Does Change in a Country's Armed Forces Personnel Signal a Potential Conflict? Evidence from World Development Indicators.

This paper investigates the unique contents of World Development indicators, and the predictability of a country’s lagged change in Armed Forces Personnel on conflict. Using World Bank data from 1960 to 2015, we show the lagged change in Armed Forces Personnel is negatively correlated to a country’s Gross Domestic Product (GDP). Applying logistic regression methods to the data, we show the lagged change in a country’s Armed Forces Personnel does predict conflict. We separate the in-sample countries into two samples. The first sub-sample is used to create a model to predict conflict, while the second sub-sample is used to test the model. After adjusting for country-specific effects, we use the model parameters to test the accuracy of the out-of-sample countries. We find a country’s lagged change in Armed Forces Personnel predicts conflict for the in-sample countries with an overall accuracy of 93.1% (first sub-sample), 86.5% overall accuracy (second sub-sample), and 94.9% overall accuracy for the out-of-sample countries.