Selecting Cases and Inferential Types in Comparative Public Policy Research

Although many scholars claim to present comparative public policy research (see the various journals and books in this field with the word ‘comparative’ in their titles), only a few scholars appear to have designed a comparative study. My critique here echoes Benoît Rihoux’ observation that, in practice, comparisons are too often ‘rather loose or not formalized’ (Rihoux, 2006: p. 681), and Ahrend Lijphart’s observation that the comparative method appears so basic and apparently simple that when applied it often ‘indicates the how but does not specify the what of the analysis’ (Lijphart, 1971: p. 682). Playing the devil’s advocate, I would like to take these observations a step further and claim that much of published comparative public policy research is ex-post facto comparative and not a priori comparative in nature. I understand this is a bold and sweeping statement, but in reading classic and contemporary comparative works, only a handful may be termed comparative by design (for example, Verba and Nie, 1972; Skocpol, 1979; Vogel, 1996; Levi-Faur, 2006b), whereas many works are comparative by outcome, or are studies that present various examples of a phenomenon of interest without actually having a rationale for comparatively studying these examples. Throughout this book many guidelines will be provided as to how to design comparative public policy research.