2011-07 Peter Orszag, How Health Care Costs Can Save or Sink America
Rising health-care costs are at the core of the United States’ long-term fiscal imbalance. The Congressional Budget Office (CBO) projects that between now and 2050, Medicare, Medicaid, and other federal spending on health care will rise from 5.5 percent of GDP to more than 12 percent. (Social Security costs, by comparison, are projected to increase from five percent of GDP to six percent over the same period.) It is no exaggeration to say that the United States’ standing in the world depends on its success in constraining this health-care cost explosion; unless it does, the country will eventually face a severe fiscal crisis or a crippling inability to invest in other areas.
The problem is not limited to the federal government. Over the past 25 years, cost increases in the national Medicare and Medicaid programs have roughly paralleled (and actually been slightly below) cost increases in the rest of the health-care system. These trends drive a wide range of problems. State governments have had to divert funds from education to health care, which is partly why salaries for professors at public universities are now often 15 to 20 percent lower than those at comparable private universities. Meanwhile, the rising cost of employer-sponsored health insurance has squeezed take-home pay for most U.S. workers at the same time as median wages have stagnated and income inequality has increased.
Another dimension of the problem involves the variation of health-care costs across the United States. A recent analysis by the Medicare Payment Advisory Commission found that spending in higher-cost areas of the United States (that is, those in the 90th percentile ranked by cost), even after controlling for various factors, was 30 percent higher than in lower-cost areas (those in the 10th percentile). This substantial variation is undesirable both because the high-cost areas unnecessarily drive up total costs and because the results are often haphazard for patients. Indeed, higher costs typically do not equal better care — and sometimes they mean the opposite.
In March 2010, the United States attempted to address these problems by passing a historic health reform act. The new law set up health exchanges through which individuals can purchase insurance, required those without health insurance to buy it, and created subsidies to offset part of the cost of insurance, especially for moderate-income households. The bill also reduced payments from Medicare and Medicaid to providers, imposed a new tax on high-cost insurance plans, and created a set of new institutions intended to bolster quality and reduce costs throughout the system.
Even before it passed, the health act became mired in political controversy, and its future remains at risk. Opponents have filed legal challenges to the law, the House of Representatives has voted to repeal it, and the funding necessary to administer it is in jeopardy. To be fair, the new law has many shortcomings — including its failure to seriously reform the medical malpractice system. It does, however, create new infrastructure that can improve the quality of treatment and cut costs. For this infrastructure to succeed, though, the tools created by the health act must be applied forcefully, rather than undermined or abandoned. Even then, more drastic measures may ultimately be needed. The United States must make a fundamental change to its health-care system, transforming it into one that emphasizes evidence and quality, one in which providers have better tools and much stronger incentives to deliver value.
STRATEGIES FOR SAVING
Health-care costs rise for a variety of reasons, and there are essentially four conceptual approaches to constraining them. The first approach is to simply reduce payments to providers — hospitals, doctors, and pharmaceutical companies. This blunt strategy can work, often quite well, in the short run. It is inherently limited over the medium and long term, however, unless accompanied by other measures to reduce the underlying quantity of services provided. If only Medicare and Medicaid payments were reduced, for example, providers would shift the costs to other patients and also accept fewer Medicare and Medicaid patients. This would make the approach politically nonviable.
The second approach is direct rationing, whereby the government decides which services will be offered and which will not. This approach is not remotely politically viable in the United States, where people have grown accustomed to access to new technologies and procedures and where antigovernment sentiment is strong.
The third approach — consumer-directed health care — could be a useful component of a cost-reduction strategy, but its benefits are often exaggerated. This approach emphasizes giving consumers more information and control over their health care and stronger financial incentives to reduce their own spending. The goal is to ensure that patients have a greater stake in keeping costs down through increased copayments and other forms of cost sharing.
If most health-care spending were driven by discretionary decisions among relatively healthy people, this approach could cut costs dramatically. But health-care costs are instead heavily concentrated among a small number of relatively sick patients. The top five percent of Medicare beneficiaries ranked by cost, for example, account for more than 40 percent of total Medicare spending, and the top 25 percent account for more than 85 percent of total costs. Financial incentives can have some effect on these people’s decisions, but under virtually all consumer-directed proposals, these patients would still be covered by generous third-party insurance for their high-cost procedures — which is, after all, the whole point of insurance.
Consumer-directed measures would have a substantial impact only if they lowered the cost of the care delivered in the most expensive cases. Yet some research suggests that consumer-directed health approaches could make high-cost cases even more expensive, because chronically ill patients facing copayments for their medicines would skip some doses, requiring even more expensive treatment later on. (Ironically, those who advocate consumer-directed reforms often oppose advance directives that spell out individuals’ care instructions for late in life — tools that might be more effective than any other consumer-directed change.) Since the share of total costs most affected by consumer-directed health-care incentives is relatively modest, no one should expect this approach to dramatically reduce overall health-care spending.
Nonetheless, the consumer-directed approach is at the heart of a reform of Medicare put forward in April by Representative Paul Ryan (R-Wis.), chair of the House Budget Committee. Under Ryan’s approach, Medicare would be transformed into a “premium support” plan, whereby the government would pay the premiums for private health insurance plans chosen by beneficiaries. Ryan’s plan appears to save substantial sums for the federal government, but it is far less clear that it would substantially reduce overall health-care costs because it may not do enough to affect high-cost cases. Indeed, a preliminary analysis of the Ryan plan by the CBO found that total costs would actually increase — by an astonishing 40-67 percent by 2030 — because the benefit of having more consumer “skin in the game” is limited and because private plans would have higher administrative costs and less negotiating leverage with providers than Medicare. The goal should not be to simply move costs around; it must be to reduce them overall.
The fourth approach, the provider-value approach, is more promising. Instead of reducing costs indirectly by having patients put pressure on doctors, the provider-value approach focuses on giving doctors more information and making changes so that payment is based on the quality of the services they provide — not the quantity. The goal is to boost the use of evidence-based medicine and narrow the variation in treatment methods across the United States, improving outcomes and lowering costs by reducing the number of expensive but unnecessary procedures. Data from the Dartmouth Atlas of Health Care suggest that the variation in treatment is greatest when there is little consensus about the appropriate treatment for a given condition, such as whether a patient with lower back pain requires surgery. The variation is much smaller when evidence-based guidelines exist, such as the recommendation that a hospital administer aspirin to a person suffering a heart attack. The underlying premise behind the provider-value approach is that in high-cost and chronic cases, which account for the bulk of overall costs, the patient typically agrees to the care recommended by the provider — so that the provider’s recommendation is most often the care that winds up being delivered. In the end, therefore, fundamentally reducing health-care costs requires that providers alter their recommendations. (Emphasizing prevention and wellness may also help reduce the incidence and severity of high-cost cases, but the evidence to date suggests limited success in reducing costs from such measures. Besides, a shift toward prevention and wellness requires many of the changes in information and incentives embodied in the provider-value approach.)
The potential for a better combination of cost and quality is not theoretical. The United States already has examples of institutions, such as the Mayo Clinic, that deliver high-quality health care at substantially lower costs than do other institutions. Such exemplary institutions tend to use information technology intensely, examine best practices rigorously, pay attention to doctors’ financial incentives, and focus on the nitty-gritty of management and the use of procedures such as checklists to minimize mistakes. All of which is easy to say and hard to do.
The health-care legislation aimed to address various gaps in insurance coverage, especially for those who were uninsured. And it aimed to do so without increasing (and, ideally, along with reducing) the budget deficit under conventional accounting methods — while putting in place the infrastructure to reduce long-term growth in health-care costs through the provider-value approach. The legislation includes three basic categories of measures aimed at containing costs. The first category involves blunt reductions in Medicare reimbursements. The legislation reduces the growth rate in provider reimbursement rates (by $196 billion over ten years), reduces payments to private insurance companies through the Medicare Advantage program (by $136 billion), and reduces payments made to hospitals for treating uninsured low-income patients (by $36 billion). These changes save money for the federal government, but they do not represent the type of structural cost containment necessary for the long term.
The second category of cost-containment measures involves private insurance. For example, the legislation made changes aimed at reducing unnecessary paperwork and moving toward uniform electronic standards to be used by all insurers (so that coding and other tasks are easier), which should yield an estimated savings of tens of billions of dollars a year. More important, the health bill includes an excise tax on “Cadillac” insurance plans — plans that will cost more than $27,500 for families or $10,200 for individuals in 2018, when the tax comes into effect. Plans exceeding these thresholds will face a 40 percent tax on the excess cost, creating a strong incentive to redesign them to be more efficient and come in under the threshold. Since the tax rate is effectively punitive, the vast majority of the tax’s projected revenue will not come from the insurance companies (who are ostensibly responsible for paying the tax). Instead, it will accrue as companies shift their compensation packages away from tax-advantaged health plans and toward taxable wages, which then generate income and payroll taxes. And since the threshold is indexed to increase with the consumer price index, which tends to rise more slowly than health-care costs, the tax will exert strong pressure on private insurance companies to keep their costs down so their premiums stay below the threshold.
The legislation’s third and arguably most important category of cost containment involves a variety of structural measures to prod Medicare to lead the way toward the provider-value model of health care. Some private insurance firms would like to move in this direction to save money, but the private market remains too fragmented for any individual firm to engineer a wholesale change in provider incentives. Given Medicare’s prominent role in the health system, it is necessary to put the program at the center of the effort to control costs. The act does so through both specific measures (such as imposing penalties through Medicare on hospitals with high rates of infection) and institutional changes (such as the creation of new bodies with the power to reduce cost growth over time without the need for new legislation).
The legislation was an impressive, perhaps even improbable, achievement, passed in an era of intense political polarization. It lays the basis for future structural cost containment while expanding coverage to tens of millions of Americans. But it is not perfect. The act’s shortcomings fall into two categories: those that have to do with appearance and those that have to do with substance.
The first mistake of messaging was made during the summer of 2009. At the time, the only bill in the public domain was the House legislation, which, although it expanded coverage substantially, did very little to contain structural costs. (It had plenty of reductions in reimbursements to providers, but again, that approach is ultimately not sustainable.) The administration nonetheless applauded the bill. The final legislation improved on the House bill’s efforts to contain structural costs, but by the time the act was passed the next year, it was too late. The damage had been done, and it proved difficult to shift the prevailing public and elite opinion that the measure failed to reduce spending.
The second such mistake involved the CBO, which is the official body charged with assessing the budgetary and economic impact of legislation. Given the complexity of reducing health costs, the CBO has been understandably reluctant to conclude that any individual measure would be hugely effective in doing so. As a result, there is essentially no policy that the CBO will score as exerting powerful downward pressure on aggregate health-care costs. (It is willing to score some policies as reducing federal health spending substantially, but mostly because they shift costs to other parts of the health-care system.) Barack Obama’s presidential campaign had promised massive cost savings from reform, including $2,500 a year per family. But such savings were never going to be confirmed by the CBO under any scenario. And since the House bill was relatively weak on cost containment anyway but was the first version to receive a public CBO analysis, the contrast between Obama’s campaign promises and the CBO’s forecast proved something of a shock to the public. These two mistakes of image may have been an inevitable part of the process of enacting the legislation; after all, getting the act passed was extraordinarily difficult. But they nonetheless fed the widespread impression that the act will do little to reduce cost growth.
The biggest substantive shortcoming of the legislation involves tort reform. The academic literature generally concludes that medical liability laws do little to raise costs, although some recent studies suggest modestly larger effects. The literature also suggests that variation in the medical malpractice laws across the United States explains very little of the variation in health-care costs. What this literature largely misses, however, is the fundamental problem with the laws’ standard of “customary practice” — the norm that protects doctors if they can be found to have treated their patients the way most other doctors in the area do. This basis for malpractice creates a strong contagion effect among doctors, because a doctor’s legal liability is minimized by doing what the doctor down the hallway is doing.
The traditional approach to tort reform involves imposing some limit on damages. The problem with such an approach, however, is that it does nothing about the customary-practice problem. A far better strategy would be to provide a safe harbor for doctors who follow evidence-based guidelines. Under this approach, a doctor would not be held liable if he or she followed the recommended course for treating a specific illness or condition under guidelines put forward by professional associations such as the American Medical Association or the Institute of Medicine. By failing to move forcefully in this direction, the health reform act missed a major opportunity.
CRITICISMS AND CONCERNS
Much of the criticism that the health legislation has attracted, however, has been misplaced. For example, one prominent critic, former CBO Director Douglas Holtz-Eakin, complained in a New York Times op-ed, “Gimmick No. 1 is the way the bill front-loads revenues and backloads spending. That is, the taxes and fees it calls for are set to begin immediately, but its new subsidies would be deferred so that the first 10 years of revenue would be used to pay for only 6 years of spending.” But the only reason one should be concerned about such an imbalance is if it created a fiscal hole in the tenth year and thereafter. That is not what the legislation does: it reduces the deficit not only over one decade but also in the tenth year alone. A more legitimate concern is that the legislated savings may be undone by a future Congress.
Another concern is that employers will drop coverage for certain employees and force them into the health insurance exchanges created by the act, thereby raising costs for the government, since coverage subsidies are available in the exchanges but not through employer-sponsored plans. The CBO has predicted that this will rarely happen: it estimated that by 2019, the legislation will reduce the number of people with employer-provided coverage by only three million. But critics have charged that the penalty the law imposes on firms that do not offer coverage ($2,000 to $3,000 per worker per year) is too small to act as a real disincentive.
Two factors suggest that this concern may be exaggerated: first, most firms consider coverage to be an important part of their compensation packages, meant to attract good workers, and second, the simple analysis ignores the effect of the tax subsidy for employer-sponsored insurance. In effect, if firms drop their coverage, they lose the tax preference on that component of their compensation packages, and that is often large enough to overcome the other incentives to drop coverage. Indeed, Massachusetts, which adopted a similar approach in order to expand coverage, saw a net increase in employer coverage. Nonetheless, how social norms develop among employers will be important. A July 2010 survey by Fidelity Investments found that two-thirds of large employers were not seriously considering eliminating their health plans because of the new law. But 36 percent of those firms said they would consider eliminating coverage if other firms did.
Although employers may not eliminate health-care plans en masse, they could start dropping high-risk workers by designing health plans that encourage these employees to purchase insurance on the exchanges. This is a legitimate concern. If employers altered their plans, this could create a spiral effect, in which those employees buying insurance on the exchanges would be disproportionately high-risk patients, raising premiums and defeating the purpose of risk sharing. The cost to the federal government of subsidizing coverage in the exchanges, in turn, could become unsustainable.
Another substantial concern involves the effect of the legislation on local hospital markets. Over the past two decades, these markets have become increasingly concentrated, raising prices as competition among providers has been reduced. The health legislation, if anything, will exacerbate this trend by inducing a new round of mergers among clinics, hospitals, and practices. According to Thomas Greaney of Saint Louis University School of Law, this process has already begun. “The risk that dominant providers and dominant insurers may exercise their market power, individually or jointly, has never been greater,” Greaney warns. Hospitals and other providers are already engaged in significant lobbying to relax a variety of older rules limiting health-care monopolies, especially in conjunction with the so-called accountable care organizations (ACOs) encouraged by the act. ACOs are meant to band doctors and hospitals together to provide comprehensive treatment for patients. In the words of Jon Leibowitz, chair of the Federal Trade Commission, “If accountable care organizations end up stifling rather than unleashing competition, we will have let one of the great opportunities for health care reform slip away.” The Justice Department and the Federal Trade Commission are trying to minimize this risk by instituting a mandatory review process to evaluate the largest proposed ACOs.
A final concern involves the CLASS Act, a voluntary national long-term-care insurance program created by the bill. There is a serious risk that healthy people may be reluctant to join the program, whereas those who most need long-term care will be eager to do so, jeopardizing the idea of a broad and stable risk pool. The only solutions may be to make the purchase of such insurance mandatory or to require employers to provide it by default unless employees opt out — a strategy that has worked well in boosting participation rates in 401(k) plans.
MOVING TO QUALITY
The health-care system of the future must be much more quality-oriented than today’s is. As the economist Victor Fuchs has underscored, accomplishing that requires changes in three areas: information, infrastructure, and incentives. When it comes to information, the U.S. health-care system is on the cusp of a dramatic development that could substantially expand evidence-based care. Over the next decade, hospitals and doctors will begin to adopt more information technology than ever before — a breakthrough that has been promised for many years and whose time is finally coming. Although many doctors still find it awkward to make the leap to electronic medical records, today’s systems based on tablets are less disruptive to their work than laptop-based ones. At the same time, the stimulus bill contains subsidies for the meaningful use of new information technology in medicine, which will be followed by penalties after four years for a failure to adopt such technologies. Systems like these give doctors more accurate and timely data on patients, protect against adverse drug interactions, and reduce paperwork.
The data produced by that technology could also expand medical knowledge about which treatments do and do not work. A new marketplace of data should develop. Promising steps toward this future have already been taken, including efforts such as the Health Data Initiative, a partnership between the Institute of Medicine (a part of the National Academies) and the U.S. Department of Health and Human Services that aims to boost the use of health data across public and private providers. In 2014, Medicare will begin releasing de-identified claims and data about doctors that will help patients more effectively select physicians and hospitals.
If the true potential of these data is to be realized, appropriate privacy protections must be put in place and the research itself must be funded. To lead the effort, the legislation created the Patient-Centered Outcomes Research Institute, a nonprofit organization that will help prioritize and fund new research into the comparative effectiveness of various treatments. It will disseminate the results of these studies to help doctors and patients make better-informed health-care decisions. Ideally, professional medical societies will increasingly rely on this research to issue more evidence-based protocols. The data gathered and the protocols based on them could then flow back into the health system through software that helps doctors make clinical decisions. Such a setup would be substantially more potent if it were combined with the type of evidence-based safe harbor under the tort laws discussed earlier: if the software could tell doctors not only what the best practices were but also that a malpractice safe harbor existed for those following such guidelines, the practice of evidence-based medicine would become much more common.
The second way to move the health-care system forward involves infrastructural reform. The most pressing need is to encourage providers to increase the coordination of care, and the leading idea for driving such coordination is ACOs. ACOs are designed to tie doctors and hospitals together financially and give them incentives to deliver better care to their patients on a coordinated basis. Many questions remain about how exactly ACOs will work, but the draft regulations governing ACOs issued by the administration in the spring of 2011 have begun to provide some of the answers. The Premier QUEST initiative, a voluntary project among hospitals focusing on evidence-driven improvements in their performance,* has highlighted the promise of information and incentives. By emphasizing evidence-based medicine and coordination across providers, the project has succeeded in narrowing the variation in practice norms, improving quality, and reducing costs.
The final prong involves incentives. The health-care system today is dominated by fee-for-service payment; the health-care system of the future needs to be dominated by fee-for-value payment. The difference is crucial: one payment system drives up quantity; the other, quality. The health bill takes some steps, albeit modest ones, toward creating a system based on paying for quality. For example, it creates penalties for hospitals with high rates of hospital-acquired infections and other avoidable conditions by reducing Medicare payments for hospitals in the top 25 percent of the distribution for such problems. It includes a variety of pilot programs involving bundled payments, which provide incentives to coordinate care for patients with chronic illnesses by paying a fixed sum for treating a specific condition rather than paying for each individual treatment. The legislation also imposes a penalty on hospitals with high rates of readmission; roughly 20 percent of Medicare patients are readmitted within 30 days after a hospital discharge. The lack of coordination in handoffs such as hospital discharges drives up costs (by increasing readmissions) and reduces quality (patients rarely prefer an unnecessary stint in the hospital).
All these measures will never be enough to substantially constrain the growth of health-care costs on their own. It would be shocking if they were, since the provider-value approach necessarily involves an ongoing, evolutionary process of continuous adjustment. That process is even more challenging in the United States’ polarized political environment, which makes it harder for legislation to respond nimbly to new developments and information. The success or failure of the health legislation in constraining costs will therefore hinge on how well a number of new institutions created by the law will work, that is, whether they can respond flexibly but forcefully to changes in the health-care system over time — and without requiring new legislation to do so.
One of those institutions is the Patient-Centered Outcomes Research Institute, which was designed to analyze drugs, medical tests, and other treatments and provide updated information on them for physicians and patients. The bill also created an organization called the Center for Medicare and Medicaid Innovation, which will develop and evaluate approaches to making Medicare and Medicaid beneficiaries’ care higher quality and less expensive. The act gives the U.S. secretary of health and human services the authority to scale to the national level pilot projects conducted by the center that prove successful, without the need for new legislation.
The new institution with the most potential by far, however, is the Independent Payment Advisory Board. President Obama fought hard for IPAB, over strong opposition from Congress, which saw the board as usurping its power. When IPAB starts up in 2014, it will comprise an independent panel of medical experts charged with devising changes to Medicare’s payment system. In each year that Medicare’s per capita costs exceed a certain threshold, IPAB will be responsible for making proposals to reduce this projected cost growth to the specified threshold. The policies will then take effect automatically unless Congress specifically passes legislation blocking them and the president signs that legislation. In other words, the default is that policies to constrain cost growth and improve quality will take effect.
These three new institutions — the Patient-Centered Outcomes Research Institute, the Center for Medicare and Medicaid Innovation, and IPAB — represent a powerful constellation in theory, especially in conjunction with ACOs. The question is whether they will prove to be so in practice — a question with critical implications for the fiscal future of the United States.
A FISCALLY RESPONSIBLE FUTURE
Despite popular impressions to the contrary, the new health legislation would significantly bend the curve of Medicare spending over the next several decades — assuming it is implemented in full. Medicare is only part of federal health spending, however. What about overall federal health expenditures, including the new subsidies to offset the cost of coverage for moderate- and middle-income families? Projections from the CBO suggest that the added cost of covering millions more Americans will initially exceed the cost reductions included in the legislation but that eventually the pattern will be reversed. The CBO projects that the transitional year will be 2028, after which the legislation will begin to modestly reduce overall health spending by the federal government.
And what about the budget deficit, including both spending and revenue? The bill includes a variety of measures that increase revenue, such as the excise tax on high-cost insurance plans. Altogether, if fully implemented, the legislation is projected to reduce the long-term fiscal gap facing the United States by roughly two to three percent of GDP, or about one-quarter to one-third of the underlying fiscal imbalance. Much of this effect will be driven by the excise tax on high-cost insurance plans and the impact of IPAB, measures that were excluded from the initial House bill. As the International Monetary Fund recently concluded, “The effects on the fiscal gap of the final healthcare legislation depend on the IPAB’s success at controlling excess growth in health spending going forward. If the IPAB is successful, the fiscal gap could be about 2 percent [of GDP] smaller. . . . However, if the IPAB fails to contain excess growth, the recent health reform will on net worsen slightly the fiscal gap, according to our estimates.”
In other words, if the legislation is implemented effectively — and especially if IPAB and the excise tax on high-cost insurance plans live up to their promise — it could significantly reduce the nation’s long-term fiscal imbalance. A big gap would remain, but the gap would be even larger without the health bill.
Major challenges remain for the health bill. The Supreme Court might find the individual mandate unconstitutional, Congress might underfund the implementation of the bill, and entities such as IPAB might have difficulty finding individuals willing to go through the Senate’s confirmation process. If implemented aggressively, however, the health bill holds the promise of moving the United States toward a better health-care system — one that not only leaves many fewer people uninsured but also, through the provider-value approach, improves quality and constrains costs. There are still many obstacles, and even stronger medicine may ultimately be necessary to limit future cost growth. And there is still room for much-needed improvements to the health bill, especially on tort reform. But much of the debate in the United States is still about the core approach adopted in the bill, not how to improve it.
The only prominent alternative that has been proposed is the consumer-directed one, and there is no doubt that this approach could supplement the provider-value one. Many opponents of the health legislation, however, are either implicitly or explicitly banking on a consumer-directed approach’s ability to fix health care by itself. That is not a plausible path forward, since such an approach would likely do little to address high-cost cases and therefore do little to contain overall costs.
In the end, there is no credible path to reducing the long-term fiscal imbalance in the United States other than directly addressing high-cost cases in health care. The best bet, then, is to implement and improve the provider-value provisions in the health legislation, not abandon them.