Homicide and social media: Global empirical evidence

摘要

This study investigates the relationship between social media and homicide in a cross section of 148 countries for the year 2012. The empirical evidence is based on Ordinary Least Squares, Tobit and Quantile regressions. The findings from Ordinary Least Squares and Tobit regressions show a negative relationship between Facebook penetration and the homicide rate. The negative relationship is driven by the 75th quantile of the conditional distribution of the homicide rate. The negative nexus is also driven by upper middle income countries and "Europe and Central Asia". Three main implications are apparent when the findings are compared and contrasted. First, established findings from OLS and Tobit regressions are driven by countries with above-median levels of homicide. Second, such above-median countries are largely associated with upper middle income countries and nations in "Europe and Central Asia". Third, modelling the relationship between Facebook penetration and homicide at the conditional mean of homicide may be misleading unless it is contingent on initial levels of homicide and tailored differently across income levels and regions of the world.