EN
This paper estimates the effect of growth on civil conflict by comparing the identification strategies and the necessary assumptions for the different instrument variable approaches. We show that different exclusion restrictions might contradict each other which cast serious doubt on the validity of their identification strategies. We find it very reasonable to believe that there are causal effects between civil conflict, quality of democratic institutions and government size.Therefore, one can doubt that the way weather variations have been used to create instrument variables in the literature leads to consistently estimated parameters for the causal effect of civil conflict, quality of democratic institutions, demography and government size on civil conflict.