Conflict and healthcare: Evidence from Gujarat

This paper examines whether the 2002 Gujarat riot in India disrupted maternal healthcare utilisation, using data from the 2002–04 District-Level Household Sur- vey (DLHS-2) and a sharp Regression Discontinuity Design around the date of the riot. While the effect is in the negative hypothesised direction, I find no statistically significant effects on institutional delivery, caesarean section rates, post-pregnancy complications, or polio drops given to newborns at birth. Subgroup analysis, how- ever, suggests a decline in hospital deliveries among women aged 30 or younger and among Muslim women statewide. NGO activity and community health net- works may have mitigated the short-run disruptions to maternal healthcare, though this study does not directly test these mechanisms. The results provide actionable lessons for policymakers and development practitioners seeking to protect maternal health in conflict-prone, low-and middle-income settings, emphasizing the value of localized health interventions and institutional preparedness. JEL Code : J16, D74, I11