Description

Abstract: The Heckman two-stage model for correcting selection bias was applied to analyze the determinants of fertility desire using DHS II (Demographic and Health Survey) data from Senegal. The study concludes that significant differences exist between women that report numeric answers and those who report non-numeric answers to a question relative to fertility desire. Therefore, excluding non-numeric answers from regression models produces biased parameter estimates. Furthermore, this study presents evidence that contraceptive methods in Senegal are primarily used to control undesired births as rural and urban residents share the desire for a large family size.