ABSTRACT
We consider sustainable development, resource dependence, and armed violence data across 104 developing countries over a five-year period, from 2003 to 2007. The sample is restricted to developing countries because we aim at analyzing whether there are systematically different development outcomes for resource-rich developing countries vs. resourcepoor ones in low-and middle-income countries, and to what extent the lack of sustainability can be explained by armed conflict and violence. The period under review is rather short and might seem arbitrary. This choice has first
been dictated by data availability. Our analysis includes armed violence as measured by the homicide rate. These data exist in relatively comprehensive fashion as of 2003, with 331 country-year pairs available over this five-year period (whereas armed conflict information is available for 525 country-year pairs). Because of an increasing prevalence of armed violence associated with extractive projects in weak states, we consider it worthwhile to use these novel data for the first time in a quantitative study on the relationship between resource dependence and sustainability outcomes despite the limiting factor imposed by missing information. Second, the period under review corresponds to a boom in commodity prices that reversed the downward trend observed during the two previous decades. Studying a period in which commodity markets boomed presents new insights into violence dynamics and extraction against the background of a global rush on commodities and the surge of Asian state-owned companies as major new actors in the market. A comprehensive list of countries and the number of observations per country can be
found in Table I. Among these 104 developing countries, roughly one-third are low-income countries according to the current World Bank classification, the remaining two-thirds are lower-middle-income and upper-middle-income countries. Thus, low-income countries are not systematically underrepresented, as may happen with novel and incomplete databases.
