In 1994, the Zapatistas took up arms claiming for indigenous people rights in Chiapas, Mexico. After 12 days of civil war, the national government called for dialogue. Nevertheless, since then, it has deployed a "low intensity war" over the self-declared Zapatista Autonomous Communities. At the same time, the Zapatistas started to implement a new set of institutions, which have allegedly enhanced their socio-economic situation. The purpose of this study is, thus, to elucidate this ambiguous theoretical effect on the wellbeing of the communities under harassment. This paper generates a unique dataset, linking socio-economic variables from the Mexican Census with different measures of conflict intensity at the locality level, based on geo-coded influence areas from the military and police positions disseminated throughout Chiapas. The present investigation controls for the endogeneity in the relationship between conflict and the socio-economic performance, instrumenting the former by the distance from each locality to a strategic military spot defined by the Zapatista Army for its uprising in 1994 on the natural boundary of the Lacandon Jungle. The results, robust to different specifications and conflict intensity definitions, imply that the impact of the Zapatista institutions has surpassed the negative effect of the civil strive, suggesting that: i) bottom-up policies carried out by grass-root organizations, even in times of conflict, might represent an appropriate path for endogenous socio-economic development; and ii) the Mexican government should recognize the Zapatista autonomy and its right for self-determination.