Malaria is a devastating disease. The mosquito-borne parasite that causes it has been eradicated in North America and Europe through a combination of swamp drainage and DDT.

But slow economic development elsewhere in the world, plus a steadily growing resistance to insecticides and to drugs used to treat the disease, means that malaria is still a widespread killer in Latin America, Africa, the Middle East and South Asia. 

A million children are expected to die from it next year in Africa alone. Millions survive, but many with liver injuries or, especially among the young, brain damage.

More people, not fewer, are suffering from malaria worldwide than thirty years ago, perhaps because global warming is extending the habitat of the anopheles mosquito.  More people are alive to be threatened too, of course.

Last week the British government announced plans to purchase 200 million to 300 million doses of malaria vaccine.  

The interesting thing about this transaction is that there is no vaccine for malaria -- not yet, not officially anyway.  Some 80 candidates are being tested around the world, 15 or so of special interest to development authorities.

Indeed, it was only in October that scientists and physicians announced their most successful test to date. A vaccine given to thousands of children in Madagascar was reported to have protected a third of them altogether, and spared another third the more serious consequences of the disease. The control group -- those not inoculated -- were carefully monitored and treated aggressively at the first signs of infection.

What the British did last week -- in concert with several non-governmental organizations -- was to offer a prize, in effect, to a successful developer.

To be sure, the malaria case is not like the example usually given -- the Longitude Prize, £20,000 offered by the British parliament early in the 18th century to the first person to reliably measure longitude within a half degree.

Then, it was expected that better star maps and more ingenious methods would be required.  Few understood that a sufficiently rugged and accurate sea-going clock would solve the problem -- fewer still had any idea how to make one.

Today, most of vaccine science is in place (though doctors have yet to conquer a complex, multi-cellular parasite like the one that causes malaria). There are abundant new tools and techniques, and the vaccine in question is in its final stages of development. 

Instead, the impediments have more to do with risks and rates of return.

Vaccines are the medicine of choice against widespread diseases, such as HIV/AIDS, malaria and tuberculosis, the three top killer diseases in the world today. The idea is to use a little bit of killed or weakened pathogen to cause bodies to develop natural immunity, similar to the vaccines that have all but eliminated smallpox, polio and a host of other diseases.

In principle, a single inoculation and occasional booster shots thereafter may be sufficient for most people to keep a targeted disease at bay.

Precisely for that reason, however, few if any pharmaceutical companies are willing to pay the huge up-front research, development and testing costs in pursuit of a successful vaccine, for which there may or may not be a market -- and never mind the liability in the inevitable few cases where the inoculation goes awry.

Hence governments have recognized for years their responsibility to step in where market mechanisms fail. Government laboratories and cost-plus contracts to private vaccine developers are nothing new.

Only recently, however, have governments (and the non-governmental organizations that advise them) begun to concentrate on the kind of mechanism that gave rise to the naval chronometer -- "pull" methods instead of more familiar "push" techniques, which include grants to academic investigators, R&D tax credits, public equity investments and government labs.

(For a lucid account of the new ideas, see Strong Medicine: Creating Incentives for Pharmaceutical Research on Neglected Diseases by Harvard University economist Michael Kremer and Rachel Glennerster.  Kremer is the man who, probably more than any other, is responsible for formulating the distinctions between push and pull approaches.

By contracting in advance to buy large numbers of doses that meet certain standards of safety and efficacy at a particular price, such methods insure that if vaccines are developed, they actually reach the people who need them. Pull methods pay for results, so that if no vaccine is developed, no public funds are spent.

That's important, because, left to their own devices, big pharmaceutical manufacturers are only too happy to pursue their work on what amounts to a cost-plus basis, pursuing long-shots and wills-o'-the-wisp longer than they would if they were hazarding their own funds.

It's one more example of why, more than ever, governments today need talented and sophisticated regulators. Technology policy has become as important as monetary policy -- in some respects, maybe more so.

The British government's largesse is surely not unrelated to the fact that the manufacturer of the malaria vaccine it promised to buy is GlaxoSmithKline Biologicals, a large British pharmaceutical concern with extensive entries in other, more lucrative markets.  

The government's investment will do more than establish a market for a commodity whose beneficiaries are too poor to fund for themselves.  It will also defray some part of the fixed costs of maintaining GlaxoSmithKline's position on the frontiers of pharmaceutical research.

But that's quite all right. If governments are going to favor the locals with subsidies, how much better that the recipients should be biotech labs and pharmaceutical firms rather than, as it has been the case for more than the last 150 years, weapons makers --  that the competition should among merchants of life rather than merchants of death?