Comportamento reprodutivo: uma análise a partir do grupo ocupacional das mulheres

Abstract

The comprehension of mechanisms that caused the substantial decline of total fertility rates (TFR) became a challenge for population studies. Ahead of this new reality, the study of fertility has demanded from social sciences in general, and from demography in particular, a series of explanations on the causes and consequences of this phenomenon. These explanations require data that allow investigating the reproductive behavior through life-time information and new approaches that associate the reproductive behavior with aspects seldom explored by the population analyses, such as the occupation of the woman. Ahead of such demands, our objective was to analyze how reproductive behavior (taking into consideration the childbearing age, the number of children, the birth spacing, and the hazard of having x+1 child) can be associated with the occupation of the woman. Our hypotheses assume that specific types of occupation can generate different strategies regarding fertility behavior, which would result in differentials in the distinct fertility measures. We used data from the 2000 Brazilian census (IBGE). We employed data from women between the ages of 30 and 45 years, whose birth history were reconstructed (about 70% of women in our sample). In the statistical analysis, we used the proportional hazards model (Cox regression models). When applying this model, we introduced a length perspective on fertility data, producing information on the risk of occurrence from a birth of order x+1. The main results showed that the type of occupation can affect the female reproductive behavior. The descriptive analysis allowed us to infer that the more qualified the occupation, the lower the parity and the TFR. On the other hand, the greater will be the delay of maternity. The statistical analysis showed the risk of parity progression is related to the occupation of the woman. This evidence is clear in models where only the occupation variable is present. As it was expected, when controlling our model for other covariates, we observed a reduction of the variability in the risks of having a child of order x+1, among occupational groups. It is important to highlight that the education variable explains a substantial part of this reduction. Nevertheless, occupation still affects the risk of parity progression, since the more qualified the occupation, the lower the risk of having a child of order x+1. In both analyses we identified that in high-skilled occupations women control fertility more intensely. Such results suggest that among these groups there are a higher incompatibility between the professional and family lives, and higher opportunity costs of childbearing.