Mining policy reform and civil conflict: Evidence from Myanmar

Abstract

This paper inverts the standard resource-conflict paradigm by examining how resource contraction, rather than expansion, affects civil violence. Exploiting a 2016 moratorium in Myanmar that halted new mining licenses, I implement a difference-in-differences strategy using a novel spatially disaggregated database linking mining activity with geo-coded conflict events from 2011–2020. The contraction led to a 69% reduction in conflict incidents — particularly violent and fatal events — in previously licensed townships. The effects were stronger in ethnic homelands, poorer areas, and remote regions. Strikingly, the analysis uncovers positive spatial spillovers: conflict also declined in neighboring non-mining areas, suggesting that reduced resource extraction diffuses peace rather than displacing violence. Evidence supports three mechanisms: (1) lower mineral rents constrained armed group financing; (2) labor reallocation to productive sectors increased the opportunity cost of violence; and (3) reduced elite rents mitigated local inequality, dampening grievance-based mobilization.