Women and the Economics of Family Migration

Abstract

N OTWITHSTANDING this early statement by Ravenstein (1885, p. 196), the separate study of geographic mobility among women has been virtually ignored by students of migration.' The reason is obvious: women are assumed to migrate because their husbands do.2 While it is undoubtedly true that most migration involves family units (the migration of husband and wife occurring jointly), the possibility that the wife's welfare is considered in the family's decision to migrate should not be ruled out. It is at least desirable to test the hypothesis that the wife's employment is considered in the migration decision and to examine the effect of that decision on women's earnings. In this paper an economic model is developed to explain the family's decision to migrate and the effect of migration on the labor market earnings of men and women. It is based on the tenet that family utility, defined operationally as the husband's and wife's labor market earnings and leisure, is maximized. The model suggests that the wife's labor market involvement is a significant consideration in a (husband-wife) family's decision to migrate. The data from the National Longitudinal Surveys (NLS) are well suited for empirical testing of this model.3 The surveys provide the opportunity to examine the change in labor market earnings of families and individuals over a five-year period. Availability of data on migratory status as well as on other personal characteristics of women and their families permits the direct testing of the model. In section I a family utility maximization model is used to derive implications with regard to the probability of migration by the family and the effect of migration on individual and family earnings. These implications are tested in section II using multiple regression analysis and the NLS data for white women who were 35 to 49 years of age in 1972. The implications of the empirical estimates for the economic welfare of women and for interpreting the observed earnings distribution are discussed in section III.