Accounting For R&D

It is a mantra among industry advocates that development of a novel drug—a new molecular entity—costs about $800 million. That figure, calculated in 2001 by the Tufts Center for the Study of Drug Development (CSDD), has now been adjusted for inflation and revised to $1.2 billion to $1.3 billion.

Pharmaceutical industry spokesmen use the figure to justify high drug prices. Even Food & Drug Administration officials repeat the number uncritically. Janet Woodcock, FDA's acting deputy commissioner for operations, says: "The cost to produce a successful new molecule has gone up very sharply and is now about $800 million."

What is the origin of this figure?

Joseph A. DiMasi, director of economic analysis at CSDD, performed the original calculation. He took 68 new molecular entities that industry had developed from scratch without any prior research by government or nonprofit organizations. After adding up the development costs for the 68 drugs, DiMasi came up with an average R&D cost of about $400 million per drug. He then asked: What could this $400 million have earned if it had been invested in the stock market for the eight to 10 years it takes to develop a typical medicine? This is known as the opportunity cost of capital. He assumed that the money would grow at 11% per year and calculated that $400 million actually cost industry about $800 million. In 2005, DiMasi took the $800 million figure and asked: "What is it worth today?" That is how he came up with an adjusted total of $1.2 billion to $1.3 billion.

Several aspects of this methodology are controversial. First, there is a dispute over whether most new molecular entities are developed without any research in academic or government labs, such as the National Institutes of Health. CSDD claims that the majority of drugs are developed entirely by pharmaceutical firms. It is only cancer and AIDS drugs that usually receive significant government-funded R&D before they are licensed to pharmaceutical companies, says Christopher-Paul Milne, assistant director of CSDD.

Many who study the pharmaceutical industry, however, say the majority of drugs are partially developed by NIH or by researchers at universities and other nonprofit institutions and then licensed to pharmaceutical firms. Some experts claim it is only the exceptional drug that is developed wholly by a drug company. In typical cases, they say, drugs are worked on for years by government or academic researchers but have not gone through clinical testing before being licensed to a large drugmaker. If significant development occurs before a pharmaceutical is licensed to a drug company, the total cost of R&D, including the license, is lower for the commercial firm because taxpayers' money has paid for a big portion of the R&D. Unfortunately, no researcher seems to have come up with precise figures on how much outside R&D goes into the average drug, says Daniel Carpenter, professor of government at Harvard University.

Another criticism of the CSDD figure is that DiMasi should have adjusted the $400 million figure downward to account for the R&D tax credit. Industry takes a federal tax credit of 20% on all R&D expenses above a certain base amount, and some states give additional R&D tax credits. Critics also say that an 11% annual growth rate for opportunity costs is unrealistically high.

Peter Lurie, deputy director of Public Citizen's Health Research Advisory Group, says the average cost of developing a drug is about $240 million. In calculating this total, he estimates that half the R&D cost of the average new molecular entity was paid for with taxpayer money. He does not compute the opportunity costs of the money that goes into drug development, because on the federal tax forms, R&D outlays are considered expenses that are paid for with that year's sales. They are not capitalized. And he accounts for the savings from the investment tax credit received as a rebate on R&D expenses.

Milne, however, defends the $1.2 billion estimate. The Federal Trade Commission critiqued CSDD's economic study and came up with an even higher figure, he says. It is standard practice to include opportunity costs in such an economic analysis, he explains.

Merrill Goozner, director of the Integrity in Science Project at the Center for Science in the Public Interest, has a different take on the issue: "Do you see any other industry that makes the argument that they are not going to invest in R&D if they don't get high prices? Every other industry says, 'We have to invest in R&D, or we don't come up with new products and processes.'"

NEXT STORY: Science - Improving Efficiency