Escalation of civil war in Nepal: The role of local poverty, inequality and caste polarization

Abstract

A growing literature examines the effects of poverty, inequality and polarization on civil war but analyses typically use spatially aggregated data. Many civil wars exhibit great spatial heterogeneity so very local effects may be obscured. We study Nepal’s civil war that escalated sharply from 2001 to contribute evidence on these local effects. We combine geo-coded data on 15 000 conflict deaths spread over a decade with small-area estimates of poverty and inequality for almost 4000 localities in Nepal. We model the likelihood of a locality being affected by conflict and the intensity of the conflict, in terms of the number of deaths. Higher local poverty rates are negatively related to an increased risk of conflict and deaths, due to a shift in strategy by the rebels to target richer middle class and urban areas for accessing resources to win the war. Local relative wealth inequality is postively associated with the escalation of conflict, suggesting that relative wellbeing affects decisions about rebellion and conflict. Caste polarization also raises conflict risk and deaths, especially where dominant caste groups were more numerous. These support an agnostic view that poverty is neither a necessary nor sufficient condition for the escalation of civil conflict. The strategy used by the rebels in Nepal saw the conflict escalate away from poor areas even as conflict was more likely in unequal and caste-polarized areas.