such a system would look like, it is worth pointing out that there is nothing intrinsically wrong with the fact that Americans spend a growing proportion of their income on health care. To begin with, the United States is a rich country. Most workers have all of their basic necessities met and can afford to spend more on things that are important, but not vital, to life. Health care certainly qualifies for that. In a country where expenditures on flat-screen televisions, pet care, and vacations are rapidly increasing, it should not be surprising that people also devote a growing fraction of their income to physical well-being. And science has certainly made this possible. A growing number of drugs and therapies now promise improvements that before were unachievable. Americans have received some benefit from these expenditures. Life expectancy has steadily risen, especially for older individuals, largely because mortality rates associated with causes like heart disease, cancer, and strokes have fallen. Also contributing to the growth of the health care sector is the fact that, demographically, Americans are getting older as the Baby Boomers mature. Since medical costs increase with age, it is almost inevitable that health care spending will rise over time.

The challenge therefore is not to slow down the growth in health care spending, but to make sure that Americans are getting adequate value for what they spend, both in terms of health benefits and in comparison to the benefits that they receive from spending in other sectors such as education, housing, and entertainment. Despite the beliefs of those whose philosophical leanings and political commitments incline them toward a single-payer system run by government and administered by a bureaucracy of experts, true cost-benefit pressure will only come about by giving individuals both the power and the incentive to make these tradeoffs for themselves. Partly because of the complexities already introduced by government, the transition to such a system will be exceedingly difficult, involving a number of separate reforms. However, there is no reason to think that health care is inherently more complex than other industries in which government plays a much smaller role. And unless we know the general direction in which we want the system to move, evaluating any single reform is very difficult.

The ideal policy for promoting both flexibility and coverage in health care would involve two reforms. The first is a refundable tax credit that would be available only if the taxpayer provided the policy number for a qualified health insurance plan that included catastrophic coverage. The coverage would have to extend to all children for whom the couple is responsible. Since the credit would only be available if an individual purchased insurance, every adult would have a strong incentive to do so even absent a government mandate. Making the full amount of the tax credit refundable even if the actual policy costs much less would ensure that taxpayers treat the money as their own. This would cause them to use the same type of cost-benefit test that they apply to spending in other parts of their lives. This change would be strongly progressive since it would effectively pay uncovered people, many of whom are poor, to obtain coverage, while holding harmless most of the people who benefit from the current tax exclusion for employer-provided insurance.

An important part of this reform would be the elimination of the income exclusion for employer-paid benefits, including health care. Employers would have to list the cost of each employee’s health care on his W-2 form. This amount would be included as income in the employee’s tax return. Employers should also be required to allow each employee to opt out of care, taking the employer’s payment share in return. This would give employees the ability to shop for other plans and increase the size of the individual insurance market.

The second major reform would be to allow much greater freedom for individuals and groups to band into insurance pools. The resistance from some quarters to allowing small businesses and groups to get together to purchase insurance for their members makes no economic sense. Similarly, the efforts of groups like AARP and many labor unions to offer insurance plans to their membership and others should be encouraged. Finally, the government should open up the Federal Employee Health Benefits Plan to all individuals. Under the federal plan, the government uses its size to negotiate terms with a wide range of health care plans. Employees can then choose whichever plan is best for them. In each pool, insurers should be required to offer the same terms to all individuals, regardless of prior history, with one exception: Insurers should be allowed to vary their price by the age and sex of the individual. This reform would keep premiums low for young people, encouraging them to participate in the pool, rather than purchase individual coverage. It would also create a better match between health care premiums and income over one’s career. (If desired, the size of the refundable tax credit could also vary by age in order to create a better match with expenses.)

The goals are to give people purchasing power and to offer them as many choices as possible. Most individuals will continue to be content to participate in a large pool. While this reduces their sensitivity to price and quality concerns, it protects them from risk. But the freedom to move between pools and to get individual coverage should ensure that competitive pressures are greater than they are now. With large pools, insurers should have greater certainty over the actual costs they will face. A reinsurance market would presumably arise so that pools could insure themselves against higher-than-expected claims. This would allow them to keep premiums stable and would minimize the possibility of an adverse cycle where high costs lead to premium increases, which drive away individuals, leading to even higher premiums.

How would such a system evolve? One outcome is that individuals are likely to assume a much greater role in their own health care. A number of people have objected to this, arguing that individuals will always lack the time or information needed to make intelligent decisions about cost and quality. As Jonathan Cohn puts it in his 2007 book Sick: “No matter how much information consumers have, who’s to say they will make good decisions?” Maybe so; there is, of course, no guarantee. However, we do know that innovations such as the Internet and support groups are making it increasingly easy for consumers to obtain detailed personalized medical information. Patients are forming sophisticated virtual communities to inform and support each other. Government agencies and health care providers use software to deliver personalized health messages and alerts. For example, a new website, GroupLoop.org, helps teenage cancer patients interact with each other online. Although consumers will not always make the correct decisions, neither will the doctors whose advice they have traditionally deferred to. Doctors are not infallible: some are young and inexperienced; others have failed to keep up with advances in their field; none has the single, self-interested focus that the patient has. For many of the cases in which a diagnosis or remedy is not clear, medicine remains as much an art as a science. Where there is doubt about the best path to pursue, who but the client should make an informed decision?

If patients are to make informed decisions, they need information. Over the last decade, the federal government has announced several initiatives to track the quality of care and the effectiveness of alternative treatments. The government should adopt policies that make it much easier for patients to get the information they need. Just as it required gas stations to post their prices, it could require hospitals to provide an all-in cost estimate prior to any medical procedure. At present, doctors often bill for their services separately, bills contain multiple itemizations that are not clearly labeled, and the bill arrives several weeks or months after the procedure. The government and employers could also continue with their efforts to collect information on the quality of outcomes, both between individual doctors and hospitals and between different procedures or treatments. This does raise some difficulties: doctors who take the most difficult cases may show higher failure rates, and different patients may react differently to the same treatments. But overall it would allow patients to make more informed choices.

Other improvements could be made in the efficiency with which services are delivered. The Veterans Health Administration has been leading an effort to implement electronic medical records for all patients. Greater use of electronic records could make it easier for doctors to look at a patient’s full medical history and could give the patient better control over his own records by putting them in the form of a single electronic record that he can carry around rather than in individual paper records scattered among the various doctors, dentists, and pharmacists that he sees throughout his lifetime.

The use of a single form to submit medical billing would also reduce administrative costs for doctors. The government’s large role in Medicare and Medicaid introduces much inefficiency into the market, but it does give the government a critical mass sufficient to impose some sensible changes into the market. By adopting a single form that other insurance agencies could voluntarily use, the government might be able to lead the market toward beneficial standardization. Since use of the form would be voluntary for private insurers, the government would have to make sure that the form met private needs to a great extent; otherwise, companies would continue to use their own forms. But if the government did implement a good form for its own programs, other insurers would face pressure to adopt it as well.

Although specialization has often led to increased costs, it also provides a possible route to greater efficiency. Centers that specialize in a limited number of procedures can often produce better results at cheaper prices due to economies of scale. Among the most important economies is the ability to fully make use of highly trained professionals who, because of their narrow focus, have a much deeper level of experience with a particular illness or procedure than does the average doctor. In India, the Aravind Eye Hospital provides one example of this model in the context of a developing country. The hospital specializes in eye surgery, delivering the world’s highest standards at a small fraction of the normal cost. Business practices are designed to make the maximum use of its most skilled professionals, while its services are closely integrated with those of suppliers and complementary products to minimize costs. If such specialized centers demonstrate greater efficiency in terms of either cost or outcomes, government policy should not discourage them in order to protect less efficient general hospitals.

Another innovation that is likely to spread if left to market forces is the emergence of small walk-in clinics that provide routine care in locations and at times that are most convenient to the average patient. Wal-Mart recently decided to place several of these centers in its stores. If the clinics do not provide enough value to individuals, they will fail. But there is no reason why we should hope that they do not succeed. They are likely to provide much better care for the type of inquiries they see. If their nature and location are widely known, they may also attract patients who might otherwise go to the emergency room for lack of a better alternative.

Government should also play an active role in addressing some of the structural rigidities that can reduce competition in the health care markets, thereby keeping costs high. Without strong competitive forces operating in all parts of the market, the drive for innovation and improvement will weaken. One issue involves determining exactly what skills are needed to provide certain treatments. Licensing requirements can make it more expensive for young people to enter professions by artificially increasing the amount of training needed to perform at a certain level. These requirements can also make professional services unaffordable to many people by giving providers the right to limit the competition they face. Within medicine there have been a number of controversies related to allowing professionals with fewer skills to perform designated services. New technology is likely to further this trend by substituting capital, including information technology, for human skills, thereby allowing lower-skilled technicians to deliver services that used to require more specialized knowledge.

Unfortunately, the context in which these disputes are settled is clouded by self-interest. The premise that government should regulate the skills needed to perform certain functions is implicitly based on the assumption that these decisions will be made in the patient’s interest, presumably following a subjective study of whether extra training is needed to produce a better outcome, and, if it is, whether the improvement in results justifies the higher costs associated with hiring higher-skilled professionals. However, threatened professions usually try to exert political pressure in order to guard their privileges. For example, radiologists have recently expressed concern about general practice doctors reading MRIs. The specialists claim that regular doctors may misinterpret the results. Anesthesiologists have opposed reforms that would allow nurse anesthesiologists to substitute for them in the operating room under general supervision. Doctors have also opposed state laws that give non-physicians the ability to write prescriptions. It is perhaps natural that in each case the threatened professionals would use the interest of patients as a cover. But it is not clear from any of these cases that the political decision of what standard of care is appropriate will be made based on the patient’s interest instead of the respective political power of the professions concerned.

Another avenue in which competition is likely to increase is “medical tourism.” The cost of care is often much lower in developing countries. Yet centers in these countries often have U.S.-trained doctors and facilities that match world standards. In addition, they can offer special amenities that patients cannot afford at home, such as private rooms, higher nurse-to-patient ratios, and special meals. There are of course some legitimate issues to be concerned about. One is the need to ensure that patients receive adequate care. Increasingly, foreign hospitals are seeking U.S. accreditation for both their facilities and their doctors, thereby ensuring that standards are the same as in America. It is also possible to negotiate ways to assure that patients who suffer from malpractice receive compensation, either by having foreign facilities affiliate with domestic institutions that can be sued, or by setting up impartial arbitration boards to hear cases and award compensation. The savings can be significant. Yet, despite the fact that treatment abroad can benefit both the health insurer and the patient, unions have opposed this trend even when willing members would be able to save thousands of dollars of their own money on the procedures.

A final improvement would be to set up an alternative mechanism for resolving disputes about medical care. The current system of medical malpractice imposes significant costs without providing timely recovery to those who are hurt by malpractice. Injured patients must initiate a private lawsuit in order to seek compensation. Unless the case is clear cut, it can often take several years before the plaintiff is able to collect anything. A recent study in the University of Michigan Law Review shows that plaintiffs win only about half of the cases that expert reviewers think they should win. In the cases where there is an award, attorneys’ fees typically consume 30 to 40 percent of it, leaving the plaintiff with only a partial recovery. At the same time, the threat of a lawsuit imposes severe costs on doctors. Medical malpractice premiums for a general surgeon in Miami can equal $174,000 per year, causing many to move to cheaper states. The result has been a shortage of specialists in some areas of the country. Although doctors frequently claim that lawsuits force them to practice defensive medicine by assigning additional procedures that are not medically indicated, a recent Congressional Budget Office report had a difficult time documenting this. Part of the reason may have been that doctors already have a financial incentive to schedule additional care or that the threat of a suit in some cases causes doctors to avoid extending care, especially to high-risk patients. Neither explanation is a sign of a healthy system.

At the same time, it can hardly be said that the medical profession has been vigilant about monitoring the quality it delivers to patients. The profession continues to suffer from a high number of medical accidents and hospital-induced infection rates. Even getting doctors to wash their hands regularly can be a large problem. There is often little pressure to conform to best known practices. But most importantly, the profession does a terrible job of weeding out those professionals who are the greatest threat to patients. Even in some cases where there is evidence that a doctor or nurse is intentionally murdering patients, hospitals have been slow to reach judgment and often merely dismiss the suspect from the hospital, leaving him free to go elsewhere.

Common Good (a nonprofit organization dedicated to reforming America’s “lawsuit culture”) and the Harvard School of Public Health have proposed setting up independent administrative law judges with medical expertise to decide cases based on liberal standards of recovery, which allow patients to obtain an award if their injury was preventable without having to prove malpractice. Such a system would be quicker and cheaper and, if truly independent, might result in more frequent awards to patients. Part of the plan calls for the information leading to the injury to be widely disclosed so that the root error causing it can be corrected. Without such a practice, it is much harder to implement a process of continuous improvement across the industry. Moreover, there is some evidence that full disclosure, followed by a quick apology and offer of a fair settlement, can often reduce total malpractice costs. When the University of Michigan implemented such a policy, its medical malpractice suits fell by over 60 percent and its cost per claim fell by more than half.

The goal of public policy should not be to lower total spending on health care, broadly defined. It may well be that in an increasingly affluent society where most individuals can easily afford the material goods they need, people will choose to spend a substantial portion of their incomes on their health. The goal is to make sure that they receive true value for whatever they spend on health care and to subject the sector to the same competitive pressures that have driven down real costs in other parts of the economy. In other words, while new technologies may come forward and justify new expenditures, we should expect the price of any given medical service to decline over time, even as its quality goes up. This can happen, but only if health care spending is forced to compete with the broad range of goods and services available to the average consumer. This in turn requires us to trust individuals with both the power and the incentive to make their own health care decisions.