How should we pay for medical research ?

Historically, the grant of patents on items such as salt and playing cards first emerged as a device for raising revenue and rewarding favourites under the Tudor and Stuart monarchies

To salt and playing cards, one might also add the right to run and collect fees for operating a lighthouse, the subject of a famous paper (and infamous Reason interview) by Ronald Coase.

I think that the government should primarily finance pharmaceutical research, then allow corporations to rip off the government research and make huge profits. In short, privatize the profits, socialize the costs.

I mean, look at what it’s done for health care in the US! It’s like a well-oiled machine!

The current patent system provides all kinds of perverse incentives. It is clearly better to develop a treatment that means dependency than it is to develop an outright cure or preventative therapy. Novel uses of existing, patent free drugs, such as the recent heart cocktail including aspirin have no route to approval. Also the pace of effective innovation is slowing despite rapidly advancing technology.

One area that I am puzzled by is whether or not it is significant that most of the costs of big pharma companies are marketing ones. Are they providing a valuable and important educational service as worthy of protection as the original innovation or is it a form of rent of no worth that only serves to keep a channel open?

I don’t get it. How is a patent equivalent to a sales tax? A sales tax doesn’t provide a sliding incentive for better products among other differences. And there is already a sales tax. And sales taxes are regressive.

Isn’t the ‘distortion’ really a sorting mechanism?

How will you determine who gets to research what? Do you believe that the government will be able to apply the same level of sorting ability between good and bad lines of research that the market provides? How will a centralized governmental system get access to the distributed knowledge needed to choose lines of research?

Does it bother you that most drugs are not discovered in universities and public research labs? If they were more efficient why don’t countries with weak patent systems and strong universities have more discoveries?

Viewing the Australian system as a reward system is perhaps a bit disingenuous since, as John says, the size of the Australian market is so small that it is hard to imagine that Pfizer is primarily (or even vaguely) motivated by a desire to win this “prize”.

I think it would be truer to say that any outcome produced with a patent could be replicated with a sales tax; this is not the same as saying it is the same as a (simple) sales tax; but I think the prime point is that both impose large welfare costs.

Still I think patents are the least best option and prizes probably the most promising – as the recent X-Prize demonstrates – the desire to win can cut the costs of putting someone on the edge of space by factors of 100.

From a practical point of view there is a constant supply of people ho would want to carry out medical research whether paid 20k or 200k – so super normal profits are probably not necessary to stimulate research – grants might therefore suffice for the funding. Similarly prizes, by stimulating the competitive instinct then probably make research more efficient.

But the problem then is research into unknown unknowns – i.e. ideas that fall outside the governments and prize setting committees comprehension; how should this be encouraged? In my view better protection of trade secrets might suffice.

“While these [free-market] methods don’t work forever, in some cases they deliver enough profits to finance a satisfactory rate of innovation. But, as far as I know, no-one seriously suggests this is the case in relation to medical research.”

A comment and two questions:

a. Your list omits private philanthropy, which has been a significant factor in medical research in the past; see e.g. the Rockefeller-funded foundations.

b. Where is the evidence that the free-market methods you list *couldn’t* deliver enough profits to finance innovation?

c. How does one decide, and more importantly *who* should decide, what a “satisfactory” rate of innovation is anyway?

“Australia is a small country, but the PBS still spends around $US 4 billion per year. Pfizer’s share, to take your example, is presumably in the hundreds of millions. ”

But thats sales – if, as you assert, the Australin system results in lower profits going to big pharma, then profits on these sales will be lower than elsewhere.

So its more probable that sales to Australia are viewed as a “bonus” and do not come into the descision making process at Pfzer. If they did consider the Australina market, then the PBS would know that they were hoping to make some a super normal profit in Australia and so would then demand a lower price. So from a game theory perspective, I think the logical thing is for Pfzer to ignore Australia

“Research grants of various kinds are the basis of most fundamental research. This category includes both project-based grants of the kind funded by national medical research agencies and the funding of universities and research institutes to undertake research without specific directions as to the content of that research.”

In fact, intramural funds are more specifically aimed than government grants. But all grant appropriations are judged on a merit basis that includes consideration for future significance. There is no funding free from politics.

A couple of other comments:
The focus in many of these posts is directed at the amount of money spent by industry in relation to their profits. However, industry expenditures on R&D make up between 1/3 and 1/2 of all medical research in the US. A large proportion of the incentive for that research is via patent protection.

It is true that universities and other institutions are searching for patent protection for basic research findings. This is a travesty. Basic research knowledge should be public domain. Likewise copy rights to research publications.

It’s absurd to point to distant advances in basic knowledge as forming a basis for present applications. The point of a patent is to support development.

Prizes for successful research would at least direct the focus back to individuals. Most patents are narrow enough for industry to circumvent and industry doesn’t pay enough for broad patents anymore.

Dan Hardie – the proportion of the difference between US/UK healthcare spending accounted for by drug price differences is much lower than you’d expect. There’ve been several posts at CT and by Matt Yglesias going over this should you have some spare time for Googling.

Dsquared could probably tell me if I’m talking through my hat here.

I could and you’re not. (iirc, MAI stood for Multilateral Agreement on Investment). But yeh, the USA has been engaged in exactly this project for the last couple of years with its bilateral agreements.

John, I don’t quite see how a patent is “equivalent” to a sales tax. I can see how it has the same effect as a particular (although impossible to determine) fixed tax, but sales taxes offer considerably more versatility in finding a more socially desirable outcome (by taxing at less than p1-p0).

One thing I’ve often thought about and find mildly annoying is the lack of marketness involved in determining patents. Given the granting of a monopoly is not socially optimal, why not hold some sort of auction before people do the research regarding the patent length? You’ll only produce this research with a 95 year monopoly, well Jones over here will do it with a 5 year one, so he get’s the patent and thanks for playing.

Jason,

But don’t you run into the problem of auctioning patents on thing that don’t yet exist? My guess is that many patents are awarded due to the creation of drugs that had not originally been planned. In other words, how would this system account for serendipitious discoveries–something that is not entirely absent for scientific development of any kind? For example, it must be common to create a drug that does X in the process of trying to create a drug that does Y.

brad delong and lawrence summers mentioned gov’t patent buyouts in their 2001 paper they presented at the annual fed symposium in jackson hole.

New institutions and new kinds of institutions–perhaps even some that have been tried before, like the French government’s purchase and placing in the public domain of the first photographic patents in the early nineteenth century (see Kremer (1998))–may well be necessary to achieve the fourfold objectives of (a) price equal to marginal cost, (b) entrepreneurial energy, (c) accelerating the cumulative process of research, and (d) providing appropriate financial incentives for research and development. The work of Harvard economist Michael Kremer (1998, 2000), both with respect to the possibility of public purchase of patents at auction and of shifting some public research and development funding from effort-oriented to result-oriented processes (that is, holding contests for private companies to develop vaccines instead of funding research directly), is especially intriguing in its attempts to develop institutions that have all the advantages of market competition, natural monopoly, and public provision.

kremer’s 1998 paper specifically sez, “Patent buy-outs may be particularly appropriate for pharmaceuticals.”

“reread the post and you’ll see that your point about the informational distinction between points and sales taxes was made there.”

Maybe I’m not using the jargon, but I don’t understand what you mean by ‘informational distinction’ at all. Aren’t you suggesting that there is no difference to the customer between a patent and a sales tax?

If so (and I realize that might not be what you are saying) that isn’t it my point. My point is to question how the research money which you posit being raised by a sales tax (or frankly any tax) is to be distributed. How will the government gain the highly distributed knowledge necessary to decide which lines of research to follow?

“…how would this system account for serendipitious discoveries—something that is not entirely absent for scientific development of any kind? For example, it must be common to create a drug that does X in the process of trying to create a drug that does Y.”

In fact the bulk, or at least a massively large percentage, of all drug discoveries are made this way. The usual process is just to synthesize millions of chemical variants of something of interest and then do gazillions of in vitro and in vivo tests until someone points out that chemical X seems to retard the action of biochemical Q which is implicated in the pathophysiology of disease-condition Y, etc.

I worked for a small bio-pharmaceutical company for a summer job once and one of their people – who used to work for the Medicines Control Agency – gave the new employees a 1 hr lecture on the usual drug development and licensing process. It’s massively more random than you think.

For a huge number of drugs they still don’t know how they actually work years after they’ve gone into widespread use. Drug development often proceeds ‘instrumentally’. This works (on some measure of ‘working’), this doesn’t kill people = this is a good drug. The how and the why are often secondary.

We’re still discovering new and interesting properties of aspirin never mind anything else.

baa: I’ll leave it to JQ to answer your specific questions (not least because I suspect I might embarrass myself by getting the answers wrong). But in the meantime, note that patents are not part of the normal functioning of the market. If the government decides to pass a law saying that in certain circumstances it will put you in prison for making and selling a particular chemical compound, then that is a political interference with the market just as much as a tax is.

Treatment protocols as well as diagnostic protocols are already being patented. They’re known as medical process patents and they raise concern in the medical community as well.

Sebastian wrote:
“How will you determine who gets to research what?”

Peer review.

“Do you believe that the government will be able to apply the same level of sorting ability between good and bad lines of research that the market provides?”

Since most marketed innovations are based on govt-sponsored research, the sorting appears to be better than the Holy Market. But hey, what’s empiricism got when stacked next to your religious beliefs?

“How will a centralized governmental system get access to the distributed knowledge needed to choose lines of research?”

Journals.

“Does it bother you that most drugs are not discovered in universities and public research labs?”

No, because your claim is false. As another poster noted, you are improperly limiting your view to drugs, too. You’re also ignoring that most new drugs are “me too” ones, with such marginal utility that they have to be advertised to death. Does any truly important drug advance need an advertising budget?

“If they were more efficient why don’t countries with weak patent systems and strong universities have more discoveries?”

But they are far more efficient. For a thorough look at how pharma companies depend on NIH-funded research and use patents to reward imitation instead of innovation, check out:

http://www.drugawareness.org/pdf/tnrdrugpiece.pdf

Is there anything to this understanding of mine wrt patents and intellectual property in the pharmaceutical industry:

The “natural” way of protecting your monopoly on a drug you invented would be to keep the formula secret for as long as you could. However government regulations have closed off this option to drug companies since they are required to reveal all before the FDA in order to get their inventions licensed for use in the US. So a patent provides a compensation for the government’s taking the drug company’s trade secret.

I don’t know if this makes any sense from a historical standpoint or for that matter from an economic one — it’s just an idea that’s been floating around in my head for a while. I’d be glad to be disabused of it or to have it confirmed. (Or as I expect is likely in this forum, both by different parties.)

Joshua, I think crack dealers and tobacco companies are one answer. Rush Limbaugh and the other prescription drug addicts might be an even better example.

At the other extreme any treatment for a disease which typically comes by surprise and which is life threatening could be sold for enormous profits since all customers would be desperate.

There are dreadful problems in India where much of the drug market is unregulated and even simple things go wrong, partial courses of antibiotics being an important example.

The final reason is that for many conditions the feedback loop will not work well. The patient may get one chance to get it right. Now this applies to other things people buy but there are extensive laws about selling everything from clothes to houses and certainly food. If similar consumer liability rules were applied to drugs I imagine that pharmaceutical companies would insist that the drugs be administered by experts.

“baa, because ideas are public goods, the term IP is completely misleading, as are any intuitions based on ordinary property.”

Isn’t this horrifically question begging? If I spend hundreds of man-hours carving and creating an intricate wood desk that is property. It is just as much property as a wood block, but worth more because of the application of skilled labor plus the desirability of the desk.

If I spend many hours creating a beautiful painting, it is property. It is just as much property as a blank canvas and paint, only (hopefully) worth more due to the application of skilled labor plus the desirability of the painting. If I make many cheap copies of the painting and devalue the painting, something has happened which will make creating paintings less likely because I can’t earn much money from it (in an area of work where it is already ridiculously difficult to earn money). Call it what you will. It is closely analogous to property in the desk sense of the word. This is why copyright exists. Extending it indefinitely is non-ideal, but getting rid of it doesn’t sound ideal either.

If I spend millions of man-hours testing various compounds until I find one that is both effective and safe against a particular disease, suddenly that isn’t property? Not even analogous to property? The lack of short-term property-like protections won’t effect the chance that people will spend millions of man-hours? I don’t believe it. You can’t jargon that away.

Sifting through even ‘public research’ and creating new drugs is one of the most intellectually difficult things done by our civilization. Universities don’t do it because they are awful at it. If you don’t believe it, I encourage you to put your money where your mouth is and invest your retirement in pharma stocks that you pick after say 5 years of intensive research.

Speaking of not understanding the jargon, this quote doesn’t mean what you think it means Jack:

The study reported that in 1998 only about 15 percent of the scientific articles cited in patent applications for clinical medicine came from industry research, while 54 percent came from academic centers, 13 percent from government, and the rest from various other public and nonprofit institutions.

That is because private research on drugs is typically not called ‘articles’ and trials are not always published in articles. Not to mention the fact that if you do your own work on something in an application for a patent that you don’t refer to your own ‘articles’ even if you publish LATER because you want to establish the patent first.

Sebastian, to what are you referring when you say: “Speaking of not understanding the jargon, this quote doesn’t mean what you think it means Jack:”?

I wrote:

My point is to question how the research money which you posit being raised by a sales tax (or frankly any tax) is to be distributed. How will the government gain the highly distributed knowledge necessary to decide which lines of research to follow?

You respond with:

“On the other hand, if the product market in question functions well in other respects (in particular, if consumers are well-informed and there are no cross-subsidies), the profit from the monopoly is a good measure of the social value of the innovation, eliminating the need for governments to make judgements on this issue.”

This precisely covers your point

It doesn’t cover my point at all but I will presume it was because I didn’t frame it correctly.

You are suggesting research grants or prizes because you don’t believe that the patent system is as effective in medical research as research grants or prizes would be. Am I tracking so far?

You suggest an analysis similarity between patent grants and a sales tax.

You then use that as a springboard to suggest that such money could be spent on grants or prizes by the government instead.

If I have any of that wrong, please tell me, because you seem to be saying that I’m not getting it and then repeating the above.

I intended to ask a question about how these grants or prizes would be distributed, but you have responded to my question as if I was asking about patents which your refer to in your response as ‘the monoploy’. I

Your response makes no sense in reference to how grants and prizes will be distributed because your response talks about ‘the monopoly’ of patents while my question is attempting to address how medical research will occur in your theoretical non-patent regime. Unless you are making a logical jump that I just can’t follow, the profit value of a patent doesn’t help us decide which government-funded research will be conducted in a non-patent regime.

So to reiterate, in your theoretical grant or prize regime how does the government gain all the specialized knowledge to decide which areas to fund and which areas not to fund? How does it harness the distributed knowledge of the tens of thousands of people with specialized knowledge who currently contribute to decisions about research. These people currently signal the market with their knowledge by investing. How will the government get all (or even a large percentage) of those inputs. It seems to me that peer review (for example) would include one or even two orders of magnitude fewer of knowledgable inputs than are currently available to sort good ideas from bad. How do you correct for that enormous information loss?

Clearly grants do exist and are allocated with some efficacy and bounties or prizes have done good things from clocks to free software. Things like the PBS also seem to work without the sky falling on anyones head and indeed beter top level indicators. It is also hard to believe that the current system of allocating resources is patently optimal. Nevertheless it does seem sensible to ask how alternative systems would bear extra responsibility. One assumes that grant giving bodies receive some proposals that are obviously worth pursuing and some which are marginal so there must be diminishing returns.

On the other hand how much more valuable would be all that research that drug companies do that doesn’t get pulished if it was? I propose the deciphering of Linear B as an example of how valuable sharing can be, as if one were needed.

It is also worth pointing out that it is not primarliy R&D expenditure that is in practice primarliy protected by patents. In most large drug companies marketing costs are IIRC roughly thre times R&D and that trend is increasing. I’m not sure if that is significant but marketing is much less emotive than R&D to my mind.

Sebastian, JQ has pointed out a number of models for allocating the money — Grants, prizes and systems like the Australian state drug buing system. All these systems certainly provide scope for entrepreneurialism prizes obviously, grants too and if you are going to allow Wal Mart so does the state drug buying model. These might or might not be as efficient as automatically granting temporary monopolies but the discussion ought at this stage to be more specific.

It should be born in mind that a change in system could still be a good idea even if the distribution of research funds were less efficient because the fund raising side is more efficient.

Also that from a free market point of view the monopoly is pretty horrid. All that research not shared. Once the patent is filed behaviour is purely rent seeking which is definitely inefficient. This is actually a broad criticism of economic defences of libertarianism or libertarian coat tail hanging tax policies. Specifically while many tax and redistribution schemes do create deadweight inefficiencies, so does accumulation of wealth and leaving wealth alone in order to preserve incentives ends up producing inefficiencies of just the same nature of its own.

Well no actually you didn’t. You disparaged my ability to understand your points, when it is now clear that I understood them just fine. My point may not have initially been expressed as an economist would have expressed it, but the very paper you point me to raises the very issues I raise. I point to the issues as raised by the very well known economists that you quote to me, and you still engage in a brush-off.

“My post discussed advantages and disadvantages of various mechanisms.”

Postively lawyer-like. You discussed what you see as the advantages of grants and prizes and the disadvantages of patents. That is technically ‘advantages and disadvantages’ of ‘various methods’ but since you failed to deal with the advantages of patents (which the Delong Paper does) and the disadvantages of your system, (which the Delong Paper does), one can hardly say that you exposed how you come to the conclusion you come to which is–“But it seems likely that the average price would be well below that prevailing at present and that the allocation of research effort would be more socially beneficial.”

If Jack sums up what would be your response, it is no response at all. I have no trouble understanding that you advocate the distribution of funds through grants or prizes. There has never been any confusion on that point. My question has always been how does the government choose which research to fund through grants or prizes and does that do better than the market. Jack’s response is that the government chooses to fund it through grants or prizes, which isn’t the issue at all. Mr. DeLong’s paper, and the paragraphs I quote from it put that very issue into economic-speak, but you still seem to suggest that answering “grants and prizes” is any kind informative response. I knew “grants and prizes” were the delivery method from the first read of your post. I want to know what the choice method is unless you think that all possible research can be infinitely funded–which I suspect you do not.

The Australian scheme may provide an interesting hybrid (though I note that you spend almost all of your time defending grants and prizes) but you brush the problems of applying it locations where the majority of the research is done as if differences in scale of application would be trivial. As noted above, Australia is almost an afterthought market for such research products. The companies would rather make marginal profit there than zero profit if Australia took the patent, but very few (if any) of the major players would expect to gain their sunk research costs from that market.

With respect to the grants and prizes which you make the centerpiece of your argument, Delong’s paper itself suggests quite the opposite from what you suggest:

Moreover, in a Schumpeterian economy it is innovation that is the principal source of wealth—and temporary monopoly power and profits are the reward needed to spur private enterprise to engage in such innovation. The right way to think about this complex set of issues is not clear, but it is clear that the competitive paradigm cannot be fully appropriate.

Note I do not snip the quote at the first sentence where it would appear the DeLong fully agrees with me, I include the sentence where he suggests that methods beyond the patent system might be necessary.

At this point it seems likely that you are one of the academics who, according to DeLong, “…long underestimated the importance of Hayekian insights into market competition as a discovery mechanism, of the entrepreneurial advantages of private enterprise, and of the administrative defects of overly-rigid systems of top-down control that come with centralized funding.” And apparently you are in good company since DeLong refers to that as “mainstream”. But really there is no need to pretend that I couldn’t understand you. I raised the very same issues that DeLong raised, and I did it before you pointed me to his article. Just say that in your economic view DeLong’s understanding above isn’t worth responding to.

Perhaps a better link is the following, sorry.

http://research.yale.edu/lawmeme/

re: information as public goods, dan kohn has a nice introductory primer and eben moglen goes through some of the legal ramifications

Sebastian, the short term monopoly referred to in the Delong & Summers article need not be an artificial one at all. It can be merely first mover advantage as works quite well enough for Amazon or Google.

If the goalpost is now how the choice mechanisms can incorporate market knowledge, you could just say so. In medicine there are clearly areas where an expertise weighted market opinion might do better than a wealth weighted one.

In fact I share your doubts about the efficacy of an institutionalised system but I don’t think the options provided by Professor Quiggin fail to address this.

It is easy to confuse patents with their purpose. Professor Quiggin has already demonstrated the deadweight inefficiencies they cause but there are plenty of other objections to them. For example they only reward the value of the innovation, not its difficulty. That leads to all sorts of spurious results and rent seeking behaviour. There are many instances of people infringing patents with no awareness of the existence of the patent holders work at all. Such chilling effects should not be underestimated. Surely a libertarian would think that the market ought to be able to solve the problem of rewarding innovators without the need for the state to intervene through patent law?

There is also the high cost of obtaining one and the extravagant cost of defending yourself against a well funded opponent. Patent law is not even mostly about Edison and lightbulbs.

My concerns are more aboout the difficulty of persuading institutions to adapt rather than setting one up that can make good decisions. Markets seem quite good at changing incrementally.

How durable do you imagine the advantage conveyed by being the first-mover would be?

I can imagine anything from three months, if you make drug companies publish the composition of their molecule and let the generic manufacturers prove their similarity to the brand-name drug, to perhaps three years if you allow the drug companies to keep their products a secret and make the generic companies do their own clinical trials. With a little industrial espionage, the market-based monopoly might disappear entirely.

In neither case is it likely to be economically worthwhile to do drug development.

I don’t see why you think patents cannot be implicated in the part I quote since just prior to those sentences is found: “However, forcing prices to be equal to marginal cost cannot be sustained because then the fixed set-up costs are not covered. Relying on government subsidies to cover fixed set-up costs raises problems of its own: it destroyes the entrepreneurial energy of the market and replaces it with the group-think and red-tape defects of admininstrative bureaucracy.” Doesn’t sound like thought about first-mover advantage.

But now you are engaging in wholesale nitpicking while avoiding the central question which is presented by the very paper you cite. Once again–how do you intend to harness the information currently available to the market, how will you evade “entrepreneurial advantages of private enterprise”, how do you evade “the administrative defects of overly-rigid systems of top-down control that come with centralized funding”, and how do you deal with the problem that “government bureaucracies have never been able to choose and assess the directions of applied research and development very well” to use Delong’s phrasing? In layman’s terms: how do you choose what research to fund?

This will never get read.

Thanks. After some careful reading and some thought(thanks) I can see your point. I think.

But I think it all hinges on my acceptance that given the correct circumstances ‘the profit from the monopoly is a good measure of the social value of the innovation’

Do you generally accept that proposition? That market value = social value? (Which I think is pretty much what that ends up meaning. But please correct me if I am wrong.)

If you do wouldn’t the better solution be to eliminate the cross subsidies and third party payers?

And if you don’t how do we haggle over social value?

I say market value doesn’t equal social value.

I say the market undervalues the social value of the innovation.

So I say keep patents since that leads to the highest price and most innovation/social value.

In fact I say increase the (real)price via subsidies. More public R&D. (We need to keep the patents because governments cannot make final products and that is an important step.)

Did I get that right? Does that make sense?

“I think we’re now agreed on the issues that need examining.”

Argh. It is too bad we couldn’t have gotten to this point four days ago instead of spending three and a half days complaining about how dense I was and how little I understood economics. But since we did and this post is about to fall off the front page, I guess we won’t get to talk about the issues.

“In the case of health care I argue (and I haven’t seen you offer any counterargument)”

Well, hmm that may have been because you spent four days pretending I didn’t understand your post and that my points were completely irrelevant until they showed up in the DeLong paper that you threw at me. And even then you dodged for almost two days and multiple posts. And still you are being unusually coy. Do you intend to merely ignore those problems and hope that they won’t be a major factor in your grants and prizes? I still don’t know.

BTW, the relevant part of the interview cited above is near the bottom, and is completely independent of the prior discussion. This should give you a flavor, but do read the rest:

“Reason: You accept no money from industry.

Ames: No. If I give a talk at Du Pont. I have them send the honorarium to charity.

Reason: So the research money comes from the government?

Ames: My research money all comes from the government. It’s fiercely competitive.

Reason: Over time, the government, especially the federal government, has acquired a near monopoly on the funding of scientific research… “