Why now for a cumulative qualitative comparison?

Designing social research involves trade-offs between complexity and generality. There remains an often overblown and unhelpful distinction in social sciences between strategies that standardly prioritize one over the other. For Ragin, ‘it is easy to exaggerate their differences and to caricature the two approaches, for example, by portraying quantitative work on general patterns as scientific but sterile and oppressive and qualitative research on small Ns as rich and emancipatory but soft and subjective’ (2000: 22). QCA sets itself up as a method that provides an avenue for assuaging some of the unhelpful tensions in this distinction.

The attractiveness of QCA as both a critical methodological approach and a novel set of techniques has led to its application across an everexpanding group of diverse disciplines and subdisciplines in the social sciences (Thiem and Dusa, 2013: 2). It has been one of the few approaches in recent times to come from outside the methodological orthodoxy that has not been quickly chased back to the margins of social science. The number of peer-reviewed articles per year using QCA has grown gradually since the first in 1984 – up to 15 in 2005, to 35 in 2011, and then more quickly to 45 in 2012 and 99 in 2013. Of those 99 articles, 16 were agenda-setting articles introducing QCA to new subdisciplines (among them, Ryan and Smith, 2012). A large and yet non-exhaustive bibliography of QCA studies on the COMPASSS website attests to the method's continued growth in popularity, containing over 1,000 articles and 80 books at the time of writing in 2020.

There is a well-worn debate in the social sciences about the value of large-N and small-N research strategies. Part of the early promise of QCA was to develop a method that makes feasible medium-N research strategies, in particular by extending the logic of small-N comparisons in a reasonable way. Ragin (2000: 25) plots the relative number of studies against the number of cases per study in the discipline producing a U-shaped curve, with several studies covering a handful of cases, several covering hundreds of cases, and a few covering case numbers that lie in between.