54FF70E3d01 - jhvhj,n PDF

Preview

Summary

jhvhj,n...

Description

Comparative Political Studies http://cps.sagepub.com/

Voting for Our Story: A Narrative Model of Electoral Choice in Multiparty Systems Tamir Sheafer, Shaul R. Shenhav and Kenneth Goldstein Comparative Political Studies 2011 44: 313 originally published online 27 October 2010 DOI: 10.1177/0010414010384372 The online version of this article can be found at: http://cps.sagepub.com/content/44/3/313 Published by: http://www.sagepublications.com

Additional services and information for Comparative Political Studies can be found at: Email Alerts: http://cps.sagepub.com/cgi/alerts Subscriptions: http://cps.sagepub.com/subscriptions Reprints:http://www.sagepub.com/journalsReprints.nav Permissions:http://www.sagepub.com/journalsPermissions.nav Citations:http://cps.sagepub.com/content/44/3/313.refs.html

>> Version of Record - Feb 20, 2011 Proof - Oct 27, 2010 What is This?

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

384372 2Sheafer et al.Co mparative Po litical Studies © T he Autho r(s ) 2011 Reprints and permis s io n: http://www. s agepub.co m/jo urnals Permis s io ns .nav

CP 4

10.1177/001041401038437

Voting for Our Story: A Narrative Model of Electoral Choice in Multiparty Systems

Comparative Political Studies 44(3) 313–338 © The Author(s) 2011 Reprints and permission: http://www. sagepub.com/journalsPermissions.nav DOI: 10.1177/0010414010384372 http://cps.sagepub.com

Tamir Sheafer1, Shaul R. Shenhav1, and Kenneth Goldstein2

Abstract Based on narrative approaches, the authors develop an empirical technique to gauge the match between stories told by political actors and voters and assess its effect on voting behavior. Even with other fundamental attitudinal and demographic factors held constant, they hypothesize that voters should prefer parties that hold and communicate similar national narratives. Using data gathered during the 2009 elections to the Israeli Knesset, the authors gauge voter attitudes about fundamental national narratives in Israel and conduct a systematic analysis of parties’ discourse in parliamentary speeches to gauge parties’ stories. Controlling for demographics and ideology in a series of logistic regressions, the authors find that voting behavior for most parties is significantly affected by the narrative proximity between voters’ stories and parties’ stories. Keywords narrative identity, narrative proximity, voting behavior, multiparty system, Israel, national stories 1

Hebrew University of Jerusalem, Israel University of Wisconsin–Madison, USA

2

Corresponding Author: Shaul R. Shenhav, Hebrew University of Jerusalem, Department of Political Science, Room 4322, Mount Scopus, Jerusalem, 91905, Israel Email: [email protected]

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

314

Comparative Political Studies 44(3)

There are various ways of telling national stories. These ways are neither a matter of truism nor a matter of scientific historiography. Rather, they embody both a retrospective and prospective perception of the collective. Consider, for example, two versions of Turkey’s national story. One version rests with the constitutional principles of “the secular Republic” and views Atatürk’s reforms as the main starting point of Turkey’s nationality.1 Another perspective launches the collective story of Turkey with Mohammed and the Koran. In a well-known speech in 1997, Recep Tayyip Erdoğan, the current prime minister of the Republic of Turkey, highlighted these competing visions by pointing to “two fundamentally different camps” in Turkey, “those who follow Atatürk’s reforms [secularists] and the Muslims who unite Islam with Shari’a” (Baykal, 2009, p. 4). Erdoğan made it quite clear where he stood when he cited and espoused the words of the Ottoman Islamist poet Ziya Gőkalp: “The mosques are our barracks, the domes our helmets, the minarets our bayonets and the faithful our soldiers” (Baykal, 2009, p. 5).2 The story of Turkey is told from a different point of view by Kemal Kılıçdaroğlu, leader of the Republican People’s Party (CHP), who promised at his first speech as the CHP leader to follow the footsteps of Atatürk, and by current president of Turkey, Abdullah Gül, who said at his acceptance speech, We, as the Turkish people, are the heirs of both a proud history and a nation that has established many great states and empires that have spanned the globe. . . . The goal, laid forth by the great Atatürk, founder of our Republic, of “raising the Republic of Turkey higher than the level of contemporary civilizations” must always be borne in mind.3 Although these quotes embody the core rift in modern Turkey, there are nuances in these competing narratives. For instance, even though Erdoğan’s national story differs from the secular tradition of modern Turkey, it does not completely echo all tenets of fundamentalist Islam. Erdoğan does not call for the implementation of Sharia and the creation of a Nation of Islam by force. Furthermore, he occasionally stresses the “boundless tolerance culture of this country”4 or “cohabitation between the Christians and Muslims.”5 There are obviously other nuances in Turkey’s narratives and the context of each speech matters. Still understanding national narratives is a rather simple and straightforward way to gauge core political positions, and similar competing narratives can be seen in other countries. For instance, if we were to ask a citizen of Italy to identify the major event in his or her national story, we would likely get a range of answers spanning from ancient Rome through the Renaissance to the Italian unification of the 19th century and

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

315

Sheafer et al.

the establishment of the Italian republic in 1946. Looking to the other side of the Mediterranean, one citizen of Egypt may date the Egyptian story from biblical times whereas another may begin the narrative with the Egyptian Revolution of 1952. Political life is fundamentally about building the future through visions of present and past, and combining visions of history with current politics has always been a basic part of political discourse. Political leaders are attuned to the constant dialogue between founding fathers and the next generations, between heritage and future developments, or between legacy and expectations for a better future. This existential political position makes the narrative mode of thought and talk, which strives to place experiences in time and place (Bruner, 1986), central for any political encounter. It is no wonder then that just as we can learn about personal identity through the stories people tell about themselves, we can understand national perceptions through the stories people tell about their nation. That said, since they always implicitly or explicitly refer to the future of the nation, national narratives are often more than historical accounts of the collective past. Accordingly, the political perception of a person who begins his or her national story with a mythical past and tells the collective future in terms of a “strong nation” will probably differ from a person who begins his or her story with the establishment of the modern state and prefers a future of “civil equality.” Moreover, it follows that both sorts of people will communicate better with and have a greater attachment to others or to a political organization that narrate the national story in a similar way. In this way, we can regard narrative similarity as a potential indicator of the political relatedness between people and political organizations. Here we see the potential causal impact of narrative similarity between voters and parties on voting behavior. Many scholars of elections in a variety of different countries have used theoretical frameworks and found strong empirical evidence that voters cast their ballots on retrospective performance (e.g., Fiorina, 1981). Of course, this work points to fairly immediate retrospection about the state of the economy, the competence of the government in a crisis, or the recent standing of the nation in the world. We have no disagreement with these findings but explore in this article whether “really” retrospective evaluations—going back hundreds or thousands of years—can have an impact on current voting behavior, even when controlling for fundamental ideological factors and recent events. Specifically, this study is the first attempt to apply narrative analysis to the field of voting behavior. Using the case of Israel, we develop an

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

316

Comparative Political Studies 44(3)

empirical measure of narrative similarity between political stories of parties and voters. Even when controlling for the fundamental, familiar, and powerful correlates of voting behavior, such as ideology and sociodemographic factors, we find that narrative similarity between voters and parties explains vote choices. We begin with a discussion of narrative theory, focusing on the concept of narrative identity and the meaning of narrative similarity in the political domain. Next, we present our empirical approach to measuring narrative proximity. We then discuss the potential contribution of the study of narrative proximity to our understanding of and ability to explain voting behavior. Finally, we demonstrate the impact of narrative proximity in the 2009 Israeli elections and discuss our findings in the larger theoretical framework of voting behavior and narrative study.

Narrative Identity and Political Ideology From the dawn of classical thought, the effect of narratives on the human mind has drawn the attention of philosophers and scholars. It was central to Plato’s concerns on the potential danger of stories and myths in his Republic, it was part of Aristotle’s Poetics and Rhetoric, and it played an important role in Quintilian’s Institutio Oratoria. Underpinning its importance for social research, Elliott (2005) pointed to three key features of narratives: “First that they are chronological (they are representations of sequences of events), second, that they are meaningful , and third, that they are inherently social in that they are produced for a special audience” (p. 4). These features, she claims, lead sociologists to understand the importance of the temporal dimension of narratives for making sense of the interrelation between individual lives and social contexts. Scholars of politics have also recognized the importance of narrative in their fields of study (e.g., Buthe, 2002; Cornog, 2004; Fischer, 2003; Fisher, 1985; Hajer, 1995; Ku, 1999; Linde, 2001; McBeth, Shanahan, Arnell, & Hathaway, 2007; McGee & Nelson, 1985; Patterson & Monroe, 1998; Roe, 1994; Shenhav, 2006; Smith & Smith, 2001; Yanow, 2006). Stemming from the basic understanding that human beings are essentially story-telling animals (MacIntyre, 1981) with a natural impulse to narrate (White, 1981), these studies put forward the unique features of the narrative form and its application to the political domain. Narratives, as explained by Patterson and Monroe (1998), “help us understand ourselves as political beings” and therefore become “an invaluable tool in navigating the myriad of sensations that bombard us daily” (p. 315).

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

317

Sheafer et al.

By and large, as suggested by Fischer (2003), the narrative mode is not only a basic form of communication but also a mode of thinking that furnishes communication out of which social and political meanings are constructed. It also allows people to assess social positions in their communities and to internalize social conventions. Fischer conceived of narrative as a distinct “cognitive scheme,” which imposes “coherent interpretation on the whirl of events and actions that surrounds us” (p. 163). Consequently “all of us come to be who we are . . . by being located or locating ourselves (usually unconsciously) in social narratives rarely of our own making” (Somers, 1994, p. 606). This recognition of the role of narratives draws strength from a growing list of work—mostly by psychologists—that examines how people think, perceive, imagine, and execute moral decisions using narrative structures (e.g., Sarbin, 1986; see review in Singer, 2004). The growing understanding of the role of narrative in the human mind resides in the concept of narrative identity (McAdams, 2001; Singer, 2004). This concept became a new subdiscipline of personal psychology (Singer, 2004, p. 437) and has also attracted the attention of other fields in the social sciences (Elliott, 2005; Somers, 1994; Whitebrook, 2001). From a psychological point of view, narrative identity can be seen as “the accumulating knowledge that emerges from reasoning about our narrative memories,” and it “yields a life story schema that provides causal, temporal, and thematic coherence to an overall sense of identity” (Singer, 2004, p. 442; also see Bluck & Habermas, 2001; Ricoeur, 1991). When it comes to the political domain, applying this concept should account for the fact that the use of narratives is undertaken at both the individual level and the aggregate level (Patterson & Monroe, 1998, p. 316). It is therefore possible to analyze the narrative identities of individuals and also the narrative identities of groups and organizations. As a concept defining an overarching outlook at the political world, narrative identity is in line with some of the broad conceptions of political ideology that emphasize the role of ideology in communicating shared beliefs, opinions, and values (Campbell, Converse, Miller, & Stokes, 1960; Converse, 1964; Downs, 1957; Jost, Federico, & Napier, 2009; Knight, 2006; M. Shamir & Arian, 1994; Zaller, 1992). Some of these conceptions of ideology specifically point to the importance of the temporal dimension, which is also central in the definition of narrative identity. For instance, according to Jost et al. (2009), “Ideologies also endeavor to describe or interpret the world as it is—by making assertions or assumptions about human nature, historical events, present realities, and future possibilities [italics added]” (p. 309).

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

318

Comparative Political Studies 44(3)

Still, operationalizations of the concept tend to adopt a more explicit political understanding of ideology and focus on a single left–right dimension (Fuchs & Klingemann, 1989; Huber, 1989; Jost et al., 2009; Knight, 2006). The explanatory power of ideology operationalized this way on voting behavior is very strong (Jost, 2006; M. Shamir & Arian, 1999). However, over the years, scholars have added other dimensions of ideology in addition to the basic left–right continuum. Examples include studies by M. Shamir and Arian (1994), J. Shamir and Shamir (2000), and Caprara, Schwartz, Capanna, Vecchione, and Barbaranelli (2006) that focused on a set of political values instead of on a single ideological dimension. Our model builds on this previous work, seeking to add a dimension of political identity derived from narrative theories to the explanation of voting behavior. We show that narrative proximity between the identities of voters and parties matters, even when accounting for ideology.

Narrative Proximity and Voting Behavior Voting behavior is an ideal context to evaluate the role of political narrative similarity. In this context, we focus on the relationship between voters and parties—with special attention to the causal relationship between the narrative identities of voters and parties when it comes to voting decisions. This expectation regarding the important role of proximity is consistent with research and theory on the role of cultural proximity in the transference of frames. According to this work, when it comes to frames, some have a natural advantage because their ideas and language resonate with the culture of the audience (Benford & Snow, 2000; Entman, 2004, 2008; Frensley & Michaud, 2006; Gamson, 1992; Miller & Riecher, 2001; Pan & Kosicki, 2001; Snow & Benford, 1988; Wolfsfeld, 1997). It is thus fair to assume that in line with the effect of cultural proximity, if the way we “story the world” (Mishler, 1995, p. 117) tells us much about the way we make meanings, then similarities or dissimilarities in the stories people hold about their nationality should influence a wider set of political behaviors such as vote choice. More specifically, the notion of similarity between narratives has been occasionally discussed under the concept of narrative proximity. This notion is typically applied to studies on nurse–patient relations, where “both physical proximity—nearness to the patient’s body—and narrative proximity—understanding the patient’s story”—have been studied (Peter & Liaschenkob, 2004, p. 219). We prefer using the concept of narrative proximity and not narrative similarity for two major reasons. First, as detailed below, our empirical and methodological goal is to measure proximities between

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011

319

Sheafer et al.

voters’ stories and parties’ stories and not merely to account for the dichotomy of similarity versus dissimilarities. Second, we are wary about the ambiguous set of potential parameters that one can utilize when referring to similarity and dissimilarity between stories. Because we employ our analyses on a particular set of narratives, we prefer this more technical term of proximity. Our model focuses on the role of proximity between the stories held by people about their nationality and the national stories held by parties. Similar to the general premise of proximity voting (i.e., voters prefer parties that are closer to them on an issue dimension), the premise of the narrative proximity approach developed here is that voters prefer parties that are closer to them on the dimension of national story or that share their national story to a greater extent. The explanation behind this assumption is that the stories of the nation perceived by people reflect collective national perceptions that are different from explicit political identities or issue preferences. We therefore expect to find narrative proximity between voters and parties to add to our current understanding of voting behavior. Specifically, we hypothesize that even with other fundamental demographic characteristics, political attachment and ideology are accounted for as follows: Hypothesis 1: The greater the proximity between a voter’s national story and a party’s national story, the greater the probability that the voter will cast his or her vote for that party, all else held constant.

Narrative Proximity in a Multiparty System Because of problems associated with modeling multiple alternatives, analysis of voting behavior in multiparty systems is more complicated than empirical analysis of voting behavior in two-party systems (Quinn, Martin, & Whitford, 1999; Whitten & Palmer, 1996). That said, fundamental explanatory factors such as sociodemographics and issue proximity still play a central role in these models. Israel, the empirical focus of this study, is a classic multiparty system with approximately a dozen parties being elected to the Knesset in a typical election. M. Shamir and Arian’s (1999) model of voting behavior in Israel emphasizes the role of collective identity in Israeli electoral competition. Their work controls for sociodemographic variables such as age, gender, education, income, ethnic background, and religiosity and for issue preferences around questions of land for peace. Bloc identification and media priming (Sheafer, 2007; Sheafer & Weimann, 2005), strategic voting (Bargsted & Kedar, 2009; Blais, Aldrich, Indridason, & Levine, 2006), and issue proximity (Bargsted &

Downloaded fromcps.sagepub.comby Hai Do on October 30, 2011