Do pharmaceutical companies discriminate against elderly U.S. consumers, charging them higher prices while offering heavy discounts to preferred consumers such as bulk American buyers as well as to retail consumers in countries such as Canada and Mexico? Two recent studies comparing pharmaceutical prices across different market segments have come to precisely this disturbing conclusion. That view, however, is unwarranted, argues Wharton’s Patricia M. Danzon, in a new review of research on pharmaceutical price differentials.

One study, known as the Minority Staff Domestic Report, has compared the retail prices of 10 branded prescription drugs with Federal Supply Schedule (FSS) prices. The FSS is a price catalog for purchases by federal agencies, and is an indication of the prices charged to most favored customers such as large insurance companies and HMOs. The average difference between the retail price and the FSS price was 106%. On the other hand, the average difference for non-pharmaceutical consumer goods was only 22%.

Another study, the Minority Staff International Report, compared the international data. Here it was found that U.S. retail consumers pay 72% more for their pharmaceutical drugs than Canadian consumers, and 102% more than Mexican consumers. The conclusion: the discounts to preferred customers in the U.S. and abroad resulted in cost-shifting to U.S. retail consumers, including the elderly.

Danzon faults these studies on several counts. First, the sample is small and unrepresentative. It considers only branded prescription drugs and excludes generic products. By ignoring the latter, which have large sales and low prices, the average pharmaceutical prices in the U.S. get overstated. Second, the price indices have been poorly constructed: appropriate weights have not been assigned to the 10 drugs considered.

Third, apples have been compared with oranges: wholesale FSS prices with retail consumer prices. Fourth, the statutory discounts which pharmaceutical companies have to offer certain bulk purchasers have not been considered. Two factors, distribution mark-ups and statutory discounts, alone can explain three-fourths of the median retail-to-FSS price differentials of 86.5%, says Danzon. Also, the entire differential can be accounted for if an average 10% to 15% discount to private consumers is considered.

The estimated difference between pharmaceutical prices in the U.S. and Canada also suffers from the same drawbacks: a biased sample and inappropriate methods. Earlier studies conducted by Danzon have concluded that Canadian prices are between 13% lower and 3% higher than the U.S., depending on the price index used. As far as Mexico is concerned, evidence suggests that prices are significantly lower than most European countries. This is not surprising, given Mexico’s low per capita income, weak patent protection before 1992 and more price-sensitive consumers.

Other studies based on much larger samples-conducted by the GAO in 1994 and the CBO in 1996-have also shown median best price discounts of 14% to 15% and weighted average best price discounts of roughly 19%, far smaller than is commonly believed. Giving discounts to price-sensitive consumers is common to many industries with significant joint fixed costs and low marginal costs: pharmaceuticals, airlines, restaurants and movie theatres, for example. Such discounting can increase overall social welfare.

Charging different prices to different markets does not imply cost shifting, says Danzon. A manufacturer who serves two different markets rationally determines the price to charge in each of the two markets independently. Giving a discount in one market does not affect the prices charged in the other market.

Yet the recent proposal in HR 664 to force pharmaceutical companies to sell their products to retail pharmacies at the lowest price available to federal government purchasers, is likely to result in higher prices to managed care and other nonfederal buyers. Recent experience points in that direction. Following the Omnibus Budget Reconciliation Act of 1990, companies were asked to give best price discounts to Medicaid. As a result, the median best price discounts given to HMOs and GPOs declined substantially.

Isn’t such a conclusion at odds with the thesis that by offering discounts to preferred customers, pharmaceutical companies shift costs to retail consumers? No, it isn’t. There it was argued, based on economic theory, that as long as the two market segments – the relatively price-inelastic retail sector and more price-elastic managed care sector – are separate, the prices in the two markets are set separately. So giving a discount to the managed care sector does not result in a price increase in the retail sector.

By contrast, once the same price is charged to both retail and bulk consumers, the two markets are no longer separate. So a price discount given to the managed care sector must be matched by a similar discount in the retail sector. A manufacturer would rationally charge a (common) price based on the average price elasticity in the two market segments. As a result, the prices to bulk buyers will increase.

Moreover, the price-sensitivity of the retail customer will not change. And retail pharmacies are not required to lower their prices. So the benefits would accrue to pharmacies in the form of higher retail margins rather than to consumers in the form of lower retail prices: to the pharmacist rather than to grandpa.