The efficacy of violence mitigation: A second look using time-series analysis

Among peace researchers and practitioners, it is generally accepted that efforts to prevent violence can be instrumental in its mitigation. So, it was distressing to read the research findings of Meier, Bond, and Bond (2007) that mitigation was positively related to organized raids in the Horn of Africa where there is pastoralist–pastoralist and pastoralist–agriculturalist violence. This article seeks to build on their research. It uses a ‘de-trending’ approach for time-series analysis that is commonly used in economics and financial studies. It reports an opposite statistically significant finding. When the data used by Meier et al. are de-trended, violence associated with organized raids is negatively correlated with mitigation. This negative correlation is similar when data on mitigation and organized raids are de-trended with time as a predictor, on the one hand, and with seasonality over time as a predictor, on the other. Implications regarding the temporal dimension of peace research and practice are presented.