The Determination of Fertility, Schooling, and Child Survival in Guatemala
构建了一个家庭生育、子女教育与儿童生存的静态选择模型，利用危地马拉数据估计需求函数，发现儿童生存并非完全外生，但其对生育和教育方程的影响很小。
This paper develops a static model of the demand for child services in which fertility child schooling and survival are choice variables. Each commodity is produced in the home with inputs of time goods and services. Demand functions are derived from the model and estimated using data for Guatemala. Significant coefficients on the variables in the survival equation provide some support for the hypothesis that survival is a choice variable. This approach departs from previous research in which survival is included as an exogenous variable shifting the demand for fertility. If however tastes for children are correlated with survival then including survival as an exogenous variable results in simultaneous equations bias. To determine the significance of this bias in fertility and schooling demand equations including child survival the reduced form demand equations for fertility schooling and survival are 1st estimated separately as functions of schooling wages wealth and village; the fertility and schooling equations are then reestimated including survival as an explanatory variable. The significant results obtained from the single equation estimation of survival supports the hypothesis that survival is not completely exogenous to the household in Guatemala. However the effects of parental schooling wages and village on survival are very small (although significant). When the fertility and child schooling regressions are estimated including the survival vairable the latter is significant only in the fertility equation. In addition the coefficients on the remaining exogenous variables in this equation register only slight changes in magnitude or significance when survival is included. Although mortality is partially determined by the exogenous variables in the model the impact is so small that the schooling wage and village coefficents only have a slight downward bias. Several other results are discussed including: the fathers and mothers value of time and level of schooling as they affect fertility child schooling and child survival; and the effect of village of residence on child services.