Epstein, the University of Chicago law professor, is a delightfully unusual scholar in that he not only covers the legal issues thoroughly, but he also is an exceptional microeconomic thinker. He has that rare ability to take an issue, apply some basic microeconomics combined with some careful assumptions about various magnitudes and, voilà, reach a conclusion that may surprise and will certainly illuminate. Indeed, his virtuosity at this is reminiscent of his retired University of Chicago colleague, economist Sam Peltzman.

Epstein displays all of these skills in Overdose. In it, he defends patent law in the drug industry while criticizing various proposals to change the law, and also argues against the extreme regulation of drugs that the FDA has engaged in for decades. I found his case for strong patent law in the pharmaceutical industry persuasive, despite my libertarian view that patent law is essentially an arbitrary assignment of monopoly power by government. If there is ever a case for patents, the drug industry is the one for which that case is strongest. Possibly more important, he makes a devastating case that the FDA should not have the power to prevent new drugs from being brought to market. That is not all; Epstein also analyzes the problems with our current tort law system for drug liability and suggests some ways out.

Drug Prices Take the old bromide about how advertising and marketing costs of drugs increase the price that drug companies must charge. Seems obvious, right? It is also incorrect, and Epstein effectively explains why. He points out that the cost structure in the pharmaceutical industry is extreme: high fixed costs — potentially in the hundreds of millions of dollars — to produce the first pill that can be legally marketed, but then low marginal costs for subsequent pills. Without large expenditures on marketing, the market will be too small to generate enough revenues in excess of variable costs to cover the fixed costs. Drug company executives, looking forward and seeing this fact, will not develop new drugs. And a nonexistent drug has, in effect, an infinite price. But if spending this money on marketing can multiply the size of the market, then the expected revenues will be enough to make developing the drug worthwhile.

Epstein uses a numerical example to clinch this case. Assume, he says, that a new product costs $1 million to produce; thus, without marketing, it must sell at least 10,000 units at $100 per unit. But assume further that at that high $100 price, only 8,000 people want to buy it. Revenue to the firm: $800,000. Result: the drug does not get produced. But now assume that for an additional $200,000 of truthful and accurate advertising (in a separate chapter, he handles ads that are not in this category), the firm can sell 100,000 units at $20 a pop. The result: the firm makes revenues of $2 million. Assuming that its variable costs are $500,000, the firm will net $300,000. The drug has been produced, and marketing has made its price fall from infinity to $20. Scale up these numbers by a factor of 100 (except for the price) and you have a fairly plausible, possibly even common, scenario in the drug industry.

Epstein's analysis of the law and the economics of the drug industry is outstanding on many levels. To give them all their due would require a small monograph. So I will focus on four: the nature of competition between "me-too" drugs, extreme FDA regulation that prevents or slows the introduction of new drugs, the issue of "reimportation," and the liability system for drugs.

One of the charges leveled by critics of the pharmaceutical industry is that it invests too much in so-called "me-too" drugs. Some have even gone so far as to say that a new drug that competes with an old one should not be allowed unless it is superior to the old one. In various speeches, I have criticized this view by pointing to an industry in which virtually everyone recognizes the value of a "me-too" good: the auto industry. You could call a Toyota Camry a me-too Honda Accord. But does anyone doubt that consumers get a better deal on Accords because Honda must compete with Toyota's Camry?

Epstein addresses the argument that drugs are too differentiated from each other to give consumers the benefit of competition. He responds: "[T]his definition of competition is too narrow for its own good: it excludes cases in which substantial but imperfect competition is present across broad product classes." Epstein points to the wine business, which is intensely competitive, but nevertheless there are distinct differences between various brands and varietals.

In the chapter titled "FDA Versus the Individual," Epstein points out something obvious but profound:

If the FDA decides to let a drug onto the market, no one is obliged to use it. Any mistake to permit the sale of a drug is therefore subject to downstream correction by individual users. But a decision to keep the drug off the market is impervious to downstream correction by individual users.

He gives an example of someone who might have a 91-percent chance of dying without taking an action, but a 90-percent chance of dying by taking the action. If the action is surgery, notes Epstein, then the FDA has no say. But if the action is to take a drug, the FDA is involved and might forbid the one-percent solution.

Epstein also notes that by placing its emphasis on the average effects of drug use in evaluating drugs for approval, the FDA "ignores the variation in individual responses." The fact that a drug does not perform as well as a placebo, the kiss of death for approval, "shows that most people should not take the drug, not that it should be banned from the market."

What is the appropriate solution for the problem of price controls that other countries' governments impose on drugs? Epstein argues that the ideal private solution is for a drug company "to attach conditions of sale forbidding any drug sold overseas to be resold for consumption in the United States." That way, arbitrage is limited and a drug company that sells in the lucrative U.S. market can still earn monopoly profits on its patent — profits that encourage it to innovate. A seller that breaches its contractual obligation would then be liable for damages. But Epstein notes that contract remedies "are difficult to come by in this international setting." What to do? The next-best strategy for drug companies is to "restrict the supplies sold in foreign markets to the estimated level of their domestic consumption to minimize the surplus available for reimportation by foreigners." Charley Hooper and I noted this in a February 24, 2004 Washington Times op-ed entitled "Hidden Drug Reimport Potential":

If many people in the United States are allowed to buy from Canada, drug companies will certainly notice. They don't want their U.S. pricing policies undercut because the U.S. market, relatively free of price controls, is the most lucrative drug market in the world. There is only one way not to have their prices seriously undercut: They will choose to limit supplies to Canada.

Unfortunately, there has been pressure from the U.S. Congress to require American firms to sell to foreign distributors in select countries whatever quantities of drugs they want to buy at a price equal to the price charged — often by government order — in the local market. Epstein points out the problem. To require these firms to sell unlimited quantities at an often-government-controlled price, quantities that could then be arbitraged to the U.S. market, would allow "foreign nations to set the prices for domestic (and foreign) sales of American drugs without any attention to whether drug companies receive a permissible rate of return once those restrictions are put in place." Epstein, always on the lookout for a government "taking" without just compensation, spots one here.

Drug R&D My fourth and final highlight from Epstein is his devastating critique of the current liability system for drugs — both in the abstract and by reference to actual legal cases, especially Vioxx. (The Vioxx discussion alone is worth the price of the book. Indeed, had I written the book, I would have led with this discussion because it shows just how dysfunctional the liability system is and how the New England Journal of Medicine, in particular, helped set up Vioxx's producer, Merck.) Instead of summarizing his complex views on the issue, I will hit the high points.

Epstein points out that the NEJM helped create suspicion of Vioxx by publishing an "expression of concern" about an earlier study of Vioxx that the journal had run. He explains, "'[E]xpression of concern' is a technical term that is tantamount to a charge of scientific fraud." He points out that the NEJM had published only three expressions of concern it its history and that, in this case, it did not take the more-normal route of referring the matter to the home institutions of the article's authors. He goes into great detail, concluding that, at most, there was evidence of procedural error, not scientific fraud.

Moreover, consider the civil case against Vioxx, in which Robert Ernst's widow was awarded $253 million. Ernst had all the indicators of someone with a high probability of dying of a heart attack whether or not he took Vioxx. Epstein writes, "Just this logic should have propelled Merck to an easy win." Moreover, because little in the current legal system "allows for the coordination of awards across different cases," even if Merck did over-promote its product, we would have been closer to the mark had the case never been brought.

Epstein cites and defends the estimate of Joseph DiMasi of Tufts University and his colleagues that the cost of bringing a new drug to market is $802 million or more. The Naderite organization, Public Citizen, criticized this study. Epstein points out that one of their main criticisms is that the estimate mistakenly included the cost of capital. Epstein argues, as would virtually any economist, that the cost of capital is a relevant cost.

Two of the drug industry's chief critics, critics who want more, not less, pharmaceutical regulation, are Arnold S. Relman and Marcia Angell, both former editors of the NEJM. Epstein criticizes other thoughts they have, but Angell and Relman's touting of Pubic Citizen's criticism is worth quoting if only because these two critics are well respected, at least by some. In "America's Other Drug Problem" (New Republic, December 16, 2002), they write:

The authors [DiMasi et al.] seem to justify this interesting account maneuver on the grounds that from the perspective of investors, a pharmaceutical company is really just one kind of investment, which they chose among other possible investment options. But while this may be true for investors, surely it is not true for the pharmaceutical companies themselves. The latter have no choice but to spend money on R&D if they wish to be in the pharmaceutical business, so they have no "opportunity costs." To add the investors' opportunity costs to the company's out-of-pocket cost of developing a drug seems rather odd. DiMasi assures us that this calculation conforms with standard economic thought and accounting practice, but recent events on Wall Street make such reassurance less comforting than it might once have been.

Wow! I mean, wow! Don't Relman and Angell understand that drug companies, to raise capital to be in the business, must give investors a return at least as high as they can earn elsewhere? And, as for "recent events on Wall Street," by which they were presumably referring to Enron, did anything that happened suggest to careful observers that companies in any business have a zero opportunity cost of capital? Relman and Angell are flailing at best.

Drugs and the FDA John R. Graham is a health economist with the Pacific Research Institute (not to be confused with John R. Graham, the financial economist at Duke University). In his new monograph Leviathan's Drug Problem, he does an excellent job of analyzing Food and Drug Administration regulation and summarizing the research on the high cost of that regulation.

Graham quotes Richard Miller Jr. and H.E. "Ted" Frech's finding (in their 2004 AEI Press book Health Care Matters) that in 1990, the lifetime cost of pharmaceutical spending to add one year of life expectancy was $15,952 for a 40-year-old woman and $14, 486 for a 60-year-old woman (with comparable numbers for men). This is well below the threshold of cost effectiveness that most health economists use to measure value for money, suggesting that pharmaceuticals offer a very good deal.

Graham also gives evidence that the FDA, by delaying and preventing the availability of drugs, kills far more people than it saves. He points out that the Prescription Drug User Fee Act (PDUFA), by charging drug companies large fees to have their drugs considered by the FDA, has reduced the drug lag. (See "How Have User Fees Affected the FDA?" Spring 2002.) But he also argues that PDUFA is not enough. He quotes the finding of the Abigail Alliance for Better Access to Experiment Drugs — which consists of families who have lost loved ones who were prohibited from taking experimental drugs — that every drug for which it advocated use as an experimental drug was later approved by the FDA.

Graham's solution? Competing approval agencies. For instance, drug certification could be undertaken by a private nonprofit approval agency, just as Underwriters Laboratories now does with thousands of products and the Snell Memorial Foundation does with helmets. But wouldn't a nonprofit be a monopoly just like the FDA? No. That is not how UL, for example, works. Although UL does not need to make a profit, it does need to attract customers and it must compete with other standards organizations. A middle ground, notes Graham, is to allow Americans to use a new drug as soon as a regulator in any developed country grants approval to that drug.