* * * * *
All of the elements of regulation by market forces that operate with hairdressers would operate in this case as well. Surely the strongest incentive at work would be the profit incentive of the drug companies. Their profitability over the long term depends not only on getting a lot of drugs to market, but also on the safety and effectiveness—and reputation for safety and effectiveness—of those drugs.
Other market actors with a strong financial incentive to give the public good information about drug safety, and to administer only safe and effective drugs, are doctors, pharmacies, and hospitals. The profitability of each of these (or the economic health of non-profit hospitals) depends on the quality of care they give their patients and customers, and on their reputation for that quality.
With tort liability always in the background—and with everyone aware of the high damages courts award for medical malpractice or negligence—doctors, pharmacies, clinics and hospitals all have strong incentives to make sensible judgments about the potential risks and benefits of different therapies and to give patients good information. And behind them stand their insurance companies, which pay damage claims. Just as insurance companies have a strong incentive to finance testing and certification of dangerous products such as electrical appliances, they also would have a strong incentive to finance the testing and certification of pharmaceuticals in a free market. In a world without an FDA monopoly on drug certification, insurance companies that insure drug companies would almost certainly require a robust third-party certification process as a requirement for insurance. Indeed, were the FDA’s monopoly to be revoked, Underwriters Laboratories seems a good candidate for one of the first institutions to set up a division for drug testing.
So market forces would regulate, but would it be good enough? When I take my students through this thought process, often one asks if it is not too dangerous to try out market regulation of drugs. What if it didn’t work? Wouldn’t the cost of the experiment in human lives and suffering be too high? It’s a fair and important question.
But behind that question lies the unchallenged assumption that the current system of FDA regulation itself works pretty well and does so at relatively low cost in human lives and suffering. That assumption does not turn out to be valid.
In deciding whether it is preferable to regulate pharmaceuticals by market forces or by the FDA, we need to contrast the reality of market regulation with the reality of FDA regulation. We need to resist the inclination to contrast the reality of market regulation, with all the imperfections it would surely have, with some idealized notion of regulation by the FDA. Economists who advocate revoking the FDA’s power do not claim that regulation by market forces would be perfect—of course it wouldn’t, because in a world where people are inescapably ignorant and biochemistry is fantastically complex, it is impossible for any person or group to forecast perfectly the effects of different compounds on different patients. We claim rather that regulation by market forces would produce much better (though still imperfect) results than FDA regulation. To understand why, we have to understand the problem of incentives in the FDA.
What are the incentives for people who work in the FDA? In the words of Professors Dan Klein and Alexander Tabarrok, “Is the FDA safe and effective?”
There are two kinds of mistakes that regulators such as the FDA can make, Type I and Type II error. Type I error is to allow a harmful drug onto the market. Type II error is to keep a beneficial drug off the market. It is not possible to eliminate both kinds of errors. In a world of inescapable uncertainty and human ignorance, to decrease the likelihood of one type of error does, necessarily, increase the likelihood of the other. The only way to eliminate all chance of letting a bad drug go through (Type I error) is to ban all drugs. After all, somebody somewhere might have an adverse reaction to almost any medicine, however beneficial it might be to the vast majority. Conversely, the only way to eliminate all chance of blocking from the market a drug that might do someone some good (Type II error) would be to ban nothing at all. While the FDA has the authority to ban drugs, it should strike a sensible balance between the two kinds of error, minimizing both together as much as possible.
Now another question for the reader: Are FDA regulators more likely to make Type I or Type II errors? Why? Consider the personal consequences to individuals working in the FDA of making each type of error. Please think it through before reading on, and consider the consequences of that kind of behavior for how well the FDA serves the public interest. When FDA personnel commit a Type I error and allow onto the market a drug that has harmful side effects, what are the consequences for them personally? On the other hand, when they commit a Type II error and keep a useful or even life-saving drug off the market for a number of months or years, what are the consequences for them personally?
* * * * *
Unless you are very unusual, you have answered that people in the FDA more carefully avoid Type I error than Type II. Letting a harmful drug out onto the market has serious, public consequences for FDA officials, especially if the drug causes actual harm. The media publicize any harm done, and the public rebukes the FDA for failing to do their duty. Politicians in Congress get involved, sometimes appointing a commission to investigate a failure to protect the public health and safety. If the harm caused by the drug is horrible enough, some of those responsible for letting it through would undoubtedly be wracked with guilt and self-reproach. In short, for those in the FDA, making a Type I error has strong, unpleasant, public consequences. Henry I. Miller, an important critic of the FDA who worked there from 1979-1994, describes the incentives in colorful language:
This kind of mistake is highly visible and has immediate consequences—the media pounces, the public denounces, and Congress pronounces. Both the developers of the product and the regulators who allowed it to be marketed are excoriated and punished in modern-day pillories: congressional hearings, television news magazines, and newspaper editorials. Because a regulatory official’s career might be damaged irreparably by his good faith but mistaken approval of a high-profile product, decisions are often made defensively—in other words, to avoid Type I errors at any cost….
The consequences for FDA personnel of a Type II error are entirely different. There are no substantial public consequences because such errors result in things not happening—medicines not prescribed, diseases not treated with them (but with some other drugs instead), and patients not cured who might have been cured. Such non-events do not attract attention. The families of patients who have suffered or died because the FDA has not (yet) authorized an effective treatment rarely blame their loved ones’ suffering on the FDA; they believe the patients are receiving the best available care and don’t stop to think that there might be effective treatments that their doctors are forbidden to use.
In brief, then, regulators in the FDA have a strong personal incentive to avoid Type I errors, whose consequences for themselves are severe. They therefore commit a lot of Type II errors, for which the consequences to themselves are negligible. These incentives are disastrous for the American public, however.
One aspect of the problem is very long delays, known as the “drug lag,” in getting a new drug to market. During this delay, patients go untreated who might have been treated; they suffer or die when they might have thrived or lived. Another aspect of the problem is that the FDA approval process is so long and expensive that only drugs that promise to sell to a large population can possibly recoup their development costs. Accordingly, drug companies tend not to research and develop drugs for more rare illnesses. In consequence, there are many diseases for which we might have treatments in a free market, for which we do not have treatments. All the research expense also means higher prices for the drugs once approved, of course, because revenue from selling the drugs must cover the costs of the approval process.
The immediate and shocking problem is thousands of avoidable illnesses and deaths every year because of FDA delays in approving
effective drugs. Figure 6.1 summarizes some examples.
These figures are staggering, aren’t they? And they document only five well-understood cases among hundreds. Extrapolate to the hundreds of other medicines delayed by the FDA—taxotere, vasoseal, ancrod, glucophage, navelbine, and many others—all denied for months or years to real patients with real families, and the horrifying magnitude of the damage suggests itself. Estimates are that, every year, tens of thousands of preventable deaths are attributable to the FDA’s over-caution.
Figure 6.1 Examples of Type II Errors by the FDA
Drug or Device
Length of
Approval Delay
Estimated
Consequences
Thrombolytic Therapy
dissolves blood clots
2 years
up to 22,000 deaths
Interleukin-2*
treats kidney cancer
3 ½ years
3,500 deaths
Misoprotol
prevents bleeding ulcers
9 ½ months
8,000-15,000 deaths
AmbuCardioPump**
not approved
7,000 deaths
annually
Home HIV test
5 years
10,000 infections
* Already available in Europe
** Available in most industrialized countries
The thoughtful reader will wonder about the countervailing benefits: Granted that lives are lost to Type II error, how many lives are saved by the avoidance of Type I error? Do the benefits of the caution exceed the costs? No. According to Klein and Tabarrok, “There is no evidence that the U.S. drug lag brings greater safety.”
The disaster is a consequence of the perverse incentives of the FDA. The individuals working there are good people, trying to do their best, like all of us. But in that institutional setting, the incentives they face are dreadful. FDA critic Sam Kazman puts it this way:
From FDA commissioners to the bureau heads to the individual NDA [New Drug Application] reviewers, the message is clear: if you approve a drug with unanticipated side effects, both you and the agency will face the heat of newspaper headlines, television coverage and congressional hearings. On the other hand, if FDA insists on more and more data from a manufacturer, and finally approves a drug which should have been on the market months or years before, there is no such price to pay. Drug lag’s victims and their families will hardly be complaining, because they won’t know what hit them. … They only know that there is nothing their doctors can do for them. From the standpoint of … politics, they are invisible.
Summary
In a free market, pharmaceuticals would not be regulated by the FDA, UL, or any particular certifier. They would be regulated by the market-imposed requirement for pharmaceutical companies to meet the standards of drug testing that would evolve as doctors, hospitals, pharmacies, and insurance companies choose among the different standards and certifications in their own efforts to please their patients and customers. Rather than the FDA—a government monopoly restricting what drugs may be sold, shielded from competition, lacking profit-and-loss feedback, sure of its customers no matter how little it innovates or how many innocents die waiting, sure of its funding no matter how badly it controls costs—we would have a dynamic process in which any organization might offer its testing and certification services. At any time among these service providers would be those who earn a good reputation for reliable processes and meaningful evaluation standards. Some might advertise the speediness of their process and others the comprehensiveness of their testing, but none could ignore one for the other and survive in the market. Doctors, hospitals, pharmacies, and insurance companies would draw on all of these certifiers and information providers, and standards would evolve with experience.
Of course, all the drug certifiers would make mistakes in their testing or their judgments (as the FDA does now). Sometimes they would certify the safety of a drug that turns out to have harmful side effects; sometimes they would withhold certification from a safe and effective drug. The importance and size of those mistakes would be judged by the doctors, hospitals, and pharmacies, as well as the drug manufacturers, insurers, and knowledgeable people (including bloggers) in the general public. Those certifiers that perform better would gain business; those that perform worse would lose it. No certifier could rest on its laurels nor ignore the tradeoffs among speed, accuracy, and low cost in coming to their decisions, because neither drug companies nor patients could be required to use their services. Best practices would tend to spread to all the different certifiers; certifiers that did not adopt best practices would tend to make losses and go out of business. Steady innovation in testing and disseminating useful information would occur.
Drug manufacturers would have a strong incentive to get evaluations from at least one, and perhaps more, of the certifiers, depending on the cost and the value of second opinions to doctors, hospitals, and pharmacies. Insurance companies would have a strong incentive to cover only therapies whose certifications they trust. Costs driven down by competition, dramatically lower than those imposed by the FDA, would allow for more drugs and therapies to be developed and tested.
As with drugs, so with all other goods and services. Market forces are not perfect regulators, but they are the best available.
Many idealize government regulation. They imagine it to be clean and effective. They imagine that government regulators have the knowledge they need to make good decisions, the wisdom to use knowledge well, and the incentive to serve the public interest rather than their own. But, in fact, all regulators are people—human beings with all their various strengths and weaknesses and inescapable ignorance.
Like everyone else, government regulators tend to put themselves first—they are self-interested. Often their main concerns, like those of their fellows in the private sector, are their career advancement, job security, or next pay raise. Some people won’t stand up to a superior who is clearly making a mistake. Some of us are corruptible, willing to take a payment or promotion for looking the other way. Some of us have simply been promoted beyond our competence.
Because of the natural, inescapable limitations of human beings, there can be no perfectly effective system of regulation. There is no way to “get it right.” Every approach will have flaws. Our goal should be the more modest one of achieving the least-flawed institutional setting—of minimizing the problems inherent in human limitations.
The least-flawed, best available approach to regulation is one that:
generates an on-going discovery of new and better knowledge and invention of new and better processes, and
gives those involved a strong incentive to create value for other people in order to benefit themselves.
A healthy process of regulation will not confer a legal monopoly on one regulatory body. Doing so insulates that regulator from the feedback it needs about how well it is performing. Healthy regulation needs profit-and-loss feedback that rewards good regulating and punishes bad. It needs a close connection of performance and funding. The criteria of good or bad performance must be criteria used by the people the regulation is meant to serve, not those of a bureaucracy with monopoly power.
A healthy process of regulation must be a market process with consumers ultimately in control, with profit-and-loss feedback flowing to both the providers of goods and services and to their certifiers, and with freedom of exchange that allows any enterprise to try its hand at providing information about the safety and efficacy of a product.
The actual regulator in this setting is not some regulatory body. It’s the process whereby 1) different certifiers strive to create valuable information and standards, 2) goods and services providers select from among the different standards, information, and certifications offered—or refuse them all—and 3) customers, at each level, reward with profit those that serve them well and punish with loss those that serve them badly.
Chapter Seven
Special Interests versus Democracy
In reading of the undesirable consequences of government intervention, some readers have undoubtedly wondered why our democratic political processes can’t solve these kinds of problems. For example, if hairdresser licensing laws really do hurt the general public by restricting our hairdressing choices and thereby making us all pay more for a haircut, then the public can simply vote out of office those who support such excessive restrictions, and vote in those who will repeal them or make them sensible. In a democracy the voters ultimately control legislation, don’t we? So why should we abandon government intervention altogether? Why not rely on the democratic process to make it work in the public interest?
This is a question studied at length in public choice economics, and at least two answers are pretty clear. One is that special interest groups tend to dominate the political processes of economic intervention because of the concentration of benefits and the diffusion of costs, a phenomenon I’ll refer to here as “the special interest effect.” The special interest effect is well understood as an elaboration of the second part of the Incentive Principle, that “government interventions tempt people to benefit themselves at others’ expense.” The other answer is that voting for particular legislators is just too indirect a way for the citizen voter to affect particular policies.
As usual, let’s begin with an example.
Federal Dairy Programs
The federal dairy programs are a tangle of laws that benefit dairy farmers at the expense of taxpayers and consumers. There are three major programs:
The Milk Income Loss Contract program “compensates dairy producers [with taxpayers’ money] when domestic milk prices fall below a specified level.”
By Federal Milk Marketing Orders the federal government sets the prices farmers will receive for their milk, thereby eliminating price competition among dairies and raising milk prices above market levels.
Free Our Markets Page 18