Abstract
This paper analyzes the influence of alternative voting technologies on electoral outcomes in multi-party systems. Using data from a field experiment conducted during the 2005 legislative election in Argentina, we examine the role of information effects associated with alternative voting devices on the support for the competing parties. We find that differences in the type of information displayed and how it was presented across devices favored some parties to the detriment of others. The impact of voting technologies was found to be larger than in two-party systems, and could lead to changes in election results. We conclude that authorities in countries moving to adopt new voting systems should carefully take the potential partisan advantages induced by different technologies into account when evaluating their implementation.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Although other studies also used experimental designs to assess the impact of different technologies on voter behavior (e.g., Herrnson et al. 2008), none of them focused on their effect on electoral outcomes.
Throughout this paper, the term “heuristics” refers to cognitive shortcuts and simplifying rules of thumb individuals resort to in order to make political judgments or inferences with considerably less than complete information about the different electoral options (Lau and Redlawsk 2006).
A party with over a 100 years of history, the Unión Cívica Radical had traditionally been the main contender of the Partido Justicialista. However, after the collapse of its last (coalition) government in 2001, which pushed Argentina on the verge of institutional chaos, UCR was harshly punished by voters. As a result, it lost ground to new political options like ARI and PRO in subsequent elections, especially in Buenos Aires.
According to the Buenos Aires Controller’s Office (AGCBA 2006), PRO’s campaign expenditures totaled 240,000 U.S. dollars, while ARI and FPV spent about 150,000 dollars each. In comparison, UCR’s expenditures were below 85,000 dollars, while most of the remaining parties spent substantially less than that.
Bielsa was President Kirchner’s Foreign Relations Minister at that time. Carrió was a rising political leader who had been a competitive presidential candidate in 2003, and Macri is a famous businessman and was the president of one of Argentina’s most popular soccer clubs at that time.
We thank an anonymous referee for bringing this point to our attention.
The percentages in Fig. 2 do not take blank ballots into account. The category “Other parties” in Fig. 2 includes 23 parties competing in the election of national representatives and 34 in the election of city legislators. Figures S.1–S.3 in the Supplementary Materials provides further details about the distribution of parties’ vote-shares across prototypes and precincts both in the pilot and in the official election.
For instance, some participants may not have had strong incentives to pay attention during the information session or to avoid voting errors, while others could have changed their choice—e.g., selecting sure losers instead of casting a strategic vote. These problems are virtually unavoidable in mock elections. Nevertheless, survey responses showed that a sizeable proportion of the participants were concerned that the pilot results would be used as an exit poll and took the task of deciding who to vote for quite seriously.
In this sense, Levitt and List (2007) note that, even though experimental results may not translate exactly to real world settings, they can provide relevant information about the underlying mechanisms that may be at work when certain data patterns are observed.
Figure S.4 in the Supplementary Materials depicts the four types of devices tested in the experiment.
In Argentina, parties are assigned an official number which is often used for advertising during the electoral campaign, especially in the case of parties—such as UCR—that have kept the same number over time.
Providing one of the DRE devices with a verifiable paper trail allowed us to test its effect on subjects’ trust in the accuracy of the voting technologies. The rate of positive answers to a survey item asking participants whether they believed that their vote had been correctly recorded was similar for Prototypes 1 (92.1%) and 2 (94.0%), with no statistically significant effect of the paper trail on the perceived reliability of the machines.
Ballots of the same party were stacked on top of each other, and the piles corresponding to the different parties were laid out on the tables next to each other, following the official list numbers.
The differences in the relative influence of alternative voting cues across prototypes are significant at the usual confidence levels (Loughin and Scherer 1998).
See Table S.1 in the Supplementary Materials.
Table S.2 in the Supplementary Materials reports the distribution of subjects by prototypes and precincts.
Table S.3 and Figure S.5 in the Supplementary Materials provide further details on the distribution of the covariates across prototypes.
Table S.4 in the Supplementary Materials summarizes the results from several covariate balance checks.
Section 4 in the Supplementary Materials describes the Dirichlet-multinomial regression model.
See Sect. 5 in the Supplementary Materials for additional details on the matching analysis.
The estimates from the Dirichlet-multinomial model are presented in greater detail in Tables S.5–S.7 of the Supplementary Materials.
In the case of Prototype 2, it is plausible that the stronger incidence of party-centric cues in the less visible election could have been countered by a larger impact of candidate-centric shortcuts in the election for Congress, thus concealing possible variations in the influence of partisan heuristics across races. However, even for the two “purely” party-centric Prototypes 1 and 4, differences in the vote-shares of ARI, FPV and PRO relative to Prototype 3 were not significantly higher in the election of city legislators.
These differences were all significant at the 0.01 level.
These nonparametric estimates are presented in greater detail in Table S.10 of the Supplementary Materials.
A PR-D’Hont formula is used in Buenos Aires to elect both national representatives and city legislators. However, the differences in district magnitudes and the fact that a party has to gather the support of more than 3% of the registered voters to achieve representation in Congress—but not in the city legislature—determines that seat allocation in the national election is less sensitive to marginal changes in relative vote-shares. See Calvo and Micozzi (2005) and the references therein for a description of Argentina’s electoral rules.
For instance, in Brazil, the first country in the world to conduct fully electronic elections, the adoption of new technologies has introduced some changes in parties’ campaign advertising tactics. In particular, the substitution of paper ballots that allowed voters to write the names of their preferred candidates to mark their preferences by voting machines that only allow them to type the candidates’ registration numbers has led parties to heavily publicize these numbers during the electoral season.
References
Alvarez, R. M., & Hall, T. E. (2008). Electronic elections: The perils and promises of digital democracy. Princeton: Princeton University Press.
Alvarez, R. M., Katz, G., & Pomares, J. S. (2009). Up to the promise? The impact of electronic voting on trust in the election process in Latin America. Working Paper 89, Caltech/MIT Voting Technology Project.
Amado, A. (2006). Argentina 2005: Repensar la comunicación política. Diálogo Político, 2, 49–77.
Ansolabehere, S., & Stewart, C., III. (2005). Residual votes attributable to technology. Journal of Politics, 67, 365–389.
Bartels, L. M. (1996). Uninformed votes: Information effects in presidential elections. American Journal of Political Science, 40, 194–230.
Bril, T. (2007). El colapso del sistema partidario de la Ciudad de Buenos Aires. Una herencia de la crisis argentina de 2001–2002. Desarrollo Económico, 47, 367–400.
Buenos Aires Controller’s Office, AGCBA (2006). Informe de Auditoría - Gastos de Campaña 2005. Technical Report, February 2006.
Calvo, E., Escolar, M., & Pomares, J. S. (2009). Ballot design and split ticket voting in multiparty systems: Experimental evidence on information effects and vote choice. Electoral Studies, 28, 218–231.
Calvo, E., & Micozzi, J. P. (2005). The governor’s backyard: A seat-vote model of electoral reform for subnational multiparty races. Journal of Politics, 67, 1050–1074.
Card, D., & Moretti, E. (2007). Does voting technology affect election outcomes? Touch-screen voting and the 2004 presidential election. Review of Economics and Statistics, 89, 660–673.
COPUB, Centro de Opinión Pública de la Universidad de Belgrano. (2005). Elecciones legislativas 2005: valoración que los ciudadanos realizan del proceso electoral. http://www.ub.edu.ar/centros_de_estudio/copub/. Accessed 8 February 2010.
Coan, T. G., Merolla, J. L., Stephenson, L. B., & Zechmeister, E. J. (2008). It’s not easy being green: Minor party labels as heuristic aids. Political Psychology, 29, 389–405.
Guimarães, P., & Lindrooth, R. C. (2007). Controlling for overdispersion in grouped conditional logit models: A computationally simple application of Dirichlet-multinomial regression. Econometrics Journal, 10, 439–452.
Herrnson, P. S., Niemi, R. G., Hanmer, M. J., Bederson, B. B., & Conrad, F. C. (2008). Voting technology: The not-so-simple act of casting a ballot. Washington, DC: Brookings Institution Press.
Herron, M. C., & Wand, J. N. (2007). Assessing partisan bias in voting technology: The case of the 2004 New Hampshire recount. Electoral Studies, 26, 247–261.
Ho, D. E., Imai, K., King, G., & Stuart, E. A. (2007). Matching as nonparametric preprocessing for reducing dependence in parametric causal inference. Political Analysis, 15, 199–238.
Lau, R., & Redlawsk, D. P. (2006). How voters decide. Information processing during election campaigns. New York, NY: Cambridge University Press.
Lechner, M. (2001). Identification and estimation of causal effects of multiple treatments under the conditional independence assumption. In M. Lechner & F. Pfeiffer (Eds.), Econometric evaluation of labour market policies (pp. 43–58). Heidelberg: Physica.
Levitt, S. D., & List, J. A. (2007). What do laboratory experiments measuring social preferences tell us about the real world? Journal of Economic Perspectives, 21, 153–174.
Loughin, T. M., & Scherer, P. N. (1998). Testing for association in contingency tables with multiple column responses. Biometrics, 54, 630–637.
Management Press. (2005). Informe cuantitativo audio-visual. Elecciones 2005. http://www.managementpress.com.ar. Accessed 15 October 2008.
Mebane, W. R., & Sekhon, J. S. (2004). Robust estimation and outlier detection for overdispersed multinomial models of count data. American Journal of Political Science, 48, 391–410.
Mondak, J. J. (1993). Public opinion and heuristic processing of source cues. Political Behavior, 15, 167–192.
Popkin, S. L. (1991). The reasoning voter: Communication and persuasion in presidential campaigns. Chicago: University of Chicago Press.
Reynolds, A., & Steenbergen, M. (2006). How the world votes: The political consequences of ballot design, innovation and manipulation. Electoral Studies, 25, 570–598.
Schaffner, B. F., & Streb, M. J. (2002). The partisan heuristic in low-information elections. Public Opinion Quarterly, 66, 559–581.
Tomz, M., & Van Houweling, R. P. (2003). How does voting equipment affect the racial gap in voided ballots? American Journal of Political Science, 47, 46–60.
Wolak, J. (2009). The consequences of concurrent campaigns for citizen knowledge of congressional candidates. Political Behavior, 31, 211–229.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendix
Appendix
Nonparametric estimates of the pairwise differential prototype effects, in percentage points. The circles are proportional to the average differential effects on the support for each of the parties competing in the two legislative elections, with the solid dots marking those estimates that are significant at the usual confidence levels (0.9 and above). The dashed vertical lines correspond to a 0 percentage point change in the expected vote share of each party
Rights and permissions
About this article
Cite this article
Katz, G., Alvarez, R.M., Calvo, E. et al. Assessing the Impact of Alternative Voting Technologies on Multi-Party Elections: Design Features, Heuristic Processing and Voter Choice. Polit Behav 33, 247–270 (2011). https://doi.org/10.1007/s11109-010-9132-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11109-010-9132-y