Using daily and monthly level night light products from National Aeronautics and Space Administration (NASA) Black Marble suite (NASA and Administration, 2199) and extrapolating hartal / strike related violence data with keyword search from geocoded Armed Conflict Location & Event Data Project (ACLED) database, we investigate the impact of such events on urban
& economic activity in Bangladesh. We focus our investigation firstly at daily level, to discern the immediate impacts of such events on daily urban activity and secondly at monthly level, to extrapolate the economic impact stemming from hartal induced shocks to key infrastructural sectors, namely transport. For daily level analysis, we utilize Autoregressive Conditional Heteroskedasticity (ARCH) estimation to factor in the deeply autoregressive nature of daily night lights, to identify immediate (within-day) effects from hartals, individually for key subdistricts. At the monthly level, to factor in the emergent consequent spatial dependence as well as suspected endogeneity from hartals, we analyze country wide dynamics in a control function
approach with bias corrected Quasi Maximum Likelihood approach as suggested by fei Lee and Yu (2010b) in the second stage. In first stage we use either lagged count of peaceful non-hartal events or 2008 election division level majority party parliamentary seat shares, interacted with subdistrict specific trends, as instruments. At daily level, over 2012-21, in the capital Dhaka, we find that daily hartals have an immediate statistically significant impact of -0.9 percent on daily night lights. However this effect does not hold uniformly across all subdistricts, and onlydoes so for a select number of subdistricts. At the monthly level, we find evidence of statistically significant country wide effects of up to 6 percent.