Little by little, the world is coming around to two self-evidently good proposals to improve global health. But there's a third, equally great proposal to which nobody pays attention. The contrast shows how corporate preferences, including largely irrational corporate preferences, distort policy toward the poor world.

Oddly, the two proposals with momentum are the ones that cost a lot. The first is designed to stimulate production of vaccines for diseases such as polio, yellow fever and hepatitis B. These vaccines have been around for decades, but poor countries can't afford them, so most manufacturers have quit. When aid people scraped together money for vaccines in the 1990s, they had trouble finding willing suppliers.

The solution has been to create a dependable source of demand for these vaccines: to simulate the "pull mechanism" created by consumers in the rich world. By pouring $1.5 billion into a vaccine purchase fund and persuading rich governments to kick in a bit, the Bill and Melinda Gates Foundation has created an incentive for manufacturers to re-enter this business. (Full disclosure: Melinda Gates is on the board of The Washington Post Co.) Next month's Group of Eight summit may boost this purchase fund's prospects further. The British government is pushing an idea to enlarge it with money borrowed on the capital markets that will be repaid with future aid flows.

The second proposal is about drugs that don't exist yet -- for example, a malaria vaccine. Again, there's no commercial incentive to invent these, since consumers in poor countries can't afford to pay much. Practically all of the world's drug development spending is targeted at the medical concerns of rich people. Hence the depressing fact that, of the 1,233 drugs licensed worldwide between 1975 and 1997, only 13 were for tropical diseases.

To fix this problem, the Gates Foundation has put its muscle behind an idea from Harvard's Michael Kremer of Harvard University and Rachel Glennerster of Massachusetts Institute of Technology: Rich countries should make an advance commitment to spend, say, $3 billion on 200 million doses of a malaria vaccine if one is invented. This commitment solves the incentive problem and could be structured as a binding contract. The Gates people, working through the Center for Global Development in Washington, have brought in a law firm to draft appropriate language. Again, this idea has political momentum. Next month's G-8 summit may get behind the Kremer-Glennerster idea.

So there's an appetite to spend taxpayers' money on buying existing vaccines and on a "pull mechanism" for new ones. But there's a third challenge in this medical battlefield: How to make drugs that have been invented for rich countries available in the poor world. Contrary to popular myth, diseases such as cancer and heart disease are big killers in developing countries; in India, heart disease causes more than a quarter of all deaths. And yet, although pills for heart disease are swallowed by millions of Americans daily, these drugs are out of reach of the world's poor.

About four years ago, when the Kremer-Glennerster purchase-commitment idea was germinating, Berkeley's Jean Lanjouw advanced a solution to the heart-pill problem. The idea would cost nothing: It merely involves drug companies giving up patent protection for heart pills and similar medicines in the poor world. Since poor countries buy almost none of these medicines anyway, giving up patent rights in those markets doesn't hurt the drug firms. But it would mean that cheap generic versions of these medicines could be distributed to poor consumers.

As well as costing nothing, the Lanjouw proposal would be simple to implement. It would require no multilateral negotiation, since it could be done with a modest fix to U.S. patent law. The fix would say that, as a condition of receiving patent protection in the U.S. market, the inventor of a drug must renounce patent rights in countries with a per capita income of less than, say, $1,000 per year.

Who could possibly object to this? The drug companies apparently fear that, if Indians are allowed to get cheap heart pills, these would find their way back onto the U.S. market. But the drug companies already accept the principle that AIDS drugs should cost less in Botswana than in Boston -- and there aren't too many reports of contraband African AIDS medicines flooding the U.S. market. Indeed, for all the fuss about seniors getting cheap drugs from Canada, these purchases represent a fraction of the U.S. drug market. Does anyone really think that Americans are going to buy masses of medicine from distant and possibly unsafe suppliers based in Mombassa or Mumbai?

Of course not. But four years after its appearance created a small flurry of excitement, the Lanjouw proposal has been forgotten. The reason is that, irrationally, drug companies want to close down all ideas involving weaker intellectual property rights, even if the rights they would forgo have no business value. The Gates Foundation, whose money derives from Microsoft's intellectual property, has no interest in challenging this world view. Hence the perverse outcome: An idea that costs nothing has less traction than two ideas that cost billions.