Communal Violence and Long-Run Educational Attainment: Evidence from India
Abstract: This paper examines how exposure to communal violence shapes long-run educational attainment in India. I focus on the 2002 Gujarat violence, widely considered one of the most severe episodes of communal conflict in post-independence India. Beginning on February 28, three days of intense violence resulted in large-scale deaths and displacement, followed by further outbreaks of violence in the following months. To estimate the causal effects of this conflict, I use data from the 2019–21 National Family Health Survey (NFHS-5) and a difference-in-differences design that exploits variation in exposure across regions and birth cohorts. I compare individuals who were of school-going age during the violence to individuals who were old enough to have already completed their schooling, in Gujarat, relative to the rest of India. I find that Grade 12 completion declined by 6 percentage points (23%), and total years of schooling fell by 0.5 years (7%). Consistent with the violence disproportionately targeting women and children, I find that the declines were more concentrated among the youngest cohorts at the time of violence. Moreover, heterogeneity analyses show that the education losses were larger among females than among males, with Muslim females experiencing the largest declines. In India, where school completion rates are relatively low, these findings highlight how communal conflict can impose persistent educational costs and exacerbate existing inequalities.
Draft available upon request.
Upcoming Presentations: Northeastern Universities Development Consortium (NEUDC); Carolina Region Empirical Economics Day (CREED); Southern Economic Association (SEA)
Working Papers:
Abstract: This paper examines the effectiveness of minimum legal drinking age (MLDA) laws in restricting alcohol consumption, and the within-household spillover effects of individuals reaching the legal drinking age. India provides a particularly useful setting for this analysis because legal drinking ages vary across states, with thresholds of 18, 21, or 25, while some states prohibit alcohol entirely. These regulations are intended to reduce the social harms associated with alcohol consumption, which are especially salient given the documented relationship between alcohol use and intimate partner violence in India. Using multiple rounds of the National Family Health Survey and an age-based regression discontinuity design, I test whether alcohol consumption changes sharply when individuals cross the legal drinking-age threshold in their state. I find no evidence of discontinuity at the ages of 21 or 25. In contrast, I document a detectable increase at age 18, regardless of the minimum legal drinking age in the individual’s state of residence. More generally, alcohol consumption rises gradually with age. These findings suggest that state-specific legal drinking-age thresholds are largely nonbinding and that raising these thresholds may have limited effectiveness in curbing alcohol use. Because age 18 marks the transition to legal adulthood in India, the discontinuity at this age is consistent with social norms surrounding adulthood playing a more important role in drinking behavior than the legal thresholds themselves.
Abstract: We examine the impact of British rule, which introduced gender reforms to abolish child marriage and promote women’s education, on women’s age at first marriage and education post-independence. We use National Family Health Survey (NFHS) data from 1998-99 and a difference-in-differences approach, comparing Madhya Pradesh (formerly British-ruled) and Rajasthan (not formerly British-ruled) across birth cohorts pre- and post-independence (1947). We find that women born in Madhya Pradesh post-independence attained higher education (0.2 more years; 5.3% fewer with no education) but married slightly earlier (0.47 years), with child marriage remaining more prevalent among Hindu women and women in rural areas.
Work in Progress:
The Effects of New School Construction on Learning Outcomes in India