More education, persistent withdrawal: Evidence from Türkiye

Withdrawal, a predominantly male-controlled contraceptive method, remains widely used in Türkiye despite near-universal awareness of and access to modern contraception. We study whether expanding female education changes contraceptive method choice, using Türkiye's 1997 compulsory schooling reform, which extended compulsory education from five to eight years, as an instrument in a regression discontinuity design. Among ever-married women, completing eight or more years of schooling reduces the probability of withdrawal use by 20.8 %age points among the treated group. The estimated effects on modern contraceptive adoption and no method are positive but statistically insignificant. Mechanism tests find little evidence for the knowledge, access, or fertility-preference channels. Instead, the evidence supports a life-cycle re-timing mechanism: additional schooling keeps women in school through the ages of earliest union formation, delaying first marriage and first birth and reducing exposure to the early-marriage environments in which withdrawal is more prevalent, without a detectable increase in women's intra-household bargaining power. Further analyses provide suggestive evidence that the reduction in withdrawal may be larger among women with lower religiosity, although the difference across religiosity groups is only marginally statistically significant. These findings suggest that schooling alone may be insufficient to increase modern contraceptive adoption without complementary interventions addressing gender norms and couple decision-making.