UC IrvineUC Irvine Electronic Theses and Dissertations

TitleSelf-Enhanced Reports of Happiness: Implications for Subjective Well-Being Research and the Ideological Happiness Gap

Permalinkhttps://escholarship.org/uc/item/3cf4k128

AuthorWojcik, Sean Philip

Publication Date2015

Copyright InformationThis work is made available under the terms of a Creative Commons Attribution-NoDerivatives License, availalbe at https://creativecommons.org/licenses/by-nd/4.0/ Peer reviewed|Thesis/dissertation

eScholarship.org Powered by the California Digital LibraryUniversity of California

UNIVERSITY OF CALIFORNIA, IRVINE

Self-Enhanced Reports of Happiness: Implications for Subjective Well-Being Research and the

Ideological Happiness Gap

DISSERTATION

To be submitted in partial satisfaction of the requirements for the degree of

DOCTOR OF PHILOSOPHY

in Psychology and Social Behavior

by

Sean P. Wojcik, M.A.

Dissertation Committee:

Peter H. Ditto, Ph.D., Chair

Linda Levine, Ph.D.

Roxane Cohen Silver, Ph.D.

2015

ii

© 2015 Sean P. Wojcik

iii

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS v

CURRICULUM VITAE viii

ABSTRACT OF THE DISSERTATION xiii

CHAPTER 1: Introduction 1

References 8

CHAPTER 2: Motivated Happiness: Self-Enhancement Inflates Self-Reported 10

Subjective Well-Being

Abstract 11

Introduction 12

Study 1 15

Figure 1: Mean social-comparative judgment scores on standard and SWB

related traits 17

Table 1: Descriptive statistics and zero-correlations for all study variables 17

Study 2 19

Table 2: Regression results and meta-analyzed effect sizes for Studies 2,

3a, and 3b 21

Study 3a 23

Figure 2: Self-reported happiness as a function of SDE and happiness

desirability condition 25

Figure 3: Satisfaction With Life Scale scores as a function of SDE and

happiness desirability condition 26

Study 3b 26

Meta-Analysis 27

General Discussion 27

Coda 30

References 31

Acknowledgements 36

CHAPTER 3: Conservative Self-Enhancement 37

Abstract 38

Introduction 39

Method 40

Results 41

Meta-analysis 41

Table 1: Effect sizes and meta-analysis of political ideology variables predicting

self-enhancement 42

Discussion 43

References 44

iv

Acknowledgements 47

Supplementary Materials 48

Supplementary Measures 48

Supplementary Table 1: Regression coefficients for general liberalism-

conservatism and demographic characteristics predicting self-enhancement 51

Supplementary Table 2: Regression coefficients for social liberalism-

conservatism and demographic characteristics predicting self-enhancement 53

Supplementary Table 3: Regression coefficients for economic liberalism-

conservatism and demographic characteristics predicting self-enhancement 55

Supplementary Table 4: Regression coefficients for moral liberalism-

conservatism and demographic characteristics predicting self-enhancement 57

CHAPTER 4: Conservatives Report, But Liberals Display, Greater Happiness 58

Abstract 59

Introduction 60

Study 1 62

Table 1: Unstandardized coefficients from mediation analysis of self-reported

life satisfaction 62

Study 2 64

Table 2: Unstandardized regression coefficients for linguistic analysis and smile

analysis for politicians in 2013 US Congressional Record 65

Study 3 67

Table 3: Odds of emotion-related words appearing in tweets of Republican Party

subscribers relative to Democratic Party subscribers 68

Study 4 68

General Discussion 69

References 72

Acknowledgements 75

Supplementary Methods and Materials 76

Study 1 76

Study 2 76

Study 3 79

Study 4 80

CHAPTER 5: Epilogue 82

References 89

v

ACKNOWLEDGEMENTS

I have always been extremely lucky. I’m sure that many years from now, when I look

back on my time in graduate school, my most prominent memories will be of the incredible

people with whom I have been fortunate enough to have shared this time. In the five and a half

years I’ve been at UCI, I’ve found myself surrounded by extremely intelligent, thoughtful,

supportive, and energizing people. I can’t possibly begin to convey how appreciative I am of the

faculty, colleagues, friends, and family who made this time so special.

But I can try.

First, I want to express my gratitude to my dissertation committee, Linda Levine, Roxy

Silver, and Pete Ditto. Thank you for your countless hours of conversation, for your creative,

theoretical, and methodological insights, and for the honor of having your names forever

associated with my dissertation. Most of all, thank you for your unconditional support and

encouragement during the last several years. Throughout my coursework, teaching, research,

dissertation, and job search, my committee has been a near-constant source of encouragement

and enthusiasm.

This is especially true of my primary advisor and dissertation chair. In 2009, I moved all

the way across the country just because I wanted to learn how to do research like Pete Ditto. But

for myself and countless others at UCI, Pete has become a role model not only as a researcher,

but also as a person. He has shown how to be a world-class scholar while also maintaining a

balance that emphasizes the things that are ultimately most important in life. He is the kind of

teacher, mentor, friend, father, and husband that I hope to become, and I’ll always be grateful to

have learned so much under his guidance. I’m also unendingly thankful to Geri Ditto (my “West

vi

Coast Mom”) for her hospitality, generosity, positivity, and for always making me feel like I’m

at home and a part of her family.

I am also indebted to a number of people who have contributed to my education and

training in ways that both directly and indirectly contributed to this dissertation. Thank you to

Virginia Kwan, who gave me my first chance in social psychology, taught me how to think (and

write) scientifically, and first piqued my interest in self-enhancement research. I would also like

to thank the PSB departmental staff, including Ellen, Suzy, Donna, Tiffany, and especially Toni,

whose warmth and positivity has epitomized the supportive culture that has made our department

such a fantastic place to work for so many years. I am also appreciative of the faculty members

who have formally and informally advised my research along the way, especially Chris Bauman

and Susan Charles (who served on my advancement committee), Eric Knowles, Joanne Zinger,

and JoAnn Prause.

I have also been very lucky in the collaborations that have enhanced my development as

a researcher—particularly in my association with the team at YourMorals.org, where I’ve been

lucky enough to learn from a number of people I deeply admire. Jon Haidt, our fearless leader,

has been a role model in theoretical ingenuity, professionalism, ambition, and kindness. Jesse

Graham has shown how a psychologist can use his cleverness and creativity not only to be

insightful and generative, but also humble and hilarious. Sena Koleva has been a terrific friend

and collaborator from day one, whose compassion and enthusiasm are contagious. Matt Motyl

has been an exemplary model of an efficient and productive young scholar. And Ravi Iyer, who

has taught me the power of using data and technology to answer complex questions about human

nature, has hugely inspired my aspirations for a career of large-scale, high impact work.

vii

It is hard to imagine going through graduate school with anyone other than the group of

friends I’ve made along the way. Thank you to my (“Dittles”) lab-mates, my cohort (the

“Platinum ‘09-‘10s”), and everyone else (you know who you are), especially Sena, Andrew,

David, Matt, Brittany, Cory, Jeff, Christina, Chris, Eric, Becky, Jennie, Karen, Jake, Liz, Tyler,

Conor, and Becca. I often think back on all of the great memories, adventures, and

misadventures we’ve shared, and feel like I have gotten away with something. No one ever said

getting a Ph.D. was supposed be so much fun.

I am also grateful to have shared so much of my time with Emily Urban. There isn’t any

short-hand way for me to express the depth of my gratitude and appreciation for Emily, or to

explain how important she has been to my success and happiness, so I’ll keep this short and

simple: Thank you, Emily, for being the best person I know.

Last but not least, I want to thank my older brother, Ryan, and my parents, Susan and

Philip Wojcik. It would be trite to distill their significance to a few short sentences, but I want to

acknowledge that they are the reason why I have always felt so lucky. Thank you Ryan, Mom,

and Dad, for being the most influential teachers, mentors, friends, and role models I’ve ever had.

viii

CURRICULUM VITAE

SEAN P. WOJCIK

Dept. of Psychology & Social Behavior Email: swojcik@uci.edu

4201 Social & Behavioral Sciences Gateway Tel: (973) 600-7216

Irvine, CA 92697 Web: seanwojcik.com

Education_____________________________________________________ Ph. D. University of California, Irvine (2015)

Major Concentration: Social/Personality Psychology

Minor Concentration: Quantitative Methods

M. A. Social Ecology, University of California, Irvine (2012)

B. S. Northeastern University (2007)

Psychology, Magna Cum Laude

Academic Honors and Awards____________________________________ 2014 Time-Sharing Experiments in the Social Sciences (TESS) Proposal Acceptance

2013 Society for Personality and Social Psychology Student Travel Award

2011, 2012 School of Social Ecology Graduate Mentor Award

2010-2014 School of Social Ecology Summer Research Fellowship

2009 Social Ecology Graduate Research Fellowship, University of California, Irvine

2003-2007 Dean’s Scholarship, Northeastern University

Research Interests______________________________________________ Motivated reasoning, judgment and decision-making, political and moral psychology, self-

enhancement, happiness, subjective well-being, Big Data.

Publications___________________________________________________ 11. Wojcik, S. P., & Ditto, P. H. (under review). Conservative self-enhancement. Manuscript

under review.

10. Wojcik, S. P., Hovasapian, A., Graham, J., Motyl, M., & Ditto, P. H. (in press).

Conservatives report, but liberals display, greater happiness. Science.

9. Ksendzova, M., Iyer, R., Hill, G., Wojcik, S. P., & Howell, R. T. (in press). The portrait of a

hedonist: The personality and ethics behind the value and maladaptive pursuit of

pleasure. Personality and Individual Differences.

8. Ditto, P. H., Wojcik, S. P., Chen, E., Grady, R., & Ringel, M. (in press). Political bias is

tenacious. Commentary on target article, Political diversity will improve social

psychological science (J. Duarte, J. Crawford, C. Stern, J. Haidt, L. Jussim, & P. E.

Tetlock). Behavioral and Brain Sciences.

7. Wojcik, S. P., & Ditto, P. H. (2014). Motivated happiness: Self-enhancement inflates self-

reported subjective well-being. Social Psychological and Personality Science, 5, 825-

834.

6. Lai, C. K., Marini, M., Lehr, S. A., Cerruti, C., Shin, J. L., Joy-Gaba, J. A., Ho, A. K.,

Teachman, B. A., Wojcik, S. P., … & Nosek, B. A. (2014). Reducing implicit racial

ix

preferences: I. A comparative investigation of 17 interventions. Journal of Experimental

Psychology: General, 143, 1765-1785.

5. Johnson, K. M., Iyer, R., Wojcik, S. P., Vaisey, S., Miles, A., Chu, V., & Graham, J. (2014).

Ideology-specific patterns of moral disengagement predict intentions not to vote.

Analyses of Social Issues and Public Policy, 14, 61-77.

4. Graham, J., Haidt, J., Koleva, S., Motyl, M., Iyer, R., Wojcik, S. P., & Ditto, P. H. (2013).

Moral Foundations Theory: The pragmatic validity of moral pluralism. Advances in

Experimental Social Psychology, 47, 55-130.

3. Kwan, V. S. Y., Wojcik, S. P., Miron-shatz, T., Votruba, A., & Olivola, C. (2012). Effects of

symptom presentation order on perceived disease risk. Psychological Science, 23, 381-

385.

2. Ditto, P. H., Liu, B., & Wojcik, S. P. (2012). Is anything sacred anymore? Commentary on

target article, Mind perception is the essence of morality (K. Gray, L. Young, & A.

Waytz), Psychological Inquiry, 23, 155-161.

1. Kwan, V. S. Y., Diaz, P., Wojcik, S. P., Matula, K. A., Kim, S. H. Y., & Rodriguez, K.

(2011). Self as the target and the perceiver: A componential approach to self-

enhancement. Psychological Studies, 56, 151-158.

Manuscripts in Preparation_______________________________________ 1. Wojcik, S. P., Gampa, A., Motyl, M., Nosek, B., & Ditto, P. H. Predictably (and politically)

illogical: Partisan belief bias in the evaluation of logical syllogisms.

2. Wojcik, S. P., Ekins, E., & Haidt, J. The moral underpinnings of support for the Tea Party

movement.

3. Ditto, P. H., Liu, B., Wojcik, S. P., Clark, C. J., Chen, E., & Grady, R. Motivated reasoning

and ideology: A meta-analysis of biased assimilation among liberals and conservatives.

4. Howell, R., Iyer, R., & Wojcik, S. P. Experiential purchasing predicts greater explicit and

implicit happiness.

Professional Presentations________________________________________ 12. Wojcik, S. P. (2015, February). Conservatives report, but liberals display, greater

happiness: Behavioral indicators of happiness from Big Data. Data Blitz presented at the

16th annual meeting of the Society for Personality and Social Psychology, Long Beach,

CA.

11. Wojcik, S. P. (2014, November). Motivated happiness: Towards a multi-method approach to

measuring subjective well-being. Colloquium presented at University of California,

Irvine.

10. Liu, B., Clark, C. J., Wojcik, S. P., Chen, E., Grady, R., & Ditto, P. H. (2014, July).

Motivated reasoning and ideology: A meta-analysis of liberals’ and conservatives’ bias

in political contexts. Paper presented at the 2014 meeting of the International Society of

Political Psychology, Rome, Italy.

9. Wojcik, S. P., Ditto, P. H. (2014, February). Self-enhancement explains the liberal-

conservative happiness gap. Poster session presented at the 15th annual meeting of the

Society for Personality and Social Psychology, Austin, TX.

8. Ekins, E. M., Wojcik, S. P., & Haidt, J. (2013, August). The moral intuitions of the Tea Party

movement: Liberty or Proportionality, or both? Paper presented at the annual meeting of

the American Political Science Association, Chicago, IL.

x

7. Wojcik, S. P., Ditto, P. H., Haidt, J., Graham, J., Koleva, S., Iyer, R., & Motyl, M. (2013,

January). Bridging the happiness gap: Self-enhancement explains ideological differences

in happiness. Poster session presented at the 14th annual meeting of the Society for

Personality and Social Psychology, New Orleans, LA.

6. Ditto, P. H., & Wojcik, S. P. (2013, January). Moral intuitionism and the politics of self-

enhancement. In D. Sherman (Chair), Situated ethics: How moral judgments and

behaviors are contaminated by situational cues. Symposium presented at the annual

meeting of the Society for Personality and Social Psychology, New Orleans, LA.

5. Ditto, P. H., & Wojcik, S. P. (2012, October). Predictably (and politically) illogical: partisan

bias in the evaluation of logical syllogisms. In L. Van Boven (Chair), Political

Polarization. Paper presented at the annual meeting of the Society of Experimental Social

Psychology, Austin, TX.

4. Miron-shatz, T., Kwan, V. S. Y., Wojcik, S. P., Votruba, A., & Olivola, C. Y. (2012,

September). Effects of symptom presentation order on perceived disease risk. Paper

presented at Medicine 2.0: Social Media, Mobile Apps, and Internet/Web 2.0 in Health,

Medicine, and Biomedical Research, Boston, MA.

3. Wojcik, S. P., Ditto, P. H., Iyer, R., Koleva, S., Haidt, J., Graham, J., & Motyl, M. (2012,

July). Overconfidence in partisan politics: Ideological correlates of self-enhancement

bias, belief bias, and logical reasoning performance. Poster session presented at the

annual meeting of the International Society of Political Psychology, Chicago, IL.

2. Arnautovic, I., Wojcik, S. P., & Ditto, P. H. (2012, May). Predictably illogical: Belief bias in

the Israeli-Palestinian conflict. Poster session presented at the 2012 Undergraduate

Research Opportunities Program Symposium, Irvine, California.

1. Wojcik, S. P., Ditto, P. H., Koleva, S., Iyer, R., Graham, J., & Haidt, J. (2011, January).

Libertarian, conservative, or something in between? The psychology of support for the

Tea Party. Poster session presented at the annual meeting of the Society for Personality

and Social Psychology, San Antonio, TX.

Teaching and Mentoring Experience______________________________ 2014, 2011 Graduate Student Mentor, University of California, Irvine, Undergraduate

Research Opportunities Program

Summer 2014 Lecturer, The Social Animal, University of California, Irvine

Summer 2014 Guest Lecturer, Research Methods, Saddleback College

Spring 2013 Guest Lecturer, Research Design, University of California, Irvine

2010, 2011 Graduate Student Mentorship Program, University of California, Irvine

2011 Mentor, UCI Undergraduate Researcher Opportunities Program

Spring 2011 Guest Lecturer, Research Design, UCI, Spring 2011

2010-2014 Teaching Assistant, University of California, Irvine, 2010 - present

Teaching Interests______________________________________________ Research methods, statistics, judgment and decision-making, political psychology, moral

psychology, social psychology.

Statistical, Methodological, and Technical Proficiencies_______________ Statistical and methodological proficiencies:

Survey, experimental, and longitudinal design

xi

Qualitative and quantitative methods and analysis

Descriptive and inferential statistics

Univariate and multivariate OLS regression

Analysis of variance

Logistic regression

Meta-analysis

Hierarchical linear modeling

Structural equation modeling

Supervised and unsupervised machine learning

Statistical software proficiencies:

R, SPSS, Stata, Comprehensive Meta-Analysis, M-Plus

Data management and miscellaneous programming proficiencies:

R, SQL, PHP, Python, Javascript, HTML, CSS

Research and Work Experience_______________________________________ 2014–Present Research Scientist, Upworthy

Providing data-driven survey and user research to assess virality of social media

content and effectiveness of content in changing user awareness, attitudes, and

behaviors; interfacing with for-profit and non-profit clients to analyze

effectiveness of media content.

2014 Graduate Student Researcher, University of California, Irvine, PI: Linda Levine

Assisted with data analysis for project on emotion regulation strategies and

emotional recall, edited manuscripts and grant proposals

2010–Present Researcher, YourMorals.org

Conducting survey and experimental design, user research, data management,

data mining, data analysis, and web development for a large public education and

research platform (~600,000 registered users) that has been featured in the New

York Times, Wall Street Journal, Science, and Nature

2014–Present Researcher, BeyondThePurchase.org

Providing web development, study design, social media integration, and user

experience research for a growing online research platform

2014–Present Creator, CongressionalWordCount.org

Created interactive psycholinguistic analysis tool that compares word use

frequency among Democratic and Republican politicians within the United States

Congressional Record over 18 year period

2008-2009 Research Specialist, Princeton University, PI: Virginia S. Y. Kwan

Designed and administered experimental and survey research related to self-

perception and medical decision-making, assisted with writing of grant proposal

and manuscripts

Professional Service____________________________________________ 2014 Ad hoc reviewer, International Journal of Political Psychology

2014 Dept. of Psychology and Social Behavior Graduate Recruitment Representative

2012-2014 Elected Council Member, Palo Verde Residence Council, Irvine, CA

2011-2013 Dept. of Psychology and Social Behavior Communications Director

2011 SPSP Student Poster Award Reviewer

xii

Honorary and Professional Memberships___________________________ 2012–present Association for Psychological Science

2012–present International Society of Political Psychology

2009–present Society for Personality and Social Psychology

2003-2007 Northeastern University Honors Program

2006 Psi Chi National Honor Society, Northeastern University

2005 The Academy International Honor Society

2005 Golden Key International Honor Society

Selected Media Coverage________________________________________ Boston Globe, The bias fighters, September 21, 2014.

Forbes, Can you overcome inbuilt bias? August 14, 2014.

National Public Radio, How to fight racial bias when it’s silent and subtle, July 19, 2014.

Jezebel, If you’re biased and you know it, clap your hands (and then do something about it), June

3, 2013

Newsweek Magazine, It’s not a tumor! The psychology behind cyberchondria, April 9, 2012.

Fox News Channel, Sunday Housecall, Should you Google your symptoms?, April 2, 2012.

The Huffington Post, How symptoms are presented online may affect whether we think we have

the disease, March 18, 2012.

xiii

ABSTRACT OF THE DISSERTATION

Self-Enhanced Reports of Happiness: Implications for Subjective Well-Being Research and the

Ideological Happiness Gap

Sean P. Wojcik

Doctor of Philosophy in Psychology and Social Behavior

University of California, Irvine, 2015

Professor Peter H. Ditto, Chair

Research on subjective well-being (SWB; i.e., happiness) relies heavily on self-report

methodologies. This dissertation explores the influence of self-enhancement, or the tendency to

evaluate oneself in an unrealistically positive way, on self-reported judgments of happiness and

life satisfaction. Further, it examines how accounting for self-enhancement can influence—and

in some ways, even reverse—our understanding of happiness differences between political

conservatives and liberals. Three studies in Chapter 2 demonstrate a distorting effect of self-

enhancement on self-reports of SWB. It is shown that both dispositional levels of self-

enhancement and experimentally manipulated self-enhancement motivation lead to a tendency to

report SWB at unrealistically favorable levels in social comparative judgment tasks, as well as

on established scales of SWB. Chapter 3 of the dissertation examines group-level differences in

self-enhancement between political conservatives and political liberals. Meta-analyses of eight

measures of self-enhancement reveal significant positive relationships with general, economic,

social, and moral dimensions of political conservatism. Chapter 4 examines whether liberal-

conservative differences in self-enhancement explain why conservatives typically report greater

SWB than liberals. Chapter 4 finds that conservatives’ stronger tendency to engage in self-

xiv

enhancement fully mediates the ideology—SWB relationship. Three additional studies assess

happiness-related behaviors among liberals and conservatives using large archives of text from

the U.S. Congressional Record and Twitter, along with photographs from the U.S. Congressional

Pictorial Directory and LinkedIn. Chapter 4 shows that liberals use more frequent positive

emotional language and less frequent negative emotional language, and also that liberals smile

more intensely and genuinely than conservatives. These findings support the hypothesis that

conservatives’ reports of SWB are attributable to a self-enhancing style of self-report. In the

Epilogue, practical implications of the research are discussed, specifically regarding the use of

self-report methods to assess and compare group-level differences in SWB for use in public

policy decision making. Theoretical implications of the research are also discussed. In particular,

the nature of the long-debated relationship between self-enhancement and well-being is

discussed, as are recommendations for measuring SWB, along with potential implications for re-

conceptualizing multiple (i.e., defensive) manifestations of happiness.

1

CHAPTER 1:

Introduction

2

Introduction

Virtually all people desire to be happy (Diener, Sapyta, & Suh, 1998; Diener, 2000).

Aristotle regarded happiness as the “supreme good”, in that it can be thought of as the desired

end to all means, and that once attained, other desires disappear (Diener, 1994). To this day,

happiness is consistently judged as one of the most important human experiences; it is a

necessary prerequisite to living the “good life” and is cherished more deeply than wealth and

even moral virtue (King & Napa, 1998). It is not surprising then, that research on the causes and

consequences of happiness has become increasingly prominent not only in psychology, but also

in related areas, including economics and public policy research. Along with this broadening

scope of scholarly attention comes a need for improved methodological rigor, and the present

dissertation aims to advance this aim by illuminating the complex relationship between

happiness and self-enhancement, as well as the implications of this relationship for

understanding group differences in happiness.

The study of subjective well-being (SWB; used interchangeably here with the more

colloquial “happiness”) involves the measurement of multiple components of happiness,

including life satisfaction (i.e., global evaluations about one’s life), satisfaction with important

life domains (e.g., relationship satisfaction, work satisfaction), and levels of positive and

negative affect (Lyubomirski, King, & Diener, 2005). Life and domain satisfaction judgments

are understood as evaluative or cognitive components of well-being, and positive-negative

affective balance is regarded as the on-line, emotional, and mood-related aspect of SWB (Diener,

Suh, Lucas, & Smith, 1999). Researchers have also differentiated these “hedonic” components of

well-being from “eudaimonic” facets of wellbeing, which emphasize living life in a full and

meaningful way (Deci & Ryan, 2008). Despite the theoretical breadth of these definitions, the

3

methodologies used to assess the various components of SWB are surprisingly uniform—that is,

SWB research relies almost exclusively on the use of self-report questionnaires.

Given the pervasiveness of self-report measures in the SWB literature, a key assumption

is that self-reports provide valid, comprehensive measurements of happiness that can be

interpreted at face value. Studies have demonstrated the predictive validity of these measures

(e.g., Sandvik, Diener, & Seidlitz, 1993), and the construction of these scales often involves

cross-validation with non-self report methods (e.g., using ratings by romantic partners, close

others, and outside observers; Sandvik, Diener, & Seidlitz, 1993; Paulhus & Vazire, 2007).

However, there are reasons to suspect that these practices may not completely eliminate the

impact of certain response biases. It is known, for example, that judgments by romantic partners

are often unrealistically positive (Brown & Han, 2012), as are judgments by knowledgeable

ingroup members (Taylor & Doria, 1981). Additionally, even seemingly objective third-party

observers may be inaccurate in their assessments of others’ happiness if deceptive manifestations

of happiness are portrayed convincingly (Von Hippel & Trivers, 2011). Clearly, verifying the

validity of SWB measures is difficult, especially because happiness lacks an objective accuracy

criterion for validation.

Of particular concern for the measurement of SWB is the common tendency to

exaggerate assessments of one’s positive qualities, known as self-enhancement. In moderation,

self-enhancement is generally understood as a typical feature of healthy, everyday thinking

(Taylor & Brown, 1988), although the degree to which individuals engage in self-enhancement

varies both between and within social groups (Taylor & Brown, 1988; Colvin, Block, & Funder,

1995; Heine, Kitayama, & Hamamura, 2007; Eriksson & Funcke, 2014; Loughnan, et al., 2011).

The variation of self-enhancement between and within groups is particularly important given the

4

frequency with which SWB researchers compare and contrast self-reports of SWB. To the extent

that reports of happiness can be distorted by self-enhancing tendencies, and to the extent that

groups may differ in these tendencies, SWB research may inadvertently capture differences in

self-enhancement, rather than differences in emotional experience.

The present dissertation aims to clarify the extent to which self-enhancement inflates

reports of subjective well-being, and it goes on to examine how self-enhancement might explain

previously observed group-level differences in SWB. These aims are accomplished in three

separate papers that are presented here as Chapters 2, 3, and 4. Two of these chapters have

already been published at peer-reviewed journals (Chapter 2, Social Psychological and

Personality Science; Chapter 4, Science), and one is currently under peer-review (Chapter 3). To

maintain the integrity of the original papers, the entirety of each manuscript is presented,

including abstracts, references, supplementary materials, tables, and figures.

Chapter 2 of this dissertation directly assesses the extent to which judgments of SWB are

inflated by self-enhancement. In the first of three studies, participants made social-comparative

self-assessments across a number of traits and abilities, in which they judged whether they

possessed each trait dimension more or less than the average person. It was found that

participants routinely rated their happiness and life satisfaction as better than average, to a degree

that was similar to judgments of other positive traits that are commonly self-enhanced to

unrealistic degrees. Further, this “happier than average” effect was more pronounced among

individuals who were dispositionally prone to self-enhancement. This latter finding could be

indicative of either of two possibilities: either that individuals engaging in stronger degrees of

self-enhancement were truly happier and more satisfied with life than others (e.g., Taylor &

Brown, 1988), or that their judgments of happiness and life satisfaction were further examples of

5

self-enhanced traits. To differentiate between these two possible interpretations of Study 1, the

motivation to self-enhance along the dimension of happiness was experimentally manipulated in

Study 2. Individuals dispositionally high in self-enhancement motivation were more affected by

the experimental manipulation than those low in dispositional self-enhancement—that is, they

not only reported higher levels of SWB when happiness was manipulated to appear desirable, but

they also reported lower levels of SWB when it was manipulated to appear undesirable. Studies

3a and 3b replicated this finding on established scales of SWB, and an internal meta-analysis of

these results confirmed that both dispositional and situationally induced self-enhancement

motivation inflate self-reports of SWB.

Chapter 2 demonstrates the potential for self-enhancement to artificially inflate reports of

SWB. But to what extent does self-enhancement confound the comparison of SWB between

groups? Chapters 3 and 4 aim to address this question within the context of the ideological

happiness gap, in which political conservatives frequently report greater SWB than political

liberals. Past research has attempted to make sense of the ideological happiness gap by arguing

that conservatives are happier than liberals because they are able to justify troubling societal

inequalities (Napier & Jost, 2008), or because conservatives have stronger senses of personal

agency, control, optimism, and transcendent moral beliefs (Schlenker, Chambers, & Le, 2012).

Interestingly, each of these explanatory factors is conceptually related to self-enhancement

motivation. Chapter 3 lays the groundwork for testing whether or not self-enhancement is at play

in this relationship by examining the relative extent to which political liberals and conservatives

engage in self-enhancement. Self-enhancement was assessed using eight measures of the

tendency to exaggerate one’s positive qualities with large samples of online survey takers. An

internal meta-analysis of these studies found that four separate dimensions of political

6

conservatism (i.e., general, social, economic, and moral conservatism) predicted stronger degrees

of self-enhancement.

Chapter 4 directly extends this research by testing whether the ideological happiness gap

can be attributed to these group-level differences in self-enhancement. In the first of four studies,

a mediation model showed that conservatives’ tendency to self-enhance fully explained why they

reported greater life satisfaction than liberals. As with Chapter 2’s first study, the causal nature

of this relationship was not immediately clear: elevated reports of SWB could either be a

beneficial consequence of conservatives’ stronger tendency to self-enhance (e.g., Taylor &

Brown, 1988), or these reports could be examples of their tendency to evaluate the self in an

overly favorable way. If the latter interpretation were true, conservatives’ greater happiness

would likely be limited to self-report measures; thus, three additional studies examined

behavioral indicators of happiness among these groups. In Study 2, the linguistic properties of

text from the United States Congressional Record revealed more frequent positive emotionality

among liberal and Democratic politicians than among conservative and Republican politicians.

Further, FACS analysis of the smiling behavior of these politicians revealed more frequent, more

intense, and more genuine smiling behavior among liberals, relative to conservatives. The third

and fourth studies replicated these findings among the general public using large archives of text

from Twitter and publicly available photographs from LinkedIn. Together, these results suggest

that conservatives’ reports of greater SWB, relative to liberals, are likely due to their stronger

tendency to make flattering self-assessments. Further, analysis of happiness-related behavior

revealed greater happiness among liberals, rather than conservatives.

The studies presented in Chapters 2-4 fundamentally challenge the way researchers are

currently interpreting the relationship between self-enhancement and SWB. As a result, a number

7

of new research questions are raised by this research, and several of these are discussed in the

Epilogue (Chapter 5). First, it is argued that comparisons of SWB between non-randomized

groups may inadvertently capture differences in self-enhancing styles of self-assessment rather

than differences in emotional experience. Implications for the “gross national happiness”

movement, which aims to directly inform policy decision-making using SWB research, are

discussed. Second, new questions are raised regarding the adaptive value of self-enhancement.

Although the research presented here does not directly discount the potential benefits of self-

enhancement (e.g., Taylor & Brown, 1988), this research raises questions about the extent to

which the self-enhancement--well-being relationship is due to self-enhancement’s actual

benefits, relative to the distorting effect that it has on reports of well-being.

Finally, recommendations are made for re-conceptualizing how psychologists understand

the construct of happiness. It is recommended that research on SWB might benefit from

following in the methodological footsteps of the self-esteem literature, which has identified

multiple independent manifestations of high self-esteem that are not equally beneficial (e.g.,

defensive, fragile, and insecure forms of self-esteem; Kernis, 2003). Given that self-reports of

happiness are artificially inflated by self-enhancing distortions, and that self-enhancing groups

display less intense and less genuine happiness-related behavior than others, it is suggested that a

more nuanced nomenclature would aid our understanding of the multifaceted, complex, and

sometimes contradictory construct of happiness.

8

References

Brown, J. D., & Han, A. (2012). My better half: Partner enhancement as self-

Enhancement. Social Psychological and Personality Science, 3, 479-486.

Colvin, C. R., Block, J., & Funder, D. C. (1995). Overly positive self-evaluations and

personality: Negative implications for mental health. Journal of Personality and Social

Psychology, 68, 1152-1162.

Deci, E. L., & Ryan, R. M. (2008). Hedonia, eudaimonia, and well-being: An introduction.

Journal of Happiness Studies, 9, 1-11.

Diener, E. (1994). Assessing subjective well-being: Progress and opportunities. Social Indicators

Research, 31, 103-157.

Diener, E. (2000). Subjective well-being: The science of happiness and a proposal for a national

index. American Psychologist, 55, 34-43.

Diener, E., Sapyta, J. J., & Suh, E. (1998). Subjective well-being is essential to well-

being. Psychological inquiry, 9, 33-37.

Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: Three

decades of progress. Psychological Bulletin, 2, 276-302.

Eriksson, K., & Funcke, A. (2014). Humble self-enhancement: Religiosity and the better-than-

average effect. Social Psychological and Personality Science, 5, 76-83.

Heine, S. J., Kitayama, S., & Hamamura, T. (2007). Which studies test whether self-

enhancement is pancultural? Reply to Sedikides, Gaertner, and Vervea. Asian Journal of

Social Psychology, 10, 198-200.

Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological

Inquiry, 14, 1-26.

9

King, L. A., & Napa, C. K. (1998). What makes a good life? Journal of Personality and Social

Psychology, 75, 156-165.

Loughnan, S., Kuppens, P., Allik, J., Balazs, K., De Lemus, S., Dumont, K., et al. (2011).

Economic inequality is linked to biased self-perception. Psychological Science, 22, 1254-

1258.

Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does

happiness lead to success?. Psychological Bulletin, 131, 803-855.

Napier, J. L., & Jost, J. T. (2008). Why are conservatives happier than liberals? Psychological

Science, 19, 565-572.

Paulhus, D. L., & Vazire, S. (2007). The self-report method. Handbook of Research Methods in

Personality Psychology, 224-239.

Sandvik, E., Diener, E., & Seidlitz, L. (1993). Subjective well‐being: The convergence and

stability of self‐report and non‐self‐report measures. Journal of Personality, 61, 317-

342.

Schlenker, B. R., Chambers, J. R., & Le, B. M. (2012). Conservatives are happier than liberals,

but why? Political ideology, personality, and life satisfaction. Journal of Research in

Personality, 46, 127-146.

Taylor, D. M., & Doria, J. R. (1981). Self-serving and group-serving bias in attribution. The

Journal of Social Psychology, 113, 201-211.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective

on mental health. Psychological Bulletin, 103, 193-210.

Von Hippel, W., & Trivers, R. (2011). The evolution and psychology of self-

deception. Behavioral and Brain Sciences, 34, 1-16.

10

CHAPTER 2:

Motivated Happiness: Self-Enhancement Inflates Self-Reported Subjective Well-Being

11

Abstract

Three studies support the contention that self-enhancement motivation distorts self-reports of

subjective well-being. Both individual differences in self-enhancement (Studies 1 & 2) and

experimental manipulations of self-enhancement motivation (Study 2) predicted an increased

likelihood of reporting subjective well-being at unrealistically favorable levels relative to others

– a “happier-than-average effect”. Studies 3a and 3b showed that both trait self-enhancement and

experimentally manipulated differences in self-enhancement motivation also affected self-reports

on established measures of subjective well-being. Specifically, individuals prone to self-

enhancement were more affected than low self-enhancers by the desirability of happiness when

reporting subjective well-being. The current studies suggest that reports of subjective well-being

are susceptible to the same self-enhancement biases that influence self-reports of other positively

valued traits. Implications and recommendations for the measurement of subjective well-being

and the use of well-being data in policy decision-making are discussed.

Keywords: Well-being, self-evaluation, self-presentation, social comparison, judgment

and decision making.

12

Motivated Happiness: Self-Enhancement Inflates Self-Reported Subjective Well-Being

People frequently evaluate themselves in overly positive ways (Alicke, 1985). In one

well-known study, 93% of American participants rated themselves as better-than-average drivers

(Svenson, 1981). Other studies have shown that most college students rate themselves as more

intelligent than average (Brown, 2012), most college professors judge themselves as above

average educators (Cross, 1977), and most relationship partners view their relationships as

superior to those of others (Rusbult, Wildschut, Yovetich, & Verette, 2000). Because these

judgments defy statistical probability, the “better-than-average effect” is widely regarded as

evidence that everyday thought is characterized by positive illusions (Taylor & Brown, 1988).

In their analysis of the pervasiveness of self-enhancing beliefs such as the better-than-

average effect, Taylor and Brown (1988) famously contended that positive illusions foster

mental health by supporting the capacity for productive work, the ability to form and sustain

satisfying relationships, resilience in response to threat, and a general tendency to feel happy and

contented. Supporting this view, unrealistically positive self-assessments have been found to

associate positively with indicators of these specific capacities and with general measures of

psychological well-being (Taylor, Lerner, Sherman, Sage, & McDowell, 2003a;b).

But the nature of the relationship between positive illusions and psychological adjustment

has also been controversial (e.g., Colvin, Block, & Funder, 1995). This is partly because the

nature of this relationship appears to hinge on a number of methodological factors, including the

operationalization of self-enhancement (Kwan, John, Kenny, Bond, & Robins, 2004), properties

of the self-assessed attributes (Paulhus & John, 1998), the evaluator of well-being (Kurt &

Paulhus, 2008), and the component of adjustment being tested (Church, Katigbak, del Prado,

Valdez-Medina, Miramontes, & Ortiz, 2006). Another challenge in understanding this complex

13

relationship is that self-reports of mental health may inadequately distinguish between genuine

and illusory reports of mental health (Shedler, Mayman, & Manis, 1993), potentially inflating the

relationship observed between self-enhancement and well-being. Importantly, this does not

preclude the possibility that self-enhancement still genuinely contributes to well-being (Taylor et

al., 2003a;b), but the notion that self-reports of positive mental health might sometimes be an

example of positive illusions rather than a downstream consequence of them, has important

implications for the literature on subjective well-being (SWB).

SWB and Self-Enhancement

The last two decades have seen a dramatic increase in research on happiness and SWB

that has spread well beyond psychology’s borders. Economists increasingly value subjective

accounts of well-being (Dolan, Peasgood, & White, 2008) and outside of academia, governments

and policy makers have called for indices of “gross national happiness” to directly inform policy

decisions (Brooks, 2008). A key assumption of this research is that self-reports provide valid

measurements of well-being that researchers and policymakers can interpret at face value

(Diener, Sandvik, Pavot, & Gallagher, 1991; Sandvik, Diener, & Seidlitz, 1993). However, self-

report methods have well-documented limitations (e.g., Schwarz, 1999), and are most useful

when combined with other non-self-report indicators (Zou, Schimmack, & Gere, 2013). Little

research, however, has directly examined the potentially confounding influence of self-

enhancement motivation on self-reports of SWB (Kim, Schimmack, & Oishi, 2012).

There are three primary reasons to suspect that self-enhancement may have a

contaminating influence in the measurement of SWB. First, several properties of SWB may

promote its susceptibility to self-enhancing distortion. People self-enhance most on trait

dimensions that are personally important to them (Brown, 2012), and happiness is a desirable

14

and subjectively important human quality with many known intra- and inter-personal benefits

(Diener, 2000). People also self-enhance most on traits that are amenable to idiosyncratic

interpretation (Dunning, Meyerowitz, & Holzberg, 1989), and the definition of happiness is

inherently subjective.

Second, related areas of research have revealed that not all expressions of positive

emotionality represent genuinely positive affective experience. Psychologists have long

recognized that reports of self-esteem can be distorted by self-enhancement motivation and thus

that explicit self-report measures should not always be accepted as valid reflections of actual

internal experience (e.g., Schneider & Turkat, 1975; Kernis, 2003). Similarly, the Duchenne

smile (i.e., smiling involving the muscles orbiting the eye; Ekman, Davidson, & Friesen, 1990)

has been linked to genuine feelings of happiness and enjoyment, as well as to beneficial long-

term psychological and physical health outcomes, but non-Duchenne smiles have not (Harker &

Keltner, 2001). Both of these literatures hint that there are meaningful differences between

genuine and superficial presentations of happiness.

Finally, some research has already implied that people report happiness at unrealistically

high levels. Using nationally representative data from 43 countries, Diener and Diener (1996)

found that mean SWB ratings were above the neutral point in 86% of countries. More directly,

studies documenting the better-than-average effect have sometimes included SWB-related items

in their self-assessment measures (e.g., “enjoy life with regard to recreation, work, and family”;

Brown & Kobayashi, 2002). Other studies have found that people tend to judge their

contentment and life satisfaction as superior to others’, but have attributed these findings to non-

motivational factors (Klar & Giladi, 1999) and/or overlooked the potential implications of self-

enhancement bias in the measurement of SWB (Vautier & Bonnefon, 2008).

15

The Present Research

This article describes three studies examining the extent to which self-reports of SWB are

contaminated by self-enhancing tendencies. Study 1 examined whether individual differences in

self-enhancement predicted the likelihood of showing a “happier-than-average effect.” Study 2

explored whether experimentally manipulating participants’ motivation to self-enhance their

SWB also predicted this tendency. Study 3 tested these effects using two frequently used

measures of SWB.

Study 1

Study 1 examined the tendency to report SWB at unrealistically favorable levels using

the social-comparative judgment format used in better-than-average effect research (e.g.,

Dunning, et al., 1989; Beer, Chester, & Hughes, 2013).

Method

Participants were 1246 visitors (mean age = 41.4, 37.3% female) to YourMorals.org, a

psychological research platform where visitors complete surveys in exchange for personalized

feedback about their results.

Participants completed two scales. The first was a standard social-comparative judgment

task in which participants indicated whether 12 traits and abilities described them less or more

than the average person (-2 = “Much less”, +2 = “Much more”). Ten of the items have been used

in past better-than-average effect research (reliable, insecure, mature, polite, maladjusted, lazy,

unpleasant, good sense of humor, good driving ability), and two items assessed primary

components of SWB (happy, satisfied with life).

The second scale was the Balanced Inventory of Desirable Responding (BIDR; Paulhus,

1991), a well-validated (Lanyon & Carle, 2007) 40-item scale measuring two forms of desirable

16

responding bias. Respondents evaluated propositions (e.g., “I am fully in control of my own

fate”) on 7-point scales (1 = “Not true”; 7 = “Very true”), which were scored using the

continuous method described by Paulhus (1994). The BIDR’s self-deceptive enhancement (SDE)

subscale assesses the tendency to report honest, privately believed, but unrealistically positive

self-assessments, and its impression management (IM) subscale measures deliberately distorted

self-assessments related to self-presentation concerns. We were primarily interested in the SDE

subscale to assess dispositional self-enhancement.

Results

Descriptive statistics and zero-order correlations for all major variables are presented in

Table 1. Participants’ mean scores combined across all 12 better-than-average ratings (α = .78;

negatively valenced items reverse-scored) were significantly higher than the scale’s midpoint (M

= .42, SD = .56, 95% CI [.39, .45]), revealing a pattern indicative of positive illusions, t(1245) =

26.67, p < .001, d = 1.51. This pattern reached significance across 9 of the 10 standard items (see

Figure 1). Importantly we found a similar better-than-average effect on the two SWB-related

items (right two bars of Figure 1) both when they were analyzed together (r = .81; M = .32, SD =

.98, t[1245] = 11.58, p < .001, 95% CI [.27, .38], d = .66) and individually (happiness: M = .30,

SD = 1.01; t[1241] = 10.28, p < .001, 95% CI [.24, .35], d = .58; satisfied with life: M = .35, SD

= 1.04; t[1231] = 11.82, p < .001, 95% CI [.29, .41], d = .67). As shown in Table 1, the better-

than-average effect across the ten standard items (α = .73) was significantly correlated with the

happier-than-average effect.

Figure 1.

Mean Social-Comparative Judgment Scores on Standard and SWB-Related Traits (Study 1)

17

Note: Error bars represent standard errors. Negative trait items reverse scored. All means differ

from zero at p < .001, except for “Lazy” (p = .053).

Table 1.

Descriptive Statistics and Zero-Order Correlations of All Study Variables

Measure Mean (SD) 95% CI 1 2 3 4

Study 1

1. SDE 4.05 (.71) [4.01, 4.09] -

2. IM 3.82 (.91) [3.77, 3.87] .32*** -

3. Standard BAE items 0.46 (.55) [.43, .49] .45*** .30*** -

4. SWB-related BAE

items 0.32 (.98) [.27, .38] .34*** .16*** .41*** -

Study 2

1. SDE 4.27 (.77) [4.16, 4.38] -

2. IM 4.11 (.95) [3.97, 4.25] .49*** -

3. Standard BAE items 0.57 (.57) [.49, .66] .63*** .40*** -

4. Comparative Happiness -0.05 (1.01) [-.20, .10] .42*** .27*** .58*** -

5. Comparative life

satisfaction -0.02 (1.13) [-.19, .14] .44*** .34*** .51*** .75***

Study 3a

1. SDE 4.40 (.89) [4.24, 4.56] -

2. IM 4.07 (1.13) [3.87, 4.27] .47*** -

18

3. Self-reported happiness 2.86 (.79) [2.72, 3.00] .38*** .13 ns -

4. SWLS 4.17 (1.55) [3.90, 4.44] .49*** .23** .77*** -

Study 3b

1. SDE 4.09 (.67) [4.04, 4.15] -

2. IM 3.80 (.86) [3.73, 3.87] .26*** -

3. Self-reported happiness 2.82 (.67) [2.77, 2.87] .31*** .15*** -

4. SWLS 4.23 (1.39) [4.12, 4.34] .38*** .21*** .71*** -

Note: SDE = Self-Deceptive Enhancement, IM = Impression Management, BAE = Better

than Average Effect, SWLS = Satisfaction With Life Scale. Study 2 and 3 statistics are

collapsed across experimental condition. ns p = .143, ** p = .010, *** p < .001.

We next tested whether individual differences in self-enhancement predicted these better-

than-average effects by examining correlations between the SDE subscale1 (α = .74), an index

combining the 10 standard items, and the index combining the 2 SWB-related items. As shown

in Figure 1, higher levels of SDE predicted both significantly larger better-than-average effects

and significantly larger happier-than-average effects.

Discussion

Study 1 participants reported being happier and more satisfied with life than the average

person, just as they reported being above average on traits and abilities that had no direct relation

to SWB. Notably, the happier-than-average effect was similar in magnitude to the better-than-

average effect observed on non-SWB-related traits, and the two effects showed a significant

positive correlation, both findings suggestive of a common underlying mechanism. Both effects

also showed a similar tendency to be more pronounced among participants with higher levels of

dispositional self-enhancement.

As is the case with all better-than-average effect research, we cannot definitively rule out

the possibility that Study 1 participants were, in actuality, happier and more satisfied with life

than the average other. However, if one accepts as genuine our participants’ positive assessments

1 Across all studies, the basic pattern of results for the SDE subscale is similar if the IM subscale is included as a

control variable. We report results without controlling for IM to facilitate simpler interpretation of effect sizes.

19

of their psychological well-being compared to others, one must either argue that this genuineness

is uniquely true of SWB-related self-assessments (despite evidence of a common mechanism), or

also accept that our participants were, in actuality, funnier, more polite, less insecure, and even

better drivers than others as well.

Still, despite recent research implicating self-enhancement as the primary driver of the

better-than-average effect (Beer et al., 2013), non-motivational interpretations of Study 1’s

results remain possible (Chambers & Windschitl, 2004). And, despite our rhetorical argument

against interpreting these self-reports at face value, Study 1’s results remain consistent with

existing conceptualizations of self-enhancement as a genuine contributor to psychological well-

being (Taylor & Brown, 1988).

Study 2

To clarify the nature of the SWB--self-enhancement relationship, Study 2 manipulated

participants’ motivation to report high levels of SWB by presenting them with a research

summary purporting either positive or negative consequences of high levels of happiness (e.g.,

Gruber, Mauss, & Tamir, 2011). To the extent that self-reports of SWB are contaminated by self-

enhancing self-report styles, increasing or decreasing the subjective desirability and importance

of happiness should increase or decrease participants’ motivation to claim happiness as a

personal trait, without affecting their actual levels of experienced happiness.

Method

Participants were 179 Amazon Mechanical Turk workers from the United States (45.3%

female, mean age = 35.5) who completed the study in exchange for a small fee.

Participants first completed the BIDR and then read a paragraph summarizing happiness

research purportedly conducted at an elite university. Participants assigned to the desirable

20

condition (n = 91) read about research demonstrating the positive correlates of happiness (e.g.,

improved psychological well-being, improved social interactions, marriage and career success),

whereas participants assigned to the undesirable condition (n = 88) read about research

demonstrating the negative correlates of happiness (e.g., increased risk taking behavior, poorer

performance in competitive situations, increased risk for mania). Participants were instructed to

write a two to three sentence research summary.2 We included a manipulation check in which

participants rated on separate 7-point scales how important happiness was to them personally and

how convincing they found the research summary (1 = “Not at all”, 7 = “Very much”). Finally,

participants completed the better-than-average effect measure from Study 1.

Results

Participants in the desirable condition (M = 5.36, SD = 1.68, 95% CI [5.00, 5.72]) found

the research summary to be more convincing than did participants in the undesirable condition

(M = 3.71, SD = 1.82, 95% CI [3.33, 4.09]; t[175] = -6.24, p < .001, d = -.94). This is not

surprising given the counterintuitive nature of the undesirable condition’s research summary.

The manipulation was nevertheless effective: participants who read the desirable research

summary judged happiness to be significantly more personally important to them (M = 6.25, SD

= 1.02, 95% CI [6.03, 6.57]) than did participants who read the undesirable research summary

(M = 5.42, SD = 1.20, 95% CI [5.17, 5.67]; t[177] = -4.99, p < .001, d = -.75).3

We also checked whether the experimental manipulation influenced judgments across the

10 standard, non-SWB judgments (α = .78), and as expected, it did not (t[177] = -1.57, p = .119,

2 In Studies 2 and 3a, no participants were removed for failing to follow instructions; in Study 3b, 86 participants

were excluded for not following this instruction. This may result from differences in paid and volunteer samples. 3 Including convincingness ratings as a covariate did not change the patterns of results observed between SWB

judgments, SDE, and experimental condition in Study 2, 3a, or 3b.

21

d = -.24). Also as expected, and replicating Study 1, the better-than-average effect on non-SWB

items was positively related to dispositional self-enhancement (see Table 1).

To examine how our experimental manipulation of self-enhancement motivation

influenced SWB judgments, we separately regressed the happiness and life satisfaction items

onto the SDE subscale score (α = .78), a dichotomous variable representing experimental

condition (0 = undesirable, 1 = desirable), and their interaction term. Results are presented in

Table 2. As shown, participants high in dispositional self-enhancement reported larger happier-

than-average effects than did participants low in dispositional self-enhancement (replicating the

results of Study 1). Importantly, participants who read that happiness had desirable effects also

showed larger happier-than-average effects than did those who read that happiness had

undesirable effects. We did not observe a significant interaction between dispositional and

experimentally manipulated self-enhancement. A similar pattern of results was observed on the

social-comparative life satisfaction item, although the effect of the experimental manipulation

did not reach significance (Table 2).

Table 2.

Regression Results and Meta-Analyzed Effect Sizes for Studies 2, 3a, and 3b

Study 2

Social-Comparative Happiness

Predictor b t p 95% CI β F (df) R2

SDE 0.60 4.99 < .001 [.36, .84] 0.46

14.93***

(3, 175) 0.20

Condition 0.31 2.31 0.022 [.05, .58] 0.16

Interaction -0.09 -0.49 0.628 [-.44, .26] -0.05

Constant -0.21 -2.17 0.032 [-.39, -.02]

Social-Comparative Life Satisfaction

Predictor b t 95% CI β F (df) R2

SDE 0.55 3.10 < .001 [.29, .82] 0.38

15.48***

(3, 175) 0.21

Condition 0.24 1.60 0.111 [-.06, .54] 0.11

Interaction 0.21 1.08 0.283 [-.18, .61] 0.10

Constant -0.14 -1.30 0.195 [-.35, .07]

Study 3a

Self-Reported Happiness

22

Predictor b t p 95% CI β F (df) R2

SDE 0.17 1.58 0.117 [-.04, .38] 0.19

9.52***

(3, 123) 0.19

Condition 0.18 1.39 0.168 [-.08, .43] 0.11

Interaction 0.32 2.20 0.030 [.03, .60] 0.26

Constant 2.77 30.31 < .001 [2.58, 2.95]

Satisfaction With Life Scale

Predictor b t p 95% CI β F (df) R2

SDE 0.56 2.90 0.004 [.18, .94] 0.25

16.83***

(3, 123) 0.29

Condition 0.54 2.31 0.023 [.08, 1.01] 0.20

Interaction 0.53 2.00 0.048 [.01, 1.05] 0.18

Constant 3.89 23.37 < .001 [3.56, 4.22]

Study 3b

Self-Reported Happiness

Predictor b t p 95% CI β F (df) R2

SDE 0.21 3.81 < .001 [.10, .31] 0.20

25.19***

(3, 617) 0.11

Condition 0.09 1.78 0.076 [-.01, .19] 0.07

Interaction 0.21 2.69 0.007 [.06, .36] 0.14

Constant 2.77 78.24 < .001 [2.71, 2.84]

Satisfaction With Life Scale

Predictor b t p 95% CI β F (df) R2

SDE 0.65 5.92 < .001 [.43, .86] 0.31

35.76***

(3, 617) 0.15

Condition 0.13 1.23 0.221 [-.08, .33] 0.05

Interaction 0.28 1.80 0.073 [-.03, .59] 0.09

Constant 4.17 57.69 < .001 [4.02, 4.31]

Meta-Analyzed Effect Sizes

Happiness

Predictor Unweighted r Weighted r p 95% CI

SDE 0.22 0.19 < .001 [.13, .25]

Condition 0.12 0.10 0.002 [.03, .16]

Interaction 0.10 0.10 0.002 [.04, .16]

Life satisfaction

Predictor Unweighted r Weighted r p 95% CI

SDE 0.26 0.25 < .001 [.19, .31]

Condition 0.13 0.08 0.003 [.02, .15]

Interaction 0.11 0.09 0.005 [.02, .15]

Note: SDE = Self-Deceptive Enhancement. Condition: 1 = Desirable, 0 = Undesirable. *** p <

.001.

Discussion

23

Study 2 replicated and extended the results of Study 1 by showing that it was not only

trait self-enhancement, but also situationally induced variation in self-enhancement motivation

that strengthened participants’ tendencies to judge themselves as happier than average.

As discussed earlier, individual differences in self-enhancement may predict SWB

judgments either because self-enhancers over-report positive qualities or because dispositional

self-enhancement genuinely promotes psychological well-being (Taylor & Brown, 1988). The

latter “positive illusions view”, however, has difficulty explaining the effects of our situational

manipulation, which showed that portraying happiness in desirable terms motivated people to

report unrealistically positive levels of SWB (but not other characteristics) relative to others.

Study 3a

Studies 1 and 2 asked participants to rate their SWB relative to others. This raises the

question of whether self-enhancement might similarly distort judgments on standard self-report

measures. Although self-report scales are empirically validated (e.g., Pavot & Diener, 1993) and

have been shown to correlate with meaningful outcomes (Lyubomirsky, King, & Diener, 2005),

self-enhancement motivation could still account for unexplained variance in these measures.

Studies 3a and 3b examined the extent to which both trait and state self-enhancement motivation

affected self-reports on two widely used measures of SWB.

Method

We recruited 128 American participants from Amazon’s Mechanical Turk (53.9%

female, mean age = 35.7). The procedure for Study 3a was identical to that of Study 2, except

that instead of completing the better-than-average effect scale following the experimental

manipulation, participants completed two standard self-report measures of SWB: 1) A single-

item measure of happiness used in a multitude of papers utilizing the World Values Surveys

24

(2006; Taking all things together, would you say that you are: 1 = “Not at all happy”, 2 = “Not

very happy”, 3 = “Rather happy”, 4 = “Very happy”); and 2) the Satisfaction With Life Scale, a

face-valid 5-item scale that is one of the most widely used and accepted measures of global life

satisfaction (Pavot & Diener, 1993).

Results

Participants in the desirable condition (M = 5.27, SD = 1.74, 95% CI [4.83, 5.70]) again

found the research summary to be more convincing than did participants in the undesirable

condition (M = 3.92, SD = 2.09, 95% CI [3.39, 4.45]; t[125] = -3.95, p < .001, d = .71).

However, the manipulation still led to the intended difference in the subjective importance of

happiness (desirable: M = 5.94, SD = 1.35, 95% CI [5.02, 5.64]; undesirable: M = 5.33, SD =

1.23; 95% CI [5.60, 6.28]; t[124] = -2.62, p = .010, d = .47).

We next performed separate regression analyses with self-reported happiness and

Satisfaction With Life Scale scores as the dependent variables. In both regressions, SDE,

experimental condition (0 = undesirable, 1 = desirable), and their interaction terms were

included as predictor variables. Results are presented in Table 2 and shown graphically in

Figures 2 and 3. For the single-item happiness measure, the main effects of SDE and

experimental condition were in the expected directions but did not reach statistical significance.

However, we observed a significant interaction between these variables. We investigated this

interaction with simple slopes analyses, and found that the relationship between SDE and

happiness judgments was stronger in the desirable condition (b = .48, t = 4.89, p < .001) than in

the undesirable condition (b = .17, t = 1.58, p = .117).

Figure 2.

Self-Reported Happiness as a Function of SDE and Happiness Desirability Condition (Study 3).

25

On the Satisfaction With Life Scale, both dispositional self-enhancement and

experimental condition significantly predicted reports of greater life satisfaction. A significant

interaction between these predictors also revealed that, as was found with the happiness measure,

life satisfaction judgments were more strongly related to dispositional self-enhancement in the

desirable condition (b = 1.09, t = 6.01, p < .001) than in the undesirable condition (b = .56, t =

2.90, p = .004).

Figure 3.

Satisfaction With Life Scale scores as a function of SDE and happiness desirability condition

(Study 3).

26

Study 3b

The interaction effects observed in Study 3a showed that reports of SWB by individuals

prone to self-enhancement were more influenced by the desirability of happiness than those by

individuals not prone to self-enhancement. However, because Study 2 revealed only main effects

on comparative SWB judgments, we replicated Study 3a with a different sample and improved

statistical power in Study 3b.

Participants were 626 visitors to YourMorals.org (48.2% female, mean age = 35.3) who

completed a study with an identical procedure to that of Study 3a. Results were generally

consistent with those of Study 3a (see Table 2). Self-reported happiness showed a significant

main effect of SDE and a marginal main effect of experimental condition. Replicating the

interaction pattern of Study 3a, we found that the relationship between SDE and self-reported

happiness was again stronger in the desirable condition (b = .41, t = 7.51, p < .001) than in the

undesirable condition (b = .21, t = 3.81, p < .001). Similarly, self-reported life satisfaction was

27

predicted by SDE, and although the condition main effect was not significant, a marginally

significant interaction pattern revealed again that dispositional self-enhancement more strongly

predicted life satisfaction judgments when happiness was portrayed as desirable (b = .93, t =

3.78, p < .001) than when it was portrayed as undesirable (b = .65, t = 5.92, p < .001).

Meta-Analysis

We meta-analyzed the results of Studies 2, 3a, and 3b to determine the robustness of the

various main and interaction effects. As shown in Table 2, both meta-analyzed main effects on

the happiness measure were significantly positive (SDE: rweighted = .19, p < .001; experimental

condition: rweighted = .10, p = .002), as was the interaction term (rweighted = .10, p = .002). Identical

results were found on the Satisfaction With Life Scale (SDE: rweighted = .25, p < .001;

experimental condition: rweighted = .08, p = .003; interaction: rweighted = .09, p = .005).

Discussion

The results of Studies 3a and 3b are notable for demonstrating that self-enhancement

concerns influence not only comparative SWB judgments (Study 2), but also responses on two of

the most relied upon self-report measures in the SWB literature. Interestingly, Study 2 showed

only main effects for SDE and experimental condition, but Studies 3a and 3b showed an

interaction pattern in which high self-enhancers were more sensitive to our happiness desirability

manipulation than were low self-enhancers. Both of these effects represent challenges to the

validity of self-reported SWB, and importantly, our meta-analysis confirmed that both were

reflected in our data at similarly modest sizes (weighted r’s of .08 to .10).

General Discussion

The current research provides evidence that, just as with driving skills, intelligence, and

professorial proficiency, people’s reports of their personal well-being can be contaminated by

28

their desire to possess traits and abilities that flatter the self. The clearest effect in our studies was

that individuals high in dispositional self-enhancement reported greater SWB than individuals

with weaker self-enhancing tendencies. SDE scores were the most consistent predictor of self-

reported SWB (significant in all but one of the reported analyses) and had the largest effect size

in the meta-analysis (see Table 2).

Still, it is possible that dispositional self-enhancement might predict higher reports of

SWB not by directly affecting self-reports, but by promoting genuine happiness and life

satisfaction, and there is evidence that it does (e.g., Taylor & Brown, 1988; Taylor et al.,

2003a;b). This “positive illusions” view alone, however, has trouble explaining the contextual

influence of our experimental manipulations, in that the positive illusions view predicts that it is

the habitual tendency to self-enhance over time that promotes behaviors that in turn promote

genuine psychological well-being. Our success in altering SWB by experimental manipulation is

important because it precludes this possibility of habitual self-enhancement over time, suggesting

that the report of SWB itself can be affected by self-enhancing motivations.

The work of disentangling the relation between positive illusions and SWB is certainly

not over. For example, a determined critic could still argue that the interaction patterns found in

Studies 3a and 3b are evidence that dispositional self-enhancers maintain genuine well-being in

part by taking particular advantage of situations that afford the opportunity for self-flattery (by

claiming desirable traits and rejecting undesirable ones). It is important to remember, of course,

that the two accounts are not mutually exclusive, and that the most parsimonious explanation of

the current findings is that, whether they flow from dispositional tendencies or situational

affordances, self-enhancing motivations have the potential to inflate self-reports of happiness and

life-satisfaction.

29

Implications for SWB Research

Our data have particular implications for programs of research that rely on self-reports to

compare SWB differences between pre-existing, non-randomized groups like nations

(Veenhoven & Ehrhardt, 1995), political groups (Brooks, 2008), and cultures (Diener & Suh,

2000). It is plausible that any two groups hypothesized to differ in happiness levels might

similarly differ in self-enhancing tendencies (Kim et al., 2012), potentially complicating our

understanding of group-level happiness differences. This possibility is particularly noteworthy in

light of suggestions that policy decisions be directly informed by national happiness indexes

(Diener, 2000; 2006; Brooks, 2008). In our studies, we found meta-analyzed effect sizes for

individual differences (happiness: r = .19; life satisfaction: r = .25) and situational variations

(ranging from r = .08 to r = .10) in self-enhancement motivation that were comparable to those

of many variables of interest to gross national happiness researchers (e.g., gender: r = .04,

employment status, r = .11, income: r = .17; Okun, Stock, Haring, & Witter, 1984). Researchers

and policymakers should recognize that differences in self-reported SWB of this magnitude may

be explained partially, or even substantially, by state or trait differences in self-enhancing styles

of self-assessment.

Although some happiness researchers have argued (following Taylor & Brown, 1988)

that self-enhancing tendencies constitute a genuine component of SWB and thus should not be

controlled for statistically (Diener et al., 1991; Kozma & Stones, 1988), many have also

advocated that self-report should constitute only one component of any comprehensive

assessment of individual or group-level SWB (Diener, 2000; Zou et al., 2013). We strongly

support this recommendation and suggest that SWB researchers might look for useful approaches

in the self-esteem literature, which has long used self-enhancement measures jointly with self-

30

report scales, as well as indirect and implicit techniques (e.g., Jordan, Spencer, Zanna, Hoshina-

Browne, & Correll, 2003; Walker & Schimmack, 2008) to identify and control for defensively

motivated reports of self-esteem (Kernis, 2003).

Coda

It makes perfect sense that if one wants to know how happy someone is, the most obvious

thing to do is ask. Indeed, the absence of any objective validation criterion to evaluate genuine

happiness leads quickly to philosophical musings about whether it is really possible to be

“implicitly” or “defensively” happy, and if so, whether and how those states differ from

“genuine” happiness. At the same time, however, it seems implausible that happiness and life

satisfaction, given the value and importance most people invest in them, are immune from the

self-aggrandizing distortions that have been shown to affect claims about myriad other personal

qualities. Researchers routinely exercise caution when evaluating the meaning of various

expressions of positive emotionality, with the knowledge that some are less secure (Schneider &

Turkat, 1975), less predictive (Harker & Keltner, 2001), and less heartfelt than others (Ekman et

al., 1990). As psychologists, economists, and policy makers become increasingly interested in

measuring SWB, we advise similar caution when interpreting self-reports of happiness.

31

References

Alicke, M. D. (1985). Global self-evaluation as determined by the desirability and controllability

of trait adjectives. Journal of Personality and Social Psychology, 49, 1621.

Beer, J. S., Chester, D. S., & Hughes, B. L. (2013). Social Threat and Cognitive Load Magnify

Self-Enhancement and Attenuate Self-Deprecation. Journal of Experimental Social

Psychology, 4, 706-711.

Brooks, A. C. (2008). Gross national happiness: Why happiness matters for America--and how

we can get more of it. New York: Basic Books.

Brown, J. D. (2012). Understanding the "Better than Average" effect: Motives (still)

matter. Personality and Social Psychology Bulletin, 38, 209-219.

Brown, J. D., & Kobayashi, C. (2002). Self‐enhancement in Japan and America. Asian Journal

of Social Psychology, 5, 145-168.

Chambers, J. R., & Windschitl, P. D. (2004). Biases in social comparative judgments: the role of

nonmotivated factors in above-average and comparative-optimism effects. Psychological

Bulletin, 130, 813-838.

Church, A. T., Katigbak, M. S., del Prado, A. M., Valdez-Medina, J. L., Miramontes, L. G., &

Ortiz, F. A. (2006). A cross-cultural study of trait self-enhancement, explanatory

variables, and adjustment. Journal of Research in Personality, 40, 1169-1201.

Colvin, C. R., Block, J., & Funder, D. C. (1995). Overly positive self-evaluations and

personality: negative implications for mental health. Journal of Personality and Social

Psychology, 68, 1152-1162.

Cross, K. P. (1977). Not can, but will college teaching be improved?. New Directions for Higher

Education, 17, 1-15.

32

Diener, E. (2000). Subjective well-being: The science of happiness and a proposal for a national

index. American Psychologist, 55, 34-43.

Diener, E. (2006). Guidelines for national indicators of subjective well-being and ill-being.

Applied Research in Quality of Life, 1, 151-157.

Diener, E., & Diener, C. (1996). Most people are happy. Psychological Science, 7, 181-185.

Diener, E., Sandvik, E., Pavot, W., & Gallagher, D. (1991). Response artifacts in the

measurement of subjective well-being. Social Indicators Research, 24, 35-56.

Diener, E., & Suh, E. M. (Eds.). (2000). Culture and subjective well-being. Cambridge, MA:

MIT Press.

Dolan, P., Peasgood, T., & White, M. (2008). Do we really know what makes us happy? A

review of the economic literature on the factors associated with subjective well-

being. Journal of Economic Psychology, 29, 94-122.

Dunning, D., Meyerowitz, J. A., & Holzberg, A. D. (1989). Ambiguity and self-evaluation: The

role of idiosyncratic trait definitions in self-serving assessments of ability. Journal of

Personality and Social Psychology, 57, 1082-1090.

Ekman, P., Davidson, R. J., & Friesen, W. V. (1990). The Duchenne smile: Emotional

expression and brain physiology: II. Journal of Personality and Social Psychology, 58,

342-353.

Gruber, J., Mauss, I. B., & Tamir, M. (2011). A dark side of happiness? How, when, and why

happiness is not always good. Perspectives on Psychological Science, 6, 222-233.

Harker, L., & Keltner, D. (2001). Expressions of positive emotion in women's college yearbook

pictures and their relationship to personality and life outcomes across adulthood. Journal

of Personality and Social Psychology, 80, 112-124.

33

Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino-Browne, E., & Correll, J. (2003). Secure and

defensive high self-esteem. Journal of Personality and Social Psychology, 85, 969-978.

Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological

Inquiry, 14, 1-26.

Kim, H., Schimmack, U., & Oishi, S. (2012). Cultural differences in self- and other-evaluations

and well-being: A study of European and Asian Canadians. Journal of Personality and

Social Psychology, 102, 856-873.

Klar, Y., & Giladi, E. E. (1999). Are most people happier than their peers, or are they just

happy?. Personality and Social Psychology Bulletin, 25, 586-595.

Kozma, A., & Stones, M. J. (1988). Social desirability in measures of subjective well-being: Age

comparisons. Social Indicators Research, 20, 1-14.

Kurt, A., & Paulhus, D. L. (2008). Moderators of the adaptiveness of self-enhancement:

Operationalization, motivational domain, adjustment facet, and evaluator. Journal of

Research in Personality, 42, 839-853.

Kwan, V. S., John, O. P., Kenny, D. A., Bond, M. H., & Robins, R. W. (2004).

Reconceptualizing individual differences in self-enhancement bias: an interpersonal

approach. Psychological Review, 111, 94-110.

Lanyon, R. I., & Carle, A. C. (2007). Internal and external validity of scores on the Balanced

Inventory of Desirable Responding and the Paulhus Deception Scales. Educational and

Psychological Measurement, 67, 859-876.

Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does

happiness lead to success?. Psychological Bulletin, 131, 803-855.

34

Okun, M. A., Stock, W. A., Haring, M. J., & Witter, R. A. (1984). Health and subjective well-

being: A meta-analysis. The International Journal of Aging and Human Development, 19,

111-132.

Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R.

Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social attitudes (pp. 17-

59). San Diego, CA: Academic Press.

Paulhus, D. L. (1994). Balanced Inventory of Desirable Responding: Reference manual for

BIDR Version 6. Unpublished manuscript, University of British Columbia, Vancouver,

Canada.

Paulhus, D. L., & John, O. P. (1998). Egoistic and moralistic biases in self-perception: The

interplay of self-deceptive styles with basic traits and motives. Journal of Personality, 66,

1025–1060.

Pavot, W., & Diener, E. (1993). Review of the satisfaction with life scale. Psychological

Assessment, 5, 164-172.

Rusbult, C. E., Van Lange, P. A., Wildschut, T., Yovetich, N. A., & Verette, J. (2000). Perceived

superiority in close relationships: why it exists and persists. Journal of Personality and

Social Psychology, 79, 521-545.

Sandvik, E., Diener, E., & Seidlitz, L. (1993). Subjective well‐being: The convergence and

stability of self‐report and non‐self‐report measures. Journal of Personality, 61, 317-

342.

Schneider, D. J., & Turkat, D. (1975). Self‐presentation following success or failure: Defensive

self‐esteem models. Journal of Personality, 43, 127-135.

35

Schwarz, N. (1999). Self-reports: How the questions shape the answers. American Psychologist,

54, 93-105.

Shedler, J., Mayman, M., & Manis, M. (1993). The illusion of mental health. American

Psychologist, 48, 1117-1131.

Svenson, O. (1981). Are we all less risky and more skillful than our fellow drivers? Acta

Psychologica, 47, 143-148.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective

on mental health. Psychological Bulletin, 103, 193-210.

Taylor, S. E., Lerner, J. S., Sherman, D. K., Sage, R. M., & McDowell, N. K. (2003a). Are self-

enhancing cognitions associated with healthy or unhealthy biological profiles?. Journal

of Personality and Social Psychology, 85, 605-615.

Taylor, S. E., Lerner, J. S., Sherman, D. K., Sage, R. M., & McDowell, N. K. (2003b). Portrait of

the self-enhancer: well adjusted and well liked or maladjusted and friendless? Journal of

Personality and Social Psychology, 84, 165-176.

Vautier, S., & Bonnefon, J. F. (2008). Is the above-average effect measurable at all? The validity

of the self-reported happiness minus others' perceived happiness construct. Applied

Psychological Measurement, 32, 575-584.

Veenhoven, R., & Ehrhardt, J. (1995). The cross-national pattern of happiness: Test of

predictions implied in three theories of happiness. Social Indicators Research, 34, 33-68.

Walker, S. S., & Schimmack, U. (2008). Validity of a happiness implicit association test as a

measure of subjective well-being. Journal of Research in Personality, 42, 490-497.

36

World Values Survey. (2006). European and World Values Surveys Four-Wave Integrated Data

File, 1981–2004, v.20060423. Retrieved September 12, 2013, from

http://www.wvsevsdb.com/wvs/WVSData.jsp.

Zou, C., Schimmack, U., & Gere, J. (2013). The validity of well-being measures: A multiple-

indicator-multiple-rater model. Psychological Assessment, 25, 1247-1254.

Acknowledgements: This chapter has been published at Social Psychological and Personality

Science, and would not be possible without the contributions of co-author Peter Ditto. The

authors received no financial support for the research, authorship, and/or publication of this

article and declared no potential conflicts of interest. Full citation for published article:

Wojcik, S. P., & Ditto, P. H. (2014). Motivated happiness: Self-enhancement inflates self-

reported subjective well-being. Social Psychological and Personality Science, 5, 525-

534.

37

CHAPTER 3:

Conservative Self-Enhancement

38

Abstract

Political conservatism has been linked to motivated forms of social cognition, sensitivity to

negativity and threat, and defensive cognitive styles. However, little research has directly

investigated whether political conservatism predicts self-enhancement bias. The present research

examined the relationship between liberal-conservative political ideology and self-enhancement

using large Internet samples across eight studies (N = 11,893). Meta-analysis of these results

revealed that self-enhancement was positively associated with general political conservatism,

social conservatism, economic conservatism, and conservative patterns of moral foundation

endorsement (.12 ≤ rrandom ≤ .13; .07 ≤ rfixed ≤ .10), even after controlling for key demographic

variables (.08 ≤ rrandom ≤ .16; .03 ≤ rfixed ≤ .09).

Keywords: Political ideology, self-enhancement, self-assessment, meta-analysis.

39

Conservative Self-Enhancement

Most people view themselves as better than average, although this phenomenon defies

statistical probability (Dunning, Meyerowitz, & Holzberg, 1989). Self-enhancement, or the

tendency to evaluate oneself in an unrealistically favorable light, is more pronounced among the

psychologically well-adjusted (Taylor & Brown, 1988), the religious (Sedikides & Gebauer,

2009), and Westerners in individualistic, unequal societies where there is greater motivation to

stand out from others (Hamamura, Heine, & Takemoto, 2007; Loughnan et al., 2011).

Past research hints that self-enhancement might also be more pronounced among political

conservatives, relative to political liberals. Historically, psychological theory has repeatedly

linked conservatism to defensive cognitive styles (Adorno, Frenkel-Brunswik, Levinson, &

Sanford, 1950; Pratto, Sidanius, Stallworth, & Malle, 1994; Jost, Glaser, Kruglanski, &

Sulloway, 2003; Hibbing, Smith, & Alford, 2014). Economic conservatism involves a

hierarchical, competitive worldview, which heightens the motivation to view the self as superior

to others (Loughnan et al., 2011). Social conservatism and conservative patterns of moral

foundation endorsement are related to greater religiosity and more favorable views of similar

others and ingroup members (Graham et al., 2011). Conservative views have also been linked to

beliefs that are indicative of self-enhancement, including flattering self-assessments of physical

attractiveness, social status, life satisfaction, optimism, and personal control (Belmi & Neale,

2014; Schlenker, Chambers, & Le, 2012).

However, the limited literature directly assessing political differences in self-

enhancement provides mixed findings. Conservatism is associated with stronger degrees of self-

deceptive enhancement and impression management (Wojcik, Hovasapian, Graham, Motyl, &

Ditto, in press; Jost et al., 2010), but liberalism has been associated with the “truly false

40

uniqueness” effect (Stern et al., 2013). The present research aimed to clarify whether, and to

what extent, political ideology predicted self-enhancement.

Method

We assessed relationships between four operationalizations of political ideology and

eight components of self-enhancement and using large Internet samples.

Participants

Participants from the United States were included in the present research if they

completed one or more self-enhancement measures from either of two online research platforms.

Five measures were posted on YourMorals.org (n = 12,605; mean age = 38.7; 48.2% female).

The remaining measures were posted on Amazon’s Mechanical Turk (n = 397; mean age = 30.0;

36.0% female).

Political Ideology

Participants separately reported general, social, and economic political ideology along 7-

point scales (1 = Very Liberal, 7 = Very Conservative). Individuals who identified as non-

political, libertarian, and other were excluded. We also assessed liberal-conservative patterns of

moral foundation endorsement among YourMorals.org participants who completed the Moral

Foundations Questionnaire (n = 8,533; Graham et al., 2011), which assesses sensitivity to five

foundations of morality. Moral liberalism-conservatism was operationalized by subtracting the

average “binding foundation” score (Harm and Fairness) from the average “individualizing

foundation” score (Loyalty, Authority, and Sanctity). Higher scores indicated conservative

patterns of morality.

Self-enhancement measures

41

Separate questionnaires assessed public and private forms of self-enhancement (Balanced

Inventory of Desirable Responding; Paulhus, 1991), socially desirable responding (Marlowe-

Crowne Social Desirability Scale; Crowne & Marlowe, 1960), direct-comparison better-than-

average effects (Dunning, Meyerowitz, & Holzberg, 1989), indirect-comparison better-than-

average effects (Hamamura et al., 2007), false-uniqueness effects (Hamamura et al., 2007),

actual-ideal self-discrepancies (Pelham & Swann, 1989), and both family member and nation-

level enhancement (Heine & Lehman, 1997).

Results

We examined the pairwise correlation between each dimension of political ideology and

each measure of self-enhancement, and then examined these relationships after controlling for

demographic characteristics. Effect sizes for each association are shown in Table 1. After

controlling for demographic indicators, general conservatism significantly predicted self-

enhancement on eight of the ten self-enhancement measures (.01 ≤ rpartial ≤ .40), social

conservatism significantly predicted seven of the ten measures (.00 ≤ rpartial ≤ .31), economic

conservatism significantly predicted seven of ten measures (.06 ≤ rpartial ≤ .29), and moral

conservatism significantly predicted four of the six measures for which data were available (-.06

≤ rpartial ≤ .26). In all cases, greater political conservatism predicted stronger self-enhancement.

Meta-analysis

We meta-analyzed the pairwise correlations and partial r-effects to calculate random and

fixed effects for each ideology measure (Table 1). Meta-analysis revealed that self-enhancement

was significantly associated with general political conservatism, social conservatism, economic

conservatism, and moral conservatism, both before and after controlling for demographic

characteristics (random effects: .08 ≤ rpartial ≤ .16; fixed effects: .03 ≤ rpartial ≤ .10).

Table 1.

Effect Sizes and Meta-Analysis of Political Ideology Variables Predicting Self-Enhancement

Political Ideology

General Liberalism-

Conservatism Social Liberalism-Conservatism

Economic Liberalism-

Conservatism Moral Liberalism-Conservatism

Samp

le N Self-Enhancement Scale

Subscal

e

Pairwise

r (p)

Partial

r (p)

Pairwise

r (p)

Partial

r (p)

Pairwise

r (p)

Partial

r (p)

Pairwise

r (p)

Partial

r (p)

YM 2,567 BIDR SDE 0.12 (<.001) 0.12 (<.001) 0.10 (< .001) 0.10 (.024) 0.14 (<.001) 0.13 (.005) 0.11 (<.001) 0.08 (<.001)

IM 0.14 (<.001) 0.07 (.001) 0.16 (<.001) 0.09 (<.001) 0.09 (<.001) 0.06 (.013) 0.07 (.002) -0.01 (.613)

YM 4,347 MC-SDS - 0.07 (<.001) 0.04 (.013) 0.06 (<.001) 0.03 (.076) 0.08 (<.001) 0.06 (.001) 0.01 (.583) -0.06 (.006)

YM 4,959 Direct Social Comparisons BAE 0.10 (<.001) 0.06 (<.001) 0.11 (<.001) 0.07 (<.001) 0.09 (<.001) 0.06 (<.001) 0.06 (<.001) 0.02 (.449)

MT 125 Indirect Social Comparisons

BAE 0.13 (.147) 0.10 (.275) 0.14 (.128) 0.11 (.222) 0.18 (.047) 0.15 (.111) - -

BROE 0.05 (.579) 0.01 (.921) 0.04 (.695) 0.00 (.965) 0.09 (.312) 0.06 (.541) - -

MT 135 Actual-Ideal Self-

Discrepancies - 0.23 (.006) 0.24 (.007) 0.20 (.020) 0.20 (.022) 0.12 (.166) 0.12 (.187) - -

YM 292 False Uniqueness Effect - 0.16 (.005) 0.40 (.001) 0.09 (.134) 0.25 (.056) 0.15 (.016) 0.29 (.025) 0.18 (.001) 0.1 (<.001)

MT 138

Family Member

Enhancement - 0.26 (.003) 0.20 (.018) 0.26 (.002) 0.21 (.015) 0.26 (.002) 0.25 (.004) - -

YM 439 Nation-Level Enhancement - 0.26 (<.001) 0.24 (<.001) 0.32 (<.001) 0.31 (<.001) 0.21 (<.001) 0.17 (.001) 0.32 (<.001) 0.26 (<.001)

Meta-Analyzed r-Effect Sizes

General Liberalism-

Conservatism Social Liberalism-Conservatism

Economic Liberalism-

Conservatism Moral Liberalism-Conservatism

Pairwise r Partial r Pairwise r Partial r Pairwise r Partial r Pairwise r Partial r

Random effects r 0.13 0.16 0.13 0.13 0.12 0.12 0.13 0.08

95% CI [.092, .175] [.087, .221] [.083, 184] [.073, .188] [.086, .155] [.063, .181] [.048, .207] [-.005, .163]

p < .001 < .001 < .001 < .001 < .001 < .001 < .001 0.065

Fixed effects r 0.10 0.09 0.10 0.08 0.10 0.07 0.07 0.03

95% CI [.086, .120] [.066, .104] [.083, .184] .058, .096] [.086, .122] [.048, .086] [.055, .094] [.006, .053]

p < .001 < .001 < .001 < .001 < .001 < .001 < .001 0.014

Note: Each partial r-effect was dertived from a separate regression equation with the political ideology variable predicting the self-enhancement measure, controlling for available demographics. Moral

liberalism-conservatism data were only available for studies conducted at YourMorals.org. Meta-analyses included one subscale per scale (SDE for BIDR, BROE for Indirect Social Comparisons). YM = YourMorals.org, MT = MTurk, BIDR = Balanced Inventory of Desirable Responding, SDE = Self-Deceptive Enhancement, IM = Impression Management, MC-SDS = Marlowe-Crowne Social

Desirability Scale, BAE = Better than Average Effect, BROE = Better than Random Other Effect.

42

43

Discussion

Eight studies including thousands of participants from two diverse populations revealed

that political conservatism consistently predicted the tendency to evaluate oneself favorably. This

phenomenon was found across multiple dimensions of political ideology and multiple

components of self-enhancement. Relative to liberals, conservatives were more likely to endorse

favorable but unlikely self-evaluations, provide socially desirable self-reports, demonstrate the

better-than-average effect, view themselves as unique, judge themselves as ideal, and rate family

members and fellow countrymen as superior to others. Meta-analysis confirmed that self-

enhancement was associated with general, social, economic, and moral dimensions of

conservatism.

An important limitation of our findings is that we constrained our investigation to non-

representative samples of Americans. Our results were consistent across qualitatively different

populations and after statistically controlling for key demographic characteristics, though future

research should investigate this phenomenon with nationally representative samples from

multiple nations.

Our results contribute to a growing body of research linking specific dimensions of

political ideology to motivational influences on social cognition (Adorno et al., 1950; Pratto et

al., 1994; Jost et al., 2003; Graham et al., 2011; Hibbing et al., 2014). Importantly, our findings

do not specify the causal direction of the association between conservatism and political

ideology, and it is not yet clear if conservatives’ stronger self-enhancement confers more

advantages or disadvantages. Past research has linked self-enhancement to both positive and

negative outcomes, and the adaptiveness of conservative self-enhancement is likely to vary as a

function of context (Kurt & Paulhus, 2008). Self-enhancement research has greatly advanced the

44

psychology of mental health (Taylor & Brown, 1988), social cognition (Dunning et al., 1989),

and culture (Hamamura et al., 2007). Its continued investigation in political psychology is likely

to yield similar dividends, and to elucidate whether political ideology is fundamentally motivated

by—or simply diagnostic of—how favorably and realistically one views the self.

45

References

Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. (1950). The

authoritarian personality. New York: Harper.

Belmi, P., & Neale, M. (2014). Mirror, mirror on the wall, who’s the fairest of them all?

Thinking that one is attractive increases the tendency to support

inequality. Organizational Behavior and Human Decision Processes, 124, 133-149.

Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of

psychopathology. Journal of Consulting Psychology, 24, 349-354.

Dunning, D., Meyerowitz, J. A., & Holzberg, A. D. (1989). Ambiguity and self-evaluation: The

role of idiosyncratic trait definitions in self-serving assessments of ability. Journal of

Personality and Social Psychology, 57, 1082-1090.

Graham, J., Nosek, B. A., Haidt, J., Iyer, R., Koleva, S., & Ditto, P. H. (2011). Mapping the

moral domain. Journal of Personality and Social Psychology,101, 366-385.

Hamamura, T., Heine, S. J., & Takemoto, T. R. (2007). Why the better-than-average effect is a

worse-than-average measure of self-enhancement: An investigation of conflicting

findings from studies of East Asian self-evaluations. Motivation and Emotion, 31, 247-

259.

Heine, S. J., & Lehman, D. R. (1997). The cultural construction of self-enhancement: An

examination of group-serving biases. Journal of Personality and Social Psychology, 72,

1268-1283.

Hibbing, J. R., Smith, K. B., Alford, J. R. (2014). Differences in negativity bias underlie

variations in political ideology. Behavioral and Brain Sciences, 37, 297-350.

46

Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as

motivated social cognition. Psychological Bulletin, 129, 339-375.

Jost, J. T., Liviatan, I., van der Toorn, J., Ledgerwood, A., Mandisodza, A., & Nosek, B. A.

(2010). System justification: How do we know it’s motivated. In The psychology of

justice and legitimacy: The Ontario symposium (Vol. 11, pp. 173-203). Erlbaum

Hillsdale, NJ.

Kurt, A., & Paulhus, D. L. (2008). Moderators of the adaptiveness of self-enhancement:

Operationalization, motivational domain, adjustment facet, and evaluator. Journal of

Research in Personality, 42, 839-853.

Loughnan, S., Kuppens, P., Allik, J., Balazs, K., De Lemus, S., Dumont, K., ... & Haslam, N.

(2011). Economic inequality is linked to biased self-perception. Psychological

Science, 22, 1254-1258.

Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R.

Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social psychological

attitudes (pp.17-59). San Diego: Academic Press.

Pelham, B. W., & Swann, W. B. (1989). From self-conceptions to self-worth: on the sources and

structure of global self-esteem. Journal of Personality and Social Psychology, 57, 672-

680.

Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A

personality variable predicting social and political attitudes. Journal of Personality and

Social Psychology, 67, 741–763.

47

Sedikides, C., & Gebauer, J. E. (2009). Religiosity as self-enhancement: A meta-analysis of the

relation between socially desirable responding and religiosity. Personality and Social

Psychology Review, 14, 17-36.

Schlenker, B. R., Chambers, J. R., & Le, B. M. (2012). Conservatives are happier than liberals,

but why? Political ideology, personality, and life satisfaction. Journal of Research in

Personality, 46, 127-146.

Stern, C., West, T. V., & Schmitt, P. G. (2013). The liberal illusion of uniqueness. Psychological

Science, 25, 137-144.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: a social psychological perspective

on mental health. Psychological Bulletin, 103, 193-210.

Wojcik, S. P., Hovasapian, A., Graham, J., Motyl, M., & Ditto, P. H. (in press). Conservatives

report, but liberals display, greater happiness. Science.

Acknowledgements: This chapter is currently under peer-review and would not be possible

without the contributions of co-author Peter Ditto. The authors received no financial support for

the research, authorship, and/or publication of this article and declared no potential conflicts of

interest. Full citation for published article:

Wojcik, S. P., & Ditto, P. H. (under review). Conservative Self-Enhancement. Manuscript

currently under review.

48

Supplementary Materials and Methods

Measures

Balanced Inventory of Desirable Responding (BIDR; Paulhus, 1991). Participants

rated the extent to which 40 statements about the self were true. Each item expressed

unrealistically positive statements about the self that were unlikely to be true, thus higher scores

indicate greater degrees of desirable responding bias. The BIDR assesses two components of

desirable responding bias: self-deceptive enhancement (SDE) and impression management (IM).

The SDE subscale reflects the extent to which participants privately believe unrealistically

favorable self-assessments, and the IM subscale reflects the extent to which participants publicly

report positive self-assessments to impress an audience. BIDR scores were computed using the

continuous scoring method outlined by Paulhus (1994).

Marlowe-Crowne Social Desirability Scale (MC-SDS; Crowne & Marlowe, 1960).

Participants indicated whether 33 self-descriptive statements consisting of positive but unlikely

self-appraisals were true or false. The MC-SDS was scored by summing the total number of

times participants responded with “true”, indicating greater socially desirability bias.

Direct social comparative judgments (direct better-than-average effect scale; Dunning,

Meyerowitz, & Holzberg, 1989). Participants rated whether 20 traits and abilities4 (polite,

observant, cooperative, creative, imaginative, perceptive, reliable, intelligent, mature, friendly,

insecure, uncivil, deceptive, maladjusted, gullible, unpleasant, mean, lazy, vain, spiteful)

describe the participant more, less, or to the same degree as the average person along 5 point

4 For some of the self-enhancement measures (direct and indirect social comparative judgments, false uniqueness,

group member and nation-level enhancement), we also included items related to our other research areas for

exploratory purposes (i.e., subjective well-being and moral foundations related items). Inclusion/exclusion of these

traits in analyses did not alter the overall pattern of results. We have decided to report results for each measure using

only the original trait/ability dimensions from the cited sources to maximize validity of the scores attributed to each

scale.

49

scales (-2 = Much less than the average person, +2 = Much more than the average person;

negative trait items reverse scored prior to analysis).

Indirect social comparative judgments (Hamamura, Heine, & Takemoto, 2007).

Individuals separately rated the extent to which 15 traits and values (competent; enjoy life in

regards to recreation, work, and family; friendly; modest; persistent; responsible; value

friendship; well-liked; affable; bold; caring; cheerful; competitive; easy-going; meticulous)

described themselves, a fictitious, specific other person, and “most other people” along 7-point

scales (1 = Not at all, 7 = Very much). These separate sets of ratings allow for the assessment of

indirect social comparisons to nonspecific others (i.e., the indirectly measured “better-than-

average effect”), indirect social comparisons to a specific other (i.e., the “better-than-a-random-

other effect”), and participants’ tendencies to provide globally positive evaluations to all specific

individuals (i.e., the “everyone-is-better-than-average” effect). The latter effect is not an

indicator of self-enhancement and is not included in analyses.

False uniqueness (Hamamura et al., 2007). Participants estimated the percentage of the

population of the same age and gender that they were superior to across ten traits, including five

independent traits (attractive, interesting, independent, confident, intelligent) and five

interdependent traits (cooperative, loyal, considerate, hard-working, dependable), using 19-point

scales ranging from the 5th to the 95th percentile. Scores on this measure indicate the extent to

which participants rated themselves as above average (computed by subtracting 50 from each

participants’ mean score).

Actual vs. ideal self-discrepancies (Pelham & Swann, 1989). Participants first rated themselves

on eight self-concept dimensions (intellectual ability, social competence, artistic ability, musical

ability, athletic ability, physical attractiveness, leadership ability, common sense), and then rated

50

where they would ideally like to fall on each dimension, on 19-point scales ranging from the 5th

to 95th percentile, relative to their peers. Actual-ideal discrepancies were assessed by

subtracting participants’ ideal self-ratings from their actual reported ratings. Self-enhancement

was operationalized as stronger correspondence between these ratings (i.e., smaller difference

scores).

Family member enhancement (Heine & Hamamura, 1997). Participants were asked to

write the initials of, and their relation to, a family member to whom they felt closest, and to

indicate how close they felt to that person on a 10-point Likert scale. They next estimated the

percentage of the population of the same age and gender that their family member was better

than across the same ten traits as the False Uniqueness measure.

Nation-level enhancement. Participants rated the extent to which each of 10 traits and

abilities (reliable, mature, driving ability, polite, sense of humor, insecure, uncivil, maladjusted,

unpleasant, lazy) described the average American more than, less than, or to the same degree as

the average non-American, on 5-point scales (-2 = Much less than the average non-American, +2

= Much more than the average non-American; negative items reverse scored prior to analysis).

51

Supplementary Table 1.

Regression Coefficients for General Liberalism-Conservatism and Demographic Characteristics

Predicting Self-Enhancement

Predictor Model

Scale

Gen. Polit.

Ideol. Age

Sex

(1=M,

2=F) Educ.

Subjective

SES

Relig.

Attend. Constant F df R sq p

BIDR-SDE

b (SE) .05 (.01) .00 (.00) -.09 (.03) -.02 (.01) .03 (.01) -.03 (.01) 3.94 (.09)

11.07 6, 1871 0.03 < .001 t 5.24 2.17 -2.93 -1.65 3.67 -3.17 42.96

p < .001 0.030 0.003 0.100 < .001 0.002 < .001

Beta 0.13 0.05 -0.07 -0.04 0.09 -0.08

BIDR-IM

b (SE) .04 (.01) .01 (.00) .13 (.04) -.02 (.02) .02 (.01) .08 (.01) 3.36 (.12)

23.25 6, 1871 0.07 < .001 t 3.22 6.81 3.23 -1.18 1.65 5.97 28.82

p 0.001 < .001 0.001 0.239 0.099 < .001 < .001

Beta 0.08 0.16 0.07 -0.03 0.04 0.14

MC-SDS

b (SE) .00 (.00) .00 (.00) .00 (.01) -.01 (.00) .01 (.00) .00 (.00) .43 (.02)

16.3 6, 3275 0.03 < .001 t 2.48 7.90 -0.68 -3.97 4.09 0.96 25.50

p 0.013 < .001 0.495 < .001 < .001 0.337 0.000

Beta 0.05 0.15 -0.01 -0.08 0.08 0.02

Direct Social Comp.

(BAE)

b (SE) .02 (.00) .00 (.00) .00 (.01) .01 (.00) .00 (.00) .02 (.00) .49 (.03)

45.46 6, 4775 0.05 < .001 t 4.16 11.37 0.13 1.66 1.25 5.53 16.13

p < .001 < .001 0.896 0.098 0.213 < .001 < .001

Beta 0.06 0.18 0.00 0.03 0.02 0.08

Indirect

Social Comp.

(BAE)

b (SE) .06 (.06) .01 (.01) -.13 (.20) - - - .23 (.36)

0.78 3, 119 0.02 0.510 t 1.10 0.63 -0.64 - - - 0.64

p 0.275 0.530 0.524 - - - 0.522

Beta 0.10 0.06 -0.06 - - -

Indirect Social

Comp.

(BROE)

b (SE) .00 (.05) .01 (.01) -.16 (.18) - - - .23 (.31)

0.39 3, 119 0.01 0.758 t 0.10 0.75 -0.90 - - - 0.74

p 0.921 0.454 0.372 - - - 0.459

Beta 0.01 0.07 -0.09 - - -

Actual-Ideal

Self-Discrep.

b (SE) 2.33 (.85)

-.05 (.12) 1.76 (2.40) - - -

-31.44

(4.39)

2.69 3, 129 0.06 0.049 t 2.74 -0.38 0.73 - - - -7.17

p 0.007 0.701 0.465 - - - 0.000

Beta 0.24 -0.03 0.06 - - -

False

Unique-ness

b (SE) 1.65 (.50) .10 (.05) 1.33 (1.55) .13 (.66) 1.22 (.46) - 6.37 (4.52)

5.44 5, 260 0.09 < .001 t 3.33 1.87 0.86 0.20 2.66 - 1.41

p 0.001 0.063 0.392 0.839 0.008 - 0.160

Beta 0.20 0.12 0.05 0.01 0.17 -

52

Family

Member Enhance-

ment

b (SE) 1.38 (.58)

.33 (.08)

-1.22

(1.91) - - -

75.83

(3.50)

8.54 3, 134 0.16 < .001 t 2.40 3.90 -0.64 - - - 21.64

p 0.018 0.000 0.524 - - - 0.000

Beta 0.19 0.32 -0.05 - - -

Nation-Level

Enhancemen

t

b (SE) .06 (.01) .00 (.00) -.06 (.04) .00 (.02) .01 (.01) .00 (.01) -.29 (.10)

8.43 6, 421 0.11 < .001 t 5.02 2.96 -1.67 -0.01 1.96 0.38 -2.83

p < .001 0.003 0.095 0.990 0.051 0.701 0.005

Beta 0.25 0.15 -0.08 0.00 0.09 0.02

Note: BIDR-SDE = Balanced Inventory of Desirable Responding: Self-Deceptive Enhancement, BIDR-IM = Balaanced Inventory of Desirable Responding: Impression Management, MC-SDS = Marlowe-Crowne Social Desirability Scale, BAE = Better than Average Effect, BROE =

Better-than-a-Random-Other Effect.

53

Supplementary Table 2.

Regression Coefficients for Social Liberalism-Conservatism and Demographic Characteristics

Predicting Self-Enhancement

Predictor Model

Scale

Social

Political

Ideol. Age

Sex (1 =

M, 2 = F) Educ. Subj. SES

Relig.

Attend. Constant F df R sq p

BIDR-SDE

b (SE) .05 (.01) .00 (.00) -.09 (.03) -.02 (.01) .04 (.01) -.03 (.01) 3.96 (.09)

9.26 6, 1834 0.03 < .001 t 4.38 1.86 -2.88 -1.61 3.95 -3.05 43.28

p < .001 0.063 0.004 0.107 < .001 0.002 < .001

Beta 0.11 0.05 -0.07 -0.04 0.10 -0.08

BIDR-IM

b (SE) .05 (.01) .01 (.00) .13 (.04) -.02 (.02) .02 (.01) .07 (.01) 3.37 (.12)

23.56 6, 1834 0.07 < .001 t 3.91 6.46 3.17 -1.24 1.81 5.29 29.09

p < .001 < .001 0.002 0.216 0.070 < .001 < .001

Beta 0.10 0.16 0.07 -0.03 0.04 0.13

MC-SDS

b (SE) .00 (.00) .00 (.00) .00 (.01) -.01 (.00) .01 (.00) .00 (.00) .44 (.02)

15.41 6, 3224 0.03 < .001 t 1.78 7.79 -0.80 -4.16 4.14 1.08 26.21

p 0.076 < .001 0.424 < .001 < .001 0.278 < .001

Beta 0.03 0.15 -0.01 -0.08 0.08

Direct

Social

Comp. (BAE)

b (SE) .02 (.00) .00 (.00) .00 (.01) .01 (.00) .00 (.00) .02 (.00) .49 (.03)

42.88 6, 4357 0.06 < .001 t 4.65 10.47 0.17 2.27 0.20 4.72 15.94

p < .001 < .001 0.867 0.023 0.842 < .001 < .001

Beta 0.08 0.17 0.00 0.04 0.00 0.08

Indirect

Social Comp.

(BAE)

b (SE) .07 (.06) .01 (.01) -.14 (.20) - - - .23 (.34)

0.9 3, 118 0.02 0.442 t 1.23 0.70 -0.71 - - - 0.68

p 0.222 0.485 0.478 - - - 0.501

Beta 0.11 0.07 -0.07 - - -

Indirect Social

Comp.

(BROE)

b (SE) .00 (.05) .01 (.01) -.16 (.18) - - - .25 (.30)

0.38 3, 118 0.01 0.767 t 0.04 0.72 -0.93 - - - 0.84

p 0.965 0.472 0.354 - - - 0.400

Beta 0.00 0.07 -0.09 - - -

Actual-

Ideal Self-Discrep.

b (SE) 1.90 (.82)

-.05 (.12) 1.94 (2.43) - - -

-29.45

(4.27)

1.94 3, 128 0.04 0.126 t 2.32 -0.45 0.80 - - - -6.90

p 0.022 0.651 0.425 - - - < .001

Beta 0.21 -0.04 0.07 - - -

False Unique-

ness

b (SE) .99 (.51) .09 (.05) 1.03 (1.57) .05 (.68) 1.29 (.47) - 9.12 (4.59)

3.6 5, 254 0.07 0.004 t 1.92 1.67 0.65 0.08 2.76 - 1.99

p 0.056 0.096 0.513 0.938 0.006 - 0.048

Beta 0.12 0.11 0.04 0.01 0.18 -

54

Family

Member Enhance-

ment

b (SE) 1.52 (.62)

.33 (.08)

-1.03

(1.91) - - -

75.68

(3.43)

8.82 3, 133 0.17 < .001 t 2.45 3.95 -0.54 - - - 22.04

p 0.015 < .001 0.524 - - - < .001

Beta 0.20 0.32 -0.04 - - -

Nation-

Level Enhance-

ment

b (SE) .08 (.01) .00 (.00) -.05 (.04) -.01 (.02) .03 (.01) -.01 (.01) -.38 (.11)

11.69 6, 380 0.16 < .001 t 6.38 2.57 -1.23 -0.51 3.44 -0.89 -3.59

p < .001 0.010 0.218 0.607 0.001 0.376 < .001

Beta 0.34 0.13 -0.06 -0.03 0.17 -0.05

Note: BIDR-SDE = Balanced Inventory of Desirable Responding: Self-Deceptive Enhancement, BIDR-IM = Balaanced Inventory of Desirable Responding: Impression Management, MC-SDS = Marlowe-Crowne Social Desirability Scale, BAE = Better than Average Effect, BROE =

Better-than-a-Random-Other Effect.

55

Supplementary Table 3.

Regression Coefficients for Economic Liberalism-Conservatism and Demographic

Characteristics Predicting Self-Enhancement

Predictor Model

Scale

Economic

Political

Ideol. Age

Sex (1 =

M, 2 = F) Educ. Subj. SES

Relig.

Attend. Constant F df R sq p

BIDR-SDE

b (SE) .05 (.01) .00 (.00) -.09 (.03) -.02 (.01) .03 (.01) -.02 (.01) 3.92 (.09)

11.16 6, 1806 0.04 < .001 t 5.49 2.14 -2.67 -1.73 3.28 -2.52 41.85

p < .001 0.032 0.008 0.083 0.001 0.012 < .001

Beta 0.13 0.05 -0.06 -0.05 0.08 -0.06

BIDR-IM

b (SE) .03 (.01) .01 (.00) .12 (.04) -.02 (.02) .02 (.01) .08 (.01) 3.41 (.12)

21.43 6, 1806 0.07 < .001 t 2.47 6.67 2.96 -1.43 1.55 6.83 28.71

p 0.013 < .001 0.003 0.152 0.122 < .001 < .001

Beta 0.06 0.16 0.07 -0.04 0.04 0.16

MC-SDS

b (SE) .01 (.00) .00 (.00) .00 (.01) -.01 (.00) .01 (.00) .00 (.00) .43 (.02)

15.62 6, 3157 0.03 < .001 t 3.24 7.33 -0.49 -4.19 3.96 1.20 24.72

p 0.001 < .001 0.624 < .001 < .001 0.231 < .001

Beta 0.06 0.14 -0.01 -0.08 0.07 0.02

Direct

Social

Comp. (BAE)

b (SE) .01 (.00) .00 (.00) .01 (.01) .01 (.00) .00 (.00) .02 (.00) .50 (.03)

39.93 6, 4260 0.05 < .001 t 3.79 10.55 0.28 1.58 0.04 6.53 15.35

p < .001 < .001 0.778 0.113 0.969 < .001 < .001

Beta 0.06 0.17 0.00 0.03 0.00 0.10

Indirect

Social Comp.

(BAE)

b (SE) .09 (.06) .00 (.01) -.09 (.22) - - - .03 (.39)

1.06 3, 114 0.03 0.367 t 1.60 0.23 -0.40 - - - 0.07

p 0.111 0.818 0.692 - - - 0.943

Beta 0.15 0.02 -0.04 - - -

Indirect Social

Comp.

(BROE)

b (SE) .03 (.05) .01 (.01) -.10 (.19) - - - .08 (.34)

0.37 3, 114 0.01 0.773 t 0.61 0.57 -0.56 - - - 0.24

p 0.541 0.568 0.578 - - - 0.810

Beta 0.06 0.06 -0.06 - - -

Actual-

Ideal Self-Discrep.

b (SE) 1.01 (.77)

-.01 (.12) 1.26 (2.48) - - -

-27.18

(4.46)

0.71 3, 127 0.02 0.550 t 1.33 -0.10 0.51 - - - -6.10

p 0.187 0.918 0.613 - - - < .001

Beta 0.12 -0.01 0.05 - - -

False Unique-

ness

b (SE) 1.00 (.44) .09 (.05) .92 (1.59) .00 (.67) 1.07 (.48) - 9.72 (4.62)

3.65 5, 249 0.07 0.003 t 2.26 1.80 0.58 0.00 2.25 - 2.11

p 0.025 0.074 0.564 0.999 0.025 - 0.036

Beta 0.14 0.12 0.04 0.00 0.15 -

56

Family

Member Enhance-

ment

b (SE) 1.58 (.53)

.35 (.08) -.85 (1.92) - - -

73.74

(3.63)

9.87 3, 132 0.18 < .001 t 2.97 4.31 -0.44 - - - 20.32

p 0.004 < .001 0.659 - - - < .001

Beta 0.24 0.34 -0.04 - - -

Nation-

Level Enhance-

ment

b (SE) .04 (.01) .00 (.00) -.06 (.04) -.01 (.02) .03 (.01) .02 (.01) -.32 (.11)

6.6 6, 374 0.10 < .001 t 3.31 2.36 -1.52 -0.64 3.10 1.23 -2.86

p 0.001 0.019 0.129 0.520 0.002 0.219 0.004

Beta 0.17 0.13 -0.08 -0.04 0.16 0.06

Note: BIDR-SDE = Balanced Inventory of Desirable Responding: Self-Deceptive Enhancement, BIDR-IM = Balaanced Inventory of Desirable Responding: Impression Management, MC-SDS = Marlowe-Crowne Social Desirability Scale, BAE = Better than Average Effect, BROE =

Better-than-a-Random-Other Effect.

57

Supplementary Table 4.

Regression Coefficients for Moral Liberalism-Conservatism and Demographic Characteristics

Predicting Self-Enhancement

Predictor Model

Scale

Moral

Political

Ideol. Age

Sex (1 =

M, 2 = F) Educ. Subj. SES

Relig.

Attend. Constant F df R sq p

BIDR-SDE

b (SE) .05 (.02) .00 (.00) -.08 (.04) -.02 (.01) .03 (.01) -.03 (.01) 4.16 (.10)

7.06 6, 1477 0.03 < .001 t 2.98 3.00 -2.25 -1.46 2.81 -2.77 42.58

p 0.003 0.003 0.025 0.143 0.005 0.006 < .001

Beta 0.08 0.08 -0.06 -0.04 0.08 -0.08

BIDR-IM

b (SE) -.01 (.02) .01 (00) .10 (.05) -.02 (.02) .02 (.01) .09 (.01) 3.46 (.12)

17.61 6, 1477 0.07 < .001 t -.51 6.87 2.17 -1.16 1.86 6.21 28.05

p .613 < .001 .030 .244 .063 < .001 < .001

Beta -.01 .19 .06 -.03 .05 .17

MC-SDS

b (SE) -.01 (.00) .00 (.00) -.02 (.01) -.01 (.00) .01 (.00) .01 (.00) .45 (.02)

12.11 6, 2061 0.03 < .001 t -2.73 6.65 -2.36 -3.23 3.46 3.04 22.16

p .006 < .001 .018 .001 .001 .002 < .001

Beta -.06 .15 -.05 -.08 .08 .07

Direct

Social

Comp. (BAE)

b (SE) .01 (.01) .00 (.00) .00 (.02) .01 (.01) .02 (.01) .01 (.01) .39 (.06)

17.06 6, 1541 0.06 < .001 t .76 6.76 -0.06 1.60 3.78 1.29 7.02

p .449 < .001 .950 .110 < .001 .196 < .001

Beta .02 .18 .00 .05 .10 .03

False Unique-

ness

b (SE) 2.29 (.59)

.08 (.04) 4.36 (1.41) -.14 (.58) 1.00 (.40) -

12.51

(3.84)

6.68 5, 290 0.10 < .001 t 3.90 1.76 3.10 -.24 2.50 - 3.26

p .000 .080 .002 .809 .013 - 0.001

Beta .22 .11 .18 -.02 .15 -

Nation-

Level Enhance-

ment

b (SE) .09 (.02) .00 (.00) -.03 (.04) .01 (.02) .03 (.01) .01 (.01) -.19 (.10)

11.73 6, 464 0.13 < .001 t 5.67 2.36 -0.83 0.47 2.92 .54 -1.90

p < .001 .019 .405 .638 .004 .588 0.059

Beta .27 .11 -.04 .02 .13 .03

Note: BIDR-SDE = Balanced Inventory of Desirable Responding: Self-Deceptive Enhancement, BIDR-IM = Balaanced Inventory of Desirable Responding: Impression Management, MC-SDS = Marlowe-Crowne Social Desirability Scale, BAE = Better than Average Effect, BROE =

Better-than-a-Random-Other Effect.

58

CHAPTER 4:

Conservatives Report, But Liberals Display, Greater Happiness

59

Abstract

Research suggesting that political conservatives are happier than political liberals has relied

exclusively on self-report measures of subjective well-being. We show that this finding is fully

mediated by conservatives’ self-enhancing style of self-report (Study 1; N = 1,433), and then

describe three studies drawing from “big data” sources to assess liberal-conservative differences

in happiness-related behavior (Studies 2-4; N = 4,936). Relative to conservatives, liberals more

frequently used positive emotional language in their speech, and smiled more intensely and

genuinely in photographs. Our results were consistent across large samples of online survey

takers, American politicians, Twitter users, and LinkedIn users. Our findings illustrate the

nuanced relationship between political ideology, self-enhancement, and happiness, and

illuminate the contradictory ways that happiness differences can manifest across behavior and

self-reports.

Keywords: political ideology, happiness, self-enhancement, linguistic analysis, FACS

analysis.

60

Conservatives Report, But Liberals Display, Greater Happiness

Are political conservatives happier than political liberals? Several recent studies have

found that greater political conservatism predicts higher levels of self-reported happiness and life

satisfaction in the United States (Brooks, 2008; Schlenker, Chambers, & Le, 2012; Napier &

Jost, 2008; Onraet, Van Hiel, & Dhont, 2013). This “ideological happiness gap” has been

attributed to a number of different psychological factors. Some explanations posit that

conservatives’ greater happiness is due to a suite of adaptive personal, social, and cultural values

(Brooks, 2008), such as high levels of personal agency, optimism, and transcendent moral beliefs

(Schlenker et al., 2012). Others portray conservatism as a protective or even defensive

mechanism that serves the palliative function of justifying troubling societal inequalities (Napier

& Jost, 2008). Although researchers disagree over the mechanism underlying the happiness gap,

meta-analytic review has confirmed that it is a small but reliable effect (r = .12; Onraet et al.,

2013).

All of the data supporting the relation between political conservatism and subjective well-

being rely on self-report measures (e.g., stated agreement with face-valid items such as “In most

ways my life is close to ideal”; Diener, Emmons, Larsen, & Griffin, 1985; Onraet et al., 2013).

This is not surprising given that self-reports are the foundation of subjective well-being research5

(Diener, 2000), but reliance on any single methodology within an area of research can result in

systematic methodological artifacts due to common method variance (Campbell & Fiske, 1959).

There are many challenges involved in self-report research (Zou, Schimmack, & Gere, 2013;

Nisbett & Wilson, 1977; Gawronski & Payne, 2010), including the influence of self-

enhancement motives, which commonly lead to unrealistically favorable self-assessments

5 The term “subjective well-being” is used interchangeably with “happiness”, and consists of life satisfaction,

positive and negative affect, and domain satisfaction (6).

61

(Brown, 2012). Self-reports of happiness and life satisfaction, much like self-assessments of

other valued characteristics, are susceptible to self-enhancing distortions (Wojcik & Ditto, 2014).

Both individual differences in, and experimental manipulations of, self-enhancement motivation

predict the tendency to report happiness and life satisfaction at unrealistically positive levels (i.e.,

the “happier-than-average effect”; Wojcik & Ditto, 2014).

Self-enhancing tendencies are not evenly distributed across populations (Heine,

Kitayama, & Hamamura, 2007; Eriksson & Funcke, 2014; Loughnan, et al., 2011), and there are

reasons to suspect that liberals and conservatives may self-enhance to differing degrees.

Conservatism6 has been characterized as an ideology grounded in ego defensiveness (Wilson,

1973), enhanced sensitivity to negativity (Hibbing, Smith, & Alford, 2014), and personality

dimensions related to defensive forms of motivated social cognition (Jost, Glaser, Kruglanski, &

Sulloway, 2003; Jost, Liviatan, Van der Toorn, Ledgerwood, Mandisodza, & Nosek, 2010). Self-

enhancement is also associated with a number of factors related to politically conservative

ideologies: it is more pronounced among individualistic cultures (Heine et al., 2007), religious

people (Eriksson & Funcke, 2014), and competitive, hierarchically-oriented groups (Loughnan et

al., 2011). Given that self-reported happiness is related to political conservatism (Onraet et al.,

2013) and self-enhancement (r = .10; Wojcik & Ditto, 2014) at comparable magnitudes, it is

possible that ideological happiness differences may simply be an example of conservatives’

stronger tendency to evaluate the self favorably.

6 Political ideology is a complex and nebulous construct that is most frequently operationalized in this literature

using self-report measures of ideology, party affiliation, or scores on psychological constructs associated with self-

reported conservatism such as authoritarianism and social dominance orientation (Onraet et al., 2013; Jost et al.,

2003). The meaning ascribed to terms like “liberal”, “progressive”, and “conservative” also varies over time and

across national contexts. Given this conceptual complexity, we have restricted our analyses to American samples

where possible and tested our effects across multiple operationalizations of liberalism-conservatism, including self-

identified ideology (Study 1) and party affiliation (Study 2), but also using behavioral measures such as

congressional voting records (Study 2), social media subscriptions (Study 3), and involvement with liberal and

conservative organizations (Study 4).

62

In Study 1, we examined whether conservatives’ reports of greater subjective well-being,

relative to liberals, could be attributed to self-enhancing tendencies. Visitors to YourMorals.org,

a psychological research web site, reported their political ideology and completed the

Satisfaction With Life Scale, the most frequently used measure of subjective well-being (Diener

et al., 1985), as well as the Balanced Inventory of Desirable Responding, a well-validated

measure of the tendency to engage in self-deceptive enhancement (Paulhus, 1994). As expected,

increasing political conservatism predicted greater reported life satisfaction (r[1433] = .10, p <

.001; adjusting for demographic characteristics: β = .09, p = .002). This happiness gap was

similar in magnitude to that found in past research (Onraet et al., 2013). Importantly, we also

found that self-deceptive enhancement was higher among conservatives than liberals (r[1433] =

.16, p < .001; adjusting for demographics: β = .18, p < .001). A bootstrapped mediation analysis

revealed that, as hypothesized, self-deceptive enhancement fully mediated the ideology-life

satisfaction association (indirect effect: b = .05, p < .001, 95% CI = [.03, .07]; see Table 1).

Table 1.

Unstandardized Coefficients from Mediation Analyses of Self-Reported Life Satisfaction.

Model 1 Model 2

Predictor Step 1 Step 2 Step 1 Step 2

Political conservatism .09 (.02) *** .04 (.02) + .07 (.01) *** .03 (.02) ns

Self-deceptive enhancement .73 (.05) *** .69 (.05) ***

SES

.02 (.01) ** .07 (.01) ***

Education

-.02 (.02) ns .21 (.03) ***

Sex (1=female)

-.03 (.04) ns .32 (.07) ***

63

Age

.00 (.00) ns .00 (.00) ns

Age squared

.00 (.00) ns .00 (.00) ***

Religious attendance

-.03 (.02) * .07 (.02) **

Constant 4.13 (.08) *** 1.35 (.22) *** 3.87 (.10) *** -.42 (.29) ns

F(df)

F(1, 1431) =

14.45 ***

F(2, 1430) =

95.63 ***

F(7, 1373) =

10.15 ***

F(8, 1372) =

46.15 ***

R2 0.01 0.12 0.05 0.21

Note. Standard errors are given in parentheses. Indirect effect for bootstrapped mediation

analysis with 5000 resamples for Model 1: b = .05, SE = .01, z = 5.57, p < .001, 95% CI = [.03,

.07]; for Model 2: b = .05, SE = .01, z = 5.47, p < .001, 95% CI = [.03, .07].

ns p > .10, + p < .10, * p < .05, ** p < .01, *** p < .001.

These results indicate that conservatives’ greater reports of life satisfaction were

attributable to their stronger tendency to engage in self-enhancement. But there are two possible

interpretations of these findings. First, conservative participants may be experiencing greater

happiness and life satisfaction than liberals as a beneficial consequence of their self-enhancing

predispositions (Taylor & Brown, 1988). Alternatively, conservatives’ elevated reports may

simply be an example of their stronger tendency to provide unrealistically favorable self-

assessments. This interpretative ambiguity highlights a major limitation of relying on self-reports

alone: they do not distinguish between genuine and superficial presentations of happiness.

Behavioral indicators of happiness provide important information that can clarify this

ambiguity. If we take self-reports of well-being at face value, rather than as examples of self-

enhanced assessments, we would expect to observe conservatives’ greater happiness in

unobtrusive measures of happiness-related behavior—for example, in the emotional content of

liberals’ and conservatives’ speech, or in the frequency and intensity of their smiling behavior.

64

Assessing smiling behavior also allows us to distinguish between genuine and superficial

expressions of happiness. Intense, genuine smiling, known as Duchenne smiling, involves the

muscles lifting the corner of the mouth as well as those orbiting the eye (Ekman, Davidson, &

Friesen, 1990). Non-Duchenne (also known as social, deceptive, or standard) smiling involves

only the muscles lifting the corners of the mouth and is less often related to genuine feelings of

happiness or enjoyment. Non-Duchenne smiling is also less predictive than Duchenne smiling of

beneficial long-term psychological and physical health outcomes (Ekman et al., 1990; Harker &

Keltner, 2001).

In Study 2, we examined happiness-related behavior among the United States’ most

salient liberals and conservatives: members of the U.S. Congress. We assessed two behavioral

indicators of happiness within this group: the use of positive and negative emotional language

from a text analysis of the 2013 United States Congressional Record, and the smiling behavior

exhibited in their publicly available photographs. For our linguistic analysis, we recorded word

use frequencies for members of the 113th U.S. Congress using relevant terms from one of the

most frequently used emotion scales in the psychological literature, the Positive and Negative

Affect Schedule: Expanded Form (PANAS-X; Watson & Clark, 1994). We assessed political

ideology using continuous liberalism-conservatism scores based on analysis of each politician’s

voting record (That’s My Congress | Legislative Scorecards, 2014) and controlled for each

member’s legislative chamber, status as a political minority/majority member in their chamber,

overall word usage, and demographic characteristics. Greater conservatism was associated with a

small but significant decrease in positive affect word use (β = -.16, p < .001). Conservatism was

not significantly associated with the use of negative affect words, joviality-related words, or

65

sadness-related words (Table 2). We replicated this analysis with political ideology defined by

party identification instead of voting record and found a similar pattern of results.

Table 2.

Unstandardized Regression Coefficients for Linguistic Analyses and Smile Analyses for

Politicians in 2013 US Congressional Record.

Linguistic Analyses Smile Analyses

Predictor

Positive

Affect

Negative

affect Joviality Sadness Overall Smile

Zygomatic

Major

(AU12)

Orbicularis

Oculi (AU6)

Political conservatism -.88 (.19) *** .04 (.05) ns -.04 (.07) ns -0.09 (.05) + -.03 (.01) * -.02 (.01) + -.03 (.02) *

Chamber (1=House) 2.79 (2.28) ns 1.07 (.58) + .08 (.91) ns

2.18 (.61)

*** .11 (.13) ns .02 (.12) ns .24 (.17) ns

Party Majority Status

(1=Majority)

7.06 (1.93)

***

-.60 (.49) ns -.21 (.77) ns .73 (.52) ns .13 (.13) ns .13 (.12) ns .13 (.16) ns

Age -.02 (.07) ns .00 (.02) ns -.06 (.03) + .00 (.02) ns .00 (.00) ns .00 (.00) ns .00 (.01) ns

Ethnicity: African

American

-.38 (3.07) ns -.23 (.79) ns -.62 (1.23) ns -.43 (.83) ns -.58 (.20) ** -.28 (.19) ns -.88 (.26) **

Ethnicity: Hispanic

American -2.38 (3.01) ns -.44 (.77) ns

3.32 (1.21)

** -.46 (.81) ns -.16 (.20) ns .00 (.18) ns -.33 (.25) ns

Ethnicity: Asian

American

-2.17 (5.58) ns .26 (1.43) ns -.50 (2.24) ns -.47 (1.50) ns -.05 (.37) ns .13 (.34) ns -.22 (.47) ns

Ethnicity: Other -5.61 (7.66) ns -.95 (1.97) ns

-1.47 (3.08)

ns

.35 (2.06) ns -.52 (.50) ns .06 (.47) ns -1.11 (.64) +

Sex (1=male) -2.38 (2.04) ns -.36 (.52) ns -.84 (.82) ns -1.05 (0.55) + -.38 (.13) **

-0.63 (0.12)

*** -.13 (.17) ns

Wordiness .00 (.00) *** .00 (.00) *** .00 (.00) *** .00 (.00) ***

Constant 2.91 (5.29) ns -.01 (1.36) ns 4.29 (2.12) *

-1.34 (1.42)

ns

2.38 (.34)

***

3.15 (.32)

***

1.60 (.44)

***

F(df) F(10, 516) =

115.69 ***

F(10, 516) =

10.01 ***

F(10, 516) =

38.64 ***

F(10, 516) =

92.04 ***

F(9, 517) =

2.44 **

F(9, 517) =

4.08 ***

F(9, 517) =

2.12 **

R2 0.69 0.16 0.43 0.64 0.04 0.07 0.04

Note. Standard errors are given in parentheses.

ns p > .10, + p < .10, * p < .05, ** p < .01, *** p < .001.

66

We also tested the reliability of the relationship over time by assessing emotional

language at the party-level across 18 years of Congressional Record data including over 432

million words. Members of the liberal leaning Democratic Party used a higher ratio of positive to

negative affect words (M = 13.65:1) than members of the conservative leaning Republican Party

(M = 11.50:1), including a higher frequency of positive affect word usage in 17 of 18 years.

This pattern of emotional expression was not limited to linguistic analyses. We next

analyzed smiling behavior in photos from the Congressional Pictorial Directory of the 113th U. S.

Congress (United States Congress Joint Committee on Printing, 2013) using the Facial Action

Coding System (FACS; Ekman & Friesen, 1978). A FACS-certified coder assessed the intensity

of two action units (AUs) associated with genuine smiling behavior: activity in the orbicularis

oculi (AU6) and the zygomatic major (AU12). We found that increasing political conservatism

predicted less intense overall smiling (β = -.12, p = .032), controlling for the same variables as

the linguistic analysis above (Table 2). We observed only marginally significant differences in

the intensity of smiling behavior in the muscles lifting the corners of the mouth (AU12: β = -.10,

p = .096), but conservatism predicted significantly less intense facial action in the muscles

around the eyes that indicate genuine happiness (AU6: β = -.13, p = .031). The odds of

displaying non-Duchenne smiles (i.e., action in AU12 but not AU6) were slightly higher for

conservatives than for liberals, but this did not reach statistical significance (controlling for

demographics: OR = 1.04, p = .206).

Study 2 revealed no evidence of greater emotionally positive behavior among

conservative elected officials. Instead, it was liberal politicians who tended to more frequently

express positive emotional language, smile more intensely, and smile more genuinely. These

67

effects were modest in size, but in the opposite direction of the previously observed self-report-

based happiness gap (Onraet et al., 2013). Of course, elected political leaders are not

representative of liberal and conservative individuals more generally, and it is unclear how well

speech and facial expressions occurring within the confines of Capitol Hill reflect similar

happiness-related behaviors in less overtly political contexts.

Accordingly, Study 3 assessed the linguistic content of 47,257 Twitter status updates

(“tweets”) from liberal and conservative members of the general public. We analyzed the

statuses of individuals who subscribed to (“followed”) the official Twitter pages of either the

Democratic or Republican Party, excluding those following both, under the assumption that users

who followed one party exclusively were likely to share that party’s political views. We assessed

the emotional content of each tweet using word lists from the PANAS-X, the Linguistic Inquiry

Word Count software (LIWC; Pennebaker, Francis, & Booth, 2001), and lists of “happy” and

“sad” emoticons. Logistic regressions predicting the presence or absence of emotion

words/emoticons at the tweet-level were conducted with political party followed as the

independent variable. Relative to Democratic Party subscribers’ updates, Republican Party

subscribers’ updates were significantly less likely to contain positive emotion words, joviality

words, and happy emoticons, and significantly more likely to contain negative emotion words

(all ps < .05; Table 3). Marginal effects also revealed similar patterns on the PANAS-X negative

affect and sadness subscales.

68

Table 3.

Odds of Emotion-Related Words Appearing in Tweets of Republican Party Subscribers

Relative to Democratic Party Subscribers.

Positive Emotion Word Lists Negative Emotion Word Lists

PANAS-X

Positive Affect

PANAS-X Joviality

LIWC

Positive Emotion

Happy Emoticons

PANAS-X

Negative Affect

PANAS-X Sadness

LIWC

Negative Emotion

Sad

Emoticons

Odds Ratio 0.88 0.80 0.95 0.83 1.44 1.32 1.07 0.77

SE 0.08 0.06 0.02 0.07 0.20 0.22 0.03 0.17

p 0.159 0.007 0.018 0.030 0.051 0.093 0.015 0.229

Note: Odds Ratios above 1.00 indicate conservative tweets contained a higher percentage than liberal tweets; values below 1.00 indicate the

opposite. PANAS-X = Positive and Negative Affect Schedule, Extended Version; LIWC = Linguistic Inquiry Word Count.

N of tweets for PANAS-X and LIWC analyses = 37,857; N for emoticon analyses = 47,257.

In Study 4, we analyzed 457 publicly available photographs of individuals from

LinkedIn, a business-oriented social networking website. We selected participants who publicly

self-identified as employees at organizations strongly associated with ideologically liberal or

conservative values (e.g., Planned Parenthood vs. the Family Research Council), under the

assumption that the majority of employees at these organizations were likely to share the

organizations' ideological views. As we found among Democratic and Republican Congress

members in Study 2, smiles were marginally more intense among employees at ideologically

liberal organizations (M = 1.98, SD = 1.35) than among employees of ideologically conservative

organizations (M = 1.75, SD = 1.24 ; t[455] = 1.91, d = .18, p = .057; controlling for gender and

ethnicity: β = -.07, p = .143). We again did not observe ideological differences in the intensity of

smiling behavior in the muscles lifting the corners of the mouth (AU12; liberals: M = 2.18, SD =

1.45; conservatives: M = 1.99, SD = 1.32; t[455] = 1.42, d = .13, p = .157; controlling for gender

and ethnicity: β = -.03, p = .458). However, individuals at conservative organizations expressed

significantly less intense facial action in the muscles around the eyes that indicate genuine

69

feelings of happiness (AU6; liberals: M = 1.78, SD = 1.43; conservatives; M = 1.50, SD = 1.38;

t[455] = 2.14, d = .20, p = .033; controlling for gender and ethnicity: β = -.09, p = .051). We also

found that the proportion of individuals displaying non-Duchenne smiles was significantly

higher at conservative organizations (n = 41 of 217, or 18.89%) than at liberal organizations (n =

24 of 240, or 10.00%; χ2 = 7.39, p = .007; controlling for demographics: OR = 2.35, p = .003).

Together, our studies found that political liberals exhibited more frequent and intense

happiness-related behavior than political conservatives. Contrary to the pattern found in self-

reports of happiness and life satisfaction (Brooks, 2008; Schlenker et al., 2012; Napier & Jost,

2008; Onraet et al., 2013), linguistic analyses of massive archives of text from both elected

politicians and the general public revealed a modest but consistent tendency for liberals to use

more positive emotional language than conservatives. In addition, we found that liberal

politicians and employees at organizations promoting liberal values smiled more intensely and

genuinely than their conservative counterparts. Although the effects in these studies were small,

they consistently revealed greater happiness-related behavior among liberals, rather than

conservatives. These behavioral data support our finding from Study 1 that differences in

liberals’ and conservatives’ reports of happiness can be attributed to conservatives’ stronger

tendency to provide flattering self-assessments.

These findings add nuance to past research suggesting that subjective well-being is a

multifaceted construct composed of both hedonic or experiential aspects (e.g., reports of current

affect, observed behavior) and eudaimonic or evaluative ones (e.g., global assessments of

happiness or life-satisfaction; Deci & Ryan, 2008; Krueger & Stone, 2014). What does it mean

when self-reports and behavioral indicators provide contradictory data? Happiness lacks an

objective “gold standard” validation criterion, and both self-report (Diener, 2000) and behavioral

70

(Harker & Keltner, 2001; Pressman & Cohen, 2012) measures of happiness have been linked to

beneficial short- and long-term outcomes. For this reason, it would be a mistake to infer from our

data that liberals are “objectively” happier than conservatives, or that conservatives’ self-

enhancing tendencies are necessarily maladaptive. It is not currently known whether claims of

happiness bolstered by self-enhancement motivation are any less advantageous than more

“genuine” reports of positive well-being. In fact, there is evidence that self-enhancement

tendencies can often facilitate psychological adjustment, including the capacity for productive

work, satisfying relationships, and resilience to psychological threat (Taylor & Brown, 1988).

This possibility raises important questions for future research, including the potential existence

of “defensive” forms of happiness similar to other forms of defensive self-evaluation, such as

those uncovered in research on self-esteem (Kernis, 2003).

Our studies did not directly test previously proposed mechanisms for the ideological

happiness gap, such as differences in ideological values (Brooks, 2008), system justification

motivation (Napier & Jost, 2008), or judgments about personal agency, optimism, and

transcendent moral beliefs (Schlenker et al., 2012). However, it seems plausible that these

explanations share common motivational mechanisms with self-enhancement (Jost et al., 2010;

Taylor & Brown, 1988). A key difference is that these previous accounts would predict

happiness-related behavior to correspond with self-report evidence of greater conservative

happiness. Our self-enhancement-based account explains this discrepancy. Future research

should continue to examine whether these motivational mechanisms truly lead to genuinely

positive emotional experiences and greater psychological adjustment, or if they primarily

motivate the superficial report of one’s positive qualities.

71

The questions raised by this research are important because of a growing interest in using

self-report measures of happiness to inform public policy (Brooks, 2008, Diener, 2000). Our

research supports those recommending caution about promoting any particular ideology or

policy as a road to happiness (Krueger & Stone, 2014). Research investigating self-report-based

happiness differences between nonrandomized groups (e.g., cultures, nations, religious groups)

may inadvertently capture differences in self-reporting styles rather than actual differences in

emotional experience. Both behavioral measures and self-reports of subjective well-being are

valuable tools, but any comprehensive assessment of subjective well-being should involve

multiple methodological approaches (Diener, 2000; Zou et al., 2013). Reliance on any single

methodology is likely to lead to an oversimplified account of not only who is happier than

whom, but also what it means to be happy at all.

72

References

Brooks, A. C. (2008). Gross National Happiness: Why Happiness Matters for America—And

How We Can Get More Of It. New York: Basic Books.

Brown, J. D. (2012). Understanding the “better than average” effect: Motives (still) matter.

Personality and Social Psychology Bulletin, 38, 209-219.

Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation of the

multitrait-multi-method matrix. Psychological Bulletin, 56, 81-105.

Deci, E. L., & Ryan, R. M. (2008). Hedonia, eudaimonia, and well-being: An introduction.

Journal of Happiness Studies, 9, 1-11.

Diener, E. D. (2000). The science of happiness and a proposal for a national index. American

Psychologist, 55, 34-43.

Diener, E. D., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction With Life

Scale. Journal of Personality and Social Psychology, 49, 71-75.

Ekman, P., & Friesen, W. V. (1978). Facial Action Coding System: Investigator’s Guide.

Consulting Psychologists Press.

Ekman, P., Davidson, R. J., & Friesen, W. V. (1990). The Duchenne smile: Emotional

expression and brain physiology: II. Journal of Personality and Social Psychology, 58,

342-353.

Eriksson, K., & Funcke, A. (2014). Humble self-enhancement: Religiosity and the better-than-

average effect. Social Psychological and Personality Science, 5, 76-83.

Gawronski, B., & Payne, B. K. (Eds.) (2010). Handbook of Implicit Social Cognition:

Measurement, Theory, and Applications. Guilford Press.

73

Harker, L., & Keltner, D. (2001). Expressions of positive emotion in women’s college yearbook

pictures and their relationship to personality and life outcomes across adulthood. Journal

of Personality and Social Psychology, 80, 112-124.

Heine, S. J., Kitayama, S., & Hamamura, T. (2007). Which studies test whether self-

enhancement is pancultural? Reply to Sedikides, Gaertner, and Vervea. Asian Journal of

Social Psychology, 10, 198-200.

Hibbing, J. R., Smith, K. B., & Alford, J. R. (2014). Differences in negativity bias underlie

variations in political ideology. Behavioral and Brain Sciences, 22, 1254-1258.

Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political conservatism as

motivated social cognition. Psychological Bulletin, 129, 339-375.

Jost, J. T., Liviatan, I., Van der Toorn, J., Ledgerwood, A., Mandisodza, A., & Nosek, B. A.

(2010). System justification: How do we know it’s motivated? In A. C. Kay et al. (Eds.),

The psychology of justice and legitimacy: The Ontario symposium. Erlbaum.

Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological Inquiry,

14, 1-26.

Krueger, A. B., & Stone, A. A. (2014). Progress in measuring subjective well-being: Moving

toward national indicators and policy evaluations. Science, 346, 42-43.

Loughnan, S., Kuppens, P., Allik, J., Balazs, K., De Lemus, S., Dumont, K., et al. (2011).

Economic inequality is linked to biased self-perception. Psychological Science, 22, 1254-

1258.

Napier, J. L., & Jost, J. T. (2008). Why are conservatives happier than liberals? Psychological

Science, 19, 565-572.

74

Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on

mental processes. Psychological Review, 84, 231-259.

Onraet, E., Van Hiel, A., & Dhont, K. (2013). The relationship between right-wing ideological

attitudes and psychological well-being. Personality and Social Psychology Bulletin, 39,

509-522.

Paulhus, D. L. (1994). Balanced Inventory of Desirable Responding: Reference manual for

BIDR Version 6. Unpublished manuscript, University of British Columbia, Vancouver,

Canada.

Pennebaker, J. W., Francis, M. F., & Booth, R. J. (2001). Linguistic inquiry and word count:

LIWC-2001 operations manual. Austin, TX.

Pressman, S. D., & Cohen, S. (2012). Positive emotion word use and longevity in famous

deceased psychologists. Health Psychology, 31, 297-305.

Schlenker, B. R., Chambers, J. R., & Le, B. M. (2012). Conservatives are happier than liberals,

but why? Political ideology, personality, and life satisfaction. Journal of Research in

Personality, 46, 127-146.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective

on mental health. Psychological Bulletin, 103, 193-210.

"That's My Congress | Legislative Scorecards." Web log post. That’s My Congress.

<http://thatsmycongress.com/index.php/legislative-scorecards> (03 Jan. 2014).

United States Congress Joint Committee on Printing, (Ed.) (2013). Congressional Pictorial

Directory: One Hundred Thirteenth Congress. Government Printing Office: Washington.

Watson, D., & Clark, L. A. (1994). The PANAS-X: Manual for the Positive and Negative Affect

Schedule-Expanded Form. The University of Iowa, Ames.

75

Wilson, G. D. (Ed.) (1973). The Psychology of Conservatism. London: Academic Press.

Wojcik, S. P., & Ditto, P. H. (2014). Motivated happiness: Self-enhancement inflates self-

reported subjective well-being. Social Psychological and Personality Science, 5, 525-

534.

Zou, C., Schimmack, U., & Gere, J. (2013). The validity of well-being measures: A multiple-

indicator-multiple-rater model. Psychological Assessment, 25, 1247-1254.

Acknowledgements: This chapter is currently in press at Science, and would not be possible

without the contributions of co-researchers Arpine Hovasapian, Jesse Graham, Matt Motyl, and

Peter Ditto. The authors received no financial support for the research, authorship, and/or

publication of this article and declared no potential conflicts of interest. We thank Jonathan Haidt

and Arthur Stone for their feedback on an earlier draft of the manuscript, and Ryan Idrogo-Lam

and Jonathan Larsen for their assistance in data collection. The data used in this paper are

available online from ICPSR at http://doi.org/10.3886/E26078V1

76

Supplementary Materials and Methods

Study 1

Participants were 1,433 visitors to YourMorals.org (45% female, mean age = 40.2).

Visitors were included in Study 1 if they reported their political ideology at registration and

subsequently completed the Satisfaction With Life Scale (Diener et al., 1985) and the Balanced

Inventory of Desirable Responding (BIDR; Paulhus, 1994). The Satisfaction With Life Scale is

one of the most widely used measures of subjective well-being, consisting of five face valid

Likert-style questions (e.g., “The conditions of my life are excellent”; 1 = “Strongly disagree”, 7

= “Strongly agree”). The BIDR measures two underlying components of unrealistically positive

self-reporting styles. Its Impression Management subscale assesses deliberately distorted self-

assessments related to self-presentation concerns, and its Self-Deceptive Enhancement subscale

assesses honestly believed, but unrealistically favorable self-assessments. We were primarily

interested in the latter scale because it has been used to predict inflated reports of subjective

well-being in past research (Wojcik & Ditto, 2014). Political ideology was coded on a 7-point

Likert scale (1 = “Very liberal”, 7 = “Very conservative”; additional options available for

libertarians and others). Due to our interest in liberal-conservative differences, those who did not

identify along the liberalism-conservatism scale were excluded from analyses.

Study 2

Linguistic analysis. We used the application programming interface (API) of

CapitolWords.org, an online service that streamlines the acquisition of word use frequency data

from the U.S. Congressional Record, to tally the frequency with which politicians used positive

and negative emotion words in 2013, which was the most recent year for which full data were

available. The CapitolWords service is provided free of charge by the Sunlight Foundation, a

77

non-profit organization providing online resources to promote government transparency. We

gathered word count data for each individual appearing in the Congressional Pictorial Directory

for the 113th Congress (United States Congress Joint Committee on Printing, 2013), including

100 Senators and 433 Representatives (mean age = 58.1, 17.5% female; 2 Congressional Seats

were vacant at the time of the directory’s publishing). Demographic information for each

participant was gathered from publicly available information found online. We assessed political

ideology with a continuous liberalism-conservatism score based on each politician’s voting

record. These scores were gathered from a popular independent political legislation blog’s

“legislative scorecards” (That’s My Congress | Legislative Scorecards, 2014) and transformed so

that they ranged from -10 to +10, with higher scores indicating increasingly conservative voting

records. Voting record scores were available for all but four of the politicians in our 2013

analysis.

We recorded word use frequencies for relevant items from the Positive and Negative

Affect Schedule: Expanded Form (PANAS-X; Watson & Clark, 1994), which has been used to

predict meaningful outcomes in previous linguistic analyses (Pressman & Cohen, 2012). Due to

the API’s inability to process word stems, regular expressions, and wildcard characters, use of

Linguistic Inquiry Word Count (LIWC; Pennebaker et al., 2001) emotion dictionaries (which

contain hundreds of items requiring wildcard characters; e.g., happi* and joy*) was not feasible

in Study 2. We recorded word use frequencies for each politician using four PANAS-X

subscales, including its positive affect, negative affect, joviality, and sadness subscale items. To

provide a more exhaustive linguistic analysis, we included alternate forms of each PANAS-X

item in our searches (e.g., for happy, we also included happily, happiness, happier, and happiest)

and excluded the word “blue”, due to its predominantly non-emotional and potentially political

78

use in this context. We also recorded a control variable that approximated each politician’s

overall wordiness, by recording the frequency with which each politician used the 10 most

common words in the English language (the, be, to, of, and, a, in, that, have, I). This step was

necessary because the CapitolWords API does not provide the total sum of all words used by a

particular politician or party. This is why Study 2 results are not presented as the percentage of

total words used by each politician. This wordiness index accounted for 2,336,370 total words, or

approximately 14% of all words appearing in the Congressional Record in 2013.

The primary linguistic analysis for Study 2 (described above) assessed emotional

language at the individual level. For our extended analysis of emotional word use over time, we

used the CapitolWords API to collect word counts at the party-level using the same lists of

positive and negative emotion words described above. For each year that complete

Congressional Record data were available (1996-2013), separate word counts were collected for

the Democratic and Republican Parties. This extended analysis investigated party-level

emotional word use, rather than individual-level word use because of the prohibitive difficulty in

acquiring comprehensive voting record analysis, demographic data, and word count data for

every Congressperson during this 18-year period of time.

Smiling behavior. Photos of each politician were gathered from the Congressional

Pictorial Directory of the 113th Congress (United States Congress Joint Committee on Printing,

2013), an official, public domain publication containing color photos of each Congressperson.

Because politicians’ photographs for the Directory are updated irregularly, photos appearing in

this edition of the Directory were not necessarily taken during or immediately prior to the current

congressional term.

79

Smiling behavior was analyzed using the Facial Action Coding System (FACS) (Ekman

& Friesen, 1978) by a FACS-certified coder. FACS is a reliable, anatomically based coding

system that identifies facial muscle movements, termed action units (AUs), relevant to the

expression of emotion. A certified FACS coder who was blind to the hypothesis and political

affiliations of the participants analyzed the occurrence and intensity of two AUs associated with

smiling behavior: activity in the orbicularis oculi (AU6) and the zygomatic major (AU12).

Action units were coded using separate 6-point intensity scales, with higher scores indicating

more intense facial activation, including a zero-point to indicate no activation. To test reliability,

a second coder analyzed a random subset (20%) of the photos. The intraclass correlation

coefficient indicated acceptable inter-rater reliability for AU6 (ICC = .70) and AU12 (ICC =

.79).

Study 3

Because Twitter does not provide publicly available political ideology information for its

users, we used the service’s API to identify 6,000 participants who subscribed to (“followed”)

the official Twitter pages for the Democratic Party (n = 3,000) and the Republican Party (n =

3,000). We excluded 1,794 users who either had not posted status updates or whose status

updates were protected by privacy settings, plus an additional 260 users who appeared in both

lists. This resulted in a total of 1,950 exclusive followers of the Democratic Party and 1,996

exclusive followers of the Republican Party who posted publicly viewable status updates. We

analyzed the 20 most recent status updates from each user (or all available statuses from users

with fewer than 20 statuses), excluding “retweets”, or shared content that was originally written

by other users. In total, we analyzed 47,257 public status updates (M = 11.98 per user) containing

585,614 words (M = 12.39 per status update). Data were collected in May 2014. Approximately

80

89% of the collected updates were published in January 2014 through May 2014. The remaining

status updates dated as far back as 2008.

We assessed the emotional content of each tweet using the PANAS-X word lists

described in Study 2, in addition to the positive and negative emotion dictionaries provided by

the LIWC software. The LIWC word lists are longer and more comprehensive than the PANAS-

X lists, were designed specifically for use in linguistic analyses, and have been validated

multiple times (Pennebaker et al., 2001). We also created separate indices of the frequency with

which “happy” and “sad” emoticons appeared in the status updates [happy: :) :-) : ); sad: :( :-( : (

]. PANAS-X and LIWC analyses were limited to status updates written in English, as assessed

by the R software’s Compact Language Detection in R (CLDR) package, and emoticon usage

was assessed across all languages.

Study 4

Participants were 504 users of LinkedIn, a business-oriented social networking website.

LinkedIn profile photos are typically clear, professional quality photographs of individuals who

are looking straight into the camera and smiling. We selected participants who publicly self-

identified as employees at four organizations associated with ideologically liberal values (the

Democratic National Committee, MSNBC, the New York Times, Planned Parenthood) and four

organizations associated with ideologically conservative values (the Republican National

Committee, Fox News Channel, the Wall Street Journal, and the Family Research Council). We

analyzed the publicly available profile photos from up to 100 employees listed for each

employer, resulting in 267 photos of liberal organization employees and 237 photos of

conservative organization employees. Of these, we excluded 47 photos that were of low

resolution or too small to be seen clearly, that contained multiple people, or that did not clearly

81

show participants’ faces, resulting in a total of 457 photos (240 liberals, 217 conservatives).

Detailed demographic information was not available, but coders estimated that 57% of

participants were female and 77% were White. As in Study 2, a FACS-certified coder who was

blind to the hypothesis and ideology of participants analyzed the smiling behavior in each

photograph along AUs 6 and 12. A second coder analyzed a random subset (20%) of the photos

to test reliability, achieving acceptable inter-rater reliability for both AU6 (ICC = .81) and AU12

(ICC = .89).

82

CHAPTER 5:

Epilogue

83

Epilogue

The research presented in this dissertation contributes to our understanding of the

relationship between self-enhancement and subjective well-being (SWB), as well as the

relationship between political ideology and happiness. Although past research has already

established an association between self-enhancement and SWB (e.g., Taylor & Brown, 1988),

the present research finds that the nature of this relationship is more nuanced than previously

conceived. Three studies presented in Chapter 2 demonstrated that, just like judgments of other

valued traits and abilities, judgments of happiness and life satisfaction are susceptible to self-

enhancing distortions. By both measuring and experimentally manipulating self-enhancement,

Chapter 2 clarified the nature of this relationship. Specifically, it showed that self-enhancing

individuals tend to report high levels of happiness only when happiness is perceived as a

desirable quality. Chapter 3 explored group-level differences in the tendency to make

unrealistically positive self-assessments, finding that multiple dimensions of political

conservatism predicted the tendency to self-enhance across eight different measures of self-

enhancement. Chapter 4 found that these ideological difference in self-enhancement explained

why conservatives reported greater SWB than liberals. Further, three additional studies in

Chapter 4 found that the long-studied “ideological happiness gap” between liberals and

conservatives was limited to self-report measures, and that when assessing happiness-related

behaviors, it was liberals, rather than conservatives, who displayed greater happiness.

The studies presented in this dissertation did not employ nationally representative data

sets, and most of the studies were conducted with American Internet users in one form or

another. It will be important to replicate these findings with broader, more representative

samples to determine the extent to which these results are generalizable. However, the results of

84

these studies still have broad implications for a number of research areas. Most obviously, these

results demonstrate that comparisons of self-reported happiness between non-randomized groups

may inadvertently capture differences in self-enhancement rather than differences in emotional

experience. This risk is likely to be particularly pronounced when interpreting “happiness gaps”

that are relatively small, because the effect of self-enhancement on judgments of SWB is

relatively small itself. Still, there is no reason to believe that the challenges outlined here are

limited to comparisons of liberals and conservatives. Many variables of interest to SWB

researchers are comparably small in magnitude (e.g., gender: r = .04, employment status: r = .11,

income: r = .17; Okun, Stock, Haring, & Witter, 1984). An intriguing question for future

research will be to determine whether or not these and other happiness gaps might be partially or

fully explained by group differences in self-enhancement.

As discussed briefly in Chapters 2 and 4, this possibility raises important concerns for the

growing “gross national happiness” (GNH) movement, which aims to inform public policy

decision making using happiness research (Brooks, 2008; Diener, 2000; Diener, 2006). A

number of meaningful social groups that are important to GNH research differ in the degrees to

which they engage in self-enhancement (e.g., religious groups, cultures, nations; Sedikides &

Gebauer, 2009; Veenhoven & Ehrhardt, 1995; Loughnan et al., 2011), and thus it is possible that

happiness differences observed between some of these groups might be limited to self-report

measures, and might be less reflective of emotional behavior and/or experience.

Further, self-enhancement is likely to not only confound between-group comparisons of

happiness, but also within-group comparisons. Advocates of the GNH movement have proposed

that public policies should be evaluated, in part, based on the changes in happiness reports that

accompany them. It is important to note, however, that negative, psychologically threatening

85

stimuli have the potential to heighten self-enhancement motivation (e.g., McGregor, 2006;

Campbell & Sedikides, 1999; VanDellen, Campbell, Hoyle, & Bradfield, 2011). As shown in

Chapter 2, situationally heightened self-enhancement motivation can lead to inflated reports of

SWB. Policies that induce feelings of psychological threat may ironically result in compensatory

reports of happiness among those affected negatively by policy change. For example, it is

conceivable that individuals who benefit from a welfare program could become motivated to

defensively report greater happiness after their access to it is removed, even if this policy change

were to reduce quality of life and happiness. Although this example is speculative in nature, it

highlights the potential for negative, psychologically threatening policies to produce ironic

effects in self-reported happiness due to compensatory self-enhancement motivation.

To address the methodological challenges of measuring happiness with minimal risk of

self-enhancing distortions, future research should continue to examine unobtrusive behavioral

measures of happiness. Other alternatives to traditional self-report methods also warrant further

study, including forced-choice measures that force participants to choose between equally

desirable self-assessments (Diener et al.,1991; Paulhus, 1991), as well as implicit and indirect

measures (e.g., Kim, 2004; Walker & Schimmack, 2008). Researchers may also benefit from

using standard self-report measures conjointly with measures of self-enhancement motivation in

order to identify and control for defensive self-reporting styles. Of course, no single

methodology is perfect or comprehensive, and as mentioned in Chapters 2 and 4, the most

convincing research on the causes and consequences of happiness will offer converging evidence

across multiple methodologies.

In addition to these public policy and methodological implications, the present research

also advances the field’s theoretical understanding of the relationship between self-enhancement

86

and well-being. Past research on this relationship has centered on whether or not positive

illusions are truly predictive of well-being (Taylor & Brown, 1988; Colvin, Block, & Funder,

1995). The present research fundamentally reconceptualizes this relationship, such that greater

reports of well-being are not viewed as a beneficial downstream consequence of self-

enhancement, but rather an example of self-enhancement. Consistent support for this

interpretation was found throughout this dissertation, though the present research does not

disprove or fully discount the potential adaptiveness of self-enhancement (e.g., Taylor et al.,

2003a;b). Instead, this research raises theoretical questions about the relative extent to which

self-enhancement genuinely improves well-being, compared to the extent to which it

superficially inflates reports of well-being. This remains an important theoretical question for

future research.

The supposition that self-enhanced reports of happiness are likely to be somehow

different from “genuine” experiences of happiness provides a somewhat novel perspective for

studying SWB, but it borrows heavily from the literature on self-esteem. Dating as far back as

the 1970s, self-esteem researchers have observed a positive relationship between self-

enhancement and high self-esteem (Schneider & Turkat, 1975). But unlike in similar research on

happiness, self-esteem researchers characterized high self-esteem that was accompanied by self-

enhancement as defensive in nature. In the following decades, researchers confirmed that not all

reports of high self-esteem were equally valuable or beneficial, and they have since distinguished

between a multitude of manifestations of high self-esteem, including defensive/genuine self-

esteem (Schneider & Turkat, 1975), secure/fragile self-esteem (Kernis, 2003), and

implicit/explicit self-esteem (Jordan, Spencer, Zanna, Hoshino-Browne, & Correll, 2003).

87

Optimal self-esteem, it has been argued, is not invariably high (Kernis, 2003), and it seems

plausible that the same is true for “optimal happiness” (Gruber, Mauss, & Tamir, 2011).

The research presented in this dissertation highlights the potential advantages of

conceptualizing happiness similarly to self-esteem. This dissertation has shown that reports of

happiness appear to be artificially inflated by self-enhancement (Chapter 2) and that self-

enhancing groups show less happiness-related behavior than others (Chapter 4). It appears likely,

based on these findings, that reports of happiness bolstered by self-enhancement are qualitatively

different than un-enhanced reports of well-being. SWB researchers frequently define SWB as a

multifaceted construct consisting of life satisfaction, domain satisfaction, and affective balance

(Diener, 2000), or as a combination of hedonic and eudaimonic forms of well-being (Deci &

Ryan, 2008). However, a further level of complexity in understanding happiness is the

implication that not all manifestations of high happiness are equivalent. Although the present

research has aimed to be value neutral by avoiding value-laden labeling of different

presentational styles of happiness (e.g., “defensive” or “insecure” happiness), future research

will likely improve our understanding of the adaptiveness of these manifestations of happiness

by borrowing the methodological insights, and perhaps even the nomenclature, of the self-esteem

literature.

Like self-esteem, happiness does not appear to be a single, unidimensional construct.

Throughout this research, happiness has revealed itself to be a surprisingly nebulous,

multifaceted, and sometimes self-contradictory phenomenon. One of the most interesting

research questions raised by this research also emerged as an unexpectedly philosophical

question: is happiness truly deserving of Aristotle’s portrayal as a “supreme good”—i.e., as the

ultimate desired end to all means? Phrased more scientifically: is happiness, with all of its

88

potential manifestations—genuine, defensive, deceptive, self-reported, behavioral, and so on—

invariably associated with beneficial outcomes? Despite its philosophical origins, this is an

empirical question worthy of further scientific research. A more nuanced conception of what it

means to be happy will allow us to better understand which forms of happiness are optimal for

human flourishing, and perhaps just as importantly—which forms are not.

89

References

Brooks, A. C. (2008). Gross national happiness: Why happiness matters for America--and how

we can get more of it. New York: Basic Books.

Campbell, W. K., & Sedikides, C. (1999). Self-threat magnifies the self-serving bias: A meta-

analytic integration. Review of General Psychology, 3, 23-43.

Colvin, C. R., Block, J., & Funder, D. C. (1995). Overly positive self-evaluations and

personality: Negative implications for mental health. Journal of Personality and Social

Psychology, 68, 1152-1162.

Deci, E. L., & Ryan, R. M. (2008). Hedonia, eudaimonia, and well-being: An

introduction. Journal of Happiness Studies, 9, 1-11.

Diener, E. (2000). Subjective well-being: The science of happiness and a proposal for a national

index. American Psychologist, 55, 34-43.

Diener, E. (2006). Guidelines for national indicators of subjective well-being and ill-being.

Applied Research in Quality of Life, 1, 151-157.

Diener, E., Sandvik, E., Pavot, W., & Gallagher, D. (1991). Response artifacts in the

measurement of subjective well-being. Social Indicators Research, 24, 35-56.

Gruber, J., Mauss, I. B., & Tamir, M. (2011). A dark side of happiness? How, when, and why

happiness is not always good. Perspectives on Psychological Science, 6, 222-233.

Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino-Browne, E., & Correll, J. (2003). Secure and

defensive high self-esteem. Journal of Personality and Social Psychology, 85, 969-978.

Kernis, M. H. (2003). Toward a conceptualization of optimal self-esteem. Psychological Inquiry,

14, 1-26.

90

Kim, D. Y. (2004). The implicit life satisfaction measure. Asian Journal of Social Psychology, 7,

236–262.

Loughnan, S., Kuppens, P., Allik, J., Balazs, K., De Lemus, S., Dumont, K., et al. (2011).

Economic inequality is linked to biased self-perception. Psychological Science, 22, 1254-

1258.

McGregor, I. (2006). Offensive defensiveness: Toward an integrative neuroscience of

compensatory zeal after mortality salience, personal uncertainty, and other poignant self-

threats. Psychological Inquiry, 17, 299-308.

Okun, M. A., Stock, W. A., Haring, M. J., & Witter, R. A. (1984). Health and subjective well-

being: A meta-analysis. The International Journal of Aging and Human Development, 19,

111-132.

Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R.

Shaver, & L. S. Wrightsman (Eds.), Measures of personality and social psychological

attitudes (pp.17-59). San Diego: Academic Press.

Schneider, D. J., & Turkat, D. (1975). Self-presentation following success or failure: Defensive

self-esteem models. Journal of Personality, 43, 127-135.

Sedikides, C., & Gebauer, J. E. (2009). Religiosity as self-enhancement: A meta-analysis of the

relation between socially desirable responding and religiosity. Personality and Social

Psychology Review, 14, 17-36.

Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective

on mental health. Psychological Bulletin, 103, 193-210.

91

Taylor, S. E., Lerner, J. S., Sherman, D. K., Sage, R. M., & McDowell, N. K. (2003a). Are self-

enhancing cognitions associated with healthy or unhealthy biological profiles?. Journal

of Personality and Social Psychology, 85, 605-615.

Taylor, S. E., Lerner, J. S., Sherman, D. K., Sage, R. M., & McDowell, N. K. (2003b). Portrait of

the self-enhancer: well adjusted and well liked or maladjusted and friendless? Journal of

Personality and Social Psychology, 84, 165-176.

Van Dellen, M. R., Campbell, W. K., Hoyle, R. H., & Bradfield, E. K. (2011). Compensating,

resisting, and breaking: A meta-analytic examination of reactions to self-esteem threat.

Personality and Social Psychology Review, 15, 51–74.

Veenhoven, R., & Ehrhardt, J. (1995). The cross-national pattern of happiness: Test of

predictions implied in three theories of happiness. Social Indicators Research, 34, 33-68.

Walker, S. S., & Schimmack, U. (2008). Validity of a happiness implicit association test as a

measure of subjective well-being. Journal of Research in Personality, 42, 490-497.