Essays in development economics
Author(s)Carvalho, Irineu E. (Irineu Evangelista), 1971-
Massachusetts Institute of Technology. Dept. of Economics.
Abhijit Banerjee and Dora Luisa Costa.
MetadataShow full item record
In 1991, a reform was passed changing the rules for social security for rural workers in Brazil. The reform consisted of a reduction in the minimum eligibility age for old-age benefits, an extension of benefit eligibility to workers who are not the heads of their households, and an increase in the minimum value for benefits. As a consequence, elderly rural workers and their households found a substantive increase in their non-labor incomes. Because old-age benefits for rural workers are not means or retirement tested, this reform may be a very useful natural experiment for studying pure income effects. I use information from surveys administered before and after the reform is implemented to identify the effects of the reform on the actual receipt of benefits by the elderly. The first chapter studies the labor supply of elderly rural males in response to the reform. I find that elasticities of labor supply with respect to benefits generosity among rural men age 60 to 64 are greater than those estimated from developed country data. I find that benefit take-up rates are greatest among the better educated, but that least-schooled workers have the largest elasticities of labor supply. I also find that husbands respond to wives' benefits by increasing their labor supply, perhaps because of bargaining considerations within the household. Last but not the least, I find that anticipated benefits do not affect the labor supply of workers close to the minimum eligibility age. The second chapter studies the choice of living arrangements of unmarried elderly females, that is never married, divorced or widowed females. The main finding is that living arrangements are responsive to benefits income: Brazilian rural elderly females value their privacy and independence, choosing not to coreside with their adult children if they can afford to do so. This result suggests that substituting the extended family for formal transfer programs by means of severe filial responsibility laws and scaling back of social security may be a very costly measure for the elderly in Brazil. Because the estimates of this paper are based on the behavioral response of unmarried elderly females in the rural areas, one may reasonably argue that those effects are underestimates of the effects for the whole sample of elderly, males and females, married or unmarried, residing in rural or urban areas. The final chapter studies the effects of increases in non-labor income at the household level on children's outcomes, particularly labor participation and school enrollment. In this chapter I study the impact of this increase in non-labor income on children of ages 10-14 living in the same household as old-age beneficiaries. Counterfactual analysis based on reduced form estimates implies that little less than 20% of the gap between 100% enrollment and counterfactual enrollment rates was closed for girls living with at least an elderly who benefited from the reform, with a smaller effect for boys. Labor force participation of boys also seem to have been effected by the reform, with a reduction in participation rate around one-tenth of counterfactual participation rates. Those results may be underestimates of the effects of overall income growth because economy-wide increases in income are likely to be associated with shifts in social norms and attitudes towards children's labor participation and schooling.
Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Economics, 2000.Includes bibliographical references.
DepartmentMassachusetts Institute of Technology. Dept. of Economics.
Massachusetts Institute of Technology