Ed Kashi/VII

A nurse consulting a doctor via Mr. Rounder the Robot, Hackensack University Medical Center, Hackensack, New Jersey, 2005


The US is facing a major crisis in the cost of health care. Corrected for inflation, health expenditures in the public sector are nearly doubling each decade, and those in the private sector are increasing even more rapidly. According to virtually all economists, this financial burden, which is now consuming about 17 percent of our entire economic output (far more than in any other country), cannot be sustained much longer. The federal share, including payments for Medicare and Medicaid, was 23 percent of the national budget in 2009 and is a prime cause of the deficit.1

There is no current prospect of raising taxes. If the federal long-term debt is to be reduced, government health expenditures on Medicare and Medicaid must be controlled. However, there is no agreement in Washington on how that can or should be done. Both parties claim to have the answer but, as I will make clear, no initiatives proposed by either party have much chance of significantly slowing the rise in federal health costs without reducing access to needed services. Major reform will be required, but that is not even under consideration. In any case, health legislation is currently stalled by a bitter political deadlock. No initiatives to improve health care will come out of Congress until after the 2012 elections and, unless the results are unexpectedly decisive, probably not even then. Still, as I will explain here, there is a chance that new developments in the way physicians are organizing themselves to deliver care might improve the currently dismal prospects for action on major reform and cost control.


In his September 8 address to Congress, the President spoke of the urgent need to control the costs of Medicare. To do this he is relying principally on the Affordable Care Act, passed by Democrats in March 2010. The Act expands insurance coverage but, as already explained in these pages,2 it is not likely to slow the rise of costs significantly. Republicans in Congress are seeking to stop it from being implemented and their potential candidates for president are demanding its repeal, while state governments that are controlled by Republicans are challenging its constitutionality in federal courts. They question the Act’s mandate that all citizens not covered by public or private insurance plans be required to purchase private insurance or incur a tax penalty. The Republican legal challenge has so far received a divided reception in lower courts and will probably reach the Supreme Court next year. The mandate is a critical part of the Act because private insurers will not offer coverage at affordable prices to all applicants, as the Act assumes, unless everyone—young and healthy included—is required to be insured. Whatever the Court’s decision, it will not have much effect on health costs because the law does very little or nothing to address some of the most important causes of the high cost of care and its rapid inflation.3

First, the Act does not replace—but expands—the investor-owned private health insurance industry. According to the actuary for the government’s Center for Medicare and Medicaid Services (CMS), this industry’s overhead and profits currently add over $152 billion to the cost of care.4

Second, the Act does not change the method of payment for most medical care, which is based on fees for each procedure and therefore encourages specialists to use expensive procedures excessively, while giving all physicians strong financial incentives to provide more services than needed.

Third, it does nothing to change the current fragmentation of medical care. This allows specialists to practice in isolation without restraints on cost, causes duplication and disorganization of services, and discourages the use of primary care physicians. The latter are essential to cost-effective care because they help patients avoid unnecessary, expensive procedures. But with specialists earning much higher incomes per unit of working time, primary care doctors are disappearing. The Act addresses this problem, but only by providing minimal financial incentives to primary care doctors; they are unlikely to make much difference.

In the absence of effective general solutions for these causes of rising costs, there is little chance that the Act will “bend the curve” of medical inflation as Democrats hope. The law does, however, propose to cover the cost of adding millions of beneficiaries to Medicaid and to private plans over ten years. It does this by cutting payments to the private “Medicare Advantage” plans now chosen by some 12 million Medicare beneficiaries, and by eliminating tax exemptions for expensive insurance coverage offered by employers to favored employees. Apart from two ill-conceived and probably unworkable exceptions, the Act offers no other broad initiatives directly aimed at reducing costs.

The first exception, in Section 3022 of the Act, proposes to share any Medicare savings with so-called accountable care organizations (ACOs). ACOs are defined as federally certified multispecialty groups of physicians that collaborate with one or more hospitals and contract with Medicare to provide comprehensive care to a designated group of patients on a fee-for-service basis. ACOs would in principle have incentives to control costs because the government would give them a share of any savings they produced by using fewer resources in delivering good care. But the 126 pages of regulations proposed to govern this program, published by the Department of Health and Human Services on April 7, 2011,5 threaten to undermine it. The regulations were considered so onerous and impractical, and the benefits to participating ACOs so dubious, that almost all multispecialty medical groups have indicated they would not participate.6


Even if the regulations are substantially modified before they are fully adopted, the CMS, which would administer the program, probably could not handle the extra work required to manage more than a relatively small number of participating ACOs. This program is therefore not likely to have much effect on health care costs.

Section 3403 of the Act is often cited as a powerful tool for controlling costs, but there is much reason to be skeptical. It establishes an Independent Payment Advisory Board (IPAB), a presidentially appointed independent committee of fifteen full-time members. If government spending rises too rapidly, the IPAB must offer proposals for cost control, and Congress must act on them or enact its own measures for equivalent savings. However, the law stipulates that the IPAB cannot reduce Medicare benefits or increase Medicare premiums, and it defers any proposed reductions in payments to hospitals for a few years. The IPAB would thus be limited to suggesting cuts in payments to physicians, payments for some ambulatory services such as emergency room visits, and payments for Medicare Advantage plans and Medicare Part D prescription drugs.

Unless the IPAB were to recommend some limitation to their total income, physicians could easily respond to reduced fee schedules by providing more services, such as performing more diagnostic tests. Physicians’ average incomes have not kept pace with the rising costs of managing an office, and there is widespread unhappiness among the profession over the threat of cuts in payment that have been mandated by past legislation. Some physicians are now refusing to see new Medicare patients for this reason, so I doubt there would be support in Congress for much further reduction in fee schedules.

Moreover, payments to physicians account for only about a fifth of total health expenditures. The other four fifths consist largely of costs of hospital and nursing home care as well as payments for drugs and other medical products, and the overhead costs of private insurers. A realistic reduction of payments to doctors could at most save one or two percent of total costs, and wouldn’t necessarily slow inflation. The other cost-saving options left open to the IPAB are also limited in scope and would have relatively little effect on “bending the curve.” Furthermore, and most importantly, there will probably be fierce Republican opposition in Congress to IPAB appointments and funding, which will hamstring its function. The IPAB’s prospects for success are therefore dim.


Republicans in Congress have pretty much limited their health policy to an unyielding opposition to “Obamacare,” to calls for reform of malpractice litigation, and to their traditional reliance on “market forces.” Their latest initiative is the proposal by Representative Paul Ryan of Wisconsin to privatize Medicare and change federal support of Medicaid by substituting fixed grants for its current open-ended commitment to paying a large share of costs to the states. Under his proposal the Medicare system covering most of the medical costs of elderly citizens would be gradually replaced by a federal voucher system. Starting in 2022 each new Medicare beneficiary would choose a private insurance plan and the government would give participants in the plan a voucher to help pay for coverage. The voucher would start at $8,000, which would grow only according to the Consumer Price Index, but with some undefined adjustment for people who are older and sicker.

However, actuarial experts predict a near doubling of total Medicare costs within a decade; so Ryan’s proposed vouchers would fall far behind actual costs, and beneficiaries would have an increasing financial burden. The Ryan plan reduces federal outlays for Medicare by shifting costs to patients. It has been passed by the Republican majority in the House, but not by the Senate, and the President strongly opposes it.

Defending Ryan’s plan, Republicans say that under current policies Medicare will soon be bankrupt. They are correct that rising Medicare costs are unsustainable, but the fact remains that the Ryan voucher plan puts an intolerable burden on the elderly and would undoubtedly reduce their access to needed care. We can predict, for example, that patients would make fewer visits to physicians for checkups and reduce their compliance with expensive drug regimens.


As for Ryan’s proposal to reduce Medicaid outlays by making fixed grants to states, the states would have considerable discretion in spending these grants, and their deficits would almost certainly force them to curtail Medicaid services in order to meet other budgetary needs. The grant idea, although favored by some state governors, has provoked wide opposition because it threatens the medical services needed by low-income citizens, and because Medicaid money now also supports nursing home care for the elderly.


If neither party is proposing effective solutions to the cost crisis, and political deadlock in Washington is preventing the consideration of new ideas, are we doomed to witness a slowly collapsing health care system that eventually will provide adequate care only to those who can afford to pay? In his latest book on health care,7 the Princeton sociologist Paul Starr, who worked on the ill-fated Clinton Health Security Plan, despairs of any political action that could bring about major reform. However, a new movement in the medical profession might help to start such reform by reconfiguring the way medicine will be practiced.

Traditionally, physicians have practiced either alone or in small partnerships with a single specialty, such as cardiology. Now a big shift to large groups containing many specialties is taking place. These multispecialty practices vary in size, but include physicians in almost all fields. Primary care physicians usually make up almost half the medical staff. They counsel patients and coordinate care by the specialists. Groups usually share common facilities, medical records, scheduling, and billing, and their physicians cooperate in the care of patients.

Many multispecialty groups pay doctors at least in part with base salaries, with the rest of their compensation derived from each physician’s collected fees. Only a few pay entirely by salary, but that may change. Most groups are owned by physicians, but many are now being organized and managed by hospitals. Well-established large groups owned by physicians include the Mayo Clinic, Cleveland Clinic, Permanente Medical Group, Intermountain Healthcare, Scott & White, Marshfield Clinic, Geisinger Health System, Lahey Clinic, and Harvard Vanguard. Many smaller physician-owned groups now exist throughout the country (though there are only a few in the Southeast), and their number is growing rapidly. So are groups owned by hospitals, including academic medical centers as well as large community hospitals. An unknown, but probably substantial, proportion of these groups are not for profit; i.e., they have no investors and use any net income to maintain and improve facilities and services.

The American Medical Group Association (AMGA) is the trade organization that represents most multispecialty physician-owned groups. The American Hospital Association (AHA) is the trade organization for almost all general hospitals. In recent communications with senior officials at the AMGA and the AHA, I learned that membership in the AMGA now stands at about four hundred groups, the highest ever, and is increasing at an unprecedented rate of about 10 percent annually. These four hundred multispecialty groups employ about 117,000 full-time medical doctors, but since many of their physicians work less than full-time, the total number of doctors now employed by such groups is probably close to 140,000, not counting the unknown number in physician-owned groups that are not members of the AMGA.

The AHA says that a 2010 survey of its member hospitals reveals that they employed roughly 120,000 practicing physicians, but the survey did not disclose how many were in multispecialty group practices. Allowing for this incomplete reporting, the total number of physicians now practicing in multispecialty groups could be close to 200,000, or about a quarter of all practicing physicians, and this number is increasing rapidly.

There are many reasons for this change. Young doctors now finish their training with substantial debt, often in the range of $150,000 or more. The prospect of assuming even more debt to start their own practice is daunting. Joining a multispecialty group means less financial risk and provides ready-made facilities and a staff to schedule patients and bill for services. Groups usually offer an income fully competitive with that to be expected in independent private practice. They often pay better fringe benefits than solo practitioners could provide for themselves.

Half of all medical students are now women, and soon they will represent nearly half of all practicing physicians. Women are more likely than men to choose some type of primary care practice, and I and others have found that many of them particularly appreciate the opportunity afforded by group practice to share responsibilities and to limit their working time. Many young male physicians are also glad to be working in a less competitive and more collegial environment, with opportunities to choose limited working hours.

In short, physicians are increasingly looking for arrangements that allow them to focus more on the practice of medicine and less on managing a business. They also want at least a little more time for personal and family life. Practice in multispecialty groups is the obvious answer. Single-specialty groups offer some similar advantages, but are not as attractive. They are smaller, have fewer resources, and will not be able to adapt as well to the new payment methods that government and private payers intend to employ.

Meanwhile, evidence has been accumulating that organized care by multi-specialty groups provides good medical service and has the potential of saving much money.8 This is because, as most clearly shown by John Wennberg and his colleagues at Dartmouth,9 much of the care prescribed and provided by independent specialists is unnecessary. Physician-owned not-for-profit groups, particularly those that pay their doctors at least partly by salaries, are not as likely to provide unnecessary services or to recommend hospitalization when it is optional. Their physicians have few financial incentives to do so, and the services of their specialists are coordinated with their primary care doctors, who usually recommend the simplest and least expensive choices consistent with good medical care.

However, today’s competitive health care market, with its private health insurance companies and fee-for-service payment system, severely limits the potential cost savings of a delivery system based on not-for-profit group practice. Private insurers contribute to high costs not only because of their profits and overhead, as already noted, but because doctors and hospitals have to pay the considerable costs of billing the insurance companies and collecting money from them. I favor a much less expensive single public payer system supported by a universal, progressive, designated health care tax. Fee-for-service payment gives doctors incentives to provide more services than needed and allows for fraudulent billing. It should be replaced by an advance payment to physicians for comprehensive care on a per capita basis.

To receive such payment, group practices would have to reimburse their physicians largely by salary. Regulations would require open enrollment of patients; to limit the risk of fixed payment for comprehensive care, groups would need public reinsurance or other financial guarantee to protect against the possibility that some patients might need expensive services.

There is much reason to believe that such a reformed system could deliver good care for all at considerably less cost and could also control cost inflation10—although such a system does not have the remotest chance of being legislated by our national government anytime soon. But what if, as seems likely, the rapid trend toward group practice that I have described were to continue, and almost all medical care were ultimately to be provided through multispecialty group practices? Despite the limitations imposed by the existing health care market, the advantages of organized care might become obvious not only to patients but to business employers and even to legislators and policymakers in Washington.

The question would then arise: If restructuring the delivery of medical care is so helpful, why not reform the insurance and payment of care as well? Supported by most of the medical profession and by most voters, major reform might finally be seriously discussed. It might become easier to pass legislation that created a system based on a tax-supported single-payer system, with prepaid comprehensive care for all. Obviously, there is no guarantee this would happen. Indeed, the odds would still be against it. But if most physicians were to be employed in multispecialty groups, the growing number of practitioners now supporting major reform would probably become a strong majority. This could well change the climate of opinion enough to influence legislation.

Dr. Francis J. Crosson, former executive director of the Kaiser-Permanente physicians’ organization and a prominent advocate for physician-managed groups becoming federally certified ACOs, believes that they will be essential for the future control of health costs.11 But unlike the broad reforms I favor, in the system he envisions these group practices would function in a competitive marketplace, in which public and private insurers would offer them financial incentives to improve care, and market forces would be expected to control prices. However, I am convinced that good affordable universal care will require major reform of all parts of the system because market forces and the current system have clearly encouraged excessive medical treatment and diagnostic testing. Insurance coverage will need to be replaced by tax-supported universal access to medical care, which should change from a competitive market to a not-for-profit social service.12

Hospitals, as we have seen, are now hiring more physicians to work in groups. As hospital employees, these physicians inevitably feel pressure to protect their employers’ income by filling beds and using hospital facilities for tests and procedures. They are less likely to control costs than physicians in independent not-for-profit groups. It is not yet clear which type of group will dominate, but those owned by physicians would probably be more supportive of reform.

The greatest resistance to reform would come from the private insurance industry. But with the savings that a reformed health system would generate, it should be possible for whatever agency collects the health care tax to buy out the industry’s investors. Some current employees of insurance companies might be usefully employed in a new single-payer system. Moreover, it seems likely that the jobs lost in health insurance companies would be offset by new jobs created through the expansion of an economy relieved of unsustainable health costs.

I do not underestimate the complexity of the task or the strength of opposition to reform by vested interests, and by conservatives who rail against a “government takeover” of health care. Reform could be delayed for years, or indefinitely. However, the status quo will soon no longer be tenable because of rising costs. New thinking about major reform and private health insurance will be necessary.

Private investor-owned health insurance appeared quite suddenly in the final few decades of the last century, reaping the profits from the health coverage that employers were offering to their employees. But except for a brief and unpopular era of cost control by “managed care” in the 1990s, the private insurance industry has not contributed anything to the health care system nearly comparable in value to the costs it has added.

In some European countries, private insurance companies are an integral part of popular and less expensive health care systems, but they are usually not-for-profit and are closely regulated by government. In Switzerland, for example, the government mandates universal not-for-profit private insurance coverage for most medical services; it also requires uniform benefits and controls costs by regulating insurance premiums and the fees paid to physicians. The government also owns most of the major hospitals. In view of our distinctive political and economic traditions, and cultural diversity, we could not easily adopt this or any other foreign system. We will have to find our own health care solutions.

Health reform might begin in the states. Vermont has recently approved a single-payer system that includes a public board to develop a global health budget and new reimbursement methods to replace fee-for-service payment. The board will help physicians to form multispecialty groups that can accept prior payments for comprehensive care of patients. Many other details remain to be worked out. Federal permission for the state to pool all government health care funds will be needed before the plan can become operative, which is estimated for 2017.13 Similar initiatives may be introduced in other states; if successful, state plans would encourage Congress to enact national reforms.


The net effect of the reformed system I have described here on the government’s costs can only be guessed at, but savings would be very large. Any initial increases in expenditures owing to expanded coverage would certainly be more than balanced by savings later on. However, government would hold the ultimate cost-controlling weapon because it would determine the rates at which it reimbursed groups for prepaid, comprehensive care, and would thereby have full control of health care expenditures. If everyone (including politicians and government officials) were in the system, there would be little chance that the government would underfund it, or allow its services to deteriorate.

Such reforms are the best—perhaps the only—hope of preserving the benefits of Medicare and Medicaid and achieving affordable universal access to care. In today’s political climate these reforms have no chance, but this could change if physicians continue to join groups and transform the organization of medical care. If encouraged by public opinion and by physicians, the members of Congress might see the clear advantages of the group practice model. They might then legislate the broad reforms in insurance and payment that would make this model much more effective. If that were to happen, the medical profession would have made an important, although unplanned, contribution to the rescue of the US health care system.