Share |






Richard A. Epstein
August 22, 2007

In a recent issue of the New Republic, Arnold Relman, a former Editor—in—Chief of the New England Journal of Medicine, offered a most unflattering review of my recent book, Overdose: How Excessive Government Regulation Stifles Pharmaceutical Innovation (Yale University Press, 2006). The Editors of the New Republic have decided not to publish any reply from me. I am therefore grateful that the Manhattan Institute has stepped forward to publish my reply, with—I should add—an open invitation to Dr. Relman to respond if he so chooses.

In dealing with a longish review, the usual approach of the aggrieved author is to take on only one or two points of the critic, and let the reader draw whatever inferences he or she chooses about those points that were not addressed. The Relman review, however, makes so many instructive errors that a different course seems preferable. Let me first start with a few general observations and then turn to a point by point rebuttal of his argument.

The most obvious difference between us is that we inhabit two different intellectual universes. He dwells from the more liberal and more complacent medical culture of Cambridge, Massachusetts, and I come from the less fashionable and more driven legal precincts of the University of Chicago. Thus two great divides separate Relman from myself.

Start with the professional training. I am a lawyer by training, with an extensive, if informal, background in economics. I have worked extensively and actively in the health field area on a wide range of topics on health care, ethics, law and medicine for the past 30 years, much of it spent in the company of health care professionals with medical training. Relman is a doctor by training who fancies himself an expert in "social medicine," who has as best I can tell no formal training in any of the collateral disciplines that bear on the comprehensive analysis of health care. In his view, mere lawyers are always out of their depth in dealing with the kinds of issues raised by the provision of health care, including of course the development, testing, and marketing of pharmaceutical products. My view is that the technical matters of medical treatment can easily be bracketed in dealing with the larger institutional issues. I do not wish to decide which drugs should be tested in what clinical trials, but, even in the absence of that information, can comment intelligently on the use of these clinical trials in dealing with FDA approvals for new drugs or with their role in litigation. Indeed, in my work as a torts scholar, I am constantly forced to examine detailed medical information that involves such hotly litigated questions as whether Bendectin causes child defects or thimerosal causes autism. In my view, there is nothing whatsoever in Relman's traditional medical background that offers him any comparative advantage in dealing with the full range of property rights, regulatory, marketing, patent, and tort issues that are examined in Overdose, which should be evident to anyone who has worked his or her way through Relman's review.

The second deep cleavage between us has to do with world views. Relman is a generation older than I, being born in 1923, while I was born a generation later in 1943. The difference in age matters, because he takes the New Deal critique of markets as a verity from which all other truths flow. Indeed, for a man so insistent on field expertise, he shows no hesitation to take on Milton Friedman for his heretical views of the role and organization of market institutions. Those sentiments are quite evident in his searing conclusion that insists that markets "do not do a good job of protecting the public interest. If we want a pharmaceutical industry and a health care system that put patients' welfare and society's needs ahead of profits, we will need more regulation, not less."

I of course am very much in the camp of Milton Friedman, which is evidenced by our parallel lives at the University of Chicago and the Hoover Institution. And I was deeply honored to speak at his memorial service held at the Hoover Institution this past January. All that does not mean that I slavishly agree with all that Friedman says-a pose that Friedman himself would have found most distasteful. It does mean, however, that I take strong issue with the simple–minded portrait that Relman paints of my views. The proper question is not whether we put society's interest ahead of profits. That formulation of the question gives no guidance as to what should be done. Clearly if we are to have pharmaceutical companies, then we must allow them to have some profits in order to attract and retain capital. Relman is not subtle enough to distinguish between a risk-adjusted competitive rate or return and monopoly profits, or to recognize that the former advances social welfare and the latter retards it. Indeed, Relman's gripe is not only with Friedman but with every neoclassical economist from Adam Smith to the present

That said, the right question to ask is how we maximize a system of overall social welfare, which cannot be done if the industry is driven out of business. On that structural question I am happy to align myself with both Smith and Friedman on a number of key points: that voluntary transactions are preferable to coerced ones, that subsidies and taxes can easily distort the choices of private actors, that state officials often have private agendas to advance even when in public office, and that the state which concentrates its effort on the control of force and fraud, the provision of public infrastructure and the control of monopoly-which means refusing to create state monopolies-will do better than the big New Deal government that sees in every market a looming imperfection, which only the wise hand of government can correct. The system of government control that Relman proposes is ruinous to the ends of patient welfare and social prosperity that he supports.

His fundamentally unsound world view does create a huge professional and philosophical chasm between us. Worse still, it infects every portion of his New Republic review. Here it is convenient to break down his arguments by into two areas. I will first treat his general critique of the economic issues raised on matters of drug, pricing, patenting and marketing. Thereafter I shall turn to health and safety, with special reference to the FDA, clinical trials and tort liability.


Corporate responsibility. Relman begins his attack by quoting Pfizer CEO Jeffrey Kindler, a lawyer by training no less, who says: "We will transform virtually every aspect of how we do business, focusing on actions that create and sustain value for our shareholders." That statement, made by the CEO of a company whose stock has been under constant pressure should be regarded as good news to shareholders, and to anyone else who is concerned with Pfizer's recent lag in performance. But not to Relman, who sees in this benign pronouncement the spread of Friedman's dreaded influence into the sacred precincts of the pharmaceutical industry. Indeed Friedman said, famously, in words that wave a red flag before Relman's eyes: "Few trends could so thoroughly undermine the very foundations of our free society as the acceptance by corporate officials of a social responsibility other than to make as much money for their stockholders as possible." Relman hears in these words not "we shall do our best by our shareholders." Rather, to him the words say: "the public be damned."

Now it turns out that Friedman (who was not a lawyer) did overclaim in this sentence. The proper analysis says that the corporation is (or is formed by) an agreement among its shareholders, so that any gifts that it chooses to make should be in accordance with the conditions of its charter, which in some cases may allow for such giving. Indeed, I know of no case where a corporate gift has been successfully attacked in court. Relman can breathe easily on this point. But that is a mere detail. Friedman's central point is that shareholders can agree on any program that maximizes share value. Once that is done then the shareholders as individuals can make whatever charitable gifts they see fit with the dividends and gains that they make from their investment. Friedman's argument, which Relman does not understand, is that private giving at the shareholder level is preferable because avoids the real conflict of interests that arises when corporations make gifts to some charities but not to other. And needless to say, Relman shows no evidence of the major corporate law qualification to this rule, namely, that "gifts" to charitable or other institutions that work to promote the good will of the corporation are consistent with the operation of a profit-making institution, as countless cases have held.

We can ignore the fine points of the analysis, because Relman is after bigger game. His view is that Kindler is engaged in some massive hypocrisy when Pfizer makes all these high-sounding statements that it is "working for a healthier world," or "putting its patients first." Pharmaceutical firms are not unique in making these remarks; Ronald Reagan remarked as the spokesman for General Electric, "Progress is our most important product." But Relman is thoroughly confused about how corporations work. The management owes duties of loyalty and care to the shareholders; after all shareholders have put their money in with the firm, and it would hardly do to allow the management to misappropriate the funds so that the shareholders go home empty-handed. But the customers of the firm are not investors, and they can walk away and buy somewhere else if they do not like what they get. The key challenge to any firm is to figure out how to secure their repeated loyalty and that can only be done by offering them goods and services that they value at more than they pay, which the corporation can produce at a cost lower than the prices they can command. Therein lies the critical concordance of interest between buyer and seller that Kindler grasps. Kindler knows that unless he can preserve the good will of his customer base, Pfizer will have to close up shop. The firm has to walk a tightrope, and in that regard it is no different from any other company in any other industry. Indeed, one major problem for any firm in a large industry is that its reputation really matters, such that a single infraction can produce losses in future sales that far dwarf any direct fine or tort judgment that could be imposed. These are good features, because we now know that large companies bind themselves to good behavior by their brand and reputation. One of the first things that Kindler has to do is to make sure that his firm does not acquire a reputation that matches that of the tobacco industry (for which, years ago, I consulted). It is very hard to do in the face of uninformed critics who see a large target that finds it difficult, even with the money it spends, to mount an effective public relations campaign. Like many corporations, pharmaceutical companies are only omnipotent in the fevered imagination of their most severe critics. In reality, a single bad headline or court ruling can erase billions in shareholder value and leave CEOs like Kindler trembling.

After his initial foray, Relman then turns to my view that regulation has hurt the industry in ways that do not help the public. Relying on the judgment of unnamed experts on the drug industry, he concludes that the industry's greatest enemy is itself, not regulation. But, put the word "greatest" to one side, and there is no either/or there. I made it quite clear in Overdose that the industry blockbuster model has not worked well, and that some of the fault lies in the management of the major companies-a point that Pfizer's Kindler obviously accepts. But one cause does not operate to the exclusion of all others, and Relman must be blind to think that the endless efforts to undermine patent protection, to increase the length of clinical trials, to impose price controls, to allow for multi-billion dollar consumer refunds and tort actions has no impact on the day-to-day operations of any pharmaceutical firm, and its long-term prospects. Any change in costs can produce negative results, and it is just unsupported allegation to say that the drug companies have a deep obsession with "financial ambitions and marketing considerations," so-his italics-"that is the root of their current problems." His argument might have some tenuous credibility if deemed to explain why some firms succeed and others do not. But if there is an industry-wide risk of stagnation, it is just wrong to assume that the management of every firm falls prey to these biases. The more plausible diagnosis for a systemwide failure is the systemwide effects of expanded regulation and liability.

Competition and Monopoly. Relman will have none of this equivocation. So that is where I come in, as an industry "apologist" who can put his lawyerly skills to defend the indefensible, which in this case is acting as handmaiden for the industry's monopolistic ambitions while preaching the virtues of free competition. He notes my consulting connections, fully disclosed, to the pharmaceutical industry, and then attacks my general view on health care, as developed in Mortal Peril: Our Inalienable Right to Health Care?, which is deeply opposed to universal health care. Oddly enough, Relman thinks that it is some kind of indictment of my position to say that excessive health expenditures through Medicare and Medicaid "has had negative, although unintended, consequences." But rather than explaining why I am wrong to his committed New Republic audience he just makes the comfortable leap: the man who cannot be trusted on health care is the man who cannot trust on pharmaceutical issues either. The nub of his argument: "prescription drugs are different from most other commercial products-so different that they warrant government regulation of their development, their manufacture, and their sale." At no point does he stop to ask where the libertarian position might call for regulation, which would surely include cases where impure drugs are sold as real, or where drugs are misrepresented as to their uses or side effects. By implication he confuses the libertarian position of small government and free competition with the anarchist position of no government at all.

Nor does he see any reason to do so because he takes the view that I will abandon my principles at the drop of the hat by arguing "for stronger government support of the industry's monopoly rights and for government protection against price competition from abroad. He wants handouts from the public without public oversight." Here, Relman has lost his intellectual compass. On the first point, "industry monopoly rights" refer to patents, which grant exclusive rights to sell particular products to individual inventors. By putting the proposition in this inflammatory way he makes it appear that I wish to block new entry by other firms to compete with firms that have already received patents on drugs already in the marketplace. In fact, in Overdose I am at pains to attack large claims that would give any firm a patent over entire areas of treatment, including my extensive defense of the judicial decision University of Rochester v. G. D. Searle & Co. 358 F.3d 916 (Fed. Cir. 2004) (a case on which I briefly worked for Searle), which refused to allow the University of Rochester to obtain a patent over all drugs that relied on any Cox-II method of inhibition. The patent system, rightly used, gives people exclusive control over their own products, which is a monopoly of sorts. But so long as there are multiple substitutes subject to different patents, then no firm in the industry exercises monopoly power. There can be, and with sound policies, should be effective competition among holders of different patents of drugs that are in the same class.

Relman's statement about "government protection from price competition from abroad" also reflects either economic ignorance or conscious deceit. The last thing that I favor is allowing any firm to induce the government to place tariff or tax barriers to entry on products from overseas. But the programs to which he refers are not efforts by pharmaceutical companies to block competition by other firms hawking their own products. Rather it has to do with the issue of parallel importation, where goods from the United States are sold to, say, Canadian firms at prices dictated by the Canadian government for internal Canadian use, with the knowledge that they will be reimported into the United States for sale at lower prices than those available domestically.

For Relman, who knows nothing about the fine print of this initiative, to call this program "free trade" is a bad joke. The defenders of parallel importation make it clear that American firms should be under a statutory duty to sell as much of any good as any buyer wants to buy at a price that is determined by the state monopolist (more precisely, monopsonist) on pain of losing their privileges to sell in the United States. Then once these goods are sold, the research pharmaceutical firms are barred from asserting their patent rights against a rival firm that wants to sell the products that it has commandeered courtesy of the United States Congress. This looks more like involuntary servitude than free trade. But then again perhaps Relman is in favor of those provisions that hold that doctors, as a condition for being able to work in the voluntary market, have to devote as much of their time to government service for the poor as the state asks, for fees that it determines.

Later on in his essay, Relman again then criticizes the industry for the "extraordinary steps that a Republican-controlled Congress took to ensure that Medicare would not be permitted to influence market prices by directly purchasing prescription drugs for beneficiaries who enroll in Part D under the so-called Medicare Modernization Act of 2003." I did not discuss this issue in the book, but it is worth noting why Relman is in such error on the point, for which the work of the Manhattan Institute's Ben Zycher is so instructive. See his The Human Cost of Federal Price Negotiations: The Medicare Prescription Drug Benefit and Pharmaceutical Innovation, Medical Progress Report, No. 3, November 2006. The first point here is that letting the government into the market on the buy-side does not advance competition, but creates a huge monopsony position capable of forcing down drug prices, which of course the industry does not like. But on this point its position and the public interest correspond. Just as monopolies set prices too high, so monopsonies set them too low. The huge fraction of the market under Medicare will lower rates of return that will reduce the funds available for funding new drugs, which will impair innovation, just as Overdose claims, and Zycher documents. No matter whether one looks at reduction in funds for research ($10 billion per annum), or new medicines delayed (107 to 220, depending on assumptions), the result is the same. In Relman's world regulation is a free good. In the world we live in, it is not.

Nor should we think that the only adverse consequences of government control lie in the future. It is worth noting that even before Relman wrote the Democratic effort to insert the government into the purchasing process started to falter. It is not that the current plan participants were worried all that much about the future. They were worried about the present. The Veteran's Administration is treated as the ideal for Medicare Part D, for its ability to control costs. But how? By limiting choice. Yet many people will prefer to pay more to get more, and the insertion of government into the purchasing process promises to bring in its wake the same restrictions on formulary that are found, say, in England and Japan, which delay the use of new drugs until it is too late for some people. There are, in a word, all sorts of principled reasons to be opposed to any expansion of the government role on the buy-side of the market.

The Cost of New Drugs. Relman then continues his critique of the industry by noting that I am wrong to make too much of the high costs and risky investments needed to bring pharmaceutical products to market. His first point is that the exact cost is a "mystery" because the company data is not public, even though it was given to a team of researchers led by Joseph DiMasi, director of economic analysis at the Tufts Center for the Study of Drug Development, in forming an estimate of the costs of drug development. See Joseph A. DiMasi et al., The Price of Innovation: New Estimates of Drug Development Costs, 22 J. Health Econ. 151-85 (2003) (available here). But this is not an insuperable obstacle. There are certainly crude measures that are publicly available that point to the high, and ever higher, cost of new drug innovation. If research budgets total around $35 billion for the industry, and we produce about that number of new products, then we have something of a rough first approximation of a billion dollars, which can to be corrected to take into account other factors. But don't count on Relman to make the right adjustments. He first notes the standard figure of $800 million, developed by DiMasi's team several years ago doubled today, only to comment: "Yet both these figures include the so-called time-cost of the money spent, a theoretical and much-disputed quantity that contributed about half of the $800 million figure." So now we have it: the "so-called" cost of capital is not a cost of production because, at least to one untutored physician, a dollar spent today is precisely offset by a dollar earned eleven years from now.

There is still the nasty question of how to recoup these costs, which present a difficult problem in any high-fixed, low-marginal cost business. Relman notes that I assert that high prices are needed "to recoup the high cost of their initial R&D, but that implication is denied by at least some of the leaders in the industry," but notes thereafter that Raymond Gilmartin, past Merck CEO, has "publicly stated that it was the 'value' of the drug to the patient that was the prime determination of price, not what it cost to bring the drug to market."

Relman's brief foray into pharmaceutical pricing reflects his total ignorance of the subject. As I explained in Overdose, high-fixed, low-marginal cost products present special challenges in market settings, because there is little tendency of the price of drugs to fall to the marginal cost of their production, which is all that Gilmartin wanted to say. The argument here depends on a variety of second-best considerations. So long as a company could recoup all its costs and normal profits in the first patented pill produced, marginal-cost pricing would be a great boon for all consumers that purchased the second to nth pills. After all, it is only that first pill that costs $800 million. But there's the rub: Who pays for that first pill? Clearly, no one consumer can cover its full marginal cost, so that if the company is to make back its costs plus a reasonable profit, it has to charge users two through ten million some figure above the marginal cost to make back something of what it lost on that first pill. But how much, and to whom?

Here is where the rubber hits the road. There is no unique algorithm which explains how much gets charged to any customer. The "value" issue arises because you cannot charge anyone more for a drug than it is worth to them, or indeed more than they can afford to pay, taking insurance into account. What makes matters more complicated is that every buyer understands the situation and hopes to negotiate a deal that allows it to pay as little above marginal cost as possible. This leads to price discrimination in these markets, where some customers who can move elsewhere pay less than those who do not. Relman notes that prices remain high at retail pharmacies, but misses the point when he treats that segment of the market as representative of the whole, when in fact they are not. Retail pharmacists and independent drugstores have to carry all lines. HMOs and other provider groups do not and can play off one supplier against another, leading to lower prices for them. Hence, one has to look at the whole market, to see how it all works. It is in this world wholly incomplete to make his observation that "the listed wholesale price of a prescription drug almost never goes down, no matter how much its sales increase." Listed prices are not where the action is. What you have to know is the number of side deals and rebates and special programs that constitute the full market, about whose structure Relman does not have a clue. And through it all, we do know this much: we should expect in general prices to rise as costs for production rise, because companies will not produce goods when they do not think that cost recovery is attainable over its useful life. The distribution of these increases is hard to predict because of the complex dynamics of market pricing, but connection between costs and prices, however hard to measure, is to this extent real.

Relman of course touches none of these issues. He notes that cancer therapies can run $50,000 per annum, but does not examine either the cost or benefit side for these drugs, or how deregulation might lower prices. He notes that the industry runs high profits, but scoffs at the notion of risk-related return, even in the face of the industry's mediocre performance in the stock market. To him the "the industry" (and never the firm) just "dictates" prices. And what does he propose? In his New Republic review, nothing. But his sympathies, like those of Marcia Angell, lie with a system of price controls that is sure to be a disaster, given the number of products, the shifts in prices, the number of different distribution paths and the like. Maybe Relman doesn't think that pharmaceutical companies should behave like ordinary firms. But he surely gives us no inkling here of what pricing regimes follow from his own enlightened view of the industry.

Advertising. Relman is equally uninformed when it comes to advertising, which in his inimitable authoritarian style he would like to limit or ban. But there are two sides to this story as well. Start with the consumer side. On many conditions, people are undermedicated because they are unaware of the risks they face or the conditions that available drugs can treat. Getting this information out is critical and can save lives. And once a patient gets to a physician, he may well take a different drug for the condition than the one for which he saw the ads. The key point is to get patients into the system, which only advertisement can do.

Continue reading >>>

home   spotlight   commentary   research   events   news   about   contact   links   archives
Copyright Manhattan Institute for Policy Research
52 Vanderbilt Avenue
New York, NY 10017
(212) 599-7000