Editor's Summary

This third issue of Economía contains papers presented at the Panel Meeting held on 21 April 2001 in Cambridge, Massachusetts. The Center for International Development and the David Rockefeller Cen-

ter for Latin American Studies at Harvard cosponsored the meeting, and they will do so again in the spring of 2002. We are grateful to them for generous financial and logistical support.

Latin America has long been a region of great income disparities. Not all kinds of inequality are created equal, however. An unequal country in which the same people (or their children and grandchildren) are always at the highest income levels is very different from one in which mobility allows the talented and skilled to rise to the top. Which kind of inequality is Latin America's? Jere Behrman, Alejandro Gaviria, and Miguel Székely, tackle this extremely important but understudied question, which has deep implications for the way we think about development in the region. Measures of mobility may well say more about fairness and equal opportunity than do standard estimates of income disparity.

Using a hundred household surveys spanning two decades and twenty countries, they reach somewhat dispiriting conclusions. Unsurprisingly, intergenerational mobility is much higher in the United States than in Latin America. In Mexico, for instance, the probability of having a white-collar job is 3.5 times higher if one's father also had such a job; the corresponding figure for the United States is 1.5 times. There are also sizable differences in mobility across countries in the region, with Brazil and some Central American countries among those showing the least mobility.

Not all the news is bad. First, mobility (measured as the inverse of the correlation coefficient of siblings' educational attainment) need not be constant over time. It increased substantially in Latin America from the mid-1980s to the mid 1990s, but it has recently declined. Second, mobility is not God given: greater expenditures in education are associated with substantially lower values of the correlation index, which suggests that public policies can increase intergenerational mobility.

A prominent public policy targeted at a related goal--reducing the transmission of poverty between parents and children--is embodied in Mexico's Education, Health and Nutrition Program, known by its Spanish acronym, PROGRESA. The program makes cash transfers, which are paid directly to mothers, that are contingent on regular family clinic attendance and children's school attendance. PROGRESA is being closely studied because it is both highly targeted and large (it now covers 2.6 million families in extreme poverty). Some analysts have hailed it as a model for antipoverty programs in developing countries.

Emmanuel Skoufias and Susan Parker analyze one concrete aspect of the program's impact: does it reduce child labor and increase school attendance? If so, among whom and by how much? PROGRESA was designed in a quasi-experimental fashion, involving communities that receive benefits earlier (and thus serve as treatment groups) and others that join the program later (the control group). Inferences can therefore be made with greater confidence than is typically the case in program evaluation exercises.

Empirical studies from other countries suggest that unconditional cash transfers have a marginal effect at best on school enrollment or child labor. The contingency element of PROGRESA seems to be important, since it introduces income effects (after the transfer the household is richer) and substitution effects (the relative price of schooling falls) that point in the same direction and can reinforce each other, causing an increase in the accumulation of human capital. The empirical results are encouraging: the program has been associated with significant increases in school attendance for boys and girls and reduced market work. Since the fall in the incidence of work is smaller than the increase in schooling, the adjustment seems to be coming mostly through leisure time, as well as through domestic work primarily in the case of girls.

The authors themselves stress that this apparent success of the program raises a number of questions. What about the work incentives of adults, which could decline...