A domestic power faces an enemy and commits terrorist atrocities to increase the likelihood of victory. A foreign patron can grant aid to the power but prefers fewer or no atrocities. The domestic power responds by acquiescing in the creation of uncontrollable paramilitaries that commit even...

We present a detailed, high-frequency dataset on the civil conflict in Colombia during the period 1988-2002. We briefly introduce the Colombian case and the methodological issues that hinder data collection in civil wars, before presenting the pattern over time of conflict actions and intensity...

We compare the treatment of Colombia in large cross-country conflict datasets with the information of the detailed micro dataset of Restrepo, Spagat and Vargas (2003). We find a general tendency of the big datasets to underestimate the magnitude of the Colombian conflict and to mischaracterize...

Group 1 holds political power. Group 2 threatens this power. Group 1 decreases the upheaval probability by co-opting some agents from Group 2 into a more benign Group 3. Improvements in upheaval technology lead to less co-optation. Increasing the relative size of Group 1 implies larger...

Analysis of our new, 16-year dataset on the Colombian civil war finds under Uribe: guerrilla and paramilitary attacks dropping sharply against long-run averages since 1988, lower for April-December, 2003; government-guerrilla clashes at all-time highs, exceeding guerrilla attacks; civilian killings...