Qualitative comparative analysis vis-à-vis regression

Discussions of Charles C. Ragin’s Qualitative Comparative Analysis (QCA) have not adequately considered the assumptions about causation on which this method depends. Yet in evaluating any method, it is important to ask the question: How many untestable, or hard-to-test, assumptions must be met for us to believe the findings it produces? Advocates of QCA claim that one of its major strengths is that it requires fewer restrictive assumptions than techniques such as regression analysis. Hence, close assessment of the assumptions that are entailed is particularly salient to evaluating QCA. This article addresses these issues by considering three of the most important kinds of assumptions discussed in the context of regression analysis: assumptions about the correct form of the relationship, missing variables, and inferring causation from association. For each assumption, the role of corresponding assumptions in QCA will be explored and illustrated through an analysis of leftparty electoral fortunes in Latin America. Regarding the correct form of causal relationships, QCA in effect builds highly demanding assumptions into measurement procedures. Concerning missing variables, whereas earlier versions of QCA require a strong assumption of no causally relevant missing variables, more recent procedures allow some kinds of missing variables, but build in mutually contradictory statistical assumptions about those variables. Resolving these contradictions essentially converts QCA into an application of regression analysis. Regarding the process of inferring causation from association, QCA makes causal inference on the basis of patterns of association purely by assumption. That is, association is assumed to have a one-to-one relationship with causation. For all three groups of assumptions, QCA is found to require assumptions that, are at least as restrictive as those employed in regression analysis.