The conventional wisdom among economists is that introducing or strengthening patents (as required by TRIPS) is very bad for developing countries. Whatever benefit there might be from an increase in domestic innovation is sure to be swamped by the great increase in payments to U.S./European patent holders. A very rough rule of thumb would say that if demand is linear, and introduction of patents causes a switch from perfect competition to complete monopoly, the loss to the developing country would be about 1.5 times the gain to the developed country holding the patent. It turns out that the situation may be considerably worse than this.

The American Economic Review has published a careful study by Chaudhuri, Goldberg and Jia which is a case study of a particular pharmaceutical product in India, Quinolones (a key molecular ingredient of several antibiotics). They estimate that the primary loss will be to consumers and not to domestic pharmaceutical producers. The domestic producers may actually gain - the increased cost of quinolones based drugs will lead consumers to switch to domestic drugs based on different molecules, increasing demand for those products. The effect on consumers depends on the extent to which India regulates the prices of imported drugs: the estimates range from $144-$450 million. By way of contrast, the gain to foreign multinationals is estimated to range from $19.6-$53 million, again depending on how much price regulation there is. That is, the loss to India is estimated to be 7-9 times the gain to the (rich) Western exporters. The reason for this is quite striking: CGJ consider in addition to the effect of increased price, patents will cause the products to become less available and accessible. It turns out that lack of availability induced by patents has very significant costs.

Whatever ones overall view of the patent system, it is hard to defend the portion of the system that taxes transfers from a poor country $7-9 so that a rich country may earn an additional $1.

The American Economic Review version of the paper is not available online, but there is a working paper version available here.

Professor Levine:

Any sensitive person has to be pained at the idea of withdrawing useful medicines from the poor and the sick. Especially when so little seems to be gained from it.

I see that you are an expert on innovation and growth, so I respect what you do very greatly. But when, in your closing line, you speak of TAXING a poor country, (rather than charging customers for services rendered) you seem to beg the very question this blog is established to debate. That's especially surprising when it follows the friendly preface: "Whatever ones overall view of the patent system,..."

Doesn't an analysis that carefully considers the costs implicit in this lack of availability...seem to need some notion of availability issues on the other side of the ledger? Are we to imagine that the $19.6-$53 million would all go to some "rich country" and disappear into party accounts of people like Paris Hilton who happen to be named Merck or Johnson?

Many goods and services provide grossly more benefit to the buyers than revenue to the sellers, don't they? Should we search for such situations more broadly and somehow resent sellers like Costco and SouthWest Air?

At a time whe drug researchers are being laid off and when price controls in Europe and Canada already leave US consumers paying most of the freight on current research....does it seem reasonable to treat existing medicines as manna from heaven?

Do you find Murphy and Topel convincing when they say that the benefits of medical research dwarf their costs in general? Is 7 to 9 times even an unusual ratio?

Your website has papers that clearly reflect on your sense of how we would finance all sorts of innovation without monopoly...can you guide me to an introductory piece that explains how we'd end up financing more good research, rather than less? Especially wrt new medicines they need so desperately in developing countries?

Dave Meleney

Sorry - you are right, that wasn't the best choice of terminology. I meant "taxing" in the figurative sense as in "carrying that load up the stairs was very taxing." Obviously it isn't literally a tax. As to exactly who gets the money going to the "rich country" I'm not sure it matters all that much: the average US resident earns about 14 times what the average Indian resident earns, so unless all the proceeds go to the homeless, it pretty clearly is money going from poor to rich. As to the returns on research - only a small fraction of pharmaceutical company revenues go to research, and the rate of return on investment in private sector pharmaceuticals isn't nearly 7 or 9, probably more like 20%. It may be that additional government/academic medical research has very high returns, but this surely isn't what the extra modest earnings of the pharmaceutical industry is going to be used for. Leaving aside the issue of whether the companies that invented (more likely participated in the process of invention) of the drug "deserve" every penny of social benefit it brings - I leave it to your judgement whether we should pass laws that increase the profits of our companies by a dollar at the expense of seven to nine dollars paid in poor countries.

The final question about "financing more good research" doesn't have an entirely satisfactory answer. The effect of eliminating patent monopoly on pharmaceutical research is complicated, and it depends on such critical things such as how we pay for clinical trials - the most significant way in which the large pharmaceutical companies contribute to the development of new products. On the one hand, the products are more useful, easier to improve and develop since research isn't hindered by existing products, and there is less wasteful "me too" research. On hand, the inventor won't earn as much. So the overall effect is unclear. We have put together what is known as best we can in our book (link here): there is a chapter specifically about pharmaceuticals, as well as a chapter discussing the broader policy issues. It is designed to be introductory.

Professor Levine:

Thanks for your thoughtful reply. Perhaps I took you over literally, sorry. Many in the "movement to rid us of patents" would indeed consider "tax" or "private tax" a fair, and in fact persuasive, term for the charge to be levied here.

The ratio of 7 or 9, if I understood it, was a ratio of lost benefit (in poor developing world consumers) vs. revenue gained (by large, rich, foreign corporations). I'm not sure how typical figures for Pharma ROI are relevant here. This ratio does, however, align well with the variables Murphy and Topel have been studying: "In that research they found that from 1970 to 2000, increased longevity and improved quality of life, much of it thanks to medical research, added nearly $3.2 trillion per year to total wealth in the United States, equal to about half the total gross domestic product" http://magazine.uchicago.edu/0612/features/murphy.shtml Measuring the Gains from Medical Research: An Economic Approach (2003), and in “The Value of Health and Longevity” (Journal of Political Economy, 2006)

I read that a lot of Chicago economists view Kevin Murphy as the smartest guy working their beat.... but he could still be way off on this. It does seem to suggest a 7 or 9 ratio is not unusually large and that the gains to consumers often dwarf the profits to those who made it possible. I was trying to suggest that this happens at Costco and South West Air also... and that we don't normally begrudge someone control of their property just because they are able to use their property (or not) to help others more powerfully than to help themselves.

If Pharma research dollars do bring consumer benefit in the US at remarkable ratios, even that wouldn't undo the unseemliness of withdrawing medicines from the terribly poor in India. But it does bring up dramatic questions of how we quit leaving so much consumer benefit (useful, joyful, human life) on the table. And whether getting rid of patents is the right direction. Wouldn't it be better to argue for getting rid of the Kefauver-Harris Amendments which seem to save a few thousand lives a decade but seem to cost hundreds of times more in drugs not developed (Peltzman's model)?

As an amateur follower of issues in drug development policy, I am amazed that most economists have ten times more interest in baseball economics than this field, though you look to be in the minority here. Thanks for what you do; if you change law such that pharmaceutical progress ensues more rapidly-- my whole family will benifit greatly, wheras my Colorado Rockies will be whupped by your Cardinals (Dodgers?) no matter how much we study them.

all the best, Dave Meleney

I went and reread Murphy and Topel for whom I have enormous respect. The JPE paper is about estimating carefully the value of medical improvements. It doesn't appear to have an estimate of a 700-900% rate of return on investment in medical research. Moreover, they consider all medical improvements, of which improved pharmaceutical products are only a portion. Moreover, also according to Murphy and Topel, the private sector pays for only about 1/3rd of biomedical research. Only about 1/3rd of the products the private sector creates are believed by the FDA to have therapeutic value (the rest being imitations of existing drugs). The big pharmaceutical companies also spend about double on marketing what they spend on R&D. So even if we assumed that the social value of a dollar of research was 7-9, a dollar going to a pharmaceutical company would result in only 1/9th of that value, so not a very good return.

The issue of property is a complicated one. You take the view that a chemical entity is the "property" of the company that invented it. In law, this is true only to a limited extent, since the patent is limited in duration to 20 years. More important - at the time these drugs were invented they could not be patented in India, so they were not invented with the expectation that they could be patented in India, nor were they property in India at that time. Indeed, the "property" belonged to Indian consumers and pharmaceutical firms - and the change in law was a "taking" of property from them and the giving of it to Western pharmaceutical companies. From an economic perspective, we award the "property," more properly "mononoply" only because of the incentive it provides (see the previous paragraph). From a moral perspective, why is an inventor of a drug entitled to property right in a monopoly over that new drug, but the inventor of a new mathematical formula is not?

If I thought that abolishing patents would have any important negative impact on advances in medicine I wouldn't support it. The reason I don't think that is the case are documented in our chapter 9.

Finally, on the Kefauver-Harris Amendments I happen to agree with you; in fact I'd like also to see the prescription system - also a form of monopoly - abolished as well.