Unintended Consequences of Lockdowns, COVID-19 and the Shadow Pandemic in India. A Reproduction Study of Ravindran and Shah

Ravindran and Shah (2023) assesses the effect of COVID-19 lockdowns on violence against women (VAM) in India. Using a dynamic fixed effects methodology exploiting spatio-temporal variation in the timing and intensity of 'stay-at-home' mandates in Indian districts, the authors find that government-mandated lockdowns increased complaints of domestic violence in districts with the strictest lockdown rules. I successfully computationally reproduce their results. Further, I test the replicability of the results by re-analyzing the results using a Poisson estimator as opposed to the ordinary least squares (OLS) estimator used by the original authors. I find that lower bound censoring of the outcome measure resulted in attenuation of the marginal effect of government-mandated lockdowns. The effect of mobility restrictions on VAW are understated in the original study. This provides support for the internal validity of Ravindran and Shah (2023)'s results by suggesting that may present lower bound estimates of an under-reported consequence of mobility restrictions during the COVID-19 pandemic.