Table 6: Cross-sectional results for conflict in Rakhine State Conflict dummy (1) (2) (3) CoverageFB Coverage Observations District dummies Spatial controls Population controls No. conflict events (4) (5) (6) −0.008 (0.009) −0.001 (0.009) 0.027∗ (0.015) −0.019 (0.015) 0.027∗ (0.015) −0.017 (0.015) 25.142 (25.587) −30.099 (21.805) −22.878 (16.556) −51.634 (42.714) −33.832 (23.837) −44.133 (34.805) 1059 X 1059 X X 1059 X X X 1059 X 1059 X X 1059 X X X ∗ p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01. Robust standard errors are reported in parentheses. The dependent variable is an indicator for conflict of particular type in a township between June 2016 until end of August 2017. Conflict data is from GDELT. Population controls: log population, log population density, dummy for below median urban rate, age (15–64 y.o.), population with no ID, population with electricity, mobile phone, landline phone, and internet at home. Spatial controls: 2nd order polynomials of distance to major town, distance to major road, distance to railway, distance to MPT transmitter, distance to other company’s transmitter, mean elevation, slope and aspect of the slope, variance of elevation and slope. The predictors are standardized. 7. Conclusions This paper has studied the effect of availability of Facebook on conflict. I exploit geographic variation in cell phone coverage together with time variation in an availability shock of Facebook to estimate whether social media affects probability or intensity of conflict. The results suggest that availability of social media may have had an average negative effect on conflict incidence. The cross-sectional analysis that exploits exogenous variation in cell phone coverage suggests that social media availability had a mitigating effect on conflict. However, this effect cannot be identified in the panel analysis. Furthermore, there is evidence that social media may have led to a slightly higher incidence of Rohingyarelated conflict. Therefore, it seems that influence of social media depends greatly on the environment in which it is introduced. Without access to Facebook content at the time, or more detailed information about internet or cell phone use, it is difficult to further disentangle these effects. There is a need for further research on the impacts of social media use in different circumstances. 21