Regression also includes dummy variables to indicate missing values for mother's schooling, father's schooling, and household shock. While women who have teen pregnancies end up with lower grade attainment than women who do not, our analysis of schooling trajectories suggests that caution must be used in assigning a causal impact to teen pregnancy. Photos are used for illustrative purposes. Much has been done to reduce mother-to-child transmission of HIV, but much more needs to be done to reduce the gender inequality and violence that women and young girls at risk of HIV often face. We also find that negative household income shocks increase the probability of sexual debut, pregnancy, and school dropout for girls. Boys tend to become sexually active at a somewhat earlier age than girls among Africans and coloureds, although this is not true for whites.
From a policy perspective, one interpretation of our results is that neither poor quality secondary schools nor pregnancy are the primary reason for school dropout in South Africa.
Sexual Behavior, Pregnancy, and Schooling among Young People in Urban South Africa
The pattern for coloured women is quite different, with much smaller proportions of women reporting that they have both had a pregnancy and are still in school. The study also found there was potential for integrated services to improve cost efficiencies at facilities but this often went unrealised. South Africa differs from this pattern, with sexual initiation occurring primarily outside marriage. We control for the number of months between interviews. We do not have a great deal of faith in the boys' reports of having made a girl pregnant, but we include males in the regressions for completeness.