Instruction des femmes et évolution du mariage en Inde

This study examines the influence of schooling on entry into marriage for women using panel data from the Indian censuses. Both schooling levels and marriage age increased in India between 1981 and 2001. While results from the cross-sectional data show that schooling is positively associated with delays in entry into marriage, results from the panel models suggest schooling to have a limited influence on marriage timing, especially during the 1981-1991 period. It had a significant, though modest, influence during the 1991-2001 period, however. The findings suggest that the association between schooling and marriage seen in the crosssectional analyses might have been due to unobserved factors influencing both schooling and marriage. Indeed, secular changes in marriage age were more important than changes in schooling levels in determining the timing of marriage before age 20. Further, the results from instrumental variable models suggest that the relationship between schooling and marriage is not endogenous: improvements in schooling levels are independent for the most part from changes in marriage age.