Relative cohort size and the risk of civil war, 1961-2001

Sarah E. Staveteig, University of California, Berkeley

Historians have posited that countries with youthful populations are more prone to war, but recent econometric studies of civil war have generally not found youth to be an important factor in conflict onset. This paper demonstrates that one likely reason for the contradictory evidence on youth and civil war is the poor conceptualization of the relationship between the two in prior research. A theory of the link between youth and conflict should involve Easterlin’s notion of relative cohort size (RCS). Using a logistic model of civil war onset worldwide, I find that, although several measures of youthful age structure are significant in the pooled cross-section analysis, first-differenced measures of youthful age structure involving RCS generally outperform other first-differenced measures of youth. The relationship between civil war onset and first-differenced RCS appears robust to a variety of controls. Hence a comprehensive understanding of civil wars would likely benefit from incorporating RCS.