Maisanta april2009 final

 News & Politics

of 32
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
1. The Price of Political Opposition:Evidence from Venezuela’s Maisanta*Chang-Tai HsiehUniversity of ChicagoDaniel OrtegaInstituto de Estudios Superiores de…
  • 1. The Price of Political Opposition:Evidence from Venezuela’s Maisanta*Chang-Tai HsiehUniversity of ChicagoDaniel OrtegaInstituto de Estudios Superiores de AdministracionEdward MiguelUniversity of California, BerkeleyFrancisco RodríguezWesleyan UniversityApril 2009AbstractIn 2004, the Chávez regime in Venezuela distributed the list of several million voters whom hadattempted to remove him from office throughout the government bureaucracy, allegedly to identify andpunish these voters. We match the list of petition signers distributed by the government to householdsurvey respondents to measure the economic effects of being identified as a Chavez political opponent.We find that voters who were identified as Chavez opponents experienced a 5 percent drop in earningsand a 1.5 percentage point drop in employment rates after the voter list was released. A back-of-the-envelope calculation suggests that the loss aggregate TFP from the misallocation of workers across jobswas substantial, on the order of 3 percent of GDP.*We thank Vanessa Chirinos, Francisco de Leon, Dan Hartley, Paulina Oliva, Reyes Rodríguez, and Sarath Sanga for superbresearch assistance. We are grateful to Alberto Alesina, Manuela Angelucci, Ernesto Dal Bo, Raj Chetty, Richard Grossman,Ricardo Hausmann, Caroline Hoxby, Pete Klenow, Dale Mortensen, Rohini Pande, Ernesto Schargrodsky, Cameron Shelton,and numerous seminar participants for helpful comments. Hsieh thanks the University of Chicago’s Initiative for GlobalMarkets and Miguel thanks the U.C. Berkeley’s Committee on Research for financial support.
  • 2. 1“1. IntroductionOver an 18 month period starting in late 2002, more than 4.7 million Venezuelans signed one or more ofthe three petitions calling for a vote to remove President Chávez from office. After two failed petitiondrives, a third petition in December 2003 was successful in forcing a recall election that took place inAugust 2004. After Chavez won the recall vote, the list of the signers of the last petition was packagedinto user-friendly software program known as Maisanta. There were soon widespread allegations thatthe Maisanta software was widely distributed throughout the public sector and used by the Chavezregime as an “enemies” list. Jatar (2006), for example, presents the stories of several individuals wholost their jobs after being identified in the Maisanta database as a Chavez opponent.This paper looks for systematic evidence that the Maisanta database was used by the Chavezregime to identify and punish the voters who had attempted to remove Chavez from office, using theMaisanta database itself in the analysis. The information in Maisanta has sufficient detail to match two-thirds of the adults in the Venezuelan national household survey to the voter lists. Using this data, wemeasure whether voters who signed petitions to recall Chavez experienced changes in income oremployment after the Maisanta lists were widely distributed.Figure 1 presents our key evidence from this data. The top panel plots the employment rate ofthe petition signers relative to that of the non-signers and the bottom panel plots their relative wage.Relative employment of the Chavez opposition was roughly constant from 1997 through 2004 beforefalling by 1.5 percentage points in 2005 and 2006. Similarly, the wage gap between the Chavezopposition and the non-signers was roughly constant until 2004, and then dropped by 5 percent in 2005-2006. The fact that there were no trends in either the employment rate or the wage prior to 2004
  • 3. 2suggests that individuals who signed a petition to recall Chavez did not do so as a reaction to worseningemployment and income after Chavez became President in 1999.This paper builds on the growing literature on the effect of political ties and conflict on economicoutcomes.1What is different about the setting we study is that political information was collected onand allegedly used to punish a large share of the population and not just political opposition leaders.Granted, the Chavez regime is not the only one that is alleged to have collected and used detailed dataon its opponents to further its political goals: witness the role of the Stasi in East Germany or the use ofpersonnel files in Communist China. But what is unique is our ability to match the database used by theChavez regime to a standard household survey, which allows us to provide precise measures of theeconomic price of political opposition for everyday Venezuelans.2. Hugo Chavez’s Venezuela and the Maisanta DatabaseHugo Chávez was elected President of Venezuela in December 1998 with the support of 56 percent ofthe electorate. Chávez capitalized on a widespread perception that Venezuela’s traditional politicalparties were corrupt and partly responsible for Venezuela’s long economic decline: Venezuela’s GDPper worker fell by 32 percent between 1978 and 1998.2Once in office, Chávez sought to remakeVenezuela’s political institutions. One of his first actions was to pass a new Constitution. The newConstitution called for Presidential and Legislative elections. The Presidential election took place inJuly 2000, which Chavez won, this time with nearly 60 percent of votes.1See Fisman (2001), Khwaja and Mian (2006), and Ferguson and Voth (2008) on the effect of political ties on economicoutcomes in Indonesia, Pakistan, and Nazi Germany, respectively. Dunning and Stokes (2007) use a subset of Maisanta toexplore political affiliation and the receipt of government social programs in Venezuela See Hirshleifer (1991) Skaperdas(1992), Alesina and Rodrik (1994), Benabou (2004), Londregan and Poole (1990), and Alesina et al. (1996) for papers on theeffect of political conflict on economic outcomes. The implications of punishments meted out to political opponents werediscussed in Kuran’s (1995) study of preference falsification, but empirical applications have been hampered by limited dataon individuals’ public political expression.2Calculations from Rodríguez (2004).
  • 4. 3Chavez used his newfound authority to pass 49 laws, including a controversial land reform billand a law that increased the taxes paid by the state-owned oil company. Venezuela’s main business andlabor organizations initiated public protests and a series of one-day national strikes to pressure Chavezto reverse course. These protests culminated in a violent confrontation with government supporters onApril 11, 2002. On that day, several high-ranking military officers announced that they had disobeyedChávez’s order to repress the opposition demonstrators and that they had removed Chavez from power.However, two days after his removal from office, Chavez was reinstated as President by his supportersin the military.Opposition groups continued to push for Chavez’s removal. They organized a nationwide strikein December 2002 that brought the economy to a standstill for two months. They also pursued a newstrategy of submitting petitions calling for a vote to remove Chavez from office.3In November 2002,opposition groups collected almost 1.6 million signatures (out of 12 million registered voters) calling fora non-binding referendum (a "Consultivo") on Chavez’s rule. The petition was accepted by theElectoral Council, but its decision was overturned by the Supreme Court with the argument that theElectoral Council had not been legally constituted. The Supreme Court then appointed a new ElectoralCouncil with a pro-government majority.Opposition groups responded by organizing a nationwide drive to collect signatures for anotherpetition. Over one day in February 2003 (the “Firmazo”), 2.8 million voters signed a petition calling fora binding vote to remove Chavez. However, because Venezuela’s Constitution stipulated that a petitionfor a binding recall vote can be scheduled only after half of the official’s term was over, the oppositionwaited until the midpoint of Chavez’s term (August 2003) to submit the petition. This second petition3The ability to petition for recall elections, if backed by enough signatures, was a novel feature of the 1999 Constitution. Forrevoking specific laws or on “matters of national interest” the threshold was 10% of voters; for a constitutional amendment,15%; and to recall an elected official, 20%.
  • 5. 4was rejected by the Electoral Council with the argument that the petition was signed before the midpointof Chavez’s term.The opposition launched a third petition drive, this time under new rules set up by the ElectoralCouncil, in which the petition signing process was directly supervised by the Electoral Council. TheElectoral Council set up 2,700 signing stations between November 28 and December 1, 2003, and voterswho wished to sign a recall petition had to show up at a signing station between these dates. This time,nearly 3.5 million voters signed yet another petition (the "Reafirmazo") calling for a binding vote toremove Chávez.The Electoral Council ruled that 375 thousand signatures were invalid and that it could not verifythe authenticity of an additional 1.2 million signatures. The voters whose signatures could not beverified had the option of appearing on May 28 to May 31 2004 at the offices of the Electoral Council toresign the petition. Over 50 percent of these voters showed up, which pushed the total number of validsignatures over the legal threshold for a recall vote (20 percent of registered voters). After 18 months ofpolitical battle, the recall referendum was finally held on August 15, 2004, which Chavez won with over59 percent of the vote.Throughout the political struggle over the recall, Chavez supporters made it clear that supportersof the recall would be publicly identified. Two months after the first petition, pro-government legislatorLuis Tascón posted the list of signers on his website. 4This website was updated with the identity of thesigners of the second and third petitions. Similar lists appeared on the website of the Electoral Council.The Chavez government also actively attempted to dissuade voters from signing the recallpetitions. For example, an advertisement (Exhibit 1 in the Appendix) entitled “Retira Tu Firma”(Withdraw Your Signature) that appeared in several newspapers in October 2003 states:4Tascón’s stated reason was to allow citizens to find out whether their signature had been forged by the opposition. SeeHernández, “MVR Asegura que 72 dirigentes opositores no firmaron solicitud,” El Universal January 15, 2003.
  • 6. 5“40% of the signatures presented by the Anti-Chavez Groups were fake...they used the IDs ofyour dead relatives, non-registered voters, the elderly, and maybe even your ID....If your ID,your friends ID, or an ID of your relative was used: YOU MUST WITHDRAW THESIGNATURE. If you signed under pressure or regret having signed: WITHDRAW YOURSIGNATURE.....Look for your signature on the lists of your voting center or on the followingwebsite" (emphasis in original)Chavez and his supporters also made explicit threats of retaliation against the petition signers. In anationally televised address on October 17, 2003, President Chavez said:"Whoever signs against Chávez… their name will be there, registered for history, because they’llhave to put down their first name, their last name, their signature, their identity card number, andtheir fingerprint.” 5A billboard on the streets of Caracas conveyed a similar message, stating: "Your Vote is Secret, YourSignature is Not." (Exhibit 2 in the Appendix).In the spring of 2004, the list of signers of the third petition was compiled into a user-friendlycomputer program that became known as "Maisanta." This program is a database of all registered votersas of March 2004 (a total of 12,394,109 voters). Exhibit 3 (in the Appendix) illustrates the informationprovided by this software. After a persons identity card number is entered (on the upper left hand sideof the screen), the entry immediately to the right of the ID indicates whether the individual signed thethird petition. Maisanta does not indicate whether the signature was challenged by the ElectoralCouncil, not does it provide information on whether the individual signed the first or the second petition.The entries in the next two rows provide information on the individuals name, birth date, and address.Finally, the bottom of the screen indicates whether the individual participated in several of thegovernment’s social programs. This last set of information makes clear that the creators of the Maisantasoftware had merged the Electoral Council’s list of voters with administrative data from thegovernment’s social programs.5“El que firme contra Chávez está firmando contra la patria,” El Universal, Oct. 17 2003. See also Ciudadanía Activa (2006).
  • 7. 6The list of the signers of all three petitions was removed from Tascons and the ElectoralCouncils websites after the August 2004 recall vote. At the same time however, the Maisanta softwarecontaining the list of voters who had signed the third petition was widely distributed throughout thepublic sector after Chavez won the recall vote. Since the identity of signers of all three petitions waspublic information at some point, the list of signers of all three petitions is the broadest definition ofwhom the Chavez regime might have considered as their opponent. However, since the widelydistributed Maisanta software only contains the list of signers of the third petition, these individualsmight have been more readily identified as political opponents by the Chavez regime after 2004.3. DataThe Maisanta database provides the list of all registered voters in Venezuela in 2003 and the list of allthe signers of the third petition. We also obtained the list of signers of the first two petitions (which waspublicly available from Tascon’s website before August 2004), which we match to the list of voters inMaisanta.Maisanta identifies the municipality and the parroquia (a small geographic unit containing anaverage of 25,000 inhabitants) of the voting center of all registered voters. Maisanta does not identifythe voter’s gender, so we impute gender from the voters name.6The combination of voting center, birthdate, and (imputed) gender uniquely identifies about 7 million individuals in Maisanta. In addition,there are 3 million voters where all the individuals with the same voting center, gender, and date of birthsigned the petitions in the same way. Including this second group of voters, we end up with a sample of10 million voters, or about 80 percent of all the registered voters, whose signing choices we can identify.We match these 10 million voters in Maisanta to the Venezuelan Household Survey (Encuesta deHogares por Muestreo) collected by Venezuela’s National Institute of Statistics. The household survey6We were able to confidently assign gender to 87% of individuals in Maisanta using lists of common first names.
  • 8. 7provides standard labor market and demographic information for a nationally representative sample. Weuse the survey waves from the first semester of 1997 to the first semester of 2006. Although theHousehold Survey is supposed to track families twice a year over three years (for a total of sixsemesters), we find that the attrition rates in the data are extraordinarily high, at 41% across threesemesters and 90% across all six semesters a household is meant to be retained in the panel. We optedto ignore the panel dimension of the data and only use the data as a repeated cross-section.The household survey provides information on each individuals municipality and parroquia ofresidence, as well as their gender and birth date. These variables uniquely identify 97% of theindividuals in the household survey. After matching this sample from the household survey to thesample of 10 million voters in Maisanta, we obtain a final sample of 145,937 individuals. Because thismatching strategy relies on the likelihood that there will be few people with the exact same birth dateand gender within a given parroquia, and because this probability varies depending on parroquiapopulation, the fraction of successful matches to the household survey varies by parroquia size. Toretain sample representativeness, we reweighted each observation in the final matched sample by thereciprocal of the match success rate calculated as the ratio of the matched population to the totalpopulation over age 18 in each parroquia.7Table 1 presents the number of petition signers in the Maisanta database (rows 1 and 2) and inour matched household data (row 3). We categorize petition signers in the following manner: those whosigned at least one petition (column 1); individuals who only signed petitions 1 or 2 (column 2);individuals who only signed petition 3 (column 3); and voters who signed petition 3 and at least one ofthe first two petitions (column 4). The Table shows that 29 percent of all voters signed at least one ofthe three petitions (column 1) and nearly 20 percent signed the third and decisive petition (columns 37The data appendix contains further discussion.
  • 9. 8and 4). We will primarily focus on voters who signed at least one petition (column 1) because theidentity of all petition signers was available to the government. Therefore, unless otherwise indicated,we define the term "petition signers" as voters who signed at least one of the three petitions. However,since the Maisanta program only contains the list of voters who signed the third petition, we also focusat times on voters who signed the third petition.Table 2 compares the labor market characteristics for petition signers ("opposition") with votersthat did not sign any of the recall petitions. The first column presents the mean of the sample (bothsigners and non-signers), the second column presents the difference in means between individuals whosigned any of the three petitions and the non-signers, and the third column presents the differencebetween signers of the third petition and the non-signers. The sample is restricted to individuals in thelabor force and to observations prior to 2002 to exclude any effect of the petition signing. The tableshows that petition signers have higher incomes than non-signers, by 9.5 percent (row 1) and similaremployment rates (row 2). Part of the higher income can be "attributed" to the fact that a larger share ofthe signers are employed in the formal sector (both public and private), by 3.4 percentage points (rows 3,4 and 5). Signers are also likely to be older (1.3 years; row 6), more educated (0.78 more years ofschooling; row 7), more likely to be female (row 8), and more likely to live in Caracas (row 9). Finally,the third column suggests that there is little difference in terms of the observables between the signers ofthe third petition and voters who signed any of the three petitions.4. Employment and Wage EffectsThis section looks for evidence that the petition signers suffered from lower employment and wagesafter Chavez prevailed in the recall election. Before we present the empirical evidence, it is useful to
  • 10. 9think about what a comparison of the employment and wages of petition signers versus non-signersmeasures. Suppose that voter (indexed by i) expected utility from signing or not signing a petition is:( )–SIGN Ci i iU T Y Pπ= +The utility gain from signing is the sum of their political distaste for Chavez Ti , the expected incomechange in the event of a Chavez victory CiYπ , and expected punishment from being identified as aChavez opponent Pπ− (where π denotes the probability of a Chavez victory). The income change inthe event Chavez was defeated in the recall election is normalized to zero. In turn, the expected utilityfrom not signing isNOT SIGN Ci iU Yπ=Note that the cost of being publicly identified as a Chavez opponent (P) is contingent on signing apetition while the expected inco
  • Related Search
    We Need Your Support
    Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

    Thanks to everyone for your continued support.

    No, Thanks