Violence and Corruption : An examination of the global effect within a causal framework

Corruption, a global problem with extensive impacts on societies and democracies,persists as a popular topic for researchers within political science, economics and statis-tics. Nevertheless, much uncertainty exists about the causal relationship between violenceand corruption. Violence is often overlooked as a primary factor to corruption, especiallyin global contexts. This paper aims to broaden awareness on the topic by investigatingif causal inference can determine global causal relationship between rising violence andincreased corruption in the subsequent year. The methodological approach taken in thispaper is causal inference through the Rubin’s causal model. Using cross sectional data from 133 countries, and adjusting for covariates of economicand political factors, propensity score nearest neighbour matching is performed. While thematching procedure achieves good covariate balance the estimation is sensitive to hiddenbias and is not found to be significant. Therefore the analysis cannot determine a significantcausal effect of the treatment on the treated. These results could originate from omittedconfounding variables, or a need for a more complex model to capture the causality ofinterest.