The Success of the Left in Latin America: Untainted Parties, Market Reforms, and Voting Behavior

Notes

Notes Introduction 1. Their dependent variable is a neat indicator called Vote-Revealed Leftism (VRL), which takes into account the ideological positions and electoral results of all parties or candidates that get votes. There is no doubt that the use of this aggregate measure has several advantages, as Baker and Greene mention; however, it does not allow us to make any inference at the micro level due to the risk of an ecological fallacy. Chapter 1. Latin American Ideological Cycles in the Postwar Era 1. In the 2009–2010 presidential election, things had changed in Chile with the victory of the right-of-center Coalición por el Cambio. 2. These three types of trends are not exhaustive. It is also possible to find random movements or fluctuations, but these kinds of movements are not usually described as trends. A “random trend” is meaningless. If a trend is random , it does not look like a trend. 3. In those elections in which there are different electoral results for the two chambers (senate and deputies), we use the results for the lower chamber (deputies). 4. Only the results of congressional elections are taken into account. The exception is the results for the Peruvian elections before 1978, which are based on presidential rather than legislative votes. 5. Alexander 1988; Coggins and Lewis 1992; Alcántara and Freidenberg 2001. 171 6. The extension of this classification was done by Germán Lodola and Rosario Queirolo while studying at the University of Pittsburgh. The following country experts were consulted: Germán Lodola and Belén Amadeo (Argentina ), Daniel Moreno Morales and Vivian Schwarz (Bolivia), Lucio Renno and Rachel Meneguello (Brazil), Francisco Díaz and Juan Pablo Luna (Chile), Laura Wills and Miguel García (Colombia), Mitchell Seligson, Juliana Martínez, and Harold Villegas Roman (Costa Rica), Grisel Lerebours (Dominican Republic), Agustín Grijalva and Pablo Andrade (Ecuador), Cynthia McClintock, Ricardo Córdova, and Margarita Correa (El Salvador), Dinorah Azpuru and Margarita Correa (Guatemala), José René Argueta and Margarita Correa (Honduras), Luis Jiménez and Juan Antonio Rodríguez-Zepeda (Mexico), John Booth and Margarita Correa (Nicaragua), Aníbal Pérez-Liñán and José Costa (Paraguay), Cynthia McClintock and Luis E. González (Perú), Juan Pablo Luna, Fernanda Boidi, and Rosario Queirolo (Uruguay), Aníbal Pérez-Liñán and Margarita López-Maya (Venezuela). 7. With the exception of the coding of new parties. 8. A table listing the political parties in each country that do not fit into the Left-Right dimension from 1980 to 2010 appears in Appendix 1. The table also indicates the percentage of the total vote that these unclassified parties obtained in the presidential elections. 9. An election is considered “fraudulent” when there is doubt that the actual electoral results are significantly different from the official ones. 10. Electoral results are taken from different sources. Coppedge’s (1997) dataset is used for the 1945–1995 period. Since 1995 the main source of electoral data has been the Political Database of the Americas (PDBA) at Georgetown University. Complementary sources were consulted to fill in blanks or verify the information: www.observatorioelectoral.org, Nohlen 2005, and the Web pages of electoral offices in each country. 11. Mexico fulfills the first three rules set out by Przeworski et al. (2000) to define a democracy, but it does not pass the “alternation rule.” 12. Figure A.1 in Appendix 1 shows Latin American ideological cycles excluding these Mexican elections. The main difference between this figure and figure 1.2 is that the predominance of the Left in the second cycle is longer and more pronounced when these controversial Mexican elections are excluded. 13. The graphs for each country are shown in Appendix 1. 14. For the same reason, Mainwaring and Scully (1995) were not able to measure ideological polarization, but taking into account the analysis of the case studies presented in their book, they categorize Costa Rica as a low polarization system; Colombia, Paraguay, and Argentina as systems with moderately 172 Notes to Pages 24–28 low polarization; Venezuela, Uruguay, Bolivia, Chile, and Ecuador as moderately high polarization; and Mexico, Peru, and Brazil as highly polarized systems . In this classification made in 1993, there are more countries with moderately high to high polarization than countries with low polarization. 15. From 1956 to 1969 and from 1976 to 1979, there are more random movements between...