Discussion

In line with the theoretical framework adopted in this study, several factors were found to act either as hindrances or as protective factors to post-secondary education enrolment in this sample of young people in South Africa. Early sexual debut (by age 17) was negatively associated with initiation of PSE for both males and females. Other variables that had a negative effect included non-contraceptive use at first sex, parenthood, illegal substance use and studying for less than an hour a day. Higher levels of parental education (except for paternal education in the female model), urban residence and higher aspirations and analogous behaviours (studying) acted as protective factors and were positively associated with PSE initiation.

While the negative relationship between early sexual debut and educational outcomes has been well researched, there is little consensus as to why this is the case. Some studies from South Africa (Marteleto et al., Reference Marteleto, Lam and Ranchhod2008; Timæus & Moultrie, Reference Timæus and Moultrie2015), as well as internationally (Frisco, Reference Frisco2008), have suggested that the negative correlation between sexual debut and educational outcomes can be explained by the risk it confers to unplanned pregnancies/parenthood. However, a UK study by Parkes et al. (Reference Parkes, Wight, Henderson and West2010) found that while early sexual debut reduced participation in tertiary education, pregnancy did not explain all the gaps between males and females in PSE initiation. This discrepancy in the findings could be attributed to differences in the levels of education studied as well as the differing contexts. Both Timæus and Moultrie’s (Reference Timæus and Moultrie2015) and Marteleto et al.’s (2008) studies focused on high school levels while Frisco’s (Reference Frisco2008) study was conducted in the United States. There are also scholars who have argued that young people who become preoccupied with sex are less likely to concentrate in class and more likely to engage in other risky behaviours (Spriggs & Halpern, Reference Spriggs and Halpern2008; Pham et al., Reference Pham, Keenan and Han2013). This is particularly so when sexual behaviours occur in the context of romantic relationships (Clark & Mathur, Reference Clark and Mathur2012; Pham et al., Reference Pham, Keenan and Han2013). The evidence from the current study seems to mirror this prior research and demonstrates a significant negative relationship between sexual behaviour, contraceptive non-use, pregnancy and educational outcomes. Further, individuals who initiated sex in early adolescence (9–14 years) were likely to spend less time studying relative to those who initiated sex later in adolescence (Table 2). Additionally, the difference in effect size between males and females suggests that pregnancy accounted for a significant amount of the gap in PSE initiation between males and females. Despite the diverse explanations in the literature, the current study, taken together with prior research, suggests that early sexual debut is a major and consequential transition in the life course of a young person, with negative implications for first birth/parenthood and academic achievement in emerging adulthood.

It may be noted that the association between enrolling in PSE and sexual debut is not as automatic as it may at first appear (Parkes et al., Reference Parkes, Wight, Henderson and West2010). Educational attainment has been conceptualized as both a consequence of sexual risky behaviour, as well as an antecedent (Marteleto et al., Reference Marteleto, Lam and Ranchhod2008; Zuilkowski & Jukes, Reference Zuilkowski and Jukes2012). For instance, school drop-out has been seen as a causal effect of teenage pregnancy (Rosenberg et al., Reference Rosenberg, Pettifor, Miller, Thirumurthy, Emch and Afolabi2015). At the same time, a higher prevalence of pregnancy has been reported amongst girls who are out of school than those who are in school (Marteleto et al., Reference Marteleto, Lam and Ranchhod2008). Furthermore, evidence emanating from South Africa has indicated that drop-out after Grade 10 is a characteristic phenomenon of the education system (Spaull, Reference Spaull2015). While the present study found that sexual debut and pregnancy or parenthood were negatively associated with PSE, the possibility that pregnancy and sexual behaviour may have followed school-leaving cannot be ignored. However, in this study drop-out was not examined as the main outcome, yet the findings suggest that high academic aspirations are beneficial for PSE enrolment. Hence, it can be inferred that young people with high aspirations are likely to persist in school. The analysis undertaken in this study does not, however, allow for casual inferences to be made, thus future studies should prioritize this.

Perhaps there is need to explore why young people with ‘aligned ambitions’, such as spending more time on school homework, expecting to pass ‘matric’ and having a career, are more likely to have better educational outcomes, despite having have initiated sex early. This finding reinforces the notion that young people with aligned ambitions are more likely to be strategic academically (Frisco, Reference Frisco2008) and that these aspirations may extend to other domains. Spriggs and Halpern (Reference Spriggs and Halpern2008) also argued that young people with aligned ambitions are more likely to postpone sexual activity as a way of safeguarding their future goals. Table 2 illustrates that young people with expectations to pass ‘matric’, or who had career expectations, were more likely to be late initiators, although the percentage difference between early and late initiators was not very great. There is a possibility, therefore, that although sexual activity was initiated early, their personal ambitions might have influenced them to abstain later on. This is only a speculative point given that the study did not examine consistency of sexual behaviour over the period of investigation. Alternatively, the small percentage difference suggests that aspirations account for much of the variation and perhaps have a more direct effect on PSE enrolment.

In South Africa, socioeconomic status is more prominent due to the history of racial disparity that was legislated under the apartheid system (Seekings, Reference Seekings2011; Mhlauli et al., Reference Mhlauli, Salani and Mokotedi2015). These racial disparities have affected every aspect of life, leaving South Africa with immense income disparities, which rank amongst the highest in the world. In this study, both sexual debut and the initiation of PSE among young people differed significantly by race and family socioeconomic status. For instance, there was high rate of early sexual debut among Africans, and a high PSE enrolment rate among White males and the children of parents with post-secondary education. Regarding parental education, there is a consensus that scholarly behaviours are passed down within families from one generation to another (Bernadi & Requena, Reference Bernardi and Requena2010; Monserud & Elder, Reference Monserud and Elder2011). The findings of this study seem to reinforce this view as parental PSE (except for parental education in the female model) was found to be the strongest predictor of enrolment in PSE. There are many explanations as to why parental education has a positive effect on educational outcomes. First, parental education is an index of socioeconomic status, which potentially leads to better educational outcomes for children (Martin, Reference Martin2012; Lundborg et al., Reference Lundborg, Nilsson and Rooth2014; Mutodi & Ngirande, Reference Mutodi and Ngirande2014). This is because parental education can generate higher incomes, which can be used to invest in children. Put simply, more educated and richer parents can provide a better environment for their children’s educational achievement (Chevalier et al., Reference Chevalier, Harmon, O’ Sullivan and Walker2013). Additionally, educated parents might have learnt positive skills and habits during their schooling that they can transfer to their children through modelling behaviour (Martin, Reference Martin2012). Recall that Table 2 shows that children of better-educated parents had later sexual debuts. By extension, these children are at a lesser risk of sexually transmitted diseases and/or teenage parenthood. Similarly, and as Lundborg et al. (Reference Lundborg, Nilsson and Rooth2014) noted, education improves productive efficiency so the more educated parents can choose a more favourable mix of inputs in the production of their children’s human capital.

One of the assumptions made in this study, in accordance with Problem Behaviour Theory, is that quality of parenting matters for children’s outcomes (Jessor, Reference Jessor1991; McCarthy & Grodsky, Reference McCarthy and Grodsky2011). Yet, unlike family structure and parental education, the parental control variable used in this study did not account for much of the variation in educational attainment. The study results suggest that parental control only accounted for 8% of the gap in PSE attainment for the females, while for the males the effect was negative. However, this finding is consistent with several prior studies, which found weak associations between parental control and educational outcomes (Crosnoe, Reference Crosnoe2001; Monserud & Elder, Reference Monserud and Elder2011). For instance, Crosnoe (Reference Crosnoe2001) found that although important, the effect of parental quality tends to decline as children transition to high school. It can be inferred from the current study that the effect of parenting quality is even weaker at post-secondary level, since at this level children are generally significantly more independent and their behaviour is influenced by diverse institutions.

Another important determinant of children’s educational outcomes is family income. This study found that family income had no effect on PSE enrolment for males, while it increased the odds for females. This was an unexpected finding given the relevance of socioeconomic and finance discrepancies in South Africa. There are several possible reasons as to why the family income variable behaved in this manner. In South Africa, children from disadvantaged families are supported through the National Student Financial Aid Scheme (NSFAS), which makes it possible for them to access higher education (Letseka & Maile, Reference Letseka and Maile2008). In addition, male children are more likely than female children to work while studying to supplement their income. Hence, at PSE level, the effect of parental income might be different. Perhaps it is important to explore deeper sociological reasons as to why family income behaved differently for males and females in this study. In patriarchal societies such as South Africa, girls and young women are often not seen as worthy of investment or protection by their families (Rarieya et al., Reference Rarieya, Sanger and Moolman2014). Recall that in Table 3 (Model 2, Females), the magnitude of the ASD coefficient decreased when family background variables were added to the model. Supplementary analyses (available upon request) also showed that family income decreased the odds for post-secondary education enrolment for early and typical sexual initiators, suggesting that the gaps in PSE enrolment are to a greater extent attributable to family income. Thus, while female children might still benefit from NSFAS, there are other constraints that might hinder their participation in education.

Overall, the findings of this study suggest that sexual debut is associated with post-secondary education enrolment and that the effect varies depending on individual aspirations and family background. Nevertheless, the results should be interpreted in light of the study’s limitations. The study population was limited to the Cape Town area of South Africa, and the socioeconomic dynamics in this province are different from those in other South African provinces. Thus, the findings, although reflective of a South African dynamic, cannot be generalized to all provinces. Second, the study did not make causal attributions to the relationship between sexual debut and enrolment in PSE. As such, the findings should be read as observed associations.

From a policy perspective, the results of this study have several implications. Given the negative effect of early sexual debut on academic outcomes, suggesting that late initiators have better outcomes, sexual education must emphasize abstinence, especially in the early teenage years. This should also be accompanied by the inclusion of research-based evidence (such as provided in this study) in the curriculum to encourage abstinence and goal-setting. For instance, the findings could be incorporated under the topic Health, Social and Environmental Responsibility in the Life Orientation Curriculum at the Intermediate Phase (Grades 7–9). Furthermore, expectations and career aspirations should form the basis of sexual education programmes in schools. Adolescence is an ideal time to focus on career development since it is during adolescence that young people face many changes and make decisions that have the potential to impact on their career aspirations.