Conceptualizing and Measuring Institutions: a View from Political Science Initial Conference Draft: Neither for Circulation nor for Citation

Introduction The following paper identifies and takes issue with a potentially troubling development in the study of comparative politics. While political scientists have traditionally deployed objective indicators of “political organization and administrative capacity” (Reynolds 1983, p. 976), including tax ratios (Organski and Kugler 1980; Benson and Kugler 1998), tax structures (Kling 1968; Krasner 1985), political participation (Przeworski and Sprague 1971), and the nature and extent of public service provision (Migdal 1988; Putnam 1992), they are beginning to import subjective indicators of “governance” from economics (Sandholtz and Gray 2003; Fish 2005; Gerring and Thacker 2005; Blake and Martin 2006; Cameron et al. 2006). In fact, the customary distinction between objective and subjective measurement is by now less salient than the various differences among the subjective or “perceptions-based” alternatives (Kaufmann et al 2005a). Whose perceptions do they capture? Of which institutions or issue areas? And to what effect? We hold that the leading perceptions-based indicators are poorly explicated; give the perceptions and interests of businesspeople (and foreign businesspeople in particular) undue influence; are riddled with measurement error and all but impossible to interpret; and should therefore be either reconsidered or replaced by improved objective indicators that take public sector inputs as well as outputs into account. We make the case for reconsideration and/or replacement in four principal sections. Section 1 offers an abbreviated introduction to the perceptions-based literature. Section 2 develops a critique of the conceptual and operational underpinnings of the leading perceptionsbased indicators. Section 3 discusses the problem of random measurement error in practice. And Section 4 discusses objective alternatives.