Medicare Reform and Risk

by Reihan Salam

Almost two years ago, I had the pleasure of reading Bring Market Prices to Medicare, an excellent book by three health economists, Robert F. Coulam, Roger Feldman, and Bryan E. Dowd, that offers an innovative approach to fixing Medicare, which the authors also described in an AEI policy brief:

 

Medicare should be an entitlement to a set of benefits, not to a particular way of financing or delivering those benefits. However, under the current structure of the program, federal payments for Medicare-covered services differ depending on whether the beneficiary is enrolled in the FFS Medicare plan or through private MA health plans. This arrangement creates inefficiency and raises Medicare costs unnecessarily.

There is a solution: full competitive pricing that applies equally to FFS Medicare and MA plans. …

Medicare faces a fiscal crisis. The 2009 annual report of the Medicare trustees warns that the Medicare hospital insurance trust fund will pay out more benefits than it receives in revenues by 2017,[2] making it crucial that the government pay only the cost of providing the Medicare entitlement benefit package from the most economical health plan in each market area. The way to do that is to have all health plans–both FFS Medicare and MA plans–submit bids for the entitlement benefit package, and then set the government’s contribution to premiums equal to the lowest bid in each market area. There is nothing particularly novel about this proposal. Many employers nationwide have used similar competitive bidding systems to induce competition among the health plans that are offered to their employees.

So that’s the basic idea. You’ll notice that there is a crucial difference between the approach offered by Coulam, Feldman, and Dowd and the Medicare premium support concept offered by House Budget Committee Chairman Paul Ryan: while Ryan transforms the Medicare program into a limited defined contribution towards the purchase of medical insurance, Coulam, Feldman, and Dowd offer a defined benefit.

That is, Ryan’s approach, which resembles that of James Capretta and Thomas Miller, will grow premium support at a certain level that might fall short of the cost of a standard benefit package for at least some Medicare beneficiaries. To be clear, Ryan would offer a higher level of premium support to older, poorer, and sicker patients, which will mitigate this problem. But the basic idea, as I understand, is that Medicare beneficiaries will respond strongly to the price signal of having to pay more for a standard benefit package, and this will encourage private insurers to organize the delivery of care more efficiently, thus restraining medical inflation overall. If medical inflation continues to gallop ahead, however, some number of Medicare beneficiaries will be exposed to risk. 

Capretta and Miller call for setting the level of premium support through competitive bidding:

Determining first the competitive price for core Medicare benefits in a relevant market for the average Medicare beneficiary. Then, applying other subsidy adjustments to deal with the peaks and valleys of variation in the income levels and/or health-risk profiles of particular collections of beneficiaries at the health plan level (that is, additional premium assistance for lower-income beneficiaries and risk adjustment for plans that attract unusually large collections of high risk or low-risk beneficiaries, whose costs of care are likely to differ substantially fromaverage levels).

In the Heritage Foundation’s “Saving the American Dream” plan, the authors use a Medicare spending cap:

During the first five years of the new Medicare program, the government’s annual contributions to enrollees’ plans are based on the weighted average premium of participating health plans’ bids on a regional basis. The plans bid to provide Medicare benefits plus catastrophic coverage and, just like the FEHBP, are weighted on plan enrollment. Thereafter, the government contribution is based on the premium bid of the lowest-cost health plan that meets the required level of quality and provides an adequate range of benefits. In both cases, the per capita government contribution on the basis of the plan bidding is set at 88 percent of the bids. By comparison, the FEHBP contribution is set at 72 percent of the national average weighted premium, and the original Medicare Part B premium contribution was set at 50 percent in 1965.

The Heritage plan also caps total Medicare spending. The spending cap is indexed annually for inflation using the Consumer Price Index plus 1 percent and Medicare population growth. If Medicare spending exceeds the cap, the government’s contribution declines from 88 percent to the percentage that complies with the Medicare spending cap, thereby pressuring the competing plans and providers to control costs more tightly. [Emphasis added]

There is a similar cap in PPACA, but breaching the cap triggers a response from IPAB. The Heritage approach doesn’t strike me as intrinsically unreasonable. But it does rest on the assumption that competition drive down prices. I believe that this strikes many of us — it certainly strikes me — as a no-brainer, but what if consolidation of medical providers, driven by the creation of ACOs and overregulation, winds up cutting in the opposite direction? Shifting the locus of responsibility is precisely what makes this approach so politically controversial. 

So what if, per Coulam, Feldman, and Dowd, we offered a defined benefit approach? The obvious downside relative to the Heritage approach is that while it limits the risk exposure of Medicare beneficiaries, it increases it for taxpayers. On the other hand, it is much easier to spread the risk across all taxpayers, as there are a lot more of them. 

Note that Coulam, Feldman, and Dowd are still creating cost pressures. Medicare beneficiaries will receive enough premium support to pay for the lowest bid in a particular market. Being the lowest bidder will guarantee lots of customers, as people will want to pay as little out-of-pocket as they can. This will encourage participating insurance providers to organize the delivery of care as efficiently as possible. 

Austin Frakt has argued for this approach:

 

One charge against the Republican plan is that beneficiaries would be on the hook for every dollar increase in health-care costs above voucher levels. That’s very different than current Medicare, and it frightens people. Competitive bidding is different. It guarantees a subsidy that is sufficient to purchase at least one plan for the standard Part B premium. That offers beneficiaries considerable protection against health care cost inflation, though it does not guarantee that the cheapest plan is one they prefer. It may not be traditional Medicare, for example.

Another charge against the Republican plan is that it puts voucher levels on the same footing as today’s MA payment rates, which have been driven sky high by the political process. Granted, nothing can stop a future Congress from doing what it can pass, but at least establishing an apolitical, market-based process for setting subsidies clearly conveys intent to depoliticize them. Putting the system within the purview of an IPAB-like board would further insulate it from politics. In this way, competitive bidding offers taxpayers protection against the type of cost run-up experienced in the MA program.

So, competitive bidding is market-based, involves private plans (hence, offers choice), and protects against program over-runs, so it should attract the interest of conservatives. Yet it includes a public option and protects beneficiaries from health care cost inflation, so it should attract the interest of liberals.

As he observes later on, however, 

Competitive bidding is not a panacea for Medicare. It cannot tell us what the standard set of benefits upon which plans bid ought to be. It would mean that the beneficiary cost of traditional Medicare coverage, as well as private plan coverage, would vary across markets, a feature some might consider inequitable. Though it has been estimated to save 8 percent of Medicare costs, it cannot, by itself, change the program’s growth rate. For that, further reforms to how traditional Medicare and MA plans pay for care would be required, as well as changes health system-wide.

I emailed Bryan Dowd, one of the co-authors, about the 8 percent savings estimate:

Though you claim modest potential savings for your proposal, do you think that your approach might actually generate very big savings as private plans organize care delivery in more efficient ways? That is, might you be underselling the upside as opposed to the downside potential of your proposal, and if so why?

Professor Dowd kindly replied:

Yes, we probably are, but we have tried to avoid speculative forecasts.  We have estimated the “static” savings from competitive pricing at around 8 percent of Medicare costs.  I expect there also would be “dynamic” savings of the type you describe once a larger proportion of beneficiaries were in price-competitive private plans and FFS Medicare was forced to respond with more aggressive purchasing strategies of its own. 

That is, Professor Dowd and his colleagues erred on the side of caution.

I’ll write more on the kind of system-wide changes we’d need to facilitate cost-control,