PLATFORM OR PERSONALITY? UNDERSTANDING THE ROLE OF LEADERS IN ELECTION OUTCOMES

by

AMANDA BITTNER

B.A. (Honours), University of Toronto, 2002 M.A., University of British Columbia, 2004

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

in

THE FACULTY OF GRADUATE STUDIES

(Political Science)

THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)

DECEMBER 2008

© Amanda Bittner, 2008 ABSTRACT

Campaign organizers and the media appear to agree that voters’ perceptions of party leaders have an important impact on the vote: substantial effort is made to ensure that leaders look good, that they speak well, and that they are up in the polls. Media reports during election campaigns focus on the horserace and how leaders are perceived in the public eye. In contrast, the academic literature is much more divided. Some suggest that leaders play an important role in the vote calculus,while others argue that in comparison to other factors (such as partisanship and the economy) perceptions of leaders have only a minimal impact.

Problematically, the literature on party leaders is diverse and non-cumulative.

Existing studies have been based primarily upon the analysis of only a single election and scholars have relied upon different survey questions in varying formats to inform their conclusions. These differences have resulted in the inconclusiveness of the literature. An effective evaluation of the role of party leaders requires a larger study, comparative across both time and space.

This study incorporates data from 35 separate election studies across a number of countries with varying institutional environments. It takes both a broad and in-depth look at evaluations of party leaders. I make five main conclusions. First, voters evaluate leaders’ traits in relation to two dimensions: characterand competence.Second, partisanship and ideology have a substantial influence on voters’ perceptions of party leaders, whereas issue attitudes and socio-demographics play a more minimal role. Third, voters perceive leaders through the lens of a partisan stereotype, in which Conservative leaders are seen to be more competent,

11 and leaders of Left parties are perceived to have more character. Fourth, political sophistication has a substantial effect on the way that voters perceive party leaders, as well as affecting the impact of those perceptions on vote choice. Fifth and finally,leaders matter— they have an influential effect on the individual vote calculus, as well as having a discernible impact on electoral outcomes.

111 TABLE OF CONTENTS ABSTRACT LIST OF TABLES vi LIST OF FIGURES ix ACKNOWLEDGEMENTS x DEDICATION xii

CHAPTER ONE: INTRODUCTION 1 Outline of Chapters 8 CHAPTER TWO: ELECTION STUDIES AND PARTY LEADERS 14 Introduction 14 The Richness of Data Availableon Party Leaders 16 KnowledgeofLeaders 17 Prospectiveand RetrospectivePeformanceEvaluations 18 PreferredLeadersb/pCandidates 18 Se/f-Reports:Most ImportantElectoralConsiderations 19 FeelingThermometers 20 Personality Traits 22 The Utility of Closed-Ended Traits Questions 25 Conclusions 27 CHAPTER THREE: CHALLENGES AND METHODOLOGY 28 Introduction 28 Analyzing Leader Evaluations on a Cross-National and Longitudinal Basis 29 ChallengesFaced in the Study 30 CodingVariablesin a CommonFormat 30 PoliticalSophistication 33 Cross-ParyLeaderEvaluations 36 Sa;wplingWezghts 38 Conclusions 40 CHAPTER FOUR: THE TWO-DIMENSIONAL STRUCTURE OF TRAITS 42 Introduction 42 The Structureof Traits as Seen in the Literature 43 .QuestionFormat 43 Method 45 Empirically Derived Structure of Traits 49 TheHurdles:Multi-Country,Multi-YearAna/ysis 49 The DimensionaliyofTraits: Convergencein ElectionStudies 51 RelationshipsBetweenTraits:AggregateVariationsin theData 53 The Underlying Structure of Traits: Character and Competence 57 CHAPTER FIVE: FACTORS AFFECTING EVALUATIONS OF PARTY LEADERS59 Introduction 59 Literature 60 The Dependent Variable: Evaluations of Leaders’Personality Traits 65 Impact of SES, PID, Ideology, and Issues on Trait Evaluations 68 The Impact of the Leader Herself: Women as an Example of the Impact of Socio-Economic Sameness75

iv The Mediating Role of Political Sophistication 79 Conclusions 84 CHAPTER SIX: THE IMPACT OF THE PARTISAN STEREOTYPE 86 Introduction 86 Cues and Stereotypic Thinking 87 WhataretheStereotypesAboutParties? 90 Impressions of Leaders’Traits 92 Conservativeleadersaremore“competent,“Left leadershavemore“character” 93 Partisan Stereotype: NotJust a Shortcut or Cue for the Least Informed 98 Conclusions 102 CHAPTER SEVEN: THE IMPACT OF LEADERS’ TRAITS IN ELECTIONS 104 Introduction 104 Theory 104 Are TraitsImportant?Whj ShouldWeFocusonLeaders? 104 HowMachDo EvaluationsofLeadersAffectVoteChoice? 108 Vote Models: The Relative Importance of Traits. in Voting for Three Main Parties 110 The Impact of Political Sophistication 118 Can Individual Leaders ActuallyMake a Difference? The Net Impact of Traits on Vote Choice: US Case Study 121 Conclusions 124 CHAPTER EIGHT: CONCLUSION 127 VotersPerceiveLeadersTraitsin TermsofCharacterandCompetence 132 PartisanshipandIdeologyAffectVoters’ImpressionsofLeaders,IssuesandDemographicsDoNot 133 VotersEvaluateLeadersin Relationtoa PartisanStereotype 133 LeadersMatter 134 PoliticalSophisticationMakesa Drence 135 TABLES AND FIGURES 138 BIBLIOGRAPHY 200 APPENDIX I: DATA,, QUESTION WORDING, AND CODING INFORMATION 219 APPENDIX II: THE CHALLENGE OF EXPLORATORY FACTOR ANALYSIS WITH MISSING DATA 250

V LIST OF TABLES

2.1 Surveys Including Questions About Party Leaders 138

2.2 Frequency of Question Types Asked About Leaders in All Surveys 139

2.3 National Election Studies Incorporating Traits Questions 140

2.4 Closed-Ended Traits Included in Surveys Across Space and Time 141

3.1 Major Election Issues 149

3.2 Political Sophistication: Measures Used to Construct Index 150

3.3 Political Sophistication Index: Multivariate Analysis 151

3.4 Relationship Between Demographics and Sophistication 152

3.5 Calculating Weights According to Total Population 153

4.1 Evaluations of Three Main Party Leaders on Trait: Strength of Leadership 154

4.2 Top “asked” Traits in Surveys: Number of Respondents Evaluating Specific Leaders’ Traits 155

4.3 Correlations of Evaluations for Traits of Two Main Party Leaders 156

4.4 Correlations of Evaluations for Traits of Three Main Party Leaders 158

4.5 Index Cohesion: Cronbach’s Alpha for Each Election Study 161

5.1 Effect of Socio-Demographics and Partisanship on Character Evaluations 170

5.2 Effect of Socio-Demographics and Partisanship on Competence Evaluations 171

5.3 Effect of SES, Partisanship, and Issue Attitudes on Character Evaluations 172

5.4 Effect of SES, Partisanship, and Issue Attitudes on Competence Evaluations 173

5.5 Evaluating Women Leaders: the Effect of Sex on Perceptions of Character and Competence 174

vi 5.6 The Role of Political Sophistication in Conditioning the Sources of Trait Evaluations of Centre-Left Party Leaders 175

5.7 The Role of Political Sophistication in Conditioning the Sources of Trait Evaluations of Conservative Party Leaders 176

5.8 Number of Respondents Evaluating Major Party Leaders, by Political Sophistication 177

5.9 Political Sophistication and Evaluations of Women Leaders: the Effect of Sex on Perceptions of Character and Competence 177

6.1 Effects of Party Label on Evaluation of Leaders’ Traits 179

6.2 Difference in Means on Evaluations of Most Frequently Asked Traits 180

6.3 Effects of Voters’ Partisanship and Leaders’ Party on Evaluations of Personality Traits 185

6.4 Effects of Voters’ Partisanship and Leaders’ Party on Trait Evaluations, by Level of Political Sophistication 187

7.1 Conservative Party Vote Choice: The Impact of Trait Evaluations 188

7.2 Centre-Left Party Vote Choice: The Impact of Trait Evaluations 189

7.3 Left Party Vote Choice: The Impact of Trait Evaluations 190

7.4 Predicted Probabilities: Changes in Probability of Vote for Party (over all other Parties) with Positive Evaluation of Same Party Leaders’ Character and Competence 191

7.5 Predicted Probabilities: Changes in Probability of Vote for Party (over all other Parties) with Changes in Explanatory Variables 192

7.6 Predicted Probabilities: The Effect of Partisanship on Changes in Probability of Vote for a Party with Positive Evaluation of Same Party Leaders’ Character and Competence 193

7.7 Vote for Conservative Party: Conditioning Role of Political Sophistication on Impact of Trait Evaluations 194

7.8 Vote for Centre-Left Party: Conditioning Role of Political Sophistication on Impact of Trait Evaluations 195

vii 7.9 Predicted Probabilities: The Effect of Political Sophistication on Changes in Probability of Vote for a Party with Positive Evaluation of Same Party Leaders’ Character and Competence 196

7.10 The Effect of Trait Evaluations on Republican Vote Choice (1972-2004) 197

viii LIST OF FIGURES

3.1 Categorization of Party Types, Based on Benoit and Layer’s (2006) PartyPolicyinModemDemocracies 148

4.1 Character and Competence Ratings of Leaders of Three Main Parties 160

6.1 Summary Statistics: Evaluations of Leaders of Three Main Parties’ Character and Competence 178

6.2 Competence Evaluations: Comparing Leaders of Three Main Parties to Average of All Leaders 181

6.3 Character Evaluations: Comparing Leaders of Three Main Parties to Average of All Leaders 182

6.4 Competence Evaluations: Comparing Leaders of Two Main Parties to Average of All Leaders 183

6.5 Character Evaluations: Comparing Leaders of Two Main Parties to Average of All Leaders 184

6.6 Summary Statistics: Evaluations of Leaders of Three Main Parties’ Character and Competence, by Level of Political Sophistication 186

7.1 Evaluations of Democratic and Republican Candidates (1972-2004) 198

7.2 Net Republican Gain from Evaluation of Candidates 199

lx ACKNOWLEDGEMENTS

In no way was this project a one-woman show. I could not have conducted the

research nor written the dissertation without the help and support of many others.

First, I would like those who helped me to access the data. They include Sophie

Holloway and Sharon Turner at the Australian Social Science Data Archive, Raphael Ventura

at the Israel Democracy Institute, Johan Fihn at the Swedish Social Science Data Service,

Statistics Sweden, and Helen Johnson at the UK Data Archive. My access to all of these data

was not guaranteed, and I appreciate all of these individuals and organizations for allowing

me to get my hands on these rich data sources. I would also like to acknowledge the financial

support of the University of British Columbia, the Isaak Walton Killam Memorial Fund, and

the Social Sciences and Humanities Research Council of Canada.

Tal Nitsan and Sascha Meyer-Diekena were an immense help, translating the Israel

and German Election Studies, respectively. Daniel Rubenson also helped with some last

minute translation of the Swedish Election Studies. Clare McGovern and Campbell Sharman provided me with not only unending entertainment with their in-depth knowledge of all things political in the UK and Australia, but also provided me with answers to “factual questions” and information about the two countries, assisting in the building of my political sophistication measure. Campbell Sharman was also instrumental in connecting me with the

Australian Social Science Data Archive, and has alwaysbeen extremely supportive of my work.

I would also like to thank colleagues at UBC and elsewhere who have been supportive, helpful, influential, and generallyinvaluable in terms of both the dissertation as

x well as my scholarly development. Gerry Baler (who first instilled my love of Canadian

Politics), Max Cameron, Greg Clarke, Gina Cosentino, Nick Dragojiovic, Lynda Erickson,

Joe Fletcher (who introduced me to the wonders of survey research), Elisabeth Gidengil,

Royce Koop, Scott Matthews, Laura Montanaro, Ben Nyblade, Mark Pickup, and Jason Roy.

Ken Carty deserves special mention for being unfailingly supportive since the moment I first

met him.

My family has also been instrumental—my dad has alwaysbeen my number one fan,

and I appreciate that he alwayslet me know it. Thanks to mom for believing in me and

teaching me the importance of going for it, and thanks to my grandparents and my sisters

and brother: Jen, Al, and Thomas, for being there and knowing that the end was in sight.

Thanks to Russell Williams, for putting up with all of my workaholic-related stress, for

reading drafts and talking with me about my research, for steadily and consistently reminding

me that my work does not define who I am, and for helping to make my life as happy as it is.

Finally, and perhaps most importantly, I would like to thank the members of my

dissertation committee, Richard Johnston, Fred Cutler, and Paul Quirk. Writing a

dissertation when you’re in a different country from your supervisor, and at least three time

zones away from the rest of your committee is not normally recommended. Through their

readiness to quickly respond to emails and phone calls, and their unceasing willingness to

read drafts and provide feedback and advice without delay, they made it not only manageable, but also enjoyable. Dick and Fred deserve extra thanks in particular, for seeing me through the MA thesis and agreeing to stay on for more. They have been the best

teachers and the most supportive mentors a person could have, and I count myself lucky to have them as colleagues.

xi DEDICATION

To my mom and dad, Olivia and Bill Robertson.

xii CHAPTER ONE: INTRODUCTION

NormativeTheorists.. . havea tendencyto think ofpolifycongruenceas the on!ylegitimatebasisfor representation,and to denigrate extrapoliy basisfor representationas “sjmbolic.“. . . It mqy be objectedthat a searchfor supportthat stressessy1isticcompatibilities.. . easi’ydegeneratesintopure image selling.And, of course,it maj. But the searchfor support that emphasies poliy coinpatibilities.. . easii degeneratesintopurepositiontaking...Positiontakingisjust as misleading to constituenciesand as manzpulativeoftheirdesiresas imageselling.It maj bejust as sjmbolicas anyform ofcandidateadvertising(Fenno1978,242).

Do leaders matter for the outcomes of elections? And what does it mean to say that

they might? Political scientists seeking answers to these questions in the past have disagreed

about the impact of leaders on voter decision-making and electoral outcomes. Johnston et al.

(1992) argue that leaders have a large effect in Canada, pointing to “the pivotal contributions

of Laurier, Diefenbaker, and Muironey,” suggesting that “each was, in Hollywood terms, cast

against type; his accession signaled that his party was now prepared to take a hitherto

neglected region or group as seriously as it possibly could. Each leader recast the entire party

system” (169).While the authors refer here to the symbolic representation of groups as imbued in the leader, Johnston’s later work (2002) goes a step further and suggests that

Canadians take leaders’ traits into account when making their decisions at the ballot box.

That leaders matter is not universally accepted, however. Other studies of Canadian elections have found that evaluations of leaders have only a weak influence on individuals’voting decisions (Blais et al. 2002). Once other factors (such as socio-demographic variables, values, beliefs, and economic perceptions) are considered, leaders themselves account for very little in voters’ decisions and the subsequent election results.

The disagreement about the impact of party leaders found in the Canadian literature is evident in other countries as well. In his recent discussion of the state of the debate,

1 Anthony Mughan notes that “at one extreme, they may have no independent effects in, say,

a highly polarized election where the partisan lines are clearly and uncompromisingly drawn.

At the other extreme, however, leaders may be the difference between victory and defeat for

their party when an election is closely fought” (2005, 1).The real issue for Mughan is that

skeptics of the notion that leaders matter rest their case on the failure of leaders “. . . to swing

the outcome of a particular election in favor of one party or another” (2005, 2).

Indeed, in his introduction to an edited volume on the effects of party leaders,

Anthony King makes this point quite clearly.He states, “if the party with the less-well

esteemed leader wins, then the outcome of that election cannothave turned on voters’

judgments of the respective qualities of the various parties’ leaders or candidates. Leadership

effects cannothave been decisive; they can, at most, only have reduced the victor’s margin of

victory. Full stop, end of story” (2002, emphasis in original). Mughan argues that this view

constitutes “an extreme and unrealistic criterion of substantive electoral relevance” (2005,2-

3). He suggests that the state of the economy may not swing the outcome of an election in

favour of one party or another, but we are happy to continue to believe that the economy

plays a critical role in determining election results.

That this more extreme criterion exists with respect to the importance of leaders is indicative that there is more to the story than a simple difficulty in measurement or mere

conflicting evidence. The bigger issue, as the above quotation from Richard Fenno hints at, is normative.That is, that in a democracy, individuals oughtto consider “higher” factors like platforms and issues, rather than base, “symbolic” factors like personality characteristics when they go to the ballot box. Indeed, after discovering that people do incorporate party leader images into their vote calculus, the assumption was that party leaders served as a sort

2 of information shortcut, helping those with less information about things like policies and

issues to make decisions. Thus it was still lamentable that people were considering party

leaders, but it was explainable because it had to be the least informed who were doing it.

Scholars then sought to compare the decision making processes among the least politically sophisticated and the most politically sophisticated segments of the population

(Cutler 2002; Glass 1985; Miller,Wattenberg, and Malanchuk 1986). Generally, the

expectation was that the less sophisticated ought to depend more heavily on impressions of leaders, while the more sophisticated, with more information and a greater ability to make more complex decisions, ought to focus more heavily on policy platforms and issues when deciding how to vote. In fact, as all of those scholars found, the more sophisticated were as likely or even more likely to consider leaders as their less sophisticated counterparts (Cutler

2002; Glass 1985; Miller, Wattenberg, and Malanchuk 1986). This finding fit with observations in the literature surrounding citizen competence, as the more sophisticated were found to generally consider a greater number of factors than the less informed overall, including factors which had been thought to serve as a “shortcut” for the least informed

(Sniderman, Brody, and Tetlock 1991).

The question then arises: why are people evaluating leaders in the first place? Is it simply an information shortcut for the least sophisticated? If so, how do we explain that the more informed also use this “shortcut”? Or, do leaders provide voters with a more complex type of information that the most sophisticated are able to tap into (Glass 1985)? Is it a bit of both, or something else entirely? And importantly, especially for those concerned with the normative question, is the evaluation of party leaders consistent with what we would desire in a political system with democratic elections?

3 The real problem is that in order to tap into the why, momentarily setting aside this

normative issue about what individuals oughtto consider when they vote, there are a host of

other questions that remain largelyunanswered. While scholars have spent significant time

and effort in assessing the role of leaders in elections, much of the literature surrounding

leaders is diverseand non-cumulative.Studies have focused on everything from the impact of

leaders’ appearance (Rosenberg et aL 1986) to the effect of television in affecting the extent

to which leaders are a focus for voters during election campaigns (Mendelsohn 1993).While

strands of this literature are substantially disconnected, there do appear to be five clear

categories or areas of study that emerge. These are:

• information sources • the issues surrounding leadership selection • leadership traits • the factors influencing candidate evaluation • the impact of leaders on electoral outcomes

The literature on information sources touches on the role of the media in priming

and framing (Mendelsohn 1994;Mendelsohn 1996), agenda-setting (Lupia 1992;McCombs

and Shaw 1972), and the role of key campaign events such as debates (Blais and Boyer 1996).

The leadership selection literature looks at the impact of leadership change and the efficacy

of changing the leader for improving a party’s success (Nadeau and Mendelsohn 1994;

Stewart and Carty 1993). On the topic of leaders’ traits, we find a rich but inconclusive

assessment of the types of individual traits that voters take note of as well as uncertain

conclusions regarding the underlying “structure” or “dimensionality” of personality traits and how voters perceive, process, and assess traits (Bartels 2002; Bean 1993;Bean and

Mughan 1989; Brown et al. 1988;Hudson 1984;Johnston 2002; Kinder 1983; Kinder 1986;

Kinder, Abelson, and Fiske 1979; Kinder et al. 1980;Rosenberg et aL 1986).

4 Related to the question of traits is the issue of where evaluations come from in the

first place—the types of factors that affect trait assessment and that influence perceptions.

This is one of the areas of scholarship on leaders that is quite inconclusive, with studies

pointing in a number of different directions. Some suggest voter demographics matter

(Cutler, 2002), others point to the importance of partisanship and ideological leanings of

both voters and leaders (Bartels 2002, Graetz & McAllister 1987),while others suggest that

policy/issue related factors are key (Weisberg & Rusk 1970, Rusk & Weisberg 1972).Some

studies note the importance of all of the above in different circumstances (Conover &

Feldman 1989). Furthermore, there is the ever-present question regarding the role of leader

evaluations: do people evaluate leaders as a way of overcoming their information shortfalls,

or are evaluations a complex task useful primarily to those who are able to translate impressions of personality into something meaningful (Glass 1985)?

The last main “category” of analysis that has emerged over the years is on the impact

of leaders on electoral outcomes. As I have already noted, scholars tend to disagree on this issue, and while some studies have been conducted, we do not really have the definitive word on the effects of patty leaders on vote choice and outcomes, as is clear from Anthony King’s

(2002) collected volume, as well as Mughan’s (2005)assessment of the state of the literature.

In fact, as King notes, we reallydo not know very much about this issue at all: “published writings are relatively sparse and sustained empirical research even more so, with the result that an assiduous student could absorb the entire relevant academic literature in less than a week” (2002: 3).

In fact, it takes more than one week to read all of the literature on party leaders, but

King’s point is quite accurate: the literature is sparse, and furthermore, the linkages between

5 the different categories of analysisare not very strong. In short, the literature is diverse and non-cumulative, and the general findings are anything but conclusive. This may seem strange, particularly given the kind of importance allocated to leaders by the media, with its constant coverage of the daily activities of party leaders, as well stories about polls and other campaign news portrayed in terms of the horserace. Furthermore, party operatives and campaigners give great weight to the image of leaders, as evidenced by all of the money spent on image consultants and speech and diction therapy.

So if it is so “obvious” that leaders are important, then we ought to have a better understanding of peoples’ evaluations of leaders and the impact of those evaluations.

Unfortunately, analyses of the role of leaders tend to be based upon a single-country cross- sectional analysis of a single election (for example, Brown et al. 1988; Conover and Feldman

1989;Mendelsohn 1994; Stewart and Clarke 1992).Longitudinal analyses have been conducted, but these results are largely based on single-country studies (Clarke,Ho, and

Stewart 2000; Gant and Sigelman 1985; Gidengil et al. 2000; Graetz and McAllister 1987;

Hudson 1984; Stokes, Campbell, and Miller 1958). Some comparative work has also been done, but this has tended to involve only one or two countries, and usually only one election in each (Bean 1993; Bean and Mughan 1989). Some scholars have conducted multi-country studies (Banducci and Karp 2000; Graetz and McAllister 1987), however these too tend to rely on a single election in each country.

It follows, then, that we lack conclusive knowledge because each of these studies have been looking at different countries, different years, different sets of leaders, and drent questionsin the election studies being analyzed. These studies have provided a lot of detailed and contextual insight into the elections fought in a particular year in a particular country

6 with particular leaders, but a broad or generaliaable idea of voters’ evaluations of party

leaders, or the impact of those evaluations across space and time, is still lacking.What is

needed is a more systematic and comparative look at:

• Which traits do people think about? • Is there an underlying structure or dimensionality to traits? • Where do leader evaluations come from? What are the respective roles of socio demographics, partisanship, party labels, ideological views, and issue attitudes? • Where do leader evaluations fit into voting models? • How much do leader evaluations affect electoral outcomes? • Why do voters use evaluations of political personalities in their vote choice? o Is it a shortcut? o Is it an easy cue? o Is it something complex with “real” information to be gleaned? • Are there differences across levels of political sophistication that might give us greater insight into the evaluations of leaders and the impact of those evaluations?

This study fillsthe gap. It does so with a cross-national and longitudinal analysisof

evaluations of leaders, using data collected in the election studies of a number of different

countries. This study builds upon existing research, by taking an assessment of the role of leaders’ traits one step further than has been done to date. It will look more systematically

and comparatively at three main strands of the literature: traits and trait dimensions; the

factors influencing leader evaluation; and the impact of evaluations on electoral outcomes.

By moving beyond a research methodology based on a single election in a single country, it is possible to better understand how and why leaders are evaluated, and what impact these evaluations have on voters’ decisions, and by extension, election results.

Before moving to an overview of chapters, an issue of terminology needs to be addressed. Throughout this dissertation, I discuss evaluations in terms of leaders.In fact, some of these leaders are candidates,specificallyfor President of the United States. Because of the greater number of party leaders in the study, I adopt the term generally and use it to refer

7 to both leaders and candidates for simplicity’ssake. In the final chapter, however, I discuss

US candidates in particular, and in that discussion I refer to them as candidates, rather than

leaders.

Some might suggest that the terminology may result in an overstatement of the

effects of evaluations of traits—arguably, in a system where individuals compete as

candidates and receive votes as candidates, evaluations of the individual are more relevant. In

parliamentary systems, however, where citizens vote for parties and sometimes do so at a

local level (e.g.in Single Member Plurality electoral systems), evaluations of the individual

leader should matter less. By combining all leaders and candidates together, therefore, we are

at risk of confusing the effects of candidate evaluations for leader effects. While the bulk of

the analyses presented in this dissertation are based on pooled data, country-by-country

analyses (results not shown) indicate that leadereffects exist outside of the United States, and

are not simply the result of overwhelming candidateeffects. The use of the term leader,

therefore, is an issue of nomenclature, and simplifies discussion of the analysespresented in

this dissertation.

Outline of Chapters

Chapter two chronicles the information that has been collected on party leaders to

date. In this chapter, I examine over 100 election studies from 19 countries, and I ifiustrate

that the problem with the current state of the literature, and its lack of clarity about voters’

evaluations of party leaders, stems from the very diversity in the types of questions that have

been asked across countries and over time. Simplyput, scholars have based their analyses on very distinct and dissimilar types of survey questions over time, which has led, I believe, to

8 the lack of comparability of conclusions. Essentially, the wide variety of data collected and analyzed has contributed to the circumstance in which scholars (all focusing on leaders) are essentially talking past one another, and largelydisagreeing with one another.

In this chapter I argue in favour of an approach that focuses on the analysisof voters’ evaluations of leaders’ characteristics, evaluations that respondents have provided in a closed-ended format. By focusing on the evaluation of traits specifically,we are able to reduce some of the risk of error and bias, including coding error (Fink 1995),interviewer bias (Shapiro 1970), and respondent misinterpretation of the question (Tourangeau and

Rasinski 1988). Furthermore, it will allow us to examine similar traits over time and across countries, with less item non-response than open-ended questions tend to 1have. Although a wide variety of different traits have been included in election studies over time, there are also some important areas of overlap. Thirty five election studies across seven countries (including Australia, Britain, Canada, Germany, New Zealand, the United States and Sweden), incorporate closed-ended questions about leaders’ traits, providing a rich set of data for comparison. A dataset of this size allows us to establish the extent to which there are similaritiesin the types of traits that matter in the evaluation of leaders, as well as similaritiesin the types of traits that matter most for election outcomes.

In the third chapter I outline the methodological challenges involved in a cross- national and longitudinal study of this magnitude, as well as the strategies employed to prevent those challenges from becoming barriers. A concern that exists for any cross- national or longitudinal study is the coding of similar but not identical questions into

I A look at the US National Election Studies Cumulative file, which merges American National Election studies conducted between 1948 and 2002, suggests that between 45 and 47% of respondents have no thoughts, either positive or negative, about Presidential candidates when asked to provide feedback in an open-ended format. The closed-ended question format provides a larger N for analysis of specific trait types.

9 common formats. Question wording changes over time and different studies ask things in

different ways, which makes it difficult to analyze more than one study at a time. This issue

was felt even more heavilygiven my spatial and temporal coverage, making it even more

likely that there would be substantial differences in question formats on the issues and

variables of interest. The issue of political sophistication is a case in point. The creation of a

measure of political sophistication where no common measure exists across studies is a task

rife with complications, and in chapter two I provide details on my efforts to use a

sophistication measure in all 35 studies.

Another difficulty intrinsic to this project is the comparison of evaluations of

different leaders, from different countries, in different years, from different parties. How is

one to compare evaluations of leaders of the German FDP with other leaders when the FDP

doesn’t exist anywhere but Germany? In Chapter Three I outline this problem in greater

detail. I employ a party classification system in order to group together “similar” parties from

different countries, facilitating the comparative analysis that is so important for our

understanding of party leaders.

Sampling weights provided an additional challenge, inherent in the analysis of so many election studies at once. How do we ensure that one country, or even one study doesn’t flood the entire sample and skew the results? For example, the 2000 National

Annenberg Election Study, a veritable goldmine of data, includes approximately 50,000 respondents, all of whom provided evaluations of party leaders. Chapter Three discusses the approach taken in order to ensure that the opinions and attitudes of these respondents did not outweigh the less than 4000 respondents in most other election studies. This chapter also includes an overview of the methodology of the study as a whole.

10 The fourth chapter plunges into the data analysis.In this chapter I ifiustrate the linkages that exist between different traits. Based on the analysis of the closed-ended traits questions, I assert that traits fallinto two dimensions: character(incorporating traits such as honesty and compassion) and competence(including traits such as intelligence and strength of leadership). These findings support those of previous scholars, in particular the pioneering efforts of Donald Kinder (1979; 1983). One of the tangible benefits of the analyses in

Chapter 3 is to put to rest the issue of how many dimensions of traits people consider when they evaluate leaders. We can then move forward, comparatively, looking at both the kinds of factors that influence voters’ perceptions of leaders on the two trait dimensions, and also the impact of these trait dimensions on vote choice.

Chapter Five begins this process. In this chapter, I look at the origins of leader evaluations. Existing research on this question, generally based on the analysisof a single election in a single country, points to divergent sets of factors that influence voters’ evaluations of leaders. This analysisclarifies the relative impacts of socio-demographics, issue attitudes, partisanship, and ideology.

By moving to a cross-national and over-time setup, a number of interesting conclusions emerge. Individuals’ personal partisanship and ideological views play the largest role in influencing their evaluations of party leaders. Attitudes towards issues play a secondary role, while socio-demographic characteristics have a more diminished impact on the perception of leaders. There does appear to be some indication, however, that a characteristic intrinsic to the leader (the leader’s sex) may play quite a substantial role in influencing evaluations. Finally, there are significant differences in the way that political sophistication conditions perceptions of party leaders: the more sophisticated consider more

11 factors overall, and tend to weigh those factors more heavily in their ratings of leaders’

character and competence.

In the next chapter I continue to assess the origins of voters’ evaluations of leaders’

traits. Here I focus more heavily on the party label of the leader, as I outline a theory of a

“partisan stereotype,” in which the leader’s party label heavilyinfluences voters’ perceptions

of that leader on the two trait dimensions. Conservative party leaders are generally rated

more positively on the competencedimension, while Left party leaders are generally given the

highest ratings on the characterdimension. This pattern exists even when we control for the

party identification of the voter, and is actually amplified among the most politically

sophisticated segments of the population.

That an enduring partisan stereotype exists, and that it exists across national borders, is one of the more exciting findings in this study, and in this chapter I discuss potential

reasons for it. I suggest that the party system itself may play an important role in shaping our perceptions of leaders, and that fundamental political values (such as attitudes towards authority) have a stable and substantial influence on the types of characteristics that the leaders of certain types of parties oughtto have.

Finally,we get to the issue that much debate centers on. How do evaluations of party leaders affect vote choice? Chapter Seven builds on the findings in previous chapters, moving evaluations of leaders’ traits from one side of the equation to the other. The chapter has a few key findings. First and foremost, leaders matter. Second, evaluations of leaders’ character affect vote choice more than evaluations of leaders’ competence. Third, there are important differences in the way that voters perceive party leaders depending on their level of political sophistication, and these differences have implications for understanding the

12 evaluation process. Finally,while leaders play an integral role in the individual vote calculus, it seems that the net effects of leaders on the popular vote range substantially across elections.

13 CHAPTER TWO: ELECTION STUDIES AND PARTY LEADERS

Introduction

Scholars and pundits alike point to the importance of party leaders. Newspaper

stories and television broadcasts of election campaigns focus on the horserace and

comparative standings of leaders, reporting leaders’ performance in polls, campaign events,

and debates. Image consultants are brought in, PR specialists provide detailed fashion

editorials in national newspapers, and leaders have been known to undergo speech therapy

between elections to improve future chances. Election results are often attributed to party

leaders specifically,rather than parties as a whole. And while the media seem convinced of

the paramount role played by leaders in elections, the scholarly literature (although sparse) is

substantially more divided.

One of the primary reasons that scholars have not come to a common understanding

of either a) the types of leaders’ characteristics that voters concentrate on; b) where their

perceptions of leaders’ traits come from; or c) the impact of those perceptions on election

results; is that they have been looking at different variables. Studies to date have focused on

different aspects of evaluations of leaders, and have come to different conclusions because

they are basing their analyses on voters’ responses to different questions. For example, some

scholars have focused on overall feelings towards leaders (thermometers) to inform their understanding of the role of leaders (Aarts and Blais 2009; Blais et al. 2002; Blais et a!.2000;

Gidengil et a!. 2000; Rahn et a!. 1990),while others have relied upon approval ratings as a

14 measure of voters’ perceptions of leaders (Clarke, Ho, and Stewart 2000; Clarke and Stewart

1995;Jones and Hudson 1996;Nadeau, Niemi, and Amato 1996).

Among those scholars who have sought to understand the dimensionality of leaders’ traits and how voters perceive leaders, many have based their analyses on open-ended questions (Bean and Mughan 1989; Glass 1985; Kinder et al. 1980),while other studies have relied upon closed-ended traits questions (Bean 1993; Gidengil, Everitt, and Banducci 2006;

Johnston 2002; Kinder 1983). Our knowledge is therefore limited by the types of questions that have been asked in surveys over time.

In this chapter I chronicle the information that has been collected to date by examining over 100 election studies from 19 different 2countries. I make three main conclusions: first, that while there is not a lot of research that has been conducted on evaluations of party leaders, election studies have asked a lot of questions about leaders over the years, providing a wealth of information available for more detailed study. Second, I point to the incredible variety in the types of questions that have been asked in all of these studies and suggest that the reason we lack conclusive evidence on the role of leaders is because we lack a consolidated approach. Third, I argue in favour of a cross-national, longitudinal study of voters’ evaluations of party leaders, based on closed-ended questions probing for evaluations of leaders’ traits, as these questions provide the best way to move forward and explore the role of leaders both broadly (across countries and over time), as well

2 Countries in the analysis include Australia, Britain, Canada, the Czech Republic, Denmark, Finland, France, Germany, Hong Kong, Israel, Italy, Japan, New Zealand, the Netherlands, Russia, Norway, Sweden, Switzerland, and the United States. Election Studies were chosen based on both whether or not they could be found, based on a search of available sources such as Michigan’s ICPSR, individual study websites, and other online archives (such as ECPR), as well as whether the codebooks and questionnaires could be obtained in either English or French. Not all of these studies incorporated questions about leaders, and some were therefore subsequently dropped from the analysis.

15 as in-depth (looking at the dimensionality of traits, the origins of trait evaluations, as well as

the impact of perceptions of leaders’ traits on vote choice).

The Richness of Data Available on Party Leaders

Anthony King observes the scarcity of scholarship on the role of party leaders in

elections. He suggests that it should not take much more than a week to learn everything

there is to know about party leaders (2002), and his estimate is not far from the truth. But

the lack of scholarship on this topic is not necessarily the result of a lack of data. Election

studies from around the world have consistently included questions about party leaders,

providing the means with which to examine the role of leaders in elections.

As a whole, national election studies provide the largest body of data upon which

scholars have based their analyses of leader evaluations. Some have conducted experiments

to investigate voters’ evaluations of party leaders (see, for example, Lodge, McGraw, and

Stroh 1989; Rahn 1993; Rosenberg et al. 1986), but the norm is to analyze election study

data. Table 2.1 lists the election studies that have incorporated at least one question about

party leaders. There is a wide variety of information collected about party 3leaders. The predominant questions include thermometers which tap respondents’ overall “feelings”

towards leaders; questions probing knowledge about party leaders; prospective and retrospective performance evaluations; respondents’ preferred leadership candidates;

respondents’ impressions of the importance of party leaders in determining their own vote

3 When compiled all together, a complete list of the questions asked about leaders in the 93 election studies consists of a document 117 pages long (single spaced).

16 choice; and open-ended and closed-ended traits questions. This section describes these

different types of questions in more detail, and highlights research that has been conducted

with these questions. I argue that because the questions are allvery different they lead us to

different kinds of findings.

KnowledgeofLeaders

Many questions found in the election studies tend to get at specific issues or aspects

of leader evaluations, including probing respondents for their level of knowledge regarding leaders. These questions can take a number of formats, including showing respondents a picture of a leader and asking them to identify the individual’sname, party, and position in government, while other surveys have asked respondents to recall the names of as many leaders as possible. Others still ask for more factual details about the individual leaders - religious affiliation, military experience, state of residence/origin, etc.

Generally these questions tend to be used not in the context of an analysis of the role of leaders, but as indicators of political knowledge. A number of studies incorporate respondents’ ablities to identify and recall party leaders into indices measuring factual knowledge (Bittner 2007; Matthews 2006; Roy 2007). One study of note which looked in greater depth at respondents’ knowledge of candidates was conducted in the Canadian context (Blais et al. 2000), but the majority of studies looking at the evaluation of leaders do not focus on knowledge.

17 Prospectiveand RetrospectivePetformanceEvaluations

Respondents were also asked fairlyfrequently to evaluate both past performance of

leaders, as well as expectations of future performance. These two performance-related

question types were coded separately, and the analysisof questionnaires indicates that

respondents have been asked to evaluate past-performance nearly twice as often as

expectations of future accomplishments. The presence of these questions in election studies

follows from a rich literature on the factors informing vote choice.

Originating from discussions by scholars such as Downs (1957) and Key (1996),who

put forth theories of prospective and retrospective voting, respectively, scholars have

extended these theories to evaluations of party leaders and the nature of the impact of

perceptions on vote choice. Nadeau, Niemi, and Amato (1996) compare retrospective and

prospective leader evaluations, and argue that the prospective (who would make the best

Prime Minister) measure outperforms the retrospective (Prime Ministerial approval) measure

in explaining vote choice. Others have surmised that voters may perceive leaders’

characteristics as an indicator of future performance (Miller,Wattenberg, and Malanchuk

1986), suggesting that voters think prospectively when forming their choices.

PreferredLeadershipCandidates

Another common question type across election studies is a question that asks individuals to indicate which leadership candidate is or was their favourite, or who would

make the best President or Prime Minister or Chancellor. Scholars have noted that these measures tend to be highly correlated with thermometers, performance-related questions, as

18 well as evaluations of specific personality traits, suggesting that although different, they are all tapping into similar perceptions of party leaders (Abelson et al. 1982; Kinder, Abelson, and Fiske 1979; Weisberg and Rusk 1970).

Sex-Reports:MostImportantElectoralConsiderations

The impact of leaders on the vote calculus is one of the issues that has seen the most debate in the literature. One of the questions incorporated fairly frequently in election studies asks respondents to reflect upon what was most important in their vote choice: the leader, the party, the candidates, or the platform. This question is asked in slightlydifferent formats (not all survey questions include candidates as an option, and not all questions include platforms as an option), but the question is surprisingly consistent across time and space, suggesting that comparative analysisis certainly feasible.

Research to date, however, suggests that retrospective evaluations of one’s

“reasoning” is not necessarily a reliable way to determine “real” thought processes, as it tends to result in post-hoc justification (Lodge and McGraw 1995). Lodge and his colleagues at Stony Brook are perhaps best known for their on-line model of candidate evaluation, suggesting that “impression-based processing implies that although citiaens generally cannot recount accurately the specific bits and pieces that originallyled to their preferences, the considerations that actually entered into the evaluative tallywere factored in at the moment of exposure, and not held in memory until some future date, when respondents would then be forced to construct an evaluation from whatever evidentiary residue still resided as memory traces” (Lodge and McGraw 1995, 113).This perspective suggests that on a more basic level, most attitudes have some origins hidden in the subconscious, which individuals

19 are not able to access retrospectively, leaving these types of responses less reliable (Macrae

and Bodenhausen 2000).

FeelingThermometers

Feeling thermometers provide one of the main sources of information to scholars

seeking to understand the role of leaders in elections. As Table 2.2 illustrates, with the

exception of “other” (a conglomeration of often country and election-specific questions), the

thermometer is the most frequently included question about leaders in election studies.

While not alwayslabeled a “thermometer” in every election study (the Germans callit a

“scalometer”), and not alwaysmeasured on the same scale (historically,scale formats include

0-100, 1-10, 1-5, and -5/+5), these questions allgenerally take a similar format, as exemplified by the British Election Study of 1996: “Using a scale from 0-10, please show how much you like or dislike the party leaders,” at which point respondents are given the names of party leaders one at a time, and are asked to give a number on the scale,where the higher the number, the more the respondent likes the leader.

The prevalence and similar formats of this question (all questions, regardless of original scale, can be recoded into a common format) facilitate both cross-national and longitudinal comparison: indeed, a forthcoming study by Kees Aarts and Andre Blais relies upon the feeling thermometer for their comparative analysis (Aarts and Blais 2009). Most scholars making use of this question tend to either do so in the context of only one election

(Clarke, Ho, and Stewart 2000;Jones and Hudson 1996), or else a longitudinal study within a single country (Gidengil et al. 2000). Thus while this question may facilitate comparison, it has not really taken place yet.

20 There is some indication, however, that while this question may be incorporated into

election studies frequently and in a relatively common format, it may not be the best way to

tap into attitudes about leaders. There is some issue over what the question actuallymeans,

especiallywhen one takes into account the scale format and question wording. Norbert

Schwar (1998) notes that the scale of the thermometer has an important effect on an individual’sperception of the question. He suggests that respondents may view a +5/-5

entirely differently than a 0-10 scale, because the former implies the existence of both positive and negative factors to be evaluated, while the latter implies finer gradations of an

overall positive evaluation. The example he gives is of evaluations of success, and suggests that the former scale implies the evaluation of both failures and successes (a bipolar dimension) where the latter may lead a respondent to perceive the question to be asking about varying degrees of success (a unipoiar dimension). Thus the scale itself may in fact have an important impact on the results of a given study.

Weisberg and Miller (1980) assessed the feeling thermometer in the 1979 NES Pilot

Study, and found that the way in which the scalewas labeled affected the nature of the response. Most importantly, they found that respondents tended to stick to the labeled points on a scale (96% of respondents, in fact, gave ratings on one of those major points), which led to a substantial difference in the responses given in the face-to-face interviews in comparison to the telephone interviews, where respondents were not and could not be given labels in the explanation of the feeling thermometer. Their conclusions led them to recommend labeling only three points on the thermometer (the two ends and the mid-point), thus bringing the measure in-line with its telephone counterpart, as well as ensuring that respondents felt freer to provide a more nuanced rating. These observations fit with

21 understandings of the way in which individuals respond to surveys, and the importance of

question phrasing (Tourangeau and Rasinski 1988; Traugott 1990; Zaller and Feldman 1987).

The thermometer is a very general measure, and it also may not be the most reliable

source of information about voters’ evaluations of leaders. As Johnston notes, the feeling

thermometer “...carries too much nonpersonality freight. Even with party identification and

the like controlled, it is still infused with party, group and policy judgments...” (2002: 174).

It seems that the best way to understand evaluations of leaders is to try to look at questions

specifically probing for impressions of traits themselves. By focusing solely on traits, we may

get a more precise picture of what voters think about when they evaluate leaders and how

those evaluations affect vote choice. Furthermore, we can gain greater insight into which

traits matter most—is it traits relating to intelligence and strength of leadership or honesty

and trustworthiness? Traits questions have been incorporated into many election studies, and

provide substantial insight into voters’ perceptions of leaders.

Personality Traits

In the 93 election studies that include at least one question about leaders in one

format or another, 60 incorporate questions about leader traits, either open or closed. Table

2.3 lists the number of national election studies that incorporate traits questions. Of these, 26

make use of open-ended questions about leaders, and 46 include closed-ended traits 4questions.

4 combined number of both types of questions exceeds 60 because of the overlap in some studies which include both open-ended and closed-ended questions. 51t is important to note that all of the estimates of the frequency of question types and trait types are on the conservative side, because the analysis is based solely on questionnaires which could be found in either English or French. The studies listed above are not exhaustive, and there are other election studies that have taken place that I

22 Closed-ended questions take a form similar to that found in the 2000 Annenberg

Election Study: “Does the phrase ‘honest’ describe George W. Bush extremely well, quite

well, or not too well at all?” Open-ended traits questions are usually asked in the format of

(as in Australia, 1990) “Is there anything in particular that you like about Bob Hawke?

yes> What is it you like about him?” and “Is there anything in particular that you don’t like

about Bob Hawke? What is it that you don’t like about him?” These open questions

allow for more variety in responses, allowing respondents to list their perceptions about

traits that have not necessarily been pre-determined by researchers.

Scholars have made use of both question formats. Some have focused on open-

ended traits questions (Bean 1993; Butler and Stokes 1969; Butler and Stokes 1974;

Campbell et al. 1960; Glass 1985; Kinder et al. 1980; Miller, Wattenberg, and Malanchuk

1986; Stewart and Clarke 1992), while others have relied upon responses to questions in the

closed-ended format (Bartels 2002b; Bean and Mughan 1989; Brown et al. 1988; Gidengil,

Everitt, and Banducci 2006; Hayes 2005; Johnston 2002; Kinder 1983; Kinder and Fiske

1986). Both methods have provided considerable insight into voters’ evaluations of party

leaders. However, the difficulty, again, lies in the conflicting conclusions that have been

drawn from the different analyses.

In their research into the effect of specific traits on electoral outcomes, Bean and

Mughan (1989) compared the impact of closed-ended traits on vote choice in Australia and

Britain. They found that four main leaders traits played a key role: effectiveness, listening to reason, caring, and sticking to principles. Four years later, Bean (1993) conducted another comparative analysis, this time using responses to open-ended questions. In this study, he do not include in the analysis. These other election studies are likely to have incorporated questions about leaders, and many probably also prompt respondents to evaluate personality traits.

23 argued that there were seven main categories of traits that had an important impact on vote

choice in Australia and New Zealand: competence, integrity, strength, harmony, general

likeability,other personal, and party/policy/group. Each study involved a comparison of the

effects of traits in two countries, but they used different types of questions, giving rise to

broadly different results. Both studies were conducted by the same scholar, suggesting that it wasn’t simply a different research perspective that led to these different conclusions.

The utility of a broad comparative analysisusing a common measure across studies is

apparent. But are open or closed-ended questions better? The literature is not conclusive.

However, some suggest that open-ended questions can be unreliable, much in the same way

that individuals’ retrospective evaluations of their own thought processes are not necessarily

accurate. A number of scholars note the potential for error in relation to open-ended

questions: they can be difficult to code and categorize (Fink 1995), and they may be subject

to interviewer bias in a way that closed-ended questions may not be (Shapiro 1970).Rahn et al. (1994) also argue that open-ended questions are not well-suited for understanding

candidate preferences because they tend to reflect rationalizations of preferences rather than the reasons that actuallyled to those feelings. Finally, since the role of interpretation plays such a fundamental role in the response process (Tourangeau and Rasinski 1988), the use of open-ended questions may be particularly problematic, because they leave open the possibility that respondents may interpret the question in a way that differs from the researchers’ intentions. This could lead to divergent answers from respondents in the same study, who may have interpreted the same question in different ways.

24 The Utility of Closed-Ended Traits Questions

For the purposes of a cross-national, longitudinal assessment of the evaluation of leaders, closed-ended traits questions provide the ideal foundation with which to move

forward. Arguably, they are best able to target specific traits, likelythey are the most robust

measures with the least coding and response error, and there is also substantial overlap in the traits that have been incorporated in election studies around the world.

Generally, the two question types (open and closed) tend to be incorporated on their own in surveys—investigators will usually choose to include one or the other. However, there are a few studies that incorporate both. The US NES is one example, as in 1972 closed-ended questions were added to the survey, and have since appeared regularlyin addition to the already-existing open-ended (likes and dislikes) questions. A number of

studies have been conducted over time to determine the efficacy of the specific closed-ended traits chosen for inclusion in election studies. Kinder (1983) made a detailed analysis of trait questions in the American NES, looking for dimensionality.

Scholars assessing leader evaluations tend to base their analyses on either open-ended or closed-ended traits exclusively.This is partly a function of the election studies themselves—usually they incorporate either one or the other, leaving researchers with little choice in the matter. Some have found unique ways to look at both, however. Brettschneider and Gabriel, in the German context, evaluated responses to both open-ended questions about the characteristics that an ideal leader should possess, in conjunction with responses to closed-ended questions evaluating the extent to which the leadership candidates possessed similar traits (Brettschneider and Gabriel 2002). In theory, the presence of both question types allows for a cross-referencing of sorts, similar to that done by Brettschneider and

25 Gabriel. Through the analysis of both open-ended and closed-ended traits questions it might

be possible to ascertain whether the traits that principal investigators choose to include in

their surveys (in a closed-ended format) are similar to those that are top of mind for

respondents.

The decision to focus on closed-ended questions in this study leaves us with a rich

Set of data. Table 2.4 lists all the traits that have been asked, indicating the election studies

and years in which that trait was included. Traits are organiaed alphabetically and similar

traits have also been grouped together with a label that reflects the common trait. For

example, “competent” and “competent leader” are listed one after another, and are labeled

collectively as “competent.” Chapter Four looks more closely at traits and will identify

clusters or groups of traits (for example, there is evidence that “knowledgeable” and

“inteffigent” are perceived similarly by respondents), but for now, traits have been left largely

untouched.

Table 2.4 indicates that there are some substantial commonalities in the types of traits

that have been incorporated into election studies over time and across countries. Most frequently included is “strength of leadership,” found 27 times in some form or another,

followed by “responsiveness to ordinary values,” included in a number of different formats,

22 times in total. Other popular trait types are “honest” (included 18 times) “knowledgeable”

(17 times), “inteffigent” (16 times), “compassionate” (15 times), “moral” (14 times),

“trustworthy” (13 times), and “arrogant” (11 times). The overlap that exists across election studies provides substantial scope for comparative analysis.Indeed, other scholars have already made use of the overlap in closed-ended questions across countries and over time in order to conduct some comparative analysis @ean and Mughan 1989) as well as longitudinal

26 analyses (Bartels 2002a; Brettschneider and Gabriel 2002;Johnston 2002) setting a precedent

for potentially broader comparison.

Conclusions

This analysis of national election study questionnaires suggests that there is substantial variation in the both the types of questions that have been asked as well as the formats of those questions, both across countries and over time. This variation has led to analyses based on different types of questions, thus providing us with an inconclusive understanding of the types of factors which may be important when voters are considering their voting options.

The analysisindicates that there is substantial overlap across time and space. The considerable commonalities provide an opportunity for both a broader and deeper analysis of the role of leader evaluations in elections. In particular, the frequent inclusion of closed- ended questions probing for perceptions of leaders’personality traits has the potential to facilitate a comparative and longitudinal study that is able to get at the heart of the role of leaders. As such, and also partially because of the limitations of other types of questions

(such as thermometers and open-ended question formats), I argue that in order to move forward, we ought to focus specificallyon questions that probe respondents’ perceptions of leaders’ traits. By focusing on traits it is possible to get a better understanding of voters’ evaluations of leaders: how leaders are perceived, the kinds of factors that influence voters’ perceptions, and the impact of those perceptions on vote choice.

27 CHAPTER THREE: CHALLENGES AND METHODOLOGY

Introduction

This project consists of three main parts, each reflecting a different facet of leader

evaluations. The first involves a deeper look at the traits themselves and the links between

them, in order to determine the extent to which traits fit together in dimensions. The second part is geared toward understanding the origins of trait evaluations—what are the factors

that influence the way that voters perceive party leaders? And finally,the third part assesses the role that evaluations of leaders play in influencing vote choice. The project is also comparative—both across time and space—in order to fill the gaps in the literature and bring us closer to a generalized understanding of the role of leaders in elections. The study therefore involves the pooling of election studies datasets and analyzes them all simultaneously.

This chapter outlines the methodological challenges inherent in a cross-national and longitudinal study of this magnitude, as well as the strategies employed to deal with them.

The chapter begins with an overview of the methodology employed in the study as a whole.

Then, I discuss four main challenges: first, the difficulty in matching questions from a diverse set of election studies in order to code them in a common format for pooled analysis.Second, and related, is the challenge of developing a measure of political sophistication across countries and years, where the same measures were not incorporated in all election studies. Third is an issue specific to the nature of this study: how do we compare the evaluations of different leaders from different countries in different years, from different parties? Fourth and finallyis the issue of sampling weights—when concatenating datasets

28 from many elections and many countries, and then running a pooled analysis,we risk

flooding the sample with the opinions of respondents from a single country or single

election.

Analyzing Leader Evaluations on a Cross-National and Longitudinal Basis

To better understand the role of leaders in elections, a comparative analysisis

necessary. I argue that in large part, the reason for the lack of agreement in the literature is

the nature of the studies that have been conducted to date—they have largelybeen based on

single elections, and studies often examine distinct survey questions. I suggest that by

looking at a common Set of variables across countries and over time, we will get more

conclusive evidence about trait dimensionality, the origins of evaluations, and the impact of

perceptions of leaders.

There are 35 election studies from a total of 7 countries with useable closed-ended traits 6questions. Concatenating these studies results in a dataset with over 186,000 respondents, and over 400 variables all coded in a similar format. In addition to evaluations

of traits, the dataset incorporates variables such as party and leader thermometers,

demographic variables, attitudes towards issues, vote choice, partisanship, media exposure,

and political sophistication.

This chapter provides an overview, and the techniques employed for each of these parts ‘willbe discussed in greater detail in future chapters. Here I want to discuss only those

6 The complete list of studies includes: the Canadian Election Study (including 1968, 1984, 1988, 1993, 1997, 2000, 2004, and 2006); the National Election Study (from the US, including 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, and 2004), the National Annenberg Election Study (US, 2000); the British Election Study (including 1983, 1987, 1992, 1997, and 2001); the Australian Election Study (including 1987, 1993, 1996, 1998, 2001, and 2004); the New Zealand Election Study (including 1999 and 2002); the German Election Study (including 1980 and 1987); and the Swedish Election Study (including 1988 and 1991).

29 issues that are common to the project as a whole, especiallythe challenges embedded in the

nature of the study.

Challenges Faced in the Study

While there were challenges that surfaced throughout the project—trying to find the

model that provided the best fit, interpreting the data analyses, and so-on—the major

challenges that were common to the whole project were largely encountered during the early

stages of the research, during the creation of the dataset.

CodingVariablesin a CommonFormat

The first main challenge is inherent to any comparative analysisinvolving several

isolated election studies. Different questions are asked in different countries and different

years, and in order to be able to analyze issues of interest across these studies, we need to

make judgments about the extent to which variables are broadly similar. The coding of two

main variable types provided the greatest challenges: leaders’ traits and issue attitudes.

Appendix I provides a complete list of the original question wording of all variables

included in the analyses, and as the list makes clear, there is indeed variety in question

wording. Even within a single country we see variety. Consider traits in Canada in 1997 versus 2000. In 1997, respondents were asked traits questions in the following format: “now we’d like to get your impressions of the party leaders. I would like you to tell me how well

the following words fit each leader. What about Jean Charest. Does ‘arrogant’describe Jean

Charest very well, fairlywell, not very well, or not at all?” In 2000, Canadians were asked a

30 similar question in a very different format “Which party leader would you describe as

arrogant?” As is fairlyobvious, these two questions are not exactlythe same, and trying to

compare trait evaluations over time could be a challenge, even if the trait itself is common to

the studies, as “arrogant” was in these two years. Preliminary investigation conducted

elsewhere (Bittner 2007a) suggests that while question format does have an impact on voters’

evaluations of party leaders, the impact is not so great as to preclude a comparative analysis.

All trait evaluations were recoded on a common 0-1 scale with evaluations separated by leader (with a value of I reflecting the most positive evaluation of a given leader on the

trait). Similar traits, regardless of original question wording, were given the same label, in order to permit large-scale analysis.To build the dataset, separate questions from separate studies that were considered to have the same root (as seen in Table 2.4) were coded under the same trait. For example, “strong leadership” and “leadership” were both recoded as

“leadership” even though the question wording was slightlydifferent in the two studies. In this fashion, it was possible to merge all of the separate studies together, in order to have many respondents evaluating leaders on the “same” traits. This led to a total of 55 separate traits to be analyzed, a process which will be discussed in greater detail in Chapter Four.

The second main coding challenge emerged with respect to respondents’ issue attitudes. Commonly included in most voting models, issue attitudes were an important component of this study, looking at the factors that influence voters’ evaluations of leaders.

If we were to determine the extent to which issue positions influenced voters’ perceptions of leaders traits on a cross-national and longitudinal basis, then it was necessary to ensure that common issues could be examined across election studies. In order to achieve this (since there was no one “issue” that appeared in all election studies), I opted to gather opinion on

31 issue dimensions, in the manner established by Kenneth Benoit and Michael Layer, in their

elite surveys on party policies and platforms (2006). Their work locates parties on two main

policy dimensions: taxesversusspending,and socialliberalism.Happily, many of the surveys

included in this study probe respondents for their attitudes about taxes versus social

spending. On the dimension of social liberalism, where possible, this dimension was based

on attitudes towards abortion. Where this was not possible, I made use of other variables, including attitudes towards immigration, attitudes towards familyvalues, and attitudes

towards treatment of violent criminals.All variables in both the taxes versus spending

dimension and the social liberalism dimension were coded on the same 0-1 scale, with 1 reflecting more left-leaning/progressive attitudes, and 0 reflecting more right- leaning/conservative attitudes.

The final dimension that I included in my analysis to tap into issue attitudes in the

election studies was a variable created to reflect opinion on the main issue in each election.

This variable was different from election study to election study, but allvariables were coded in the same direction as the other two issues (0-1,with 1 reflecting more left- leaning/progressive attitudes). Table 3.1 lists the major issue of each election, and the exact question used from each election study can be found in Appendix I. Compiling a list of all the major election issues was a challenge in itself and the list was generated from a number of different sources. Starting in 1992, the EuropeanJournalofPoliticalResearchbegan to publish a yearbook with information about elections that have taken place over the year. This was the main resource for all elections where it was available (Bale and Van Biezen 2007; Katz

2003; Katz and Koole 1997; Katz and Koole 1999; Katz and Koole 2002; Koole and Katz

32 1998; Koole and Katz 2000; Koole and Katz 2001; Koole and Mair 1993; Koole and Mair

1994; Van Biezen and Katz 2005).

For the remaining elections (pre-1992), a general search for election-specific research proved useful, and information about the important issues in these elections was obtained from journal articles and books discussing specific elections (Arter 1989; Blais et al. 2002;

Clarke et al. 1984; Crewe and Harrop 1986; Crewe and Harrop 1989; Gant and Lyons 1993;

Irving and Paterson 1981; Irving and Paterson 1987;Johnston et al. 1992; Kinder, Adams, and Gronke 1989; Lockerbie 1992; Mendelsohn and Nadeau 1999; Wiffiamsand Wilson

1977; Wilson 1985; Worlund 1992). Finally, in two Canadian elections (1997 and 2000), a clear issue did not emerge in the campaign. For these two elections, I used attitudes toward a long-standing issue in the Canadian context, attitudes towards Quebec, as a substitute.

Readers may question some of these choices, but I argue that by incorporating attitudes on the three issue dimensions (as imprecise as the measures may be), we are able to tap into voters’ issue attitudes on a comparative level, and even maintain some contextual/time- specific information through the inclusion of the major issue variable.

PoliticalSophistication

Many of the questions that remain unanswered regarding the evaluation of party leaders relate to mechanisms—how do voters consider party leaders when making decisions at the ballot box? Is it an information shortcut, helping those who lack information about policies and platforms to come to a decision, or is it something else? By incorporating political sophistication into the analyses, we can get a little closer to an answer.

33 The challenge is compounded with the difficulty of concatenating separate election

studies, often with very different measures of political sophistication. Scholars suggest that

the best measure of political sophistication is one that taps into factual knowledge about

politics (DeffiCarpini and Keeter 1996;Price and Zaller 1993). Unfortunately, not all of the

studies included in this analysisincorporate questions of factual political knowledge into the

survey. Some heavy tradeoffs were involved. Table 3.2 lists the type of measure used for

each election study in the creation of the overall sophistication index.

Where possible, an index based on responses to factual questions was used, with one

exception: the US election studies. Factual questions were not asked in all of the election

studies, therefore to keep the measure consistent within the country, the interviewer rating

(common to all studies in the NES) was utilized for the US election studies, including the

2000 Annenberg Election Study. After the knowledge index, it has been shown that

measures of general political interest are the best indicators of political sophistication

(Matthews 2006), and indeed, in the election studies which included more than one type of

measure, the general interest measure was most highly correlated with factual knowledge 7indexes. Other scholars have pointed to the utility and validity of the interviewer rating

(Bartels 1996; Zaller 1985), and this was the preferred measure after political interest. Finally, when the interviewer rating was not available, an index was created based on media

consumption in both television and print. Appendix I lists all of the variables included in all

of the measures for each country, expanding on the information in Table 3.2. All variables

Results not shown.

34 were recoded and combined to fit a 0-I scale, and were then merged when all of the election studies were pooled together.

In order to test the political sophistication measure, two main strategies were employed. First, I regressed the index on a series of dummy variables, one for each of the different sophistication measures that were combined to create the overall index. The reference category (constant) was the knowledge index. Table 3.3 depicts the results of the analysis,and illustrates that both the general interest measure and the interviewer rating did not differ substantially from the knowledge index. The media exposure variable was the most different from the other variables, and the coefficient suggests that where the media variable was used as the sophistication measure, the level of sophistication of the respondent was 0.21 points higher on the 0-1 scale than the sophistication level of the reference group.

This measure was only used for two studies, suggesting that the level of bias in the measure overall was minimal.

The second main strategy was another regression analysis,in which the sophistication measure was regressed on a series of socio-economic and demographic variables, to determine the extent to which levels of sophistication and common demographic variables were related as we might expect. Table 3.4 depicts the results of this analysis. University graduates tended to have a higher level of political sophistication (coefficient of 0.187), as did those with higher incomes (coefficient of 0.064), those who were older (coefficient of

0.02), and those who were employed (coefficient of 0.017). Women tended to have lower levels of political sophistication (coefficient of -0.074) as did respondents of visible minority groups (with a coefficient of -0.041). All of these fit with our common perceptions of who in society is more politically sophisticated and informed (Bartels 1996; Deli Carpini and Keeter

35 1996; Popkin 1991; Sniderman, Brody, and Tetlock 1991). These results suggest that the

political sophistication measure performs well given the methodological challenges involved

with creating a common measure across this many election studies.

Cross-Parfrj’LeaderEvaluations

Comparing evaluations of party leaders across countries comes with considerable

methodological challenges, in terms of grouping parties into “types.” Because, for example, the

German FDP does not exist in Canada, and because the British Labour Party does not exist in

the US, using each country’s party label for cross-national analysis is not feasible. However, there

are commonalities among parties from different countries, and scholars have expended

considerable effort to develop cross-national comparisons between parties based on party

policies and platforms. Examples include expert surveys (Benoit and Layer 2006) and the

Comparative Manifesto Project (Budge et al. 2001). See McDonald, Mendes, and Kim (2007)

for a review of these efforts. My analysis makes use of Benoit and Layer’s extensive (2006) work, which maps parties along two dimensional lines: stances on social liberalism and on taxes versus

spending. By relying upon Benoit and Layer’s placement of parties, I was able to group similar parties together, creating a common label for the group. This permitted cross-national, over-time comparison of leaders’ traits.

I categorized parties according to their placement along two issue dimensions, grouping parties with similar locations into the same group or category. Figure 3.1 illustrates the placement of parties, and the groups into which they were gathered, as it amalgamates the individual country-based figures that Benoit and Layer include in their book for the countries in this study. Party categories included “Conservative” (including the Canadian Conservatives, the

36 American Republicans, the British Conservatives, the German CDU and CSU, the Australian

Liberals and the Swedish Moderate Party); “Centre-Left” (including the Canadian Liberals,

British Labour, New Zealand Labour, Australian Labor, the German SPD, and the US

Democratic party); “Left” (which includes the Canadian NDP, the Swedish Left Party, the New

Zealand Alliance, the British LibDems, and the Australian Democrats); “Centre-Right” (includes

the Swedish Centre Party, the New Zealand National Party, New Zealand ACT, the Swedish

Peoples Party, and the German FDP); and “Right” (includes New Zealand First, Australian

ONE, Australian Nationals, Swedish Christian, Swedish NDP, and the Canadian Reform Party).

There were two exceptions to the grouping of parties according to their locations on the two dimensions: Green parties (includes the Swedish, Canadian, New Zealand, Australian, and

German Green Parties) and Sectional parties (which includes the Canadian Bloc Quebecois, the

Scottish National Party and the Welsh Plaid Cymru). Some may argue that these two party types

could have been included in either Left or Centre-Left party categories. I felt that due to their very specific focuses, either on environmental issues (Green) or nationalist platforms (Sectional), it made sense to separate these parties into their own categories.

Some may quibble with the location of any of the parties along the two dimensional lines: for example, it may be argued that the Canadian Green Party might be located further to the right on the taxes versus spending dimension, certainly in comparison to other Canadian, and even non-Canadian, parties. While this may be true, and it is up for debate, Benoit and

Layer’s conclusions were relied upon as a whole because of their unified and systematic formulation of party locations, as well as the fact that their placements are based upon elite surveys of more than simply a few individuals. I am content to rely on their results, because they provide a consistent and defensible way to conceive and organize party groups.

37 After having grouped parties into categories, the data were recoded to ensure that trait evaluations were comparable along these lines. Thus the trait, “leadership” was labeled with the party category label, for example, “Conservative.” “Conservative leadership” therefore included evaluations of the strength of leadership trait for leaders of the US Republican Party, the

German CDU/CSU, the British Conservative Party, the Australian Liberal Party, the Canadian

Conservative Party, and the Swedish Moderate Party. In total, this brought the total number of respondents evaluating this trait for Conservative leaders to 73,000.

By aggregatingparties into these groups, it was then possible to analyze evaluations of the leaders of these different parties en masse, because of the new common nomenclature.

Evaluating leaders from this many parties and this many countries would otherwise not be possible.

SamplingWeights

The total number of respondents in each of the election studies varied, from less than 2000 (in the NES, for example) to approximately 50,000 (in the case of the Annenberg

Election Study). The risk when attempting to conduct an analysis with pooled data is that the attitudes of respondents in a particular country or a particular election would flood the sample and skew the results. In a sense, what we want is to mirror a random sample across countries and elections. Without any adjustments, however, some people have higher or lower probabilities of selection (Americans in 2000 have a much higher probably of selection) 8 In order to prevent this from happening, and to ensure that the results of the

8 For a more detailed discussion of sampling, see (Kalton 1983; Kish 1965; Sudman 1983), who provide both theoreticai and practical information about various aspects of sampling.

38 analyses more accurately reflect the true relationships between variables, a system of

sampling weights was created.

In order to determine sampling weights, I obtained each country’s population in the

year 1988. The year 1988 was chosen because it represents an approximate mid-point in the

1968-2006 time spread in the 35 election studies in this analysis,and as well, it is one of the

two years in which the largest number of elections took place (three election studies were

conducted this year, and three took place the year before). Population data were obtained

from gapminder.org, a non-profit organization that works closelywith both the United

Nations and Google, to promote achievement of UN Millennium Development Goals. The total population of the seven countries was calculated, as was each country’s proportion of the total. At the same time, the total number of observations in the dataset was obtained as was the proportion of each country’s presence in the 35 country dataset.

Table 3.5 lists all of the total numbers, as well as the proportions of each country within the total. Weights were generated by dividing each country’s proportion of the population in the real world by its proportion of the number of observations within the dataset. In order to ensure that any one dataset within a particular country did not flood the sample, each election study was weighted to ensure that it had an equal position within the country’s total sample. These two sets of weights were multiplied by one another, and the product was the weight assigned to each election study. In effect, then, each individual’s presence in the dataset more closely reflected a random sample than it did before I generated these sampling 9weights. 9This weighting scheme does weight the US, non-parliamentary context heavier than the parliamentary context because of the size of the American population compared to the other countries. The risk, then, is that the US case is swamping the others, thereby skewing the analyses. Other weighting formulas were considered—for example, an alternative was to give each country equal weight, but then the US 2000 NABS would overwhelm the American

39 By creating these sampling weights, and incorporating them into the analyses conducted, we are able to ensure that no single election study (e.g. Annenberg 2000, with over 50,000 respondents) swamps the dataset and the analyses, and that each respondent’s weight within the larger dataset reflects the country’s population in relation to the total population of all of the countries in the study.

Conclusions

The primary goal of this study is to update and unify a diverse and inconclusive literature on the evaluation of party leaders. In order to do this, it is imperative that a large, cross-national and longitudinal analysis be conducted, because the bulk of research that has taken place to date has involved the analysis of single elections or a few elections in a single country. While existing studies have provided us with important insight into the role of leader evaluations, many questions remain.

This study combines data from 35 election studies across 7 countries, and as such involves a massive amount of data management, recoding, as well as troubleshooting in order to make the project workable. Many of the challenges that have arisen are the kinds of challenges inherent in any comparative project involving survey data, where different

population. If we were to give each election study equal weight, then those countries with a larger number of election studies (the US and Canada) would dominate the sample. If we gave each country equal weight and then each study equal weight within the country, then there would still be some election studies weighted more heavily than others—namely, those countries with fewer election studies in the sample (Germany, Sweden, New Zealand). Regardless of the weighting scheme chosen, tradeoffs are involved. To ensure that the results of the analyses were not dominated by the more weighty American and non-parliamentary context, they were all re-run without the American Election Studies included. The results were encouraging. The patterns described in coming chapters continued to be present with the US case removed from the sample. In many instances, the patterns observed were actually stronger without the opinions of American respondents. These results suggest, first, that the weighting scheme deployed does not result in an overstatement of the impact of party leaders (or in fact, making the evaluations of candidates skew our understanding of the evaluations of leaden).Second, the results also give further credence to the fact that evaluations of party leaders play a role in Parliamentary systems, where the role of leaders has not been thought to be as pronounced.

40 questions are asked in different countries and different years. Some of the challenges were specific to the issue of the comparative analysis of leader evaluations. This chapter described the challenges that were applicable to the study as a whole—other analysis-specifichurdles will be addressed in the discussion related to those analyses.

This chapter has also provided an overview of the project as a whole. The study will proceed in three main parts. The first comprises an in-depth look at personality traits themselves, in particular the linkages between traits, in order to better understand the underlying dimensionality and structure of trait evaluations. In the second part, I look more closely at the origins of trait evaluations—the types of factors that affect the way that voters perceive party leaders and influence how they judge leaders’ personality traits. The third part takes the project to the final step, and looks at the effect of evaluations of leaders’ personalities on vote choice. The common coding of many datasets, the development of a political sophistication measure, the creation of a party classification system, and the inclusion of sampling weights have all ensured that the comparative analysiscould actually take place.

41 CHAPTER FOUR: THE TWO-DIMENSIONAL STRUCTURE OF TRAITS

Introduction

Over the years, election studies have collected respondents’ reactions to over 150

different specific traits of party leaders. For allintents and purposes, it is difficult to make

any headway into understanding voters’ evaluations of party leaders if we base our analyses

on all of these traits separately. Other studies have tended to collapse the multitudes of

individual traits into a series of dimensions, partially to facilitate further analysis,but also

because there is a substantial amount of socio-psychological evidence that people reallydo

think about traits in terms of dimensions (Bean 1993;Bean and Mughan 1989;Johnston

2002; Kinder 1983; Kinder, Abelson, and Fiske 1979; Kinder and Fiske 1986; Kinder et al.

1980;Miller, Wattenberg, and Malanchuk 1986;Shabad and Andersen 1979).

Taken as a whole, studies of leaders’ traits are divided with respect to the number and

types of dimensions into which traits fall. Some suggest two (Johnston 2002; Stewart and

Clarke 1992), others suggest as many as twelve (Brown et al. 1988),with most studies falling

below the half-way point. The diversity in the number of dimensions that have been put

forth to date is probably largely the result of scholars’ reliance upon different datasets,

different questions, and different question formats in the formation of their understanding

of trait dimensionality over the years.

This chapter builds upon past research by evaluating the underlying structure of traits in two ways. First, it assesses the state of the literature on person perception and the evaluation of traits, determining the extent to which past studies have converged on a

42 particular trait typology. Second, the chapter presents new research and results on trait

dimensionality, based on the comparative analysis of 35 election studies. The combination of

existing theory and broad-based data analysispoints to a two-dimensional trait typology. The

two dimensions, characterand competence,will provide the basis for further research into both

the origins and effects of leader evaluations, when we move forward in chapters to follow.

The Structure of Traits as Seen in the Literature

Across previous studies, there is variation in the type of data that is used, with some

scholars relying upon responses to open-ended questions about likes and dislikes of party

leaders, while other scholars base their analyses on closed-ended trait questions. In addition

to the variation in the data being analyzed, there is also variation in the methods that are

used to generate the typologies, including mainly a) factor analysis;or b) the creation of

typologies based on the researcher’s own intuitions and the face validity of the links made

between groups of traits. Not surprisingly, there is variation in the conclusions that have

been drawn about the nature of the underlying structure of personality traits.

,QuestionFormat

From the early I970s-onward, a number of different trait typologies have been

advanced in the literature. Winham and Cunningham (1970),based on an analysis of open-

ended questions in a survey conducted in the 1968 Canadian federal election campaign,

suggested that there were eight dimensions of party leaders’ personality traits. They divided

all open-ended responses into eight categories, labeled “youthful,” “personal maturity, age,”

43 “initiative, drive,” “steady, serious,” “sincere, honest,” “capable, competent,” “speaking

ability,” and “independent, free-thinker” (1970, 42), and then subdivided them yet again into

positive and negative traits in each category.

A couple of decades later, Clive Bean (1993)grouped open-ended responses into

seven main categories: competence, integrity, strength, harmony, general likeability,other

personal, and policy/party/group. In the American context, David Glass (1985) examined

likes and dislikes of leaders, and decided that the most common responses about a leader’s

personal characteristics fall into three main categories: competence, character, and personal

attraction. Numerous others have also conducted analyses based on open-ended assessments

of leaders, and as a whole, have advanced a wide spectrum of common categories of traits.

Among those employing closed-ended traits questions, a wide variety of dimensions

have also been put forth. Bean and Mughan (1989) assessed all of the nine traits that are

common to both the British Election Study and the Australian Election Study, and found

that the same four characteristics are most important in both countries: effectiveness, listening to reason, caring, and sticking to principles. Stewart and Clarke (1992) examined closed-ended traits in the British context, and found that competence and responsiveness are the two main dimensions that voters use when evaluating party leaders. In the Canadian context, based on the evolution of traits included in the Canadian Election Study, it appears that the Election Study team has decided that there are two main factors that are important in the evaluation of party leaders: competence and honesty (these are the only two traits that appear in the 2006 CES, after years of having a multitude of traits included in the study).

Kinder (1983) conducted an analysisinto the dimensionality of closed-ended leaders’ traits, and argued that there are four main categories of traits that we ought to consider and

44 incorporate into our analyses more commonly: competence, leadership, integrity, and

empathy. Furthermore, he recommended that the American NES incorporate a total of

twelve traits regularlyinto its election studies, including a few from each of the four

categories. Indeed, his recommendations were followed, thus beginning a long legacy of the

inclusion of specific traits, fitting these four dimensions, in the NES. Funk (1999) later

assesses the links between the traits in Kinder’s dimensions, as well as examining their

respective impacts on overall evaluations (thermometers), and argues that it makes more

sense to think about three dimensions rather than four. She argues in favour of collapsing

the leadership and competence dimensions into one combined category, and considering perceptions of those traits alongside of Kinder’s two other dimensions, integrity and

empathy. In fact, it does appear that with a few exceptions, the literature is converging on

between two and four dimensions, either by collapsing leadership, competence, integrity, and

empathy, or using them individually.

Method

To a large extent, this convergence is occurring regardless of the method employed by scholars. Generally speaking, there are two main methods that have been used in the past to develop an understanding of the underlying trait structure. The first is the use of factor analysisto determine the extent to which traits load onto specific factors or dimensions. The second, used with nearly identical frequency is the generation of trait typologies based on scholars’ own intuition and the face validity of the trait groupings. Both methods come with challenges of their own, and neither has provided the authoritative answer about trait structure: both methods have generated numerous and varied perspectives on trait

45 dimensions, but if we look a little closer we can see that they are largely oscifiatingaround the dimensions as Kinder identified them.

The important thing to note with all of these studies, regardless of which method they use, is that they all come up with slightly different conclusions. Among those who employ the use of factor analysis,Miller and Miller (1976) look at open—endedsurvey questions, and find that the data support five trait dimensions: competence, trust, reliability, leadership appeal, and appearance. Nimmo and Savage (1976) find four main trait dimensions: leadership, partisanship, dramatic performance, and personal qualities.

Additionally, they find that there is variation in trait dimensions depending on the individual leader being evaluated—some dimensions exist for some leaders, some exist for others. Also, through factor analysis,Kinder has found support for both two dimensions, namely competence and integrity (Kinder, Abelson, and Fiske 1979), as well as four dimensions: competence, integrity, leadership, and empathy (Kinder 1983).

A few years later, Miller, Wattenberg, and Malanchuk (1986) found support for five dimensions by employing factor analysis on open-ended traits, sorting them into the categories of competence, integrity, reliability,charisma, and personal. Stewart and Clarke

(Stewart and Clarke 1992) find two dimensions, competence and responsiveness, in their study, and as mentioned, Funk (1999)uses factor analysis and finds three main dimensions.

In the same year, Pancer, Brown, and Barr also find three dimensions based on factor analysis:charisma, competence, and integrity (Pancer, Brown, and Barr 1999).

That many scholars have opted to generate their own typologies rather than employ factor analysis has not led to drastically different conclusions—in many cases, the dimensions put forth are very similar to those identified through factor analysis.Perhaps the

46 main difference in the conclusions of the two methods is that in some cases, manj more dimensions are put forth. Winham and Cunningham (1970)put forth eight dimensions.

Shabad and Andersen (1979) find six dimensions (personality, background, competence, leadership, reliability,and trust). Brown et al. (1988)based their analysis on twelve dimensions (competence, dynamism, integrity, empathy, responsibility, personal style, political skills,episodic judgments, social background attributes, party preferences, political positions, and ‘unclassified).

Twelve appears to be the largest number of trait dimensions that have been suggested, but contemplating the dimensionality or structure of traits without factor analysis has not necessarily led to a large number of dimensions. Kinder et al. (1980) suggest two

(personality and performance), Johnston (2002) also suggests two (character and competence), Bean (1993) and McCurley and Mondak (1995) also point to two (competence and integrity), and a few others suggest three: Funk (1996)with competence, integrity, and warmth qualities; Glass (1985)with competence, character, and personal attraction; and

Lavine and Gschwend (2006)with experience and ability,leadership qualities, and personal qualities.

While one might think that the use of factor analysisprovides more robust and reliable conclusions about trait dimensionality, the plethora of different findings provides one indication that factor analysisis not a panacea. Johnston (2002) observes that traits tend to load onto one overarching “partisan” factor, and he opts instead to generate his own two dimensional typology, based on the face validity of the links between traits and a detailed examination of the data.

47 Fundamentally, regardless of the method in which typologies are generated, patterns have emerged in the literature. A wide reading suggests that about 28 different typologies have been developed. Among these, 20 have included “competence” as a dimension, twelve have included “integrity,” ten have included “leadership” or “strength of leadership,” ten have included “empathy” or “warmth,” five have included “charisma,” and three have included “character.” A closer look at all of the traits within these dimensions suggests that

“integrity” and “character” are reallythe same thing with different names, and if we combine them, then integrity/character feature in fifteen different typologies. With a few exceptions, the literature appears to converge somewhere between two and four dimensions, generally incorporating some combination of “competence,” “leadership,” “integrity/character,” and

“empathy.” A lot of the variation appears to stem from nomenclature—some scholars group leadership and competence together to create “competence” and will group integrity and empathy together and call the category either integrity or character. Kinder (1986) confirms the reliabilityof four separate dimensions, but finds that competence and leadership are highly correlated, and that integrity and empathy are also highly correlated, suggesting that two dimensions or four dimensions may both be about right. Indeed, at varying times, he has identified a two-dimensional structure (1974; 1980) and a four-dimensional structure

(1983; 1986). Furthermore, his two-dimension structures generally reflect a “collapsed” version of his four dimensions.

48 Empirically Derived Structure of Traits

Is there an underlying trait structure or dimensionality that applies outside the

bounds of a single country? Thus far, very few scholars have attempted to look at trait

structure on a comparative basis (see, however, Bean and Mughan (1989) and Ohr and

Oscarsson (2003)). Certainly, nobody has attempted to examine the dimensionality of traits

on as large a scale as the present study. While a study of this magnitude provides a rich set of

data from which to determine the structure of traits, the scale of the project also presents a

number of difficulties. This section begins by outlining the challenges present in this

undertaking, and will follow with a discussion of the patterns that emerged from the data.

TheHurdles:Multi-Coun1y,Multi-YearAna/ysis

The size of the dataset and the extent to which some traits have been asked

frequently in the studies being looked at allow us to make links between traits and specific

dimensions based on a larger number of respondents than would be possible with only one

election study. As noted in Chapter Two, closed-ended questions probing respondents for their perceptions of party leaders’ strength of leadership can be found 25 times in the 45

election studies containing closed-ended questions about leaders. When we look at evaluations of this trait within the studies included in this project, the total number of respondents evaluating leaders on this trait is reallyhigh. Table 4.1 lists the distribution of ratings of leaders from the three main party types, Conservative, Centre-Left, and Left, on the 0-1 rating scale. The table shows that there are slight differences in the distribution of ratings of leaders by party type, but it is also important to note the massive number of

49 observations included—as many as 81, 401 respondents evaluated Centre-Left party leaders’

strength of leadership!

While the sheer size of the dataset is one of its greatest assets, it also leads to a

number of methodological challenges. The foremost among them is ensuring comparability

from country to country and year to year. Even where questions are similar, coding all

election studies in a common format for analysispresented a massive challenge.

Related to the issue of different questions in different election studies, the largest

challenge was conducting a single statistical analysis to determine the structure of trait

evaluations for all of the 35 election studies at the same time. Not all respondents evaluated

leaders on all traits. Indeed, not a single respondent evaluated all of the traits. For example,

81,000 respondents evaluated “leadership,” and this was the most frequently asked trait. One

of the main methods used by scholars was going to be impossible: conducting a factor

analysiswould not only be a challenge, but was actuallynot possible give the extent to which

there are missing data in this dataset. Effectively, any factor analysisis left with no

observations, because there are missing data on every trait, for every leader.’°This ruled out

the option of running a “giant” factor analysis on these data. It is possible, however, that

even had there not been so much missing data, the results of the factor analysismay not have been conclusive. Similarlyto Johnston’s (2002)observation that traits tended to load

first and foremost on a “partisanship” factor, the exploratory factor analysismay have yielded unusable results.

The solution, then, was to evaluate the relationship between traits without factor analysis,and to consider trait dimensions based on aggregate patterns found in the

‘OAppend II provides a detailed explanation of the missing data problem, the various steps that were taken to attempt to resolve the problem, as well as potential avenues for future research in this area.

50 evaluation of leaders. I ran two sets of pairwise correlation analyses: first, between trait evaluations for the leaders of the three most common party types, Conservative, Centre-Left, and Left. Second, I ran a stacked 11analysis, running pairwise correlations for traits common to all three sets of party leaders, and traits common to just Conservative and Centre-Left leaders (respondents as a whole evaluated 55 traits for both Conservative and Centre-Left leaders, and 34 traits for all three leaders). By stacking the analysis the party label was removed, and the correlation analysis looked simply at the links between traits, regardless of which of the three parties was being evaluated.

TheDimensionaliyofTraits:Convergencein ElectionStudies

Table 4.2 lists the ten traits that were evaluated by the largest number of respondents.

The number of Conservative, Centre-Left, and Left Party leaders’ trait evaluations are listed separately, illustrating the difference in numbers for each. These three party categories received the largest number of evaluations overall, as well as the largest variety of specific trait-types. Conservative, Centre-Left, and Left Parties appear to be some of the most common party types across countries, and researchers have tended to isolate these parties and ask respondents to evaluate their party leaders more often than the leaders of other types of parties.

11 Stacking the data involved changing the nature of the data matrix. The original format reflected a setup in which each observation was one respondent, and evaluations for each individual leader on each individual trait were separate variables. By stacking the data, individual trait evaluations for individual leaders were no longer separate variables. So for example, a respondent’s individual ratings of Centre-Left, Left, and Conservative leaders on the trait “honesty” were stacked one on top of another under the variable name “honesty,” and thus each respondent was associated with three separate observations instead of just one. This facilitated the analysis of the variable “honesty” as a single entity, rather than three separate variables “conservative honesty,” “centre-left honesty,” and “left honesty.”

51 The traits that are included in election studies most frequently include “leadership,”

“cares,” “knowledgeable,” “intelligent,” “inspiring,” “honest,” “compassionate,”

“trustworthy,” “arrogant,” and “moral.” For Conservative and Centre-Left Party leaders, we have evaluations from between 20,000 respondents and 80,000 respondents. The numbers are not as large among those evaluating Left party leaders, and range from just under 2,000 respondents to 21,000. The “top” traits also feature fairlyprominently in Kinder’s (1983) typology of four trait dimensions: leadership, competence, integrity/character, and empathy.

The legacy of Kinder’s research into the instrumentation of candidate evaluation can be seen in the time path of trait indicators. Once the NES incorporated traits from Kinder’s four dimensions, the frequency with which these traits have been included in election studies— worldwide—increased substantially. In fact, the traits that Kinder recommends for inclusion in the NES are more frequently asked across all election studies than other traits, and as a group, they achieve their peak in the late 1980s and early 1990s. The data suggest that scholars of election studies have broadly operated within the framework introduced by

Kinder, even if in an “amateurish” way—that is, by developing and including traits questions based on the NES (as scholars readily admit to having done—see Johnston et al. (1992) as an example). As a result, a general understanding of which traits “matter,” or at least which traits ought to be included in election studies, has been reached, even if not all scholars have conducted extensive research into trait dimensionality.

It makes sense then, that we have seen some convergence in the literature on trait dimensions—especially among those who are basing their analyses on closed-ended questions. If Kinder identified between two and four dimensions, and then recommended the inclusion of specific traits which fit into those dimensions, we should expect that

52 analyses based on election studies which incorporate those same traits ought to make similar findings. Indeed, when we look at the results of the pairwise correlation analysesin this study, Kinder’s legacy becomes even more evident.

Reiationshz],sBetweenTraits:AggregateVariationsin theData

These pairwise correlations provide further support for the work Kinder has done over the years, and suggest that his assessment of the underlying structure of traits is fairly accurate. The linkages he observed between traits within two separate dimensions (as collapsed from four) stand up to the scrutiny of these data. The analyses suggest that the largest amount of internal cohesion exists when we conceive of traits as fallinginto two dimensions, characterand 12competence. Tables 4.3 and 4.4 depict the results of stacked pairwise correlation analyses, conducted on evaluations of leaders of the three Party types which received the most trait evaluations 13overall. The list of traits assessed for Conservative and Centre-Left Party leaders are identical, and include approximately the same number of respondents evaluating each pair of correlations. The Left Party list has fewer traits overall (34 instead of 55),

12The nomenclature is similar to Johnston’s (2002). Other scholars have put forth two-dimensional typologies, labeled competence and responsiveness (Stewart and Clarke 1992); competence and integrity (Kinder et al. 1979); and personality and performance (Kinder et al. 1980). The contents of these dimensions are largely the same—that is, the competence/performance category includes traits such as intelligence and strength of leadership, as derived from the separate “competence” and “leadership” categories in Kinder’s four-point typology (1983). The character/integrity/responsiveness/personality category, regardless of its exact nomenclature, includes traits such as honesty and compassion, as derived from the “integrity” and “empathy” categories in Kinder’s four-point typology. I have chosen the labels “competence” and “character” because the literature has converged on competence and because character seems like a wider umbrella label than either integrity, personality, or responsiveness. Character is a more inclusive label, extending across both integrity-related traits and empathy-related traits. 3Not shown are additional unstacked analyses, looking at the pairwise correlations between trait evaluations for the leaders‘ of each of the three main types of parties, separately. The results between the stacked and unstacked analyses were nearly identical, thus the stacked analyses alone are presented, since they are more generalized, and the results are not particular to any one party.

53 although all of the traits for which Left Party leaders were evaluated were also evaluated for

the leaders of the other two Party types. There are two stacked analyses presented: the first

for correlations between trait evaluations of Conservative and Centre-Left leaders, and the

second for correlations between trait evaluations of the leaders of all three major parties.

Thus the second table, Table 4.4, contains fewer traits overall, but assesses the links between

evaluations of more leaders.

The level of missing data in the trait evaluations can be seen in the two tables, since

in the pairwise correlation all respondents who did not evaluate both halves of the pair of

traits were eliminated from the analysis,and in many cases, those who evaluated a leader on

one trait did not evaluate that same leader on another. This occurrence is represented by a

“.“ in the tables, and all other cells of the table contain the correlation values obtained in the

analysis.Nearly allvalues reach traditional levels of statistical significance,with the few

exceptions noted at 0.00 in order to differentiate those from the missing data in the table.

A number of patterns emerge from the data. First, the strongest correlations exist

within traits that might be thought of as fitting in the “character” dimension: honest and

trustworthy are highly correlated—hovering in both the two-party and three-party setups at

around 0.8. Other traits that we might think of as fitting into this dimension are also highly

correlated with these two traits: “keeps promises,” “sensible,” and “reliable” are some

examples. That traits within this category are the most highly correlated is indicative of a high level of internal cohesion between these traits: perhaps respondents have a high level of

comprehension when interpreting questions regarding these traits. Also important, however, is the high level of correlation between the traits in this category and traits that can be

54 thought of as fitting in Kinder’s (1983) “empathy” category—traits such as cares,

compassionate, and kind.

In the two-party setup, the average inter-item correlation between traits in the

category I label character (in the bottom right shaded section of the correlation matrix) is

0.41, while the average inter-item correlation between these traits and those relating to

competence (bottom left, unshaded section of the correlation matrix) is 0.32. These numbers

are 0.35 and 0.3, respectively, in the three-party setup, and indicate a higher level of cohesion

within dimensions than across dimensions. The correlation between the traits in these two

categories (character and empathy) is generally stronger than the correlation between these

traits and traits related to leadership or competence. In fact, the same can be said for the links between traits in the competence dimension. The average inter-item correlation within

this category is 0.43 and 0.47 for the two and three-party setups, respectively. There are

strong links between traits spanning across this category: among others, strong leadership is highly correlated with commands respect (0.62), intelligent (0.51), and knowledgeable (0.56).

These traits are also fairly highly correlated with one another. While no pair of traits achieves as high a correlation as a couple of traits in the character dimension (0.8), the average inter- item correlation is higher than in the character dimension.

In some instances, the traits in the competence category are more strongly correlated with traits in the character category than they are with each other. These correlations suggest that individuals do connect traits across dimensions as well as within dimensions. The links between traits in different dimensions suggest that there may be something more to consider than simply how we might fit traits into dimensions, and how dimensions might relate to one another. It raises the issue of what I term “anchor” versus “supplementary” traits. Are there

55 some traits that play a central role in both a given dimension, as well as an overall evaluation? Is the fact that “intelligent” is more highly correlated with “honest” (correlation 0.55) than with

“commands respect” (correlation 0.48) illustrative of a deeper positioning of a given trait in the mind of the voter? Are “inteffigent” and “honest” anchor traits, while “commands respect” simplyplays the role of supplementary trait in the mind of the voter? One might think researchers believe this to be so, given the frequency with which “intelligent” and “honest” have been included in election studies over other, less frequently asked traits (like“commands respect”). The extent to which this can reallybe investigated with the current dataset is limited, given the level of missing data on some of these traits. More research is needed.

Generally, however, it appears as though traits within a particular category (whether character or competence) tend to be the most highly correlated. The strong correlations within each of the dimensions suggest that that two dimensions: character,which encompasses traits from the character and empathy categories; and competence,which encompasses traits from the competence and leadership categories, do make sense as cohesive units. The extent to which missing data prevents us from looking deeper into the relationships between all of the traits further points to the appropriateness of considering traits within only two dimensions rather than four. While there are linkages across these two dimensions (traits within character are correlated with traits in the competence category)— and logically, this ought to be so—the patterns in the data do suggest that respondents evaluate the two dimensions separately.

Figure 4.1 gives further evidence to the importance of distinguishing between the two dimensions. This figure graphs mean character ratings against mean competence ratings for all of the leaders of the three main parties. All ratings range across the 0-1 scale. The figure

56 clearlyillustrates that in the aggregate, across both leaders as well as time and space, evaluations on one dimension do not simply track evaluations on the other. If this was so, all dots would be located on a straight line. Further, the graph illustrates that voters distinguish between leaders and between parties. The placement of leaders on these two dimensions is quite scattered, albeit with a densely populated centre, indicating that respondents make distinctions between leaders and trait dimensions.

The Underlying Structure of Traits: Character and Competence

Since the analyses in the following chapters will rely on these two characterand competencedimensions, it is important that we have confidence in them. I tested the links between traits in these dimensions, and the results indicate that the two dimensions are quite cohesive, and can be thought of as distinct facets of perceptions of party leaders. I also again note the importance of Kinder’s influence in shaping our understanding of traits, and I argue that future Election Study teams oughtto develop trait batteries with his recommendations in mind, because they form the most robust indices of trait evaluations.

In the same way that a “giant” factor analysis could not be run on all of the traits simultaneously due to the nature of the dataset and the extent of missing data, a “giant”

Cronbach’s Alpha could also not be generated to test the reliabilityof the two dimensions.

Still,it is important to feel that the indexes generated for each of the two dimensions were fairlyrobust. Once traits were combined in the dataset into two indexes, one for character and one for competence, a separate Cronbach’s alpha was computed for each election study, looking at the strength of the links between traits from each study that were included in the particular dimension. Table 4.5 lists the results for each election study.

57 With a couple of exceptions, the aiphas are quite high—reaching as high as 0.94—

suggesting that overall, the traits fit together rather well. When we consider these aiphas in

conjunction with the pairwise correlation coefficients presented, it becomes even more clear

that the two dimensions are fairlyreliable, and that they are measuring broadly the same

thing, whether character or competence.

Second, the aiphas are highest where the traits included in the election study are most

consistent with the traits that Kinder (1983)proposed for inclusion. The 1984 NES included

16 traits, and the aiphas for each dimension range from 0.86 to 0.94. All NES surveys after

1984 are based to some extent on this same battery, and have alpha scores above 0.72, with

most scores in the mid 0.8 range. The 1988 Canadian Election Study, whose principal investigators based their trait battery on the NES (Johnston et al. 1992), also has high alphas

(ranging from 0.76 to 0.84), while the remaining Canadian studies stray further from the

Kinder battery, and have lower alpha scores. Beginning in 1993, Australia’s trait battery is

more similar to Kinder’s, and alpha scores are higher for that country’s election studies

between 1993 and 2004. Thus Kinder’s legacy proves important not only for influencing the

trajectory of traits included in election studies, but also because of the cohesiveness of the

corresponding trait dimensions. Future studies would do well to include the Kinder trait battery, rather than branching out on their own, not only to facilitate cross-national

comparison, but also because these traits appear to most closely correspond to what voters actuallyevaluate leaders on. Using these two trait dimensions, it becomes possible to move

forward to understand the origins and impact of evaluations on a comparative basis.

58 CHAPTER FIVE: FACTORS AFFECTING EVALUATIONS OF PARTY LEADERS

Introduction

Why do voters evaluate party leaders? And where do perceptions come from? A

number of scholars have pointed to the overwhelming influence of the media in shaping our political perceptions more generally, and have suggested that one of the reasons that party leaders weigh as heavilyin the vote calculus as they do is the result of media focus and priming, placing the horse race first and foremost in the minds of voters (Crewe and King

1994; Gidengil et al. 2000; Mendelsohn 1994; Mendelsohn 1996;Mughan 2000). Others have

suggested that we evaluate leaders because it is an “easy” thing to do, since it mimics a process we do every day in our regular lives, as we encounter new individuals and decide how we feel about them (Cottrell, Neuberg, and Li 2006; Rahn et al. 1990). Others still suggest that we can use how we feel about leaders as a “shortcut” to help with vote choice when we lack other types of information—policy platforms and performance history, for example (Sniderman, Brody, and Tetlock 1991), and that personality traits may provide cues about how a given individual will perform once in office (Glass 1985; Popkin 1991;

Sniderman, Brody, and Tetlock 1991).

While some or even all of these “reasons” for leader evaluation are entirely plausible, what they do not tell us is where perceptions come from in the first place: what types of factors influence the way we evaluate party leaders? Existing studies, generally based upon the analysis of a single election year in a single country, point to divergent sets of factors that influence voters’ evaluations of leaders. Some suggest voter demographics matter (Popkin

59 1991; Cutler 2002), others point to the importance of partisanship and ideological leanings of

both voters and leaders (Bartels 2002; Graetz & McAllister 1987),while others suggest that

policy/issue related factors are key (Weisberg & Rusk 1970;Rusk & Weisberg 1972). Some

studies note the importance of all of the above in different circumstances (Conover &

Feldman 1989).

I suggest that there are indeed some generalizations that can be made about the

sources of perceptions across these seven countries. First, individuals’ personal partisanship

and ideological views play the largest role in influencing evaluations of party leaders.

Attitudes towards issues play a secondary role, while socio-demograpbics have a smaller impact on perceptions of leaders. Second, there does appear to be some indication that a

characteristic intrinsic to the leader (the leader’s sex) may play a more substantial role than anything else in determining how voters rate the leader. Finally, there are significant

differences in the way that political sophistication mediates perceptions of party leaders: the more sophisticated consider more factors overall and weigh those factors more heavilyin their ratings.

Literature

A substantial proportion of the literature on leader evaluations suggests that voters’ perceptions of leaders play a discernible role in influencing vote choice. Scholars suggest that leader evaluations are quite proximate to the vote, and are thus often “entered” as one of the last variables in voting models (Campbell et al. 1960, Miller and Shanks 1996).This practice suggests that other, prior, variables have an impact on leader evaluations as well as having an independent impact on the vote itself. These “prior” variables are usually socio

60 demographics (such as age, gender, and employment status), partisanship, ideology, and attitudes towards issues.

Studies of voters’ evaluations of leaders suggest that these types of “prior” variables may act as cues and heuristics, shaping perceptions and evaluations of candidates, both in terms of personality as well as their positions on issues (Conover and Feldman 1989, Rahn

1993).These scholars suggest that people will use the information available to them, making inferences based on whatever information they have, even if they have next to no information at all. Indeed, socio-psychological experiments suggest that without any other sources of information, subjects will make inferences about quality of leadership and ideology based on looks alone (Riggleet al. 1992).Most suggest, however, that with increasing amounts of information, voters tend to use other, “higher,” factors when making judgments.

Conover and Feldman (1989) find that at different times, issue positions, partisanship, and the ideology of the individual affect his or her impressions of candidate positions on issues. They suggest that campaigns and the media prime some cues over others at different times, and that as a result, different combinations of factors may come together to shape perceptions. They find that early in the campaign, when there may be less information about candidates available to voters, there will be more false consensus and projection effects, based on the individual’s own positions. As more information surfaces over the course of the campaign, this projection happens less, and individuals tend to base their perceptions more on the partisanship and ideology of the leader.

The notion that partisanship plays a key role is supported and emphasized by experimental research (Rahn 1993),which suggests that not only is the individual voter’s

61 partisanship a factor, but that the party label of the candidate has a major impact on voters’

perceptions. Rahn argues that “partisan stereotypes” act as heuristics, affecting the nature of

information processing, and that voters tend to use the heuristic even when more

information is available. She finds that when party and ideological cues are not available,

people will listen to candidates’ policy statements and make inferences based on what they

hear. However, when both candidate statements andparty labels are available,subjects prefer

to rely on heuristics, making judgments based on impressions of what it means to be a

Republican or a Democrat, regardless of what policy statements individuals actuallymake.

While Rahn doesn’t talk about issue ownership as such, her findings give support to the notion that voters have expectations about candidates’ issue positions based on the party banner.

The role of the partisan stereotype is different from the role of the voter’s own partisanship in affecting evaluations of party leaders, but points to the role of the party in shaping perceptions. Aldrich et. al (1999) regressed trait dimensions on demographics and partisanship for each party leader in a number of US elections, and found that partisanship and race were the strongest predictors of trait evaluations. Their observations fit neatly with previous conclusions made by Conover (1981),who suggests that voters are more likelyto use a cue to differentiate between candidates if they feel strongly about it, and that the characteristics voters use to define their own political selves are more likelyto resonate strongly when evaluating candidates than are “non-self attributes.” Race and partisanship are two fairly strong “identifications” in the American context, thus it is not surprising that these two factors influenced perceptions of party leaders so strongly.

62 The impact of socio-demographics (such as race) have also been found to be integral in perceptions of party leaders. Cutler (2002)paints perhaps the clearest picture of the link between voter characteristics and leader evaluations, with his finding that “closeness” matters. The more similar a leader is to a voter in terms of socio-demographic characteristics, the more positively disposed that voter is towards the leader. Cutler incorporates sex, religion, region of residence and language into his analysis,and his findings point to the importance of basic demographic characteristics. The broader literature on the role of sex, however, is mixed.

While there is a substantial literature on the gender gap, suggesting that women and men make different choices and have different opinions (Bittner 2007b, Conover 1988,

Gidengil et al. 2003, Inglehart and Norris 2000, Sapiro and Mahajan 1986, Sapiro and

Conover 1997), the effects of gender (of both the voter andthe candidate) on evaluations of leaders are not as clear. Zipp and Plutzer (1985) suggest that the gender of the candidate does not have a substantial impact on perceptions. While they focus more on vote choice than leader evaluations per se, their study finds that the votes of men and women are not a major reflection of the sex of the candidate. Later, however, they find that the sex of the voter is in fact related to voting for women candidates (1996). In fact, they suggest that gender identity competes with partisanship as a cue for voting behaviour, as the effect of women voting for women is amplified for Democratic female candidates who are perceived to be the most feminist. Banducci and Karp (2000) suggest, however, that the effects of gender identity are mixed, finding that the link between women voting for women is not concrete. With a few notable exceptions (Banducci & Karp 2000; O’Neill 1998), these

63 previous studies tend to look at congressional candidates rather than party leaders, and our

knowledge of the impact of leaders’ sex is therefore limited.

In general, it seems as though our understanding of the influence of all of these

“background” factors on perceptions of leaders is not concrete. Nor is our understanding of

why they have an impact on evaluations. Much of the literature on the impact of

demographics looks at the effect of the demographic characteristics of the candidate or

leader, suggesting that they provide a cue for voters, especiallythe least sophisticated voters

(Dolan 1998; Popkin 1991; Sigelman et at. 1995;Tate 1993; Zipp & Plutzer 1985). Scholars

often suggest that these demographics not only provide cues, but may also provide an

indication about the likelihood that a leader will represent the interests of the voter.

The links with evaluations of leaders are less clear. McDermott (1997)suggests that

voters see these kinds of demographic cues and then evaluate the leader in terms of an

associated stereotype—be it gender, race, or something else. The effect is similar to the

impact of the leader’s ideological stance, issue positions, or partisanship. That is, voters glean information from the leaders’ own positions and then incorporate that into their perceptions

of the leader.

The influence of the ideological stances, issue positions, and partisanship of the voter is one step ftwther removed. Voters’ own attitudes and positions have an impact on their perceptions of leaders as well. We would expect, for example, that individuals will perceive the leaders of their own party more favourably and that more conservative voters will perceive more liberal leaders less favourably. Similarly,voters in favour of easier access to abortion wifilikelyview more conservative or right-wing leaders less favourably than leaders on the left side of the political spectrum. Based on their own political values and attitudes,

64 therefore, it seems reasonable to expect that voters will evaluate leaders more or less

positively. This could conceivably extend to the impact of socio-demographics as well.

Research has shown that social cleavagesand socio-demographic characteristics

affect attitudes and vote choice (Campbell et al. 1960;Berelson et al. 1954).The effects of

religion (Blais2005; Irvine & Gold 1980;Johnston 1985); ethnic origin (Blais2005; Blais et

aL2002); education (Johnston et at. 1996);and sex (Conover 1988, Gidengil et al. 2003,

Inglehart and Norris 2000, Sapiro and Mahajan 1986, Sapiro and Conover 1997) have been demonstrated. We might therefore expect these attitudinal differences to extend to perceptions of party leaders as well—women, traditionally more left-leaning, might be expected to perceive Left leaders more positively, for example.

At different times, scholars have found differing combinations of factors (socio demographic, partisanship, partisan cues, ideology, and issue attitudes) to affect perceptions and evaluations most strongly. It is not at all clear why these different factors have the influence they do, or what the process of evaluation reallyis. Does the impact of partisanship or ideology differ when individuals are more or less politically sophisticated?

Does gender matter more or less according to level of political sophistication? Do attitudes towards policy issues affect evaluations to the same extent as either socio-demographic variables or partisanship? It is to these questions that I now turn.

The Dependent Variable: Evaluations of Leaders’ Personality Traits

The bulk of the discussion of the impact of leaders has focused on the effect of overall “feelings” towards leaders on election outcomes: essentially,the net effect of “feeling thermometers” on vote choice. And while some scholars have looked at the role of traits

65 themselves (Gidengil et al. 2006,Johnston et al. 1992,Johnston 2002), a more systematic

assessment is necessary in order to reallyunderstand how people evaluate leaders.

Furthermore, a look at the evaluations of actual traits rather than feeling thermometers may

help to clarify the role of leaders in elections, since (as noted earlier) not only is the

thermometer a very general measure, it also may not be the most reliable source of

information about voters’ evaluations of leaders. By focusing solely on traits, we may get a

more precise picture of what voters think about when they evaluate leaders.

There is a rich literature on “person perception” in cognitive and political

psychology. Scholars in these fields have spent years assessing perception of those traits in

others, as part of a larger study of the human psyche and our perceptions of our

environment. MacRae and Bodenhausen (2000) suggest that the perception of personality

traits in others takes place as a part of general “categorical thinking” on the part of

individuals. That is, in order to make the world ordered, meaningful, and predictable, we

think categoricallyabout others. Much of this is subconscious, and the result of the way in which our mind stores and processes information. Cognitive psychology research suggests

that information is stored in our minds in what are known as “schemata” (Lau and Sears,

1986). Schemata can be likened to a series of hierarchical storage cabinets in our minds, each

cabinet essentially reflecting a different category or topic, with links in between categories.

Each schema in the mind affects the way in which we gather new information, as well as how we call up old information (Lodge and Hamill 1986). Scholars have suggested that schemata play an important role in how we perceive and interact with the world: they provide categories for labeling people, events, and places, they influence what information gets both stored in and recalled from memory, and they allow us to integrate what we already

66 know into our interpretations of new circumstances where we lack a complete picture—

essentially,in new situations, they allow us to “fill in the blanks” with information we already

have (Conover and Feldman 1989, Lodge and Hamill 1986).

It has been suggested that the process of evaluating political candidates and party

leaders fits within the schemata framework. Kinder et al. (1980) suggest that voters agree on

the traits that an ideal president should possess. They argue that voters then use this

“presidential prototype” or schema as a shortcut to decision-making. The idea is that voters apply existing categories (the prototype) to the leadership candidates and evaluate the

candidates based on the traits that an ideal candidate should possess. It is as if the voter opens up his/her filing cabinet, takes out the file labeled “presidential prototype” and checks to see whether the candidate’s traits match the traits inside the file.A comparison is made between the individual candidates running in the election and the ideal candidate.

Brown et al. (1988) assess the extent to which the concept of schemata applies to the way Canadian voters evaluated the traits possessed by party leaders during the 1984 Federal election. They find that schemata or prototypes of leaders get used repeatedly, as voters consider the same types of factors when evaluating all of the leaders. This finding suggests that looking more closely at traits is a useful exercise, as it would allow us to get a deeper insight into the way in which voters think “categorically” about leaders. Rather than just looking at overall feelings towards a leader, by looking at traits we might be able to gain a deeper understanding of how people evaluate leaders.

In the previous chapter, I suggested that it makes the most sense to think about traits as falling within two main “umbrella” dimensions: competenceand character.A typology based on these two dimensions refines the existing literature: the competencedimension broadly

67 includes traits fallingin both the “leadership” and “competence” dimensions, and the

characterdimension includes traits which had previously been thought to belong in both the

“integrity” and “empathy” dimensions. The new labels themselves do not signify substantively different understandings of the way in which voters perceive party leaders: they represent a collapsing of the four previous dimensions into two, based on patterns and correlations in the evaluations of party 14leaders. Because the dimensions themselves do not change, even if the specific traits within them might differ slightly from year to year, looking at traits in this way allows us to consider evaluations of leaders’ character and competence, regardless of the changes that have taken place in the question format.

Impact of SES, PID, Ideology, and Issues on Trait Evaluations

In order to assess the factors influencing the evaluations of leaders, evaluations of leaders’ character and competence were regressed on a slate of independent variables traditionally found to have an effect on the way voters perceive party leaders. These independent variables include socio-demographics, partisanship, ideological self-placements, and issue attitudes. Tables 5.1 through 5.4 present these regression analyses.Tables 5.1 and

5.2 look at the respective roles of socio-demographics and partisanship, while Tables 5.3 and 5.4 also incorporate voters’ ideological self-placements and issue 15attitudes. The regression

Organizing and conceptualizing traits into overarching dimensions is particularly valuable in that it facilitates longitudinal4 assessment of the evaluation of leaders’ traits where it otherwise would not be possible. One of the main‘ difficulties with assessing voters’ evaluations of leader traits over time is the extent to which surveys change over time. These changes are characteristic of nearly all election studies. Question formats change, the types of traits that respondents are asked to evaluate change, and (obviously) leaders change, which makes isolating and examining patterns in evaluations a real challenge. 15 models were run separately because of a phenomenon in the German Election Studies where individuals who were asked to evaluate leaders’ personality traits were not also asked to provide their ideological perspectives and issue placements. While German respondents didanswer these questions, it appears as though the sample was split on which respondents were asked which questions, thus the models that incorporate issues attitudes and ideological

68 analyseswere run for the leaders of the three most common parties, the Centre-Left leaders,

Conservative leaders, and Left party leaders, and not for the remaining party leaders, because

of the substantial drop in sample size. As Table 5.1 indicates, the sample size for evaluations

of Left party leaders is substantially lower than that for evaluations of the other two main

party leaders.

The model includes five socio-demograpbic variables: marital status, sex, education,

employment status, and age, because these are the only five demographic variables that were

universal to all of the 35 election studies. Unfortunately other variables which one might

expect to be important (income, union membership, and race/ethnicity, for example) were

not present in each of the election studies, and thus cannot be incorporated into the large

pooled analysis.As illustrated in Table 5.1, the impact of the demographic variables on

evaluations of leaders’ character ranges from 0.003 to 0.054 on a scale from 0-1 (a fraction of

a percentage point to just over five percentage points when converted to a 0-100 scale), and

the impact of the variables varies by the leaders’ political party. Most of the differences are

quite small. Interestingly, regardless of which leader the individual is evaluating, marital

status appears to have one of the stronger effects among demographic variables, with married individuals giving Centre-Left party leaders lower ratings, and Conservative and Left leaders slightly higher ratings on the character dimension.

Employment status also has a slightlylarger coefficient than other demographic variables, as employed individuals give lower ratings to Centre-Left leaders than the unemployed, but give higher ratings to Conservative and Left party leaders. Age has the views were run without the German study, and weights were recalculated and applied to reflect the exclusion of the country from the analysis. While the sample size may be smaller, the relationship between variables is not substantively different. Re-running the analyses in Tables 5.1 and 5.2 without those respondents who were dropped from the analyses in Tables 5.3 and 5.4 does not substantially change the coefficients in Tables 5.1 and 5.2. The relationships hold, even with a reduced and selective sample.

69 largest impact on evaluations of Conservative Party leaders, as with each step in the age

category (each category reflects 10 years), individuals give higher ratings to the leader by

0.015 points. Education has the largest impact on evaluations of Left party leaders’ character,

with individuals having obtained a university degree giving higher ratings to the Left leader

by 0.054 points. University graduates also give Centre-Left leaders slightlyhigher ratings (by

0.014 points) than non-graduates.

Table 5.1 also indicates that the impact of partisanship substantially dwarfs that of

the included socio-demographic variables. Centre-Left partisans give their party leader nearly

0.14 points higher than the reference group (non-partisans). Conservative Partisans give more positive ratings to the Conservative leader, just over 0.17 points higher than the reference group, while Centre-Left partisans give the most negative evaluations to the

Conservative leader, with ratings just over 0.1 points lower than the reference group. Left partisans give highest ratings to their leader on the character dimension, with ratings nearly

0.17 points higher than non-partisans.

The effect of variables on evaluations of leaders’ competence, as seen in Table 5.2, is similar to that seen in Table 5.1. That is, marital status has slightly larger coefficients, and individuals who are married tend to rate Centre-Left leaders more negatively and

Conservative leaders more positively than do non-married individuals. Employment status affects only evaluations of the Centre-Left leaders’ competence, with employed individuals giving lower ratings than others do. Education again has a large impact on evaluations of

Left leaders, with university graduates giving Left leaders higher ratings than non-graduates, and giving Conservative leaders lower ratings than non-graduates. Age appears to have less

70 of an impact than it does for evaluations of character, as the coefficient for age is less than

half the size it is for evaluations of Conservative leaders’ character.

Interestingly, sex has a different effect for evaluations of competence than character

as well: in terms of evaluations of character, women tend to give slightlyhigher (less than

0.01 points) ratings to Centre-Left and Conservative leaders, and slightlylower (again,less

than 0.01 points) ratings to Left leaders. This appears to conflict with the gender gap that

exists in voting and attitudes in general, where women appear more to the left (Conover

1988, Gidengil et al. 2003, Inglehart and Norris 2000, Sapiro and Conover 1997) and thus

would be expected to evaluate Left party leaders more positively. However, women evaluate

Left leaders more positively on the competence dimension than men do, and to a greater

extent than their evaluations of the other two types of party leaders, perhaps making up for

the lower evaluations of Left leaders’ character. ‘3Vhydo women rate Left party leaders less

positively on character than competence? Is it that their expectations are higher on this trait

dimension? More research is needed to better understand this dynamic.

Partisanship continues to have a substantially larger impact on evaluations than do socio-demographic 16variables. Centre-Left partisans tend to rate Centre-Left leaders’ competence more highly than non-partisans. Conservative partisans evaluate Conservative party leaders more positively on the competence dimension, while Centre-Left partisans are the most negative about Conservative leaders. Left partisans evaluate Left leaders the most positively on the competence dimension.

16 It is quite possible that these models understate the importance of socio-demographic variables, due to the limitations of the dataset. Some of the socio-demographic variables which we might expect to influence evaluations (income, union membership, race and ethnicity) were not universally available and therefore not included. In this sense the model is underspecifled and while the coefficients for socio-demographics are much smaller than the other variables, this may not reflect the “real” impact of socio-demographic variables overall, compared to the other variables in the model. Consideration of the tables and values presented should bear this in mind.

71 Tables 5.3 and 5.4 introduce individuals’ideological self-placements and attitudes

towards policy issues into the model. Three types of issue attitudes were chosen for inclusion

into the analysis, two of which were based on the nature of party policy dimensions as

identified and utili2ed by Benoit and Layer (2006),and which served as the basis for situating

parties and placing them into groups (as discussed in Chapter 2).

The two main issue attitudes which served as dimensions for Benoit and Layer were

party positions on taxes versus spending and social liberalism. Thus variables getting at

respondents’ views on these issues were coded in as similar a fashion as possible across

election studies. All issue attitudes and the ideology variable were coded on a 0-1 scale, where I reflects a more “liberal” or “progressive” attitude, and 0 reflects a more

“conservative” or “reactionary” attitude. Thus, for example, individuals believing that access

to abortion should be the most readily available (of the choice set presented to them in the

election study where they were interviewed) were given a I on the social liberalism variable, while those believing most strongly in a claw back in access to abortion, or in limiting the

circumstances under which abortions could be obtained, were coded a 0 on the scale. We would expect these coefficients to be positive for evaluations of Left leaders (as more left- leaning respondents should be more likelyto evaluate left-leaning leaders more positively),

and negative for evaluations of Conservative leaders. In fact, this is the case in Table 5.3.

When we introduce the ideology and issue attitude variables, the size of the sample included in the analysis drops substantially. As mentioned earlier, the German sample was

split, and those individuals evaluating character and competence were not also asked for their ideological views or issue attitudes, thus Germany was dropped from the analysis.In addition, not all studies within a given country included an ideology variable (for example,

72 Canada 1993), thus some respondents were dropped from the analysis,even if the country as

a whole was not dropped. Finally, the other main source of attrition occurred where some issue attitude variables were included in a post election study rather than the campaign study,

thus the number of observations dropped where those individuals whose socio-demographic

data were collected in a pre-election study were subsequently not included in the post

election study, and therefore were not asked for their opinions. The number of observations, while substantially smaller than it was in the analysespresented in Tables 5.1 and 5.2, is still quite large (as high as 75,000 respondents!) and weights are still incorporated so as to ensure that no one study floods the analysis.

Broadly, Tables 5.3 and 5.4 illustrate that socio-demographics have the smallest impact on evaluations of character and competence. Ideological self-placements and partisanship have the largest impact on evaluations, followed by attitudes towards policy issues. Furthermore, ideology rivals partisanship for largest impact on evaluations. Now, this may be an artifact of the way the variables are coded: a person either is or isn’t a Centre-Left partisan. But the ideology variable, though also coded in terms of 0-1, is a scale, where the difference between 0 and I is more extreme. Thus the impact of ideology, as measured in the analysis, ought to have more of an impact than that of partisanship, because individuals who consider themselves to be very “right-wing” are quite different from those individuals who consider themselves to be very “left-wing.”

While the exact sizes of coefficients change somewhat with the inclusion of issue attitudes and ideology into the model, the effects of socio-demograpbics or partisanship do not change much. Centre-Left partisans still give highest ratings to their own leader.

Conservative partisans rate their own leaders most highly,while Centre-Left partisans are

73 most negative about Conservative leaders. Left partisans rate their own leaders most

positively. Attitudes towards issues have a larger impact on evaluations than demographics,

but a smaller impact than either partisanship or ideology.

Attitudes on the issue of taxes versus spending had the greatest impact on

evaluations of the leaders’ character, for all three major parties. After taxes versus spending,

the major campaign issue appears to have the largest impact on evaluations of character, for

the Centre-Left and Conservative leaders, while attitudes on the social liberalism dimension

affect evaluations of Left leaders’ character more than the major campaign issue (although

the coefficient on social liberalism is larger for evaluations of the other two major parties,

suggesting that in general, issue attitudes appear to influence evaluations of character for

Centre-Left and Conservative leaders more than Left leaders).

When it comes to evaluations of leaders’ competence, the dynamic is slightly

different. While issues still affect evaluations more than demographics, attitudes towards

taxes versus spending are not the most influential issue for evaluations of the competence of

all three leaders. The taxes versus spending dimension has the largest impact on evaluations

of Centre-Left and Left leaders, but not Conservative leaders. Perceptions of Conservative leaders’ competence appear to be more influenced by the major campaign issue and by attitudes towards social liberalism (with equally-sized coefficients) than taxes versus

spending. For Centre-Left and Left party leaders, the major campaign issue has the second largest impact on evaluations of competence, and the coefficients on attitudes towards social liberalism trail further behind. For the most part, then, the data suggest that of this group of issues, the taxes versus spending dimension and attitudes towards the major campaign issue have the largest impact on perceptions of leaders’ traits generally.

74 The data suggest that while socio-demographics have the smallest effect on perceptions of leaders’ traits, age and sex appear to have the smallest effects among demographic characteristics, while marital status and education appear to affect evaluations a little more. The data as presented in Tables 5.1 through 5.4 do not actually account for characteristics possessed by the leaders however, or the extent to which voters may be affected by the “similarity” between themselves and the party leader (beyond the effects of partisanship and, more loosely, ideology). Cutler (2002)notes the effect of socio demographics, in particular the extent to which leaders are similar to voters, in influencing vote choice, and thus it seems worth exploring a bit in the context of evaluations of party leaders.

The Impact of the Leader Herse1f Women as an Example of the Impact of Socio Economic Sameness

The difficulty with the task of matching socio-demographic characteristics between leaders and voters in the comparative context is the sheer number of leaders evaluated in the

35 election studies (105 in total), and the difficulty in finding biographical information about all of them, to then match it to information about voters—information that is not consistently available across all election studies. I collected information about the leaders’ age, sex, religious background, ethnicity, and language (in the Canadian context), and merged this information into the larger dataset. The difficulty, of course, is that the only two sociodemographic variables of all of these that were common to respondent information in all election studies were age and sex, thus narrowing down the ability to “match” leaders’ and voters’ characteristics quite a bit, at least in the context of a large pooled analysis.

75 As such, I opted to focus solely on the socio-demographic characteristic that appears

to attract the most discussion and analysisin the comparative literature, sex. Scholars have

explored the effects of both leaders’ and candidates’ sex on vote choice, examining the

extent to which voters perceive women leaders differently than men. As mentioned earlier,

the theories are two-fold: first, that women might prefer women because they perceive that

women will be more likelyto look out for their interests (Popkin 1991;johnston et al. 1992)

because of shared experiences and understanding. The second perspective is linked to

stereotyping, in that women are perceived to hold particular ideological views and attitudes

(McDermott 1997), therefore certain voters will be positively inclined towards women

leaders because they will act in a certain way and advance the policies associated with the

stereotype. Scholars have examined these theories empirically,and the results are mixed.

The literature is not conclusive, as to whether or not the sex of the leader affects voters’ perceptions, nor is it conclusive as to whether women are more likelyto vote for women candidates (Banducci and Karp 2000, Gidengil, Everitt, and Banducci 2006, Plutzer

and Zipp 1996, Shabad and Andersen 1979, Zipp and Plutzer 1985).By looking at this issue on a larger scale, with a total of 10 woman party leaders (of 105) evaluated by respondents in the election studies, it may be possible to gain greater insight into the impact of the leaders’ sex on perceptions.

Table 5.5 depicts the results of the regression analyses, in which evaluations of leaders’ character and competence for each of the three main party types were regressed on five demographic variables, as well as partisanship and the variables of interest: the leader’s sex as well as the interaction between the leader’s sex and the sex of the voter. Only the coefficients of the latter two are presented in the table. Men give lower character ratings to

76 women leaders versus men leaders of all three major parties. For Centre-Left leaders, women

leaders receive ratings nearly 0.4 points lower (on a 0-1 scale) than men leaders.

This effect is slightlyreversed after we account for women voters. The interaction

coefficient (woman voter*woman leader) is 0.032, suggesting that while women leaders are

still evaluated more negatively than men, by -0.35 points, women tend to view women

Centre-Left leaders slightlymore positively than men do. Conservative women leaders are

also evaluated more negatively than Conservative men leaders, with men giving women

leaders ratings 0.48 points lower than men leaders. Unlike the effect seen for Centre-Left

leaders, however, women voters also perceive women Conservative leaders more negatively,

with the sum of the main effect and the interaction coefficient suggesting that women

Conservative leaders are rated nearly 0.5 points lower than men Conservative leaders are.

This suggests that women do not necessarily evaluate a leader more positively simply

because she is a woman, and the results may actuallyhave more to do with the findings of

traditional gender gap literature: perhaps women evaluate women Conservative leaders more

negatively because they are Conservative, rather than because they are women. Indeed, the

fact that women voters evaluate women Left leaders more positively gives further credence

to this notion. Women Left leaders are evaluated slightly more negatively than men leaders

by men voters (coefficient of -0.04), while women voters evaluate them slightlyless

negatively (the sum of the interaction and the main effect is -0.03 and statisticallysignificant).

On the competence dimension, men again tend to view women leaders more negatively than men leaders, although to a lesser extent than they did on the character dimension. Men give competence ratings just over 0.1 points lower to women Centre-Left leaders than men leaders, and the effect of women voters’ perspectives slightlyreverses this

77 effect: women voters evaluate women Centre-Left leaders slightlyless negatively than men

do, with a ratings nearly 0.09 points lower than they give to men leaders. For Left leaders,

men also view women candidates more negatively on the competence dimension, with

ratings over 0.02 points lower than men candidates receive. The effect of women’s

perceptions actually reverses this effect, however, as the coefficient for the interaction effect

is positive and larger than the main effect: women evaluate women Left leaders more

positively than men leaders.

The effect of the leaders’ sex on perceptions of Conservative leaders’ competence is

not statistically significant, and the sizes of the coefficients are quite small. This may have

something to do with Margaret Thatcher specifically:in the 1987 British election, 88% of

voters gave her the most positive evaluation on the competence dimension. This is the

highest competence rating for any of the leaders of any of the parties in any of the 35

election studies in this analysis.Thatcher’s mean competence rating in 1987 was 0.95 (out of

1),while her average character rating was approximately half of that: this is a substantial

difference, and sets her apart from other leaders in terms of perceptions of her competence.

She was no ordinary Conservative leader, regardless of her sex.

Women leaders are generally perceived more negatively than men on both personality dimensions, regardless of party. Interestingly, both men and women view women leaders less negatively on the competence dimension than they do on the character dimension, indicating that they evaluate the two traits very differently. Women voters tend to evaluate women leaders slightly more positively than men do, but with the exception of the evaluation of Left leaders on the competence dimension, women do not evaluate women leaders more positively than they evaluate men leaders, they simply rate them a littlelessnegatively.

78 The Mediating Role of Political Sophistication

Partisanship and ideology appear to affect evaluations of leaders’ traits most,

followed by attitudes towards issues, followed finally by voters’ own socio-demographic

characteristics. It appears, however, that if we consider the size of coefficients, the variable

that has the largest impact on voters’ evaluations of party leaders is the sex of the leader

herself: that is, men and women alike perceive women leaders more negatively than men

leaders. While this information helps us to understand the sources of evaluations, it does not

actually assist in determining the mechanism: why do people evaluate leaders the way they

do? Do personal demographic characteristics matter more when a voter is less informed and

less able to make judgments based on other factors? Do people with more political

sophistication evaluate leaders differently, based on more “sophisticated” factors? Do ideological and policy-based considerations matter more? And perhaps even more interestingly, are there differences in how voters perceive women candidates based on their levels of political sophistication? Are the less sophisticated more or less predisposed to rate women leaders negatively?

A rich literature has emerged on the role of political knowledge, political information, and political sophistication (Alvarez 1997, Bartels 1996, Bittner 2007, Converse 1964, Cutler

2002, Delli Carpini and Keeter 1996, Hayes 2004, Lau and Sears 1986, Luskin 1987, Luskin

1990, Roy 2007, Sniderman, Brody, and Tetlock 1991, Zaller 1992). It is completely reasonable to expect that more sophisticated individuals will perceive party leaders very differently from their less sophisticated counterparts. Perhaps the more sophisticated base their evaluations on a separate Set of considerations than the less sophisticated. Or perhaps

79 the more sophisticated wifi still consider the same number or same types of factors, but will

base their evaluations more heavily on some rather than others.

In this section I look at the types of factors that affect evaluations of leaders’

character and competence, among a partial sample of all of the respondents interviewed in

these studies: the least sophisticated and the most sophisticated. Within each election study,

respondents were divided into quartile groups according to their level of political

sophistication: roughly the 25% who were most sophisticated were grouped together, and

the 25% who were the least sophisticated were grouped together, to create an overall low

and high sophistication group. In total, when all election studies are combined, there are

37,905 respondents in the least sophisticated group and 37,026 in the most sophisticated group, of the 156,307 respondents for whom there is information about their political sophistication.

Character and competence evaluations were then regressed on all of the same variables as in the previous tables, only this time the analysistook place within sophistication groups rather than in the sample as a whole. Tables 5.6 and 5.7 give the results of these regression analyses, for the Centre-Left and Conservative parties 17only. The first thing that one might notice is the substantial difference in the number of observations in the low and high sophistication groups. While both groups started out at approximately the same size, the numbers dropped off substantially when the other variables were entered into the mix.

Indeed, even when looking simply at the breakdown of individuals who evaluated the traits of Centre-Left and Conservative party leaders, the drop is 18notable.

17Left party leaders’ evaluations are not examined according to low/high levelsof sophistication because of the substantially smaller sample size: we are left with less than 5000 respondents in those analyses. 18 5.8 lays Out the number of individuals evaluating the leaders’ traits of these two parties. As immediately evident, there are substantially fewer than 37,000 respondents in any of the cells, and furthermore, there are

80 So are there differences in the factors that influence perceptions according to level of

political sophistication? The quick answer is yes. As Tables 5.6 and 5.7 illustrate, first, the

number of coefficients that are statisticallysignificant is generally larger in the high

sophistication group than in the low sophistication group, for evaluations of both trait types

for leaders of both parties. It appears as though highsophisticatesconsideragreaternumberof factorsoverall.

In addition, however, the amount of emphasis placed on certain types of factors is greater among the highly sophisticated. While ideology, partisanship, and issue attitudes are more important than socio-demographics among both groups of respondents, the size of the coefficients are substantially larger among the highly sophisticated group in comparison to their less sophisticated counterparts. For example, with respect to evaluations of Centre-Left party leaders’ character (Table 5.6), the ideology coefficient is 0.07 for those in the less sophisticated group, and twice as large (0.154)in the more sophisticated group: double the size!In evaluations of Centre-Left leaders’ competence, a similar effect exists, but is even more pronounced: the ideology coefficient is 0.055 among the less sophisticated, and 0.152 for the more sophisticated.

approximately 4000 fewer respondents in the iow sophistication group than in the high sophistication group, regardilessof which party’s leader is being evaluated. There are also fewer individuals who evaluated Conservative leaders’ personality traits than Centre-Left leaders’ personality traits. And this all takes place before we add the other variables into the analysis. When evaluations are regressed on all of the other variables laid out in Tables 4.6 and 4.7, we are left with just over approximately half and one third the number of respondents that began in each of the two groups of political sophistication. The substantial drop-off in the low-sophistication group is somewhat predictable, for a couple of reasons. First, there is the effect of coding: aUrespondents providing an answer of “don’t know” on a trait evaluation question were coded as missing. Since people who are less politically sophisticated are more likely to provide “don’t know” responses to questions than those who are more politically sophisticated (Berinsky 2004), the drop in sample size is not surprising. Second, there is a significant link between participation in surveys (and especially re-interviews) for respondents who are less interested and knowledgeable (again, see Berinsky 2004). These individuals are less likely to agree to be interviewed in the first place, and the re-interview rate is substantially lower among the less sophisticated compared to those individuals with higher levels of political sophistication.

81 Partisanship has, for the most part, a greater impact on perceptions of character and

competence among the more sophisticated rather than the less sophisticated. With the

exception of Centre-Left partisans evaluating Centre-Left leaders’ competence, the more

sophisticated tend to perceive the leaders of their own party more positively than the less

sophisticated do. Furthermore, partisans of the “opposing” party tend to view the leader

even more negatively in the highly sophisticated group than in the less sophisticated group:

more sophisticated Conservative partisans view the Centre-Left leader more negatively than less sophisticated Conservative partisans, and more sophisticated Centre-Left partisans perceive the Conservative leader more negatively than less sophisticated Centre-Left partisans.

Finally, socio-demographic variables generally appear not to have much of an effect regardless of political sophistication: only employment status is significant with respect to evaluations of the Centre-Left leaders’ competence, and marital status and education achieve statistical significance only with respect to evaluations of Conservative leaders’ character.

The effect of socio-demographics is dwarfed by the effects of partisanship, ideology, and issue attitudes, with one exception: the sex of the party leader.

Much as when looking at the effects of the leaders’ sex on attitudes among the whole sample, this variable continues to provide the greatest amount of explanatory power for both low and high sophisticates’ evaluations of leaders’ traits. Table 5.9 mirrors the setup in Table

5.5, breaking out the coefficients for the leader’s sex, only this time it does so by level of political sophistication.

Overall, low sophisticates evaluate Centre-Left women leaders’ character more negatively than high sophisticates do. Furthermore, while the interaction variable (woman

82 voter*woman leader) has a larger coefficient for the less sophisticated, less sophisticated women voters still give Centre-Left women leaders more negative evaluations on character than more sophisticated women do (i.e.,when we sum the coefficients). With regards to evaluations of Centre-Left leaders’ competence, the effect is similar:both groups evaluate women leaders more negatively than men leaders, and the less sophisticated evaluate women leaders more negatively than the more sophisticated. Once the coefficients are combined, it becomes clear that the two groups of women evaluate women leaders similarly(with a coefficient of -0.211).

For evaluations of Conservative party leaders’ character, both less and more sophisticated men evaluate women leaders equallynegatively. Ratings of Conservative leaders’ competence are quite different with the more sophisticated giving less negative evaluations than the less sophisticated. Among women voters, high sophisticates tend to view women leaders less negatively than low sophisticates, for both trait dimensions. While the interaction variables (woman voter*woman leader) are not statisticallysignificant, the sums of the two coefficients are statistically significant, with p values of less than 0.001.

Broadly, then, the evidence suggests that political sophistication affects the way that both women and men perceive women leaders. While women voters do evaluate women leaders more negatively than they evaluate men leaders of the two major parties (all coefficients are negative), more sophisticated women voters tend to evaluate women leaders slightlyless negatively than less sophisticated women voters do.

83 Conclusions

Overall, it appears that we can in fact make some general statements about the types of factors that influence evaluations of party leaders. First, partisanship and ideology have a greater impact on ratings than other variables do. After these two influential factors, attitudes towards issues have the largest impact, and socio-demographic background characteristics of voters appear to exert the smallest influence on evaluations (although not so small that the effect is non-existent, and also recall that this setup likelyunderstates the overall effect of socio-demographic variables). The effect of the leader’s sex is interesting, in that it (by far) has the greatest impact on ratings of character and competence. Voters of all types evaluate women party leaders more negatively than men leaders. This finding requires further research, in order to ascertain the extent to which other leaders’ characteristics (religion, education, socio-economic status, etc.) may have an influence on individuals’perceptions of leaders’ personality traits as well. Is the magnitude of the effect of the leader’s sex simply a general effect of leaders’ characteristics overall, where we could substitute sex for religion or language and get a similarlylarge effect? Or is there something about sex itself that matters so much? Perhaps running a series of country-specific analyses (where the data exist) might shed some additional light on this issue.

What these results show us, however, is that there are substantial differences in the way that different individuals weigh “background” factors when evaluating party leaders’ personality traits. The more sophisticated tend to consider more factors overall (as evidenced by the greater number of statistically significant coefficients), they tend to give more weight to the factors that they do consider (as evidenced by the larger size of most coefficients), and they tend to rely more heavily on their ideological views than their partisanship in forming

84 their perceptions. Finally,more sophisticated women tend to evaluate women leaders a little less negatively than less sophisticated women voters do, suggesting that the negative view of women leaders may not be a permanent endemic, but may shift as a) more women take a leadership role and people become more comfortable with the idea of women leaders; and b) education rises in society and more individuals in general become more politically knowledgeable and sophisticated.

85 CHAPTER SIX: THE IMPACT OF THE PARTISAN STEREOTYPE

Introduction

While the assessments of the origins of leader evaluations as presented in the last chapter included the role of partisanship of the voter, what we have not yet examined is the role of the partisanship of the leader. To what extent do voters perceive Conservative leaders in a certain way, simply because that leader is Conservative?

There is reason to believe that the party label provides a cue that imbues meaning and information to the voter, which the voter may then use to form his or her impressions of the party leader. Past studies have shown that individuals will use what information is made available to them, and that even in contexts of nearly no information they will find a way to assess party leaders, even if that means basing the assessment on the individual’s looks alone (Riggleet al. 1992). Scholars have suggested that the party label can act as a cue, which voters with less information can use as a heuristic or shortcut to assist with vote choice (Conover and Feldman 1989;McDermott 1997; McDermott 1998).

In this chapter I argue that the party of the leader affects voters’ evaluations, by acting as a stereotype that voters apply when confronted with party leaders. I begin by assessing research to date on the role of stereotypes and cues in perceptions of the political world (including the evaluations of party leaders), to provide background as to why we might expect a partisan stereotype to influence perceptions. I then present the results of a series of data analyses ifiustrating that a partisan stereotype does indeed exist—a stereotype that crosses national boundaries—in which leaders of Conservative parties are perceived to have

86 a strength on the Competencedimension and leaders of Left parties are perceived to have a

strength on the Characterdimension. This partisan stereotype shapes the way that voters

perceive party leaders. However, it is not the case that only those in low-information

conditions use the partisan stereotype as a cue. Respondents with higher levels of political

sophistication evaluate leaders in a way that conforms most with the partisan stereotype,

suggesting that it is not simply a heuristic or tool that only the least informed use to

compensate for a lack information. More research is needed in order to assess exactlywhy

the partisan stereotype has such a large effect on the evaluations of party leaders, and why

this effect is stronger among those segments of the population with the highest levels of

political sophistication.

Cues and Stereotypic Thinking

In one of the earliest studies of voting behaviour, Berelson et al. (1954) observed the lack of interest and knowledge among the majority of voters, and found that both were

closely associated with education—that is, that the more educated tended to be more interested in and more knowledgeable about politics (1954:25).These findings were not only confirmed, but also reinforced in Converse’s (1964) seminal work, in which he found that the mass public had little understanding of basic political concepts (e.g.left and right dimensions), and that the political ideas that voters did possess lacked constraint or consistency both horizontally (across ideas) and longitudinally (over time). The implications of these findings were devastating for notions of democracy which expected individuals to have some basic understanding of politics in order to be able to articulate their own interests: what is the point of democracyif citizens lack coherent attitudes and beliefs?

87 Political scientists’ discussions of citizens’ and voters’ processes of reasoning rely heavily on two main arguments or factors affecting reasoning. The first is drawn from

cognitive psychology, and suggests that “people have well-defined cognitive limits” (Lau and

Sears 1986, 349). Essentially, people do not have a very large active memory, and they generally focus their attention narrowly in many aspects of the environments they are in.

This suggests that even if individuals were to make concerted efforts to incorporate more information into their considerations of political events and issues, they may not ever be able to achieve the levels of information that would satisfy traditional views of democratic requirements. Thus it is not simply the case that individuals aren’t trying hard enough, but it may be the case that it is not possible to achieve high levels of knowledge.

The second main factor or consideration, drawn from economic and social science literature, is the notion that people are rational individuals who seek to maximize their benefits while minimizing their costs (Downs 1957).Jeffery Mondak summarizes this perspective nicely, suggesting that “correct decisions are preferable, but precision brings inefficiency; the citizen can form reliable judgments while simultaneously conserving valuable cognitive resources” (Mondak 1993, 168). Thus regardless of whether or not citizens are capable of incorporating more information into their decisions, it is probably not rational to focus heavily on information gathering. It is more rational to try to come to a conclusion based on as little information (and as little effort) as possible.

It is out of these two literatures that political scientists have come to some understanding of how voters cope with their cognitive limits, as well as the necessities of efficiency, to make reasoned decisions. Scholars have turned to concepts developed in cognitive psychology with regards to the storage and retrieval of information. Pamela

88 Conover and Stanley Feldman (1989) suggest that because of their limited capabilities of dealing with information, people use information that is already stored to arrive at their decisions. Indeed, many scholars suggest that individuals are able to make use of information shortcuts, or heuristics, to come to reasoned decisions (Brady and Sniderman 1985;Lau and

Redllawsk1997).

Among the many heuristics that individuals may make use of (see Lau 2003, for a detailed description of different types of heuristics), two are particularly pertinent in the evaluation of party leaders. First is the application of partisan and ideological schemata— which suggests that individuals will categorize candidates with whom they are relatively unfamiliar according to political schemata, and will assume that new information is consistent with existing schemata, and then will apply a “category-based affect” (Lau 2003,

45). Second, individuals may also apply person stereotypes, in which factors such as age, gender, race, and the way that a person looks willinform the individual’simpression of the candidates. Thus, “since women are traditionally seen as more compassionate than men, women candidates are often seen to be more competent on, or more concerned with, compassion issues such as helping the poor or advocating for children” (McDermott 1998,

899).

Rahn suggests that “our notions about what groups are like strongly influence how we appraise individual members of these groups” (1993,474), and that “in partisan elections, the most powerful cue provided by the political environment is the candidate’s membership in a particular political party. Even if voters know nothing else about a candidate, the ballot provides them with one important piece of information” (1993, 473). The party label, therefore, provides information to voters that will assist in the decision-making process. This

89 partisan stereotypeis different from the effect of the voter’s own partisanship. Conover and

Feldman (1989) note that voters will rely upon their own partisanship to make inferences

about candidates’ positions as well. This projection effect occurs in extremely low information settings, where voters will project their own issue positions (and/or partisanship) onto their preferred candidate.

The effect of the partisan stereotype is different from a projection effect, in that the label of the candidate provides information to the voter, information that the voter then uses to ascribe issue positions to that candidate. Thus the idea is that in low-information settings, that is, where voters lack the information that would allow them to wade through competing candidates’ issue positions and platforms in order to decide who to vote for, they will rely upon readily available cues (including the party label of the candidate) in order to be able to assign issue positions and make decisions (Conover and Feldman 1989; Kinder 1978;

McDermott 1997; McDermott 1998;McDermott 2005; Rahn 1993).

WhataretheStereotypesAbout Parties?

Rahn suggests that the use of partisan stereotypes may be a fairlyreliableway to

“simplify the political environment” (1993, 474), because parties differ in what are largely predictable ways. Hayes notes that party leaders are perceived to have strengths in certain personality traits in the American context (Hayes 2005), and puts forth a theory of “trait ownership,” much like theories of “issue ownership” that have emerged in voting behaviour literature both inside and outside the United States (Budge and Farlie 1983; Petrocik 1996;

Petrocik, Benoit, and Hansen 2003; Rabinowitz and MacDonald 1989).The idea is that certain executive characteristics have policy content, and much as parties tend to “own”

90 issues, their party leaders tend to “own” related traits. Petrocik describes the division of issue ownership in the United States:

Democrats are seen as better able to handle welfare problems. Perceptions of the parties on social issues (e.g. crime and protecting moral values) favor the GOP. The data also document the GOP’s hold on foreign policy and defense through the late I 980s. Opinions were mixed on economic matters, but were generally a GOP asset (by an average of about 13 points). Government spending, inflation, and taxation were also Republican issues (Petrocik 1996).

Because parties have an advantage in their issue areas, candidates will emphasize their party’s issues (Petrocik 1996; Petrocik, Benoit, and Hansen 2003; Sides 2006; Simon 2002), and because of the heavy focus on leaders in the media (Gidengil and Everitt 2000;

Mendelsohn 1993; Mendelsohn 1994;Mendelsohn 1996), the issue ownership is translated into the ownership of related personality traits. Unless shown otherwise, “voters will usually assume that a Democratic candidate is more liberal than conservative, that he/she favors social programs over defense programs, while Republicans are, for the most part, defense

“hawks” who support lower taxes and smaller government” (McDermott 1998).Brought to the level of specific personality traits, “Republicans appear to own leadership and morality, while Democrats own compassion and empathy” (Hayes 2005, 913).

These perceptions about the types of characteristics leaders ought to possess may have their origins in basic political attitudes. John Ray notes that Conservatives “...believe that man is not naturally good, are superstitious, and prefer hierarchical social structures.

They think highly of order, authority, and duty” (1973, 23). Thus, that Conservative leaders are perceived in greater numbers to possess traits related to order, authority, and duty suggests something about the way that individuals link parties to basic values and attitudes.

In TheAuthoritarianPersonality,Adorno and his colleagues note the susceptibility of certain types of personalities towards authoritarian values. They suggest that”.. .it will be granted

91 that opinions, attitudes, and values depend on human needs, and since personality is

essentially an organization of needs, then personality may be regarded as a determinantof

ideological preference” (Adorno et al. 1950, 5, emphasis in original).While they do not

specificallyaddress the nature of our evaluations of political party leaders, their observation

does point to an important link that can be drawn between individual political views and the way that we perceive party leaders—in particular, the way that we feel leaders shouldbe.

Right parties, therefore, also ought to be seen as more conservative than liberal, and more supportive of lower taxes and smaller government. Taking this argument one step

further, then, we ought to expect right of centre party leaders to have strengths in leadership in morality, while left of centre leaders should possess traits like compassion and empathy.

There is no reason to believe that the partisan stereotype should apply only in the American context. Parties play up their strengths in election campaigns around the world, and thus a

similar effect ought to exist across national boundaries. Finally,we ought to expect that the less sophisticated voters should evaluate leaders according to the partisan stereotypes the most: “since this information is readily availablewhile other political information is costly, we would expect voters in low information conditions to use these cues when voting”

(McDermott 1998, 898).

Impressions of Leaders’ Traits

The data indicate that indeed, individuals do perceive leaders through a partisan lens, and that this partisan stereotype is not merely an American phenomenon. Voters perceive the leaders of Conservative and Left parties as having party-specific personality strengths, and these perceptions are based solely on the party label of the leader. Conservative leaders

92 are rated more positively on the Competencedimension, while Left party leaders are rated more positively on the Characterdimension. This partisan stereotype exists even when we control for the partisanship of the voter, thus it is not simply a projection effect where those feeling an affinity to the Left party imagine that their leaders must display these characteristics. Partisansofdferentpartiesseetheleadersin thesamestereotypicwqy.

Conservativeleadersaremore“competent,“Left leadershavemore“character”

The data indicate that indeed, individuals do see party leaders based on a partisan stereotype. Leaders of Left parties are rated more highly on the character dimension, and leaders of Conservative parties are rated more highly on the competence dimension. Figure

6.1 illustrates this trend, through the use of box plots: the left hand side of the figure depicts evaluations of competence of the leaders of the three main party types, while the right side depicts evaluations of character. The line in the middle of each box represents the median evaluation of the leaders’ competence or character. As the figure makes clear, there are distinct differences in how respondents perceive leaders on the two dimensions. While the lines display medians, the mean evaluations of leaders on the two dimensions tell the same story. The average competence rating is 0.584 for Conservative leaders, 0.572 for Centre-

Left leaders, and 0.534 for Left leaders. The average character rating is 0.537 for

Conservative leaders, 0.542 for Centre-Left leaders, and 0.644 for Left leaders.

Table 6.1 expands on the trends in Figure 6.1. The Table depicts the results of a stacked regression analysis,in which evaluations of the character and competence of leaders of the three main party types were regressed on dummy variables indicating the party label of the leader. The coefficients are statisticallysignificant, and illustrate that Conservative leaders

93 are rated slightlymore negatively than Centre-Left leaders on the Character dimension, while

Left leaders are rated substantially more positively than Centre-Left leaders on this same

dimension. At the same time, Conservative leaders are rated more positively on the

Competence dimension, and Left leaders are rated more negatively than Centre-Left leaders

on this dimension.

Furthermore, analysis of the traits within dimensions suggests that these relationships

are not simply inflated by perceptions of leaders on a particular trait. Table 6.2 lists the

results of T-Tests performed to determine the difference in means between leaders on the

ten most frequently asked personality traits. All coefficients are statistically significant with

the exception of one, and indicate that not only do voters assess leaders within a partisan

stereotype on dimensions as a whole, but that they do so with regards to specific traits as well. Each column in the table represents a comparison between leaders of two of the three

party types. The third column, which compares evaluations of Conservative leaders with

evaluations of Left leaders, displays perceptions of the stereotypes most clearly.That is,

Conservative leaders score substantially higher than Left leaders on the traits leadership, knowledgeable, intelligent, inspiring, and arrogant, while Left leaders are rated more positively than Conservative leaders on the traits cares, honest, compassionate, trustworthy, and moral. Evaluations of Centre-Left leaders generally fall somewhere in between the two, where Left leaders still rank more highly on character traits, and Conservative leaders rank more highly on some competence traits, but not all of 19them.

19 This result might provide some insight as to the lack of historical “success” of centre-parties, as noted by Blais (2005). If voters’ evaluations of party leaders do play an important role in the electoral success of a party, then Centre parties might be at a disadvantage on all fronts—Party leaders on either side appear to be evaluated more positively on some trait, leaving the Centre Party with a lack of overwhelmingly supportive evaluations of its leadership. This is something that should be smdied in greater detail in the future.

94 That Left leaders are perceived more positively on the character dimension and

Conservative leaders are perceived more positively on the competence dimension is not

simply a pattern of one or two elections in one or two countries. Indeed, generally speaking, where respondents were asked to evaluate leaders from all three party types, the pattern holds for nearly all elections, as evidenced by Figures 6.2 and 6.3. These Figures depict the ratings of leaders of the three major parties, in comparison to the grand mean for ratings of the leaders of all the parties included in each election study. Figure 6.2 graphs evaluations of leaders’ competence, while Figure 6.3 graphs evaluations of leaders’ character. Elections are ordered in relation to the extent to which the ratings of the leaders fit the partisan stereotype: that is, in Figure 6.2 moving towards the far right of the graph, Conservative leaders were perceived most strongly on the Competence dimension in comparison to all other leaders. In Figure 6.3, as we move to the right of the graph, Left leaders were perceived most strongly on the Character dimension in comparison to all other leaders. The graphs illustrate that evaluations of the leaders of Left and Conservative leaders generally fit the partisan stereotype, when respondents are asked to evaluate all three.

When Left leaders are not included in the choice set, the pattern is not quite as stark.

Indeed, as Figures 6.4 and 6.5 illustrate, the leaders of the Conservative and Centre-Left parties are more closely matched in their likelihood to be perceived more positively than other leaders on both dimensions. This applies even when leaders of Left parties form part of the choice set in the election but respondents in the election study are not asked to evaluate them: in the 1968 Canadian election, respondents were not asked to evaluate the competing leader of the New Democrat Party (a Left party), and in that election respondents perceived the Liberal (Centre-Left) Pierre Trudeau more positively on the competence

95 dimension, and the Conservative Robert Stanfield more positively on the character dimension. It appears that the act of comparison might make a difference in activating respondents’ perceptions of the partisan stereotype. The act of comparison is an explicit part of the activation of cues, according to Conover (1981).She suggests that voters willlook at a field of candidates and note the obvious differences between them, including differences of partisanship. She states: “contextual factors such as the minority status of one candidate as compared to others—be it ideological, partisan, racial, or sexual in nature—may encourage

voters to apply the stereotype associated with the minority group to the individual. . . the

“conservative label”, for example, should be a more salient cue in a field of candidates in which one is a conservative and the rest are moderates and liberals” (1981:433).

These findings support others, who suggest that the act of comparison is key. Rahn et al. (1990)run their candidate models in two ways: first, separately for each candidate; and second, using comparative scores for judgments. They find that the comparative model is more accurate and suggest that “the entire judgmental process appears to be comparative”

(1990: 119).While Rahn et al.’suse of a comparative measure of evaluations to understand voter perceptions is reasonable in the American two-party system, in a multi-party system it makes more sense to code traits in relation to one leader at a time, as I have done in this study.

It is not entirely dear what exactly is activating this partisan stereotype. What is clear, however, is that this partisan stereotype is not simply partisanship by another name. In fact, when we control for the partisanship of the voter, the extent to which individuals perceive leaders within the partisan stereotype framework becomes even more evident. Table 6.3 lists the results of a stacked regression analysis,in which evaluations of character and competence

96 of the three main party leaders were regressed on a series of “partisan” dummy variables. A dummy variable was created for the party label of each leader, and this variable was interacted with dummy variables for the partisanship of the voter. There are five groups of partisans—Centre-Left, Conservative, Left, Non-Partisans (those claiming to be either

Independents or to have no partisan affiliation), and Partisans of “Other” Parties. The coefficients in the table represent the ratings of the leaders of each of the three main party types, among groups of partisans.

First, there is an overwhelming effect of individual-level partisanship. That is, all partisans view the leader of their own party most positively. Individuals identifying with a

Conservative party view Conservative leaders most positively on both dimensions, those claiming an affiliation with a Centre-Left party rate Centre-Left leaders most positively on both dimensions, and Left partisans perceive their own party leaders most positively on both dimensions. After their own leader, however, partisans of the three main parties rate

Conservative leaders more highly on competence, and Left leaders more highly on character.

So while voters still perceive the leader of their own party most positively, they rate the remaining leaders according to the partisan stereotype.

Among both non-partisans and “other” partisans, perceptions of leaders follow the stereotype most clearly.Non-partisans rate Left leaders more positively on the Character dimension, and Conservative leaders more positively on the Competence dimension.

Partisans of other parties do the same. That this dynamic exists among non-partisans and

“other partisans” provides the greatest indication that voters reallydo perceive leaders according to a partisan stereotype, and that this is not simply about voters perceiving the leaders of their own parties in a positive light. Furthermore, these models include sampling

97 weights and fixed effects in order to ensure that no one study is skewing the results. The

results of this analysis tell us that generally speaking, voters believe that Conservative leaders

are the most competent, and that Left leaders have the most character.

Partisan Stereotype: Not Just a Shortcut or Cue for the Least Informed

The literature on stereotyping and the use of cues, including partisan cues, suggests

that voters will tend to make use of them in situations of low-information. McDermott

suggests that “even in low-information elections, voters inadvertently obtain basic

information about the candidates such as party identification and incumbent/challenger

status. Cues such as these can help voters make decisions in an otherwise uncertain situation.

Through past experience and stored knowledge, voters can make reasonable assumptions

about the ideology of a candidate based on associations with salient political or social groups.

In other words, voters use candidate cues as cognitive shortcuts...” (1997, 271). While

McDermott suggests that the shortcuts will allow voters to guess theviewsof candidates, it is

reasonable to assume that voters will also employ the shortcuts to allow them to interpret

traits as well.

Thus the less knowledgeable should rely on the use of shortcuts more—that is, individuals with lower levels of political sophistication ought to perceive leaders through the lens of the partisan stereotype more than those with higher levels of political sophistication,

because they lack the information that would allow them to distinguish leaders from the party label. In fact, this is not the case. These data show that the more politically sophisticated perceive leaders more in line with the stereotype than do the less sophisticated.

These results suggest that the partisan stereotype is not simply an information shortcut or

98 heuristic assisting less informed voters to wade through competing candidates, but that something different is taking place.

Figure 6.6 replicates the graphs in Figure 6.1, this time according to the political sophistication of the voter. The graph on the left compares competence ratings of the three main party leaders among the least politically sophisticated 25% of voters, and the most politically sophisticated 25% of voters. The graph on the right does the same for ratings of the leaders’ character. The box plots indicate that those with a higher level of political sophistication are more likelyto perceive the leaders as conforming to the partisan stereotype than the least sophisticated. The lines in the middle of the box plots mark the median rating for each leader on this dimension, and the mean ratings fit the same pattern.

Among the least politically sophisticated, respondents gave Conservative leaders an average of rating of 0.582, Centre-Left leaders an average rating of 0.591, and Left leaders an average rating of 0.552. Among the most sophisticated, respondents gave Conservative leaders an average rating of 0.615, Centre-Left leaders an average rating of 0.562, and Left leaders an average rating of 0.488. The perceived differences between party leaders are stronger, and more clearlyifiustrate that Conservative leaders are most highly rated on competence.

The same dynamic can be seen in the left side of the figure as well. Within the least politicallysophisticated group of respondents, the Conservative leader gets an average character rating of 0.529, the Centre-Left leader gets an average of 0.547, and the Left leader gets an average of 0.604. Among the most politically sophisticated, the distinctions between leaders are greater: Conservative leaders get an average rating of 0.526, Centre-Left leaders get an average rating of 0.533, and Left leaders get an average rating of 0.648. Thus while both groups (less sophisticated and more sophisticated) rate the Left leader most highly on

99 character, the more sophisticated rate the Left leader more highly, and they rate the leaders of the other two parties slightlylower than do the least sophisticated.

These results hold when we control partisanship as well: respondents with higher levels of political sophistication perceive party leaders through the lens of the partisan stereotype more than those with lower levels of political sophistication. Table 6.4 replicates the stacked regression analysespresented in Table 6.3, this time by level of political sophistication. As the table indicates, partisans still view the leader of their own party most favourably, regardless of their level of political sophistication. Furthermore, the impact of partisanship is stronger among the more politically sophisticated: those with higher levels of political sophistication rate the leader of their own party even more favourably than do those with lower levels of political sophistication.

While the effect of partisanship is stronger among the more sophisticated, the effect of the party label is also stronger. Table 6.4 includes a series of arrows indicating where the size of coefficients conforming to the partisan stereotype grows as we move from the less sophisticated group to the more sophisticated group. These arrows indicate that more sophisticated respondents almost alwaysgive higher stereotypic ratings to party leaders than do less sophisticated respondents. For example, among non-partisans, the less sophisticated respondents give Left party leaders a rating 0.065 points higher than the reference group

(ratings of the Centre-Left leader among “other” partisans). Meanwhile, the most sophisticated non-partisans give Left party leaders a rating 0.132 points higher than the reference group. Similarly,the most sophisticated non-partisans give a rating 0.086 points higher than the reference group, while the coefficient for the least sophisticated non partisans is 0.010 (and does not reach traditional levels of statistical significance).With few

100 exceptions, the more sophisticated are more likelyto rate leaders consistent with the partisan

stereotype.

Among Left partisans the effect is particularly interesting. Although all rate their own

leader most positively on the character dimension, regardless of level of political

sophistication, on the competence dimension something very different happens. Less

politically sophisticated Left partisans give their own leader the highest rating on

competence. More sophisticated Left partisans give their own leader a lower rating than less

sophisticated partisans, but they also give the highest rating to the Conservative leader. This

suggeststhat theimpactofthestereotypeisevenlager thanpartisanshiyamong the most

sophisticated Left partisans!

That the most politically sophisticated tend to perceive leaders in a manner most

consistent with the partisan stereotype suggests that the stereotype is not simply a tool utilized by the least informed in order to be able to distinguish between party leaders. Rahn

(1993) does note that individuals are likelyto continue to rely upon the partisan label even when other types of information about candidates are available, suggesting that increasing levels of information do not necessarily change the preferred sources of information among those evaluating candidates. Furthermore, these findings conform with findings of previous scholars who have noted that the more sophisticated tend to use information shortcuts just as much as or even more than the least sophisticated (Cutler 2002; Sniderman, Brody, and

Tetlock 1991). It is possible, however, that the party leaders themselves actually do conform to the partisan stereotype: that the most sophisticated, possessing greater amounts of information about these leaders, are able to assign ratings that more closely match the

“truth,” and that this truth just happens to fit the stereotype. Arguably, for example, parties

101 might select certain “types” of people as leaders—these types may fit the stereotype. More research is needed before we can really determine what exactly is happening. What seems fairlyclear, however, is that the partisan stereotype is not simply a shortcut for the least informed.

Conclusions

Past studies of the origins of leader evaluations have pointed to the influence of the party label as a cue to be used in low-information settings, allowingindividuals to ascribe issue positions to candidates when actual positions are not known (Conover and Feldman

1989). Other research on stereotypes and heuristics has suggested that individuals will use information available to them (including partisan, racial, and gender stereotypes) in order to formulate opinions and make decisions when other types of information are not available

(McDermott 1997; McDermott 1998; Rahn 1993).As such, it seemed reasonable to expect that voters, especiallyless sophisticated voters, would rely upon the party label in order to evaluate party leaders.

This chapter has ifiustrated that indeed, the party label does transmit information to voters, and a partisan stereotype does exist: respondents perceive leaders of Left parties more positively on the character dimension, and they perceive leaders of Conservative parties more positively on the competence dimension. Generally speaking, even when controlling for the partisanship of the voter, individuals continued to perceive party leaders as examples of the partisan stereotype. In contrast to expectations, however, the partisan stereotype is not simply a tool used by the least informed in order to compensate for a lack of information. The most politically sophisticated segment of the sample evaluates party

102 leaders in a fashion that conforms most strongly with the partisan stereotype, providing Left leaders with even higher character ratings and Conservative leaders with even higher competence ratings than the less sophisticated group. What is not known is why. If it is not simply an information shortcut, what exactly is going on? Why do more informed individuals make use of the stereotype more than less informed individuals? More research is needed.

103 CHAPTER SEVEN: THE IMPACT OF LEADERS’ TRAITS IN ELECTIONS

Introduction

This chapter discusses the effect of evaluations of party leaders on vote choice. I present a few key findings. First and foremost, leaders matter. Second, evaluations of leaders’ character affect vote choice more than evaluations of their competence. Third, there are important differences in the way that voters perceive party leaders depending on their level of political sophistication. Finally,while leaders play an integral role in the vote calculus, the actual net effects of leaders on the popular vote range substantially from election to election and from place to place.

Theory

Are TraitsImportant?IVIyShouldWeFocusonLeaders?

Dating from the earliest years (Berelson, Lazarsfeld, and McPhee 1954; Campbell et al. 1960), most studies of voting behaviour have pointed to the critical role played by long- term forces: party identification, ideological beliefs, and the socio-econoniic or demographic characteristics of voters. Authors suggest that who we are as people—characteristics intrinsic to how we grew up and how we were socialized—affects how we vote. Thus, for example, gender affects vote choice and issue attitudes (Almond and Verba 1963; Gidengil et al. 2003;

Inglehart and Norris 2000); as does partisanship (Campbell et al. 1960; Green, Palmquist, and Schickler 2002); and other socio-demographics (Bartels 1996; Cutler 2002).

104 While the importance of stable and long-term forces is fundamental to understanding

voting behaviour, a comprehensive look at voters’ decisions must also consider “short-term”

forces. Stokes, Campbell and Miller (Stokes, Campbell, and Miller 1958) note that short-term

fluctuations in vote choice and preferences cannot be accounted for by long-standing

predispositions: we do not usually switch our sex between elections, nor do we suddenly

have a different ethnic background, therefore these things can’t really explain why our

preferences change. Partisanship, while it moves a little bit, is a fairlystatic and long-term identification as well (Green, Palmquist, and Schickler 2002;Johnston 2006). Stokes and his

colleagues (1958) suggest that candidates and issues can account for change where factors such as partisanship and gender cannot.

Stokes (1966) later echoes these earlier comments, suggesting that party identification is not sufficient to explain shifts in vote choice, because it does not reallychange, while the evaluation of leaders is constantly shifting, and thus has greater potential for explaining fluctuations in the vote. Miller and Shanks (1996) support these arguments for the inclusion of short-term forces in vote models with their updated version of Campbell et al.’s(1960)

“funnel of causality,” the block recursive model, in which both long-term and short-term forces have their proper places in models of vote choice. The block recursive model they advance organizes explanatory variables into a series of themes (blocks). These themes are organized and entered into vote models in order from most removed from the vote

(demographics) to most proximal to the vote (evaluations of party leaders). The idea is that by entering groups of variables in their proper order, we can allocate an appropriate role to variables for both their direct and indirect effects on the vote. Miller and Shanks put leaders evaluations in the final block of their model of vote choice.

105 While all of these observations point to reasons why we ought to include evaluations

of leaders in vote models, they do not necessarily explain how much evaluations of leaders

affect the vote, or voters focus on leaders in the first place. One often-cited reason for

the importance of leaders in vote choice is the media focus on party leaders and the

“horserace” during election campaigns. A number of scholars have observed that media coverage of election campaigns tends to prime leaders, which has the effect of encouraging voters to base their attitudes more heavily on leaders than other factors (Mendelsohn 1993,

1994, 1996; Gidengil et al. 2000; Gidengil and Dobrzynska 2003). The idea is that since the media focus so heavily on leaders’ personalities—what they are doing, saying,and where they are in the “race”—it is natural that as consumers of the media, voters are also likelyto focus heavily on party leaders when making their choice at the ballot box.

In addition to the effect of media priming, it has also been suggested that deciding how we feel about others is a relatively “easy” process. People evaluate others regularlyin everyday life: Cottrell et al. (2006:2) suggest that “humans, as discriminately social creatures, make frequent judgments about others’ suitability for interdependent social relations.” They suggest that as individuals, we only have so much time, and cannot be friends with everybody; therefore we must make judgments as to whether or not we think others are worth our time. Perhaps making judgments about party leaders is much the same thing:

Rahn et al. (1990) suggest that even for those voters who are not terribly interested or involved in politics, forming candidate images ought to be fairlysimple, since it mirrors activities that we perform on an everyday basis. We do not need to develop an entirely new skill to be able to decide how we feel about others, so this makes it more likelythat we wifi evaluate leaders, and that those evaluations will factor into vote choice.

106 In addition to the “ease” or inevitability of the evaluation of leaders, and the fact that we are primed to think about leaders during election campaigns, it is conceivable that considering leaders is also actually a reliable way to make inferences about a candidate’s future performance in office. Popkin (1991) argues that the behaviour displayed by a leader can give clues about the kind of representation he or she will provide. Voters can gain insight into future behaviour. Glass (1985) suggests that the evaluation of candidates and leaders may in fact be a response to the complexity of political life. Circumstances may change, and a candidate may need to adopt new policies, but he/she is unlikely to be able to change his or her personality. Perhaps, then, personality is more reliable than platform or party as an indicator of how the individual will act in office. Rosenberg et al. (1986) put forth a complementary argument, suggesting that image and physical appearance matter because they provide clues about a candidate’s character and fitness for public office. How a leader presents him or herself provides us with some indication of the individual’s ability to do the job.

This reliance upon perceptions of leaders to make voting decisions may be related to conditions of low information. That is, maybe voters are using leaders as a “shortcut” to help them make the vote choice. Less informed voters, who lack the knowledge or political sophistication required to make voting decisions based on policy, decide whether or not they like the party’s leader, and vote for the party largely on that basis. This mirrors the notion put forth by Sniderman et al. (1991),who suggest that people can figure out what they oppose or support if they can simplify their options, and that among the less educated, affect

(or how you feel towards something or someone) plays a significant role in explaining policy preferences. Essentially, you may not know a lot about a candidate, but with relative ease you

107 can choice

How

as choice conclusions leaders

pivotal the

looks effects

the 0.5%

advantage source

very small.

distribution may

noted

decide

1988

net

Much

have

often”

to

at

They

by

for

Exactly

on

of Later, had

role

impact

Battle

the 3.5%.

earlier,

Canadian

acting

dynamism Do

resulted

the 21% whether

ranging

a

in

are

effect suggest (2002:

of

small

Johnston Evaluations

and vote

the

In

what of

leaders’

of

mixed.

limited,

on

an leaders

election.

voters

Crewe of

in anywhere

election, 93).

net

range or

that

that

kind

earlier

in

leaders’

the

not

impact

Blais

Bean traits

elections,

(2002)

and feeling.

of “leadership

in

“difference

anywhere traits

note

of

you

Leaders

piece,

the

and

where

impact

et

from can and

traits

on like

in examines

on at.

2000

suggest

affect the

Mughan

and

the

Stokes

(2002) Affect

the

2% him

in

scholars

from

leaders

effects

British

between

Canadian

vote.

American

that

vote

to

or

the

Vote that

the

noted

(1966) 0-2%,

8%.

her,

the

(1989),

Johnston

varies

balance

have

have

may

impact

context

perceptions

Choice?

victory

impact

and

election,

that

depending

elections

suggested

well

on

attempted

by

however,

you

of

leaders of

vote

that

election,

et

matter,

of

and

the

traits

can

at.

but

candidates

the

choice

of

from

party

defeat”

(1992)

that

on

therefore played

argue

to

competence in

suggest

effect

but

and

the

answer Canada

leadership

1980

vote.

is

not

assess

that

ranges

for election.

not

a

of

led

at crucial

to

by simplify

They

the

personal

this the clear.

the

and

to

2000,

the

very

may

from

Australian

provides

a

variation

same

question,

Bartels

suggest finds

role

net

role

The

much

and

have

your

approximately

traits

partisan

in

of time

that

literature

finds

vote

(2002)

played vote

traits

that the

and

in

Labor

was

that

the

main

that

traits

not

in

net

a

is,

108 party in 1987. Their research found that traits gave a net 5.8% vote advantage to the British

Conservatives and a net 3.7% advantage to Labor in Australia. Taken as a whole, the results of these analyses indicate that leaders’ personality traits can have an impact anywhere from 0 to 8%, and that the net impact on the electoral result varies depending on the importance of traits in the given election, as well as the distribution of perceptions of leaders’ traits in that year: a leader may have been perceived way ahead of the group on a particular trait, but if voters are focusing mainly on an issue during the election, the leader’s advantage on that trait may not have an impact at all. On the other hand, if traits are reallyimportant in a given election, even the slightest perceived advantage can translate into votes.

This brings us to the question of the relative importance of traits compared to other factors affecting vote choice. In the French context, Pierce (2002) suggests that the leader is one factor among many, and downplays the importance of leaders. Winham and

Cunningham (1970) find in the Canadian context that leader preferences and evaluations had a greater impact on vote choice among voters not identifying with a particular party. In contrast, they find that partisans generally tend to vote for their party, regardless of their evaluations of the party leaders. Bean (1993) makes similar findings, suggesting that leader effects are small compared to partisanship and other factors, but argues that in a close race, leader effects can be decisive.

In contrast to those claiming that leaders usually do not have a large impact,

Banducci and Karp (2000) find that evaluations affect vote choice, even when other factors are controlled for. Glass looks at US elections, and his findings take the importance of leaders one step further. He states, “in short, looking at all voters, personal attributes had as large or a larger impact than policies on their vote in five out of seven elections” (1985: 530).

109 It

relative

given Stewart

Vote

vote

partisanship,

leaders’ the

competence; outcome

independent with

logit

analyses. fact-finding patterns

20 anything regression model have probit

seems

Scholars

air.

choice,

argued Models: credit

vote

models,

models

from

to

Scholarship

and

The

To

as

traits

from

analyses, (which other of

in Most

choice though

my

that

gauge

Clarke

for.

the

analyses

and elections.

as

simple mission.

issue

and variables.20

do

with

“toolkit.”

The logit

well.

past

“decisive”

importantly,

affect

the

I

and

as

the second,

leaders will

Ordinary attitudes, (1992) have models

different

thus

Relative

the

montecarlo

question

presented

impact

A

Furthermore,

the describe

utilized

dependent

number

far

I are

argue

may

that

Least vote

factors

ran

is

more

and

“combinations”

while

of

a Importance

of inconclusive

simulations.

evaluations standard have number

Squares in

in calculus,

leaders’

that

how

socio-demographic appropriate of

a

this

like

a

number the

variable,

interesting

moment)

an

in

of much

linear

chapter partisanship,

sizes

fact,

important

logit

different

traits

Since

and

of

of

for

regression

of

as

leaders scholars of leaders

and

models, character

the of party

modeling

on Traits

were to

leads

coefficients

findings

statistical

party

dependent

the

traits

vote

role

leaders

(Dow

about

are issue models, affect

to

relative

variables

in

multinomial

vote vote

choice,

in

(among

two

has

more

models

emerged Voting

and

influencing

attitudes, variable

choice the

electoral

to

have in

a

differed

main

Endersby

effect

weightier logit,

important

the

same, to it

are

other

than

was

uncover is for

a

conclusions:

from

probit,

dependent

binary, discernible accounted

logit

of

and

outcomes

Three

are across

suggesting

necessary

2003;

explanatory vote

leader

impact probit

these

the

multinomial ideological

models,

I

than

eliminated

Kropko

impact choice,

models,

Main

models,

evaluations

preliminary variable,

for

they

effect

once

than first,

to

that

and of

2008;

is

factors) run

Parties

logit,

even

the are

leaders,

views.

still

so

the

that

the on

stacked

linear

Long I

analyses

as being

stacked

eliminated

general up

the

on

a

using

as

OLS in 1997)

110 conclusions I draw from this cross-national dataset are about right. The model I am most

comfortable with is a logit model where the dependent variable coded as 0/I: where vote for

the party in question is I and vote for any other party is coded as 0. I ran three main analyses with vote for the three main party types as dependent variables, looking at vote choice for

the Conservative Party/Other; Centre-Left Party/Other; and Left Party/Other.

This model is preferable to a multinomial logit model because of the drastically

reduced sample size in the MNL setup: all those not voting for one of the parties embedded in the dependent variable (either Conservative, Centre-Left, or Left) are dropped from the

analysis,and therefore the corresponding number of empty cells was substantially higher in

the MNL setup. Furthermore, having a setup with all of the parties across all of the countries

in the model gets messy and makes less sense, since no voter faced all of those choices when

they went to the ballot box. Finally, the sizes of traits coefficients were largest in the IVINL

setup (whether two parties or three parties made up the DV), perhaps because of the nature

of the choice between specific parties. The coefficients in the three logit models were

smaller, thus providing a more conservative estimate of the impact of leaders. I prefer to

understate rather than overstate the impact of traits. The estimations are depicted in Tables

7.1 through 7.3, respectively.

Similarlyto the setup in Chapter Five, which examined the sources of evaluations, the

model is based on the Miller and Shanks (1996) update of the “funnel of causality.”While

the results presented in Tables 7.1 through 7.3 are not really from a pure block recursive

model, sets of variables were added to the logistical model in blocks, from the most long

term to most short-term forces, to control for prior variables and to illustrate the effects of

the more proximal variables. Thus column 1 lists coefficients and standard errors for socio

111 demographic variables and partisanship, column 2 incorporates the addition of ideological self-placements and issue-attitudes, column 3 includes evaluations of the leader of the party in question, column 4 includes evaluations of a second party leader, and column 5 includes evaluations of the third party leader. The models include leader evaluations of only the three most prevalent party types, mainly because of the drastic reduction in sample size if we incorporate any additional leaders into the 21model.

What becomes immediately evident from the tables is that leader evaluations matter.

All variables (with the exception of age) are coded on a 0-1 basis. Socio-demographic and partisanship variables are binary, while ideology, issue attitudes, and leader evaluations are scalar and range from 0-1, as described in Appendix A. Comparing coefficients at their face value is largely feasible, as a result. The coefficients for leader evaluations are alwayslarger than those for both socio-demographics and issue attitudes, and are often even larger than ideology and partisanship. Even with all of these other variables incorporated in the model, leader evaluations have a sizable impact on vote choice.

Furthermore, and perhaps more interestingly, the data indicate that not all traits matter in the same way: character evaluations appear to have a larger impact on vote choice than evaluations of leaders’ competence. This may disappoint those who would have voters consider nobler factors like policy stances and platforms when deciding which party to vote for. It is perhaps problematic that leaders’ traits appear to matter more than issues, but surely a leader’s competence ought to have a larger effect on vote choice than his or her character?

These data do confirm earlier results based on fewer election studies (Johnston 2002),

21 As is immediately obvious, even incorporating evaluations of the leader of the third most common party type, the Left party, leads the number of observations included in the model to drop considerably. There are substantially more respondents across the election studies incorporated in this analysis who evaluated the Conservative and Centre-Left Party leaders but did not evaluate Left Party leaders.

112 suggesting that character is more important than competence. Johnston et al. (1992) also find that character evaluations influence overall “feelings” towards a leader more than competence evaluations.

Table 7.4 illustrates more clearlythe relative impacts of leaders’ character and competence on vote choice. The values listed are changes in the probability of vote for the party with a change in the evaluation of the leaders’ character and competence from one standard deviation below the mean to one standard deviation above the 22mean. Thus when holding all other variables in the models at their means, voters were 46% more likely to vote for the Conservative party if they evaluated the Conservative party leader one standard deviation above the mean Conservative rating on character, compared to those giving the leader a rating one standard deviation below the mean. A similar change in the competence rating led to a 32% increase in the likelihood of voting conservative, while a similar change in the evaluations of both traits led to nearly a 70% increase in the probability of a

Conservative vote.

The effects of evaluations have a nearly similar impact on vote for the Centre-Left parties. A change in the evaluation of the Centre-Left party leaders’ character from one standard deviation below the mean to one standard deviation above the mean leads to a 43% increase in the probability of voting for the Centre-Left party, while a similar change in the evaluation of the leaders’ competence leads to a 20% increase in the probability of a Centre-

Left vote. The combination of the two leads to a 70% increase in the probability of a Centre

Differences were generated using Long and Freese’s add-on for Stata, SPost (Long 1997; Long and Freese, 2005). SPost was used over Clarify to calculate first differences because the data and corresponding analyses were weighted, and Clarify does not allow the use of probability/sampling weights in its simulations. AUvariables in the model were held at their means, including partisanship. The resulting estimate of the effects of leaders’ traits are therefore simulations based on a fictional “multi-party” partisan, made up of all PID variables in the model. Later, in Table 6.6, simulations are run for different groups of partisans.

113 Left vote. These changes occur while holding partisanship, issue attitudes, and ideological self-placements at a constant—while these are hypothetical simulations, they ifiustrate that evaluations of party leaders have an effect on vote choice, as well as showing the over arching influence of the leaders’ character.

Not all traits have an equal impact, nor do the traits of different leaders. With respect to voting for the Conservative and Centre-Left party leaders, positive evaluations of the opposing leaders’ character and competence lead to a reduction in the probability of vote for the party in question. However, positive evaluations of the Conservative leader have a greater impact on vote for the Centre-Left party than positive evaluations of the Centre-Left leader on vote for the Conservative party. In particular, evaluations of the Conservative leaders’ competence weigh in a lot more heavily than evaluations of the Centre-Left leaders’ competence.

That evaluations of leaders have differing impacts on vote choice becomes even more clear when we look at the impact of evaluations of Left party leaders on the vote. As seen in Table 7.4, the predicted probability of a vote for the Left party with a change in ratings of the leaders of any of the three main parties (including the Left party leader!)is substantially diminished (i.e.the values in column 3 are considerably smaller than the values in the first two columns).

These results support earlier findings suggesting that leaders have a larger impact on vote choice for “main” parties, and less of an impact for third or smaller parties (Johnston

2002). And indeed, it does appear that these results are about the size of parties, and do not simply reflect the fact that Leftleaders in particular matter less. Similarvote models run for

114 Right parties have equally small (and even smaller) predicted probabilities for evaluations of leaders’ traits, as do vote models for centre-right and sectional 24parties. Returning to the values presented in Tables 7.1 through 7.3, it appears that not only do evaluations of leaders have an impact on vote choice, but they also appear to have an effect that rivals partisanship and individuals’ideological views. When looking at the overall size of coefficients, it is clear that non-moving socio-demographic variables have very little effect on vote choice compared to other variables. Furthermore, most of the socio demographic variables do not approach traditional levels of statistical significance.The variables as shown are not really significant, but recall that this complement of socio demographic variables is not ideal and probably understates the effect of core socio demographic indicators.

After socio-demographic variables, issue attitudes play the second least important role in affecting vote choice. While most of the coefficients are statisticallysignificant, they are quite small,with a much smaller impact on vote choice. The impact is not negligible, however, and Table 7.5 illustrates the isolated effect of changes in issue attitudes (from the most conservative position (0) to the most liberal position (1)) on the probability of vote for the three main parties. The impact of issue attitudes ranges from a 3% difference in the probability of voting for a party (the Left party) to just under 13% change in the probability of voting for a party (Conservative).

Ideology, partisanship, and evaluations of party leaders have the largest impact on vote choice. As Table 7.5 illustrates, when we compute changes in these independent variables, while holding all others at their means, the difference in probability of vote for one

24Results not shown, but changes in the predicted probabilities of vote for these minor parties range from 0.04 to 0.12 with positive evaluations of party leaders’ traits.

115 of the three major parties is striking. Ideology appears to have an impact greater than

Conservative partisanship (when compared to non-partisanship), but this is probably largely

the result of a measurement artifact: ideology is coded on a 0-i scale, but the difference between 0 and I (most right and most left) is substantially larger than the difference between

oand I (non-partisan and Conservative partisan) on the Conservative PID variable. There are many who consider themselves Conservative partisans that would not place themselves on the furthest point to the tight of the ideological scale. Indeed, when we compare the effects of identification with a Conservative Party to identification with a Centre-Left party, the effect of partisanship is clearer: Centre-Left partisans are 42% less likely to vote for the

Conservative party than Conservative partisans, all else being equal. Similarly,Left partisans are nearly 73% less likely than Centre-Left partisans to vote for the Centre-Left party. Party

Identification plays a major role in determining vote choice.

The most important comparison to be made, however, is a comparison between the values in Table 7.5 and the values in Table 7.4. The predicted increase in the probability of vote for a Centre-Left Party, for example, is nearly 64%, with a two-standard deviation increase in the rating of the Centre-Left leaders’ traits, while all other variables (including partisanship) are held at their means. The comparable for the effect of partisanship is the predicted probability of vote for the Centre-Left Party among Conservative identifiers versus

Centre-Left identifiers. Centre-Left partisans are 60% more likelyto vote for their party than are Conservative partisans, with all other variables (including leader ratings) held at their means. These values suggest that a large jump in ratings of a leader can have an effect on as large a scale as partisanship. In the real world, however, the number of people who rate a leader two standard deviations apart, all other things equal, is probably substantially smaller

116 than the number of people identifying with two different parties. So while the effect of

leader ratings looks like it might be similar to the effect of partisanship, the real effect is

likelymuch smaller.

Partisanship also has the effect of intensifying the influence of party leaders: partisans

who positively evaluate the leader of their own party are even more likelyto vote for their

own party than are partisans who evaluate the leader of their own party poorly. Table 7.6 illustrates this effect. In the first block of column one, for example, we see that Conservative partisans who evaluate the Conservative leader one standard deviation above the mean

character rating and the mean competence rating are 73% more likelyto vote Conservative

than are Conservative partisans who rate their leader one standard deviation below the mean

on the two traits. In the second block of column two, we see that Centre-Left partisans giving their leader a rating one standard deviation above the mean rating on both traits are

70% more likelyto vote for the Centre-Left party than are Centre-Left partisans who rate their leader one standard deviation below the mean on the two traits.

This pattern continues in the third block of the third column, where Left partisans evaluating their leader one standard deviation above the mean rating on the two traits. The comparable figures for Centre-Left and Left partisans are 70% and 37%, respectively.

Among non-partisans, the effects of ratings of Conservative and Centre-Left leaders’ character and competence are even stronger: those giving Conservative leaders higher ratings on both dimensions are nearly 79% more likelyto vote for the Conservative party, and those rating Centre-Left leaders higher on both dimensions are just over 72% more likelyto vote for the Centre-Left party.

117 The Impact of Political Sophistication

Do leaders act as information short-cuts, helping the less knowledgeable decide how to vote? Or do they provide a source of information about performance in office that only

the more sophisticated can truly integrate adequately into vote choice? It looks like it might actually be a bit of both: the impact of leader evaluations varies by level of political sophistication, and the patterns suggest that leader evaluations may act as a shortcut for the least politically sophisticated, and may also provide a complex and “higher” set of information for the most sophisticated.

The analyses presented in Tables 7.1 to 7.3 were re-run separately by respondents’ levels of political sophistication. Because of the substantial drop in sample size, and the limited impact of leader evaluations on vote for the Left party, the analyseswere limited to two dependent variables: vote choice for a Conservative Party and vote choice for a Centre-

Left party. Tables 7.7 and 7.8 list the results of these analyses.

Those with a higher level of political sophistication evaluate a larger number of factors overall. This comports with research suggesting that the more politically sophisticated consider a greater number of factors in the vote calculus (Cutler 2002; Sniderman, Brody, and Tetlock 1991). In Table 7.7, the number of coefficients achieving traditional levels of statistical significance among the less sophisticated is ten, compared to twelve among the more sophisticated. In Table 7.8, the numbers are eight and twelve, among the less and more sophisticated, respectively.

Second, for most variables, coefficients are larger among the more sophisticated than the less sophisticated, suggesting that not only do the more sophisticated consider more factors overall, the factors that they do consider also play a more weighty role in the voting

118 calculus. Among both groups, socio-demographic variables continue to have a minimal

effect on vote choice, but partisanship matters more among the more sophisticated, ideology

and issue attitudes matter more among the more sophisticated, and leaders also matter more

among the more sophisticated.

To isolate and better understand the impact of leader evaluations within the two groups, I show first-difference calculations in Table 7.9, exactlylike those in Table 7.4 for

the full sample. All independent variables were held at the sample mean, and then the change in predicted probability of vote was generated, based on a change in evaluations of leaders’

character and competence, from one standard deviation below the mean rating to one

standard deviation above the mean. Essentially, I ran hypothetical changes in evaluations through the models in Tables 7.7 and 7.8. Table 7.9 lists the changes in probability of vote

for either the Conservative Party or Centre-Left party, according to level of political

sophistication.

For both Conservative and Centre-Left party vote choice, the simultaneous change in evaluations of both character and competence had a larger impact on the probability of vote among the most informed. For example, among the highly informed, those giving the

Conservative leader a rating one standard deviation above the mean on both character and competence were nearly 70% more likelyto vote for the Conservative party than those giving the leader a rating below the mean. This is compared to a change in the probability of vote for the Conservative party of 62% among the least informed. The pattern holds across the board. Overall, traits affect vote choice more among the more informed.

Among all groups, character still has a greater impact on vote choice than competence, much as it does when we look at the sample as a whole. However, there are

119 subtle differences between sophistication groups as to the extent to which the traits

influence vote choice: evaluations of the Conservative leaders’ character have a greater

impact among the less sophisticated, while evaluations of the Centre-Left leaders’ character

have a greater impact among the more sophisticated. Evaluations of the Conservative

leaders’ competence have a greater impact among the more sophisticated, while evaluations

of the Centre-Left leaders’ competence have a greater impact among the less sophisticated.

Thus it seems that the partisan stereotype identified earlier, in which Conservative leaders are evaluated more positively in terms of competence, and Left leaders are evaluated

more positively in terms of character, with Centre-Left leaders somewhere in between the

two (thus Centre-Left leaders are viewed more positively on character than are Conservative leaders), actually gets translated into an associated effect on vote choice, amongthemost

informed.The most informed are not only more likely to evaluate leaders according to these partisan stereotypes, but they are also more likely to translate these partisan stereotypes into

the voting decision. While character has a greater influence on vote choice than competence

for all voters, the more sophisticated are more influenced by the Conservative leaders’ competence than are the less sophisticated, and are also more influenced by the Centre-Left leaders’ character than are the less 25sophisticated. Although all respondents appear to think in terms of the partisan stereotype, it is the most sophisticated that really apply it to vote choice. Among the less sophisticated, however, leader evaluations still matter a great deal. It appears that the less informed are also making

25 Similar analyses were run with vote for the Left Party as the dependent variable, to ascertain the effects of evaluations of leaders on the Left vote, but half of the values were not statistically significant, leaving it difficult to find any patterns in the data. It seems that the main factor influencing the Left vote is identification with the Left party.

120 note of leaders’ personalities when deciding which party to vote for, although perhaps not in

as sophisticated a fashion as those voters with higher levels of political sophistication.

Can Individual Leaders Actually Make a Difference? The Net Impact of Traits on Vote Choice: US Case Study

The results and observations thus far are based on comparative, pooled data. These

data help us to determine the extent to which patterns may apply generally, but when we

want to think about how much leaders matter in specific elections, we cannot really

extrapolate from the larger picture. In order to examine the “real” impact of party leaders on

elections, then, it makes sense to look at specific leaders in specific elections.

In order to examine the role of individual leaders frirther, I ran the same logit

analyses as those performed above, for each of the American elections incorporated into this

study (1972-2004), using the National Election Study data. The dependent variable was vote

for the Republican candidate, coded where 1 was a vote for the Republican and 0 was a vote

for anybody else (usuallythe Democratic candidate, but sometimes also a third candidate). I chose this setup over a pure Republican/Democrat choice to maintain continuity with the

setup used for the cross-national sample as a whole. Table 7.10 reports the results of the analyses, presenting only the coefficients for trait 26variables. As the table clearlyillustrates, character is generally more important than competence. Of 18 character evaluations (one for each party, in nine elections), twelve had larger coefficients than the associated competence evaluation. Thus the cross-national patterns hold for specific elections as well.

Although the remainder of the variables (socio-demographics, partisanship, ideology and issue attitudes) are not shown, they were included in the model and their impact on vote choice was broadly the same as that seen in the model for the cross-national sample as a whole: socio-demographics had a minimal effect, issue attitudes had minimal effects, and partisanship played an important role, although the PID coefficients were smaller than those of trait evaluations.

121 Figure 7.1 depicts the mean trait evaluations for each party leader in each of the 9

election years in the analysis. 1972 saw the largest gap between the Republican and

Democrat candidates, with voters overwhelmingly viewing Richard Nixon more positively

on both character and competence than his Democratic competitor, George McGovern.

Also notable in this graph is the distance between character and competence lines for each

of the two parties: voters perceive the same leader differently on each trait dimension,

suggesting again that they are indeed capable of differentiating between different personality

traits, and they aren’t merely giving a particular leader either a positive or negative evaluation

on both traits.

These differences in trait evaluations translate into election results. In order to calculate the net effect of candidates’ traits in each of the elections, the average character and competence ratings were calculated for each year, thus establishing a fictional “neutral” or baseline candidate. Then, using the results of the logit analyses presented in Table 7.10, simulations were run to determine the change in probability of voting for the Republican candidate based on the mean ratings each of the two leaders were actuallygiven in each election—essentially, comparing the real leader to the fictional 27baseline. Figure 7.2 tracks the results, illustrating the Republican candidate net gain from evaluations of the Republican candidate’s traits in each year, from evaluations of the Democratic candidate’s traits in each year, and the sum of the two, the total net effect of candidates’ traits. Where the lines dip below the horizontal zero line, the Republican candidate lost ground as a result of evaluations of the party leaders.

27 See Johnston (2002) for a similar setup.

122 The net effect of candidates’ personalities differs in each election, ranging from just

over 0 in 1976 to 10% in 1972. These results are on par with conclusions made by scholars in the past, who have made attempts to calculate the net effect of leaders in elections in the

US (Bartels 2002b), Britain (Graetz and McAllister I987a), and Canada (Johnston 2002). Did leaders play a decisive role in any of the nine US elections? Not usually.In 1972,where the net gain for the Republican candidate as a result of positive evaluations of Nixon and negative evaluations of McGovern was highest at 10%, the Republican candidate received

23.2% more of the popular vote than the Democratic candidate. Had the margin been closer, personality may have mattered more, but in this election, it did not play a decisive role. In elections where the margin was small and candidates may have played a more decisive role, such as 1976, where the Democratic candidate received 2.1% more of the popular vote; and in 2004, where the Republican candidate received 2.4% more of the popular vote, the net effect of candidates’ traits was less than 1%. In both years, evaluations of candidates’ traits were quite closely matched.

These results do suggest, however, that evaluations of leaders maj have been decisive in one (infamous) election where the margin was close, and the Republican advantage was actuallya net loss rather than a gain: the Bush/Gore election of 2000. In this election, Gore won the popular vote by 0.5%, and as a result of evaluations of the two leaders’ personalities, the Republican candidate experienced a net loss of approximately 4.5% of the vote, as evaluations of Gore’s character and competence were both more positive than Bush’s ratings. Thus had it not been for Gore’s positive evaluations, Bush may have won the election more decisively (not simply in the electoral college).

123 I equivocate, however, and say that leaders only mqy have been decisive for a number

of reasons. First, is the fact that actually, Bush became President, not Gore, so in the end,

Gore’s “winning” personality did not win him the presidency. Second, the means in my

analysis are based on the 2000 NES sample as a whole, and do not track evaluations over

time. Johnston et al.’s (2004) close examination of evaluations of the two leaders’

personalities over the course of the election campaign suggests that in fact, Gore lost a lot of

ground, especially on ratings of his character, as the campaign progressed. Indeed, had he

managed to uphold the positive perceptions that the public had in the early days, leaders

personalities and the campaign may have been more decisive (and in his favour). Third,

Bartels’ (2002) analysis of the 1980-2000 elections suggests that in fact, candidates’

personalities only had a net effect of 0.4% in the 2000 election, suggesting that personality,

while it played a role, was not as decisive as my estimate of 4.5% suggests it was. He used a

slightly different method in his calculation of the net effects, which may provide some

explanation as to why our conclusions are different in this election 28year. 29

Conclusions

Leaders affect election outcomes. The data presented in this chapter suggest that voters’ evaluations of leaders’ personalities play a major role in influencing vote choice.

28 Bartels ran a probit model to estimate the probability of voting for the Republican candidate (versus voting for the Democratic candidate—votes for other candidates were excluded) in each election. He then compared these values with counterfactual probabilities in which the candidates were viewed equally favourably on all of the traits, as seen from the perspective of what he labeled a “neutral observer”—that is, an individual stripped of partisan attachments, ideological leanings, and either positive or pessimistic views of the economy. He also incorporated a smaller selection of traits, using only those specific traits that were common to all of the election studies he examined. 29 Although, notably, my estimates of the net effects of leaders in three election yeats (1980, 1984, and 1988) were identical to Bartels’ (2002) findings, and my estimates in two election years (1992 and 1996) were more conservative than his (by 0.8 percentage points and 1.4 percentage points, respectively). Only in 2000 do we have very different estimates of the net effects of the party leaders.

124 Furthermore, the analyses show that voters’ perceptions of leaders play a role that is larger

than the effect of either socio-demographic characteristics or attitudes towards policy issue

areas, and are more on par with the effects of ideology and partisanship. That is, if we take

the simulations at face value, leaders play one of the most fundamental roles in determining

how we vote, whether we look at voters in just one country, or from seven different countries over 35 separate election studies.

I suspect, however, that the “real” effect of party leaders is more muted than the changes in predicted probabilities, as presented in Tables 7.4 through 7.6, suggest. The values in the tables suggest that evaluations of leaders’ traits have an effect on vote choice of a similar magnitude to that of partisanship. The size of that impact is based on a simulated change in ratings from one standard deviation below to one standard deviation above the mean trait rating: an unlikely jump in the real world. This explains the difference in the simulated effects in the pooled analysesversus the estimated effects for the actual

(American) elections: simulated effects in the pooled model led to a change in the probability of vote choice by as much as 70% while the US estimates of the net effects of leaders’ traits were less than ten percent. The net effects in the US calculations were based on substantially smaller changes in ratings of the candidates’ traits, changes that were nowhere near the size of changes in the pooled simulations.

The value of combining the two exercises is the illustration of what we might consider the “maximal” impact versus the more “real” effect of voters’ evaluations of party leaders. The first demonstrates the incredible impact leaders could have, if perceptions changed dramatically in an election, while the second reminds us that while dramatic changes may be unlikely,leaders traits can and do still have a discernible impact on the distribution of

125 votes in the real world. The analyses presented in this chapter suggest that leaders matter, and that under certain circumstances, they can matter a lot.

126 CHAPTER EIGHT: CONCLUSION

‘Vemocraycannotsucceedunlessthosewhoexpresstheirchoicearepreparedtochoosewise/y.The realsafeguardofdemocray,therefore,is education”- Franklin D. Roosevelt (Beilenson,1982,29)

From minimalist definitions of democracy (Schumpeter, 1943) to more elaborate

definitions involving complex conceptions of human rights and representation, there appears to be a wide variety of understandings of what it means to have a successful democracy.

Recently, Lau and Redllawskstated that “...citizens, voting for leaders who best represent

their views, and holding those leaders (or their political parties) accountable for their performance in office at the next election, make democracy work” (2006, 3). Their comment provides a useful starting point for a number of reasons. First, it hits home that when heading to the ballot box, people are going to vote for somebodjto lead the country. At the end of the day there is an actual person who holds the title of President, Prime Minister, or

Chancellor. This suggests that the individual him or herself ought to be important in the vote calculus. Second, their comment makes note of the fact that in a working democracy, citizens should be voting for leaders who bestrepresenttheirviews.This assumes that individuals ought to be able to sort out who it is that best represents their views.

In fact, as scholars have noted, very few citizens have consciously sorted out what their views actually are: “The major chord in the analysis of public opinion, endlessly repeated, is how little attention they pay to politics, how rarely they think about even major issues, and how often they have failed to work through a consistent or genuine position on them” (Sniderman, Brody, and Tetlock 1991, 1). The assumption is therefore that those who are better educated or who have more knowledge about politics ought to be able to make a

127 better assessment of the extent to which political parties and leaders best represent their

views. Furthermore, the assumption is often made that those who are more informed will

base their assessments on “better” criteria—such as policy platforms and issue attitudes—

than those who are less informed.

Fenno suggests that indeed, policy concerns are often considered to be superior to

other aspects of representation, especiallyby normative theorists. He states that “the

traditional focus of political scientists on the policy aspects of representation is probably

related to the traditional focus on activity in the legislature. So long as concentration is on what happens in Washington, it is natural that policymaking will be thought of as the main

activity of the legislature and representation will be evaluated in policy terms. . . The

extrapolicy aspects of representational relationships have tended to be dismissed as

symbolic—as somehow less substantial than the relationship embodied in a roll call vote in

Washington—because what goes on at home has not been observed” (1978, 243). Within the community that Fenno describes, the notion that an individual voter would decide in

favour of a particular party or leader without primary reference to policy platforms is particularly problematic.

Many lament the modern media focus on the “horserace” during election campaigns, for drawing the attention of voters away from policy and toward “inferior” concerns such as party leaders. The media isn’t solely responsible for the focus away from discussions of policy: the leaders themselves do it (Savoie 1999). In Canada, former Prime Minister Kim

Campbell will forever be remembered for her suggestion (during the 1993 Canadian election) that election campaigns were “no time to discuss serious issues.” If the media does not focus heavily on policies and issues, and if party leaders themselves shy away from a serious policy

128 discussion, then how are individuals, regardless of their levels of education or political sophistication, able to ascertain who best represents their views?

Popkin (1991) suggests that the leaders themselves can in fact transmit a lot of information to the voter. At the beginning of his book on citizen reasoning processes, he describes what many saw as a major political gaffe: Gerald Ford tried to eat a tamale with the husk still on. Popkin argues that it would be reasonable for Mexican-Americans to glean from this that Ford lacked sensitivityand understanding of their culture and interests. A man who did not know how to eat a tamale was unlikely to prioritize their needs once in office.

This observation raises an important question about how voters conceptualize and understand representation, and how they can best determine if a leader is the best person to represent them.

While some, such as the normative theorists Fenno refers to, might argue that focusing on party leaders is a lamentable act that illustrates a lack of awareness or political sophistication, others have argued that in fact, leaders might actuallyprovide a complex source of information that the most educated and most politicallyaware segments of the population are best able to tap into (Glass, 1985). Glass’ suggestion goes beyond even that of scholars who suggest that party leaders may provide information that the least educated and sophisticated groups can use to compensate for a lack of deeper, policy-related information, in the fashion generally articulated by the information shortcuts and heuristics literature

(Cutler 2002; Fiske 1986;Hamill, Lodge and Blake 1985;McDermott 1997; McDermott

1998; Sniderman and Brady 1985; Sniderman et al. 1991).Taken to its fullest, his suggestion actuallyimplies that voters would do well to factor party leaders into the vote calculus,

129 because leaders can provide a rich source of information that will help them to do a better

job of determining who will best represent them.

The fact of the matter is, the scholarly literature on the role of party leaders is

undecided as to whether or not party leaders actual/ymatterin thefirstplace,never mind

agreeing on who it is that evaluates leaders, why, and what it might mean for our normative

understandings of democratic processes. While the media focus might suggest that leaders

matter, and the leaders’ own preoccupations with image, style, and fashion might also

suggest that leaders matter, academics are less clear on exactlywhat is happening. The

scholarly literature on party leaders is not very big, nor is it by any means conclusive. I have

argued that the lack of agreement in the literature stems from the fact that most conclusions

have been based on the analysis of a single election study in a single country, and that

because election studies over time and around the world have asked different sorts of

questions, scholars have drawn different conclusions from the analysis of different variables.

In this study, I have built upon and extended the work of previous scholars, by

conducting an analysis of voters’ evaluations of party leaders on a much larger scale that

what has been done in the past. This study has looked systematicallyand comparatively at

three main strands in the existing literature: a) traits and their underlying structure or

dimensions; b) the factors that influence evaluations of party leaders; and c) the impact of

those evaluations on vote choice.

By conducting a cross-national and longitudinal analysis of evaluations of leaders, using data collected in the election studies of a number of different countries, it has been possible to make a number of observations about voters’ perceptions of party leaders on a more general level. The conclusions provide us with a substantial new insight into the nature

130 and importance of trait evaluations, in a way that we have not had in the past, because of the

three-pronged nature of the analysis. Studies in the past have generally focused on single

aspects of leader evaluations, whether traits, origins, impact, or something else, while this

study has taken a more holistic approach, examining all three.

Five main conclusions are worth highlighting. First is the issue of trait structure:

voters evaluate leaders traits in terms of two main dimensions, characterand competence.

Specific traits fall within these two categories, and these categories can then be examined in

order to inform our understanding of both the origins and effects of leader evaluations.

Second, partisanship and ideology have a substantial effect on voters’ perceptions of party

leaders, whereas issue attitudes do not. Third, voters perceive leaders within the framework

of apartisan stereoyp6-.that is, the party label of the leader imbues meaning, in which voters

see Conservative leaders as possessing strength on the competence dimension, and Left leaders as particularly strong on the character dimension. Fourth, and perhaps most importantly for putting one of the major debates to bed, leadersmatter. 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. The fifth and final conclusion gets at the issue of political sophistication and decision-making: this dissertation has shown that there are noticeable and consistent differences in the perception of party leaders based on an individual’slevel of political sophistication. While all voters evaluate party leaders and consider leaders in the vote calculus, themoresophisticateddoit most.The sum total of these differences among the least and most sophisticated suggests that the evaluation of leaders is not necessarily a lamentable process, and that it is not simply a tool to assist the least educated and least politically aware in the decision-making process.

131 VotersPerceiveLeadersTraitsin TermsofCharacterand Competence

Scholars have assessed the dimensionality of traits in a number of ways over the

years, generally looking at either open-ended or closed-ended traits questions, and generating

dimensions either through the use of factor analysis or by creating intuition-based typologies.

As a result of the diversity of methods, question-formats, election studies, and traits

themselves, a number of different dimensions have been articulated over the years, ranging

from as few as two to as many as twelve.

By pooling data in a common format, it is clear that traits falllargelyinto two

categories or dimensions: character,generally including traits such as honesty and compassion; and competence,including traits such as strength of leadership and intelligence. These findings are broadly consistent with the bulk of findings in the literature, which have converged on the existence of between two and four dimensions. The analysis conducted in this study suggests that if scholars focus on incorporating traits from these two dimensions into the design of election studies, in keeping broadly with the recommendations made by Kinder

(1983) to the NES, then we will continue to have a wealth of data available to evaluate the evolution of perceptions of leaders over time and across space. The existence of two dimensions allows us to evaluate the origins of voters’ evaluations of traits within the dimensions, as well as the effects of character and competence on vote choice.

132 PartisanshipandIdeologyAffectVoters’ImpressionsofLeaders,IssuesandDemographicsDo Not

In fact, the identification of the two dimensions facilitated a cross-national and

longitudinal study of the origins of traits that would not otherwise have been possible.

Different election studies have incorporated different traits, but even where traits vary, the

dimensions they fit into remain largelythe same, allowing us to examine them comparatively.

This comparative analysisilluminated a hierarchy of sorts in the types of factors that

influence voters’ evaluations of party leaders. That is, voters’ partisanship and ideological

views play a major role in influencing the way that they perceive party leaders, with a larger

impact on ratings of leaders’ character and competence than do either socio-demographic

variables or attitudes towards policy issues.

In fact, the substantively larger impact of partisanship and ideology holds true even

for those with higher levels of political sophistication—it isn’t simply that the less informed

are basing evaluations on their partisanship while the more informed are making more

careful considerations of public policy. All voters are influenced primarily by partisanship and ideology in their perceptions of party leaders.

VotersEvaluateLeadersin Relationto a PartisanStereo’ype

Observations regarding the role of political sophistication hold true for the other main factor influencing voters’ perceptions of party leaders as well: the partisan stereotype.Past research has suggested that the party label of the leader or candidate carries meaning, and can act as a cue or shortcut for those in low-information elections (Conover and Feldman

1989, McDermott 1997, McDermott 1998, Rahn 1993).The idea is that voters will think of

133 leaders as members of a partisan group, and will translate that stereotype into impressions of

leaders’issue and ideological stances, even without information about their actual positions.

The analyses in this study confirmed that indeed, voters do perceive leaders within

the framework of a partisan stereotype. Furthermore, this partisan stereotype is not isolated

to a particular country. Generally speaking, voters perceive Conservative leaders to have a

particular strength on the competence dimension, while character is a perceived strong suit

among Left party leaders. One of the more interesting findings in these analyses is that the

partisan stereotype does not appear to be a cue used only by the least informed. In fact, the

more politically sophisticated respondents were more likely to perceive leaders within the

partisan stereotype, thus raising important questions about the role of the stereotype in

decision-making processes.

LeadersMailer

In Chapter Seven, I established the important role played by impressions of character

and competence in the vote calculus. The data analysesindicated that evaluations of party

leaders were influential, and furthermore, that not only do voters incorporate evaluations of party leaders into the vote calculus, but that moresophisticatedvotersdoit more.Ratings of

character and competence were more influential in determining vote choice among the more

sophisticated than among the least sophisticated, indicating that leaders are not something

that only the least aware and least capable focus on.

This chapter also presented an assessment of the net effects of leaders’ evaluations on the distribution of the vote in elections, focusing on the US elections from 1972 to 2004.

The data analyses indicated that evaluations of leaders’ personalities had varying effects on

134 the distribution of votes, and accounted for as much as 10% of the popular vote in the 1972

election. The data indicate therefore, that leaders canhave a decisive effect, where the

difference in perceptions of the party leaders is greater than the margin of victory in a

particular election.

PoliticalSophisticationMakesa Drence

That leaders matter in the vote calculus is one of the most important findings in this

study. That this finding emerged on a cross-national and longitudinal basis is even more

important, as it suggests that the findings are relativelyrobust, and generalizable beyond a

particular election or institutional setting. It puts to bed the on-going debates about the

extent to which evaluations of party leaders play a role in elections. Even more importantly,

the finding that individuals with higher levels of political sophistication incorporate

evaluations of leaders more heavily into their vote choice suggests that evaluations of leaders

are not simply a tool to help the least informed come to some sort of decision at the ballot

box. Leaders are not just a heuristic for those segments of society that lack the capacity to

make more “informed” decisions based on “higher” factors.

In fact, there are discernible differences throughout this study in the way that individuals perceive party leaders. These differences suggest that the more sophisticated

consider a greater number of factors overall; that they incorporate these factors more deeply into their decision-making processes; and that all of the “tools” that we assume must exist to help the less sophisticated actually tend to be used more heavily by the more sophisticated.

These findings indicate that we need to do a lot more research on the decision-making processes of individuals, at all levels of political sophistication. They suggest that some of

135 our assumptions about the impact of education and political knowledge need more

development, or that our assumptions about what constitutes “good” decision-making

criteria are not entirely accurate.

These findings relate to the general assumptions embedded in Roosevelt’s comment

above. His comment assumes that those individuals who are better educated, and more

prepared to “choose wisely”will base their choices on different criteria from those who are

less educated and less prepared. In fact, what this study has shown, is that while it is the case

that the decision-making processes differ among the least sophisticated and most

sophisticated segments of the population, the main difference is that the more sophisticated

rely more heavily on those tools that scholars have long considered to be aids for the less informed.

Roosevelt suggests that there is something about the education of the populace that will raise the ability of citizens to participate fully in the electoral process, and in so doing, will ensure the success of democracy. In fact, what appears to be happening is that with increased ability to participate fullyin the electoral process, voters are focusing more heavily on the types of factors that normative scholars find deplorable for their understanding of a healthy democracy. That personality should matter more than platform in determining the vote choice of not only the least informed segments of society, but especiallythe most informed, supports one of two possible alternatives: first, that even our most informed, most sophisticated members of society are not realljcapable of making an informed decision (see

Kuklinski and Quirk (2001) for the point that even the best informed are not very well informed); or second, that actually,generating impressions about future performance in office from the personality traits of party leaders is a difficult process, and a process that can

136 lead to greater insight than other types of information (such as policy platforms). Scholars

have suggested that individuals generate opinions and impressions about people they

encounter every day, therefore evaluating party leaders ought to be an easy process, since it

mirrors an activity that we perform on a regular basis (Cottrell et al. 2006, Rahn et al. 1990).

The question is: with limited cognitive capacities and resources (liketime), do we ever really

get enough information to judge political leaders accurately?Maybe it is only the most

sophisticated who actually gather enough information to translate their impressions of party

leaders into indicators of performance potential. They rely on these impressions most,

because these impressions actually mean something. A person cannot usuallychange their

personality, even though circumstances and policy stances may change, therefore their

personality might constitute a good indicator of how the individual will deal with the

challenges and stresses of political office.

What has clearly emerged from this study, however, is that leaders have an important influence on vote choice, across countries and over time. Furthermore, the evaluation of party leaders is not only a heuristic for those segments of society who lack the ability and information to consider other, “higher” factors. They may assist those individuals in deciding which party to vote for, but they also feature heavilyin the vote calculus of the most

sophisticated segments of the population. For allvoters, leaders matter.

137 TABLES AND FIGURES

Table 2.1 Surveys Including Questions about Party Lenders Country Year Australia 1987, 1990, 1993, 1996, 1998, 2001, 2004 Britain 1974, 1978, 1983, 1987, 1992, 1997, 2001, 2005 Canada 1963, 1968, 1972, 1974, 1979, 1980, 1984, 1988, 1993, 1997, 2000, 2004, 2006 Finland 1991 France 1967, 1968, 1988, 1997 Germany 1969, 1972, 1976, 1980, 1987, 1990, 1998 Hong Kong 1998, 2000 Israel 1969, 1977, 1981, 1984, 1988, 1992, 1996, 1999, 2001, 2003 Italy 1968, 1972 Japan 1967 Netherlands 1967, 1971, 1981, 1983, 1989, 1994, 1998 New Zealand 1993, 1996, 1999, 2(X(1, 20(12 Sweden 1956, 1960, 1964, 1968, 1979, 1985, 1988, 1991 Czech Republic 1996 Russia 1995-96

US - NES 1952, 1956, 1960, 1964, 1968, 1972, 1976, 1980, 1984, 1988, 1992, 1996, 2000, 2004

US - Annenhcrg 2000, 2004

00 Table 2.2 Frequency ot Question Types Asked About Leaders in All Surveys Type of Question Pl)xirna Frequency Feelings Thermometers 1(11 Closed—ended Traits 67 Knowledge 58 Open-ended Traits 42

Ability in office/ability to tackle specific tasks (future) 39 Performance evaluation (past 37

Preference for PM/Prez/Chanccllor (favourite candidate) 3()

Most important factor in vote choice (party/leader/candidate/issues) 24

Assessments of issue attitudes of leaders 23

iffect towards leaders (specific feelings - angry, etc,, not just overall feelings) 16

Where would place leader on left/right scale or liberal/conservative scale 13 Other (variety: questions specific to country, ear, specific event, etc.) 229 Table 2.3 National Election Studies Incorporating Traits Questions Number of Studies Incorporating Traits National Election Study

7 Australian Election S tLldy 6 British Election Stud 11 Canadian Election Study 7 German Election Study 6 Israeli Election Study 2 Dutch Election Study 2 New Zealand Election Study 2 Swedish Election Study I Russian Election Study

14 US - National Election Study

2 US - National Anncnber Election Study Table 2.4 Closed-Ended Traits Included in Surveys Across Space and Time Root Trait Traits Countries/years able to stand up to Quebec demands Canada 1968 adventurer (daring)/balanced and sensible Israel 1981 aggressive Canada 1993 Australia 1996, Canada 1968, Canada 1984, Canada 1993, Canada 1997, Canada 2000, Germany 1980, Germany 1987, NZ 1999, NZ 2002, Annenberg arrogant 2004 can really speak for women Canada 1993, NZ 1999 capable of solving English-French problems in Canada Canada 1968 Australia 1987, Britain 1983, Britain caring 1987, Britain 1992 charming Canada 1968 commands respect Canada 1984, US 1984 pro-communist Canada 1968 Australia 1993, Australia 1996, Australia 1998, Australia 2001, Australia 2004, Canada 1988, Canada 1993, Canada 1997, Canada 2000, Netherlands 1983, NZ 1999, US 1984, US 1988, US 1992, compassionate US 1996 competent competent Canada 2004, Canada 2006 competent leader Britain 2005, Canada 1984 willing to compromise Netherlands 1983 conservative conservative/modern Germany 1980, Germany 1987 views are too conservative Canada 1968 too conservative Annenberg 2004 credible in TV appearances/not credible in TV appearances Israel 1981 would control crime US 1972 Australia 1993, Canada 1984, US 1984, decent US 1988 Australia 1987, Britain 1983, Britain 1997, Israel 1981, Israel 1992, Israel 1996, Israel 1999, Netherlands 1983, decisive decisive Annenberg 2004 Root Trait Traits Countries/years Australia 1993, Australia 1996, Australia dependable 1998 determined Australia 1987, Britain 1983 diligent Germany 1998 dull Canada 1984 dynamic/hesitating Germany 1980, Germany 1987 experienced Israel 1996, 1999 extreme extreme Canada 2000 extreme/moderate Britain 1987, Britain 1992 Stephen Harper is just too extreme Canada 2004, Canada 2006 fair fair US 1984 fair-minded Canada 1968 fresh tired/fresh Israel 1981 young leader/old leader Israel 1981 a leader whose time has passed/a leader whose time has not passed Israel 1981 his health does not permit him to continue in his role/his health permits him to continue in his role Israel 1981 represents change Canada 1984 having new ideas Canada 2000 gets things done gets things done US 1992, US 1996 good at getting things done/bad at getting things done Britain 1987 effective & gets things done Annenberg 2004 initiates and moves things Israel 1992, Israel 1996, Israel 1999 hardworking US 1984 is helpful to his country/is damaging to the country Israel 1981 Australia 1996, Australia 1998, Australia 2001, Australia 2004, Canada 1968, Canada 2004, Canada 2006, Israel 1981, US 1988, US 1992, US 1996, Annenberg honest honest 2000, Annenberg 2004 is an honest and trustworthy person Russia 1995/96 Canada 2000, US 1980, US 2000, US dishonest 2004 Root Trait Traits Countries/years hypocritical Annenberg 2000 ill-tempered, lacking control/calm, matter of fact Germany 1980, Germany 1987 Australia 1993, Australia 1996, Australia 1998, Australia 2001, Australia 2004, Sweden 1988, Sweden 1991, US 1980, US 1984, US 1988, US 1992, US 1996, inspiring Annenberg 2000, Annenberg 2004 in touch with the times Canada 1997 Australia 1993, Australia 1996, Australia 1998, Australia 2001, Australia 2004, Canada 1984, Canada 1993 Canada 2000, US 1984, US 1988, US 1992, US intelligent intelligent 1996, US 2000, US 2004 highly intelligent Canada 1968 is an intelligent and knowledgeable person Russia 1995/96 man of great integrity Canada 1968 kind US 1984 Australia 1993, Australia 1996, Australia 1998, Australia 2001, Australia 2004, Canada 1988, Sweden 1988, Sweden 1991, US 1980, US 1984, US 1988, US 1992, US 1996, US 2000, US 2004, knowledgeable Annenberg 2000, Annenberg 2004 too liberal Annenberg 2004 Australia 1987, Britain 1983, Britain likeable likeable as a person 1987 he’s easy to like as a person Annenberg 2004 listens to reason Australia 1987, Britain 1983 given to moods/not given to moods Israel 1981 Australia 1993, Australia 1996, Australia 1998, Canada 1988, US 1980, US 1984, US 1988, US 1992, US 1996, US 2000, moral moral US 2004 which candidate best reflects high moral or religious standards US 1972 would bring moral and religious standards to government US 1976 has adequate concern for public moral standards Canada 1968 Root Trait Traits Countries/years nervous Canada 1984 Israel 1992, Israel 1996, Israel 1999, good negotiator Israel 2003 optimistic Annenberg 2004 Israel 1992, Israel 1996, Israel 1999, places benefit of country before the party Israel 2001, Israel 2003 power-hungry Paul Martin only cares about staying in power Canada 2006 power-hungry US 1980 progressive Canada 1968 keeps someone who keeps his promises/breaks his promises promises Britain 2001 too quick to make promises Canada 1968 reckless Annenberg 2004 Australia 1993, Australia 1996, Australia 1998, Germany 1998, Israel 1992, Israel 1996, Israel 1999, Israel 2001, Israel 2003, Netherlands 1983, Sweden 1988, reliable Sweden 1991 religious US 1984 Germany 1980, Germany 1987, responsible/irresponsible Germany 1998 responsive to ‘ordinary’ values shares my values Annenberg 2004 listens to the views of people in (name province) Canada 1984 Paul Martin only cares about big business Canada 2004 jack Layton only cares about minorities Canada 2004 in touch with ordinary people US 1984 looks after one class/looks after all classes Britain 1987, Britain 1992 understands people like you US 1984 Out of touch with ordinary people US 2000 Out of touch with people like me Annenberg 2004 cares about people like me Annenberg 2004 Canada 1988, Russia 1995, US 1984, US 1988, US 1992, US 1996, US 2000, US really cares about people like you 2004 knows the thoughts and opinions of ordinary people Sweden 1988, Sweden 1991 Root Trait Traits Countries/years responds to voters concerns Britain 2005 too rigid Canada 1968 ‘right’ leadership better suited for provincial than for federal skills politics Canada 1968 a bad Prime Minister/an excellent Prime Minister Israel 1981 a bad Defense Minister/an excellent Defense Minister Israel 1981 has the kind of personality a President ought to have US 1972, US 1976 has the right kind of experience to be President Annenberg 2004 his place is in the opposition/his place is in government Israel 1981 ruthless Canada 1984 Australia 1993, Australia 1996, Australia sensible 1998, Australia 2001, Australia 2004 sets a good example US 1984 shallow Canada 1984 shrewd Australia 1987, Britain 1983 sincere Australia 1987, Canada 1984 slick slick/straight Israel 1981 slick Canada 1984 manner is too slow Canada 1968 too soft on French Canada Canada 1968 stable/unstable Israel 1981 steady Annenberg 2004 can stand stress/cannot stand stress Israel 1981 sticks to Australia 1987, Britain 1983, Britain principles sticks to principles 1997 says one thing but does another Annenberg 2004 someone who will make the tough decisions despite political pressure Annenberg 2004 Israel 1992, israel 1996, Israel 1999, can stand up under pressure Israel 2001 changes his mind for political reasons Annenberg 2004 flip-flops on the issues Annenberg 2004 Root Trait Traits Countries/years not willing to admit when he makes a mistake Annenberg 2004 strong leader strong leader Canada 1997, Russia 1995, Annen 2004 Australia 1993, Australia 1996, Australia 1998, Australia 2001, Australia 2004, Canada 1988, Canada 1993, NZ 1999, NZ 2002, US 1984, US 1988, US 1992, US 1996, US 2000, US 2004, Annenberg provides strong leadership 2000 strong/weak Israel 1981, Israel 1992 Israel 1992, Israel 1996, Israel 1999, has a leadership characteristic Israel 2003 capable of being a strong leader/not capable of Britain 1987, Britain 1992, Britain 1997, being a strong leader Britain 2001 a weak leader Canada 2000 weak US 1980 stubborn Annenberg 2004 sympathetic Germany 1998 sure of himself Canada 1984 can work on a team/cannot work on a team Israel 1981 tough tough Australia 1987, Britain 1983 tough-minded Canada 1968 Australia 2001, Australia 2004, Britain 2005, Canada 1988, Canada 1993, Canada 1997, Canada 2000. Germany 1980, Germany 1987, NZ 1999, NZ trustworthy trustworthy 2002, Annenberg 2000, Annenberg 2004 could be trusted US 1972, US 1976 Jack Layton cannot be trusted Canada 2006 Giles Duceppe cannot be trusted Canada 2006 likely to unite the nation/divide the nation Britain 1987 he has a clear vision of where he wants to lead has vision the country Annenberg 2004 has his own vision of the country’s future Russia 1995/96 man of vision Canada 1988 under Dalton Camp’s influence Canada 1968 warm Canada 1984 weighs his words carefully/does not weigh his words carefully Israel 1981 Root Trait Traits Countries/years will know how to fight terror Israel 2001 will negotiate decisively in the negotiation on peace and terror will lead to real peace with Arabs Israel 2001 will reduce the division within Israeli society Israel 2001 would reduce the tension between religious and seculars Israel 1999 would preserve the rule of law Israel 1999, Israel 2001, Israel 2003 would reduce social gaps Israel 1999 would bring peace in Vietnam US 1972 Figure 3.1: Categorization of Party Types, Based on Benoit and Layer’s (2006) Partj Policji in Modern Democracies

SWNDP NZ First SW Christian Can Reform AU Nat US Rep AU One CDU!CSU UK Cons AU Lib Can PC

SW Mod

Social Liberalism

NZ Nat NZ ACT UK SNj__ Centre-Left Can Lib Conservative UK Plaid UK Lab NZ Lab Left Lab CEJ Right SW CY Ger SPD USDem__— Centre-Right Sectional SW Green Green NZ4JI Can breen.an oc ‘u Lioem SW Left Peoples Ger FDP Can NDP, AU Dem AU Green Ger Green

Taxes vs Spending Table 3.1 Major Election Issues Country/Year Major Election Issue Country/Year Major Election Issue Canada Britain 1968 Quebec separatism 1983 Economy 1984 Patronage & corruption 1987 Defense 1988 Free Trade 1992 I-lealthcare 1993 Debt and Deficit 1997 1-lealthcare 1997 Quebec 2001 Currency 2000 Quebec Australia 2004 Party Financing 1987 Economy/Recession 2006 Party Financing 1993 GST, Economy United States 1996 Economy, Recession 1972 Vietnam War 1998 GST 1976 Watergate, Pardon of Nixon 2001 Immigration, Asylum Seekers 1980 Iran 1-lostages 2004 Terrorism, Defense Spending 1984 Nuclear \Vcapons, Defense New Zealand 1988 Willie Horton 1999 Law and Order 1992 Economy 2002 Healthcare, genetic engineering 1996 Economy (;errnan 2000 Moral Decay of Society, Clinton Scandals 198(1 Economy, Recession, jobs 2004 Iraq, War on Terror, Defense 1987 Economy, Recession, jobs Sweden 1988 Large inefficient government 1991 Large inefficient government Table 3.2 Political Sophistication: Measures Used to Construct Index General Knowledge Intcrviewct Exposure Interest Index Rating to Media CA 1968 X CA 1984 X CA 1988 X CA 1993 X CA 1997 X CA 2001) X CA 2004 X CA 2006 X US 1972-2004 X US 2000 (a) X UK 1983 X UK 1987 X UK 1992 X UK 1997 X UK 2001 X AU 1987 X AU 1993 X AU 1996 X AU 1998 X AU 2001 X AU 2004 X NZ1999 X NZ 2002 X GE 1980 X GE 1987 X SW 1988 X SW 1991 X Table 3.3 Political Sophistication Index: Multivariate Analysis

General interest 0.004 (0.003) interviewer Rating 0.016 0.002) Media Exposure 0.210 (0.004) Constant 0.540 0.002) Observations 156301 R-s9uarcd 0.023 Ordinary Least S.1uarcs Regression Analysis Table 3.4 Relationship Between Demographics and Sophistication University Graduate 0.187 0.004) Income 0.064 (0.002) Woman -0.074 (0.003) Non-\Vhitc -0.041 (0.004) I lomcowncr 0.002 (0.004) Age 0,02 (0.001) Employed 0.017 (0.004) Constant 0.339 (0.008) Observations 40835 R-squared 0.12 Ordinary Least Squares Regression Analysis Table 3.5 Calculating Weights According to Total Population A B C D B Population Proportion of Total Sample in ES Proportion of total Weight Population ES sample (B/I)) NZ 3,331,000 0.007649962 11,755 0.0630 17326 0.121394582 AU 16,320,000 0.037939769 12,326 0.066078398 0.574162976 SW 8,469,001) 0.0 19449873 7,626 (1(140882189 (1.475754206 (;E 78,031,000 0.179205699 26,361 0.141318566 1.268097353 UK 57,160,000 0.13127344 22,749 0.121955011 1.076408736 (;A 26,895,00() 0.061766955 29,264 0.156881245 0.3937179(14 US 245,021,000 0.562714301 18,082 0.096935712 5.805025709 58,373 0.312931552 1.7982025(16 Total 435,427,00() 1 186,536 * Population Numhcrs from the ear 1988, since 1987/1988 had the largest number of Election Studies, and since it represents the mid-point in the timespan of the election studies in the analysis (1968—2006) ** Data obtained from http://gapminder.org, a non-profit organization that works closely with the UN and Google to promote achievement of United Nations Millennium 1)cvelopmcnt Goals Table 4.1 Evaluations of Three Main Party Leaders on Trait: Strength of Leadership Rating Conservative (entrc-Lcft Left 0 10,349 12,702 4,704 (1.25 346 735 277 0.33 14,965 16,808 3,088 (1.5 627 797 1,528 0.67 26,248 28,684 4,274 0.75 1,136 666 810 1 19,653 21,009 6,650 Number of Observations 73,324 81,401 21,331 Table 4.2 Top Ten “asked” Traits in Surveys: Number of Respondents Evaluating Specific Leaders’ Traits Conservative Centre-Left Lcfr N= Lcadcrship 73,324 81,401 21,331 Cares 83,325 80,890 19,027 knowlcdgeahle 77,059 47,892 4,877 Inspiring 71,247 69,531 2,887 Honest 70,855 69,744 8,409 Intelligent 33,124 34,968 11,928 Trustworthy 32,653 42,158 14,796 Compassionate 26,527 33,081 14,159 Arrogant 21,016 31,159 19,52% Moral 20,997 20,336 1,994 ((.23

3i,bIr 4..) C,mn,6a,onn ‘6 Ovahc,,inn for 19niof To,, Mom Pomp 1odnr,

2 3 .1 5 (0 H (2 LI (I (0 (6 I” 10 (9 .01 21 22 2.’ 2.1 25 26 r 1.1 29 .1(1 Il .1.1 .1.1

4662 I 1 (((((oF (654 (563 (66 Coinpc’tcncc’

- V V (48 ‘crOpI ln,e,4wnlcorreIalio,, nl343

- ,ko,olc ((Ill - . ((43 (44. I

S(I1idiIOc;O ((.5! 64(1 ((.45 (638 V V 9 (.l,llooltcnl’Ie ((96 4(4% (64 (655 V ((.6 I

III Cnfl3(c110( • 1156 ...... I

II tots lImo’ Inc (66 . . (154 049 (641 (145 ((.42 146 .

((42 VVVVV ((45 V

V I,andnndaoo (643 0.56 (64! ‘V (6411 (65 V V

((.54 V - VV -I 001(1(6. 019 0.69 I13’( V V V

.. (54 0.4 . . . (645 (62’ . . V V 1 Compownot and Cl,oratcr ‘4,041, ((o(n..sIroe(I (Sea)

(6 (1060(1 ((.240.3 0.29 . . . (12! A.nIaI1, I,,(er-((e,,,c,orc’InSinn 0.32

0’ ((.42 ...,.. 0.114 122 .

III I0fl(mII1(l(O(IlI (62.1 (1.29 ((II ... (665 ((46 . I

I’( p0101601(’.II V V .51 ((((1(I) ,(( V _ .

2) .11(1 (III . . . (‘((I

- V V 23 3(0010(1 (I_Il . . - 2.1 .I,ok . . . I Is - (104 Cl,.o..e,,

(1(1 (I_U’ - (6_’l(O.lO.lr (((S ((.11 (II, (103 . Ill . - (124 (1(116 (126 ((.12 649 I A,eroge In,er-iwn,eorrnl44i,,,, 00.41 (61,5 I

- .11 ‘(1-I . - - ((.20 (121 . ... 16(12 162 ((2) . 29 (161100% (I(S) ((00.12 . ((3.4 I 0(1091, (653 ((.51 . . (((-I .61 (1.41 ((01 ((VII (1.12 (I (I, - 03.1 61-I (.62 1.2 (6(2 . . . - (((.9 .0.11 . V

.12 -- . oil - Un V V (lii ..... I (130 (II (6(2 (LI_I (1(6 (II . (III , ((.23 .0.42 .11 (o(n’oo( II

. ((I (III (LIII ((35 - (U ((.30 V V

16 ((62 . . ((.32 ((.56 (-(0 . - V V V (635

.13 (Il1I(V01L’O ((3.I II 2’( , (6.1 - 111(0 , ‘(.1 I . 16.11 . ((.1(6 (64 ((43 V V V (644

, 62 . , (1.1 ((II (6211 (61’ (SM, 0.35 (623 . Ill - ((.6 (126 , , .11.10 ((II - or (63 (625 .. (III, (Ir , (643 46.32 (63% ((.2! ((((8 .16.15 —II (i—lo,,o,,c .0.18 39 (625

- . -(2 (III ((020.06 (1(11 - (62 110(1 1(3 , 0.06 (3 .

II ((.610, ((.35 1.-Il (I.1 . 11.0, (1(3 0.41 ((.22 (1550.62 . V ((.193 (1211 (62(1 (12 (6(63, (540

- - (1.10 II. ((5 - (150 (153 (131 . (16) ... - (6.31 ((SI .

.ISnn,oo.,nh, ((.35 - (65 ((36 (6-12 (6-I’) . (61,1 ((((1 (6.1,0 ((.5) ((.25 0 ..02 l1% (1% ((55 . ((S ,lqnooI,6Ic’ ((-(‘I . (((‘I - - . (I .0.2) V (L’S .

-IS roII,l6Io (U, (6 I.( . - (144 ((.3 ((SI V (623 (629

(‘1 1116 V (.06) ((2 - ‘(52 . (-(6 - 601

ILIO (630 o.r 05) - i-I ((‘III (((2 163,0601 - . lUll (or . . (623 . . . . (64.5 . .. 0.11 - .1(0 — - . ((.(.( . . ((61(1622 (6.29 . (62.1 (619 . . .

(VU 11-lI - - 65.1 ((So (13.( . -, V 33 61(61 (UK (1.—I (II’) ((.1,2 ((.06 (I_I ‘11 I 0.41 II. (‘1 ((2) ((.3-I . (60’(O..I.( (Ill ((.25 .03! III (61

. 3.1 ((56 (U (655 II. or - (‘I ((.3) . . V (l4 Ills (I_il . ((‘II (1.1 ((II (1(19 (U’l ((.10 . ((.12 . ((.5) . . . (OIl ((SI . . (122 . . . . (642 . (614 Table 4.3, continued Correlations of Evaluations for Traits of Leaders of Two Main Parties 34 35 35 37 30 39 -51 41 42 43 44 45 46 47 40 49 514 51 52 53 54 55 I strong leaderthip 2 comnsandr eespect 3 itsitnol vision 4 inspiring 5 determined 6 ntngh 7 decisive 9 intelligent 9 hot seledgealsle III competent II gets things dccc 42 ‘credit mit’ IS hordwethtng 14 sensible iSneab lb nbresvd 17 intlecisiee IS gets most oat itt a react 19 prntklentiol pctsttttalae 214 ds-namie 21 ideas 22 ctseethcngrv 25 aggressIve 24 sueS

25 nrtttgact 27 calm 25 doll 29 nerecais 3(1 respsinsiltle 31 mcml 32 keeps pntmises 33 principled 34 hypocritical I

36 otto ginnl esample 0.48 I

37 cot teustssstrtlsi - - I

30 dtslrstnest - - I 39 slsafltise 411 sincere 37 1 41 listenc vi reantn 0.49 I 42 litaacthle 0.42 11.41

. 43 mnttetf vueb 0.4 - -

44 decent 1144 0.37 3 15149 - 45 leanest 0.37 34 1161 46 rrasneortlsv (1.22 11.79

47 dependohle 1660 11.72 -

40 reliohte 11.7 0.77 - 0.8 49 bind 11.46 1550

59 conapotsionate 0.44 0.58 - 0.16 0.63 11.62 0.59 1664 11.64 0.66 51 ocrnt (125 11.94 1A - 52 fnie t548 14.63 1155 11.41 - . - 11.62 ((.66 -

53 codes 0.31 1146 1565 1523 4143 1151 1193 0.54 1134 (592 1561 0.65 .. 0.6 (163 - 1173

- 54 oneleestocalr 11.45 1162 5 - . . - 11.57 11.62 . (172 (103

- 55 in toceb 1143 1152 39 0.4 1137 - 1143 - (151 - 0.47 0.52 0.32 - 0.68 (146 0.75 Table 4.4 Correlations of Evaluations for Traits of Three Main Party Leaders 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 1 strong leadership 2 commands respect 0.62 1

3 man of vision 0.53 .

4 inspiring 0.63 0.6 . 1 Competence Traits

5 decisive (1.45 . . . Average Inter-item correlation = 0.47

6 intelligent 0.51 0.48 0.46 0.59 . 7 knowledgeable 0.56 ((45 0.46 0.55 0.66 1 Competence and Character Traits (non-shaded area)

8 competent 0.54 . . Average Inter-item correlation 0.30 9 sureofsclf 0.42 ((.43 1

10 sure of self 0.55 . . 0.6 0.6 0.6

11 weak (1.54 . . 0.4 (1.15 (1.27 Character Traits 12 ideas (1.09 0.03 1 Average Inter-item correlation = 0.35

13 aggressivc -0.2 . . -0.12

14 slick —0.08 . . . . . —0.05 —0.21

15 ruthless 0.09 . . . . . ((.11 —0.04 ((.39 1

16 arrogant (1.12 (1.15 . —0.07 ((.23 —0.04 —0.07 (1.16 —0 ((17 —0.1 ((.05 0.3 (1.35 0.48 1

17 dull ((.23 . . . . . (1.26 1)22 0.05 (1.2 ((.19 1

18 nervous 0.2 . . . 0.24 ((37 0.03 (1.14 (1.18 ((31 1

19 moral . (1.35 ((44 ((.41 ((.19 ((.32 ((.33 ((32 ((2 -0.31 . .

20 keeps promises ((.48 . . . ((.33 ((32

21 principled ((.37 . . . ((.41 ((.37 22 not trustworthy ((.25 (1.35

23 dishonest ((34 . . . ((.29 0.03 ((.3 ((.05 ((.05 . ((.42 . . ((.44 24 shallow (1.23 ((29 ((13 0.32 (1.37 (1.37 0.34 ((.26 25 sincere (1.45 (1.26 ((46 ((.29 (1.16 (1.31 0.36 0.2 0.13

26 listens to reason ((.33 . . . ((.22 . (1.42

27 decent ((.42 . . ((.39 ((.37 ((.45 (1.46 ((.4 ((.26 (1.62 . . (1.08 0.27 (1.28 0.18 0.09 (1.47

. . ((57 . 28 honest . . (1.58 0.55 ((.48 ((.68 ((62 (1.3 . . ((51

29 trustworthy ((.54 . ((.56 . ((.51 ((.44 ((.5 ((61 ((08 ((.24 0.03 (1.2 . . ((58

30 reliable . . . 045 ((.42 ((.48 ((.49 ((64 . . 0.23 . . (1.29

31 compassionate ((.38 . ((.39 (1.47 0.41 ((.4 ((.43 ((6 (1.02 ((.17 0.1 0.21 . . ((.45

32 warm (1.42 . . . . . 0.43 (1.34 0.02 0.23 ((.3 0.22 (1.19 33 cares ((.44 ((.4 ((.5 ((.62 ((24 ((.4 ((.5 ((.28 . . (1.51

34 intouch ((.45 ((.31 . (1.39 ((.4 ((.1 ((.35 0.21 . . (1.42 Table 4.4, cont’d - Correlations of Evaluations for Traits of Leaders of Three Main Parties 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 1 strong leadership 2 commands respect 3 man of vision 4 Inspiring 5 decisive 6 intelligent 7 knowledgeable 8 competent 9 sureofself 10 sure of self 11 weak 12 ideas 13 aggressive 14 slick 15 ruthless 16 arrogant 17 dull 18 nervous 19 moral 20 keeps promises 21 principled 0.43 1 22 not trustworthy I 23 dishonest 24 shallow 25 sincere 0.45 0.37 1 26 listens to reason 0.4 0.3 0.45 27 decent 0.3 0.58 28 honest 0.32 0.61 1 29 trustworthy 0.17 0.79 30 reliable 0.7 0.77 31 compassionate 0.19 (1.63 (1.62 0.59 (1.64 32 warm (1.25 0.54 0.45

33 cares 0.4 0.33 0.23 0.43 0.51 0.54 0.52 0.6 0.65 0.63 .

34 intouch 0.16 . 0.39 0.34 0.45 0.51 ((.47 0.52 . 0.4 Figure 4.1

Character and Competence Ratings of’ Leaders of Three Main Parties

Margaret Thaicher Pierre Triidciu • • I lulen C1ark • I .d Br( )idI)eflt Ja rgare I .1 • • Tiatcher (3) • • • Padd \shd vn . • • • • • •. • • • •.• . - • • Competence . . • • •:s.1.. Jean • (hret ten (I I( ) • • • \eil kinnnck • C)hfl . I Loward ($‘7 • • •

\lichael L()( )I

T 4 .6 C’ h arac t er

• Conservative (‘entre-Ldfi • Ldfi C F Table 4.5 Index Cohesion: Cronbach’s Alpha for each Election Study

Traits in Study Alpha for Character and Competence Competence Traits Character Traits

Australia 1987

Determined Cares Conservative Leader Character = 0.7963 Shrewd Likeable Conservative Leader Competence = 0.4748 Tough Listens to Reason Centre-Left Leader Character = 0.8333 Decisive Principled Centre-Left Leader Competence = 0.6054 Gets Things Done Sincere Right Leader Character 0.7106 Gets Most out of a Team In Touch Right Leader Competence = 0,5516

Australia 1993 Intelligent Moral Conservative Leader Character 0.8700 Sensible Compassionate Conservative Leader Competence = 0.8740 Strong Leadership Decent Centre-Left Leader Character = 0.9249 Knowledgeable Reliable Centre-Left Leader Competence = 0.9093 Inspiring Dependable

Australia 1996

Intelligent Moral Conservative Leader Character = 0.8302 Sensible Compassionate Conservative Leader Competence = 0.8811 Strong Leadership Honest Centre-Left Leader Character = 0.8324 Knowledgeable Reliable Centre-Left Leader Competence = 0.8865 Inspiring Dependable Arrogant

aN Traits in Study Alpha for Character and Competence Competence Traits Character Traits Australia 1998

Intelligent Moral Conservative Leader Character = 0.9207 Strong Leadership Compassionate Conservative Leader Competence 0.8773 Knowledgeable Sensible Centre-Left Leader Character 0.9182 inspiring Honest Centre-Left Leader Competence = 0.8925 Reliable Right Leader Character = 0.9258

Dependable Right Leader Competence = 0.8931

Australia 2001

Intelligent Compassionate Conservative Leader Character = 0.8983 Strong Leadership Sensible Conservative Leader Competence = 0.9101 Knowledgeable Honest Centre-Left Leader Character = 0.8728 Inspiring Trustworthy Centre-Left Leader Competence = 08643

Right Leader Character = 0.9010

Right Leader Competence = 0.9015

Left Leader Character = 0.9025

Left Leader Competence = 0.9066

Australia 2004

Intelligent Compassionate Conservative Leader Character = 0.9089 Strong Leadership Sensible Conservative Leader Competence 0.8937 Knowledgeable Honest Centre-Left Leader Character 0.8758 Inspiring Trustworthy Centre-Left Leader Competence = 0.8703

Britain 1983 Gets Things Done In Touch Conservative Leader Character 0.6475 Determined Cares Conservative Leader Competence 0.5008 Tough Shrewd Centre-Left Leader Character 0.6807 Decisive Likeable Centre-Left Leader Competence = 0.4808

Listens to Reason Left Leader Character = 0.6732

Principled Left Leader Competence = 0.5774 Traits in Study Alpha for Character and Competence Competence Traits Character Traits Britain 1987

Gets Things Done Cares Conservative Leader Character = 0.7034 Strong Leadership Likeable Conservative Leader Competence = 0.5471

Centre-Left Leader Character = 0.5693 Centre-Left Leader Competence 0.7283

Left Leader Character = 0.5581

Left Leader Competence = 0.6624

Britain 1992 Strong Leadership Cares

Britain 1997

Strong Leadership Cares Conservative Leader Character = 0.6107 Decisive Principled Conservative Leader Competence 0.6911

Centre-Left Leader Character = 0.6044

Centre-Left Leader Competence = 0.6831

Britain 2001

Strong Leadership Keeps Promises Conservative Leader Character = 0.7318 Decisive Arrogant Conservative Leader Competence = 0.4561 Cares Centre-Left Leader Character = 0.7814 Principled Centre-Left Leader Competence 0.5604 Listens to Reason Left Leader Character 0.73 12

Left Leader Competence = 0.4134

Sectional Leader Character = 0.7802 Sectional Leader Competence 0.5134 Sectional2 Leader Character 0.8008

Sectional2 Leader Competence = 0.5142 Traits in Study Alpha for Character and Competence Competence Traits Character Traits Canada 1968 Intelligent Fair Conservative Leader Character 0.4084

Honest Centre-Left Leader Character = 0.5900

Canada 1984

Competent Arrogant Conservative Leader Character = 0.7521 Commands Respect Ruthless Conservative Leader Competence = 0.6441 Sure of Self Nervous Centre-Left Leader Character = 0.7472 Decent Centre-Left Leader Competence 0.6762 Slick Left Leader Character 0.7313 Sincere Left Leader Competence = 0.6787 Shallow Dull Warm Canada 1988 Intelligent Trustworthy Conservative Leader Character 0.8410 Man of Vision Compassionate Conservative Leader Competence = 0.7784 Knowledgeable Moral Centre-Left Leader Character = 0.7988 Strong Leadership Cares Centre-Left Leader Competence = 0.7774 Left Leader Character 0.8057

Left Leader Competence = 0.7590 Traits in Study Alpha for Character and Competence Competence Traits Character Traits Canada 1993

Inteffigent Arrogant Conservative Leader Character = 0,6080 Strong Leadership Trustworthy Conservative Leader Competence = 0.6395 Compassionate Centre-Left Leader Character = 0.5644

Aggressive Centre-Left Leader Competence = 0.6435

Left Leader Character = 0.4931 Left Leader Competence 0.5328

Sectional Leader Character = 0,6707

Sectional Leader Competence = 0.6447

Right Leader Character = 0.6114

Right Leader Competence = 0.6704

Canada 1997 Strong Leadership Trustworthy Conservative Leader Character 0.6634 Arrogant Centre-Left Leader Character = 0.7184 Compassionate Left Leader Character 0.6351

In Touch Sectional Leader Character = 0.6255

Right Leader Character = 0.7302

Canada 2000

Man of Ideas Arrogant Conservative Leader Character = 0.3680 Intelligent Trustworthy Conservative Leader Competence = 0.2256 Weak Dishonest Centre-Left Leader Character = 0,6239 Compassionate Centre-Left Leader Competence 0.2423

Left Leader Character = 0.4089 Left Leader Competence 0.2219

Sectional Leader Character = 0.3298

Sectional Leader Competence = 0.1707

Right Leader Character = 0.5753

Right Leader Competence = 0.3119 Traits in Study Alpha for Character and Competence Competence Traits Character Traits Canada 2004

Intelligent Honest Conservative Leader Character = 0.5960 Arrogant Centre-Left Leader Character = 0.6237

Cares Left Leader Character = 0.5053

Not Trustworthy Sectional Leader Character = 0.179 1

Canada 2006 Competent Honest

Germany 1980

Dynamic Responsible Conservative Leader Character = 0.8141 Arrogant Centre-Left Leader Character = 0.8327 Trustworthy Calm Germany 1987

Dynamic Responsible Conservative Leader Character = 0.8750 Arrogant Centre-Left Leader Character = 0.8610 Trustworthy Calm New Zealand 1999 Strong Leadership Trustworthy Centre-Left Leader Character 0.7138 Arrogant Centre-Right Leader Character 0.8150

Compassionate Right Leader Character = 0.7636

Left Leader Character = 0.8001

New Zealand 2002

Strong Leadership Trustworthy Centre-Right Leader Character = 0.6276 Arrogant Centre-Left Leader Character 0.6627 Traits in Study Alpha for Character and Competence Competence Traits Character Traits Sweden 1988

Inspiring Reliable Conservative Leader Character = 0.6436 Knowledgeable In Touch Conservative Leader Competence 0.5455

Centre-Left Leader Character = 0.6292

Centre-Left Leader Competence = 0.5451

Centre-Right Leader Character = 0.5760

Centre-Right Leader Competence = 0.4568 Left Leader Character 0.6 123

Left Leader Competence = 0.5338

Sweden 1991

Inspiring Reliable Conservative Leader Character = 0.6236

Knowledgeable In Touch Conservative Leader Competence = 0.4767

Centre-Left Leader Character = 0.6032

Centre-Left Leader Competence = 0.5291

Right Leader Character = 0.6163

Right Leader Competence = 0.6218

US —2000 (Annenberg)

Inspiring Cares Conservative Leader Character (1) 0.7621 (2) = 0.8516

Knowledgeable Honest Conservative Leader Competence = 0.8265

Strong Leadership Hypocritical Centre-Left Leader Character = (1) = 0.7733 (2) = 0.8627 Trustworthy Centre-Left Leader Competence 0.7927

(1) = with hypocritical; (2) = with trustworthy (no overlap between respondents who were asked these two traits) US 1972 Presidential Personality Trustworthy Traits in Study Alpha for Character and Competence Competence Traits Character Traits US 1976 Presidential Personality Trustworthy

US 1980

Knowledgeable Moral Conservative Leader Character = 0.6580 inspiring Dishonest Conservative Leader Competence 0.7656 Strong Leadership Power Hungry Centre-Left Leader Character 0.6147 Centre-Left Leader Competence 0.7909

US 1984

Hard Working Decent Conservative Leader Character = 0.9385 Commands Respect Compassionate Conservative Leader Competence = 0.8795 Intelligent Moral Centre-Left Leader Character = 0.9232 Inspiring Kind Centre-Left Leader Competence = 0.8602 Sets Good Knowledgeable Example Strong Leadership Cares Understands Fair In Touch Religious US 1988

Intelligent Compassionate Conservative Leader Character = 0.8757 Inspiring Moral Conservative Leader Competence = 0.8101 Strong Leadership Decent Centre-Left Leader Character 0.8593 Knowledgeable Cares Centre-Left Leader Competence = 0.7847 Honest

00 Traits in Study Alpha for Character and Competence Competence Traits Character Traits US 1992 Intelligent Compassionate Conservative Leader Character 0.8473

Inspiring Moral Conservative Leader Competence = 0.8332 Strong Leadership Cares Centre-Left Leader Character = 0.8297 Knowledgeable Honest Centre-Left Leader Competence = 0.8514 Gets Things Done

US 1996

Intelligent Compassionate Conservative Leader Character = 0.8093 Inspiring Moral Conservative Leader Competence = 0.8101 Strong Leadership Cares Centre-Left Leader Character = 0.8922 Knowledgeable Honest Centre-Left Leader Competence 0.8497 Gets Things Done

US 2000

Knowledgeable Moral Conservative Leader Character = 0.7229 Strong Leadership Cares Conservative Leader Competence = 0.8177 Intelligent Dishonest Centre-Left Leader Character 0.7846

Out of Touch Centre-Left Leader Competence = 0.7848

US 2004

Strong Leadership Moral Conservative Leader Character = 0.8018 Knowledgeable Cares Conservative Leader Competence = 0.7941 Intelligent Dishonest Centre-Left Leader Character 0.7349

Indecisive Centre-Left Leader Competence = 0,7532 Table 5.1 Effect of Socio-Demographics and Partisanship on Character Evaluations Centre-Left Conservative Left Leaders Leaders Leaders ,Socio-Dtvuographiec Married -0.015 0.033 0.020 (0.003) 0.003) (0.004) \X’oman 0.009 0.007 -0.008 ).003) (0003) (0.003) University Degree 0.014 0.002 0.054 (I )J H(S) (0.003) (0.) H (4) Lrnployed -0.012 0.008 0.017 (0.003) (((.0)3) (0.(H (4) Age 0.003 0.015 0.007 (0.) 8)1) (0.001) (0.001) Pu/t/sa/Ih/p (;entre-Left PID 0.136 -0.107 0.021 (0.) ((3) (0.) (03) (0.) ((5) Conservative PID -0.082 0.173 0.019 (0.004) 0.((03) (0.005) Left ND -0.014 -0.034 0.166 ((.0)7) (0.006) (((.006) Other P11) -0.062 -0.056 0.046 (0.) 8)6) (0.008) (0.007) Observations 122874 115378 36640 R-squared 0.23 0.26 0.27 OLS [stimation, robust standard errors in parenthess Mode! includes controls for fiNed effects., with dummy variables for each election study Coefficients statistically significant at minimum of 55, level in bold Table 5.2 Effect of Socio-Demographics and Partisanship on Competence Evaluations Centre-Left Conservative Left Leaders Leadcrs Leaders .focth-D,sgraphiec Married -0.014 0.022 ((MO (0.003) (0.0(3) (U8(4) \Voman 0.005 ((.001 0.011 0.O02) (0.0(3) (0.004) University Degree 0.001 -0.031 0.029 (0.( (03) 0.003) O.( (04 Employed —0.016 0.005 0.001 0.0((3) (0.0( (3) (0.004) Age -0.002 0.006 0.005 (0.( (01) (0.( ((1) (Out) 1) Porisa,iship Centre-Left PID 0.127 -0.084 0.012 (0.( 03) ( (.( (03) (((((((6) Conservative PID -0.075 0.142 -0.029 0.003) (((((((3) (0.006) Left PID -0.017 -0.018 0.180 (0.007) (((((((6) (0.( ((17) Other PlO -0.036 -0.030 0.015 (0.( (((6) (0.007) (((((((8) Observations 122809 114720 35314 R-squared ((.26 ((.23 ((.25 OLS Estimation, robust standard errors in parenthess Model includes controls for fixed effects, with dummy variables for each election study Coefficients in bold significant at minimum of 5/ level Table 5.3 Effect of SES, Partisanship, and Issue Attitudes on Character Evaluations (;entre-Left Conservative Left Leaders Leaders Leaders .Socio-I)wiopaphi:r Married -0.006 0.025 0.024 (0.0)4) (0.004) (0.1)06) Woman 0.006 0.010 -0.006 (0.003) (0.1)04) (0.005) University Degree 0.005 0.008 0.044 (0.) (04 (0.004 (0.106) Employed —0.008 0(X) 1 0.016 (0.004) (0.004) (0.1)06) Age 0.004 0.012 0.013 ((.001) (((.001) (0.002) Parlisimship Centre-Left PID 0.112 -0.075 -0.003 (0.004) (((.0(15) (0.1)1(8) Conservative PID -0.053 0.133 1)013 0.005) (0.005) (0.009) Left P1D -0.038 -0.032 0.166 (0.010) (0(108) (0.011)) Other PID -0.061 -0.034 0.039 (0.008) (0.010 (0.01))) Ideo/ogj Left-Right Self-Placement 0.145 -0.190 0.082 (0.009) (0.009) (.013) Issue /111,1/Ides: Major Campaign Issue 0.063 -0.040 0.013 (0.005) ).005) (0.009) Taxes versus Spending 0.077 -0.066 0.076 (0.006) (0.006) (((.009) Social Liberalism 0.024 -0.034 0.020 (0.005) (((.005) Observations 74865 67899 15536 R-squared 0.2$ 0.34 0.21 OLS Estimation, robust standard errorsin parentheses Model includes controls for fixed effects, with dummy variables for each election stud Coefficients in bold significant at minimum of 5% Table 5.4 Effect of SES, Partisanship, and Issue Attitudes on Competence Evaluations Centre-Left Conservative Left Leaders Leaders Leaders .th-Deiuo,grapliiec Married -0.010 0.018 -0.001 (0.003) (0.004) (ft( (06) Woman 0.001 0.002 0.016 ((L()03) (003) (0006) University Degree -0.006 -0.023 0.021 0.003) (1)004) (0.006) Employed -0.014 0.006 0.012 (0.004) (0.004) (0.006) Age —0.001 0.002 0.005 (0.001) (0.01(1) (0)11)2) Pai’tLranship Centre-Left PID 0.099 -0.065 0.020 (0.004) (0.005) (0.008) (;onservati’e PID -0.042 0.107 0.002 (0.004) (11.004) (0,009) Left PID -0.038 -0.022 0.156 (0.008) ).( (09) 0.((12) Other PID -0.048 -0.015 0.012 0.00$) (((((10) (0.011) Ideology Left-Right Self-Placement 0.121 -0.171 0.063 (0.008) ((.(1((9) 0.014) Issue /1/ti//ide. Major Campaign Issue 0.049 -0.045 0.036 (((.01(5) (((.0)15) (0.010) Taxes versus Spending 0.067 -0.039 0.053 0.005) (0.005) (0.010) Social Liberalism ((.008 -0.045 0.023 ).0( (4 (0.005) (((.11)19) Observations 75019 67771 15140 R-squarcd 0.35 0.26 0.33 OLS Estimation, robust standard errorsin parentheses Model includes controls for fixed effects, with dummy variables for each election studs’ Coefficients in hold significant at minimum 5% level Table 5.5 Evaluating Women Leaders: the Effect of Sex on Perceptions of Character and Competence Evaluations of Party Leaders Centre-Left Conservative Left Character Woman Leader (Main Effect) -0.382 -0.479 -0.038 (0.007) (0008) (0.009)

Woman Votcr * Woman Leader (Interaction) 0.032 -0.017 0.01 ft006) (0.006) (0.007) Total -0.349 -0.496 -0.028 (0.007) (((.008) (0.009) Number of Observations 12(1918 113419 36640 (ompeIe)ice ‘voman Leader (Main Effect) -0.112 -0.00$ -0.024 (((.18)6) (0.007) (0.009) \Voman Voter * Woman Leader (Interaction) 0.028 0.0(16 0.035 0.006) (((((((8) Total -0.085 -0.002 ((.011 0.006) (0.007) (0.009) Number of Observations 12(1855 112767 35314 OLS Estimation, robust standard errors in parentheses Demographic and Partisanship Controls, as well as fixed effects included in model, coefficients not shown Coefficients in bold significant at 5% level Table 5.6 The Role of Political Sophistication in Conditioning the Sources of Trait Evaluations of Centre-Left Party Leaders Cliaracter Competence Low High Low High Sophistication Sophistication Sophistication Sophistication .Vocio-Demogoiptiics Married —((.012 0.001 —(1.012 —0.0(12 (0.009) 0.007) (((.009) (0.006) Woman ((.013 0.002 —((.006 —0.005 0.010) (0.007) (((.1)1(9) (0.006) University Degree ((.007 0.0(17 (1.002 —0.010 (0.012) 0.007) (0,011) 0.006) Employed 0.000 -0.005 —0.020 -0.013 (0.010) (0.008) (0.009) (0.1(07) Age 0.003 0.004 —0.000 0.000 (((.1(1(3) 0.002) (0.003) 0.002)

Centre-Left PID 0.118 0.134 0.100 0.078 (0.011) (0.010 0.011) (0.008) Conservative PID -0.015 -0.071 -0.009 -0.047 (((.1(13) (0.01(1) (0.012) 0.009) Left PID -0.064 -0(123 -0.1(11 -0.052 (0.024) (0.015) (((.1126) (((.1)11) Other PID -0.058 -0.050 -((.1)16 -0.051 (((.1(14) (((((13) (0.014) (0.010) IdeaJo Left-Right Self-Placement 0.070 0.154 0.055 0.152 (0.022) (0.017) (0.022) 0.0i5) issue Attitudes Major Campaign Issue 0.046 0.058 0.037 0.061 (0.014) (0.011) (11.1)13) (0.009) Taxes vetsus Spending 0.070 0.079 0.058 0.074 ((1.1)16) (11.1)11) (((.1(14) (0.009) Social Liberalism -0.1)11 0.029 -((.013 0.019 (((.1)13) (0(110) (((.012) (((.0(18) Observations 11753 19307 1178(1 19310 R-s9uarcd ((.21 (1.37 0.22 ((.55 OLS Estimation, tol,ust standard errors in parentheses Fixed Effects (Dummy Variahles for each election study) included hut coefficients not shown Cocfficients in bold significant at minimum of S% level Table 5.7 The Role of Political Sophistication in Conditioning the Sources of Trait Evaluations of Conservative Party Leaders Character Competence Low High Low High Sophistication Sophistication Sophistication Sophistication .S’oc/o-Deiiig,ipic Married 0.011 0.027 0.016 (1.013 (0.1 10) (0.008) (0.009) (0.017) Woman 0,006 —0.007 —OX) 10 —0.0(14 (0010) (0.008) 0.(II 0) (0.0(18) University Degree 0.033 0.016 0.008 -0.011 (11.1112) 0.007) ((1.1)11) (0.1)1)7) Employed —(1.0(16 (1.003 0.008 0.003 (((.1(11) ((1.0(18) 0.011) (0.008) Age (1.005 0.014 0.004 —0,002 ((1.1)1)3) (1)002) 0.(l03) 0.002 Pa’/istm.chip (;cntre-Lcft PID -0.045 -0.088 -0.057 -0.076 (0.012) 0.0I1) ).012) (0.010) Conservative PID 0.123 0.149 0.091 0.114 (((.1)13) (0.01() (0.012) (0.01(9) Left PID -0.027 -0.046 -0(149 -0.006 (0.025) 0.0l2) (0.027) ((.014) Other PID —0.003 —0.041 —0.018 —0.010 (0.1)17) (((.1)16) (RU 18) (((.1116) Ideo/o,g)’ Left-Right Self-Placement -0.102 -0.214 -0.076 -0.191 (0.024) CR019) (0.024) (((.018) Issue /1/fit//des Major Campaign Issue -0.022 -0.031 -0.019 -0.043 (((.1(14) (0.011) ((((((4) (((.1(1(1) Taxes versus Spending -0.058 -0.072 -0.013 -0.052 (0.015) (((.1(11) 0.015) (0.010) Social Liberalism -0.039 -0.045 -0.040 -0.040 (((.1(14) 0.01(() (0.013) (0.010)

Observations 9888 16620 ‘ 9794 16557 R-sc1uared ((21 0.49 (1,17 (1.38 OLS Estimation, robust standard errors in parentheses Fixed Effects (Dummy Varjal,les for each election study) included but coefhcienrs not shown Coefficients in bold significant at a minimum of 5% ______

Table 5.8 Number of Respondents Evaluating Major Party Leaders, by Political Sophistication Ccntrc-Lc ft Conservative C haractcr Competence Character Competence

Low 27,533 27,342 24,663 24,183 High 31,584 31,553 28,437 28,320

Table 5.9 Political Sophistication and Evaluations of Women Leaders: the Effect of Sex on Perceptions of Character and

Competence

Evaluations of Parts’ Leaders

Ccntrc-Le ft Conservative

Chaiac1r Low High 1.0W High Woman Leader (Main Effect) -0.369 -0.192 -0.691 -0.691 (0.021) 0.014) (0.023) f).013) \X’man \!oter * \Voman Leader (Interaction) 0.039 0.021 —0.029 —0.020 (0.015) (0.011) ((1.029) (0.013) Total -0.330 -0.172 -0.720 -0.711 0M2l) (0.015) 0.025) 0.0l3) Number of Observations 11753 19307 9888 16620 Coi,ipe/ellce iman Leader (Main Effect) -0.252 -0.243 -0.267 -0.232 (0.020) (0.0 13) (0.029) 0.014) \Xman Voter * \Xoman Leader (Interaction) 0.041 0.032 -((.024 -0.012 (0.014) (0.010) (((.033) (0.014) Total -0.211 -0.211 -0.292 -0.243 (0.019) 0.0l3) (0.029) 0.014) Number of Observations 1178(1 19310 9794 16557

OLS Estimation, robust standard errors in parentheses

Demographic anti Partisanship Controls, as well as fixed effects included in model, coefficients not shown Coefficients in bold significant at 5% level Figure 6.1 Summary Statistics: Evaluations of Leaders of Three Main Party Leaders’ Competence and Character

Comparing Major Parties: Ratings of Competence Comparing Major Parties: Ratings of Character

Conserv5v (n=124, 914) Cntr-Lft (n=133, 954) Left (n4O, Coservative (=12 506) Cen1reLeft (n=134, 068) EEl Left (n=42,

00 Table 6.1 Effects of Party Label on Evaluation of Leaderts Traits Character Cornpcrencc Conservative Leader -0.004 r 0.028 (0.001) r (0.001)

Left Leader 0.109 “ -0.040 r (0.002) (( ).0( 2)

Centre-Left Leader (reference) -- -- r Nurnhcr of Observations 301649 r 299463 Clusters r 138079 r 137816 R-squarcd 0.12 0.11 Stacked Ordinary Least Scuares Regression Analysis Robust Standard errors in parentheses (clustered on the individual) Fixed Effects (Dummy Variables for Each Elcerion Study) included in the model Coefficients significant at minimum 5% level in bold Table 6.2 Difference in Means on Evaluations of Most Frequently Asked Traits (T-Tests) Conservative vs Centre-Left vs Conservative vs Centre-Left Left Left Lcadcrship 0.047 0.016 0.150 Cares -0.077 -0.091 -0.254 Knowledgeable -0.052 0.022 0.043 intelligent -0.008 0.059 0.062 inspiring 0.034 0.030 0.015 Honest 0.056 —0.079 —0.024 Compassionate -0.013 -0.041 -0.139

Trustworthy 0.011 — —0.057

Arrogant (1.011 0.170 0.152 Moral —0.034 —0.122 —0.165

QO © ______

181

Rating p p p p M F- C i- F’..) (A) -1.

I

2o — —

:4 2004 — C (2o- (b 4 CI) —. C 19(

C. — )-t g 1’L H — 1984

k 1992 — 2o

1987

‘983 Figure 6.3

Character Evaluations: Comparing Leaders of Three Main Parties to Average of All Leaders

0,2-

0.15

: .111111111111 r’ -0.15

-0.2

-0.25

- C) j-.. c - t- t- cc ‘-. cc cc “1 c cc cc cc CN C C’ (‘4 -, (‘4 —. — (‘4 - — —, <‘4

C) cc L) ( () D

• Cnnservrve • (‘,enri’e-Te.ft • Left

00 Figure 6.4

Competence Evaluations: Comparing Leaders of Two Main Parties to Average of All Leaders (No Left Party Leader Evaluated)

0.25 0.2 0.15 0.1 0.05 0 -0.05 -0.1 -0.15 -0.2

N- N- 00 N- .D C N o o - N 03 Q 00 N- \D 03 03 00 03 N- CN C\ — (\j cD , ‘—‘ —, -_4 —_ — _• ,, ‘__4 —j _—4

Conservative Centre-Left

00 Figure 6.5

Character Evaluations: Comparing Leaders of Two Main Parties to Average of AU Leaders (No Left Party Leader Evaluated)

0.15

0.1

0.05

0 j:4ji

-0.05

—0.1

-0.15

cc -, - cc N - C\ C CC ‘D rj ‘.D CC cc c c cc cc cc cc c

- ,-_ _-. __ - — ‘ (N @4 @4 t—, C

Conservative Centre-Left

00 Table 6.3 Effects of Voters’ Partisanship and Leaders’ Party on Evaluations of Personality Traits Character Competence Centre-Left Partthm. Centre-Left PID & Conservative Leader -0.072 0.004 ([(.002) (0.0)13) Centre-Left PID & Left Leader 0.144 0.002 (0.003) (0.003) Centre-Left P11) & Centre-Left Leader 0.163 0.142 (0.002) (((.003) (m.re,7’ati,s I’GltfldtIJ Conservative PID & Conservative Leader 0.177 0.204 (0.0(13) (((.003) Conservative PID & Left Leader 0.149 -0.037 (0.003) (([.004) Conservative P11) & Centre-Left Leader -0.043 -0.05 (0.003) (0)103) Left Pan’,.ca,is Left PID & Conservative Leader -0.039 0.079 (0.005) ([(.01)5) Left PID & Left Leader 0.289 0.161 (0.005) (0.005) Left P1D & Centre-Left Leader 0.019 -0.013 (0.005) (0.005) ou-J’artiems No PID & (;onser’ative Leader 0.007 0.054 (0.003) (0.003) No PID & Left Leader 0.087 -0.02 (0.003) ((1(103) No PID & Centre-Left Leader 0.006 0.021 (0.003) (0.01(3) Other Part/sans Other PID & Conservative Leader -0.067 0.045 (0008) (((.009) Other PID & Left Leader 0.135 0.021 0.01) (((.011) Other P1D & Centre-Left Leader (Reference Observations 185242 182668 Number of Clusters 80441 80014 R-squared 0.24 0.2 Stacked Ordinary Least Squares Regression Analyses Robust standard errors in parentheses (Clustered on the Individual) Fixed Effects (Dummy Variables for Each Election Studs’) included as controls but coefficients nor shown Coefficients significant at minimum 5% level in bold _H______H_

Figure 6.6 Summary Statistics: Evaluation of Leaders of Three Main Parties’ Character and Competence, by Level of Political Sophistication

Comparing Major Parties: Ratings of Competence Comparing Major Parties: Ratings of Character by Level of Political Sophistication by Level of Political Sophistication TTT TLT :TTT, TTT H H 111 i1! ‘iiI 11 Low Sophistication High Sophistication Low Sophistication High Sophistication

II I Conservative - Centre-Left I I Left I I Conservative I I Centre-Left Left

00 ______

Table 6.4 - Effects of Voters’ Partisanship and Leaders’ Party on Trait Evaluations, by Level of Political Sophistication Character Competence

Low Sophistication High Sophistication Low Sophistication High Sophistication (en/ic-f-‘1/ Pa,tiscms Centre-Left PID & Conservative Leader -0.059 -0.164 -0.038 -0.049 (I (.0(6) (((008) (0.0 (7) (((.0(17) Centre-Left PID & Left Leader 0.125 0.149 -0.003 -0.111 (0.0))))) (((.007) (0.1)10) (((.007) Centre-Left PID & Centre-Left Leader 0.130 0.157 0.102 0.103 (11.1)07) (0.007) fl.0O6) (((.006) (s,i.ce,i’c,ü,e Part,scn,.c Conservative PID & Conservative Leader 0.154 0.177 0.132 0.218 (0.1(1)7) (00)07) (((((((7) (0.1(06) Conscrvatis’e PID & Left Lcadcr 0.145 0.144 -((.009 -0.188 (0.11)8) (0.008) (0.010) (0.01(7) Conservative P11) & Centre-Left Leader -0.017 -0.110 -0.041 -0.134 (0.008) (((.1(07) (0,008) (0.006) 14/ Pa,’/j.ccms Left PID & Conservative Leader -0.019 -0.138 -0.011 0.157 (0.015) (0.010 (((((18) (0.010) Left PID & Left Leader 0.246 0.324 0.158 0.06 (((.1(12) (0.009) (((.1(16) (0.011) Left P11) & Centre-Left Leader -0.010 0.025 -0.1(24 -0.131 (((((15) (((.1(11) (((.1(17) (0.010)

No PID & Conservative Leader 0,0(11 —0.020 ((.1(1(1 p0.086 (((.009) (0(110) (0.006) (0.0(0) No PID & Left Leader 0.065 0.132 -0.045 -0.106 (0.11)9) (((.11)9) (((.1(1 1)) (((.008) No PID & Centre-Lefr Leader ((.007 -0.032 ((.002 -0.059 (0.009) (0.009 (((.1)1)8) (R00(7) 0/he,’ Pa,lisa,,s

Other PID & Conservative Leader -0.046 —0.032 0.002 (1.046 (((.031) (0(128) (L032) (0.031) Other PID & 1.eft Leader 0.127 0.209 ((.01(6 -0.107 (((031) (ft(L2) 40035) (0.025) Other P1D & Centre-Left Leader (Reference) Observations 44169 53880 42594 53 Number of Clusters 19872 22122 19552 22076 R-squared ((.16 0.30 ((.1(1 0.29 Stacked Ordinary Least Squares Regression Analyses Robust standard errors in parentheses (Clustered on the Individual) Fixed Effects (Dummy Variables for l.(ach [(lection Study) included as controls but coefficients not shown

Coefficients in bold significant at a minimum level of 5’Yn Table 7.1 Conservative Party Vote Choice: The Impact of Trait Evaluations (1) (2) .focio-Dr///oç’rapbics Married 0.305 0.222 0.114 ((.090 0043 (0.03( (0.( (57) (0.1 64) ((1.1)68) (0.080) ‘coman —0.069 0.038 0.017 0.038 ((.099 (((.036) (0(153) (0.059) (0(163) (0.073) University Degree -0.037 0.101 0.146 0.182 0.164 (0.042) ((1.059) (0.064) (0.068) (11.080) Employed 0.087 1)1)69 ((((22 -0.025 0.133 (((.1(41) (((.1(60) (0.067) (((.072) (((.1(81) Age 0.077 0.038 -0.013 -((((((2 0.057 ft1(12) (((((18) (0.020) (((((22) (0.025) Pu1i.caiis/iip Conservative PID 2.117 1.716 1.267 1.192 1.165 (((.049) (((.069) (((.1(77) (((((81) (0.1 ((9) Centre-Left PID -1.808 -1.560 -1.393 -1.101 -0.617 (0.046) (0.069) (((.075) (0.1(79) 0.112 Left PID -1.750 -1.884 -2.062 -2.466 -2.076 (0.1(77) (((.127) (0.143) (((.169) (((.184) Other PID -1.391 -1.514 -1.735 -2.200 -2.474 (((.1)91) (0.141) (((.163) (0.181) (0.199) Ideo/5g( Left-Right Self-Placement -2.944 -2.346 -1.841 -1.779 (0.133) (((.151) (0.165) (0.206) Iss,,e /l////ndes Major Campaign Issue -0.757 -0.576 -0.369 -0.775 (0.073) (((((84) (0.1(89) (0.120) Social Liberalism -0.742 -0.648 -0.599 -0.310 (((.083) (((.092) (0.099) (0.114) Taxes VerSUS Spending -0.955 -0.868 -0.559 -0.904 (((.1(95) (((.110) (((.121) (((.124) •I 15111 ! Oil/kit/OIlS Conservative Leader Character 3.670 3.993 3.846 (0.164) (((.189) (0.195) Conservative Leader Competence 2.491 2.722 1.832 (((.153) (0.173) (0.173) c:entre-Lca Leader Character -2.875 -2.484 (((.164) (((.157) Ccntre-1..eft Leader Competence -2.138 -1.504 (0.169) (((.187) Left Leader Character -0.633 (((.167) Left Leader Competence -0.479 (((.162) Observations 62500 32011 29107 28764 9979 Logistical Regression Analysis (oefficients Reported; Robust standard errors in parentheses Model includes controls for fixed effects, with dummy variables for each election study Coefficients in bold significant at a minimum level of 5% Table 7.2 Centre-Left Party Vote Choice: The Impact of Trait Evaluations (1) (2) (3) (4) (5)

Married -0.289 -0.205 -0.200 -0.094 0.087 (0.037) (0.1(54) (O.06( I) (0.( (66) (0.078) Woman 0.132 0.038 0.025 0.053 -0.032 (0.035) (0.051) (0.( (57) (I ).( 62) (0.071) University Degree -(1.028 -0.128 -0.107 -0.143 -0.383 (((.1(41) (((.1(57) (0.062) (((.068) (0.078) hmploved —0.068 —0.029 0(158 0.136 —(1.134 (((.040) (((.1(58) (0.1(64) (((((72) (((((78) Age -((.1)19 0.032 ((((25 0.081 -0.009 (0,1(11) (((017) (0,( 19) (((.022) (((((24)

(;()nserv(itive PID -1.926 -1.532 -1.490 -1.006 (1.11)2 (((.1(52) (0.073) (0.1(81) (((.1(89) (0.115) Centre-Left P11) 1.986 1.742 1.509 1.375 1.833 (0.042) (((.061) (0.069) (0.074) (0.105) Left PID -1.887 -2.373 -2.473 -2.954 -1.860 (0.080) (((.126) (0.144) (((.184) (0.197) Other PID -1.642 -1.860 -1.728 -2.251 -1.734 (((.1(83) (((.143) (((.146) (0.213) (0.204) Ideology Left-Right Self-Placement 2.588 2.098 1.387 1.111 (((.128) (((.145) (((.169) (((.185) Levis /11tlt,,dec Major Campaign Issue 0.863 0.654 0.503 1.005 (0M73) (0.082) (0.09(1) (((.117) Social Liberalism 0.559 0.547 0.378 0.098 (0.080) (0.090 (0.11)))) (0.112) Taxes versus Spending 0.974 0.782 0.575 0.357 (0.091) (0.104) (((.119) (0.117) .1 rail f:i’alnalioiec c:cntre-Left Leader Charicter 2.853 3.431 3.351 ()).l40 (0.165) (0.152) Centre-Left Leader Competence 2.749 2.903 1.942 (0.151) (0.174) (0.179) Conservative Leader Character -3.280 -2.416 (((.174) (((.175) c;c,iscrative 1..eader Competence -2.406 -1.586 (((.163) (((.155) Left Leader Character -0.872 (0.165) Left Leader Competence -0.393 (0.160) Observations 71512 38636 35741 28764 9979 Logistical Regression Analysis Coefficients Reported; Robust standard errors in parentheses Model includes controls for fixed effects, with dummy variables for each election study Coefficients in bold significant at a minimum level of 5V Table 7.3 Left Party Vote Choice: The Impact of Trait Evaluations (1) (2) (3) (4 Soc/s-I )e,,/ogrep/occ Married —0.027 —0.039 -0.058 —0.076 (0.043) (0.065) (0.082) (((.086) \‘omafl 0.023 0.035 0.1(04 —0.001 9(041) (0.061) (0.078) (0.081) University Degree 0.362 0.304 0.176 0.221 (0.045) 9(067) (0.084) (((.089) Lmplovcd (1.047 0.104 0.102 0.106 (0.045) (((((68) (0.088) (0,092) Age -0.032 -0.027 -0.054 -0.043 (0.013) (0.020) (0.027) (0(128) PrF/t/s(fus/vp (;c)nservathc P1D -1.648 -1.556 -1.772 -1.655 (((((66) (0.099) (0.126) (((.131) Centre-Left PID -1.164 -1.128 -1.079 -1.023 (((((58) (((.085) (0.111) (((.1(9) Left PID 2.617 2.629 2.199 2.194 (((.1(65) (0(198) (11.124) (((.132) Other PID -0.196 -0.270 -0(162 -0.063 (((.068) (((.11)2) (11.128) (((.139) Ideo/o,gy Left-Right Self-Placement 1.216 0.942 0.820 (((.143) (((.192) (0.209) 1sf/se ‘lttguc/es Major Campaign issue 0.102 -((.119 -(1.139 (0.103) (0.125) (0.130) Social Liberalism 0.161 0.166 (1.200 (((.1(99) (((.129) (0.136) Taxes versus Spending 0.580 0.542 0.520 (0.098) ((.135) (((.144) I’n’,/? I- iu/iuiiiou.. Left Leader Character 1.887 1.980 (0.180) (0.186) Lcft Leader Competence 0.676 0.776 (((.154) (0.161) Conservative Leader Character -1.198 91.186) Conservative Leader Competence (1.037 (((.174) Centre-Left Leader Character -0.521 (0.158) Centre-Left 1,.eadcr Competence -0.262 (((.18 1) Observations 32739 29353 13233 9979 Logistical Re’ression Analysis Coefficients Reported; Robust standard errors in parentheses Model includes controls for fixed effects, with dummy variables for each election study Coefficients in bold significant at a minimum level of 5% Table 7.4 Predicted Probabilities: Changes in Probability of Vote for Party (over all other Parties) with Positive Evaluation of Leader’s Character and Competence Conservative Vote Centre-Left Vote Left Vote Conservative Leaders Character 0.4615 -0.4073 -0.047

Conservative Leaders Competence 0.3239 -0.3051 - Conservative Leaders Traits 0.6955 -0.6399 -0.045

Centre-Left Leader’s Character -0.3486 0.4319 -0.0204

Centre-Left Leader’s Competence -0.1108 0.1946 - Centre-Left Leader’s Traits -0.5672 0.6965 -0.0305

Left Leader’s Character 0.075 Left Leader’s Competence 0.0349 Left Leader’s Traits 0.113 * values represent change in probahility of vote for Party with change from one standard deviation below the mean trait evaluation to one standard deviation above the mean evaluation after conducting logistical regression model. Evaluations of Left Party Leader included only in Left Vote model Table 7.5 Predicted Probabilities: Changes in Probability of Vote for Party (over all other Parties) with Changes in Explanatory Variables Conservative Vote Centre-Left Vote Lcft Vote Ideology (most Left to most Right) (0-1) —0.364 0.3066 0.049 Issue Attitude: Taxes versus Spending (0-1) -0.1181 0.1269 0.0323

issuc Attitude: Major Campaign Issue (0-1) -0.0775 0.112 -

Issue Attitude: Social Liberalism (0-1) -0.1274 0.0837 - Conservative P1D (0-1) 0.2683 -0.2054 -0.0826 Centre-Left PID (0-1) -0.2144 0.3151 -0.0598 Left PID (0-1) -0.2752 -0.3265 0.3156 (..onservative PID to Centre-Left PID -0.4218 0.6017 0.0423 Conservative P11) to Left PID -0.464 -0.1273 0.6694 Centre-Left PID to Left PID -0.0424 -0.7295 0.6271 values represent change in probability of vote for Parts’ with change in independent variable (as indicated), with all other variables (including trait evaluations) held at their means, after conducting logistical regression model. Evaluations of Left Party Leaders arc included only in Left Vote Model. Table 7.6 Predicted Probabilities: The Effect of Partisanship on Changes in Probability of Vote for a Party with Positive Evaluation of Same Party Leader’s Character and Competence Conservative Partisans Centre-Left Partisans Left Partisans Non-Partisans Total Impact of Csa.re,,’atii’e Leader au Co,iserraiire I ‘ole: Conservative Leaders Character 0.4994 0.3034 0.1093 0.6289 0.5795 Conservative Leader’s Competence 0.3551 0.202 0.0685 0.534 0.4867 Conservative Leader’s Traits 0.7304 0.5239 0.2299 0.7854 0.7526 Impact of Cenlre-Le/1 Leader on centre-Left I Centre-Left Leader’s Character 0.2644 0.4372 0.0549 0.524 0.5877 Centre-Left Leader’s Competence 0.0981 0.2431 0.0173 ((.3196 0.3727 Centre-Left Leader’s Traits 0.5008 0.7032 0.1332 0.7209 0.7688 Impact of Left Leader on Left I ‘ole: Left Leader’s Character 0.0342 0.0604 0.2574 0.2259 0.1178 Left Leader’s Competence 1)0158 0.028 0.1236 0.1805 0.0915 Left Leader’s Traits 0.0522 (>0913 0.3669 0.2729 (>1465 * values represent change in probability of vote for Party with change from one standard devianon below the mean trait evaluation to one standard deviation above the mean evaluation after conducting logistical regression model. Evaluations of Left Party Leaders are included only in the Left Vote Model.

t) Table 7.7 Vote for Conservative Party: Conditioning Role of Political Sophistication on Impact of Trait Evaluations Lo\v Sophistication High Sophistication

Married ((.034 0.117 .Sis-Demg;tphiec (0.164) (0.144) Woman 0.174 0.221 (0169) (0.124) University Degree (1.137 ((.234 ((1.2(10) (0.136) F.rnpl ned —0(139 0. I 74 (((.155) (((.139) Age 1)049 ((((26 (0.052) (((((46) Pmiism.hip Conservative PID 0.962 1.524 U.2((7) (0,16(J) Centre-Left PID -1.319 -1.288 (0.210) (0.173) Left PID -2.607 -2.242 (((.487) (0.246) Other PID -1.907 -2.705 (0.334) (0.248) ides/ogy Left-Right Self-Placement -1.197 -2.735 (0.377) (0.334) issue /11/iludes Major Campaign Issue -0.263 -0.592 (0,2321 (0.197) Social Liberalism -0.652 -1.034 (0.274) (((.267) Taxes versus Spending -0.386 -0.550 (0.263) (0.198) •i,ml b,’a/,wt,o,,s Conservative 1_eader Character 1767 3.930 (0.437) (0.392) Conservative Leader Competence 1.844 3.405 0.37l) (0.330) Centre-Left Leader Character -2.090 -2.862 (((.4(16) (0.277) Centre-Left Leader Competence -1.675 -1.954 (((384) (0.342) Observations 4287 9562 Logistical Regression Analysis Coefficients Reported; Robust standard errors in parentheses Model includes controls for fixed cffccts, with dummy variables for each election study Coefficients in bold significant at a minimum level of 5°/n Table 7.8 Vote for Centre-Left Party: Conditioning Role of Political Sophistication on Impact of Trait Evaluations Lo\v Sophistication High Sophistication .Socth-Deiiiopap/its Married 0.048 0.012 (0.156) (0.135) Womnfl -0.031 -0.068 (0.157) (0.126) University Degrec -0.227 -0.217 (0.194) ((1.129) Lrnplovcd (1.060 (1.166 (0.156) (((.141) Age —(1.009 (1.060 (((.049) (0.046) Pm/t.rmislnp Conservative PID -11.266 -1.351 (0.220) (0.199) Centre-Left PID 1.935 1.634 (((.192) (((155) Left PID -1.806 -3.121 (0.444) (0.303) Other PID -1.761 -1.775 (0.313) (((.256) ideo/o.gy Left-Right Self-Placement 0.549 2.07 8 (((.371) (0.337) issue /l/I/t,,des Major (ampaign Issue ((.122 0.822 (0.228) (((.195) Social Liberalism 0.789 0.756 (((.271) (0.234) Taxes versus Spending ((.087 0.466 (((.252) (((.21(7) 1 ru/f fpu/,,fjo,,s (:onserative Leader Character -3.153 -3.279 (((.4(19) (((.360) Conservative Leader Competence -1.763 -3.044 (((.386) (0.322) Centre-Left Leader Character 2.627 3.647 (0.393) (((.312) Ccnrte-Lcft Leader Competence 2.461 3.272 (0.397) (0.352) Observations 4287 9562 Logistical Regression Analysis Coefficients Reported; Robust standard errors in parentheses Model includes controls for fixed effects, with dummy variables for each elecden study Coefficients in bold significant at a minimum level of 5% Table 7.9 Predicted Probabilities: The Effect of Political Sophistication on Changes in Probability of Vote for a Party with Positive Evaluation of Same Party Leader’s Character and Competence Conservative Vote Centre-Left Vote Low 1-ugh Low High Conservative Leader’s Character 0.4468 0.4416 -0.403 -0.3888 Conservative Leader’s Competence 0.228 0.3584 -0.2328 -0.3618 Conservative Leader’s Traits (1.6199 0.6963 -0.5847 -0.6694

Centre-Left Leader’s Character -0.2634 -0.3103 0.3676 0.436 Centre-Left Leader’s Competence -0.1051 -0.0363 0.196 (1.1773 Centre-Left Leader’s Traits -0.4531 -0.5033 0.6068 0.7184 values represent change in probability of vote for Party with change from one standard deviadon below the mean trait evaluation to one standard deviation above the mean evaluation after conducting logistical regression model. Evaluations of Left Party leaders arc not included in model. Table 7.10 The Effect of Trait Evaluations on Republican Vote Choice (1972-2004) • 1972 1976 198(1 1984 1988 1992 1996 2000 2004 Republican Leader Character 2.637 4.137 2.385 9.053 6.573 3.787 3.708 5.513 5.124 (0.435) (0.641) (0.66(l) (1.079) (1.082) (0.617) (0.843) (1.250) (0.937) Republican Leader Competence 1.306 1.701 5.700 3.378 7.051 3.975 3.162 5.934 3.852 (0.418) (0.514) 0.730) (0.945) (1.157) (0.707) (((.91(1) (1.166) (1.1113) Democrat Leader Character -1.625 -2.239 -1.330 -4.284 -5.019 -3.225 -4.938 -4.857 -4.767 ((1.448) (((.58(l) (0.631) (1.141) (1.179) (((.601) (0.811) (1.1(16) (1.(t18) Democrat Leader Competence -1.596 -2.594 -1.849 -5.437 -4.473 -2.111 -2.761 -3.106 -2.864 (0.463) (0.543) (((.696) (1.1(92) (1.079 (11.695) (0.938) (1.032) (1.145) Observations 856 674 745 1090 728 1331 967 537 768 Logistical Regression Analysis Coefficients Reported; Standard errors in parentheses Coefficients in bold significant at a minimum level of 5% ______-

Figure 7.1

Evaluations of Democratic and Republican Candidates (1972-2004)

0.7

Os:

0.5

0.45

0.4 1972 1976 1980 1984 1988 1992 1996 200() 2004

Rcpiiblican (]1ar1ctcr Rej,ubhcan Conpctcncc Democrat CharacLer Democrat Competence

00 Figure 7.2

Net Republican Gain from Evaluation of Candidates (1972-2004)

0.11

0.09

0.07

0.05

0.03

0.01

0.0 I 1972 1976 198(1 I 984

-0.03

Republican net gain fri,m Republican evaluation Republican net gain from 1)cmocrar evaluation ‘fital Republican net gain from evaluations

-a BIBLIOGRAPHY

Aarts, Kees, and André Blais. Forthcoming, 2009. Pull or Push? Positive and Negative Leader Evaluations and Vote Choice. In PoliticalLeadersandDemocraticElections,edited by K. Aarts, A. Blais and H. Schmitt. Oxford: Oxford University Press. Abelson, Robert P., Donald R. Kinder, Mark D. Peters, and Susan T. Fiske. 1982. Affective and Semantic Components in Political Person Perception. JournalofPersonalityand SocialPsjchology42:619-30.

Adorno, T.W., Else Frenkel-Brunswik, DanielJ. Levinson, and R. Nevitte Sanord. 1950. The AuthoritarianPersonali!y.New York: Harper & Row. Aldrich, John H., Paul Gronke, andJeffGrynaviski. 1999. Policy, Personality, and Presidential Performance. Paper read at Midwest Political Science Association Annual Meeting, at Chicago, Illinois.

Almond, Gabriel, and Sidney Verba. 1963. TheCivicCulture:PoliticalAttitudesandDemocrafyin FiveNations.Princeton: Princeton University Press. Aispach, Steven E., and George F. Bishop. 1991. Question-Order Effects of Presidential Approval Ratings on Gubernatorial Approval Ratings: A Research Note. SocialForces 69 (4):1241-1248. Alvarez, R. Michael. 1997. Information“ Elections.Michigan: University of Michigan Press. Alvarez, R. Michael, and Jonathan Nagler. 1998. When Politics and Models Collide: Estimating Models of Multiparty Competition. AmericanJournal ofPoliticalScience 42:55-96. Alwin, Duane F. 1976. Attitude Scales as Congeneric Tests: A Re-Examination of an Attitude-Behavior Model. Sociomety39 (4):377-383. Anderson, Norman H. 1968. Likeableness Ratings of 555 Personality-Trait Words. Journalof PersonalityandSocialPychologv9 (3):272-279. Arter, David. 1989. A Tale of Two Carlssons: The Swedish General Election of 1988. ParliamentaryAffairs42 (1):84-101. Bale, Tim, and Ingrid Van Biezen. 2007. Political Data in 2006. EuropeanJournalofPolitical Research46:853-886. Banducci, Susan A., and Jeffrey A. Karp. 2000. Gender, Leadership and Choice in Multiparty Systems. PoliticalResearchQuarter!y53 (4):815-848. Bartels, Larry M. 1988. PresidentialPrimariesandtheDjinamicsofPublicChoice.Princeton: Princeton University Press. Bartels, Larry M. 1996. Uninformed Votes: Information Effects in Presidential Elections. AmericanJournalofPoliticalScience40 (1):194-230. Bartels, Larry M. 2002. The Impact of Candidate Traits in American Presidential Elections. In Leaders’PersonalitiesandtheOutcomesofDemocraticElections,edited by A. King.

200 Oxford: Oxford University Press. Bartels, Larry M. 2002. Beyond the Running Tally: Partisan Bias in Political Perceptions. PoliticalBehavior24(2):117-50. Battle, John, and Ivor Crewe. 2002. The Impact of Party Leaders in Britain: Strong Assumptions, Weak Evidence. In Leaders’PersonalitiesandtheOutcomesofDemocratic Elections,edited by A. King. Oxford: Oxford University Press. Bean, Clive. 1980. Political Leaders and Voter Perceptions: Images of Muldoon and Rowling at the 1975 and 1978 New Zealand General Elections. PoliticalScience32:55-75. Bean, Clive. 1993. The Electoral Influence of Party Leader Images in Australia and New Zealand. ComparativePoliticalStudies26:111-32. Bean, Clive, and Anthony Mughan. 1989. Leadership Effects in Parliamentary Elections in Australia and Britain. AmericanPoliticalScienceReview83:1165-79. Beilenson, Peter. 1982. TheWisdomand Wit ofFranklin D. RooseveltPeter Pauper Press. Benoit, Kenneth, and Michael Layer. 2006. PartyPoli’yinModernDemocracies.London: Routledge. Berelson, Bernard, Paul Lazarsfeld, and William McPhee. 1954. Voting.Cambridge: Harvard University Press.

Berinsky, Adam J. 2004. SilentVoices:PublicOpinionandPoliticalParticzjationinAmerica.New Jersey: Princeton University Press. Bittner, Amanda. 2007. Competence or Charisma? The effects of measurement on why some leaders’ traits matter more than others. Paper read at Canadian Political Science Association Annual Meeting, at Saskatoon, SK. Bittner, Amanda. 2007. The Effects of Information and Social Cleavages: Explaining Issue Attitudes and Vote Choice in Canada. CanadianJournalofPoliticalScience40 (4):935- 968. Bittner, Amanda. Forthcoming, 2008. Personality Matters: The Evaluation of Party Leaders in Canadian Elections. In CanadianVotingBehaviour,edited by L. Stephenson and C. Anderson. Vancouver: University of British Columbia Press. Blais, André. 2005. Accounting for the Electoral Success of the Liberal Party in Canada Presidential Address to the Canadian Political Science Association London, Ontario June 3 2005. CanadianJournalofPoliticalScience38 (4):821-840. Blais, André, and M. Martin Boyer. 1996. Assessing the Impact of Televised Debates: The Case of the 1988 Canadian Election. BritishJournalofPoliticalScience26 (2):143-164. Blais, André, Elisabeth Gidengil, Richard Nadeau, and Neil Nevitte. 2002. Anatomj ofa Liberal Victory:MakingSenseofthe Votein the2000 CanadianElection.Canada: Broadview Press, Ltd. Blais, André, Neil Nevitte, Elisabeth Gidengil, and Richard Nadeau. 2000. Do People Have Feelings Toward Leaders About Whom They Say They Know Nothing? ThePublic OpinionQuarter/y64 (4):452-63.

201 Blalock, H.M. 1970. Estimating Measurement Error Using Multiple Indicators and Several Points in Time. AmericanSociologicalReview35 (1):101-111. Bosch, Agusti, and Guillem Rico. 2003. Leadership Effects in Regional Elections: The Catalan Case. In WorkingPaper.Barcelona: Institut de Ciencies Politiques i Socials (ICPS). Brady, Henry, and Paul Sniderman. 1985. Attitude Attribution: A Group Basis for Political Reasoning. AmericanPoliticalScienceReview79:1061-1078. Brettschneider, Frank, and Oscar W. Gabriel. 2002. The Nonpersonalization of Voting Behavior in Germany. In Leaders’PersonalitiesandtheOutcomesofDemocraticElections, edited by A. King. Oxford: Oxford University Press. Brody, Richard A. 1986. Candidate Evaluations and the Vote: Some Considerations Affecting the Application of Cognitive Psychology to Voting Behavior. In Political Cognition,edited by R. R. Law and D. 0. Sears. Hillsdale, NJ: Lawrence Erlbaum Associates. Brody, Richard A., and Benjamin I. Page. 1973. Indifference, Alienation and Rational Decisions: The Effects of Candidate Evaluations on Turnout and the Vote. Public Choice15:1-17. Brown,J. Andrew. 1992. The Major Effect: Changes in Party Leadership and Party Popularity. ParliameatayAffairs45:545-64. Brown, Steven D., Ronald D. Lambert, BarryJ. Kay, and James E. Curtis. 1988. In the Eye of the Beholder: Leader Images in Canada. CanadianJournalofPoliticalScience21:729- 55.

Budge, Ian, and Dennis J. Farlie. 1983. ExplainingandPredictingElections.Boston: Allen and Unwin. Budge, Ian, Hans-Dieter Klingemann, Andrea Volkens, Judith Bara, Eric Tanenbaum, Richard C. Fording, DerekJ. Hearl, Hee Mm Kim, Michael D. McDonald, and Silvia Mendes. 2001. MappingPolifyPreferences.Estimatesfor Parties,Electors,and Governments, 1945-1998. Oxford: Oxford University Press. Butler, David, and Austin Ranney. 1992. Electioneering:AComparativeStudjiofContinui’yand Change.Oxford: Clarendon Press. Butler, David, and Donald Stokes. 1969. PoliticalChangein Britain:ForcesShapingElectoral Choice.New York: St. Martin’s Press. Butler, David, and Donald Stokes. 1974. PoliticalChangein Britain:TheEvolutionofElectoral Choice.2 ed. London: MacMillan.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. 1960. The American Voter.Chicago: John Wiley & Sons, Inc. Campbell, Angus, Gerald Gurin, and Warren E. Miller. 1954. TheVoterDecides.Illinois: Row, Peterson.

202 Campbell, CoHn. 1980. Political Leadership in Canada: Pierre Elliot Trudeau and the Ottawa Model. In PresidentsandPrimeMinisters,edited by R. Rose and E. N. Suleiman. Washington D.C.: American Enterprise Institute. Clarke, Harold, and Moshe M. Czudnowski, eds. 1987. PoliticalElitesinAnglo-American Democracies.DeKaib: Northern Illinois University Press. Clarke, Harold, Jane Jenson, Lawrence LeDuc, and Jon H. Pammet. 1979. PoliticalChoicein Canada.Toronto: McGraw-Hill Ryerson. Clarke, Harold, and Marianne Stewart. 1998. The Decline of Parties in the Minds of Citizens. Annual ReviewofPoliticalScience1:357-78. Clarke, Harold, and Motoshi Suzuki. 1994. Partisan Dealignment and the Dynamics of Independence in the American Electorate, 1953-88. BritishJournalofPoliticalScience24 (1):57-77. Clarke, Harold D., Karl Ho, and Marianne C. Stewart. 2000. Major’s Lesser (Not Minor) Effects: Prime Ministerial Approval and Governing Party Support in Britain Since 1979. ElectoralStudies19:255-73. Clarke, Harold D., Jane Jenson, Lawrence LeDuc, and Jon H. Pammett. 1984. Absent Mandate:ThePoliticsofDiscontentin Canada.Toronto: Gage. Clarke, Harold D., and Marianne C. Stewart. 1995. Economic Evaluations, Prime Ministerial Approval and Governing Party Support: Rival Models Considered. BritishJournalof PoliticalScience25:145-70.

Conover, Pamela J. 1981. Political Cues and the Perception of Candidates. AmericanPolitics Quarter/y9:427-45.

Conover, Pamela J. 1988. Feminists and the Gender Gap. TheJournalofPolitics50 (4):985- 1010. Conover, Pamela Johnston, and Stanley Feldman. 1986. The Role of Inference in the Perception of Political Candidates. In PoliticalCognition,edited by R. R. Lau and D. 0. Sears. Hilisdale, NJ: Lawrence Eribaum Associates. Conover, PamelaJ., and Stanley Feldman. 1989. Candidate Perception in an Ambiguous World: Campaigns, Cues, and Inference Processes. AmericanJournalofPoliticalScience 33 (4):912-940.

Converse, Philip E. 1964. The Namre of Belief Systems in Mass Publics. In Ideologyand Discontent,edited by D. E. Apter. New York: Free Press. Converse, Philip E., and Georges Dupeux. 1966. De Gaulle and Eisenhower: The Public Image of the Victorious General. In ElectionsandthePoliticalOrder,edited by A. Campbell, P. E. Converse, W. E. Miller and D. E. Stokes. New York: John Wiley. Cottrell, Catherine A., Steven L. Neuberg, and Norman P. Li. 2006. What Do People Desire in Others? A Sociofunctional Perspective on the Importance of Different Valued Characteristics. JournalofPersonalityandSocialPsychologyforthcoming. Courtney, John C. 1973. TheSelectionofNationalPartyLeaders.Toronto: Macmillan.

203 Craig, Stephen C., Michael D. Martinez, and James G. Kane. 1999. The Structure of Political Competition: Dimensions of Candidate and Group Evaluation Revisited. Political Behavior21 (4):283-304.

Crewe, Ivor, and Martin Harrop, eds. 1986. PoliticalCommunications:TheGeneralElection Campaignof 1983.Cambridge: Cambridge University Press. Crewe, Ivor, and Martin Harrop, eds. 1989. PoliticalCommunications:TheGeneralElection Campaignof 1987.New York: Cambridge University Press. Crewe, Ivor, and Anthony King. 1994. Are British Elections Becoming More ‘Presidential’? In Electionsat HomeandAbroad:Essqysin Honorof7Warren E. Miller,edited by M. K. Jennings and T. E. Mann. Ann Arbor: University of Michigan Press. Crewe, Ivor, and Anthony King. 1994. Did Major Win? Did Kinnock Lose? Leadership Effects in the 1992 Election. In Labour’sLast Chance?The 1992ElectionandBçyond, edited by A. Heath, R. Jowell and J. Curtice. Aldershot: Dartmouth. Crewe, Ivor, Bo Sarlvik, and James Alt. 1977. Partisan Dealignment in Britain, 1964-1974. BritishJournal ofPoliticalScience7 (2):129-190. Curtice, John, and André Blais. 2001. Follow my Leader? A Cross-National Analysis of Leadership Effects in Parliamentary Democracies. In WorkingPaper Centre for Research Into Elections and Social Trends. Cutler, Fred. 2002. The Simplest Shortcut of All: Sociodemographic Characteristics and Electoral Choice. JournalofPolitics64 (2):466-490. Dalton, Russell, Scott Flanagan, and Paul Beck. 1984. ElectoralChangeinAdvancedIndustrial Democracies:RealignmentorDealignment?Princeton: Princeton University Press. Dalton, Russell, Ian McAllister, and Martin P. Wattenberg. 2000. The Consequences of Partisan Dealignment. In PartiesivithoutPartisans:PoliticalChangeinAdvancedIndustrial Democracies,edited by R. Dalton and M. P. Wattenberg. Oxford: Oxford University Press. Dayhoff, Signe A. 1983. Sexist Language and Person Perception: Evaluation of Candidates from Newspaper Articles. Sex Roles9 (4):1983.

De]li Carpini, Michael, and Scott Keeter. 1996. WhatAmericansKnowAbout Politicsand W4yit Matters.New Haven: Yale University Press. Dolan, Kathleen. 1998. Voting for Women in the ‘Year of the Woman’. AmericanJournalof PoliticalScience42 (1):272-293. Dow, Jay K., and James W. Endersby. 2003. Muhinomial Probit and Multinomial Logit A Comparison of Choice Models for Voting Research. ElectoralStudies23 (1):107-122. Downs, Anthony. 1957. An EconomicTheoryofDemocra’y.New York: Harper & Brothers. Enders, Craig K. 2001. The Performance of the Full Information Maximum likelihood Estimator in Multiple Regression Models with Missing Data. Educationaland P.ychologicalMeasurement61 (5):713-740.

204 Enders, Craig K. 2001. A Primer on Maximum Likelihood Algorithms Available for Use with Missing Data. StructuralEquationModeling8:128-141. Esaiasson, Peter, and Donald Granberg. 1993. Hidden Negativism: Evaluation of Swedish Parties and Their Leaders Under Different Survey Methods. InternationalJournalof PublicOpinionResearch5 (3):265-277. Evans, Geoffrey, and Robert Anderson. 2005. The Powerful Impact of Party Leaders: How Blair Cost Labour Votes (But Not Too Many). Paper read at EPOP, at University of Essex. Everitt, Joanna. 2003. Media in the Maritimes: Do Female Candidates Face a Bias? Atlantis 27 (2):90-98. Fenno, Richard. 1978. HomeStyle:HouseMembersin TheirDistricts.Boston: Little Brown. Fink, A. 1995. HowtoAsk Sureejuestions. Thousand Oaks, CA: Sage Publications.

Fiske, Susan T. 1986. Schema-based versus Piecemeal Politics. In PoliticalCognition:The 19th Annual CarnegieSjmposiumonCognition,edited by R. Lau and D. Sears. New Jersey: Lawrence Eribaum Associates. Foti, Roseanne, S. Fraser, and R. Lord. 1982. Effects of Leadership Labels and Prototypes on Perceptions of Political Leaders. JournalofAppliedPychology67:326-33. Franklin, Charles H. 1991. Eschewing Obfuscation? Campaigns and the Perception of U.S. Senate Incumbents. AmericanPoliticalScienceReview85 (4):1193-1214.

Funder, David C. 2001. Personality. Annual ReviewofPsjicbology52:197-221. Funk, Carolyn L. 1996. The Impact of Scandal on Candidate Evaluations: An Experimental test of the Role of Candidate Traits. PoliticalBehavior18 (1):1-24.

Funk, Carolyn L. 1999. Bringing the Candidate into Models of Candidate Evaluation. The Journal ofPolitics61 (3):700-720. Gant, Michael M., and William Lyons. 1993. Democratic Theory, Nonvoting, and Public Policy: The 1972-1988 Presidential Elections. AmericanPoliticsResearch21 (2):185-204. Gant, Michael M., and Lee Sigelman. 1985. Anti-Candidate Voting in Presidential Elections. PokEy18 (2):329-339. Gidengil, Elisabeth, André Blais, Richard Nadeau, and Neil Nevitte. 2000. Are Party Leaders Becoming More Important to Vote Choice in Canada? Paper read at Annual Meeting of the American Political Science Association, at Marriott Wardman Park, Washington DC. Gidengil, Elisabeth, André Blais, Richard Nadeau, and Neil Nevitte. 2003. Women to the Left? Gender Differences in Political Beliefs and Policy Preferences. In Womenand ElectoralPoliticsin Canada, edited by M. Tremblay and L. Trimble. Oxford: Oxford University Press. Gidengil, Elisabeth, André Blais, Neil Nevitte, and Richard Nadeau. 2000. Do Election Campaigns Prime Leadership? Evidence From Recent Canadian Elections. Paper read at ECPR 28th Joint Session Workshops, at Copenhagen.

205 Gidengil, Elisabeth, and Agniesaka Dobrzynska. 2003. Using a Rolling Cross-Section Design to Model Media Effects: The Case of Leader Evaluations in the 1997 Canadian Election. Paper read at Annual Meeting of the American Political Science Association. Gidengil, Elisabeth, and Joanna Everitt. 2000. Filtering the Female: Television News Coverage of the 1993 Canadian Leaders’ Debates. Women& Politics21 (4):105-131. Gidengil, Elisabeth, and Joanna Everitt. 2003. Conventional Coverage/Unconventional Politicians: Gender and Media Coverage of Canadian Leaders’ Debates, 1993, 1997, 2000. CanadianJournalofPoliticalScience36 (3):559-577. Gidengil, Elisabeth, Joanna Everitt, and Susan A. Banducci. 2006. Gender and Perceptions of Leader Traits: Evidence from the 1993 Canadian and 1999 New Zealand Elections. Paper read at Conference on Women and Leadership, at University of Toronto. Glass, David P. 1985. Evaluating Presidential Candidates: Who Focuses on Their Personal Attributes? ThePublicOpinionQuarter/y49 (4):517-534. Graetz, Brian, and Ian McAllister. 1987. Popular Evaluations of Party Leaders in the Anglo- American Democracies. In Politica/ElitesinAnglo-AmericanDemocracies,edited by H. D. Clarke and M. M. Czudnowski. DeKalb, Illinois: Northern Illinois University Press. Graetz, Brian, and Ian McAllister. 1987. Party Leaders and Election Outcomes in Britain, 1974-1983. ComparativePoliticalStudies19:484-507. Granberg, Donald. 1985. An Anomaly in Political Perception. ThePublicOpinionQuarter/y49 (4):504-516. Green, Donald, Bradley Palmquist, and Eric Schickler. 2002. PartisanHeartsandMinds: PoliticalPartiesandtheSocialIdentitiesofVoters.New Haven: Yale University Press. Gronke, Paul. 1997. Partisan Strength and Uncertain Presidential Evaluations. Paper read at American Political Science Association Annual Meeting, at Washington, D.C. Haddock, Geoffrey. 2003. Making a Party Leader Less of a Party Member: The Impact of Ambivalence on Assimilation and Contrast Effects in Political Party Attitudes. PoliticalPsjichology24 (4):769-780. Hamill, Ruth, Milton Lodge, and Frederick Blake. 1985. The Breadth, Depth, and Utility of Class, Partisan, and Ideological Schemata. AmericanJournalofPoliticalScience29:850- 870. Harman, Harry H. 1976. ModernFactorAnaysis. Chicago: University of Chicago Press. Hayes, Danny. 2004. Accentuating the Personal: Media Exposure, Political Sophistication, and Evaluation of Presidential Candidate Traits. JournalofPublicAffairs26:86-97. Hayes, Danny. 2005. Candidate Qualities Through a Partisan Lens: A Theory of Trait Ownership. AmericanJournalofPoliticalScience49 (4):908-923.

206 Hellweg, Susan A. 1995. Campaigns and Candidate Images in American Presidential Elections. In CandidateImagesandPresidentialElections,edited by K. L. Hacker. Westport: Praeger. Hilton, James L., and William von Hippel. 1996. On the Question of’Why?’:the Context- Dependent Function of Stereotypic Thinking. Annual ReviewofPsjchology47:237-271. Holbrook, Allyson L., Jon A. Krosnick, Penny S. Visser, Wendi L. Gardner, and John T. Cacioppo. 1999. The Formation of Attitudes Toward Presidential Candidates and Political Parties. Unpublished Manuscript. Hudson, John. 1984. Prime Ministerial Popularity in the UK: 1960-81. PoliticalStudies32:86- 97. Ikuo, Kabashima, and Imai Ryosuke. 2002. Evaluation of Party Leaders and Voting Behaviour-An Analysis of the 2000 General Election. SocialScienceJqpanJournal5 (1):85-96. Inglehart, Ronald, and Pippa Norris. 2000. The Developmental Theory of the Gender Gap: Women’s and Men’s Voting Behavior in Global Perspective. InternationalPolitical ScienceReview21 (4):441-63. Irvine, William P, and H Gold. 1980. Do Frozen Cleavages Ever Go Stale? The Bases of the Canadian and Australian Party Systems. BritishJournalofPoliticalScience10 (2):187-218. Irving, R.E.M., and W.E. Paterson. 1981. The West German Election of 1980: Continuity Preferred to Change. ParliamentaryAffairs34 (2):191-209. Irving, R.E.M., and W.E. Paterson. 1987. The West German General Election of 1987. ParliamentaryAffairs40 (3):333-356. Jenssen, Anders Todal, and Toni Aalberg. 2006. Party-leader Effects in Norway: A Multi- Methods Approach. ElectoralStudies25:248-269. Johnston, Richard. 1985. The Reproduction of the Religious Cleavage in Canadian Elections. CanadianJournalofPoliticalScience18 (1):99-113. Johnston, Richard. 2002. Prime Ministerial Contenders in Canada. In Leaders’Personalitiesand theOutcomesofDemocraticElections,edited by A. King. Oxford: Oxford University Press. Johnston, Richard. 2006. Party Identification: Unmoved Mover or Sum of Preferences? Annual ReviewofPoliticalScience9:329-351. Johnston, Richard. forthcoming. Survey Methodology. In O.ford HandbookofPolitical Methodology,edited byJ. Box-Steffensmeier, H. E. Brady and D. Collier: Oxford University Press.

Johnston, Richard, André Blais, Henry E. Brady, and Jean Crete. 1992. LettingthePeople Decide:Djnamicsofa CanadianElection.Stanford: Stanford University Press. Johnston, Richard, André Blais, Elisabeth Gidengil, and Neil Nevitte. 1996. TheChallengeof DirectDemocracy.Montreal & Kingston: McGill-Queen’s University Press.

207 Johnston, Richard, and Henry E. Brady. 2002. The Rolling Cross-Section Design. Electoral Studies21:283-95.

Johnston, Richard, Michael G. Hagen, and Kathleen Hall Jamieson. 2004. The2000 PresidentialElectionandtheFoundationsofParEyPolitics.Cambridge: Cambridge University Press. Jones, Philip, and John Hudson. 1996. The Quality of Political Leadership: A Case Study of John Major. BritishJournalofPoliticalScience26 (2):229-244. Jordan, Nehemiah. 1965. The ‘Asymmetry of ‘Liking’ and ‘Disliking’:A Phenomenon Meriting Further Reflection and Research. ThePublicOpinionQuartery29 (2):315-322. Kaase, Max. 1994. Is There Personalization in Politics? Candidates and Voting Behavior in Germany. InternationalPoliticalScienceReview15:211-30. Kahneman, Daniel, and Amos Tversky. 1979. Prospect Theory: An Analysis of Decision under Risk. Econometrica47 (2):263-292.

Kalton, Graham. 1983. IntroductiontoSurvejSampling.Newbury Park, CA: Sage. Kamakura, Wagner A., and Michel Wedel. 2000. Factor Analysis and Missing Data. Journal of MarketingResearch37:490-498.

Katz, Richard S. 2003. Political Data in 2002. EuropeanJournalofPoliticalResearch42:873-879. Katz, Richard S., and Ruud Koole. 1997. Political Data in 1997. EuropeanJournalofPolitical Research32:291-300. Katz, Richard S., and Ruud Koole. 1999. Political Data in 1998. EuropeanJournalofPolitical Research36:307-315. Katz, Richard S., and Ruud Koole. 2002. Political Data in 2001. EuropeanJournalofPolitical Research41:885-896. Kelley, Stanley Jr., and Thad W. Mirer. 1974. The Simple Act of Voting. AmericanPolitical ScienceReview68 (2):572-591. Kenney, PatrickJ., and Tom W. Rice. 1988. Presidential Prenomination Preferences and Candidate Evaluations. AmericanPoliticalScienceReview82 (4):1309-1319. Kemell, Samuel. 1978. Explaining Presidential Popularity. How Ad Hoc Theorizing, Misplaced Emphasis, and Insufficient Care in Measuring One’s Own Variables Refuted Common Sense and Led Conventional Wisdom Don the Path of Anomalies. AmericanPoliticalScienceReview72 (2):506-522. Key, V.0. 1996. TheReiponsibkElectorate.Cambridge, MA: Harvard University Press. Kinder, Donald R. 1978. Political Person Perception: The Asymmetrical Influence of Sentiment and Choice on Perceptions of Presidential Candidates. JournalofPersonality and SocialPycholo

208 Kinder, Donald R. 1983. Presidential Traits: Pilot Study Report to the 1984 NES Planning Committee and NES Board. Kinder, Donald R. 1986. Presidential Character Revisited. In PoliticalCognition,edited by R. R. Lau and D. 0. Sears. Hifisdale, NJ: Lawrence Erlbaum Associates. Kinder, Donald R., Robert P. Abelson, and Susan T. Fiske. 1979. Developmental Research on Candidate Instrumentation: Results and Recommendations. In R4ort Submittedto BoardofOverseers,NationalElectionStudies.Ann Arbor, Michigan: University of Michigan.

Kinder, Donald R., Gordon S. Adams, and Paul W. Gronke. 1989. Economics and Politics in the 1984 American Presidential Election. AmericanJournalofPoliticalScience33 (2):491-515. Kinder, Donald R., and Susan T. Fiske. 1986. Presidents in the Public Mind. In Political Psjchologp:ContemporayProblemsandIssues,edited by M. G. Hermann. San Francisco: Jossey-Bass. Kinder, Donald R., and Thomas E. Nelson. 1990. Experimental Investigations of Opinion Frames and Survey Responses. In A ReporttotheNationalElectionStudiesBoard.Ann Arbor, Michigan: University of Michigan. Kinder, Donald R., Mark D. Peters, Robert P. Abelson, and Susan T. Fiske. 1980. Presidential Prototypes. PoliticalBehavior2:315-37. King, Anthony, ed. 2002. Leaders’PersonalitiesandtheOutcomesofDemocraticElections.Oxford: Oxford University Press.

King, Anthony. 2002. Do Leaders’ Personalities Really Matter? In Leaders’Personalitiesandthe OutcomesofDemocraticElections,edited by A. King. Oxford: Oxford University Press. Kish, Leslie. 1965. SurvejSampling.New York: Wiley. Kitschelt, Herbert, and Anthony J. McGann. 1995. TheRadicalRzghtin WesternEurope:A ComparativeAna/ysis.Ann Arbor: The University of Michigan Press. Klingemann, Hans-Dieter, and Martin P. Wattenberg. 1992. Decaying Versus Developing Party Systems: A Comparison of Party Images in the United States and West Germany. BritishJournalofPoliticalScience22:131-49. Koole, Ruud, and Richard S. Katz. 1998. Political Data in 1997. EuropeanJournalofPolitical Research34:329-338.

Koole, Ruud, and Richard S. Katz. 2000. Political Data in 1999. EuropeanJournalofPolitical Research38:303-312.

Koole, Ruud, and Richard S. Katz. 2001. Political Data in 2000. EuropeanJournalofPolitical Research40:223-232.

Koole, Ruud, and Peter Malt. 1993. Political Data in 1992. EuropeanJournalofPoliticalResearch 24:361-367.

Koole, Ruud, and Peter Mair. 1994. Political Data in 1993. EuropeanJournalofPoliticalResearch 26:221-229.

209 Kropko, Jonathan. 2008. Choosing Between Multinomial Logit and Multinomial Probit Models for Analysis of Unordered Choice Data. Paper read at Midwest Political Science Association Annual Meeting, at Palmer House Hilton, Chicago. Krosnick, Jon A. 1999. Survey Research. Annual ReviewofPycholgy 50:537-67. Kuklinski, James H., and Paul J Quirk. 2001. Conceptual Foundations of Citizen Competence. PoliticalBehavior23 (3):285-311. Ladd, Jonathan, and Gabriel Lena. 2004. Emotions and Voting Behavior: A Critique. Paper read at Midwest Political Science Association Annual Meeting, at Chicago, Illinois. Lane, Robert E. 1978. Interpersonal Relations and Leadership in a “Cold Society”. ComparativePolitics10 (4):443-459.

Lau, Richard. 2003. Models of Decision-Making. In OxfordHandbookofPoliticalPychology, edited by D. Sears, L. Huddy and R. Jervis. Oxford: Oxford University Press.

Lau, Richard, and David Redlawsk. 1997. Voting Correctly. AmericanPoliticalScienceReview 91:585-598.

Lau, Richard, and David Redlawsk. 2006. How VotersDecide:InformationProcessingDuring ElectionCampaigns.New York: Cambridge University Press. Lau, Richard R. 1982. Negativity in Political Perception. PoliticalBehavior4:353-77. Lau, Richard R. 1986. Political Schemata, Candidate Evaluations, and Voting Behavior. In PoliticalCognition,edited by R. R. Lau and D. 0. Sears. Hilisdale, NJ: Lawrence Eribaum Associates. Lau, Richard R., and David 0. Sears. 1986. Social Cognition and Political Cognition: The Past, the Present, and the Future. In PoliticalCognition:The19thAnnual Carnegie Sjvmposiumon Cognition,edited by R. Lau and D. Sears. New Jersey: Lawrence Erlbaum Associates. Layer, Michael, and Kenneth A. Shepsie. 1990. Coalitions and Cabinet Government. AmericanPoliticalScienceReview84 (3):83-890. Layer, Michael, and Kenneth A. Shepsle. 1998. Events, Equilibria, and Government Survival.AmericanJournalofPoliticalScience42 (1):28-54. Lavine, Howard, and Thomas Gschwend. 2006. Issues, Party and Character: The Moderating Role of Ideological Thinking on Candidate Evaluation. BritishJournalofPoliticalScience 37:139-163. Lee, Sik-Yum, Xin-Yuan Song, and John C. K. Lee. 2003. Maximum Likelihood Estimation of Nonlinear Structural Equation Models with Ignorable Missing Data. Journalof EducationalandBehavioralStatistics28 (2):111-134. Letaring, Tera D., Shannon M. Wells, and David C. Funder. 2006. Quantity and Quality of Available Information Affect the Realistic Accuracy of PersonalityJudgment. Journal ofPersonalityandSocialP.’ychologyforthcoming.

210 Upset, Seymour Martin, and Stein Rokkan. 1967. Cleavage Structures, Party Systems and Voter Alignments: an Introduction”. In PartySjistemsand VoterAlignments:Cross- National Per.rpectives,edited by S. M. Upset and S. Rokkan. New York: The Free Press. Little, Roderickj.A., and Donald B. Rubin. 2002. StatisticalAnalysiswithMissingData. Hoboken, New Jersey: John Wiley & Sons, Inc. Liu, Chuanhal, and Donald B. Rubin. 1997. Maximum Likelihood Estimation of Factor Analysis Using the ECME Algorithm with Complete and Incomplete Data: Unpublished Paper. Lockerbie, Brad. 1992. Prospective Voting in Presidential Elections, 1956-1988. American PoliticsResearch20 (3):308-325. Lodge, Milton. 1981. Magnitude Scaling: Quantitative Measurement of Opinions. In QuantitativeApplicationsin theSocialSciences(a SageUniversi’yPaper).Newbury Park: Sage Publications. Lodge, Milton, and Ruth Hamill. 1986. A Partisan Schema for Political Information Processing. AmericanPoliticalScienceReview80:505-520.

Lodge, Milton, and Kathleen M. McGraw. 1995. PoliticalJudgment:StructureandProcess.Ann Arbor: University of Michigan Press. Lodge, Milton, Kathleen M. McGraw, and Patrick Stroh. 1989. An Impression-Driven Model of Candidate Evaluation. AmericanPoliticalScienceReview83 (2):399-419. Lodge, Milton, Marco R. Steenbergen, and Shawn Brau. 1985. The Responsive Voter: Campaign Information and the Dynamics of Candidate Evaluation. AmericanPolitical ScienceReview89 (2):309-26.

Long, J. Scott. 1997. RegressionModelsfor CategoricalandLimitedDependentVariables.Vol. 7, AdvancedQuantitativeTechniquesin theSocialSciences:Sage Publications. Long, J. Scott, and Jeremy Freese. 2005. RegressionModelsfor CategoricalOutcomesUsingStata. Second ed. College Station, TX: Stata Press. Lupia, Arthur. 1992. Busy Voters, Agenda Control, and the Power of Information. American PoliticalScienceReview86:390-404.

Luskin, Robert C. 1987. Measuring Political Sophistication. AmericanJourualofPoliticalScience 31 (4):856-899. Luskin, Robert C. 1990. Explaining Political Sophistication. PoliticalBehavior12 (4):331-361. Macrae, C. Neil, and Galen V. Bodenhausen. 2000. Social Cognition: Thinking Categorically About Others. Annual ReviewofPsychology51:93-120.

Marcus, George E., Russell W. Neuman, and Michael MacKuen. 2000. AffectiveIntelligenceand PoliticalJudgment.Chicago: University of Chicago Press. Markus, Gregory B. 1982. Political Attitudes During an Election Year: A Report on the 1980 NES Panel Study. AmericanPoliticalScienceReview76 (3):538-560.

211 Matthews,J. Scott. 2006. The Campaign Dynamics of Economic Voting. Doctoral Dissertation, Department of Political Science, University of British Columbia, Vancouver, BC. McAllister, Ian. 1996. Leaders. In ComparingDemocracies:Electionsand Votingin Global Per.rpective,edited by L. LeDuc, R. G. Niemi and P. Norris. Thousand Oaks: Sage Publications Inc. McAllister, Ian, and Clive Bean. 2005. Leaders, Social Infrastructure, the Economy or Iraq? Voting Behaviour in the 2004 Australian Federal Election. UnpublishedManuscrbt. McCann, James A. 1990. Changing Electoral Contexts and Changing Candidate Images During the 1984 Presidential Campaign. AmericanPoliticsQuarter!y18 (2):123-140. McCombs, Maxwell E., and Donald L. Shaw. 1972. The Agenda-Setting Function of Mass Media. PublicOpinionQuarter 36:176-87. McCurley, Carl, and JeffreyJ. Mondak. 1995. Inspected by #1184063113: The Influence of Incumbents’ Competence and Integrity in U.S. House Elections. AmericanJournalof PoliticalScience39 (4):864-885. McDermott, Monika L. 1997. Voting Cues in Low-Information Elections: Candidate Gender as a Social Information Variable in Contemporary United States Elections. American JournalofPoliticalScience41 (1):270-283. McDermott, Monika L. 1998. Race and Gender Cues in Low-Information Elections. Political ResearchQuarter/y51 (4):895-918.

McDermott, Momka L. 2005. Candidate Occupations and Voter Information Shortcuts. The JournalofPolitics67 (1):201-219. McDonald, Michael D., Silvia M. Mendes, and Myunghee Kim. 2007. Cross-temporal and cross-national comparisons of party left-right positions. ElectoralStudies26 (1):62-75. McGinniss, Joe. 1969. TheSellingofthePresident1968.New York: Trident Press. Mendelsohn, Matthew. 1993. Television’s Frames in the 1988 Canadian Election. Canadian JournalofCommunication19 (2):149-171. Mendelsohn, Matthew. 1994. The Media’s Persuasive Effects: The Priming of Leadership in the 1988 Canadian Election. CanadianJournalofPoliticalScience27:81-97. Mendelsohn, Matthew. 1996. The Media and Interpersonal Communications: The Priming of Issues, Leaders, and Party Identification. TheJournalofPolitics58 (1):112-25. Mendelsohn, Matthew, and Richard Nadeau. 1999. The Rise and Fall of Candidates in Canadian Election Campaigns. TheInternationalJournalofPress4 (2):63-76. Miller, Arthur H., and Warren Miller. 1975. Issues, Candidates and Partisan Divisions in the 1972 American Presidential Election. BritishJournalofPoliticalScience5 (4):393-434. Miller, Arthur H., and Warren E. Miller. 1976. Ideology in the 1972 Election: Myth or Reality-A Rejoinder. AmericanPoliticalScienceRenew70 (3):832-849.

212 Miller, Arthur H., and Martin P. Wattenberg. 1985. Throwing the Rascals Out: Policy and Performance Evaluations of Presidential Candidates, 1952-1980. AmericanPolitical ScienceReview79:359-72. Miller, Arthur H., Martin P. Wattenberg, and Oksana Malanchuk. 1986. Schematic Assessment of Presidential Candidates. AmericanPoliticalScienceReview80:521-40. Miller, Warren. 1991. Party Identification, Realignment, and Party Voting: Back to the Basics. TheAmericanPoliticalScienceReview85 (2):557-568. Miller, Warren E., andJ. Merrill Shanks. 1996. TheNewAmericanVoter.Cambridge, Mass.: Harvard University Press.

Mischel, Walter. 2004. Toward an Integrative Science of the Person. Annual Reviewof P.ychology55:1-22. Mondak, Jeffrey. 1993. Public Opinion and Heuristic Processing of Source Cues. Political Behavior15:167-192. Mughan, Anthony. 1993. Party Leaders and Presidentialism in the 1992 Election: A Post War Perspective. In BritishElectionsandPartiesYearbook,1993,edited by D. Denver, P. Norris, D. Broughton and C. Railings. Hemel Hempstead: Harvester Wheatsheaf. Mughan, Anthony. 2000. Mediaand thePresidentialiationofParliamentaryElections.Basingstoke: Paigrave. Mughan, Anthony. 2005. On the Conditionality of Leader Effects. Paper read at EPOP Annual Meeting, 9-11 September, at University of Essex. Nadeau, Richard, and André Blais. 1993. Explaining Election Outcomes in Canada. Canadian JournalofPoliticalScience26:775-90. Nadeau, Richard, and Matthew Mendelsohn. 1994. Short-Term Popularity Boost Following Leadership Change in Britain. ElectoralStudies13 (3):222-228. Nadeau, Richard, Richard G. Niemi, and Timothy Amato. 1996. Prospective and Comparative or Retrospective and Individual? Party Leaders and Party Support in Great Britain. BritishJournalofPoliticalScience26 (2):245-258. Nimmo, Dan, and Robert L. Savage. 1976. Candidatesand TheirImages.Pacific Palisades: Goodyear. O’Neill, Brenda. 1998. The Relevance of Leader Gender to Voting in the 1993 Canadian National Election. InternationalJournalofCanadianStudies17:105-130. Ohr, Dieter. 2004. Physical Attractiveness and Family Life as Determinants of Candidates’ Images and the Vote. Paper read at Annual Meeting of the International Communication Association, at New Orleans. Ohr, Dieter, and Henrik Oscarsson. 2003. Leader Traits, Leader Image and Vote Choice. Paper read at Meeting of the European Consortium for Political Research, at Marburg, Germany. Page, Benjamin I. 1978. ChoicesandEchoesin PresidentialElections.Chicago: University of Chicago Press.

213 Page, Benjamin I., and Calvin C. Jones. 1979. Reciprocal Effects of Policy Preferences, Party Loyalties, and the Vote. AmericanPoliticalScienceReview73 (4):1071-1089. Pancer, S. Mark, Steven D. Brown, and Cathy Widdis Barr. 1999. Forming Impressions of Political Leaders: A Cross-National Comparison. PoliticalPsjichology20 (2):1999. Peterson, David A.M. 2004. Certainty or Accessibility: Attitude Strength in Candidate Evaluations. AmericanJournalofPoliticalScience48 (3):513-520. Petrocik, John R. 1996. Issue Ownership in Presidential Elections, with a 1980 Case Study. AmericanJournalofPoliticalScience40 (3):825-850. Petrocik, John R., William L. Benoit, and GlennJ. Hansen. 2003. Issue Ownership and Presidential Campaigning, 1952-2000. PoliticalScienceuarter!y 118 (4):599-626. Pierce, Roy. 1995. ChoosingtheChiçftPresidentialElectionsin Franceandthe UnitedStates.Ann Arbor: University of Michigan Press.

Pierson, Paul. 2004. Politicsin Time:History,Institutions,andSocialAna!ysis.New Jersey: Princeton University Press. Plutzer, Eric, and John F. Zipp. 1996. Identity Politics, Partisanship, and Voting for Women Candidates. ThePublicOpinionQuarter/y60 (1):30-57. Popkin, Samuel L. 1991. TheReasoningVoter.Chicago: University of Chicago Press. Popkin, Samuel L.,John W. Gorman, Charles Phillips, and Jeffrey A. Smith. 1976. Comment: What Have You Done for Me Lately? Toward an Investment Theory of Voting. AmericanPoliticalScienceReview70 (3):779-805. Price, Vincent, and John Zaller. 1993. Who Gets the News? Alternative Measures of News Reception and Their Implications for Research. ThePublicOpinion,Quartery57 (2):133-164. Rabinowitz, George, and Stuart MacDonald. 1989. A Directional Theory of Voting. American PoliticalScienceReview83:93-122. Rahn, Wendy M. 1993. The Role of Partisan Stereotypes in Information Processing about Political Candidates. AmericanJournalofPoliticalScience37 (2):472-496. Rahn, Wendy M.,John H. Aldrich, Eugene Borgida, and John L. Sullivan. 1990. A Social- Cognitive Model of Candidate Appraisal. In InformationandDemocraticProcesses,edited byJ. A. Ferejohn andJ. H. Kuklinski. Urbana: University of lilinois Press. Rahn, Wendy M.,Jon A. Krosnick, and Marijke Breuning. 1994. Rationalization and Derivation Processes in Survey Studies of Political Candidate Evaluation. American JournalofPoliticalScience38 (3):582-600. Ray,John J. 1973. Conservatism, Authoritarianism and Related Variables a Review and Empirical Study. In TheP’ychologyofConservatism,edited by G. D. Wilson. New York: Academic Press.

Riggle, Ellen D., Victor C. Ottati, Robert S. Wyer, James H. Kuklinski, and Norbert Schwarz. 1992. Bases of Political Judgments: The Role of Stereotypic and Nonstereotypic Information. PoliticalBeharior14 (67-87).

214 Rosenberg, Shawn W., Lisa Bohan, Patrick McCafferty, and Kevin Harris. 1986. The Image and the Vote: The Effect of Candidate Presentation on Voter Preference. American JournalofPoliticalScience30:108-27. Roy, Jason. 2007. When Less is More: No Information Versus Low Information and the Application of Heuristics in Canadian Context. Paper read at Canadian Political Science Association Annual Meeting, May 30-June 1, at Saskatoon, Canada. Rummel, R.J. 1970. AppliedFactorAna!ysis.Evanston: Northwestern University Press. Rusk, Jerrold G., and Herbert F. Weisberg. 1972. Perceptions of Presidential Candidates: Implications for Electoral Change. MidwestJournalofPoliticalScience16:388-410. Sanders, David, Harold D. Clarke, Marianne C. Stewart, and Paul Whiteley. 2006. Does Mode Matter for Modelling Political Choice? Evidence from the 2005 British Election Study. Unpublished Manuscript. Sapiro, Robert Y, and Harpreet Mahajan. 1986. Gender Differences in Policy Preferences: A Summary of Trends from the 1960 to the 1980s. ThePublicOpinionQuarter/y50 (1):42- 61. Sapiro, Virginia, and Pamela Johnston Conover. 1997. The Variable Gender Basis of Electoral Politics: Gender and Context in the 1992 US Election. BritishJournalof PoliticalScience27 (4):497-523.

Savoie, Donald J. 1999. GoverningFromtheCenfre:TheConcentrationofPowerin CanadianPolitics. Toronto: University of Toronto Press. Schumpeter, Joseph A. 1943. Capitalism,SocialismandDemocracy.London: Allen & Unwin. Schwara, Norbert. 1998. Warmer and More Social: Recent Developments in Cognitive Psychology. Annual ReviewofSociology24:239-264. Sears, David 0., and Richard R. Lau. 1979. Personal Impact, Governmental Performance, and Policy Attitudes: A Preliminary Analysis of the 1979 CPS Pilot Study. In Reportto theNationalElection Studj Board.Ann Arbor, Michigan: University of Michigan. Shabad, Goldie, and Kristi Andersen. 1979. Candidate Evaluations by Men and Women. The PublicOpinionQuarter/y43 (1):18-35.

Shapiro, Michael J. 1970. Discovering Interviewer Bias in Open-Ended Survey Responses. PublicOpinionQuarter!y34 (3):412-415. Sides, John. 2006. The Origins of Campaign Agendas. BritishJournalofPoliticalScience36:407- 436. Sigelman, Carol, Lee Sigelman, Barbara Walkosz, and Michael Nitz. 1995. Black Candidates, White Voters: Understanding Racial Bias in Political Perceptions. AmericanJournalof PoliticalScience39 (1):243-265. Sigelman, Lee, and Michael M. Gant. 1989. Anticandidate Voting in the 1984 Presidential Election. PoliticalBehavior11:81-92. Simon, Adam F. 2002. TheWinningMessage:CandidateBehavior,CampaignDiscourse,and Democray.Cambridge: Cambridge University Press.

215 Sniderman, Paul, Richard Brody, and Philip Tetlock. 1991. Reasoningand Choice:Explorationsin PoliticalPsjvchology.New York: Cambridge University Press. Stewart, David K., and R. K. Carty. 1993. Does Changing the Party Leader Provide an Electoral Boost? A Study of Canadian Provincial Parties: 1960-1992. CanadianJournal ofPoliticalScience26:313-30. Stewart, Marianne C., and Harold D. Clarke. 1992. The (Un)Importance of Party Leaders: Leader Images and Party Choice in the 1987 British Election. JournalofPolitics54:447- 70. Stokes, Donald E. 1966. Some Dynamic Elements of Contests for the Presidency. American PoliticalScienceReview60:19-28. Stokes, Donald E., Angus Campbell, and Warren E. Miller. 1958. Components of Electoral Decision. AmericanPoliticalScienceReview52:367-87. Sudman, Seymour. 1983. Applied Sampling. In HandbookofSurvejResearch,edited by P. H. Rossi, J.D. Wright and A. B. Anderseon. New York: Academic Press. Sullivan, Denis G., and Roger D. Masters. 1988. “Happy Warriors”: Leaders’ Facial Displays, Viewers’ Emotions, and Political Support. AmericanJournalofPoliticalScience32 (2):345-368. Sullivan, John L.,John H. Aldrich, Eugene Borgida, and Wendy M. Rahn. 1990. Candidate Appraisal and Human Nature: Man and Superman in the 1984 Election. Political Pychology3:459-84.

Tate, Katherine. 1993. FromProtesttoPolitics:TheNewBlackVotersinAmericanElections. Cambridge: Harvard University Press. Thelen, Kathleen, and Sven Steinmo. 1992. Historical Institutionalism in Comparative Politics. In StructuringPolitics:HistoricalInstitutionalismin ComparativeAnavsis, edited by S. Steinmo, K. Thelen and F. Longstreth. New York: Cambridge University Press. Tourangeau, Roger, and Kenneth A. Rasinski. 1988. Cognitive Processes Underlying Context Effects in Attitude Measurement. PychologicalBulletin103 (3):299-314. Traugott, Mike. 1990. Report on Pilot Study Items. Ann Arbor, Michigan: Center for Political Studies! Institute for Social Research!University of Michigan. Van Biezen, Ingrid, and Richard S. Katz. 2005. Political Data in 2004. EuropeanJournalof PoliticalResearch44:919-928. van der Brug, Wouter. 2004. Issue Ownership and Party Choice. ElectoralStudies23:209-233. van der Brug, Wouter, and Anthony Mughan. 2007. Charisma, Leader Effects and Support for Right-Wing Populist Parties. PartyPolitics13 (1):29-51. van der Eijk, Cees, Wouter van der Brug, Martin Kwh, and Mark Franklin. 2006. Rethinking the Dependent Variable in Voting Behavior: On the Measurement and Analysis of Electoral Utilities. ElectoralStudies25:424-447.

216 Vetter, Angelika, and Oscar W. Gabriel. 1998. Candidate Evaluations and Party Choice in Germany, 1972-94: Do Candidates Matter? In StabilityChangein GermanElections:How ElectoratesMerge,Converge,or Collide,edited by C. J.Anderson and C. Zelle. Westport: Praeger.

Wattenberg, Martin P. 1991. TheRiseofCandidate-CenteredPolitics:President.ialElectionsofthe 1980s.Cambridge: Harvard University Press. Wattenberg, Martin P. 2006. Elections: Reliability Trumps Competence: Personal Attributes in the 2004 Presidential Election. PresidentialStudiesQuarter’y36 (4):705-713. Weisberg, Herbert F., and Arthur H. Miller. 1980. Evaluation of the Feeling Thermometer. In A ReporttotheNationalElectionStudjiBoardBasedonDatafromthe 1979PilotStudji. Ann Arbor, Michigan: University of Michigan. Weisberg, Herbert F., and Jerrold G. Rusk. 1970. Dimensions of Candidate Evaluation. AnzericanPoliticalScienceReview64:1167-85. Western, Bruce. 1998. Causal Heterogeneity in Comparative Research: A Bayesian Hierarchical Modelling Approach. AmericanJournalofPoliticalScience42 (4):1233-1259. Western, Bruce, and Filiz Garip. 2005. Bayesian Analysis of Comparative Survey Data. UnpublishedManuscrbt. Williams, Philip, and Graham K. Wilson. 1977. The 1976 Election and the American Political System. PoliticalStudies25 (2):182-200. Wilson, Graham. 1985. The American Election Campaign. PoliticalQuarter/y.47-55. Winham, Gilbert R., and Robert B. Cunningham. 1970. Party Leader Images in the 1968 Federal Election. CanadianJournalofPoliticalScience3 (1):37-55. Worlund, Ingemar. 1992. The Swedish Parliamentary Election of September 1991. ScandinavianPoliticalStudies15 (2):135-143. Wright, Vincent. 1989. TheGovernmentandPoliticsofFrance.London: Unwin Hyman. Yuan, Ke-Hai, Linda L. Marshall, and Peter M. Bender. 2002. A Unified Approach to Exploratory Factor Analysis with Missing Data, Nonnormal Data, and in the Presence of Outliers. Psjichometrika67 (1):95-122.

Zajonc, R.B. 1980. Feeling and Thinking: Preferences Need No Inferences. Anzerican Psychologist35 (2):151-175. Zaller, John. 1985. Pre-Tesling Information Items on the 1986 NES Pilot Survey: Report to the National Election Study Board of Overseers. Zaller, John. 1990. Experimental Tests of the Question-Answering Model of the Mass Survey Response. In ReporttotheNationalElectionStudiesBoardofOverseers.Ann Arbor, Michigan: University of Michigan. Zaller, John. 1992. TheNature and OnginsofMass Opinion.Cambridge: Cambridge University Press.

217 Zaller, John, and Stanley Feldman. 1987. Frame of Reference and the Survey Response. Zipp, John F., and Eric Plutzer. 1985. Gender Differences in Voting for Female Candidates: Evidence From the 1982 Election. ThePub/icOpinion,Quarter/y49 (2):179-197.

218 APPENDICES

APPENDIX I: DATA, QUESTION WORDING, AND CODING INFORMATION

Partisanship

Canada, 1968 Generally speaking, do you usually think of yourself as Liberal, Conservative, NDP, Creditiste, Social Credit, Union Nationale, or what?

Canada, 1984 Thinking of federal politics, do you usually think of yourself as a liberal, Progressive Conservative, NDP or what?

Canada, 1988 Thinking of federal politics, do you usually think of yourself as a Liberal, Conservative, NDP, or none of these?

Canada, 1993, 1997 Thinking of federal politics, do you usually think of yourself as a liberal, Conservative, NDP, (Reform/Bloc Quebecois) or none of these?

Canada,2000 Respondents were randomly assigned to one of two party identification questions. Responses to both questions were combined to create the PID variable. • Thinking of federal politics, do you usually think of yourself as a liberal, Bloc Quebecois, Alliance, Conservative, NDP, or none of these? • Generally speaking, in federal politics, do you usually think of yourself as a Liberal, Bloc Quebecois, Alliance, Conservative, NDP, or do you usually think of yourself as not having a general preference?

Canada,2004, 2006 Respondents were randomly assigned to one of two party identification questions. Responses to both questions were combined to create the PID variable. In federal politics, do you usually think of yourself as a liberal, Conservative, NDP, Bloc Quebecois or none of these? • In federal politics, do you usually think of yourself as a liberal, Conservative, NDP, Bloc Quebecois, another party, or no party?

219 UnitedStates, 1972-2004 (NationalElectionStudj) Generally speaking, do you usuallythink of yourself as a Republican, a Democrat, an Independent, or what?

UnitedStates,2000 (NationalAnnenbeigElectionStudj) Generally speaking, do you usually think of yourself as a Republican, a Democrat, an independent, or something else?

Britain, 1983, 1987, 1992, 1997, 2001 Generally speaking, do you think of yourself as a Conservative, Labour, Liberal, Social Democrat, (Nationalist/Plaid Cymru), or what?

Australia, 1987, 1993, 1996, 1998, 2001, 2004 Generally speaking, do you usually think of yourself as Liberal, Labor, National, or what?

New Zealand, 1999 Generally speaking, do you usually think of yourself as Labour, National, NZ First, Alliance, Act or some other, or don’t you think of yourself in this way?

New Zealand,2002 Generally speaking, do you usually think of yourself as National, Labour, Act, Greens, New Zealand First, or some other, or don’t you usually think of yourself in this way?

Germanj, 1980, 1987 Many people in the federal republic lean toward a particular party for a long time, although they may occasionally vote for a different party. How about you? Do you in general lean to ward a particular party? If so, which one?

Sweden,1988 Many people consider themselves as supporters of a specific party. But there are many others who don’t feel such an attachment to any of the parties. Do you generally think of yourself as a supporter of for example, the People’s Party, Social Democrats, Moderate Party, Vpk, Green Party or Kds? Or do you feel no such attachment to any of the parties? Which party do you like best?

Sweden,1991 Many people consider themselves adherents of a specific party. But there are also many others who do not have such an attachment to any of the parties. Do you usually think of yourself as, for example, a folkpartist, socialdemocrat, moderat, centerpartist, vansterpartist, miljopartist, kds:are or nydemocrat? Or do you not have such an attachment to any of the parties? Which party do you like best?

For inclusion in the analyses, based on these questions, dummy variables were created to capture partisanship for the following categories of parties: Centre-Left, Conservative, Left, Other, and None.

220 Ideology

Canada, 1968 Now I have a different kind of question for you. I’m going to show you some word pairs. Each pair is separated by seven boxes like this. For example, if you think that the particular political party is very modem, you would put a check mark in the box on the right end of the scale. If you think it is very out-of-date, you would check the box on the left end of the scale. Or you might rate it somewhere between these two extremes. (Respondent is asked to check where the “ideal party” would be located on a 7 point left/right scale)

Canada, 1984 For the next few questions I would like you to use this scale which goes from left to right, with “1” being the most to the left and “7” being the most to the right. When you think of your own political opinions, where would you put yourself on this scale?

Canada, 1988 Sometimes in Canada people use the labels ‘left’or ‘left-wing’and ‘tight’or ‘right-wing’to describe political parties, politicians and political ideas. Do you sometimes use these labels? (Y/N) if Yes: Where would you place yourself: very much to the left, somewhat to the left, in the centre, somewhat to the right, or very much to the right?

Canada, 1997 In politics people sometimes talk of left and right. Where would you place yourself on the scale below? (0-10 scale)

Canada, 2000 Some people talk about political parties being on the LEFT or the RIGHT. And finally,you personally. Would you say: on the left, on the right, in the centre, not sure.

Canada,2004, 2006 In politics people sometimes talk of left and right. Where would you place yourself on the scale below? (0-10 scale)

UnitedStates, 1972-2000 (NES) We hear a lot of talk these days about liberals and conservatives. Here is (:1972/1974: I’m going to show you) a 7-point scale on which the political views that people might hold are arranged from extremely liberal to extremely conservative. Where would you place yourself on this scale, or haven’t you thought much about this?

UnitedStates,2004 (NES) Please tum to page 17 of the booklet. In politics people sometimes talk of left and right. Where would you place yourself on a scale from 0 to 10 where 0 means the left and 10 means the right?

221 UnitedStates,2000 (NAES) Generally speaking, would you describe your political views as very conservative, conservative, moderate, liberal, or very liberal?

Britain, 1983 Respondents’ Left/Right self-placement, on -10 to + 10 scale.

Britain, 1997 In politics people sometimes talk of left and right. Where would you place yourself on a scale from 0 to 10, where 0 means the left and 10 means the right?

Britain,2001 In politics, people sometimes talk of left and right. Using the scale from 0 to 10, where would you place <5 parties>. And finally, where would you place yourself?

Australia, 1987 In political matters, people talk about the ‘left’ and the ‘right’. Generally speaking, where would you place your views on the scale? (1-10 scale)

Australia, 1993 In politics, people talk about the ‘left’and the ‘right’. Generally speaking, where would you place your views on the left-right scale? (1-10 scale)

Australia, 1996, 1998, 2001, 2004 In politics, people sometimes talk about the ‘left’ and the ‘right’.Where would you place yourself on a scale from 0 to 10, where 0 means the left and 10 means the right?

New Zealand,1999, 2002 In politics, people sometimes talk about the ‘left’ and the ‘right’.If you can, where would you place yourself on a scale from 0 to 10, where 0 means the most left and 10 means the most right?

Gerwanji,1980 There are various terms which are frequently used when talking about the political parties, such as “left” and “right.” Please mark this scale where you would place the SPD, the CDU, the FDP, the CSU, and the Greens. And where would you place the party which would be the ideal party for you? (1-11 scale)

Germanj,1987 There are various terms which are frequently used when talking about the political parties, such as “left” and “right.” Please mark this scale where you would place the SPD, the CDU, the FDP, the CSU, and the Greens. And now we would like to know from you where you would place yourself on the scale. (1-11 scale)

222 Sweden,1988 The parties are sometimes thought of as being ordered from left to right. On this card I have a kind of scale. I would like you to place the political parties on the scale. The further to the left you think a party is, the lower the number. The more on the right you think a party is, the higher the number. Where on the scalewould you place yourself?

Sweden,1991 The parties are sometimes thought of as being ordered from left to right. On this card I have a kind of scale. I would like you to place the political parties on the scale. The further to the left you think a party is, the lower the number. The further to the right you think a party is, the higher the number. Where on the scale would you place yourself?

All ideology variables coded on common 0-1 scale, where 0 represents most right-wing and I represents most left-wing.

Issue Attitudes: Major Campaign Issue

Canada, 1968 There has been quite a bit of talk recently about the possibility of Quebec separating from the rest of Canada and becoming an independent country. Are you in favour of separation or opposed to it? Please tell me whether you are strongly in favour of separation, slightlyin favour, undecided, slightlyopposed, or strongly opposed to separation.

Canada, 1984 Many people in the Federal Government are dishonest. Strongly agree, somewhat agree, disagree somewhat, strongly disagree, neither agree nor disagree, no opinion.

Canada,1988 As you know, the Mulroney government has reached a Free Trade Agreement with the United States. All things considered, do you support the agreement or do you oppose it? (support/oppose/don’t know)

Canada, 1993 Respondents were randomly assigned to one of two questions regarding the deficit. Responses to both questions were incorporated into the measure. • On the deficit, which comes closest to your own view? One: we must reduce the deficit even if that means cutting programmes. Or Two: Governments must maintain programmes even if that means continuing to run a deficit. On the deficit, which comes closest to your own view? One: governments must maintain programmes even if that means continuing to run a deficit. Or Two: We must reduce the deficit even if that means cutting programmes.

Canada, 1997 How much do you think should be done for Quebec: MORE, LESS or ABOUT THE SAME as now?

223 Canada,2000 How much do you think should be done for Quebec: much more, somewhat more, about the same as now, somewhat less, much less?

Canada,2004, 2006 To accept contributions or issue tax receipts, local riding associations of political parties must now register with Elections Canada and file yearly financial reports. Is this: a good thing, a bad thing, or do you have no opinion on this?

NES 1972 Do you think we did the right thing in getting into the fighting in Vietnam or should we have stayed out? Yes, did the right thing/No, should have stayed out/Depends.

NES 1976 Shortly after taking office, President Ford pardoned Richard Nixon for any wrong-doings he may have committed while he was President. Do you think that Ford should have pardoned Nixon? Yes/No.

NES 1980 Do you approve or disapprove of Jimmy Carter’s handling of the crisis brought about by the taking of Americans as hostages in Iran? Strongly approve/not strongly approve/not strongly disapprove/strongly disapprove.

NES 1984 Some people believe that we should spend much less money for defense. Others feel that defense spending should be greatly increased. Where would you place yourself on this scale, or haven’t you thought much about this? (7 point scale)

NES 1988 There is much concern about the rapid rise in medical and hospital costs. Some people feel there should be a government insurance plan which would cover all medical and hospital expenses for everyone. Others feel that all medical expenses should be paid by individuals, and through private insurance plans like Blue Cross or other company paid plans. Where would you place yourself on this scale, or haven’t you thought much about this? (7 point scale)

NES 1992, 1996 Next, I am going to ask you to choose which of the two statements I read comes closer to your own opinion. You might agree to some extent with both, but we want to know which one is closer to your views. One, we need a strong government to handle today’s complex economic problems; or two, the free market can handle these problems without government being involved? Strong government/Free market/Both, Depends.

224 NES 2000 This country would have many fewer problems if there were more emphasis on traditional family ties. (Do you agree strongly, agree somewhat, neither agree nor disagree, disagree somewhat, or disagree strongly with this statement?)

NES 2004 What about war on terrorism. Should federal spending on the war on terrorism be increased, decreased, or kept about the same?

NAES 2000 Trying to stop discrimination against homosexuals—should the federal government do more about this, the same as now, less, or nothing at all?

Britain, 1983 Which of the three statements on this card comes closest to your own views on the Common Market? Britain should leave the Common Market without further ado. Britain should stay in the Common Market provided we can get better terms. Britain should stay in the Common Market anyway.

Britain, 1987 Please use this card to say whether you think the government should or should not do each of the following things, or doesn’t it matter either way? Spend less on defense?

Britain, 1992 Please use this card to say whether you think the government should or should not do each of the following things, or doesn’t it matter either way? Put more money into the National Health Service?

Britain, 1997 And here are three statements about the future of the pound in the European Union. Which one comes closest to your view? Replace the pound by a single currency. Use both the pound and a new European currency in Britain. Keep the pound as the only currency for Britain.

Britain,2001 Thinking of the Single European Currency, which of the following statements on this card would come closest to your own view? Definitely join as soon as possible. Join if and when the economic conditions are right. Stay out for at least the next four or five years. Rule out on principle.

Australia, 1987 Pleas say whether you agree or disagree with each of these statements. Income and wealth should be distributed towards ordinary working people. Strongly agree, agree, not sure, disagree, strongly disagree.

225 Australia, 1993 Here are some statements about economic issues. Please saywhether you strongly agree, agree, disagree, or strongly disagree with each statement. Income and wealth should be redistributed towards ordinary working people.

Australia,1996 Please say whether you strongly agree, agree, disagree, or strongly disagree with each statement. Income and wealth should be redistributed towards ordinary working people.

Australia, 1998 Please say whether you strongly agree, agree, disagree, or strongly disagree with each statement. Income and wealth should be redistributed towards ordinary working people.

Australia, 2001 Do you think the number of immigrants allowed into Australia nowadays should be reduced or increased? Increased a lot, increased a little, remain about the same as it is, reduced a little, reduced a lot.

Australia, 2004 Do you think that the government should spend more or spend less on defence? Send much more on defence, spend some more on defence, about right at present, spend less on defence, spend a lot less on defence.

New Zealand, 1999 Listed below are various areas of government spending. Please show whether you would like to see more or less government spending in each area. Remember that if you say “more” or “much more”, it might require a tax increase to pay for it. Much more, more, same as now, less, much less, don’t know.

New Zealand,2002 Generally, do you think it should be or should not be the government’s responsibility to provide or ensure free health care for everyone? Definitely should, should, shouldn’t, definitely shouldn’t, don’t know.

Germanj,1980 Now I would be interested to know what importance some problems have for you personally. Here is a ladder: on the top rung would be the problem which is most important to you. At the bottom would be a problem which may be significant otherwise but which has no importance for you personally. Please show me on the ladder how important --- is for you. SecureJobs (1-10 scale)

Germanj,1987 Here we have put together several tasks and goals which are being talked about in the Federal Republic. Please tell us for each of these issues whether you personally consider it to be very important, important, not so important, or completely unimportant. Fight unemployment.

226 Sweden,1988 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairly good proposal, neither good nor bad proposal, fairly bad proposal, very bad proposal. Reduce the public sector.

Sweden,1991 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairly good proposal, neither good nor bad proposal, fairly bad proposal, very bad proposal. Reduce the public sector.

All issue attitudes coded on common 0-1 scale, where 0 represents most conservative/reactionary perspective and I represents most liberal/progressive perspective.

Issue Attitudes: Taxes Versus Spending Canada, 1968 In deciding what to do about voting in this election, was medicare extremely important to you, very important, fairly important, or not too important?

Canada, 1984 Here are some opinions that different people hold. As I read each statement, I would like you to tell me whether you agree or disagree with it, and how much. However, tell me if you have no opinion on the matter. It is not the responsibility of government to ensure jobs for unemployed Canadians. Strongly agree/Somewhat agree/Disagree somewhat/Strongly Disagree/Neither Agree nor Disagree.

Canada, 1988 Most experts argue that if we want more government services we must increase taxes and if we want lower taxes we must reduce services. If you HAD TO CHOSE, should the level of taxes and services be MUCH HIGHER, SOMEWHAT HIGHER, ABOUT THE SAME, SOMEWHAT LOWER, or MUCH LOWER than now?

Canada, 1993 Respondents were randomly assigned to one of three questions. Responses to all three questions were combined to create the taxes versus spending variable. • Would you personally be willing to pay higher taxes to maintain our social programmes? • Would you personally be willing to pay higher taxes to maintain our social programmes such as health care? • Would you personally be willing to pay higher taxes to maintain our social programmes such as welfare?

227 Canada, 1997 We face tough choices. Cutting taxes means cutting social programmes and improving social programmes means increasing taxes. If you had to choose, would you cut taxes, increase taxes, or keep taxes as they are?

Canada,2000 How would YOU spend the federal budget surplus? Would your FIRST priority be improving social programs, cutting taxes, or reducing the debt?

Canada,2004, 2006 And now taxes. Keep in mind that cutting taxes means spending less in some areas. Should personal income taxes be increased, decreased, or kept about the same as now?

NES, 1972, 1976, 1980 Some people feel that the government in Washington should see to it that every person has a job and a good standard of living. Others think the government should just let each person get ahead on his/their own. And of course, some other people have opinions somewhere in between. Where would you place yourself on this scale, or haven’t you thought much about this? (7 point scale)

NES, 1984, 1988, 1992, 1996, 2000 Some people think the government should provide fewer services, even in areas such as health care and education, in order to reduce spending. Other people feel that it is important for the government to provide many more services even if it means an increase in spending. Where would you place yourself on this scale, or haven’t you thought much about this? (7 point scale)

NES, 2004 Do you favour cuts in spending on domestic programs like medicare, education and highways in order to cut the taxes paid by ordinary Americans? Yes/No.

NAES, 2000 Which do you personally think is more important, cutting taxes or strengthening the Social Security system?

Britain, 1983 Please look at this card. People who are convinced that we should put up taxes a lot and spend much more on health and social services will put a tick in the last box on the left, while those who are convinced that we should cut taxes a lot and spend much less on health and social services will put a tick in the last box on the right. So, just as before, people who hold views that come somewhere in between those two positions will tick a box somewhere along here or somewhere along here. First, would you tick the box anywhere along the scale that comes closest to your own views about taxes and government spending? (20 point scale)

228 Britain, 1987 Please look at this card. Some people feel that government should put up taxes a lot and spend much more on health and social services. These people would put themselves in Box A. Other people feel that government should cut taxes a lot and spend much less on health and social services. These people would put themselves in Box K. And other people have views somewhere in-between, along here or along here. In the first row of boxes, please tick whichever box comes closest to your own views about taxes and government spending? (10 point scale)

Britain, 1992 Please look at this card. Some people feel that government should put up taxes a lot and spend much more on health and social services. These people would put themselves in Box A. Other people feel that government should cut taxes a lot and spend much less on health and social services. These people would put themselves in Box K. And other people have views somewhere in-between, along here or along here. In the first row of boxes, please tick whichever box comes closest to your own views about taxes and government spending? (11 point scale)

Britain, 1997 Suppose the government had to choose between these three options, which do you think it should choose? Reduce taxes and spend less on health, education and social benefits, or keep taxes and spending on these services at the same level as now, or increase taxes and spend more on health, education and social benefits?

Britain, 2001 On a scale from 0 to 10, where 0 means government should cut taxes a lot and spend much less on health and social services, and 10 means government should raise taxes a lot and spend much more, where would you put the views of yourself?

Australia, 1987 If government had a choice about reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

Australia, 1993 If the government had a choice between reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

Australia, 1996 If the government had a choice between reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

229 Australia, 1998 If the government had a choice between reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

Australia, 2001 If the government had a choice between reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

Australia, 2004 If the government had a choice between reducing taxes or spending more on social services, which do you think it should do? Strongly favour reducing taxes, mildly favour reducing taxes, it depends, mildly favour spending more on social services, strongly favour spending more on social services.

New Zealand, 1999 ONE represents the view that the government should reduce taxes, and SEVEN the view that there should be a tax increase so government can spend more money on health and education. Where would you place your view? (1-7 scale)

New Zealand,2002 ONE represents the view that the government should reduce taxes and people should pay more for their own health and education, and SEVEN the view that there should be a tax increase so the government can spend more money on health and education. Where would you place your view? (1-7 scale)

Germanj, 1980 Now I would be interested to know what importance some problems have for you personally. Here is a ladder: on the top rung would be the problem which is most important to you. At the bottom would be a problem which may be significant otherwise but which has no importance for you personally. Please show me on the ladder how important --- is for you. Reducing Taxes. (1-10 scale)

Germany,1987 Here we have put together several tasks and goals which are being talked about in the Federal Republic. Please tell us for each of these issues whether you personally consider it to be very important, important, not so important, or completely unimportanL Reduce Taxes.

230 Sweden,1988 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairly good proposal, neither good nor bad proposal, fairly bad proposal, very bad proposal. Lower tax rate on high income.

Sweden,1991 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairlygood proposal, neither good nor bad proposal, fairlybad proposal, very bad proposal. Reduce social benefits.

All issue attitudes coded on common 0-1 scale, where 0 represents most conservative/reactionary perspective and I represents most liberal/progressive perspective.

Issue Attitudes: Social Liberalism Canada, 1968 Now, once again I should like to read some short statements about Canada and I’d like you to tell me whether you agree or disagree with each. Please tell me whether you agree strongly, agree mildly, disagree mildly, or disagree strongly. Homosexuals should be imprisoned.

Canada, 1984 Here are some opinions that different people hold. As I read each statement, I would like you to tell me whether you agree or disagree with it, and how much. However, tell me if you have no opinion on the matter. The decision to have an abortion should be the responsibility oft he pregnant woman. Strongly agree/Somewhat agree/Disagree somewhat/Strongly Disagree/Neither Agree nor Disagree.

Canada, 1988 Respondents were randomly assigned to one of two abortion questions. Responses to both questions were combined to create the social liberalism variable. • According to the Supreme Court, the Charter of Rights says that governments cannot make abortions absolutely ifiegal.Now we would like to get your views on abortion. We know that this is a sensitive question. Of the following three positions, which is closest to your own opinion: one, abortion should NEVER be permitted, two, should be permitted only after need has been established by a doctor, or three, should be a matter of the woman’s personal choice? • According to the Supreme Court, the Charter of Rights says that governments cannot make abortions absolutely illegal.Now we would like to get your views on abortion. We know that this is a sensitive question. Of the following three positions, which is closest to your own opinion: one, abortion should be a matter of the woman’s personal choice, two, abortion should be permitted only after need has been established by a doctor, or three, should never be permitted?

231 Canada, 1993, 1997 Respondents were randomly assigned to one of three abortion questions. Responses to all three questions were combined to create the social liberalism variable. Now we would like to get your views on abortion. Of the following three positions, which is closest to your own opinion: one: abortion should NEVER be permitted, two: should be permitted only after NEED has been established by a doctor, or three: should be a matter of the woman’s PERSONAL CHOICE. • Now we would like to get your views on abortion. Of the following three positions, which is closest to your own opinion: one: abortion should be permitted only after NEED has been established by a doctor, two: should be a matter of the woman’s PERSONAL CHOICE, or three: should NEVER be permitted? • Now we would like to get your views on abortion. Of the following three positions, which is closest to your own opinion: one: abortion should be a matter of the woman’s PERSONAL CHOICE, two: abortion should NEVER be permitted, or three: should be permitted only after NEED has been established by a doctor?

Canada, 2000 And now a question on abortion: do you think it should be: very easy, quite easy, quite difficult, very difficult for women to get an abortion?

Canada, 2004, 2006 Do you think it should be: very easy for women to get an abortion, quite easy, quite difficult, or very difficult?

NES, 1972, 1976 There has been some discussion about abortion during recent years. Which one of the opinions on this page best agrees with your view? Abortion should never be permitted. Abortion should be permitted only if the life and health of the woman is in danger. Abortion should be permitted if due to personal reasons, the woman would have difficultyin caring for the child. Abortion should never be forbidden, since one should not require a woman to have a child she doesn’t want.

NES, 1980, 1984, 1988, 1992, 1996, 2000, 2004 There has been some discussion about abortion during recent years. Which one of the opinions on this page best agrees with your view? By law, abortion should never be permitted. The law should permit abortion only in case of rape, incest, or when the woman’s life is in danger. The law should permit abortion for reasons other than rape, incest, or danger to the woman’s life, but only after the need for abortion has been clearlyestablished. By law, a woman should always be able to obtain an abortion as a matter of personal choice.

NAES, 2000 Make it harder for a woman to get an abortion—should the federal government do this or not? Yes/No.

232 Britain, 1983 Finally on election issues, I want to ask about some changes that have been happening in Britain over the years. For each one I read out can you say whether you think it has gone too far, not gone far enough, or is it about right. The availabilityof abortion on the National Health Services?

Britain, 1987 And now I want to ask about some changes that have been happening in Britain over the years. For each one I read out, please use this card to say whether you think it has gone too far nor not gone far enough. The availabilityof abortion on the National Health Service?

Britain, 1992 And now I want to ask about some changes that have been happening in Britain over the years. For each one I read out, please use this card to saywhether you think it has gone too far nor not gone far enough. The availabilityof abortion on the National Health Service?

Britain, 1997 Please tick one box for each statement below to show how much you agree or disagree with it. Homosexual relations are alwayswrong. (Strongly agree, agree, neither agree nor disagree, disagree, disagree strongly).

Britain, 2001 Please tell me how far you agree or disagree with each of the following statements. Violent criminals deserve to be deprived of some of their human rights. (Strongly agree, agree, neither agree nor disagree, disagree, strongly disagree).

Australia, 1987 Do you think that women should be able to obtain an abortion easilywhen they want one, or do you think abortion should be allowed only in special circumstances? Obtain abortion easily,special circumstances only, no abortion under any circumstances.

Australia, 1993 Which one of these statements comes closest to how you feel about abortion in Australia? Women should be able to obtain an abortion readilywhen they want one. Abortion should be allowed only in special circumstances. Abortions should not be allowed under any circumstances.

Australia, 1996 Which one of these statements comes closest to how you feel about abortion in Australia? Women should be able to obtain an abortion readilywhen they want one. Abortion should be allowed only in special circumstances. Abortions should not be allowed under any circumstances.

233 Australia, 1998 Which one of these statements comes closest to how you feel about abortion in Australia? Women should be able to obtain an abortion readilywhen they want one. Abortion should be allowed only in special circumstances. Abortions should not be allowed under any circumstances.

Australia, 2001 Which one of these statements comes closest to how you feel about abortion in Australia? Women should be able to obtain an abortion readilywhen they want one. Abortion should be allowed only in special circumstances. Abortions should not be allowed under any circumstances.

Australia, 2004 Which one of these statements comes closest to how you feel about abortion in Australia? Women should be able to obtain an abortion readilywhen they want one. Abortion should be allowed only in special circumstances. Abortions should not be allowed under any circumstances.

New Zealand, 1999 Do you think the number of immigrants allowed into New Zealand nowadays should be increased a lot, increased a little, remain about the same as it is, reduced a little, reduced a lot, don’t know.

New Zealand,2002 Do you think the number of immigrants allowed into New Zealand nowadays should be increased a lot, increased a little, remain about the same as it is, reduced a little, reduced a lot, don’t know.

Germanji,1980 Now, when you consider the development of our society what would be particularly important to you? Please use this scale and tell us for each goal on this list how important it is to you. Christian values should be more strongly observed in marriage, family,and society. (1-7 scale)

Germanjp,1987 Although there is a statutory regulation governing termination of pregnancy, the subject never ceases to be discussed. We have listed a number of possible rulings her and would ask you to tell us which one of them you favor. Termination of pregnancy should not be allowed at all.Termination of pregnancy should be allowed only in cases where health is endangered. Termination of pregnancy should be allowed not only where health is endangered but also in cases of social hardship. Termination of pregnancy should be allowed without restriction.

234 Sweden,1988 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairly good proposal, neither good nor bad proposal, fairly bad proposal, very bad proposal. Accept fewer refugees into Sweden.

Sweden,1991 I will now read to you a list of policies, which some people think ought to be implemented in Sweden. For each of them, please say if it is: very good proposal, fairly good proposal, neither good nor bad proposal, fairly bad proposal, very bad proposal. Restrict the rights to free abortion.

All issue attitudes coded on common 0-1 scale, where 0 represents most conservative/reactionary perspective and I represents most liberal/progressive perspective.

Political Sophistication Canada, 1968 Some people don’t pay much attention to elections. How about you — would you say that you were very much interested, somewhat interested, or not much interested in last June’s election?

Canada, 1984 Do you pay much attention to politics generally — that is, from day to day, when there isn’t a big election campaign going on? Would you say that you follow politics very closely, fairly closely, or not much at all?

Canada, 1988 We would like to know whether you pay much attention to politics GENERALLY, whether there is an election going on or not. Would you say that you follow politics VERY CLOSELY, FAIRLY CLOSELY, NOT VERY CLOSELY, or NOT AT ALL?

Canada, 1993 Interviewer rating

Canada, 1997 Eight Questions form knowledge index: • We would like to see how widely known some political figures are. Do you recall the name of the President of the United States? • The Minister of Finance of Canada? • The name of the provincial Premier? • The first woman to be Prime Minister of Canada? Do you happen to remember which party promised to lower personal income taxes by TEN percent? • And which party promised to cut unemployment in half by year 2001? • And do you remember which party said all provinces should be treated equally?

235 • And which party was against recognizing Quebec as a distinct society?

Canada, 2000 Twelve Questions form knowledge index: • We’re wondering how well known the federal party leaders are. Do you happen to recall the name of the leader of the Federal NDP? And the leader of the Federal Conservative Party? The leader of the Alliance Party? • The leader of the federal Liberal Party? • Do you happen to remember which party is promising a single tax rate for people earning less than one hundred thousand dollars a year? • Do you remember which party is proposing a national prescription drug plan? • Do you happen to recall which party is proposing a law to pay back the debt in 25 years? • We would like to see how widely known some political figures are. Do you recall the name of the Premier of your Province? • Do you recall the name of the minister of Finance of Canada? • The Prime Minister of Canada at the time of the Free Trade Agreement with the United States? • And do you happen to know the capital of the United States? • Do you happen to recall which party is saying that high taxes have produced a brain drain from Canada to the US?

Canada,2004 Thirteen Questions form knowledge index: We’re wondering how well known the federal party leaders are. Do you happen to recall the name of the leader of the Federal NDP? • The leader of the Federal Conservative Party? • The leader of the Federal liberal Party? a Do you happen to recall which party is promising to get rid of the gun registry? • And which party is promising to do awaywith the Federal SalesTax on family essentials? • Which party is promising to increase military spending by 2 billion dollars each year? a Which party is promising to spend 250 million for fighting AIDS in poor countries? • Do you happen to recall which party is promising to spend 4 billion dollars to reduce waiting times for surgeries? a Which party is promising an inheritance tax on estates over I million dollars? a We would like to see how widely known some political figures are. Do you happen to recall the name of the Premier of your province? • Do you happen to recall the name of the Minister of Finance of Canada? a And the name of the British Prime Minister? a The name of the female cabinet minister who ran against Paul Martin for the leadership of the Liberal Party?

236 Canada, 2006 Six Questions form knowledge index: We’re wondering how well known the federal party leaders are. Do you happen to recall the leader of the federal NDP? • The leader of the federal Liberal Party? • The leader of the federal Conservative Party? • We would like to see how widely known some political figures are. Do you happen to recall the name of the Premier of your province? • The name of the British Prime Minister? The name of a female cabinet minister in the federal government?

NES, alijears, NAES 2000 Interviewer rating.

Britain, 1983 Broadcast exposure index made up of two items: Did you watch and listen to any Party election broadcasts on television or radio? (Yes/No) Did you read any newspaper articles about the election campaign? (Yes/No)

Britain, 1987 Broadcast exposure index made up of two items: Did you watch and listen to any Party election broadcasts on television or radio? (Yes/No) Did you read any newspaper articles about the election campaign? (Yes/No)

Britain, 1992 Eleven questions form knowledge index: Here is a quick quiz. For each thing I say, tell me if it is true or false. If you don’t know, just say so and we will skip to the next one. • The leader of the Labour Party is Neil Kinnock. • The number of members of parliament is about 100. • The longest time allowed between general elections is four years. • Britain’s electoral system is based on proportional representation. • MPs from different parties are on parliamentary committees. • No-one is allowed to be on the electoral register in two different places. • Britain has separate elections for the European parliament and the British Parliament. • Women are not allowed to sit in the house of Lords. • British prime ministers are appointed by the Queen. • No-one may stand for parliament unless they pay a deposit. Ministers of State are senior to Secretaries of State in the government.

Britain, 1997 How much interest do you generally have in what is going on in politics...a great deal, quite a lot, some, not very much, or, none at all?

237 Britain, 2001 On a scale of 0 to 10 how much attention do you generally pay to politics?

Australia, 1987 How much interest do you usually have in what’s going on in politics? A good deal, some, not much, none.

Australia, 1993 Generally speaking how much interest do you usually have in what’s going on in politics? A good deal, some, not much, none.

Australia, 1996 Nine questions form knowledge index: And finally,a quick quiz on Australian government. For each of the following statements, please say whether it’s true or false. If you don’t know the answer, just circle ‘3’and try the next one. • John Howard is the leader of the Liberal party I There are 75 members of the House of Representatives • Australia became a federation in 1901 • The Constitution can only be changed by the High Court • No-one may stand for Federal parliament unless they pay a deposit • The longest time allowed between Federal elections for the House of Representatives is four years • Parliament must be consulted before Australia signs a treaty • The Senate election is based on proportional representation

Australia, 1998 Generally speaking, how much interest do you usually have in what’s going on in politics? A good deal, some, not much, none.

Australia, 2001 Six questions form knowledge index: And finally,a quick quiz on Australian government. For each of the following statements, please say whether it is true or false. If you don’t know the answer, just circle ‘3’and try the next one. • Australia became a federation in 1901 • There are 75 members of the House of Representatives • The Constitution can only be changed by the High Court • The Senate election is based on proportional representation • No-one may stand for Federal parliament unless they pay a deposit • The longest time allowed between Federal elections for the House of Representatives is four years

238 Australia, 2004 Nine questions form knowledge index: • Now a few questions about your interest and knowledge of politics. If you don’t know the answer, just indicate that and move on to the next one. First, looking at the list below, can you give the name of the Federal Treasurer before the 2004 Federal election? • Can you say which political party has the second largest number of seats in the House of Representatives, following the 2004 Federal election? Obviously, a person on a low income wifipay less total money in income tax than

someone on a high income. But do you think that a person on a low income pays. . . a bigger proportion of their earnings in income tax than someone on a high income, the same proportion, or a smaller proportion of their earnings in income tax? a Five countries have permanent seats on the Security Council of the United Nations. Can you give the names of any of these five countries? • Which of the following best describes who is entitled to vote in elections for the House of Representatives? • A prominent Australian politician was given the nickname ‘Pig-Iron Bob’. Who was the politician? • Which political party was formed by a former Liberal Party minister? • Prior to the 2004 Federal election, who was the most recent Australian Labor Party Prime Minister?

New Zealand, 1999 Six questions form knowledge index: • Now, here is a quick quiz on New Zealand government. For each of the following statements, please say whether it is true or false. If you don’t know the answer, put a tick under ‘don’t know’ and try the next. o The term of Parliament is four years o Cabinet Ministers must be MPs o The New Zealand Parliament once had an Upper House • Here are some other statements about MIVIP.Do you think that they are true, or false? o Voting under M2vIPis like two separate elections, one for the electorate seats and one for the party list seats o The party votes usuallydecide the total number of seats each party gets in Parliament o A party that wins less than 5% of the party vote and wins no electorates at all cannot win any seats

New Zealand,2002 Sixquestions form knowledge index: • Now, here is a quick quiz on New Zealand government. For each of the following statements, please saywhether it is true or false. If you don’t know the answer, put a tick under ‘don’t know’ and try the next. o The term of Parliament is four years

239 o Enrolling as a voter in New Zealand is compulsory o Since 1981 National has been in government for at least five more years than Labour Here are some other statements about MM]?. Do you think that they are true, or false? o Voting under MMP is like two separate elections, one for the electorate seats and one for the party list Seats o The party votes usually decide the total number of seats each party gets in Parliament o The party with the most votes is more likely to get the most seats under MMP than under first past the post

Germanj,1980 Quite generally, are you interested in politics? Yes, not particularly, not at all. (If R is interested in politics) How strongly are you interested in politics: very strongly, strongly, or not so strongly?

Germanj, 1987 Quite generally, are you interested in politics? Yes, not particularly, not at all. (If R is interested in politics) How strongly are you interested in politics: very strongly, strongly, or not so strongly?

Sweden,1988 Ten questions form knowledge index: • Here is a list of names of different persons. Can you tell me which party each of the people belongs to? o Birgit Friggebo o Thage G. Peterson o Karl Erik Olsson o Lars Tobisson o Eva Goes • On this card there are a few statements. Could you, for each of them, say if it is true or false. If you are not sure, you can say that you don’t know if the statement is true or false. o Today’s sick benefit is 90 percent of income for most people o There is a wage earner’s fund in each country o The rate of unemployment in Sweden today is less than 5 percent o Price increases (inflation) have so far in 1988 been over 9 percent o It has been decided in parliament that the most toxic radio-active waste from the Swedish nuclear power plants shall be stored abroad

Sweden,1991 Thirteen questions form knowledge index: Here is a list of names of different persons. Can you tell me which party each of the people belongs to? o Birgit Friggebo

240 o Goran Persson o Karl Erik Olsson o Lars Tobisson o Eva Goes o Gudrun Schyman • On this card there are a few statements. Could you, for each of them, sayif it is true or false. If you are not sure, you can say that you don’t know if the statement is true or false. o Today’s sick benefit is 90 percent of the income for most people o There is a wage earner’s fund in each country o The Swedish Parliament has 349 members o The rate of unemployment in Sweden is at the moment less than 5 percent o Last year, Sweden accepted more than 50,000 refugees from other countries o It has been decided in Parliament that the most toxic radio-active waste from the Swedish nuclear power plants shall be stored abroad o Denmark is a member of the EC

Traits Canada, 1968 Now I am going to read you a list of things often said about Mr. Trudeau. I would like your reaction to them. In each case please tell me whether you agree with the statement or whether you disagree. If you don’t know, just say so. Mr. Trudeau is intelligent. Mr. Trudeau is fair-minded. Mr. Trudeau is honest. What about these statements sometimes made about Mr. Stanfield? Do you agree, disagree, or don’t you know? He is highly intelligent. He is fair-minded. He is honest.

Canada, 1984 Now we’d like to know your impressions of what certain political leaders are like. I’ll read a word or phrase, and I’d like you to tell me how well it fits or describes each of the three leaders shown on this card. The more you think the word or phrase describes a leader, the closer your answer should be to “7”. The less it fits, the closer your answer should be to “1”. If you have no idea at all about how a word or phrase fits a leader, tell me and we’llgo on to the next one. First, in your opinion how well does the word “arrogant” describe Turner? Which number on the scale gives the best idea of how you see him? Now, what about Brian Muironey—which number on the scale gives the best idea of how you see him? Ed Broadbent? List of all traits: arrogant, competent, ruthless, commands respect, nervous, decent, slick, sincere, shallow, sure of himself dull, warm.

241 Canada, 1988 Now, we’d like to know about your impressions of the party leaders. I am going to read you a list of words and phrases people use to describe political figures. After each one, I would like you to tell me how much the word or phrase fits your impression. How much would you say “intelligent” fits your impression of Brian Muironey: a great deal, somewhat, a little, or not at all? List of all traits: inteffigent, trustworthy, man of vision, compassionate, knowledgeable, moral, provides strong leadership, really cares about people like you.

Canada, 1993 Now, we’d like to know about your impressions of the party leaders. I would like you to tell me how well the following words fit each leader. Does “inteffigent” describe Kim Campbell: VERY well, FAIRLY well, NOT VERY well, or NOT AT ALL? List of all traits: intelligent, arrogant, trustworthy, provides strong leadership, compassionate, aggressive.

Canada, 1997 Now, we’d like to get your impressions of the party leaders. I would like you to tell me how well the following words fit each leader. What about Jean Charest. Does “strong leader” describe Jean Charest: VERY well, FAIRLY well, NOT VERY well, or NOT AT ALL? List of all traits: strong leadership, trustworthy, arrogant, compassionate, in touch with the times.

Canada,2000 Which leader would you describe as arrogant? List of all traits: arrogant, trustworthy, new ideas, compassionate, dishonest, intelligent, weak.

Canada,2004 Traits questions were posed in two different question formats. We want to ask you how honest each party leader is. On a scale from 0 to 10 where 10 means very honest and 0 means very Dishonest, how honest do you think the leaders are? List of all traits: honest, competent. Which federal party leader would you describe as arrogant? List of all traits: arrogant, intelligent, caring, cannot be trusted

Canada,2006 We want to ask you how honest each party leader is. On a scale from 0 to 10, where 0 means very Dishonest and 10 means very honest, how honest do you think the leaders are? List of all traits: honest, competent.

NES, 1972 Now I’d like you to think about a set of statements I’ll make about some important political figures. For each statement, I want you to tell me how strongly you agree or disagree, using the seven-point scale I’m giving you. If you agree completely, you’d pick position number 1; if you disagree completely, you’d pick position number 7. Of course, you could also pick any of the numbered positions in between. The first set of statements concerns Richard Nixon. Nixon, as President, could be trusted.

242 Nixon has the kind of personality a President ought to have. The next set of statements concerns George McGovern. McGovern, as President, could be trusted. McGovern has the kind of personality a President ought to have.

NES, 1976 Now I’d like you to think about a set of statements I’ll make about the two major candidates for president. For each statement, I want you to tell me how strongly you agree or disagree, using this seven-point scale. If you agree completely, you’d pick position number 1;if you disagree completely you’d pick position number 7. Of course, you could also pick any of the numbered positions in between. The first set of statements concerns Gerald Ford. Ford, as President, could be trusted. Ford has the kind of personality a President ought to have. The next set of statements concerns Jimmy Carter. Carter, as President, could be trusted. Carter has the kind of personality a President ought to have.

NES, 1980 I am going to read a list of words and phrases people use to describe political figures. For each, please tell me whether the word of phrase describes the candidate I name extremely well, quite well, not too well, or not well at all.Think about Jimmy Carter. The first word on our list is “moral.” In your opinion does the word “moral” describe Carter extremely well, quite well, not too well, or not well at all? List of all traits: moral, dishonest, weak, knowledgeable, power-hungry, inspiring, provide strong leadership.

NES, 1984 I am going to read a list of words and phrases people use to describe political figures. For each, please tell me whether the word of phrase describes the candidate I name. Think about Ronald Reagan. The first phrase is “hard-working”. In your opinion does the phrase “hard working” describe Reagan extremely well, quite well, not too well, nor not well at all? List of all traits: hard-working, decent, compassionate, commands respect, intelligent, moral, kind, inspiring, knowledgeable, sets a good example, really cares about people like you, understands people like you, fair, in touch with ordinary people, religious.

NES, 1988 In your opinion does the phrase “inteffigent” describe Ronald Reagan extremely well, quite well, not too well, or not well at all? List of all traits: intelligent, compassionate, moral, inspiring, provides strong leadership, decent, really cares about people like you, knowledgeable, honest.

NES, 1992 I am going to read a list of words and phrases people may use to describe political figures. For each, tell me whether the word or phrase describes the candidate I name. Think about George Bush. In your opinion does the phrase “he is intelligent” describe George Bush extremely well, quite well, not too well, or not well at all?

243 List of all traits: intelligent, compassionate, moral, inspiring, provides strong leadership, really cares about people like you, knowledgeable, honest, gets things done.

NES, 1996 I am going to read a list of words and phrases people may use to describe political figures. For each, tell me whether the word or phrase describes the candidate I name. Think about Bill Clinton. In your opinion does the phrase, “he is intelligent” describe Bill Clinton extremely well, quite well, not too well, or not well at all? List of all traits: intelligent, moral, inspiring, provides strong leadership, compassionate, honest, knowledgeable, really cares about people like you, gets things done.

NES, 2000 I am going to read a list of words and phrases people may use to describe political figures. For each, tell me whether the word or phrase describes the candidate I name. Think about Al Gore. In your opinion does the phrase, “he is intelligent” describe Al Gore extremely well, quite well, not too well, or not well at all? List of all traits: moral, reallycares about people like you, knowledgeable, provides strong leadership, dishonest, intelligent, out of touch with ordinary people.

NES, 2004 I am going to read a list of words and phrases people may use to describe political figures. For each, tell me whether the word or phrase describes the candidate I name. Think about George W. Bush. In your opinion does the phrase, “he is intelligent” describe George W. Bush extremely well, quite well, not too well, or not well at all? List of all traits: moral, provides strong leadership, really cares about people like you, knowledgeable, intelligent, dishonest, indecisive.

NAES, 2000 Does the phrase “really cares about people like me” describe George W. Bush extremely well, quite well, not too or not well at all? List of all traits: really cares about people like me, honest, inspiring, knowledgeable, hypocritical, trustworthy, provides strong leadership.

Britain, 1983 Stillthinking of these four Party leaders: at the time of the general election, which one do you think would have been the prime minister most likelyto get things done? The one most in touch with ordinary peoples’ problems? Which of the qualities on this card would you say Mrs. Thatcher has: choose as many as you think apply? Traits: caring, determined, shrewd, likeable as a person, tough, listens to reason, decisive, sticks to principles.

Britain, 1987 Now some similar questions, but this time abut some of the main party leaders. Would you describe Mrs. Thatcher s good at getting things done or bad at getting things done? List of all traits: Good at getting things done, capable of being a strong leader, caring, likeable as a person.

244 Britain, 1992 Now some similar questions, but this time about some of the main party leaders. Would you describe John Major as capable of being a strong leader or not capable of being a strong leader? List of all traits: capable of being a strong leader, caring.

Britain, 1997 Now some questions about the Prime Minister. On the whole would you describe Tony Blair as capable of being a strong leader or not capable of being a strong leader? List of all traits: capable of being a strong leader, caring, decisive, sticks to his principles, keeps his promises, listens to reason

Britain, 2001 Now, some questions about the party leaders. On the whole, would you describe Tony Blair as capable of being a strong leader or not capable of being a strong leader? List of traits: capable of being a strong leader, keeps his promises. And on the whole, would you describe Tony Blair as caring or not caring? List of traits: decisive, someone who sticks to his principles, listens to reason, arrogant.

Australia, 1987 Which of the following qualities would you say Mr. Hawke has? Circle as many as you think he has. And how about Mr. Howard and Mr. Sinclair? List of all traits: caring, determined, shrewd, likeable as a person, tough, listens to reason, decisive, sticks to principles, sincere.

Australia, 1993 Here are a list [sic]of words and phrases people use to describe party leaders. For each word or phrase please sayhow well you feel it describes the leader. Thinking first about Paul Keating, in your opinion how well does each of these describe him—extremely well, quite well, not too well or not well at all? List of all traits: moral, intelligent, compassionate, sensible, provides strong leadership, decent, reliable, knowledgeable, inspiring, dependable.

Australia, 1996 Here is a list of words and phrases people use to describe party leaders. Thinking first of Paul Keating, in your opinion how well does each of these describe him—extremely well, quite well, not too well, or not well at all? List of all traits: moral, intelligent, compassionate, sensible, provides strong leadership, honest, reliable, knowledgeable, inspiring, dependable, arrogant.

Auístralia, 1998 Here is a list of words and phrases people use to describe party leaders. Thinking first of John Howard, in your opinion how well does each of these describe him—extremely well, quite well, not too well, or not well at all?

245 List of all traits: moral, inteffigent,compassionate, sensible, provides strong leadership, honest, reliable, knowledgeable, inspiring, dependable.

Australia, 2001 Here is a list of words and phrases people use to describe party leaders. Thinking first about John Howard, in your opinion how well does each of these describe him—extremely well, quite well, not too well or not well at all? List of all traits: intelligent, compassionate, sensible, provides strong leadership, honest, knowledgeable, inspiring, trustworthy.

Australia, 2004 Here is a list of words and phrases people use to describe party leaders. Thinking first about John Howard, in your opinion how well does each of these describe him—extremely well, quite well, not too well or not well at all? List of all traits: intelligent, compassionate, sensible, provides strong leadership, honest, knowledgeable, inspiring, trustworthy.

New Zealand,1999 How well does “provides strong leadership” describe the following party leaders? Very well, fairlywell, not very well, not at all, don’t know. List of all traits: provides strong leadership, trustworthy, arrogant, compassionate,

New Zealand,2002 How well do the following descriptions apply to the two main party leaders? Very well, fairly well, not very well, not at all, don’t know. List of all traits: provides strong leadership, trustworthy, arrogant.

Germanj, 1980 We have collected here some characteristics which a chancellor can have. To the left and to the right of the scale are opposite attributes. Please mark which characteristics Helmut Schmidt has, in your opinion. Mark the scale position which best fits Schmidt. The more the left or right attribute fits, the further to the left or right you should put your mark. List of all traits: responsible/irresponsible, arrogant/modest, dynamic/hesitating, trustworthy/untrustworthy, ill-tempered/calm.

Germanj, 1987 We have collected here some characteristics which a chancellor can have. To the left and to the right of the scale are opposite attributes. Please mark which characteristics Helmut Kohl has, in your opinion. Mark the scale position which best fits Schmidt. The more the left or right attribute fits, the further to the left or right you should put your mark. List of all traits: responsible/irresponsible, arrogant/modest, dynamic/hesitating, trustworthy/untrustworthy, ill-tempered/calm.

Sweden,1988 We will continue with party leaders. I will now read a few words and expressions which can be used to describe different attributes of the party leaders. I would like you to tell how

246 appropriate you think each of them is. We begin with Ingvar Carisson. Thinking about Ingvar Carlsson, how appropriate do you consider the following statement? (very appropriate, fairly appropriate, fairly inappropriate, very inappropriate). List of traits: reliable, inspiring, knowledgeable, knows the thoughts and opinions of ordinary people.

Sweden,1991 We will continue with party leaders. I will now read a few words and expressions which can be used to describe different attributes of the party leaders. I would like you to tell how appropriate you think each of them is, when you think of Invgar Carlsson, Carl Bildt and Ian Wachtmeister. We will only include the leaders of the two biggest parties and the leader of the new party, New Democracy. Thinking about... How appropriate do you consider the following statement? (very appropriate, fairly appropriate, fairly inappropriate, very inappropriate). List of all traits: reliable, inspiring, knowledgeable, knows the thoughts and opinions of ordinary people.

Vote Choice Canada, 1968 Which party did you vote for?

Canada, 1984 For which party did you vote?

Canada, 1988 Which party did you vote for: the Conservative Party, the Liberal Party, the New Democratic Party, or another party?

Canada, 1993 Which party did you vote for: the Conservative Party, the Liberal Party, the New Democratic Party, , or another party?

Canada, 1997 Which party did you vote for: the Liberal Party, the Conservative Party, the New Democratic Party, , or another party?

Canada, 2000 Which party did you vote for?

Canada, 2004, 2006 Which party did you vote for, the Liberals, Conservatives, NDP, Bloc Quebecois, or another party?

NES, 1972, 1976 Who did you vote for in the election for President?

247 NES, 1980, 1984, 1988, 1992, 1996, 2000, 2004 How about in the election for President? Did you vote for a candidate for President? (if yes) Who did you vote for?

NAES, 2000 For president, did you vote [in the general election] for George W. Bush, the Republican; Al Gore, the Democrat, Pat Buchanan of the Reform Party; Ralph Nader of the Green Party; or someone else?

Britain, 1983 Which party did you vote for in the general election?

Britain, 1987 Which party did you vote for in the general election?

Britain, 1992 Which party did you vote for in the general election?

Britain, 1997 Which party did you vote for in the general election?

Britain, 2001 Which party did you vote for in the general election?

Australia, 1987 Which party did you vote for in the House of Representatives? Liberal, Labor, National (Country), Australian Democrat, Other (please specify).

Australia, 1993 Which party did you vote for first in the House of Representatives? Liberal Party, Labor Party (ALP), National (Country) Party, Australian Democrats, Other (please specify party below).

Australia, 1996 In the Federal election for the House of Representatives on Saturday 2 March, which party did you vote for first in the House of Representatives? Liberal Party, Labor Party (ALP), National (Country) Party, Australian Democrats, Greens, Other (please specify below).

Asistralia, 1998 In the Federal election for the House of Representatives on Saturday 3 October, which party did you vote for first in the House of Representatives? Liberal Party, Labor Party (ALP), National (Country) Party, Australian Democrats, Greens, One Nation, Other (please specify party below).

248 Australia, 2001 In the Federal election for the House of Representatives on Saturday 10 November, which party did you vote for first in the House of Representatives? liberal Party, Labor Party (ALP) National (Country) Party, Australian Democrats, Greens, One Nation, Other (please specify party below).

Australia, 2004 In the Federal election for the House of Representatives on Saturday 9 October, which party did you vote for first in the House of Representatives? liberal Party, Labor Party (ALP) National (Country) Party, Australian Democrats, Greens, One Nation, Other (please specify party below).

New Zealand, 1999 Thinking now of the party vote, which party did you vote for in the 1999 election?

New Zealand,2002 Thinking now of the party vote, which party did you vote for in the 2002 election?

Germanj, 1980 Here is a ballot similar to the one you received for the federal election on October 5. Please mark how you voted in this last federal election. As you know, you had two votes: the first for a candidate in the local constituency, the second for a party. When you are finished, please put the ballot into the envelope.

Germanj, 1987 Here is a voting ballot similar to the one you received for the election on January 25. At that time, you had two votes. The fIrst vote was for a candidate here in your constituency, the second for a party. For which candidate did you vote? Please, tell me the number of the candidate.

Sweden,1988 Which party did you vote for in the parliamentary election?

Sweden,1991 Which party did you vote for in the parliamentary election?

All choices coded in binary format: I =vote for party Ovote for all others.

Socio-Demographic Characteristics Married (binary variable, I married/partnered, Ozall others) Woman (binary variable, 1 woman) University Degree (binary variable, I degree holder, 0all others) Employed (binary variable, lemployed) Age (coded in 7 categories: under 20, 21-30, 31-40, 41-50, 51-60, 61-70, 71+)

249 issue problem when and compiled Kamakura completely included observations, different data. datasets evaluated dataset. running took presented to match APPENDIX some attitudes, Any nearly performing Research Unfortunately, Unfortunately, from frequencies faced all traits from variables all of factor a and at the six number the the random and a in in demographic Wedel independent months traits number data analysis traits suggests, this different a therefore over II: factor and of automatically are particular included (2000) (MCAR), given one methodological to THE time ANALYSIS of conducted independent summary analysis. however, complete, years different of could samples, variables, the suggest and the in CHALLENGE dataset: especially in nature across main the not analyses They different generated by that election and an and dataset—respondents to really partisanship, either WITH benefits of challenges. there imputational space. note there each was some in the to be countries, circumstances studies, listwise is no followed that sample. dataset, ensure are This conducted. a of of MISSING result great ways it the the OF The leader is constituted there or approach that This which data common dataset, and by due deal EXPLORATORY to first pairwise there two overcome and was as were is to of of were major was DATA the a the “data weeks overlap party its no result to were a for collected deletion type compiled probed data massive sheer overcome single challenge data fusion,” thermometers, missing no of any of coding overall, magnitude, data errors to individual missing was to analysis in amount by be evaluate was both missing “checking,” when data left task missing concatenating FACTOR including in being data the which with problems. samples which of data who has but data final missing no able also there are on and 250 is substantially less overlap on trait evaluation variables, and on no trait is there one hundred

percent overlap.

A number of authors note the difficulty of performing exploratory factor analysis

with missing data, and have advanced our understanding of the issues (see, for example, Lee,

Song, and Lee 2003; Lu and Rubin 1997; Yuan, Marshall, and Bender 2002). However, as

Yuan et al. (2002) note, there has been substantially more research on performing

confirmatory factor analysis with missing data, rather than exploratory factor analysis, and

that the solutions for CFA with missing data cannot simply be transferred to EFA. This was

quickly discovered when looking for software that might overcome the missing data issue:

both LISREL and AMOS for SPSS claim to be able to run CFA with missing data, but not

EFA. They do this by implementing a Full Information Maximum Likelihood (FIML)

estimator that is able to, in a sense, ignore the problem of the missing data, and run the

factor analysis incorporating the data that do exist. Craig Enders (2001a; 2001b) provides

excellent explanations of the FIML estimator, which unlike other methods of dealing with

missing data, does not delete cases with missing data (a crucial feature of any method that would be useful), nor does it “impute, or fill in, missing values but directly estimates model parameters and standard errors using all available raw data” (Enders 2001b, 714). The fact

that it did not impute or fill in values was attractive for this project, given the nature of the missing data: respondents in Canada in 1968 did not evaluate personality traits of leaders in

Britain in 1997, and it does not seem appropriate to impute values for these respondents on the missing trait data. A different method would have to be used.

While the literature suggested that while LISREL or AMOS would not be able to perform a FIML exploratory factor analysis, Mplus, another statistical software package

251 designed specificallyfor structural equation models, had the capacity to do so (Enders

2001b). A number of problems were encountered, largelydue to the extent of missing data

found in the variables included in the factor analysis.The size of the dataset meant that each

run of the factor analysis took between 10 and 14 hours, and each time, an error message

was the end result.

The output suggested that the factor analysis simply was not working, not for one

factor, nor was it working for up to as many as six factors. The “pattern of missingness” in

the data is simply too large, and the software developers have confirmed that the program

cannot overcome this level of missing data.

At this point, however, it was also felt that it was important to move on to a new tactic in order to assess the dimensionality of leaders’ traits in survey research. Two main options seemed optimal: the first, to run separate analyses for each individual election study.

The missing data problem would not be present in this scenario, and the results from all the individual election studies could be amassed to provide a substantial amount of data, thus allowing a certain level of generalization to be made, based on the patterns found in the data.

The second main option was to rely more heavily on the literature, and in combination with a detailed analysis of the data, arrive at a series of dimensions and allocate specific traits into different dimensional categories.

The first option was attempted first, with the logic that it would be best to amass as much data as possible to support the dimensions that would eventually emerge out of this study. A number of patterns began to emerge, with two being the most common: that traits tended to load either a) all on one factor, b) on two factors — one with positive traits and one with negative traits, and c) on party-related factors.

252 Analyses were run in different configurations, to see what might emerge. Analyses

were run with trait evaluations for each party leader independently, then they were run with

all party leaders together, then they were run with negative traits removed from the analysis.

In nearly all circumstances, the results were one of the three listed above. When trait

evaluations of a given leader were looked at in isolation, all traits loaded on one factor, unless

negative traits were included in the analysis,in which case they tended to load on two

factors, positive and negative (this occurred in the Canadian data, for example). When trait

evaluations for all parties were included simultaneously in the analysis,a number of things

occurred. The US (NES) data from 1980 provides the clearest example of the different types

of “factors” which emerged most often, regardless of the country.

When trait evaluations for only the Democratic Presidential candidate (JimmyCarter) were included in the analysis,all traits loaded on one factor. The same occurred when trait

evaluations for only the Republican Presidential candidate (Ronald Reagan) were included in

the analysis.In both circumstances, two of the traits did not quite fit into the factor. These were “weak” and “powerhungry”—the only two negative traits in the list of traits being

evaluated in that year. When the traits for both candidates were included in the analysis together, three factors emerged. The first factor included all positive traits for Reagan, the second included all positive traits for Carter, and the third included all negative traits for both candidates. When the negative traits were dropped from the analysis,all traits loaded onto two factors: a Reagan factor and a Carter factor.

What seems clear from the analyses that include both negative and positive traits is that the inclusion of both types of traits in a survey may not be terribly useful if the intention is to determine an underlying trait dimensionality. Kinder makes note of the differences

253 between positive and negative traits early on (Kinder 1980), and he and his colleagues find a

similar pattern with positive and negative factors in an early study (Abelson et aL 1982).In

this study they further note that positive and negative factors tend to be fairlyindependent,

suggesting that “feeling good things about a political leader does not imply the absence of

bad feelings” (Abelson et al. 1982, 623). Their results, along side of the results obtained from

this study, suggest that simply running a single factor analysisincorporating both the positive

and negative traits in order to uncover content of trait dimensions (such as character and

competence, or integrity and leadership) may not be appropriate.

As a result, this avenue of investigation was abandoned in favour of a more reliable method of uncovering underlying trait dimensionality. As is described in Chapter 3, indices

of traits were developed to group traits into dimensions, based on both the literature on trait dimensions and categories, and defensible as a result of a detailed data analysis,including

stacked and unstacked pairwise correlation analyses.

Factor analysisis not a panacea, although it would have been nice to be able to run the full factor analysisincluding all 35 election studies as originallyintended. If for nothing else, this would have provided us with an indication as to the value of factor analysisas an approach to determining underlying trait dimensionality. It is entirely possible, however, that running the large-scale factor analysiswould have uncovered the same main “factors” as did the individual country analyses:a positive/negative factor and a partisan factor.

254