How Insurers' Bargaining Power Affects Drug Prices in Medicare Part D

The Medicare Modernization Act of 2003 expanded Medicare to include a prescription drug benefit, known as Medicare Part D. One of the most controversial features of the law is that it tasked private insurers with negotiating drug prices with retail pharmacies and drug manufacturers, rather than having Medicare negotiate a single price on behalf of all beneficiaries as some legislators would have preferred.

In "Insurer Bargaining and Negotiated Drug Prices in Medicare Part D" (NBER Working Paper 15330), researchers Darius Lakdawalla and Wesley Yin explore whether greater concentration among private insurers offering Part D plans allows them to obtain lower prices for their members.

This is an interesting twist on the usual situation. Increasing the market power of a firm is normally viewed as undesirable, as it may allow the insurer to charge their members a higher price. But in this case, private insurers may use their ability to deny pharmacies or drug manufacturers access to their members to extract some of the profits normally earned by these companies and may pass these on to members in the form of lower premiums. Under this scenario, greater market power may be good for consumers - the more enrollees an insurer has, the more bargaining power they have with pharmacies and drug manufacturers - so long as the market does not get so concentrated that insurers are able to charge consumers a large mark-up over their costs.

From a policy perspective, this mechanism creates spillover effects of Part D. Insurers that experience Part D-related increases in enrollment may negotiate lower drug prices, which extend to both their Part D plan enrollees and their population of non-Part D commercial enrollees. The sheer number of non-Part D enrollees in the marketplace magnifies the importance of these spillovers, compared to the direct effect of the program on its own enrollees.

The authors use pharmacy claims data from a large national retail pharmacy in their analysis. The data set includes the drug prices negotiated between the pharmacy and every insurer with which it contracts. The claims data covers all prescriptions filled between September 2004 and April 2007 for a 5 percent sample of the pharmacy's patients. Medicare Part D was implemented on January 1, 2006, in the middle of the sample period.

The authors' basic empirical strategy is to examine whether those insurers who experienced larger increases in enrollment following the implementation of Part D also had larger decreases in the prices they paid to pharmacies. This specification controls for any characteristics of the drug, insurer, or geographic market in which the insurer operates.

Turning to the results, the authors find that insurers that experience larger enrollment increases due to Part D implementation negotiate lower drug prices with pharmacies. Specifically, enrolling 100,000 additional members is associated with a 2.5 percent decrease in drug prices and a 5 percent decrease in pharmacy profits earned on prescriptions filled by enrollees of that insurer.

Insurers also negotiate directly with drug manufacturers, as manufacturers may offer rebates in exchange for including their drugs in the insurer's formulary (set of covered drugs) or placing them on a more desirable "tier" (assigning them a lower copayment). The authors do not directly observe drug manufacturers' prices and profits. However, economic theory indicates that when an insurer obtains greater bargaining power, the profits and mar-kups of both pharmacies and drug manufacturers will move in the same direction. Thus their estimates of the effect of enrollment on pharmacies represent a lower bound on the effect of enrollment on total drug costs.

Next the authors explore whether their results differ by type of drug. They find that for both generic drugs and branded drugs with therapeutic substitutes, enrollment increases lead to larger declines in prices and profits - a 7 to 9 percent drop in profits, for example, as compared to a 5 percent average drop in profits from all drugs. However, the effect on non-competitively supplied branded drugs, which account for roughly half of all drug expenditures in the U.S., is close to zero. These results suggest that insurers use their increased bargaining power to extract rents from pharmacies for drugs that are supplied competitively, but have little ability to leverage additional enrollments into lower prices for drugs with few substitutes.

When an insurer negotiates lower prices with a pharmacy as a result of additional Part D enrollees, the lower prices may spillover to all of their members, including the population of non-Part D commercial enrollees. Indeed, the authors find that a Part D enrollment increase of 100,000 reduces prices by 2 percent for non Part D members. The authors suggest that these external effects have the potential to affect many outside the Part D program. Given the enormous size of the commercial enrollment population, the external effects may rival the direct benefits of Part D to the population of now-insured seniors who lacked drug coverage prior to Part D. They note that if the government were to negotiate drug prices for all Medicare beneficiaries, there would be no such external benefit to younger consumers or to seniors enrolled in commercial plans.

The authors acknowledge financial support from the National Institute on Aging and the Robert Wood Johnson Foundation.