205 N otes Chapter 1 Democracy, Economic Policy, and Political Corruption in Comparative Perspective 1. See Gerring and Thacker (2004) for a recent review of the literature. 2. Note that this is a purposefully broad definition of political corruption, a term that some might interpret more narrowly as corruption within the electoral or legislative spheres. But corruption within the government more generally, including the bureaucracy, is certainly a political phenomenon, as it involves public officials who have either been elected themselves or have been appointed by elected politicians . For these reasons, I prefer the term “political corruption,” which encompasses both the narrow concept of electoral or legislative corruption as well as what others may call “bureaucratic” or “administrative” corruption. This broad definition of political corruption is also consistent with the empirical measures used in this chapter (see Gerring and Thacker 2005). 3. Other, more intermediate-level institutions such as federalism/unitarism and presidentialism/parliamentarism also influence corruption (see Gerring and Thacker 2004). But there is insufficient variation in these institutional arrangements in Latin America to get empirical leverage on them in a regional context. 4. This segment of the chapter draws upon Gerring, Thacker, and Alfaro (2006). 5. For example, see Alt and Lassen (2008) for a treatment of the role of elected (as opposed to appointed) judges in reducing corruption. 6. Democracy is not a necessary condition for the presence of a strong civil society . But it is likely that, all else being equal, civil society networks will be stronger in a democratic system than a nondemocratic one. 7. Institutional decay can occur under democratic rule, too, as the cases of Venezuela and Peru have demonstrated in recent years. The present argument is probabilistic : old democracies are more likely than new democracies and authoritarian regimes to have strong institutions. 8. For an excellent summary and intellectual history of neoliberalism, see Colclough (1991). 9. This section draws upon Gerring and Thacker (2005). 10. The privatization process, for example, has been fraught with abuse in many countries. See also La Palombara (1994, 340); Snyder (1999); Tanzi (1998, 563). 11. The classic Stigler-Peltzman theory of regulation, in which utility-maximizing 206 n otes to pages 33–34 interest groups offer political support to regulators in exchange for favorable treatment , is apropos here. See Stigler (1971) and Peltzman (1976). 12. The World Bank calculates this indicator every other year, beginning in 1996. The present study covers the four observations from 1996 through 2002. This indicator is available for two additional observation years (2004 and 2006), but insufficient coverage of some of the control variables precludes an analysis of those data points. 13. The World Bank indicator and Transparency International’s more wellknown Corruption Perceptions Index are very highly correlated (r = 0.92). Because of the World Bank index’s greater coverage, I focus here on that measure. For discussion of the relative merits of cross-national corruption polls, see Elliott (1997); Gerring and Thacker (2004); Heywood (1997); Jain (1998); Johnston (2000); and Robinson (1998). 14. This index is a poll of polls. The polls (of businesspeople, country residents, and experts) ask respondents to rate the general level of corruption among public officials, the effectiveness of anticorruption initiatives, the frequency of additional payments necessary to “get things done,” and corruption as an obstacle to foreign investment and domestic business enterprise (for further details, see Kaufmann, Kraay, and Mastruzzi 2003). Recent scholarship has explored the relationship and the differences between the perception of corruption and actual corruption (see Morris’s contribution to this volume for one example). One might question whether perceptions of corruption as judged by survey responses accurately reflect an empirical reality (actual, existing corruption). It could be that perceptions measure precisely that— perceptions—and nothing more. There is no question that corruption is to some extent , here and everywhere, a matter of perception. Yet, at the same time, it is likely that perceptions of corruption have some substantial basis in reality. Olken (2007) has shown, for example, that while perceptions of corruption do typically contain valid information about levels of corruption, that information is often incomplete. At the same time, perceptions themselves can matter; if foreign investors believe a country to be corrupt, that belief itself may alter their investment patterns (Treisman 2007). In this sense, the perception of corruption may be at least as important as its reality. In employing the World Bank perception-based measure of political corruption , this chapter does not claim to provide definitive tests of the...