World governments might be more successful in removing nearly 100 million children from the labor market by working to increase adult wages and employment rates rather than pursuing legislative action against child labor, which could be effective only in certain countries, say two Cornell University economists.

These findings from Kaushik Basu, a professor of economics at Cornell, and Pham Hoang Van, a Cornell doctoral student in economics, appear in their paper "The Economics of Child Labor" in the June issue of The American Economic Review.

Large numbers of children under age 15 often work long hours, sometimes in dangerous industrial environments, in many Asian, African and South American countries. In their journal article, Basu, the Carl Marks Professor of International Studies, and Van argue that most parents send their children to work as a response to poverty. Consequently, higher wages for, and lower unemployment among, adults will lead most parents to keep their children from working, they argue.

"If we agree that sending children out to work is an act of desperation on the part of the parents, it seems reasonable to expect that parents would not send their children to work if their own wages were higher or employment prospects better," the authors say in their paper.

Governments should occupy themselves with improving the adult labor market, since its condition can dictate the size of the child labor market, Basu and Van say. Trying to solve the child labor problem first, by banning it outright, could do more harm than good in the poorest countries, such as Ethiopia, they say. Because there is little reason to think the adult labor market can be speedily improved in these countries, they argue, barring children from work would plunge many families into starvation.

But in some countries, such as India and China, where the standard of living is higher -- though still low overall -- a government ban on child labor could act as a corrective factor, keeping the country's economy stable while actually raising the standard of living, the Cornell researchers say.

Here's how their model works: Imagine a stable economy in which wages are very low. To escape hunger and deprivation, parents find work for their children. But if the government decrees that children cannot work, the demand for adult labor should rise, assuming that adults can do the work the children were doing. With this increased demand for labor comes an increase in wages, which may raise family incomes to a level at which there is no need to send children to work.

Basu and Van warn that economists must study a particular country's economy thoroughly before advising this course of action. If there are any indications that the process they describe will not apply to a certain country, they recommend that no ban on child labor be imposed and that economic development efforts be targeted at the adult labor market.

The researchers say that current political efforts to curb child labor are driven by emotions rather than the results of scholarly investigation and that their contribution helps fill the void of serious studies of child labor.

"This is a field of study where prescription has outstripped analysis by a wide margin," they say in their paper.

And that can be most damaging to the children who work, they point out. For example, Basu and Van refer to a United States senator's proposal last year to make illegal the importing of goods produced with child labor. Shortly after the senator suggested the law, a Bangladesh manufacturers' association fired its child employees, who then ended up in more debilitating jobs and even prostitution. It is believed, Basu and Van say, that the firings resulted from the association's fear of a U.S. ban on their products.

To avoid this kind of outcome, which harms children instead of helping them, careful study of child labor on a country-by-country basis must precede policy decisions, the researchers say.