Research in Progress
Education, Social Norms, and Intimate Partner Violence Attitudes
Abstract. This paper examines cross-country variation in the relationship between women’s education and reported acceptance of intimate partner violence (IPV). I combine harmonized microdata from Multiple Indicator Cluster Surveys, conducted between 2005 and 2023, with country-level educational and other national measures. I first estimate country-specific education gradients in acceptance of at least one justification for IPV. In a seven-country subsample, I add controls for parental and interparental violence and assess sensitivity to unobserved confounding. I then relate these gradients to male and female basic education completion ratios. The female education gradients are predominantly negative but highly heterogeneous: 58 of 73 estimates are negative and statistically significant. A 10-percentage-point increase in male attainment is associated with a gradient that is 1.8 percentage points more negative, whereas the same increase in female attainment is associated with a gradient that is 2.2 percentage points less negative. These findings provide comparable cross-country evidence that the female education gradient in IPV acceptance is context-dependent and are consistent with the hypothesis that social norms shape how education translates into expressed opposition to IPV.
The Firstborn Daughter Effect and Confucian culture: Evidence from Korea
Abstract. This paper studies how having a firstborn daughter shapes parents’ attitudes and behaviors in South Korea, a society influenced by Confucian family norms. Using nationally representative survey data from the Korean General Social Survey and the Korean Time Use Survey, I estimate both the overall firstborn-daughter effect and the effect net of subsequent fertility responses. The results show that daughter effects are selective rather than broadly liberalizing. Having a firstborn daughter increases the probability of identifying crime as the top social-safety concern by 3.7 percentage points (17.5 percent of the baseline mean) for fathers and 3.4 percentage points (11.2 percent of the baseline mean) for mothers, while non-crime safety concerns show little robust evidence of change. In the marriage-and-childbearing domain, robust effects are concentrated on attitudes aligned with traditional family-formation norms: having a firstborn daughter increases opposition to cohabitation among both parents and opposition to nonmarital childbearing among fathers. Latent class analysis shows that these item-level effects also appear in broader attitudinal profiles, especially those characterized by traditional family-formation norms. Overall, the findings suggest that child-to-parent influence can be shaped by cultural norms and may reinforce specific conventional family-formation boundaries.