Last week Sen. Hillary Rodham Clinton and Rep. Nita Lowey, her fellow New York Democrat, unveiled a bill that would make universal education in the poor world a U.S. government objective. Yesterday James Wolfensohn, the World Bank's president, spoke extensively about this goal in his speech at the bank's annual meeting. Both the bill and the speech are pushing the right way, since schooling offers an escape from poverty and since the dearth of schools financed by Western aid creates a vacuum to be filled by hate-preaching madrassas. But these calls for universal education, like most grand development targets, need to be treated with caution. If people believe that development advocates are promising to get all children into school by some certain date, they may turn against aid when these targets prove unattainable.

Why is universal education unattainable in the next decade or so? The advocates of foreign assistance tend to stress the lack of aid resources. "Three point six billion dollars in additional aid flows is needed each year, for the next seven to eight years, to ensure that all children complete primary school," Wolfensohn said yesterday. "That comes to $1,200 per class of 40 children to pay for the teacher, books and classroom." But the truth about education is more complex. To get children into school, you need to address not just the supply of schools but also the demand for them.

What drives this demand? You need a healthy economy that rewards education financially: If there's a war, or if the economy is too dysfunctional to create jobs, parents won't see much benefit in pressing their children to attend school. The circumstances of households influence demand as well. Poor parents may be unable to forgo child labor. Uneducated parents may have difficulty inculcating the curiosity and habits of learning that children need in order to absorb education. Extra foreign assistance can probably increase the number of teachers and school rooms -- though there are plenty of countries where even this has proved tough. But foreign assistance can change these demand factors only slowly and gradually.

What's more, demand seems to matter more than the supply of schooling. Michael Clemens of the Center for Global Development has reviewed a pile of economic literature on this question, and he finds that the odds of a child attending school depend more on parental income and education levels than on the proximity of a school or whether it is free. One study -- conducted, ironically, at the World Bank -- considers the effect of halving the distance to school for children in 21 countries. In rural Chad, that huge investment in school infrastructure would boost enrollment by less than 5 percent. In the other 20 countries, the effect would be even smaller.

So more aid for education, while desirable, cannot bring about 100 percent enrollment in the next seven or eight years, as Wolfensohn implied; nor can it achieve that by 2015, the target laid down by the Millennium Development Goals, which were embraced by all the world at a U.N. summit in 2000. Rather, universal education will arrive gradually, along with economic growth and the slow accumulation of a stock of educated adults -- and there's surprisingly little that policymakers can do to accelerate this progress. The experience of the past half-century, as analyzed by Clemens, shows that countries pursuing a wide variety of education policies have progressed toward universal education at roughly the same rate -- a finding consistent with the point that demand for education matters more than supply does.

None of this means that aid is useless, nor that policy is unimportant. To start out on the road toward higher school enrollment, countries need to get their policies right -- and they can be encouraged to do so by enlightened aid donors. A large pile of economic literature -- as large, roughly, as the pile on education that Clemens reviews -- finds that aid boosts growth in countries that are decently governed; and faster growth means higher family incomes, which in turn boost school enrollment. But aid, however constructive, cannot achieve utopian targets like universal education except in the very long term. And targets that wind up not being met tend to fuel the cynicism of aid critics.