Choosing cases for case studies: A qualitative logic

Abstract

David Freedman in his essay has called into question some of the advice given by Fearon-Laitin and Gerring in their chapters in the Oxford Handbook of Political Methodology. I would like to second those criticisms and extend them in various ways. In particular, Freedman does not address in much detail the regression or logit model which underlies, explicitly or implicitly, both these chapters—and more generally Gerring's book on case study methodology (2007). This "regression approach" to case studies (to give it a name) informs much discussion about case studies and qualitative methods, going back to King, Keohane, and Verba (1994) and more recent works such as Lieberman (2005). In these few pages I can but sketch a rationale for choosing cases following a different logic of research. In the first part of the essay I address the choice of case studies from a qualitative logic of research. In the second part, I briefly describe a "descriptive–causal" approach to case study selection which is different from the regression logic of Fearon, Laitin, and Gerring.