Abstract
The analysis of voter transitions is an important area of electoral studies. A main strategy is to use aggregate data provided by the offices of statistics regarding districts, precincts, communities etc. and to rely on ecological inference. Ecological inference, however, is plagued by the well-known indeterminacy problem. In this article, we present the so far most extensive systematic empirical comparison of commonly used approaches for ecological inference of the analysis of voter transitions. Our evaluation is based on diverse simulations for multiple assumptions and scenarios. Based on recent election data for the German metropolitan city Munich, we are able to show that an application of the hierarchical multinomial-Dirichlet model, which is implemented in the R-library eiPack, exhibits the best overall estimation performance. Other prominent approaches frequently used by practitioners, e.g. the Thomsen logit approach, proved to be inconsistent. Furthermore, we demonstrate that appropriate data preprocessing is crucial for achieving reliable results.
Similar content being viewed by others
References
Achen, C.H., Shively, P.W.: Cross-Level Inference. University Of Chicago Press, Chicago (1995)
Ambühl, M.: Methoden zur Rekonstruktion von Wählerströmen aus Aggregatdaten. Bundesamt für Statistik (BFS), Neuchâtel (2003)
Andreadis, I., Chadjipadelis, T.: A method for the estimation of voter transition rates. J. Elections Public Opin. Parties 19(2), 203–218 (2009)
Cho, W.K.T.: Iff the assumption fits..: a comment on the king ecological inference solution. Political Anal. 7(1), 143–163 (1998)
Cho, W.K.T., Manski, C.F.: Cross-Level/Ecological Inference. In: Box-Steffensmeier, J.M., Brady, H.E., Collier, D. (eds.) The Oxford Handbook of Political Methodology, pp. 547–569. Oxford University Press, New York (2008)
Deming, W.E., Stephan, F.F.: On a least squares adjustment of a sampled frequency table when the expected marginal totals are known. Ann. Math. Stat. 11(4), 427–444 (1940)
Duncan, O.D., Davis, B.: An alternative to ecological correlation. Am. Soc. Rev. 18(6), 665–666 (1953)
Elff, M., Geschwend, T., Johnston, R.J.: Ignoramus, ignorabimus? On uncertainty in ecological inference. Political Anal. 16(1), 70–92 (2008)
Goodman, L.A.: Ecological regression and the behavior of individuals. Am. Soc. Rev. 18(6), 663–664 (1953)
Greiner, D.J., Quinn, K.M.: RxC ecological inference: bounds, correlations, flexibility and transparency of assumptions. J. Royal Stat. Soc. A 172(1), 67–81 (2009)
Grofman, B., Merrill, S.: Ecological Regression and Ecological Inference. In: King, G., Rosen, O., Tanner, M.A. (eds.) Ecological Inference: New Methodological Strategies, pp. 123–143. Cambridge University Press, Cambridge (2004)
Kellermann, T.: Vom Wahlergebnis zur Wählerwanderung: Welche Wähler wechselten wie ihre Entscheidung. Stadtforsch. Stat. 2011(1), 34–40 (2011)
King, G.: A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data. Princeton University Press, Princeton (1997)
King, G., Rosen, O., Tanner, M.A.: Binomial-beta hierarchical models for ecological inference. Soc. Methods Res. 28(1), 61–90 (1999)
King, G., Rosen, O., Tanner, M.A. (eds.): Ecological Inference: New Methodological Strategies. Cambridge University Press, Cambridge (2004)
Lau, O., Moore, R.T., Kellermann, M.: eiPack: RxC ecological inference and higher-dimension data management. R News 7(2), 43–47 (2007)
Lau, O., Moore, R.T., Kellermann, M.: eipack: Ecological inference and higher-dimension data management, 2012. http://CRAN.R-project.org/package=eiPack. R package version 0.1-7
Park, W.-H.: Ecological inference and aggregate analysis of elections. Ph.D. thesis, The University of Michigan (2008)
R Core Team: A language and environment for statistical computing, 2013. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org
Robinson, W.S.: Ecological correlations and the behavior of individuals. Am. Soc. Rev. 15(3), 351–357 (1950)
Rosen, O., Jiang, W., King, G., Tanner, M.A.: Bayesian and frequentist inference for ecological inference: the RxC case. Stat. Neerl. 55(2), 134–156 (2001)
Schoen, H.: Wählerwandel und Wechselwahl: Eine vergleichende Untersuchung. Westdeutscher Verlag, Wiesbaden (2003)
SORA. SORA Wählerstromanalysen für die Nationalratswahl 2013 in Österreich, 2013. http://www.sora.at/themen/wahlverhalten/wahlanalysen/waehlerstromanalysen.html. (19.03.2014)
Statistisches Amt der Landeshauptstadt München. Amtliches Endergebnis der Bundestagswahl am 18.09.2005 Erst- und Zweitstimmenergebnisse der Landeshauptstadt München, 2005. http://www.mstatistik-muenchen.de/themen/wahlen/wahlberichterstattung/btw2005/index.html. (08.06.2015)
Statistisches Amt der Landeshauptstadt München. Amtliches Endergebnis der Bundestagswahl am 27.09.2009 Erst- und Zweitstimmenergebnisse der Landeshauptstadt München, 2009. http://www.mstatistik-muenchen.de/themen/wahlen/wahlberichterstattung/btw2009/index.html. (08.06.2015)
Statistisches Amt der Landeshauptstadt München. Amtliches Endergebnis der Landtagswahl am 15.09.2013 Erst-, Zweit- und Gesamtstimmenergebnisse der Landeshauptstadt München, 2013. http://www.muenchen.de/rathaus/Stadtinfos/Statistik/Wahlen/Landtagswahl.html. (20.03.2014)
Statistisches Amt der Landeshauptstadt München. Die Umzüge, Zuzüge, Wegzüge und Wanderungssalden nach Stadtbezirken 2013, 2014. http://www.muenchen.de/rathaus/Stadtinfos/Statistik/Bev-lkerung/Wanderungen.html. (04.11.2014)
Thomsen, Søren R.: Danish elections, 1920–79: a logit approach to ecological analysis and inferencea. Politica, Aarhus (1987)
Thomsen, S.R., Frandsen, A.G., Kristmar, T., Lauristsen, P., Sørensen, M.B.: Ecol: Version 3, 1995. Aarhus University
van Elsas, E.J., Lubbe, R., van der Meer, T.W.G., van der Brug, W.: Vote recall: a panel study on the mechanisms that explain vote recall inconsistency. Int. J. Public Opin. Res. (2013)
Wakefield, J.: Ecological inference for 2x2 tables. J. Royal Stat. Soc. A 167(3), 385–445 (2004)
Acknowledgments
We would like to thank Uta Thien-Seitz and Sibel Aydemir for helpful discussions during the project. We are also grateful for the valuable comments of the anonymous referees.
Author information
Authors and Affiliations
Corresponding author
Additional information
For the analysis we received funds from the city of Munich, Department of Statistics.
Appendix
Appendix
1.1 Model description Greiner and Quinn
As distributions at the district level for the Greiner and Quinn (2009) model several multinomial distributions are assumed:
Equation (14) presents the assumed multinomial distribution for one row. Instead of the margins like in the Rosen et al. (2001) proposal, the inner cells, e.g. \((\mathrm{CSU}_{1},\mathrm{CSU}_{2})_{i}\), of the table are considered and assumed to be dependent from the parameters
and the corresponding row-margins \(\mathrm{CSU}_{2,i}\). For each of the R-rows, one independent multinomial distribution is assumed. To hold the restrictions defined by the margins, indicator functions for the row and column sums are used.
For the second level of the models, the parameters are logit transformed using a reference column (party). If the non-voters would be used as reference party, the transformed \(\beta \)-parameters from (14) would have the following form.
The parameters of each multinomial distribution (row) are transformed. Therefore, the \(R\times C\) \(\beta \)-parameters are reduced to \(Rx(C-1)\) \(\omega \)-parameters. As common distribution for the district \(\omega \)-parameters, a normal distribution is assumed at the second level.
As hyperpriors for the \(\mu \) a normal distribution and for the \(\varSigma \) an inverse wishart distribution are proposed by Greiner and Quinn (2009).
1.2 Description of the 2005 and 2009 Federal elections in Munich
See Fig. 4.
1.3 Additional model evaluation results
Rights and permissions
About this article
Cite this article
Klima, A., Thurner, P.W., Molnar, C. et al. Estimation of voter transitions based on ecological inference: an empirical assessment of different approaches. AStA Adv Stat Anal 100, 133–159 (2016). https://doi.org/10.1007/s10182-015-0254-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10182-015-0254-8