Specialty Grand Challenge ARTICLE
A Research Agenda in Elections and Voting Behavior in a Global and Changing World
- Department of Political and Social Sciences, Universitat Pompeu Fabra, Barcelona, Spain
Electoral democracies have expanded rapidly worldwide in the last decades. According to the dichotomous measure of democracy by Boix et al. (2013), while in 1975 there were 47 democracies in the world (the 30.92 percent of the existing countries), in 2015 the number of democracies rose by 2.5–117 (the 60.62 percent of existing countries). This spread of democracy around the world has been particularly noticeable in Africa: only two countries (Botswana and Gambia) were democracies in 1975, but 16 in 2015. However, elections are not restricted to democracies nowadays. In 2018 111 national elections (Assembly, Head of Government, Head of State, Lower House, Upper House, and Referendum) were held in the world, at least 30 in non-democracies (Election Guide, http://www.electionguide.org) The purpose of elections in non-democracies is, of course, different than in democracies (Gandhi and Lust-Okar, 2009).
Electoral democracy rests on a straightforward idea: citizens elect their fellow citizens to represent their interests. Research on elections and representation has focused on four dimensions of this relationship between those who have the right to elect and those who want to be elected. First, why and how the supply of candidates or parties varies across elections and countries. Before the election occurs, political parties have to decide whether they enter the race alone, engage in some form of pre-electoral coordination or stay out. Second, when parties decide to enter, they have to define their campaign strategies to influence voters' decisions, that is, select policy positions, define the salience of issues, allocate their resources and select their candidates. Third, citizens have to decide whether they vote or not and those who vote have to choose a given party. Finally, citizen preferences are aggregated and converted into seats to form a government. The four dimensions respond to the interaction between institutions, in particular electoral systems, and basic assumptions about the electorate's political abilities—knowledge and interest—and the decision-making process.
In what follow I will highlight some limitations of existing research that constitute in my view the research agenda for the next years and the focus of the Elections and Representation section of Frontiers in Political Science.
1) Emphasis on theory building: The most glaring limitation of the literature on elections and voting behavior is the scarce emphasis on theories. With the advent of the high-speed computer and statistical sampling theory, the use of rich sources of data and quantitative techniques has made possible a huge accumulation of knowledge in the form of verifiable or falsifiable statements. However, at the same time this causal inference based on large-N analyses has fostered the development of a variable-centered type of theory in electoral behavior that only devotes a scant attention to explanatory mechanisms. Unfortunately, case-studies, which are more useful than cross-unit study when insight into causal mechanisms is more important than insight into causal effects (Gerring, 2004, p. 352), have been absent from the major journals in the field. Clearly, I believe that more importance should be given to theory building rather than simply to find reliable effect. The literature is crying out for middle-range theories (Merton, 1967) such as the M+1 rule formulated by Cox (1997)1 when explaining the number of parties we should be expect in a given election, particularly when accounting for voter behavior. I do not intend to under-value empirical, confirmatory research, but explaining is more than identifying causal effects: the causal generative mechanisms have to be provided as well. A deep understanding of how things behaved in the past and how they will behave under new circumstances entails the formulation of causal arguments succesfully traveling across time and space.
2) A global focus: The literature on voting behavior is clearly skewed westward. Even in the literature which may well be the largest in all of political science (Fiorina, 1997, p. 391), empirical research is overwhelmingly focused on Western countries. This western bias is due to the political history of countries, the availability of data, and the production of research almost exclusively in western universities. The democratization waves in Central and Eastern Europe, Asia and Africa have put non-western politics in the spotlight and question whether the body of knowledge generated by studies in voting behavior is highly culturally or geographically specific. For instance, in a seminal analysis of the determinants of the number of parties in African democracies, Mozaffar et al. (2003, p. 387) found that district magnitude, widely acknowledged as the decisive institutional variable in shaping the structure of party systems, has no independent effect on the number of parties. Similarly, elections in non-democracies, more frequent than what it is presumed, deserve more attention in the research agenda in electoral behavior. Do electoral rules matter in non-democracies? How do voters make decisions when elections are not free and fair?
3) New methodological approaches: Most of the classic studies on the consequences of electoral systems and voting behavior have been based on cross-sectional observational data. Democratization waves, on the one hand, and the greater availability of survey data, on the other hand, have allowed increasing the number of observations and countries in the analyses. The big peril in this type of research is the omitted-variable bias and thus spurious relationships. It could be argued that time-series analyses are more compelling than cross-sectional research when it comes to ascertaining the impact of electoral system and studying voter decision making processes. However, there are very few such time-series analyses when examining the role of electoral systems, while panel surveys are limited to a short time span. Longitudinal analyses with multiple waves (years) should be fostered. Additionally, spatial modeling techniques capturing the mechanisms underlying spatial patterns deserve more attention. The next logical approach is to move from observational data to experimental or quasi-experimental designs. There have been a few quasi-experimental studies when exploring how the consequences of electoral systems (for instance, Pellicer and Wegner, 2014) and many more experimental studies, especially in the lab (for instance, Blais et al., 2016). Unsurprisingly when talking about elections, field experiments are scarce. Clearly, the use experimental designs and big data should be strongly encouraged. Finally, let me stress again that a good balance between empirics and theory development in the filed demands the complementarity of single-unit and cross-unit research designs.
4) Interdisciplinary dialogue: Research on elections should be clearly interdisciplinary. The close and long-lasting dialogue between political science and economics has been very fruitful for the field in terms of accumulation of knowledge and using quantitative methods and experimental designs. Yet, the understanding of complex phenomena such as electoral outcomes requires multiple perspectives. We need a closer dialogue with history, demography, and psychology. First, when examining the impact of electoral systems or even the behavior of political parties, particularly with observational data, we should engage in more historically sensitive studies. As there is no path independence, a necessary condition for causal identification, endogeneity is an elephant in the room (Przeworski, 2008). As Kreuzer (2010, p. 385) rightly argues, “it would be beneficial to first do the more nuts-and-bolts work of using historical knowledge to improve the quantitative study of institutional origin, … [to] give us a better understanding of qualitative history's potential ontological incompatibilities with statistical methodologies.” Second, advanced societies have experienced deep changes across the last decades such as the emergence of one-person households whose effect on voting behavior seems unquestionable. The literature on elections and voting behavior cannot be rendered blind to demographic studies. Traditional models of voter choice based on sociological factors and party identification (the Columbia and the Michigan models) are time-specific and do not travel well-across generations. Social influences are changing and explanations of voting behavior should adapt to those changes. Finally, the literature on voting behavior would benefit from greater attention to research in psychology. Given that voting is a complex decision-making process in which many cross-cutting variables are considered, any explanation is ultimately grounded on behavioral assumptions about voters. More psychology would help the field to refine assumptions about how persuasion or framing become effective or to conflate emotions and political cognition.
To conclude, the research on elections and voting behavior we promote from the Elections and Representation section of Frontiers in Political Science is focused on theory building rather than simply to find reliable causal effects. I dream of a research on elections and voting behavior beyond the current the state of the art and open to case studies and time-series, experimental, and bid data, in close dialogue with demography and psychology and not only with economics, and with a diverse and global empirical scope, including both democracies and non-democracies.
The author confirms being the sole contributor of this work and has approved it for publication.
The author acknowledges support from the Spanish Minister of Science, Innovation and Universities (Grant number: AEI/FEDER CSO2017-85024-C2-1-P) and ICREA under the ICREA Academia programme.
Conflict of Interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
1. ^The number of viable parties or candidates in a given district is limited by an upper bound of M+1 (where M is equal to district magnitude).
Mozaffar, S., Scarritt, J. R., and Galaich, G. (2003). Electoral institutions, ethnopolitical cleavages and party systems in africa's emerging democracies. Am. Political Sci. Rev. 97, 379–39. doi: 10.1017/S0003055403000753
Keywords: elections, representation, voting behavior, political parties, theory building, demography
Citation: Lago I (2019) A Research Agenda in Elections and Voting Behavior in a Global and Changing World. Front. Polit. Sci. 1:1. doi: 10.3389/fpos.2019.00001
Received: 06 November 2019; Accepted: 22 November 2019;
Published: 06 December 2019.
Edited and reviewed by: Ferran Martinez i Coma, Griffith University, Australia
Copyright © 2019 Lago. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Ignacio Lago, Ignacio.email@example.com