cantly lower wages than those who gave birth after age 20. The issue of the potential endogeneity of adolescent childbearing and wages was not addressed, however; in addition, the analysis failed to control for the fact that women who work in this setting are a selective group. Buvinic also noted that adolescent childbearing in Mexico and Guatemala (for which she analyzed retrospective data on 850 women collected in 1967, 1974, and 1988) was significantly associated with poverty status measured by an index of housing quality in the two countries and socioeconomic status in Mexico. Although economic status as a child was controlled, it is not clear whether economic status immediately prior to childbirth was accounted for.
It is interesting that in none of the four countries Buvinic studied did adolescent childbearing affect marital status. Those who gave birth early were as likely to marry in subsequent years, although in Chile (where this issue was explored) those who gave birth in their 20s were significantly more likely to be living with the father of the child five years after the birth.
Another study in Nigeria compares the life course of adult Yoruba women who had a birth in adolescence regardless of marital status with those who delayed childbearing into their 20s (Omololu, 1994). Compared with women who had their first child in their 20s, women who had their first birth during their adolescent years were more likely to have more children by the time of the survey, more likely to be second wives, and more likely to receive assistance from parents. The author argues that in many cases, partners and families of orientation were unsupportive, contributing to the higher levels of poverty among adolescent mothers compared with others. Social adjustment during pregnancy and in the first year of birth in southwest Nigeria was characterized by an end to schooling, difficult parental relations, and peer difficulties for the adolescent mother.
A prospective study in Forteleza, Brazil of young women experiencing early pregnancy compares those who sought prenatal care from a clinic with those who sought postabortion care. Preliminary findings suggest that while sociodemographic characteristics of the two groups were relatively similar, adolescents who continued the pregnancy were more likely than those who opted for abortion to have dropped out of school as a result of the pregnancy. Yet those who underwent induced abortion reported significantly lower levels of self-esteem, a difference attributed in part to the circumstances surrounding the unwanted pregnancy and abortion seeking in this setting, in which abortion is legally restricted. While these differences continued to be manifested at one and five years following the pregnancy, they had narrowed considerably by year five; about three-fifths of young women in both groups reported high self-esteem, and about one-third were still students (Bailey and Bruno, 2001).
While the findings from these studies are suggestive, they are, as noted above, far from definitive. In the absence of study designs that control for