Special Data Feature Funding rebellion: The Rebel Contraband Dataset James Igoe Walsh, Justin M Conrad, Beth Elise Whitaker & Katelin M Hudak Department of Political Science and Public Administration, University of North Carolina at Charlotte Abstract We introduce a new dataset measuring if and how rebel groups earn income from the exploitation of natural resources or criminal activities. The Rebel Contraband Dataset makes three contributions to data in this area. First, it covers a wide range of natural resources and types of crime. Second, it measures rebel engagement in these activities over time. Third, it distinguishes among different strategies that rebel groups employ, such as extortion and smuggling. Theory suggests that reliance on natural resource wealth should lead rebels to mistreat civilians, but cross-group research using existing data does not find support for this relationship. We replicate an earlier study using data from the Rebel Contraband Dataset and conclude that there is a consistent relationship between natural resource exploitation and civilian victimization. Future research can use the dataset to explore questions about the onset, location, severity, and outcomes of civil conflicts. Keywords civil war, crime, natural resources, resource curse Introduction Extensive literature establishes a relationship between natural resources and conflict, including its onset (Coll- ier & Hoeffler, 2002; Fearon & Laitin, 2003; Ross, 2012), duration (Collier, Hoeffler & So ¨derbom, 2004; Fearon, 2004; Lujala, 2010), intensity (Addison, Le Bil- lon & Murshed, 2002; Lujala, 2008), and recurrence (Doyle & Sambanis, 2000; Rustad & Binningsbø, 2012). Scholars have highlighted how resource charac- teristics (such as ‘lootability’) and motivating factors (such as ethnic exclusion) increase the likelihood of vio- lence (LeBillon, 2001; Lujala, Gleditsch & Gilmore, 2005; Ross, 2006; Snyder, 2006; Asal et al., 2016). A related body of research shows that the resource endow- ments of rebel groups influence their strategies and beha- vior (Weinstein, 2005), with groups that rely on natural resources or foreign support being more likely to abuse civilian populations (Weinstein, 2007; Cohen, 2013; Wood & Kathman, 2015). Despite the wealth of research on resources and con- flict, knowledge in this field has been limited by data availability. Many studies rely on state-level data about the presence and/or volume of natural resources within a country and assume that armed groups operating there have access to these resources to fund their rebellions. A few more focused datasets make important contributions by providing information about the exploitation of resources in specific conflicts (Fearon, 2004; Rustad & Binningsbø, 2012) or the location of high-value resources vis-a `-vis conflict sites (Lujala, 2010), but they are limited in scope and/or temporal variability. To this point, it has not been possible to analyze the relationship between resources and conflict dynamics with global time-series data at the level of the rebel group. We build on the work of scholars such as Rustad & Binningsbø (2012), Lujala (2010), and Fearon (2004), and introduce the Rebel Contraband Dataset (RCD), which measures rebel groups’ exploitation of natural resources and their involvement in organized criminal Corresponding author: jwalsh@uncc.edu Journal of Peace Research 2018, Vol. 55(5) 699–707 ª The Author(s) 2018 Reprints and permission: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0022343317740621 journals.sagepub.com/home/jpr