• Text smaller
  • Text bigger

During the past 50 years, malaria, tuberculosis and HIV have killed
several times as many people as have all wars. Five million people
succumb to these diseases each year, mostly in developing countries. The
breakdown: Malaria kills 1.1 million people annually, tuberculosis 1.9
million and AIDS 2.3 million. Each year, an additional 5.8 million
people, 70 percent of whom are sub-Saharan Africans, are newly infected.

What options are there to halt this epidemic? Most might say more
government-funded research is the answer. But Rachel Glennerster, a
staff member at the International Monetary Fund, and Michael Kremer, a
Harvard University economics professor, say no in their article, “A
Better Way to Spur Medical Research and Development,” in the Summer 2000
Regulation magazine, a publication of the Washington, D.C.-based Cato
Institute.

Government-funded research has a record of mixed results and quite a
few spectacular disasters such as: the Carter administration’s synthetic
fuels program and the Clinch River Breeder Reactor. In 1980, the U.S.
Agency for International Development (USAID) committed over $60 million
to promote the development of a malaria vaccine. In 1984, USAID
announced a “major breakthrough in the development of a vaccine against
the most deadly form of malaria in human beings. The vaccine should be
ready for use around the world … within five years.” Fifteen years
later, the world’s still waiting; The USAID program was a monumental
failure.

Government-funded research produces the wrong incentives.
Government-funded recipients have incentives to be overly optimistic;
that’s how they get the money. Government project directors have
incentives to fund unpromising research — after all it’s not their
money. Recipients of government-funded research get paid before
delivering a product. As such, they may be tempted to divert resources
away from the contracted research toward activities that promote their
personal careers, such as publishing professional articles.

Glennerster and Kremer recommend several alternatives to
government-directed research of vaccines for AIDS, malaria and
tuberculosis. A particularly fascinating alternative is to offer prizes
for, say, a vaccine against AIDS.

Prizes for research have produced results. Napoleon needed better
ways to feed his troops; he established a prize that led to the
development of canning. In 1959, British industrialist Henry Kremer
offered a 50,000-pound prize for the first substantial flight of a
human-powered airplane. In 1977, Paul MacCready’s Gossamer Condor won
the prize. The next year, MacCready won a 100,000-pound prize by flying
the Gossamer Albatross across the English Channel entirely under human
power. More recently, several electric utilities established a
$30-million prize for the most energy-efficient refrigerator. Whirlpool
won with a line of refrigerators that were 70 percent more efficient
than their competitors.

Why prizes for research? Prizes provide strong incentives. Unlike
government-funded research, researchers get money only if their research
succeeds. Glennerster and Kremer suggest that prizes might be the way to
go in finding safe and effective vaccines for malaria, tuberculosis and
AIDS. They propose that once a vaccine is developed, philanthropic
foundations or the government might play a role in its distribution to
people in poverty-stricken nations. Patents might be bought out with
lump-sum payments. Also, the government or a foundation might provide
incentives by committing to the purchase of a certain quantity of the
vaccine at a certain price.

Glennerster and Kremer have produced a brilliant proposal for ways to
save millions of the world’s poor. What makes their proposal an
attractive alternative to government-funded research is that it pays
attention to human incentives. With government-funded research,
researchers get paid whether they deliver or not. In the case of prizes,
no delivery, no prize. One need not be a rocket scientist to figure out
which method is more likely to deliver the goods.

  • Text smaller
  • Text bigger
Note: Read our discussion guidelines before commenting.