Statement of Michael J. O'Grady, Ph.D.,
Senior Research Director, Project HOPE, Bethesda, Maryland

Testimony Before the Subcommittee on Health
of the House Committee on Ways and Means

Hearing on the Medicare+Choice: Lessons for Reform

May 1, 2001

Problems That Still Plague the Medicare+Choice Program:

· The program still has too few incentives for plans to compete for beneficiaries by providing the best care at the most competitive price.

· The program still overpays in some counties and underpays in others.

· The program still relies heavily on fee-for-service experience to determine payments to plans. In many of our most urban and most rural counties, fee-for-service experience is not a very good indicator of appropriate plan payments.

How We Got Here:

The Medicare+Choice (M+C) program was created as part of the Balanced Budget Act of 1997 (BBA) to correct problems that existed in Medicare's Risk HMO program. The risk plans had moved into some urban counties, but few rural counties. Some of the risk plans enrolled healthier, less expensive beneficiaries, but their payments were not reduced enough to adjust for their healthier mix of beneficiaries.

Under the BBA, payments to low payment counties were increased by the creation of a guaranteed minimum amount for every county. Payments to high payment counties were not cut, but their growth rates were limited to no more than two percent per year. In addition, the Administration was to develop a "risk adjustment" formula that would more accurately adjust plan payments to account for the mix of healthier and sicker people. The BBA also included a number of across the board cuts to plans that had nothing to do with improving the payment formula, but were necessary to meet the budget reduction targets of the bill.

By 1999 it became clear that the BBA cut Medicare spending much more than the Congressional Budget Office had originally estimated. Providers of all sorts, hospitals, physicians, nursing homes and Medicare+Choice plans, were arguing strenuously to restore some of the BBA cuts.

The congressional response was the Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act (BBRA) of 1999, and the Medicare, Medicaid, and SCHIP Beneficiary Improvement and Protection Act (BIPA) of 2000. The BBRA and BIPA went a long way to offset some of the "deeper than expected" BBA cuts. However, there is still a need to improve the way Medicare+Choice plans are paid.

Essential Points in Considering Changes to the Medicare+Choice Program - Getting the Incentives Right:

There are several essential points that need to be taken into consideration in thinking about how to improve the Medicare+Choice program. Most of these points center on a common theme: getting the incentives right. A well-designed Medicare+Choice program will have compelling incentives for plans to provide the highest quality benefits in the most cost-efficient way. It will also have incentives for beneficiaries to be prudent consumers of their own health care, so beneficiaries will "shop" among plans for the best care at the most competitive price. As beneficiaries act in their own interest, they exert pressure on the plans to keep premiums as low as possible. This competitive pressure slows spending growth, which results in savings for taxpayers.

These sorts of incentives are found in the vast majority of consumer decisions Americans make every day. The vast majority of prices are set, not by government formula, but by market competition. In the Medicare+Choice program the overall price is not set by government formula, but the government contribution is set by formula. A formula that does not always do a very good job of providing strong incentives for plans to offer quality coverage at competitive prices.

The current formula bases the government contribution on the county level spending by the traditional government operated fee-for-service (FFS) plan. FFS spending does not provide a good barometer of private plan costs in certain types of counties. Using FFS spending results in paying too much in some counties and too little in others.

In many rural counties there are serious problems with a shortage of medical providers. This shortage affects the Medicare and non-Medicare populations alike. It creates severe difficulties in gaining access to needed medical care. The result is that many rural counties have low FFS spending, lower than it would be if the people in those counties were receiving the medical care they needed. Using FFS spending as the measure of "appropriate" spending and setting the government contribution based on that level significantly reduces the likelihood that a private plan will find the government contribution adequate to offer care in these counties. Compounding this problem is the fact that the Medicare+Choice program and its predecessor the Medicare risk program have traditionally been exclusively health maintenance organizations (HMO). HMO plans have seldom had much success in rural areas, especially those with provider supply and patient access problems.

In some urban counties using FFS spending as the measure of "appropriate" spending results in Medicare+Choice plans receiving overly generous government contributions. There are a number of different reasons FFS spending would be significantly higher than Medicare+Choice plan spending in these counties. For example, some counties have an oversupply of providers. Hospitals have empty beds and physicians, especially specialists, have excess capacity. The FFS plan pays providers on a piecework basis. The more procedures they perform, the more they are paid. Managed care plans tend to negotiate with providers for discounts. Beneficiaries in managed care plans pay much less if they go to "in-network" providers. As a result, in-network providers see a higher volume of patients. Providers who wish to be part of the network often have to offer the managed care plan a significant discount. The result is that the exact same set of circumstances -- oversupply and excess capacity -- tends to increase FFS spending and decrease managed care spending. The FFS incentives are to do more procedures to be paid more. The managed care incentives are to offer greater discounts to fill more beds and waiting rooms.

The Medicare+Choice plans respond rationally to the incentives they find themselves confronted with. Plans move into counties were the payment is adequate or more than adequate, for example Dade and Los Angeles counties, and stay out of counties where the payment is less than adequate, for example most rural counties.

The current Medicare+Choice incentive structure result in the following:

· In under payment counties -- plans cannot afford to operate.

· In over payment counties -- plans offer significant additional benefits, reduced cost sharing, but have few incentives to be cost efficient.

· In counties where the payment is about right -- plans offer some additional benefits and reduced cost sharing, if they are more efficient than FFS.

It is not particularly constructive to blame the plans for responding rationally to the incentives built into the Medicare+Choice program. If the incentives change, plan behavior will change.

In all fairness to the designers of the current formula, it is extremely difficult to design an efficient, effective formula for hundreds of plans in thousands of counties. And, the government does have to have some way of determining what is an appropriate contribution. However, in Medicare+Choice the problem is that the current program relies too heavily on an inefficient formula-driven approach without the assistance of strong consumer incentives for the beneficiaries and effective negotiating authority for the government.

Why worry about getting the incentives right?

While there are obvious reasons to try to improve Medicare+Choice's incentive structure, there are also a few less obvious reasons. Two of the more obvious reasons are:

· To be careful with the taxpayers' money and

· The current system results in geographic inequities in benefits.

Over paying in some counties results in additional benefits and lower cost sharing for beneficiaries. On the one hand, if beneficiaries are willing to give up some choice of providers by entering a managed care plan; they should enjoy the rewards of additional benefits and reduced cost sharing. This can be particularly attractive to lower-income beneficiaries who cannot afford Medigap coverage and are not eligible for employer-provided retiree coverage. Still, this means that lower-income beneficiaries have this alternative in some counties and not in others. This is particularly problematic in rural areas where the traditional industries, such as agriculture, have an almost nonexistent track record at offering employer-provided retiree coverage.

Less obvious reasons to try to improve Medicare+Choice's incentive structure would include:

· The baby boom is still coming. Making Medicare as efficient as possible reduces future pressure to either reduce benefits or increase taxes.

· Protect the work-related nature of the Medicare entitlement. The greater the financing from general tax revenues the less Medicare becomes a benefit people earn and the more it becomes subsidized assistance.

Without a doubt it is better to have a budget surplus, than a budget deficit. However whatever the budget situation, the demographics of the American population have not changed. The baby boom is still approaching retirement and with it will come severe financial pressure on the Medicare program. Future Congresses will face difficult choices. Raise the taxes of the subpopulation still working, cut the benefits of future Medicare beneficiaries, or both? These will be very difficult decisions to face. But before confronting such difficult choices it seems incumbent on those responsible for the Medicare program to make Medicare as efficient a program as possible.

Most seniors view their Medicare benefits as something they have earned during decades of hard work and paying taxes. It is not clear if they are aware that in reality, payroll taxes and premiums only pay for 70 percent of Medicare. The rest is subsidized through general tax revenues. As the baby boom retires, the percentage of Medicare paid by the beneficiaries is projected to shrink from 70 percent to about 40 percent. The idea that Medicare is something beneficiaries have earned and paid for will be diluted considerably over the next few decades. A more efficient Medicare program with a slower growth rate helps maintain the work-related aspect of the Medicare entitlement.

How to improve the incentives to provide the highest quality health care at the most competitive price?

· Have plans compete based on offering competitive premiums, not just more benefits and lower cost sharing.

· Improve the accuracy of the payments formula to minimize under and over payments to plans.

· Move toward a process that involves negotiation and bidding, rather than relying solely on administrative procedures.

Before BIPA, plans were required to pass on all savings to the beneficiaries in the form of additional benefits or reduced beneficiary cost sharing. Their profits were limited to no more than the percentage they made on their commercial business. Unlike much of their commercial business, these incentives put plans in competition with one another to provide more benefits, rather than lower premiums.

BIPA allows plans to compete based on premium price, as well as benefits. Beginning in 2003, plans can offer beneficiaries a rebate on part, or the entire, Part B premium. Some Medicare+Choice plans are currently called zero-premium plans, but this only means the beneficiary pays nothing beyond their current $600 Part B premium. Under BIPA, if plans provide all the Medicare benefits and still have savings, they can rebate the entire $600 premium to the beneficiary.

Table 1 illustrates how premium competition can work in practice. The column labeled "Before Enrollees Switched" shows what the annual premium increase would have been if Federal workers and retirees had stayed in the same plan they had been enrolled in during the previous year, i.e., if no one switched plans. The column labeled "After Enrollees Switched" shows what the annual premium increase actually was, after workers and retirees chose their new plan for the upcoming year. In each of the years the Bipartisan Commission found data available, the effect on premium growth was the same - workers and retirees acted as prudent consumers and the result was to slow the growth rate of the program.

The FEHBP and Medicare populations are not identical and the effect might be different for Medicare. FEHBP is only about 40 percent retirees and younger workers are probably more likely to switch plans than older workers or retirees. However, even if this consumer behavior yielded only half the savings found in FEHBP the effect on Medicare's financial well being would be substantial.

As mentioned above, BIPA has already taken the first step in this direction by allowing plans to rebate the beneficiaries' Part B premium. Premium price competition is a core component of most recent Medicare reform proposals including the Bipartisan Commission Chairmen's proposal and President Clinton's Competitive Defined Benefit proposal.

The introduction of premium price competition is a good first step, but plan payments are still based on an inefficient formula that overpays some counties and under pays others.

What options are available for increasing the accuracy of plan payments?

· Premium support strategies - base payments on the average premiums of all plans, public and private, like FEHBP.

· BBA-style strategies - set payment minimums to bring up the low payment counties and set payment maximums to slow the growth of high payment counties.

One of the key design components of premium support strategies involves using an average of all plans, both public (FFS) and private (M+C), to determine the level of government contribution. The most common example of this type of contribution formula is the one used in FEHBP where the contribution is set at 72 percent of the average premium⁽¹⁾. The average premium calculation is adjusted to take into account the relative size of the different plans' enrollments.

During the deliberations of the Bipartisan Commission this design caused concern among some Commissioners who feared the traditional FFS plan would not be able to compete effectively and the result would be higher premiums for beneficiaries remaining in the FFS plan. Using this type of formula does mean that all plans, including the government-run FFS plan, would have to operate efficiently to keep their premiums competitive.

In President Clinton's response to the Bipartisan Commission, this design feature was modified. Rather than using average premiums, the Clinton plan stayed with using FFS spending to determine private plan contributions. This modification has the effect of protecting the FFS premium. This protection means beneficiaries who remain in FFS will not see upward pressure on premiums, but it also means that the incentives to operate efficiently will be significantly weakened. The FFS plan, which currently comprises about 85 percent of the program, will not face the same competitive pressure as the Medicare+Choice plans.

Another strategy was employed in the BBA and modified in BIPA. Rather than trying to replace the underlying formula with a more accurate one, the high and low ends of the payment distribution were truncated. The BBA used a two percent growth cap for counties with the highest payment levels. This did not cut payments to this counties, it attempted to limit the growth rate until lower payment counties could catch up. The BBA also truncated the low end of the distribution as well. It set a minimum amount that Medicare would pay plans regardless of the level of FFS spending in the county.

Under BIPA the idea of a floor payment was expanded significantly. A single floor payment amount was replaced with urban and rural floors. Floor payments were increased from $415 per month to $475 per month for rural counties and $525 per month for urban counties.

For rural counties this is a significant increase, but payment floors are a short-term strategy for a problem that goes well beyond Medicare. An under supply of providers and the resulting lack of access for both Medicare and non-Medicare people, will not be solved by additional money to private Medicare plans.

While the use of payment floors and ceilings may provide some short-term relief, they do not really attempt to determine what the appropriate contribution should be. It seems quite unlikely that all the low payment counties should be brought up to these higher levels or that all the high payment counties should be constrained to no more than two percent growth per year.

Rather than choosing arbitrary minimum and maximum payments, or tying payments to only one plan, using data from all the plans, public and private, has the advantage of capturing the experience of all plans. If private plans are operating at significantly lower costs, then the contribution will come down. If private plans are operating at significantly higher costs then the contribution will increase. In addition, it ensures that all plans, public and private, face the same set of competitive incentives.

Enhance the government's ability to act as a prudent consumer through negotiation and bidding.

BIPA again made a first step in this direction. Pre-BIPA, plans filed their benefit and premium offerings with HCFA through the Adjusted Community Rate (ACR) process. HCFA determined the completeness of the application, but made no judgement on the appropriateness of the estimates and calculations involved. After the fact, the ACR's were audited by HCFA, but by that time benefits were already received and premiums paid by the beneficiaries. BIPA did not replace the ACR process, but enhanced its effectiveness by adding a review of all applications by the HCFA actuaries. Before accepting a plan's application, the government actuaries will review the data and assumptions used by the plan to set premiums and benefits. This additional review is similar to the review the OPM actuaries make of FEHBP plans premium and benefit applications.

Most reform proposals, including the Bipartisan Commission, President Clinton's Competitive Defined Contribution and the Breaux-Frist proposals, would go further than an actuarial review of data and assumptions. They all move toward approaches that rely on some combination negotiation and bidding.

The core element of all these proposals is to attempt to construct a set of incentives where plans do well if they provide high quality benefits and a competitive price. The government itself can become a more prudent consumer of health care and in doing so provide more benefits to beneficiaries at a lower cost to tax payers. Still that process is greatly strengthened if the incentives are such that the beneficiaries also become prudent consumers of their own health care. Beneficiaries' shopping to find the best value at the most reasonable price creates substantial competitive pressure on plans to offer the best benefits at the lowest price.

These two changes, basing payment on the experience of all plans, not just the fee-for-service plan and moving toward negotiation and bidding will go a long way toward getting the incentives right. These changes move the Medicare+Choice program away from a system that rewards plans for gaming a payment formula to system that rewards plans who best meet the beneficiaries' needs for the best benefits at the most reasonable price.

1. FEHBP also has a provision that the government contribution will not exceed 75 percent of the total premium.