Poverty of Democracy: The Institutional Roots of Political Participation in Mexico

Appendix A

217 The majority of the statistical analysis in this book relies on the results of the 2003 Mexico wave of the CSES survey (see http://www.cses.org). In addition to the standard CSES survey questions, which are the same for every country, each country collaborator can add additional questions to the end of the questionnaire. For all the regression analysis and much of the rest of the statistical analysis, I rely on the full Mexico version of the 2003 CSES survey which is housed at the CIDE’s library (Centro de Investigación y Docencia Económica), which I refer to as CSES-CIDE 2003 (the datasets for 2000, 2003, 2006 can be accessed free of charge at http://www.biiacs.cide.edu/ consulta).The principal advantage of this version is that it has a much richer set of political participation questions than the standard CSES questionnaire and than the 2006 CIDE version. Some of the analysis, particularly in chapter 4, used four additional datasets: the 1990–2005 Mexico waves of the World Values Survey, and the standard CSES dataset for Mexico from 1997 and 2000 (referred to as CSES 1997 and CSES 2000). This appendix contains descriptions of all of the variables used in the analysis for this book. In some cases I provide the exact wording of the survey question, and in others I describe how the variables were constructed. Unless otherwise noted, the variables were taken from the CSES-CIDE 2003 dataset. Political Participation 1. Voting. “Did you vote in the past elections held on July 6 for federal deputies?” 2. Campaign activity. “I am going to read a few activities that people do during elections. Please tell me if you did any of these activities during the appendix a. Survey Questions and Variables holzner text-3.indd 217 8/6/10 10:52 AM 218  Appendix A. Survey Questions and Variables past elections: Did you show support for a party or candidate by attending meetings or campaign rallies, distributing posters or bumper stickers , or in any other way?” 3. Political contacting. “During the past five years, have you have you contacted any politician or government official in person, by writing a letter or in any other way?” 4. Protest. “Please tell me if during the past five years you’ve participated in any manifestation, protest or march?” 5. Overall activity index. For 2000 data the political participation index is an additive scale of fifteen different political acts: voting in the 2000 elections, contacting a congressional representative, signing a petition, organizing a complaint or petition against the government, protesting, participating in a sit-in, occupying government offices, participating in land invasions, blocking streets or roads, participating in an armed uprising against the government, distributing campaign propaganda, attending a campaign meeting or rally, donating money to a political party or candidate, working for a political campaign, party or candidate, and trying to persuade others to vote for a party or candidate. One point was given for each act; the scale runs from 0 to 15 (Cronbach alpha = 0.6797). For analyses using 2003 data, the political participation index is an additive scale of eight different acts: persuading others to vote for a party or candidate, voting in 2003 elections, participating in campaign activities or events, participating in marches or protests, contacting government officials , organizing informally with others to solve a common problem, occupying government offices, blocking roads or highways. One point was given for each act; the scale runs from 0 to 8. (Cronbach alpha = 0.556) Demographic Variables 1. Income. The WVS measures income on a 10-point scale, with monetary values varying across waves. In all cases, I coded those who reported incomes ranging from 1 to 3 points as low income, 4–6 points as middle income, and those reporting incomes ranging from 7 to 10 points as high income. Though not identical, the coding of income categories for all datasets turned out to be quite consistent. The CSES-CIDE 2000 and 2003 coded household income according to multiples of the monthly Mexican minimum wage, with the following frequency distributions: In addition, I divided individuals into three income categories: holzner text-3.indd 218 8/6/10 10:52 AM Appendix A. Survey Questions and Variables  219 Low income = 1 to 3 times minimum wage ($0–$3,930 for 2003) Middle income = 4 to 7 times minimum wage ($3,931–$9,170...