Killing as Capital: Perverse Effects of Truce Negotiations on Gang Violence in El Salvador

In March 2012, the government of El Salvador brokered a truce between MS‐13 and Barrio 18, the two largest and most notorious gangs in El Salvador. The truce was designed to decrease homicides in exchange for more lenient treatment of incarcerated gang leaders. Despite early, promising results, the truce was short‐lived. From May 2013 to February 2015, the government walked back its concessions under the truce. Homicides increased steadily during the rescission period and exponentially in the wake of full revocation. Economic theory suggests negotiating with gangs may achieve short‐term gains, but may also cause long‐term losses, particularly when the government reneges: once negotiations are on the table, gangs may use killing to increase their political capital and induce the government to re‐enter negotiations and make greater concessions. Using Salvadoran crime data, we deploy a difference‐in‐differences model to estimate the effect of the truce on homicides. We estimate the truce resulted in 1130 fewer homicides during its implementation and 551 fewer homicides during its piecemeal revocation. However, we also estimate 2250 more homicides occurred after full revocation, a perverse net effect of 569 more lives lost compared to what would have happened had the truce not been negotiated. These findings demonstrate negotiating with gangs may be an effective means to curb gang violence, but, if negotiations result in an unstable truce, they also introduce a perverse incentive structure that may result in long‐term harms that exceed short‐term gains.