A strange reversal is underway in the development business.

A young blond Dane, who shows up at business meetings in a black T-shirt and sneakers, is convening a conclave of high-powered economists today in Copenhagen. His goal is to determine, by means of mathematical regressions, the proper development priorities for poor countries. You've heard of the Washington Consensus, the set of economic prescriptions imposed on poor countries by the International Monetary Fund and the World Bank. Now comes the Copenhagen Consensus.

Meanwhile, the ex-Wall Streeter who runs the World Bank is off in the opposite direction. Last week he showed up at Rome's Circus Maximus for a concert featuring the garbage-can clangers from the Broadway show "Stomp"; he was introduced by the actress Angelina Jolie, which was a pretty good reason to go. Tomorrow he convenes a grand development conference in Shanghai. It will eschew prescriptions, and there won't be much math. It will be Copenhagen inverted.

The young blond Dane, Bjorn Lomborg, is hitherto known for disputing pessimistic environmental projections in a book called "The Skeptical Environmentalist." His latest project is even more presumptuous. Lomborg has divided the problems of poor countries into 10 categories -- disease, financial instability, conflict and so on -- and commissioned a paper from a top economist on each. The economists are supposed to consider the cost-effectiveness of three interventions in their assigned categories. Within the health category, for example, donors could prioritize malaria or HIV or a general scaling up of health services.

In Copenhagen this week, an international committee of economic superstars, featuring four Nobel laureates, will consider all the development interventions for which Lomborg has collected cost-benefit estimates. The panel will cross-examine the 10 authors about their statistical techniques, then issue a league table of the 30-plus interventions, ranked scientifically according to value for money. Aid donors will be expected to shift their dollars to the top ones -- or face questions from the Copenhagen crew about why they are wasting money.

The first objections to this exercise are not exactly crushing. Yes, cost-benefit analysis is an inexact science, involving heroic projections far into the future. But it's better to attempt such estimates than to spend aid dollars blindly. Yes, donors waste money in ways that are already known to be dumb -- they tie it to purchases from their own companies, for example -- so it's doubtful whether new knowledge will improve their behavior. But the best shot at reducing waste is surely to expose it.

The real problem with the Copenhagen Consensus is that aid investment, like most investment, is subject to diminishing returns. Imagine that aid donors really did respond to the Lomborg league table and piled all their money into the top three interventions on his list. Pretty soon, those problems might be absorbing as much money as they usefully could -- and value for money in other cash-starved areas might quickly become greater. We've seen something like this happen with the recent explosion in AIDS funding. Two or three years back, the lack of money to fight HIV was scandalous. Today most decent AIDS projects are awash in cash, and other causes such as vitamin deficiency and malaria are underfunded.

Because of the law of diminishing returns, development priorities will shift at the global level. But they will also vary by country. Perhaps Brazil's most urgent need right now is slum-upgrading projects, whereas Uganda needs transport and Indonesia needs to fight corruption; don't forget that, if all three countries pursue these priorities diligently for a few years, the list will soon look different. A static global list of priorities won't capture that variety. Which leads to the conclusion that despondent development theorists have reached before: The search for silver bullets is doomed. No amount of brilliant top-down analysis can solve the problems of world poverty.

What can solve them? The answer is pragmatic leaders who wrestle with development challenges as they come up, paying less attention to the theoretically perfect solution than to effective implementation. So long as a country gets some basic things right -- so long as it targets exports, protects private property and avoids ruinous inflation -- the precise policy it chooses matters less than its ability to put it in place and then to adapt as circumstance demands, so that there's constant learning by doing. Places as diverse as China, Vietnam, India and Uganda, each pursuing somewhat different strategies, have made marvelous progress. Thanks to their lead, the number of people living on less than $1 a day has declined by 375 million since 1981 -- without any Copenhagen league table.

The conference that opens tomorrow in Shanghai, convened by World Bank President James Wolfensohn, will celebrate this pragmatism. Rather than treating development as a quasi-science, it will present it as an art, in which the heroes are the resourceful managers who implement programs. In place of top-down analysis by first-world econometricians, it will feature bottom-up case studies presented by third-world practitioners. Just about every speaker in the two-day program will come from the developing world; the delegates from the rich North will be there to listen to them. It will be hard to turn the conference into newspaper headlines. But it will have captured the real spirit of development.