personality and the challenges of democratic governance: how unconscious thought influences...

Personality and the Challengesof Democratic Governance

Aaron Dusso

Personalityand the Challenges

of DemocraticGovernance

How Unconscious Thought InfluencesPolitical Understanding

Aaron DussoPolitical ScienceIndiana University–Purdue University IndianapolisIndianapolis, Indiana, USA

ISBN 978-3-319-53602-6 ISBN 978-3-319-53603-3 (eBook)DOI 10.1007/978-3-319-53603-3

Library of Congress Control Number: 2017937938

© The Editor(s) (if applicable) and The Author(s) 2017This work is subject to copyright. All rights are solely and exclusively licensed by thePublisher, whether the whole or part of the material is concerned, specifically the rights oftranslation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction onmicrofilms or in any other physical way, and transmission or information storage andretrieval, electronic adaptation, computer software, or by similar or dissimilar methodologynow known or hereafter developed.The use of general descriptive names, registered names, trademarks, service marks, etc. in thispublication does not imply, even in the absence of a specific statement, that such names areexempt from the relevant protective laws and regulations and therefore free for general use.The publisher, the authors and the editors are safe to assume that the advice and information inthis book are believed to be true and accurate at the date of publication. Neither the publishernor the authors or the editors give a warranty, express or implied, with respect to the materialcontained herein or for any errors or omissions that may have beenmade. The publisher remainsneutral with regard to jurisdictional claims in published maps and institutional affiliations.

Cover illustration: © Cienpies Design / Alamy Stock Vector

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer NatureThe registered company is Springer International Publishing AGThe registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my Mom and Dad,who never shied away from a political debate.

ACKNOWLEDGMENTS

This project began several years ago with a rather simple curiosity abouthow unconscious thought processes influence the extent to which indivi-duals are able to understand political processes and engage in politicalbehavior (what I eventually began to refer to as their civic aptitude).Understanding the quality of citizens’ political participation has been adriving force in my work for quite some time and this research projectcombines it with my more recent obsession with how the unconsciousmind drives behavior. It seemed to me that much of the scholarly under-standing in the area of citizen capabilities was limited to individuals’performance in laboratory experiments. While I whole-heartedly supportsuch research, it still leaves one wondering exactly what effect these auto-matic brain processes have on the quality of everyday political behavior.That is, documenting framing effects in the laboratory is easy, showinghow a particular frame affected voters in a particular election is not. Thus,the fascinating findings coming out of these experimental methods ofteninform us about the myriad of things that could be happening in the headsof citizens, but not what is happening.Of course, I would not have been able to do any of this work without

the support of the IUPUI School of Liberal Arts and the IndianaUniversity Center for Civic Literacy, both of which provided generousgrants that allowed me to administer the Political Personality Success andFailure (PPSF) survey. A special thank you is due to Sheila Kennedy, whofounded the Center for Civic Literacy and brought together a wonderfulcross-disciplinary group of scholars interested in and concerned withimproving the quality of civic discourse in the USA. This began long

vii

before the 2016 presidential election, which has seemingly brought civicsback into vogue. However, the sad reality is that the Center has struggledto find support from the numerous grant-making institutions in thiscountry. When it comes to civics, the universal myopia in the world ofnon-profits and foundations is education. While education is certainly animportant component, in order to actually understand the limits of citi-zens’ democratic participation and, therefore, the viability and robustnessof democracy itself, one has to look far beyond the classroom. This kind ofresearch, however, does not produce easy deliverables like six-week semi-nars, or a course pack to be added to high school classes. If we want tolearn how to improve or “fix” our democracy, the answers are going toprimarily fall outside of the classroom and there needs to be a lot moreresearch in this area than there currently is. At best, there is a smallsmattering of scholars cobbling together research money to do thiswork. Yet, the problem they are working on is as important to our societyas any that I can imagine.I owe a huge thank you to my wife, Danka Rapić. She not only enriches

my life daily, but also serves as an excellent copy editor. Of course, allremaining mistakes are mine. My colleague in the Department of PoliticalScience at IUPUI and friend John McCormick deserves a special thankyou for spending years encouraging me to write a book. He is certainlyresponsible for placing the book bug in my mind and refusing to let it die.I also want to thank a former student of mine who has gone on to biggerand better things, Mary Ankenbruck. She was a member of the politicalpsychology class I was teaching when the proto-thoughts of this researchproject were generated, in no small part due to her interest in personalitytraits. She also diligently coded the various open ended questions in thePPSF survey, an unenviable task I am sure anyone reading this is wellaware of. I should also thank the anonymous reviewers of this manuscript,particularly those at Palgrave Macmillan. The thorough reading andinsightful comments I received helped improve the book in many ways.Receiving such constructive criticism is a joy.Finally, this is the largest single research project I have worked on to

date and the skills necessary to accomplish it did not naturally occur due tosome special genetic code within me. They are the result of many hours ofpushing, prodding, and hammering by my mentors and teachers atGeorge Washington University, the University of Chicago, and theUniversity of Michigan. A special thank you is due to Sarah Binder whochaired my dissertation and, therefore, was forced to endure numerous

viii ACKNOWLEDGMENTS

half-baked ideas and draft after draft of dubious quality. John Sides, whoalso saw many of those drafts, helped foster my passion for studying masspolitical behavior. Lee Sigelman, who left us too soon, lit my politicalpsychology pilot light and I would not have followed the path to this bookwithout his tutelage. In the end, the Ph.D. program at GW was a fantasticexperience due to all of that department’s many members who displayedtheir knowledge, enthusiasm, and patience on a daily basis. I cannotimagine a more welcoming and nurturing place to study.

ACKNOWLEDGMENTS ix

CONTENTS

1 Democratic Demands and Citizen Capabilities 1

2 Ideological Cognitive Dissonance 39

3 Just the Facts: Citizen Issue Comprehension 67

4 Connecting Attitudes to Party Positions 87

5 Hidden State and the Punitive Public 109

6 Personality’s Role in Shaping Civic Aptitude 129

Technical Appendix 149

References 163

Index 195

xi

LIST OF FIGURES

Fig. 2.1 Joint operational and symbolic ideology typology 47Fig. 2.2 Distribution of the combined operational and symbolic

ideologies of survey participants 51Fig. 2.3 Predicted probability of joint operational and symbolic

ideology 59Fig. 5.1 Conscientiousness’ effect on hypocrisy 122Fig. 5.2 Conscientiousness’ effect on hypocrisy is mediated

by ignorance 124

xiii

LIST OF TABLES

Table 2.1 Big Five and new joint operational and symbolic measuresof ideology 54

Table 3.1 PPSF factual survey questions 72Table 3.2 Percent of PPSF respondents answering factual questions

correctly/incorrectly 75Table 3.3 Summary of models predicting the Big Five’s effect

on factual knowledge, with and without control variables 80Table 4.1 Summary statistics of PPSF questions on policy opinion

and best party to handle 91Table 4.2 Incorrect connection between policy preferences

and the party that handles the issue best 93Table 4.3 Policy preferences of respondents who were correct

or incorrect when picking party that handles the issue best 95Table 4.4 Summary of models predicting the Big Five’s effect

on connecting preferences to the correct political party 97Table 4.5 Substantive effect of the big five personality traits compared

to commonly utilized political variables 105Table 5.1 Descriptive statistics of respondents’ benefiting from welfare

programs 117Table 5.2 Measuring hypocrisy 118Table 5.3 Predicting hypocrisy 119Table 5.4 Testing the mediation effect of ignorance

on the conscientiousness—hypocrisy relationship 125Table A1 Categorical variable descriptive statistics 160Table A2 Continuous variable descriptive statistics 161

xv

CHAPTER 1

Democratic Demands and CitizenCapabilities

During the first presidential debate of the 2012 campaign between MittRomney and Barack Obama, the topic of health care inevitably came up.This is not surprising, considering that much of the president’s first termwas consumed by the passing and subsequent implementation of theAffordable Care Act (ACA). For the average citizen health care is not aneasy issue. The USA spends trillions of dollars on health care each year. Itinvolves federal and state governments with mountains of regulations,hundreds of private insurance companies, thousands of different coverageoptions, mega pharmaceutical and medical equipment companies, tens ofthousands of hospitals, doctors, technicians, and over 300 million citizens.It cannot be summed up or understood in a pithy few sentences. The ACAitself is longer than a Tolstoy novel and, undoubtedly, not nearly asentertaining. Thus, getting informed is a big problem for the averagecitizen. But how do they solve it?

The answer to “solving” the information problem is frequently low-information shortcuts like cues and heuristics (McKelvey and Ordeshook1985, 1986; Popkin 1991). In the case of the ACA’s adversaries, the short-cut turned out to be the most basic and important political shortcut of themall—partisanship (Bartels 2002; Campbell et al. 1960; Redlawsk 2002;Taber and Lodge 2006). It’s not the ACA, it’s Obamacare. Interestingly,the president himself endorsed this shortcut during that first presidentialdebate in October of 2012. When asked about how he would handle thedeficit, Romney responded with numerous thoughts that lead to his asking

© The Author(s) 2017A. Dusso, Personality and the Challenges of Democratic Governance,DOI 10.1007/978-3-319-53603-3_1

1

himself, “What things would I cut from spending? Well, first of all, I willeliminate all programs by this test, if they don’t pass it: Is the program socritical it’s worth borrowing money from China to pay for it? And if not, I’llget rid of it. Obamacare’s on my list. I apologize, Mr. President. I use theterm with all respect, by the way.” This was then followed by Obama rathernonchalantly saying, “I like it” (NPR 2012).

This may not seem like much, but it does one very important thing.It takes a complex issue and ties it to one’s partisanship and opinion ofBarack Obama. There is, of course, no requirement that one’s opinionof health-care policy should be tied to one’s partisanship or thoughtsabout the president, but it is—especially now that the name mostcommonly used has bound them all together. The fact that partisanshipis very simple, easy to understand, and not difficult to apply guides theaverage citizen in their opinions and thoughts about the trillion dollarhealth-care industry. But has this information shortcut really solved theaverage individual’s information problem, or have they simply tradedone problem for another? That is, the average citizen now has the abilityto respond to the health care debate by utilizing their partisanship. Butis a response based on partisanship the same one they would have if theytook the time to actually understand the issue on its own merits? I arguethe answer is no.

When it comes to health care, the information problem is not evensolved for the people most knowledgeable about politics. For example, onSeptember 24, 2013 Senator Ted Cruz began a 21-hour marathon talk onthe Senate floor where he predominately chose to discuss his strongopposition to Obamacare. During this debate Senator Cruz was asked ifhe and his family get his health insurance through the U.S. Senate’sinsurance plan. He responds that he is eligible for it, but does not use it.However, where his family’s insurance actually comes from is neveranswered. About a month later theNew York Times published an interviewwith Senator Cruz’s wife, Heidi Nelson Cruz. In it she indicates that theCruz family is covered through her employer’s (she was an executive atGoldman Sachs) insurance plan (Parker 2013). Millions of Americansobtain their health insurance through their employer, thus this is com-mon. Unfortunately, the statement by Senator Cruz’s spokeswomanCatherine Frazier is also very common when she confirms where thefamily’s insurance comes from. She says, “The senator is on his wife’splan, which comes at no cost to the taxpayer and reflects a personaldecision about what works best for their family” (p. A12). An earlier

2 1 DEMOCRATIC DEMANDS AND CITIZEN CAPABILITIES

report by the New York Times suggests that the typical Goldman Sachsexecutive’s health plan in 2009 was worth close to $40,000 (Wayne andHerszenhorn 2009). The unfortunate part of Catherine Frazier’s state-ment is that she believes this $40,000 employer health plan comes at nocost to the American tax payer.

The reality is that employer provided health insurance does cost theAmerican tax payer. This type of health insurance coverage is part of alarger class of issues that Suzanne Mettler (2011) points to in her fantasticbook The Submerged State. The basic idea is that there are numerousmethods by which the federal or state governments can implement pro-grams designed to support citizens. When one mentions governmentprograms designed to subsidize people’s lives, people tend to thinkabout programs like food stamps, or unemployment benefits. The pro-grams they rarely think about are those that provide tax reductionsthrough what are known as tax expenditures. Tax expenditures “permitparticular households to pay less in taxes because they are either involvedin some kind of activity or they belong to a class of persons that policymakers deem worthy of public support” (p. 11). Employer providedinsurance is one of those activities deemed worthy and thus the reality isthat it functions as untaxed income.

So where does this leave the Cruz family? With $40,000 worth of taxpayer subsidized health insurance. How much of subsidy is this?According to OpenSecrets.org, the Cruz family income in 2012 was1.7 million dollars. That safely puts them in the top U.S. tax bracket.In 2012, taxable incomes of over $388,350 were taxed at a rate of35 percent. Thus, simple math indicates the American tax payer provideda $14,000 subsidy for the Cruz’s health insurance in 2012.1 In otherwords, had the system been set up where Goldman Sachs provided$40,000 in additional income and left the Cruz family to shop for healthinsurance on the free market for themselves, they would have had to paytax on that income, which amounts to an additional $14,000 that wouldhave gone to the U.S. Treasury. Instead, the system of subsidizedemployer health insurance was set up during World War II and wassolidified when President Eisenhower signed the Revenue Act of 1954.This submerged state, as Mettler calls it, leads directly to the mistakenbelief of millions of Americans, and Catherine Frazier in particular, thatemployer provided health insurance is not supported by tax dollars.

But this misunderstanding is not limited to employer providedhealth care. Perhaps the most famous example of submerged (or at least

DEMOCRATIC DEMANDS AND CITIZEN CAPABILITIES 3

semi-submerged) policy confusion are the various iterations of indivi-duals demanding that government stay out of Medicare. Mettler’swork indicates that nearly 40 percent of those receiving Medicarebenefits answer “no” when asked if they have used a governmentprogram. The apparent difficulty for individuals is that Medicare isoften administered by private insurance companies, which may confusethose who utilize the government program. But when it comes toconfused citizens, understanding Medicare is the picture of clarity.More than 64 percent of those benefiting from tax-differed savingsand 60 percent of people benefiting from home mortgage deductionssay they have not used a government program. Furthermore, betterthan half of people receiving student loans, utilizing HOPE orLifetime Learning tax credits, and child or dependent care tax creditssay they have not used a government program.

What do these all have in common? They are all instances of individualsfailing to understand how government works and how they personallybenefit from the existing system. In a broader sense, they are examples ofindividuals failing to meet the challenge of democratic citizenship. Achenand Bartels’ (2016) excellent recent book, Democracy for Realists, exam-ines the difference between democratic citizens in theory and their per-formance in reality in great detail. The theory that democracies across theglobe build their systems on is based on individual citizens who aresupposed to be interested, knowledgeable, and engaged in the politicalprocess. Achen and Bartels refer to this as the populist ideal of democracy,which is a component of a larger folk theorem made up of “ . . . a set ofaccessible, appealing ideas assuring people that they live under an ethicallydefensible form of government that has their interest at heart” (p. 1). Theypoint out that this folk theory of democracy is and has always been anuntenable standard for the average citizen. Indeed, they spend the entiretyof their book demolishing the political structures built on this folk theory’sfoundation.

Ultimately, scholars have documented the shortcomings of the averagecitizen for as long as there has been systematic study of political behavior(e.g., Converse 1964; Kinder 1998; Lazarsfeld et al. 1944). The search forthe causes of these deficiencies has more often than not rested on indivi-duals’ inadequate education, knowledge, and interest in political affairs.On average, people are far more interested in the personal lives of pop starsthan the finer points of public policy. In this book I move beyond theusual suspects and argue that the failures of the average citizen can be


traced (at least in part) to the functioning of their personalities, which areconsistent, durable, mental processes that are not generally under ourdirect control. The decisions we make and the things we believe are heavilyinfluenced by brain processes that occur without our conscious awareness.

Thus, I am going to eschew the standard search for answers external tothe individual in a person’s environment, for example, education, socialcontacts, or media exposure. Instead I am going to turn inward and lookinto the minds of individuals for answers. This is not wholly new inpolitical science. Concepts such as schemas and heuristics have beenreferenced for decades (Kuklinski and Quirk 2000; Popkin 1991). Theidea is that individuals develop plans over the course of their lives abouthow to deal with common tasks, such as getting to work, as well as withcomplex ones, like whom to vote for. These tasks become automated andthe mind performs them without much conscious thought going into itfrom day to day.

While schemas and heuristics are undoubtedly a part of the functioningof the unconscious mind, work in the psychological, neuro, and geneticsciences over the past 30 years has begun to paint a far more active pictureof the unconscious mind and its influence on the decisions, attitudes, andchoices individuals make from day to day. This work has been embracedby numerous political science scholars who have begun to document theimportant effects of unconscious, affective, and cognitive processes alongwith their genetic antecedents and subsequent physiological responses(Alford et al. 2005; Fowler and Dawes 2008; Fowler et al. 2008;Hibbing et al. 2014b; Lodge and Taber 2013; Marcus et al. 2000,2007). Marcus (2013) sums up the interactive relationship between theconscious and unconscious by saying: “The brain is actively involved in theconstruction of reality; indeed without the active engagement we wouldnot be able to perceive the world at all. Perception is more than the passiverecording of sense-data. Perception is the joint result of sensory data andan active and theorizing brain” (137). The result is that in order tounderstand the causes of political behavior, it takes more than an explora-tion of consciousness because our very consciousness is predicated on theactions of our unconscious mind. The two are inextricably linked. Lodgeand Taber (2013) state the imperative to incorporate the functioning ofthe unconscious rather bluntly: “After five decades of well-replicatedresearch, it is simply no longer tenable for those interested in understand-ing political attitudes, public opinion, campaigns, media, or votingdecisions to ignore the effects of automaticity” (42).


The challenge for scholars when dealing with the unconscious is thedifficulty of measuring something that one cannot “see” happening.While researchers have recently begun using fMRI and other technicaladvances to identify brain activity and physiological changes in thebody (Hatemi and McDermott 2012), the vast majority of researchattempting to document the effects of automatic unconscious brainactivity has been based on experimentation in social and cognitivepsychology. This book builds on a large body of research into indivi-duals’ durable personality traits as the observable manifestations ofindividuals’ consistent unconscious processing of stimuli. There hasbeen a recent surge in scholarship focused on laying the fundamentalconnections between what are known as the Big Five personalitytraits (openness, conscientiousness, extraversion, agreeableness, andemotional stability) and basic political behavior (Dietrich et al. 2012;Gerber et al. 2012a, 2012b, 2010; Ha et al. 2013; Mondak 2010;Mondak et al. 2011; Mondak and Halperin 2008; Mondak et al. 2010;Schoen and Schumann 2007; Schoen and Steinbrecher 2013).

My thesis is that an individual’s personality is a durable and measurablemanifestation of unconscious brain processes that systematically influenceindividuals’ civic aptitude. The first part of this thesis is not in questionhere. There has been a veritable mountain of research over the last 80 yearsreaching across numerous disciplines documenting and measuring humanpersonality (see below). However, the second part—how personality influ-ences civic aptitude—(i.e., the quality of one’s engagement with thepolitical world) is still an open question. In particular, this book examinesthe relationship between the Big Five personality traits and how theycontribute to the successes and failures citizens have in understandingand participating in politics in four separate areas.

First, this research documents how personality contributes to the ideo-logical cognitive dissonance found when individuals express a belief thatthey are conservative, but prefer liberal policies (Ellis and Stimson 2012).Thus, I am interested in explaining personality’s connection to ideologicalconfusion, not its connection to a particular ideology. Second, I examinethe personality correlates of failing to understand facts about public policy.This failure is often a form of motivated reasoning where individuals shapetheir understanding of the political world in a biased partisan manner,even to the point of bending objective factual information to conform totheir partisan reality (Bartels 2002; Gaines et al. 2007; Jerit and Barabas2012; Jost and Amodio 2012; Kahan et al. 2013).


Third, the folk theory of democracy suggests that individuals should beable to connect their own personal policy preferences to the party orcandidate that best represents them. Indeed, Estlund (2008) argues thatdemocratic institutions begin to lose their legitimacy when citizens cometo their opinions unthinkingly. Unfortunately, research suggests that indi-viduals are woefully ignorant and rather than choosing parties or candi-dates based on their policy positions, individuals tend to either projecttheir own policy positions onto their preferred candidates or simply adoptthe preferences espoused by these candidates as their own (Jensen 2009;Lenz 2012; Levendusky 2009; Ross et al. 2012; Tomz and VanHouweling 2009). I build on this work and show how personality affectsthe ability of individuals to connect their own subjective policy preferencesto the proper political party. Finally, fourth, I look specifically at sub-merged welfare state policies and how individuals’ personalities contributeto a situation where one receives tremendous “hidden” benefits from thestate, but strongly opposes providing the more visible benefits, such asfood stamps, to others (Mettler 2011).

In what follows, I first survey research on the capabilities of democraticcitizens. Are they able to live up to the demands that democratic citizen-ship places on them? In other words, do they have a functional under-standing of politics, its issues, and its players? While the folk theory ofdemocracy may have always included an untenable conception of theaverage citizens’ capabilities, the reality of the institutional structure ofthe U.S. political system is that it depends on those very citizens. Thus,documenting the distance between theory and reality is incredibly impor-tant. Without spoiling the suspense for many readers, the simple answer tothese questions is no.

Given citizens’ inability to understand polices in anything approachingthe classic folk theorem’s comprehensive rational manner, I look toresearch attempting to explain why people fail so miserably to live up tothe democratic ideal. Variables like partisanship and lack of interest inpolitics are undoubtedly a culprit here, but I want to move further insidethe mind of individuals. I look to research in psychology, genetics, andneuroscience to come to grips with the reality that perhaps the bulk of thereason people often fail to be good democratic citizens is because we relyso heavily on our unconscious mind. Thus, I would argue that it is not the“fault” of individuals that they do not resemble the citizens contained inthe democratic folk theorem. People who do not vote are not exhibitingsome kind of moral failing. The reality is that the choice to gain political


knowledge and participate in elections is heavily influenced by brainprocesses occurring long before any conscious deliberation occurs.

The result is that the work I present here on the Big Five personalitytraits and other works documenting the role of unconscious behavioralinfluences absolves citizens of their culpability. The political system withinwhich they live was designed for people that have never existed. If I wereto walk into a home with a staircase where each step was ten feet tall,I would not blame the home owners for failing to ever visit the secondfloor. I would blame the architect. Thus, if a finger must be pointed for theproblems manifesting themselves in the American political system, itshould be pointed at the people who wrote the rules governing it. Thisis a topic I return to in the final chapter as I discuss changes to thestructure of democratic systems that take into account how people actuallyunderstand the world, rather than the pseudo-humans inhabiting classicfolk democratic theory.

Finally, after surveying research in psychology, genetics, and neuros-ciences, I turn my focus specifically to how scholars have begun toincorporate unconscious processes into models predicting political beha-vior. This leads directly to a class of variables that I refer to as theobservable manifestations of unconscious mind processes, with the BigFive personality traits being prime examples of such variables.

UP TO THE TASK OF DEMOCRATIC CITIZENSHIP?Mettler’s work may demonstrate that people often fail to understand whenthey are receiving help from government, but that doesn’t necessarilymean that they are not up to the task of being good democratic citizens.After all, failing to understand the delivery method of government pro-grams does not mean one cannot have a well-informed opinion of theprogram itself. But do they have a well-informed opinion? Unfortunately,70 years of research suggests they do not.

In 1940, a group of researchers led by Paul Lazarsfeld (Lazarsfeld et al.1944; see also Berelson et al. 1954; Hyman and Sheatsley 1947) set out todocument the effect of campaigns and the media on voters. The assump-tion at that point in time was that voters look to campaigns to learn aboutthe candidates, weigh their options, and choose the candidate they likedbest. What did they find? Contrary to expectations, they found that formost voters, their minds were made up well before the campaign had fullybegun. Voters relied on their standing decision rules, which in this case


primarily amounted to supporting their favorite party’s candidate, withlittle change, as the campaign progressed. Perhaps most disturbing, peopleappeared to be impervious to contrary argumentation and their under-standing of the issue positions of the candidates seemed to be based moreon wishful thinking than actual knowledge.

The work of another set of scholars from the University of Michiganalso proved seminal in the development of the study of mass politicalbehavior. With the publication of The American Voter in 1960,Campbell, Converse, Miller, and Stokes developed a theory of votingknown as the funnel of causality. They argued that in order to understandthe causes of an individual’s vote, one must begin with the psychologicalattachment to political parties. Partisanship is the most proximate of thesecauses of vote choice. Other variables exert less and less influence as onemoves outward from this psychological attachment. Individuals develop apartisan perceptual screen early in life through which they view the poli-tical world. As a result, everything they see and hear about in politics isfiltered through this screen. This echoes the earlier findings of Lazarsfeldet al. Citizens approach politics more like lemmings than independentindividuals.

Subsequent research on the role of partisanship has confirmed itsimportance as one of the best, if not the best, predictors of attitudes andvote choice (Flanigan and Zingale 2010; Goren 2005). But, partisanshipand ideology are not the same thing (Ellis and Stimson 2012). Theperceptual screen theorized by Campbell et al. does not preclude thepossibility that individuals fail to understand the important political issuesof the day entirely. Converse (1964) addresses political sophisticationdirectly by analyzing whether citizens use an ideological lens to under-stand and interpret politics. His focus is on ideology, or what he callsbeliefs systems, rather than partisanship. For Converse, in order to do thisone must, at a minimum, demonstrate constraint and interconnectednessamong one’s beliefs. In other words, thinking ideologically requires one tohold consistent beliefs and if one were to change one’s thoughts on aparticular topic, it would necessarily have an effect on other similar beliefs.Does Converse find that citizens analyze politics through an ideologicallens? The simple answer is no. Converse sums it up thusly:

For the truly involved citizen, the development of political sophisticationmeans the absorption of contextual information that makes clear to himthe connections of the policy area of his initial interest with policy

UP TO THE TASK OF DEMOCRATIC CITIZENSHIP? 9

differences in other areas; and that these broader configurations of policypositions are describable quite economically in the basic abstractions ofideology. Most members of the mass public, however, fail to proceed sofar. Certain rather concrete issues may capture their respective individualattentions and lead to some politically relevant opinion formation. Thisengagement of attention remains narrow however; Other issue concernsthat any sophisticated observer would see as ‘ideologically’ related to theinitial concern tend not to be thus associated in any breadth or number.The common citizen fails to develop more global points of view aboutpolitics. A realistic picture of political belief systems in the mass public,then, is not one that omits issues and policy demands completely nor onethat presumes widespread ideological coherence; it is rather one thatcaptures with some fidelity the fragmentation, narrowness, and diversityof these demands (pp. 246–247).

Converse’s conclusion was rather upbeat compared to that of JosephSchumpeter (1950). Schumpeter finds that “ . . . the typical citizen dropsdown to a lower level of mental performance as soon as he enters thepolitical field. He argues and analyzes in a way which he would readilyrecognize as infantile within the sphere of his real interests. He becomes aprimitive again” (p. 262). All told, the result of these early works was topaint a rather bleak picture of the American electorate. Philosophers hadspent centuries debating the virtues and pitfalls of democratic governance,but hard data was a rarity. For the first time systematic scientific methodswere being applied to the question of the competence of the averagedemocratic citizen and the results were frightening. Unfortunately, workmeasuring civic knowledge over the next 60 years has largely confirmedthese results (Bennett 1995; Caplan 2007; Converse 2000; Delli Carpiniand Keeter 1991, 1996; Shaker 2012).

While the lack of knowledge displayed by the average citizen isviewed as a problem by some, not everyone sees it that way (Page andShapiro 1992; Popkin 1991; Stimson et al. 1995). V.O. Key (1966) inhis defense of the American electorate famously opined, “ . . . that votersare not fools” (7). He rests his belief largely on aggregate correlationsbetween the voters and candidate positions. Key also conducted inter-views with voters that suggest a strong sense of retrospection on thepart of individuals when they enter the voting booth. A vast literaturehas been built on the concept of retrospective voting, where voters relyon a sense of how well things are going for them and the country as awhole rather than on specific issue knowledge (e.g., Downs 1957;


Kramer 1971; Jackson 1975; Fiorina 1981). These thoughts and feel-ings are then translated into a vote meant to either reward or punish theincumbent.

The strongest evidence to support the reward versus punishment dynamiccomes from work on the effects of the economy and individuals’ pocket-books (Killian et al. 2008). There appears little doubt that good economictimes help the incumbent president and his party, while bad economic timeshurt them (Kramer 1971; Lewis-Beck 1988; Lewis-Beck et al. 2008; Lewis-Beck and Paldam 2000; Lewis-Beck and Stegmaier 2000; Markus 1988;Nannestad and Paldam 1994; Norpoth et al. 1991). The effect of theeconomy on voting has been found to be true not just in the USA but acrossthe globe (Choi 2010; Cohen 2004; Duch 2001; Duch and Stevenson2008; Gélineau 2007; Gomez and Wilson 2006; Pacek and Radcliff 1995).

However, strong aggregate results like these often appear to be builton a rather rickety foundation once one looks beyond the magic ofaggregation and directly at the individuals behind it (Caplan 2007).Indeed, voters are often found to be myopic and shortsighted witheconomic perceptions heavily tinged by partisanship (Achen and Bartels2008; Bartels 2008; Enns et al. 2012; Gerber and Huber 2010; Lenz2012; Zaller 1992). Additionally, it often appears that in their zeal tohold politicians accountable for things going wrong, they may punishincumbents for things that they have absolutely no control over like theweather and shark attacks (Achen and Bartels 2016). Bartels (2008)points out, “real voters’ errors are quite unlikely to be random andstatistically independent . . .When thousands or millions of voters mis-construe the same relevant fact or are swayed by the same vivid cam-paign ad, no amount of aggregation will produce the requisite miracle—individual voters’ ‘errors’ will not cancel out in the overall electionoutcome” (47). Indeed, even a strong philosophical advocate for self-rule like Rousseau (1988) recognized the potential problem that largegroups of individuals myopically following a party or leader can posefor a democracy. He states, “But when factions, partial associations, areformed at the expense of the greater one, the will of each of theassociations becomes general with respect to its members, and particu-lar with respect to the state; it may then be said that there are nolonger as many voters as men, but only as many as there are associa-tions” (101). If the goal of research using aggregate correlations is toassuage concerns about the political ignorance of the mass electorate, itleaves a lot to be desired.


Fixing the Problem?

The great thing about political ignorance is that, at least theoretically, itcan be fixed. A more knowledgeable citizenry can be achieved througheducation. Scholars have advocated and documented the connectionbetween education, political knowledge, and a whole host of other civicvirtues (Converse 1972; Fishkin 1991; Galston 2001; Niemi and Junn1998; Putnam 2000; Shields and Goidel 1997; Verba et al. 1995;Wolfinger and Rosenstone 1980), although not all education is equallyup to the task (Hillygus 2005). However, the problem for democraticgovernance is more insidious than a simple lack of knowledge may sug-gest. This more serious problem was identified in Lazarsfeld et al.’soriginal work. Citizens do not just lack knowledge, they often appearwillfully ignorant and strongly resistant to anything that might contradicttheir beliefs. Indeed, Plato identified this particular problem with theaverage individual more than 2000 years ago.

In The Republic (2007) Plato tells an allegory about a group ofindividuals who have spent their entire lives chained up in a cave facinga blank wall. There is a fire behind them that they cannot see. Theirentire reality consists of the shadows they see dancing across the wall.Unbeknownst to the captives, their captors use puppets to project theseshadows. With this setup, Plato has created several layers between theprisoners’ reality of shadows on a wall and the reality outside the cave.But the prisoners do not know that. For them the word tree refers tothe shadow on a wall, not a green leafy thing with a brown trunkgrowing out of the ground. As one would expect, the prisoners filltheir lives with made-up stories about the shadows. They even conveyhonors and prestige on the individuals who are best at predicting theoccurrence of shadows. One day one of the prisoners escapes the caveand sees the real world for the first time, a real tree and a real house.He returns to the cave. For him, the honors and knowledge thisprimitive society had produced had lost all meaning, since they werebased on ignorance of reality. The returned prisoner tells of the newreality he had seen, but the remaining prisoners think him a fool andlaugh at him. They resist the new information that challenges theirbeliefs about the world. In fact, in Plato’s story the remaining prisonersdecide that they should kill anyone who attempts to show them thisnew reality. That is how strongly they hold to their incorrect beliefs.Plato understands how resistant ignorance can be to enlightenment.


The modern conception of this problem has many facets and typicallyfinds its roots in psychological theories of the mind. Much of this workexposes how individuals actively shape the content of information theyseek out, receive, understand, and remember (Zaller 1992). Cognitiveconceptions such as dissonance theory (Festinger 1957), selective percep-tion (Abelson et al. 1968; Sherrod 1971–72), and balance theory (Heider1958), all theorize about how attitudes are formed and when they aremore or less likely to change. One common thread in this is that indivi-duals prefer consistency in their thought processes. Inconsistent beliefsand information cause a sort of cognitive agitation within the individualthat ultimately needs to be resolved. This is how attitudes can potentiallychange. The problem is that individuals actively avoid information andthoughts that are inconsistent with what they already think and know. Inother words, people are motivated to protect their current beliefs andavoid contrary information (Druckman and Bolsen 2011; Kunda 1987,1990; Lebo and Cassino 2007; Leeper and Slothuus 2014; Lewandowskyet al. 2013, 2012; Lodge and Taber 2013; Nir 2011; Redlawsk et al.2010; Rudolph 2006; Slothuus and De Vreese 2010; Taber and Lodge2006).

People do not randomly select bits of information from all that isavailable, or passively listen to whatever happens to come their way.They choose what they see (Frey 1986; Lau and Redlawsk 2006; Lavineet al. 2005; Sweeney and Gruber 1984). But they do not seek outinformation in a rational manner in order to have the best and mostcomplete understanding of an issue. They seek out information that con-firms what they already know (Nickerson 1998; Olson and Zanna 1979).Even when they do get new information, they do not update their beliefs,but stick with their original understanding (Druckman et al. 2012). Thatis, unless they have the proper partisan cue, at which point people turninto the Borg and rapidly assimilate the new information that is consistentwith their partisanship without critical thought (Cohen 2003; Smith et al.2012a).

When individuals accidentally hear something that contradicts whatthey believe, they disregard it (Taber et al. 2009). If they hear some-thing that they are unsure of, they do not do more research in order tofully understand the issue. They simply interpret the new information sothat it is consistent with what they already know (Fazio andWilliams 1986;Lord et al. 1979; Vidmar and Rokeach 1974), or simply adopt theirpreferred candidate or party’s position (Brader 2012; Bolsen et al. 2014;


Cohen 2003; Druckman et al. 2013; Lenz 2012; Petersen et al. 2013).When canvasing their memories, individuals do not take a representativesample. They select memories that are consistent with their current attitude(Zanna and Olson 1982). When voters face ambiguity in the policy posi-tions of their favored candidate, they happily fill in the specifics to theirliking (Sherrod 1971–72; Tomz and Van Houweling 2009). Importantly,education and knowledge do not solve these problems. Highly knowledge-able individuals bend reality when asked factual questions to suit theirpartisan beliefs (Kahan et al. 2013, 2012). When new political controver-sies and scandals break, it is the politically sophisticated who rely on simplepartisanship to guide their understanding of the issue (Wagner et al. 2014).

The mental gymnastics that people engage in so that they can maintaintheir illusions of cognitive integrity or balance are astounding. This is notlimited to how they structure their own thinking, but includes what theythink about the opinions of others. Concepts such as looking glass percep-tion, pluralistic ignorance, and the third person effect all help individualscreate their own special worlds of make-believe (Fields and Schuman1976; Davison 1983; Shamir and Shamir 1997; Taylor 1982). All ofthese concepts suggest a tendency on the part of individuals to holdmistaken beliefs about what others think. As one might suspect from thename, looking glass perception is self-serving. To the extent that indivi-duals make this mistake, they assume others think like them (Van Bovenet al. 2012). Indeed, Fields and Schuman conclude, “ . . . people appear tolook out into the world and somehow see their own opinions reflectedback” (437). Pluralistic ignorance is the tendency to overestimate supportfor social norms. This can be a big problem when it allows for thepersistence of harmful norms (Bicchieri and Fukui 1999; Young andWeerman 2013). Finally, the third-person effect documents the hubrisof individuals as they assume others are affected by the media more thanthey are. Unfortunately, they are not (Andsager and White 2007; Cohen2003; DeLorme et al. 2006). Indeed, the less one knows, the more likelyone is to overestimate their own expertise (Kruger and Dunning 1999)with individual extremism often supported by the illusion of understand-ing (Fernbach et al. 2013).

In summary, research suggests that the average individual lacks politicalknowledge; actively avoids information that might contradict what theybelieve; ignores or rationalizes away the disconfirming information theydo receive; retrieves information from memory in a biased and self-servingmanner; tends to have mistaken beliefs about what others think; and


overestimates the difference between how much they are personally affectedby new information versus how much others are affected. Indeed, Shermanand Cohen (2006) call it a “psychological immune system” (184) whenfaced with information that may be threatening to our self-integrity.Importantly, almost all of this is occurring without people consciouslythinking about it. In other words, what unifies so many of the documentedfailures of democratic citizens is that most of our interaction with andunderstanding of the world occurs automatically (Bargh 1997; Bargh andMorsella 2008). Unfortunately, that means relying on preconsciousmechanisms, such as heuristics, that are designed for quick and effortlessresponses to common occurrences. This can be helpful in many ways, butvery harmful in others, particularly when it drives intolerance for people orgroups that have different opinions (Brandt et al. 2014; Chambers andMeInyk 2006; Henry and Reyna 2007).

WE KNOW WHAT YOU’RE THINKING: THE MOVEMENT

TO THE UNCONSCIOUS MIND

There has been a strong push among some political science scholarsto focus research more directly on the important role that the uncon-scious mind plays in political decision making. The revolution thatthis portents has, perhaps, not been fully grasped by many scholars.For instance, the psychological concept of priming suggests that thestandards by which individuals evaluate some object (such as the pre-sident) can be altered through priming (Althaus and Kim 2006; Kinderand Iyengar 1987). In other words, the bits of information that comeeasily to mind change as a result of this process. Increased news cover-age of military conflicts primes the concept of military conflict in theminds of individuals. As a result, when asked to evaluate the president’sjob performance, these thoughts will be more easily accessible thanother thoughts about the president and, therefore, they will be morelikely to evaluate the president’s job performance based on how he hashandled military conflicts (Miller and Krosnick 2000). The scholarlytreatment of priming in political science leaves one with the impressionthat the individual’s conscious mind is still in charge and in controlof choosing, albeit choosing among a different set of options thanmay have otherwise been the case had priming not occurred. Butthat may not necessarily be the case.

WE KNOW WHAT YOU’RE THINKING . . . 15

The most fascinating and cutting edge research on the relationshipbetween the unconscious and conscious mind suggests that perhaps it’sthe unconscious mind that has the upper hand. This is a revolutionaryconception of how the mind works. Patrick Haggard (2008) summarizedthis relationship thusly, “Modern neuroscience rejects the traditionaldualist view of volition as a causal chain from the conscious mind or‘soul’ to the brain and body. Rather, volition involves brain networksmaking a series of complex, open decisions between alternative actions”(944). Research in neuroscience has spent the last 40 years examining theactivation of neural activity in temporal relation to when individuals areconsciously aware of making a decision. The simple idea is that if the brainstarts to work before we are consciously aware of making a decision, thenwe are not consciously in control of our decisions.

Researchers can test this in multiple ways, but the classic test comesfrom Libet et al.’s (1983) work where participants’ brain activity is mon-itored as they voluntarily choose to move their finger. Participants are alsoasked to indicate the moment they become consciously aware of theirdecision to move their finger. Since this is voluntary movement, nobodyoutside of the participant knows when the choice to move the finger willoccur, but researches can watch the brain activity and see the brain changein order to move the finger. The startling finding is that neural indicatorsof an intention to act are present about 350 milliseconds prior to anindividual being consciously aware of making a decision to act (Haggardand Eimer 1999; Libet 1985). Subsequent work has stretched the timeinterval between brain activity and conscious awareness to several seconds(Bode et al. 2011; Haynes 2011; Schlegel et al. 2013; Soon et al. 2008,2013). The implication of these findings is that behavior is not consciouslycontrolled.

Much of this work has been based on rather rudimentary tasks such asthe decision to move a finger. But the question still remained, does brainactivity also precede more complex judgments, such as if one likes ordislikes a particular piece of art? Colas and Hsieh (2014) investigatedthis exact question. They used fMRI brain imaging while participantsviewed one of nearly 300 abstract images for 500 milliseconds, that is,just barely long enough to consciously perceive that the image is there.Subjects were then asked to choose “like” or “dislike.” The authors foundthat brain activity occurring up to 6 seconds prior to the presentation ofthe image significantly predicted individuals’ aesthetic response to theimage. In addition, when subjects were asked after the experiment why


they liked or disliked an image, they mentioned features of the images thatthey liked or disliked. One interpretation of this result is that participants’aesthetic responses were determined by brain states prior to viewing theimage and then they rationalized their responses after the fact.

Another important demonstration of the influence of the unconsciouson conscious decision making has been built around what is known as theIowa Gambling Task (IGT). The impetus behind Bechara et al.’s (1994)work was the observation that individuals with lesions on the ventromesialfrontal lobes of their brains appeared to behave normally. The only pro-blem was that they could not learn from their past mistakes. The immedi-ate reward always appeared to trump any long-term concerns (Becharaet al. 2000). They hypothesized that the missing piece for these individualswere unconscious brain mechanisms that help to incorporate certain typesof experiences into current decision making. In order to test this, theauthors came up with a rather ingenious card game. First, they gaveparticipants $2000 in play money and sat them in front of four decks ofcards. The goal for participants was to gain as much money as possible bychoosing cards from any of the four decks in any order without knowledgeof a limit to the number of times they would be choosing cards (ultimatelychoosing was stopped after 100 cards). The payoffs and penalties variedacross decks. The first two decks had winnings of $100, while the secondtwo decks hand winning of only $50. The problem was that the first twodecks also had higher penalties than the second two decks to go along withtheir higher rewards. As a result, after pulling ten cards from the first twodecks, individuals would have won $1000 but incurred $1250 in penalties,resulting in a net loss of $250. After pulling ten cards from the second twodecks, the subject would have won $500 with only $250 in penalties,resulting in $250 in winnings. Thus, the second two decks provided betterlong-term payoffs.

So where does one start when faced with a blank set of cards and noinformation about what they contain and the only directive is to winmoney?As onemight suspect, people start randomly drawing cards. Since the playershave no way of predicting when penalties will occur and no idea how longthey will be drawing, there is no way for them to calculate with any precisionthe odds of gaining or losing. As a result, one might suspect that people willcontinue to draw cards randomly. But they don’t. The individuals withhealthy brains began to gravitate toward the two decks with better long-term payoffs. Indeed, they produce anticipatory skin conductance responses(SCRs) whenever they ponder selecting a card from the high loss/high

WE KNOW WHAT YOU’RE THINKING . . . 17

reward deck long before they become consciously aware of the quality of thedecks. These SCRs began to appear in as few as ten draws. After 20 draws,participants all claimed to have no idea what was going on. Yet, their bodyhad already responded. After 50 cards, those individuals with healthy brainsclaimed to have a hunch about the bad decks and by 80 draws they claimedknowledge of why the decks varied. Interestingly, a few of the healthy brainparticipants never claimed to understand what was going on, yet they stillgravitated to the more advantageous decks. The individuals with lesions ontheir ventromesial frontal lobes never generated SCRs and never stoppedselecting cards from the bad decks (Bechara et al. 1997).

The intriguing result here is that it appears that our unconscious is a bettergambler than our conscious. Our unconscious is making risk assessmentsoutside of our awareness (Huang et al. 2014) and then producing bodilyresponses (i.e., SCRs) that help our conscious selves catch up. The researchof Peatfield et al. (2012) and Paynter et al. (2010) shows a similar importantrole for the unconscious in problem solving. The value (or perhaps super-iority) of unconscious decision making has been demonstrated in numerousways and is particularly the case with complex decisions (Evans et al. 2003;Dijksterhuis 2004; Dijksterhuis et al. 2006; Nisbett and Wilson. 1977;Newell and Shanks 2014; Usher et al. 2011; Wilson 2002).

However, it is not just decisions about how to choose between cardsthat appear to be heavily influenced by the unconscious mind. Creativityhas also been the subject of study and recent work has begun to unravelthe important contribution that unconscious mental processes play (Kühnet al. 2013; Ritter and Dijksterhuis 2014). Other research has documen-ted unconscious word manipulation (Armstrong and Dienes 2013; VanGaal et al. 2014) and the significant effect of implicit memory activationon thought processes (Voss et al. 2012). Indeed, Pérez (2016) movesoutside the laboratory and documents the varied effects of implicitthought processes on attitudes toward both legal and illegal immigration.In addition, the unconscious has been shown to be a good predictor ofelection outcomes based solely on the faces of the candidates (Olivola andTodorov 2010; Todorov et al. 2005).

All told, the inescapable conclusion is that the unconscious mind is anactive participant in constructing our reality (Stroop 1935). It works quicklyto interpret sensory or other data, even when this data is present for far tooshort a time for our conscious minds to be aware of it, or so incidental thatone would assume that it could not have an important effect (Ansorge,Kunde, Kiefer 2014; Britz et al. 2014; Izatt et al. 2014; Molden 2014;


Montoro et al. 2014; Weibel et al. 2013). For instance, it is not surprising tolearn that individuals prefer the familiar. However, it is the mere exposure to astimulus that can make it familiar. One does not even need to be consciouslyaware of the exposure for the stimulus to have an effect (Kunst-Wilson andZajonc 1980; Yoshimoto et al. 2014; Zajonc 1968). This is an unequivocaldemonstration of the incredibly important role that the unconscious mindplays in driving behavior. Conscious awareness is not always necessary, per-haps rarely necessary, but unconscious brain functions are. Countless studiesof individuals with damage to their brains of one sort or another havedocumented how our conscious selves are entirely dependent on properbrain functioning (Demertzi et al. 2013; Gainotti 2012; Giacino et al.2014; Herbet et al. 2014). In the end, our consciousness is not a magicalconstruct separate from the tangible physical connections of our brains.

THE UNCONSCIOUS APPROACH TO POLITICS

When the work of cognitive, social, and political psychology are combinedwith neuroscience’s exploration of brain activity, one begins to build a morecomplete understanding of the process of human decisionmaking. The simplefact is that the conscious mind cannot be separated from the unconsciousmind. Recent research into the unconscious mind in political science hasbegun to produce exciting results. But how have scholars approached theintegration of unconscious processes into models of political behavior?Research tends to take one of two paths to answer this question. The first isthe newest and is a direct examination of genetic, neurological, and physiolo-gical processes and activity. The second approach is focused onmeasuring theobservable manifestations of unconscious mental processes, which comes outof the classical psychological approach to brain research. Ultimately, thefollowing discussion of these two lines of research will lead to an elucidationof a class of variable that I have referred to as observable manifestations ofunconscious brain processes, which are important connector variablesbetween the hidden activity in our brain and the attitudes and behaviors wesee every day.

Genetic, Neurological, and Physiological Influences and Politics

The first approach to understanding the unconscious mind’s effect onpolitical behavior looks directly at physiological structures of the brainand their genetic precursors. This research has been spearheaded by the

THE UNCONSCIOUS APPROACH TO POLITICS 19

works of Fowler and Dawes (2008) and Hibbing et al. (2014b; see alsoAlford et al. 2005; Fowler and Dawes 2013; Gruszczynski et al. 2013;Hatemi et al. 2010, 2011; Hibbing and Smith 2007; Smith et al. 2012b,2011). They move well away from the basic assumption that people areconsciously in control of how they feel, think, and act. These authors pointto the significant influence of genetics and physiology on attitudes andchoice, but they are not determinative. Genetics and physiology work inconjunction with psychological processes and predispose people to certaintypes of responses. These predispositions vary from person to person. Forexample, some people are strongly predisposed to participate in politics,while others are not. That does not mean that the latter will neverparticipate in politics, but simply that they are less likely. As a result,more effort would be required to coax these individuals out of their shells.This is not just important information to political campaigns, which spenda tremendous amount of time and resources trying to get people to thepolls. It also speaks to the likely distribution of preferences being givenvoice through voting or any other means of influencing political out-comes. Not everyone is equally likely to participate, and a large part ofthe reason has nothing to do with conscious deliberation.

Much of the work documenting the influence of genetics on politicalattitudes and behavior is based on the study of twins (Alford et al. 2005;Bouchard et al. 2003; Friesen and Ksiazkiewicz 2015; Verhulst et al.2012). Thus, it does not provide a direct observation of brain activity orexplicit connection of genetic information to political attitudes andbehavior. Twin studies work by leveraging a comparison between identical(monozygotic) and fraternal (dizygotic) twins. The logic is that if identicaltwins (who share 100 percent of their genetic material) are more similar onsomething like ideology than fraternal twins (who typically only shareabout 50 percent of their genetic material), then the difference is due togenetics rather than environment. Critics of the twin design have pointedout the potential problem that it assumes that identical twins and fraternaltwins experience similar environments (Joseph 2013), although others havepointed out that these studies do in fact account for gene-environmentinteraction (Verhulst and Hatemi 2013). Political ideology has seen quite abit of scholarly focus in this area. Hatemi et al.’s (2014) recent massive crossnational twin study provides strong evidence that political ideology isdriven, at least in part, by one’s genetic makeup (see also Bell andKandler 2015; Fazekas and Littvay 2015). However, twin studies leaveopen the question of how genetic makeup is ultimately translated into


political attitudes and behavior. They also have little to say about con-scious control of one’s relation to politics.

Works drawing a more direct connection between genetics and brainstructures have shown intriguing results, too. The work of Kanai et al.(2011) has documented a connection between conservatism and the sizeof various regions of the brain. Using fMRI brain imaging, scholars havebegun mapping the divergent brain activity of individuals while readingliberal or conservative statements (Rule et al. 2010; Zamboni et al. 2009).The interesting result is that regardless of one’s own ideology, liberalstatements activated different regions of the brain than conservativeones. Takeuchi et al. (2016) show similar differences in gray matterassociated with nationalism and patriotism. In another fascinating study,Kaplan et al. (2007) document a more active role for the mind whenviewing pictures of presidential candidates. They show that brain activity isconditional on one’s partisanship, thus capturing a motivated brainresponse to familiar political faces (Knutson et al. 2007). In a similarvein, Gozzi et al. (2010) show differential brain activity among thosewho are interested in politics when compared to those who are not.Westen et al. (2006) capture motivated partisan reasoning in action.Partisans facing information that is inconsistent with what they alreadybelieve produces brain activity in areas of the brain associated with painand negative emotions. This would seemingly make it very difficult forpartisans to treat this information with an even hand when it activatesthese negative areas of the brain automatically. Moving away from the useof fMRI, Dodd et al. (2012) document that those on the right direct moreattention and have a stronger physiological response to aversive images(see also Hibbing et al. 2014a; Joel et al. 2014). Finally, Schwabe et al.(2016) document the interaction between genetic and environmentalinfluences on ideology.

What all these works have in common is the recognition that theseeffects are generally outside of the control of individuals. We do notconsciously choose which areas of the brain are activated when we see apolitical figure or consciously choose to focus more on negative imagery.They are automatic reactions to the stimulus built into our unconsciousmind. The key insight from all this work is that these automatic brainprocesses have an important effect on how people understand the politicalworld. Focusing directly on the activity of the unconscious mind, Tuscheet al.’s (2013) recent work provides powerful documentation of thisprocess at work. They utilize fMRI imaging in connection with


participants’ reaction to politicians and parties. They find “ . . . that brainresponses can reflect individual political preferences—for politicians andassociated parties—in the absence of conscious deliberation and attention”(179). In other words, the unconscious mind reacts in a partisan manneron its own. Of course, these works are still at the very earliest stages of thiskind of research (Krastev et al. 2016), but they represent the cutting edgeof research on the causes of mass political behavior. Even limited to theseearly findings, the presumption that political actors are consciously con-trolling their behavior, subject only to external environmental influences,is simply untenable.

Classic Psychological Approaches to Politics

Research employing the second methodological approach to understand-ing the connection between the unconscious mind and political behavioris much older than the first and has been based heavily on social andcognitive psychology. The vast majority of research on the effects ofmotivated reasoning discussed above provides an example of thisapproach. In these works, scholars do not or simply cannot directlyobserve brain activity and the schemata or heuristics constructed in indi-viduals’ minds. However, they can theorize that, for example, givenscholarly knowledge of partisanship and motivated reasoning, if partisan-ship behaves as a perceptual screen, then we should see individuals inter-preting their world to fit with their preferred party, even to the point ofbending facts to fit one’s partisanship (Bisgaard 2015; Claassen and Ensley2016; Kahan et al. 2013; Nawara 2015). As a result, scholars have beenable to capture the effects of unconscious brain activity, even though theyare not directly observing it. The exciting connection between thisapproach and the genetic/neuroscience approach is that many of thepsychological brain processes theorized over the last century are nowfinding validation in brain imaging (Jost et al. 2014).

Lodge and Taber (2013; see also Taber and Lodge (2016); for asuccinct distillation of their ideas) have developed a comprehensivemodel of conscious and unconscious thought processes that connectsthe two through continuous on-line processing of new information com-bined with associative memory. They refer to this as a dual-process modelof information processing where automatic brain processes and consciouscontrolled processes interact in a continuous autoregressive dance.Importantly, unlike much past work, they do not give primacy to the


conscious mind saying: “Rather than controlling our actions, conscious-ness often functions to interpret and rationalize the actions and processesthat have already been carried out unconsciously” (47–48). Hence thename of their book, The Rationalizing Voter. They argue that peopleautomatically incorporate information into their existing beliefs, but theydo it in a biased fashion. “The process is autoregressive: the summaryjudgment is not based on a fresh look at all the evidence, but rather on theOL [on-line] tally and the biased assessment of messages at the instantthey are received” (56). The key, of course, is that this is an unconsciousprocess. Our minds do it for us and, unfortunately, they do not do an evenhanded job of it.

All the millions of bits of information we receive every day are tagged,catalogued, and connected to our existing attitudes and beliefs by ourunconscious minds. This can produce a “snowball” effect as we head downone particular attitude path. The more information we have tagged in ourminds pointing us in one direction, the more likely any new informationwill be processed in a manner that continues to point us in that samedirection, regardless of the content of that new information. Our uncon-scious minds have essentially locked us onto one particular path. Theunsettling result is that those with the most knowledge about a topic arealso the most likely to be driven by the unconscious mind. The politicallysophisticated have spent their entire lives applying biased partisan affectivetags to new information. This makes it nearly impossible for them to seeany alternative. One cannot help but see a connection between Lodge andTaber’s model and the work of Tusche et al. (2013) that, essentially,documents this unconscious, partisan, associative brain process at work.

Marcus et al.’s (2000) theory of affective intelligence also falls withinthe rather broad umbrella of scholarly work focused on measuring theobservable manifestations of unconscious mental processes, but again withstrong ties to work in neuroscientific research on brain activity. They arguethat our mental engagement with the world is largely governed by adispositional system that is largely unconscious with automatic processingof common occurrences. This is governed by an emotional sense ofenthusiasm. As long as everything is going as expected, we move throughlife without needing to expend too much conscious mental effort. Forexample, millions of people drive to work every day. Undoubtedly, thisrequires some mental processing of information, but once one has donethe same drive a few hundred times, it becomes automatic and unlesssomething odd occurred on the way to work, it is forgotten. However,


should something odd occur on the way to work, for example, a caraccident, our surveillance system, also known as the behavioral inhibitionsystem (Gray and McNaughton 2000), activates our sense of anxietywhich wakes us from our mental torpor and stimulates our consciousmental processes. The surveillance system is designed to monitor andrespond to threats. The interesting relationship between the dispositionaland surveillance systems is that the surveillance system is working whollyon its own in unconscious neural networks (Gray 1985), with recentresearch documenting the genetic and physiological predictors of theseprocesses (Davies et al. 2013; Fox et al. 2005; Hogan et al. 2007).

Ted Brader’s (2006; see also Groenendyk and Banks 2014; MacKuenet al. 2010; Miller et al. 2009; Valentino et al. 2011) work catches theseprocesses in action in relation to campaign advertising. Campaign advertis-ing nearly always includes some type of emotional appeal to voters. Bradershows how it is the anxiety raising advertising that has the greatest possibilityof making people deviate from their normal voting routine, while enthusias-tic appeals simply reinforce them. In other words, voters’ surveillance sys-tems are monitoring the political landscape for novelty. An advertisementcan present new negative information to the viewers. The surveillance systemthen sounds the alarm and voters’ consciousness is then asked to deal withthis new information. They may simply argue against this new informationthat contradicts what they already think, as research on motivated reasoningsuggests (Taber et al. 2009). But at least they are thinking. If they arethinking, there is a chance that they may change their mind. This is exactlywhat Brader finds. Anxiety raising ads are the ones that have the potential tochange minds, even if it is only a small chance.

Of course, the potential that anxiety may actually be helpful in apolitical context (see MacKuen et al. 2010) has been countered byAlbertson and Gadarian (2015) who show in their work that anxietydoes indeed lead to an increased interest and search for new information.However, that search is biased toward negative information and leads to adesire for protective policy. Furthermore, Weeks (2015) shows that anxi-ety can leave individuals susceptible to misinformation. Thus, anxiouscitizens do not magically become ideal democratic citizens seeking outobjective information in order to obtain the best possible understandingof the issue causing them anxiety. They seek out further confirmation oftheir anxiety in negative information and look to perceived experts to solvethe problem. These newly engaged citizens are quite predictable and in noway are they open minded.


Observable Manifestations of Unconscious Brain Processes

I have referred to variables that I call observable manifestations of theunconscious mind numerous times, but have not explicitly outlined what Imean by this concept. At this point, I have laid enough of the groundwork to address this idea directly. As outlined above, there is a bounty offascinating works on the influence of the unconscious brain on ourthoughts and behaviors across numerous academic fields, the implicationsof which are far reaching. However, what I am most interested in is howthese mental processes influence the political world in general and speci-fically in regards to civic aptitude. The problem for those interested indocumenting the influence of the unconscious brain on real world politics,is that observing mental processes in action is difficult. It typically requiresthat subjects be brought into a laboratory where scholars can be reason-ably certain that the behaviors and attitudes being expressed by subjectsare the result of the experimental manipulation. The obvious problem thatarises in these situations is one of external validity. The best laboratoryexperiments only demonstrate what can happen, they do not documentwhat does happen.

I want to know how much the unconscious mind influences civicaptitude, for example, how well people understand the important politicalissues of the day, or how well people can connect their own policypreferences to the party that best represents them. To put it anotherway, experimental research has documented that exposing people to animage of a happy face so fast that they cannot have become consciouslyaware that they saw it, leads to an increase in positive associations withpolitical figures they are asked about later. While intriguing, this informa-tion does not tell one anything about when, where, or how much of aneffect this kind of automaticity has on people making political decisions inthe real world. To be clear, in one sense scholars know a lot. They knowthat the unconscious mind affects every person, every day, all the time. Butin knowing this, scholars essentially know nothing. It is not until we canmeasure differences between people and how their individual unconsciousminds work that we can begin to leverage this knowledge as a variablepredicting things like votes or policy positions.

What can researchers interested in documenting the unconsciousmind’s influence on political behavior do? The tools we have right noware connector variables, which I refer to as the observable manifestationsof unconscious brain processes. These are things like emotions such as


anxiety, enthusiasm, and anger. They are also concepts like personality andpartisanship. What all these variables have in common is that there isevidence that they are influenced (not wholly determined) by our uncon-scious brains and, in turn, that these variables influence political behaviorand understanding. Thus, these variable make the connection betweenactivity in the unconscious mind and real world political engagement. Thisis important because these variables give scholars the ability to documenttheir effects the same way political scientists have documented the effectsof common variables like age, income, race/ethnicity, and gender. Thus,while in this book I may be measuring one’s personality and demonstrat-ing its effect on various aspects of civic aptitude, I would argue that what isactually being documented are the effects of predictable patterns ofunconscious brain activity. The Big Five is a convenient name for five ofthese patterns.

PERSONALITY TRAITS, POLITICS, AND THE OBSERVABLE

UNCONSCIOUS

Another major line of research designed to measure the observable effectsof unconscious processes is focused on personality traits. Recent years haveseen a relative boom in the number of scholars incorporating measures ofpersonality into their mass political behavior models (Dietrich et al. 2012;Gerber et al. 2012a, 2012b, 2010; Ha et al. 2013; Jost et al. 2003a;Mondak 2010; Mondak et al. 2011; Mondak and Halperin 2008; Mondaket al. 2010; Schoen and Schumann 2007; Schoen and Steinbrecher 2013).In his recent book on personality and politics, Jeffery Mondak (2010)defines, “ . . . personality as a biologically influenced and enduring psycho-logical structure that shapes behavior” (6). The key observations here are,first, that individuals’ personalities have their roots in genetic and physio-logical processes. Second, that they are relatively stable, although person-ality traits do evolve over the course of one’s lifetime, with the typicalpattern being an emergent solidification of an individual’s personality inadolescence, an increase in the stability of personality traits through mid-dle adulthood and then decreasing as one enters old age (Bleidorn et al.2014; Cobb-Clark and Schurer 2012; Gerber et al. 2013a; Hopwoodet al. 2013; Kandler 2012; McAdams and Olson 2010; Roberts andMroczek 2008; Roberts et al. 2006). While it is clear that one’s personalitytraits are not inviolable, they do not change rapidly and are not under the


conscious control of individuals. Thus, personality is not unlike other com-mon political variables like partisanship, income, age, or ideology. Variablesdo not need to be static throughout one’s life for them to be useful.

The scientific investigation of the genetic origins of personality hasprovided interesting results, although they may not be as conclusive asmany had hoped. Research in Behavioral Genetics, relying primarily ontwins studies, has estimated the heritability of personality traits to rangefrom 30 to 60 percent (Bouchard and Loehlin 2001; Cattell et al. 1981;Eysenck 1990; Krueger and Johnson 2008; Mann et al. 2015; Savitz andRamesar 2004; Turkheimer 2000). This work has left little doubt thatgenetics plays a significant role in driving personality. The genesis of thisresearch path long predates the ability to examine minute genetic pro-cesses and the role of specific genes. With the sequencing of the humangenome completed in 2003, there was much scholarly hope that research-ers would be able to pinpoint the contributions of specific genes tocomplex phenotypes like personality. However, that hope has provenrather naive. As it turns out, the difficulty involved in finding associationsbetween complex behavior and genetic information has proven extremelytough to overcome. The thought that there is one or a handful of genesresponsible for something like personality is now understood to be foolish(Turkheimer 2011; Turkheimer et al. 2014).

This is not just true for personality, where some people may still find thegenetic connection dubious. It is also true for something as seeminglyuncontroversial as the heritability of height. There is no debate regardingthe effect that genetics has on one’s height (Turkheimer 2011). But justlike with personality, genetics researchers have found it difficult to pin-point the exact genes responsible for causing this process. The reason forthis seeming failure is the fact that there isn’t a “height gene” any morethan there is a “personality gene.” There are presumably infinite combina-tions of genes in the human genome that could potentially affect theseprocesses. To date, scholars have begun to narrow down a few potentialcandidates (or loci), but the tests they use to document the effects areoften unable to detect extremely small individual genetic contributions(Luciano et al. 2012; Montag and Reuter 2014). As a result, phenotypicprocesses that are generated from complex genetic interactions will beexceedingly difficult to pinpoint, at least with the capabilities scientistsnow possess.

For example, in their recent meta-analysis of research attempting to finda genetic association with personality, De Moor et al. (2012) find a

PERSONALITY TRAITS, POLITICS, AND THE OBSERVABLE UNCONSCIOUS 27

potential connection when it comes to openness to experience and con-scientiousness, but fail to find significant results relating to neuroticism,extraversion, or agreeableness. They conclude that “the findings of thisstudy show that large-scale collaborative studies with combined samplesizes in the order of thousands still have difficulties in identifying commongenetic variants that influences complex phenotypes such as personalitytraits . . .Larger GWA [Genome-Wide Association] studies may revealthese variants, as has been already successfully shown for human heightin a larger meta-analytic GWA study of over 180,000 individuals, in whichat least 180 loci were identified together explaining about 10% of thevariation in height” (345). It took anN of over 180,000 to explain just 10percent of variation in height. If it takes that kind of work to produceresults explaining height, than the work needed to explain something farmore complex, like personality, is hard to comprehend. Despite the diffi-culty of the task, scholars continue to take baby steps toward a completeunderstanding of personality’s genetic underpinnings (e.g., De Moor et al.2015; Kim et al. 2015; Okbay et al. 2016; Power and Pluess 2015; Smithet al. 2016; Van Den Berg et al. 2016).

Following their own path to drawing connections between person-ality and physiological brain processes, scholars using neuroimagingtechnologies have been successful in mapping differential brain activitiesassociated with different personality types (Canli 2004; Grimm et al.2012; Schaefer et al. 2012, 2013; Servaas et al. 2013). For example,Aghajani et al. (2014) show a connection between extraversion andneuroticism and the resting-state functionality of the amygdala, suggest-ing the biological origins of one’s susceptibility and/or resilience toemotional disorders, which manifest in personality. These types of neu-rological studies are not limited to simple differences in the processingof the brain but also have documented variability in the physical struc-ture of gray matter, the amygdala, and medial orbitofrontal cortex in thebrain in relations to personality differences (DeYoung et al. 2010;Kapogiannis et al. 2013; Omura, Todd, and Canli 2005; Wright et al.2006). Indeed, Taki et al. (2013) find that being high in openness toexperience plays an important role in preserving gray matter and cogni-tive functioning. In addition, the recent work of Sampaio et al. (2014)documents specific neuroanatomic correlates with each of the Big Fivepersonality traits and Beaty et al. (2016) find a biological basis for thetrait openness to experience that suggests it may influence the veryarchitecture of the brain (see also, Ikeda et al. 2014; Lei et al. 2015).


All told, the combined efforts of scholars across numerous disciplineshave produced significant positive results. Of course, some areas have hadmore success than others. Behavioral Genetics appears all but certain ofthe genetic connection to personality, while geneticists examining thehuman genome directly are still coming to grips with its vast complexityand the monumental task of connecting it to complex phenotypes.Neuroscientists have had tremendous success in making the connectionbetween personality traits and brain structure and activity, but are leery ofmaking claims about causation, that is, whether the observed brain activitycauses one’s personality, or whether one’s personality causes the observedbrain activity. Therefore, much work in these areas is left to do, but even atthis extremely early stage in the process, there is reason to be confidentthat biology plays a significant role in the production and functioning ofindividual personalities.

The Big Five Personality Traits and Politics

Of course, the concept of personality long predates any attempts todocument their biological origins. Research on personality has provideda cornucopia of different individual traits to study since scholars began toexamine the concept in earnest nearly 80 years ago (Allport 1937; Allportand Odbert 1936; Cattell 1943; Fiske 1949). Fortunately, the nebulousconstellation of traits began to coalesce in the late 1970s and early1980s around a hierarchical representation of traits with a set offive factors representing the highest level (Goldberg 1990, 1995; Johnet al. 2008; McCrae and John 1992). The Five Factor Model orBig Five personality traits consist of (1) Extraversion, (2) Agreeableness,(3) Conscientiousness, (4) Neuroticism/Emotional Stability, and(5) Openness to experiences. Research into each of these traits individuallyand collectively has produced a considerable amount of knowledgeabout their expected effects on how individuals understand and engagewith the world.

Extraversion is the oldest and perhaps most extensively studied of theBig Five traits (Eysenck 1947) and is associated with adjectives like active,assertive, energetic, enthusiastic, talkative, gregarious, expressive, out-going, social, and positive emotionality. Extraverted individuals tend toexpress themselves with more comfort and have larger social networks(Gosling et al. 2003; Hibbing et al. 2011). When it comes to politics, themost obvious effect of being more of an extravert than an introvert is on


participation. Research has successfully documented the positive relation-ship between an increase in extraversion and increase in the likelihood ofpolitical participation (Vecchione and Caprara 2009). It has been shownto increase one’s likelihood to vote, attend political rallies or meetings,speak at meetings, volunteer and donate to campaigns, online politicalengagement, and likelihood of contacting officials (Caprara et al. 2010;Cooper et al. 2013; Gallego and Oberski 2012; Gerber et al. 2011a; Ha,Kim, Jo 2013; Mattila et al. 2011; Mondak 2010; Mondak and Halperin2008; Mondak et al. 2010; Quintelier and Theocharis 2013). It has alsobeen shown to increase one’s sense of civic duty (Weinschenk 2014) andtrust (Dinesen et al. 2014). Extraverts also appear to be stronger partisans(Gerber et al. 2012b), although this does not appear to point in any onepartisan direction (Cooper et al. 2013; Gerber et al. 2010, 2012b;Mondak 2010).

Agreeableness is inherently social and is aimed at developing andmaintaining positive relationships (Graziano and Eisenberg 1997),while judging prosocial behavior more highly and antisocial behaviormore harshly than others (Kammrath and Scholer 2011). It is associatedwith the adjectives appreciative, forgiving, generous, kind, sympathetic,trusting, warm, modesty, straightforward, and altruistic. It also appearsto rely heavily on conscious mental processes as it seeks to suppressselfish desires and negative affect (Jensen-Campbell and Graziano 2001;Jensen-Campbell et al. 2007). Political participation is, once again, oneof the first areas in which scholars have tested the potential effects ofagreeableness, producing mixed results. On the one hand, Mondak andHalperin (2008) find a negative relationship with voting and a positiverelationship with attending meetings, signing petitions, contacting offi-cials and writing letters to the editor. On the other hand, Schoen andSteinbrecher (2013) find a positive association with turnout and Gerberet al. (2011a) find a negative relationship with contacting officials andspeaking at meetings (Ha et al. 2013). Interestingly, Mondak et al.(2010) found very little connection between agreeableness and partici-pation at all. As was the case with extraversion, Dinesen et al. (2014)find a strong positive association between agreeableness and a compo-site score of general civic participation (see also Cooper et al. 2013;Weinschenk 2014), although those scoring high in the trait do notappear to like large numbers of referenda (Freitag and Ackermann2016). Finally, when it comes to partisanship, Cooper et al. (2013)find a negative relationship between liberal ideology and agreeableness,


Schoen and Schumann (2007) and Dirilen-Gümüş et al. (2012) find apositive relationship between the two, and Gerber et al. (2012b) fail tofind a significant connection at all.

Conscientiousness found its way into the Big Five personality traits as aresult of the obvious need for a concept that captures variance in indivi-duals’ levels of self-control or self-governance (McCrae and John 1992). Ithas grown to also incorporate concepts like diligence and thoroughness.Thus, conscientiousness is associated with adjectives like efficient, orga-nized, reliable, responsible, thorough, dependable, productive, dutiful,and deliberate. Like agreeableness, conscientiousness operates throughexecutive control of the conscious mind through self-regulation (Ahadiand Rothbart 1994; Rothbart et al. 2001). Jensen-Campbell et al. (2007)provide some of the first evidence documenting conscientiousness’ role ineffortful anger control. However, the executive control associated withboth conscientiousness and, potentially, agreeableness does not mean thatthe unconscious mind does not make an important contribution to theproduction of conscientious behavior. Kapogiannis et al. (2013) havedocumented variance in the size of various parts of the brain in relationto conscientiousness. Forbes et al. (2014) have detected a connectionbetween damage to a specific part of the brain resulting in higher neuroti-cism and lower conscientiousness, suggesting a biological relationshipbetween the two. The undiscovered country here is the likely directionof the causal arrow. Does genetics lead to brain structures that lead tocertain personality types or do personality types alter brain structures?There is ample evidence that our behavior can alter the structure of ourbrains (Draganski et al. 2014; Lövdén et al. 2010). Thus, this is not assimple a question as it may appear. Ultimately, the point to be made here isthe same that Hibbing et al. (2014b) make regarding biology predisposingus to certain behaviors. That is, even with personality traits that are heavilybased on conscious thought mechanisms, one’s ability to engage theseeffortful thought processes is likely dictated by brain processes out of ourcontrol.

Beyond the conscious versus unconscious control debate, the person-ality trait of conscientiousness has been associated with higher grades,better performance at work, being less accident prone, better health, andbetter financial practices (Barrick and Mount 1991; Bogg and Roberts2013; Digman and Inouye 1986; Letkiewicz and Fox 2014; Vedel 2014).Research has failed to find much of a connection to political participation,with the few significant findings pointing in different directions (Gerber


et al. 2011a; Mondak and Halperin 2008). Mondak (2010) finds a nega-tive correlation between conscientiousness and political knowledge.Finally, some of the strongest and most consistent findings relating con-scientiousness to politics are in its relation to ideology and partisanship.Increased conscientiousness tends to be correlated with conservatism andidentification with the Republican Party in the USA (Carney et al. 2008;Cooper et al. 2013; Gerber et al. 2011a; Gosling et al. 2003; Mondak2010; Mondak and Halperin 2008; Sibley et al. 2012; Osborne and Sibley2012).

Neuroticism or emotional stability is focused primarily on the ability tocontrol one’s emotions. Those who demonstrate an inability to do so or atendency to exhibit negative affect are said to be high in neuroticism(Schoen and Schumann 2007). Neuroticism is associated with adjectiveslike anxious, tense, touchy, unstable, worrying, hostile, and an obsessionwith inadequacy. The political science literature tends to prefer to use theterm emotional stability rather than neuroticism. Research has, onceagain, produced mixed results regarding emotional stability’s effect onpolitical participation, if it has any at all (Gerber et al. 2011a; Hibbinget al. 2011; Mondak 2010; Mondak and Halperin 2008; Mondak et al.2010; Vecchione and Caprara 2009; Caprara, Barbaranelli, and Zimbardo1999). Dinesen et al. (2014) find an increase in neuroticism is associatedwith a decrease in one’s general level of trust. Weinschenk (2014) finds noconnection between emotional stability and one’s sense of civic duty andMondak finds (2010) no connection to political efficacy. When it comes topartisanship, emotional stability/neuroticism has produced results sug-gesting a positive relationship between emotional stability and conserva-tives. Mondak (2010) indicates that an increase in emotional stability isassociated with an increase in conservative ideology (in two of the threesamples) and Republican identification (in one of three samples). Gerberet al. (2012b) and Carney et al. (2008) find similar results. However,Cooper et al. (2013) fail to find a significant relationships between neu-roticism and ideology, partisanship, or efficacy (see also Dirilen-Gümüşet al. 2012).

Finally, openness to new experience has had the most difficulty findingits footing in the Big Five taxonomy (Connelly et al. 2014; Mondak2010). Originally it was labeled cultured or intellect with a general inqui-sitiveness associated with scoring high in the category (Cattell 1957;McCrae and Costa 1983; Fiske 1949; Goldberg 1992). Woo et al.(2014) suggest, “at the highest levels of abstraction, openness to


experience concerns an individual’s preference for variety and novelty”(29). Openness to experience is associated with adjectives like artistic,curious, imaginative, insightful, original, introspective, and unconven-tional. Perry and Sibley (2013) document the moderating effect thatopenness has on how individuals interpret information about danger andsocietal threat, with those low in openness being particularly susceptible tosuch messages.

To the extent that a connection to political participation has beenshown, it has been a positive one (Gerber et al. 2011a; Ha et al. 2013;Mondak 2010; Mondak et al. 2010; Vecchione and Caprara 2009).Weinschenk (2014) finds a positive correlation between openness and asense of civic duty to vote and keep informed. Dinesen et al. (2014) finda similar increase in one’s sense of civic duty as one’s level of opennessincreases, although this positive association holds for “never cheat ontaxes,” “watch public authorities,” “active in associations,” and “under-stand others,” but emphatically not when the dictum is “follow the rules.”In this final case, an increase in openness is associated with a decline in one’swillingness to follow the rules. Of course, given these same respondents’lack of interest in cheating on their taxes as their openness increased, theirclaims about bucking the system and not following the rules appear hollow.Indeed, it is typically low conscientiousness that is associated with criminalbehavior (Ozer and Benet-Martínez. 2006), not openness to experience. Inanother interesting work, Gerber et al. (2013a) perform an experiment inorder to test for a relationship between the Big Five and the persuasivenessof campaign mobilization efforts. They find that those high in opennesswere most susceptible to these types of appeals.

The strongest and most consistent finding regarding openness toexperience is that high levels are associated with a liberal ideology(Barnea and Schwartz 1998; Carney et al. 2008; Gerber et al. 2010,2012b; Gosling et al. 2003; Jost et al. 2003a, 2003b; McCrae 1996;Mondak 2010). In their recent study of the openness and ideology con-nection, Roets et al. (2014) find a small bit of nuance in this relationship.They document the expected relationship for Western Europeans, but findthe opposite is true for Eastern Europeans, who appeared to be more likelyto place themselves on the right side of the political spectrum. This resultdemanded more scrutiny, thus, they split the Eastern Europeans into twocategories representing those raised during communist rule versus thoseraised after the fall. The results showed that the young Eastern Europeanswere much more like their Western counterparts, although slightly


weaker, whereas the older communist-raised Eastern Europeans were theones more likely to place themselves on the right.

The large body of work on the political implications of the Big Fivepersonality traits has done a tremendous job laying a foundation uponwhich to build more complex research.2 Here I leverage this work in anexploration of the role that personality plays in driving the successes andfailures citizens have when engaging with and understanding politics. Icontend that the ability of individuals to understand politics from theabstract to the concrete is conditional on their particular personality traits,which are the observable manifestations of unseen physiological processesproduced through a combination of conscious and unconscious thought.Perhaps “unseen” is an incorrect characterization. As I have documentedabove, there is actually quite a bit of work focused on trying to “see” theseprocesses in action. The problem is, right now, the picture we are “seeing”is rather blurry and we are not entirely sure what we are looking at.Fortunately, we do not have to be able to see something to understandits effects. Astrophysicists cannot observe black holes, dark matter, or darkenergy directly, but they can be confident in their existence because theycan see their effects. They can make prediction and provide a morethorough understanding of the universe despite never actually “seeing”the whole process in action. In a similar manner, scholars cannot yetobserve the biological origins of an individual’s personality, but theymost certainly can observe the effects of this process. These effects arethe culmination of the interaction of unconscious and conscious mentalprocesses that produce predictable behavior that will help scholars under-stand how citizens comprehend and engage in the political process.

In the remainder of this book I explore the effects of personality on thequality of citizen political engagement and understanding, that is, civicaptitude, in four separate areas. First, in the next chapter I examine theoccurrence of a type of cognitive dissonance found among individualswhen they are asked to think about their own ideology. Ellis andStimson (2012) Ideology in America examines the paradoxical relationshippeople have with the concept. Americans want to identify and labelthemselves with the word conservative, yet when asked about their parti-cular policy preferences, they tend to prefer liberal policies. This translatesto a preference for symbolic conservatism and operational liberalism. Whydo people do this? I argue that this combination of contradictory prefer-ences is driven, in part, by personality. In particular I point to one’s level ofagreeableness. The more agreeable one is, the more they desire positive


relationships. This leads to the simple desire, at least in the abstract, tosupport what they perceive to be the majority’s opinion.

In Chapter 3 I focus directly on personality’s role in causing individuals’failure to understand basic facts about important political issues.Traditional accounts of this type of ignorance point to education, interest,and partisan motivated reasoning. The conception of partisanship as aperceptual screen dates back to some of the earliest works in mass politicalbehavior (Campbell et al. 1960). In other words, individuals filter realitythrough their partisan lens, which results in a self-serving biased represen-tation of reality. This is not limited to subjective opinions about issues, buthas been shown to occur with alarming regularity when consideringfactual questions and even appears to affect one’s ability to do mathcorrectly (Bartels 2002; Jerit and Barabas 2012; Kahan et al. 2013). Inthis chapter I examine responses to factual survey questions in five policyareas—(1) food stamps, (2) same-sex marriage, (3) health insurance sub-sidies, (4) drug testing welfare recipients, and (5) U.S. oil production. Asone would expect, partisanship plays an important role in driving incorrectunderstandings about what is happening in these policy areas. However,partisanship is not the only cause. What past works have been missing ispersonality. While a substantial amount of personality research has spenttime attempting to draw connections between one or the other Big Fivepersonality traits and political ideology and partisanships, I document herethat personality does not simply work through its effects on citizens’partisanship. It has its own independent effect on how people understandpolitical issues, even highly partisans ones. This effect can lead to incorrectunderstandings of objective factual reality in various political issue areas.

In Chapter 4 I look beyond one’s ability to identify factual policy informa-tion and to citizens’ ability to connect their personal policy preferences to theparty that best represents them. One basic necessity for quality democraticparticipation is the capability to connect one’s own policy preferences to theproper party. As was outlined earlier in this chapter, there is ample reason tosuspect that the average citizen fails to make these simple connections withregularity. Indeed, the works of Lau and Redlawsk (2006; Lau, Andersen,and Redlawsk 2008) indicate that somewhere between 20 and 30 percent ofvoters chose the wrong candidate during past presidential elections. I exam-ine how personality correlates with how well individuals are able to connecttheir attitudes to the correct party in the same five issues areas as in Chapter 3.The findings here point to particularly important roles for extraversion andopenness to experience, although in opposite direction. Extraversion


contributes to failing to connect one’s policy preferences to the correct party,while openness contributes to success in making this connection.

In Chapter 5 I return to Mettler’s (2011) submerged state policies.Mettler’s work documents the reality that citizens often fail to recog-nize instances where government programs support their lives.Programs such as tax deductions for employer provided health insur-ance and retirement savings deliver direct benefits to some individualsand not others, but are not perceived as government “helping” thesepeople, whereas programs like food stamps are viewed in an entirelydifferent light. They are understood as direct subsidies to some peopleat the expense of others. In this chapter I examine the policy attitudesof individuals who benefit from three of the largest submerged stateprograms (home mortgage interest, retirement savings, and healthinsurance tax deductions) and show how holding hypocritical attitudestoward more visible welfare policies depends on personality. In parti-cular, I demonstrate how conscientiousness leads these individuals toexpress a desire to reduce or eliminate more visible government pro-grams like welfare in general and food stamps in particular, whilebenefiting themselves from submerged welfare policies. Interestingly,agreeableness appears to have the opposite effect.

Finally, in Chapter 6 I evaluate the results reported here in relation topersonality’s effect on the ability of individuals to understand and engagein politics in a manner consistent with the demands placed upon them asdemocratic citizens. Education, political knowledge, partisanship, andinterest in the subject are not enough to mitigate the effects that certainpersonality traits have on the likelihood of one holding consistent ideolo-gical beliefs; knowing simple facts about an issue area; connecting theirown policy preferences to the correct party; or wanting to dismantle visiblewelfare state programs, while they personally benefit from numerous sub-merged programs themselves. Of course, the news is not all bad. In theissue areas examined here, openness to experience often leads to a morepositive impression of citizen capabilities, while extraversion and conscien-tiousness pose the greatest challenge to civic aptitude.

In addition to summing up the empirical results in the chapter, I alsodiscuss where our political system can go from here. Calls for more andbetter education will continue to have little effect on the average citizen’soverall political knowledge and interest. This is not because they consciouslychoose to disregard politics. The choice was made for them long before theybecome consciously aware that the option to engage with the political world


is even present. Thus, my story is not one of blame. It is one of under-standing. Personality is but one factor found in the hidden depths of theunconscious mind influencing political behavior. As we continue to gainmore knowledge of all the processes that go into generating highly attuned,civically engaged citizens, we can begin to make refinements to our politicalsystem that are based on how people actually behave, rather than a folktheory of democracy (Achen and Bartels 2016) devoid of real human beings.

Thus, I take the baton from Achen and Bartels and present my ideas onhow the electoral process should be altered in order to accommodate theway our individual minds really work. Of course, I am not the first person tooffer such suggests, but in doing so, most scholars have tended to focus onpost-election fixes. That is, they point to a need to level the playing fieldduring the policy-making process in Congress or to simply ignore activelyignorant individuals when scientific consensus points clearly to a high-qualitypolicy solution. These suggestions, while potentially helpful, largely ignorethe reality of a political system that is built on citizen participation throughelections. Instead of focusing on what to do after elections, I directly addressthe electoral process and how it could be altered to accommodate thelimitations of a flawed electorate (which the author counts himselfamong). When we structure the electoral process with an eye toward coaxingthe best out of voters, I would expect to see significant down-streamimprovements, too. It is much more difficult for interest groups to influencethe policy-making process, when the electorate is likely to be aware of thisinfluence and can hold candidates accountable in the voting booth. The waythe U.S. political system is structured now, voters largely cannot do this. Butif the system is altered to demand the time and attention of voters, they can.

NOTES

1. Lest one get confused by large numbers and think that would be a signifi-cant burden to place on the Cruz family. This $14,000 amounts to 0.008percent of their income. The equivalent for a family making $40,000 inincome a year would be an extra $320 in tax per year. Well, actually only$240 extra since the family making $40,000 a year would be taxed at thelower rate of 25 percent. Thus, not insubstantial, but hardly a burden. Andalso an example of how regressive this type of government subsidy is. Bothfamilies need health insurance and the family making 1.7 million gets$14,000 from the U.S. tax payer, while the family making $40,000 receives240 dollars.

NOTES 37

2. It is important to note that recent work by Hatemi and Verhulst (2015) callsinto question the causal connection between personality traits and politicalattitudes. They find over the course of 10 years, changes to one’s personalitydo not correspond to changes in one’s political attitudes. Their variousgenetics models also fail to find a significant causal connection betweenpersonality traits to political attitudes. Of course, their findings are limitedto the connection between personality and political attitudes and, therefore,have no bearing on research like that presented here, which is focused on theconnection between the Big Five and one’s ability to understand the poli-tical world correctly.


CHAPTER 2

Ideological Cognitive Dissonance

The September 17, 2014 edition of the New York Times carried a storyabout the state of Kentucky’s Senate race, which pitted longtimeRepublican Senator Mitch McConnell against Democratic challengerAlison Lundergan Grimes (Goodnough 2014). In it, a Kentucky residentis asked about her new health insurance coverage. Her replay is simple,“I’m tickled to death with it.” But when she is asked about PresidentObama, she is not quite so chipper. “Nobody don’t care for nobody nomore, and I think he’s got a lot to do with that.” And finally, as a born andraised Republican, she indicates she would support Senator McConnell inthe election. Of course, this particular voter’s confusion is rather common-place. She loves the health care she is now eligible for due to changes inMedicaid insurance rules that came about as part of the health-care over-haul ushered in by the passage of the ACA. She dislikes, or at leastdistrusts, the person largely credited with spearheading the change thatresulted in the new rules governing Medicaid in Kentucky, that is,President Obama. On top of that, she plans to support a person forSenate who has stated numerous times that he would like to repeal thevery law that she likes so much. For example, in response to a questionabout what he would do about Obamacare if Republicans regained con-trol of the Senate after the 2014 midterm elections, McConnell said, “pullit out root and branch and start over” (Youngman 2014).

The difficulty this one voter has connecting her opinion of the parti-culars of a policy to a candidate and her abstract thoughts about the policy


39

area is not unique to her or to Kentucky. When it comes to the ACA, theentire country has tended to dislike the idea of it, but to love the parti-culars. Polling data from April 2010 to September 2016 shows that thepercentage of Americans opposed to the law has rarely ever fallen belowthe percentage in favor of it (Kirzinger et al. 2016), yet Kaiser Health’stracking poll of opinion on individual portions of the law shows thatpeople are heavily in favor of every provision of the law except theindividual and employer mandates that everyone must have health insur-ance (Kaiser Family Foundation 2013, 2014). The ACA has produced aparticularly acute case of a “hate the law, love the results” way of thinkingand is an example of a classic disconnect between symbolic politics andoperational politics in the minds of many individuals (Free and Cantril1967; Stimson 2004; Popp and Rudolph 2011). Americans have a strongdistaste for government (the federal government in particular) in theabstract, but are strong supporters of the particular things governmentprovides them. Ellis and Stimson (2012) Ideology in American drives thisdisconnect in the minds of individuals home and makes a strong case thatany study or use of ideology as a variable that fails to incorporate bothoperational and symbolic conceptions lacks validity. Looking at a halfcentury’s worth of survey data, they document a general preferenceamong citizens for the conservative symbolic label, while simultaneouslyholding liberal policy views.

In this chapter I examine the role that the Big Five personality traitsplay in explaining how well individuals are able to connect their perso-nal policy preferences to the larger more abstract concept of ideology.The failure of anything more than a small stratum of our society tohold consistent attitudes about policy has vexed scholars and interestedobservers of politics for decades (Converse 1964; Schumpeter 1950).Here I do not examine the minutia of the various specific policyattitudes of individuals for consistency, but focus on the connectionbetween the abstract and the specific. If someone supports increasedspending on welfare, scientific research, and social security, do theyknow that they are liberal? Are they willing to label themselves that?The idea here is rather simple. In the ideal world of the democratic folktheory, citizens should be able to understand politics in this manner.Of course, according to Ellis and Stimson, most citizens are not able todo this successfully, but it is not a blowout. Around 45 percent ofpeople are actually able to make the connection between their specificpolicy preferences and an abstract ideology. In the following sections

40 2 IDEOLOGICAL COGNITIVE DISSONANCE

I outline how failure to hold consistent operational and symbolicideological beliefs is a type of cognitive dissonance and argue thatindividuals’ durable personality traits contribute to how well they areable to make this connection. I then outline a new ideology typologythat incorporates both operational and symbolic ideologies and presentdata measuring their occurrence and demonstrating their connection tothe Big Five personality traits.

IDEOLOGICAL DISSONANCE AND PERSONALITY

The disconnect between symbolic and operational ideology in the mindsof individuals is an example of widespread cognitive dissonance among thepopulous. Leon Festinger’s (1957) A Theory of Cognitive Dissonanceposits that when individuals hold two or more bits of related contradictoryknowledge, they enter a state of dissonance. Dissonance is a kind of mentaland emotional agitation that demands resolution. Festinger’s dissonancetheory is, perhaps, the most well-known theory produced during the mid-twentieth century’s flowering of various cognitive consistency theories,with the common theme being that individuals desire consistency orbalance among their various cognitions. When inconsistencies appear,individuals are motivated to resolve them. However, finding a resolutionis not required. Simply holding contradictory beliefs may never agitate oneenough to seek a resolution in situations where the individual is disinter-ested in the area or subject of the discordant cognitions. What constitutesan actionable inconsistency and what counts as resolving it have been thedriving questions of research in this area ever since (Aronson 1992;Harmon-Jones et al. 2009).

Scholars have applied dissonance theory in numerous ways in the realmof politics. One of the most common demonstrations has been in connec-tion to voting. Dissonance theory suggests that an individual’s behaviorproduces a cognition about that behavior. Individuals will then interpretnew information related to that behavior, or re-interpret old informationabout that behavior, so that it is consistent. In the case of voting, the act ofvoting for one candidate will cause individuals to align their attitudes sothat they are consistent with that vote choice, thus avoiding the agitativestate known as dissonance (Beasley and Joslyn 2001; Bølstad et al. 2013;McGregor 2013; Mullainathan and Washington 2009). Moving beyondvoting, Collins (2011) shows that justices on the U.S. Supreme Court aremore likely to write a separate opinion when they decide a case that is

IDEOLOGICAL DISSONANCE AND PERSONALITY 41

inconstant with their attitudes so that they can justify to themselves, aswell as to interested observers, the seeming inconsistency.

The widespread dissonance found in the American public observed byEllis and Stimson (2012) does not appear to be of the ephemeral sort.They show how the New Deal era ushered in “ . . . a pattern, continued tothe present day, in which the name for an ideology that supports highlypopular programs is itself unpopular” (61). In other words, going back80 years, how one chooses to measure ideology goes a long way indetermining what one will find. If one uses a symbolic measure, one willfind that most people identify as conservatives. If one uses an operationalmeasure, one will find that most people are liberals. Ellis and Stimsonargue that the primary cause of this result is the conflicted conservative,that is, individuals who identify as conservatives, but hold liberal policybeliefs. About a third of all conservative identifiers fall into this category.The same is not true of liberal identifiers, with very few self-proclaimedliberals holding conservative policy positions. Ellis and Stimson explain theoverall preference for conservatism as a function of the word’s meaning inother contexts of life. People prefer the word “conservative” to the word“liberal.” This is not the case because of its specific policy content, butbecause in other parts of life, such as in religious contexts, it is viewedfavorably. Since people are more familiar with the term in these otherfavorable contexts, they simply transfer use of the term and its positiveassociation to the less familiar context of politics. If this is the case, thenwhy do people, in the aggregate, tend to support liberal policies? Therather blunt answer is that people are self-interested. Therefore, whenlooking at specific government policies, they support those that helpthem and others around them.1

In their examination of conflicted conservatism, Ellis and Stimson cometo the rather puzzling conclusion that it “ . . . represents an ideology all itsown, not an inconsistency in need of resolution, but a real and systematicset of political beliefs” (173). This is odd because they then go on todescribe this new ideology as “ . . . a large segment of the population that isboth liberal and conservative, depending on the frame of reference”(173). Framing effects have been studied for decades across numerousdisciplines and they are almost always viewed in negative terms. Framing isthe easy manipulation of individuals’ attitudes based on little more thanhow the question or problem is presented. It can even occur withoutconscious awareness, as outlined in the previous chapter. Framing is anexample of individuals failing to live up to expectations of rational


engagement with whatever subject is being studied (e.g., Iyengar andKinder 1987; Tversky and Kahneman 1981). The persistent presence ofa large group of individuals whose attitudes and behavior are conditionalon the framing of the alternatives is certainly a violation of the democraticfolk theory. To call it a “systematic set of political beliefs” does not sit wellwhen the only thing systematic is that their attitudes are malleable. That isnot an ideology in the sense that Converse (1964) defined it, whichrequires constraint and interdependence in one’s attitudes. Conflictedconservatism is the exact opposite. There is no constraint to their ideologyand altering one’s attitude in response to a particular frame has no effecton the other idea elements in their head.

Therefore, I approach this disconnect between symbolic and opera-tional political beliefs as a set of inconsistent cognitions that do, in fact,need resolution. But this will only be true for individuals who are inter-ested enough in politics to care. For those only marginally interested therewill be little need to resolve the operational versus symbolic contradiction.These individuals are interested enough to vote, but not interestedenough to learn how various policy areas interconnect. Thus, they willhappily move through life holding inconsistent political beliefs, which canhave a significant impact on election outcomes. Ellis and Stimson showthat without conflicted conservatives, the 1988, 2000, and 2004 electionsmay have turned out differently. In order to understand the failures ofideologically inconsistent individuals, one must also understand the causesof ideological consistency found among a large portion of the electorate.Not everyone is conflicted. The causes of successful political engagementare as important as the causes of failure. In the next section I discuss howpersonality contributes to how well individuals are able to integrate theiroperational and symbolic ideologies.

Personality and the New Ideology

As outlined in the previous chapter, the connection between personalityand political ideology has been the focus of numerous studies. Mondak(2010) uses the standard symbolic measure of ideology and finds thatconscientiousness (all three data sets) and emotional stability (two of threedata sets) are strongly related to conservatism, while openness to experi-ence (two of three) and agreeableness (one of three) are related to liberal-ism. Mondak also presents results connecting a “moral traditionalism”

scale to the Big Five, which is similar to an operational ideology, although


limited to only social/moral issues. He once again finds that conscien-tiousness is predictive of a more conservative ideology, while openness toexperience predicts liberal leanings. Agreeableness and emotional stabilitydo not fare as well in this context, failing to reach standard levels ofstatistical significance. Looking only at symbolic ideology, Mondak andHalperin (2008) find similar results. Gerber et al. (2010) look at bothsymbolic and operational ideology, although they split the latter intoeconomic policy and social policy attitudes. They find consistent resultsfor conscientiousness, emotional stability, and openness. When it comes toagreeableness, they find it has a significant connection to liberalism, butonly for their operational measurements of ideology, not the symbolic self-placement. Their unusual finding is that extraversion is related to conser-vative ideology in both symbolic and operational measures.

Furthermore, Osborne and Sibley’s (2012) work only looks at symbolicideology and finds, once again, that conscientiousness and emotionalstability (they use the term neuroticism) are correlated with conservatismand openness and agreeableness are correlated with liberalism. Their find-ings also support Gerber et al.’s finding that extraversion is correlated withconservatism. Finally, in their meta-analysis of 73 published and unpub-lished studies, Sibley et al. (2012) focus on the connection of personalityto symbolic ideology and find further corroboration for a connectionbetween conscientiousness and emotional stability (they use neuroticism)and conservatism, on the one hand, and the openness to experienceconnection to liberalism, on the other. Interestingly, they fail to find asignificant correlation between either agreeableness or extraversion.2 Insum, there is ample reason to believe that conscientiousness and emotionalstability are connected with conservatism, while openness to experience isconnected to liberalism.3 This has been consistently found across justabout every study of ideology. However, in the case of agreeablenessand extraversion, it is harder to draw a definitive conclusion. The tentativeresult, though, appears to be that if they are connected to ideology at all,agreeableness is likely related to liberalism, while extraversion may beconnected to conservatism.

While these works often acknowledge that there are many ways tomeasure ideology, none of them have grappled with the ideological cog-nitive dissonance found with regularity among the general public. Thisleaves a significant gap in our understanding of the connection betweenpersonality and ideology. It has been understood for at least 60 years thatthe average citizen does not think about politics in strict ideological terms


(Converse 1964) and is not often highly engaged with the day-to-dayhappenings of the political process (Delli Carpini and Keeter 1996; Zaller1992). Thus, it is unlikely that the observed dissonance between opera-tional and symbolic ideological orientations is going away anytime soon.The connections to personality traits found by previous research provideonly a partial treatment of the whole ideological story. The same is true ofscholars studying ideology who have ignored the role of personality. Theiraccount is, at best, only partial. All told, there are two lessons to be learnedfrom the personality and operational versus symbolic approaches to ideol-ogy. First, variance in the durable personality traits of individuals has animportant effect on one’s ideological approach to politics. Second, one’stotal ideology must be understood as the combination of both operationaland symbolic conceptions of ideology, which I refer to as their jointideology. Here I combine these two in order to understand personality’seffect on the joint symbolic and operational ideologies of individuals.

The first step I take in order to fully understand the dynamics at playbetween personality and ideology is to change the approach to the pro-blem. Ellis and Stimson start with symbolic ideology, that is, how indivi-duals place themselves on a single dimension. They then compare this totheir specific issue beliefs. Therefore, conflict or congruence is found whenindividuals fail to live up to their chosen symbolic ideology. This results inthem labeling individuals who are symbolic conservatives and operationalliberals as “conflicted conservatives.” There is no inherent reason whysymbolic ideology needs to be the starting point. Presumably this hasbeen the starting point because a half century of research has continuallypointed out that the average citizen does not live up to ideologicalstandards of academics. However, this seems like the opposite of theapproach one should take in this case. Abstract ideological concepts aremuch harder for individuals to grasp than specific policy questions. Thus,if either of the two types of ideology are apt to misrepresent individuals’“true” ideology, it is likely their claims about their abstract symbolicideology. Ellis and Stimson’s explanation for why Americans tend to saythey are conservative rather than liberal recognizes this. They say it isbecause people do not understand the meaning of the concept whenapplied to politics, but they do when used in other contexts. As a result,individuals incorrectly apply the label to their political attitudes. Of course,there are no such cognitive mistakes posited by Ellis and Stimson whenlooking at individuals’ operational policy attitudes. Thus, the logical start-ing point is with citizens’ operational ideology.


Why is the starting point important? Because it allows for a properlabeling of the problem. It is not conservatives that are conflicted orconfused, it is liberals. It is the large number of operational liberals whofail to understand the definition of the symbolic label they use for them-selves that are making the mistakes. Operational conservatives do not dothis nearly as often. Thus, a more apt label for these individuals would be“confused liberals.” In addition, it is not clear that work to this point hasproperly defined all the potential combinations of operational and sym-bolic ideology. Moderates often get left by the wayside. Popp andRudolph (2011) indicate that “moderates receive less attention on thegrounds that they lack clear symbolic attachments, making it impossible toidentify those who are cross pressured by symbolic and principled con-siderations” (814). I disagree with this particular claim. Operational mod-erates should symbolically label themselves as moderates.

The standard American National Election Studies (ANES) symbolicideology question gives respondents the option to choose “moderate;middle of the road.” ANES then follows that question up with a push toget respondents to choose either liberal or conservative, if they origin-ally chose moderate. This is a good idea, if one is interested in predictingsomething like vote choice, as Keith et al. (1992) clearly demonstratedin the case of partisan self-placement. People are not really independentwhen it comes to their partisanship, they are closet partisans. How do weknow this? Because once the questions get respondents to admit thatthey lean one way, we see that these “leaners” behave very much like thestrong partisans who answered with their partisanship right away to theopening question. Thus, from a behavioral perspective, these individualsare no different than any other partisan even though their originalchoice was independent or moderate. Maybe people are not reallymoderates when it comes to their symbolic ideology either. However,the point of this research is not to predict behavior, but to understandwhy people fail to choose the proper ideological label. Therefore, scho-lars should not be involved in helping respondents choose their correctsymbolic ideological label through the structure of the ideological ques-tions asked of them.

People like to label themselves as moderate, which may or may not haveanything to do with their actual policy preferences, which is also true ofpeople who choose the conservative label. In order to understand thecauses of individuals’ choice of symbolic label, one cannot disregardthose who want to label themselves as moderate by utilizing a follow up


question designed specifically for them that is essentially saying “we don’taccept your first answer. Try again.” Scholars do not do this for those whochose liberal or conservative. They accept it even though Ellis and Stimsonand others have demonstrated conclusively that this is not necessarily anymore accurate for people choosing these labels than it is for people whochoose moderate. Indeed, in Ellis and Stimson’s data, 56 percent ofpeople who label themselves as extreme conservatives were actually opera-tional liberals. The fact that many people prefer to use the moderate labelrather than liberal or conservative is meaningful and needs attention.Therefore, a full accounting of the possible combinations of operationaland symbolic ideologies must contain moderate, which makes it a three bythree grid rather than a two by two grid.

Figure 2.1 displays the possible combinations of operational and sym-bolic ideology. The category Ellis and Stimson and others have tended to

Symbolic

Liberal Moderate Conservative

ConservativeShy ConservativeConfusedConservative

Con

serv

ativ

e

Operational

Mod

erat

eLi

bera

l

Moderate Liberal

Liberal Shy Liberal

ModerateModerateConservative

Confused Liberal

Fig. 2.1 Joint operational and symbolic ideology typology


focus on is in the bottom right hand corner where an individual isoperationally liberal, but labels themselves as symbolically conservative.They have called these conflicted conservatives, but I refer to them asconfused liberals. This is because people do not start with an ideology andthen adopt specific policy attitudes to fit the ideology (Converse 1964).People have some, perhaps not many, policy attitudes. They may neverthink in terms of abstract ideology. This is something imposed on them byacademics. However, when asked in a survey what they would call them-selves, they pick the words they like best regardless of the word’s actualpolicy meaning. The reason scholars have tended to focus on confusedliberals is because the upper left hand corner happens so infrequently, thatis, someone holding conservative policy attitudes while simultaneouslylabeling themselves as liberal. What has been left entirely unexplored arethe various combinations of moderate. The top and bottom boxes in themiddle column I label as shy liberals and shy conservatives. These areindividuals who hold either liberal or conservative operational policybeliefs, but choose not to symbolically label themselves as either liberalor conservative. Why do people do this? I argue that it is due, at least inpart, to their unconscious minds as manifested by a combination of BigFive personality traits.

Based on previous research, the easiest personality predictions withregard to ideology are in relation to standard liberals and conservatives.First, as one’s level of openness to experience increases, one should bemore likely to fall in the bottom left hand corner of Fig. 2.1, simplycalled liberal. The same goes for agreeableness, although extentresearch has been less consistent when it comes to the connectionbetween agreeableness and liberalness. Second, as one’s level of con-scientiousness and/or emotional stability increases, the more likely oneshould fall in the upper right hand corner of Fig. 2.1, that is, standardconservatives. A weaker prediction can also be made of extraversion.Some research has suggested that it is connected with conservatism,but it has been far from conclusive. Thus, if extraversion is going tohave an effect, it should be to increase the probability of falling in thestandard conservative box. The other boxes are more difficult to pre-dict since there has yet to be any personality research focused on thesetypes of ideologies.

Confused liberals (a.k.a., conflicted conservatives) are a combinationof operational liberalism and symbolic conservatism. What is uniqueabout this group is that they fall in the majority category for both


symbolic and operational ideology. That is, Ellis and Stimson clearlydemonstrate that most people choose the symbolic conservative labeland most people have liberal policy positions. Thus, the confusedliberal is an individual who prefers the labels and policies that mostother people prefer. This sounds like a person who is looking to agreewith as many people as possible. Positive relationships are often basedon agreement, particularly when the issues are perceived as importantto them. I hypothesize that as individuals’ level of agreeablenessincreases, they are more likely to desire political positions that theyperceive as being what “most people” want. In the context of U.S.politics, increased agreeableness leads to support for operational liber-alism and symbolic conservatism. The key is that this may very well bedifferent in other countries and political contexts. The common threadwill be that individuals high in agreeableness will count themselvesamong the majority both operationally and symbolically regardless ofwhether these two positions actually contradict each other like they doin the USA.

The last set of operational and symbolic ideology combinationsincludes moderates. Two of the more intriguing combinations are whatI call “shy liberals” and “shy conservatives.” These are individuals who areoperationally liberal or conservative, but prefer to label themselves asmoderates. Even in research that has focused on the interaction of opera-tional and symbolic ideology, this particular combination has receivedlittle attention. Why do individuals do this? I argue that this is due totheir combination of personality traits. Individuals who are high in multi-ple Big Five personality traits that would normally lead them to be liberalor conservative will be more likely to call themselves moderate. In otherwords, these individual are cross-pressured by their own personality traits.For example, an individual who is high in openness would be expected tobe liberal, while someone high in conscientiousness would be expected tobe conservative. But what happens when someone is high in both? Mysupposition is that they will be more likely to symbolically label themselvesas moderate.

TESTING THE EFFECT OF PERSONALITY ON NEW JOINT IDEOLOGY

I utilize the Political Personality Success and Failure (PPSF)4 survey Iadministered in July of 2014 in order to get at how well individuals areable to successfully connect their policy preferences to the abstract concept

TESTING THE EFFECT OF PERSONALITY ON NEW JOINT IDEOLOGY 49

of ideology. The PPSF allows for the production of an operational ideol-ogy measure based on responses to seven spending questions. Thesefollow ANES’ standard question wording when asking respondentswhether they would like to see spending in a particular area “increase,”“decrease,” “kept about the same,” or “don’t know.” Respondents to thePPSF survey were asked about (1) welfare spending, (2) social security, (3)public schools, (4) science and technology, (5) dealing with crime, (6)protecting the environment, and (7) health care. In order to build theoperational ideology variable, each individual question was converted intoa three point scale, with −1 meaning the individual answered liberally (i.e.,said increase), 0 meaning the individual answered either “kept about thesame” or “don’t know,”5 and 1 meaning individuals answered conserva-tively (i.e., said decrease). I then summed the seven spending questions,resulting in a variable that ranges from −7 (the most liberal) to 7 (the mostconservative). Symbolic ideology is once again measured using ANES’standard ideology question wording that requests respondents to placethemselves on the standard seven point left to right political spectrum. Ithen take the operational and symbolic ideology measures and collapsethem into simple three point scale indicating liberal, moderate, or con-servative. Finally, in order to produce the new joint measure of ideology, Icombine each individual’s operational and symbolic ideology and placeeach individual in the nine different ideological combinations found inFig. 2.1.

Figure 2.2 displays the distribution of respondents after I combine opera-tional and symbolic ideology. This produces nine total joint ideologicalcategories. As can be seen, most respondents in this data are operationalliberal (72 percent). The big difference among these respondents is how theychoose to label themselves symbolically. Of those who are classified asoperational liberals, 38 percent label themselves as liberals (“liberal”), 40percent call themselves moderates (“shy liberals”), and 22 percent refer tothemselves as conservatives (“confused liberals”). The operationalmoderatesmake up 11.7 percent of the respondents. Of these individuals, 18 percentlabel themselves as liberals (“mod. liberals”), 47 percent think of themselvesas moderates (“moderate”), and 36 percent label themselves as conservatives(“mod. conservatives”). Finally, 16.4 percent of all respondents are classifiedas operational conservatives. Of these individuals, just 11 percent think ofthemselves as liberals (“confused conservatives”), 24.6 percent thinkof themselves as moderates (“shy conservatives”), and 64 percent callthemselves simply conservatives (“conservative”).


In order to test the effects of the Big Five personality traits on thenew joint measures of ideology, I begin with an analysis of thetraditional ideology measures. This will provide a baseline for com-parison to other scholarly work in this area and to the analysis of thenew joint measures that follows. The first step is creating the depen-dent variables. I utilize the operational and symbolic measures ofideology separately by creating four dichotomous variables. First isoperational liberalism, which is coded 1 if the respondent scored lessthan zero on the operational ideology scale and coded 0 otherwise.Second, operational conservatism is coded 1 if respondents scoredmore than zero and coded zero otherwise. The third and fourthdependent variables are built from the symbolic measure of ideology.Symbolic liberals are coded 1 if they scored less than zero and codedzero otherwise and, finally, symbolic conservatives are coded 1 if theyscored greater than zero and coded 0 otherwise. In addition to mea-sures of each of the Big Five personality traits, I used numerous control

010

2030

Per

cent

Liberal

Shy Liberal

Confused Liberal

Mod. Liberal

Moderate

Mod. Conservative

Confused Conservative

Shy Conservative

Conservative

Fig. 2.2 Distribution of the combined operational and symbolic ideologies ofsurvey participants


variables, which include standard demographic measures (male, white,age, education), political knowledge, partisanship, and a measure of howfundamentalist one’s religious views are. Specific coding rules can befound in the Technical Appendix’s Tables A1 and A2.

While previous works have utilized both symbolic and operationalmeasures of ideology, the predicted effect of personality has not made adistinction between them. In order to conserve space, full model resultsare available from the author. The results after estimating four indepen-dent logit models predicting one’s symbolic (liberal and conservative) andoperational (liberal and conservative) ideology are clear. As has beenfound in past works, openness to experience and liberalness move togetherin all but one model (predicting operational conservatism), while emo-tional stability is strongly aligned with conservatism in all four models. Theinteresting result is the failure of conscientiousness to predict operationalideology in either case (i.e., being operationally liberal or conservative).Agreeableness has the opposite effect, in that it only predicts operationalideology (both liberal and conservative) and not symbolic. This suggeststhat the predictive power of these traits is conditional on how ideology ismeasured—not just the differences found when measuring symbolic andoperational ideology, but also differences found when measuring opera-tional ideology itself. Unlike symbolic ideology, which almost always relieson the self-placement of respondents on a seven point ideology scale,operational ideology has not been standardized. The constellation ofpositions used by scholars in a particular piece of research is typicallyunique to that piece of research. This can make comparison difficult.

For instance, Gerber et al. (2010) find a significant connectionbetween conscientiousness and their measures of operational ideology.They use two separate measures, one focused on economic policy atti-tudes and one on social policy. The economic policy measure is com-prised of two questions about health-care provisions and tax policy, whilethe social policy measure is produced by two questions about abortionand same-sex marriage. Thus, the effect of conscientiousness is sensitiveto the specific operational measures being used. Gerber et al. find asignificant connection, while the results presented here do not.However, this may not be the case with agreeableness. Gerber et al.find the exact same result as I did when looking at the effect of agree-ableness. They fail to find a connection between symbolic ideology andagreeableness, but do find a connection with both operational ideologymeasures. Indeed, when it comes to agreeableness and the standard


symbolic ideology measure, even Sibley et al.’s (2012) meta-study fails tofind a significant connection. There have been some exceptions, how-ever. One of the three surveys used by Mondak (2010) produces asignificant result, as do Osborne and Sibley (2012).

Moving beyond the simple measures of ideology, Table 2.1 presentsresults after estimation of five logit models predicting the effect of the BigFive on the new joint measures of ideology found in Fig. 2.2.6 Each ofthese models is estimated with the same set of control variables as used inthe previous models, with one exception. I include a measure of politicalinterest in this case because there is reason to suspect that the effect ofcognitive dissonance is conditional on how interested one is in the topic.The expectation is that an increase in interest should make holding contra-dictory operational and symbolic ideologies less likely.

Looking first at the results associated with predicting “liberal,” whichmeans the respondent is both symbolically and operationally liberal, I findexactly what would be expected. As one’s openness to experienceincreases, one’s likelihood of being a consistent liberal also increases.The significant and negatively signed coefficients associated with conscien-tiousness and emotional stability are also as expected. The result is thatbeing low in conscientiousness and emotional stability (or high in neuroti-cism, if one prefers the alternative terminology), while also being high inopenness helps individuals successfully understand the connectionbetween their own policy attitudes and political symbolism. The samecan be said when looking at “conservative,” which means the respondentis both symbolically and operationally conservative. The higher one’semotional stability, the more likely one is to be consistently conservative,while openness and agreeableness have the opposite effect. In combina-tion, the odd result here is that increased conscientiousness makes being aconsistent liberal less likely, but does not make being a consistent con-servative more likely. The opposite is true with agreeableness. Increasedagreeableness makes being a consistent conservative less likely, but doesnot help make being a consistent liberal more likely.

Turning now to the “confused liberal” category, which representsrespondents who are operationally liberal, but symbolically conservative.These are the individuals that Ellis and Stimson refer to as conflictedconservatives. Past work provides little guidance as to what one shouldexpect to see here. I suggest that since these individuals are choosing tostand with the majority in both symbolic and operational ideology,increased agreeableness could account for this desire for social acceptance


Tab

le2.1

Big

Five

andne

wjointop

erationala

ndsymbo

licmeasuresof

ideo

logy

Variables

Liberal

Shylib

eral

Confusedlib

eral

Moderate

Conservative

Big

Five

Extraversion

0.01

7−0.01

5−0.02

20.07

8*−0.04

2(0.021

)(0.019

)(0.025

)(0.037

)(0.031

)Agreeablene

ss0.04

8−0.01

30.11

0***

−0.06

8−0.13

0***

(0.029

)(0.026

)(0.033

)(0.051

)(0.040

)Con

scientiousne

ss−0.07

5**

0.04

7*0.04

2−0.08

2*−0.01

5(0.029

)(0.026

)(0.035

)(0.046

)(0.046

)Emotionalstability

−0.05

4*−0.02

1−0.02

50.03

80.12

8**

(0.024

)(0.021

)(0.028

)(0.040

)(0.041

)Openn

essto

experien

ce0.09

5***

0.04

3*0.00

6−0.07

0−0.07

6*(0.029

)(0.024

)(0.031

)(0.044

)(0.039

)Control

variables

Male

−0.15

1−0.05

90.10

6−0.15

00.05

1(0.126

)(0.109

)(0.142

)(0.205

)(0.186

)White

0.14

8−0.27

2**

−0.12

70.21

50.48

0*(0.128

)(0.113

)(0.146

)(0.223

)(0.219

)Age

−0.01

2**

0.00

40.01

6***

−0.00

10.00

2(0.004

)(0.003

)(0.004

)(0.006

)(0.005

)Edu

catio

n0.12

0**

−0.13

5***

−0.01

6−0.12

10.04

6(0.044

)(0.040

)(0.049

)(0.078

)(0.061

)Po

litical

know

ledg

e0.22

9***

−0.13

7***

−0.12

9***

−0.22

5***

0.29

9***

(0.033

)(0.019

)(0.036

)(0.054

)(0.053

)


Religious

fund

amen

talism

−0.50

2***

0.13

5*0.35

5***

−0.16

10.43

0***

(0.088

)(0.075

)(0.092

)(0.135

)(0.129

)Interestin

politics

−0.00

2−0.03

2*0.03

0−0.00

70.00

7(0.018

)(0.015

)(0.019

)(0.030

)(0.026

)Repub

lican

−1.11

6***

−1.40

4***

1.93

0***

−0.83

4**

2.30

1***

(0.289

)(0.180

)(0.289

)(0.290

)(0.463

)Dem

ocrat

1.78

4***

−0.84

3***

0.34

5−0.87

2***

−1.44

0*(0.239

)(0.162

)(0.287

)(0.251

)(0.592

)Log

pseu

dolik

elihoo

d−90

6.36

−1,15

5.7

−80

7.03

−41

5.78

−46

2.70

Pseu

do-R

20.26

90.08

90.13

30.08

70.34

5N

2,11

72,11

72,11

72,11

72,11

7

Notes:R

obuststandard

errors

inparenthe

ses.*p

≤0.05,

**p≤0.01

,***

p≤0.00

1,on

e-tailed

The

depend

entvariable

isadu

mmyvariable

inallfi

vemod

els,thus

logitmod

elswereestim

ated

ineach

case.


by the largest number of individuals possible in each category, regardlessof the inherent conflict. As one can see in Table 2.1, agreeableness is theonly Big Five personality trait to reach statistical significance.Agreeableness increases one’s likelihood of being a confused liberal, anddecreases one’s likelihood of being a consistent conservative. What isinteresting here is that in both of these categories the respondent claimedsymbolic conservatism, but increased agreeableness drives individuals tosupport liberal spending priorities. This result is consistent with what pastworks have found regarding the agreeableness and operational ideologyconnection. However, while it appears that people may at times getconfused about the meaning of abstract symbolic ideological labels, agree-ableness still leads to support for liberal policy.

“Shy liberals” are respondents who are operationally liberal, but refuseto label themselves as either liberal or conservative. The results afterestimation of the logit model predicting holding this combination ofideologies finds both conscientiousness and openness reaching conven-tional levels of significance. The odd part is that they both have positivesigns. This means that increased conscientiousness and openness lead to anincrease in the probability of one being a shy liberal. This is a strikingfinding because, to my knowledge, there has never been a test of the BigFive’s relationship with political ideology that has found these two traits tohave the same effect. They are almost always opposites. This suggest thekey variable for operationally liberal individuals is their level of conscien-tiousness. If they are low in it, they will simply call themselves liberals, andif they are high in it, they won’t choose liberal or conservative at all.

Finally, results estimating the joint ideology labeled “moderate,” thatis, those respondents who were operationally moderate and symbolicallylabeled themselves as such, find extraversion with a positive sign andreaching significance. This indicates that the more extraverted individualare, the more likely they are to be moderate. This is the first time extra-version has been significantly related to any of the ideological measures Ihave utilized. In addition, consistent moderates also appear to depend onlow levels of conscientiousness, which is significant and negatively signed.Thus, high extraversion and low conscientiousness increases the probabil-ity of one holding operationally moderate attitudes combined with apreference for the moderate symbolic label.

In sum, the most detailed analysis these results can provide comeswhen looking at operational liberals. Shy liberals, confused liberals, andconsistent liberals all hold operationally liberal attitudes, that is, prefer


increased spending. What drives their choice of symbolic ideologicallabel? The results in Table 2.1 provide strong support for the notionthat one of the key drivers of this choice is found in their personalitytraits. If operationally liberal individuals are high in agreeableness, theywill tend to choose conservative (thus becoming confused liberals). Ifthey are high in openness and high in conscientiousness, they will tendto call themselves moderate (thus becoming shy liberals). If they arehigh in openness and low in conscientiousness, they will simply labelthemselves liberals.

Of course, the complete story of what causes individuals’ joint ideologyalso includes the effects of many of the control variables in Table 2.1.Political knowledge operates exactly as one would expect. It is highlysignificant in each model with a positive sign when predicting consistentliberals and conservatives, and a negative sign otherwise. Thus, it is clearthat increased political knowledge helps individuals align their specificpolicy attitudes with the correct symbolic ideology. The partisan dummyvariables also produce results in line with what would be expected. In thiscase, the meaning of the coefficients for Republican and Democrat are theeffects of each of these in relation to the dropped category, which isindependent. The religious fundamentalism variable also produces theexpected results when it comes to predicting being a consistent liberal orconservative, that is, as fundamentalism increases, one is more likely to bea conservative and less likely to be a liberal. What is interesting is thatincreased religious fundamentalism also increases the probability of beinga “shy liberal” and “confused liberal.” The upshot here is that no matterwhat one’s operation policy attitudes are, increased religious fundament-alism makes it much less likely for one to apply the liberal label tothemselves. This result is consistent with Ellis and Stimson’s argumentsabout the effect of religion on how people view the meaning of the wordsconservative and liberal. The word liberal is toxic to highly religiousindividuals.

Finally, looking at the variable “Interest in Politics,” which is based ontwo questions asking respondents how many days the last week they (1)watched television news and (2) discussed politics. Contrary to expecta-tions based on dissonance theory, this variable only reaches significanceonce. The more interested in politics one is, the less likely one is to be a“shy liberal.” While that result is consistent with the expectation thatincreased interest should lead to less cognitive dissonance, the failure ofit to reach significance in any of the other models cast doubt on its


importance. However, political knowledge is often viewed as a proxy forpolitical interest. Not shown here, but given the potential for multicolli-nearity between these two variables, I re-estimated each of the modelswithout political knowledge (i.e., everything is the same as in Table 2.1,except Political Knowledge was dropped). When I do this, the politicalinterest variable is a significant positive predictor of being a consistentliberal and consistent conservative exactly as dissonance theory wouldpredict. It is still negatively correlated with shy liberals, but it does notpredict consistent moderates or confused liberals (i.e., conflicted conser-vatives). All told, dissonance theory does find support in this data when itcomes to the effects of interest on the likelihood of holding consistentpolitical beliefs.

Substantive Meaning of Personality’s Effect on Operational Liberals’Choice of Symbolic Ideology

Finding statistically significant results is only halfway to a complete under-standing of the importance of personality in helping individuals success-fully understand and engage with politics. Scholars in their zeal to reportstatistical significance often forget to examine the substantive meaning oftheir results. Unless one is adept at thinking in terms of logged-odds,looking directly at logit coefficients does not provide much help in under-standing the substantive meaning of these variables. Therefore, Fig. 2.3presents results documenting how the probability of operational liberalslabeling themselves as symbolic liberals, moderates, or conservativeschanges as their levels of openness, agreeableness, and conscientiousnesschange.

Part A of Fig. 2.3 depicts the marginal effect of openness to experienceon the probability of an operational liberal labeling themselves as a sym-bolic liberal, holding all other variables at their mean. Openness rangesfrom 0 to 12 and moving from a score of 4 (a score of 4 is in the 5thpercentile of this variable) on openness to 10 (75th percentile) increasesthe probability of being a consistent liberal by 0.078. As a reference point,political knowledge ranges from 0 to 9, and moving from a score of 1 (5thpercentile) to a score of 6 (75th percentile) increases the probability ofbeing a consistent liberal by 0.16. Religious fundamentalism has a similarlysized effect. Moving from saying the Bible is a fairytale to saying it is theliteral word of God decreases an operational liberals’ likelihood of labelingthemselves as liberal by 0.144. Partisanship, as one might suspect, is the


Part A

Part B

.1.1

5.2

.25

.3.3

5

Pro

babi

lity

of L

iber

al Id

eolo

gy

0 2 4 6 8 10 12Openness to Experience

Predictive Margins with 95% CIs.0

5.1

.15

.2.2

5

Pro

babi

lity

of C

onfu

sed

Libe

ral I

deol

ogy

0 2 4 6 8 10 12Agreeableness

Predictive Margins with 95% CIs

Fig. 2.3 Predicted probability of joint operational and symbolic ideology


largest player here. Moving from identifying as Republican to Democratincreases the probability of being a consistent liberal by 0.39. Finally,conscientiousness and emotional stability have significant, but smallersubstantive effects than openness. A similar increase in conscientiousnesscorresponds to a decreased probability of choosing liberal by 0.065, whilean increase in emotional stability leads to a 0.045 drop in probability. Ofcourse, these small results are based on looking at each of the personalityvariables independently, but the combined effect of the Big Five traits is aslarge as any other variable besides partisanship, (i.e., openness 0.078 +conscientiousness 0.065 + Emotional Stability 0.045 = 0.188), althoughopenness is pulling in the opposite directions of the other two.

Part B of Fig. 2.3 shows the marginal effect of agreeableness on theprobability of operational liberals calling themselves conservatives. Onceagain, these are Ellis and Stimson’s conflicted conservatives. As operationalliberals’ level of agreeableness increases from 5 (5th percentile) to 10(75th percentile), their probability of labeling themselves as liberalincreases by 0.071. Putting this value in context, the same type of increase

Part C

.2.2

5.3

.35

.4P

roba

bilit

y of

Shy

Lib

eral

Ideo

logy

6 8 10 12Openness to Experience

Conscientiousness = 6 Conscientiousness = 12


Fig. 2.3 Continued


in political knowledge is associated with a 0.079 decrease in the prob-ability of being a confused liberal, while an increase in religious funda-mentalism from a non-believer to a fundamentalist believer increases thisprobability by 0.083. Interestingly, age is also a significant predictor ofthis type of confusion. Moving from a 21-year-old (5th percentile) to a60-year-old (75th percentile) increases the probability of making thismistake by 0.07. Of course, partisanship has the largest effect. Movingfrom a Democratic identifier to a Republican identifier increases theprobability of an operational liberal identifying themselves as conservativeby 0.211. All told, agreeableness has as large an effect on one holdingthese contradictory beliefs as political knowledge does, albeit in the oppo-site direction.

Finally, Part C of Fig. 2.3 documents the marginal effect of openness toexperience on the probability of being a shy liberal at two different levelsof conscientiousness. I present the effects of openness and conscientious-ness together to emphasize their joint effect. The 95 percent confidenceintervals of each line overlap at each value of openness. In order toimprove one’s ability to make comparisons between the various pointestimates, I have marked a horizontal dashed line at 0.278 indicating thetop of the confidence interval around the first “x” point estimate whereopenness and conscientiousness both equal 6. The predicted probability ofbeing a shy liberal at this point is 0.24. Moving from a value of 6 and 6 inopenness and conscientiousness to a 10 and 10, increases this by 0.063.Increasing political knowledge from a score of 1 to 6 decreases thisprobability by 0.134, while moving from non-believer to religious funda-mentalist increases it by 0.049. An increase in political interest from 2 (5thpercentile) to 11 (75th percentile) decreases the probability of being a shyliberal by 0.054. Finally, changing from a Democrat to a Republicanincreases this probability by 0.099. Therefore, the combined effect ofconscientiousness and openness is lower than partisanship and politicalknowledge, but larger than the effect of religious beliefs and interest inpolitics.

SUCCESS, FAILURE, AND THE UNCONSCIOUS MIND

When it comes to the capability of the average citizen to understandand engage in politics intelligently and consistently, research has typi-cally found that citizens are simply not up to the task (Delli Carpiniand Keeter 1996). Those who have attempted the seemingly herculean

SUCCESS, FAILURE, AND THE UNCONSCIOUS MIND 61

mission of defending the average citizen, and by extension the folktheory of democracy, almost always abandon the individual for thecomfort of the aggregate polity (Page and Shapiro 1992). In theaggregate, one can take solace in the knowledge that the country’sgeneral policy mood appears to have an effect on the aggregate set ofpolicies produced by Congress, the president, and even the SupremeCourt (Erikson et al. 2002). But here I am interested in the individual,in the chaos that is the average person trying to live their life whilebeing asked by a political system to be an expert in dozens of differentpolicy areas; to judge the past actions of incumbents against theirclaims of credit for the good and disassociation from the bad; and tolook at challengers whose history and qualifications they inflate withvague policy positions. It is this democratic citizen that fascinates andconfounds. This citizen who is asked not to vote in one election, butdozens; not asked to choose just between candidates, but also a myriadof referenda from social policy to tax policy because elected officialsrefuse to make politically difficult decisions themselves. When looked atclosely, given the demands placed on the average democratic citizen, itis amazing that they are able to function at all.

But they do function, albeit frequently falling short of democratic ideal.They continue to show up to the voting booths despite recent efforts tomake it harder to do so. They contact their representatives, they donate tocampaigns, and they join groups, even if it is less common than it oncewas. So there is success out there among the citizenry, which is too oftenoverlook by scholars focused on the rather large shadow of the failures.Here I have built on work on ideology and personality in order to furtherour understanding of the roots of both success and failure to understandpolitics consistently. These roots are in the unconscious mind and manifestthemselves in the form of personality traits. Personality is not somethingwe choose or are generally consciously aware of, yet it affects our under-standing of the external world and influences every decision we make.

Scholarly work documenting the connection between personality andideology have found a strong connection between liberalness and open-ness to experience and a somewhat weaker connection with agreeableness.It has also found a strong connection between conservativeness and twoBig Five traits, conscientiousness and emotional stability. These findingswere largely supported by the data presented in Table 2.1. But scholars ofpolitical personality have yet to explore the possibility that individualsmake mistakes when they label themselves as liberal or conservative.


Works by Ellis and Stimson (2012) and Popp and Rudolph (2011) clearlydemonstrate that individuals do make mistakes when attempting to con-nect their specific policy preferences (referred to as their operationalideology) to the abstract label scholars would give this set of policypreferences (referred to as their symbolic ideology). Importantly, peoplewho hold liberal operational policy preferences and choose to label them-selves symbolic liberals behave differently than if those same operationalliberals choose to call themselves conservatives. Thus, understanding whysome people successfully align their operational and symbolic ideologiesand others do not has important behavior implications.

What leads to success in this case? The obvious answers are the oldest.Political knowledge and partisanship both are important contributors tosuccess here. Highly knowledgeable partisans are very good at connectingspecific policy preferences to abstract ones. But this is only a partial answer.The Big Five traits openness to experience and emotional stability play asignificant role in helping individuals successfully make this connection. Asone’s level of openness increases and level of emotional stability decreases,the more likely one is to successfully label themselves as liberal when theyhold operationally liberal policy preferences. The opposite is true forconsistent conservatives. As one’s level of openness decreases and level ofemotional stability increases, the more likely one is to call themselvesconservative when they hold conservative policy attitudes.

The story does not end there, however. Each of these successful ideol-ogies has a third Big Five trait that contributes to this success. In the caseof consistent liberals, low levels of conscientiousness are important(because high levels push them to be shy liberals). In the case of consistentconservatives, it is low levels of agreeableness. Thus, the complete story ofsuccess is one that points to a combination of Big Five traits, each havingan independent effect that in combination is comparatively quite large(i.e., as large as political knowledge). Finally, the often forgotten wayindividuals can be ideologically consistent is when they are operationalmoderates and label themselves as moderates. The results from Table 2.1point to an interesting combination of high extraversion and low con-scientiousness. Extraversion is rarely a significant predictor of ideology inpast works. This is perhaps because these works so rarely consider moder-ates. Of course, why extraverts are more likely to be moderate is beyondthe scope of this work, but certainly worth further exploration.

What leads to failure to connect operational and symbolic ideology?The answer is that it depends on what kind of inconsistent ideology is

SUCCESS, FAILURE, AND THE UNCONSCIOUS MIND 63

being considered. The confused liberal (i.e., conflicted conservative) hasreceived the most attention. The analysis presented here suggests a per-son’s level of agreeableness is a key to understanding why some individualswho are operational liberals call themselves conservatives. The more agree-able one is (i.e., the more one values positive social relationships), themore likely one is to support increases in spending on specific issues, whilesimultaneously thinking of themselves as a conservative. Ellis and Stimsonindicate that which side of a particular policy debate these individuals willfall on is often dependent on the framing of that particular debate. Thissusceptibility to framing effects is not surprising given these individuals arelikely high in agreeableness. Agreeable individuals look for consensus,which is exactly what alternative framings of an issue can alter. Beyondthe Big Five, religious fundamentalism contributes to making this kind ofmistake. The more fundamentalist individuals are, the more they willrefuse to label themselves as liberals, regardless of the content of theiractual policy attitudes. In this case, partisanship is not a help to individuals.When one is both a Republican and an operational liberal, the internalconflict is obvious, and these individuals will be far less likely to labelthemselves as liberal.

Finally, when looking at shy liberals, we see individuals who areoperationally liberal, but call themselves moderate. These individualsare high in openness to experience, which would normally lead themto a liberal symbolic ideology, but they are also high in conscientious-ness, which typically points them in the opposite direction. The result isan individual who is less likely to identify themselves as either liberal orconservative. Thus, for operational liberals, the key difference betweenthem being a consistent liberal and a shy liberal is their level of con-scientiousness. Low-scoring individuals are more likely to call themselvesliberals, while high-scoring individuals tend to call themselves moder-ates. In addition to the effects of the Big Five, religious fundamentalismonce again pushes individuals away from using the word liberal, which issimply unacceptable to them. Partisanship, political knowledge, educa-tion, and interest in politics all contribute to reducing the likelihood ofthis kind of confusion.

In summary, the lesson from this chapter is that when it comes to ideolo-gical labels, be they operational or symbolic, individuals’ durable personalitytraits are significant drivers of how they understand the political world. In thenext chapter I move beyond the abstract connection people have with politicsthrough ideology. I examine how the Big Five contribute to the successes and


failures individuals have understanding five specific policy areas: health care,energy production, same-sex marriage, food stamps, and drug testing welfarerecipients. A large body of research has begun to document the effects ofmotivated reasoning, that is, the idea that individuals do not approach politicswith an eye toward objective understanding, but with the goal of confirmingwhat they already think. This research has shown that individuals are often soblinded by partisanship that they often get factual questions wrong in anattempt to make reality fit the partisan fiction in their mind. I examine howwell individuals can answer factual questions about these five policy areas andshow that success in this case depends on more than political knowledge.

NOTES

1. The authors do offer a more complex study of the dimensionality of policypreferences, but ultimately come to the conclusion that there are twocorrelated dimensions of thought that correspond to standard conceptionsof economic and social/cultural issue areas. Of course, documenting thatthere are two related dimensions does not provide an explanation for whyAmericans prefer liberal policy in the aggregate.

2. Although, agreeableness is properly signed (i.e., negatively correlated withconservatism) with a z-score of −1.86, which would reach one-tailedsignificance.

3. I use the term connected rather than cause here because there is reason to besuspicious of the causal link (Hatemi and Verhulst 2015).

4. See the Technical Appendix for a full discussion of the contents and admin-istration of the Political Personality Survey.

5. I coded “don’t know” as simply maintaining the status quo, which isfunctionally the same as affirmatively stating that one would like to seespending kept about the same.

6. Unfortunately, shy conservatives, confused conservatives, moderate liberals,and moderate conservatives were not estimated due to their being so fewcases of each of these joint ideologies.

NOTES 65

CHAPTER 3

Just the Facts: Citizen Issue Comprehension

In the summer of 2014 a video appeared of Jonathan Gruber, an MITeconomist who had played a role in helping to create the ACA. In it, hereferred to the ignorance of the average American voter as a key to passingthe health-care overhaul (Freyer 2014). This caused a bit of an uproaramong some political commentators. Apparently, pointing out the ignor-ance of the average American when it comes to public policy is taboo incertain contexts. This can be somewhat disconcerting for those of usteaching courses on American politics. Discussions of what citizens knowand don’t know about the U.S. political system comprise the bulk of thefirst month of my introduction to American Politics class and the conclu-sion is never a favorable one. The uproar is odd, of course, since theaverage Americans’ low level of political knowledge is not a questionanymore, it’s an axiom. Seventy plus years of scholarly research has con-firmed this as an enduring fact of American political life.

As a result of this ignorance, scholarly attention has turned to trying tounderstand its causes, effects, and developing ways to “fix” it. While muchwork has been done connecting the Big Five personality traits to holdingspecific policy preferences, little has been done to examine the role thatpersonality plays in driving the correct understanding of political facts.I find this to be an important omission. As I have argued, personality traitsare observable manifestations of unconscious brain processes. Ha andLau’s (2015) recent work documents that unconscious brain processesin the form of the Big Five personality traits have a real effect on one’s


67

likelihood of voting correctly during the 2008 presidential election. Theyshow that openness and conscientiousness moderate the traditionallyunderstood relationship between partisanship and voting. Therefore, ifone is interested in “fixing” the problem of political ignorance in theUSA, then finding that significant contributors to such ignorance arepersonality processes that are outside of our control (like Ha and Lau’swork and what I present here) will demand fundamentally different solu-tions than those proposed if simple apathy and disinterest were the causes.While it may be possible to overcome individuals’ apathy and lack ofeducation, it is not as clear how one overcomes their personality.

Here I examine citizen comprehension in five specific policy areas:health care, energy production, same-sex marriage, food stamps, anddrug testing welfare recipients. In each of these areas, I asked factualquestions of the respondents in order to tap their understanding of eachtopic. For my purposes, it is important these issues be specific rather thanvague abstract notions like liberalism and conservativism for two reasons.First, I need to be able to identify when a respondent answered incorrectlyand, second, as was learned from the results of the last chapter, genericsymbolic concepts such as liberal and conservative do not necessarily haveanything to do with one’s specific policy attitudes. Too often scholars andpolitical commentators make the leap from an individual’s indication oftheir symbolic ideology in a survey to assuming the respondents’ attitudeson specific policies (Pierce 2015). This is simply wrong. Finally, I specifi-cally chose a wide range of topics because I am interested in how person-ality affects one’s knowledge in general. Thus, this chapter’s analysis buildsfrom a foundation that looks at the effects of personality in each individualarea to an analysis where all the areas are combined to see how the Big Fivepredict the quantity of incorrect answers.

MOTIVATED IGNORANCE

One of the most important movements in the scholarly understanding ofthe causes of political ignorance is the concept of motivated reasoning(Bartels 2002; Bolsen et al. 2014; Claassen and Ensley 2016; Gaines et al.2007; Jerit and Barabas 2012; Jost and Amodio 2012; Leeper andSlothuus 2014; Lodge and Taber 2013). The simple idea is that indivi-duals are motivated to seek out and interpret information in a manner thatconfirms what they already believe. This is not a new idea, of course.Scholars have pointed to concepts such as selective perception for decades.

68 3 JUST THE FACTS: CITIZEN ISSUE COMPREHENSION

But, the more recent push to look beyond ignorance as the result of apathyand/or lack of education forces scholars to come to grips with the realitythat highly educated, knowledgeable, and interested individuals are notliving up to the ideal of the democratic citizen either. Joseph Schumpeterappears to have recognized the twin problem of motivated reasoning andhighly intelligent citizens (even if he would never have used the currentterminology) as far back as the 1940s. As quoted in the first chapter, hefamously pointed out that when citizens engage in politics, they turn intoprimitives. Too many people assume he must have been talking about theuneducated masses, but he was not. The example he used was an attorneywho knows every detail and complex legal argument involved with aparticular case, but happily accepts childlike assertions when engaging inthe political realm.

One of the best examples of motivated reasoning at work comesfrom Kahan et al. (2013). What makes this piece so fascinating is thatthe authors add a third component to the traditional connectionbetween partisanship and issue knowledge. In it they develop an experi-ment that tests the effects of partisanship on one’s willingness to domath correctly when the math would contradict one’s partisan beliefsabout the issue of gun control. People vary substantially in their abilityto handle math problems. The authors leverage this fact to test whetherpartisans who are really good at math will be motivated to ignore thecorrect answer and instead give the answer that is most consistent withtheir partisanship.

What they do is, first, measure participants’ numeracy and partisanship.Then they provide them with fictitious data on the effects of changes togun control laws on crime rates. This consists of giving respondents simplecounts of the number of cities that saw either an increase or a decrease incrime after either banning or not banning carrying a concealed handgun.For example, one of the sets of data provided indicated that of cities thatbanned concealed handguns, 223 saw an increase in crime while 75 saw adecrease; and of cities that did not ban concealed handguns, 107 saw anincrease in crime and 21 saw a decrease in crime. The job of the respon-dent is to determine whether banning concealed weapons produces betterresults than not banning them. Of course, this requires some math skills,since one first needs to convert the raw numbers into percentages beforecomparing them. Thus, looking at the data in the example, 25 percentof the cities that banned concealed handguns saw a decrease in crime(75/298 = 0.25), while only 16 percent of cities that did not ban

MOTIVATED IGNORANCE 69

concealed handguns saw a decrease (21/128 = 0.16). As a result, in thisexample, the correct answer is that banning concealed weapons reducescrime more.

Importantly, Kahan et al. offered multiple versions of these numbers,sometimes banning concealed handguns was better at reducing crime,while other times not banning them was better at reducing crime. Inother words, sometimes the numbers supported Democrats’ beliefs andother times they supported Republicans’ beliefs. One would expect thatthe better at math one is, the better one will be at answering thesequestions correctly regardless of the substantive meaning of the numbers.But that is not what they found. They found that when the numberssupported the participant’s partisanship, numeracy had a big effect, thatis, the more skilled at math one is, the more likely he or she is to get thequestion correct. However, when the results contradicted the participant’spartisanship, numeracy had no effect. Highly numerate individuals eitherrefused to report what the numbers told them, or questioned their ownmathematical abilities when faced with contradictory results.

When it comes to motivated reasoning, the question still remains as towhether it is a conscious or unconscious process. The word “motivated”suggests agency and conscious deliberation on the part of individuals. Butthat does not have to be the case. There is ample reason to suspect that theeffect of partisanship is an automatic unconscious process. Campbell et al.(1960) thought of partisanship as a perceptual screen through whichpeople view the world, thus, coloring everything they see. More recentresearch has further supported the notion that partisanship is somethingthat happens automatically outside of conscious brain processing (e.g.,Dodd et al. 2012; Knutson et al. 2007; Westen et al. 2006). Regardless ofwhether the though process was deliberative or automatic, the result ofKahan et al.’s work is an unequivocal demonstration of the role partisan-ship has in creating one’s reality. Importantly, this effect is not limited tothe ignorant uneducated masses—it is universal.

This fact has been amply demonstrated by Taber and Lodge (2016).They document the unconscious mind at work through a serious ofexperiments that utilize treatments that occur too quickly for the indivi-dual to become consciously aware of them. They then show that positiveor negative symbols influence attitudes toward hypothetical politicalcandidates. Importantly, they demonstrate that the effect can last for upto 45 minutes and that the more politically sophisticated the participant,the larger the effect of the unconscious trigger. Bolsen et al. (2014) and


Kahan (2013) using different data and approaches also document thatmotivated partisanship is not a simple cue for low-information, disinterestedvoters, but the very path by which the most civically engaged citizens processpolitical information. The overall result of works like these suggest thatmotivated reasoning may not be as universal as previously thought, butthat it might be the exclusive domain of the most politically engagedcitizens.1

Here I move away from the motivating force of partisanship in creatingone’s political reality and ask what role personality plays in either helpingor hindering the generation of a correct understanding of political issues.To be sure, partisanship is still an important contributor to this under-standing. Indeed, the data I present below provides further support for theeffects of motivated partisan reasoning. I argue that personality’s effect,like that of partisanship, is universal and outside of the control of indivi-duals. That is, regardless of education or political knowledge, one’s per-sonality affects the probability that one will hold a correct understandingof various political issues.

PERSONALITY AND POLITICAL FACTS: THE GOOD AND THE BAD

The PPSF survey asked seven factual questions across the five issueareas. These questions are listed in Table 3.1. The answers to each ofthese questions are not always straightforward. They often require verydetailed knowledge of particular policies and issue areas. It might beclaimed that these questions are unfair, since scholars and interestedobservers of mass political behavior know all too well that there are notgoing to be large numbers of people who are capable of answeringthese questions correctly. This is undoubtedly true. However, the folktheories of democracy, upon which our very real political structures arebased, do ask citizens to know these policy areas. Citizens are asked toweigh these complex issues and adjudicate their success or failureduring elections.

For example, after signingMichigan’s new law creating a pilot program in2015–2016 that allows for the testing ofwelfare recipients who are suspectedof using drugs, Governor Rick Snyder said, “This pilot program is intendedto help ensure recipients get thewrap-around services they need to overcomedrug addiction and lead successful lives. We’ll then have opportunityto assess effectiveness and outcomes” (Michigan Government 2014).Presumably, it is not just the governor himself who will be evaluating the

PERSONALITY AND POLITICAL FACTS: THE GOOD AND THE BAD 71

effectiveness of this government program. The average Michigander will beasked to judge this program’s effectiveness, too. Of course, there is no doubtthat the vastmajority of themwill have little to no knowledge of the programwhen the next election comes around, yet this very ignorance of the detailsand results of this and other programs just like it will be viewed as a tacitendorsement of them when an incumbent is reelected.2 Thus, while it iscertainly true that democratic folk theories have always contained a falseimpression of the average citizen, the very structure of our political systemis based on those false impressions.

As a result, the type of factual questions presented in the PPSF surveyare fair because these are the policy details that democratic governmentsproduce and if one cannot expect the average voter to have any under-standing of these policies, than the primary mechanism through whichdemocratic governments and politicians are held accountable is broken.The all too common refrain at this point is that individuals do not need toknow because they rely on cues and opinion leaders to do what they want(e.g., Popkin 1991). I would refer those making this argument to the

Table 3.1 PPSF factual survey questions

Drug testing welfare recipients:•Do you think people who receive some type of welfare assistance are more likely to abusedrugs than the average U.S. citizen?

• Members of which party are more likely to support requiring drug testing for peoplereceiving welfare?

Health insurance:• Do you get help paying for your health insurance by either the state or federalgovernment?

Same-sex marriage:• In states where same-sex marriage is legal, can religious organizations like churches orsynagogues legally refuse to marry same-sex couples?

Energy policy:• Has U.S. oil production gone up or down since President Obama took office inJanuary 2009?

Food stamps:• About what percent of federal food stamp benefits do you think go to individuals livingin households that have income from a job?

• About what percent of federal food stamp benefits do you think go to individuals who areworking age but do not work a paid job and are not living with children, elderly, ordisabled individuals?


previous chapter. Indeed, one of the most commonly pointed to cue isone’s partisanship, but my own work shows that individuals low in poli-tical information are not helped by having strong partisanship, but areactually more likely to vote for the wrong candidate when they are highlypartisan (Dusso 2015). When respondents to a survey say they are liberalor conservative that does not mean they support liberal or conservativepolicies. In fact, there is little reason to assume that a person who labelsthemselves as a conservative actually supports specific conservative policies.Any work that bases the success of American democracy on the connectionbetween these abstract concepts and specific policy outcomes is mean-ingless. It is a kind of ecological fallacy. One knows no more about thespecific policy positions of a specific individual who happens to live in a“red” state than one does of a person who labels themselves a conservative.

Looking at Table 3.1, the first two questions concern drug testingwelfare recipients. The first question asks if welfare recipients are morelikely than the general populous to take drugs. This answer is no (Grantand Dawson 1996; Yacoubian and Urbach 2002). Indeed, after Floridaimplemented mandatory drug testing in 2011 and 2012, just 2.7 percentof welfare recipients tested positive (Klas 2014), which is lower than thenational average of 9 percent (NIH: National Institute on Drug Abuse2015). Michigan’s pilot program lauded by Gov. Snyder as a way to helppoor drug users has not identified a single one, as of June 2016 (MichiganGovernment 2014; Felton 2016). The second question asks respondentsto identify which political party is most in favor of requiring drug testing ofwelfare recipients. In this case the answer is the Republican Party.According to the National Conference of State Legislatures, as ofDecember 29, 2014 (the PPSF survey was administered in July of2014), there have been 12 states to pass some form of law regarding thistype of testing. These states are Alabama, Arizona, Florida, Georgia,Kansas, Michigan, Mississippi, Missouri, North Carolina, Oklahoma,Tennessee, and Utah. In every case both houses of the state legislaturewere majority Republican and in only two cases, Missouri and NorthCarolina, were the governors Democrats. However, even these twoinstances of Democratic complicity are too many since Governor BeverlyPerdue vetoed the bill and the North Carolina state legislators overrodethat veto. There is not a single instance where a state controlled byDemocrats has passed this type of legislation.

The next question is on health insurance and asks respondents if theyhave received any help paying for their health insurance. In previous


questions respondents were asked where they get their health insurance, ifthey have any. In this case, the responses should be yes, if the respondenthas health insurance. As Mettler’s (2011) work on the submerged welfarestate has indicated, everyone receives some kind of benefit from either thestate or federal government when it comes to health insurance, even if it is“just” a reduction in one’s taxable income. Next is a question about theeffects of legalized same-sex marriage on individual religious institutions’practices. In this case the answer is yes, religious organizations can legallyrefuse to marry same-sex couples. States have either explicitly put into lawexceptions for religious organizations and clergy or courts have foundsuch exceptions, such as in both California and Iowa (Masci 2013). Thenext question asks about energy production and if U.S. oil production hasgone up or down since President Obama took office. The answer here is,once again, yes. Indeed, as of August 2014, the U.S. Energy InformationAdministration indicates that U.S. crude oil production is at levels notseen since the mid-1980s (EIA 2014).

The last two questions in Table 3.1 are regarding food stamp recipients.These questions were designed to get at knowledge of facts regarding thelives and activities of food stamp recipients versus the often fictitiouscharacterization of them as lazy individuals abusing the system (see, e.g.,the special investigative report by Brett Baier, “The Great Food StampBinge” which originally aired on Fox News in August of 2013).3 Theofficial name of the food stamp program is the Supplemental NutritionAssistance Program (SNAP), but the more colloquial term “food stamps”was used in order to not unnecessarily confuse respondents. I am notinterested in whether respondents know the official name, but in theirperceptions of the individuals who receive benefits from the program. Inaddition, these questions are, perhaps, a bit more difficult than the pre-vious questions, since they ask for percentages. The first question asksabout the percentage of food stamp benefits that go to individuals inhouseholds with income. According to the U.S. Department ofAgriculture’s February and December 2014 reports, 42 percent of SNAPparticipants in fiscal year 2012 lived in households with earnings and fiscalyear 2013 saw that number rise to 43 percent (USDA February 2014b;December 2014b). I was generous when coding these responses, with anyanswer between 37 and 47 percent being coded as correct. The secondquestion gets directly at the notion of “lazy” welfare recipients. It askswhat percentage of recipients are of working age, but do not work and arenot caring for dependents. Once again, according to the U.S. Department


of Agriculture’s full reports for fiscal years 2012 and 2013, the answer is13.8 percent in 2012 and 14.5 in 2013. I coded any answer from 9 to 19percent as correct.

Table 3.2 shows the percent of respondents who answered each of thefactual questions correctly or incorrectly. As expected, the questionsrespondents struggled with the most were those asking them to estimatethe proportion of food stamp recipients either living in a household withan income or who are of working age but do not work. Also as predictedby Mettler’s work on the submerged welfare state, less than 20 percent ofrespondents with health insurance thought of themselves as getting helpto pay for it. This number drops to just 8 percent when looking only atrespondents who indicated that they get their health insurance throughtheir employer. This is the case even though the cost of exemptingemployer provided health insurance from one’s taxable income is esti-mated by the Joint Committee on Taxation to be 143 billion dollars in2014 and to rise to 172 billion in 2018 (Joint Committee on Taxation2014). In other words, people are virtually unaware of a 150 billion dollargovernment program that directly benefits them. The amount spent onthis program is far more than is spent on SNAP benefits, which was lessthan 80 billion annually from 2012 to 2014.4

Respondents did much better when asked about whether religiousorganizations could be legally required to marry same-sex couples and

Table 3.2 Percent of PPSF respondents answering factual questions correctly/incorrectly

Question Correct Incorrect Don’t know

Welfare recipients not more likely to be on drugs 36.7 40.0 23.3Republicans more likely to support drug testing 55.5 27.3 17.3Get help paying for health insurance 19.2 62.2 18.6U.S. oil production has increased since Pres.Obama took office

44.0 20.5 35.5

Religious Organizations can refuse to marrysame-sex couples

56.5 18.5 25.0

Percent of food stamp recipients in householdswith an income

9.5 90.5

Percent of food stamp recipients not livingwith dependents, capable of working,but do not earn an income

16.6 83.4


which party is the primary supporter of drug testing welfare recipients. Inboth cases better than half answered correctly. Republican identifyingrespondents did particularly well when answering these two questions,with 68 percent knowing that religious organizations are exempt fromsame-sex laws and 65 percent knowing that the Republican Party is theprimary supporter of drug testing welfare recipients. However, when itcomes to drugs and welfare recipients, only 36.7 percent of respondentsknew that welfare recipients are no more likely to use drugs than nonwel-fare recipients. Interestingly, the more important an individual thinks theissue of drug testing welfare recipients is, the more likely he or she is to getthis question wrong. And, finally, 44 percent of respondents were able tocorrectly indicate that oil production has increased since Obama tookoffice.

The Big Five and Political Understanding

What should be the expected effect of each of the Big Five personalitytraits when it comes to factual political knowledge? The simplest answermight be to expect the traits that have been found to be highly correlatedwith partisanship to behave in a similar manner. In this case, openness toexperience would be analogous to identifying as a Democrat, while con-sciousness and emotional stability would be akin to Republican identifica-tion. This seems intuitive and reasonable. However, it may also be toosimple. The Big Five personality traits were developed wholly indepen-dently of the meaning of party labels found in the USA in the early twenty-first century. Thus, even though some correlations between them andpartisanship have been found, it is unlikely that they would have thesame effect on answering public policy questions correctly. This potentialconnection will be further explored as I analyze the results of the fullmodels presented below, but a more thorough examination of each trait isin order to determine more specific expectation.

Looking at each of the Big Five personality traits individually, there maybe some clues as to what one might expect of them in this context.When it comes to a simple question of getting things right or wrong,the Big Five trait that jumps out is openness to experience. The genesisof this trait were attempts to get at cultural openness and the intellect(Cattell 1957; McCrae and Costa 1983; Fiske 1949; Goldberg 1992).There is quantitative support for this too. Openness has routinelybeen found to be a strong predictor of general knowledge and intelligence


(e.g., Chamorro-Premuzic et al. 2006; Furnham et al. 2007). Therefore, areasonable expectation might be that the more open to experience anindividual is, the more likely he or she will be able to answer factualquestions correctly. There is some evidence to support this notion inpolitical science research as well (Mondak 2010). Looking at the con-sumption of political information, Gerber et al. (2011b) find that open-ness and emotional stability are correlated with the broadest range ofinformation consumption practices. Given the consumption of politicalinformation, it does not seem too far of a stretch to assume the individualwill be better at answering factual questions too. Ultimately, to the extentthat openness represents some form of intellectual curiosity and generallevel of knowledge, it should improve one’s ability to answer thesequestions.

The expectations of the remaining four traits are less straightforward,since they are not necessarily tied to one’s intelligence. Conscientiousnessis associated with being responsible, productive, and dutiful. It has beenshown to predict doingwell in school and high performance at work (BarrickandMount 1991; Noftle and Robins 2007; O’Connor and Paunonen 2007;Sackett and Walmsley 2014; Vedel 2014). However, there has not been aconsistent finding regarding political participation (Gerber et al. 2011a;Mondak 2010; Mondak and Halperin 2008). Lack of participation suggestsa lack of interest and thus a lack of knowledge, althoughWeinschenk (2014)reports an increase in one’s sense of civic duty as conscientiousness goes up.However, Mondak (2010) finds a negative correlation with political knowl-edge. Thus, while conscientiousness helps one dutifully perform theirresponsibilities at school and work, there is reason to be less sanguineabout consciousness’ effect on political knowledge.

Emotional stability is another trait that has failed to find a consistentconnection to political engagement, a sense of civic duty, or politicalefficacy (Mondak 2010; Weinschenk 2014). However, Gerber et al.(2011b) find that those high in emotional stability are more likely toengage with political information, although this finding is weaker thanwhat they find for openness to experience. Chamorro-Premuzic et al.(2006) find a positive relationship between emotional stability and theirmeasure of general knowledge, but this is not a consistent finding in theliterature (e.g., Chamorro-Premuzic and Furnham 2003; Furnham et al.2007). Overall, if emotional stability has an effect, it appears that it shouldbe similar to openness. However, this is a weak expectation based onconflicting past research.


Agreeableness is another trait for which research has failed to paint aconsistent picture. When it comes to political participation, agreeablenessmay or may not have an effect (e.g., Gerber et al. 2011b; Mondak et al.2010; Mondak and Halperin 2008; Schoen and Steinbrecher 2013). WhileGerber et al. (2011b) find that agreeableness does not necessarily turnpeople off from the inherently conflictual process, it does not appear tostrongly attract them either. Work on the Big Five connection to generalknowledge also fails to find a consistent result, or any significant result at all.Thus, there is little reason to suspect that being high or low in agreeablenesswill have a systematic effect on answering policy questions correctly.

Extraversion is associated with being outgoing, energetic, active, andexpressive. Being high in extraversion is associated with an increasedprobability of participating in politics in numerous ways (Caprara et al.2010; Cooper et al. 2013; Gallego and Oberski 2012; Gerber et al. 2011a;Ha, Kim, Jo 2013; Mattila et al. 2011; Mondak 2010; Mondak andHalperin 2008; Mondak et al. 2010; Quintelier and Theocharis 2013;Vecchione and Caprara 2009). Scoring high in extraversion is also asso-ciated with a stronger sense of civic duty and stronger partisanship (Gerberet al. 2012b; Weinschenk 2014). But does this participation and partisan-ship translate into actual political knowledge? Gerber et al. (2011b) findthat extraverts were more likely to watch local and national news duringthe height of the 2008 election, but not at other times. They suggest thisis because the social interactions that extraverts engage in are more likelyto include something about the upcoming election. Thus, they feel theneed to watch the news. Mondak (2010) either finds a significant negativecorrelation with political knowledge or none at all. Outside of a specificfocus on politics, Chamorro-Premuzic et al. (2006) find a negative rela-tionship between extraversion and general knowledge. O’Connor andPaunonen (2007) meta-analysis of the connection between the Big Fiveand post-secondary educational performance found a negative relationshipbetween extraversion and performance (see also Chamooro-Premuzic andFurnham 2003). Interestingly, Jacobs, Szer, and Roodenburg (2012; seealso Bratko et al. 2006) analyze the difference between individuals’ self-estimates of their intelligence versus test-based measures of their intelli-gence and find that the more extraverted an individual is, the moreinaccurate is their self-assessment of their intellectual capacity. All told,when it comes to extraversion’s effect on correctly answering policy ques-tions, there appears to be reason to believe that it may actually decrease theprobability of being correct.


Predicting Incorrect Answers

Table 3.3 presents a summary of results after estimating models predictinggetting each of the individual questions wrong and also the results ofmodels predicting the total number of incorrect responses. The tablealso includes results of models with and without the numerous controlvariables.5 The dependent variable in each of the first seven models is adummy variable coded 1 if the respondent was incorrect and 0 if he or shewas correct. “Don’t know” answers were coded as incorrect.6 The tablealso presents results after estimation of a model predicting the quantity ofincorrect responses. In this case, I summed the number of questions eachrespondent answered incorrectly. This produced a dependent variableranging from 0 (meaning the individual answered all the questions cor-rectly) to 7 (meaning every response was incorrect). I then estimated atobit model with right censoring, due to the possibility that some observa-tions may have only been stopped from getting more than seven incorrectanswers because they were not asked any more questions. Tobit canaccount for this possibility.

For each of these models, the PPSF survey’s question wording, codingrules, and descriptive statistics can be found in the Technical Appendix.The first set of control variables are standard demographic measurements,that is, gender, race, age, and education.7 These are quite common andthus need little introduction. Second, I include standard political variables,all of which have been found to be important contributors to politicalbehavior and understanding. These are partisanship, ideology (which isthe joint ideology scale I generated in the previous chapter), politicalknowledge, and political interest. I also include a measure of religiousfundamentalism.

Finally, I include two less common measures. First is a measure of theimportance of the particular issue to the individual respondent. Since I amtrying to predict one’s ability to answer factual questions, the assumptionis that one’s level of interest in the particular topic will have an effect onhow closely one follows that particular topic. Second, I include a variablemeasuring the ambiguity of one’s thoughts on a particular issue. Forexample, when it comes to U.S. oil production, it could be that somepeople are strongly in favor of increasing fossil fuel production, whileothers are strongly in favor of developing new alternatives. Thus, whenthey think about the issue, it is clear in their minds. But it is also possiblethat some people have more ambiguous thoughts as they see both the pros


Tab

le3.3

Summaryof

mod

elspred

ictin

gtheBig

Five’seffect

onfactualk

nowledg

e,with

andwith

outcontrolvariables

Question

Extraversion

Agreeableness

Conscientiousness

Em.stability

Openn

ess

w/o

With

w/o

With

w/o

With

w/o

With

w/o

With

Welfare

recipien

tsmorelik

elyto

beon

drug

s*+

*

*

*

*

*

Which

partymorelik

elyto

supp

ortdrug

testing

*+

*+

*

*

*

*

*

Get

help

paying

forhe

alth

Insurance

*+

*+

*+

U.S.o

ilprod

uctio

nhasincreasedsincePres.

Obamatook

office

*+

*+

*

*

*

*

*

Religious

Organizationcanrefuse

tomarry

same-

sexcoup

les

*

Percen

tof

food

stam

precipien

tsin

househ

olds

with

anincome

*

*

Percen

tof

food

stam

precipien

tsno

tlivingwith

depend

ents,c

apable

ofworking

butno

tearning

anincome

*+

*

Sum

ofincorrectrespon

ses(0–7)

*+

*+

*

*

*

Notes:*

=traitwas

asign

ificant

pred

ictorin

themod

el;+

=apo

sitiv

erelatio

nship;

=ane

gativ

erelatio

nship


and cons of either option. This type of ambivalence has been shown to bean important contributor to how individuals understand issues (e.g.,Zaller 1992; Holbrook and Krosnick 2005; Lavine et al. 2012; Tormalaand DeSensi 2008; Mulligan 2013).

There are two pretty clear patterns to be found in Table 3.3. The first isthat openness to experience always has a negative sign, meaning that asone’s level of openness increases, the less likely one is to answer thesequestions wrong. This is true in four of the seven individual questions andalso true in the count model predicting the volume of incorrect responses.The effects of openness to experience are also found to be quite robust. Ineach case, whether the control variables are included in the model has noeffect on the statistical significance of these results. Substantively, open-ness to experiences results are also illuminating. Looking at the first modelpredicting whether respondents know if welfare recipients are more likelyto use drugs, a move from 5 (10th percentile) to 11 (90th percentile)8 onthis scale is associated with about a 12 percent decline in the probability ofgetting the question wrong.

Openness is a significant predictor of the next question too, that is,knowing which party is most likely to support drug testing. In this case, anincrease in openness from the 10th to the 90th percentile is associatedwith about an 18 percent decrease in the probability of getting thatquestion wrong. When it comes to knowing if U.S. oil production hadincreased during the Obama administration, the same sized increase inopenness leads to about an 11 percent decrease in getting the questionwrong. When it comes to predicting the percent of the households receiv-ing SNAP benefits that also have an income, an increase in opennessdecreases the probability of getting that question wrong by about 3.5percent. Finally, the overall model predicting the volume of results issignificant. However, the substantive meaning of the model’s resultsindicated that making a six-point jump on the openness scale onlydecreases the number of incorrect responses by about 6.3 percent.

The second consistent pattern is with the performance of extraver-sion. It is significant in half the models and always has a positive sign,which includes the overall model predicting the volume of incorrectresponses. This indicates that as extraversion increases, the probabilityof answering these factual questions incorrectly also increases. Focusingon the results of the models with control variables, the first modelextraversion research significance in is when respondents are asked whichparty is most likely to support drug testing. In this case, moving from


the 10th percentile (an extraversion score of 2) to the 90th percentile(an extraversion score of 10) is associated with about a 16 percentincrease in the probability of getting this question wrong. The samesized movement in extraversion when predicting knowing whether theUSA has increased oil production is associated with an 8.9 percentincrease in answering that question incorrectly. Turning to the modelpredicting if respondents know the proportion of food stamp recipientsthat are of working age, not living with any dependents, but still do nothave any income, extraversion also has a significant effect. Moving froman introvert (score of 2) to an extravert (score of 10) on this scaleincreased the probability of getting this question wrong by nearly 5percent. Finally, the volume of incorrect responses is also significantlyinfluenced by one’s level of extraversion. In this case, moving from the10th to the 90th percentile increases the number of questions one getswrong by 6.2 percent.

The size of the extraversion and openness’ effect on the overall volumeof incorrect answers may not seem that large on their own. For compar-ison purposes, moving from the 10th percentile score in political knowl-edge (score of 1) to the 90th percentile (score of 7) is associated with a20.6 percent decrease in the number of factual questions one gets wrong.However, when thinking about the influence of personality, I would arguethat one should look at the effect as a whole when comparing to othertraditional variables like political knowledge. In this case, moving fromscoring low in openness and high in extraversion (i.e., a close-mindedextravert) to scoring high in openness and low in extraversion (i.e., anopen minded introvert) decreases the total number of questions one ispredicted to get wrong by 12.1 percent. In other words, the joint effect ofopenness and extraversion is larger than every other variable except poli-tical knowledge.

Beyond extraversion and openness, the other three personality traits areless conclusive. Emotional stability reaches significance in six of the eightmodels without the control variables, but then only twice when they areincluded. Conscientiousness is significant twice, but the sign is negativewhen it comes to knowing which party supports drug testing, meaning anincrease in this trait decreases the likelihood of getting this wrong. Butwhen conscientiousness reaches significance again the sign is positive. Thismeans that an increase in this trait increases the likelihood of not knowingwhether one gets help paying for health insurance. Agreeableness onlyreaches significance once when the control variables are included, when


asked about health-care insurance. The more agreeable one is, the morelikely one is to get this question wrong.

Given the multitude of works, past and present, pointing to the impor-tance of group identity when it comes to individuals’ relationship topolitics (e.g., Achen and Bartels; Converse 1964; Gamm 1989;Hochschild and Einstein. 2015), I further examine the connection ofextraversion and openness to party identification. In doing so, I focusonly on the last model, which predicts the total number of incorrectanswers respondents gave to the factual questions. As stated above, extra-version and openness are both significant with a combined effect that candecrease the number of incorrect responses by 12 percent (when movingfrom a close-minded extravert to an open minded introvert). Partisanshipalso reached significance in this model. Moving from identifying as aDemocrat to a Republican increases the number of questions one ispredicted to get wrong by just under 4 percent. Thus, it is not nearly aslarge an effect as that of either extraversion or openness. But one mightwonder if the effect of these personality traits is conditional on one’spartisanship. I have cited dozens of works already that document theimportance of partisanship as a perceptual screen influencing every poli-tical thought we have.

In order to examine the potential that the relationship between one’spersonality and one’s ability to answer factual questions correctly is mod-erated by one’s partisanship, I re-ran the same tobit model predictingrespondents’ total number of incorrect responses to these questions,with the same control variables. However, this time I included a three-way interaction between extraversion, openness, and partisanship. Three-way interactions can be difficult to interpret, but I accomplished this byasking two simple questions. First, if one is an introvert (score of 2 on thescale) and a Democrat or Republican, does it matter if one is high or low inopenness to experiences? And second, if one is an extravert (score of 10 onthe scale) and a Democrat or a Republican, does it matter if one is high orlow in openness to experiences? The interesting answer is that if one is anintrovert—either a Democrat or a Republican—one’s level of openness toexperience is not relevant.

However, when it comes to extraverts the story is more complex. Forextraverts who identify as Republicans, their level of openness does notmatter, that is, there is not a statistical difference between how an extra-verted Republican high in openness or low in openness responds to thesequestions. But if one is an extraverted Democrat, there is a big difference.


An extraverted Democrat high in openness (score of 11 on the scale) ispredicted to have a score nearly 14 percent lower than an extravertedDemocrat low in openness (score of 5 on the scale). Of course, a simpletwo-way interaction between Democrat and openness finds a similarresult, but it misses the fact that this effect appears to be confined toextraverted Democrats. A two-way interaction between extraversion andDemocrat is no help here either, since it is not significant. Thus, thisfinding only appears when one examines the three-way interaction and itindicates that introverted Democrats do not rely on the openness person-ality trait in order to understand the political world, but extraverted onesmost certainly do.

SUCCESSES AND FAILURES TO ANSWER FACTUAL QUESTIONS

I have argued here that the effect of the Big Five is similar to that ofpartisanship in that they are both automatic brain processes that influencehow people see the world. Like partisanship, personality has been shownto develop over the course of early childhood, eventually coalescing andremaining relatively stable throughout adulthood. Of course, some peoplenever develop a strong sense of partisanship, but that too is a value on thepartisanship scale, that is, none. Thus, partisanship can be strong or weakor somewhere in-between, just like one can be extraverted or introverted,emotionally stable or highly neurotic, open or closed minded. And whereone falls on any of these scales matters, be it a personality or partisanmeasure. Highly partisan individuals bend the political world to fit theirbeliefs, while those with little to no partisanship simply ignore it. Thepreceding analysis provides some answers about how variance in the BigFive personality traits influences political perception.

After looking at each of these questions independently and jointly,with or without control variables, a clear picture of the role of the BigFive personality traits begins to emerge. When it comes to factualunderstanding of public policy issues, the more extraverted individualsare, the less factual knowledge they are likely to have. Why this is thecase beyond the scope of this analysis. It is a very intriguing findingthat deserves further exploration. Perhaps the time extraverts spend onsocial interaction presents a type of opportunity cost that simply leavesthem without sufficient time to spend on learning about public policy.The second strong finding is that openness to experience has a signifi-cant positive effect on knowing political facts. This result is far less


surprising than that of extraversion due to the fact that the opennessscale has routinely shown individuals that are high in openness to alsobe higher in political knowledge and interest. Indeed, the scale’s rootsare found in attempts to capture general openness to cultural experi-ence and intelligence.

The performance of emotional stability surprised a bit as well. Mysurvey of the research failed to find a strong indication that emotionalstability has a consistent effect. At best, I could predict only weakly thatits effect would follow that of openness. However, the results of theanalysis of these seven factual policy questions indicate a stronger, moreconsistent positive effect on knowing political facts than was expected,although it appears to be somewhat mitigated when I include all thecontrol variables. Finally, agreeableness and conscientiousness do notdemonstrate a consistent pattern. Indeed, they appear to only rarelyhave an effect and the direction of that effect is inconsistent. Forexample, on the one hand, those high in conscientiousness were morelikely to know which party supports drug testing welfare recipients. Onthe other hand, those same highly conscientious individuals were lesslikely to know that they are getting help paying for health insurance.Thus, the effect of conscientiousness on political knowledge does notappear to be strong or consistent, but rather conditional on the parti-cular topic.

In the next chapter, I move away from citizens’ ability to know factsabout public policy issues to how well they are able to connect their ownsubjective policy preferences to the proper political party. Not all issues arepartisan, but the five issue areas discussed in this chapter (health care,energy production, same-sex marriage, food stamps, and drug testingwelfare recipients) do have a clear partisan divide. I leverage this fact inthe PPSF survey to ask respondents both about their personal thoughts oneach of these issues, but also which party they think is best at handling theparticular issue. Failure to connect one’s own views to the correct party isanother instance of democratic citizens failing to live up to the democraticideal. But is this failure random? The answer, of course, is no. As has beendemonstrated in this chapter, the Big Five play an important role indriving factual knowledge. As demonstrated in the next chapter, the BigFive play an equally important role in driving individuals’ ability to makethe most important and basic connections democratic governance requiresof them, that is, between their personal preferences and the party theythink represents them.

SUCCESSES AND FAILURES TO ANSWER FACTUAL QUESTIONS 85

NOTES

1. As a further demonstration that motivated reasoning is the domain of themost politically engaged, on September 16, 2016, the Republican presiden-tial nominee Donald Trump held an event in order to declare whether hepersonally believed the unequivocal fact that President Obama was born inthe USA. Belief in facts is now a subjective policy position to be declared bya presidential candidate.

2. As of June 2016, not one welfare applicant or recipient in the state ofMichigan has tested positive for drugs (Felton 2016).

3. The report is no longer available on the Fox News website, but can be foundon Real Clear Politics (Real Clear Politics. 2013).

4. See the U.S. Department of Agriculture summary statistics (U.S.D.A. 2017).5. Full results tables for each model are available from the author.6. I also estimated ordered logit and multinomial logit models with a three-

point-dependent variable coded 0 if correct, 1 if answered don’t know, and2 if incorrect. While there are some minor shifts in the results, the overallpicture is the same. The results of these other models are available uponrequest.

7. The demographic information for each respondent was provided by SurveySampling International.

8. In order to make all the discussions regarding the substantive meaning ofthese results easily comparable, I always produce these values by movingfrom a score at the 10th percentile of the variable to the 90th percentile ofthe variable and base them on the model that includes all of the controlvariables.


CHAPTER 4

Connecting Attitudes to Party Positions

One of the favorite pastimes of political commentators and late night talkshow hosts is pointing out the general ignorance of the average American.In October 2014, market research company Ipsos MORI published yetanother demonstration of this (Ipsos Mori 2014). Their report finds analarming or amusing (depending on one’s perspective) disconnectbetween perception and reality. For instance, Americans apparently thinkthat nearly one in four teenage girls between the ages of 15 and 19 ispregnant. The real number is just 3 percent. Indeed, of the 14 countriesincluded in the study, the USA did better than just one (Italy) in their“Index of Ignorance.” Ignorance is not limited to political issues, how-ever. It flourishes across a wide range of topics. Thus, failing to live up tothe ideal of a democratic citizen found in the various folk theories ofdemocracy is to be expected. But that does not mean it is not importantto understand what causes variance in how far individuals fall short of thisideal and what that means for a democratic process that relies on citizenparticipation. The effect of this failure is not benign if the befuddledmasses’ opinions are influencing policy. Barber (1992) cautions:

Democracy is not a natural form of association; it is an extraordinary and rarecontrivance of cultivated imagination. Empower the merely ignorant andendow the uneducated with a right to make collective decisions and whatresults is not democracy but, at best, mob rule: the government of privateprejudice and the tyranny of opinion—all those perversions that liberty’s


87

enemies like to pretend (and its friends fear) constitute democracy. For truedemocracy to flourish, however, there must be citizens. (5)

Barber’s words could not have been more prescient. The limitations ofcitizens to understand and engage in politics would be comical, if theirparticipation wasn’t so critically important for democratic decision mak-ing. One need only turn on the nightly news to see stories where thebaseless opinions of individuals are held up as legitimate counterpoints toscientific knowledge (e.g., the importance of measles vaccines). IsaacAsimov probably put it best in his essay “A Cult of Ignorance” (1980).“There is a cult of ignorance in the United States, and there always hasbeen. The strain of anti-intellectualism has been a constant thread windingits way through our political cultural life, nurtured by the false notion thatdemocracy means that ‘my ignorance is just as good as your knowledge.’”Research has shown public opinion does influence policy, particularlywhen the issue is highly salient (Burstein 2003, 2014; Shapiro 2011).The result of this influence in the case of vaccines are numerous states thatallow parents to easily opt out of vaccinating their children, which endan-gers the whole community. This is Barber’s tyranny of ill-informed opi-nion at work.

What is the ideal democratic citizen? I have purposely avoided provid-ing a concrete definition of the ideal democratic citizen here because onceI do that, the focus may likely shift to a debate about the definition ratherthan what I am most concerned about in this book. I am in full agreementwith Achen and Bartels (2016) in their characterization of most conceptsof the ideal democratic citizen as comprising a kind of folk theory thatposits the existence of something that never truly resembles the humanbeings that live on this planet. I would think most of these ideas aboutcitizen competence would agree that, at a minimum, the ideal democraticcitizen has an accurate understanding of political issues and that whateverone’s political preferences, one is able to connect them to the party thatbest represents those preferences.

Dahl (1989) develops a far more erudite definition of the ideal demo-cratic system. He argues, in part, that it includes sufficient and equalopportunities for citizens to contribute to decision making through vot-ing, develop an understanding of public policy, influence the agenda, andexpress their policy preferences. All of this happens through a continuousdialogue between governors and the governed. Of course, Dahl recog-nized this as an ideal type and that all known currently functioning

88 4 CONNECTING ATTITUDES TO PARTY POSITIONS

democracies fall short. This failure is due, at least in part, to the expecta-tions that democratic institutions place on the citizenry. Citizens areexpected to know the best course of action and then be able to judgethe results for a seemingly infinite number of issues. Indeed, the verypeople elected to enact such policy fail to live up to that standard as theyrely heavily on their colleagues and outside interests for information (e.g.,Hall and Deardorff 2006; Kingdon 1989). The unfortunate reality is thatdespite scholars routinely pointing out that the average citizen always fallsshort of the folk theories ideals, our system is still built on their fictitiousbacks. Thus, we are left with a very real political system that needs a well-oiled (i.e., knowledgeable and engaged) citizenry to work, but breaksdown at every turn because it never has enough lubrication.

Up to this point, I have taken two steps in the process of analyzingpersonality’s influence on individuals’ ability to understand politics in acoherent way. In Chapter 2, I showed how high levels of agreeablenessand low levels of openness contribute to individuals failing to make aconnection between the abstract concept of ideology and their own setof policy beliefs. In Chapter 3, I demonstrated that the amount of indivi-duals’ factual knowledge about important public policy areas is affected bypersonality. In this case, high levels of extraversion and low levels ofopenness significantly affect factual knowledge. Now I move on to whatI believe is one of the most fundamentally important elements of thedemocratic citizen. Given individuals’ struggles with abstract notions ofideology and their lack of factual knowledge, can they at least connect thepolicy thoughts and preferences they do have (no matter how flawed theymay be) to the correct party?

Every citizen is free to believe whatever they want to believe, but thatdoes not make everyone’s beliefs equally correct. For one’s preferences toinfluence the political processes correctly, the ideal democratic citizenmust know which party agrees and which party disagrees with thesepreferences. In each of the five issue areas covered in the previous chapter(drug testing welfare recipients, U.S. energy policy, same-sex marriage,health insurance, and welfare spending), the PPSF survey asks two simplequestions of respondents. The first elicits the respondent’s opinion on thesubject and the second asks which party the respondent thinks does thebest job handling the particular issue. As was documented in the lastchapter, each of these issue areas have clear dividing lines between thetwo major political parties. Unfortunately, the hope that individuals willmake the proper connection is often not met. Choosing the wrong party

CONNECTING ATTITUDES TO PARTY POSITIONS 89

will be the third separate demonstration of how the Big Five contribute toindividuals’ understanding of the reality of politics.

YOU’RE DOING IT WRONG

Political scientists have spent a great deal of time documenting low levelsof political knowledge and individuals’ inability to answer factual ques-tions. However, they are often reluctant to label many beliefs or behaviorsas simply being wrong. This may be because in many, perhaps most,situations there can be arguments made that one’s actions or beliefs arenot wrong from a certain point of view. For example, scholars rarely labelthe decision to abstain from voting as wrong. Indeed, there is theoreticalreason to think of the choice to stay home as rational, while it is the voterswho are acting foolishly (Downs 1957; Riker and Ordeshook 1968).

The one significant exception to this is the relatively recent set of worksfocused on voting correctly (Bartels 1996; Dusso 2015; Ha and Lau 2015;Lau and Redlawsk 1997, 2006; Lau et al. 2008; Sokhey and McClurg2012). In these works, scholars label individuals’ vote choice as wrongwhen they fail to choose the candidate that best represents their personalpolicy preferences. The ideal in every one of these cases is that correctvoting is based on policy preferences, and not on something like theheight of the candidates. Here I expand on this idea, but rather thanlooking at an aggregation of numerous policy preferences in relation to asingle vote, I look at individual issue preferences in relation to which partyrespondents believe handles the issue best. For example, if one thinks thatsame-sex marriage should be banned and also thinks Democrats are hand-ling the issue best, that person is wrong.

Within each of the five issue areas covered by the PPSF survey, Iincluded questions designed to tap respondents’ subjective opinionsabout the issue and which party they believe is handling the issue best.Thus, this is an entirely subjective connection. My only requirement is thatrespondents make the proper connection between their preferences andthe party that, currently, best represents those preferences. The exact ques-tion wording of each question can be found in the Technical Appendix.Table 4.1 provides response summary statistics for each question.Respondents appear to have a mix of support for both conservative andliberal policy. On the conservative side, respondents are overwhelmingly infavor of drug testing welfare recipients and a plurality support decreasedspending on food stamps. On the liberal side, majorities favor government


Table 4.1 Summary statistics of PPSF questions on policy opinion and best partyto handle

Supportdrug testingwelfarerecipients

Yes71.7

No18.4

Don’t know9.9

Which partyhandles best

Republican22.9

Democrat18.5

Neither36.6

Don’t know22.0

Support gov.supportpaying forhealthinsurance

Yes61.9

No24.8

Don’t know13.3


Republican17.3

Democrat36.8

Neither33.5

Don’t know12.4

Supportsame-sexmarriage

Yes59.0

No29.9

Don’t know11.2


Republican11.0

Democrat39.0

Neither35.1

Don’t know15.0

Support newsources ofenergy orproductionof existing

New65.5

Existing22.7

Don’t know12.8


Republican20.4

Democrat27.5

Neither35.4

Don’t know16.7

Supportmore or lessspending onfood stamps

More19.9

Less42.6

KeepSame32.9

Don’t know4.6


Republican16.7

Democrat30.3

Neither34.4

Don’t know18.6

Notes: All numbers represent the percent of the total for each question

YOU’RE DOING IT WRONG 91

support for health-care insurance, same-sex marriage, and researching newsources of energy production. Thus, it would appear that, on average, thissample of individuals leans liberal, but are not particularly big fans ofwelfare programs or those who use them.

It is also important to note the significant number of respondents whosimply believe that “Neither” party does a better job handling these issues.In each case, slightly more than a third of respondents think neither partydoes a better job despite the vast differences in each party’s approach tothe subject. These non-committal answers pose a problem for this analysisbecause I cannot simply label these responses as wrong, even if therespondents gave a substantive answer to the subjective policy preferencequestion. This is because there may be legitimate reasons why the respon-dent feels both parties are failing to handle the issue well. For example, arespondent may have answered that they support government programsdesigned to help cover the cost of health insurance. If that same individualthen went on to answer that neither party is “best” at dealing with it, Icould code him or her as wrong because clearly Democrats are closer tothis person’s opinion than Republicans are. However, if this individualfavors a fully government run health system, then one could make theargument that this individual is not wrong since neither party is proactivelytrumpeting that particular solution. I want to keep this analysis as clean aspossible, thus I will only label a choice as wrong when an individualchooses a policy position and then also indicates the party least supportiveof that position is best at handling it, for example, the respondent supportsdrug testing welfare recipients and believes that Democrats are doing thebest job dealing with the issue.

Table 4.2 displays the percentage of respondents who indicated apreference on the issue and then failed to match their preference to thecorrect party. As can be seen, respondents were best at connecting theirviews on same-sex marriage to the correct party, with only 12.7 percentgetting this wrong. Unfortunately, they were far less successful in makingthis connection when thinking about the other issues, where nearly aquarter to a third of respondents failed to do this correctly each time.The table also documents the total number of respondents that were ableto both express a preference and choose a party that handles it best. In allthese cases, less than half of respondents were able to do this. It istroubling that so few individuals have both an opinion and can pick aparty that handles the issue best. These issues are important public policyconcerns that receive political news coverage and are referenced during


campaigns; they are not the fine minutia of a U.S. trade agreement. Yet,less than half of respondents even attempt to live up to the democraticideal, that is, have an opinion and know which party represents thatopinion.

For example, in the case of drug testing welfare recipients, Table 4.1showed that nearly 72 percent of respondents were happy to impose thisdemeaning requirement on those getting benefits even though (as dis-cussed in the previous chapter) they are no more likely to be on drugs thanthe average American. Of these individuals, only 44 percent were willingto choose one of the two parties as the party that handles the issue best,despite the fact that there are vast differences in how the two parties haveapproached this issue. Of those who did choose a party, only 19 percentwere able to correctly choose the Republican Party. Indeed, despite over-whelming popular support for drug testing in this survey, Republicanswould only get 23 percent of the vote based on this issue. Of course, thissample is not representative of the USA as a whole. One can hope that theaverage American is much better at making this connection than thissample indicates, but I doubt it.

Does it matter if people are making these kinds of mistakes? One of themore common responses to this question is simply “no.” The argument isthat the American electorate is made up of millions of people and that themistakes they make cancel each other out. This argument is based onCondorcet’s jury theorem that argues that the collective choice of agroup is better than the choice any one individual would make on his orher own (Congleton 2007; Ladha 1992). In other words, the wisdom of

Table 4.2 Incorrect connection between policy preferences and the party thathandles the issue best

Issue Percent making incorrectconnection between preferenceand party

Total

Drug testing welfare recipients 32.5 907Government health insurance subsidies 20.5 1,165Same-sex marriage 12.7 1,093U.S. energy production 29.5 1,030Spending on food stamps (SNAP) 25.8 709

Notes: The percent of respondents who indicated a policy preference on the issue, but did not match theirpreference to the correct party.


collective choice mitigates the ignorance and mistakes that individualsmake. Is this true? Not according to the works of Bartels (1996) andLau and Redlawsk (2006; Lau et al. 2008). These scholars have docu-mented incorrect voting large enough, and one sided enough, to swingpresidential election outcomes. This suggests the errors people make arenot random, but systematic.

The data utilized here finds the same result. In each of the five issueareas, the policy preference of the majority of respondents is found inTable 4.1. It also shows that the party that actually supports themajority’s policy preference never receives the majority’s endorsement.At best, it receives a plurality of the “votes” in the cases of healthinsurance and same-sex marriage. This is because most people have noidea what the parties stand for and answer either “don’t know” or“neither” when asked which party is handling the issue best. But look-ing at the raw numbers is not entirely satisfying as a response becausewhat the Condorcet based argument is saying is that of those who doshow up to vote, the mistakes that they make will cancel each other outand the outcome will be the same. When it comes to the surveyrespondents, there are a large number who have functionally chosennot to show up to “vote” by failing to either express an opinion orrefusing to choose a party as the one that is best at handling the issue.Thus, the raw data is not really testing the argument that mistakescancel each other out.

A better test of the assumption that the mistakes of the electoratecancel each other out is to look only at those who were able to bothexpress an opinion and choose a party. The supposition is that thosewho make mistakes in this case should be equally supportive of bothpolicy positions. For example, when looking at supporting or not sup-porting drug testing welfare recipients, the respondents who choose thewrong party should be equally likely to be for or against the policy. Theresult is that the respondent who professes to support drug testing,while incorrectly choosing the Democratic Party, is cancelled out bythe respondent who does not support the policy, but incorrectly choos-ing the Republican Party. Thus, one incorrect “vote” for the Democratsis cancelled out by one incorrect “vote” for the Republicans. If thosewho choose the wrong party fall heavily on one side of the issue, thenthat side will have a harder time getting the correct party “elected.” Itmay even turn a majority’s policy opinion into majority support for aparty that would do the opposite.


Table 4.3 presents the policy preferences of both the respondents whomade the correct connection to a party and those who did not. As can beseen, in nearly all the cases, the incorrect respondents did not balance eachother out, but overwhelmingly fall on one side of the issue. For instance,of those respondents who could not connect their preference on drugtesting welfare recipients to the correct party, 86.8 percent said theysupport drug testing, which means that these same people also thinkDemocrats are the party that handles the issue best. Thus, when mistakeswere made on this issue, they were overwhelmingly mistakes made bypeople favoring drug testing and choosing the Democratic Party. Thiscould not be more wrong. Only when looking at the issue of same-sexmarriage does there appear to be behavior that is consistent with the“mistakes cancel each other out” hypothesis. In the case of drug testing,health insurance, and energy production, super majorities favoring oneside of the issue see these majorities whittled away as large portions of theirfellow supporters fail to connect this preference to the correct party.Finally, in the case of food stamp spending, there is majority support forspending less on food stamps that actually produces a majority believingthat Democrats are best at handling the issue. If this was an election, notonly would the mistaken respondents not cancel each other out, they

Table 4.3 Policy preferences of respondents who were correct or incorrect whenpicking party that handles the issue best

Correct party Incorrect party

Support drug testing welfarerecipients

Yes77.5

No22.6

Yes86.8

No13.2

Support gov. support payingfor health insurance

Yes77.9

No22.1

Yes66.5

No33.5

Support same-sex marriage Yes82.1

No19.9

Yes51.8

No48.2

Support new sources of energyor production of existing

New69.2

Existing30.9

New68.8

Existing31.3

Support more or less spendingon food stamps

More48.3

Less51.7

More17.5

Less82.5

Notes: All numbers represent the percent of the total for each question


would actually help put the wrong party in charge of the issue. All told,there is little evidence in the real world for the claim that the mistakes thatindividuals make cancel out in the aggregate. Thus, understanding thecauses of these mistakes is critically important, with individuals’ personal-ities playing a big role in driving this process.

Personality and the Preference Party Connection

At this point, the expected effect of the Big Five personality traits hasbecome clear. Openness to experience should have a significant negativeeffect on the probability of incorrectly connecting one’s personal policypreferences to the correct party. Increased emotional stability should alsohave a negative effect in this case. Given extraversion’s performance whenpredicting the ability to answer factual policy questions, it is an easy jumpto hypothesize that it will have a similar effect on one’s knowledge of partyplatforms. Finally, when it comes to agreeableness and conscientiousness,it is less clear. On the one hand, agreeableness was a significant predictorof conflict between one’s symbolic and operational ideologies inChapter 2. On the other hand, it did not appear to have a consistent effecton one’s factual knowledge. Since the ability to connect one’s preferencesto the correct party relies heavily on knowledge, it seems safe to predictthat agreeableness will perform closer to how it did in Chapter 3 than inChapter 2. Conscientiousness has failed to demonstrate a consistent effectthroughout this analysis. However, in Chapter 3 it did have a significantnegative effect on the probability of not knowing which party supportsdrug testing welfare recipients. Obviously, knowledge of which partysupports drug testing is directly applicable to knowing which party sup-ports one’s own policy preferences. Thus, I would expect conscientious-ness to have a significant effect in this one issue area.

Table 4.4 presents a summary of results after estimating models pre-dicting the ability of respondents to connect their personal policy prefer-ences to the correct party. A quick technical note is in order here. Asshown in Table 4.2, less than half of respondents were able to givesubstantive responses to the policy questions (i.e., they did not say“don’t know”) and to choose a party that best represents the issue (i.e.,they did not say “neither” or “don’t know”). The problem that ariseswhen so many respondents essentially “opt out” of this part of the studybecause of their responses, is that what is left is likely not due to a random


process. In other words, there are likely systematic differences betweenthose who remain in the analysis and those who are now excluded.

In order to account for this, I chose to estimate heckman probit modelsthat allow a two-stage process in the generation of these data. In the firststage, there are variables that make it more or less likely that one is goingto be able to give both a substantive opinion on the issue and then alsochoose a party that best represents the issue. These are likely variables thattap political interest and knowledge, that is, the more interest and knowl-edge an individual has, the more likely they are to be in the sample. In thesecond stage, the model predicts whether the respondents are able tomake the correct connection. Table 4.4 presents results after estimatingheckman probit models for each issue area. For the first stage I includedevery variable I had available that tapped political interest and knowledge,including the variable asking specifically about how important the

Table 4.4 Summary of models predicting the Big Five’s effect on connectingpreferences to the correct political party

Drugtestwelfare

Health-caresubsidies

Same-sexmarriage

Energyproduc.

Foodstampspending

Count #wrong τ

ExtraversionSelection model * + * + * + * + * +Logit w/controls * + * + * + * +

AgreeablenessSelection modelLogit w/controls

ConscientiousnessSelection model * * Logit w/controls *

Emotional stabilitySelection modelLogit w/controls

OpennessSelection model * + * * * * Logit w/controls * *

Notes: * = trait was a significant predictor in the model; + = a positive relationship; = a negativerelationshipτA tobit model was estimated to predict the number of incorrect responses


respondent felt the specific issue was. The second stage contains the Big Fivepersonality variables. I performed likelihood ratio (LR) tests asking if there isa difference between the two-stage heckman model and the single-stagestandard model I might have estimated. In all but the drug testing welfarerecipients’ case, there is a significant difference. Thus, the heckman model ispreferred.However, once I added in all of the control variables to the secondstage of the heckmanmodel, the LR test failed to find a significant differencebetween it and the standard logit model. Therefore, when utilizing all of thecontrol variables, I return to the simpler logit models.

Since none of the Big Five personality variables appear in both theselection and outcome stages of the model, the coefficients can be inter-preted in the same manner as a standard probit model and thereforeinterpretation of the summary table here is no different than it was in theprevious chapter. Looking at the top of Table 4.4, the story is similar to theone from the previous chapter. Both extraversion and openness to experi-ence appear to have the most consistent effect. Extraversion is significantwith a positive sign in 9 of the 12 models. In other words, in every one ofthese models, the higher respondents’ score on the extraversion scale, themore likely they are to fail to connect their personal policy preferences tothe correct party. Substantively, when looking at the selection models, ineach case going from an extraversion score of 2 to 10 increases the prob-ability of being wrong by about 5 to 6 percent.1 Openness to experience isalso significant with the expected sign in all but the food stamp equation,when looking at the selection models, but only twice when looking at themodels with controls (including the tobit models estimating the volume ofincorrect responses). Substantively, as openness increases from 5 to 11, theprobability of getting this connection wrong goes down by about 4 percentin the case of same-sex marriage, 6 percent in the case of health insurance,and 7.6 percent in the case of energy production.

Continuing to look at Table 4.4, the results from the model predict-ing the ability to connect personal opinions about drug testing welfarerecipients to the correct party requires more analysis because it bucksthe trend in several ways. First, the LR test indicates that the selectionmodel is no different than a standard probit model. Second, extraver-sion fails to reach significance. Third, while openness is significant, ithas the wrong sign. In other words, an increase in openness to experi-ence in this case corresponds to an increase in the probability of gettingthis wrong. This is the first time that openness has had a significant positiveeffect on the probability of being wrong, suggesting that something


different is going on with this issue and that difference appears to be drivenby party identification.

As Table 4.1 documents, the drug testing welfare recipients questionproduced the most lopsided results of any substantive question, that is,nearly 72 percent of respondents said that welfare recipients should betested. In addition, Table 4.2 indicates that this question also proved themost difficult for respondents to connect their personal policy preferencesto the correct party, with nearly 33 percent of respondents failing to do so.Why is this the case? The answer is driven by Democratic identifiers. Nearly46 percent of the respondents that said that welfare recipients should betested identified as Democrats. Of those, 26 percent said that Democratsare handling the issue best. This is compared to only 8 percent ofRepublican identifiers who had answered positively to testing. Republicanidentifiers correctly identified Republicans as being best 87 percent ofthe time compared to just 62 percent of Democrats. Not shown here,I re-estimated the model with the addition of partisanship in the outcomestage and found that it is highly significant and substantively indicates thatmoving from being a Republican identifier to a Democratic identifierincreases the probability of getting this wrong by 37 percent. Addingpartisanship also affects openness, which is no longer significant. Thus, itwould appear that the significant positive effect for openness observed inTable 4.4 is an artifact of omitting the partisanship variable. Importantly,this omission does not affect any of the other personality traits.

Finally, in the drug testing model conscientiousness reaches significancewith the predicted negative sign. This result is consistent with the finding inthe previous chapter regarding factual knowledge of which party is more infavor of drug testing welfare recipients. I am tempted to write somethingabout how conscientiousness makes good dutiful citizens who are morelikely to know party platforms, but that statement is simply not supportedby any of the other results. It is also not a party identification story, since theaddition of party identification to the model has no effect on conscientious-ness’ result. Substantively, moving from a score of 6 to 12 on the conscien-tiousness scale in the selection model decreases the probability of failing tomake the proper connection between personal preferences and the party thatbest supports them by 16.4 percent. This is one of the single largest effectsfor any of the Big Five traits that I have found in this analysis. The size of thiseffect is 9.3 percent when looking at the logit model with controls.

Yet, conscientiousness fails to reach significance in any of the othermodels. This suggests that there is something different about this issue


than any of the others. Interestingly, this is not a welfare policy concerneither, since I ask about welfare spending and conscientiousness doesnot have the same effect. It could be the case that using drugs isparticularly galling to highly conscientious individuals, thus they aremuch more aware of this issue. But recall from the previous chapter,those high in conscientiousness were no more or less likely to knowthat welfare recipients use drugs at the same rate as non-welfare reci-pients. Thus, this is not simply a story about the issue of drug use ingeneral. It is about knowing the party platforms on this particular issuemore than any of the others. Unfortunately, I do not have an answer asto why this is the case. Obviously, the study of personality in relation topolitical knowledge and attitudes is still in its infancy, with only themost generic connections having been explored to this point. Thisfinding indicates that there is a far more interesting and nuancedstory to be told.

When looking at the results of the logit models with all the controlvariables, openness to experience reaches significance in the case of same-sex marriage, while falling just short with the issue of spending on foodstamp programs. This is a much different result than was observed in theselection models and is a great example of why it is always important topresent results with varying model specifications. Throughout this book Ihave utilized just about every potential control variable I can think of. Thisincludes the variables tapping the specific importance of the issue to theindividual and the ambiguity of the issue in the respondent’s mind, whichare often overlooked. The effect of extraversion has stood up regardless ofthe specification, while in this case openness falters a bit with the additionof the control variables.

Substantively, the size of extraversion’s effect is fairly consistent acrosseach of the logit models. In the case of same-sex marriage, moving from ascore of 2 to 10 on the extraversion scale is associated with an 8.6 percentincrease in the probability of making the wrong connection between partyand preference. The same sized move in the model predicting health careresults in a 7.9 percent increase, while in the food stamp model the changeis 8.2 percent. In comparison, the size of extraversion’s effect is nearly thesame as that of political knowledge. That is, a move from the 10th to the90th percentile in political knowledge decreases the probability of gettingthis wrong by 9.3 percent. Openness is the only other of the Big Five toreach significance in this model. Its substantive effect is a 5.3 percentdecrease in the probability of getting this wrong. Interestingly, as a


comparison religious fundamentalism is associated with an increase in theprobability of getting this connection wrong in every model. In the same-sex marriage model, moving from a non-believer to a fundamentalistbeliever sees a 12.6 percent increase. This is similar in size to the 13percent increase associated with a move from identifying as a Republicanto Democrat.

When it comes to the issue of health-care subsidies, the 7.9 percentpredicted increase in the probability of getting this wrong is somewhatsmaller in magnitude than that of political knowledge, which is predictedto decrease this probability by about 11 percent. However, it is similar insize to the effect of religious beliefs, with a move from a non-believer to afundamentalist associated with an 8 percent increase in the probability ofgetting it wrong. The importance of the issue of health-care insurance tothe respondents is also a significant contributor to the likelihood of gettingthis wrong. Moving from feeling it is “slightly important” to “extremelyimportant” results in a 12 percent decrease in the probability of beingwrong. All of these are dwarfed by the effect of partisanship, where movingfrom identifying as a Democrat to Republican is predicted to increase thisprobability by nearly 31 percent.

The issue of spending on food stamps sees extraversion increase theprobability of incorrectly connecting one’s preferences to a political partyby 8.2 percent. In this case, political knowledge has a much larger effecton this probability. Moving from a score of 1 to 7 on this scale decreasesthe likelihood of getting this wrong by nearly 20 percent. Religiousfundamentalists are 10 percent more likely to get this wrong than non-believers, while moving from thinking that the issue is only “slightlyimportant” to “extremely important” decreases this probability by morethan 18 percent. Ultimately, the probability of an extraverted individualwho is low in political knowledge, religiously fundamentalist, and finds theissue of food stamps to only be slightly important is 58 percent. This is instark contrast to the low likelihood of just 4.5 percent of an introvertednon-believer, who is high in political knowledge and finds the issueextremely important getting it wrong.

Having looked at the effect of the Big Five in each of these issue areasindependently, I now consider their effect on the volume of incorrectpreference-party connections. Table 4.4 also contains results after estimat-ing models predicting the volume of incorrect connections between one’spreferences and the party that best represents it. The first point to noteabout these new models is the small number of cases. Recall, I only count


respondents as being correct or incorrect if they gave a substantiveresponse to the subjective policy question and also chose one of the twoparties as best at handling the issue. Thus, more than half of respondentswere lost because they did not do one or both of these two things. In thiscase, I am counting the number of times (out of 5) where respondentsmake the wrong connection. There were only 260 respondents whofulfilled the criteria in all five issue areas. As a result, only a small portionof the overall sample remains. Therefore, these results as more suggestivethan conclusive. Additionally, for those keeping score on the technicalmodel choices in each of these tables, here I ran tobit models with andwithout controls. I also estimated heckman selection models in each case,but the LR tests indicated there was not a significant difference betweenthe results of the selection models and more traditional ones.

As can be seen in Table 4.4, the results of the models with and withoutcontrols find extraversion and openness once again reaching significancewith the expected signs. When it comes to significance, the differencebetween the two is that conscientiousness is only significant in themodel without control variables. Substantively, the effect of the Big Fivevariables is quite large. Moving from a score of 2 to 10 on the extraversionscale increased the predicted number wrong by 0.49, while moving from ascore of 5 to 11 on the openness scale decreases the predicted numberwrong by 0.47. In comparison, moving from a score of 1 to 7 on thepolitical knowledge scale is associated with a predicted drop in the numberwrong by 0.74. Moving from a non-believer to a fundamentalist believer ispredicted to increase this score by 0.47. Ideology also reaches significanceand sees a move from a consistent liberal to a consistent conservativeassociated with an increase in one’s predicted score by 0.54. Overall,holding every variable at their mean, the predicted number wrong wouldbe 1.37. However, an individual who is extraverted, not open to experi-ence, low in political knowledge, holds fundamentalist religious beliefs,while being a consistently conservative Republican identifier would bepredicted to score about 2.37 on this scale. In comparison, an introvert,who is open to experience, high in political knowledge, a non-believer,and a consistently liberal Democratic identifier, would be predicted toscore just 0.33 on this scale.

In the end, the important takeaway point with regard to the effect ofthe Big Five on the volume of incorrect preference-party connections isthat it is as large as that of political knowledge and partisanship. This pointcannot be stressed enough. More than half a century of work in political


science has documented the important roles of knowledge and partisan-ship in virtually every conceivable area of mass political behavior.Scholarship is at the point where these two concepts are well understoodand fully accepted as key contributors to behavior. Indeed, models that failto account for them would be pilloried as incomplete, at best, and down-right foolish, at worst. Thus, the numerous findings here that the Big Five,in combination, have important effects similar in magnitude to those ofwell-established variables, is significant. Yet, the inclusion of personalityvariables, or any psychological variables for that matter, in models predict-ing political behavior is still the exception rather than the rule.

THE IDEAL DEMOCRATIC CITIZEN?I began this chapter with a short discussion of what it means to be the idealdemocratic citizen. While I did not wish to wade too deeply into thephilosophical debate surrounding the minutia of a specific definition ofthis concept, I do argue that, at a minimum, the structure of our politicalsystem stands on a basic assumption that citizens formulate policy opi-nions and are capable of knowing which political party best representsthose opinions. This minimal suggestion is so weak that I do not evenrequire that the ideal democratic citizen formulate coherent policy opi-nions, just that they know who supports and who opposes them.Unfortunately, most of the respondents to the survey were not even will-ing or able to provide both a policy preference and a belief about whichparty is handling the issue best. When they did provide both, they madethe wrong connection nearly a quarter to a third of the time in most cases,with one’s personality proving to be an important contributor to whypeople make this mistake.

Ultimately, the overall story has been consistent across that last twochapters. Extraversion tends to increase the probability of getting thingswrong, while openness tends to decrease it. Of course, this is not true inevery case, but it has been true in both attempts to measure the volume ofincorrect understandings found in Tables 4.4 in this chapter and 3.3 in theprevious one. Thus, in combination, the last two chapters have providedsolid evidence that these two personality traits are particularly importantcontributors to citizens’ proper understanding of public policy and thepolitical process. The effect of openness is not too surprising given itsorigins and connection to intelligence. However, the performance ofextraversion is surprising, especially because it is typically viewed as a

THE IDEAL DEMOCRATIC CITIZEN? 103

positive quality, which is why the trait is labeled extraversion and notintroversion. These results clearly paint this personality trait in a badlight. Conscientiousness is the only other trait to play a role in this chapter,much like in Chapter 3 where emotional stability joined extraversion andopenness. However, in this case conscientiousness does not appear to haveas broad an effect as was observed with emotional stability.

Finally, I have documented the important substantive effect of person-ality in relation to other variables like political knowledge and partisanshipthroughout this analysis. I want to conclude this chapter with a look atTable 4.5, which summarizes the joint effect of the Big Five personalitytraits compared to the size of the effect of these other well-known politicalvariables. This table is based on the models in Table 4.4. In each case, thepercent change represents how a move in the variable (with the Big Fiveconsidered as a single variable) contributes to an increase in the probabilityof getting the connection wrong. This was accomplished by moving onlythe variable from its 10th to its 90th percentile score or 90th to 10thpercentile score (depending on whether the other predicted relationshipbetween the variable and the probability of getting the connection wrongis positive or negative), while holding all other variables at their mean. Ineach case, the variable was included if it reached the 0.10 level ofsignificance.

Table 4.5 explains what I have discussed at length in this chapter. Theeffect of the Big Five varies from one issue to the next, which is also true ofall the other variables. The Big Five have their largest effect in the twoareas dealing with welfare. The issues of health care and same-sex marriagesee slightly smaller effects, while the issue of energy policy is considerablysmaller. Overall, the Big Five behave in the same manner as all the other,much more common, political variables. For instance, the lack of politicalknowledge is a consistent predictor of failing to make this connection.However, the magnitude of its effect varies substantially from a high ofnearly 26 percent (drug testing welfare recipients) to a low of just 9 percent(same-sex marriage). Partisanship appears to be the single largest predictorof getting the connection between preferences and party wrong, exceptwhen it comes to spending of food stamps, or the case of same-sexmarriage where the size of its effect is little different than that of the BigFive or religious beliefs.

The takeaway point from Table 4.5 and this chapter as a whole is thatwhen it comes to the failures and the successes individuals have in living up tothe democratic ideal, the effects of the unconscious mind in the form of one’s


Table 4.5 Substantive effect of the big five personality traits compared tocommonly utilized political variables

Marginal effect on the probabilityof incorrect preference-partyconnection

Drug testing welfare recipientsBig Five (Cons. , Em. St. +) 16.4Political knowledge ( ) 25.5Partisanship (Dem. +) 28.2Ideology ( ) 15.0Religious (+) 9.5

Health-care subsidiesBig Five (Extraversion +, Cons. ) 12.9Political knowledge ( ) 11.1Partisanship (Dem. ) 30.8Ideology ( ) 5.7Religious (+) 8.0

Same-sex marriageBig Five (Extraversion +, Openness ) 13.9Political knowledge ( ) 9.3Partisanship (Dem. ) 13.1IdeologyReligious (+) 12.6

U.S. Energy PolicyBig Five (Extraversion +) 6.8Political knowledge ( ) 19.7Partisanship (Dem. ) 30.2Ideology ( ) 6.0Religious (+) 8.2

Food stamp spendingBig Five (Extraversion +, Openness ) 15.3Political knowledge ( ) 20.3Partisanship (Dem. +) 7.1Ideology ( ) 10.5Religious (+) 9.7

Notes: Values represent the percent change in the probability of getting the connection wrong. Eachvariable is followed in parentheses by the specific variable (if necessary) and sign. Values of each variablewere moved from their 10th to their 90th or 90th to their 10th percentile score depending on the signof the relationship, for example, political knowledge always moved from 90th to 10th percentilebecause a decrease in knowledge always accompanies an increase in the probability of getting theconnection wrong.

THE IDEAL DEMOCRATIC CITIZEN? 105

personality cannot be overlooked. Yes, the standard predictors matter too.Scholars and well intentioned individuals advocating for a more active andengaged electorate are right, but their vision is limited. The fact that some-thing outside of our control, like our personalities, plays such a large role inour understanding of politics demands that reformers begin to think of newsolutions to these problems. More education does not appear to have aneffect, as evidenced by the fact that the measure of educational attainmentused in these models rarely ever reaches significance. I would surmise that atleast part of the reason this is the case is because the specific content of one’seducational path is voluntary. Therefore, it is easy to achieve academic successwithout coming into contact with the political world. This means that thepotential value of education is blunted by a system that gives everyone equalopportunity to participate without taking seriously the job of creating goodcitizens, that is, the very citizens Barber (1992) argues democracy dependson. But it is also questionable how much civics education would be necessaryto have an effect or what the magnitude of that effect would be. After all,these models all indicate two things when it comes to political knowledge.First, that the more of it one has, the better one is at getting things right.Second, that even when people are highly politically knowledgeable, variancein their personality will affect their ability to be correct.

If one cannot change people’s personalities, how does one mitigatepersonality’s potential harmful effects? Unfortunately, I do not have a simpleanswer to that question. At this point, it is simply a matter of changing thediscourse. In Chapter 6 I outline my ideas for changes to our political systemthat account for the limitations of real human beings. But for now, it isimportant to note that the myopic focus on low political knowledge andpartisanship as the twin scourges of quality democratic discourse is limiting.The problems of low levels of civic engagement and understanding are farmore complex than that. Solutions begin with a recognition that thedemands of democratic citizenship are far greater than virtually anyonewill ever be capable of handling. This is the case, not because of a lack ofeducation and not because the media has failed to inform “the people,” butbecause of factors which are outside of our conscious control. Our person-alities make people different, interesting, and wonderful. But our personal-ities also alter our individual abilities to understand and engage in politics.Once we recognize that, we can begin to contemplate new methods ofincorporating individuals’ opinions into the political process that do notassume we are all equally capable of casting a correct vote.


NOTE

1. As was the case in the previous chapter, unless otherwise stated, all thepredicted marginal effects presented in this chapter are based on a movefrom the 10th to the 90th percentile score in that variable with all othervariables held at their mean.

NOTE 107

CHAPTER 5

Hidden State and the Punitive Public

In May 2015, The Charlotte Observer reported on the story of a man fromSouth Carolina who has had several mini-strokes in his eyes and is nowlosing his sight (Helms 2015). He needs various surgeries to correct theproblem. The unfortunate part is that he does not have health insurance.He has worked as a handyman and has prided himself on paying hismedical bills as necessary throughout his life. But now he has had tostop working due to his loss of sight and does not have the money topay for the corrective surgery, and makes too little to qualify for federalsubsidies through the ACA (although he had already missed the signupdeadline). Fortunately, individuals like him, whose income is less than thefederal poverty line, are handled through Medicaid programs.Unfortunately for him, the state of South Carolina has declined theopportunity to expand Medicaid in the state, so he is out of luck.

Who does his wife blame for this situation? President Obama andCongressional Democrats saying, “(My Husband) should be at the frontof the line because he doesn’t work and because he has medicalissues . . .We call it the Not Fair Health Care Act” (Helms 2015). Ofcourse, what the story does not mention is that this man and his familyhave benefited from the submerged welfare state their entire lives. He mayvery well have prided himself on paying his own medical bills, but if theyever reached more than 10 percent of his adjusted gross income, he wouldqualify for a tax deduction. The story also mentions the $300,000 homehe lives in. It seems likely that this family has been receiving tax subsidies


109

to pay for that mortgage for years. The happy ending to this story comesfrom the charity of strangers. After setting up a page on the websiteGoFundMe.com, he received more than $12,000 in donations in orderto pay for the eye surgery he needed.

While it might be tempting to dismiss this story as yet another exampleof the failure of an individual to know what is in their own best interest, Ifind it to be more instructive. The story documents an all too commonphenomenon. This is the failure of citizens to understand their relation-ship with the federal, state, and local governments that make society work.He has a belief that he is self-sufficient and has not benefited fromgovernment welfare programs his entire life. This has led him to makemonstrously bad decisions regarding his own health. These decisionswould likely have cost him his sight had he not received the publicityfrom the story in the newspaper. He was lucky.

Another story of an individual who does not want to be seen asdependent on government services comes from the Washington Post(Sun and Chokshi 2015). This time it is the story of a woman in Texaswho had always opposed the ACA as another wasteful government hand-out. But, then after a divorce saw her lose her health insurance and findinga new job that does not provide such benefits, she was diagnosed with arare medical condition. She could not afford health insurance on her ownand therefore turned to health plans provided through the ACA’s federalmarketplace. Fortunately, she qualified for a subsidy and now has anaffordable monthly payment of $89. As one would expect, she does notwant to lose this and points out, “I’m not trying to go to Cancun or carrya Michael Kors bag. I drive a 2009Mazda, and I’m just trying to make it inmy little apartment and not be on government assistance” (Sun andChokshi 2015).

Of course, the upshot of this woman’s story is that she reluctantlyturned to a medical plan provided through a program she did not likebecause she thought of it as a government handout. Now that she isbenefiting from it she is scared of losing it and asks, “If they’re [Texas’political leaders] not going to participate in Obamacare and I’m not goingto have these financial benefits, which will force me to pay $220 a monthfor coverage, do you know if Greg Abbott, our governor, has any plan tooffer something comparable? I understand and support his efforts to putWashington back in its place. I just don’t want that to come at the cost ofhard-working Texans and their ability to maintain medical coverage” (Sunand Chokshi 2015). What she does not understand in her attempts to

110 5 HIDDEN STATE AND THE PUNITIVE PUBLIC

http://GoFundMe.com

differentiate herself and the benefits she is receiving through the ACA andthose “lazy” people on welfare is that she has always received governmenthelp paying for her insurance, as her husband did and continues to, andthe politicians in Austin do as well.

While it is undeniable that the man in South Carolina has made baddecisions for himself and the woman in Texas feels significant anxietyreceiving help from a visible government program, the larger concern fora democratic society moves well beyond any one individual’s health-carechoices. What happens when whole segments of citizens make these samemistakes; when support for policies is based on the false premise that somepeople are takers, while others have to work for everything they have? InThe Submerged State Suzanne Mettler (2011) argues that “for citizens topossess the political power to rule themselves, they need to understandtheir own relationship to government and the role it does and does notplay in their lives. As early as the nineteenth century, American publicofficials chose to direct public funds through private channels, and thatpattern has become increasingly common in our own time . . .Yet sucharrangements leave citizens oblivious to how much government itself isresponsible for the vibrancy of whole sectors of the market and for pro-tecting Americans’ well-being” (123). The problem, therefore, is thatthere is bias built into the system by which government services areprovided to citizens that systematically hides the benefits that somereceive, while amplifying the visibility of others. The effect of this bias isoften to benefit the majority, with a corresponding political discourse thatdoes not recognize the very programs that help the majority for what theyare, that is, welfare. At the same time, the term welfare is limited to certaintypes of programs, like food stamps, where the recipients are vilified as lazyand undeserving (Stanley 2015).1

Scholarly research in the USA suggests that the perceptions of govern-ment welfare are biased against minority groups. What constitutes govern-ment welfare and what does not is not value neutral. The long history ofresearch on welfare policy and how it is understood by citizens hasdocumented a strong racially charged bias in public opinion (e.g., Gilens1999; Mendelberg 2001; Schram et al. 2003). As a result, the damagecaused by the incomplete understanding of what government does foreach citizen, which Mettler points to, is amplified by the long dreadfulracial history of the United States. This ignorance and racism produce acocktail of failure that I argue is as large a failing on the part of democraticcitizens as any I have pointed to so far in this book.Majorities benefit through

HIDDEN STATE AND THE PUNITIVE PUBLIC 111

hidden welfare state policies, which in the USA tends to be middle- andupper-class white citizens, at the expense of minority populations.

After having explored the Big Five’s effect on individuals’ capacity tounderstand abstract notions like ideology (Chapter 2), basic knowledge ofpolicy facts (Chapter 3), and ability to connect their own policy prefer-ences to the correct party (Chapter 4), I now return to the importance ofthe submerged state and how variance in individuals’ personalities altertheir susceptibility to holding hypocritical policy positions where theypersonally benefit from submerged welfare policies (e.g., tax expendi-tures), while advocating for the reduction of visible social spending (e.g.,budgetary appropriations). I do this, first, through an exploration ofhypocrisy in the areas of welfare, food stamps, and healthcare insurance.This will take the form of multiple models that test the effect of the BigFive in these areas individually and in terms of the volume of hypocrisyrespondents exhibit. Second, I examine the relationship between person-ality and ignorance of submerged welfare state policy in relation to healthinsurance policy hypocrisy. The interesting result is that personality (par-ticularly conscientiousness) drives both ignorance and policy hypocrisy inthis area.

TAX EXPENDITURES AND THE HIDDEN WELFARE STATEResearch on governmental spending is dominated by a focus on directbudgetary appropriations (e.g., Gilens 1999; Jacoby 1994, 2000; Stimson2004; Wlezien 1995). However, recent work has begun to focus on all thevarious methods governments use to deliver support to citizens. The mostcommonly overlooked service delivery method is through the use of taxexpenditures (Collado and Iturbe-Ormaetxe 2010; Faricy 2011; Hacker2002; Howard 1997, 2007; Mettler 2008). Of course, these two types ofspending are functionally equivalent. They both reduce the U.S. treasuryand provide benefits to selected groups of citizens. Indeed, theCongressional Budget and Impoundment Control Act of 1974 (Pub. L.No. 93–344, sec. 3(3)) defines a tax expenditure as “revenue lossesattributable to provisions of the Federal tax laws which allow a specialexclusion, exemption, or deduction from gross income or which provide aspecial credit, a preferential rate of tax, or deferral of tax liability.”

Research on tax expenditures has found an interesting combination ofbeliefs among the public. First, individuals tend to favor governmentalprograms that are implemented through tax expenditures to those that are


distributed through direct budgetary appropriations (Faricy and Ellis2014). Second, those who benefit from these government programs arelargely ignorant of these benefits (Hacker 2002; Mettler 2011). Third,many such programs are regressive in their implementation, that is, thehigher on the economic ladder people are, the greater they benefit fromthese types of programs (Burman et al. 2008).

For example, the USA has determined that it is important to helpindividuals purchase homes. In order to facilitate this behavior, the federaltax code includes a provision for deducting the interest paid on a homemortgage from one’s taxable income. The mortgage interest deductioncan be used on homes totaling up to 1 million dollars. Depending on theexact terms of the mortgage, an $800,000 mortgage on a million dollarhome with a mortgage rate of 5 percent will run about $25,000 annually2

in interest over the course of 30 years. Assuming that this individual falls inthe highest income tax bracket of 39.6 percent with an income of$500,000, this governmental assistance program will reduce the U.S.treasury by nearly $8,600 in order to help with the purchase of thishome.3 By contrast, a married couple buying a $150,000 home afterputting 20 percent down will take out a $120,000, 30 year mortgage at5 percent, which will run about $3,700 in interest per year. If this couple isin the 15 percent tax bracket (income between $18,650 and $75,900 in2017), this mortgage will reduce the U.S. treasury by $555. Thus, thisprogram helps with the purchase of a million dollar home to the tune ofabout 12 thousand dollars a year, while kicking in less than 600 dollars ayear to help the lower income family. Of course, this is assuming themarried couple itemize their deductions on their tax returns. Anotherway this program is regressive in its effect is that it only helps those whoitemize their deductions. The lower on the economic ladder one is (evenamong those who own a home), the less likely one is to do this. Mettlershows that in 2004, 69 percent of the benefit of this tax deduction wasclaimed by the top 15 percent of the income distribution.

Not only is the homemortgage interest deduction regressive in its effect, italso disproportionally benefits whites. According to recent research based onthe long running U.S. Census Bureau’s Survey of Income and ProgramParticipation, as of 2011, 73 percent of white households own their homecompared to just 45 percent of black and 47 percent of Latino households(Sullivan et al. 2015). Ultimately, this disparity is one of the primary drivers ofthe wealth gap found in the United States between whites and minorities.Thus,while a tax expenditure like themortgage interest deduction is ostensibly

TAX EXPENDITURES AND THE HIDDEN WELFARE STATE 113

neutral, when it comes to race and ethnicity, it is not neutral in its effect.Moreover, this bias is further baked into our tax code by the choice to not taximputed rent in theUSA, yet it is counted as part ofGDPby theU.S.BureauofEconomic Analysis (U.S. Bureau of Economic Analysis 2017). What isimputed rent? Let’s say that a married couple rents a home for $2000 amonth. Their landlord would have this $2000 a month ($24,000 a year)taxed as income. If the landlord then sells the home to the couple rentingthe house, the landlord no longer pays tax on the rental income.Of course, thenew homeowners do not pay tax on the rental value of the home that they arepaying themselves rather than the landlord either. They pay property taxeswhere the interest paid is tax deductible. In places like the Netherlands,property owners pay tax on the rental income potential of the home evenwhen they are “renting” the property to themselves. This is imputed rent, thatis, rent that is being paid by homeowners to themselves, rather than being paidby tenants to the homeowner. The income doesn’t disappear, it just comesfrom a different source. Indeed, imputed rent accounts for a little over a trilliondollars of U.S. GDP. Of course, if instead of investing in a home, this couplehad invested that money in something like a U.S. Treasury bond, they wouldbe taxed.

The result of providing government services through tax expenditures is adevastating combination where individuals get to believe they are self-suffi-cient, while demonizing those who receive governmental benefits that aremore visible to the public. For example, in early 2015, Kansas’ state legislaturepassed legislation outlining a demeaning list of items that recipients of thestates’ TANF funds (Temporary Assistance for Needy Families) cannot spendtheir benefits on. This includes things like spending while onboard cruiseships, on movies, or swimming pools. Of course, micro-managing the spend-ing of government money provided to the poor seems perfectly reasonable tosome. However, it is unclear why that does not also apply to all the membersof the state legislature and other citizens of the state who receive governmentsupport to pay for their home mortgages, student loans, child and dependentcare, health insurance, and retirement funds. After all, retired folks like to takecruises too and a good portion of the funds to purchase such trips will likelycome from their lifelong taxpayer subsidized retirement accounts.

Submerged Welfare Policy and the Big Five

When it comes to the Big Five personality traits, the hidden welfare statecan be particularly problematic for a trait like conscientiousness. Being


high in conscientiousness is not only about setting high standards foroneself, but also for others. Indeed, Hofstee, de Raad, and Goldberg’s(1992) AB5C measure of conscientiousness includes questions like, “Sethigh standards for myself and others” and “Demand quality.” And, ofcourse, research has documented the connection between conscientiousnessand adherence to the norms of citizenship and sense of duty (Dinesen et al.2014). If public knowledge of the help that governmental programs provideto citizens is biased toward only certain types of programs, for example,welfare spending through budgetary expenditures, then it will leave manypeople with the impression that they themselves are not being helped. Thiswill be particularly problematic for those high in conscientiousness, whobelieve they are playing by the rules and working hard without help fromgovernment. As a result, I would expect that they will be significantly morelikely to be welfare policy hypocrites, that is, benefiting from hidden welfarestate programs while opposing programs that visibly help those who aredemonized as lazy freeloaders (Lepianka et al. 2009).

On the other hand, agreeableness is predicated on feelings of sympathy,kindness, affection, warmth, and not finding fault with others.Interestingly, one of the few times those high in agreeableness are dis-approving is when they observe antisocial behavior (Kammrath andScholer 2011). Agreeableness is also associated with helping others anddonating to charity (Graziano and Eisenberg 1997). Thus, it is not toomuch of a stretch to hypothesize that increased agreeableness will beassociated with less hypocrisy and general support for welfare policiesregardless of the method used to deliver the benefits.

It is less clear what to expect from the remaining traits—extraversion,emotional stability, and openness. Extraversion is, of course, about socialinteraction with individuals scoring high in the trait thriving in socialsituations. However, there is little in the trait to suggest that these indi-viduals are particularly judgmental of others. Openness to experiences isfocused on being imaginative, intellectual, and having wide interests.Once again, this trait does not appear to be connected to beliefs abouthow others should behave. Finally, emotional stability is associated withhaving an even temper, lacking mood swings or anxiousness, and beingcomposed and confident. This trait is inwardly directed and thus also lacksany particular expectations of others.

The PPSF survey contains multiple avenues to examine the relationshipbetween the Big Five and hypocritical attitudes toward welfare policy. Itasks respondents if they are paying off a mortgage or have a 401(k) or


Individual Retirement Account (IRA) through their employer. It also asksif they have health insurance and where it comes from. According toCongress’ Joint Committee on Taxation estimates, tax expenditures forthe deduction of mortgage interest for owner-occupied residence will cost74.8 billion in 2015 and rise to 93.2 billion in 2018. The various types ofIRAs are estimated to cost 19.5 billion in 2015, rising to 24.4 billion in2018. Social Security costs an additional 39.3 billion in 2015 and anestimated 46.8 billion in 2018. This does not include tax deferrals asso-ciated with 401(k) contribution because this money is taxed as incomewhen it is withdrawn after retirement. But since people tend to be in lowertax brackets after they retire, this method of providing government help tofacilitate retirement savings is not revenue neutral (Brady 2012). Finally,as one might guess, the cost of health insurance related tax expendituresdwarfs all of these. The tax exclusion of employer contributions forhealthcare, health insurance premiums, and long-term care insurance pre-miums is estimated to be 150.6 billion in 2015 and to rise to 172 billion in2018. This does not include the additional 33.2 billion associated withexclusions for Medicare benefits in 2015, which is estimated to increase to40.8 billion in 2018. All told, there are hundreds of billions of dollars inbenefits going to citizens, most of whom do not think they are gettinghelp from government.

The PPSF survey also asks respondents about their substantive prefer-ences about spending on welfare in general and food stamps specifically.This provides the opportunity to examine these attitudes in a new light.Typically research will focus on what drives these attitudes directly.However, here I am not interested in simply examining the Big Five’seffect on a policy area, but rather on how one’s personality can causeopposition to these polices while simultaneously benefiting from hiddenwelfare state programs. I refer to this as hypocrisy. The survey also asks aseries of questions about health insurance. First, it asks if respondents havehealth insurance. If respondents answer yes, it then asks them where theyget it. Then it asks if they get help paying for their health insurance.Finally, it asks all respondents if they support federal or state governmentprograms designed to help individuals pay for their health insurance. Thisbattery of questions allows me to not only look at hypocrisy amongrespondents but also examine the effects of ignorance of the governmentalprograms helping respondents pay for their health insurance. Whether therespondent is aware of it, virtually everyone who has health insurance isgetting help paying for it.


Table 5.1 presents descriptive statistics associated with each of thesesurvey questions. As can be seen, about 39 percent of respondents arepaying off a mortgage. This is consistent with the 2014 U.S. CensusBureau’s American Community Survey (USCB 2015), which found thatabout 40 percent of households are owner-occupied with a mortgage.In addition, nearly half of respondents are taking advantage of an IRA or401(k) through their employer. Nearly 85 percent of respondents indicatedthat they have health insurance. This also appears to be consistent withnational trends. According to anOctober 2014Gallup poll, 13.4 percent ofAmericans are uninsured (Levy 2014). Of course, the interesting part ofTable 5.1 is that only 23 percent of those individuals who have healthinsurance believe they are getting help paying for it. At the same time,nearly 62 percent of all respondents support government programs to helpindividuals pay for insurance. Thus, there are many respondents whowouldlike to see government programs to help people pay for insurance that donot realize they are already benefiting from such programs.

I have measured hypocrisy in three ways with this data. First, I focus onopinions on welfare and food stamp spending. In this case, I look at those

Table 5.1 Descriptive statistics of respondents’ benefiting from welfare programs

Yes No Don’t know

Paying mortgage 39.41 60.59Have 401(k) or IRA 48.55 51.45Currently have health insurance 84.96 13.44 1.60Do you get help paying for health insurance 22.74 73.63 3.63Do you support programs to help pay for healthinsurance

61.91 24.77 13.32

Health insurance details

Where do you get your primary health insurance? %

Through employer 50.72Medicare 23.24Medicaid 7.05Exchange setup as part of Affordable Care Act 4.95State health insurance program 2.20Directly from a health insurance provider 8.22Other 3.52


who say they are either paying off a mortgage or have an IRA or 401(k)retirement accounts through their employer and thus are benefiting fromtwo of the largest submerged welfare state programs, but would like to seespending on welfare and food stamps reduced. This produces two vari-ables. The first indicates hypocrisy when it comes to welfare policy ingeneral and the second indicates hypocrisy when it comes to food stampsin particular. The third hypocrisy measure is focused specifically on healthinsurance coverage and combines those who have health insurance, andtherefore receive help paying for it, but do not favor governmental pro-grams to help pay for health insurance. Table 5.2 displays descriptivestatistics for each of these variables. As can be seen, hypocrisy on welfarespending in general is the single largest group, with nearly a third ofrespondents who benefit from submerged welfare state programs like themortgage interest deduction and retirement savings programs wanted tosee less spending on welfare. This number drops a bit when focusedspecifically on food stamp spending. The issue of health insurance seesthe fewest hypocrites, with just over 21 percent having health insurancebut against programs that help pay for it.

Table 5.3 presents results after estimation of four models predictingthe different types of hypocrisy individually (Part A) and the volume ofhypocrisy (Part B).4 Looking first at Part A’s model predicting welfarehypocrisy, conscientiousness is the only one of the Big Five that reachessignificance with the predicted sign. This will become a recurringtheme with the rest of the models. As individuals’ level of conscien-tiousness increases, their probability of being a hypocrite also increases.Indeed, the effect is quite large. An individual high in conscientious-ness is nearly 15 percent more likely to be a hypocrite than one who islow in this trait. In comparison, whites are only 8 percent more likely

Table 5.2 Measuring hypocrisy

Yes No

Benefit from mortgage and/or retirement programsSpend less on welfare 32.15 67.85Spend less on food stamps 27.51 72.49

Have health insuranceAgainst programs to help pay for health insurance 21.23 78.77Don’t think they get help paying for it 62.54 37.46


to be hypocrites than minorities and those who have a college degreeand some post-graduate work are nearly 14 percent more likely to behypocrites than those who did not finish high school. As one mightexpect, ideology plays a large role here. Consistent conservatives are

Table 5.3 Predicting hypocrisy

Personally benefiting from the welfare state while not support it

Part A Welfare spending Food stamp spending Health insurance

Big FiveExtraversion 0.031τ

(0.020)0.036*(0.021)

0.020(0.024)

Agreeableness −0.039τ

(0.027)−0.057*(0.028)

−0.098**(0.032)

Conscientiousness 0.148***(0.031)

0.183***(0.033)

0.110**(0.035)

Emotional stability 0.006(0.024)

0.011(0.026)

0.041τ

(0.029)Openness −0.035τ

(0.026)−0.030(0.027)

−0.027(0.031)

Part B

Predicting the amount of hypocrisy

Count Zero (inflate)

Big FiveExtraversion −0.003 −0.065*

(0.011) (0.031)Agreeableness −0.010 0.087*

(0.014) (0.040)Conscientiousness 0.051** −0.144***

(0.018) (0.044)Emotional stability 0.008 −0.006

(0.014) (0.041)Openness 0.008 0.097**

(0.014) (0.041)

Notes: Part A, Logit models were estimated in all three cases. Part B, Zero-inflated negative binomialmodel was estimated (Vuong = 41.46, pr. = 0.00). The “Zero” model is the inflation model predictingzero hypocrisy. The “Count”model predicts non-zero responses. Scale runs from 0 to 3. Both portions ofthe model contain the same variables. Robust standard errors in parentheses. Control variables not shown.τ p < 0.10, * p < 0.05, ** p < 0.01; one-tailed


nearly 42 percent more likely to be hypocrites than consistent liberals.This is far larger than the effect of partisanship, where a Republicanidentifier is 12 percent more likely to be a hypocrite than a Democraticidentifier. In the end, conscientiousness is one of the largest predictorsof hypocrisy in this area.

The second model in Table 5.3 predicts hypocrisy regarding foodstamp programs. Obviously, this is a more specific type of welfare programand thus can draw direct connection between those who benefit fromgovernmental programs associated with home mortgages and retirementsavings. As can be seen, conscientiousness once again reaches significancewith the expected sign. In addition, both extraversion and agreeablenessfind significance in this model. Agreeableness has the expected negativesign, which indicates that the more agreeable people are, the less likelythey are to be hypocrites. On the other hand, extraversion has a positivesign indicating extraverts are more likely to be hypocrites than introverts.Substantively, the effect of conscientiousness is much larger than the othertwo traits. Those high in conscientiousness are 16 percent more likely tobe hypocrites on this issue than those low in the trait. By contrast, extra-verts are only 4 percent more likely to be hypocrites than introverts, whilethose high in agreeableness are 5 percent less likely to be hypocrites thanthose low in the trait. The control variables’ (ideology, partisanship, andrace/ethnicity) effects are similar to those of the previous model. Thereare two new additions to this model. These are the inclusion of variablestapping how important individuals feel the issue is and the ambiguity oftheir thoughts on it. Interestingly, the importance of food stamps to therespondent reaches significance, but the ambiguity variable does not.Those who find the issue extremely important are nearly 8 percent lesslikely to be hypocrites than those who do not find it important at all.

The final model in Table 5.3 predicts hypocrisy in relation to healthinsurance. Both conscientiousness and agreeableness are once again sig-nificant with the expected signs. Substantively, an individual moving frombeing low to highly conscientious is associated with a 9 percent increase inthe probability of being a hypocrite on health care. Going from low tohigh in agreeableness has the opposite effect, decreasing the probability ofbeing a hypocrite by about 8 percent. As with the other two models,ideology has the single largest effect, with a move from a consistent liberalto a consistent conservative associated with nearly a 39 percent increase inthe likelihood of being a hypocrite. Partisanship’s effect is less than halfthat, with a move from being a Democrat to a Republican increasing this


probability by more than 16 percent. In sum, Table 5.3 documents aconsistent positive effect of conscientiousness on hypocrisy, regardless ofthe issue area, with only ideology having a larger, more consistent effect.

Each one of the models in Part A of Table 5.3 predicts one type ofhypocrisy, but one might wonder about the volume of hypocrisy. Being ahypocrite on one issue is not uncommon. I would guess that most of ushave been a hypocrite in one way or another when it comes to publicpolicy. But there may be something unique about those individuals whoare consistently hypocrites, that is, on numerous issues. With this in mind,I summed the three different instances of hypocrisy into a single variablethat indicates the total number of times a respondent was a hypocrite onthese issues. The result is a scale that runs from 0 to 3. Part B of Table 5.3presents results after estimation of a model that both predicts the prob-ability of being a zero on this scale (i.e., the respondent was not ahypocrite in any of the models in Part A) and the volume of none zeroresponses.5 Thus, there are two separate outputs for the one model. Theleft-hand side of Part B depicts the effect that the Big Five and controlvariables have on the total number of times one is a hypocrite, while theright-hand side of the table indicates the effect of the variable on theprobability of not being a hypocrite in any of the three previous models.

Looking first at the right-hand model predicting the probability of azero, one can see that nearly all of the Big Five appear to contribute to thisprobability. Extraversion and conscientiousness have a negative effect,while openness and agreeableness have a positive effect. A simple way tolook at this is that all four of these personality traits have a significant effecton the probability that someone will be a hypocrite, extraversion andconscientiousness increase it, while openness and agreeableness decreaseit. Substantively the effect of personality is quite large. A person who islow in openness and agreeableness, but high in extraversion and cons-cientiousness is 40 percent less likely to be a hypocrite than one who ishigh in openness and agreeableness, but low in extraversion andconscientiousness.

The more interesting result comes when we look at the count model onthe left-hand side. Increases in both extraversion and conscientiousnesshave a significant negative effect on the probability that one will score azero on this scale (i.e., not be a hypocrite), but only conscientiousness hasa significant positive effect on the amount of hypocrisy one is likely toexhibit. Figure 5.1 depicts this relationship quite clearly. The top portionof the figure represents the effect that conscientiousness has on the


.35

.4.4

5.5

.55

.6

Pro

babi

lity

of Z

ero

on H

ypoc

risy

Sca

le

6 7 8 9 10 11 12Conscientiousness

Predictive Margins with 95% CIs.4

.6.8

11.

2

Pre

dict

ed A

mou

nt o

f Hyp

ocris

y

6 7 8 9 10 11 12

Conscientiousness


Fig. 5.1 Conscientiousness’ effect on hypocrisy


probability of scoring a zero on this scale, while the bottom portiondepicts change in the expected amount of hypocrisy.

The more highly conscientious individuals are, the more hypocriticalthey are likely to be. The other variables that appear to significantlycontribute to the amount of hypocrisy are education, religious fundament-alism (weakly), and ideology. Education has a positive effect here, mean-ing that more educated individuals are greater hypocrites than the lesseducated. Ideology is also positive meaning that as one moves from beinga consistent liberal to a consistent conservative, the larger the volume ofhypocrisy one would expect to see on these issues. For example, a highlyeducated, consistent conservative who is highly conscientious would beexpected to score about 2.1 out of 3 on this scale, while a poorly educated,consistent liberal who is low in conscientiousness would score just 0.2 onthe scale.

Health Care and the Hidden Welfare State

The models presented in Table 5.3 do not allow for the examination of theeffect of ignorance, that is, whether one understands that they are benefit-ing from tax expenditures designed to help cover the costs of healthinsurance. A primary contention of the submerged welfare state thesis isthat individuals do not think of tax expenditures as government programsdesigned to help individuals. Fortunately, the survey also asks respondentswhere they are getting their health insurance and whether they are gettinghelp paying for it. This provides an opportunity to further explore theeffect of hidden welfare state on hypocrisy. Of course, it would be easy tosimply use a variable indicating ignorance in the model predicting health-care hypocrisy in Part A of Table 5.3. I did just that and unsurprisingly, itis highly significant. Those who are ignorant of the submerged welfarestate in this area are significantly more likely to be hypocrites. While this isan interesting result, I do not think it fully captures the relationshipbetween conscientiousness, ignorance, and hypocrisy. I therefore examinethe mediating role that ignorance plays in the connection that conscien-tiousness has to hypocrisy.

Recall from Table 5.2 that nearly 2 out of every 3 respondents whohave health insurance do not think they are getting help paying for it. Tothis point, I have focused almost exclusively on the direct effect of the BigFive on the political understanding of individuals. However, there isreason to believe that this only captures part of the influence that


personality has on something like hypocrisy (Mondak et al. 2010). Whatdoes it mean when I say that ignorance of whether one is getting helppaying for health insurance mediates the relationship between conscien-tiousness and hypocrisy? Fig. 5.2 provides a graphical depiction. The idea isthat both conscientiousness and ignorance of the submerged welfare stateaffect the likelihood that one is a hypocrite on this issue. But, conscien-tiousness also affects the likelihood that one is ignorant of the fact that theyare getting help paying for health insurance. Thus, conscientiousness notonly has a direct effect on hypocrisy but it also has an indirect effect onhypocrisy that works through ignorance. Therefore, in order to provide amore complete understanding of the total effect that conscientiousness hason hypocrisy, I provide a more thorough examination in what follows.

There are essentially four steps to documenting a mediation effect(Baron and Kenny 1986; Judd and Kenny 1981; James and Brett 1984).First, demonstrate that the causal variable (in this case conscientiousness) iscorrelated with the dependent variable (hypocrisy). Second, show that thecausal variable is correlated with the mediating variable (ignorance). Third,show that the mediating variable is also correlated with the dependentvariable. Finally, fourth, parse the direct from the indirect effect of thecausal variable in order to establish the total effect of the causal variable.

Beginning with the first step, one can see in Part A of Table 5.3 thatconscientiousness is clearly correlated with hypocrisy on the issue of healthinsurance. Next, in order to fulfill step two, I estimate a model predictingthe probability that one is ignorant of the fact that they are getting helppaying for health insurance, with the Big Five and all the other controlvariables (not shown). In this case, conscientiousness is also a strongpredictor of this ignorance. Finally, step three is to show that ignoranceis a predictor of hypocrisy. As stated above, when I re-estimate the Health

Ignorance

HypocrisyConscientiousness

Fig. 5.2 Conscientiousness’ effect on hypocrisy is mediated by ignorance


Insurance model in Part A of Table 5.3 with a dummy variable thatcaptures whether the respondent believes that they get help paying forhealth insurance, the variable is highly significant.6 Thus, the first 3 stepsof documenting the mediating relationship have been demonstrated.

The final step of the process is to test the mediated effect of thevariables. Here I perform this test and display the results with boot-strapped confidence intervals in Table 5.4. The indirect, direct, and totaleffects are shown. As can be seen, in each case these relationships arestatistically significant, that is, none of the 95 percent confidence intervalscontain zero. This lends strong support to the notion that ignorance playsa mediating role in the conscientiousness-hypocrisy relationship.Substantively, these results show that about 29 percent of the effect thatconscientiousness has on hypocrisy is filtered through ignorance.Importantly, the coefficient corresponding to the total effect of conscien-tiousness is 0.139, which is about 26 percent larger than the 0.110observed in Table 5.3. Thus, Table 5.3 underestimates the total effect ofconscientiousness and also suggest that many of the findings in this bookmay be underestimating the importance of the Big Five. Ultimately, thismakes conscientiousness second only to ideology in terms of the size of itseffect on the probability of being a hypocrite on health insurance.7

THE PERSONALITY OF THE HIDDEN WELFARE STATEThe problem that the submerged or hidden welfare state poses for demo-cratic governance is that it causes citizens to have an incomplete or simplyincorrect understanding of what government does for them personally.This results in a perceptual divide where hundreds of billions of dollars ingovernment spending in the form of tax expenditures is not understood as

Table 5.4 Testing the mediation effect of ignorance on the conscientiousness—hypocrisy relationship

Coefficient Bootstrapped std. errors Bias-corrected 95% C.I.

Indirect effect 0.040 0.012 0.019–0.063Direct effect 0.099 0.039 0.029–0.183Total effect 0.139 0.040 0.063–0.222

Notes: The binary_mediation command in Stata was used to produce these results; bootstrapped standarderrors after 500 replications using the percentile method for bias-corrected confidence intervals.

THE PERSONALITY OF THE HIDDEN WELFARE STATE 125

welfare, while programs administered through budgetary expendituresare. The bias this produces is one where individuals hypocritically opposespending for others while reaping the rewards of government spending forthemselves. In many instances this hypocrisy occurs because it benefits amajority. But as Madison argued in Federalist 10, simply following the willof the majority can be a big problem for democratic societies.

Ultimately, there are three questions regarding the effects of the hiddenwelfare state that this chapter can provide answers to. First, does knowl-edge that one is benefiting from a government program affect one’s like-lihood of being a policy hypocrite? Second, are whites (racial majority)more likely than minorities to by hypocrites? Third, how does one’spersonality affect their propensity to be policy hypocrites? Because thisbook is centrally focused on the personality correlates of political knowl-edge and understanding, the previous sections have focused heavily on thethird question.

As has been shown, the answer to the first question is yes, at least in thecase of health insurance. Those who do not think they are getting helppaying for their health insurance are significantly more likely to be hypo-crites. Importantly, the answer to question three is also yes and has a directeffect on whether one knows they are benefiting from government help.In this case, conscientiousness is a cause of this knowledge, that is, themore conscientious one is, the less likely one is to think one is getting helpfrom government programs. This is then a cause of policy hypocrisy.Unfortunately, I cannot test the effect of ignorance in the case of welfareor food stamp policy, but my supposition is that ignorance has the sameeffect in other contexts.

However, I was able to test the direct effect of the Big Five on being ahypocrite and once again conscientiousness proves to be a strong predictorof hypocrisy. Of course, while conscientiousness has received the lion’sshare of the discussion in this chapter, it is important to remember thatagreeableness has nearly as consistent an effect on hypocrisy, but in theopposite direction. Agreeableness’ effect is also not as large as conscien-tiousness’, but when one combines the effect of both of these personalitytraits, one is struck by how large an effect they have. For example, anindividual who is high in agreeableness and low in conscientiousness isnearly 23 percent less likely to be a hypocrite on food stamps than one whois low in agreeableness and high in conscientiousness. Nothing else in themodel except ideology comes close to the size of this effect.


To this point, I have left the second question I mention above unexplored.Fortunately, Tables 5.3 provides some answers. One of the control variablesincluded in each of the models is a dummy variable indicating whether one iswhite. This provides a rudimentary test of the supposition that those in themajority will be more likely to be hypocrites than those in the minoritybecause hidden welfare policies disproportionally benefit their group. Thefull model results are not show, but in the case of welfare generally and foodstamps particularly, whites are significantly more likely to be hypocrites on thisissue. That is, of all the individuals who are receiving government benefitsthrough tax expenditures on home mortgage deductions and retirementsavings, whites are more likely to want to reduce spending on welfare andfood stamps than minorities. Interestingly, this effect does not show up whenlooking at the issue of health insurance.8 This is likely the case because publicdiscourse on healthcare lacks a racial component.

Given these strong results in the case of welfare and food stamps, I wascurious as to the possible interactive effect that the Big Five might havewith race/ethnicity. As an exploratory test, I re-ran the models in Part A ofTable 5.3 with interactions between each of the Big Five and the variable“white.” The results failed to show a significant interaction between thesevariables in all but one case. When openness to experience is interactedwith race, it shows in both the welfare and food stamp models that anincrease in openness reduces the probability of being a hypocrite forminorities, but not for whites. In other words, ceteris paribus, whitesthat score high in openness to experience are significantly more likely tobe hypocrites on these issues than minorities that score high in this trait.

All told, as was the case in the previous chapter, the importance of one’spersonality in driving hypocrisy should not be underestimated. In thiscase, it was conscientiousness that played the biggest role, with agreeable-ness having a smaller effect. Conscientiousness is often viewed as a goodtrait in which to score high. It contributes to success at work and inschool. However, the results of this chapter suggest it can also lead to anincomplete understanding of one’s own relationship with governmentservices and the incorrect perception that one is wholly self-sufficient.This results in one hypocritically supporting policies that help oneself,while opposing those that help others. And perhaps even more impor-tantly for individual citizens, it can result in them cutting off their nose tospite their face, as was the case with the man from South Carolina and thewoman from Texas whose stories began this chapter.

THE PERSONALITY OF THE HIDDEN WELFARE STATE 127

NOTES

1. See, for example, Wisconsin Representative and Speaker of the House PaulRyan’s comments about the inner city culture of men not wanting to work(Volsky 2014).

2. Of course, this is a generalization. The exact amount of interest woulddepend on various terms and vary over the length of the loan. That is, theamount of interest could be as high as $30,000 the first year and diminish to$20,000 by year 15.

3. How much an individual can deduct phases out as incomes increase over$166,800, so an individual with an income of $500,000 would only be ableto claim about $21,700 of the $25,000 in interest they pay.

4. All of the same control variables were include in these models. Table 5.3focuses only on the results of the Big Five variables in order to conservespace. Full model results are available upon request.

5. Since this is count data with over dispersion, I estimated a zero-inflatednegative binomial model (Vuong = 41.46, pr. = 0.00). The Vuong testindicates that the zero-inflated version is preferred to a simple negativebinomial model.

6. The results of some of these models are not shown in order to conservespace, but are available upon request.

7. It should be noted that conscientiousness is often found to be a cause ofone’s ideology. In Chapter 2, it was shown to be negatively correlated withbeing a consistent liberal, however, it was not associated with being aconsistent conservative. As a result, I tested the potential mediating effectthat ideology might have on the connection between conscientiousness andhypocrisy, but found it to not be significant.

8. One might think that this difference is due to the fact that (unlike withwelfare and food stamp policy) this measure of hypocrisy does not usehaving a mortgage or IRA/401(k) as the hidden welfare state benefit, butrather having some type of government subsidized health insurance.However, this is not the case. If I measure health insurance hypocrisy asthose who have a mortgage or IRA/401(k) benefit but oppose governmentsubsidized health insurance, the result is the same. Conscientiousness is astrong positive indicator, agreeableness is a strong negative indicator, andbeing white has no discernible effect.


CHAPTER 6

Personality’s Role in Shaping Civic Aptitude

In his magisterial history of the nineteenth century, Jürgen Osterhammel(2014) discusses the challenges scholars face when attempting to definewhat they mean with the label “nineteenth century.” Where is it locatedboth in time and space? This may come as a surprise to anyone with aglobe and a calendar, but as with many ideas we take for granted (for noother reason than that’s the way it has been done for as long as we canremember), defining time and space is rather challenging. One of theproblems is that time definitions are arbitrary, that is, large swaths of theglobe did not think the year 1800 (defined by Christian European coun-tries) was the year 1800. There is nothing magical about the year 1800either. Many of the most important processes and events began well beforeor after that particular year. Space is problematic, too, since asOsterhammel points out, “The map of the world looks different accordingto the place you take as your systematic observation point” (p. 78). In theend, the problem is one of perspective and the first step to addressing thesescholarly challenges is to recognize that one’s perspective induces unin-tentional (and sometimes intentional) biases into one’s analysis. For exam-ple, calling China, Japan, and Korea the “Far East” is a subjective labelwholly dependent on one’s perspective. It is only the Far East if you startin Europe.

My thesis throughout this work has been that personality is a durableand measurable manifestation of unconscious brain processes that system-atically influences individuals’ civic aptitude. Thus, the perspective for this


129

book is the psychology of personality. One of the primary consequences ofthis choice is that it looks inward into the minds of individuals for causes ofbehavior rather than simply relying on the more typical external causes,such as the socio-economic status of the individual. This was a deliberatechoice in reaction to the wave of new research across psychological, neuro,and genetic sciences documenting, with more precision than ever before,the active role that unconscious brain processes play in driving behavior.This revolution in our understanding has been picked up by scholars andapplied in political contexts to great effect (Alford et al. 2005; Fowler andDawes 2008; Fowler et al. 2008; Hibbing et al. 2014b; Lodge and Taber2013; Marcus et al. 2000, 2007).

The work presented here specifically focuses on civic aptitude as theability of individuals to understand the political world in both an abstract(i.e., ideology) and concrete (i.e., knowing facts about issues; connectingpreferences to parties; and properly understanding one’s own relationshipto government benefits) way. I find these basic connections, which the“ideal” democratic citizen should be able to make, a perfect test arena forthe role of the unconscious mind because they test the limits of the averagecitizens’ ability to live up to this ideal. This ideal may have its roots in folktheories of democracy that have never had an accurate conception of howhumans understand their world (Achen and Bartels 2016), but our poli-tical system is structured based on these false assumptions. Theorists andphilosophers have debated the viability of democratic governance forcenturies (Estlund 2008; Plato 2007; Przeworski 2010; Rousseau 1988)and social scientists finally have the tools to measure and test these ideas.

When it comes to civic aptitude, for half a century civic reformers haveseen the problem of the ignorance of democratic citizenry as one that canbe fixed through education and campaigns to encourage people’s interestin politics (i.e., external variables). While these individuals are certainlywell-meaning, the myopic focus on what can be done to “fix” the problemhas caused a blind spot to develop among these researches and advocacygroups. The blind spot omits an understanding of the causes of citizens’failures that are not obviously fixable (i.e. variables active in the uncon-scious mind). This produces an incomplete picture of the problem and,ultimately, distorts the importance of variables like education and interest.

I have spent the last four chapters documenting how the Big Fivepersonality traits contribute to shaping citizens’ ability to engage withthe political world. But how does one “fix” personality? The reality isthat a variable like personality often has a larger effect on people than their

130 6 PERSONALITY’S ROLE IN SHAPING CIVIC APTITUDE

education or interest in a particular issue or politics in general. My choiceto look to the internal workings of the mind for answers presents ideas onhow to better design our political system to account for how humansactually behave rather than how the democratic folk theorem supposesthey do. Thus, I present my thoughts on where democratic politicalsystems can alter their process in order to try and coax the best out oftheir citizens. Before I get to those, I review the findings from the previouschapters. To do this I ask a simple question of each of the Big Fivepersonality traits: How does the trait affect the quality of individuals’civic aptitude?

THE BIG FIVE EFFECT

Extant research suggests that individuals who score high in extraversionare active, assertive, energetic, outgoing, social, and talkative. These indi-viduals have large social networks and thrive is social settings. Researchindicates that being high in extraversion is correlated with political parti-cipation of all sorts, consuming more political news, and with a strongersense of civic duty. Extraverts also appear to be stronger partisans,although not necessarily drawn to any one particular party. How doesextraversion affect the quality of individuals’ political understanding?Given previous findings that extraverts are more likely to participate inpolitics, the results presented here are disheartening. The simple fact isthat extraverts make mistakes, a lot of them. When it comes to factualknowledge, the more extraverted one is, the more likely one is to believethat welfare recipients are more likely to use drugs than the overallpopulation. They are less likely to know which party supports drug testingwelfare recipients and are less likely to know that U.S. oil production hadgone up since President Obama took office in January 2009. In the end,the more extraverted one is, the larger the volume of incorrect answersthey give to factual questions. At no point in any of the numerous modelspresented in Chapter 3 did an increase in extraversion have a positive effecton respondents’ ability to answer factual questions. Thus, extraversionmay be correlated with increased participation, but the results presentedhere suggest these individuals are far more likely to be ignorant of basicfacts about important political issues than more introverted individualswho are, apparently, more likely to stay home.

Those high in extraversion are also more likely to fail to connect theirpersonal policy preferences to the correct political party. This is true across

THE BIG FIVE EFFECT 131

a wide variety of issues. They are less likely to know which party supportsthem on health-care subsidies, same-sex marriage, energy production, andfood stamp spending. Overall, as extraversion increases, the total volumeof incorrect connections between preferences and party also increases.This is particularly disturbing in light of research by Gerber et al.(2011b) and Mondak (2010), which indicates that extraverts claim toconsume more political news than introverts do. If this is true, they arefailing to understand this news. The issues covered in Chapter 4 are notesoteric, rarely covered trade policy minutia, but highly visible topics likesame-sex marriage. Yet, time and again extraverts fail to make thisconnection.

While not quite as damning as the results of Chapters 3 and 4, extra-verts do not perform well when it comes to policy hypocrisy either. Thehigher one scores in extraversion, the more likely one is to benefit fromsubmerged welfare programs (like the home mortgage tax deduction ortax deferred retirement savings programs), while opposing spending on avisible welfare program such as food stamps. The best that can be said inthis case is that at least extraverts did not always appear to be hypocrites. Ifound no connection between the trait and being a hypocrite on healthinsurance subsidies. Finally, there is also a bit of optimism to be found inChapter 2 with regards to being consistent with one’s symbolic andoperational ideology. In this case, the more extroverted one is, the morelikely one is to be a consistent moderate. Perhaps extraverts’ need forsocial contact and interaction causes them to opt for moderate politicalviews so that they can appeal to the broadest swath of individuals possible.

Openness to experience proves to be a much more positive contributorto individuals’ civic aptitude. This is not too surprising given its originsand close connection with general intellect. Openness is associated withadjectives like curious, imaginative, insightful, original, introspective, andartistic. Extent research has consistently found a connection betweenopenness and liberal ideology, which was echoed here in Chapter 2. Themore open to experience individuals are, the more likely they are to beconsistent or shy liberals. Recall that “shy liberals” are those individualswho hold operationally liberal policy views, but are reluctant to labelthemselves as liberal symbolically, and instead call themselves moderates.As expected, openness was also negatively associated with being a consis-tent conservative.

Beyond ideology, existing research has produced mixed results withregard to participation and civic engagement. Importantly, when


statistically significant results are found, they indicate that increased open-ness leads to an increase in political participation and sense of civic duty.Interestingly, Gerber et al. (2013b) found that those high in opennesswere most susceptible to campaign mobilization efforts. However, Perryand Sibley (2013) show that these same high scoring individuals are lesssusceptible to messages and information regarding danger and societalthreat. This suggests that highly open individuals are paying attention tothe political world, but are not myopically influenced by it. That is, theyare able to discriminate between the content of various messages. Thissuggests that these individuals should do well with the questions found inthe PPSF survey. But do they?

The answer to that question is an emphatic yes. Of all the ways I testedthe civic aptitude of individuals in the previous four chapters, openness hasthe most consistent positive effect. Scoring high in openness is associatedwith an increased likelihood of knowing political facts, such as which partyis more likely to support drug testing welfare recipients; that welfarerecipients are no more likely to be on drugs than the general populous;that U.S. oil production has increased since January 2009; knowing thepercentage of food stamp recipients living in households with an income;and, of course, an overall decrease in the volume of incorrect answers tofactual policy questions.

In addition, scoring high in openness to experience also increasesthe probability that one can connect their personal policy preferencesto the party that best represents those preferences. This is the caseacross a large diversity of issues, for example, welfare policy, health-caresubsidies, same-sex marriage, and energy production. It should benoted that some of these effects are sensitive to model specification.However, in the case of predicting the total volume of correct connec-tions, model specification did not matter. The higher one scores inopenness, the fewer incorrect connections that individual is likely tomake. Finally, when it comes to the hypocrisy found in Chapter 5,openness did not play much of a role. The one instance where it showsa significant effect is in predicting the total amount of hypocrisy(Table 5.4). The results of that model indicate that scoring higher inopenness increases the likelihood of scoring a zero (i.e., not being ahypocrite on any issue). Ultimately, this is still a positive result, whichis consistent with the results from the other chapters. In the end,scoring high in openness to experiences has a positive effect on one’soverall knowledge and civic mindedness.


The Big Five trait conscientiousness produced more varied results thaneither extraversion or openness. Recall that conscientiousness is about self-control or self-governance. It is associated with adjectives like efficient,organized, reliable, thorough, dependable, and dutiful. Given this con-stellation of adjectives, it is no surprise that research has found those highin conscientiousness to perform better at work and school. They alsoappear to have better financial and health practices, while being lessaccident prone. Although research has failed to find a connection betweenconscientiousness and political participation or civic engagement, therehas been some evidence to suggest a negative correlation between the traitand political knowledge. Overall, the strongest finding has been betweenconscientiousness and political ideology and partisanship. The higher onescores in conscientiousness, the more conservative and the stronger oneidentifies with the Republican Party (in U.S. samples).

Given that one of the more consistent findings with regard to con-scientiousness is that it is correlated with conservative ideology, the resultsof Chapter 2 are quite illuminating. None of the previous research on thisconnection has taken into account the potential contradiction betweenone’s symbolic ideology and operational ideology. This has led to a bit of amisunderstanding with regards to how highly conscientious individualsrelate to the political world. Data used in Chapter 2 showed the typicalpositive relationship between the trait and symbolic conservatism. Butonce I utilized the more accurate nine-point joint ideology scale, whichcombines both symbolic and operational ideology, one can see that con-scientiousness is not associated with being a consistent conservative.Indeed, it is negatively correlated with being a consistent liberal or con-sistent moderate, but only positively associated with being a shy liberal. Inother words, those high in conscientiousness appear to feel the need tosymbolically identify as moderate, while holding liberal operational policybeliefs. For some reason, highly conscientious individuals find it difficultto self-identify as liberal. Thus, contrary to extant research, the story withregard to conscientiousness’ connection to ideology is that conscientious-ness is not associated with actual conservative operational policy beliefs. Itsimply causes a strong reluctance to identify as liberal. Of course, a dislikefor the term “liberal” is not the same as actually being conservative inpolicy terms, although it may result in respondents choosing conservativewhen the survey question demands it.

Beyond ideology, how well did conscientiousness perform in myother tests of political understanding? When it comes to knowledge


of political facts, it did not play much of a role. In only two instancesdid it reach significance. An increase in conscientiousness appeared tohelp individuals know which party supported drug testing welfarerecipients, while it hindered knowing whether one is getting helppaying for health insurance. Thus, it appears conscientiousness doesnot have much of an effect on knowing policy area facts. But does ithelp one better connect their personal policy preferences to the correctparty? Chapter 4 suggests there is little connection between the traitand this ability. There were two exceptions to this general rule. Anincrease in conscientiousness decreased the likelihood of connecting thewrong political party to one’s preference on drug testing welfare reci-pients. The other instance occurred in the model predicting the totalnumber, or volume, of incorrect connections. Increased conscientious-ness appeared to help in this regard too, that is, reducing the totalnumber of times one makes this mistake. However, this was only truein the model without control variables. Thus, it is not a strong a resultas one might hope.

Finally, when looking at the results of Chapter 5’s examination ofpolicy hypocrisy, conscientiousness performs particularly poorly. It doesnot appear that civic mindedness goes very well with scoring high inconscientiousness. It was the only one of the Big Five traits to consistentlypredict an increase in the probability of holding hypocritical views. Ofthose respondents who benefited from submerged welfare state programs,only those scoring high in conscientiousness also wanted to decreasespending on welfare in general, food stamps in particular, and opposedgovernment health insurance subsidies.

Importantly, the effect of conscientiousness is quite pervasive in thisinstance. Chapter 5 showed that its effect on hypocrisy is mediated bywhether individuals are ignorant of the fact that they are getting help froma government program. In other words, increased conscientiousness hasboth a direct effect on hypocrisy (i.e., increased conscientiousness leads toincreased hypocrisy) and an indirect effect where it causes an increasedlikelihood of ignorance that then also increases the probability of hypoc-risy. Thus, the total effect of conscientiousness in this case is larger thanjust about every other variable. In the end, conscientiousness is an inter-esting and complex personality trait. On the one hand, research has shownit to be quite valuable for individuals as it leads to better performance atwork and school, but, on the other hand, it appears to create individualswho are less likely to understand their relationship with government


programs, which turns them into policy hypocrites that take benefits forthemselves while denying them to others.

Agreeableness is the Big Five trait most focused on developing andmaintaining positive social relationships. It is associated with adjectives likeappreciative, forgiving, generous, kind, trusting, warm, modesty, straight-forward, and altruistic. Overall, extant research has not found a strongconsistent role for agreeableness in the political sphere. Work has shownmixed results with regards to political participation, although there isevidence of agreeableness being associated with civic engagement. Thereare also conflicting findings with regards to its connection to ideology.The results in Chapter 2 can add some clarity to this. First, using standardmeasurements of ideology, an increase in agreeableness is positively asso-ciated with operational liberalism and negatively associated with opera-tional conservatism. However, the more nuanced investigation intorespondents’ joint ideology reveals that scoring higher in agreeablenessis positively associated with being a confused liberal. This means that themore agreeable one is, the more they claim to be a symbolic conservativeand operational liberal. This is interesting because, as Ellis and Stimson(2012) show in their work, when measuring symbolic ideology, conser-vatism is the majority response and when measuring operational ideology,liberalism is the most common response. Thus, it appears that in eithercase those scoring high in agreeableness are happy to simply agree withwhat most people around them would answer.

When it comes to knowing facts about public policy, agreeablenessdoes not consistently reach significance. However, there were a fewinstances where increased agreeableness did reduce the likelihood ofbeing wrong. Those scoring high in the trait were less likely to incorrectlythink they were not getting help paying for health insurance (when con-trols were included in them model), less likely to think that people onwelfare are more likely to be on drugs (without controls), and less likely tothink that U.S. oil production had not increased since January 2009(without controls). Thus, when agreeableness did have a significant effect,it was to help reduce the probability of being wrong. This is a positiveresult, but it is unfortunate it did not appear more consistently acrossissues and model specification.

Increased agreeableness is also a positive predictor with civic mind-edness in Chapter 5. Indeed, agreeableness appears to be the strongestcounterpoint to conscientiousness in this context. Scoring high in agree-ableness is associated with a decreased probability of being a hypocrite


when it comes to food stamp spending and health insurance subsidies.When it comes to the volume of hypocrisy, the higher one scores inagreeableness, the higher the probability that they will hold zero hypocri-tical policy positions. This, of course, is a very good result for thoseinterested in creating a sense of community.

The one place where one’s level of agreeableness does not appearto have any effect at all is in the ability to match one’s personal policypreferences to the proper political party. Agreeableness failed to reachsignificance in every single model I estimated in Chapter 4. This is acurious result, given the significant role it plays in the other areas ofpolitical understanding I examined. It appears that agreeablenessleads to a general inclusive and positive view of public policy, butthis does not translate to knowing which party one should support ifone wants to see their preferences put into law. When it comes to thefunctioning of democratic institutions, this is problematic since it isnot clear that these individuals are capable of navigating the complexpolitical world successfully. This can then lead to a systematic processby which their policy preferences are improperly translated into publicpolicy.

Finally, the last of the Big Five personality traits is emotional stability(i.e., neuroticism). This trait captures one’s ability to control their emo-tions. It is associated with concepts like anxious, tense, touchy, unstable,hostile, and worrying, and is labeled neuroticism when one is high in theseconcepts and emotionally stable when one is low. Extent research hasfound mixed results with regards to emotional stability’s effect on politicalparticipation and has failed to find a connection to a sense of civic duty orpolitical efficacy. In terms of ideology and partisanship, when a significantresult has been found (which is not always the case), scoring high inemotional stability is associated with increased conservatism and identifi-cation with the Republican Party (in U.S. samples).

The results of Chapter 2 lend further support to the notion thatemotional stability is a driver of ideology, not just the simple connectionbetween scoring high in emotional stability and increased conservatism,but also an ideological clarity. The results here show that as emotionalstability increases, it increases the likelihood of being a consistent con-servative and decreases the probability of being a consistent liberal. Theonly other trait that has this type of effect on ideology is openness toexperience, which is the opposite. However, in the case of openness, thosescoring high in the trait also appear to be a bit “shy” about their liberalness


in that they would often symbolically label themselves moderates ratherthan liberal. This is not the case with emotional stability.

Emotional stability proved to be a positive predictor of the ability toanswer factual policy questions as well. In Chapter 3, an increase inemotional stability was associated with a decreased probability of incor-rectly thinking welfare recipients are more likely to use drugs than thegeneral populous; not knowing which party supports drug testing welfarerecipients; not knowing that U.S. oil production had increased sincePresident Obama took office; not knowing that religious organizationscan refuse to marry same-sex couples; and not knowing the percentage offood stamp recipients of working age, but not working. It is important tonote that only in two cases did these results remain significant once controlvariables were added to the model (in the case of U.S. oil production andwelfare recipients’ comparative drug use).

Beyond ideology and factual knowledge, emotional stability failed toproduce significant results in the remaining tests. That is, when it comes tothe ability to connect one’s personal policy preferences to the correctpolitical party, much like agreeableness, emotional stability failed toreach significance in every model. Finally, while emotional stability didnot appear to significantly reduce the probability of one being a policyhypocrite in Chapter 5, it did not increase this likelihood either. Perhapsthe failure to find significant results in either of these chapters is not toosurprising given that previous research has also failed to find a connectionbetween emotional stability and civic mindedness, political efficacy, orpolitical knowledge. Thus, these “null” results can be added to that pile.

Overall, the story of emotional stability’s effect in the political sphereappears to be that it is limited to the realm of ideology. It helps individualsdevelop coherent conceptions of policy effects and how they connect toone another, which leads to a general tendency to favor conservative policypositions. The inability of these individuals to then connect their policypositions to the proper party is troubling. It would appear that thesehighly emotionally stable individuals need a healthy dose of openness toexperience to help them make that connection. Being high in both ofthese traits might however cause an existential crisis, since they pull inopposite ideological directions.

In the end, what is the ideal democratic personality? Based on thevarious civic aptitude tests examined here, the ideal democratic personalityis one that is high in openness to experience, agreeableness, and emotionalstability, with moderate to low levels of conscientiousness, and also low in


extraversion. Openness is the most consistently influential trait across eachof the areas I examined in this book. Extraversion plays a big role when itcomes to knowledge related to political facts and party issue positions, butdisappears when it comes to hypocrisy. Agreeableness finds a small role inhelping understand some facts, but is a big player when it comes tohypocrisy. Conscientiousness is also only a bit player when it comes toknowing facts, but is the major counterpoint to agreeableness when con-sidering hypocrisy. Finally, emotional stability is probably the least impor-tant of the Big Five traits. It is limited to helping with ideological claritythat, apparently, does not translate to knowing party platforms.

HUNTING WHERE THE DUCKS ARE: ADAPTING THE DEMOCRATIC

PROCESS TO ACCOMMODATE REAL PEOPLE

Democratic political systems are built on a rickety foundation of untenableassumptions about how citizens understand and engage with politics. Thisrealization is not new. Parts of the problem have been clear from the veryfirst treatises on government, that is, that most people will never ascend tothe heights of the ideal democratic citizen. Unfortunately, the last cen-tury’s worth of empirical research has not only confirmed this pessimismregarding the average citizen but it has documented that not even themost engaged citizens can reach those lofty heights. As a result, many, ifnot most, scholars have long abandoned the notion of the ideal demo-cratic citizen found in the various components of the folk theory ofdemocracy outlined so masterfully by Achen and Bartels (2016). Yetdespite this scholarly move away from untenable assumptions about gen-eral civic aptitude, we are left with real world political systems that arestructured based on the participation of the civic equivalent of a unicorn.This reality and the all too common refrain by civic reformers to the falseconceptions of citizens found in the folk theory are not politically neutralin their effects. Achen and Bartels state this as succinctly as anyone could:

“In current thinking about democracy, bad ideas dominate public discourse.Romantic notions of democracy sound good and have emotion on theirside. Let the public rule! Generations of thoughtful Americans have pro-moted with genuine sincerity reforms deriving from the folk theory ofdemocracy. But they tend to be badly flawed in practice, primarily becausethey make life all too easy for special interests. Especially at the state level,proponents of mind-numbing clichés about giving power to ordinary people

HUNTING WHERE THE DUCKS ARE: ADAPTING THE DEMOCRATIC . . . 139

bear considerable responsibility for the domination of government by nar-rowly self-interested groups. In reforming government, good intentions andhigh-sounding rhetoric are not enough. In the end, it is the folk theory thatprops up elite rule, and it is unrepresentative elites that most profit from theconvenient justifications it provides for their activities” (327).

References to freedom, liberty, and other aspects of folk theories ofdemocracy rely on the positive unconscious tagging done through “hotcognition” in the brain prior to individuals’ becoming consciously awareof a topic (Lodge and Taber 2013). We have no defense against this.Whatever follows the romantic claims espoused by advocates will auto-matically have a positive feeling attached to it. Of course, that does notguarantee success on the part of the advocate, but it does make it moredifficult to refute these claims because refuting them requires effort on thepart of the individual, rather than happy acquiesces to positive feelings.

If folk theories of democracy are wrong, where do democratic politicalsystems go from here? Some scholars have offered suggestions for altera-tions to the political system that might help account for the non-existenceof ideal democratic citizens. Achen and Bartels readily admit that they donot have a new theory of democracy to offer. Their one concrete sugges-tion is a need to level the playing field during the policy-making process inCongress and state legislatures, where well financed interests tend todominate. This is hard to argue with. However, it largely obviates thecentral problem that democracies face, that is, that they are based on theparticipation of citizens, through voting and elections, who are not up tothe task.

Of course, others have examined the many shortcomings of citizens andoffered suggestions for change. For instance, in their book Do FactsMatter? Information and Misinformation in American Politics,Hochschild and Einstein (2015) identify four types of democratic citizens:(1) those that are active and informed, (2) those who are inactive andinformed; (3) those that are active and misinformed; and (4) those that areinactive and misinformed. They demonstrate that the biggest problem fordemocracies are the active and misinformed, while the inactive andinformed are a missed opportunity. One solution they offer is to ignorethe active and misinformed. They point to real world issues like addingfluoride to public water supplies. Experts and experience have demon-strated this to be a significant benefit to the public. Yet, public opinion andreferenda are dominated by the active misinformed, which results in public


defeat of many attempts to add fluoride to potable pubic water supplies.Thus, the best solution has been to stop asking the public if a town, city, orstate should add fluoride to their water and just do it. They also point toregulatory agencies like the FDA as watchdogs with regard to misinforma-tion in health-related advertising. Another suggestion is to look directly atthe information being provided by political elites and others and provideincentives for them to convey factual information. They suggest this couldtake the form of some kind of public shaming of those who willfully pedalincorrect information. While none of these suggestions are necessarilywrong, like Achen and Bartels, the only one that is aimed at helpingcitizens do better is public shaming of individuals giving false information,but scholarly work on motivated reasoning and the 2016 presidential racesuggest that this is unlikely to have much of an effect. The others are toeither eliminate public participation or fix the problems this participationcreates after the fact.

In their book Predisposed: Liberals, Conservatives, and the Biology ofPolitical Differences (2014b), Hibbing, Smith, and Alford broach thetopic of institutional change by first pointing out what is never going tochange. That is, “ . . . longing for a political system devoid of ideologicaland partisan differences is pointless” (258). And that while representativedemocracy is not going to solve the problem of factions, attempts atreform that focus on direct democracy are worse. They suggest reformsin the primary system that would help more moderate candidates competein a context where voters are typically more ideologically extreme than theaverage general election voter. They also decry the redistricting processthat leads to uncompetitive elections, suggesting the need to provideincentives, beyond blind partisan gain, to those drawing the map.Ultimately, they conclude by saying, “the message of the predispositionsargument to those seeking a form of low-conflict politics based on mutualcooperation, interest and goals is this: Grow up” (261).

That may be true, but I would add that we also need to wake up. Thestructure of our electoral process has been chosen. The rules have beenchosen. They have not magically manifested themselves from our geneticor physiological makeup. As it stands now, our elections are free-for-allswhere the loudest and most financially robust can flood the arena withvirtually no limits to the volume or veracity of their claims. This does nothave to be the case. The way elections are conducted today are not the waythey were conducted 100 years ago (not that they were better, justdifferent) and there is no reason this has to be the way they are conducted


100 years from now. The electoral process we have today was designedbased on the false premise of participation by ideal democratic citizens. Itis not natural. It has not been bestowed upon us from on high. TheFounding Fathers did not have any concept of what a modern electoralprocess would entail.1 Indeed, in the past, if you wanted to be president,you campaigned from your front porch. Recognizing that ideologicalconflict has and will always be with us does not mean that the process bywhich we choose our representatives cannot be structured in such a waythat it helps to reduce the influence of partisan rancor on the choices thatindividuals make rather than exacerbates it.

The steps that can be taken, should Americans have the political will todo so, should be focused on the information environment both in the timeleading up to the election and in the voting booth itself. The scholarlywork on automaticity and the effects of the unconscious mind can lead oneto a rather hopeless view of the democratic political process. But there isreason for hope. There is research that suggests that individuals can breakthrough their partisan blinders and automatic information processing incertain contexts, such as when an issue is particularly important or salient(Bolsen et al. 2014; Cuik and Yost 2016; Druckman and Lupia 2016;Kahan 2013; Mullinix 2016; Mummolo 2016; Sides 2016). And there isevidence that even strong partisans are willing to let go of their partisanopinions when enough information is provided to counter their beliefs(Redlawsk et al. 2010; Weeks 2015). The way individuals receive informa-tion now is haphazard; dependent on interest, prior beliefs, and accident.It encourages information providers to be as superficial as possible to fittheir message in sound bites and 30 second commercials paired withevocative music and visuals. Journalists do not have the time, expertise,or inclinations to play the role of arbitrators. The electoral informationenvironment, as it stands now, is fertile ground for implicit biases andselective perceptions driven by the automatic processes occurring in ourunconscious brains. Nothing is demanded of citizens, passive observanceof attention grabbing spectacle is the norm. But implicit bias can becountered by our conscious selves and that is what a democracy mustdemand of its citizens.

Rebuilding a democracy based on the capabilities of real human beingsrequires a fundamental restructuring of the electoral process. Votingcannot be a right, it must be a duty (Maskivker 2016). And that duty isnot a half thought out vote cast one day every couple of years. Theoverhaul must being with how information is provided to citizens and


how they are required to engage with it. The first suggestion I would makeis that a structured information environment should be created prior tovoting. This could take the form of a three month election season. Duringthis time, the information candidates, parties, and other groups who wishto communicate must be structured in such a way that maximizes under-standing. Issues are presented in context. The claims being made must bespecific, supported by evidence, and describe proposed solutions withevidence to back up the proposed policy. In other words, if one wants toclaim there is a problem with crime, one must say more than it’s a disaster.Pithy claims of this sort leave it up to the listener to fill in the blanks andcitizens are not capable of doing that accurately. They do not have thepolicy expertise and our unconscious brains fill our thoughts with biasedinformation. Importantly, I am not suggesting a limitation of the topicscandidates can discuss. The key is not what is discussed, but how.

Additionally, during this period the playing field must be level.Communication of information cannot depend on a candidate’s wealthor how much money one can raise. The number of different electionscitizens will vote in at one time should be limited. This would likely entaileliminating direct democracy referenda and elections for positions likejudges, sheriffs, and prosecutors. The push for direct democracy effortslike ballot initiatives and electing judges and other officials in the publicsafety field have been fueled by references to false folk theories of democ-racy in reaction to a system of representation that has failed becausecitizens themselves are not capable of holding their elected officialsaccountable. The result has been to turn judges into politicians who touttheir conviction rates when running for office and sheriffs who implementcrime fighting policies not because they have been found to be bestpractices within the profession, but because they look good to voterswho are wholly ignorant of the best crime fighting procedures (Berdejóand Yuchtman 2013; Berry 2015; Bandyopadhyay and McCannon 2014;Canes-Wrone et al. 2014; Cohen et al. 2015; Shepherd 2009, 2013).Indeed, recent work by McCannon (2013) finds that “ . . . the popularelection of prosecutors results in inaccurate sentences, wrongful convic-tions, and, consequently, successful appeals” (1). Too much informationoverwhelms people and limiting the number of elections will help citizensmake higher quality choices in the ones that remain. When this is com-bined with an electoral process designed to help voters understand ratherthan play to their biases, voters will be better equipped to choose amongcandidates, be they incumbents or challengers. This should have a down-


stream effect on the choices elected official make. That is, when politiciansknow that their constituents will understand their behavior, their behaviorchanges.

The controlled structuring of an electoral season is one change thatshould be made, but it is not the only one. The process of voting itselfshould also change. Right now people show up to vote only if they want toand receive a sterile ballot with very little information on it. None of theinformation that is available in the voting booth is helpful to making aninformed decision. The party label does not help low information voterschoose the candidate they would choose if all they knew were the candi-dates’ policy positions (Dusso 2015). And the most highly informedvoters are also the strongest partisans and most susceptible to believingpartisan based falsehoods and bending reality to fit their partisan beliefs(Bartels 2008, 2016, 2008; Claassen and Ensley 2016; Enns et al. 2012;Gerber andHuber 2010; Kahan et al. 2013, 2012; Lenz 2012;Wagner et al.2014; Zaller 1992). Thus, the day of the election also needs to change.

First, voting can no longer be thought of as just a right, it must be aduty. Everyone must participate. Previous works on personality havesuggested that voluntary participation will likely draw extraverts and thehighly conscientious, yet the evidence I present here clearly demonstratesthat introverted individuals are better at performing the basic tasks thatdemocratic elections require. This work also indicates that highly con-scientious individuals fail to see how they benefit from governmentalprograms, while supporting punitive measures to deny governmentalassistance to others. In addition, voluntary participation results in a woe-fully unrepresentative group of people participating. Whites, older indivi-duals, and the economically well off are always overrepresented inelections. For a society that clings to the ideals found in folk theories ofdemocracy, we are certainly more than happy to turn a blind eye to thereality of our unrepresentative elections. Thus, not only is the processbiased against certain groups but it also reduces the participation ofsome of our most capable citizens.

Second, the process of voting should change to involve learning aboutthe candidates while casting one’s vote. What I mean by this is that there isno reason to believe that citizens will engage with the process during theelection season. While the presentation of information during this seasonwould be significantly altered (as I suggested above), it would not requireanyone to actually pay attention. Our long experience with electionsclearly indicates that large portions of the electorate would not spend


much time learning about their options even if the information available isbetter structured to preform that exact task. In addition, while altering thepresentation of information would help to mitigate some of our implicitand explicit biases, it would not eliminate them. Thus, there is no reasonto redesign how information is presented during an election and not takeadvantage of the captive audience that shows up to cast a ballot.

As a result, I would strongly suggest that we structure the process ofvoting in such a way that individuals learn about their options whilechoosing. This would be required, with no option for voters to skip pastthe reading. For instance, advances in learning technologies have turnedbooks into dynamic learning applications. Smart books engage my stu-dents dynamically with the materials I assign. The readings present infor-mation along with periodic questions to probe students understandingwith follow-up material when students get questions wrong. The end goalis to not only learn and comprehend the material but also prepare studentsto do well on quizzes and tests. Elections themselves are a test. They are atest of our ability to live up to standard of the ideal democratic citizen.There is no reason that the process of voting could not adopt this type oftechnology to provide citizens with information to improve their knowl-edge and comprehension while in the voting booth itself. Thus, learningand voting happen together. One would think that the choice of presidentdeserved at least as much time as it takes my students to complete anonline chapter assignment with a smart book. The idea would be tochange the voting process from a sterile information poor environmentto a dynamic information rich environment. Of course, turning electionsinto dynamic learning and voting processes would require considerablymore time on the part of voters than it does now. If participation is arequirement of citizenship, there would also be a significant increase in thetotal number of voters. Thus, my suggestions would necessitate extendingthe time of an election from a single day to at least a few days, if not a weekor more. It would also require standardization across the country andsignificant financial investment to make the technology available in suffi-cient quantity at each voting precinct.

Ultimately democracy has to demand more of its citizens. My pro-posals are focused specifically on changes to the process of elections. Atits base, my argument is that we need to get rid of the show andspectacle of elections. People cannot be dazzled by sparking lights ifthey are not there. This would require taking away campaigns’ abilityto package their own messages. Our society has allowed virtually


unrestricted freedom on the part of those trying to influence voters forcenturies and it is not working. Campaigns have abused this freedom.We have created a political process that makes it impossible for anyoneto try and run a civil, principled, issue based campaign. Glitzy spectacleand scary black and white commercials with ominous music that play tovoters unconscious automatic processing of information will alwaysoverwhelm the often boring presentation of policy discussions. Thus,not unlike an unregulated sport where anyone can use performanceenhancing drugs, the choice to stay clean is really no choice at all. Youeither give up your dream of playing or take the performance enhan-cers. In our political system, even the best, most principled individualsmust wade into the political mud pit, if they ever want to win officebecause our electoral process gives them no choice.

Conclusion

The impetus for this work has been a concern for democratic systems ofgovernance that expect a lot, perhaps too much, from their citizens. Scoresof works in political science and elsewhere have documented that thegeneral populous has little knowledge of or interest in politics and publicpolicy. This deficit in knowledge has been understood for long enough nowin academic circles that it is taken for granted and often brushed aside. Asnew research continues to document the ugly effects of this ignorance, forexample, the wrong person being elected president (Bartels 1996; Lau andRedlawsk 2006), ignoring the problem becomes a more and more unten-able position. Beyond the ignorance of the average citizen, the most dis-turbing research is that which shows the willful ignorance of the highlyeducated, knowledgeable, and involved citizens (Kahan et al. 2013). Thismotivated reasoning on the part of the very individuals who are supposed tobe the foundation upon which democratic governance is built, challengesthe very legitimacy of these institutions (Estlund 2008).

When it comes to engaging with the political world (what I havereferred to as civic aptitude), my idea has been that successes and failuresare caused by far more than education, knowledge, or interest. It is some-thing that comes out of who we are as individuals. Some people arepolitical animals, while others prefer never to hear about politics at all.The causes of this difference are as much internal to the individual, as theyare external in their environment. The difficulty for those interested in


studying the effects of the unconscious mind on political behavior hasalways been the challenge of measurement.

Here I focus on the Big Five personality traits, which have a long historyof analysis across numerous disciplines. This research has documented themanifestation of these traits that comes, largely, out of our unconsciousmind, along with their genetic and physiological antecedents. As a result, Irefer to the Big Five as one of a class of psychological variables that are theobservable product of an active unconscious mind. That is, the effect thatone’s personality has on one’s behavior is generally considered outside of theindividual’s conscious control. Thus, personality’s effect on political beha-vior is the result of brain activity initiated by something other than ourconscious perception. This is incredibly important because one’s personalityhas a big effect on citizens’ civic aptitude. I document the important andvarying effects of each of these traits in the abstract and the concrete, across awide variety of public policy concerns. Ultimately, the task for civic refor-mers is large, much larger than we realize. Increased education and interestin politics can only have a limited effect on the quality of civic discoursebecause the very brain processes that have lead these individuals to careabout politics do not exist in the minds of the individuals they are trying toenergize. In order to affect real change, one either needs to alter citizens’unconscious minds, which is still in the realm of science fiction (Dick 1968),or alter the system in such a way that it meets citizens where they are, whichis nowhere near the idealized democratic citizen found in folk theories.

NOTE

1. To be fair, the founding generation saw quite plainly the limits of individualparticipation, which is why only the House of Representatives was directlyelected. It has been the high minded civic reformers of subsequent genera-tions, mentioned by Achen and Bartels, which have refined our politicalsystem in such a way that it is more and more dependent on the participationof the ideal democratic citizen.

NOTE 147

TECHNICAL APPENDIX

The PPSF survey was designed to connect respondents’ personality to thequality of their understanding of five different issue areas. As such, thesurvey contained two key elements. First, is a measure of the Big Fivepersonality traits and, second, a series of in depth questions about five issueareas. These issue areas included (1) food stamps, (2) same-sex marriage,(3) health insurance subsidies, (4) drug testing welfare recipients, and (5)U.S. energy policy. Each of these five issue areas contained questionsabout (1) the importance of that particular issue area to the respondent,(2) the level of ambiguity contained within respondents’ thoughts aboutthe particular area, (3) a factual question about the issue area, (4) at leastone subjective question about the issue area, and (5) a question aboutwhich political party they felt best handled that particular issue. Finally, thePPSF survey also contained numerous important variables known to con-tribute to how individuals understand and engage with politics, for exam-ple, standard demographic information (age, race/ethnicity, gender, andeducation), political knowledge, religiosity, partisanship, and ideology. Inwhat follows I document how the survey was administered, explain each ofthese measures, their specific question wording, and present descriptivestatistics on the distribution of respondents in each measure.

© The Author(s) 2017A. Dusso, Personality and the Challenges of Democratic Governance,DOI 10.1007/978-3-319-53603-3

149

SURVEY IMPLEMENTATION

The PPSF survey was designed and hosted using Qualtrics’ online surveyplatform. Sampling was done through Survey Sampling International(SSI). The survey was administered July 8–10, 2014. SSI was founded in1977 and maintains 34 ongoing panels to form the core of its onlinesample. Knowing the potential problem with selection bias inherent in anopt-in online survey, SSI improves the quality of its sample by furtherrecruiting participants from online communities, social networks, andwebsites of all types. In doing so, SSI can potentially reach anyone onlinethrough their network of relationships with these online communities,websites, and social media groups. SSI ensures the representativeness ofthe sample and minimizes the risk of bias through a three-stage randomi-zation process. First, participants are randomly selected from SSI’s panelsto be invited to take a survey, and these participants are combined withothers entering SSI’s sampling platform after responding to online messa-ging (through invitations of all types, including e-mail, SMS, text, tele-phone, banner ads, and messaging on websites and online communities).Second, a set of profiling questions (never affirmation questions) is ran-domly selected for them to answer. Third, upon completion, participantsare matched with a survey they are likely to be able to take, using a furtherelement of randomization.For this work, a sample of 2,314 individuals were surveyed. The survey

itself is not a representative sample of the U.S. population. A quota systemwas used to approximate the distribution of U.S. residents in race/ethnicityand gender. Thus, this data is not appropriate for answering questions aboutthe overall U.S. population, such as what percent of the U.S. populous is prosame-sex marriage? Or, what percent of the U.S. population is extraverted?These are questions that would require a representative sample of the U.S.population. Of course, here I am interested in understanding the connectionbetween personality and the successes and failures people have understand-ing and engaging in politics. I cannot think of a confounding variable thatmight drive this relationship that is associated with the opt-in nature of thissurvey. That is, a variable that causes someone to be more or less likely toparticipate in this survey, which also affects the relationship between person-ality and political successes and failures.Of course, personality may have an effect on whether one chooses to

participate in this survey, but as will be seen, there is still significantvariance in each of the Big Five personality traits. For one to make the

150 TECHNICAL APPENDIX

claim that the results presented here are biased in some way because of thesurvey’s opt-in nature, they need to show that the variable they surmise iscausing participation in the survey (and is biasing results) is related to boththe dependent and independent variables of interests. For example, inChapter 2, I show that the personality trait “openness to experiences” ispositively related to having a liberal ideology. If one wants to claim thatthis result is caused by the survey’s implementation, one needs to showthat the variable they claim is causing participation in the survey is drivingboth the respondent’s level of openness to experience and the respon-dent’s ideological liberalness. The vast majority of works in psychologystudying personality do not use representative samples of the U.S. pre-cisely because there is, typically, no reason to suspect that the choice toparticipate in the research is driving the results and the purpose of theresearch is not to make a claim about the overall population.

SURVEY QUESTIONS AND CODING

Part A: Ten-Item Personality Inventory (TIPI)In order to measure the Big Five personality traits, I utilized the Ten-ItemPersonality Inventory (TIPI) designed by Gosling et al. (2003). This shortversion of a Big Five personality measure relies on two items to assesspersonality in each of the Big Five traits. It has been utilized in numerousstudies (e.g., Carney et al. 2008; Gerber et al. 2010, 2013a). Ultimately,the TIPI was chosen for the same reason that just about any scholar mightchoose it rather than the numerous other options. It is short and, there-fore, has the ability to fit into a 20-minute survey that also needs to capturenumerous other demographic and political variables. It is simply notfeasible to include a measure of personality containing fifty to sometimesseveral hundred questions, when one’s entire survey consists of onlyseventy to eighty questions. For example, the Revised NEO PersonalityInventory (NEO PI-R) contains 240 questions. While I would have pre-ferred the precision of the NEO PI-R, it is more than three times longerthan my entire survey. The TIPI question wording is as follows:Here are a number of personality traits that may or may not apply to

you. Please select a number next to each statement to indicate the extentto which you AGREE or DISAGREE with the statement. You should ratethe extent to which the pair of traits applies to you, even if one character-istic applies more strongly than the other.

TECHNICAL APPENDIX 151

(Traits were presented one at a time in random order, each with the7 point scale)

(1) I see myself as: Extroverted, enthusiastic

• Disagree strongly• Disagree moderately• Disagree a little• Neither agree nor disagree• Agree a little• Agree moderately• Agree strongly

(2) I see myself as: Critical, quarrelsome(3) I see myself as: Dependable, self-disciplined(4) I see myself as: Anxious, easily upset(5) I see myself as: Open to new experiences, complex(6) I see myself as: Reserved, quiet(7) I see myself as: Sympathetic, warm(8) I see myself as: Disorganized, careless(9) I see myself as: Calm, emotionally stable

(10) I see myself as: Conventional, uncreative

Each of the Big Five traits was measured with two items.Extraversion: Extraverted, enthusiastic and Reserved, quite (reversed)Agreeableness: Sympathetic, warm and Critical, quarrelsome(reversed)Conscientiousness: Dependable, self-disciplined and Disorganized, care-less (reversed)Emotional Stability: Calm, emotionally stable and Anxious, easily upset

(reversed)Openness to Experience: Open to new experiences, complex and

Conventional, uncreative (reversed)Cronbach’s α = 0.62The correlation between each pair of items:Extraversion = 0.39Agreeableness = 0.24Conscientiousness = 0.35Emotional stability = 0.49Openness = 0.30


These correlations are in line with previous works and not particularlyinformative of the items’ reliability (see Gerber et al. 2010 footnote 11).

Part B: Five Issue Area Question SetsThe PPSF survey contains five sets of questions corresponding to fivedifferent issue areas. These issue areas included (1) food stamps, (2) same-sex marriage, (3) health insurance subsidies, (4) drug testing welfare reci-pients, and (5) U.S. energy policy. The order in which the issue areas werepresented to respondents was randomized. Each issue area contained ques-tions about the importance of the issue to them, a question about howambivalent their thoughts are about the issue, and a question asking whichparty is best at handling the issue. The importance questions followed theANES’ template as follows (See ANES Pre-Election Questionnaire):

How important is the issue of ___________ to you personally?

• Not important at all• Slightly important• Moderately important• Very important• Extremely important

The ambivalence question follows the work of Feldman and Zaller (1992),Holbrook andKrosnick (2005), Tormala andDeSensi (2008), andMulligan(2013). Rather than counting the individual positive and negative thoughtsabout a particular issue (e.g., Lavine et al. 2012), this measure is a singlequestion designed to tap subjective feelings about being pulled in multipledirections on a particular issue. The generic question wording is as follows:Some people feel that there are only good things or bad things about

________. Their feelings are consistent. Other people feel that that thereare both good things and bad things about __________. Their feelings areINconsistent. Thinking about your own views, would you say that yourfeelings about _____________ are:

• Extremely consistent• Very consistent• Somewhat consistent• Somewhat INconsistent• Very INconsistent• Extremely INconsistent


The generic question asking which party handles each issue also followsANES’ template. The question template is as follows:Thinking of _____________: Which party do you think is best in dealing

with it?(Republican and Democrat were randomized)

• Republican• Democrat• Neither• Don’t know

The remaining questions for each issue area contained at least one factual andat least one subjective question. I list each of the remaining questions by issuearea.

Drug Testing Welfare Recipients(Subjective) Should welfare recipients be required to pass a drug test inorder to receive benefits, or do you think they should not be tested? (Firsttwo options randomized)

• They should be tested• They should NOT be tested• Don’t know

(Factual) Do you think people who receive some type of welfare assistanceare more likely to abuse drugs than the average U.S. citizen? (First twooptions randomized)

• Yes• No• Don’t know

(Factual) Members of which party are more likely to support requiringdrug testing for people receiving welfare? (First two options randomized)

• Republican• Democrat• Neither• Don’t know


Healthcare InsuranceDo you currently have health insurance? (First two optionsrandomized)


If respondent answered “yes” to the previous questions, they wereasked:

Where do you get your primary health insurance?

• Through my or a family member’s employer• Medicare• Medicaid• Health-care exchange setup as part of the Affordable Care Act• A state health insurance program• Directly from a health insurance provider• Other ____________________

If respondent answered “yes” indicating they had health insurance, theywere also asked:

(Factual) Do you get help paying for your Health Insurance by either thestate or federal government? (First two options randomized)


(Subjective) Do you support federal or state government programsdesigned to help individuals pay for health insurance? (First twooptions randomized)



Same-Sex Marriage

(Subjective) Should same-sex couples be allowed to marry, or do you thinkthey should not be allowed to marry? (First two options randomized)

• Allowed to marry• NOT allowed to marry• Don’t know

(Factual) In states where Same-Sex Marriage is legal, can religious orga-nizations like churches or synagogues legally refuse to marry same-sexcouples? (First two options randomized)


(Subjective) Do you think the federal or state governments should makelaws regarding who religious organizations can and cannot marry? (Firsttwo options randomized)


Energy Policy

(Subjective) Some people feel the USA should dedicate its resources todeveloping new sources of energy, such as wind and solar, while others feelthe USA should dedicate its resources to the production of existingsources of energy, such as oil and coal. Which side do you tend to agreewith? (First two options randomized)

• Developing new sources of energy• Production of existing sources of energy• Don’t know

(Factual) Has U.S. oil production gone up or down since PresidentObama took office in January of 2009? (First two options randomized)

• Up• Down• Don’t know


If answered “Don’t know” to previous question, responded was asked:What’s your best guess? (Options randomized)

• Up• Down

Food Stamps

(Factual) About what percent of federal Food Stamp benefits do youthink go to individuals living in households that have income from a job?(Open ended)(Factual) About what percent of federal Food Stamp benefits do you

think go to individuals who are working age but do NOT work a paidjob and are not living with children, elderly, or disabled individuals?(Open ended)(Subjective) If you had a say in making up the federal budget this year, for

which of the following programs would you like to see spendingINCREASED and for which would you like to see spending DECREASED?Spending on Food Stamps should be . . . (First three options randomized)

• Increased• Decreased• Kept about the same• Don’t know

Part C: Remaining Questions Measuring Important Political Variables

Political knowledge—this is measured following ANES general politicalknowledge questions. I will just list the topic of the question rather thanexact wording here in order to save space. See ANES 2012 pre-electionquestionnaire for exact wording.

1) ID party controlling House of Representatives2) ID party controlling Senate3) Current unemployment rate4) Properly ID Republican Party as more conservative than

Democratic Party5) ID John Boehner6) ID Joe Biden7) ID David Cameron8) ID John Roberts


Partisanship—this is measured following ANES question sequence inorder to avoid “hidden” partisans (Keith et al. 1992). From this series ofquestions, one builds the partisan seven point scale.

1) Generally speaking, do you usually think of yourself as a Republican,a Democrat, an Independent, or what?

• Republican• Democrat• Independent• Other party• No Preference• Don’t know

2) If Democrat or Republican selected, ask about strength of connec-tion (i.e., “Strong” or “Not very Strong”).

3) If anything besidesDem. orRep. selected, ask “Doyou think of yourselfas closer to the Republican or Democratic Party?” with answers:

• Closer to the Republican• Closer to the Democratic• Other ____________________

Ideology—this is measured in two ways, both using ANES standardquestions.

Symbolic ideology:

We hear a lot of talk these days about liberals and conservatives. When itcomes to politics, do you usually think of yourself as extremely liberal, liberal,slightly liberal, moderate or middle of the road, slightly conservative, con-servative, extremely conservative, or haven’t you thought much about this?

• Extremely Liberal• Liberal• Slightly Liberal• Moderate; Middle of the Road• Slightly Conservative• Conservative• Extremely Conservative• Haven’t thought much about this


Operational Ideology:

The following generic questions setup, with Welfare, Social Security,Public Schools, Science and Technology, Dealing with Crime,Protecting the Environment, and Healthcare used to fill in the blanks.(These were presented in random order)If you had a say in making up the federal budget this year, for which of

the following programs would you like to see spending INCREASED andfor which would you like to see spending DECREASED? (First twooptions randomized)Spending on __________ should be . . .

• Increased• Decreased• Kept about the same• Don’t know

Religious fundamentalism—this is measured using a single General SocialSurvey (GSS) question (Ellis and Stimson 2012).Which of these statements comes closest to describing your feelings

about the Bible?

• The Bible is the actual word of God and is to be taken literally, wordfor word

• The Bible is the inspired word of God but not everything in it shouldbe taken literally, word for word

• The Bible is an ancient book of fables, legends, history, and moralprecepts recorded by men

Interest in politics—measured based on two standard ANES questions.First, “How many days in the past week did you watch news on TV?” and,second, “How many days in the past week have you discussed politics withsomeone else?” Each question ranges from 0 to 7. These two are thencombined with a range of 0–14.Finally, SSI provided the basic demographic information for each

respondent: Age, Race/Ethnicity, Gender, and Education. Tables A1and A2 present descriptive data for each of the variables.


Table A1 Categorical variable descriptive statistics

Variable Percent of total Regression coding

Race/ethnicity (N = 2,317) Dummy, Coded 1 ifWhite 62.24 White, 0 otherwiseAfrican-American 13.16Hispanic 17.18Asian 6.39Other 1.03

Gender (N = 2,317) Dummy, Coded 1 ifFemale 50.8 Male, 0 otherwiseMale 49.2

Education (N = 2,255) 7 point scaleSome high school 2.71 corresponding toHigh school grad 17.01 the 7 education levelsSome college 28.78 0 = some high schoolCollege grad 29.58 6 = Ph.D., J.D., M.D.Some post-college grad 4.79Master’s Degree 13.39Ph.D., J.D., M.D. 3.68

Religious fundamentalism (N = 2,309) 3 point scaleBible is literal word of God 28.24 0 = Bible fableBible is inspired word of God 46.30 2 = literal BibleBible is fable, legend, history 25.47

Partisanship (N = 2,317) 3 dummy variablesRepublican 36.34 for each partisanDemocrat 52.70 type.Independent 10.96

Note: Not all respondents answered every question, thus the N varies for each variable.


Table A2 Continuous variable descriptive statistics

N Mean Std. Dev. Min. Max.

Big FiveExtraversion 2,292 5.91 2.80 0 12Agreeableness 2,303 8.41 2.34 0 12Conscientiousness 2,296 9.61 2.28 0 12Emotional stability 2,298 7.98 2.80 0 12Openness to experience 2,297 8.15 2.43 0 12Political knowledge 2,317 4.21 2.22 0 8

IdeologySymbolic 2,314 −0.05 1.54 −3 3Operational 2,265 −2.04 2.65 −7 7Interest in politics 2,297 8.27 3.71 2 16Age 2,317 46.6 17.4 18 96

Note: Not all respondents answered every question, thus the N varies for each variable.


REFERENCES

Abelson, Robert P., Elliot Aronson, William J. McGuire, Theodore M. Newcomb,Milton J. Rosenberg, and Percy H. Tannenbaum. 1968. Theories of CognitiveConsistency: A Sourcebook. New York: Rand McNally.

Achen, Christopher H., and Larry Bartels. 2008. “Myopic Retrospection andParty Realignment in the Great Depression.” Working Paper: https://my.vanderbilt.edu/larrybartels/files/2011/12/Myopic_Retrospection.pdf

Achen, Christopher H., and Larry Bartels. 2016. Democracy for Realists: WhyElections Do Not Produce Responsive Government. Princeton, NJ: PrincetonUniversity Press.

Aghajani, Moji, Ilya M. Veer, Marie-José Van Tol, André Aleman, Mark A. VanBuchem, Dick J. Veltman, Serge A. R. B. Rombouts, and Mie J. Van Der Wee.2014. “Neuroticism and Extraversion are Associated with Amygdala Resting-State Functional Connectivity.” Cognitive, Affective, & BehavioralNeuroscience 14: 836–848.

Ahadi, Stephan A., and Mary K. Rothbart. 1994. “Temperament, Development, andthe Big Five.” In The Developing Structure of Temperament and Personality fromInfancy to Adulthood, edited by Charles F. Halverson Jr., Geldolph A. Kohnstamm,and Roy P. Martin, 69–90. New York: Lawrence Erlbaum Associates.

Albertson, Bethany, and Shana Kushner Gadarian. 2015.Anxious Politics: DemocraticCitizenship in a Threatening World. New York: Cambridge University Press.

Alford, John R., Carolyn L. Funk, and John R. Hibbing. 2005. “Are PoliticalOrientations Genetically Transmitted?” American Political Science Review 99:153–167.

Allport, GordonW. 1937. Personality: A Psychological Interpretation. New York:Holt.Allport, Gordon W., and Henry S. Odbert. 1936. “Trait-Names: A Psycho-Lexical

Study.” Psychological Monographs 47.


163

https://my.vanderbilt.edu/larrybartels/files/2011/12/Myopic_Retrospection.pdf

https://my.vanderbilt.edu/larrybartels/files/2011/12/Myopic_Retrospection.pdf

Althaus, Scott L., and Young Mie Kim. 2006. “Priming Effects in ComplexInformation Environments: Reassessing the Impact of News Discourse onPresidential Approval.” Journal of Politics 68: 960–976.

Andsager, Julie L., and H. Allen White. 2007. Self versus Others: Media Messagesand the Third Person Effect. Mahwah, NJ: Lawrence Erlbaum.

Ansorge, Ulrich, Wilfried Kunde, and Markus Kiefer. 2014. “Unconscious Visionand Executive Control: How Unconscious Processing and Conscious ActionControl Interact.” Consciousness and Cognition 27: 268–287.

Armstrong, Anna-Marie, and Zoltan Dienes. 2013. “Subliminal Understanding ofNegation: Unconscious Control by Subliminal Processing of Word Pairs.”Consciousness and Cognition 22: 1022–1040.

Aronson, Elliot. 1992. “The Return of the Repressed: Dissonance Theory Makes aComeback.” Psychological Inquiry 3: 303–311.

Asimov, Isaac. January 21, 1980. “A Cult of Ignorance.” Newsweek. p. 19.Bandyopadhyay, Siddhartha, and Bryan C. McCannon. 2014. “The Effect of the

Election of Prosecutors on Criminal Trials.” Public Choice 161: 141–156.Barber, Benjamin R. 1992. An Aristocracy of Everyone: The Politics of Education

and the Future of America. New York: Ballantine Books.Bargh, John A. 1997. “The Automaticity of Everyday Life.” In Advances in Social

Cognition, edited by Robert Wyer, 1–61. Mahwah, NJ: Erlbaum.Bargh, John A., and Ezequiel Morsella. 2008. “The Unconscious Mind.”

Perspectives on Psychological Science 3: 73–79.Barnea, Marina F., and Shalom H. Schwartz. 1998. “Values and Voting.” Political

Psychology 19: 17–40.Baron, Reuben, M., and David A. Kenny. 1986. “The Moderator-Mediator

Variable Distinction in Social Psychological Research: Conceptual, Strategic,and Statistical Considerations.” Journal of Personality and Social Psychology 51:1173–1182.

Barrick, Murray R., and Michael K. Mount. 1991. “The Big Five PersonalityDimensions and JobPerformance: AMeta-Analysis.”Personnel Psychology44: 1–26.

Bartels, Larry M. 1996. “Uniformed Voters: Information Effects in PresidentialElections.” American Journal of Political Science 40: 194–230.

Bartels, Larry M. 2002. “Beyond the Running Tally: Partisan Bias in PoliticalPerceptions.” Political Behavior 24: 117–150.

Bartels, Larry M. 2008. “The Irrational Electorate.” The Wilson Quarterly 32: 44–50.Beasley, Ryan K., and Mark R. Joslyn. 2001. “Cognitive Dissonance and Post-

Decision Attitude Change in Six Presidential Elections.” Political Psychology 22:521–540.

Beaty, Roger E., Scott Barry Kaufman, Mathias Benedek, Rex E. Jung, Yoed N.Kenett, Emanual Jauk, Aljoscha C. Neubauer, and Paul J. Silvia. 2016.“Personality and Complex Brain Networks: The Role of Openness to Experiencein Default Network Efficiency.”Human Brain Mapping 37: 773–779.

164 REFERENCES

Bechara, Antione, Antonio R. Damasio, Hanna Damasio, and Steven W.Anserson. 1994. “Insensitivity to Future Consequences following Damage tothe Human Prefrontal Cortex.” Cognition 50: 7–15.

Bechara, Antione, Hanna Damasio, Daniel Tranel, and Antonio R. Damasio.1997. “Deciding Advantageously before Knowing the AdvantageousStrategy.” Science 275: 1293–1295.

Bechara, Antione, Hanna Damasio, and Antionio R. Damasio. 2000. “Emotion,Decision Making and the Orbito Frontal Cortex.” Cerebral Cortex 10: 295–307.

Bell, Edward, and Christian Kandler. 2015. “The Origins of Party Identificationand Its Relationship to Political Orientations.” Personality and IndividualDifferences 83: 136–141.

Bennett, Stephen Earl. 1995. “Comparing Americans’ Political Information in1988 and 1992.” The Journal of Politics 57: 521–532.

Berdejó, Carlos, and Noam M. Yuchtman. 2013. “Crime, Punishment andPolitics: An Analysis of Political Cycles in Criminal Sentencing.” Review ofEconomics and Statistics 95: 741.

Berelson, Bernard R., Paul F. Lazarsfeld, and William N. McPhee. 1954. Voting: AStudy of Opinion Formation in A Presidential Campaign. Chicago: Universityof Chicago Press.

Berry, Kate. 2015. “How Judicial Elections Impact Criminal Cases.” BrennanCenter for Justice. Available: http://www.brennancenter.org/sites/default/files/publications/How_Judicial_Elections_Impact_Criminal_Cases.pdf

Bicchieri, Cristina, and Yoshitaka Fukui. 1999. “The Great Illusion: Ignorance,Information Cascades, and the Persistence of Unpopular Norms.” BusinessEthics Quarterly 9: 127–155.

Bisgaard, Martin. 2015. “Bias Will Find a Way: Economic Perceptions,Attributions of Blame, and Partisan-Motivated Reasoning during Crisis.” TheJournal of Politics 77: 849–860.

Bleidorn, Wiebke, Christian Kandler, and Avshalom Caspi. 2014. “The BehavioralGenetics of Personality Development in Adulthood—Classic, Contemporary,and Future Trends.” European Journal of Personality 28: 244–255.

Bode, Stefan, Anna Hanxi He, Chun Siong Soon, Robert Trampel, Robert Turner,and John-Dylan Haynes. 2011. “Tracking the Unconscious Generation of FreeDecisions Using Ultra-High Field FMRI.” Plos One 6: 1–13.

Bogg, Tim, and Brent W. Roberts. 2013. “The Case for Conscientiousness:Evidence and Implications for a Personality Trait Marker of Health andLongevity.” Annals of Behavioral Medicine 45: 278–288.

Bolsen, Toby, James N. Druckman, and Fay Lomax Cook. 2014. “The Influence ofPartisanMotivatedReasoning onPublicOpinion.”Political Behavior36: 235–262.

Bølstad, Jørgen, Elias Dinas, and Pedro Riera. 2013. “Tactical Voting and PartyPreferences: A Test of Cognitive Dissonance Theory.” Political Behavior 35:429–452.

REFERENCES 165

http://www.brennancenter.org/sites/default/files/publications/How_Judicial_Elections_Impact_Criminal_Cases.pdf

http://www.brennancenter.org/sites/default/files/publications/How_Judicial_Elections_Impact_Criminal_Cases.pdf

Bouchard Jr., Thomas J., and John C. Loehlin. 2001. “Genes, Evolution, andPersonality.” Behavior Genetics 31: 243–273.

Bouchard Jr., Thomas J., Nancy L. Segal, Auke Tellegen, Matt McGue, MargaretKeyes, and Robert Krueger. 2003. “Evidence for the Construct Validity andHeritability of the Wilson-Patterson Conservatism Scale: A Reared-Apart TwinsStudy of Social Attitudes.” Personality and Individual Differences 34: 959–969.

Brader, Ted. 2006. Campaigning for the Hearts and Minds How EmotionalAppeals in Political Ads Work. Chicago, IL: University of Chicago Press.

Brader, Ted. 2012. “Following the Party’s Lead: Party Cues, Policy Opinion, andthe Power of Partisanship in Three Multiparty Systems.” Comparative Politics44: 403–420.

Brady, Peter. 2012. The Tax Benefits and Revenue Costs of Tax Deferral(September). Washington, D.C.: Investment Company Institute.

Brandt, Mark J., Christine Reyna, John R. Chambers, Jarret T. Crawford, andGeoffrey Wetherell. 2014. “The Ideological-Conflict Hypothesis; IntoleranceAmong Both Liberals and Conservatives.” Current Directions in PsychologicalScience 23: 27–34.

Bratko, Denis, Tomas Chamorro-Premuzic, and Zrnka Saks. 2006. “Personalityand School Performance: Incremental Validity of Self- and Peer- Ratings overIntelligence.” Personality and Individual Differences 41: 131–142.

Britz, Juliane, Laura Diaz Hernandez, Tony Ro, and Christopher M. Michel.2014. “EEG-Microstate Dependent Emergence of Perceptual Awareness.”Frontiers in Behavior Neuroscience 8: Article 163.

Burman, Leonard E., Christopher Geissler, and Eric J. Toder. 2008. “How BigAre Total Individual Income Tax Expenditures, and Who Benefits fromThem?” American Economic Review: Papers & Proceedings 98: 79–83.

Burstein, Paul. 2003. “The Impact of Public Opinion on Public Policy: A Reviewand an Agenda.” Political Research Quarterly 56: 29–40.

Burstein, Paul. 2014. American Public Opinion, Advocacy, and Policy in Congress:What the Public Wants and What It Gets. New York: Cambridge University Press.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.1960. The American Voter. New York: John Wiley & Sons, Inc.

Canes-Wrone, Brandice, Tom S. Clark, and Jason P. Kelly. 2014. “JudicialSelection and Death Penalty Decisions.” American Political Science Review108: 22–39.

Canli, Turhan. 2004. “Functional Brain Mapping of Extraversion andNeuroticism: Learning from Individual Differences in Emotion Processing.”Journal of Personality 72: 1105–1132.

Caplan, Bryan. 2007. The Myth of the Rational Voter: Why Democracies Choose BadPolicies. Princeton, NJ: Princeton University Press.

Caprara, Gian Vittorio, Claudio Barbaranelli, and Philip G. Zimbardo. 1999.“Personality Profiles and Political Parties.” Political Psychology 20: 175–197.

166 REFERENCES

Caprara, Gian Vittorio, Donata Francescato, Minou Mebane, Roberta Sorace, andMichele Vecchione. 2010. “Personality Foundations of Ideological Divide: AComparison of Women Members of Parliament and Women Voters in Italy.”Political Psychology 31: 739–762.

Carney, Dana R., John T. Jost, Samuel D. Gosling, and Jeff Potter. 2008. “TheSecret Lives of Liberals and Conservatives: Personality Profiles, InteractionStyles, and Things They Leave Behind.” Political Psychology 29: 807–840.

Cattell, Raymond Bernard. 1943. “The Description of Personality: Basic TraitsResolved into Clusters.” Journal of Abnormal and Social Psychology 38: 476–506.

Cattell, Raymond Bernard. 1957. Personality and Motivation Structure andMeasurement. Oxford: World Book.

Cattell, Raymond Bernard, D. C. Rao, J. M. Schuerger, and D. S. Vaughan. 1981.“Unitary Personality Source Traits Analyzed for Heritability.” HumanHeredity 31: 261–275.

Chambers, John R., and Darya Melnyk. 2006. “Why Do I Hate Thee? ConflictMisperceptions and Intergroup Mistrust.” Personality and Social PsychologyBulletin 32: 1295–1311.

Chamooro-Premuzic, Tomas, and Adrian Furnham. 2003. “Personality Traits andAcademic Examination Performance.” European Journal of Personality 17:237–250.

Chamorro-Premuzic, Tomas, Adrian Furnham, and Phillip L. Ackerman. 2006.“Ability and Personality Correlates of General Knowledge.” Personality andIndividual Differences 41: 419–429.

Choi, Eunjung. 2010. “Economic Voting in Taiwan: The Significance ofEducation and Lifetime Economic Experiences.” Asian Survey 50: 990–1010.

Claassen, Ryan L., and Michael J. Ensley. 2016. “Motivated Reasoning and Yard-Sign-Stealing Partisans: Mine Is a Likable Rogue, Yours Is a DegenerateCriminal.” Political Behavior 38: 317–335.

Cobb-Clark, Deborah A., and Stefanie Schurer. 2012. “The Stability of Big-FivePersonality Traits.” Economic Letters 115: 11–15.

Cohen, Geoffrey L. 2003. “Party over Policy: The Dominating Impact of GroupInfluence on Political Beliefs.” Journal of Personality and Social Psychology 85:808–822.

Cohen, Jeffrey E. 2004. “Economic Perceptions and Executive Approval inComparative Perspective.”.” Political Behavior 26: 27–43.

Cohen, Alma, Alon Klement, and Zvika Neeman. 2015. “Judicial Decision Making:A Dynamic Reputation Approach.” Journal of Legal Studies 44: S133–S159.

Colas, Jaron T., and Po-Jang Hsieh. 2014. “Pre-Existing Brain States PredictAesthetic Judgments.” Human Brian Mapping 35: 2924–2934.

Collado, M. Dolores, and Iñigo Iturbe-Ormaetxe. 2010. “Public Transfers to thePoor: Is Europe Really Much More Generous than the United States?”International Tax and Public Finance 17: 662–685.

REFERENCES 167

Collins, Paul M. 2011. “Cognitive Dissonance on the U.S. Supreme Court.”Political Research Quarterly 64: 362–376.

Congleton, Roger D. 2007. “Informational Limits to Democratic Public Policy:The Jury Theorem, Yardstick Competition, and Ignorance.” Public Choice 132:333–352.

Connelly, Brian S., Deniz S. Ones, and Oleksandr S. Chernyshenko. 2014.“Introducing the Special Section on Openness to Experience: Review ofOpenness Taxonomies, Measurement, and Nomological Net.” Journal ofPersonality Assessment 96: 1–16.

Converse, Philip E. 1964. “The Nature of Belief Systems in Mass Publics.” InIdeology and Discontent, edited by David E. Apter, Glencoe, IL: Free Press.

Converse, Philip. 1972. “Change in the American Electorate.” In The HumanMeaning of Social Change, edited by Angus Campbell and Philip Converse,New York: Russell Sage Foundation.

Converse, Philip E. 2000. “Assessing the Capacity of Mass Electorates.” AnnualReview of Political Science 3: 331–353.

Cooper, Christopher A., Lauren Golden, and Alan Socha. 2013. “The Big FivePersonality Factors and Mass Politics.” Journal of Applied Social Psychology 43:68–82.

Cuik, David J., and Berwood A. Yost. 2016. “The Effects of Issue Salience, EliteInfluence, and Policy Content on Public Opinion.” Political Communication33: 328–345.

Dahl, Robert A. 1989. Democracy and Its Critics. New Haven, CT: YaleUniversity Press.

Davies, Patrick T., Dante Cicchetti, Rochelle F. Hentges, and Melissa L. Sturge-Apple. 2013. “The Genetic Precursors and the Advantageous andDisadvantageous Sequelae of Inhibited Temperament: An EvolutionaryPerspective.” Developmental Psychology 49: 2285–2300.

Davison, P.W. 1983. “The Third-Person Effect in Communication.” PublicOpinion Quarterly 47: 1–15.

De Moor, Marleen H. M., P. T. Costa, A. Terracciano, R. F. Krueger, E. J. C DeGeus, T. Toshiko, B.W.J.H. Penninx, T. Esko, P. A. F. Madden, J. Derringer,N. Amin, G. Willemsen, J.-J. Hottenga, M. A. Distel, M Uda, S. Sanna,P. Spinhoven, C.A. Hartman, P. Sullivan, A. Realo, J. Allik, A.C. Health,M. L. Pergadia, A. Agrawal, P. Lin, R. Grucza, T. Nutile, M. Ciullo, D.Rujescu, I. Giegling, B. Konte, E. Widen, D.L. Cousminer, J. G. Eriksson, A.Palotie, L. Peltonen, M. Luciano, A. Tenesa, G. Davies, L.M. Lopez, N.K.Hansell, S. E. Medland, L. Ferrucci, D. Schlessinger, G. W. Montgomery, M. J.Wright, Y.S. Aulchenko, A.C.J.W. Janssens, B. A. Oostra, A. Metspalu, G.R.Abecasis, I.J. Deary, K. Räikkönen, L. J. Bierut, N.G. Martin, C. M. Van Duijn,and D.I. Boomsma. 2012. “Meta-Analysis of Genome-Wide AssociationStudies for Personality.” Molecular Psychiatry 17: 337–349.

168 REFERENCES

De Moor, Marleen H. M. et al. 2015. “Meta-Analysis of Genome-WideAssociation Studies for Neuroticism, and the Polygenic Association withMajor Depressive Disorder.” JAMA Psychiatry 72: 642–650.

Delli Carpini, Michael X., and Scott Keeter. 1991. “Stability and Change in the U.S. Public’s Knowledge of Politics.” Public Opinion Quarterly 55: 583–612.

Delli Carpini, Michael X., and Scott Keeter. 1996. What Americans Know aboutPolitics and Why It Matters. New Haven, CT: Yale University Press.

DeLorme, Denise E., Jisu Huh, and Leonard N. Reid. 2006. “Perceived Effects ofDirect-To-Consumer (DTC) Prescription Drug Advertising on Self andOthers: A Third-Person Effect Study of Older Consumers.” Journal ofAdvertising 35: 47–65.

Demertzi, Athena, Audrey Vanhaudenhuyse, Serge Bredart, Lizette Heine, Caroldi Perri, and Steven Laureys. 2013. “Looking for the Self in PathologicalUnconsciousness.” Frontiers in Human Neuroscience 7: Article 538.

DeYoung, Colin G., Jacob B. Hirsh, Matthew S. Shane, Xenophon Papademetris,Nallakkandi Rajeevan, and Jeremy R. Gray. 2010. “Testing Predictions fromPersonality Neuroscience: Brain Structure and the Big Five.” PsychologicalScience 21: 820–828.

Dick, Philip K. 1968. Do Androids Dream of Electric Sheep? New York: Doubleday.Dietrich, Bryce, Scott Lasley, Jeffery J. Mondak, Megan L. Remmel, and Joel

Turner. 2012. “Personality and Legislative Politics: The Big Five TraitDimensions among U.S. State Legislators.” Political Psychology 33: 195–210.

Digman, John M., and Jillian Inouye. 1986. “Further Specification of the FiveRobust Factors of Personality.” Journal of Personality and Social Psychology 50:116–123.

Dijksterhuis, Ap. 2004. “Think Different: The Merits of Unconscious Thought inPreference Development and Decision Making.” Journal of Personality andSocial Psychology 87: 586–598.

Dijksterhuis, Ap, Maarten W. Bos, Loran F. Nordgren, and Rick B. Van Baaren.2006. “On Making the Right Choice: The Deliberation-Without-AttentionEffect.” Science 311: 1005–1007.

Dinesen, Peter Thisted, Asbjørn Sonne Nørgaard, and Robert Klemmensen.2014. “The Civic Personality and Democratic Citizenship.” Political Studies62: 134–152.

Dirilen-Gümüş, Özlem, Susan E. Cross, and Ali Dönmez. 2012. “Who Voted forWhom? Comparing Supporters of Obama and Mccain on Value Types andPersonality Traits.” Journal of Applied Social Psychology 42: 2879–2900.

Dodd, Michael D., Amanda Balzer, Carly M. Jacobs, Michael W. Gruszczynski,Kevin B. Smith, and John R. Hibbing. 2012. “The Political Left Rolls with theGood and the Political Right Confronts the Bad: Connecting Physiology andCognition to Preferences.” Philosophical Transactions of the Royal Society B:Biological Sciences 367: 640–649.

REFERENCES 169

Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harperand Row.

Draganski, Bogdan, Ferath Kherif, and Antoine Lutti. 2014. “ComputationalAnatomy of Studying Use-Dependent Brain Plasticity.” Frontiers in HumanNeuroscience 8: article number 380.

Druckman, James N., and Toby Bolsen. 2011. “Framing, Motivated Reasoning,and Opinions about Emergent Technologies.” Journal of Communication 61:659–688.

Druckman, James N., Jordan Fein, and Thomas J. Leeper. 2012. “A Source ofBias in Public Opinion Stability.” American Political Science Review 106:430–454.

Druckman, James N, Erik Peterson, and Rune Slothuus. 2013. “How ElitePartisan Polarization Affects Public Opinion Formation.” American PoliticalScience Review 107: 57–79.

Druckman, James N., and Arthur Lupia. 2016. “Preference Change inCompetitive Political Environments.” Annual Review of Political Science 19:13–31.

Duch, Raymond M. 2001. “A Developmental Model of HeterogeneousEconomic Voting in New Democracies.” American Political Science Review95: 895–910.

Duch, Raymond M., and Randy Stevenson. 2008. The Economic Vote: HowPolitical and Economic Institutions Condition Election Results. Cambridge:Cambridge University Press.

Dusso, Aaron. 2015. “Incorrect Voting in the 2012 U.S. Presidential Election:How Partisan and Economic Cues Fail to Help Low-Information Voters.”Electoral Studies 37: 50–62.

EIA: U.S. Energy Information Administration. 2014. “Crude Oil andLease Condensate Production at Highest Volume Since 1986.” http://Eia.gov. www.eia.gov/todayinenergy/detail.cfm?id=18831

Ellis, Christopher, and James A. Stimson. 2012. Ideology in America. New York:Cambridge University Press.

Enns, Peter K., Paul M. Kellstedt, and Gregory E. McAvoy. 2012. “TheConsequences of Partisanship in Economic Perceptions.” Public OpinionQuarterly 76: 287–310.

Erikson, Robert S, Michael B. Mackuen, and James A. Stimson. 2002. The MacroPolity. New York: Cambridge University Press.

Esterling, Kevin M., Michael A. Neblo, and David M. J. Lazer. 2011. “Means,Motive, and Opportunity in Becoming Informed about Politics: A DeliberativeField Experiment with Members of Congress and Their Constituents.” PublicOpinion Quarterly 75: 483–503.

Estlund, David M. 2008. Democratic Authority. Princeton, NJ: PrincetonUniversity Press.

170 REFERENCES

http://Eia.gov

http://Eia.gov

http://www.eia.gov/todayinenergy/detail.cfm?id=18831

Evans, Jonathan St. B. T., John Clibbens, Allegra Cattani, Anita Harris, and IanDennis. 2003. “Explicit and Implicit Processes in Multicue Judgment.”Memory & Cognition 31: 608–618.

Eysenck, H. J. 1947. Dimensions of Personality. London: Routledge & Kegan Paul.Eysenck, Hans J. 1990. “Genetic and Environmental Contributions to Individual

Differences: The Three Major Dimensions of Personality.” Journal ofPersonality 58: 245–261.

Fahimi, Mansour, Dale Kulp, and J. Michael Brick. 2009. “A Reassessment of List-Assisted RDD Methodology.” Public Opinion Quarterly 73: 751–760.

Faricy, Christopher. 2011. “The Politics of Social Policy in American: The Causesand Effects of Indirect versus Direct Social Spending.” Journal of Politics 73:74–83.

Faricy, Christopher, and Christopher Ellis. 2014. “Public Attitudes Toward SocialSpending in the United States: The Differences Between Direct Spending andTax Expenditures.” Political Behavior 36: 53–76.

Fazekas, Zoltan, and Levente Littvay. 2015. “The Importance of Context inthe Genetic Transmission of US Party Identification.” Political Psychology 36:361–377.

Fazio, Russell H., and Carol J. Williams. 1986. “Attitude Accessibility as aModerator of the Attitude-Perception and Attitude-Behavior Relations: AnInvestigation of the 1984 Presidential Election.” Journal of Personality andSocial Psychology 51: 505–514.

Feldman, Stanley, and John Zaller. 1992. “The Political Culture of Ambivalence:Ideological Responses to the Welfare State.” American Journal of PoliticalScience 36: 268–307.

Felton, Ryan. 2016. “Michigan’s Drug-Testing Welfare Program Has YieldedZero Positive Results so Far.” Theguardian.com. https://www.theguardian.com/us-news/2016/jun/21/michigan-welfare-drug-testing-program

Fernbach, Philip M., Todd Rogers, Craig R. Fox, and Steven A. Sloman. 2013.“Political Extremism Is Supported by an Illusion of Understanding.”Psychological Science 24: 939–946.

Festinger, Leon. 1957. A Theory of Cognitive Dissonance. Redwood City, CA:Stanford University Press.

Fields, James M., and Howard Schuman. 1976. “Public Beliefs about the Beliefs ofthe Public.” Public Opinion Quarterly 40: 427–448.

Fiorina, Morris P. 1981. Retrospective Voting in American National Elections.New Haven, CT: Yale University Press.

Fishkin, James S. 1991. Democracy and Deliberation: New Directions forDemocratic Reform. New Haven, CT: Yale University Press.

Fiske, Donald W. 1949. “Consistency of the Factorial Structures of PersonalityRatings from Different Sources.” Journal of Abnormal and Social Psychology 44:329–344.

REFERENCES 171

https://Theguardian.com

https://www.theguardian.com/us-news/2016/jun/21/michigan-welfare-drug-testing-program

https://www.theguardian.com/us-news/2016/jun/21/michigan-welfare-drug-testing-program

Flanigan, William, H., and Nancy H. Zingale. 2010. Political Behavior of theAmerican Electorate 12th Ed. CQ Press: Washington, D.C.

Forbes, Chad E., Joshua C. Poore, Frank Krueger, Aron K. Barbey, JeffreySolomon, and Jordan Grafman. 2014. “The Role of Executive Function andthe Dorsolateral Prefrontal Cortex in the Expression of Neuroticism andConscientiousness.” Social Neuroscience 9: 139–151.

Fowler, James H., and Christopher T. Dawes. 2008. “Two Genes Predict VoterTurnout.” Journal of Politics 70: 579–594.

Fowler, James H., and Christopher T. Dawes. 2013. “In Defense ofGenopolitics.” American Political Science Review 107: 362–374.

Fowler, James H., Laura A. Baker, and Christopher T. Dawes. 2008. “GeneticVariation in Political Participation.” American Political Science Review 102:233–248.

Fox, Nathan A., Heather A. Henderson, Peter J. Marshall, Kate E. Nichols,and Melissa M. Ghera. 2005. “Behavioral Inhibition: Linking Biology andBehavior within a Developmental Framework.” Annual Review of Psychology56: 235–262.

Free, Lloyd A., and Hadley Cantril. 1967. The Political Beliefs of Americans. NewBrunswick, NJ: Rutgers University Press.

Freitag, Markus, and Kathrin Ackermann. 2016. “Direct Democracy andInstitutional Trust: Relationships and Differences across Personality Traits.”Political Psychology 37: 707–723.

Friesen, Amanda, and Aleksander Ksiazkiewicz. 2015. “Do Political Attitudes andReligiousity Share a Genetic Path?” Political Behavior 37: 791–818.

Frey, Dieter. 1986. “Recent Research on Selective Exposure to Information.” InAdvances in Experimental Social Psychology, edited by L. Bekowitz, Vol. 19,pp. 41–80. New York: Academic Press.

Freyer, Felice J. “Health Care Law Debate Heats Up: MIT Professor StirsControversy.” The Boston Globe, July 25, 2014.

Furnham, Adrian, Andrew N. Christopher, Jeanette Garwood, and G. NeilMartin. 2007. “Approaches to Learning and the Acquisition of GeneralKnowledge.” Personality and Individual Differences 43: 1563–1571.

Gaines, Brian, James H. Kuklinski, Paul J. Quirk, Buddy Peyton, and JayVerkuilen. 2007. “A Disconfirmation Bias in the Evaluation of Arguments.”Journal of Personality and Social Psychology 71: 5–24.

Gainotti, Guido. 2012. “Unconscious Processing of Emotions and the RightHemisphere.” Neuropsychologia 50: 205–218.

Gallego, Aina, and Daniel Oberski. 2012. “Personality and Political Participation:The Mediation Hypothesis.” Political Behavior 34: 425–451.

Galston, William A. 2001. “Political Knowledge, Political Engagement and CivicEducation.” Annual Review of Political Science 4: 217–234.

172 REFERENCES

Gamm, Gerald H. 1989. The Making of New Deal Democrats: Voting Behaviorand Realignment in Boston, 1920–1940. Chicago: University of ChicagoPress.

Gélineau, Franҫios. 2007. “Presidents, Political Context, and EconomicAccountability: Evidence from Latin America.” Political Research Quarterly60: 415–428.

Gerber, Alan S., and Gregory A. Huber. 2010. “Partisanship, Political Control,and Economic Assessments.” American Journal of Political Science 54: 153–173.

Gerber, Alan S., Gregory A. Huber, David Doherty, Conor M. Dowling, andShang E. Ha. 2010. “Personality and Political Attitudes: Relationships AcrossIssue Domains and Political Contexts.” The American Political Science Review104: 111–133.

Gerber, Alan S., Gregory A. Huber, David Doherty, Conor M. Dowling, ConnorRaso, and Shang E. Ha. 2011a. “Personality Traits and Participation in PoliticalProcesses.” Journal of Politics 73: 692–706.

Gerber, Alan S., Gregory A. Huber, David Doherty, and Conor M. Dowling.2011b. “Personality Traits and the Consumption of Political Information.”American Political Research 39: 32–84.

Gerber, Alan S., Gregory A. Huber, David Doherty, and Conor M. Dowling.2012a. “Disagreement and the Avoidance of Political Discussion: AggregateRelationships and Differences across Personality Traits.” American Journal ofPolitical Science 56: 849–874.

Gerber, Alan S., Gregory A. Huber, David Doherty, and Conor M. Dowling.2012b. “Personality and the Strength and Direction of Partisan Identification.”Political Behavior 34: 653–688.

Gerber, Alan S., Gregory A. Huber, David Doherty, and Conor M. Dowling.2013a. “Assessing the Stability of Psychological and Political SurveyMeasures.” American Politics Research 41: 54–75.

Gerber, Alan S., Gregory A. Huber, David Doherty, Conor M. Dowling, andCostas Panagopoulos. 2013b. “Big Five Personality Traits and Responses toPersuasive Appeals: Results from Voter Turnout Experiments.” PoliticalBehavior 35: 687–728.

Giacino, Joseph T., Joseph J. Fins, Steven Laureys, and Nicholas Schiff. 2014.“Disorders of Consciousness after Acquired Brain Injury: The State of theScience.” Nature Reviews of Neurology 10: 99–114.

Gilens, Martin. 1999. Why Americans Hate Welfare: Race, Media, and the Politicsof Antipoverty Policy. Chicago: University of Chicago Press.

Goldberg, Lewis, R. 1990. “An Alternative ‘Description of Personality’: The Big-Five Factor Structure.” Personality Processes and Individual Differences 59:1216–1229.

REFERENCES 173

Goldberg, Lewis, R. 1992. “The Development of Markers for the Big-Five FactorStructure.” Psychological Assessment 4: 26–42.

Goldberg, Lewis, R. 1995. “What the Hell Took So Long? Donald W. Fiske andthe Big-Five Factor Structure.” In Personality Research Methods, and Theory: AFestschrift Honoring Donald W. Fiske, edited by Patrick E. Shrout and Susan T.Fiske, 29–44. New York: Psychology Press.

Gomez, Brad T., and J. Matthew Wilson. 2006. “Cognitive Heterogeneity andEconomic Voting: A Comparative Analysis of Four Democratic Electorates.”American Journal of Political Science 50: 127–145.

Goodnough, Abby, “Where Health Law Helps Voters but Saps Votes,” New YorkTimes 17 September, 2014.

Goren, Paul. 2005. “Party Identification and Core Political Values.” AmericanJournal of Political Science 49: 881–896.

Gosling, Samuel D., Peter J. Rentfrow, and William B. Swann Jr. 2003. “A VeryBrief Measure of the Big-Five Personality Domains.” Journal of Research inPersonality 37: 504–528.

Gozzi, Marta, Giovanna Zamboni, Frank Krueger, and Jordan Grafman. 2010.“Interest in Politics Modulates Neural Activity in the Amygdala and VentralStraitum.” Human Brain Mapping 31: 1763–1771.

Grant, Bridget F., and Deborah A. Dawson. 1996. “Alcohol and Drug Use,Abuse, and Dependence among Welfare Recipients.” American Journal ofPublic Health 86: 1450–1454.

Gray, Jeffrey, A. 1985. “The Neuropsychology of Anxiety.” In Stress and Anxiety,edited by C.D. Spielberger, 201–227. Washington, D.C.: Hemisphere.

Gray, Jeffrey A., and McNaughton. Neil. 2000. The Neuropsychology of Anxiety: AnEnquiry into the Functions of the Septo-Hippocampal System. 2nd New York:Oxford University Press.

Graziano, William, and Nancy H. Eisenberg. 1997. “Agreeableness: ADimension of Personality.” In Handbook of Personality Psychology, edited byRobert Hogan, John Johnson, and Stephen Briggs, 795–824. San Diego, CA:Academic Press.

Grimm, Simone, Florian Schubert, Maren Jaedke, Juergen Gallinat, and MalekBajbouj. 2012. “Prefrontal Cortex Glutamate and Extraversion.” SocialCognitive and Affective Neuroscience 7: 811–818.

Groenendyk, Eric W., and Antoine J. Banks. 2014. “Emotional Rescue: HowAffect Helps Partisans Overcome Collective Action Problems.” PoliticalPsychology 35: 359–378.

Gruszczynski, Michael W., Amanda Balzer, Carly M. Jacobs, Kevin B. Smith, andJohn R. Hibbing. 2013. “The Physiology of Political Participation.” PoliticalBehavior 35: 135–152.

Ha, Shang E., and Richard R. Lau. 2015. “Personality Traits and Correct Voting.”American Politics Research 43: 975–998.

174 REFERENCES

Ha, Shang E., Seokho Kim, and Se Hee Jo. 2013. “Personality Traits and PoliticalParticipation: Evidence from South Korea.” Political Psychology 34: 511–532.

Hacker, Jacob S. 2002. The Divided Welfare State: The Battle Over Public andPrivate Social Benefits in the United States. New York: Cambridge UniversityPress.

Haggard, Patrick. 2008. “Human Volition: Toward a Neuroscience of Will.”Nature Reviews Neuroscience 9: 934–946.

Haggard, Patrick, and Martin Eimer. 1999. “On the Relation between BrainPotentials and the Awareness of Voluntary Movement.” Experimental BrianResearch 126: 128–133.

Hall, Richard L., and Alan V. Deardorff. 2006. “Lobbying as Legislative Subsidy.”American Political Science Review 100: 69–84.

Harmon-Jones, Eddie, David M. Amodio, and Harmon-Jones. Cindy. 2009.“Action-Based Model of Dissonance: A Review, Integration, and Expansion.”Advances in Experimental Social Psychology 41: 119–166.

Hatemi, Peter K., and McDermott. Rose. 2012. “Broadening PoliticalPsychology.” Political Psychology 33: 11–25.

Hatemi, Peter K., and Brad Verhulst. 2015. “Political Attitudes DevelopIndependently of Personality Traits.” Plos One 10: 1–24.

Hatemi, Peter K., John R. Hibbing, Sarah E Medland, Matthew C. Keller, John R.Alford, Kevin B. Smith, Nicholas G.Martin, and Lindon J. Eaves. 2010. “Not byTwins Alone: Using the Extended FamilyDesign to InvestigateGenetic Influenceon Political Behavior.” American Journal of Political Science 54: 798–814.

Hatemi, Peter K., Nathan Gillespie, Lindon Eaves, Brion Maher, Bradley Webb,Andrew Heath, Sarah Medland, David Smyth, Harry Beeby, Scott Gordon,Grant Montgomery, Ghu Zhu, Enda Byrne, and Nicholas Martin. 2011. “AGenome-Wide Analysis of Liberal and Conservative Political Attitudes.”Journal of Politics 73: 1–15.

Hatemi, Peter K., Sarah E. Medland, Robert Klemmensen, Sven Oskarsson,Levente Littvay, Christopher T. Dawes, Brad Verhulst, Rose McDermott,Asbjørn Sonne Nøgaard, Casey A. Klofstad, Kaare Christensen, MagnusJohannesson, Patrick K. E. Magnusson, Lindon J. Eaves, and Nicholas G.Martin. 2014. “Genetic Influences on Political Ideologies: Twin Analyses of19 Measures of Political Ideology from Five Democracies and Genome-WideFindings from Three Populations.” Behavior Genetics 44: 282–294.

Haynes, John-Dylan. 2011. “Decoding and Predicting Intentions.” Annals of theNew York Academy of Science 1224: 9–21.

Heider, Fritz. 1958. The Psychology of Interpersonal Relations. New York: Wiley.Helms, Ann Doss. “Who Should Save Sight of S.C. Man Who Can’t Afford

Surgery?” The Charlotte Observer, May 12, 2015. http://www.charlotteobserver.com/news/business/health-care/health-care-challenge-blog/article20696283.html

REFERENCES 175

http://www.charlotteobserver.com/news/business/health-care/health-care-challenge-blog/article20696283.html



Henry, P.J., and Christine Reyna. 2007. “Value Judgments: The Impact ofPerceived Value Violations on American Political Attitudes.” PoliticalPsychology 28: 273–298.

Herbet, Guillaume, Gilles Lafargue, Nicolas Manjot De Champfleur, Sylvie Moritz-Gasser, Emmanuelle Le Bars, Francois Bonnetblanc, and Hugues Duffau. 2014.“Disrupting Posterior Cingulate Connectivity Disconnects Consciousness fromthe External Environment.” Neuropsychologia 56: 239–244.

Hibbing, John R., and Kevin B Smith. 2007. “Introduction: The Biology ofPolitical Behavior: An Introduction.” Annals of the American Academy ofPolitical and Social Science 614: 6–14.

Hibbing, Matthew V., Melinda Ritchie, and Mary R. Anderson. 2011.“Personality and Political Discussion.” Political Behavior 33: 601–624.

Hibbing, John R., Kevin B. Smith, and John R. Alford. 2014a. “Differences inNegativity Bias Underlie Variations in Political Ideology.” Behavioral andBrain Sciences 37: 297–350.

Hibbing, John R., Kevin B. Smith, and John R. Alford. 2014b. Predisposed:Liberals, Conservatives, and the Biology of Political Differences. New York:Routledge.

Hillygus, D. Sunshine. 2005. “TheMissing Link: Exploring the Relationship betweenHigher Education and Political Engagement.” Political Behavior 27: 25–47.

Hochschild, Jennifer L., and Katherine Levine Einstein. 2015. Do Facts Matter?Information and Misinformation in American Politics. Norman, OK:University of Oklahoma Press.

Hofstee, Willem K. B., Boele de Raad, and Lewis R. Goldberg. 1992. “Integrationof the Big Five and Circumplex Approaches to Trait Structure.” Journal ofPersonality and Social Psychology 63: 146–163.

Hogan, Alexandra M., Elinor L. Butterfield, Luke Phillips, and Julia A. Hadwin.2007. “Brain Response to Unexpected Novel Noises in Children with Low andHigh Trait Anxiety.” Journal of Cognitive Neuroscience 19: 25–31.

Holbrook, Allyson L., and Jon A. Krosnick. 2005. “Meta-Psychological versusOperative Measures of Ambivalence.” In Ambivalence and the Structureof Political Opinion, edited by Stephen C Craig and Michael D. Martinez,73–103. New York: Palgrave Macmillan.

Hopwood, Christopher J., Leslie C. Morey, M. Brent Donnellan, Douglas B.Samuel, Carlos M. Grilo, Thomas H. McGlashan, M. Tracie Shea, Mary C.Zanarini, John G. Gunderson, and Andrew E. Skodol. 2013. “Ten-YearRank-Order Stability of Personality Traits and Disorders in a ClinicalSample.” Journal of Personality 81: 335–334.

Howard, Christopher. 1997. The Hidden Welfare State: Tax Expenditures andSocial Policy in the United States. Princeton, NJ: Princeton University Press.

Howard, Christopher. 2007. The Welfare State Nobody Knows: Debunking Mythsabout U.S. Social Policy. Princeton, NJ: Princeton University Press.

176 REFERENCES

Huang, Yu-Feng, Chu Siong Soon, O’Dhaniel A. Mullette-Gillman, and Po-JangHsieh. 2014. “Pre-Existing Brain States Predict Risky Choices.” Neuroimage101: 466–472.

Hyman, Herbert H., and Paul B. Sheatsley. 1947. “Some Reasons WhyInformation Campaigns Fail.” Public Opinion Quarterly 11: 412–423.

Ikeda, Haruka, Eiji Ikeda, Kazumasa Shiozaki, and Yoshio Hirayasu. 2014.“Association of the Five-Factor Personality Model with Prefrontal ActivationDuring Frontal Lobe Task Performance Using Two-Channel Near-InfraredSpectroscopy.” Psychiatry and Clinical Neurosciences 68: 752–758.

Ipsos Mori. 2014. “Perceptions are Not Reality: Things the World Gets Wrong.”Ipsos-mori.com. https://www.ipsos-mori.com/researchpublications/researcharchive/3466/Perceptions-are-not-reality-Things-the-world-gets-wrong.aspx

Iyengar, Shanto, and Donald R. Kinder. 1987. News That Matters: Television andAmerican Opinion. Chicago, IL: University of Chicago Press.

Izatt, Gregory, Julien Dubois, Nathan Faivre, and Chistof Koch. 2014. “A DirectComparison of Unconscious Face Processing under Masking and InterocularSuppression.” Frontiers in Psychology 5: Article 659. doi: 10.3389/fpsyg.2014.00659.

Jackson, John E. 1975. “Issues, Party Choices and Presidential Votes.” AmericanJournal of Political Science 19: 161–185.

Jacobs, Kate E., Dion Szer, and John Roodenburg. 2012. “The Moderating Effectof Personality on the Accuracy of Self-Estimates of Intelligence.” Personalityand Individual Differences 52: 744–749.

Jacoby, William G. 1994. “Public Attitudes Toward Government Spending.”American Journal of Political Science 38: 336–361.

Jacoby, William. 2000. “Issue Framing and Public Opinion on GovernmentSpending.” American Journal of Political Science 44: 750–767.

James, Lawrence R., and Jeanne M. Brett. 1984. “Mediators, Moderators, andTests for Mediation.” Journal of Applied Psychology 69: 307–321.

Jensen, Thomas. 2009. “Projection Effects and Strategic Ambiguity in ElectoralCompetition.” Public Choice 141: 213–232.

Jensen-Campbell, Lauri, and William G. Graziano. 2001. “Agreeableness as aModerator of Interpersonal Conflict.” Journal of Personality 69: 323–362.

Jensen-Campbell, Lauri, Jennifer M. Knack, Amy M. Waldrip, and Shaun D.Campbell. 2007. “Do Big Five Personality Traits Associated withSelf-Control Influence the Regulation of Anger and Aggression?” Journal ofResearch in Personality 41: 403–424.

Jerit, Jennifer, and Jason Barabas. 2012. “Partisan Perceptual Bias and theInformation Environment.” Journal of Politics 74: 672–684.

Joel, Samantha, Caitlin M. Burton, and Jason E. Plaks. 2014. “ConservativesAnticipate and Experience Stronger Emotional Reactions to NegativeOutcomes.” Journal of Personality 82: 32–43.

REFERENCES 177

http://Ipsos-mori.com

https://www.ipsos-mori.com/researchpublications/researcharchive/3466/Perceptions-are-not-reality-Things-the-world-gets-wrong.aspx

https://www.ipsos-mori.com/researchpublications/researcharchive/3466/Perceptions-are-not-reality-Things-the-world-gets-wrong.aspx

http://dx.doi.org/10.3389/fpsyg.2014.00659

http://dx.doi.org/10.3389/fpsyg.2014.00659

John, Oliver P., Richard W. Robins, and Lawrence A. Pervin. 2008. Handbook ofPersonality: Theory and Research. 3rd New York: Guilford Press.

Joint Committee on Taxation, “Estimates of Federal Tax Expenditures for FiscalYears 2014–2018.” By committee staff. August 5, 2014. Washington, D.C.Report No. JCX-97-14.

Joseph, Jay. 2013. “TheUse of theClassical TwinMethod in the Social and BehavioralScience: The Fallacy Continues.” Journal of Mind and Behavior 34: 1–39.

Jost, John T., and David M. Amodio. 2012. “Political Ideology as MotivatedSocial Cognition: Behavioral and Neuroscientific Evidence.” Motivation andEmotion 36: 55–64.

Jost, John T., Jack Glaser, Arie W. Kruglanski, and Frank J. Sulloway. 2003a.“Political Conservatism as Motivated Social Cognition.” Psychological Bulletin129: 339–375.

Jost, John T., Jack Glaser, Arie W. Kruglanski, and Frank J. Sulloway. 2003b.“Exceptions that Prove the Rule: Using a Theory of Motivated SocialCognition to Account for Ideological Incongruities and Political Anomalies.”Psychological Bulletin 129: 383–393.

Jost, John T., H. Hannah Nam, David M. Amodio, and Jay J. Van Bavel. 2014.“Political Neuroscience: The Beginning of a Beautiful Friendship.” PoliticalPsychology 35: 3–42.

Judd, Charles M., and David A. Kenny. 1981. “Process Analysis: EstimatingMediation in Treatment Evaluations.” Evaluation Review 5: 602–619.

Kahan, Dan M. 2013. “Ideology, Motivated Reasoning, and CognitiveReflection.” Judgment and Decision Making 8: 407–424.

Kahan, Dan M., Ellen Peters, Maggie Wittlin, Paul Slovic, Lisa LarrimoreOuellette, Donald Braman, and Gregroy N. Mandel. 2012. “The PolarizingImpact of Science Literacy and Numeracy on Perceived Climate ChangeRisks.” Nature Climate Change 2: 732–735.

Kahan, Dan M., Ellen Peters, Erica Cantrell Dawson, and Paul Slovic. 2013.“Motivated Numeracy and Enlightened Self-Government.” CulturalCognition Project Working Paper No. 116.

Kaiser Family Foundation. 2013. “Kaiser Health Tracking Poll: March 2013.” Kff.org. http://kff.org/health-reform/poll-finding/march-2013-tracking-poll/

Kaiser Family Foundation. 2014. “Kaiser Health Policy Tracking Poll: December2014.” Kff.org. http://kff.org/health-reform/poll-finding/kaiser-health-policy-tracking-poll-december-2014/

Kammrath, Lara K., and Abigail A. Scholer. 2011. “The Pollyanna Myth: HowHighly Agreeable People Judge Positive and Negative Relational Acts.”Personality and Social Psychology Bulletin 37: 1172–1184.

Kanai, Ryota, Tom Feilden, Colin Firth, and Geraint Rees. 2011. “PoliticalOrientations are Correlated with Brain Structure in Young Adults.” CurrentBiology 21: 677–680.

178 REFERENCES

http://Kff.org

http://Kff.org

http://kff.org/health-reform/poll-finding/march-2013-tracking-poll/

http://Kff.org

http://kff.org/health-reform/poll-finding/kaiser-health-policy-tracking-poll-december-2014/

http://kff.org/health-reform/poll-finding/kaiser-health-policy-tracking-poll-december-2014/

Kandler, Christian. 2012. “Nature and Nurture in Personality Development: TheCase of Neuroticism and Extraversion.” Current Directions in PsychologicalScience 21: 290–296.

Kaplan, Jonas T., Joshua Freedman, andMarco Iacoboni. 2007. “Us versus Them:Political Attitudes and Party Affiliation Influence Neural Response to Faces ofPresidential Candidates.” Neuropsychologia 45: 55–64.

Kapogiannis, Dimitrios, Angelina Sutin, Christos Davatzikos, Paul Costa Jr., andSusan Resnick. 2013. “The Five Factors of Personality and Regional CorticalVariability in the Baltimore Longitudinal Study of Aging.” Human BrainMapping 34: 2829–2840.

Keith, Bruce, E., David B. Magleby, Candice J. Nelson, Elizabeth A. Orr, andMark C. Westlye. 1992. The Myth of the Independent Voter. Berkeley CA:University of California Press.

Key Jr., V.O. 1966. The Responsible Electorate: Rationality in Presidential Voting1936–1960. Cambridge, MA: Harvard University Press.

Killian, Mitchell, Ryan Schoen, and Aaron Dusso. 2008. “Keeping up with theJoneses: The Interplay of Personal and Collective Evaluations in VoterTurnout.” Political Behavior 30: 323–340.

Kim, Bo-Hye, Han-Na Kim, Seung-Ju Roh, Mi Kyeong Lee, Sarah Yang,Seung Ku Lee, Yeon-Ah Sung, Hye Won Chung, Nam H. Cho, CholShin, Joohon Sung, and Hyung-Lae Kim. 2015. “GWA Meta-Analysisof Personality in Korean Cohorts.” Journal of Human Genetics 60:455–460.

Kinder, Donald R. 1998. “Opinion and Action in the Realm of Politics.” In D. T.Gilbert, S.T. Fiske, and G. Lindzey, The Handbook of Social Psychology,778–867, Boston: McGraw-Hill.

Kinder, Donald R., and Shanto Iyengar. 1987. News that Matters: Television andAmerican Opinion. Chicago, IL: University of Chicago Press.

Kingdon, John W. 1989. Congressmen’s Voting Decisions. 3rd Ann Arbor, MI:University of Michigan Press.

Kirzinger, Ashley, Bryan Wu, and Mollyann Brodie. 2016. “Kaiser HealthTracking Poll: September 2016,” kff.org. http://kff.org/health-costs/report/kaiser-health-tracking-poll-september-2016/

Klas, Mary Ellen “Court Rejects Florida Gov. Rick Scott’s Drug Testing of WelfareApplicants,” Miami Herald, December 3, 2014.

Knutson, Kristine M., Jacqueline N. Wood, Maria V. Spampinato, and JordanGrafman. 2007. “Politics on the Brain: An Fmri Investigation.” SocialNeuroscience 1: 25–40.

Kramer, Gerald H. 1971. “Short-Term Fluctuations in U.S. Voting Behavior:1896–1964.” American Political Science Review 65: 131–143.

Krastev, Sekoul, Joseph T. McGuire, Denver McNeney, Joseph W. Kable,Dietlind Stolle, Elisabeth Gidengil, and Lesley K. Fellows. 2016. “Do

REFERENCES 179

http://kff.org

http://kff.org/health-costs/report/kaiser-health-tracking-poll-september-2016/

http://kff.org/health-costs/report/kaiser-health-tracking-poll-september-2016/

Political and Economic Choices Rely on Common Neural Substrates? ASystematic Review of the Emerging Neuropolitics Literature.” Frontiers inPsychology 7: 1–10.

Krueger, Robert F., and Wendy Johnson. 2008. “Behavioral Genetics andPersonality: A New Look at the Integration of Nature and Nurture.”Handbook of Personality: Theory and Research 3rd ed., edited by Oliver P.John, Richard W. Robins, and Lawrence A Pervin, New York: Guilford Press.

Kruger, Justin, and David Dunning. 1999. “Unskilled and Unaware of It:How Difficulties in Recognizing One’s Own Incompetence Lead toInflated Self-Assessments.” Journal of Personality and Social Psychology77: 1121–1134.

Kühn, Simone, Simone M. Ritter, Barbara C. N. Müller, Rick B. Van Baaren,Marcel Brass, and Ap Dijksterhuis. 2013. “The Importance of the DefaultMode Network in Creativity—A Structural MRI Study.” Journal of CreativeBehavior 48: 152–163.

Kuklinski, James H., and Paul J. Quirk. 2000. “Reconsidering the RationalPublic: Cognition, Heuristics, and Mass Opinion.” In Elements of Reason:Cognition, Choice, and the Bounds of Rationality, edited by Arthur Lupia,Mathew D. McCubbins, and Samuel L. Popkin, New York: CambridgeUniversity Press.

Kunda, Ziva. 1987. “Motivated Inference: Self-Serving Generation andEvaluation of Causal Theories.” Journal of Personality and Social Psychology53: 636–647.

Kunda, Ziva. 1990. “The Case for Motivated Reasoning.” Psychological Bulletin108: 480–498.

Kunst-Wilson, William Raft, and Robert B. Zajonc. 1980. “AffectiveDiscrimination of Stimuli that Cannot Be Recognized.” Science 207: 557–558.

Ladha, Krishna K. 1992. “The Condorcet Jury Theorem, Free Speech, andCorrelated Votes.” American Journal of Political Science 36: 617–624.

Lau, Richard R., and David P. Redlawsk. 1997. “Voting Correctly.” TheAmerican Political Science Review 91: 585–598.

Lau, Richard R., and David P. Redlawsk. 2006. How Voters Decide: InformationProcessing During Election Campaigns. New York: Cambridge University Press.

Lau, Richard R., David J. Andersen, and David P. Redlawsk. 2008. “AnExploration of Correct Voting in Recent U.S. Presidential Elections.”American Journal of Political Science 52: 395–411.

Lavine, Howard, Milton Lodge, and Kate Freitas. 2005. “Threat,Authoritarianism, and Selective Exposure to Information.” Political Psychology26: 219–244.

Lavine, Howard G., Christopher D. Johnston, and Marco R. Steenbergen. 2012.The Ambivalent Partisan: How Critical Loyalty Promotes Democracy. New York:Oxford University Press.

180 REFERENCES

Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet. 1944. The People’sChoice: How the Voter Makes up His Mind in a Presidential Campaign. NewYork: Columbia University Press.

Lebo, Matthew J., and Daniel Cassino. 2007. “The Aggregated Consequences ofMotivated Reasoning and the Dynamics of Partisan Presidential Approval.”Political Psychology 28: 719–746.

Leeper, Thomas J., and Rune Slothuus. 2014. “Political Parties, MotivatedReasoning, and Public Opinion Formation.” Advances in Political Psychology35: 129–156.

Lei, Xu, Tianliang Yang, and Taoyu Wu. 2015. “Functional Neuroimaging ofExtraversion-Introversion.” Neuroscience Bulletin 31: 663–675.

Lenz, Gabriel S. 2012. Follow the Leader? How Voters Respond to Politicians’Policies and Performance. Chicago: The University of Chicago Press.

Lepianka, Dorota, Wim Van Oorschot, and John Gelissen. 2009. “PopularExplanations of Poverty: A Critical Discussion of Empirical Research.”Journal of Social Policy 38: 421–438.

Letkiewicz, Jodi C., and Jonathan J. Fox. 2014. “Conscientiousness, FinancialLiteracy, and Asset Accumulation of Young Adults.” Journal of ConsumerAffairs 48: 247–300.

Levendusky, Matthew. 2009. The Partisan Sort: How Liberals BecomeDemocrats and Conservatives Become Republicans. Chicago: University ofChicago Press.

Levy, Jenna. 2014. “In U.S., Uninsured Rate Holds at 13.4%.” gallup.com.http://www.gallup.com/poll/178100/uninsured-rate-holds.aspx

Lewandowsky, Stephan, Ullrich K. H. Ecker, Colleen M. Seifert, NorbertSchwarz, and John Cook. 2012. “Misinformation and Its Correction:Continued Influence and Successful Debiasing.” Psychological Science in thePublic Interest 13: 106–131.

Lewandowsky, Stephan, Klaus Oberauer, and Gilles E. Gignac. 2013. “NASA FakedtheMoon Landing—Therefore, (Climate) Science Is a Hoax: An Anatomy of theMotivated Rejection of Science.” Psychological Science 24: 622–633.

Lewis-Beck, Michael S. 1988. Economics and Elections. Ann Arbor: University ofMichigan Press.

Lewis-Beck, Michael S., and Martin Paldam. eds. 2000. “Economics andElections: An Introduction.” Electoral Studies 19(Special Issue): 113–121.

Lewis-Beck, Michael S., and Mary Stegmaier. 2000. “Economic Determinants ofElectoral Outcomes.” Annual Review of Political Science 3: 183–219.

Lewis-Beck, Michael S., Richard Nadeau, and Angelo Elias. 2008. “Economics,Party, and the Vote: Causality Issues and Panel Data.” American Journal ofPolitical Science 52: 84–95.

Libet, Benjamin, Curtis A. Gleason, Elwood W. Wright, and Dennis K. Pearl.1983. “Time of Conscious Intention to Act in Relation to Onset of Cerebral

REFERENCES 181

http://gallup.com

http://www.gallup.com/poll/178100/uninsured-rate-holds.aspx

Activity (Readiness-Potential): The Unconscious Initiation of a FreelyVoluntary Act.” Brain 106: 623–642.

Libet, Benjamin. 1985. “Unconscious Cerebral Initiative and the Role ofConscious Will in Voluntary Action.” Behavioral and Brain Science 8: 529–566.

Lodge, Milton, and Charles S. Taber. 2013. The Rationalizing Voter. New York:Cambridge University Press.

Lord, Charles G., Lee Ross, and Mark R. Lepper. 1979. “Biased Assimilation andAttitude Polarization: The Effects of Prior Theories on SubsequentlyConsidered Evidence.” Journal of Personality and Social Psychology 37: 2098–2109.

Lövdén, Martin, Lars Bäckman, Ulman Lindenberger, Sabine Schaefer, andFlorian Schmiedek. 2010. “A Theoretical Framework for the Study of AdultCognitive Plasticity.” Psychological Bulletin 136: 659–676.

Luciano, Michelle, Jennifer E. Huffman, Alejandro Arias-Vásquez, Anna A.E.Vinkhuyzen, Christel M. Middeldorp, Ina Giegling, Antony Payton, GailDavies, Lina Zgaga, Joost Janzing, Xiayi Ke, Tessel Galesloot, Annette M.Hartmann, William Ollier, Albert Tenesa, Caroline Hayward, MaaikeVerhagen, Grant W. Montgomery, Jouke-Jan Hottenga, Bettina Konte, JohnM. Starr, Veronique Vitart, Pieter E. Vos, Pamela A.F. Madden, GonnekeWillemsen, Heike Konnerth, Michael A. Horan, David J. Porteous, HarryCampbell, Sita H. Vermeulen, Andrew C. Heath, Alan Wright, OzrenPolasek, Sanja B. Kovacevic, Nicholas D. Hastie, Barbara Franke, Dorret I.Boomsma, Nicholas G. Martin, Dan Rujescu, James F. Wilson, Jan Buitelaar,Neil Pendleton, Igor Rudan, and Ian J. Deary. 2012. “Genome-WideAssociation Uncovers Shared Genetic Effects Among Personality Traits andMood States.” American Journal of Medical Genetics Part B: NeuropsychiatricGenetics 159B(6): 684–695.

MacKuen, Michael, Jennifer Wolak, Luke Keele, and George E. Marcus. 2010.“Civic Engagements: Resolute Partisanship or Reflective Deliberation.”American Journal of Political Science 54: 440–458.

Malhotra, Neil, and Yotam Margalit. 2010. “Short-Term Communication Effectsor Longstanding Dispositions? the Public’s Response to the Financial Crisis of2008.” The Journal of Politics 72: 852–867.

Mann, Frank D., Daniel A. Briley, Elliot M. Tucker-Drob, and Paige K. Harden.2015. “A Behavioral Genetic Analysis of Callous-Unemotional Traits and BigFive Personality in Adolescence.” Journal of Abnormal Psychology 124: 982–993.

Marcus, George E., W. Russell Neuman, and Michael MacKuen. 2000.Affective Intelligence and Political Judgment. Chicago, IL: University ofChicago Press.

182 REFERENCES

Marcus, George E., W. Russell Neuman, Michael MacKuen, and Ann N. Crigler.2007. The Affect Effect: Dynamics of Emotion in Political Thinking andBehavior. Chicago, IL: University of Chicago Press.

Marcus, George E. 2013. Political Psychology: Neuroscience, Genetics, and Politics.New York: Oxford University Press.

Markus, Gregory. 1988. “The Impact of Personal and National EconomicConditions on the Presidential Vote: A Pooled Cross-Sectional Analysis.”American Journal of Political Science 32: 137–154.

Masci, David. “States that Allow Same-Sex Marriage Also Provide Protections forReligious Groups and Clergy Who Oppose It.” Pew Research Center,November 20, 2013.

Maskivker, Julia. 2016. “An Epistemic Justification for the Obligation to Vote.”Critical Review 28: 24–247.

Mattila, Mikko, Hanna Wass, Peter Söderlund, Sami Fredriksson, Päivi Fadjukoff,and Katja Kokko. 2011. “Personality and Turnout: Results from the FinnishLongitudinal Studies.” Scandinavian Political Studies 34: 287–306.

McAdams, Dan P., and Bradley D. Olson. 2010. “Personality Development:Continuity and Change over the Life Course.” Annual Review of Psychology61: 517–542.

McCannon, Bryan. 2013. “Prosecutor Elections, Mistakes, and Appeals.” Journalof Empirical Legal Studies 10: 696–714.

McCrae, Robert R. 1996. “Social Consequences of Experiential Openness.”Psychological Bulletin 120: 323–337.

McCrae, Robert R., and Paul T. Costa Jr. 1983. “Joint Factors in Self-Reports andRatings: Neuroticism, Extraversion and Openness to Experience.” Personalityand Individual Differences 4: 245–255.

McCrae, Robert R., and Oliver P. John. 1992. “An Introduction to the Five-Factor Model and Its Applications.” Journal of Personality 60: 175–215.

McGregor, R. Michael. 2013. “Cognitive Dissonance and Political Attitudes: TheCase of Canada.” Social Science Journal 50: 168–176.

McKelvey, Richard D., and Peter C. Ordeshook. 1985. “Sequential Elections withLimited Information.” American Journal of Political Science 29: 480–512.

McKelvey, Richard D., and Peter C. Ordeshook. 1986. “Information, ElectoralEquilibria, and the Democratic Ideal.” Journal of Politics 48: 909–937.

Mendelberg, Tali. 2001. The Race Card: Campaign Strategy, Implicit Messages,and the Norma of Equality. Princeton, NJ: Princeton University Press.

Mettler, Suzanne. 2008. “Reconstituting the Submerged State: The Challengesof Social Policy Reform in the Obama Era.” Perspectives on Politics 8:803–824.

Mettler, Suzanne. 2011. The Submerged State: How Invisible Government PoliciesUndermine American Democracy. Chicago, IL: University of Chicago Press.

REFERENCES 183

Michigan Government. 2014. “Gov. Rick Snyder: Pilot Program Intended toRemove Barriers to Employment, Independence.” Michigan.gov. http://www.michigan.gov/snyder/0,4668,7-277-57577-344374–,00.html

Miller, Joanne M., and Jon A. Krosnick. 2000. “News Media Impact on theIngredients of Presidential Evaluations: Politically Knowledgeable Citizensare Guided by a Trusted Source.” American Journal of Political Science 44:295–309.

Miller, Daniel A., Tracey Cronin, Amber L. Garcie, and Nyla R. Branscombe.2009. “The Relative Impact of Anger and Efficacy on Collective Action IsAffected by Feelings of Fear.” Group Processes & Intergroup Relations 12:445–462.

Molden, Daniel C. 2014. “Understanding Priming Effects in Social Psychology:What Is ‘Social Priming’ and How Does It Occur?” Social Cognition 32:1–11.

Mondak, Jeffery J. 2010. Personality and the Foundations of Political Behavior.New York: Cambridge University Press.

Mondak, Jeffery J., and Karen D. Halperin. 2008. “A Framework for the Studyof Personality and Political Behavior.” British Journal of Political Science 38:335–362.

Mondak, Jeffery J., Matthew V. Hibbing, Damarys Canache, Mitchell A. Seligson,and Mary R. Anderson. 2010. “Personality and Civic Engagement: AnIntegrative Framework for the Study of Trait Effects on Political Behavior.”The American Political Science Review 104: 85–110.

Mondak, Jeffery J., Damarys Canache, Mitchell A. Seligson, and Matthew V.Hibbing. 2011. “The Participatory Personality: Evidence from LatinAmerica.” British Journal of Political Science 41: 211–221.

Montag, Christian, and Martin Reuter. 2014. “Disentangling the MolecularGenetic Basis of Personality: From Monoamines to Neuropeptides.”Neuroscience and Biobehavioral Review 43: 228–239.

Montoro, Pedro R., Dolores Luna, and Juan J. Ortells. 2014. “Subliminal GestaltGrouping: Evidence of Perceptual Grouping by Proximity and Similarity inAbsence of Conscious Perception.” Consciousness and Cognition 25: 1–8.

Mullainathan, Sendhil, and Ebonya Washington. 2009. “Sticking with Your Vote:Cognitive Dissonance and Political Attitudes.” Applied Economics 1: 86–111.

Mulligan, Kenneth. 2013. “Variability or Moderation? the Effects of Ambivalenceon Political Opinions.” Political Behavior 35: 539–565.

Mullinix, Kevin J. 2016. “Partisanship and Preference Formation: CompetingMotivations, Elite Polarization, and Issue Importance.” Political Behavior 38:383–411.

Mummolo, Jonathan. 2016. “News from the Other Side: How Topic RelevanceLimits the Prevalence of Partisan Selective Exposure.” Journal of Politics 78:763–773.

184 REFERENCES

http://Michigan.gov

http://www.michigan.gov/snyder/0,4668,7-277-57577-344374%E2%80%93,00.html

http://www.michigan.gov/snyder/0,4668,7-277-57577-344374%E2%80%93,00.html

Nannestad, Peter, and Martin Paldam. 1994. “The Vp-Function: A Survey of theLiterature on Vote and Popularity Functions after 25 Years.” Public Choice 79:213–245.

National Public Radio. 2012. “Transcript And Audio: First Obama-RomneyDebate,” NPR.org, http://www.npr.org/2012/10/03/162258551/transcript-first-obama-romney-presidential-debate

Nawara, Steven P. 2015. “Who Is Responsible, the Incumbent or the FormerPresident? Motivated Reasoning in Responsibility Attributions.” PresidentialStudies Quarterly 45: 110–131.

Newell, Ben R., and David R. Shanks. 2014. “Unconscious Influences on DecisionMaking: A Critical Review.” Behavioral and Brain Sciences 37: 1–61.

Nickerson, Raymond S. 1998. “Confirmation Bias: A Ubiquitous Phenomenon inMany Guises.” Review of General Psychology 2: 175–220.

Niemi, Richard, and Jane Junn. 1998. Civic Education: What Makes StudentsLearn. New Haven: Yale University Press.

NIH: National Institute on Drug Abuse. 2015. “Drug Facts.” Drugabuse.gov.www.drugabuse.gov/publications/drugfacts/nationwide-trends

Nir, Lilach. 2011. “Motivated Reasoning and Public Opinion Perception.” PublicOpinion Quarterly 75: 504–532.

Nisbett, Richard E., and Timothy DeCamp Wilson. 1977. “Telling More than WeCan Know: Verbal Reports on Mental Processes.” Psychological Review 84:231–259.

Noftle, Erik E., and Richard W. Robins. 2007. “Personality Predictors ofAcademic Outcomes: Big Five Correlates of GPA and SAT Scores.” Journalof Personality and Social Psychology 93: 116–130.

Norpoth, Helmut, Michael S. Lewis-Beck, and Jean-Dominique Lafay, edited by1991. Economics and Politics: The Calculus of Support. Ann Arbor: University ofMichigan Press.

O’Connor, Melissa, and Sampo V. Paunonen. 2007. “Big Five PersonalityPredictors of Post-Secondary Academic Performance.” Personality andIndividual Differences 43: 971–990.

Okbay, Aysu. et al. 2016. “Genetic Variants Associated with Subjective Well-Being, Depressive Symptoms, and Neuroticism Identified through Genome-Wide Analyses.” Nature Genetics 48: 624–636.

Olivola, Christopher Y., and Alexander Todorov. 2010. “Elected in 100Milliseconds: Appearance-Based Trait Inferences and Voting.” Journal ofNonverbal Behavior 34: 83–110.

Olson, James, and Mark Zanna. 1979. “A New Look at Selective Exposure.”Journal of Experimental Social Psychology 15: 1–15.

Omura, Kazufumi, R., Todd Constable, and Turhan Canli. 2005. “AmygdalaGray Matter Concentration Is Associated with Extraversion andNeuroticism.” Neuroreport 16: 1905–1908.

REFERENCES 185

http://NPR.org

http://www.npr.org/2012/10/03/162258551/transcript-first-obama-romney-presidential-debate

http://www.npr.org/2012/10/03/162258551/transcript-first-obama-romney-presidential-debate

http://Drugabuse.gov

http://www.drugabuse.gov/publications/drugfacts/nationwide-trends

Osborne, Danny, and Chris G. Sibley. 2012. “Does Personality Matter? OpennessCorrelates with Vote Choice, but Particularly for Politically SophisticatedVoters.” Journal of Research in Personality 46: 743–751.

Osterhammel, Jürgen. 2014. The Transformation of the World: A Global History ofthe Nineteenth Century. Princeton, NJ: Princeton University Press.

Ozer, Daniel J., and Verónica Benet-Martínez. 2006. “Personality and the Predictionof Consequential Outcomes.” Annual Review of Psychology 57: 401–421.

Pacek, Alexander C., and Benjamin Radcliff. 1995. “The Political Economy ofCompetitive Elections in the Developing World.” American Journal ofPolitical Science 39: 745–759.

Page, Benjamin I., and Robert Y. Shapiro. 1992. The Rational Public: Fifty Yearsof Trends in Americans’ Policy Preferences. Chicago, IL: University of ChicagoPress.

Parker, Ashley, “A Wife Committed to Cruz’s Ideals, but A Study in Contrasts toHim,”New York Times, October 23, 2013. http://www.nytimes.com/2013/10/24/us/politics/a-wife-committed-to-cruzs-ideals-but-a-study-in-contrasts-to-him.html?hp&_r=1&

Paynter, Christopher A., Kenneth Kotovsky, and Lynne M. Reder. 2010. “Problem-Solving Without Awareness: An ERP Investigation.”Neuropsychologia 48: 3137–3144.

Peatfield, Nicholas A., Oliver H. Turnbull, John Parkinson, and James Intriligator.2012. “Quick as a BLINK: An Ultrarapid Analogue of Iowa Gambling TaskDecision Making.” Journal of Clinical and Experimental Neuropsychology 34:243–255.

Pérez, Efrén O. 2016.Unspoken Politics: Implicit Attitudes and Political Thinking.New York: Cambridge University Press.

Perry, Ryan, and Chris G. Sibley. 2013. “Seize and Freeze: Openness toExperience Shapes Judgments of Societal Threat.” Journal of Research inPersonality 47: 677–686.

Petersen, Michael Bang, Søren Serritzlew, Martin Skov, and Thomas Ramsøy.2013. “Motivated Reasoning and Political Parties: Evidence for IncreasedProcessing in the Face of Party Cues.” Political Behavior 35: 831–854.

Pierce, Douglas R. 2015. “Uniformed Votes? Reappraising Information Effectsand Presidential Preferences.” Political Behavior 37: 537–565.

Plato. 2007. The Republic. New York: Penguin Classics.Popkin, Samuel L. 1991. The Reasoning Voter: Communication and Persuasion in

Presidential Campaigns. Chicago, IL: University of Chicago Press.Popp, Elizabeth, and Thomas J. Rudolph. 2011. “A Tale of Two Ideologies:

Explaining Public Support for Economic Interventions.” The Journal ofPolitics 73: 808–820.

186 REFERENCES

http://www.nytimes.com/2013/10/24/us/politics/a-wife-committed-to-cruzs-ideals-but-a-study-in-contrasts-to-him.html?hp%26_r=1%26



Power, Robert A., and Michael Pluess. 2015. “Heritability Estimates of the BigFive Personality Traits Based on Common Genetic Variants.” TranslationalPsychiatry 5: 1–5.

Przeworski, Adam. 2010. Democracy and the Limits of Self-Government. NewYork: Cambridge University Press.

Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of AmericanCommunity. New York: Simon and Schuster.

Quintelier, Ellen, and Yannis Theocharis. 2013. “Online Political Engagement,Facebook, and Personality Traits.” Social Science Computer Review 31:280–290.

Real Clear Politics. 2013. “FOX News Special: “The Great Food Stamp Binge.”Realclearpolitics.com. http://www.realclearpolitics.com/video/2013/08/10/fox_news_special_the_great_food_stamp_binge.html

Redlawsk, David. 2002. “Hot Cognition or Cool Consideration.” Journal ofPolitics 64: 1021–1044.

Redlawsk, David P., Andrew J. W. Civettini, and Karen M. Emmerson. 2010.“The Affective Tipping Point: Do Motivated Reasoners Ever ‘Get It’?”Political Psychology 31: 563–593.

Riker, William H., and Peter C. Ordeshook. 1968. “A Theory of the Calculus ofVoting.” American Political Science Review 62: 25–42.

Ritter, Simone M., and Ap Dijksterhuis. 2014. “Creativity—The UnconsciousFoundations of the Incubation Period.” Frontiers in Human Neuroscience8: 1–10.

Roberts, Brent W., and Daniel Mroczek. 2008. “Personality Trait Change inAdulthood.” Current Directions in Psychological Science 17: 31–35.

Roberts, Brent W., Kate E. Walton, and Wolfgang Viechtbauer. 2006. “Patternsof Mean-Level Change in Personality Traits Across the Life Course: A Meta-Analysis of Longitudinal Studies.” Psychological Bulletin 132: 1–25.

Roets, Arne, Ilse Cornelis, and Alain Van Hiel. 2014. “Openness as a Predictorof Political Orientation and Conventional and Unconventional PoliticalActivism in Western and Eastern Europe.” Journal of Personality Assessment96: 53–63.

Ross, Lee D., Yphtach Lelkes, and Alexandra G. Russell. 2012. “How ChristiansReconcile Their Personal Political Views and the Teachigns of Their Faith:Projection as a Means of Dissonance Reduction.” Proceedings of the NationalAcademy of Sciences 109: 3616–3622.

Rothbart, Mary K., Kean H. Chew, and Maria Amy Gartstein. 2001. “Assessmentof Temperament in Early Development.” In Biobehavioral Assessment of theInfant, edited by Lynn Twarog Singer, Philip Sanford Zeskind, and Lynn T.Singer, 190–208. New York: Guilford Press.

REFERENCES 187

http://Realclearpolitics.com

http://www.realclearpolitics.com/video/2013/08/10/fox_news_special_the_great_food_stamp_binge.html

http://www.realclearpolitics.com/video/2013/08/10/fox_news_special_the_great_food_stamp_binge.html

Rousseau, Jean-Jacques. 1988.Rousseau’s Political Writings. Edited by Alan Ritterand Julia Conaway Bondanella, translated by Julia Conaway Bondanella. NewYork: W.W. Norton & Company.

Rudolph, Thomas J. 2006. “Triangulating Political Responsibility: TheMotivated Formation of Responsibility Judgments.” Political Psychology 27:99–122.

Rule, Nicholas O., Jonathan B. Freeman, Joseph M. Moran, John D. E. Gabrieli,Reginald B. Adams, and Nalini Ambady. 2010. “Voting Behavior Is Reflectedin Amygdala Response Across Cultures.” Social Cognition and AffectiveNeuroscience 5: 349–355.

Sackett, Paul R., and Philip Walmsley. 2014. “Which Personality Attributes areMost Important in the Workplace?” Perspectives on Psychological Science 9:538–551.

Sampaio, Adriana, Jose Miguel Soares, Joana Coutinho, Nuno Sousa, and OscarGoncalves. 2014. “The Big Five Default Brian: Functional Evidence.” BrianStructure and Function 219: 1913–1922.

Savitz, Jonathan B., and Rajkumar S. Ramesar. 2004. “Genetic VariantsImplicated in Personality: A Review of the More Promising Candidates.”American Journal of Medical Genetics Part B: Neuropsychiatric Genetics131B: 20–32.

Schaefer, Michael, Hans-Jochen Heinze, and Michael Rotte. 2012. “Touch andPersonality: Extraversion Predicts Somatosensory Brain Response.”Neuroimage 62: 432–438.

Schaefer, Michael, Michael Rotte, Hans-Jochen Heinze, and Claudia Denke.2013. “Mirror-Like Brain Responses to Observed Touch and PersonalityDimensions.” Frontiers in Human Neuroscience 7: 1–9.

Schlegel, Alexander, Prescott Alexander, Walter Sinnott-Armstrong, AdinaRoskies, Peter U. Tse, and Thalia Wheatley. 2013. “Barking up the WrongFree: Readiness Potentials Reflect Processes Independent of Conscious Will.”Experimental Brain Research 229: 329–335.

Schoen, Harald, and Siegfried Schumann. 2007. “Personality Traits, PartisanAttitudes, and Voting Behavior: Evidence from Germany.” PoliticalPsychology 28: 471–498.

Schoen, Harald, and Markus Steinbrecher. 2013. “Beyond Total Effects:Exploring the Interplay of Personality and Attitudes in AffectingTurnout in the 2009 German Federal Election.” Political Psychology 34:533–552.

Schram, Sanford F., Joe Soss, and Richard C. Fording. 2003. Race and the Politicsof Welfare Reform. Ann Arbor, MI: University of Michigan Press.

Schumpeter, Joseph. 1950. Capitalism, Socialism, and Democracy. 3rd New York:Harper & Row.

188 REFERENCES

Schwabe, Inga, Wilfried Jonker, and Stephanie M. Van Den Berg. 2016. “Genes,Culture, and Conservativism-A Psychometric-Genetic Approach.” BehaviorGenetics 46: 516–528.

Servaas, Michelle N., Jorien Van Der Velde, Sergi G. Costafreda, Paul Horton,Johan Ormel, Harriette Riese, and Andre Aleman. 2013. “Neuroticism andthe Brian: A Quantitative Meta-Analysis of Neuroimaging StudiesInvestigating Emotion Processing.” Neuroscience and Biobehavioral Reviews37: 1518–1529.

Shaker, Lee. 2012. “Local Political Knowledge and Assessments of CitizenCompetence.” Public Opinion Quarterly 76: 525–537.

Shamir, Jacob, and Michal Shamir. 1997. “Pluralistic Ignorance Across Issues andOver Time: Information Cues and Biases.” Public Opinion Quarterly 61: 227–260.

Shapiro, Robert Y. 2011. “Public Opinion and American Democracy.” PublicOpinion Quarterly 75: 982–1017.

Shepherd, Joanna M. 2009. “The Influence of Retention Politics on Judges’Voting.” Journal of Legal Studies 38: 169–203.

Shepherd, Joanna M. 2013. “Justice at Risk: An Empirical Analysis of CampaignContributions and Judicial Decisions.” American Constitution Society for Lawand Policy. Available: https://assets.documentcloud.org/documents/1307928/justice-at-risk.pdf

Sherman, David K., and Geoffrey L. Cohen. 2006. “The Psychology of Self-Defense: Self-Affirmation Theory.” Advances in Experimental SocialPsychology 38: 183–242.

Sherrod, Drury R. 1971–72. “Selective Perception of Political Candidates.” ThePublic Opinion Quarterly 35: 554–562.

Shields, Todd G., and Robert K. Goidel. 1997. “Participation Rates,Socioeconomic Class Biases, and Congressional Elections: A Cross-Validation, 1958–1994.” American Journal of Political Science 41: 683–691.

Sibley, Chris G., Danny Osborne, and John Duckitt. 2012. “Personality andPolitical Orientation: Meta-Analysis and Test of a Threat-Constraint Model.”Journal of Research in Personality 46: 664–677.

Sides, John. 2016. “Stories of Science? Facts, Frames, and Policy Attitudes.”American Politics Research 44: 387–414.

Slothuus, Rune, and Claes H. De Vreese. 2010. “Political Parties, MotivatedReasoning, and Issue Framing Effects.” The Journal of Politics 72: 630–645.

Smith, Kevin B., Douglas R. Oxley, Matthew V. Hibbing, John R. Alford, andJohn R. Hibbing. 2011. “Linking Genetics and Political Attitudes.” PoliticalPsychology 32: 369–397.

Smith, Colin Tucker, Kate A. Ratliff, and Brian A. Nosek. 2012a. “RapidAssimilation: Automatically Integrating New Information with ExistingBeliefs.” Social Cognition 30: 199–219.

REFERENCES 189

https://assets.documentcloud.org/documents/1307928/justice-at-risk.pdf

https://assets.documentcloud.org/documents/1307928/justice-at-risk.pdf

Smith, Kevin, John R. Alford, Peter K. Hatemi, Lindon J. Eaves, Carolyn Funk,and John R. Hibbing. 2012b. “Biology, Ideology, and Epistemology: How DoWe Know Political Attitudes Are Inherited and Why Should We Care?.”American Journal of Political Science 56: 17–33.

Smith, D. J. et al. 2016. “Genome-Wide Analysis of Over 106000 IndividualsIdentifies 9 Neuroticism-Associated Loci.” Molecular Psychiatry 21:749–757.

Sokhey, Anand Edward, and Scott D. McClurg. 2012. “Social Networks andCorrect Voting.” The Journal of Politics 74: 751–764.

Soon, Chun Siong, Marcel Brass, Hans-Jochen Heinze, and John-Dylan Haynes.2008. “Unconscious Determinants of Free Decisions in the Human Brain.”Nature Neuroscience 11: 543–545.

Soon, Chun Siong, Anna Hanxi He, Stefan Bode, and John-Dylan Haynes.2013. “Predicting Free Choices for Abstract Intentions.” Proceedings of theNational Academy of Sciences of the United States of America 110:6217–6222.

Stanley, Jason. 2015. How Propaganda Works. Princeton, NJ: PrincetonUniversity Press.

Stimson, James A., Michael B. MacKuen, and Robert S. Erikson. 1995. “DynamicRepresentation.” American Political Science Review 89: 543–565.

Stimson, James A. 2004. Tides of Consent: How Public Opinion Shapes AmericanPolitics. New York: Cambridge University Press.

Stroop, John Ridley. 1935. “Studies of Interference of Serial Verbal Reactions.”Journal of Experimental Psychology 18: 643–662.

Sullivan, Laura, Tatjana Meschede, Lars Dietrich, Thomas Shapiro, Amy Traub,Catherine Ruetschlin, and Tamara Draut. The Racial Wealth Gap: Why PolicyMatters. Demos and Institute for Assets & Social Policy, Brandeis University.New York, 2015. Available from: http://www.demos.org/publication/racial-wealth-gap-why-policy-matters

Sun, Lena H., and Niraj Chokshi. 2015. “Millions at Risk of Losing Coverage asJustices Take up Challenge to Obamacare.” The Washington Post February 16,2015. http://www.washingtonpost.com/national/health-science/millions-at-risk-of-losing-coverage-in-supreme-court-health-law-case/2015/02/16/0597f6aa-ae50-11e4-ad71-7b9eba0f87d6_story.html

Sweeney, P.D., and K. L. Gruber. 1984. “Selective Exposure: Voter InformationPreferences and the Watergate Affair.” Journal of Personality and SocialPsychology 46: 1208–1221.

Taber, Charles S., and Milton Lodge. 2006. “Motivated Skepticism in theEvaluation of Political Beliefs.” American Journal of Political Science 50:755–769.

Taber, Charles S., Damon Cann, and Simona Kucsova. 2009. “The MotivatedProcessing of Political Arguments.” Political Behavior 31: 137–155.

190 REFERENCES

http://www.demos.org/publication/racial-wealth-gap-why-policy-matters

http://www.demos.org/publication/racial-wealth-gap-why-policy-matters

http://www.washingtonpost.com/national/health-science/millions-at-risk-of-losing-coverage-in-supreme-court-health-law-case/2015/02/16/0597f6aa-ae50-11e4-ad71-7b9eba0f87d6_story.html



Taber, Charles S., and Milton Lodge. 2016. “The Illusion of Choice inDemocratic Politics: The Unconscious Impact of Motivated PoliticalReasoning.” Advances in Political Psychology 37: 61–85.

Takeuchi, Hikaru, Yasuyuki Taki, Atsushi Sekiguchi, Rui Nouchi, Yuka Kotozaki,Seishu Nakagawa, Carlos Makoto Miyauchi, Kunio Lizuka, Ryoichi Yokoyama,Takamitsu Shinada, Yuki Yamamoto, Sugiko Hanawa, Tsuyoshi Araki, HiroshiHashizume, Keiko Kunitoki, Yuko Sassa, and Ryuta Kawashima. 2016.“Differences in Gray Matter Structure Correlated to Nationalism andPatriotism.” Scientific Reports 6: 1–10.

Taki, Yasuyuki, Benjamin Thyeau, Shigeo Kinomura, Kazunori Sato, Ryoi Goto,Kai Wu, Ryuta Kawashima, and Hiroshi Fukuda. 2013. “A Longitudinal Studyof the Relationship Between Personality Traits and the Annual Rate of VolumeChange in Regional Gray Matter in Healthy Adults.” Human Brain Mapping34: 3347–3353.

Taylor, Shelley E. 1982. “The Availability Bias in Social Perception andInteraction.” In Judgment under Uncertainty: Heuristics and Biases, editedby Daniel Kahneman, Paul Slovic, and Amos Tversky, 190–200. New York:Cambridge University Press.

Todorov, Alexander, Anesu N. Mandisodza, Amir Goren, and Crystal C. Hall.2005. “Inferences of Competence from Faces Predict Electoral Outcomes.”Science 308: 1623–1626.

Tomz, Michael, and Robert P. Van Houweling. 2009. “He Electoral Implicationsof Candidate Ambiguity.” American Political Science Review 103: 83–98.

Tormala, Zakary L., and Victoria L. DeSensi. 2008. “The Perceived InformationalBasis of Attitudes: Implications for Subjective Ambivalence.” Personality andSocial Psychology Bulletin 34: 275–287.

Turkheimer, Eric. 2000. “Three Laws of Behavior Genetics and What TheyMean.” Current Directions in Psychological Science 9: 160–164.

Turkheimer, Eric. 2011. “Still Missing.” Research in Human Development 8:227–241.

Turkheimer, Eric, Erik Pettersson, and Erin E. Horn. 2014. “A Phenotypic NullHypothesis for the Genetics of Personality.” Annual Review of Psychology65: 515–540.

Tusche, Anita, Thorsten Kahnt, David Wisniewski, and John-Dylan Haynes. 2013.“Automatic Processing of Political Preferences in the Human Brain.”NeuroImage 72: 174–182.

Tversky, Amos, and Daniel Kahneman. 1981. “The Framing of Decisions and thePsychology of Choice.” Science 211: 453–458.

U.S. Bureau of Economic Analysis. 2017. “FAQ: What Is Rental Income ofPersons?” bea.gov. https://www.bea.gov/faq/index.cfm?faq_id=64

U.S. Department of Agriculture, Food and Nutritional Service, Office of PolicySupport, “Characteristics of Supplemental Nutritional Assistance Program

REFERENCES 191

http://bea.gov

https://www.bea.gov/faq/index.cfm?faq_id=64

Households: Fiscal Year 2012—Summary,” by Kelsey Farson Gray and EsaEslami. February 2014a. Washington, D.C.

U.S. Department of Agriculture, Food and Nutritional Services, Office of PolicySupport, “Characteristics of Supplemental Nutrition Assistance ProgramHouseholds: Fiscal Year 2012.” By Kelsey Farson Gray and Esa Eslami.February 2014b. Washington, D.C., Report No. SNAP-14-CHAR.

U.S. Department of Agriculture, Food and Nutritional Service, Office of PolicySupport, “Characteristics of Supplemental Nutritional Assistance ProgramHouseholds: Fiscal Year 2013—Summary,” by Kelsey Garson Gray.December 2014a. Washington, D.C.

U.S. Department of Agriculture, Food and Nutritional Services, Office of PolicySupport, “Characteristics of Supplemental Nutrition Assistance ProgramHouseholds: Fiscal Year 2013.” By Kelsey Farson Gray. December 2014b.Washington, D.C., Report No. SNAP-14-CHAR.

U.S.C.B. 2015. “Mortgage Status.” Factfinder.census.gov. https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_14_SPL_K202508&prodType=table

U.S.D.A. 2017. “Supplemental Nutrition Assistance Program Participation andCosts.” Fns.usda.gov. https://www.fns.usda.gov/sites/default/files/pd/SNAPsummary.pdf

Usher, Marius, Zohar Russo, Mark Weyers, Ran Brauner, and Dan Zakay. 2011.“The Impact of the Mode of Thought in Complex Decisions: IntuitiveDecision are Better.” Frontiers in Psychology 2: 1–13.

Valentino, Nicholas A., Ted Brader, Eric W. Groenendyk, Krysha Gregorowicz,and Vincent L. Hutchings. 2011. “Election Night’s Alright for Fighting: TheRole of Emotions in Political Participation.” Journal of Politics 73: 156–170.

Van Boven, Leaf, Charles M. Judd, and David K. Sherman. 2012. “PoliticalPolarization Projection: Social Projection of Partisan Attitude Extremity andAttitudinal Processes.” Journal of Personality and Social Psychology 103: 84–100.

Van Den Berg, Stephanie M. et al. 2016. “Meta-Analysis of Genome-WideAssociation Studies for Extraversion: Findings from the Genetics ofPersonality Consortium.” Behavior Genetics 46: 170–182.

Van Gaal, Simon, Lionel Naccache, Julia D. I. Meuwese, Anouk M. Van Loon,Alexandra H. Leighton, Laurent Cohen, and Stanislas Dehaene. 2014. “Canthe Meaning of Multiple World Be Integrated Unconsciously?.” PhilosophicalTransactions of the Royal Society B: Biological Sciences 369(1641): 1–11.

Vecchione, Michele, and Gian Vittorio Caprara. 2009. “Personality Determinantsof Political Participation: The Contribution of Traits and Self-Efficacy Beliefs.”Personality and Individual Differences 46: 487–492.

Vedel, Anna. 2014. “The Big Five and Tertiary Academic Performance: ASystematic Review and Meta-Analysis.” Personality and Individual Difference71: 66–76.

192 REFERENCES

http://Factfinder.census.gov

https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_14_SPL_K202508%26prodType=table



http://Fns.usda.gov

https://www.fns.usda.gov/sites/default/files/pd/SNAPsummary.pdf

https://www.fns.usda.gov/sites/default/files/pd/SNAPsummary.pdf

Verba, Sidney, Kay Lehman Schlozman, and Henry Brady. 1995. Voice andEquality: Civic Voluntarism in American Politics. Cambridge, MA: HarvardUniversity Press.

Verhulst, Brad, and Peter K. Hatemi. 2013. “Gene-Environment Interplay inTwin Models.” Political Analysis 21: 368–389.

Verhulst, Brad, Lindon J. Eaves, and Peter K. Hatemi. 2012. “Correlation NotCausation: The Relationship between Personality Traits and PoliticalIdeologies.” American Journal of Political Science 56: 34–51.

Vidmar, Niel, and Milton Rokeach. 1974. “Archie Bunker’s Bigotry: A Study inSelective Perception and Exposure.” Journal of Communications 24: 36–47.

Volsky, Igor. “Paul Ryan Blames Poverty on Lazy ‘Inner City’ Men.”ThinkProgress. March 12, 2014. http://thinkprogress.org/economy/2014/03/12/3394871/ryan-poverty-inner-city/

Voss, Joel L., Heather Lucas, and Ken A. Paller. 2012. “More than a Feeling:Pervasive Influences of Memory without Awareness of Retrieval.” CognitiveNeuroscience 3: 193–207.

Wagner, Markus, Jessica Tarlov, and Nick Vivyan. 2014. “Partisan Bias in OpinionFormation on Episodes of Political Controversy: Evidence from Great Britain.”Political Studies 62: 136–158.

Wayne, Leslie, and David M. Herszenhorn, “A Bid to Tax Health Plans ofExecutives,” New York Times, July 26, 2009. http://www.nytimes.com/2009/07/27/health/policy/27insure.html

Weeks, Brian E. 2015. “Emotions, Partisanship, and Misperceptions: How Angerand Anxiety Moderate the Effect of Partisan Bias on Susceptibility to PoliticalMisinformation.” Journal of Communication 65: 699–719.

Weibel, Sebastien, Anne Giersch, Stanislas Dehaene, and Caroline Huron. 2013.“Unconscious Task Set Priming with Phonological and Semantic Tasks.”Consciousness and Cognition 22: 517–527.

Weinschenk, Aaron C. 2014. “Personality Traits and the Sense of Civic Duty.”American Politics Research 42: 90–113.

Westen, Drew, Pavel S. Blagov, Keith Harenski, Clint Kilts, and Stephen Hamann.2006. “Neural Bases of Motivated Reasoning: An Fmri Study of EmotionalConstraints on Partisan Political Judgment in the 2004 U.S. PresidentialElection.” Journal of Cognitive Neuroscience 18: 1947–1985.

Wilson, Timothy D. 2002. Strangers to Ourselves: Discovering the AdaptiveUnconscious. Cambridge, MA: Belknap Press.

Wlezien, Christopher. 1995. “The Public as Thermostat: Dynamics of Preferencesfor Spending.” American Journal of Political Science 39: 981–1000.

Wolfinger, Raymond, and Steven J. Rosenstone. 1980. Who Votes? New Haven:Yale University Press.

Woo, Sang Eun, Oleksandr S. Chernyshenko, Andrew Longley, Zhi-Xue Zhang,Chi-Yue Chiu, and Stephen E. Stark. 2014. “Openness to Experience: Its

REFERENCES 193

http://thinkprogress.org/economy/2014/03/12/3394871/ryan-poverty-inner-city/

http://thinkprogress.org/economy/2014/03/12/3394871/ryan-poverty-inner-city/

http://www.nytimes.com/2009/07/27/health/policy/27insure.html

http://www.nytimes.com/2009/07/27/health/policy/27insure.html

Lower Level Structure, Measurement, and Cross-Cultural Equivalence.”Journal of Personality Assessment 96: 29–45.

Wright, Christopher I., Danielle Williams, Eric Feczko, Lisa Feldman Barrett,Bradford C. Dickerson, Carl E. Schwartz, and Michelle M. Wedig. 2006.“Neuroanatomical Correlates of Extraversion and Neuroticism.” CerebralCortex 16: 1809–1819.

Yacoubian, George, and Blake J. Urbach. 2002. “A Comparison of Drug Usebetween Welfar-Receiving Arrestees and Non-Welfare-Receiving Arrestees.”Health & Social Work 27: 290–293.

Yoshimoto, Sanae, Hisato Imai, Makio Kashino, and Tatsuto Takeuchi. 2014.“Pupil Response and the Subliminal Mere Exposure Effect.” Plus One 9:e90670.

Young, Jacob T.N., and Frank M. Weerman. 2013. “Delinquency as aConsequence of Misperception: Overestimation of Friends’ DelinquentBehavior and Mechanisms of Social Influence.” Social Problems 60: 334–356.

Youngman, Sam. 2014. “Mitch Mcconnell Outlines Future Opposition to‘Obamacare’ if GOP Wins Senate Majority,” Lexington Herald-Leader,http://www.kentucky.com/2014/01/10/3027635/mitch-mcconnell-outlines-obamacare.html

Zajonc, Robert B. 1968. “Attitudinal Effects of Mere Exposure.” Journal ofPersonality and Social Psychology 9: 1–27.

Zaller, John R. 1992. The Nature and Origins of Mass Opinion. New York:Cambridge University Press.

Zamboni, Giovanna, Marta Gozzi, Frank Krueger, Jean-René Duhamel, AngelaSirigu, and Jordan Grafman. 2009. “Individualism, Conservatives, andRadicalism as Criteria for Processing Political Beliefs: A Parametric FmriStudy.” Social Neuroscience 4: 367–383.

Zanna, Mark P., and James M. Olson. 1982. “Individual Differences in AttitudinalRelations.” In Consistency in Social Behavior: The Ontario Symposium Volume 2,edited by Mark P. Zanna, E. Tory Higgins, and C. Peter Herman, 75–104.Hillsdale, NJ: Erlbaum.

194 REFERENCES

http://www.kentucky.com/2014/01/10/3027635/mitch-mcconnell-outlines-obamacare.html

http://www.kentucky.com/2014/01/10/3027635/mitch-mcconnell-outlines-obamacare.html

INDEX

AAbelson, Robert, 13Achen, Christopher, 4, 11, 37, 71, 83,

88, 130, 139, 140, 141, 147n1,166n1

Ackerman, Phillip, 30Affective Intelligence, 23Affordable Care Act (ACA), 1, 157Aghajani, Moji, 28Agreeableness, 6, 28–31, 34, 36,

43–44, 48, 49, 52–58, 60–64,65n2, 78, 82, 85, 89, 96, 115,120–121, 126–127, 128n8,136–139, 154, 162n1, 166n8

Ahadi, Stephan, 31Albertson, Bethany, 24Alford, John, 5, 20, 130, 141Allport, Gordon, 29Althaus, Scott, 15American National Election Studies

(ANES), 46, 50, 143, 155, 156,159, 160, 161

Amodio, David, 6, 68Andersen, David, 35Andsager, Julie, 14Ansorge, Ulrich, 18Anxiety, 24, 26, 111Armstrong, Anna-Marie, 18Aronson, Elliot, 41

Asimov, Isaac, 88Automatic, 6, 21–23, 70, 84, 142, 146

BBalance Theory, 13Bandyopadhyay, Siddhartha, 143Banks, Antoine, 24Barabas, Jason, 6, 35, 68Barbaranelli, Claudio, 32Barber, Benjamin, 87, 88, 106Bargh, John, 15Barnea, Marina, 33Baron, Reuben, 124Barrick, Murray, 31, 77Bartels, Larry, 1, 4, 6, 11, 35, 37, 68,

71, 83, 88, 90, 94, 130, 139,140–141, 144, 146, 147n1,166n1

Beasley, Ryan, 41Beaty, Roger, 28Bechara, Antoine, 17, 18Behavioral genetics, 27, 29Bell, Edward, 20Benet-Martínez, Verónica, 33Bennett, Stephen Earl, 10Berdejó, Carlos, 143Berelson, Bernard, 8Berry, Kate, 143


195

Bias, 111, 114, 125, 126, 142, 151Bicchieri, Cristina, 14Big Five personality traits, 6, 8, 28,

29–37, 40, 41, 48, 49, 51, 67, 76,84, 96, 104, 105, 114, 130, 131,137, 147, 149, 152, 153

Biology, 29, 31, 141Bleidorn, Wiebke, 26Bode, Stefan, 16Bogg, Tim, 31Bolsen, Toby, 13, 68, 70, 142Bølstad, Jørgen, 41Bouchard, Thomas, 20, 27Brader, Ted, 13, 24Brady, Peter, 116Brain, 5–6, 8, 16–23, 25–26, 28–29,

31, 67, 70, 84, 129–130, 140,142–143, 147

Brandt, Mark, 15Bratko, Denis, 78Brett, Jeanne, 74, 124Britz, Juliane, 18Budgetary appropriations, 112, 113Burman, Leonard, 113Burstein, Paul, 88

CCampaign, 1, 5, 8–9, 11, 20, 24, 30, 33,

62, 93, 130, 133, 142, 145, 146Campbell, Angus, 1, 35, 70Canes-Wrone, Brandice, 143Canli, Turhan, 28Cantril, Hadley, 40Caplan, Bryan, 10, 11Caprara, Gian Vittorio, 30, 32, 33, 78Carney, Dana, 32, 33, 153Cassino, Daniel, 13Cattell, Raymond Bernard, 27, 29,

32, 76Chambers, John, 15Chamorro-Premuzic, Tomas, 77, 78

Choi, Eunjung, 11Chokshi, Niraj, 110Civic, 129–147Civic Aptitude, 6, 25, 26, 34, 36,

129–147Claassen, Ryan, 22, 68, 144Cobb-Clark, Deborah, 26Cognition, 41, 43, 140Cognitive dissonance, 6, 34, 39–65Cohen, Alma, 143Cohen, Geoffrey, 13, 14Cohen, Jeffrey, 11Colas, Jaron, 16Collado, Dolores, 112Collins, Paul, 41Condorcet, 93, 94Conflicted conservative, 42, 43, 45,

48, 53, 58, 60, 64Confused conservative, 50, 65n6,

164n6Confused Liberal, 46, 48, 49, 50, 53,

56, 57, 58, 61, 64, 136Congleton, Roger, 93Connelly, Brian, 32Conscientiousness, 6, 28, 29, 31–33,

36, 43–44, 49, 52–58, 60–64, 68,77, 82, 85, 96, 99–100, 102,104, 112, 114–115, 118,120–126, 128n7, 128n8,134–136, 138–139, 154, 165n7

Conscious, 5, 8, 15–24, 26, 30, 31,34, 42, 70, 106, 142, 147

Converse, Philip, 4, 9, 10, 12, 40, 43,45, 48, 83

Cooper, Christopher, 30, 32, 78, 141Costa, Paul, 32, 76Cuik, David, 142

DDahl, Robert, 88Davies, Patrick, 24

196 INDEX

Davison, P. W., 14Dawson, Deborah, 73Deardorff, Alan, 89Delli Carpini, Michael, 10, 45, 61DeLorme, Denise, 14Demertzi, Athena, 19Democracy, 4, 7, 11, 37, 62, 71, 73,

87–88, 106, 130, 139–145Democratic ideal, 7, 62, 85, 93, 104de Moor, Marleen, 27, 28DeSensi, Victoria, 81, 155de Vreese, Claes, 13DeYoung, Collin, 28Dick, Philip K., 147Dienes, Zoltan, 18Dietrich, Bryce, 6, 26Digman, John, 31Dijksterhuis, Ap, 18Dinesen, Peter Thisted, 30, 32,

33, 115Dirilen-Gümüş, Özlem, 31, 32Dissonance, 6, 13, 34, 39–65Dissonance theory, 13, 41, 57, 58Dodd, Michael, 21, 70Downs, Anthony, 10, 90Draganski, Bogdan, 31Druckman, James, 13, 14, 142Drug Testing, 35, 65, 68, 73, 76, 81,

82, 85, 89, 90, 92–95, 96, 98–99,104, 131, 133, 135, 138, 149,155–156

Duch, Raymond, 11Dunning, David, 14Dusso, Aaron, 73, 90, 144

EEconomic Perception, 11Economy, 11Eimer, Martin, 16Einstein, Katherine Levine, 83, 140Eisenberg, Nancy, 30, 115

Ellis, Christopher, 6, 9, 34, 40, 42, 43,45, 47, 49, 53, 57, 60, 63, 64,113, 136, 161

Emotional stability, 6, 29, 32, 43–44,48, 52–53, 60, 62–63, 76–77, 82,85, 96, 104, 115, 137–139, 154

Enns, Peter, 11, 144Ensley, Michael, 22, 68, 144Erikson, Robert, 62Estlund, David, 7, 130, 146Evans, Jonathan St. B. T., 18Eysenck, Hans, 27, 29

FFaricy, Christopher, 112Fazekas, Zoltan, 20Fazio, Russell, 13Feldman, Stanley, 155Fernbach, Philip, 14Festinger, Leon, 13, 41Fields, James, 14Fiorina, Morris, 11Fishkin, James, 12Fiske, Donald, 29, 32, 76Five Factor Model, 29Flanigan, William, 9fMRI brain imaging, 16, 21Folk Theorem, 4, 7, 131Folk theory, 4, 7, 37, 40, 43, 62, 88,

139, 140Food Stamps, 3, 7, 35, 36, 65, 68, 72,

74, 85, 90, 95, 101, 104, 111,112, 116, 118, 120, 126–127,132, 135, 149, 155, 159

Forbes, Chad, 31Fowler, James, 5, 20, 130Framing, 42, 43, 64Free, Lloyd, 40Frey, Dieter, 13Freyer, Felice, 67Fukui, Yoshitaka, 14

INDEX 197

Funnel of causality, 9Furnham, Adrian, 77, 78

GGadarian, Shana Kushner, 24Gaines, Brian, 6, 68Gainotti, Guido, 19Gallego, Aina, 30, 78Galston, William, 12Gamm, Gerald, 83Gélineau, Franҫios, 11Genetics, 7, 8, 20, 21, 27, 29, 31,

38n2, 163n1Gerber, Alan, 6, 11, 26, 30–33, 44, 52,

77, 78, 132–133, 144, 153, 155Giacino, Joseph, 19Gilens, Martin, 111, 112Goidel, Robert, 12Goldberg, Lewis, 29, 32, 76, 115Gomez, Brad, 11Goodnough, Abby, 39Goren, Paul, 9Gosling, Samuel, 29, 32, 33, 153Government, 1, 3–4, 8, 36, 37n1, 40,

42, 71, 72–75, 87, 90, 92–93,110–118, 120, 123, 125–127,128n8, 130, 135, 139–140, 144,157, 158, 163n1, 166n8

Gozzi, Marta, 21Grant, Bridget, 73Gray, Jeffrey, 24Graziano, William, 30, 115Grimm, Simone, 28Groenendyk, Eric, 24Gruber, K. L., 13, 67Gruszczynski, Michael, 20

HHa, Shang, 6, 26, 30, 33, 67, 68, 78, 90Hacker, Jacob, 112, 113

Haggard, Patrick, 16Hall, Richard, 89Halperin, Karen, 6, 26, 30, 32, 44,

77, 78Harmon-Jones, Eddie, 41Hatemi, Peter, 6, 20, 38n2, 65n3,

163n2, 164n3Haynes, John-Dylan, 16Healthcare, 112, 116, 123,

157, 161Heckman probit model, 97Heider, Fritz, 13Helms, Ann Doss, 109, 143Henry, P. J., 15Herbet, Guillaume, 19Heritability, 27Herszenhorn, David, 3Heuristics, 1, 5, 15, 22Hidden welfare state, 112, 114, 115,

116, 123, 125–128, 166Hillygus, Sunshine, 12Hochschild, Jennifer, 83, 140Hogan, Alexandra, 24Holbrook, Allyson, 81, 155Hopwood, Christopher, 26Hot Cognition, 140Howard, Christopher, 112Hsieh, Po-Jang, 16Huang, Yu-Feng, 18Huber, Gregory, 11, 144Hyman, Herbert, 8

IIdeal democratic citizen, 24, 88, 89,

103–107, 130, 139, 140, 142,145, 147n1, 166n1

Ignorance, 11, 12, 14, 35, 67–69, 72,87–88, 94, 111–112, 116,123–126, 130, 135

Ikeda, Haruka, 28Inouye, Jillian, 31

198 INDEX

Insurance, 1–4, 35, 36, 37n1, 39–40,73–75, 82–83, 85, 89, 92–95, 98,101, 109, 110–112, 114,116–120, 123–127, 128n8, 132,135–137, 149, 155, 157, 163n2,166n8

Insurance Subsidies, 35, 93, 132, 135,137, 149, 155

Iowa Gambling Task (IGT), 17Iturbe-Ormaetxe, Iñigo, 112Iyengar, Shanto, 15, 43Izatt, Gregory, 18

JJackson, John, 11Jacoby, William, 112Jensen, Thomas, 7Jensen-Campbell, Lauri, 30, 31Jerit, Jennifer, 6, 35, 68Jo, Se Hee, 30, 78Joel, Samantha, 21John, Oliver, 29Johnson, Wendy, 27Joint Committee on Taxation, 75, 116Joseph, Jay, 20Joslyn, Mark, 41Jost, John, 6, 22, 26, 33, 68Judd, Charles, 124Junn, Jane, 12

KKahan, Dan, 6, 14, 22, 35, 69, 70, 71,

142, 144, 146Kahneman, Daniel, 43Kammrath, Lara, 30, 115Kanai, Ryota, 21Kandler, Christian, 20, 26Kaplan, Jonas, 21Kapogiannis, Dimitrios, 28, 31Keeter, Scott, 10, 45, 61

Keith, Bruce, 46, 160Kenny, David, 124Key, V.O., 10Kiefer, Markus, 18Kim, Bo-Hye, 28Kim, Seokho, 78Kinder, Donald, 4, 15, 43Kingdon, John, 89Klas, Mary Ellen, 73Knutson, Kristine, 21, 70Kramer, Gerald, 11Krastev, Sekoul, 22Krosnick, Jon, 15, 81, 155Krueger, Robert F., 27Kruger, Justin, 14Kühn, Simone, 18Kuklinski, James, 5Kunda, Ziva, 13Kunde, Wilfried, 18Kunst-Wilson, William Raft, 19

LLadha, Krishna, 93Lau, Richard, 13, 35, 90, 94, 146Lavine, Howard, 13, 81, 155Lazarsfeld, Paul, 4, 8, 9, 12Lebo, Matthew, 13Leeper, Thomas, 13, 68Lei, Xu, 28Lenz, Gabriel, 7, 11, 14, 144Lepianka, Dorota, 115Letkiewicz, Jody, 31Levendusky, Matthew, 7Lewandowsky, Stephan, 13Lewis-Beck, Michael, 11Libet, Benjamin, 16Littvay, Levente, 20Lodge, Milton, 1, 5, 13, 22, 23, 68,

70, 130Loehlin, John, 27Looking glass perception, 14

INDEX 199

Lord, Charles, 13Lövdén, Martin, 31Luciano, Michelle, 27Lupia, Arthur, 142

MMacKuen, Michael, 24Mann, Frank, 27Marcus, George, 5, 23, 130Markus, Gregory, 11Masci, David, 74Mattila, Mikko, 30, 78McAdams, Dan, 26McCannon, Bryan, 143McClurg, Scott, 90McCrae, Robert, 29, 31, 32, 33, 76McDermott, Rose, 6McGregor, Michael, 41McKelvey, Richard, 1McNaughton, Neil, 24Media, 5, 8, 14, 106, 151Medicare, 4, 116, 157MeInyk, Darya, 15Mendelberg, Tali, 111Mettler, Suzanne, 3, 4, 7, 8, 36, 74,

75, 111, 112, 113Mind, 5, 7, 8, 13, 15, 16, 18–19,

21–26, 31, 37, 61, 62, 65, 70,100, 104, 121, 130, 131, 139,142, 147

Moderate, 46–48, 49–50, 56–57, 63,64, 65n6, 68, 132, 134, 138,141, 160, 164

Molden, Daniel, 18Mondak, Jeffrey, 6, 26, 30, 32, 33, 43,

44, 53, 77–78, 124, 132Montag, Christian, 27Montoro, Pedro, 19Morsella, Ezequiel, 15Motivated reasoning, 6, 22, 24, 35, 65,

68–71, 86n1, 141, 146, 164n1

Mount, Michael, 31Mroczek, Daniel, 26Mullainathan, Sendhil, 41Mulligan, Kenneth, 81, 155Mullinix, Kevin, 142Mummolo, Jonathan, 142

NNannestad, Peter, 11Nawara, Steven, 22Neuroticism, 28–29, 31, 32, 44,

53, 137Neuroscience, 7, 8, 16, 19, 22Newell, Ben, 18Nickerson, Raymond, 13Niemi, Richard, 12Nir, Lilach, 13Nisbett, Richard, 18Noftle, Erik, 77Norpoth, Helmut, 11

OObamacare, 1, 2, 39, 110O’Connor, Melissa, 77, 78Oberski, Daniel, 30, 78Odbert, Henry, 29Oil Production, 35, 72, 74, 76, 79, 81,

82, 131, 133, 136, 138, 158Okbay, Aysu, 28Olivola, Christopher, 18Olson, James, 13, 14Omura, Kazufumi, 28On-line, 22, 23Operational Ideology, 41, 43–45,

49–56, 63, 132, 134, 136, 161Ordeshook, Peter, 1, 90Osborne, Danny, 32, 44, 53Osterhammel, Jürgen, 129Ozer, Daniel, 33

200 INDEX

PPacek, Alexander, 11Page, Benjamin, 10, 62Paldam, Martin, 11Parker, Ashley, 2Paunonen, Sampo, 77, 78Paynter, Christopher, 18Peatfield, Nicholas, 18Pérez, Efrén, 18Perry, Ryan, 33, 133Personality, 29, 37, 43, 49, 58, 62,

65n4, 96, 153Petersen, Michael Bang, 14Pierce, Douglas, 68Plato, 12, 130Pluess, Michael, 28Pluralistic ignorance, 14Policy Beliefs, 42, 48, 89, 134Policy Preferences, 7, 25, 34, 35–36,

40, 46, 49, 63, 65n1, 67, 85, 88,90, 93, 95, 96, 98–99, 112, 131,133, 135, 137–138, 164

Political Behavior, 4–6, 8–9, 19, 22,25–26, 35, 37, 71, 79, 103, 147

Political Knowledge, 12, 14, 32, 36,52, 57, 58, 61, 63–65, 67, 71,76–79, 82, 85, 90, 100–102,104, 106, 126, 134, 138,149, 159

Political Party, 7, 73, 85, 101, 103,131, 135, 137, 138, 149

Political Personality Success andFailure (PPSF), 49–50, 71,72–73, 75, 79, 85, 89, 90, 91,115, 116, 133, 149, 151, 155

Political Psychology, 19Popkin, Samuel, 1, 5, 10, 72Popp, Elizabeth, 40, 46, 63Power, Robert, 28Przeworski, Adam, 130Psychological approaches to

politics, 22–24

Psychological immune system, 15Public Policy, 4, 6, 67, 76, 84, 85,

88, 89, 92, 103, 121, 136, 137,146, 147

Putnam, Robert, 12

QQuintelier, Ellen, 30, 78Quirk, Paul, 5

RRadcliff, Benjamin, 11Ramesar, Rajkumar, 27Redlawsk, David, 1, 13, 35, 90, 94,

142, 146Reuter, Martin, 27Reyna, Christine, 15Riker, William, 90Ritter, Simone, 18Roberts, Brent W., 26, 31, 159Robins, Richard, 77Rokeach, Milton, 13Romney, Mitt, 1Rosenstone, Steven, 12Ross, Lee, 7Rothbart, Mary, 31Rousseau, Jean-Jacques, 11, 130Rudolph, Thomas, 13, 40, 46, 63Rule, Nicholas, 21

SSackett, Paul, 77Same-Sex Marriage, 35, 52, 65, 68,

74, 85, 89, 90, 92, 94, 95,98, 100, 104, 132, 133, 149,152, 158

Sampaio, Adriana, 28Savitz, Jonathan, 27

INDEX 201

Schaefer, Michael, 28Schema, 5Schlegel, Alexander, 16Schoen, Harald, 6, 26, 30, 31, 32, 78Scholer, Abigail, 30, 115Schram, Sanford, 111Schuman, Howard, 14Schumann, Siegfried, 6, 26, 31, 32Schumpeter, Joseph, 10, 40, 69Schurer, Stefanie, 26Schwabe, Inga, 21Schwartz, Shalom, 33Selective Perception, 68, 142Servaas, Michelle, 28Shaker, Lee, 10Shanks, David, 18Shapiro, Robert Y., 10, 62, 88Sheatsley, Paul, 8Shepherd, Joanna, 143Sherman, David, 15Sherrod, Drury, 13, 14Shields, Todd, 12Shy Conservative, 48, 49, 50, 65n6,

164n6Shy Liberal, 48, 49, 50, 56–58, 61,

63–64, 132, 134Sibley, Chris, 32, 33, 44, 53, 133Sides, John, 142Slothuus, Rune, 13, 68Smith, D. J., 28Smith, Kevin, 20Sokhey, Anand Edward, 90Soon, Chun Siong, 16Stanley, Jason, 111Stegmaier, Mary, 11Steinbrecher, Markus, 6, 26, 30, 78Stevenson, Randy, 11Stimson, James, 6, 9, 10, 34, 40,

42–43, 45, 47, 49, 53, 57, 60, 63,64, 112, 136, 161

Stokes, Donald, 9Stroop, John Ridley, 18

Submerged Welfare state, 7, 74, 75,109, 112, 118, 123, 124, 135

Sullivan, Laura, 113Sun, Lena, 110Sweeney, P. D., 13Symbolic Ideology, 44–47, 49–50,

52, 57, 58, 59, 63, 64, 68, 134,136, 160

TTaber, Charles, 1, 5, 13, 22–24, 68,

70, 130Takeuchi, Hikaru, 21Taki, Yasuyuki, 28Tax expenditure, 3, 112–114, 116,

123, 125, 127Theocharis, Yannis, 30, 78Third person effect, 14Todd, Constable, 28Todorov, Alexander, 18Tomz, Michael, 7, 14Tormala, Zakary, 81, 155Turkheimer, Eric, 27Tusche, Anita, 21, 23Tversky, Amos, 43Twin studies, 20

UUnconscious, 5–8, 15–19, 21–26, 31,

34, 37, 48, 61, 62, 67, 70, 104,129–130, 140, 142–143,146–147

Urbach, Blake, 73Usher, Marius, 18

VValentino, Nicholas, 24Van Boven, Leaf, 14

202 INDEX

van den Berg, Stephanie, 28van Gaal, Simon, 18Van Houweling, Robert, 7, 14Vecchione, Michele, 30, 32, 33, 78Vedel, Anna, 31, 77Verba, Sidney, 12Verhulst, Brad, 20, 38n2, 65n3,

163n2, 164n3Vidmar, Neil, 13Volsky, Igor, 128n1, 165n1Voss, Joel, 18Voting/voters, 5, 9, 10–11, 20, 24,

30, 37, 41, 62, 68, 88, 90, 94,140, 142–145

WWagner, Markus, 14, 144Walmsley, Philip, 77Washington, Ebonya, 41Wayne, Leslie, 3Weeks, Brian, 24, 142Weerman, Frank, 14Weibel, Sebastien, 19Weinschenk, Aaron, 30, 32, 33, 77, 78Welfare, 7, 35, 36, 40, 50, 65, 68, 71,

73–76, 81, 85, 86n2, 89–90,92–100, 104, 109–112,114–120, 123–127, 128n8,

131–133, 135–136, 138, 149,155–156, 161, 164n2, 166n8

Westen, Drew, 21, 70White, Allen, 14Williams, Carol J., 13Wilson, J. Matthew, 11Wilson, Timothy D., 18Wlezien, Christopher, 112Wolfinger, Raymond, 12Woo, Sang Eun, 32Wright, Christopher, 28

YYacoubian, George, 73Yoshimoto, Sanae, 19Yost, Berwood, 142Young, Jacob, 14Youngman, Sam, 39Yuchtman, Noam, 143

ZZajonc, Robert, 19Zaller, John, 11, 13, 45, 81, 144, 155Zamboni, Giovanna, 21Zanna, Mark, 13Zimbardo, Philip, 32Zingale, Nancy, 9

INDEX 203

personality and the challenges of democratic governance: how unconscious thought influences...

Documents