By R. Michael Alvarez, Francisco Cantu, and Sebastian Saiegh
In anticipation of Argentina’s mid-term elections to be held on Sunday, 27 October 2013, Political Analysis co-editor R. Michael Alvarez (Caltech) discussed some of the most important things that we need to know about this contest with Francisco Cantu (University of Houston) and Sebastian Saiegh (UCSD), authors of “Fraudulent Democracy? An Analysis of Argentina’s Infamous Decade Using Supervised Machine Learning.”
Your paper uses Argentina as a laboratory for studying whether social scientists can detect election fraud. Why Argentina?
During the decade of 1930s in Argentina, elections were marked by two patterns. First, despite no significant differences in turnout rates, elections in that period showed dramatic shifts in electoral support for the Radical and Conservative parties. Second, the period has a checkered history of electoral fraud, which identified this period in Argentine politics as the “infamous decade.” The motivation for our article was to provide empirical evidence for the manipulation allegations made by the opposition at that time and to identify the specific elections contaminated by fraud.
What are the implications of your paper for the upcoming elections in Argentina?
As Alston and Gallo (2010) argue, the widespread fraud during the “infamous decade” in Argentina marked the precedent for the current party system and the lack of checks and balances in the country. Regarding the specific implication of our paper, we provide evidence for a specific type of fraud that the opposition denounced: the informal disenfranchisement of opposition voters by brute force and intimidation and the replacement of those votes with ballots supporting the incumbent. Given the modern conditions to monitor elections and that modern malpractices in competitive elections appear in a decentralized and scattered way, such as the municipal elections in Japan or the gubernatorial elections in Mexico, that type of blatant manipulation is unlikely to appear nowadays. However, as long as it is possible to depict a model for the specific type of fraud that may occur during the forthcoming elections, the method we propose can be adjusted to test modern forms of electoral fraud, as the recent research in Colombia and Zambia.
What are the most important issues regarding the transparency of the national elections in Argentina?
Argentina’s voting system is quite outdated. Citizens vote with slips of paper that carry the names only of a given party’s candidates, like the coupon ballots used in the nineteenth-century United States. Each political party prints, distributes and supplies its own ballots during Election Day. This voting “technology” allows party operatives, to potentially condition any benefits (or punishments) targeted to individual voters on their receipt of the party printed ballot. It also implies that political parties often require thousands of poll-watchers to avoid ballot theft and to monitor the count in each of the polling stations. Therefore, the existing system tends to favor incumbents as well as political parties with territorial outreach and resources, namely president Cristina Fernández de Kirchner’s Front for Victory (Frente para la Victoria, FPV), the Peronist Party (Partido Justicialista, PJ) and the Radical Civic Union (Unión Cívica Radical, UCR).
How can the management of the country’s elections be improved?
The capacity to monitor individual vote choice could be reduced by the adoption of the so-called Australian ballot (AB). The recent experiences of Cordoba and Santa Fe, two provinces that experimented with the AB seem quite promising. Researchers at the Observatorio Electoral Argentino of the Center for the Implementation of Public Policies Promoting Equity (CIPPEC) found that voters were highly satisfied with the AB. The introduction of electronic voting could also make voting in Argentina a simpler and more pristine affair. Nonetheless, there are reasons to believe that one should be cautious in this respect. A recent study by Alvarez et al. (2013) measured the causal effect of replacing traditional voting technology with e-voting on the voting experience using data from an e-voting ﬁeld experiment in Salta, the only Argentinean province where 100% of the registered voters are expected to cast an electronic ballot in the upcoming October elections. The authors found that while e-voters perceived the new technology as easier to use, it also raised some concerns about ballot secrecy. In addition, Katz et. al (2011) found that alternative voting technologies may favor some parties to the detriment of others and could lead to changes in election results. Therefore the adoption of new voting systems should take these effects into account when evaluating the implementation of different technologies.
What are the most interesting things that we might learn from this month’s elections in Argentina?
President Cristina Fernandez de Kirchner’s Frente para la Victoria is one of the few parties with a large enough territorial organization to field candidates everywhere. As such, it is the favorite to carry the nationwide vote. Nonetheless, the results of the Open, Compulsory and Simultaneous Primaries (PASO) held on 11 August 2013 indicate that her party’s candidates will be defeated in five critical districts: the City of Buenos Aires, the province of Buenos Aires, Cordoba, Santa Fe, and Mendoza. The victorious leaders in these districts will be very well positioned to contend the presidential candidacies of their respective parties for the 2015 elections: Mauricio Macri (PRO), Elisa Carrio (UNEN), Sergio Massa (Frente Renovador), Jose Manuel de la Sota (Partido Justicialista), Hermes Binner (Frente Amplio Progresista), Julio Cobos (Union Civica Radical). In the case of the incumbent party, the most likely presidential candidates to emerge from this contest will be Daniel Scioli and Sergio Urribarri. The former is the governor of Buenos Aires, a province that concentrates 37.3% of the country’s registered voters. So, despite being the runner-up in his district, he could still stake a claim on the FPV’s mantle. The latter, is the governor of Entre Rios. If his chosen candidates repeat their performance in the August primaries, he could become the preferred choice of president Kirchner, who is term-limited and has never been too enthusiastic about Daniel Scioli.
Finally, you begin your paper asking, “How can we distinguish an electoral landslide from a stolen election?” In a nutshell, what’s the answer?
As Beber and Scacco (2009) suggest, “the devil is in the digits.” Assessing the integrity of an election by only looking at the final outcome is a challenging task, so we propose to analyze the electoral results at a more disaggregated level and to think in which way fraud should affect particular features of the vote tallies; in our case, we focused on the distribution of the first significant digits of the vote counts. If the particular type of fraud has an expected effect on the electoral results, then it is possible to simulate vote tallies with and without electoral manipulation to then assess whether the real data resemble a clean or a manipulated election. The extension of our method to other elections only relies on the plausibility of modeling the way in which fraud perpetrators cheat.
Dr. Francisco Cantu is an Assistant Professor in the Political Science Department at the University of Houston. Sebastian Saiegh is an Associate Professor in the Political Science Department at the University of California, San Diego. He is currently a Visiting Scholar at the Research Department of the Inter-American Development Bank. They are the co-authors of “Fraudulent Democracy? An Analysis of Argentina’s Infamous Decade Using Supervised Machine Learning” (available to read for free for a limited time) in Political Analysis.
Political Analysis chronicles the exciting developments in the field of political methodology, with contributions to empirical and methodological scholarship outside the diffuse borders of political science. It is published on behalf of The Society for Political Methodology and the Political Methodology Section of the American Political Science Association. Political Analysis is ranked #5 out of 157 journals in Political Science by 5-year impact factor, according to the 2012 ISI Journal Citation Reports. Like Political Analysis on Facebook and follow @PolAnalysis on Twitter.