The Case Against Premium Support
On December 16, 2011, the Brookings Institution's project on Budgeting for National Priorities hosted a discussion of the proposal for Medicare premium support developed by former Office of Management and Budget Director Alice Rivlin and former Senator Pete Domenici.  That proposal, and the markedly similar proposal advanced by House Budget Committee Chairman Paul Ryan (R-WI) and Senator Ron Wyden (D-OR), would restructure Medicare by establishing a Medicare exchange in which beneficiaries would be given a "premium support" payment, or voucher, to help purchase a private health insurance plan or traditional Medicare.  The amount of the voucher (that is, the federal contribution) would be tied to the cost of the second cheapest plan in each area, as determined by competitive bidding. Growth in federal payments (either in aggregate or per beneficiary) would be capped at the rate of growth of gross domestic product plus one percentage point.
At the event, the following critique of premium support and these proposals was presented by:
- Judy Feder, Urban Institute Fellow and professor and former dean, Georgetown Public Policy Institute,
- Paul Van de Water, senior fellow, Center on Budget and Policy Priorities, and
- Henry Aaron, Bruce and Virginia MacLaury Senior Fellow, the Brookings Institution.
Medicare's Superior Record
Supporters of premium support are asking us to replace long-standing public insurance through Medicare with vouchers for private insurance and Medicare—something new and untried. Wouldn't you think their case would be based on evidence that, on the one hand, private insurers (and competition among them) have successfully guaranteed equitable access to affordable health care, while, on the other, that Medicare has done a pretty poor job?
Funny—the record shows exactly the opposite. Let me be clear: Medicare's record is far from what it could be. Its payment mechanisms—like those for all of health care—need reform. But relative to private insurance—which is the alternative being proposed—Medicare's performance is quite impressive.
First, Medicare does a terrific job at what any successful insurance plan must do. It pools risks without regard to people's health status. Private insurers, in contrast, make money, which is just what their shareholders have a right to expect, by serving the healthy and avoiding the sick. With competition among private insurance plans, it takes aggressive government intervention and oversight (which, as Paul will say more about, we have far from mastered) to assure that doesn't happen. Advocates of premium support themselves often oppose that regulation—calling its likely adequacy into serious question. In Medicare, that risk pooling just comes naturally.
Tens of millions of purchasers in a single pool also give Medicare the edge in dealing with providers, who are increasingly concentrated and therefore effective in driving up payments where they can. Medicare pays hospitals about 30 percent less than private insurers do; it pays physicians about 20 percent less. In overwhelming numbers, providers accept what Medicare pays because they can't live without the Medicare business. But that's not true for private insurers, who increasingly confront providers with close-to-monopoly power. Whether because insurers lack the clout (or, in some cases, the market pressure), private insurers have been markedly ineffective in resisting provider pressure to increase payment rates. If the goal is cost containment—or value for the dollar—it's hard to see why giving up Medicare's considerable purchasing advantage—divvying it up among relatively weak private insurance plans— would make any sense.
The difference between Medicare and private insurance is not just in the level of spending, but also in rates of growth. Historically, Medicare per capita spending has grown a bit slower than the private sector's. But when Medicare is serious about containing costs—as demonstrated by the Affordable Care Act—the difference really shows. The Congressional Budget Office (CBO) finds that Medicare premiums, currently estimated to be 11 percent lower than private insurance premiums for the same benefit package, will be about 30 percent lower by the end of the next decade.
It's important to remember that overall Medicare cost growth reflects two components: growth in costs per capita and growth in "capitas" or enrollees. Looking out over the next 10 years, the combination of the two—per capita costs times the number of capitas—leads to an overall or aggregate growth rate of 6.7 percent—faster than overall GDP. It's that aggregate growth rate that Medicare's critics focus on. But it's actually the growth in the "capitas" that's driving total spending above GDP growth—not surprising, since this year Medicare starts adding about a million and a half baby boomers to its rolls every year.
But when it comes to what health care costs per person, Medicare's growth rate is remarkably low. As a result of payment changes in the Affordable Care Act, Medicare per capita spending is projected to grow at an average rate of about 3 percent per year—as much as a point below per capita GDP growth (or, with a fix in the SGR for physician payment, at or just above the rate of growth in per capita GDP.) The ACA's reduction of $500 billion in projected Medicare spending (by slowing growth in payment rates to providers and reductions in the substantial overpayments made to the private plans in Medicare) has a big impact!
Not only did the ACA bring this growth rate below the Rivlin-Domenici plan's target of GDP+1; the Affordable Care Act has already set up a mechanism to enforce it. If Medicare spending growth exceeds GDP+1 per beneficiary, the ACA's Independent Payment Advisory Board is charged with proposing changes to lower the growth rate, which are then fast-tracked through Congress. Interestingly, the cost reductions the ACA has already put in place slow spending so much that CBO finds it unlikely that IPAB will even be activated in the coming decade.
Why, then, is this argument to shift from strong public insurer to vouchers for private insurance on the table? There are many reasons—not the least of which is ideology. But despite the fact that some of its advocates claim not to care anymore about its contribution to cost containment, it's at least partly because we'd like more cost containment than we've already got.
That's a good goal—but the way to achieve it is not to abandon the public insurance that is demonstrably superior to the alternative that's being proposed--undermining Medicare's risk-pooling and market power (its equity and efficiency) in one fell swoop; and establishing vouchers easily ratcheted down to shift costs to beneficiaries. It's to make that public insurance work better and to extend the improvements that its market power allows it to make across the whole health care system. That's what the payment and delivery reforms in the Affordable Care Act are all about. And that's the path that will allow efficiency, equity and affordable health care in the future.
Issues in Designing a Premium Support Plan
Paul Van de Water
If premium support is to have any chance of working, health insurance plans must be pressed to compete on the basis of providing value for money rather than by attracting healthy enrollees and deterring sicker ones. In this respect the Rivlin-Domenici proposal is well intentioned but still falls short in critical areas.
- First is low-income protection. The proposal says that current Medicare beneficiaries with low incomes will be guaranteed access to traditional Medicare with no additional premiums. That's fine as far as it goes, but it doesn't apply to new beneficiaries. And it doesn't apply to anyone whose income is greater than 135 percent of the federal poverty threshold. Thus, elderly or disabled individuals with incomes as low as $15,000 or couples with incomes as low as $20,000 could face increased premiums.
- Second is the scope of benefits. Today, Medicare Advantage (MA) plans can reduce the scope of some benefits if they increase others, or reduce certain cost sharing, as long as they provide the same actuarial value as traditional Medicare. The Wyden-Ryan plan explicitly adopts an even weaker actuarial value standard, and the Rivlin-Domenici plan apparently does so as well. And even if a plan nominally covered the same services as traditional Medicare, it could still fall short in the adequacy of its network, waiting times, customer services, and other features.
Although it might be possible to add elements to a premium support plan to make it seem acceptable on paper, these additional elements are impossible to enact or implement in the current political climate.
- Proponents of premium support acknowledge that risk adjustment is critical to its success. But there is good evidence that current risk-adjustment technology is inadequate to the task. Furthermore, insurance companies are trying to withhold data necessary to assure that risk adjustment under the Affordable Care Act meets its potential. 
- Any premium support system should also discourage risk selection by standardizing insurance offerings. The Affordable Care Act requires that plans offered through the health insurance exchanges provide an "essential health benefits package" and empowers the federal government to define its components. But here, too, insurers are trying to undercut the law by proposing that the essential health benefits package be defined in terms of a dollar value rather than a specific set of covered services. Both the Rivlin-Domenici and Wyden-Ryan proposals allow insurers to engage in cream-skimming through plan design.
- To compete with private plans on a comparable basis, traditional Medicare would need authority to offer an integrated benefit package (including drug and supplemental coverage), to update that package in response to changes in the health care system and the insurance market, and possibly to offer various benefit options, such as preferred provider organizations. Congress is unlikely to provide that authority, in light of the opposition to including a strong public option in health reform.
- Even if a premium support plan included the necessary consumer protections, monitoring and enforcement would be difficult. State insurance regulators complain that the federal government does not adequately protect consumers from deceptive practices by private Medicare plans. They have sought authority to enforce state laws on marketing practices. But insurers have succeeded in blocking this proposal and will likely continue to do so.
Traditional Medicare is a justly popular program, for reasons that Judy Feder has explained. The program has a strong record of innovation and cost control. Before substantially restructuring it, we need to be confident that the new arrangements will be allowed to work even beter. This is particularly important since Medicare beneficiaries face many more challenges than those of working age in dealing with a competitive, choice-based system. Many have physical or cognitive impairments that make it difficult or impossible for them to assess alternative plans or to cope with limited provider networks or other restrictions. For this group, a poorly implemented premium support plan would impose particularly high costs. Only when the ACA exchanges are up and running effectively should we consider introducing a similar arrangement in Medicare.
Judy, Henry, and I all share Senator Domenici's and Alice Rivlin's concern with constraining cost growth in Medicare, but we have a different vision of how best to achieve that goal. The Affordable Care Act takes important steps to slow the growth of health care costs through a host of delivery system reforms. As a backstop, it creates an Independent Payment Advisory Board that will make sure that spending growth per beneficiary is limited to the growth of GDP per capita plus one percentage point — the same growth rate promised by the Rivlin-Domenici and Wyden-Ryan plans. And, unlike premium support, it does so without reducing benefits or shifting costs to beneficiaries. We should not abandon the approach embodied in the ACA before giving it every chance to work.
The Changing Health Policy Environment
My text today is the remark attributed to John Maynard Keynes: 'When the facts change, I change my mind. What do you do, sir?'
In 1995 when Bob Reischauer and I coined the term 'premium support' and described its characteristics, the environment for health policy was rather different from what it is today. The Clinton health plan had failed. Hope for systemic health reform was dead. Medicare spending per person was rising faster than other health care spending. Several groups had endorsed or come close to endorsing replacing Medicare with vouchers that people could use to buy private insurance. We thought that simply dropping money on the Medicare population and asking them to fend for themselves in the famously dysfunctional small-group insurance market was a recipe for disaster.
We put forward three conditions for such a shift:
- The voucher should be linked to an index that grows as fast as overall per person health care spending;
- Insurance offerings and selling arrangements should be aggressively regulated; and
- Risk adjustment had to be good enough to make cream-skimming by insurance companies unprofitable.
Linkage to a health cost index was critical. Savings had to come from genuine efficiencies, not off-loading costs onto the elderly and disabled.
Aggressive regulation was essential because insurers have the bad habit of overloading customers with so many insurance plan variations that no one could possibly choose rationally among them. Furthermore, plan offerings and sales methods can subtly—or not so subtly—abet competition based on risk selection. "We offer sports medicine benefits. Hear about our plan after the dance. It starts at 8:00 PM, ends at 11:00. The venue is on the second floor. Sorry, no elevator."
And risk adjustment had to be good enough so that insurers could make money only by competing on what counts — better service and higher quality of care.
Well, sixteen years have passed since Bob and I wrote. Some things have changed. Alas, some haven't.
The most important change is basic health care policy. But we have also gained a deeper understanding of what we don't know how to do. And we have learned the limits of what elected officials are willing to do.
On policy...first, Medicare has changed. Most notably, for purposes of today's discussion, the sort of competitive system that voucher advocates say they want to create already exists. The average Medicare enrollee today may choose among an average of 24 plans, in addition to traditional Medicare, including 10 health maintenance organizations. Furthermore, Medicare spending per person is slated, at least for the next decade and under current law, to rise less than the targets set under the Rivlin-Domenici or Wyden-Ryan plans.
Second, and most important, systemic health reform is no longer a pipedream. It is the law of the land. The Affordable Care Act sets in motion a process of experimentation and change—long overdue, but with the potential to revolutionize how the United States pays for and delivers health care: ACOs, bundled care, comparative effectiveness research, a center for innovation, to name just some of the ways. It will also lower reimbursements to providers curbing Medicare spending by $500 billion over the next decade.
To be sure, implementation of the Affordable Care Act will be hard. States are now discovering that enrolling 29 million people in health insurance exchanges and avoiding competition based on risk selection won't be easy. They are working on those problems and I think they will solve them. But dealing with the nearly 50 million Medicare enrollees would be vastly harder. Why? Because per person spending under Medicare is three times that on those who will be served under the ACA, and variation in spending is correspondingly larger. The profit from cream-skimming is that much greater.
Furthermore, the Medicare population contains many people with mental disabilities and early or advanced mental decline. The recently announced Wyden-Ryan plan promises to provide voucher recipients with 'clear and easy to understand information' on various plans. Has any of you actually read the clear and easy to understand (!) information that Medicare and private insurers now distribute to enrollees? To think that providing 'clear and easy to understand information' equips those with mental disabilities or early-state dementia to deal with competing insurers is delusional.
So, what is the take-away?
First, contrary to the allegations of critics, Medicare works. In fact, it works better than typical private insurance plans do—providing benefits fairly and at lower overall cost.
Second, the plans put forward so far are poorly designed. They lack the regulatory teeth necessary to make premium support even worth considering.
Third, even well-designed premium support plans are not ready for prime time. We have work to do in implementing the Affordable Care Act. Only when that is done, should we decide whether to extend similar arrangements to the Medicare population.
Fourth, there are changes to be made in Medicare that will improve its operation. Among these changes would be to give CMS the administrative resources it needs to reduce improper payments and to enforce coverage guidelines. Reforms of supplementary insurance are also overdue.
Finally, controlling overall health care costs is a top health and budget priority. The urgency of that problem is why the nation's top health policy priority is to implement the Affordable Care Act, not to replace Medicare—a well-functioning and popular program—with an untried alternative distressingly similar to current plans that do a poorer job of serving the elderly and disabled.
 Pete Domenici and Alice Rivlin, "The Domenici-Rivlin Premium Support Plan," in Premium Support: A Primer, Brookings Institution, December 2011, http://www.brookings.edu/papers/2011/1216_premium_support_primer.aspx .
 Ron Wyden and Paul Ryan, Guaranteed Choices to Strengthen Medicare and Health Security for All, December 2011, http://budget.house.gov/UploadedFiles/WydenRyan.pdf .
 Jason Brown, Mark Duggan, Ilyana Kuziemko, and William Woolston, How Does Risk Selection Respond to Risk Adjustment? Evidence from the Medicare Advantage Program, National Bureau of Economic Research, Working Paper No. 16977, April 2011; Edwin Park, Ensuring Effective Risk Adjustment, Center on Budget and Policy Priorities, May 18, 2011.
 Edwin Park, Allowing Insurers to Withhold Data on Enrollees' Health Status Could Undermine Key Part of Health Reform, Center on Budget and Policy Priorities, December 12, 2011.
 National Association of Insurance Commissioners, White Paper on Regulation of Medicare Private Plans, Approved by the Health Insurance and Managed Care (B) Committee, September 10, 2008.