Mexico s Evolving Democracy. A Comparative Study of the 2012 Elections. Edited by Jorge I. Domínguez. Kenneth F. Greene.

Transcription

1 Mexico s Evolving Democracy A Comparative Study of the 2012 Elections Edited by Jorge I. Domínguez Kenneth F. Greene Chappell Lawson and Alejandro Moreno Johns Hopkins University Press Baltimore i

2 2015 Johns Hopkins University Press All rights reserved. Published 2015 Printed in the United States of America on acid-free paper Johns Hopkins University Press 2715 North Charles Street Baltimore, Maryland <CIP data to come> A catalog record for this book is available from the British Library. Special discounts are available for bulk purchases of this book. For more information, please contact Special Sales at or specialsales@press.jhu.edu. Johns Hopkins University Press uses environmentally friendly book materials, including recycled text paper that is composed of at least 30 percent post-consumer waste, whenever possible. ii

3 Figures for online appendix A-3.1. Parties yearly public subsidy A-3.2. Fairness through the eyes of a regulator A-7.1. Three party configuration of competition in Mexico 2012 elections A-7.2. Cartogram of Mexico 2012 elections A-7.3. Moran I and plots of the spatial autocorrelation in dependent variables iii

4 Tables for online appendix A-4.1. Multinomial logistic regression results A-5.1. Survey question wordings used in constructing mood estimates A-5.2. Panel model of vote choice in 2006 Mexican presidential election A-6.1. A model of vote choice in Mexico s presidential election, 2000 A-6.2. A model of vote choice in Mexico s presidential election, 2006 A-6.3. A model of vote choice in Mexico s presidential election, 2012 A-6.4. Types of campaign effects in presidential elections, A-6.5. Panel response model A-8.1. Baseline summary statistics and orthogonality tests, list experiment iv

5 Figure A-3.1. Parties yearly public subsidy Superscripts p and m indicate presidential and midterm election years, respectively. Columns report yearly total (ordinary and campaign) subsidies to parties. Black portion of columns in election years shows amount of total subsidy corresponding to radio and television advertisement spending by parties in 1997, 2000, 2003, and White dotted portions report an attempt to valuate unaccounted advertisement subsidies in 2009 and Numbers in parentheses below columns report the number of registered parties eligibale for subsisies, numbers above 2003 and subsequent columns report the six-year percentage change in total subsidy. The exchange rate used is US$1=Mx$12.50.

6 Figure A-3.2. Fairness through the eyes of a regulator The spreadsheet indicates how one specific network in the state of Hidalgo (XHPAH-TV, channel 3) must schedule campaign spots from the parties in six commercial breaks of each hour between 6AM and 2PM on a daily basis. The range portrayed corresponds to the official campaign, between May 9 and June 9, Source: downloaded from

7 Table A-4.1. Multinomial logistic regression results: coefficients and (standard errors) Presidential Voting JVM/EPN AMLO/EPN Other, NV/EPN Retrospective Assessment of Unified Governance under the PRI -.45 (.27) -.56 (.27) -.30 (.17) Conventional Retrospective Evaluations National Economy.01 (.19) -.13 (.20) -.09 (.15) Personal Finances.35 (.21).24 (.20).12 (.16) Crime Fighting -.06 (.21) -.30 (.21) -.19 (.15) Approval rating for Calderón.04 (.17).17 (.15).02 (.11) Partisanship PAN identification.85 (.20).01 (.24).22 (.18) PRI identification (.27) -.51 (.20) -.70 (.14) PRD identification -.77 (.50).77 (.24).12 (.25) Candidate Evaluations Josefina Vázquez Mota 1.58 (.24).15 (.23).23 (.16) Enrique Peña Nieto (.24) (.26) (.17) Andrés Manuel López Obrador -.20 (.20) 1.60 (.20).17 (.14) Political Ideology On the left.08 (47).61 (.43) -.19 (35) On the right -.07 (.42) -.25 (.48) -.28 (.32) In the center.01 (.40) -.13 (.41) -.53 (.29) Socioeconomic and Demographic Variables Gender (Female).20 (.33) -.34 (.35) -.13 (.25) Age (Years).01 (.01) -.01 (.01) -.02 (.01) Education -.14 (.08) -.02 (.08).01 (.07) Household affluence.08 (.11) -.11 (.11) -.09 (.09) Attendance at religious services.08 (.14) -.01 (.14) -.03 (.10) Constant Term (1.36) 1.82 (1.82) 5.29 (1.00)

8 Table A-4.1, continued. Legislative Voting PAN/PRI PRD/PRI Other, NV/PRI Retrospective Assessment of Unified Governance under the PRI -.85 (.23) -.32 (.22) -.58 (.18) Conventional Retrospective Evaluations National Economy -.27 (.18) -.01 (.18) -.25 (.15) Personal Finances.17 (.19).19 (.19).07 (.16) Crime Fighting -.18 (.19) -.31 (.19) -.23 (.15) Approval rating for Calderón.07 (.15) -.04 (.14).02 (.12) Partisanship PAN identification.86 (.18) -.23 (.26) -.05 (.15) PRI identification (.59) -.56 (.21).79 (.18) PRD identification.09 (.28).75 (.22).11 (.22) Candidate Evaluations Josefina Vázquez Mota.98 (.21) -.26 (.20) -.20 (.16) Enrique Peña Nieto -.59 (.22) (.22) -.54 (.17) Andrés Manuel López Obrador -.30 (.16) 1.11 (.18).17 (.14) Political Ideology On the left.03 (42).51 (.41).03 (35) On the right -.12 (.38).18 (.42) -.18 (.32) In the center -.06 (.36) -.19 (.42) -.85 (.32) Socioeconomic and Demographic Variables Gender (Female).13 (.29) -.18 (.28) -.14 (.24) Age (Years).01 (.01) -.01 (.01) -.01 (.01) Education -.08 (.08) -.23 (.08) -.12 (.06) Household affluence.24 (.10).11 (.10).14 (.09) Attendance at religious services.07 (.13) -.01 (.12).07 (.11) Constant Term.59 (1.28) 2.79 (1.35) 5.31 (1.04)

9 A-5.1. Survey question wordings used in constructing mood estimates Responses that are tallied as liberal are shaded in grey. (1) Attitudes toward privatization, Privatization of the petroleum sector (Latinobarometer). (1995, 1998, 2007, 2008) From the following list of activities, which do you think should be majority-owned by the state and which do you think should be majority-owned by private hands? Petroleum. Privatization of the electricity sector (Latinobarometer). (1995, 1998, 2007, 2008) From the following list of activities, which do you think should be majority-owned by the state and which do you think should be majority-owned by private hands? Electricity. Privatization of the telephone sector (Latinobarometer). (1995, 1998, 2007, 2008) From the following list of activities, which do you think should be majority-owned by the state and which do you think should be majority-owned by private hands? Petroleum. Privatization of the water sector (Latinobarometer). (1995, 1998, 2008) From the following list of activities, which do you think should be majority-owned by the state and which do you think should be majority-owned by private hands? Petroleum. Privatization of retirement pensions (Latinobarometer). (1995, 1998, 2008) From the following list of activities, which do you think should be majority-owned by the state and which do you think should be majority-owned by private hands? Retirement Pensions. Price and quality of privatized services are good (Latinobarometer). (2003, 2004, 2005, 2006, 2010, 2011) State-owned services like water, light, etc., have been privatized. Taking into account the price and the quality, are you currently much more satisfied, satisfied, less satisfied or much less satisfied with privatized services?

10 Privatization is good for the country (Latinobarometer). (1998, 2000, 2001, 2002, 2003, 2005, 2007, 2009, 2010, and 2011) The privatization of state-owned enterprises has been beneficial to the country. Do you strongly agree, agree, disagree, or strongly disagree with this statement? State should yield to private sector (Latinobarometer). (1998, 2001, and 2002) The state should leave productive activity to the private sector. Do you strongly agree, agree, disagree, or strongly disagree with this statement? Private ownership should be increased (World Values Survey). (1990, 1995, 2000, 2005, and 2012) Now I d like you to tell me your views on various issues. How would you place your views on this scale? 1 means you agree completely with the statement on the left; 10 means you agree completely with the statement on the right; and if your views fall somewhere in between, you can choose any number in between. Sentences: Private ownership of business should be increased vs Government ownership of business should be increased Private ownership of business and Government ownership of business and industry should be increased. industry should be increased. State should own the most important industries (Latin American Public Opinion Project). (2008, 2010, and 2012) The Mexican state, instead of the private sector, should be the owner of the most important industries in the country. Indicate how much you agree with this sentence Strongly disagree Strongly agree Sale of state-owned enterprises (Office of the Technical Advisor to the President of Mexico). (1989). During the government of President Salines, various businesses that were property of

11 the government have been sold. Do you agree or disagree with this policy? (1994) The current government sold most of its businesses (banks, telephones, highways) to private interests. Do you agree or disagree with this policy? Private investment in the electricity sector (Mexico Panel Study 2006). (2005 and 2006).Do you believe that 1) More private investment should be allowed in the electricity sector or that 2) the electricity sector should remain almost completely in hands of government (2) Attitudes toward foreign investment Foreign investment is good for the country (Gallup). (1988 and 1991). Considering all of the consequences, good and bad, in your opinion is investment in Mexico by foreign companies positive or negative for the country? Foreign investment should be encouraged (LB). (1998 and 2001). Foreign investment should be encouraged. Do you strongly agree, agree, disagree, or strongly disagree with this statement? Large foreign companies are good (Pew). (2002 and 2007) As I read a list of groups and organizations, for each, please tell me what kind of influence the group is having on the way things are going in Mexico. Is the influence on Mexico very good, somewhat good, somewhat bad, or very bad: large companies from other countries. Foreign investment helps the country (Office of the Technical Advisor to the President of Mexico). (1993) The economic policy of commercial opening to foreigners has attracted investment. Based on what you know, has the establishment of foreign businesses in Mexico helped the economy of the country or not? (1994) Based on what you know, has the establishment of foreign businesses in Mexico helped the economy of the country or not?

12 (3) Attitudes toward NAFTA and international trade. Country benefits from NAFTA. (Latinobarometer). (1996, 1997, 1998, 2001 and 2003) All things considered, would you say your country benefits a lot, quite a bit, a little or not at all for being part of NAFTA? Support for NAFTA (Office of the Technical Advisor to the President of Mexico). (1991, 1992 and 1993). Removing taxes and barriers to trade in products between countries is the goal of NAFTA. Do you agree or disagree with a treaty of this type between Mexico, the USA, and Canada? (4) Diffuse values about the market and capitalism Fairness of unequal wages (Latinobarometer). (1995, 2002, 2007) Imagine the following situation: two people, of the same age, work as computer programmers doing the same work. One earns 20,000 pesos (equivalent to U$50) more than the other one, but he does his work more quickly, efficiently, and is more trustworthy than the other. In your opinion, do you consider it fair that, in this situation, one programmer is paid more than the other or do you consider it unfair? Personal versus government responsibility (Latinobarometer). (2004 and 2006) On a scale from 1 to 10, where 1 is each person should be in charge of their own well-being and 10 is the government should assume responsibility for people s well-being, where do you place yourself? Each person should be in charge of their own well-being. The government should assume responsibility for people s well-being

13 Personal versus government responsibility (World Value Survey). (1990, 1995, 2000, 2005, and 2012) Now I'd like you to tell me your views on various issues. How would you place your views on this scale? 1 means you agree completely with the statement on the left; 10 means you agree completely with the statement on the right; and if your views fall somewhere in between, you can choose any number in between. Sentences: People should take more responsibility to provide for themselves vs The government should take more responsibility to ensure that everyone is provided for The government should take more People should take more responsibility. responsibility. People are better off in a free market economy (Pew). (2002, 2007, 2009, 2010, 2012) Please tell me whether you completely agree, mostly agree, mostly disagree or completely disagree with the following statements. Most people are better off in a free market economy, even though some people are rich and some are poor. State is responsible for well-being (Latin American Public Opinion Project). (2008, 2010, and 2012). The Mexican state, more than individuals, should be principally responsible for ensuring the well-being of the people. Indicate how much you agree with this sentence Strongly disagree Strongly agree State is responsible for job creation (Latin American Public Opinion Project). (2008, 2010, and 2012). The Mexican state, more than the private sector, should be principally responsible for creating jobs. Indicate how much you agree with this sentence

14 Strongly disagree Strongly agree State should redistribute wealth (Latin American Public Opinion Project). (2008, 2010, and 2012). The Mexican state should implement firm policies to reduce the inequality in incomes between the rich and the poor. Indicate how much you agree with this sentence Strongly disagree Strongly agree Elements in the Economic Liberalism index from the 2006 Mexico Panel Study A. Do you believe that 1) More private investment should be allowed in the electricity sector 2) The electricity sector should remain almost completely in hands of government B. What would you prefer: commercial relations between Mexico and the United States increase, decrease, or remain the same? 1) Increase 2) Remain the same 3) Decrease C. In general, what would you prefer? 1) That the government be responsible for the economic well being of individuals 2) That individuals be responsible for their own economic well being D. In your opinion, what should the government do to reduce poverty? 1) Give money to the poor and raise taxes to the rich 2) Promote private investment and leave taxes as they are

15 Table A-5.2. Panel model of vote choice in 2006 Mexican presidential election Variable Calderón Madrazo Others Abstain or DK Economic liberalism t AMLO.341* AMLO.206* AMLO AMLO.203 (.097) (.104) (.145) (.119) Economic liberalism t (.098) (.100).290* (.141) (.118) Presidential Approval Approval of Fox t.541* (.083).176* (.080) (.093).138 (.092) Approval of Fox t (.072) (.076) (.095).136 (.085) Party ID PANista t 1.676* (.225).674* (.261) (.355) (.338) PANista t (.184).134 (.230).110 (.313).259 (.240) PRDista t * (.276) * (.330) * (.578) * (.391) PRDista t (.201) (.245) (.300).098 (.297) PRIista t.103 (.234) 1.706* (.207) (.328).134 (.285) PRIista t (.194).317 (.174) (.270).255 (.232) Lagged DVs Calderón vote t-1.714* (.245) (.332) * (.379) * (.322) AMLO vote t * (.223) * (.264) * (.292) * (.283) Madrazo vote t (.290) 1.274* (.286) * (.347) * (.300) Others, abstain, or DK t-1 0 (0) 0 (0) 0 (0) 0 (0) Constant (.304) (.310).653 (.366) (.360) Entries are multinomial logit coefficients with robust standard errors in parentheses. Standard errors are corrected for clustering by respondent. * p<.05, one-tailed test. 1,721 distinct respondents. T=3. Overall N=2,824.

16 Table A-6.1. A model of vote choice in Mexico s presidential election, 2000 Variable Labastida vs. Fox Cárdenas vs. Fox Coef SE Sig Coef SE Sig PAN ID, Feb *** * Δ PAN ID, Feb-July *** *** PRD ID, Feb *** Δ PRD ID, Feb-July *** PRI ID, Feb *** Δ PRI ID, Feb-July *** Privatization policy preferences, Feb Δ Privatization policy preferences, Feb-July Democracy assessment, Feb Δ Democracy assessment, Feb-July Labastida (PRI) honesty, Feb *** Δ Labastida (PRI) honesty, Feb-July *** Fox (PAN) honesty, Feb *** *** Δ Fox (PAN) honesty, Feb-July *** *** Cárdenas (PRD) honesty, Feb *** Δ Cárdenas (PRD) honesty, Feb-July *** Retrospective evaluations, Feb Δ Retrospective evaluations, Feb-July Cárdenas probability of victory, April Constant Multinomial regression models. The dependent variable is reported vote choice in the July survey. N = 932, pseudo-r2 =.599. Fox is the excluded category. * p<.1, ** p <.05; *** p <.01, two-tailed tests.

17 Table A-6.2. A model of vote choice in Mexico s presidential election, 2006 Variable Madrazo vs. Calderón López Obrador vs. Calderón Coef SE Sig Coef SE Sig PAN ID, Oct ** *** Δ PAN ID, Oct -July *** PRD ID, Oct ** Δ PRD ID, Oct -July *** PRI ID, Oct *** Δ PRI ID, Oct-July *** Retrospective evaluations, Oct ** ** Δ Retrospective evaluations, Oct-July *** *** Economic policy preferences, Oct Δ Economic policy preferences, Oct-July Calderón (PAN) competence, Oct *** ** Δ Calderón (PAN) competence, Oct-July * * López Obrador (PRD) competence, Oct *** Δ López Obrador (PRD) competence, Oct-July *** Madrazo (PRI) competence, Oct *** Δ Madrazo (PRI) competence, Oct-July *** Madrazo probability of victory, Oct Δ Madrazo probability of victory, Oct-July Constant % vote choices correctly predicted w/o campaign 69.9% % vote choices correctly predicted w/ campaign 89.3% Multinomial regression models. N = 391, pseudo-r2 =.695. Calderón is the excluded category. * p<.1, ** p <.05; *** p <.01, two-tailed tests. Note: The dependent variable is reported vote choice in the July survey. The percent correctly predicted without the campaign was generated by setting change scores to zero. The percent correctly predicted with the campaign from October to July was generated with the full model.

18 Table A-6.3. A model of vote choice in Mexico s presidential election, 2012 Peña Nieto vs. Vázquez Mota López Obrador vs. Vázquez Mota Variable Coef SE Sig Coef SE Sig PAN ID, April Δ PAN ID, April-July * ** PRD ID, April *** Δ PRD ID, April-July *** PRI ID, April *** Δ PRI ID, April-July *** Economic policy preferences, April Δ Economic policy preferences, April-July * Vázquez Mota (PAN) competence, April * **** Δ Vázquez Mota (PAN) competence, April-July *** *** López Obrador (PRD) competence, April *** Δ López Obrador (PRD) competence, April-July *** Peña Nieto (PRI) competence, April ** *** Δ Peña Nieto (PRI) competence, April-July *** *** Retrospective evaluations, April Δ Retrospective evaluations, April-July ** Drug war policy preferences, April Δ Drug war policy preferences, April-July Vázquez Mota probability of victory, April Constant Multinomial regression models using weights to adjust for demographics and panel-related attrition. Models use Taylor-linearized variance estimation. N = 724. Vázquez Mota is the excluded category. * p<.1, ** p <.05; *** p <.01, two-tailed tests.

19 Table A-6.4. Types of campaign effects in presidential elections, Panel Wave 1 vote intention Consistent with pre-campaign dispositions Inconsistent with pre-campaign dispositions Undecided July vote choice Consistent with Inconsistent with pre-campaign dispositions pre-campaign dispositions Reinforcement (home) Conversion away Conversion home Reinforcement (away) Activation Partial conversion

20 Vote Choice Models for the 2012 Elections As described in the text, the models for 2012 use an alternative vote choice variable to correct for bandwagoning (false reports of voting for the winner) and weighting to adjust for sampling procedures and panel-related attrition. If voters bandwagon and provide false reports of which candidate they supported on election day, then we need an alternative measure of vote choice. I use feeling thermometers of the three main candidates. When one candidate is strictly preferred to the other two, the voter is said to have voted for the preferred candidate. When two candidates are preferred over a third and the two candidates are rated equally, or when all three candidates are rated equally, I break ties using ratings of the candidates traits. The weighting scheme proceeds in two steps. To adjust for sampling, I weight by gender. Wave 1 includes the same proportion of women as in the total population but the voter registration rolls are weighted toward men. So I adjusted based on voter rolls. In the second step, I adjusted for panel-related attrition by running the logistic regression model below where the dependent variable is responding to Wave 2 or failing to respond. The weight is the inverse of the probability of responding to Wave 2 as recovered from the model. I then multiplied the sampling weight by the attrition weight (assuming independence) to arrive at the final weight. Table A-6.5. Panel response model Variable Coef SE Sig Education *** Female *** SES *** April vote intention for Peña Nieto ** Constant Logistic regression. N=998. Dependent variable is coded as responding to Wave 2 of the panel survey (1) or falling out of the panel (0). Explanatory variables for each Mexican state represented in the panel study are included but not shown. * p<.1, ** p <.05; *** p <.01, twotailed tests.

21 Figure A-7.1. Three party configuration of competition in Mexico 2012 elections

22 Figure A-7.2. Cartogram of Mexico 2012 elections

23 Figure A-7.3. Moran I and plots of the spatial autocorrelation in dependent variables

24 Table A-8.1: Baseline summary statistics and orthogonality tests, list experiment Means and Standard Deviations Treatment Control Orthogonality tests Women (0.49) (0.49) Age (15.91) (15.58) Catholic (0.38) (0.39) Religiosity (1.19) (1.19) Std. Poverty Index (1.06) (1.09) Education: None (0.18) (0.19) Unfinished Primary (0.31) (0.31) Primary (0.38) (0.37) Unfinished Sec (0.20) (0.22) Secondary (0.44) (0.44) Unfinished high (0.23) (0.24) High (0.34) (0.35) Unfinished college (0.23) (0.23) College (0.31) (0.29) Orthogonality test 0.98 Notes: This table reports baseline summary statistics for the list experiment. Column (1) presents the means (and standard deviations in parentheses) for the treatment group, while column (2) reports the same summary statistics for the control group. The last column presents the p-values of an F-test from a regression of each baseline characteristic on treatment. The last row in each panel shows the p-values of an F-test from regressions of treatment on baseline covariates.