Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Politics. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 25 April 2024

Candidates and Voting Choicefree

Candidates and Voting Choicefree

  • Ian McAllisterIan McAllisterDepartment of Political Science, Australian National University


The role of candidates in shaping voting choice has generated much research—and at least as much controversy—since modern electoral behavior research began in the 1960s. Much of the controversy surrounds the personalization of politics and whether political systems—and especially parliamentary systems—are becoming more leader-oriented. Three fundamental changes in electoral behavior underpin the study of candidates and voting choice behavior: the declining impact of social structure on the vote; partisan dealignment, with voters drifting away from their traditional party attachments; and the decline in the mass memberships of political parties. Researchers argue that because of these changes, fostered by the growth of television, candidates have assumed a greater role in structuring the vote. While there is impressionistic evidence that leaders have become more important, empirical evidence of an underlying change in voter behavior is more difficult to identify. Accordingly, this essay focuses mainly on changes in the political context within which candidates operate, since we expect this to be the source of any change.

The design of political institutions shapes the level of attention that candidates receive, and that is especially the case with electoral systems. Electoral systems with fewer parties are more likely to focus voters’ attentions on candidates when compared to systems with larger numbers of parties. Weak party organizations coupled with partisan dealignment within the electorate can also alter the role and profile of candidates, although their impact is difficult to quantify. Changes in the mass media—and particularly the advent of television in the 1960s and the visual images on which it relies—are often viewed as the major cause of the personalization of politics. A new disruptive technology, the Internet, looks likely to stimulate additional political change for candidates and voting in the 21st century. Finally, what voters look for in their candidates appears to be stable both over time and cross-nationally and can be reduced to two overarching qualities: character and competence.


  • Political Behavior


The decline of social structure as an influence on voting choice, especially in European democracies, has led to much debate about what may be replacing it. One explanation sees voting as becoming more issue-based, with parties and candidates emphasizing issues that they believe will attract support as well as promoting their ability to resolve them (Clarke et al., 2009). Another explanation is that social structure is being replaced as a voting cue by evaluations of the individual candidates and the party leaders. In this view, the rise of television has focused the public’s attention on the personalities of the leaders, which are then evaluated by voters as a guide to voting choice. Such a trend is further enhanced by the increasing importance of election campaigns in determining election outcomes, where personalities can exert a sometimes crucial short-term influence (Schmitt-Beck & Farrell, 2003). This greater attention toward candidates’ personalities is called the “personalization of politics” (Karvonen, 2010; McAllister, 2007).

The trend toward the personalization of politics is usually traced to the 1980s, especially the Thatcher government in Britain and the Reagan administration in the United States.1 In each case, the governments were named after their leader rather than their party, on account of their strong free market policies. Nor is the trend toward the personalization of politics restricted to presidential systems, such as the United States; the popular focus on leaders is now commonplace across almost all of the major parliamentary systems. Indeed, the focus on leaders within parliamentary systems has been so marked since the 1980s that it has spawned a large literature, which has variously labeled it the “presidentialization of politics” (Poguntke & Webb, 2005), “institutional presidentialization” (Maddens & Fiers, 2004), or “presidential parliamentarism” (Hazan, 2005).

There is therefore much impressionistic evidence that candidates and leaders have become more important,2 but cross-national, longitudinal empirical analysis is largely lacking. In a comprehensive literature review combined with an analysis of selected parliamentary democracies, Karvonen, while seeing no general trend toward personalization, concludes that “there are many indications that persons have become more prominent in both electoral and executive politics in many countries” (2010, p. 106). Bittner, who conducts a cross-national longitudinal analysis of seven countries, concludes that “leaders play an important role in the individual vote calculus, and they also have a discernible effect on the distribution of votes in an election” (2011, p. 139). In a study of elections in Britain, Germany, and the Netherlands, Garzia (2014) comes to a similar conclusion. Single-country studies, such as those in Britain (Mughan, 2000; Heffernan & Webb, 2005) and Germany (Poguntke, 2005), found a general shift toward personalization, but not a consistent linear trend.3

There is, then, general agreement that the personalization of politics is occurring but disagreement about how far it has progressed. Arriving at an answer to the second question is complicated by three considerations. First, longitudinal analyses are necessarily contaminated by changes in the candidates, so it is not possible systematically to compare like with like. Second, along with personnel changes, the institutional context often varies, once again stymieing systematic comparisons. Third, studies frequently use different data sources. For example, mass media studies typically examine who appears in the media and how they are reported, while voting studies are anchored in surveys of citizens. These different considerations have led scholars to conclude that any systematic evaluation of the changing role of candidates and leaders is highly problematic. Kaase puts it succinctly when he says that “it is the specific combination of candidates and political context which defines the candidate impact for each individual election” (1994, p. 222). This suggests that we must examine whether the context has changed in order to favor the role of candidates.

This article reviews the research on candidates and voting with the aim of understanding how the changing political context may shape the role of candidates in the electoral process. The first section outlines the possible contribution of institutional design to the personalization of politics, emphasizing constitutional arrangements and the mechanics of the electoral system. The second section covers political parties and the contribution of partisan dealignment. The third section analyses the mass media with a particular emphasis on the role of television in directing attention to individual personalities. The fourth and final section discusses candidate traits and identifies which traits or images are most likely to influence electoral choice.

Institutional Design

The design of political institutions is central to determining whether a political system is candidate-centered or party-centered. The major distinction is between presidential and parliamentary systems, although a few countries—France, Switzerland, and several Latin American democracies—sit between the two systems and are often classified as semi-presidential. The defining characteristic of a presidential system is the popular election of the president, thus directing public attention to the individual who occupies the position. Further making presidential systems candidate-centered is the fact that party discipline is usually weak, since the president’s political survival does not depend on the unity of the governing party. By contrast, parliamentary systems encourage collective responsibility, so that the executive is both dependent upon the confidence of the members of legislature and accountable to them. The operation of parliamentarism also encourages party government, so that in contrast to presidential systems, party discipline becomes a primary factor in maintaining executive authority.

Presidential systems have fixed terms for their leaders, so retaining office is not dependent upon the day-to-day confidence of the legislature. This permits presidents greater flexibility in formulating and implementing policy without the risk of an unexpected election. The survival of the executive in parliamentary systems depends upon the confidence of the legislature; the executive can be removed at any time by the legislature following the passing of a vote of no confidence. In practice, this means that a prime minister must emphasize retaining the confidence of his or her party colleagues. These two institutional arrangements mean that presidential systems are by nature more candidate-centered, while parliamentary systems are more party-centered.

Parliamentary systems have begun to move closer to the presidential model over the past several decades, however. British prime ministers, for example, have accumulated considerable power and authority. In part, this developed as a consequence of the increased complexity of modern decision-making, which has produced a large cadre of advisors. In part, too, prime ministerial power has been increased by the prime minister’s ability to shape ministerial careers, a crucial tool in ensuring compliance and centralizing authority. This trend led Michael Foley (2000; but see Dowding, 2013), among others, to talk of the presidentialization of British politics. Other work has extended the concept to other parliamentary systems (Poguntke & Webb, 2005). While the debate about presidentialization continues, the general view is that political power has shifted from the legislature to the executive, but how much change has taken place and with what consequences remain a matter of dispute.4

A second institutional effect over which there is little disagreement is electoral system design. Electoral systems are easily manipulated by politicians and parties since they are rarely constitutionally embedded, unlike presidentialism or parliamentarism. As a consequence, electoral systems are often subject to repeated reform by party elites, especially in newly democratizing countries. In Russia, for example, the parliamentary election law was amended no less than 26 times between its adoption in 2005 and 2012, with some of the reforms representing major changes. Most changes were designed to benefit the ruling party elite (McAllister & White, 2014).5 The frequency of electoral law reforms seems to be increasing in established democracies as well.

From the perspective of candidates and voting, electoral systems can be classified according to the degree to which either the candidate or the party is the most important consideration for the individual voter. Electoral systems that permit voters to discriminate between candidates have most potential to generate a personal vote. Ireland’s single transferable vote system falls into this category (Martin, 2010). At the other end of the scale, closed party list systems, such as those used in Belgium and Denmark, where parties determine the order of the candidates standing for election, give more prominence to parties. The main distinction is therefore between ordinal systems where voters have more choice, either by being able to select multiple candidates or by rank-ordering them, and categorical ballots, where voters have fewer choices in determining the fate of individual candidates.

Figure 1 demonstrates the importance of the electoral system is shaping whether or not a voter could name any of the candidates standing in their electoral district. Using data from the first module of the Comparative Studies of Electoral Systems project, the results show that in countries with plurality systems and single-member districts, 66% could name one or more of the local candidates, falling to 38% in plurality systems with multi-member districts. In proportional representation systems based on a party list, just under half could name a candidate. The evidence suggests, then, that electoral system design does indeed shape the voters’ information about their local candidates.

Figure 1 Electoral systems and knowledge of local candidates.

(Source: Comparative Study of Electoral Systems, module 1.)

Has there been a general move to adopt more candidate-centered electoral systems, thus enhancing the electoral role of the candidate? Examining electoral reforms across the established democracies from the 1980s onward, Karvonen concluded that there has been no clear pattern. He argues that the most that can be said is that there has been “a development towards a compromise between party and candidate-centeredness” (2010, pp. 37–39).On the question of which systems are more likely to profile candidates rather than party, there is little disagreement. Examining six countries, Bittner identifies a clear pattern in which “more proportional systems and fewer parties campaigning place leaders more prominently in the vote calculus” (2011, p. 136). The electoral system, as much as the country’s constitutional arrangements, thus represents a key factor that helps to shape the importance of candidates to electoral outcomes.

Political institutions clearly matter in the level of attention they can direct toward a candidate. There is little disagreement, for example, about the key role of the electoral system. But to what degree institutions matter and, perhaps more importantly, how they are mediated by a plethora of related factors, such as the size of the party system, and level of polarization, or the level of political sophistication of the mass public, remains a matter of considerable disagreement.

Candidates and Party Decline

How citizens view candidates is closely associated with how they view political parties. Popular feelings about parties are strongly associated with the direction and strength of partisan attachments. The link between partisan attachment and support for a particular candidate is strongest in parliamentary systems. Some of the earliest voting studies in Britain found that citizens’ views of party leaders were associated with the popular images of the parties to the extent that they were almost indistinguishable. Partisanship can change popular views of candidates in two ways. First, partisanship can shape how voters view candidates and the candidate traits and qualities they regard as important. Second, as partisanship declines, it can provide candidates with more independence, freed from the necessity of following party rules and policies.

Across all but a handful of democracies—usually the newer democracies of central and eastern Europe—party membership has declined dramatically from the high point recorded just after the end of the Second World War (Scarrow, 2014). At one level, this decline may have few negative consequences, since party linkage remains crucial to the operation of modern democracy and there is no obvious alternative to the party (Dalton, Farrell, & McAllister, 2011). At another level, the mass party model is less relevant to modern politics, particularly with the rise of the Internet and its unrivaled ability to segment the electorate (Sunstein, 2009). With the possible exception of the Nordic countries, political parties have been remarkably slow to adapt to the challenge that the Internet poses to traditional forms of political activism and involvement.

From a candidate perspective, weakening party organization makes it more difficult to mobilize voters. This has often shifted voters’ attentions away from local election campaigns, where parties have few resources, and toward the national political stage, where parties can use the mass media to promote their appeal to the widest number of voters (McAllister, 2013). This shift from local candidates to national political leaders promotes candidates who the parties believe will exercise the maximum geographical and social appeal to voters. The national party leaders themselves thus become more central to campaigns. In addition, Cutler (2002) showed, for example, that voters prefer leaders who are closest to themselves in terms of their socio-demographic background. This builds on research demonstrating that gender influences the choice of candidate, which is at least partly the result of gender stereotypes (Sanbonmatsu, 2002).

How voters view leaders through the prism of partisanship has also attracted scholarly attention. Declining levels of party identification across the electorate should mean that leaders and candidates attract more attention from voters, who evaluate them as surrogates for—or at the very least complementary to—party (Garzia, 2014). Karvonen found no empirical support for this argument, and stated that “citizens with marked party identities are more prone to like or dislike party leaders more intensively than citizens with weak or no party identities” (2010, p. 78). Similarly, Bittner showed that partisan stereotypes are important, particularly among the most informed: “the most politically sophisticated segment of voters evaluates party leaders in a fashion that conforms most strongly to the partisan stereotype” (2011, p. 90). While both studies use a restricted range of countries to reach these conclusions, the results suggest that parties and candidates are not alternatives in the electoral equation but dependent upon one another.

The widespread partisan dealignment in almost all of the advanced democracies in the past several decades has had profound effects on citizen voting behavior (Dalton, 2000). Partisan dealignment stimulates greater electoral volatility, and in the absence of strong links between the social structure and the party system, voters are more likely to switch their party support between elections. In these circumstances, weaker party attachments may enhance the role of the candidate in both the mobilization and conversion of the vote. With weaker party cues, voters will rely more heavily on the short-term appeal of the candidates to decide their vote. This should be particularly the case in new democracies, where parties are weaker generally and the parties’ contact with voters is lower.

One indicator of the changing balance between parties and candidates in the voting decision asks citizens what they regard as more important in their voting decision—the candidate or the party. Such data exist over time for three countries—Australia, Finland, and Ireland. Figure 2 shows a pattern of decreasing reliance on party in the voting decision and an increasing reliance on the candidates. Ireland most neatly fits this pattern; in 1979 around 5% more respondents mentioned the party than mentioned the candidate, but by 2007 around 19% more respondents mentioned the candidate rather than the party. Finland also fits this pattern until 2011, when there was unexpected increase in party over candidate mentions. The broad trend in Australia, with some fluctuations, also shows a modest trend toward favoring the candidate over the party in the voting decision.

Figure 2 The importance of party versus candidate in the voting decision.

(Sources: Ireland, Karvonen (2010, p. 56) and updates; Finland, Karvonen (2010, p. 51) and updates; Australia, Australian Election Study 1998–2013.)

The decline of political parties—in mass membership and organization role and in their day-to-day salience for the ordinary voter—is clearly another important element in shaping how citizens view candidates. But once again, in evaluating the impact of parties on candidates, much depends on the circumstances of the country under scrutiny. In particular, research has most frequently examined the established European democracies, and we know little about non-European countries or about those that are in the process of democratization.

The Mass Media

The mass media represents the arena within which political competition takes place. For most of the 20th century that arena was dominated by newspapers and by radio. Since the 1960s, television has become the dominant communications medium, and this has forced political actors to adjust their priorities in ways that would have been unthinkable just a decade earlier. The ubiquity of television has profound implications for political competition generally, and for the role of candidates and leaders more specifically (Lau & Redlawsk, 2006). More recently, the Internet has made substantial inroads into political communication, particularly among younger people and via the social media. While the eventual political impact of the Internet is unclear, it looks likely to have consequences for candidates and voting that are at least as profound as those of television half a century earlier.

In the early years of its development, television devoted relatively few resources to the coverage of politics because the medium was regarded as ill-suited to politics. That view rapidly changed as television’s potential to market politics to voters through the personalities of the candidates became apparent. In the 1952 U.S. presidential election campaign, Dwight D. Eisenhower, the successful candidate, made extensive use of television advertising for the first time. His television exposure projected him as a warm and friendly personality, in contrast to his opponent, Adlai Stevenson, who refused to follow suit and was viewed as aloof and detached. Stevenson reacted by claiming that the media were biased against him, a charge that would be repeated by almost every candidate at every subsequent presidential election.

While television had an early role in U.S. politics, its potential developed more slowly in the major parliamentary systems. Nevertheless, by the 1960s the television coverage of politics—and especially political leaders—was widespread, and television began to influence the way that voters viewed their leaders. In Britain, the 1964 general election was the first to be systematically covered by television; it was also the first election in Britain where the term “presidential” was used to describe the character of the campaign (Mughan, 2000). The link between television and personalization expanded in other parts of Europe as well, although the effects are uneven and often contingent on the characteristics of the campaign. By the late 1960s, television was an indispensable tool for modern election campaigning in virtually all of the established democracies (Lau & Redlawsk, 2006).

Television also stimulated the first campaign debates between the major political leaders. The first nationally televised debate was held in the United States between John F. Kennedy and Richard Nixon during the 1960 presidential election campaign and was credited with winning Kennedy the presidency. The next U.S. presidential debate did not take place until the 1976 election, when Gerald Ford debated with Jimmy Carter. Since then, the leaders’ debate has been a major event in every presidential election campaign. The U.S. experience helped to propagate the idea of a leaders’ debate in the established parliamentary democracies; by the mid-1990s most held a leaders’ debate in the national election. The last established democracy to relent was Britain; in 2010 three debates were held involving all three major party leaders. The debates were credited with giving the Liberal Democrats an immediate surge in support (Clarke et al., 2011).

Television changed the relationship between voters and candidates by using visual images, rather than the printed word or voice, to communicate information. This has necessarily focused attention on the personalities of candidates, since it is easier for television to disseminate information through a familiar personality rather than through an abstract document or an institution. For television, candidates represent a convenient visual shortcut to capture and retain the viewer’s attention, particularly if the information overlaps with the leader’s personality. In addition, parties find it easier to promote political choices to voters through a familiar personality, who can present the party’s policies more effectively in comparison to the printed word. And not least, voters prefer to hold an individual accountable for government performance (or, occasionally, for the performance of the opposition), rather than an abstract institution or a political ideal.

The rise of the Internet as a communication medium adds a further layer of complexity to the relationship between voters and candidates. Early research suggested that the major parties, with their greater command of resources, would dominate the new medium, resulting in offline patterns of competition being replicated in the virtual world. However, the expansion of social media produced a different pattern; voters now have an unprecedented choice in the political information that they choose to access. They may seek large amounts of political information or, equally, they may eschew all politics altogether (Prior, 2007).

With the unprecedented segmentation (or “micro-targeting”) of the media, voters and candidates can communicate with one another in a virtual one-to-one relationship. This has had two consequences. First, it favors minor parties, who have been able to use the relatively low resource threshold of the social media to directly communicate with potential voters, thus equalizing the communications arena (Gibson & McAllister, 2014). Second, social media further focus attention on the personalities and activities of candidates and leaders. Facebook and Twitter, for example, allow voters to follow the daily activities and shifting views of candidates and permit candidates to profile themselves to voters and to communicate directly with them (De Vreese, 2007). Such interactivity bypasses the role of the party and instead personalizes the relationship between voter and candidate, thus enhancing political involvement. In an experimental study, Kruikemeier et al. (2013) showed that “interactive, personalized online communication has a positive effect on citizens’ feelings of having the opportunity to come into contact with politics, and citizens’ feelings of closeness to politics” (p. 8). While research is at an early stage, the existing studies suggest that the Internet will increase personal contacts between voters and candidates and hence political participation generally.

Technological change is constantly changing the opportunities for candidates to communicate with voters through the mass media. Since the 1960s, the trajectory of change has been in the direction of focusing the public’s attention on candidate profiles and personalities, above and beyond their policy platforms. The visual images on which television relies started this process, and it had continued with the interactivity that is the hallmark of the Internet. These technological changes have enabled candidates to rely less on their party affiliation for support and more on their personal appeals and images conveyed through the electronic media.

Candidate Traits

The images or traits that voters associate with candidates and leaders enable them to make an informed judgment about the personal qualities that would make the person an effective representative. This is the type of judgment that people perform regularly in their daily lives; in the political context, such an assessment acts as a useful cue in guiding voting choice. The importance of candidate traits in shaping the vote first came to prominence in the 1960s, when researchers observed that while party identification did not change much, election outcomes did. They attributed this in part to the changing evaluations of leaders that influenced the vote. Scholars have therefore focused attention on leaders as a core variable in the voting decision and, as the previous section demonstrated, devoted considerable resources to analyzing how the media portrays them.

Which aspects of candidate images or traits matter most in shaping the vote? Identifying the most salient traits has been a rich vein in electoral research. While there is great diversity in the range of findings—based on country, question format, time period and methodology—two major conclusions emerge from this research. First, the basic structure of traits remains relatively stable overtime. Wattenberg (1991) and others have shown that how U.S. voters view presidential candidates has remained remarkably stable over an extended period. McAllister (2011) also found considerable stability in voters’ trait evaluations in Australia. In short, regardless of the changing political context or the country, voters evaluate the same traits in their leaders.

A second conclusion from the research on candidate images is the number of dimensions that emerge from the numerous traits included in the surveys. Bittner’s (2011) thorough review identified 28 different typologies. After taking into account overlapping traits, many based on labeling, she suggested that research points to four core dimensions: competence, leadership, character/integrity, and empathy. Kinder (1986) similarly identified four dimensions broadly corresponding to these labels; he also suggested that these four dimensions could be further refined by combining competence and leadership, and character/integrity and empathy. Bittner (2011) largely confirmed the existence of these two underlying dimensions—character and competence—by combining 35 election studies covering more than 81,000 respondents to measure candidate traits.

Which traits are more likely to shape overall candidate evaluations, and in turn the vote? In general, traits associated with competence and leadership appear more important than character and empathy. Fridkin and Kenney (2011) found that in U.S. Senate elections, leadership is most important in predicting incumbent evaluations and ranks second in importance in predicting challenger evaluations. They also showed that the media plays a significant role in mediating trait evaluations; in particular, negative trait messages, coming either from challengers or from the media, disproportionately shaped voters’ impressions of the main incumbents. Other research suggests that both broad traits matter in congressional elections but become more important when there are fewer ideological differences between candidates (Buttice & Stone, 2012).

Evidence from outside the United States has generally confirmed the predominance of leadership evaluations in shaping the vote. An analysis of the 2004–2010 Australian federal elections found that leadership is generally more important than character in shaping the vote (McAllister, 2011). Cross-national research has also demonstrated the importance of traits in forming overall evaluations of candidates and in turn the importance of these evaluations in predicting the vote. Comparing Australia, the United States, and Sweden, Ohr and Oscarsson found that there is a consistent effect for leader traits regardless of the country in question. They concluded that “politically relevant and performance-related leader traits are important criteria for voters’ political judgments and decisions” (2011, p. 212).

A further line of research on candidate traits examines women candidates and leaders and the extent to which certain images are gender-specific. Research in the United States suggests that personality traits are of limited importance in shaping support for women candidates, although views on stereotypical male and female policy issues are important (Dolan, 2010). For example, voters prioritizing issues such as ethics in government, often characterized as a female trait, are more likely to support a female candidate; those seeing foreign policy as their main priority, a trait dominated by men, favor a male candidate (Dolan, 2004). While the popular tendency to favor a candidate based solely on gender has largely disappeared, changing issue priorities and political contexts can indirectly influence the electoral fortunes of women candidates.

Another theme examines women political leaders. The consensus view of a large number of studies of women who have achieved national office is that a leader’s gender is a significant consideration among voters. For example, a study of the 2005 German election found that Angela Merkel’s gender played a considerable role in framing media stories, a pattern that was absent prior to her election as chancellor (Semetko & Boomgaarden, 2007). Denemark, Ward, and Bean (2012) found substantial gender differences in voters’ evaluations of Australia’s first female prime minister, Julia Gillard, with women evaluating her far more positively than men. Her popularity among women voters remained high even several years after her initial election, when the novelty of a female prime minister might have been thought to have dissipated.

The images and traits that voters associate with candidates and major political leaders clearly do matter. These traits contribute to a summary evaluation of a candidate’s suitability for election, which in turn has a tangible impact on the vote that she attracts. Moreover, the traits that emerge as important are stable overtime and consistent in their structure, falling into two broad categories, character and competence. More recent research in this area focuses on the issue of gender, and how and in what ways voters respond to female candidates. Again, this research demonstrates a consistent finding: namely, that voters do not discriminate against female candidates, but they do see them as more effective in dealing with certain issues rather than others.

Any summary judgment of a candidate by voters is obviously the accumulation of many smaller evaluations about a candidate’s personality, background, policies, and much else. These evaluations may also be shaped by dramatic events, which then influence popular views of a candidate’s personality. Examples of such key events include Margaret Thatcher during the 1982 Falklands War, Helmut Kohl following the collapse of the Berlin Wall in 1989, and George Bush in the immediate aftermath of 9/11. In each case the leaders benefited from their incumbency during a time of crisis and what has been termed the “rally round the flag” effect. For example, George Bush’s popularity surged following the 9/11 attacks, since he was the incumbent leader charged with countering a physical threat to the United States, although his popularity subsequently declined.


Few debates in political science have generated as much controversy as the impact of candidates and leaders on voting. Impressionistic evidence, often derived from mass media studies, leads to the conclusion that candidates have become more important overtime, as other factors, such as social structure, have declined. Yet the results of empirical studies concerning the greater importance of personality on voting are, as Karvonen (2010) noted, decidedly mixed. There is little disagreement that candidates are significant short-term influences on the vote—to a degree—but the data to address the key question of whether they have increased in importance are largely absent.

This article has highlighted some of the complexities involved in evaluating the electoral impact of leaders. In part, this stems from the fact that leaders are intertwined with all levels of the political process, from the design of institutions to their presentation within the mass media. At all stages of the process, particular circumstances may converge to either direct or deflect attention from a candidate. A further complication is that leaders are constantly being replaced, so any longitudinal analysis must confront changing personalities and voter popularity. Nevertheless, we can arrive at four general conclusions from the research that has been conducted to date.6

First and most crucially, leaders and candidates do matter in determining electoral outcomes, but their net effect is less than many observers might expect given the attention they receive in election campaigns and other events. Estimates of its effect normally place it at around plus or minus 2 or 3% of the vote. In Canada, Johnston (2002) suggested the net effect on the vote to be up to 2%, while in the United States, Bartels (2002) saw the effect as slightly larger, at just over 3%. Australian estimates suggest a net effect of around plus or minus 2%, depending on the leader in question (McAllister, 2011), and similar results are reported in Britain (Clarke et al., 2011). These empirical estimates are much lower than popular wisdom, but they still have the potential to change the outcome in a closely fought election. Second, institutions matter in directing attention to or from candidates, with the electoral system exercising a pivotal and consistent role in the process. Electoral systems with fewer parties and single member districts are likely to focus more attention on candidates compared to other systems. Third, voters—regardless of the country or time period under study—evaluate candidates based on two overarching criteria: character and competence. In turn, these evaluations can have significant effects on the voting decision. Fourth, voters view candidates through the prism of party, although changes in partisanship and in party membership may dynamically affect this relationship.

Finally, it is important to mention the mass media and political parties, which are covered in this chapter but not in the list of conclusions above. While we know that the mass media matters in the candidate–voter equation, and we know that the growth of television has had a profound influence on the nature of the relationship, we do not know if the mass media is a cause or a consequence of whatever changes are underway. There is a similar lack of clarity about the role of political parties and their relationship to candidates. In an era when parties have weaker social structural bases, and with party identification in steep decline, one hypothesis is that candidates should fill this gap. But once again, empirical evidence to support such a prediction is lacking. Perhaps one area about which the main protagonists in the debate about the role of candidates and voting agree is that further research is needed.


  • Bartels, L. M. (2002). The impact of candidate traits in American presidential elections. In A. King (Ed.), Leaders, personalities and the outcomes of democratic elections (pp. 44–68). Oxford: Oxford University Press.
  • Bittner, A. (2011). Platform or personality? The role of party leaders in elections. Oxford: Oxford University Press.
  • Buttice, M. K., & Stone, W. J. (2012). Candidates matter: Policy and quality differences in congressional elections. Journal of Politics, 74, 870–887.
  • Clarke, H. D., Sanders, D., Stewart, M. C., & Whiteley, P. (2009). Performance politics and the British voter. Cambridge, U.K.: Cambridge University Press.
  • Clarke, H., Sanders, D., Stewart, M., & Whiteley, P. (2011). Valence politics and electoral choice in Britain, 2010. Journal of Elections, Public Opinion and Parties, 21, 237–253.
  • Costa Lobo, M., & Curtis, J. (Eds.). (2015). Personality politics? The role of leader evaluations in democratic elections. New York: Oxford University Press.
  • Cutler, F. (2002). The simplest short of all: Sociodemographic characteristics and electoral choice. Journal of Politics, 64, 466–490.
  • Dalton, R. J. (2000). The Decline of Party Identification. In R. J. Dalton & N. P. Wattenberg (Eds.), Parties without partisans: Political change in advanced industrial democracies (pp. 19–36). Oxford: Oxford University Press.
  • Dalton, R. J., Farrell, D., & McAllister, I. (2011). Political parties and democratic linkage: How parties organize democracy. Oxford: Oxford University Press.
  • Denemark, D., Ward, I., & Bean, C. (2012). Gender and leader effects in the 2010 Australian election. Australian Journal of Political Science, 47, 563–578.
  • De Vreese, C. H. (2007). Digital renaissance: Young consumer and citizen? Annals of the American Academy of Political and Social Science, 611, 207–216.
  • Dolan, K. (2004). Voting for women: How the public evaluates women candidates. Boulder, CO: Westview.
  • Dolan, K. (2010). The impact of gender stereotyped evaluations on support for women candidates. Political Behavior, 32, 69–88.
  • Dowding, K. (2013). The prime ministerialisation of the British prime minister. Parliamentary Affairs, 66, 617–635.
  • Foley, M. (2000). The British presidency: Tony Blair and the politics of public leadership. Manchester, U.K.: Manchester University Press.
  • Fridkin, K. L., & Kenney, P. (2011). Variability in citizens’ reactions to different types of negative campaigns. American Journal of Political Science, 55, 307–325.
  • Garzia, D. (2014). Personalization of politics and electoral change. London: Palgrave.
  • Hazan, R. Y. (2005). The failure of presidential parliamentarism: Constitutional versus structural presidentalization in Israel’s parliamentary democracy. In T. Poguntke & P. Webb (Eds.), The presidentialization of politics: A comparative study of modern democracies (pp. 93–110). Oxford: Oxford University Press
  • Heffernan R., & Webb P. (2005). The British prime minister: much more than ‘first among equals.’ In T. Poguntke & P. Webb (Eds.), The presidentialization of politics: A comparative study of modern democracies (pp. 26–62). Oxford: Oxford University Press.
  • Johnston, R. (2002). Prime ministerial contenders in Canada. In A. King (Ed.), Leaders, personalities and the outcomes of democratic elections (pp. 158–183). Oxford: Oxford University Press.
  • Kaase, M. (1994). Is there personalization in politics? Candidates and voting behavior in Germany. International Political Science Review, 15, 211–230.
  • Karvonen, L. (2010). The personalization of politics: A study of parliamentary democracies. Colchester, U.K.: ECPR Press.
  • Kinder, D. R. (1986). Presidential character revisited. In R. R. Lau & D. O. Sears (Eds.), Political cognition (pp. 233–255). Hillsdale, NJ: Lawrence Erlbaum.
  • Kruikemeier, S., et al. (2013). Getting closer: The effects of personalized and interactive online political communication. European Journal of Communication, 28, 53–66
  • Lau, R. R., & Redlawsk, D. P. (2006). How voters decide: Information processing during election campaigns. Cambridge, U.K.: Cambridge University Press.
  • McAllister, I. (2007). The personalization of politics. In R. J. Dalton & H-D. Klingemann (Eds.), Oxford handbook of political behavior (pp. 571–588). Oxford: Oxford University Press.
  • McAllister, I. (2011). The Australian voter: Fifty years of change. Sydney: UNSW Press, 2011.
  • McAllister, I. (2013). The personalization of politics in Australia. Party Politics.
  • McAllister, I., & White, S. (2014). Electoral integrity and support for democracy in Belarus, Russia and Ukraine. Journal of Elections, Parties and Public Opinion, 25, 78–96.
  • Maddens, B., & Fiers, S. (2004). The direct PM election and the institutional presidentialisation of parliamentary systems. Electoral Studies, 23, 769–793.
  • Martin, S. (2010). Electoral rewards for personal vote cultivation under PR-STV. West European Politics, 33, 369–380.
  • Mughan, A. (2000). Media and the presidentialization of parliamentary elections. London: Palgrave.
  • Ohr, D., & Oscarsson, H. (2011). Leaders traits, leader image, and vote choice. In K. Aarts, A. Blais, & H. Schmitt (Eds.), Political leaders and democratic elections (pp. 187–219). Oxford: Oxford University Press.
  • Poguntke, T. (2005). A presidentializing party state? The Federal Republic of Germany. In T. Poguntke & P. Webb (Eds.), The presidentialization of politics: A comparative study of modern democracies (pp. 187–219). Oxford: Oxford University Press.
  • Poguntke, T., & Webb, P. (Eds.). (2005). The presidentialization of politics: A comparative study of modern democracies. Oxford: Oxford University Press.
  • Prior, M. (2007). Post-broadcast democracy. Cambridge, U.K.: Cambridge University Press.
  • Sanbonmatsu, K. (2002). Gender stereotypes and vote choice. American Journal of Political Science, 46, 20–34.
  • Scarrow, S. (2014). Beyond party members: Changing approaches to party mobilization. Oxford: Oxford University Press.
  • Schmitt-Beck, R., & Farrell, D. (2003). Do political campaigns matter? Campaigning effects in elections and referendums. London: Routledge.
  • Semetko, H. A., & Boomgaarden, H. (2007). Reporting Germany's 2005 Bundestag election campaign: Was gender an issue? Harvard International Journal of Press-Politics, 12, 154–171.
  • Sunstein, C. R. (2009). Going to extremes: How like minds unite and divide. New York: Oxford University Press.
  • Wattenberg, M. P. (1991). The rise of candidate-centered politics. Cambridge, Mass.: Harvard University Press.
  • Webb, P., & Poguntke, T. (2013). The presidentialisation of politics thesis defended. Parliamentary Affairs, 66, 646–654.


  • 1. The earliest postwar manifestation of a leader gaining particular the popularity was Pierre Trudeau, the Canadian prime minister first elected in 1968.

  • 2. The terms “candidate” and “leader” are used interchangeably. Some leaders (particularly in the United States) are also candidates. A candidate may also be someone standing at the local level.

  • 3. There have also been vigorous academic debates on the topic, the most recent of which are Dowding (2013) and Webb and Poguntke (2013).

  • 4. Other aspects of institutional design that have potential effects on how citizens view candidates are whether the constitution is unitary or federal, and unicameral or bicameral (see Bittner, 2011, pp. 16ff).

  • 5. A particularly important change was the elimination of single-member constituencies and with them the possibility that independent candidates would stand for election who could articulate the interests of local elites—and sometimes of ordinary citizens—against the federal authorities.

  • 6. Bittner (2011, pp. 139–140) identifies six broad conclusions. See also Karvonen (2010, pp. 105–107).