This paper investigates labour supply of the wives of the heads of households in Mexico City, with a focus on the impact of family structure. A static neoclassical structural model is used. We assume that each woman chooses her labour supply and corresponding income so that her utility is maximized, conditional upon her husband’s labour supply and earnings. We use a direct translog specification, and include family composition variables as taste shifters. Also taken into account are fixed costs of working, nonlinear taxes, unobserved preference variation, prediction errors in wages of nonworkers, and potential endogeneity of wages. The models are estimated by smooth simulated maximum likelihood using data from Mexico’s Urban Employment Survey drawn in 1992. We find income elasticities of labour supply of about -0.35, and wage elasticities of about 0.5. The latter is substantially overestimated if wage rate endogeneity is not taken into account. The results are robust with respect to other specification choices. We find that the impact of family structure variables on participation is different from that on hours worked, so that their total effect is ambiguous.
|Place of Publication||Tilburg|
|Number of pages||32|
|Publication status||Published - 1997|
|Name||CentER Discussion Paper|
- Female labour supply
- family structure
- discrete regression