Thesis Summary Title: Split-Ticket Voting in Mixed-Member Electoral Systems: A Theoretical and Methodological Investigation. Author: Carolina Plescia (carolina.plescia@univie.ac.at) Supervisors: Prof. Michael Marsh and Prof. Kenneth Benoit A thesis submitted for the degree of Doctor of Philosophy, Trinity College Dublin, Department of Political Science, 2013. Mixed-member electoral systems give people the opportunity of voting for the same representative body by casting two votes: one for a national party under proportional rules and one for a local candidate under majoritarian rules. 1 Today, at least 20% of voters vote under mixed-member electoral systems (Carter and Farrell, 2010). In those countries where scholarly investigations exist, findings show that many voters split their ticket voting for a party and a candidate that are not linked. Moreover the levels of split-ticket voting have often been found to have an impact on the electoral outcome (Bawn, 1999; Karp et al., 2002; Benoit et al., 2006). Despite several explanations exist today, although still much debated, our knowledge of how people vote under mixed systems is still limited. This is due to methodological as well as substantive issues. Concerning the substantive issues, mixed systems have attracted attention in the past primarily because they afford an opportunity for voters to engage in strategic behavior. While the question of strategic behavior is important, up to this point the literature on split-ticket voting has largely been concerned with this particular strategic moment. Moreover most of the studies are country-based and no comparative research exists. This thesis challenges the current understanding of strategic voting in split-ticket voting and develops an original account. Also, the thesis places split-ticket voting - and mixed systems more generally - in a wider framework 1 In some mixed systems, voters receive two separate ballot papers, one with a list of candidates and one with a list of parties (e.g. Hungary, Japan, Russia, Scotland); in some other cases voters find the parties and candidates listed on the same ballot paper (e.g. Germany, New Zealand, Venezuela). 1
from within which to understand electoral behavior. The second contribution of this thesis is methodological. The secret ballots hinders the possibility of measuring split-ticket voting directly and both the use of surveys and aggregate data is not satisfactory; the reliance of the literature on exclusively one type of data leads often to contradictory findings. The thesis reviews the use of both individual and aggregate data in the existing literature, identifies gaps and limitations and proposes improved alternatives. This thesis consists of a purely methodological analysis based on the New Zealand and Scottish parliamentary elections, a comparative study across 10 mixed systems, and it also looks in depth at case studies of Japan and Italy. The methodological analysis explains in detail why both individual and aggregate data are needed to measure split-ticket voting and how to improve current uses. This is done using the New Zealand and Scottish case studies for which, exceptionally, actual levels of split-ticket voting are available, thus, inference using both individual and aggregate data can be empirically evaluated. When it comes to aggregate data, the methodological analysis tests the estimation methods proposed by Rosen et al. (2001), Greiner and Quinn (2009) and the Maximum Entropy originally applied to electoral data by Johnston and Hay (1982). The methodological chapter also reviews the use of individual-level data in the literature and introduces new ways of using surveys to analyse split-ticket voting. The discussion shows ways of corroborating current usage by employing additional matrix formats, and through the merging of aggregate-level district features to individual-level information. A comparative investigation is then conducted to address the fundamental limitation with the current literature in that it only provides country-based analyses. The comparative study applies an original approach based on the idea of examining the two votes separately rather than in conjunction as it is the practice in the existing literature. Scholars generally measure splitticket voting as a deviation from a straight vote assuming that voters like a single party. Moreover the party vote, because it is cast under proportional rules, is thought to mirror voters preferences better than the candidate vote cast under majoritarian rules (Pappi and Thurner, 2002; Gschwend, 2007). The aim of the comparative analysis is to relax these assumptions and assess whether or not substantial differences exist between the two votes when it comes to the influence of sincere, strategic and institutional factors. Some support is found for the standard argument that a vote under proportional rules is more sincere than the one under majoritarian rules. Despite this, the evidence indicates that the two votes are used by voters to express a more 2
nuanced electoral choice which cannot be accounted for by using a simple strategic versus sincere voting explanation. The Japanese case allows us to investigate further the question of sincere and strategic voting because of the availability in this context of detailed survey questions able to capture the rationale behind vote choice. The analysis also focuses on the separation of sincere versus strategic preferences controlling for candidates availability on the majoritarian ballot. This is an innovation on previous analyses, which focus exclusively on voter preferences assuming that these are all available on the ballot paper. This assumption is problematic because, under mixed rules, there are always fewer candidates than parties running for elections. The Japanese findings suggest that split-ticket voting is rarely a consequence of strategic behaviour as the current literature holds confirming findings from the comparative chapter. Often, when not forced by the unavailability of the candidate on the majoritarian ballot, the split depends on two sincere preferences: one for a party and one for a candidate that just happens to run for another party. The peculiarity of the Italian case permits us to explore additional aspects of split-ticket voting. First, the Italian regional elections use a unique mixed system which allows pre-electoral coalitions to be displayed on the electoral ballot. Second, each region is characterized by similar but not identical mixed rules. Furthermore, Italy is a peculiar setting in which socio-geographical variation has always been directly linked to voting behaviour (e.g. Pasquino, 1995). These features allow a first comprehensive analysis of the influence of long-term cleavages when compared to the impact of short-term factors such as party appeal, candidate availability and districts features. The analysis shows that the electoral rules and the electoral supply play only a limited role in explaining split-ticket voting. With regard to long-term features instead, they strongly influence the way people vote. The general conclusion to be derived from the work conducted in this thesis is that voters act upon the electoral competition and the availability of candidates and parties on the electoral ballot. Despite this however, sincere voter preferences explain most of the variation. Contrary to conventional wisdom, split-ticket voting is not always or even mostly a strategic phenomenon but rather represents, in most cases, the expression of sincere preferences. The broader picture is the one of a simple voter with sophisticated preferences. Mixed systems allows voters to express a more nuanced electoral choice and from the analysis conducted in this thesis, mixed systems appear to be the best of both worlds (Shugart and Wattenberg, 2001) for voters; at least 3
for those who have sophisticated preferences. Another important result regards the rationale of the voting act. Despite vote choice under mixed systems being complicated, there is no evidence supporting the conclusion that ticket-splitting is a consequence of voters confusion with the electoral rules. Remarkable similarities are found across countries and some interesting differences across types of mixed systems and level of experience with the electoral rules. The work conducted in this thesis raises some intriguing issues which merit further investigation moving beyond mixed systems. To start with, the thesis indicates that vote choice cannot be accounted for by using a simple strategic versus sincere voting explanation. The broader picture suggests a simple voter with relatively sophisticated preferences. The proposed line of investigation should consider to what extent these sophisticated preferences are indeed strategic and to what extent voters are party-centred and/or candidate-centred when compared to coalition-centred while paying close attention to policymaking institutions in each country. The thesis has raised also some interesting questions about the link between voters and elected officials when voters are asked to vote for candidates as opposed to when they are asked to vote for parties. This can potentially provide the basis for a future line of investigation that tries to establish the degree of congruence between voters and elected officials and how different this congruence is when people vote directly for parties rather than for candidates. Despite congruence between voters and parties being very important for representative democracy (Powell, 2009; Golder and Stramski, 2010), little is known today about whether the link between voters and elected officials is different across countries adopting diverse electoral rules. References Bawn, K. (1999), Voter responses to electoral complexity: Ticket splitting, rational voters and representation in the Federal Republic of Germany, British Journal of Political Science, 29(3): 487-505. Benoit, K., Giannetti, D. and Laver, M. (2006), Voter strategies with restricted choice menus, British Journal of Political Science, 36(3): 459-485. Carter, E. and Farrell, D. (2010) Electoral Systems and Election Management, in LeDuc, L., Niemi, R. and Norris, P. (2010), Comparing democracies. Elections and voting in the 21st century, Oxford University Press. 4
Golder, M. and Stramski, J. (2010), Ideological congruence and electoral institutions, American Journal of Political Science, 54(1): 90-106. Greiner, J. D. and Quinn, K. (2009), RXC ecological inference: bounds, correlations, flexibility and transparency of assumptions, Journal of the Royal Statistical Society: Series A (Statistics in Society), 172(1): 67-81. Gschwend, T. (2007), Ticket-splitting and strategic voting under mixed electoral rules: Evidence from Germany, European Journal of Political Research, 46(1): 1-23. Johnston, R. J. and Hay, A. M. (1982), On the parameters of uniform swing in singlemember constituency electoral systems, Environment and Planning A, 14(1): 61-74. Pappi, F. U. and Thurner, P. W. (2002), Electoral behaviour in a two-vote system: Incentives for ticket splitting in German Bundestag elections, European Journal of Political Research, 41(2): 207-232. Pasquino, G. (1995), La politica italiana: dizionario critico, 1945-95, Editori Laterza. Powell, G. B. (2009), The ideological congruence controversy the impact of alternative measures, data, and time periods on the effects of election rules, Comparative Political Studies, 42(12): 1475-1497. Rosen, O., Jiang, W., King, G. and Tanner, M. (2001), Bayesian and frequentist inference for ecological inference: The RXC case, Statistica Neerlandica, 55(2): 134-156. Shugart, M. and Wattenberg, M. (2001), Mixed-member electoral systems: the best of both worlds?, Oxford University Press. 5