Deprivation theory is tested with multivariate time-series regression models, with controls for both security force levels and industrial production. It was found that although Northern Ireland suffers from high unemployment, increases in fatal violence are not related to rises in unemployment, and that violence has a tendency to perpetuate itself. It was also found that the conflict is not representative of clashes between State and insurgent forces in that most of the fatalities in it are not attributable to the security forces, and that further explanations should incorporate both political factors and the mechanisms that contribute to the escalation process. The paper explores several possibilities. 6 figures, 2 tables, 75 references. (Author abstract modified)