A new discussion paper by Niaz Asadullah and Zaki Wahhaj, KDPE 1602, February 2016
Non-technical summary
A third of women in developing countries around the world marry before the age of 18, and about one in nine before the age of 15. A large literature argues that early marriage disrupts the accumulation of human capital among adolescent girls due to early school drop-out, withdrawal from labour markets and adverse effects on health from early childbearing.
International development agencies, national governments and NGO’s have made concerted efforts in recent years to lower the incidence of early marriage through new legislation on child marriage, improved enforcement of existing laws and interventions aimed at adolescents.
In this paper we investigate whether female early marriage is a conduit for the transmission of social norms, specifically norms relating to gender roles and rights within the household (henceforth called `gender norms’). Gender norms are believed to play an important role in perpetuating gender inequalities in child survival, education, control over assets and economic participation in a wide range of developing countries.
The challenge for empirical research on the consequences of early marriage is that girls who marry early tend to be poorer, have less educated parents, and to be born in rural areas; but these background characteristics can have a direct effect on their opportunities and subsequent life choices (such as schooling, fertility and employment). A recent set of studies have used variation in the timing of menarche across women to estimate the impact of early marriage on future outcomes.
To investigate the effects of early marriage on gender norms, we introduce an innovation to this approach by making use of a unique dataset with first-hand information on the age of onset of menarche and marriage timing of sisters in rural Bangladesh. Specifically we use the age of menarche as a source of exogenous variation in marriage timing between sisters. To the extent that sisters are raised within the same household by the same parents, this allows us to control for beliefs and attitudes that are transmitted from parents to children, and identify the effects of the social consequences of reaching menarche and early marriage. We provide evidence that early marriage affects a woman’s attitudes towards traditional gender norms and the characteristics of her social network. In particular, early marriage reduces the likelihood that a woman in her social network has made a non-traditional life choice (completed secondary school, used contraception before the birth of her first child, engaged in an income-generating activity) and increases her agreement with statements supportive of gender bias in the allocation of resources and traditional gender roles.
We find that the woman’s own schooling, her husband’s schooling, and her social network account for, at most, one-third of the estimated effect of early marriage on gender norms. Furthermore, using a sample of adolescents, we find no evidence that early onset of menarche directly leads to increased agreement with traditional gender norms. Taken together, the evidence suggests that the key pathway of norm transmission is early marriage itself or, more specifically, the socialisation of young girls within the marital household.