What's the worth of a promise? Evaluating the long-term effects of a programme to reduce early marriage in India

This paper examines the long-term effects of ‘Apni Beti Apna Dhan’ (ABAD), a unique conditional cash transfer programme implemented in the north Indian state of Haryana from 1994 to 1998. The programme provided bonds to female beneficiaries at birth, redeemable on attaining 18 years of age, if they remained unmarried. In particular, we examine the programme effects on prevalence of early marriage and then on educational attainment. Exploiting the exogenous variation in the birth year of females born when the programme was implemented, we find a significant reduction in the likelihood of early marriage before age 18 among ABAD beneficiaries. Additionally, the prevalence of marriage by age 16, 17 or 18 is found to reduce, though the likelihood of being married by age 19 remains unchanged. We further observe robust improvements in their educational attainment. Despite reduction in early marriage and years of education, we do not find any discernible evidence of improvements in labour market participation, time allocation patterns, women empowerment indicators or intergenerational spillover. We provide suggestive evidence of parents utilizing the educational gains to ensure marital association with families of higher social status.