Fertility behavior and labor force participation: a model of lexicographic choice.

Evidence exists that a smaller family size is usually associated with female employment and that fertility rises with family income and the wife's education at relatively low levels of income and education. Only at higher levels is there the generally expected relationship that fertility declines with more education or income. Due to the fact that a woman's labor force participation and her fertility are aspects of behavior of the same person (or couple), they should be explained by a model of choice. Such a model is presented, and empirical evidence is cited. In particular, the model allows for a fertility decline even before a decline in mortality during the demographic transition. The model of choice involves threshold values of education and income, such that the marginal effects of these variables on fertility and labor supply are qualitatively different below and above the threshold. The model is in conformity with cross-section regressions using Philippine data and appears to explain why various studies give positive, zero, or negative regression coefficients relating fertility to education and income when standard linear regression specifications are used. Such results would depend on the proportions of families falling below and above the thresholds in the sample of observations. The model also implies that the fertility effects of a child mortality decline on those proportions, meaning that one could have lower mortality without affecting fertility levels. From a policy perspective, the broad implications of the model are distrubing. Development that raises very low income and education levels would increase fertility and so would a more egalitarian distribution of the same low aggregate income. It is necessary to shift the underlying functions so that the thresholds become as low as possible, but general economic development may be too slow for this purpose.