Territory is one of the main drivers of international and subnational conflicts. Explanations of intrastate conflict suffer from two major limitations: First, they rely on a non-random sample in analyzing only violent conflict, limiting out understanding why conflicts develop. Second, most accounts of intrastate violence rely more or less explicitly on ethnicity as the main predictor, while considerable definitional unclarity remains. This paper offers an explanation for claims for territorial autonomy focusing on territorial value and testing the assumptions derived from the conflict literature against the complete set of subnational territorial claims. I analyze geo-coded data on ethnic groups and the territorial conditions they face for the period between 1995 and 2005 to understand the predictors of demands for autonomy. Initial results find mixed support for the assumptions derived from the established conflict literature. Settlement patterns, political status, strategically important terrain, alluvial diamond deposits, as well as peripheral location are found to be associated with territorial claims. To understand how the sources of violent and non-violent demands systematically differ, the two subsets are explicitly distinguished in future research.
