Criminal Order, Ports, and Coca Booms: The Onset, Amplification, and Persistence of Ecuador’s Territorial Homicide Divide (2014–2025)

Abstract This article examines homicidal violence in Ecuador by distinguishing three moments of a trajectory—onset, amplification, and persistence—and by asking how a transnational supply shock becomes territorially differentiated. Using administrative microdata (2014–2025; N = 39,819), we estimate difference-in-differences event studies on four spatiotemporal panels that combine provinces and cantons with monthly, quarterly, and annual frequencies. With 2019 as the reference year, joint Wald tests of all pre-treatment coefficients do not reject parallel trends in any panel. The Coast–interior divide opens in the last quarter of 2020—before the assassination of the leader of Los Choneros and before the change of government in May 2021—amplifies through 2023, and persists without reconvergence through December 2025; using 2020 as the reference, a robustness check, delays the dated onset to April 2021 because it absorbs the late-2020 rise into the baseline. Interacting coastal exposure with the log of Colombian coca cultivation shows that Colombia's two coca booms are not equivalent: during the first (2014–2019) the Coast's homicide trajectory is unrelated to coca supply, whereas during the second (2020–2025) it tracks it closely (1.8–2.2 log points per unit of ln coca across scales). Proximity to the nearest seaport reproduces the pattern at the cantonal scale, but coastal membership absorbs most of the effect. The results are consistent with a supply shock whose territorial transmission depended on the prior collapse of the criminal order.