Thursday, October 12, 2006

Drugs for Africa: A Modest Proposal

My previous post mentioned the issue of how to make medical drugs for conditions such as AIDS available in poor countries. On the face of it, there should be an easy solution. Sellers want to make money; you cannot make money trying to sell someone something at a price he cannot afford to pay. One would expect drug companies to be happy to charge lower prices in poorer countries, as long as they get enough to more than cover the actual production cost of the drug.

The problem is resale. How can the drug company be sure that the drugs it sells at a cheap price in Zambia don't get resold in France or the U.S., reducing the quantity it can sell in those countries at a high price? One possible solution is to control distribution; once the drug is in the patient it cannot be resold. But that may be difficult to do in poor countries, lacking the infrastructure to monitor what happens to the drugs once they get there.

I have another solution. Let charitable donors in rich countries buy out the patent on the second best AIDS drug or combination of drugs and public domain it—let anyone who wants make it. Buying the second best drug should not be that expensive, since it probably is not making much money any more. And even if the same company owns the first best drug, it should not lose too many sales, since most customers who can afford the best drug will keep taking it.

This proposal has one large advantage over the usual alternative of forcing drug companies to make their drugs available at a low price in poor countries, with the threat that if they do not the countries in question will simply refuse to enforce their patents. That proposal makes the development of new drugs less profitable and so buys a short run gain in availability at the long run cost of slowing the development of new drugs. It could be a very large long run cost if the practice spreads from very poor countries up to less poor countries.

My proposal, on the other hand, makes the development of drugs more profitable. You can not only make money on your drug until a competitor brings out something better, you can even get a little more money at that point by selling it to the Gates Foundation or some similar organization.

While on the subject, I have a second suggestion, this one intended to make drugs more available for both rich people and poor people. FDA rules on testing should be designed to encourage drug companies to make not yet approved drugs available abroad in order to use the information so generated to meet the requirements for approval in the U.S. That would bring down the cost of finding out whether new drugs are safe and getting them approved. At the same time it would provide low cost—albeit somewhat risky—drugs for people in poor countries.

My previous post discussed a conflict between good economics and good rhetoric, between an argument that was right and one that was persuasive. These proposals face similar problems. Opponents will argue that it is unjust for rich people to get the best drugs and poor people the second best—even if the realistic alternative is poor people not getting any drugs at all. They will make good demagogic use of the idea that it is wicked to use human beings as guinea pigs for potentially dangerous drugs—despite the fact that using humans as guinea pigs is the only way we have of finding out whether or not drugs are safe.

16 comments:

Anonymous said...

My very loose understanding, from a piece on NPR, is that more and more drug companies do the bulk of their testing abroad. If I recall the piece correctly, even keeping to the same standards as they would in the US, it is a lot cheaper to do testing in India.

Ah, here's a link to some Washington Post articles on the topic: http://www.washingtonpost.com/wp-dyn/world/issues/bodyhunters/. It is quite slanted, but has some data about current practices.

Anonymous said...

As much economic sense as it makes to give the 'second best' drug to the poor abroad (and domestically as well), it so violates political correctness that I would think the idea has no chance (I wouldn't mind being proved wrong). To people with a certain moral philosophy your idea translates to saying that those poor people are second class citizens, of lower worth. To them this idea would not be a convenient solution, but an abomination akin to overt racism.

I shouldn't have to point out that the non-FDA approved drugs are even worse in this regard.

Steve said...

Sorry. I'm pretty sure that wouldn't work. There just isn't a universal best and second-best drug for anything: differences in metabolism and previous medical history alone will see to that.

A variation on this idea did work for academic texts when I was a student. Here in Britain, you could have bought either the International edition or the American one, but the price difference (I think the International edition of the organic synthesis text I'm thinking of was printed in the Philipines, so transport costs were high for both) was smaller than the difference in quality (stitched signatures as opposed to perfect binding - just plain silly for 900+ pages, and gluing on a cardboard cover over a previous paper one...)

Trent McBride said...

This idea is interesting, of course, in an general sense, but may fail specifically, as described, with regard to AIDS drugs. These drugs are necessary to be given in cocktails because of the problem of resistance development. Often, single drugs can be wasteful or even harmful if they create a stronger virus. This is not to say your proposal could not be modified around this problem.

Anonymous said...

Buying the second best drug should not be that expensive, since it probably is not making much money any more.

The price of the patent should be closely related to the price of the drug. What you suggest is not very different from buying and donating doses of the drug. It does solve the problem that those donated doses might head back to expensive markets, but it doesn't sound cheap.

Anonymous said...

An idea I read on someone else's blog (I think it was Educated Guesswork, but can't find the article there) is even simpler, but a bit more radical.

Basically, discounted versions of drugs would come with a disclosed share of a toxin known to increase cancer risk or something to that effect. Presumably, rich-world patients would prefer paying the premium price for a toxin-free drug, while people whose life expectancy wasn't long enough to even consider cancer would be better off with the relief from the cheaper drug.

Of course, this is not very feasible politically.

Anonymous said...

What about a long term insurance scheme. The probabilities of a person dying within 20 years are (based on the US):

Age and probability of dying within 20 years

0 0.0140
1.0000 0.0076
2.0000 0.0080
3.0000 0.0086
4.0000 0.0093
5.0000 0.0100
6.0000 0.0107
7.0000 0.0114
8.0000 0.0122
9.0000 0.0130
10.0000 0.0138
11.0000 0.0146
12.0000 0.0155
13.0000 0.0165
14.0000 0.0176
15.0000 0.0186
16.0000 0.0196
17.0000 0.0205
18.0000 0.0214
19.0000 0.0223
20.0000 0.0232
21.0000 0.0243
22.0000 0.0255
23.0000 0.0269
24.0000 0.0286
25.0000 0.0305
26.0000 0.0325
27.0000 0.0349
28.0000 0.0375
29.0000 0.0403
30.0000 0.0433
31.0000 0.0465
32.0000 0.0501
33.0000 0.0543
34.0000 0.0583
35.0000 0.0632
36.0000 0.0686
37.0000 0.0746
38.0000 0.0808
39.0000 0.0880
40.0000 0.0959
41.0000 0.1044
42.0000 0.1138
43.0000 0.1241
44.0000 0.1354
45.0000 0.1478
46.0000 0.1613
47.0000 0.1762
48.0000 0.1919
49.0000 0.2093
50.0000 0.2285
51.0000 0.2489
52.0000 0.2717
53.0000 0.2965
54.0000 0.3245
55.0000 0.3544
56.0000 0.3866
57.0000 0.4212
58.0000 0.4597
59.0000 0.5008
60.0000 0.5468
61.0000 0.5970
62.0000 0.6542
63.0000 0.7170
64.0000 0.7855
65.0000 0.8618

There is a glitch for 0 due to infant mortality. However, if you make it to 1, then the odds of dying within 20 years increase exponentially as time passes.

People would buy insurance every year, but get overlapping cover (i.e. you buy 1/20 of a 20 year policy rather than 100% of a 1 year policy every year). This is different from the current system where people buy for just 1 year with the rule that they can renew even if they get sick.

Children have very low mortality. This means that a low premium will give very high cover (or more to the point they can build up to a large cover very quickly)

Also, if you end up getting sick, you will still have nearly all your cover as the previous 20 years will still be active.

Anyway, the benefit is the insured people have now passed their interest in living longer to the insurance companies who can aggregate the costs. This would only be death benefit (there would be a need for general insurance for regular higher probability treatments). Also, the reason that the person died would not matter, the payment would happen anyway (except for suicide probably).

Insurance companies could then offer prizes for curing certain diseases. They could also trade patients. A company might buy up lots of people who are at high risk of heart disease and then offer a prize to reduce heart disease (and then sell off their contracts as the price will have shifted upwards due to the prize .. hopefully for them :) )

Also, insurance companies would try to give their clients accurate advice on what treatments to take. Their interests are aligned in having the client extend their lifespan as then some of the payments will lapse.

Also, they will only cover treatments that are economically efficient. The premiums that a client selects effectively set the amount the company will pay for 1 extra year of life for the patient.

Anyway, the real problem is that people don't want to apply economics to health care. They would rather just not think about it ... and definately not have to answer questions like "how much is 1 more year of life worth to you ?"

Anonymous said...

It does solve the problem that those donated doses might head back to expensive markets

No, it doesn't even solve that problem! It just ignores it by donating drugs to everyone. (One could also directly ignore it -- keep intellectual property in the developed world and public domain it elsewhere.) It does have the advantage of getting intermediaries (donors) out of the price system.

Anonymous said...

Because Japan has a socialized medical system, they're desperate to cut costs in any way possible. One way they do it is to use second-rate drugs. If you get a prescription from a doctor in Japan, check it out on the Internet. You sometimes find that it was obsoleted by a new drug in the U.S. The U.S. pharmaceutical company then sold the rights to a Japanese pharmaceutical company, and the obsolete drug is manufactured and routinely prescribed in Japan. Japanese doctors complain that they can't get access to the best drugs. It means there's a fairly strong demand even for out-of-date U.S. drugs.

Anonymous said...

What is a patent anyway?
The only legitimate way a pharmaceutical company can enforce patent right is by binding the seller of every unit of medicine by a contract prohibiting reverse engeenering. Thus it would have to prove that a competitor didn't come up with the formula himself but breached the buying contract. They could also further replicate the patent system by also making the buying contrat a non-compete agreement. Having to behave that way to enforce "patents" makes it just an easy extra step to segment the market.

Mike Huben said...

The "no patent in the third world" policy could work well if drug companies were guaranteed that their first world customers still paid full price to the patent owner. That could easily be accomplished with socialized medicine: if only the first world governments were paying for the drugs. Countering the patent monoploy with roughly equal monopsonistic power would prevent some abuses and still retain market development of drugs.

David Friedman said...

Mike Huben suggests that with socialized medicine one could guarantee that first world customers still paid full price.

That would work if there were only one country in the world. But with multiple countries, each has an incentive to insist on getting the drug at a low price. If the seller refuses, the country is in a rhetorically strong position when it announces its own policy of compulsory licensing: Who can object to taking a strong line against a selfish firm that would rather see people die than accept a reduction in its profits?

Or in other words, giving drug companies an incentive to develop new drugs faces a public good problem, where the "public" is, in Mike's world, made up of governments.

Mike Huben said...

David, your argument would work just as well to show that drug patents would not be honored, because nations have incentives to allow their citizens to buy at a low price. Yet some how drug patents are honored in first world nations.

The same sort of treaty as we have for international recognition of patents could also apply to free licensing to third world nations as long as first world nations are not cheating.

David Friedman said...

Mike writes:

"David, your argument would work just as well to show that drug patents would not be honored, because nations have incentives to allow their citizens to buy at a low price. Yet some how drug patents are honored in first world nations."

It is indeed a puzzle why first world nations that consume drugs but don't develop them honor drug patents. One possible answer is that each is afraid that if it is the first to defect, others will follow, and its loss from many countries failing to honor drug patents will be larger than its gain.

That becomes a less convincing arguments if some countries have been allowed to defect. After all, there isn't a sharp line between rich countries and poor countries. Brazil, for example, could decide it was on the poor side of the line for this purpose. Then Greece and Portugal. Then ... .

Think of it as a Schelling point problem.

Mike Huben said...

I don't understand your quick recourse to Schelling points when there are plenty of alternative explanations, such as relatiation for violation of treaty. Perhaps it's not such a puzzle.

In any event, you haven't given much reason why my suggestion would not work if you can't even explain why international patent protection works.

Anonymous said...

Weren't 3rd world countries threatened with WTO sanctions if they didn't comply with the patents or something ?