A Stochastic Dynamic Analysis of Parental Sex Preferences and Fertility

Abstract

The author has developed a stochastic dynamic model of fertility to examine assumptions which form the basis of often used econometric tests for parental sex preferences. These assumptions included the use of fertility data to test for sex preferences and to test for son preference based on the effect of the number of males on fertility. The Heckman and Willis stochastic model of reproduction was used to formulate this model. This new model allowed some tractable and testable predictions. It was able to predict income and wage effects without sex preferences. Further it can accurately describe life cycle dynamic fertility behavior. The model demonstrated the existence of a quantity quality interaction which influence parents decisions on fertility. For example higher quality measured by the effective number of children substituted for a higher quantity (number of children). The model proved that sex preferences can indeed be tested but it could not differentiate empirically between son preference and daughter preference from fertility data. Perhaps data on intrafamily allocation of resources as used by Rosenzweig and Schultz; Behrman Pollak and Taubman; and Behrman would better be able to test for son versus daughter preference since allocation is based on existing children. The results of this newly formulated stochastic dynamic model of fertility challenge the validity of traditional econometric tests for son preference.