Abstract
Studying the development of stable political attitudes, political scientists have argued that repeated voting for a political party reinforces initial party preferences, in a seemingly mechanistic process of habit-formation. However, the empirical evidence is scarce and the theoretical framework underdeveloped. Does the act of voting for a party improve an individual’s evaluation of this party? If so, is this effect simply due to habit-formation, or a more complex psychological mechanism? Drawing on cognitive dissonance theory, we examine the act of voting as a choice inducing dissonance reduction. We go beyond existing research, by focusing on tactical voters—a group for which the notion of habitual reinforcement does not predict an effect. The analyses reveal a positive effect of the act of voting tactically on the preferences for the parties voted for and may thus call for a revision of the traditional understanding of the role of voting in shaping party preferences.
This is a preview of subscription content, access via your institution.
Notes
The findings from these early studies provided some evidence, albeit contested (Jackson 1975), about the role of vote choice on change in party identification. Later work on the dynamics of party identification, however, challenged the idea that partisan identities respond to such short-term factors, suggesting that this seemingly volatile picture stemmed from measurement error that attenuated short-term shifts in partisan orientations (Green and Palmquist 1990, 1994). Experimental evidence on the endurance of party identification against new information about parties’ ideological positions provided further evidence to the Michigan view of partisanship as a stable predisposition toward the political parties (Cowden and McDermott 2000).
See Stone (2000) for a review of both critical and supportive essays related to the book.
Cognitive dissonance has frequently been used to explain rationalization bias in studies of voting behavior, for example with regard to economic voting (Anderson et al. 2004; Evans and Andersen 2006; Anderson 2007) and partisans’ placements of parties on issues where they disagree (Eiser 1992; Sherif and Hovland 1961), but see van der Brug (2001) for a null-finding on the latter topic. Plumb (1986) uses cognitive dissonance to explain bias in voters’ statements regarding the timing of their changing parties before the election, while (Groenendyk, forthcoming) argues partisans counter new negative information about their party by depicting the other party in an even more negative light, thus justifying their alignment.
Another theory of attitudinal change, balance theory, could also be of interest here, because it anticipates that people try to maintain balance between their attitudes and preferences. We do not refer to this theory in the main text because it does not directly relate to the impact of behavior. Balance theory is a more general theory of how people update their perceptions so as to preserve balance in light of new information about the stimulus of interest (Heider 1958).
In our view, there are at least two other situations in which some sort of discomfort can be expected to emerge in the context of democratic elections: first, in a two-round majority runoff system, when the party/candidate preferred by the voter is eliminated in the first round; and second, in all types of electoral systems, when the party/candidate preferred by the voter strategically abstains from entering the race because they have zero chances of obtaining a seat/winning the election. Compared to these two groups, tactical voters are preferable because it is easier to rationalize—and thus creates less of a need to reduce dissonance—a vote for a party different than one’s most preferred party when the latter does not participate in the election than when it takes part but bears little chance of winning in the constituency.
The single-member district and plurality rule used in Britain is often seen as the archetypal context within which tactical voting is likely to flourish. Given that only one candidate is elected from each district under this system, voters often have incentives to switch from their first preferences to candidates who have better chances of beating a less favored candidate. Although the exact way to measure this kind of voting behavior is still relatively disputed in the existing literature (compare for example Niemi et al. 1993 with Fisher 2004), aggregate-level studies on the basis of constituency vote returns (Shively 1970; Spafford 1972; Johnston 1981; Curtice and Steed 1988; Johnston and Pattie 1991; Evans et al. 1998; Kiewiet 2009) and studies using survey data (Cain 1978; Heath et al. 1991; Evans 1994; Alvarez and Nagler 2000; Fisher 2004; Fisher and Curtice 2006) alike have established that tactical voting is a stable feature of British elections.
In 2001, the election was held on June 7, the pre-election wave was conducted from March 3 to May 14, and the post-election wave started immediately after the election, on June 8, and lasted until July 30. In 2005, the election was held on May 5, the pre-election wave was conducted from February to April 12, and again the post election wave started immediately after the election, on May 6, and lasted until July 4.
The other alternatives are: “The party had the best policies,” and “The party had the best leader.”
See Fisher (2004) for a theoretical and empirical justification of this measurement strategy.
This may, however, involve a risk of excluding the voters experiencing the most cognitive dissonance. If a voter initially prefers a small third party, but ends up voting for a one of two larger parties, cognitive dissonance may be so strong as to make this voter believe she even preferred the chosen party before the election. If she thus fails to state that hers was a tactical vote, we will not identify her as a tactical voter, and not register this effect. Another potential problem is that individuals may over-report having voted for the winner, but it seems less likely this would influence our results, and, in both cases, the potential bias would most likely serve to conceal rather than exaggerate the effect we seek to identify.
We have reversed the number assigned to these alternatives, however, in order to ease interpretation.
The alternative would be dropping this variable from the analysis, which we find less attractive. The new indicator is 1 if either of the two initial indicators are 1, and 0 otherwise.
Age is included as a continuous measure; for the gender dummy, 1 is male; education is coded as follows: four dummies have been constructed corresponding to the following categories: finished education at the age of 16 or lower; finished education at the age of 17 or 18; finished education at the age of 19 or higher; still in education. The first of these categories serves as the reference category in the parametric analysis.
The non-voters have been identified by the question: “Talking with people about the general election […], we have found that a lot of people didn’t manage to vote. How about you, did you manage to vote in the general election?’’
The predictions for both plots are based on the regression of tactical voting versus non-voting, because regular voters are so similar to tactical voters, a regression based on them hardly yields any significant estimates. This means that the plotted probabilities of tactical voting in the plot on the right are higher than they would have been if they were predicted based on a regression including sincere voters (which also makes the two plots more comparable), but the important point to notice is the balance between tactical and sincere voters, which would be evident regardless of which regression was used to produce the predictions.
Further, to make the variable work in a logical manner, when tactical voters vote for a party that is tied with another (say, both were rank 1), the tactically voted preference is ranked below the other (in this case, 2), as the fact that the voter sees the vote as tactical implies the other party was actually preferred. According to a similar logic, for sincere voters, when the party voted for is tied, the other party is ranked below the party voted for, as the vote implies that this other party was less attractive.
To further ensure that we focus on the right preferences, we also exclude from the analysis preferences that are zero (minimum) before the election. Such preferences indicate that parties are “strongly disliked,’’ which means they are extremely unlikely to form the basis of a vote, whether tactical or non-tactical. In other words, few or none of the treated preferences are expected to be of this kind, while some are to be expected among the potential control preferences. This would be a source of imbalance, and as such preferences would be of little interest in the analysis, they are excluded from both the treated and the control group. We also exclude preferences that are ten (maximum) before the election. The reason it is possible for treated preferences to be ten is that we have allowed tactical voters to vote for a party tied with their first preference. However, such tactical voters are not the most interesting ones, especially when they vote for a party for which they express a maximum preference before the election—suggesting that they might as well consider their vote non-tactical. Thus, again having as our motivation to maximize balance—there is no point in which the dependent variable has been taken into any consideration thus far—we focus our analysis on relevant preferences, excluding from both the treated and the control group those that have a value of ten before the election.
We could use only one party preference for each individual. However, to supply more cases that could be used as controls in the matching procedure, in our estimation presented in the next section, we have allowed each individual in the group of sincere voters to be represented with more than one party*individual combination. For instance, a sincere voter may supply two observations in the control group, one related to her second and one related to her third preference. Again, the motivation for this decision is to facilitate the process of finding balance between the control and the treated groups in terms our set of observed pre-treatment characteristics. We have also tested the sensitivity of our results, by only selecting one party*individual combination for each individual in our control group is chosen. The effects are almost identical to those presented here.
Genetic matching is a generalization of propensity score and Mahalanobis distance matching that uses an evolutionary search algorithm to determine the weight given to each baseline covariate (Sekhon and Mebane 1998). Genetic matching has shown to outperform in terms of Mean Squared Error alternative matching methods both when the Equal Percent Bias Reduction assumption holds and when it does not. In the latter case, it also performs much better in terms of bias (Diamond and Sekhon 2008).
This test has some desirable properties: it culminates in a single test statistic and p value; it is based on an χ2 approximation which seems to work well with small samples; and it appraises balance not only on the set of covariates listed in Fig. 3, but also on all linear combinations of them (Hansen and Bowers 2008). To implement the test we used the Randomization Inference tools package (RItools) in R (Hansen and Bowers 2008).
We estimate the LATE here rather than the ATE because we essentially have focused our analysis on a subset of units whose potential outcomes are affected by the treatment assignment mechanism, i.e. voted preferences of tactical voters and similar, but non-voted preferences of sincere voters. In other words, we estimate a LATE, defined by the selection of our treatment and control groups. This is analogous to a regression discontinuity (RD) design, although we do not use such a design here. In RD, only observations around a given threshold are examined because cases far from this cut-off point are deemed to differ from those near the threshold in various observable and unobservable ways. Here, this applies for sincere voters’ party preferences, among which we only select those most similar to the treated preferences.
Note, however, that the LATE we identify is further complicated by the fact that we use the matched dataset, which already focuses the analysis on a particular subset of the population.
The LARF estimation has been implemented by using Abadie’s cls.m function for linear response in Matlab: http://www.hks.harvard.edu/fs/aabadie/cls.m.
To be more precise, let us summarize the estimation procedure by following the notation of Hansen and Bowers (2009). Let r i denote constituency i’s observed proportion of party preferences higher than 5 and let C denote the group of non-marginal constituencies, so that r i denotes the same proportion in the absence of marginal constituencies. The observed outcome can be written \(\sum\nolimits_{U}r_{i}\), where U denotes all constituencies, while the proportion of high party preferences one would extrapolate only from the group of non-marginal seats can be written \(\sum\nolimits_{i\in U}\hat{r}_{ci}(\hat{\beta})\), where \({\hat{\beta}}\) represents the logistic regression coefficients from the logistic regression of the outcome on the covariates among the control group. The final estimate comes from the difference between \(\sum\nolimits_{U}r_{i}\) and \(\sum\nolimits_{i\in U}\hat{r}_{ci}(\hat{\beta})\). The 95 % confidence bands result from a calculation of the variance of \(\bar{\hat{r_{C}}}\), i.e. \(\hat{V}(\bar{r}{}_{C})=(1-n/N)s^{2}[r_{i}:i\in C]/n\), where n = |C|, N = |U| and \(s^{2}[(r_{1},\ldots,r_{J}])=(J-1)^{-1}\sum\nolimits_{1}^{J}(r_{j}-\bar{r})^{2} \).
References
Abadie, A. (2003). Semiparametric instrumental variable estimation of treatment response models. Journal of Econometrics, 113(2), 231–263.
Abadie, A., & Imbens, G. (2006). Large sample properties of matching estimators for average treatment effects. Econometrica, 74(1), 235–267.
Alvarez, R. M., & Nagler, J. (2000). A new approach for modeling strategic voting in multiparty elections. British Journal of Political Science, 30, 57–75.
Anderson, C. J., Mendes, S. M., & Tverdova, Y. V. (2004). Endogenous economic voting: Evidence from the 1997 British Election. Electoral Studies, 23(4), 683–708.
Anderson, C. J. (2007). The end of economic voting? Contingency dilemmas and the limits of democratic accountability. Annual Review of Political Science, 10, 271–296.
Angrist, J. D., Imbens, G., & Rubin, D. (1996). Identification of causal effects using instrumental variables. Journal of the American Statistical Association, 91(434), 444–472.
Aronson, E., Fried, C., & Stone, J. (1991). Overcoming denial and increasing the intention to use condoms through the induction of hypocrisy. American Journal of Public Health, 81(12), 1636–1638.
Arulampalam, W., Booth, A. L., & Taylor, M. P. (2000). Unemployment persistence. Oxford Economic Papers, 52(1), 24–50.
Bem, D. (1972). Self-Perception Theory. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 6, pp. 1–62). New York: Academic Press.
Besley, R., & Joslyn, M. (2001). Cognitive dissonance and post-decision attitude change in six presidential elections. Political Psychology, 22(3), 521–540.
Brader, T., & Tucker, J. A. (2001). The emergence of mass partisanship in Russia, 1993–1996. American Journal of Political Science, 45(1), 69–83.
Brehm, J. W. (1956). Post-decision changes in desirability of alternatives. Journal of Abnormal and Social Psychology, 52(3), 384–389.
Brody, R. A. (1991). Stability and change in party identification: Presidential to off-years. In P. M. Snidman, R. Brody, & P. E. Tetlock & (Eds.), Reasoning and choice: Explorations in political psychology (pp. 179–205). Cambridge: Cambridge University Press.
Butler, D., & Stokes, D. E. (1974). Political change in Britain (2nd edn.). London: MacMillan.
Cain, B. (1978). Strategic voting in Britain. American Journal of Political Science, 22(3), 639–655.
Campbell, A., Converse, P. E., Miller, W. E., & Stokes D. E. (1960). The American Voter. New York: John Wiley & Sons, Inc.
Converse, P. E. (1969). Of time and partisan stability. Comparative Political Studies, 2(2), 139–171.
Converse, P. E. (1976). The dynamics of party support: Cohort-analyzing party identification. Beverly Hills: Sage Publications.
Cowden, J. A., & McDermott, R. M. (2000). Short-term forces and partisanship. Political Behavior, 22(3), 197–222.
Curtice, J., & Steed, M. (1988). Appendix 2: analysis. In D. Butler & D. Kavanaugh (Eds.), The British General Election of 1987 (pp. 316–362). New York: St. Martin’s Press.
Daniel, S. R., Armstrong, K., Silber, J. H., & Rosenbaum, P. R. (2008). An algorithm for optimal tapered matching, with application to disparities in survival. Journal of Computational and Graphical Statistics, 17(4), 914–924.
Diamond, A., & Sekhon, J. S. (2008). Genetic matching for estimating causal effects: A general multivariate matching method for achieving balance in observational studies, working paper. Berkeley: University of California.
Dobson, D., & St Angelo, D. (1975). Party identification and the floating vote: Some dynamics. American Political Science Review, 69(2), 481–490.
Egan, L. C., Santos, L. R., & Bloom, P. (2007). The origins of cognitive dissonance: Evidence from children and monkeys. Psychological Science, 18(11), 978–983.
Eggers, A. C., & Hainmueller, J. (2009). MPs for sale? Returns to office in postwar British politics. American Political Science Review, 103(4), 513–533.
Eiser, R. J. (1992). Attitudinal judgement: The to and fro of assimilation-contrast. In D. Granberg & G. Sarup (Eds.), Social judgement and intergroup relations. Essays in honor of Muzafer Sherif (pp. 129–146). New York: Springer.
Evans, G. A. (1994). Tactical voting and Labour’s prospects. In A. Heath, R Jowell, & J. Curtice & (Eds.), Labour’s last chance?: The 1992 Election and beyond (pp. 65–84). Aldershot: Dartmouth.
Evans, G. A., & Andersen, R. (2006). The political conditioning of economic perceptions. Journal of Politics, 68(1), 194–207.
Evans, G. A., Curtice, J., & Norris, P. (1998). New Labour, new tactical voting? The causes and consequences of tactical voting in the 1997 General Election. In D. Denver, J. Fisher, P. Cowley, C. Pattie (Eds.), British Elections and Parties Review vol. 8: The 1997 General Election (pp. 65–79). London: Frank Cass.
Evans, G. A., & Heath, A. F. (1993). A tactical error in the analysis of tactical voting: A response to Niemi, Whitten and Franklin. British Journal of Political Science, 23(1), 131–137.
Festinger, L., Riecken, H. W., & Scachter, S. (1956). When prophecy fails. Minneapolis: University of Minnesota Press.
Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203–210.
Festinger, L. (1957). A Theory of cognitive dissonance. Evanston: Row, Peterson.
Fiorina, M. P. (1981). Retrospective voting in American Elections. New Haven: Yale University Press.
Fisher, S. D. (2004). Definition and measurement of tactical voting: The role of rational choice. British Journal of Political Science, 34, 152–166.
Fisher, S. D., & Curtice, J. (2006). Tactical unwind? Changes in party preference structure and tactical voting in Britain between 2001 and 2005. Journal of Elections, Public Opinion and Parties, 16(1), 55–76.
Franklin, C. (1984). Issue preferences, socialization and the evolution of party identification. American Journal of Political Science, 28(3), 478–479.
Franklin, C., & Jackson, J. E. (1983). The Dynamics of party identification. American Political Science Review, 77(4), 957–973.
Gerber, A. S., & Green, D. P. (2000). The effects of canvassing, telephone calls, and direct mail on voter turnout: A field experiment. American Political Science Review, 94(3), 653–663.
Gerber, A. S., & Green, D. P. (2005). Correction to Gerber and Green (2000), replication of disputed findings, and reply to Imai (2005). American Political Science Review, 99(2), 301–313.
Green, D. P., & Palmquist, B. (1994). How stable is party identification?. Political Behavior, 16(4), 437–466.
Green, D. P., & Palmquist, B. (1990). Of artifacts and partisan instability. American Journal of Political Science, 34(3), 872–902.
Grimmer, J., Hersh, E., Feinstein, B., & Carpenter, D. (2011). Are close elections random? Unpublished Manuscript. http://www.stanford.edu/∼jgrimmer/CEF.pdf.
Groenendyk, E. (forthcoming). Justifying party identification: A case of identifying with the “lesser of two evils”. Political Behavior.
Grofman, B., Wayman, F., & Barreto, M. (2009). Rethinking partisanship: Some thoughts on a unified theory. In J. Bartle & P. Bellucci (Eds.), Political parties and partisanship: Social identity and individual attitudes (pp. 60–73). Routledge: Ruhr University Bochum.
Hansen, B. B., & Bowers, J. (2008). Covariate balance in simple, stratified and clustered comparative studies. Statistical Science, 23(2), 219.
Hansen, B. B., & Bowers, J. (2009). Attributing effects to a cluster-randomized get-out-the-vote campaign. Journal of the American Statistical Association, 104(487), 873–885.
Heath, A., Jowell, R., Curtice, J., Evans, G. A., Field, J., & Witherspoon, S. (1991). Understanding political change: The British voter 1964–1987. Oxford: Pergamon Press.
Heider, F. (1958). The psychology of interpersonal relations. New York: John Wiley Sons.
Howell, S. (1980). The behavioral component of changing partisanship. American Politics Quarterly, 8(3), 279–302.
Howell, S. (1981). Short term forces and changing partisanship. Political Behavior, 3(2), 163–180.
Huber, G., Gerber, A., & Sempolinski, J. (2009, 3–6 September). Using ballot order to test for cognitive dissonance: Results from a natural experiment. Paper presented at the 2009 APSA Annual Meeting, Toronto.
Imai, K. (2005). Do get-out-the-vote calls reduce turnout? The importance of statistical methods for field experiments. American Political Science Review, 99(2), 283–300.
Jackson, J. E. (1975). Issues, party choices, and presidential votes. American Journal of Political Science, 19(2), 161–185.
Johnston, R. (1981). Regional variations in British voting trends—1966-1979: Tests of an ecological model. Regional Studies, 15(1), 23–32.
Johnston, R., & Pattie, C. (1991). Tactical voting in Britain in 1983 and 1987: An alternative approach. British Journal of Political Science, 21, 95–128.
Kern, H. L., Hainmueller, J. (2009). Opium for the masses: How foreign media can stabilize authoritarian regimes. Political Analysis, 17(4), 377–399.
Kiewiet, R. (2009, 28–30 August). The ecology of tactical voting in Britain, 1983–2005. Paper presented at the 2009 EPOP conference, University of Strathclyde, Glasgow.
Knoke, D. (1976). Change and continuity in American politics: The social basis of political parties. Baltimore: Johns Hopkins University Press.
Larsen, R. J., & Buss, D. M. (2005). Personality psychology: Domains of knowledge about human nature. Boston: McGraw Hill.
Lu, B., & Rosenbaum, P. R. (2004). Optimal pair matching with two control groups. Journal of Computational and Graphical Statistics, 13(2), 422–434.
Markus, G. B., & Converse, P. E. (1979). A dynamic simultaneous equation model of electoral choice. American Political Science Review, 73(4), 1055–1070.
McPhee, W., & Ferguson, J. (1962). Political immunization. In W. N. McPhee & W. A. Glaser (Eds.), Public opinion and congressional elections (pp. 155–179). Glencoe: The Free Press.
Meredith, M. (2009). Persistence in political participation. Quarterly Journal of Political Science, 4(3), 186–208.
Mullainathan, S., & Washington, E. (2009). Sticking with your vote: Cognitive dissonance and political attitudes. Applied Economics, 1(1), 86–111.
Niemi, R., Whitten, G., & Franklin, M. N. (1993). People who live in glass houses. A response to Evans and Heath’s critique of our note on tactical voting. British Journal of Political Science, 23, 549–563.
Plumb, E. (1986). Validation of voter recall: Time of electoral decision making. Political Behavior, 8(4), 301–312.
Robinson, G., McNulty, & J. E., Krasno, J. S. (2009). Observing the counterfactual? The search for political experiments in nature. Political Analysis, 17(4), 341–357.
Rosenbaum, P. R. (2010). Design of observational studies. New York: Springer.
Sekhon, J. S., & Mebane, W. R. Jr. (1998). Genetic optimization using derivatives: Theory and application to nonlinear models. Political Analysis, 7(1), 189–203.
Sherif, M., & Hovland, C. I. (1961). Social judgement: Assimilation and contrast effects in communication and attitude change. New Haven: Yale University Press.
Shively, W. P. (1970). The elusive ‘psychological’ factor: A test for the impact of electoral systems on voters’ behavior. Comparative Politics, 3(1), 115–125.
Spafford, D. (1972). Electoral systems and voters’ behavior: Comment and a further test. Comparative Politics, 5(1), 129–134.
Stone, J. R. (2000). Introduction. In J. R. Stone (Eds.), Expecting armageddon: Essential readings in failed prophecy (pp. 1–30). New York: Routledge.
Tavits, M. (2005). The development of stable party support: Electoral dynamics in post-communist Europe. American Journal of Political Science, 49(2), 283–298.
Van der Brug, W. (2001). Perceptions, opinions and party Preferences in the face of a real world event: Chernobyl as a natural experiment in political psychology. Journal of Theoretical Politics, 13(1), 53–80.
Zanna, M. P., & Cooper, J. (1974). Dissonance and the pill: An attribution approach to studying the arousal properties of dissonance. Journal of Personality and Social Psychology, 29(5), 703–709.
Acknowledgment
The authors are grateful for comments by three anonymous reviewers, the editors of this journal, Juan A. Mayoral, and the participants of the Elections, Public Opinion and Parties (EPOP) Annual Conference at the University of Essex, September 10–12, 2010. The usual disclaimer applies.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bølstad, J., Dinas, E. & Riera, P. Tactical Voting and Party Preferences: A Test of Cognitive Dissonance Theory. Polit Behav 35, 429–452 (2013). https://doi.org/10.1007/s11109-012-9205-1
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11109-012-9205-1
Keywords
- Party preferences
- Partisanship
- Party identification
- Cognitive dissonance
- Tactical voting
- Genetic matching
- Multiple control groups