December 2003
The Price of Pharmaceuticals: International Comparisons and the Effects of Controls
Are drug prices in the United States higher than in other countries? The answer to this question is not as simple as it may seem. Patricia Danzon and her colleague Michael Furukawa at the University of Pennsylvania's Wharton School have recently published three articles that explore the complex issues involved in making international price comparisons and measuring the effects of price control schemes used in several foreign countries. At an AEI conference on December 12, Danzon explained how price comparisons are affected by foreign exchange rate fluctuations and the selection of products that are used in the comparisons. She presented her own estimates of how prices of pharmaceuticals in the United States compare to Canada and other countries and how the use of price controls is affecting the availability of new products.
Patricia Danzon
University of Pennsylvania
Most of the clamor that pharmaceutical prices are unfairly low in foreign countries is based solely on anecdotal evidence. In contrast, this study makes international price comparisons using a broad market basket of pharmaceuticals characteristic of the United States and estimating its retail value in the largest European markets, Mexico, and Chile. We analyzed several factors contributing to overall price differences, such as varying utilization patterns.
Defining an appropriate basket was complicated by the availability of multiple presentations of a drug. One molecule is generally prescribed in different forms and dosages, and is often manufactured by more than one company. Licensing out patents, parallel importation, and generic availability further obscure the identification of a particular compound. Because pharmaceutical markets are structured so differently around the world, no sample can be fully representative. Still, most other international price comparisons tend to include too few products, use only one presentation of a drug, and even exclude generic drugs entirely. Previous comparisons have generally ignored differences in utilization and taken unweighted averages of drug prices. Finally, retail pharmaceutical prices overstate the prices actually received by manufacturers in the United States since they include wholesaler and pharmacist markups and neglect discounts given by manufacturers to managed care and Medicaid.
The original data included 350 leading molecules by volume in the United States in 1999. We then eliminated any compounds available in fewer than four countries and added all molecules approved by the Food and Drug Administration after 1992, so our data is slightly biased toward new products. Our final sample contained 249 molecules sold in the United States. We used IMS data on national sales in 1999 of all presentations of each molecule at manufacturer-level prices. By tracking the active ingredient of a drug and not a particular presentation of it, we achieved more matching across countries. Had the molecules instead been matched on presentation, the sample would have been only half as large. The U.S. sales data were adjusted for off-invoice discounts given to pharmacy benefit managers (PBMs) and Medicaid rebates, reducing U.S. prices by an average of 8 to 10 percent. The calculated price indices compared the cost of the U.S. market basket of drugs at foreign prices relative to U.S. prices. The price ratio was then weighted by the molecule's share of U.S. sales, making the comparison most relevant to the United States and its policies.
The sample used represents 61 percent of pharmaceutical sales in the United States, the United Kingdom, and Canada, and between 30 and 40 percent in Chile, France, Germany, Italy, Japan, and Mexico. Again, if the study identified drugs by their presentation instead of their active ingredient, matching across countries would have been substantially less. Another interesting variant among the markets analyzed is the much larger unbranded generic share by volume in the United States as compared to other countries. Branded products are able to retain their market share abroad even after their patent expires.
Converting currencies with exchange rates, we find the highest price index in Japan, where the U.S. market basket sells for 127 percent of its cost in the United States. However, this finding is largely a function of the exchange rate. The basket sells more cheaply in all other countries investigated, ranging from 33 percent less in Canada to 6 percent less in the United Kingdom. When the data is broken down between originator single-source drugs and generics, we find that generics are uniformly cheaper in the United States because of a more competitive retail sector. Originator drugs, on the other hand, are generally less expensive abroad than in the United States.
Because exchange rates reflect financial flows and the cost of living, we repeated the analysis using purchasing power parities (PPPs), specifically economy-wide GDP PPPs, and medical PPPs. Medical PPPs allow us to look at the cost of pharmaceuticals relative to other medical services in a country. Pharmaceutical price indices mostly climbed above 100 percent, suggesting that drugs are in fact cheaper abroad when compared to the cost of other medical services. It is important to note that these data are not adjusted for quality, so they match one hospital day in the United States to one hospital day in France, though the care delivered can be quite different. Using PPPs is also advantageous because manufacturers set their prices at the launch date, after which exchange rate fluctuations put price changes largely out of their control. Using launch date exchange rates, Canada's pharmaceutical prices become only 14 percent lower than those in the United States.
A molecule's age on the market has significant impacts on utilization. While per capita total pharmaceutical consumption abroad is generally comparable to the United States, new products are used much less. During the first two years after a therapy is globally launched, individual consumption in Chile is as low as 2 percent of U.S. consumption.
Pharmaceutical price differentials across countries roughly reflect differences in income. Adjusting the price indices by per capita income suggests that drug price levels are actually slightly higher in other major markets. For the basket used in this study, drug prices in Canada are 4 percent higher than in the United States and are 25 percent higher in the United Kingdom. In Chile and Mexico, prices are nearly 430 percent higher than found in the U.S. when normalized for income. While the American public protests that drug prices abroad are too low, the data suggest that we in fact are asking low-income countries to contribute more for pharmaceuticals than they can reasonably afford.
Price differentials in the pharmaceutical market are necessary to sustain the research and development (R&D) of new therapies. Pharmaceutical R&D is a large, globally-joint fixed cost, where optimal pricing is a function of demand elasticities. Using income as a proxy for elasticity, we conclude that higher drug prices for high-income countries are both efficient and equitable. For the United States to artificially dampen prices through reimportation or price controls would create inappropriate price uniformity and lower social welfare.
Another study I conducted with colleagues at the University of Pennsylvania investigates the impact of price regulation on the launch of new drugs. We analyzed stalled and aborted drug launches for eighty-five new drugs in twenty-five countries between 1994 and 1999. In theory, manufacturers may rationally delay or cancel a drug's arrival to market if the low price threatens to undermine the price of the drug in other markets. This spillover effect can occur through parallel trade or price regulation based on foreign prices. In larger markets, the opportunity cost of delaying a drug launch is higher.
Only 55 percent of all potential launches actually occurred, most of these in the United States, Germany, and the United Kingdom. Japan, Portugal, and New Zealand not only had the fewest drug launches but also experienced the longest delays in bringing new therapies to market. Within each country, the number of new chemical entities launched was generally a function of the expected price, and to a lesser degree of the expected volume. Price controls stifle launch probability because of the potential for parallel exportation. Originator firm characteristics can also positively influence the introduction of new drugs.
John Calfee
AEI
Patricia Danzon has now analyzed international pharmaceutical price differentials twice, using data from 1992 and 1999. The news media demonstrate a studied ignorance of her findings, but these results should be central to any discussion of drug pricing.
Danzon's finding that newer drugs experience slower uptake in foreign markets has obvious policy implications. Among drugs available for two years or less, Canadians consume one-fourth of what Americans do per capita. Canada's price controls have lowered not only drug prices but also access to medicines. Looking beyond Canada, we can see that price controls create drug price differentials that bear a crude relationship with income disparities, but Danzon argues persuasively that free markets would, at least in theory, produce this outcome as well and not reduce access to drugs or endanger a nation's health. Moreover, highly regulated retail sectors are extremely inefficient, which raises drug prices. While the United States has cut drug costs through its booming generic industry, foreign markets do not yet capture substantial savings from generic pharmaceuticals.
Judith Wagner
Institute of Medicine
The finding that international drug price differentials are smaller than the media portray is critically important. While the United States has significantly reduced its drug costs by using generics, branded drugs remain more expensive here than abroad. While Ramsey optimal pricing is based on demand elasticities or incomes, the current price differentials did not develop because the pharmaceutical industry deliberated on how much countries could afford to pay. Rather, they are rigorously imposed by foreign governments through price controls and aggressive pushback systems. These mechanisms exist in the United States as well-tiered copayment structures and formularies, for instance-and are driven largely by employers worried about the expense of retiree drug coverage. Still, U.S. drug purchasers are much less successful in negotiating drug prices than foreign governments have been. First, the Medicaid best price law impedes payers from negotiating effectively with manufacturers, who would have to extend the rebate to Medicaid, which buys 15 percent of outpatient prescription drugs. Eliminating the Medicaid best price rebate law would increase price disparities initially, but they would fall over time, as private payers became more aggressive negotiators. Instead of worrying about drug prices in France or Germany, we should focus on untying the hands of private plans and enabling them to become prudent purchasers.
AEI research assistant Ximena Pinell prepared this summary.
