Explaining Research – How much do drugs really cost to develop?

Buckle in, because this will be a long one…

Although this does not seem like a fact that most people would commit to memory, somehow the average American has come to know that, not only is it insanely expensive to create a new drug and bring it to market, but it is expensive to the tune of $800 million dollars. Why is this important?  It is the reason most often cited for why medications are so expensive in the United States.  And no one can doubt that they are expensive; we spent over $950 per person on drugs in 2007, and that number is going up every year. Many drug company executives claim that if we didn’t pay so much, then the $800 million dollars needed to create each new drug would prevent us getting any new scientific advances.  Even if you didn’t remember this number, it is reinforced repeatedly in advertisements, news reports, and campaign speeches.  It is so widely accepted that few people challenge the figure, at least not in public.  The truth, however, is a different story.

In order to understand why this is a myth, we must first understand how this number came to be.  In 2001, a group known as the Tufts Center for the Study of Drug Development released the results of a landmark study performed by Joseph DiMasi, MD and colleagues.  They claimed that, after exhaustive work and unprecedented access to company data, the average cost of developing a new drug was $802 million. Unfortunately, the methods of that study were not released until sometime after that, so no one could analyze their figures for over a year. When the paper describing the work was finally released, the $800 million number had already become branded into the American psyche.  So much so that a number of subsequent books and studies attacking the findings not yet been able to change the public consciousness.

So what is wrong with the figure?  A number of things.  First, let’s just approach the study from an economic standpoint.  The actual out of pocket costs for developing a drug in the Tufts study were about $400 million, half of the final number.  How did it get to $800 million then? Through a fancy piece of economic sleight of hand known as “opportunity costs of capital.” You see, when a business chooses to invest money in one area instead of another, it takes a gamble.  The drug industry could spend $400 million on research or, for instance, put it in the bank.  If they put money in the bank instead, they would have earned interest on it.  That interest income was sort of “lost” since they didn’t get to put it in the bank.  Perhaps they could have made even more money if they had invested that $400 million in the stock market.  Or in baseball cards.  The Tufts authors calculated that the drug industry “lost” another $400 million in potential money that they could have earned if they had done something different with the money.  So, they added that up and arrived at $800 million.

The problem with this line of thinking is that research and development is not an investment option for the pharmaceutical industry.  It is a necessary expense — at least it is if they want to stay profitable.  If they want to take their money and invest it in cars that’s fine, but then they are in the auto industry, not the drug industry.  Ironically, when they do their taxes, research is categorized as an expense, not an investment, so any argument about capitalizing costs is rendered moot by the actions of the drug companies themselves.  Research is not a cost to be capitalized.

Which brings us to the other economic trick.  Research is an expense; it is listed as such on tax forms and is, therefore, tax deductible.  This means that the $400 million bill found in the Tufts study was in pre-tax dollars.  Since it is tax-deductible, and the corporate tax rate was about 35%, the actual after-tax cost would be less than $270 million.

While that’s still a lot of money, it is far less than $800 million.  But there are even more problems with the Tufts study.  When drugs go in for approval from the U.S. Food and Drug Administration (FDA), they are classified according to their perceived importance and novelty.  A small percentage of the drugs seeking approval are what are known as “New Molecular Entities” (NMEs.)  Most of the drugs the FDA reviews are just slightly different versions of drugs that already exist on the market.  It makes intuitive sense that it would cost more to make an NME than to merely change an existing drug into a slightly different molecule.  Not only did the Tufts study only look at NMEs, it only included NMEs where the entire cost of development was born by the drug company itself.

As I discussed yesterday, the government, not the drug companies, supports many of the costs of developing drugs. A very, very small number of drugs are developed entirely by industry.  So by taking only NMEs and taking only NMEs that were entirely and solely paid for by companies, the study selected the most expensive drugs for analysis.  This does not yield an “average” cost, no matter what the New York Times says. It provides a cost for the most expensive and rarest types of drugs around — NMEs developed entirely in house.

Even that result is suspect, however.  To this day, no one outside of the study has been permitted to examine the data that the drug companies provided secretly to the Tufts group.  And while we would like to believe the best about everyone, we would be naive to ignore the desire of the pharmaceutical industry to make the costs of research and development seem as high as possible.

Are there any other studies out there to help us arrive at the true cost?  Of course there are.  In 2001, Public Citizen, a consumer safety group known best for its founder Ralph Nader, published a detailed analysis attacking the findings of the Tufts group.  They made a compellingly simple argument: to calculate the average research and development costs to bring a drug to market, you only need to know the total amount spent on research and development and the total number of drugs that obtained FDA approval. Dividing the costs by the number of drugs yields the average cost per drug.

Since total costs are available through publicly available data, and drug approvals are available through the FDA, the group was able to analyze data from 1994 to 2000. They found that the average cost to develop a drug all the way through FDA approval was $161 million pre-tax, and $110 million in after tax dollars. Again, not chump change, but not $800 million.

Even if you think Ralph Nader should not be trusted any more than the drug companies, anyone can reproduce this experiment using more recent data.  According to the 2005 Pharmaceutical Research and Manufacturers in America (PHRMA) Industry Profile, drug companies spent about $121 billion on research and development in 2000-2003. In that same time period, 314 drugs obtained FDA approval. This comes to about $385 million per drug in pre-tax dollars, and assuming a corporate tax rate of 35%, a total of $250 million per drug.

All of this information may leave you with a spinning head. After all, we want the drug companies to spend more time developing NMEs because odds are that most breakthroughs in healthcare will come from these types of drugs.  Far too many new drugs are just copies of drugs that already exist and are really no better than the originals.  It is also likely that NMEs cost more to create than copycat drugs.  However, since NMEs represent such a small percentage of new drugs, it is unfair and dishonest to represent the costs of these drugs as the average costs of new drug development.

Is $250 million a more accurate, and up-to-date, average cost for bringing a drug all the way to market?  Likely, yes.  Some will argue with this.  In fact, Bain & Company, a consulting firm, made headlines by claiming that the cost per drug is really $1.7 billion. However, they argue that the costs of marketing and advertising, which are 2 to 3 times higher than research and development costs, should be included in the cost of creating a new drug.  That’s ridiculous; truly amazing new drugs need no marketing or advertising, only copycat drugs do.

Ultimately, the truth will never be known until the drug companies open up their books and allow us to examine their data ourselves.  That’s not likely to occur soon.

I learned so much of this from Marcia Angell’s The Truth About the Drug Companies: How They Deceive Us and What to Do About It that I’m going to plug it again.  Go buy it.

Hidden information below


Email Address*