Making the Cases Visible: Revealing Genres of Insurgency by Turning Regression Inside Out

• In Event: 1523 - Mixed Methods Research Approaches to Social Inquiry

Abstract

While cross-national quantitative analyses provide useful ways to think about the role of ethnicity in subnational conflict, case-based approaches—those emphasizing the context, history, and qualitative features of particular conflicts—offer a more nuanced perspective for understanding the circumstances that increase the likelihood of conflict, given a range of possible realities relating to a nation’s level of ethnic diversity. In this paper we call for a quantitatively inspired case-based approach to the comparative study of insurgency. We show that, inside the regression modeling that Fearon and Laitin (2003) and many other quantitative researchers have been employing, there are important opportunities for using the variables to learn about the distinctive features of the cases, rather than the usual applications of regression analyses in which, as comparativist researchers have long noted (Shalev 2007), the cases become invisible, and the analysts’ sole attention is focused on the variables. In this sense, we turn regression modeling “inside out” so as to exploit the variables as a means to let the cases be seen. We show that, with respect to their landmark 2003 paper, Fearon and Laitin’s own data reveal important and heretofore unreported lessons with regard to the role of ethnicity in the instigation of civil conflict. In doing so, we reveal how important lessons can indeed be gleaned from the use of ethnic fractionalization, even while the theoretical logic behind such a use remains contestable. We use our approach to look at statistical relationships in a qualitative way – noting that cases are at the same time affecting the regression coefficient in oftentimes contradictory ways.

Authors

• Andrew Davis, University of Arizona

• Ronald L. Breiger, University of Arizona