Ethnic Conflict and Secessionism

Abstract

AbstractThrough statistical analysis, this chapter tests the generalizability of its argument regarding the effect of decentralization and regional parties on ethnic conflict and secessionism. The analysis draws on an original dataset, known as the constituency‐level elections (CLE) dataset, which provides election results for national (lower and upper house) elections and regional elections from 1945 to 2002 at the constituency‐level of government, to measure systematically the electoral strength of regional parties. To measure ethnic conflict and secessionism, the chapter draws on the Minorities at Risk dataset (1985–2000), which it corrects for selection bias. The analysis supports the chapter's argument that decentralization diminishes the strength of ethnic conflict and secessionism, while regional parties intensify it, and that the ability of decentralization to reduce conflict decreases as the electoral strength of regional parties increases. The chapter controls for a number of factors that may affect regional party strength (e.g. ethnolinguistic heterogeneity, economic development, democracy, and the executive and electoral system). The chapter also uses instrumental variable regression in this chapter to disentangle the causal relationships between decentralization, regional parties, and ethnic conflict and secessionism.