ABSTRACT
This chapter discusses the mechanics of corruption indicator measurement with the statistics of attempting to measure something unobservable and difficult to quantify indirectly. It looks at measurement error, sample size, reverse causation, and spurious correlation as sources of problems in model endogeneity. Throughout the discussion that follows, emphasis is placed upon the Kaufmann-Kraay model (KK) measures, the Business Environment and Enterprise Performance Survey (BEEPS) measures, which are based on some of the same methodology, and the TI measures that fall victim to some of these general problems. In the Kaufmann model, there is cause to believe several of the independent variables may be correlated with democracy, as may be the outcome of better governance. Endogenous in economics refers to an internal process, such as income rising when education increases. Model uncertainty refers to the problem of what is included in the model and what is not.
