Integration, Cost, and Efficiency in Pharmaceutical Distribution

Considerable controversy exists over the role of vertical integration in pharmaceutical distribution. This research establishes that an insurance plan pays less for brand-name drugs dispensed at PBM-captive mail-order pharmacies compared to the same drugs dispensed at retail pharmacies. For generic drugs, the relative costs are reversed. That is, generics cost more at mail-order than retail pharmacies. Integration may affect the price that pharmacy benefit managers (PBMs) can negotiate with drug makers. The preferred status set by PBM can provide significant demand enhancement to a brand-name drug, so the brand's producer is willing to offer discounts to PBMs and their captive mail-order pharmacies. The lower insurance plan payments of mail-order distribution could be the results of either cost advantages arising from negotiation or from saving generated by efficient distribution. When drug patents expire, the competitive landscape changes. No generic maker will be willing to pay for preferred status, since other generics could free ride on that status. Empirical results show that insurance plan payments for generics dispensed at captive mail-order pharmacies are more than that they pay to competitive retail pharmacies. Since the generic drug price comparisons should be free of negotiated cost advantages, this comparison suggests that retail pharmacies are at least as efficient compared to mail-order. This means that the price difference between mail and retail for brand-name drugs understates the true effects of PBM negotiations. Legislation introduces an odd intervening case in the form of the Hatch-Waxman Act, which gives the first generic maker to file an Abbreviated New Drug Application (ANDA) a 180-day window following patent expiration in which other generic entry is prohibited. The producer of the brand-name is permitted to authorize a second generic producer, and hence three players (brand maker, authorized generic, and first ANDA generic producer) compete during the 180-days. But these three are not sufficient to induce Bertrand competition. Instead, they respond to strong incentives to reach agreement to share the market. As a result, while consumers are charged low generic copayments during the 180-day window, the price of the generic falls by less than the decline in the copayment. Hence the insurance plan actually pays more for the generic than it had been charged previously for the branded drug. While the lower copayment ensures that patients clamor for the generic drug, the lack of competition in the window means that the plan is worse off. The 180-day window is thus very costly for the insurance plan. The three players gain substantial rents during the window, though of course smaller than the rents to patent protection. The ANDA first filer generic obtains a portion of these rents which is, for the drugs in our sample, sufficient to induce it to enter and incur the litigation costs of challenging the patent holder (the motivation for the 180 day protection). But in addition, the brand's owner receives additional rents subsequent to its patent expiration both from any brand name sales it makes and also from it agreement to authorize a generic of its own.