Our "Protective" FDA

Thom Lambert —  23 August 2007

The FDA, it seems, is rejecting more new drugs. The agency approved only 61 percent of 2007 drug applications through mid-August, down from 73 percent in the same period last year. A new report by James Kumpel of Friedman, Billings, Ramsey & Co. shows that FDA approvals of drugs made from new chemical compounds are at their lowest level in a decade.

This should not be surprising. The FDA is still taking heat for its approval of Vioxx, which manufacturer Merck voluntarily withdrew from the market. The agency is thus flexing its protective biceps, being extra careful not to approve overly risky drugs.

The problem is, federal law forces all of us to live according to the FDA’s risk/benefit preferences, even though they may not mirror our own. Until the FDA determines that a newly discovered drug is, in its view, safe and effective — a process that takes, on average, 14.2 years — the rest of us aren’t allowed to have it.

This would be troubling enough if the FDA’s risk preferences mirrored those of the average person on the street. In reality, the FDA tends to be significantly more risk-averse than the general population. That’s because the consequences of FDA misjudgments are incommensurate. If the FDA approves a drug that’s “too risky” and people are hurt, the media are all over it and there’s a huge public outcry. By contrast, if the FDA errs in the direction of conservatism (i.e., it fails to approve a relatively safe and effective drug, thereby preventing currently sick people from getting drugs that could help them), there’s usually not much of a news story. (See here for a good explanation of the FDA’s incentives.)

And what about terminally ill people whose only hope is a drug the FDA has not yet approved? Too bad, so sad. The D.C. Circuit recently held that the Constitution does not protect those folks’ right to obtain even those drugs that have passed the first (but not all) stages of FDA approval. In the court’s words, there is no constitutional right for “a terminally ill patient with no remaining approved treatment options to decide, in consultation with his or her own doctor, whether to seek access to investigational medicines that the FDA concedes are safe and promising enough for human testing.”

I must say, I agree with Judge Rogers (and Judge Ginsburg) that

it is startling that the oft-limited rights to marry, to fornicate, to have children, to control the education and upbringing of children, to perform varied sexual acts in private, and to control one’s own body even if it results in one’s own death or the death of a fetus have all been deemed rights covered, although not always protected, by the Due Process Clause, but the right to try to save one’s life is left out in the cold despite its textual anchor in the right to life.

But such is our system — a system that diligently seeks to prevent market failure, with no apparent concern for government failure.

Thom Lambert


I am a law professor at the University of Missouri Law School. I teach antitrust law, business organizations, and contracts. My scholarship focuses on regulatory theory, with a particular emphasis on antitrust.

13 responses to Our "Protective" FDA

    market failure, right here 29 August 2007 at 4:07 pm

    I should add that it’s a mistake to assume that individual patients are perfectly rational decisionmakers in this context.

    I think the FDA, detached from the emotions, fear, and desperation that many terminally ill patients and their families experience, are much better positioned to make determinations as to what drug therapies are safe and effective.

    market failure, right here 29 August 2007 at 4:01 pm

    I didn’t realize Hodak was such a Marxist–endorsing the idea of economic coercion and all. Well then Hodak, I’m assuming you’ll be campaigning for stronger worker protections and union organizing. Welcome to the team!

    As for reframing Paul’s analysis here, he’s essentially appyling Neal Komesar’s idea of comparative institutional analysis, and suggesting that we can choose between two regimes: 1) anyone has access to any experimental drug (in consultation with a physician); or 2) open access to drugs applies only to those therapies that are deemed safe and effective by the FDA.

    Paul’s arguing, correctly in my opinion, that we would prefer regime 2 over regime 1. In regime 2, we sacrifice some access in return for greater certainty that the treatments we receive are actually beneficial. We also gain some measure of cost control. In regime 1, we gain access to more meds, with less certainty that any of them will work.

    I think it’s also worth considering that many of these drugs come with serious, sometimes life-threatening side-effects. So at a 9% probability of success, you’re still suffering a 100% chance of enduring side-effects.

    Finally, how about the Hippocratic Oath? Isn’t the first responsibiliity of a heathcare practioner to do no harm? Once we start throwing darts at the board with Phase II drugs, I think we’ve violated that time-honored maxim. (The pharmacist in Madame Bovary provides a nice literary construction of the ‘mad pharmacist tinkerer’)


    No problem. It is my view that granting the right would slow medical progress. Whether our current regime is most effective at promoting medical progress, I don’t have a view on. But you’re characterization of my view with regard to the right is on target. No hard feelings. I didn’t think we were making headway.



    I really try to be civil in these discussions, and sincerely do not wish to engage in ad hominem. I apologize if my comment regarding your position on the current policies overreached in my interpretation of your views.

    I based my comment on what I understood was your sense that absent the exercise of the FDA’s regulatory powers, or something like it, unhindered by a right to access that enabled one to avoid their regulations, we could not achieve the pace of drug development we see today.

    However, even if this were a proper characterization of your position, I should not have used the phrase “people like you” which, in hindsight, is clearly term loaded with derision, even though I merely meant it to denote your position (as I understood it) as being in the majority. You don’t have to respond to this, but I am sorry about that.


    M. Hodak-

    This will be my last post on this subject because you either refuse to address my actual argument or because you just don’t understand it. Perhaps you think you’re addressing it because you mention Rawls or the veil of ignorance, but my argument in no way appealed to Rawls or a veil of ignorance. And you just stomped your foot with regard to this idea of “sacrificing” individuals without addressing my argument. The fact that the government says that its policy coerces or sacrifices people in no way effects what I am saying. The government just may be wrong about why it is right that judges should refuse to find in the Constitution the Abigail Alliance right. And moreover, your definition of “coercion” is silly. On your definition, there is coercion with the fact that you can’t buy chemical weapons or heroin, assuming you want them, at the grocery store. With regard to heroin, perhaps the government should not criminalize its possession or sale. But if it shouldn’t, the moral problem with its prohibition can’t be settled by labeling that prohibition “coercion.”

    And finally, instead of actually addressing my argument, you resort to an ad hominem attack, that it is because of people like me that the government has policies that slow down medical progress. The only thing I’ve argued for is that it would be unjustified to find the right asserted by the Abigail Alliance plaintiffs in the Constitution. So you resort to an ad hominem attack based on barely any knowledge. And thus, this is my last response.



    My definition of coercion does not depend on what I think is justified. If someone threatens a potential provider with fines or prison to prevent them from providing me something I want, it’s coercion–whether or not the threatening agent is “justified.”

    I truly don’t get your logic that there is no coercion here, or that individuals are not sacrificed by current policy. Even the government doesn’t take that position. The FDA acknowledges it’s policy is coercive and that some individuals are being sacrificed for the common good. They (and you) may feel this is justified. I would allow as to how sometimes collective well-being requires coercive restraints on certain individuals. But I see no benefit to blurring the distinction between “coercion” and “justification.”

    As to the justification, I get the logic of the scientific method. I get that some people behind a Rawlsian veil of ignorance are better off with the current policy than with a constitutional right to be free to choose experimental drugs that might, but won’t necessarily, prolong their lives. I just don’t buy the premise behind that reasoning.

    In a free society, when we talk about curbing the rights of individuals (which is necessarily, if justifiably, coercive), it’s because they provably pose some threat to other individuals if left unconstrained. They might be a criminal, or they might be infectious. Generally, those rights are compromised by due process.

    But we aren’t talking epidemiology, here. We’re talking about the regulatory partitioning of actual individuals with (supposed) rights for no purpose other than experimental expediency. We’re talking about using one group of human beings, i.e., those afflicted with deadly disease, to help another group of human beings, i.e., the rest of us (hopefully) not yet so afflicted. We’re talking about dividing the afflicted between “test” and “control” groups. I get the scientific reasoning. It still comes down to the collective good versus individual rights. You don’t get to label that as anything other than utilitarian maximization.

    And we have this regime because people like you can’t conceive that market processes, without such constraints, could possibly devise a better and faster path to medical innovation.


    M Hodak-

    Your defense of your coercion claim begs the very question at hand. You say that there is coercion once you decide you want the experimental drug because the government won’t let you have it. But you’re just saying that that represents coercion because the government is unjustified in restricting your access. That begs the question of whether the government is justified in restricting access, which is the question at hand. So your claim regarding coercion can only be a conclusion regarding the question we’re discussing; not support for the conclusion.

    Second, that 9% of drugs get passed on to Phase II does not mean that 9% of those drugs will help people “save their life.” For oncology drugs taken in phase I studies, I’ve been informed that about 1% of the drugs reduce tumors by at least 30%. So that’s 1% passing that particular threshold, which itself doesn’t imply improved quality of life. And the percent of phase I oncology drugs that actually save someone’s life = 0%. So it is really not justified to keep saying that individuals are being “sacrificed” by the government policy. It’s just not true.

    And finally, I feel like I keep repeating myself – but my point is just misinterpreted. Let me say clearly: i am not a utilitarian and my primary argument was not utilitarian. I mentioned that the challenged policy was justified on utilitarian grounds in case readers were more sympathetic to utilitarianism than i am.

    My argument is that each person – including someone who one day will want a phase-I-passed exprimental drug – is better off not having this constitutional right and better off with the FDA challenged policy because without the policy the very drugs that you might one day want won’t even be there. There is no presumption that your life belongs to the government. The presumption is that we need some rule – whether it’s the FDA policy, the constitutional right, or somethign else – and it would be irrational for anyone to prefer a rule that grants us this right because that rule would make that person worse off. That is not a utilitarian argument. No appeal is made to maximizing overall welfare. Appeal is made to the interests of each individual.

    And finally I have in no way defended FDA policies generally. I have only defended the position that finding this constitutional right would be unjustified.


    Paul- You made four points:

    “I would need to hear why that is coercion. Sure doesn’t look like coercion…”

    That’s because you’re looking at the wrong decision node. The decision to enroll is a later point on the decision tree. The prior node is the decision to try an experimental drug proved “safe” (at least relatively non-toxic–I grant that Phase II drugs are rarely safe in the normal sense). That decision is truncated–the government won’t let you take that drug. That is the moment of coercion. And the government’s explicit purpose for this coercion is that otherwise it would not get enough people to make the second choice (i.e., whether or not to enter a drug trial). The second choice doesn’t “look like coercion,” but it clearly arises from a coercive antecedent. Even the government doesn’t deny that there is coercion involved in this regulatory regime.

    “you don’t back up the claim that that would maximize overall well-being.”

    All that is required for medical progress is, as I put it earlier, two sets of bodies–a test group and a control group. In my comment, I said a control group could be coerced into existence. Just take the pool of people eligible for the drug and randomly sequester a portion of them. Sure, many of the ‘unsequestered’ taking experimental drugs will suffer negative consequences, but I think they must, as a group, still better off for having a (presumably informed) choice to risk those consequences. From a purely utilitarian perspective, society is certainly better off–you still have a test group (via the those in the coerced control group), so science progresses; and more people end up with a choice, which increases social welfare.

    For anyone uncomfortable with the idea of a blatantly coerced control group, the only difference between that and what we have now is simply the decision point at which the coercion takes place. The coercion may be more visible in the regime I’m suggesting, but there is less overall coercion.

    “Moreover, if you think, say, that oncology drugs that pass phase I studies have as high as a nine percent chance of significantly extending and improving your life at that point, you’re deeply mistaken.”

    Aside from the adjective “significantly,” I don’t think I am mistaken. I believe that nine percent is about right in the proportion of drugs that pass Phase II testing. That means that with no other information, a Phase I drug is likely to be good enough to approve beyond Phase II nine percent of the time. The purpose of Phase II trials is to resolve whether or not the actual benefit is zero or positive to some degree (no one said anything about “significant”). The Abigail Alliance claims that all four drugs denied their original co-defendants–all of whom have since died–were eventually approved by the FDA. The odds, whatever they are, are not that long.

    “You seem to regret that we even conduct research.”

    I don’t regret it at all. I’m a big fan of research. I see research being done all the time, and I celebrate it. It’s just that very few instances of research involve the presumption that your life or my life belong to the government. As Thom mentioned in his comment, there are innumerable ways that individuals could be sacrificed in overall utility-maximizing ways. We don’t allow people to be thrown to lions for public entertainment. I don’t see why we would allow individuals who are about to die to be denied the chance to take medical risks that make it more difficult (but by no means impossible) to produce generalizable results.

    By the way, I strongly believe that the FDA bureaucracy has the net effect, via myriad costly restrictions less onerous than the denial of drugs to terminally ill patients, of significantly slowing down the scientific process of hypothesis generation, testing, and validation with regards to patient treatments. If you believe that the current regime is the best we can do for valid results, you are way underestimating the speed of the innovation in a much lower-friction, albeit less rigorous market.


    M. Hodak–

    So all people who enroll in research, knowing that there is a chance they will be in a control group, are coerced? I would need to hear why that is coercion. Sure doesn’t look like coercion – I think if people are informed, they can make a decisions for themselves. And if you’re worried that it is unduly paternalistic to deny someone a drug that has only passed a phase I study, why isn’t it unduly paternalistic to claim that it is surely coercion to ask someone to enroll in a randomized controlled trial?

    And you also say that a utilitarian justification would allow “people to take whatever safe treatments they and their doctors choose.” But that (a) begs the question of what is safe (a drug that passes a phase I study is not, thereby, found to be safe), and (b) you don’t back up the claim that that would maximize overall well-being. It seem quite clear that it will not, given the problems it would cause for medical progress. If you want to reject the argument from medical progress, you can’t do so on utilitarian grounds.

    And if the day comes when I am not allowed access to a drug that has only passed phase I studies and docs have no empirical support to show that it will be effective at all in treating my disease, I, for one, will not be indignant because that would be irrational: the chances that there will be some treatment for whatever I have will be greatly increased, at this point, with a policy that allows physicians to conduct research in the research setting, instead of experimenting on people in the clinical setting where results will not be generalizable and medical progress will not be made. Moreover, if you think, say, that oncology drugs that pass phase I studies have as high as a nine percent chance of significantly extending and improving your life at that point, you’re deeply mistaken.

    You seem to regret that we even conduct research. You do realize that prohibiting controlled research will not mean that medical experimentation will cease. Risks will just increase in the medical care setting for a host of reasons – experimentation will take place there, but without the kind of monitoring that goes on in the research setting, and without the aspects of research that allow results to be generalizable.



    I replied to Mr. Hodak, but the post doesn’t seem to appear.

    But anyway, you ask: “If the text, which refers to “life” and “liberty,” supports a right to engage in homosexual sodomy, wouldn’t you think it would support a right to try to save your life by taking an unapproved drug of last resort?”

    Well, no. I think perhaps I didn’t do a good job explaining my point. Here it is: from the perspective of each individual, there is more reason to endorse the challenged FDA policy than a constitutional right. I am not arguing that I prefer the challenged FDA policy because it maximizes overall welfare. I am arguing that each individual has more reason to *reject* the finding of an individual right here. *Even from the perspective of someone who will one day want an unstudied drug* there is more reason to reject the finding of an individual right. So why should we think there is an individual right if *each individual* has more reason to reject the granting of the right?

    You might say, “but how about from the people who right now want the unapproved drug?” Well, first, as a sidenote, there are ways to get it. There are ways to petition for an unapproved drug. But putting that aside, in thinking whether they have a right to the drug, they must think about the implications for other individuals if the right is found. If the right is granted, she might have improved her chances of significantly extending her life by an extremely minimal amount, and at the same time, the finding of that right will burden at least some other individuals in that it will prohibit physicians and researchers from finding effective treatment for their disease. In fact, I’d argue that the people who want the unapproved drug now have *already been made better off by the fact that we haven’t had this right.”* Since they’ve been made better off by not having the right, I can’t see how they are entitled to the right and depriving every one else of the benefit they have accrued. The very experimental drug that they want might not have even existed if this right had been granted in the past!

    Look – I’m very sympathetic to the concern that it is unduly paternalistic to tell someone, “too bad, so sad, this is too risky for you.” (as I said to Mr. Hodak, it seems unduly paternalistic to me to say that it is necessarily coercive to ask someone to enroll in a randomized study). But notice that I’m not saying that the law should be paternalistic in that way. It’s not the law should tell us not to take stupid risks. It’s that each one of has more reason to reject the right – each one of us is more likely to be better off.

    Now I won’t give you a satisfactory answer here with regard to comparing the finding of the Abigail Alliance right and the Lawrence right since I don’t agree with your view as to how the Fourteenth Amendment should be interpreted. So I will just say that the impact on any individual’s autonomy of a rule prohibiting access to an unapproved drug that has an incredibly puny chance of doing any good for the person is not even comparable to the impact on autonomy of allowing the state the power to criminalize private consensual gay sex. And, to slightly add to the dead horse I’ve killed, I’d argue that there is no infringement on autonomy posed by the challenged FDA rule because each person has more reason to endorse the policy than to endorse the Abig. Alliance right.



    As I think you realize (based on your first comment), my main point was about the perverse incentives facing the FDA. I’m neither a con law scholar nor a philosopher, and I’m certainly not one to find lots of rights hiding in the penumbrae of the Fifth and Fourteenth Amendments.

    Yet, I will stick by my guns here and say that a law that says “you shall not have access to an at least minimally promising drug of last resort when you otherwise will die” is, to use your words, “disrespectful of the individual.” Indeed, it says to the individual, “You must be sacrificed so that we can save others.”

    Now, you could likely justify such a rule on utilitarian grounds. But counting the collective’s utils would seem to be an odd means of adjudicating individual constitutional rights. If 95% of Americans would experience extreme pleasure if Linday Lohan were publicly flogged for her recent DUI (so that utility would be maximized by such a flogging), should we say that Lindsay has no Eight Amendment right to be free from such punishment? Surely not.

    I suppose you could take a “rule utilitarian” (rather than “act utilitarian”) tack. But even that leads to outcomes that conflict with our considered moral/legal judgments. Suppose it could be shown (a la Robert Fogel) that slavery was efficient, so that a rule that permitted Africans to be held as slaves was a rule that led to the greatest good. Wouldn’t we still say that the rule was violative of the slaves’ rights?

    Of course, all this fancy philosophizing goes out the door when we start looking at positive constitutional rights. There, the question seems to be whether the right at issue is really in the text (read in the light of history, tradition, etc.). If the text, which refers to “life” and “liberty,” supports a right to engage in homosexual sodomy, wouldn’t you think it would support a right to try to save your life by taking an unapproved drug of last resort?


    That’s a good rendering of the government’s position, Paul, but it still comes down to this: the government needs two sets of bodies, and we’re here to supply them.

    A utilitarian justification would just as well allow most people to take whatever safe treatments they and their doctors choose, except for a coerced few who are to be locked up and prevented from using those treatments–you know, as a control group.

    As someone who believes in science, I think our society has made an unfortunate choice in treating people as medical specimens. Paul, I hope the day doesn’t come when a bureaucrat has to explain to you that you need to be sacrificed because that life-extending treatment you are being denied has only a nine percent of being effective.



    I’m not commenting on your concern about the FDA’s particular standards for safety and efficacy – that seems separate from your comments on the Abigail Alliance case.

    But on that case … you can’t just dismiss the DC Circuit’s decision – or the challenged FDA policy – as disrespectful of the individual, as you seem to do. Unlike some of the other substantive due process contexts mentioned in the dissenting quote above (e.g., right to fornicate in private, right to refuse medical treatment even if it will result in your own death), finding some constitutional right here would have very harmful consequences for all of us, including those of us who will one day want some life-saving drug. Allowing people to have drugs for which we have only done phase I studies with a very limited number of research subjects and for which we can have little confidence in any risk/benefit profile will be very detrimental to research, thereby slowly medical progress to a significant extent. There would be no incentive for anyone to enroll in a randomized trial when they’d have a right to receive the drug they want in the clinical care setting. And this fact comes with great costs for all of us. Zeke Emanuel points out in an article on this topic that we’ve tried giving this kind of access before. Though there was very little evidence that bone-marrow transplants would be effective in treating breast cancer, states passed laws requiring insurance companies to pay for the transplants so that people could receive them outside of research trials. More than 20,000 women got the transplants – which, mind you, were accompanied by serious risks of pain, death, etc., in addition to the perceived potential for benefit – while it took years for researchers to enroll just 1,000 women in controlled studies. The research results? Transplant not effective compared to chemo. Time wasted (medical progress slowed), unnecessary suffering, and “millions, if not billions, of dollars spent on a substandard treatment.”

    If your concern is for each individual’s interest in saving his or her own life, each one of us has greater reason to endorse the challenged FDA policy over the kind of constitutional right that the plaintiffs sought in Abigail Alliance. And, as a side note, in case you base your overall moral/political outlook on utilitarian considerations, it seems quite clear that granting a right to people to have these drugs before adequate safety/efficacy research is done would not maximize utility.