ABSTRACT
Several of the most widely used meas ures of corrup tion are, at least in part, based upon percep tions of corrup tion. For example, Transparency International’s Corruption Perceptions Index, prob ably the most widely used measure of corrup tion, is expli citly a measure of the perceived level of corrup tion (Andersson and Heywood 2009). Intrinsically, there is nothing wrong with a percep tion-based eval u ation of corrup tion. Citizens of a country have direct lived exper i ences of the level of corrup tion in their own country; country experts have an argu ably broader famili ar ity, although when they are not resid ent in the country in ques tion this famili ar ity is perhaps less ‘deep’. Moreover, since many cases of corrup tion will go unrepor ted to the author it ies, corrup tion percep tions can almost inev it ably consider more cases than an ‘object ive’ analysis of, say, offi cial corrup tion reports. However, despite such virtues, there remains a legit im ate fear that percep tual meas ures of corrup tion are not simply imperfect, as all meas ures are, but are instead an inad equate record of the level of corrup tion within a country (for a discus sion, see Treisman 2007: 241).
