11/16/2016 change of mind notice

I am no longer confident that the FDA harms more lives than it saves, and thus I no longer endorse setting fire to it, pending further investigation. What changed my mind was not the commenters disparaging me as a “101 economist” in the comments. What changed my mind was my mom, a long-time professional in the pharmaceutics industry, explaining in detail the rules the FDA plays by and the exact procedures and standards they follow.

I still think that the GDA is a cool idea that should be explored, and there’s a good chance that it will do better than the FDA (measured by number of lives saved) when properly set up.

0 – Escaping the mind kill

My political-economic philosophy is best described as Bowmanite-Neoliberalism, but the word “neoliberal” is too loaded. Most people will call me a consequentialist-libertarian instead. Also, it’s no secret that Slate Star Codex is my favorite blog. So of course, when SSC published a sharp consequentialist-libertarian rant about the EpiPen price controversy, I had a lot of fun reading it.

Too much fun, in fact. The kind of fun you have when you’re landing a cracking punch on the enemy, not when you’re mindfully thinking about complex policy.

So much fun that I neglected to actually read the original Vox piece that Scott railed against. It turns out that the original piece (and the follow-up) was penned by Sarah Kliff. She’s no idiot and no socialist, a couple weeks ago I gave high praise to her article on the gender wage gap. So what’s going on?

In particular:

  1. Who’s right, Scott or Sarah?
  2. What’s wrong with the FDA?

1 – Buyer, seller or regulator?

Scott is talking about the US Government making EpiPens expensive by blocking the entry of competitors, thus ensuring that “the pharmaceutical industry is part of a highly-regulated cronyist system that works completely differently from chairs and mugs”. Sarah blames the US Government for not making Mylan charge a lower price for EpiPens. But “the US Government” isn’t really a single actor in this case. It’s a terrifying hydra with many heads, and each head has its own deal with EpiPens.

In a normal world, there are two types of actors in each market. There are sellers, who set the prices. There are buyers, who decide whether to take the price or not. When left to themselves, buyers and sellers have a remarkable tendency to converge on a price that makes everyone happy. That tendency is the main reason why we get to have good stuff.

Occasionally, there’s a third type of actor: regulators. They sometimes do useful things, like protecting third parties harmed by buyers and sellers. They sometimes do appalling things, like arbitrarily constraining prices and causing terrible tragedies.

It’s important not to confuse sellers (who make the prices), buyers (who take the prices) and regulators (who fuck with the prices).

Back to Sarah and Scott. Scott says that in the case of EpiPens, regulation is causing massive harm. Here’s the quote from Sarah that makes it seems like she’s saying regulation is good and we need more of it:

The story of Mylan’s giant EpiPen price increase is, more fundamentally, a story about America’s unique drug pricing policies. We are the only developed nation that lets drugmakers set their own prices, maximizing profits the same way sellers of chairs, mugs, shoes, or any other manufactured goods would.

This paragraph is so maddening (oh no, God save us from profits!) that I wonder if Scott just threw something heavy at his monitor before he had the chance to read past it. The most unfortunate thing about this paragraph is that it has nothing to do with the rest of Sarah’s article, which doesn’t actually talk about drug pricing but about drug buying, and the harms caused by the regulation thereof.

America doesn’t have a problem with drug pricing policies because America doesn’t price EpiPens, Mylan does. America runs Medicare, which is a buyer of EpiPens. And if it were a buyer free like all buyers to take prices or leave them, it would tell Mylan that at any price above $100 it would switch to Adrenaclicks and that’s that. The next day EpiPens would cost $99.99.

In countries like the UK, there is just one allowed buyer of medicine (aka “single payer”) so the sole buyer is de facto the regulator. But the US Government has been clever enough to create a regulator (FDA), a separate buyer (Medicare) and, astoundingly, a separate federal regulation of Medicare itself. From Sarah again:

In the United States, there’s no such negotiation process to speak of. Federal law bars Medicare, the country’s largest insurance plan, from even trying to negotiate bulk discounts with drugmakers. Once a pharmaceutical company sets its price, the government-run plan that insures 49 million seniors is required to accept it.

Negotiating discounts isn’t “regulation”, and it’s not “pricing policy”. It’s what consumers do. The problem is with the ludicrous law that prevents Medicare from being a free buyer and drags all other smaller consumers along with it. If Mylan can charge $500 to Medicare and they have to charge everyone the same price, they’ll charge $500 even if they lose some other customers to cheaper alternatives.

I’m honestly dumbfounded how that law could even exist. It’s opposed by 83% of Americans, including both Trump and Hillary. Even that 83% surprisingly low: I didn’t know that a whole 17% of Americans were employed by pharmaceutical lobbying companies, and I can’t think of any other reason to think it’s a good idea.


The Medicare non-negotiation policy is like if I walked into the only bar in town and announced that I will buy all their whiskey at whatever price they charge. The bartender quickly raises the price of Jack from $20 per glass to $20,000, and the other customers in the bar are forced to leave in disgust, cursing my name. I spend half a million dollars getting drunk on overpriced whiskey and fall under the table, shitfaced and broke. Just before I pass out, I call the cops to complain angrily about the evil, price gauging, bartender.

2 – The Good Drug Administration

Q: You didn’t really talk about the FDA in the previous section, is Scott wrong to blame them?

A: No. If anything, Scott is underselling the harm done by the FDA.

Q: But the FDA saves people from bad drugs!

A: It does, on occasion. Compared to other developed countries, the FDA saves about 1,000 people per year by preventing them from getting bad drugs. In the two centuries of US history before the FDA got expanded powers in 1962, the worst drug-related disaster killed a mere 107 people. Every death is a tragedy, but 107 people is a small enough number to be on the same scale as casualties due to under-regulation of chairs.

In contrast to the 1,000 saved, the FDA kills 10,000 people each year by delaying their access to good drugs. These numbers are from a large scale collaborative academic report, and the 1:10 estimate is from 1985. It has gotten much worse since 1985. Today it takes a new generic drug four years just to get to the review stage due to back-log.

Q: That’s horrible! So everyone in the FDA is an evil murderous villain?

A: NO! I’m pretty sure that not a single one of the FDA’s 14,824 employees is there trying to kill people. The reason why the FDA is a tragedy is simply: bad incentives plus the power to coerce. In fact, I think that “bad incentives plus the power to coerce” explain the majority of historical tragedies, from the slave trade to Stalin. The FDA is punished when they approve a bad drug, but not when they delay a life saving one, so their incentives are screwed. And since they have the coercive power to prohibit the sale of any drug in the country, their bad incentives kill people.

Q: I get it, we need to reform the FDA! We’ll align their policies with Europe, give them enough budget to clear the back-blog…

A: The problem with the FDA isn’t the back-log or the policies, these are mere symptoms. The only problem with the FDA is that it has the power to coerce the market and block people from buying and selling drugs. All the other problems stem from that.

Q: So what should we do with the FDA?

A: Burn it to the ground.


Q: But you just said that the FDA’s employees aren’t bad people!

A: That’s right, they’re excellent people. I’ll make sure they stand at a safe distance while the FDA burns and I’ll give them marshmallow to roast on the embers. Then, I’ll hire them to work for my new start up: the Good Drug Administration, or GDA.

Q: So you’re saying that if the FDA didn’t exist, we would need to invent it? How Voltairean!

A: The GDA will not have the power to coerce, that’s the difference. And be thankful that I didn’t name it Good Overseer of Drugs.

Here’s how the GDA will work:

I’ll go to all the pharma companies and tell them that I charge 2% of the price of each drug to put the GDA stamp of approval on the package. I employ the experts who know how to test drugs, so consumers will prefer to buy drugs with the GDA seal of approval. Pharma companies want to sell more drugs, so they’ll happily pay the 2% if they know their drug is safe and effective enough to pass my test.  If even half of them agree, the 2% charge will earn the FDA’s old budget of $4 billion in revenue.

Some manufacturers will sell drugs without the GDA stamp, but consumers will be wary of those and the manufacturers will have to lower their prices. The only people buying non-GDA drugs will be the poor (those who can’t afford alternatives) and the desperate (those who don’t have alternatives). But these people weren’t buying FDA drugs anyway since there weren’t FDA drugs that could help them at a price they could afford!

Everybody who was buying FDA approved drugs before will now be buying GDA approved drugs. Pharma innovation will skyrocket, since manufacturers know that even if their new drug doesn’t get GDA approval right away (or at all) they can still make some money of it. And since the GDA only makes money when the drugs are sold, it has every incentive to test the drugs quickly. Can you imagine a private company with a four year back-log on innovation review?

The GDA’s main asset is its reputation among consumers. Its reputation will suffer both if it approves a bad drug, and if people buy non-GDA drugs that turn out to be safe and helpful. Without the power of coercion, the GDA would have to find the perfect balance between scrutiny and approval just to maximize its profits, its only incentive is to do what the FDA is supposed to do today, but better. If it doesn’t, someone will create the Better Drug Administration and will eat my lunch.

That’s literally what we do for every great product we have, from chairs to mugs to t-shirts. Usually it’s retailers who are the approving intermediary. Bloomingdales checks the clothes they sell for quality, because the reputation hit to Bloomingdales from selling a single bad shirt is much bigger than their profit from a shirt. Clothing chains that have lower standards of quality control sell to people who want cheaper stuff that isn’t tested as rigorously.

Somebody can create the Strict Drug Administration that will adhere to the FDA’s old standard and only sell and high-end retailers, and the Loose DA that will do minimal testing on cheap drugs. Customers will decide.

Here’s a similar proposal to my GDA idea, with some more detail.

Q: But if the GDA makes money on every drug sold, they have the incentive to sell Americans too many drugs!

A: Well, it is called the good drug administration, administering drugs is what it’s about.

Jokes aside, the job of the GDA (or FDA) isn’t to control the quantity of medicine Americans use, it’s to control the quality. Asking someone to do both is as silly as asking your central bank to manage both inflation and employment at the same time.

Look, Apple puts its stamp on approval on every app in its App Store. They control the quality. Of course, they also want you to buy a lot of iPhone apps, but you’re not forced to buy them. You know that Apple’s incentives are selling apps, so even when they advertise to you, you don’t spend all your money on iPhone apps. The GDA’s job is providing Americans with a way to acquire great, safe medicine at a mere 2% price increase. The job of preventing Americans from using all this medicine to live long and healthy lives falls to the homeopaths.

Q: Who is against your system?

A: All people who suck at their jobs.

The good FDA employees will be happy to get a pay raise to work for the GDA, and the GDA will have to pay them because it competes with all the other DAs. The employees who suck at their job (of testing drugs) will not be. The good pharmaceutical companies who know they can invent profitable drugs will be happy with the GDA, but the ones who suck at their job (of making new drugs) will lose the existing profits from their FDA-protected market capture. Seriously, the only ones benefitting under the current system are the incompetent.

Q: So why are we stuck with the FDA instead of switching to the much better GDA?

A: Because unfortunately, the set of people who suck at their jobs includes 99% of American politicians.


24 thoughts on “EpiPenomenon

  1. The GDA sounds like a terrible idea, mostly because witch doctors are so great at selling quack remedies, and because it’s much worse to get bad or useless medication for your body than an app that crashes your phone. This seems like a special case of this http://krugman.blogs.nytimes.com/2016/04/20/101-boosterism/?_r=0 – in general, using economic 101isms for solving real-world problems and forgetting to qualify the conclusions.

    Liked by 1 person

    1. Quack remedies today are not subject to FDA approval. Faith healers can make lots of money today. Homeopathy is the first angle of attack to any ailment for millions of people.

      The GDA is not perfect. There will be cases where the FDA would have prevented a death that the GDA let happen. Judging from the various certification authorities who do things much like the GDA would do for all kinds of stuff, those work out surprisingly well on average. If it were possible to live in a GDA world with like 5 competing DAs, I’d try it. Obviously this will never happen, but one can dream.

      Liked by 1 person

    2. Endorse.

      There seems to be a failure mode amount rationalists to map complex problems onto simple models, slap numbers on said models, and then pretend to have solved the problem. And not just rationalists, its just taken to the extreme in this community.


    3. Bloomingdale’s actually once sold me a piece of furniture whose poor design injured multiple people, but I’ll try not to put too much weight on my anecdotal evidence :p


    4. 101 boosterism is more specific and subtle than that–it’s when econ 101 says there’s some effect, and you overstate how large the effect is.

      I think the implicit syllogism goes like this:
      1) This market has many problems.
      2) A free market would not have problems.
      3) Regulation X makes this market unfree.
      4) Therefore, Regulation X must cause all of this market’s problems.

      While econ 101 might say that regulation X could at most cause at 1% deadweight loss.


  2. I don’t really see how the GDA scores better than the FDA-without-ban-powers-plus-time-for-cultural-shift. Any organization big enough to do drug approvals over the entire US is probably going to suffer organizational slowdown. A smaller one won’t be able to handle the caseload, especially if they’re doing the studies themselves (which the FDA largely does not). The incentives seem to stack up about the same, with the modified FDA leaning more in favor of blocking drugs that have flashy rare side-effects, while the GDA will lean in favor of approving drugs that kill patients slowly and not in a noticeable or attributable way, or which don’t even kill them at all but do make their lives less good than they could be.


    1. Taking away the ban powers is my main proposal, I think that alone takes you 90% of the way to not killing a bunch of people. Do we disagree on whether the xDA’s revenue should come from the government or from private sources? I still think that allowing competition between private xDAs will let the best ones reach prominence, but it’s not a slam dunk. Allowing the FDA to compete with “drugs the FDA didn’t approve but people still use” is the more important shift.

      By the way, by “drugs that kill patients slowly and not in a noticeable or attributable way, or which don’t even kill them at all but do make their lives less good than they could be” I assume you mean sugar, starches and junk food? My GDA will never approve twinkies :)


      1. Yeah, that’s it exactly. I think switching to a private service is a bit more of a trade-off than dropping the ban powers, and you’ve sort of elided that. As a person who has read SSC, I agree that the ban powers should probably go, or at least be reduced in some way.


  3. The poll (methedology which is sorely lacking in detail) demonstrating why (presumably) 17% of people oppose letting medicare negotiate drug prices presents some reasons why this number would be this high.

    The same poll found that 74% of people thought it would be effective and reducing prescription drug prices meaning (presumably) 26% of people thought it would be ineffective. It appears that some people want medicare to negotiate prices based on ideological grounds even thought they thought it wouldn’t work, thought it wouldn’t work but leave room for themselves to be wrong and wanted to try it anyway, and/or something else. Also, there were probably people who thought it would work but didn’t think it should be done. Assuming that half of the people polled would not be able to make a good quality professional opinion on the effectiveness of drug negotiation but they would get it wrong evenly both ways and assuming that drug negotiation is effective is consistent with this data and can explain why some people would be opposed to the policy.

    The same poll also found people pay less attention to news stories about medicine then more popular news stories (down in the section “Kaiser Health Policy News Index: August 2015”). This is to be expected but ignoring the comparison the most popular news story of those that are listed (“Profits made by insurance companies”) found that 9% followed this story “very closely,” 21% followed this story “fairly closely,” 27% followed this story “Not too closely,” and 42% followed this story “Not at all closely.” This suggests that it is plausible that fewer then 83% of people considered whether or not medicare should negotiate prescription drug prices before this question was asked to them in this poll and when asked a public policy question on the spot, at-least some people would make a different answer at-least some of the time then when given time to think about it. This phenomenon could also explain why some people would not want to see drug negotiations.

    I can think of three reasons why people might be ideologically opposed to such a policy. I am sure there are other reasons but these are the three I thought of: 1) the belief that if medicare would negotiate drug prices then that would amount to medicare setting drug prices (either because of the market share of drugs that medicare purchases leading to a power imbalance in the negotiation and/or because of a belief that powerful governments don’t negotiate, they dictate) and that governments should not do such things. 2) a signaling reaction to the debate in that it appears that medicare negotiating drug prices is an evil liberal plot (because they only know of democrats advocating for such a thing and/or it sounds like something that these evil liberals would do) and someone not an evil liberal should be opposed to it. 3) That if medicare negotiated drug prices then drugs would be less profitable, if drugs are less profitable then drug companies would do less innovation and the pace of medical advancement would slow which is bad so medicare shouldn’t negotiate drug prices.

    Regardless of how I, or you, or some other people evaluate these reasonings and underlying assumptions, if people do think in these ways it can explain how someone could be opposed to this policy. Most people are not experts (most people aren’t even even amateur experts) in analyzing public policy and when asked to make a decision on public policy should be expected to give a wider spread of results then experts would. I can identify a variety of factors that would lead to this conclusion: lack of information (and the even worse misinformation), ideological predilection towards a given answer (that is they go with their predilection and first impression without really considering the issue), attempt (valid or not) at group signaling, different assumptions (or at-least a wider set of assumptions) than experts hold, different reasoning (or at-least a wider set of reasoning approaches) than experts hold.

    That said 83% of the public being in favor of a public policy in a poll like this is a rather large number. If this was the only determining factor in what public policy gets implemented, it should be implemented. There must be a reason why it isn’t implemented. I suspect that pharmasutical companies are doing some effective lobbying.

    I’m sorry if I’m making too much on a minor point of if I’m confusing feigned confusion for humorous purposes with real confusion.


  4. “If I present a one-sided case, readers call me a clueless ideologue. If I present all the caveats and counterarguments I can possibly think of, the posts would be 15,000 words each and come out twice a year”

    I don’t expect a caveat filled piece, just a basic understanding of the problem you’re supposed to be discussing. I don’t think you’ve demonstrated that here – you haven’t even considered the question “how are drugs like and not like other commodities”. There are many things to consider – two important ones being:

    1) Physical dependence: I’m not sure if you think that consumers should have free and unfettered access to methamphetamines, opiates etc. – the FDA prevents this and the GDA would not.
    2) The fact that you often can’t tell if a drug works by trial and error (in contrast to chairs or machines). You need stronger tools (randomized controlled trials) that the FDA requires for companies to prove not just that a drug is safe, but that it does something at all. Frankly the idea that in without requiring drug companies to prove that their products actually do something, they would not only do so on their own, but we would somehow end up with the same knowledge of which drugs work and how well 5 years sooner … I’m at a loss for words. I’ll point out that the herbal supplement industry exists, and does not suffer in the slightest for not having to prove that their products do something (they mostly don’t).

    Also, I think it is bordering on unethical to call a paper from the Cato institute “a large scale collaborative academic report” as though it were some uncontroversial thing – I doubt most readers know that is where you got the 10,000 number from.


    1. Don’t you wish you posted this thoughtful and informative comment before you randomly shit-talked the rationalist community a few comments up?

      To your points, in reverse order.

      1. The 10,000 is the only number I could find that was backed by any academic research at all. It’s from a Stanford PhD dissertation that was then published in the Cato journal, not a hit piece commissioned by the Cato Institute. It would be nice if the government audited the FDA for total utilitarian impact, but that doesn’t seem likely to happen. I do agree that the fact that this number was selected by Cato for publications means it’s more likely to be biased towards the high side. How low would the number need to be to justify the current FDA in your opinion?

      2. The herbal supplement industry suffers a lot. The size of the entire vitamin and supplement industry is $18 billion in the US, compared to $177 just for brand name pharmaceuticals (excluding generics drugs, non-pharma medicine etc.) There are less idiots than you think. If Vitamin C cured cancer we would all know about it and start popping Vitamin C pills years before any government agency approved it.

      3. I’m not sure it would be in the GDA’s commercial interests to approve metamphetamines, but that’s beside the point.

      To your question: I’m OK with fettered access to metamphetamines. If people use them too much, the government can tax them at a high rate, or make you pass a test to get a metamphetamine-user license, or relegate them to special stores with no visible advertising and plaster scary photos of meth addicts on the packages (like on cigarettes). I would need to see some strong evidence that “fettered” meth causes much more harm than prohibited meth before I would prohibit it. Again, there are less idiots than you think: you and I and our friends and families will not do meth for the same reason we don’t eat all our meals at McDonald’s.

      Prohibition is an extreme weapon, to be used in extreme cases. Right now, it’s used on cheap EpiPens.


      1. I regret the tone of both of my comments and I apologize.

        Let me be a bit more constructive/direct. I don’t think that banning dangerous drugs are the major benefit of having an FDA. I think it is a benefit, but I’m not interested in discussing it right now (I am very relieved to hear that you are opening to the idea of restricting people’s access to drugs by means other than taxation though).

        My criticism of the 10,000 figure stands. Regardless of who wrote it, the Cato journal would only publish it if it agreed with their ideological agenda. I think it is wrong because their model doesn’t have the right parameters. It leaves out the question of which drugs we would use if you got rid of the proof of efficacy (more on that in a bit). The GDA would have no power to enforce the same level of proof that drug companies currently meet.

        Also, The GDA would not allow us to have drugs that are extremely expensive to develop, but are also dirt cheap to produce. Without a regulatory mechanism, a company could not profit out of developing a drug with these properties (because a competitor would simply undercut them and sell the drug for slightly more than dirt cheap).

        Basically I’m frustrated that this is a question of how drug development should work, and you haven’t really addressed the question of how drugs get developed/should get developed ie. what the GDA would actually do to shorten the drug development process. Would they get rid of Phase III? Lower our sample sizes? Get rid of Phase I? If you change any of those parameters you can no longer assume the same benefits of the drugs being approved – because we would not have the same knowledge of them. In fact even the same drug can give different results depending on the pharmacokinetics (Lantus vs Toujeo for instance).

        I brought up the herbal supplements industry because of the fact that they are able to make that much profit out of a large number of products that do practically nothing. They shouldn’t be on par with pharmaceuticals because their products, when actually tested, do almost nothing.

        “There are less idiots than you think. If Vitamin C cured cancer we would all know about it and start popping Vitamin C pills years before any government agency approved it.”

        In a more realistic world – if vitamin C reduced the odds of cancer by 1%, we not not all know about it and we would not start popping vitamin C pills before proper testing with rigorous methodology. This is why drug development is hard. Even harder – if vitamin C reduced mortality when given in doses of 500mg – 750mg units but not higher or lower we definitely not know about it before rigorous testing with proper methodology (there are many drugs like this: vancomycin, theophylline – any drug with a low therapeutic index)

        I haven’t mentioned generics because I don’t know much about them. The approval process could probably better, but before assuming that a private agency will always do a better job than a government agency – it would be useful to actually look at the approval process and think about what would change.


        1. The GDA would not allow us to have drugs that are extremely expensive to develop, but are also dirt cheap to produce. Without a regulatory mechanism, a company could not profit out of developing a drug with these properties.

          The mechanism that protects the right of companies to profit from inventions is patent and intellectual property laws, not the FDA. If you make and sell a drug that’s on patent, you get sued in court for breaking patent, the GDA or FDA have nothing to do with it. That’s the subject of Scott’s last post too.

          Currently, the patent is for 20 years, but it takes 10 whole years to get FDA approval so that companies only have 10 years to profit. I have no idea if that’s too strict or too loose, but I know that the GDA would make the outcome better. Even if the GDA took exactly as long to test and certify each drug (10 years), companies could begin selling it from day one, albeit at reduced prices because they’re not certified yet and most people are afraid to buy them. Even if the manufacturer gets 20% of the yearly profit before the drug is certified, they would get “2 years” of profit in the first decade, whereas right now they get 0.

          This is a huge win: we can either change the patent law to 18 years, so companies have the same profit incentives but generics are available two years sooner or we can keep the 20 years law and see more innovation because the profit incentive is 20% higher.


          1. Patent laws are certainly part of what allow for innovation in drug development, however, the FDA also plays a role via market exclusivity and also the fact that a company can be assured that anyone else trying to say develop a drug for a specific purpose will also have to adhere to a rigorous standard of proof before entering the marketplace.

            I don’t know why you continue to assume that in the absence of regulation – the standards of proof would remain unchanged and we would eventually have the same drugs and the same level of evidence at the end of the day. Regulatory power is a huge part of what drives (forces) clinical trials to happen.


          2. I don’t assume we’ll have the same drugs and level of evidence, I assume we’ll have the drugs and drug research that consumers desire to pay for. Before I bought a used car, I took it to a mechanic I trusted who checked that it works fine. I paid him $200 and got a fine vehicle. In another world, the car would cost $5,000 more and have to go through four years of Federal Car Administration clinical trials. The GDA would be the mechanic, and it will charge whatever people are willing to pay for whatever level of evidence they want to see.

            As to your first point, that’s perverse. If profits are created by a monopoly, companies would invest in protecting the monopoly. If profits are created by competitive markets, companies would invest in making a better product. I want companies to optimize their profits by creating better drugs, not by finding ways to game the system by creating drugs that would be hard for competitors to pass through the FDA,


  5. What sort of market failure gets in the way of a functioning market in drug administration, then? Why bother with a public GDA at all? Don’t the same market forces that efficiently allocate medicine also efficiently allocate stamps of approval?


  6. Two quick points:

    One: not being able to negotiate a bulk discount is not the same as refusing to select the cheapest provider. Given the size of medicare it is arguably justified to prevent them from using their monopsonist power in ways that might be anticompetitive but that is a long discussion I won’t get into.

    Two: I think you (and other proponents) of a non-mandatory FDA like body don’t take a hard enough look at the incentives that such a regime would create (it might still be a net improvement but needs more analysis).

    First, the GDA incentives will likely strongly favor great conservatism in drug approval. As people will no longer be denied drugs because of a failure to approve the GDA won’t face much blowback for taking a conservative time consuming approach and (unless they are rewarded for GDA profits which creates other incentive problems) the agency leaders will primarily want to avoid any public screw up. This means that many important good drugs won’t have GDA approval for quite some time.

    Second, Drug trials and data analysis are very costly and time consuming. Drug companies will only engage in these trials if they have an incentive to do so, i.e., if being GDA approved in your scenario means substantially more sales. However, we already know people are quite willing to use medicines for off label uses and (if this switch makes any difference at all) a substantial number of people must be taking non-GDA drugs. Anecdotal evidence of the effectiveness of these drugs will accumulate in the medical community and the longer important useful drugs remain non-GDA approved the more comfortable people will be buying and using them. So suppose a large number of widely used non-GDA drugs exist should a drug maker pay for the studies needed to get GDA approval?

    Probably not. The drug maker knows there is always a chance the studies they fund will expose the fact the drug doesn’t really work as well as doctors believe or reveal some hidden side effect. That is a huge risk and the boost from GDA approval would have to be massive to risk that profit stream. This likely means even more drugs would be sold long term without GDA approval. Indeed, the only people who really have an incentive to get GDA approval are those whose drugs are wrongly believed to be more dangerous or less effective than the drug marker estimates they are. Worse asymmetric information means that any drug maker who suspects a hidden flaw or lack of effectiveness will most definitely not pursue approval. This means those drugs will stay on the market and never undergo rigorous large tests to find flaws or substantiate effectiveness.

    I mean look at the market for cars. Sure, some people look at the consumer reports safety rating but it hardly dominates the market and car accident deaths are certainly comparable to deaths from poor drug choices. Also lots of people buy ridiculous herbal remedies based on rumor. Why would you expect the market for drugs with a GDA to be any different? Ohh, and don’t count on insurance companies or other actors to intermediate and substitute their better judgement. They want to maximize profit not prolong life and generally end of life care like cancer drugs massively increases costs to the insurer by increasing the length of time the patient lives.

    IMO we need some kind of intermediate between the FDA and GDA. A agency with some level of coercive power but not quite so much as the FDA. Maybe a rule that prohibits Medicare or other federal insurance from covering non-GDA approved drugs. Or a 50% tax on all non-GDA approved drug revenues.


    1. One: Medicare covers one sixth of Americans, hardly a monopsony.

      Two: the point isn’t to invent the perfect GDA, is to let markets come up with as many DAs as we need. If the first one is too strict, someone will start up a looser DA to run quick and dirty tests on the new drugs and just eliminate the obviously horrible one. GDA too loose? Consumers will lose trust and switch to the stricter DA. People using non-GDA drugs, these drugs work great and they tell their friends? Perfect! Now drugs are finally working like a normal market and people get the medicine they need with no hassle.

      I’m not trying to optimize for GDA profits, I’m trying to optimize people’s access to medicine.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s