A proposal to require manufacturers to pay a minimum rebate on drugs covered under Medicare Part D for those beneficiaries who receive the program’s Low-Income Subsidy (LIS) has received considerable attention during the current debate over the federal budget.  Through mandating discounts in the form of rebates, this proposal would ensure lower drug prices than currently negotiated by private Part D drug plans.

A study by the HHS Inspector General in August 2011 found that rebates negotiated by private plans average about one-third the size of those received in the Medicaid program, where most LIS beneficiaries received drug coverage prior to the implementation of Part D.  The proposal is scored by the Congressional Budget Office as achieving savings of $137 billion over 10 years (2013-2022) or about $15 billion in the first full year of implementation.

The rebate approach has been part of Medicaid drug pricing for over 20 years and includes three components.  The rebate is set as the greater of (1) a minimum rebate, currently set at 23.1 percent of the average manufacturer price (AMP) for most brand-name drugs and (2) the difference between the AMP and the “best price,” defined as the lowest manufacturer price paid for a prescription drug by any private purchaser.  In addition, if a brand drug’s AMP rises faster than inflation, the minimum rebate is increased based on the change in AMP compared to the consumer price index.

It has come to our attention that a number of interested parties have signed a letter, sponsored by the Council on Affordable Health Coverage, expressing concerns about this budget proposal. The analysis upon which the concerns are based is that by Douglas Holtz-Eakin and colleagues. They offer two arguments against this policy, which they claim would increase Part D premiums by 20 to 40 percent and shift costs to employers, Medicaid, and non-LIS Medicare beneficiaries. They are:

Two Flawed Arguments Against The Medicare Part D LIS Drug Rebate Proposal

The first argument: price distortions for private purchasers. The first argument against the Medicaid drug rebate proposal for Part D LIS enrollees concerns distortions in the prices paid by private purchasers as a result of the rules governing the Medicaid “best price” method of price determination.  Research by Fiona Scott Morton (“The Strategic Response by Pharmaceutical Firms to the Medicaid Most-Favored-Customer Rules,” RAND Journal of Economics, 1997) addressed the idea that regulating Medicaid prices by linking Medicaid to the best private price inhibits the granting of rebates by manufacturers to their private plan customers.

The empirical evidence suggests that in fact, some small distortion in private prices does occur as a result of the Medicaid rules.  However, Scott Morton concludes that the incentive effects of Medicaid pricing incentives “are visible, though small.”  She estimates that, on average, prices for drugs that face generic competition increased by 4 percent because of the Medicaid rebate provision.

It is important to note that since Scott Morton conducted her study, the rate of generic penetration has increased dramatically.  Many more drugs have lost patent protection, and the share of Medicare Part D prescriptions filled by generics has exceeded 75 percent (although representing just 25 percent of drug spending).  Typically, about 80 percent of sales shift from branded to generic products within six months of patent expiration.  Because the number of branded sales is lower today than in 1991, when Medicaid drug prices were linked to the best private price, the impact of any price increases will be smaller than shown in her study.

The central policy question is: are these price distortions large enough to offset the savings realized in a segment of the market (Medicare Part D) where prices currently exceed those that would result in a well-functioning market?  Scott Morton’s estimates, together with the changes noted in the market for prescription drugs, suggest that they would not be.  CBO’s analysis of the budgetary savings from this proposal is consistent with this view.

The second argument: price distortions for other Part D purchasers. The second argument, offered by Holtz-Eakin and Michael Ramlet, extends the analysis offered by Scott Morton to an area not covered by her research.  They state the following:  “Such an impact [of the Medicaid rebate on private payer prices] would likely be felt in the employer based insurance market, as well as in government programs including Part D.”  The core of this argument also relies on the “best price” piece of the rebate approach.

The first part of the statement made by Holtz-Eakin and Ramlet was addressed in the previous section in the discussion of concerns about the impact on private payer prices.  The second part of their statement suggests that not only will there be price distortions in private markets but that other Part D beneficiaries will also pay higher prices because discounts negotiated by Part D plans will go down.

The implications here are two-fold.  First, they imply that the “best price” formula will affect Part D prices.  Since the rebates negotiated by Part D plans are explicitly excluded from the “best price” criterion, it is not clear what such a mechanism could be.  Second, they imply that manufacturers are not pursuing profit-maximizing prices for Medicare Part D plans in today’s marketplace.

Because Part D prices are excluded from the Medicaid formula, the assertion by Holtz-Eakin and Ramlet implies that manufacturers could have set Part D prices higher than they do under “current law” but chose not to, thereby leaving money on the table.  The notion that the industry consciously has been pricing so as not to maximize profits flies in the face of decades of economic analysis of the industry and Holtz-Eakin’s and Ramlet’s own conception of this industry.  Therefore this concern does not likely pose a meaningful risk to the effectiveness of the Medicaid drug rebate proposal.

Furthermore, Holtz-Eakin and Ramlet project their largest Part D premium increase (40 percent) by assuming that Part D plans will segment negotiations between LIS and non-LIS offerings and that the impact of higher prices will fall solely on the non-LIS offerings.  This logic fails on several grounds.  First, plan sponsors are not permitted to restrict plan offerings to either group of beneficiaries.  If premiums were increased for plans focused on the non-LIS population, these beneficiaries could switch to the LIS-focused plans.

Second, CMS guidance requires that the premium of each separate plan is actuarially sound on its own.  As a result, CMS would reject plan bids that allocate costs differently across different plan offerings.  Finally, even if plan sponsors could take this segmented approach, it seems reasonable to assume that they would prefer to raise premiums in their LIS-focused plans since the federal government pays those premiums for most LIS enrollees.  This would result in higher government costs and would have lowered CBO’s scoring for this proposal and would preserve their ability to offer plans at an attractive price to non-LIS enrollees.

In the end, the analysis by Holtz-Eakin and colleagues fails on both empirical and logical grounds. The projections they make are not grounded in the existing evidence base, and the economic logic is inconsistent with everything we know about the industry.