Show Summary Details

Page of

date: 19 November 2019

# Issue Voting: Modern and Classic Accounts

## Summary and Keywords

Political scientists have long assumed that issues were at the heart of vote choice and a causal determinant of it—that is, citizens came to politics with clear issue preferences and they voted for the party or candidate that best represented those interests. Many would say democracy demands this kind of link between issue preferences and vote choice. And yet, studies have consistently shown that surprisingly few voters measure up. Many voters know remarkably little about politics, including even basic facts about their own political system. They have little conception of issues, party, and candidate positions on those issues or how issues relate to one another as part of a coherent political ideology. As a result, they often have unstable and ephemeral preferences. Worse, among the fraction of voters who are engaged and well-informed, many appear susceptible to persuasion and possibly manipulation of their issue opinions from the media and their partisan leaders. This raises questions about the viability of representative democracy and, at its most pessimistic, the possibility that elites are largely free to pursue personal goals unchecked and independent of the public good.

Some of the most innovative work on issue voting is focusing on partisan bias, including its limitations and how it relates to other social divisions. When mapped onto existing social divisions in society, such as those arising from race, religion, and immigration, issues can indeed matter for elections because they tap into and stimulate the same psychological and affective processes that make partisanship so powerful in the first place—in-group and out-group bias. Cross-national research has also increasingly pointed to the role of such issues as part of an emerging political cleavage related to globalization that is transforming elections and party systems across Europe and other postindustrial democracies. The causal determinants of issue voting is a promising avenue for future research.

Keywords: issues, ideology, partisanship, voting, elections, immigration

# Introduction

Postmortem accounts of elections written by journalists and pundits often portray campaigns as living or dying on the specific policy positions politicians chose to broadcast to the electorate. If Candidate A hadn’t put forward Policy B that was unpopular with Group X, surely the outcome would have been different! Implied in such accounts is that issues in some way determine vote choice causally. Despite this popular narrative, electoral scholars have yet to fully coalesce around a unified understanding of the role of policy issues during election campaigns. To be sure, a lengthy literature beginning with Anthony Downs (1957) has shown that issues and ideology are often correlated with vote choice. However, more recent work, especially in the American context, has questioned whether voters’ issue positions are exogenous to long-term political variables such as party identification and social identities, and hence, truly causal to vote choice (e.g., Bartels, 2002; Zaller, 1992).

In Europe, there has never been much emphasis on issue voting in this sense. Instead, the issue voting research has been largely on how long-standing social cleavages, such as class, religion, language, and the urban–rural divide, structure the vote and how they are tied to issue opinions (Budge, Crewe, & Farlie, 1976). Starting in the 1980s, there was a long running debate about the weakening of traditional social cleavages and its possibility for increasing issue voting (Dalton, Flanagan, & Beck, 1984). More recently, there is evidence of an emerging cultural cleavage that ties together a range of issues related to globalization (Kriesi et al., 2012; Hellwig, 2014). Still, even here, there has been little effort to disentangle issues from social divisions and cleavages. Thus, much of this article focuses on American literature before integrating comparative studies in its final section.

In the American case especially, spatial models of political competition driven by issue voting have long held a powerful influence over the way political scientists have understood the behavior of voters, political elites, and the nature of representative democracy. Beginning with rational choice models of political competition (Downs, 1957; Key, 1966), the core motivation of party selection was believed to be the relative utility derived from each party’s set of issue positions. Arranged in one- or multidimensional space, voters selected from the represented ideologies that were closest to their ideal point. Further extensions of this work (e.g., Fiorina, 1981) focused on more simple but powerful applications of the basic rational choice framework in the form of retrospective voting—the simple evaluation of the incumbent government’s past performance on valence issues.

Despite the intuitive appeal of this approach, later works began to probe the stability and coherence of voters’ policy attitudes and perceptions themselves (Converse, 1964; Delli Carpini & Keeter, 1996). Most voters, it turned out, have little conception of issues, party issue-staking, or even abstract ideology. Instead, long-run forces such as partisanship effectively guide voters’ decisions on Election Day (e.g., Campbell, Converse, Miller, & Stokes, 1960). Indeed, to the extent that many voters possess attitudes at all, later works focused on the ability of political elites to broadcast issue positions that were subsequently adopted by and motivating to in-group partisans (Lenz, 2012; Zaller, 1992).

While attitude adoption may vary by levels of political sophistication or whether the attitude concerns “core” values, the dominant argument in American political behavior continues to emphasize the endogeneity of issues to the political process and partisanship especially (Achen & Bartels, 2016). Further, decades of work in social psychology and political science suggest that even “objective” political perceptions can be strongly conditioned by voters’ prior attitudes or partisanship (Bartels, 2002; Evans & Pickup, 2010; Lodge & Taber, 2013).

This article reviews the research concerning issue voting with the aim of understanding the sources and consequences of policy attitudes in the mass electorate. Are issue positions endogenous to partisanship? Are all issues for all voters determined by political variables? Can information “shortcuts” and heuristics allow people with limited political information to approximate informed issue voting? How do relatively “set” social and ethnic groupings affect these processes?

# The Concept of Issue Voting

Though often caricatured as putting forth a vision of electoral politics filled with utility maximizers, the underlying argument of rational choice models focuses on how voters’ relatively exogenous issue preferences drive electoral outcomes. This simple insight has led to decades of work investigating the exact relationship between issues and voters’ final decisions on Election Day. At its most crude, the major cleavages in this line of work separate those who believe that citizens choose based on an ordered and relatively “flat” dimension of policy alternatives (the proximity model) and those who suggest voters choose one of two sides of each issue.

The proximity model makes three fundamental assumptions about voters: they have clear exogenous issue preferences; they know the positions of competing parties or candidates on the issues; and they vote for the party closest to their own position. It follows from this that parties should converge on the median voter position to maximize their vote share, always in two-party systems but also in multiparty systems if the distribution of voters is normal—that is, follows a bell-shaped curve (Downs, 1957).

In response to the reality that parties rarely converge on the median voter even in two-party systems, Macdonald and Rabinowitz (1989) developed the directional voting model. They argued voters do not have a very clear sense of parties’ policy positions, at least not sufficiently to make informed, utility-maximizing decisions as Downs imagined. According to the directional model, voters effectively choose “sides” of an issue and maintain the belief with varying levels of intensity. They vote for the party that most clearly signals they are on “their” side. As empirical models, the proximity and directional accounts of issue voting have been widely applied, modified with various linear and quadratic utility-loss functions, and tested against one another (e.g., Macdonald & Rabinowitz, 1989).

# The Issues With Issues: Cognitive Demands

These two models have long dominated the scholarly debate on vote choice, particularly as it relates to position issues and ideology. Issue voting, whether proximity, directional, or otherwise, however, works from the premise that voters have clear issue preferences and they can differentiate parties or candidates on these issues. Some scholars have long questioned this more fundamental assumption, and the drumbeat has grown increasingly louder in recent years that voters are generally ill informed, have few “real” attitudes, and are not very capable of distinguishing between competing parties on different issues.

The seminal studies of political behavior from the 1950s were among the first to raise serious questions about a well-informed electorate (see also Lippmann, 1922). Per Berelson and colleagues (1954, p. 308), “The democratic citizen is expected to be well informed about political affairs. He is supposed to know what the issues are, what their history is, what the relevant facts are, what alternatives are proposed, what the party stands for, what the likely consequences are. By such standards the voter falls short.” The authors of The American Voter reached a similar conclusion: “many people know the existence of few if any of the major issues of policy” (Campbell et al., 1960, p. 170) and “large portions of an electorate do not have meaningful beliefs” regarding a wide range of political issues and debates (Converse, 1964, pp. 243–245).

Further research has only confirmed this. Indeed, it is difficult to overstate how little Americans seem to know about politics. Even basic political facts seem to escape many citizens, including an inability to name just one branch of government or the party affiliation of their state’s governor (Delli Carpini & Keeter, 1996). When applied to voters’ policy attitudes and ideology and their (lack of) familiarity with major party positions on political issues, the outlook for issue voting looks remarkably bleak.

Lacking not only in basic political information, few voters can boast a full ideological or issue-driven palette. In his study of American voters from the 1950s, for example, Converse (1964) found that just 2.5% of the electorate were “ideologues,” meaning they correctly applied a coherent, ideological belief system in the assessment of political issues and objects. Another 9.0% of respondents used ideological concepts, but did not apply them consistently. Similarly, there is precious little evidence that issue positions fit together in any ideologically coherent whole for many voters.

Even worse, a remarkable number of voters fail to take the same position on the same issue at different points over time; that is, they lack stability in their own policy preferences. Using panel surveys, where voters are interviewed multiple times, Converse and Markus (1979) find only moderate correlations between the positions that voters take at different times on the key issues of the economy and foreign affairs. There is greater consistency on issues related to civil rights and race relations, but even this pales in comparison to the stability of partisanship. Further works using more complex empirical models accounting for possible endogeneity find similar results (e.g., Tesler, 2016).

Furthermore, despite rising education levels, the emergence of mass media, and even increased polarization, there is little evidence that this situation has improved in the nearly 60 years since The American Voter was published. As Kinder and Kalmoe (2017, pp. 123–124) summarize in their recent update of Converse’s seminal findings on the United States: “Americans (most Americans)’s . . . opinions are ideologically incoherent: some are liberal, some conservative, and some, when examined closely, seem not to be opinions at all. Most know little about what is happening in politics and even less why.”

# The “Panacea” of Aggregation and Heuristics

Scholars now largely accept that much of the electorate has limited political knowledge and unstable issue and ideology positions; however, not all consider this alarming or inherently problematic for representative democracy. Defenders of representative democracy have long offered two possible mechanisms for how citizens could still hold governments accountable and ensure public policy represents public opinion: “aggregate level rationality” and the use of cognitive heuristics or cues. Both responses take as a starting point the same central assumption that there is a share of the public that is well informed with real preferences and is thus capable of ensuring democratic responsiveness.1 It is important to examine the assumptions and evidence for both, because, as will be shown, far from being a panacea, they suggest that the challenges to representative democracy run deeper than just the electorate’s limited knowledge of politics and issues.

First, proponents of aggregate-level rationality argue that few voters actually need to be well informed or have stable policy preferences for representative democracy to function well. As long as a sufficient number do have real preferences and are assumed to be uninformed, and unstable preferences are randomly distributed and independent, then aggregate public opinion will be both stable and meaningful. In this way, statistical aggregation is thought to purge the more “undesirable” attributes of public opinion, such as random variation and measurement error, so that voters’ policy attitudes collectively still matter (Erikson, MacKuen, & Stimson, 2002; Page & Shapiro, 1992).

In aggregating preferences, the argument goes, a great deal of “error” is effectively cancelled out. As nicely summed by Page and Shapiro (1992, p. 15) before a subheading aptly titled “From Collective Ignorance to Collective Wisdom”:

The simple process of adding together or averaging many individuals’ survey responses . . . tends to cancel out the distorting effects of random errors in the measurement of individuals’ opinions. Similarly, statistical aggregation tends to eliminate the effects of real but effectively random (i.e., offsetting) opinion changes by individuals. And social processes involving division of labor and collective deliberation mean that collective opinion responds—more fully and attentively than most individuals can hope to do—to new events and new information and arguments.

While this argument is elegant and has normative appeal, its key assumptions are problematic. Indeed, most troubling is the assumption that the errors in opinion change are independent of one another with a mean of 0. In other words, the error that one person makes is unrelated to the error made by another person in their connection between their particular policy beliefs and those of parties or politicians.

As Bartels argues (1996, p. 199), this intuitively seems unlikely: “If one uninformed voter is inappropriately swayed by a rhetorical flourish in a televised debate or advertisement, another may be equally swayed in the opposite direction; but it seems more likely that the second ‘error’ would reinforce rather than mitigate the first.” In this case, such errors would not have a mean of 0 and would thus bias aggregate measures of political attitudes in systematic ways. Bartels (1996), for example, finds strong information effects in U.S. presidential voting behavior and argues that if everyone had “perfect” knowledge, electoral outcomes would look much different.

Second, many scholars have argued that conventional tests of political knowledge hold citizens to an unrealistic ideal and citizens do not require perfect rationality or full information to make effective political decisions. For example, Lupia (2016) argues that conventional survey “factual recall” methods drastically understate political knowledge levels in the electorate. In a series of experiments, he finds political knowledge increases when respondents are sufficiently motivated. This suggests that otherwise disinterested citizens can acquire the knowledge to make effective political decisions in the right conditions, possibly including a high-stakes election.2

Cognitive “shortcuts” or heuristics may help citizens do this more effectively. Starting with the seminal studies of Popkin (1991) and Sniderman, Brody, and Tetlock (1991), there has been extensive work on voters’ use of political heuristics and how they help voters reach at least a minimum level of rationality. Do voters use such cognitive shortcuts? A large literature primarily based in the United States offers a resounding “yes.” Simple rules of thumb such as accepting or rejecting a message based on whether the speaker is a Democrat or Republican, White or Black, or liberal or conservative appear to help voters make more ideologically coherent and consistent political choices (Brady & Sniderman, 1985; Popkin, 1991; Lupia, 1994). The logic is people do have relatively fixed, albeit possibly latent policy positions, and they use their affective feeling toward the speaker or group as a shortcut to ascertain how close a speaker’s positions are to their own (Brady & Sniderman, 1985).

Do these cognitive shortcuts make for a more informed electorate that can better hold leaders accountable? That is less clear. One reason is that the primary users of heuristics appear to be the already more informed voters, not those with less political knowledge. In other words, heuristics may increase the gap in political sophistication between the “haves” and “have nots” of political information (e.g., Lau & Redlawsk, 2001). The less informed, meanwhile, appear to use heuristics in a haphazard and uneven way.

A second and more important reason is there is a great deal of evidence that heuristics do not work as advocates have assumed (Kuklinski & Quirk, 2000). Instead, they appear to often increase bias and error in decision making. This point is expanded on in greater detail in “The Issue with Issues: Cues, Motivated Reasoning, and Partisan Bias,” but, in short, researchers must be at least somewhat skeptical that aggregate-level rationality or cognitive shortcuts are effective mechanisms for overcoming the public’s limited political knowledge and ensuring policy representation in representative democracies.

# The Issues With Issues: Valence Issues and Retrospective Voting

Before turning to how cognitive heuristics may bias opinion formation and policy positions, the major alternative theory is reviewed of democratic accountability, which emerged in response to the early findings on voters’ limited political sophistication—the “retrospective theory” of voting (Key, 1966; Fiorina, 1981). This theory shifts the focus of voting behavior from policy proximity to performance on valence issues. Valence issues are issues where everyone shares a common preference, such as economic growth, lower unemployment, or less crime. As Fiorina (1981, p. 5) argued, valence issues and retrospective voting lower the cognitive bar for voters significantly since all voters “typically have one comparatively hard bit of data: they know what life has been like during the incumbent's administration.” At the ballot box, they simply need to decide to reward or punish incumbent governments based on their past performance.

It is not within the scope of this article to provide a full review of this extensive research (see Hellwig & Marinova, 2017). Rather, highlighted here is that despite its lower cognitive bar, retrospective voting appears to suffer from many of the same limitations as policy voting. Indeed, there is growing evidence that voters struggle to determine even such simple tasks as whether the economy has improved or stagnated. At minimum, voters show a tremendous amount of recency bias, often judging the incumbent party on just the last 4 to 6 months of its term (Achen & Bartels, 2016). This behavior hardly qualifies as rational; worse, it gives incumbents an incentive to manipulate the business cycle, even to the harm of the long-term economy (Erikson, 1989).

Other performance issues are likely to suffer from similar challenges. Voters are also prone to punishing incumbents for random events outside of their control, such as droughts, floods, and even shark attacks and possibly losses of one’s favorite sports team (Achen & Bartels, 2016, Chapter 5; Malhotra & Kuo, 2008; Malhotra, Healy, & Mo, 2010). Simply put, these findings further underscore the challenges facing voters in acquiring the knowledge necessary to make effective political choices and ensure a representative and responsive government.

# The Issue With Issues: Cues, Motivated Reasoning, and Partisan Bias

It is understandable if one is left with the impression that democracy is best left to the informed and engaged. However, even if this were possible, there are very good reasons to be skeptical this would solve the problem of democratic accountability. This is because there is little evidence that even informed voters make more informed choices—at least in the sense of being able to process competing viewpoints, formulate policy attitudes independently, and update their opinions given dissonant information.

Specifically, decades of social psychology research point to the critical role of prior attitudes and identities on processing and responding to information (e.g., Kahneman, 2003). This research finds most people do not seek out competing cues to become better informed about policy issues and their political choices. Instead, theories of “motivated reasoning” suggest people have a particular cognitive destination in mind when processing new information (Kunda, 1990). Rather than treat new information of all kinds equally or update existing beliefs according to Bayes’ Rule, innate desires to preserve attitude consistency induce a tendency to downplay contradictory information (disconfirmation bias) and uncritically accept consistent information (confirmation bias).3

The motivation to engage in motivated reasoning in the political realm generally results from partisan goals in the processing of new information. This motivation stands in contrast to the goal of accuracy, which is what is necessary for informed and effective policy or performance issue voting. As Taber and Lodge (2013, p. 43) note, the dominance of partisan goals in processing new information is due to the combination of affect with political ideations in memory: “all political leaders, groups, issues, symbols, and ideas thought about and evaluated in the past become affectively tagged—positively, negatively, or both—and with repeated coactivation an evaluative charge is linked directly to the concept in long-term memory.” In the political case, partisanship effectively imbues in-group political objects with positive affect and out-group with negative, creating what is known as affect-laden or “hot” cognition. To match the existing affect attached to political objects, motivated reasoners will expend large amounts of cognitive energy to maintain internal consistency with these value labels.

Going further, recent research has questioned altogether the direction of causality between voters’ policy attitudes and their political partisanship and support for specific candidates (e.g., Lenz, 2009, 2012). These studies suggest political attitudes largely originate among political elites and diffuse through the public on the basis of party affiliation. In short, political elites lead and (many) fellow partisans follow (Lenz, 2009).

Partisan bias also appears to plague the evaluation of seemingly objective facts (Bartels, 2002; Gaines et al., 2007; Jacobson, 2010; Taber & Lodge, 2006). A recent finding from the Pew Research Center’s annual survey on the state of the economy illustrates this bias well. They found in the lead-up to the 2016 U.S. presidential election just 31% of Republican identifiers thought the economy was doing well. Immediately after the election, nearly twice as many (61%) thought so. Evans and Pickup (2010, p. 1236) highlight similar trends during the 2000–2004 U.S. election cycle, finding that “economic perceptions are consistently and robustly conditioned by political preferences.”

It appears then that rather than grapple with information that may damage their view of “their” party, for many the motivating force of partisanship works to downplay or outright distort information that does not reinforce existing partisan beliefs. While there are no doubt limits to motivated reasoning and partisan bias (e.g., Redlawsk, Civettini, & Emmerson, 2010), partisans are profoundly biased toward rejecting information that contradicts their partisan identification and accepting information that supports it (Taber & Lodge, 2006). Politically sophisticated partisans, in turn, are skilled at parroting the messaging from (their) partisan elites and helping to disseminate issue arguments throughout the electorate.

This research not only turns conventional wisdom on its head; it undermines what for many was the one possible saving grace for ensuring democratic accountability—citizens who are politically knowledgeable and engaged. It is precisely this group who are most likely to toe the party line and follow partisan leaders in adopting issue positions. To see why, it is useful to review Zaller’s (1992) influential two-mediator Receive-Accept-Sample (RAS) model of public opinion formation and dissemination.

For Zaller (1992), the likelihood of a political message influencing a person’s attitudes is a function of two separate but interconnected processes: message reception and acceptance. To receive a political message, a person must be relatively aware of politics itself. Those who are not chronically attuned to politics are much less likely to receive political arguments of any kind as compared to those who regularly think about and consume political news. In this way, political awareness has a strong positive influence on the probability of message reception.

The second stage of this process, acceptance, represents the likelihood that a person hearing a message will either accept it or reject it as incompatible with their belief system. The key individual-level traits for this step are political awareness and a person’s predispositions or underlying value systems. Awareness mediates the impact of predispositions on the probability of message acceptance: Less aware voters cannot effectively connect their predispositions to incoming arguments and are thus more likely to accept messaging that may not conform with their underlying values. Highly aware persons, meanwhile, understand the various connections between political debates, actors, and values and are much less vulnerable to accepting a political message that might substantively change their beliefs.

Important to the effect of party identification, political elites and the media often connect (new) information to (old) political cues in their issue messaging. Political cues like partisanship play a critical role in filtering new information and activating the affective predisposition to support or reject that information or argument (Zaller, 1992; Converse, 1964). In this sense, the “persuasiveness” of a message largely depends on the messenger or the partisan labels attached to a specific message, not the message itself. Citizens appear to have limited ability or interest in evaluating messages, including policy proposals, on their own merit and in terms of the objective benefit of the policy for society or themselves.

In short, far from being a panacea for democracy, cognitive cues appear to undermine thoughtful, rational political decision making. Many knowledgeable voters use heuristics in a way that simply reinforces their preexisting beliefs and partisan leanings—they do not become more sophisticated in the sense of better understanding key issues and how they relate to their interests; they become more sophisticated in the sense that they are better able to argue and defend the viewpoints of their partisan “team.” Again, the central idea here turns a great deal of research concerning issue voting on its head: Vote choice or partisanship determines voters’ policy attitudes.

Taken together, it is clear that a large share of the electorate is either disengaged from politics and lacks meaningful policy attitudes or engaged but highly susceptible to persuasion from partisan elites. Both reinforce the power of leaders to pursue policies and personal interests independent of the public good. Even if the extent to which elites do this in practice isn’t known, especially in established democracies with a strong rule of law tradition, it nonetheless challenges the basic understanding of how democracy does function and how democracy should function.

# Mobilizing Voters and the Activation of Social Group Divisions

The discussion thus far may give the strong feeling that the prospects for issue voting are slim—at least as envisioned in the ideal classical model of representative democracy and as it relates to certain types of issues. At the same time, this does not necessarily mean issues do not matter at all for electoral politics or all of the time. Campaigns certainly behave as if issues affect voters, strategically emphasizing certain issues over others and devoting significant resources to identifying persuadable voters (Hillygus & Shields, 2008).

Moreover, if we take a step back, electoral outcomes do change from one election to the next and sometimes dramatically so. This simple fact is important as it does not square well with perspectives that fixate on relatively static factors like partisanship or the more pessimistic conclusions about democracy such as those of Achen and Bartels (2016). To be sure, some component of elections may be driven by random or ephemeral concerns, but elections and public opinion do appear to broadly respond to political stimuli, at least to some degree (e.g., Erikson, MacKuen, & Stimson, 2002).

One possibility is that the research on partisan bias overestimates the number of true partisans in democracies. The decline of party identification since the 1950s in the United States and across other established democracies is well known (Dalton & Weldon, 2007; Dalton, 2016). Indeed, self-reported Independents constitute the largest category of voters in almost all democracies as of 2018, including the United States. Even if partisanship is trending up again with growing polarization (Bafumi & Shapiro, 2009) or it is assumed “leaners” share many similarities to self-professed partisans (Green, Palmquist, & Shickler, 2002), true Independents still constitute a significant share of democratic electorates. Moreover, as Dalton (2012, 2016) shows, nonpartisans are younger, better educated, and more engaged than in the past, suggesting that partisan bias in the electorate as a whole may also be lower.

Perhaps the biggest caveat, though, is that almost all of the research on partisan bias has focused on the United States, a country with higher levels of partisanship than most democracies (Dalton, 2016). An expanded research agenda that focused on other democracies would improve understanding of partisan bias, how the institutional and political context shapes it, and the prospects for issue voting.

Beyond this, there remains evidence that policy issues might matter for electoral politics in at least two ways. The first is the saliency of issues in an electoral campaign may work to mobilize and demobilize certain voters to turn out on Election Day. In a reanalysis of Lenz’s (2009, 2012) key findings on partisan learning, Matthews (2017), for example, finds evidence for “issue priming” once respondents are included in the analysis who were nonresponsive or decided to abstain or support a minor party—a group that Lenz excluded from his analyses.

Parties and the media also often talk about the importance of “energizing the base,” and in this sense issues may play a bigger role in affecting voters’ decisions to turn out than it does in persuading them to switch their vote. Kinder and Kam’s (2010) study of ethnocentrism in the United States appears to support this. They found ethnocentric voters voted at much higher rates in states where gay marriage initiatives were on the ballot in the 2004 presidential election. Indeed, they went from a group that was less likely to turn out than the general electorate to a group that was more likely.

Kinder and Kam’s (2010) study, however, also points to a way that issues can affect actual vote choice—primarily by activating existing social and political group divisions in the electorate. We know partisan divisions play a key role in structuring issue opinions, but some recent research suggests issues also can have the opposite effect and weaken partisan ties. Hillygus and Shields (2008), for example, examine the role of so-called wedge issues in persuading partisans to vote for a different party. They find partisans are more likely to abandon their partisan leanings the more they disagree with their party on different issues. Moreover, parties appear to strategically emphasize precisely the issues that cause internal conflict for other parties’ partisans—that is, those issues where the partisan holds a different position from that of their party—in an effort to cross-pressure these voters.

Yet, what types of issues actually cause the most internal conflict for partisans? Clearly not all of them, because we know partisans often resolve this conflict rather easily by switching their issue positions to bring them in line with those of their party. In short, we know partisans are amazingly adept at reducing cognitive dissonance in the presence of issue conflict, and they seem rarely to do this by bringing their partisanship or vote choice in line with their issue positions (Lenz, 2009). In order to establish the presence of issue voting, this is precisely what is needed—voters who bring their partisanship or vote choice in line with their issue positions rather than those who bring their issue positions in line with their partisanship.

The key to understanding when and why this is most likely to happen is believed to lie in understanding partisanship’s power as a social identity—one that stimulates psychological and affective processes to protect and enhance one’s own partisan in-group (e.g., Greene, 1999). Partisanship, however, is not the only social identity or division that can structure political conflict, or even the most important necessarily, and issues that map onto salient social divisions in a society are more likely to affect voting behavior. In turn, the extent to which an issue genuinely cross-pressures voters depends on how much a voter’s position on such an issue differs from that of its current preferred party. Put differently, issues need to be identified that activate cross-cutting social cleavages. It is suspected that issues that do this benefit parties because they tap into and stimulate the same psychological and affective processes that make partisanship so powerful in the first place, especially in-group and out-group bias (see also Achen & Bartels, 2016; Tesler, 2016).

Students of U.S. politics have long recognized the power of racial divisions to structure political competition and how White racial identity and ethnocentrism, in particular, relate to positions on a range of issues (Carmines & Stimson, 1989; Key, 1949; Kinder & Kam, 2010). Over time, through repeated association, an issue also can become so affectively laden with race that it can implicitly prime voters on its own. Gilens (1999) finds this with crime and social welfare for many White Americans. Black–White racial identities, however, are already closely connected to partisan divisions in the United States, and hence, may have less power to cross-pressure voters, even if the two have become more closely connected since the Obama era (Tesler, 2016).

In recent decades, another social division has become increasingly salient in the United States and nearly all advanced democracies—that between citizens and immigrants. Immigration has long been a challenge for many countries, but its pace has accelerated recently amid globalization, increased mobility, declining fertility rates, and rapidly aging populations. With migrants coming from around the world, societies are becoming increasingly diverse in terms of race, ethnicity, and religion. Such diversity has put immigration squarely on the political agenda, and indeed, there is little doubt that it is fundamentally transforming politics in these countries. This is most evident in the rise of radical right parties; anti-immigrant sentiment appears to have also played a key role in fueling the Brexit vote and Donald Trump’s presidential election in 2016 (Inglehart & Norris, 2017).

Much of the recent research on issue voting in Europe has focused on this division, either immigration specifically or immigration along with European integration and other aspects of increasing globalization as part of an emerging cultural cleavage (Green-Pedersen, 2007; Hellwig, 2014; Kriesi et al., 2012). Yet, just as scholars of American politics have failed to take into account European focused research on social divisions and cleavages, these scholars have largely failed to take into account findings on social identity and in-group and out-group bias as a driver of issue opinions. The authors believe both groups of scholars can benefit from each other to help improve the understanding of issue voting.

Immigration is such a powerful and transformative political issue partly because, much like Black–White relations in the United States, it increasingly maps onto the division between the mostly White native populations and visible minority migrants. Religious differences only reinforce this, which is seen in the radical right’s anti-Muslim appeals. Depicting migrants as threatening in some way, either to physical safety, the allocation of material goods, traditional cultural values, or social status, only seems to strengthen the in-group and out-group divide. This is most evidently in Trump’s rhetoric. Indeed, few have done more to push the link between threat and ethnic minority immigration, from his pleas to build a wall on the Mexico–U.S. border, to repeatedly calling Mexicans rapists and Muslims terrorists, to his incendiary claim in January 2018 that the United States needs more migrants from Norway and fewer from “shithole countries” like Haiti, El Salvador, and African countries.

Immigration, however, also maps onto another powerful social division: that between citizens and non-citizens, creating state-sanctioned in-groups and out-groups. Not surprisingly, these overlapping social divisions and identities have helped radical right political entrepreneurs connect immigration to a wide range of other salient issues, including terrorism and security, crime, jobs, taxes, and the social welfare state. Moreover, immigration largely cuts across the traditional left–right economic spectrum, opening the door for issue voting and the reshaping of electoral politics. This is evident in observational studies of party competition (Kriesi et al., 2012) and voter preferences (Hellwig, 2014). Experimental research also supports this premise. Craig and Richeson (2014), for example, find that when told Whites were soon to become a demographic minority in California, White Americans of all political leanings more strongly endorsed conservative policy positions. They connect this to the status and social threat from increased immigration, especially from Latinos.

# Conclusion

The ideal, issue-driven voter as envisioned in classical accounts of democracy seems to find little support in political reality. Myriad studies demonstrate the vast majority of voters hold uninformed, unstable issue positions that are disconnected from a coherent ideology. Many even fail to know seemingly basic facts about their own political system. Furthermore, there is little evidence the voting public is becoming more ideologically informed, despite rising education levels and the emergence of mass media.

Among the fraction of voters who are engaged and well informed, many appear susceptible to persuasion and possibly manipulation of their issue opinions from the media and their partisan leaders. In short, there is a significant share of the electorate that either is too uninformed or too persuadable to engage in issue voting. This certainly raises questions about the viability of representative democracy and at its most pessimistic, the possibility that elites are largely free to pursue personal goals unchecked and independent of the public good. The authors are not this pessimistic.

Building on the insight about what makes partisanship such a powerful predictor of issue opinions, namely, social identity and in-group–out-group bias, one possible avenue has been identified for observing issue voting. Issues that map onto existing social divisions, especially race, religion, and immigration, appear to tap into the same psychological and affective processes and can work to cross-pressure voters. This is seen with the rise of the radical right, which now has a stable foothold in almost all European countries, and Trump’s election in the United States (Inglehart & Norris, 2017). However, there are other social cleavages not pitting the largely White majority against visible minorities that also could tap into these same cognitive processes, including class, which has long played a role in structuring democratic political competition.

Still, the darker elements of the radical right movement hardly inspire confidence in voters as tolerant and open democratic citizens, and that presents new challenges. As of late 2018, the limits of anti-immigrant, anti-Muslim, and racial resentment appeals are not known. On the one hand, Trump’s dismal approval ratings compared to the economic health of the country display a public with fairly cohesive beliefs regarding the limits of White ethnonationalism and political scandal. On the other, the “base” of Trump support remains fairly buoyant despite waves of negativity—the limits of such buoyancy remain unclear.

# Acknowledgments

The authors would like to express their appreciation to Vanna Lodders for her research assistance on this article. They also would like to thank the two reviewers for their thoughtful comments to improve the contribution.

## References

Achen, C. H., & Bartels, L. M. (2016). Democracy for realists: Why elections do not produce responsible government. Princeton, NJ: Princeton University Press.Find this resource:

Bafumi, J., & Shapiro, R. Y. (2009). A new partisan voter. Journal of Politics, 71(1), 1–24.Find this resource:

Bartels, L. M. (1996). Uninformed votes: Information effects in presidential elections. American Journal of Political Science, 40(1), 194–230.Find this resource:

Bartels, L. M. (2002). Beyond the running tally. Political Behaviour, 24(2), 117–150.Find this resource:

Berelson, B. R., Lazarsfeld, P. F., & McPhee, W. N. (1954). Voting: A study in opinion formation in a presidential campaign. Chicago, IL: University of Chicago Press.Find this resource:

Brady, H., & Sniderman, P. (1985). Attitude attribution: A group basis for political reasoning. American Political Science Review, 79, 1061–1078.Find this resource:

Budge, I., Crewe, I., & Farlie, D. (Eds.). (1976). Party identification and beyond. New York, NY: Wiley.Find this resource:

Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. Chicago, IL: University of Chicago Press.Find this resource:

Carmines, E. G., & Stimson, J. A. (1989). Issue evolution: Race and the transformation of American politics. Princeton, NJ: Princeton University Press.Find this resource:

Converse, P. E. (1964). The nature of belief systems in mass publics. New York, NY: Free Press.Find this resource:

Converse, P. E., & Markus, G. (1979). A dynamic simultaneous equation model of electoral choice. American Political Science Review, 73(4), 1055–1070.Find this resource:

Craig, M., & Richeson, J. (2014). On the precipice of a “majority-minority” America: Perceived status threat from the racial demographic shift affects white Americans’ political ideology. Psychological Science, 25(6), 1189–1197.Find this resource:

Dalton, R. J. (2012). The apartisan American: Dealignment and changing electoral politics. Thousand Oaks, CA: CQ Press.Find this resource:

Dalton, R. J. (2016). Party identification and its implications. In Oxford research encyclopedia. New York, NY: Oxford University Press.Find this resource:

Dalton, R. J., Flanagan, S., & Beck, P.A. (Eds.). (1984). Electoral change in advanced industrial democracies: Realignment or dealignment? Princeton, NJ: Princeton University Press.Find this resource:

Dalton, R. J., & Weldon, S. (2007). Partisanship and party system institutionalization. Party Politics, 13(2), 179–196.Find this resource:

Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it matters. New Haven, CT: Yale University Press.Find this resource:

Downs, A. (1957). An economic theory of democracy. New York, NY: Harper.Find this resource:

Erikson, R. S. (1989). Economic conditions and the presidential vote. American Political Science Review, 83(2), 567–573.Find this resource:

Erikson, R. S., MacKuen, M. B., & Stimson, J. A. (2002). The macro polity. New York, NY: Cambridge University Press.Find this resource:

Evans, G., & Pickup, M. (2010). Reversing the causal arrow: The political conditioning of economic perceptions in the 2000–2004 U.S. presidential election cycle. Journal of Politics, 72(4), 1236–1251.Find this resource:

Fiorina, M. P. (1981). Retrospective voting in American national elections. New Haven, CT: Yale University Press.Find this resource:

Gaines, B. J., Kuklinski, J. H., Quirk, P. J., Peyton, B., & Verkuilen, J. (2007). Same facts, different interpretations: Partisan motivation and opinion on Iraq. Journal of Politics, 69(4), 957–974.Find this resource:

Gilens, M. (1999). Why Americans hate welfare: Race, media and the politics of antipoverty policy. Chicago, IL: University of Chicago Press.Find this resource:

Green, D., Palmquist, B., & Schickler, E. (2002). Partisan hearts and minds: Political parties and the social identities of voters. New Haven, CT: Yale University Press.Find this resource:

Greene, S. (1999). Understanding party identification: A social identity approach. Political Psychology, 20(2), 393–403.Find this resource:

Green-Pedersen, C. (2007). The growing importance of issue competition: The changing nature of party competition in Western Europe. Political Studies, 55(3), 607–628.Find this resource:

Hellwig, T. (2014). The structure of issue voting in postindustrial democracies. Sociological Quarterly, 55(4), 596–624.Find this resource:

Hellwig, T., & Marinova, D. M. (2017). Dead ends and new paths in the study of economic voting. Oxford research encyclopedia. New York, NY: Oxford University Press.Find this resource:

Hillygus, D. S., & Shields, T. G. (2008). The persuadable voter: Wedge issues in presidential campaigns. Princeton, NJ: Princeton University Press.Find this resource:

Inglehart, R., & Norris, P. (2017). Trump and the populist authoritarian parties: The silent revolution in reverse. Perspectives on Politics, 15(2), 443–454.Find this resource:

Jacobson, G. C. (2010). Perception, memory, and partisan polarization in the Iraq war. Political Science Quarterly, 125(1), 31–56.Find this resource:

Kahneman, D. (2003). A perspective on judgement and choice: Mapping bounded rationality. American Psychologist, 58(9), 697–720.Find this resource:

Key, V. O. (1949). Southern politics in state and nation. New York, NY: Vintage.Find this resource:

Key, V. O. (1966). The responsible electorate: Rationality in presidential voting 1936–1960. Cambridge, MA: Harvard University Press.Find this resource:

Kinder, D. R., & Kalmoe, N. P. (2017). Neither liberal nor conservative: Ideological innocence in the American public. Chicago, IL: University of Chicago Press.Find this resource:

Kinder, D. R., & Kam, C. D. (2010). Us against them: Ethnocentric foundations of American opinion. Chicago, IL: University of Chicago Press.Find this resource:

Kriesi, H., Grande, E., Dolezal, M., Helbling, M., Hölinger, D., Hutter, S., & Wueest, B. (2012). Political conflict in Western Europe. Cambridge, U.K.: Cambridge University Press.Find this resource:

Kuklinski, J. H., & Quirk, P. J. (2000). Reconsidering the rational public: Cognition, heuristics, and mass opinion. In A. Lupia, M. McCubbins, & S. Popkin (Eds.), Elements of reason: Cognition, choice, and the bounds of rationality. New York, NY: Cambridge University Press.Find this resource:

Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480.Find this resource:

Lau, R. R., & Redlawsk, D. P. (2001). Advantages and disadvantages of cognitive heuristics in political decision making. American Journal of Political Science, 45, 951–971.Find this resource:

Lenz, G. S. (2009). Learning and opinion change, not priming: Reconsidering the priming hypothesis. American Journal of Political Science, 53(4), 821–837.Find this resource:

Lenz, G. S. (2012). Follow the leader? How voters respond to politicians, policies, and performance. Chicago, IL: University of Chicago Press.Find this resource:

Lippmann, W. (1922). Public opinion. New York, NY: Macmillan.Find this resource:

Lodge, M., & Taber, C. S. (2013). The rationalizing voter. Cambridge, U.K.: Cambridge University Press.Find this resource:

Lupia, A. (1994). Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review, 88(1), 63–76.Find this resource:

Lupia, A. (2016). Uninformed: Why people seem to know so little about politics and what we can do about it. New York, NY: Oxford University Press.Find this resource:

Macdonald, S. E., & Rabinowitz, G. (1989). A directional theory of issue voting. American Political Science Review, 83(1), 93–121.Find this resource:

Malholtra, N., Healy, A. J., & Mo, C. H. (2010). Irrelevant events affect voters’ evaluations of government performance. Proceedings of the National Academy of Sciences, 107(29), 12804–12809.Find this resource:

Malhotra, N., & Kuo, A. G. (2008). Attributing blame: The public’s response to Hurricane Katrina. Journal of Politics, 70(1), 120–135.Find this resource:

Matthews, J. S. (2017). Issue priming revisited: Susceptible voters and detectable effects. British Journal of Political Science. First View, online publication.Find this resource:

Page, B. I., & Shapiro, R. Y. (1992). The rational public: Fifty years of trends in American policy preferences. Chicago, IL: University of Chicago Press.Find this resource:

Popkin, S. L. (1991). The reasoning voter: Communication and persuasion in presidential campaigns. Chicago, IL: University of Chicago Press.Find this resource:

Redlawsk, D. P., Civettini, A. J., & Emmerson, K. M. (2010). The affective tipping point: Do motivated reasoners ever “get it”? Political Psychology, 31(4), 563–593.Find this resource:

Sniderman, P. M., Brody, R. A., & Tetlock, P. E. (1991). Reasoning and choice: Explorations in political psychology. New York, NY: Cambridge University Press.Find this resource:

Taber, C. S., & Lodge, M. (2006). Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science, 50(3), 755–769.Find this resource:

Taber, C. S., & Lodge, M. (2013). The rationalizing voter. Cambridge, U.K.: Cambridge University Press.Find this resource:

Tesler, M. (2016). Post-racial or most-racial? Race and politics in the Obama era. Chicago, IL: University of Chicago Press.Find this resource:

Zaller, J. (1992). The nature and origins of mass opinion. Cambridge, U.K.: Cambridge University Press.Find this resource:

## Notes:

(1.) This is not to say that the use of heuristics or cues necessarily entails rational or informed decision making on policy issues.

(2.) Still, it is unclear the extent to which the treatment condition of incentivizing knowledge is applicable to real-world politics (Achen & Bartels, 2016, p. 37).

(3.) Bayes’ Rule understands the processing of new evidence as partially independent from existing priors. The general form of Bayes’ Rule assumes the following: $P(A|B)=P(B|A)P(A)P(B)$, where B represents the weight of prior beliefs and A is new evidence. P(A|B) is thus one’s belief about the likelihood that A would occur if B (one’s existing beliefs) were true. Theories of motivated reasoning posit that the updating of posterior beliefs is either wholly determined by B rather than A or that any dissonant A has so little weight as to render it effectively meaningless.