Teenage marriages and births continue to be high on the agenda in many countries, rich and poor. A recent report notes that 12 million girls get married each year before they turn 18, a figure that translates into 650 million women who are alive today (International Center for Research on Women, 2018).

Turkey is among the middle-income countries that face significant levels of teenage marriages and births. According to the 2013 Turkish Demographic and Health Survey (TDHS), 22% of women aged 25-49 were married before they were 18.

Teenage marriage and births are associated with various adverse outcomes that include poorer education and labour market outcomes, higher welfare dependency and crime. The children of teenage mothers also carry a higher risk of teenage motherhood and have worse labour market, education and health outcomes.

Evidence suggests a strong negative correlation between teenage marriage and education. In the case of Turkey, among women aged 25-49 in 2013, while the median age of marriage was 18.9 for those with no school degree, it was 24.6 for those with a high school diploma or above.

But does this observation imply causality? In other words, would an increase in women’s schooling decrease teenage marriage and births? If this is indeed the case, compulsory schooling policies that mandate children to stay in school up to a certain age or for a pre-determined number of years may work to reduce teen marriages and births.

An increase in years of education may increase the age at marriage and first birth for two reasons. The first is the ‘compulsory attendance effect’, which arises because schooling and marriage are incompatible events. Requiring teens to stay in school longer would delay their entry into the marriage market and this would delay childbearing, particular in conservative societies, where giving birth out of wedlock is not socially acceptable. A rigid sequence of events of completion of education, marriage and birth of the first child that is observed in various countries (Blossfeld and De Rose, 1992) is also the case in Turkey, where the average time between marriage and first birth is 1.6 years.

The second reason for more years in education raising women’s age at marriage and first birth is the ‘human capital effect’. With increased schooling, the opportunity cost of marriage and raising children increases for women because of their higher market wages. Other human capital-related factors include more effective use of contraceptives, higher bargaining power in fertility decisions, and changes in marriage and fertility preferences of women with higher levels of schooling.

In an attempt to understand how compulsory schooling affects teenage marriages and births, our recent study (Kırdar et al, 2018) analyses data from the 2008 and 2013 TDHS. We study the 1997 schooling reform that extended compulsory schooling from five to eight years, and which resulted in a substantial increase in the number of students in grades 1 to 8. The completion of grade 8 would earn the students a basic education diploma, which would entitle them to continue to high school.

In an earlier study (Kırdar et al, 2016), we establish that the policy not only increased completion rates in the newly mandated schooling levels but also had ‘spillover effects’ onto high school, with the result that women’s completed years of schooling increased by more than a year.

The 2008 and 2013 TDHS cover women aged 15-49. Using these two data sets, we construct histories of ‘ever-married’ status and ‘ever-given-birth’ status until age 19 or until the current age at the time of the survey, whichever comes earlier. In our empirical analysis, we study how the policy has affected the ever-married status, ever-given-birth status, time to first birth after marriage, and the number of children ever born to women at each age.

Our results indicate that the extension of compulsory schooling reduces the probability of teenage marriages: the probability of girls getting married at ages 12-14 – the years in which they are mandated to remain in school (grade levels 6 to 8) – is reduced, which is likely to be due to the compulsory attendance effect.

But the probability of marriage is also reduced for 15 and 16-year-olds, who are not mandated to stay in school, suggesting a human capital effect. The size of the policy effect is quite substantial: the probability of marriage by age 16 is reduced by about 50%.

When we examine the marriage hazard rates, we find a lower rate at age 15. (The hazard rate for 16-year-olds is generally lower as well but statistically insignificant.) But the probability of marriage at age 17 conditional on not getting married until that age increases due to the policy, pointing to a catching-up effect that was not present at ages 15 and 16. As a result, the probability of being ever-married reverts to its pre-policy level after age 17.

Since nearly all births are to married women, the reduction in marriage probability for girls younger than 17 translates into a reduction in first birth for 15 to 17-year-olds. The probability of giving birth by age 17 is reduced by almost 50% due to the policy. But we do not find a policy effect beyond age 17.

Furthermore, we find that once a woman (who is almost certainly out of school) is married, there is no policy effect on the time to first birth. We interpret these findings as suggesting a strong compulsory attendance effect, but a relatively short-lived human capital effect of compulsory schooling.

If girls are postponing getting married and giving birth to their first child due to the extension of compulsory schooling, the number of children born to these women must also fall. Indeed, this is what we find.

The number of children born to a woman by age 18 is reduced. We estimate the size of the effect to be on the order of 0.07 to 0.10. Due to the relatively recent application of the reform, we are yet to determine whether the reform will lead to a reduction in the total number of children ever born these women by the end of their reproductive years.

But our finding that the total number of children is not reduced for women a little older than 18 (for example, at age 20) suggests that perhaps the total number of children born will remain the same.

In summary, we find that the extension of compulsory schooling – which was motivated by concerns other than the welfare of teenage mothers and their children (see Kırdar et al. 2016) – has had an unintended effect of reducing teenage marriage and the number children born to teenage women. These changes, in turn, are likely to have positive welfare effects at the country level.

Further reading

Blossfeld, HP, and A De Rose (1992) ‘Educational Expansion and the Changes in Entry into Marriage and Motherhood: The Experience of Italian Women’, Genus 48(3-4): 73-89.

International Center for Research on Women (2018) ‘Child Marriage Facts and Figures’.

Kırdar, GM, M Dayıoğlu and I Koç (2016) ‘Does Longer Compulsory Education Equalize Schooling by Gender and Rural/Urban Residence?’, World Bank Economic Review 30(3): 549-79.

Kırdar, GM, M Dayıoğlu and I Koç (2018) ‘The Effects of Compulsory-Schooling Laws on Teenage Marriage and Births in Turkey’, Journal of Human Capital 12(4): 640-68.