External and Internal Conflict Behavior Among Nations, 1966-1967

Abstract

In this article we attempt to replicate the hypothesis tested by Rummel and others that external and internal conflict are unrelated. We do this using data for 125 nations for the years 1966-1967. As did Rummel, we use exploratory factor analysis and regression analysis; in addition, we also employ confirmatory factor analysis. Results from confirmatory factor analysis contradict Rummel and reveal moderately strong correlations between internal and external conflict factors. Regression analysis and partial correlations, however, show that zero-order cross-country correlations between internal and external conflict are reduced to insignificance when a control variable, population size, is introduced in the analysis.