Thursday, April 21, 2011

What Can We Learn about the Ryan Medicare Plan from German Experience?

Last week Republicans in the US House of Representatives, following the lead of Representative Paul Ryan, endorsed a far-reaching plan to reform Medicare, the nation's health care system for the elderly. Since it began in 1965, Medicare has been a government-run, single-payer system that reimburses private doctors and hospitals for the health care services they provide. Under the Ryan plan, it would be transformed into a system in which seniors would choose from a list of  competing private insurance plans, with the premiums paid partly by government and partly by the beneficiaries themselves.

Supporters of the Ryan plan see several benefits. An open letter, posted on the web site of the American Enterprise Institute and signed by a list of prominent physicians and economists, puts it this way:
Having more control over their health care spending would encourage consumers and patients to make better health care choices. It would stimulate more innovative and accountable competition by health care providers and give them incentives to better coordinate the care of their patients. Enhanced competition could offer seniors relief from rising Medicare premiums. Just as important, this reform could begin to ease the crushing tax burden imposed by the current program on our children and grandchildren.
Critics fear that in its zeal to ease the burden on taxpayers, the Ryan plan would make Medicare-equivalent health care unaffordable for many, if not most seniors. Under the plan, the value of government payments would be capped at the rate of growth of the Consumer Price Index. If medical costs continued to grow faster than the CPI, as they have in the past, more and more of the financial burden of health care would be shifted over time to beneficiaries. A study from the Center for Economic and Policy Research, using assumptions from the Congressional Budget Office, claims that by 2022, a senior citizen at the median income would have to pay 35 percent of that income to obtain coverage equivalent to Medicare, with the figure rising to 68 percent by 2050.

The idea of competing private health insurance plans, with premiums split between beneficiaries and the government, is far from new. It has been used for many years in Germany, among other places. What insights regarding the likely effects of the Ryan plan can we get by looking at the German experience?

Under the German system, seniors choose insurance coverage from among a list of approved, nongovernmental "sickness funds" (Krankenkassen). As in the United States, those insurers, in turn, pay for health care provided by private physicians and hospitals. Beneficiaries and the government each pay a share of health care premiums. In these respects, the German system is closer to the Ryan plan than to the current version of Medicare.

Does the German system work? Broadly, the answer is yes. A variety of international comparisons rank the performance of the German health care system above that of the United States. That good performance is achieved at a cost of 10.5 percent of GDP, compared with 16 percent here. In dollar terms, Germany spends less than half as much per capita on health care. (See the slideshow attached to this post for details.)

So far, so good. Germany has competing private health insurers with premiums paid in part by the government and in part by beneficiaries. It achieves better health care outcomes at a lower cost than the United States. But before we jump to the conclusion that the German experience supports the claims of the Ryan plan, we need to look more closely not just at the results the German system achieves, but at how it achieves them.

In contrast to the rhetoric of Ryan-plan supporters, the German system does not achieve its results primarily by unleashing the forces of competition. In fact, both the German system and the Ryan plan explicitly prohibit one of the main forms of competition among insurance companies, namely, the use of experience rating, that is, the practice of differentiating premiums according to demographics, health status, past health care use, and similar factors. Experience rating in health insurance leads to "cherry picking," in which insurers compete to lower their premiums by excluding all but the healthiest customers. Under such a system, the very elderly and those with chronic illnesses are likely to find that insurance is unaffordable or completely unavailable.

Unfortunately, prohibiting experience rating also weakens the extent to which competition encourages people to make better health care choices. Lifestyle decisions like smoking, diet, and exercise no longer affect the availability of insurance or its price. Instead market forces must operate indirectly, through competition among insurers to devise packages that offer consumers the best value for money. One way to do so is to assemble provider networks of doctors and hospitals that offer good quality services at reasonable prices. Another is to cover only treatments that are known to be cost effective. Still another is to encourage preventative care, when doing so reduces total health care costs.

Would those forms of competition among insurers be enough to achieve the slowdown in growth of health care costs that supporters of the Ryan plan hope for? Competition based on cost-effectiveness already exists in some sectors of the US health care market, for example, in the provision of employer-paid group health plans and in the Medicare Advantage program. Health care costs have continued to rise faster than the CPI in both of those cases, and many observers are skeptical that the Ryan plan would lead to a different outcome.

Germany, too, encourages sickness funds to compete in providing patient-friendly and cost-effective care. However, it backs up the effects of competition with an array of proactive cost-control regulations, including the following:
  • National and regional budgets to cap total quarterly health-care outlays.
  • Diagnosis-related groups to reimburse hospitals for care of patients with specific conditions, rather than paying for each test and procedure performed.
  • A reference-price system for drugs that provides for similar payments for drugs with equivalent therapeutic effects, favors use of generic drugs, and encourages pharmaceutical companies to concentrate on innovation rather than developing follow-on drugs with little added benefit.
  • Disease management programs for chronic conditions like diabetes and heart disease, which use evidence-based guidelines for cost-effective treatment protocols and focus on preventing high-cost complications.
  • A national institute that assesses the effectiveness of medical treatments and products; encourages speedy introduction of new treatments that offer added value; permits the use of new treatments that provide similar value to existing ones, but without higher reimbursement; and disallows reimbursement for new treatments that have no demonstrated added value.
Germany's health care system, which dates back to the rule of Otto von Bismarck in the 19th century, did not always have these cost-control features. They were introduced gradually, many of them over just the last 20 years, in an attempt to curtail a tendency for medical costs to grow faster than the rest of the economy. They have not been completely successful, but the excess growth of health care expenditures has been less than in most other developed economies. In the decade from 1999 to 2008, health care costs grew only from 10.3 to 10.5 percent of GDP in Germany, compared with an increase from 10.1 to 11.2 percent in France and from 13.4 to 16 percent in the United States. (In the preceding decade, before many of the cost-control regulations were implemented, German health care expenditures had grown from 8.3 percent to 10.2 percent of GDP.)

There is one more difference between the Ryan plan for Medicare and the German system. The former is what is popularly described as a "voucher" program. It provides beneficiaries with a fixed payment toward the purchase of health insurance that increases only with the general cost of living. Any excess growth of health care costs beyond the CPI is borne by the individual. That formula assures that as the economy grows, the government's share of health care costs decreases as a share of GDP regardless of what happens to total costs. At the same time, the share of a typical senior's retirement income that goes to health care would increase over time unless total health care costs decreased steadily as a share of GDP, something that has never happened in any developed country.

In contrast, the German system uses a premium support approach. Beneficiaries and the government share the cost of health insurance, but the individual's share is capped as a percentage of his or her income. Any excess increase of health care costs relative to the CPI or to GDP is borne by the government. Long-time US backers of the premium support concept, like Henry Aaron of the Brookings Institution, strongly object to applying that term to the Ryan plan. So does Alice Rivlin, who worked with Ryan on an earlier Medicare reform plan, but vehemently rejects the version that appears in the current Republican Plan for Prosperity.

Taking all these similarities and differences into account, then, what do we learn about the Ryan plan from the German experience? I see three lessons.

First, there is nothing inherently unreasonable in the idea of replacing today's single-payer, government-run version of Medicare with a system that offers beneficiaries a choice among competing private plans. The Germans do it, and it works.

Second, the German experience suggests that it would be unrealistic to rely on unregulated market competition to hold the rate of growth of health care costs to the rate of CPI inflation. That has never happened anywhere, not in any foreign country, and not in any sector of the US health care system where competition has been tried. The point is an important one because Republicans tend to have an ideological aversion to government regulation in all forms. One cost control measure, medical malpractice reform, does pass the Republican litmus test, but, although helpful, it solves only a small part of the problem. If the Ryan plan is to have any hope of achieving its promised results, it will have to embrace a full-court press of cost control regulations.

Third, if the Ryan plan were implemented, and if health care costs did continue to grow faster than the CPI, the burden of health care for seniors would increasingly shift to individuals. If that happened, there is no conceivable way the American political system could resist the resulting political pressure to increase benefits. There are just too many senior voters. Anyone with the slightest sense of realism must realize that a clash between growing health care costs and lagging government benefits would be resolved through annual Congressional approval of benefit increases, just as used to happen with Social Security before benefits were indexed, and just as happens now with the alternative minimum tax.

In short, the Ryan plan, as it stands, does not represent a realistic path forward. Either it will not be enacted, or it will be modified to look more like the German system before it is enacted, or it will be enacted first and modified later. Meanwhile, it will remain more of a political platform than a serious policy initiative.

Follow this link to view or download a brief slideshow with additional information on the German health care system.


    1. "it would be unrealistic to rely on unregulated market competition to hold the rate of growth of health care costs to the rate of CPI inflation. That has never happened anywhere [...] not in any sector of the US health care system where competition has been tried."

      Lasik surgery?

    2. Clarification: Yes, certainly costs for specific procedures have fallen after an experimental procedure becomes routine, or after a technological advance (sometimes), or after a patent expires allowing generics. By "sector" I mean, not in the employer group sector, the Medicare Advantage sector, the individual policy sector, etc.

    3. Ed: Can you clarify on healthcare cost growth pre-medicare (pre-1965) vs. CPI inflation at that time - was it also excessive?

    4. Ed, this is a most excellent post. I have similar thoughts about the Ryan Medicare plan. Structurally it's not really all that radical, as you point out, but the lack of real cost control measures combined with the CPI indexing makes it unworkable.

      If we want market forces to work in health care, I suspect we'd have to get rid of insurance altogether and make patients pay for health care directly. Health insurance suffers from moral hazard, as the presence of a third-party payer means that patients and doctors don't have any incentive to choose efficient treatments that could save money. In theory, insurance companies could force providers to organize in a way that changes those incentives, but in practice they don't seem to have the leverage to do so.

      Perhaps some combination of mandatory health savings accounts combined with government subsidies and even a single-payer system for certain treatments could do the trick. Even then, additional cost control regulations would probably be necessary.

    5. There must be some kind of Medicare reform put in place in the near future.

    6. As cost is shifted to individuals that necessarily reduces demand for some things. Lower demand, ceteris paribus, leads to lower prices.

      Reducing the amount of money people can spend freely reduces pricing. The marginal consumer foregoes some consumption and moderating demand.

      Ultimately this is a supply and demand question, and cost shifting to individuals is a demand management approach.

      Your analysis assumes no effect there, which I think misses the point.

      One element of the Ryan plan should be refundability. If you underutilize consistently you should get a premium refund on a portion.

      That would encourage some economizing.

      Further, I am fairly sure the Ryan plan meets the standard of premium support. The support level is progressive with level of need, as far as I know.

      The only major difference from the Rivlin-Ryan approach was the ability to opt back into an actuarily similar plan offered by Medicare.