This article analyzes the "risk factors" of secessionism at the substate, regional level. It seeks to answer the question, What regions are more likely to support more successful secessionist parties? Using new data in cross-sectional regression analysis, the author finds that secessionism involves unique factors not common to other kinds of ethnic conflict. Specifically, in addition to "identity" variables such as regional language and history of independence, the following variables explain secessionist strength: lack of irredentist potential, relative affluence, geographical noncontiguity, population, and multiparty political system. These factors generally serve as activators of ethnic identity rather than a substitute for the same, although there are important cases of nonethnic secessionism.

This article introduces a new, public database of U.S. state and local public policies, now available at www.statepolicyindex.com. The database covers more than 170 different state or local policies, coded at the state level as of December 31, 2006, in most cases. We use principal components analysis and derive two orthogonal measures of state policy ideology, which we label policy liberalism and policy urbanism. Our policy liberalism measure passes several reliability and validity checks, while policy urbanism is strongly predicted by urbanization rate, percentage of African Americans in the population, and percentage of Christian adherents in the population.

No abstract

Several possible relationships between natural resources and civil conflict have been hypothesized and tested in the literature. The impact of resources on conflict should depend on the circumstances of the group that (potential) rebels see themselves as representing and depend upon for support. While ‘lootable’ resources such as alluvial diamonds have been shown to increase the likelihood of insurgency, among territorially concentrated ethnic groups looting by rebels recruiting from the group is counterproductive because it imposes negative externalities on the rebel constituency. However, local mineral abundance could encourage rebellion indirectly, by promoting the development of secessionist objectives, since autonomy or independence would allow the rebel constituency to enjoy a larger share of the benefits flowing from mineral revenues. On the other hand, mineral abundance could encourage the government to exercise greater surveillance and control over potentially restive minority populations. On balance, then, mineral abundance should affect ethnoregional conflict primarily by encouraging ethnic rebels to adopt limited, territorial-autonomy objectives as opposed to governmental objectives. This hypothesis is tested with a new, global dataset of substate mineral production. Local mineral resource abundance is indeed negatively associated with governmental conflict among ethnoregional groups and positively related to secessionist or territorial conflict. Moreover, it is the total value of mineral production that matters, not specific types of minerals such as oil or diamonds. The net effect of mineral abundance on the total risk of intrastate conflict onset among ethnoregions is essentially zero.

No abstract