The politics of social security: on regressions, qualitative comparisons, and cluster analysis

Abstract

Recent debates in macrosociological and comparative methodology have revolved around the respective merits of the “variable-oriented” and “case studies” approaches. In its extreme form, the methodological debate has presented these two approaches as mutually antagonistic strategies and even as epistemologically distinct ways to interpret social reality. They have been contrasted by exemplifying them as nomothetic versus idiosyncratic brands of social sciences, or “inquiry from the outside” versus “inquiry from the inside” (see Bryman 1988). Indeed, there are essential differences that seem to legitimate the sharp distinction. The variable-oriented approach attempts to apply standard quantitative methods in comparative analysis on a wider sample of countries, while the case approach is more focused on qualitative and historical analysis of individual cases. As a rule, the variable-oriented approach is theory-centered (see Ragin 1987, p. 55; Kiser and Hechter 1991). Hypotheses are derived from theories in order to test the relevance of the theory by submitting it to empirical examination that is usually conducted with the help of multivariate techniques. Thus, the approach is hypothetical-deductive in character and is employed in order to obtain results that can legitimately be generalized to the broadest possible sample. By contrast, case studies are not aimed at testing theories per se; instead, the formulation and evaluation of theories proceeds simultaneously with historical narration.