Our project will examine how the incentives created by government procurement of prescription drugs influence pricing and innovation by pharmaceutical firms. Our current research focuses on two potential effects of the reimbursement rules in the Medicaid program. The first is the direct impact on price levels due to the provision of insurance to a large fraction (19%) of the market. The price Medicaid pays for a pharmaceutical product is a function of private market prices; as a firm raises its price to non-Medicaid customers, it will receive a higher price for all its Medicaid sales. Quantity decline in the Medicaid segment is minimal because consumers have inelastic demand due to full insurance. We empirically investigate whether the Medicaid reimbursement rules lead firms to increase private market prices above what they otherwise would be, and find preliminary evidence that this is the case.
We also examine whether a rule limiting price increases on Medicaid products causes pharmaceutical firms to introduce new versions of their treatments. Prices for pharmaceuticals have increased much more rapidly than the CPI, and Medicaid's share of prescription drug sales increased substantially during the 1990s. A firm might therefore want to increase prices to the Medicaid segment faster than growth in the CPI, but the rules disallow this. However, a firm can introduce a new version (coating, capsule, dosage) of its drug and launch it at a higher price. We test this hypothesis on a dataset of new product (NDC) introductions in the late 1990s and find evidence that firms do introduce new products in response to Medicaid share when the branded drug has large sales to Medicaid.
We will build on our first set of findings by exploring the robustness of our results with a variety of specification checks detailed below, such as estimating our model on generic drugs and on particular drugs with plausibly exogenous increases in Medicaid demand. We will also compare prices to some external benchmark that is totally unrelated to Medicaid, such as drug prices in New Zealand, where the central government aims to set prices based on the health benefits of the drugs. If the incentives we describe above are important, we expect to find a larger deviation between US and foreign prices for drugs with higher Medicaid market shares.
Increased profitability in the Medicaid market may lead more firms to conduct research on drugs used differentially by Medicaid recipients and thus increase the number of available drug treatments. This innovation effect could to some extent (perhaps fully) offset the loss in consumer surplus incurred by private sector consumers and perhaps reduce firm profits from the Medicaid sector. Our initial strategy is to examine the effect of the large and unanticipated increase in the Medicaid market share that occurred in 1984 and 1990 as a result of expansions in eligibility for the federal Supplemental Security Income (SSI) program. The growth in SSI differed substantially across diagnosis groups and thus certain categories of prescription drugs were more affected than others. We can then use information from the FDA and other sources to estimate the number of clinical trials and other measures of innovation by drug category and year. Determining the magnitude of this innovation effect will allow us to provide a more comprehensive estimate of the effect of Medicaid on social welfare.
In contrast to Medicaid, the federal Medicare program does not provide coverage for prescription drugs, though this is set to change as a result of recent legislation. Beginning in January of 2006, Medicare will give its 45 million beneficiaries the option to purchase insurance from tightly regulated private firms. These firms will have some discretion over which drugs to include in their formularies, though not over the fraction of costs that they must cover. Individuals who are dually eligible for Medicare and Medicaid will shift into these plans and thus Medicaid's share of the drug market will decline. While "dual eligibles" account for just 13% of Medicaid recipients, they account for 60% of Medicaid prescription drug spending. The change in Medicaid's market share will therefore be substantial and will vary both across and within therapeutic categories. We will exploit two natural experiments: the decline in the Medicaid market share and the change in the insurance coverage of Medicare enrollees. This latter effect is complicated by several factors - Medicare recipients may choose not to purchase the insurance, the coverage may crowd out insurance that would otherwise have been purchased, and private insurers may have incentives that affect the prices firms set for their drugs. Our research will investigate the effect of this policy change on pharmaceutical prices, innovation, and other outcome variables of interest.
Many have argued that U.S. pays too much for prescription drugs. The feedback from public to private prices described above may partially explain why pharmaceutical prices in the US are relatively high. While any method of procurement has its problems, in our view, up to this point policy makers have not focused sufficient attention on the effect that linking public and private markets has on the prices paid by consumers in the private sector. Our results take on additional policy significance when one considers that government involvement in this sector will soon expand by a great deal due to the new Medicare prescription drug benefit. Understanding the effects of the program and its procurement structure are extremely important for designing policies that control spending on healthcare programs, protect current consumer surplus, and allow for socially beneficial innovation.
