University of Southern California Dissertations and Theses

Information, public opinion, and international relations

(USC Thesis Other)

Information, public opinion, and international relations

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INFORMATION, PUBLIC OPINION AND INTERNATIONAL RELATIONS
by
Thomas Jamieson
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLITICAL SCIENCE AND INTERNATIONAL RELATIONS)
August 2018
Copyright 2018 Thomas Jamieson
Dedication
To my late grandparents,
Gweneth & Donald Johnston and Marjorie & Hugh Jamieson.
Thank you for your wisdom, kindness, and love.
You inspired me to pursue my passions,
and your love and encouragement made me believe it was possible.
I regret I finished too late for you to read this dissertation,
but your profound influence on me endures.
ii
Acknowledgments
The dissertation represents the culmination of an education that began with looking
over my patient older sister’s shoulder as my parents taught her to read.
Along the way, I benefited from support and encouragement from teachers at
ElstreeKindergarten, GlenholmePrimarySchool, OverseasFamilySchool, Southwell
School, andSt. Paul’s CollegiateSchool. Iwould especiallyliketothank theteachers
who inspired me, pushed me, and provided opportunities to extend my learning.
AtVictoriaUniversityofWellington, Iwasluckytohaveaparticularlysupportive
group of professors, without whom I would not have had the opportunity to pursue a
PhD in the first place. I would especially like to thank Roy Shuker, Peter Thompson,
and Doug Van Belle for challenging me and supporting me as I pursued postgraduate
study at home and abroad. I really cherish our friendship, and I am looking forward
to hopefully many more debates over a drink in the future!
Once at USC, I benefited from excellent professors who broadened my interests,
introduced me to some amazing scholarship, and were very patient as I grappled with
learning the literature and the methods of the discipline. Thank you Pablo Barberá,
Dennis Chong, Andrew Coe, Ben Graham, Ange-Marie Hancock-Alfaro, Pat James,
Lian Jian, Jane Junn, David Kang, Morris Levy, Gerry Munck, Sheila Murphy, Brian
iii
Rathbun, Jefferey Sellers, J. Ann Tickner, Nicholas Weller, and Carol Wise, I really
enjoyed learning from you all.
I would also like to thank the excellent staff in the School of International Rela-
tions, the Department of Political Science, and the Center for International Studies
at USC. Thank you Cathy Ballard, Ashley Bonanno-Curley, Cort Brinkerhoff, Made-
line Brown, Chandra Caldwell, Linda Cole, Gina Hakim, Yaminah McKessey, Indira
Persad, Aurora Ramirez, and Karen Tang for your tireless efforts behind the scenes
to make everything work, and for helping make USC such a great place to be.
Veri Chavarin deserves special praise as the graduate advisor of the Political
Science and International Relations (POIR) PhD program. Thank you for your
calm assistance with crisis management, funding, and for generally being the glue
binding the whole PhD program together!
I was also very fortunate to have a range of mentors at USC, including people
outside of my committee who I could rely on for support, encouragement, advice, and
frank conversations. Thank you Pablo Barberá, Christian Grose, Jane Junn, Saori
Katada, Morris Levy, James Lo, Gerry Munck, and Carol Wise for your support, I
really appreciate it.
MydissertationbenefitedfromfundingsupportfromthePOIRPhDprogram, the
SchoolofInternationalRelations,andtheDornsifeGoldFamilyGraduateFellowship.
Thank you very much for the funding. The funding gave me the resources to conduct
theexperiment, andthetimetowriteuptheresultswithouthavingtoseekadditional
funding to make rent.
I had a wonderful dissertation committee, and they deserve all of the praise for
what this dissertation has become, and none of the blame for any errors that remain.
iv
Dennis Chong’s insight challenged me to think bigger, to aim higher, and to ex-
pand my horizons. While I could not adopt all of his suggestions in this dissertation,
they form the basis for the book I hope to write over the next few years.
Phil Seib brought an incredible attention to detail to the project, and his sugges-
tions encouraged me to be more precise throughout the dissertation. His constructive
feedback has been invaluable not only for the dissertation, but several other projects
as well.
Nick Weller has been an incredible source of support over the past five years.
In my dissertation and in all my other projects too, Nick has been a constant font
of constructive criticism which has helped me advance many projects, and just as
importantly, to discard some of the less promising ones.
Thank you all for your critical feedback that led to improvements in my research,
and for providing excellent advice and support for my future beyond the PhD.
I cannot imagine a better Chair than Pat James. Over the past five years he has
provided consistent support, and given me excellent and prompt feedback on dozens
of projects. He gave me opportunities for me to develop research skills, publish my
research, and meet prominent members of the field. He has also been someone I
could always turn to for advice and encouragement when I needed it most. Thank
you for being an amazing adviser, mentor, and treasured friend.
My friends have made the past five years a wonderful experience, but they also
make leaving LA very difficult. From spontaneous drinks to the stammtisch, Ko-
rean BBQ to Patty’s tacos at York & Avenue 56 every week, I have loved hanging
out with you all. I would especially like to thank Therese Anders, Jiun Bang, Justin
Berry, Ashley Bonanno-Curley, Tyler Bonanno-Curley, Fabián Borges-Herrero, Cody
v
Brown, Victoria Chonn Ching, Youssef Chouhoud, Juve Cortés, Taylor Dalton, Nico-
las de Zamaroczy, Joe Gancayco, Gina Hakim, Whitney Hua, Joey Huddleston, Josh
Jacobson, Rebecca Kimitch, Brian Knafou, Jillian Medeiros Pérez, Matt Mendez,
Rosie Murphy, Paul Orner, Mark Paradis, Adam Perrotta, Simon Radford, Michael
Pérez, Jennifer Roglá, Justin Rowley, Sara Sadhwani, Güez Salinas, Cailyn Shin,
Keira Stearns, Mao Suzuki, Scott Wilbur, Phil Wilcox, and Ming-min Yang for be-
ing such great companions throughout my time at USC. I am excited to see you at
the next conference, wedding, or karaoke night!
I have been fortunate to have the continued support of friends from New Zealand.
IespeciallywouldliketothankNeilBadenhorst, JarredFoster, UmeshGovind, Andy
Nyce, Shaun Raumati, and Paul Richards. I truly cherish our friendship, and I am
excited to catch up with you soon, wherever in the world that may be!
Above all, my thanks begins and ends with family.
Over the past few years, Juve Cortés has become a brother to me. I will truly
miss hanging out with you all the time. También me gustaría agradecer a la familia
Cortés por ser mi segunda familia en Los Ángeles. Me encantó conocerlos durante los
últimos cinco años, y espero seguir en contacto sin importar dónde termine. Gracias
por toda su hospitalidad y amabilidad, tienen un lugar muy especial en mi corazón.
I would like to thank the Lunnon clan for making me feel so welcome, and for sup-
porting Jess and I through our unusual journey. Brenda, Mike, Kayla, Ali, Charlie,
Bianca, Danita, and Roy, thank you for all of your love and support.
I have been blessed with an amazing family, and it is no exaggeration to say I
could not have completed this without them. This dissertation is dedicated to the
vi
memory of my grandparents, all of whom were generous, selfless people who spent
their lives helping others, including myself.
My family’s unbending encouragement and love gave me the belief that I could
do it. Throughout my life, my parents have supported me through every challenge,
providing assistance wherever possible, and giving me every chance to achieve my
goals. Thank you Mum, Dad, Jemima, and Chris for all that you have done for me.
I will be forever grateful.
Most of all, I would like to thank my wife, Jess. You sacrificed so much to give me
a chance to live my dream. Through the countless texts, Snapchats, glitchy phone
calls, fleeting visits, the ecstatic hellos and the painful goodbyes, you have been my
inspiration, my rock, and my guide. I am excited to see what our future holds, and
I hope I can help you live your dreams too. I love you heaps.
vii
Contents
Dedication ii
Acknowledgments iii
List of Tables xiii
List of Figures xvii
Abstract xx
1 Introduction 1
1.1 Preview of Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2 Information, Public Opinion, and Foreign Policy 9
2.1 Public Opinion about Foreign Policy . . . . . . . . . . . . . . . . . . 11
2.2 How People Receive Information about International Affairs . . . . . 16
2.2.1 The Supply of Information about International Affairs . . . . 17
2.2.2 The Demand for Information about International Affairs . . . 25
2.3 Towards an Understanding of Information, Public Opinion, and For-
eign Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.3.1 Attention, Information, and Political Knowledge . . . . . . . . 31
2.3.2 Attention, Information, and Political Predictions . . . . . . . . 32
2.3.3 Attention, Information, and Foreign Policy Attitudes . . . . . 33
2.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3 A Theory of Incentives, Information, and Public Opinion about
Foreign Policy 36
3.1 Theoretical Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . 37
viii
3.2 Incentives, Information, and Public Opinion about Foreign Policy . . 38
3.3 Implications of the Model . . . . . . . . . . . . . . . . . . . . . . . . 45
3.3.1 Incentives and Information Search . . . . . . . . . . . . . . . . 45
3.3.2 Incentives and Political Knowledge . . . . . . . . . . . . . . . 49
3.3.3 Incentives and Political Predictions . . . . . . . . . . . . . . . 51
3.3.4 Incentives and Foreign Policy Attitudes . . . . . . . . . . . . . 53
3.4 Definition of Key Constructs . . . . . . . . . . . . . . . . . . . . . . . 56
3.4.1 Incentives for Accuracy . . . . . . . . . . . . . . . . . . . . . . 57
3.4.2 Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.4.3 Information Search and Information Avoidance . . . . . . . . . 60
3.4.4 Attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4.5 Political Knowledge . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4.6 Political Predictions . . . . . . . . . . . . . . . . . . . . . . . 62
3.4.7 Political Attitudes . . . . . . . . . . . . . . . . . . . . . . . . 62
3.5 Operationalization of Variables . . . . . . . . . . . . . . . . . . . . . 63
3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4 Motivated Learning and Public Opinion about Foreign Policy 67
4.1 Experimental Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
4.1.1 Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.1.2 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2.1 Incentives and Information Search . . . . . . . . . . . . . . . . 76
4.2.2 Incentives and Political Knowledge . . . . . . . . . . . . . . . 78
4.2.3 Incentives and Political Predictions . . . . . . . . . . . . . . . 83
4.2.4 Incentives and Foreign Policy Attitudes . . . . . . . . . . . . . 87
4.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
5 The Microlevel Foundations of Attention to International Affairs 95
5.1 The Conscientious Voter . . . . . . . . . . . . . . . . . . . . . . . . . 97
5.2 Data and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.1 Case Selection: U.S. Public Opinion about Military Action in
Syria, 2013 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.2.3 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.2.4 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
5.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
ix
6 Perceptions of Impact and Public Opinion about Foreign Policy 111
6.1 Perceptions of Impact and Public Opinion about Foreign Policy . . . 112
6.2 Data and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.2.1 Case Selection: U.K. Public Opinion about the E.U. Referen-
dum, 2016 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
6.2.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.2.3 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
6.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
6.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
7 Crisis, Information Search, and Public Opinion about Foreign Pol-
icy 134
7.1 Crisis, Information Search, and Public Opinion about Foreign Policy . 135
7.2 Empirical Strategy of the Plausibility Probe . . . . . . . . . . . . . . 138
7.2.1 Observable Implications of Theory . . . . . . . . . . . . . . . 140
7.3 U.S. Public Opinion about Military Action in Iraq, 2002-2008 . . . . 140
7.3.1 Case Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.3.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
7.3.3 Evidence from Public Opinion Polls . . . . . . . . . . . . . . . 144
7.4 U.S. Public Opinion about Military Action in Syria, 2012-2017 . . . . 149
7.4.1 Case Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
7.4.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
7.4.3 Evidence from Public Opinion Polls . . . . . . . . . . . . . . . 153
7.5 U.S. Public Opinion about Military Action against ISIS, 2013-2016 . 158
7.5.1 Case Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
7.5.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
7.5.3 Evidence from Public Opinion Polls . . . . . . . . . . . . . . . 162
7.6 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
8 Conclusion 169
8.1 Summary of Arguments and Evidence . . . . . . . . . . . . . . . . . . 170
8.2 Policy Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
8.3 Future Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
8.3.1 Incentives for Accuracy . . . . . . . . . . . . . . . . . . . . . . 174
8.3.2 Attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
8.3.3 Information Search . . . . . . . . . . . . . . . . . . . . . . . . 176
8.3.4 Information and Social Media . . . . . . . . . . . . . . . . . . 177
8.3.5 The Public as Democratic Constraint . . . . . . . . . . . . . . 178
x
8.3.6 Beyond Foreign Policy: Incentives, Information, and Public
Opinion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
8.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
A The Full Experimental Design 181
A.1 Demographics 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
A.2 Experimental Treatments . . . . . . . . . . . . . . . . . . . . . . . . . 184
A.2.1 Common introduction . . . . . . . . . . . . . . . . . . . . . . 184
A.2.2 Control Condition . . . . . . . . . . . . . . . . . . . . . . . . . 185
A.2.3 Bonus Treatment . . . . . . . . . . . . . . . . . . . . . . . . . 185
A.2.4 Random Bonus Treatment . . . . . . . . . . . . . . . . . . . . 185
A.2.5 Lottery Treatment . . . . . . . . . . . . . . . . . . . . . . . . 186
A.3 Political Knowledge Questions . . . . . . . . . . . . . . . . . . . . . . 187
A.4 Prediction Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
A.5 Information Search Questions . . . . . . . . . . . . . . . . . . . . . . 193
A.6 Answers to Political Knowledge Questions (Randomly presented to
participants before/after political attitudes questions) . . . . . . . . . 196
A.7 Political Attitudes Questions . . . . . . . . . . . . . . . . . . . . . . . 202
A.8 Answers to Prediction Questions . . . . . . . . . . . . . . . . . . . . . 204
A.9 Demographics 2 (Randomly assigned to participants before/after De-
mographics 3) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
A.10 Demographics3(Randomlyassignedtoparticipantsbefore/aftertreat-
ments) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206
B Appendix for Chapter 4 209
B.1 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
B.2 Full Experimental Results on Political Knowledge using OLS Regression214
B.3 Full Experimental Results on Political Knowledge using Logistic Re-
gression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
B.4 Full Experimental Results on Political Predictions using OLS Regression235
B.5 Full Experimental Results on Political Predictions using Logistic Re-
gression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245
B.6 The Determinants of Time Spent on the Survey . . . . . . . . . . . . 254
B.7 Randomization Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
B.8 Causal Mediation Sensitivity Analysis . . . . . . . . . . . . . . . . . . 259
B.9 Expected Foreign Policy Attitudes in Experiment . . . . . . . . . . . 265
xi
C Appendix for Chapter 5 267
C.1 Descriptive Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 268
C.2 Polls about U.S. Military Action in Syria . . . . . . . . . . . . . . . . 271
D Appendix for Chapter 6 272
D.1 Codebook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
D.1.1 Dependent Variables: Vote Intention . . . . . . . . . . . . . . 272
D.1.2 Independent Variables . . . . . . . . . . . . . . . . . . . . . . 272
D.1.3 Political Control Variables . . . . . . . . . . . . . . . . . . . . 273
D.1.4 Risk and Uncertainty Control Variables . . . . . . . . . . . . . 276
D.1.5 Region Control Variables . . . . . . . . . . . . . . . . . . . . . 276
D.1.6 Demographic Control Variables . . . . . . . . . . . . . . . . . 277
D.2 Descriptive Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
D.3 Full Results using Logistic Regression . . . . . . . . . . . . . . . . . . 282
D.4 Alternative Specification of the Model using Probit Regression . . . . 284
D.5 Full Results with Alternative Coding of the Dependent Variable . . . 286
E Appendix for Chapter 7 288
E.1 Polls about U.S. Military Action in Syria . . . . . . . . . . . . . . . . 288
E.2 Polls about U.S. Airstrikes against Islamic Militants . . . . . . . . . . 300
E.3 Polls about U.S. Troops to Fight Islamic Militants . . . . . . . . . . . 307
Bibliography 319
xii
List of Tables
3.1 Definition of Key Constructs. . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Operationalization of Constructs across Experimental and Observa-
tional Studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.1 Topics and Statement Wording for Foreign Policy Attitude Questions. 74
4.2 Direct Treatment Effects and Mediation Effects of Incentives on the
Number of Correct Answers for Individual Political Knowledge Ques-
tions. Mediating Variable: Information Search. . . . . . . . . . . . . . 80
4.3 Direct Treatment Effects and Mediation Effects of Incentives on the
NumberofCorrectAnswersforPredictionQuestions. Mediating Vari-
able: Information Search. . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.4 Attitudes about Foreign Policies related to Security. . . . . . . . . . . 89
4.5 Attitudes about Foreign Policies not related to Security. . . . . . . . . 90
4.6 Direct and Mediated Effects of the Bonus Treatment on Political At-
titudes (Mediating Variable: Political Knowledge). . . . . . . . . . . . 91
4.7 Direct and Mediated Effects of the Random Bonus Treatment on Po-
litical Attitudes (Mediating Variable: Political Knowledge). . . . . . . 91
xiii
4.8 Direct and Mediated Effects of the Lottery Treatment on Political
Attitudes (Mediating Variable: Political Knowledge). . . . . . . . . . 91
5.1 Determinants of Attention to the Syria News Story in 2013. . . . . . 108
7.1 Operationalization of Constructs. . . . . . . . . . . . . . . . . . . . . 139
7.2 Crisis, Debate, and Events in the U.S. Military Action in Iraq. . . . . 142
7.3 Crisis, Debate, and U.S. Military Action in Syria. . . . . . . . . . . . 150
7.4 Crisis, Debate, and U.S. Military Action against ISIS. . . . . . . . . . 160
B.1 Descriptive Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
B.2 Descriptive Statistics of Individual Political Knowledge Questions. . . 211
B.3 Descriptive Statistics of Individual Prediction Questions. . . . . . . . 212
B.4 Descriptive Statistics of Individual Foreign Policy Attitudes. . . . . . 213
B.5 The Determinants of Information Search. . . . . . . . . . . . . . . . . 215
B.6 The Determinants of Political Knowledge. . . . . . . . . . . . . . . . 216
B.7 Correct No. Deaths from Jihadist Terrorism. . . . . . . . . . . . . . . 217
B.8 Correct No. Ratifications of Paris Agreement. . . . . . . . . . . . . . 218
B.9 Correct No. Unauthorized Immigrants. . . . . . . . . . . . . . . . . . 219
B.10 Correct No. Ukraine Deaths. . . . . . . . . . . . . . . . . . . . . . . . 220
B.11 Correct Special Counsel. . . . . . . . . . . . . . . . . . . . . . . . . . 221
B.12 Correct International Support for Iran Deal. . . . . . . . . . . . . . . 222
B.13 Correct N. Korea Missile Range. . . . . . . . . . . . . . . . . . . . . . 223
B.14 Correct U.S.-China Deficit. . . . . . . . . . . . . . . . . . . . . . . . . 224
B.15 The Determinants of Information Search. . . . . . . . . . . . . . . . . 226
B.16 Correct No. Deaths from Jihadist Terrorism. . . . . . . . . . . . . . . 227
xiv
B.17 Correct No. Ratifications of Paris Agreement. . . . . . . . . . . . . . 228
B.18 Correct No. Unauthorized Immigrants. . . . . . . . . . . . . . . . . . 229
B.19 Correct No. Ukraine Deaths. . . . . . . . . . . . . . . . . . . . . . . . 230
B.20 Correct Special Counsel. . . . . . . . . . . . . . . . . . . . . . . . . . 231
B.21 Correct International Support for Iran Deal. . . . . . . . . . . . . . . 232
B.22 Correct N. Korea Missile Range. . . . . . . . . . . . . . . . . . . . . . 233
B.23 Correct U.S.-China Deficit. . . . . . . . . . . . . . . . . . . . . . . . . 234
B.24 Political Predictions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
B.25 Correct Prediction of Laptop Ban on Domestic Flights. . . . . . . . . 237
B.26 Correct Prediction of Arctic Sea Ice Extent (Area) vs. 2016. . . . . . 238
B.27 Correct Prediction of Border Wall Construction. . . . . . . . . . . . . 239
B.28 Correct Prediction of Lifting of Russia Sanctions. . . . . . . . . . . . 240
B.29 Correct Prediction of Impeachment of Trump. . . . . . . . . . . . . . 241
B.30 Correct Prediction of U.S. Sanctions on Iran for Nuclear Deal. . . . . 242
B.31 Correct Prediction of Lethal Confrontation with N. Korea. . . . . . . 243
B.32 Correct Prediction of U.S. Sanctions on China for Trade. . . . . . . . 244
B.33 Correct Prediction of Laptop Ban on Domestic Flights. . . . . . . . . 246
B.34 Correct Prediction of Arctic Sea Ice Extent (Area) vs. 2016. . . . . . 247
B.35 Correct Prediction of Border Wall Construction. . . . . . . . . . . . . 248
B.36 Correct Prediction of Lift Russia Sanctions. . . . . . . . . . . . . . . 249
B.37 Correct Prediction of Impeachment of Trump. . . . . . . . . . . . . . 250
B.38 Correct Prediction of U.S. Sanctions on Iran for Nuclear Deal. . . . . 251
B.39 Correct Prediction of Lethal Confrontation with N. Korea. . . . . . . 252
B.40 Correct Prediction of U.S. Sanctions on China for Trade. . . . . . . . 253
xv
B.41 The Determinants of Time Spent on the Survey (Seconds). . . . . . . 255
B.42 Randomization Check: F-test of Joint Orthogonality. . . . . . . . . . 257
B.43 Randomization Check: T-tests of the Differences of Means between
Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
B.44 Expected and Actual Foreign Policy Attitudes Based on Party ID. . . 265
B.45 Expected Foreign Policy Attitudes Based on Political Ideology. . . . . 266
C.1 Descriptive Statistics for Variables in the Pew Poll, April 25-28, 2013. 268
C.2 DescriptiveStatisticsforVariablesinthePewPoll,August28-September
1, 2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
C.3 Descriptive Statistics for Variables in the Pew Poll, September 12-15,
2013. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
D.1 Descriptive Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
D.2 Logit Regression: The Determinants of Voting to Leave the European
Union. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
D.3 Logit Regression: The Determinants of Voting to Leave the European
Union. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
D.4 ProbitRegression: TheDeterminantsofVotingtoLeavetheEuropean
Union. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
D.5 OLS Regression: The Determinants of Voting to Leave the European
Union, including Don’t Know and Nonresponses. . . . . . . . . . . . . 287
xvi
List of Figures
4.1 The Experimental Design. . . . . . . . . . . . . . . . . . . . . . . . . 72
4.2 Determinants of Information Search. . . . . . . . . . . . . . . . . . . 76
4.3 Determinants of Political Knowledge. Dependent Variable: Number
of Correct Answers for Political Knowledge Questions. . . . . . . . . . 79
4.4 Determinants of Individual Political Knowledge Responses. Depen-
dent Variable: Correct Answers for Individual Political Knowledge
Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.5 Determinants of Correct Predictions. Dependent Variable: Number of
Correct Answers for Prediction Questions. . . . . . . . . . . . . . . . 85
5.1 Attention to the Syria News Story Over Time. . . . . . . . . . . . . . 101
5.2 The Determinants of Attention to the Syria News Story. . . . . . . . 107
6.1 Voter Turnout in UK Parliament General Elections and the E.U. Ref-
erendum, 1979-2017. . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.2 Determinants of Voting to Leave the E.U. Dependent Variable: The
Predicted Probability of Voting to Leave the European Union. . . . . . 126
xvii
6.3 The Marginal Effects of Perceptions of Impact and Attention on Vot-
ing to Leave the E.U. Dependent Variable: The Predicted Probability
of Voting to Leave the European Union. . . . . . . . . . . . . . . . . . 130
7.1 U.S. Public Support for the Iraq War, 2002-2008 . . . . . . . . . . . . 146
7.2 U.S. Public Support for the Iraq War by Party, 2002-2008. . . . . . . 147
7.3 Support for U.S. Military Action in Syria, 2012-2017. . . . . . . . . . 155
7.4 Support for U.S. Military Action in Syria by Party, 2012-2017. . . . . 156
7.5 Public Support for U.S. Airstrikes against ISIS, 2013-2016. . . . . . . 164
7.6 Public Support for the Deployment of U.S. Troops to Fight against
ISIS, 2013-2016. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
B.1 Average Causal Mediation Effect of the Bonus Treatment on Political
Knowledge as a Function of the Degree of Violation of Sequential
Ignorability Assumption. . . . . . . . . . . . . . . . . . . . . . . . . . 259
B.2 Average Causal Mediation Effect of the Random Bonus Treatment
on Political Knowledge as a Function of the Degree of Violation of
Sequential Ignorability Assumption. . . . . . . . . . . . . . . . . . . . 260
B.3 Average Causal Mediation Effect of the Lottery Treatment on Political
Knowledge as a Function of the Degree of Violation of Sequential
Ignorability Assumption. . . . . . . . . . . . . . . . . . . . . . . . . . 261
B.4 Average Causal Mediation Effect of the Bonus Treatment on the Num-
ber of Correct Predictions as a Function of the Degree of Violation of
Sequential Ignorability Assumption. . . . . . . . . . . . . . . . . . . . 262
xviii
B.5 Average Causal Mediation Effect of the Random Bonus Treatment on
the Number of Correct Predictions as a Function of the Degree of
Violation of Sequential Ignorability Assumption. . . . . . . . . . . . . 263
B.6 Average Causal Mediation Effect of the Lottery Treatment on the
NumberofCorrectPredictionsasaFunctionoftheDegreeofViolation
of Sequential Ignorability Assumption. . . . . . . . . . . . . . . . . . 264
xix
Abstract
This dissertation contributes towards a more complete understanding of informa-
tion, public opinion, and international relations. I demonstrate that scholars must
understand the conditions under which people perceive incentives for accuracy to
understand how people learn, think, feel, and behave regarding international affairs.
Through a combination of an incentivized online survey experiment and analysis of
public opinion polls, I find that people’s attitudes and behavior are contingent on
their perception of incentives for accuracy. If people consider the stakes are high and
it is important to make the correct decision, they are motivated to: 1) search for
information to help them make the correct decision; 2) become more knowledgeable
about the foreign policy debate; and 3) update their attitudes about the foreign
policy debate. The dissertation has important implications for the study of public
opinion and international relations, suggesting that people are capable of effective
decision making when it matters most.
xx
Chapter 1
Introduction
A large body of international relations research examines how people make sense of
international affairs. Yet, we know surprisingly little about how people think about
the decisions that matter the most.
First, scholarship focuses on how people think about decisions of huge conse-
quence, including the use of nuclear weapons, debates over the use of force, entry
into transformational trade agreements, or the punishment of leaders for their deci-
sions. Yet, scholarship relies on data from surveys and experiments where the stakes
for participants could not be any lower. In effect, survey responses are essentially
“cheap talk,” but previous scholarship mistakenly uses them as the basis for the
entire understanding of public opinion about foreign policy.
Second, the understanding of public opinion about foreign policy is built on stud-
ies with low information. However, if information was once scarce, it is now abun-
dant. People form judgments about international affairs, and politics more broadly,
with more information at their disposal than they could ever hope to consume. Yet,
1
public opinion scholarship relies on studies with low information environments that
are drastically removed from the information environment individuals face in the real
world.
Third, public opinion scholarship assumes that people’s attention to international
affairs is stable over time. Studies rely on inferences from surveys and experiments
that do not consider the conditions under which people pay attention to international
affairs, how attention influences people’s attitudes about foreign policy, and how
attention influences the public’s willingness to constrain foreign policymakers.
In short, previous scholarship has worked towards developing an understanding
of public opinion about foreign policy. However, the questions asked, the methods
used, and the answers found in previous studies indicate a failure to satisfactorily
consider the context in which decisions are made. As such, our understanding of
public opinion about foreign policy is incomplete.
In this dissertation, I aim to contribute towards a more complete understanding
of information, public opinion, and international relations. I argue that scholarship
focused on public opinion about foreign policy must shift its gaze towards the con-
ditions under which people perceive incentives for accuracy, and examine how this
influences their demand for information to understand how people learn, think, feel,
and behave regarding international affairs.
1.1 Preview of Argument
My argument is simple and intuitive, but it constitutes a significant departure from
existing scholarship in International Relations.
2
First, I argue that to understand how the public forms attitudes about foreign
policy, it is vital to consider the context in which they are asked to make decisions
about foreign policy. Prior scholarship does an excellent job in explaining public
opinion about foreign policy when the stakes are low, and when the public feels
little pressure to make correct decisions. However, this is at odds with the decisions
we most care about in international relations – decisions where the stakes are high
such as the decision to enter conflict, use nuclear weapons, or enter international
agreements.
I argue that people think and behave differently when considering international
affairsdependingontheirperceptionofincentivesforaccuracy. Ifpeopleconsiderthe
stakes are high and it is important to make the correct decision, they are motivated
to 1) search for information to help them make the right decision; 2) become more
knowledgeable about the foreign policy debate; and 3) update their attitudes based
on motivated learning about the foreign policy debate.
Differences in perception of incentives for accuracy can occur between individuals
can emerge on the basis of their perception of the impact of their decisions, or their
innate sense of civic duty. Differences in perception of incentives for accuracy can
also occur within individuals over time. In particular, the onset of crisis and a period
of public debate about foreign policy can lead to reassessments of the consequences
of decisions based on individuals’ reassessment of the context in which they make
decisions about foreign policy.
3
Second, I argue that information is more available than ever before, and people
can cheaply and quickly find whatever information they seek, with very few excep-
tions.
1
As a result, I argue that it is critical to understand the conditions under
which people demand information, instead of focusing exclusively on the supply of
information.
I argue that people search for information when they perceive there are incentives
for accuracy, when it is important to make correct decisions about foreign policy.
During these periods, people dramatically change their information consumption
behavior. Because people are motivated to learn about international affairs as they
search for information, they use this information to learn about international affairs,
and they update their attitudes about foreign policy.
Third, I argue that attention to international affairs is episodic. People cannot
keep abreast of every development in international affairs, and it would not be in
their interests to do so. However, as people respond to the context under which
they are asked to form judgments about foreign policy, they become more attentive
to international affairs. This has important consequences for their attitudes about
foreign policy, and their willingness to ensure democratically elected officials act
according to their preferences.
1
For example, some classified information and private briefings may not be available online.
However, even these documents are available through official information requests in most demo-
cratic states. Websites such as Wikileaks also publish information sourced from intelligence and
diplomatic communities that was previously only available to people within those communities.
As data protection and cybersecurity measures have not kept pace with hackers’ ability to ex-
tract information from insecure networks, any information gap that exists between political elites,
bureaucrats, and the public continues to narrow.
4
1.2 Overview
The dissertation consists of nine further chapters as I develop and test my theory of
information, public opinion, and international relations.
Chapter 2 reviews the state of the literature concerning public opinion about
foreign policy, before introducing my theoretical argument. In this chapter, I argue
that an improved understanding of the dynamics of attention and information search
is essential to better understand public opinion about foreign policy.
To address this gap in the understanding of public opinion and foreign policy, I
introduce a theory of incentives, information, and public opinion about foreign pol-
icy in Chapter 3. This decision theoretic model outlines the conditions under which
people attention to international affairs, and how this influences political knowledge,
political predictions, and ultimately political attitudes about foreign policy. I de-
fine and operationalize the constructs and discuss the observable implications of the
theory for the dependent variables of interest: information search, political knowl-
edge, political predictions, and political attitudes. This theory forms the basis of the
argument which is elaborated upon in subsequent chapters.
Chapter 4 is the initial test of the theory introduced in Chapter 3. This chapter
uses an incentivized survey experiment to examine the effects of financial incen-
tives on information search, political knowledge, political predictions, and political
attitudes. I find that people respond to incentives by searching for information
and improving their political knowledge. The results also indicate that individuals
who received the lottery treatment become less likely to make accurate predictions,
5
demonstrating how the uncertainty of payoffs demotivates individuals. In particu-
lar, the chapter illustrates how people can become less motivated to make accurate
decisions about international affairs if they have little expectation of receiving any
reward for their efforts. Finally, the chapter suggests that people search for informa-
tion, improve their political knowledge, and update their foreign policy attitudes as
the stakes of making correct decisions increase. This has important implications for
the study of public opinion and international relations, suggesting that people are
capable of effective decision making when it matters most.
Chapters 5, 6, and 7 then shift attention from experimental evidence to observa-
tional data to examine the implications of variation in perception of incentives for
accuracy on attitudes about foreign policy. Chapters 5 and 6 examine the effects
of between-subject differences in perceptions of incentives for accuracy in debates
over U.S. military action in Syria and the Brexit Referendum respectively. Chapter
7 examines the implications of within-subject variation in perceptions of accuracy
on information search and attitudes about foreign policy across three major foreign
policy debates in the U.S.
In Chapter 5, I use multilevel regression to estimate how perceptions of incentives
foraccuracy andpolitical behaviorvarysystematically by votertype. Iillustrate how
voter type influences individual attention to foreign policy debates over U.S. military
action in Syria. The chapter demonstrates that registered voters are systematically
more likely to pay attention to international affairs than other individuals. The re-
sultssuggestthatpreviousresearchhasnotpaidenoughattentiontotheimplications
of heterogeneity of voter type for political behavior, and it also underestimates the
willingness of voters to fulfill their democratic obligations.
6
Chapter 6 argues that individuals’ foreign policy attitudes are contingent on their
perceptionsofincentivesforaccuracy. Itestthistheorythroughanalysisofdatafrom
the British Election Survey immediately prior to the referendum on the membership
of the United Kingdom in the European Union [E.U.]. The results indicate that
individuals who perceived the impact of the vote as being high were significantly less
likelytovotetoleavetheE.U., holdingpotentialconfoundingvariablesconstant. The
results demonstrate the importance of perceptions of accuracy on political attitudes
and political behavior in a public referendum on foreign policy.
In Chapter 7, I examine how within-individual variation in perceptions of in-
centives for accuracy influence information search and foreign policy attitudes over
time. I illustrate how the onset of crisis leads people to search for information about
international affairs, and to update their foreign policy attitudes in response to new
information. Using evidence from debates about U.S. foreign policy towards Iraq,
Syria, and ISIS, I demonstrate the plausibility of the theory that dynamics of cri-
sis and information search explain variation in public opinion about foreign policy.
The evidence presented here suggests that people can fulfill their democratic obliga-
tions as an informed citizenry that can produce informed decisions using available
information when it matters most.
The dissertation ends with a concluding chapter that discusses its implications.
I summarize the arguments and the evidence presented in the dissertation. I discuss
the primary policy implications of the research – that if political actors want people
to change their mind about foreign policy, they must convince people that there are
incentives for accuracy. If people perceive that it is important to make the correct
7
decision, and that their opinion might matter, they are more likely to search for
information and update their attitudes.
I also outline the opportunities for future research to build on the research pre-
sented here in the pursuit of an improved understanding of information, public opin-
ion, and international relations. In particular, I argue that an important extension
of this research is to examine how incentives for accuracy and information search
influence the public as a constraint on foreign policymakers in democratic states. I
then conclude with a brief discussion of the contribution of the research to the un-
derstanding of public opinion about foreign policy, and international relations more
broadly.
8
Chapter 2
Information, Public Opinion, and
Foreign Policy
Previous research has contributed towards a sophisticated understanding about how
peopleformattitudesaboutforeignpolicy. However, twosignificantlimitationserode
the foundations upon which this body of scholarship rests.
First, previous research has built a thorough understanding of how the supply of
information about international affairs to the public influences individuals’ attitudes
about foreign affairs (Baum and Potter 2008, 2015). Yet, comparatively little re-
search has examined the implications of the conditions under which people demand
information about international affairs, and how this influences public opinion about
foreign policy. It is vital to study processes of information search to gain a complete
picture of public opinion about foreign policy.
Second, there has been a deficit of attention to attention. If people have lim-
ited attention available (Broadbent 2013; Lupia and McCubbins 1998; Popkin 1994;
9
Schneider and Shiffrin 1977; Simon 1971; Stroud 2017), it is important to understand
the conditions under which people pay attention to international affairs, and how this
affects individuals’ attitudes about foreign policy.
In this chapter, I argue that an improved understanding of the dynamics of at-
tention and information search is essential to better understand public opinion about
foreign policy. While scholars have developed a thorough understanding of how peo-
ple form political attitudes in low-information environments, this scholarship has
not considered when and how people might seek information about international
relations, and how this influences people’s attitudes about foreign policy.
I outline avenues to address this gap in the understanding of public opinion and
foreign policy. The chapter advocates for the recalibration of public opinion scholar-
ship in international relations, so that greater attention is given to the implications
of changes in individuals’ demand for information about international affairs. This
research has the capacity to reshape the subfield, providing the basis for innovative
research within an already substantial research program.
The chapter proceeds as follows. In the first section, I review the state of the
literatureinpublicopinionaboutforeignpolicy. Second, Ireviewhowpreviousschol-
arship has addressed how the supply of information influences public opinion about
foreign policy. Yet, I discuss how this leaves a critical gap in the understanding
of public opinion about foreign policy, where scholars have not developed an un-
derstanding of the consequences of changes in individuals’ demand for information
about international affairs.
10
The third section outlines directions for future research in the pursuit of an im-
proved understanding of information, public opinion, and foreign policy, with rec-
ommendations for research into effects of changes in the demand for information on
several dependent variables of interest. Finally, the chapter concludes with a brief
summary of the implications of this research agenda for the study of public opinion
and foreign policy, and IR scholarship more broadly.
2.1 Public Opinion about Foreign Policy
The study of public opinion about foreign policy has evolved into a significant sub-
field within International Relations, with a significant amount of research having
been conducted into how people form attitudes about foreign policy. As the area
of inquiry has developed, insights from political science, psychology, behavioral eco-
nomics, communication, and sociology have contributed towards advancements in
theory, while increasingly sophisticated methods have led to increased rigor in the
field. In this section, I provide a potted overview of developments in the subfield,
before I outline how the study of public opinion about foreign policy could be im-
proved further by increased attention to the conditions under which people demand
information about international affairs.
Early studies of public opinion about foreign policy characterized the mass opin-
ion as emotive, unstructured, and irrational because people were easily persuaded to
adopt positions about foreign policies they knew little about (Almond 1960; Holsti
1992; Lippmann 1927). This pessimistic vision suggested that public opinion was so
unstructured and unstable it was hardly worth paying attention to. Simultaneously,
11
scholarship about international politics avoided studying public opinion, instead fo-
cusing on states as ’black boxes’ that act out of self-interest in the international
system (Morgenthau 1948; Waltz 1959, 1979). Over time, what became known as
the “Almond-Lippmann consensus” was supplanted by more nuanced approaches to
the study of public opinion about foreign policy.
Although scholarship in public opinion and foreign policy has shifted away from
such pessimistic visions of the mass public, few studies expect individuals to be
well informed on foreign policy. Instead, one branch of scholarship argues that atti-
tudes are structured around time invariant core values, morality, mood, or partisan-
ship (Gries 2014; Hatemi and McDermott 2016; Hurwitz and Peffley 1987; Jentleson
1992; Kertzer 2013; Kertzer and Rathbun 2015; Kertzer et al. 2014; Kupchan and
Trubowitz2007;MilnerandJudkins2004;Rathbun2004,2011;Rathbunetal.2016).
As a result, foreign policy preferences remain stable irrespective of new information,
and it is assumed that individuals’ knowledge about international affairs is consistent
over time.
To overcome their lack of knowledge about politics, another research program
suggests people use information shortcuts to form opinions consistent with their be-
liefs (Lupia and McCubbins 1998; Page and Shapiro 1992; Popkin 1994; Simon 1978).
For example, domestic publics in democracies constrain foreign policymakers from
using force against other democratic states to resolve disputes (Maoz and Russett
1993; Mesquita et al. 1999; Russett 1994; Tomz and Weeks 2013). Party ID and
political ideology might explain people’s attitudes towards foreign policy options
(Abramowitz 2010; Bartels 2002; Brady and Sniderman 1985; Fiorina and Abrams
2008).
12
People also use shortcuts such as elite cues (Berinsky 2007, 2009; Zaller 1992);
social cues (Kertzer and Zeitzoff 2017); moral values (Kertzer and Rathbun 2015;
Kertzer et al. 2014; Rathbun et al. 2016); regime type (Tomz and Weeks 2013);
numbers of casualties (Gartner and Segura 1998, 2000; Gelpi, Feaver, and Reifler
2009;Mueller1973); outcomesofconflict(Eichenberg2005;Gelpi, Feaver, andReifler
2006); or rally behind leaders ahead of conflict (Baum 2002a; Brody 1992; Gadarian
2010; James and Rioux 1998; Kam and Ramos 2008; Lian and Oneal 1993; Mueller
1973) to help them form reasoned choices about foreign policy. As such, people are
capable of making informed decisions about politics using limited information (Page
and Shapiro 1992).
Recently, the “behavioral revolution” has also led to advances in the study of
public opinion about foreign policy. Scholars have also developed a sophisticated
understanding of the effects of psychological biases and heuristics that people in-
voke on to make decisions about politics (Gaines et al. 2007; Gigerenzer et al. 1999;
Kahneman 2013; Kahneman and Tversky 1979; Simon 1978; Tversky and Kahneman
1981). People often fall short of meeting ex ante expectations of rational decision
making, placing too much weight to the desirability of outcomes ahead of feasibility
(Rapport 2012). People are susceptible to framing effects (Chong and Druckman
2007b; Kahneman and Tversky 1979; Tversky and Kahneman 1981); using mislead-
ing analogies to make decisions (Khong 1992); rationalizing contradictory evidence
(Koopman, Snyder, and Jervis 1990); and becoming overconfident as war becomes
more likely (Johnson and Tierney 2011).
Similarly, scholarly attention has shifted to the complex interactions between
emotion, cognition and political attitudes (Albertson and Gadarian 2015; Bargh,
13
Chen, and Burrows 1996; Forgas 1995; Huddy et al. 2005; Lazarus 1991; Lodge and
Taber2013;Marcus, Neuman, andMacKuen2000;Neuman2007;PettyandWegener
1998; Petty and Cacioppo 1986; Sears and Freedman 1967; Taber and Lodge 2006;
Zajonc 1980). Although scholars disagree about the order of affective responses to
stimuli and cognition, this research emphasizes the importance of emotion for models
of attitude formation.
Affect is particularly significant when considering partisan politics, where emo-
tions and loyalty drive attitudes (Clifford 2017). People often come to interpret
events with preconceptions, which bias the processing of information across a range
of issues in the U.S. and elsewhere (Gaines et al. 2007; Lodge and Taber 2013; Taber
and Lodge 2006).
Biased information processing occurs even when new information challenges an
individual’s prior attitudes (Suhay et al. 2015), and anger about political informa-
tion can trigger biased information assimilation (Suhay and Erisen n.d.). Individuals
double down on their prior beliefs when confronted with competing frames and con-
flictual information (Chong and Druckman 2007a,c); and resist attempts at persua-
sion (Tormala and Petty 2002, 2004). If people are unresponsive to new information,
this undermines their ability to make reasoned judgments about politics, especially
when confronted with contradictory information.
Increasingly, scholars have sought to understand how partisanship affects atti-
tudes towards foreign policy. While previously people thought that partisanship
may stop at the water’s edge, recent evidence demonstrates there are important dif-
ferences between the foreign policy attitudes of Democrats and Republicans in the
United States (Gries 2014; Kupchan and Trubowitz 2007; Milner and Judkins 2004;
14
Rathbun 2004, 2011). If people see international affairs through partisan lenses, it is
possiblethatpeoplemightbeasunresponsivetonewinformationaboutinternational
affairs as they are to domestic issues.
Collectively, previous research has contributed much towards our understanding
about public opinion and foreign policy. However, this scholarship suffers from im-
portant limitations that mean that we only have a partial understanding of attitude
formation in the domain of international politics.
First, prior research almost exclusively relies on low-information environments to
measureandtestpublicopinion. Masssurveysrelyontop-of-the-headresponsesthat
do not necessarily correspond to true attitudes (Zaller and Feldman 1992). While ex-
perimentsisolatethecausaleffectsoftreatmentsonoutcomes, themajorityofstudies
rely on artificially low-information environments that do not replicate patterns of in-
formation acquisition in the real world, with a few notable exceptions (Bolsen and
Leeper 2013; Druckman, Fein, and Leeper 2012; Leeper 2014). If information was
once scarce, that is not the case today. Yet, existing scholarship does not take this
into account.
A second critical limitation of prior research is that it has not paid enough at-
tention to attention. Prior scholarship largely assumes that people’s patterns of
information consumption are consistent and stable over time. Measures of politi-
cal knowledge and news consumption manage to capture the nature of information
search at a given point in time, but the absence of longitudinal studies of these
patterns over time make generalization from these results over time problematic.
15
Asaresult, scholarshipcouldbeimprovedwithincreasedattentiontoinformation
consumption. For an improved understanding of public opinion about foreign policy,
it is critical to consider how people receive information about international affairs.
2.2 How People Receive Information about Inter-
national Affairs
To understand how people think and feel about international affairs, it is impor-
tant to consider how people receive information about international affairs. Where
information is included in models of public opinion about foreign policy, scholars
overwhelmingly focus on the supply of information to the public, including the anal-
ysis of the news media environment and democratic constraints as explaining public
attitudes towards foreign policy.
Unfortunately, this focus only tells part of the story. Individuals’ demand for
information also may play an important role in influencing public opinion about
foreignpolicy. Inthissection, Ireviewtheliteratureonhowthesupplyofinformation
influences public opinion about foreign policy, before highlighting how there remains
a significant absence of knowledge about how the demand for information augments
public opinion. I then borrow from work in other disciplines to suggest how the
demand for information about international relations likely affects public opinion
about foreign policy.
16
2.2.1 The Supply of Information about International Affairs
As people do not experience most events first hand, they rely on providers of infor-
mation to learn about events and debates in the world outside of their own limited
experience. As most people’s primary source of information about the world, it
is important to consider when the news media cover international affairs, how the
news media cover international affairs, and how news coverage of international affairs
influences public opinion about foreign policy.
News organizations provide a mediated version of reality, one that is shaped by
providers’ incentives and constraints. Accordingly, “the typical conception of the me-
dia’s role...as reporter-reflector-indicators of an objective reality ‘out there’” is false
(Molotch and Lester 1974, 105). To understand public opinion about foreign policy,
it is important to acknowledge the processes through which mass media produce
information about international affairs, and how this shapes public opinion about
foreign policy.
When the News Media Covers International Affairs
The news is the product of a series of decisions made by journalists, editors, and
producers. Gatekeeping processes affect what makes it into the finished product,
as editors select stories to include in the final bulletin at the expense of others
(Bennett, Lawrence, and Livingston 2007; Shoemaker 1996; Tuchman 1980; White
1950). Values of newsworthiness also influence the selection of events, and how they
are reported as news producers attempt to maximize their audience (Galtung and
Ruge 1965; Hamilton 2006; Harcup and O’Neill 2001). News producers do their
best to anticipate consumer demand for stories about particular topics, and provide
17
information that people are willing to purchase from the producer (Hamilton 2006,
18-19). Of all topics, foreign policy crises are especially newsworthy.
As a result, it is no surprise that news media cover crises extensively. As visually
compelling and dramatic events, “wars and civil conflicts are, therefore, good news
for 24/7 networks” (Thussu 2003, 123-124). For example, network news coverage
of the 1991 Gulf War was characterized by “an initial period of saturation coverage
followed by a steady state of heavy news that culminated in 2 months of virtual
saturation coverage” (Iyengar and Simon 1993, 374). Similarly, the 2003 Iraq War
spanned over 4,000 minutes of network news coverage in 2003, before declining over
time (Pérez-Peña 2008).
The volume and frequency of news coverage of international affairs is significant
because press attention can set the agenda for discussions of foreign policy at home.
Cohen(1963,13)arguedthatthenewsmedia“maynotbesuccessfulmuchofthetime
in telling people what to think, but it is stunningly successful in telling its readers
what to think about.” Press coverage “forces attention to certain issues” through
repeated coverage that details “what individuals in the mass should think about,
know about, have feelings about” (Lang and Lang 1966, 468). Repeated studies
have found the news media’s agenda setting function to be consistently powerful
across different platforms, setting the terms of political debate through emphasis on
specific events or issues (Iyengar and Kinder 2010; McCombs and Shaw 1972).
Scholars have gone further to postulate that the proliferation and popularity
of 24-hour news channels has effects on public opinion and foreign policy. For
instance, the “CNN effect” literature assesses “the effect of a worldwide array of
print and broadcast media capable of instantaneous, real-time coverage of breaking
18
events”(Livingston and Eachus 1995, 413). Theoretically, the constant stream of
information on cable channels leads to shifts in public opinion in response to crises
that are increasingly visible to the public. As people are exposed to images of hu-
man suffering, the public pressures policymakers to respond. There may also be
direct effects of coverage on elite opinion as policymakers respond to news coverage.
However, recent evidence suggests the CNN effect is limited, and policy uncertainty
and elite divisions have more explanatory power than news coverage of distant events
(Bahador 2007; Drury, Olson, and Belle 2005; Gilboa 2005a,b; Livingston 1997, 2011;
Potter and Van Belle 2008; Robinson 1999, 2000, 2001). Collectively, scholars have
built a thorough understanding of news media attention to international affairs, par-
ticularly during periods of crisis.
How the News Media Covers International Affairs
Scholars have also examined how the news media cover international affairs, and
how this influences public opinion about foreign policy. Much research focuses on
the interaction between political elites and the news media.
One recurring debate concerns whether the news media performs as a “watchdog”
or as a “lapdog” to either enable or constrain governments (Kim, Whitten-Woodring,
and James 2015; Kovach and Rosenstiel 2007; Robertson 2004; Whitten-Woodring
2009; Whitten-Woodring and James 2012). Some scholars argue that the media is
overlyinfluencedbyeliterhetoric, wherepoliticalactorsdirectlyinfluencethecontent
of the news and the news media merely serve as a conduit for elites to get their mes-
sages to the mass public (Bloch and Lehman-Wilzig 2002; Brody 1991; Malek 1996;
Zaller 1992). This is reflected in the consistent finding that the media “index” their
19
coverage of domestic political debates according to the extent to which elites disagree
(Bennett 1990; Bennett, Lawrence, and Livingston 2007; Hallin 1989). Because jour-
nalists rely on political elites as sources to verify accounts of events, provide facts,
and offer opinions, elites exert some influence over news reports (Bennett and Paletz
1994; Davis 2007, 2010; Livingston and Eachus 1995). However, the relationship
between elites and the news media is best characterized as interdependent.
There are “strange mutual dependencies (punctuated by occasional bursts of
antagonism) between reporters and officials” (Bennett, Lawrence, and Livingston
2007, 3). While journalists rely on elites as sources, political actors also rely on
the news media to communicate to the public. A prominent example of this is
the omnipresence of campaign surrogates and administration officials on cable news
channels to disseminate their positions to the mass public.
1
News channels require
official sources to provide free content for their news shows, filling up airtime by
providing their version of events as authoritative sources. For political actors, the
platform of news channels is an opportunity to frame the debate and to spin events
according to their own objectives to the mass public. As political actors and news
organizations depend on each other, they typically avoid antagonizing the other to
the extent to which they would endanger access to the other party.
1
A further demonstration of the interdependence between official sources and journalists is
the practice of “embedding” journalists within military units during conflict (Cortell, Eisinger,
and Althaus 2009). Some critics of the program during the 2003 Iraq War feared that the military
exerted too much control over reports from the war, and that they received more favorable coverage
than they would otherwise (Biernatzki 2003; Brockus 2009; Der Derian 2008; Haigh et al. 2006;
Pfau et al. 2004; Robertson 2004). However, subsequent analysis suggests these concerns were not
borne out in embedded reporters’ coverage of the war (Aday, Livingston, and Hebert 2005; Fox and
Park 2006)
20
In the contemporary information environment, some scholars have accused news
organizations of a “race to the bottom” in the wake of technological changes and
increased access, where news providers prioritize high-rating entertainment news and
“soft news” above news oriented towards the public interest. Baum (2002b) argued
thattheincreasedfrequencyofsoftnewshassomepositiveeffectsonpeople’spolitical
knowledge, as people who would not otherwise have read the news are exposed to
current events through this type of news. In turn, this can help people vote according
totheirpreferencesdespitenotbeinginterestedinpolitics(BaumandJamison2006).
However, other scholars contend that there is insufficient evidence to suggest that
people learn much from soft news at all (Prior 2003).
Citizens also exert some control over content as consumers of news that tells
them what they need to know as quickly as possible, preferring coverage of elite
competition than of consensus because it is more interesting, and they wish to know
what the government intends to do (Zaller 1999). As such, journalists are compelled
to provide such content to their audience, even if they wish to smuggle in more
sophisticated content than their audience desires (Zaller 1999). As a result, coverage
ofinternationalaffairsisdeterminedbyacombinationof: 1)eliteopinionastheyseek
to influence mass opinion; 2) journalists’ ability to gather facts and independently
verify elite accounts of international events; and 3) the public’s appetite for news
coverage of international affairs.
Although the incentives of political actors, the news media, and the public may
remain stable, their ability to achieve their objectives changes over time. Baum and
Groeling (2010) argued that “reality reasserts itself” over time, as the elite ability to
shape the narrative about foreign affairs decreases over time as the information gap
21
narrows between elites and the public, and contradictory evidence emerges. As more
facts emerge, the media is increasingly able to challenge political elites, resulting in
less reliance on official sources. As a result, we should expect to see that as a foreign
policy crisis endures, news coverage of the crisis should become more negative over
time, especially relating to the incumbent government involved in the crisis. As a
result, thenatureofcoverageofforeignpolicycrisesisnotjustafunctionofattention,
but also is reliant on the availability of information that journalists can use to verify
official accounts.
Another important constraint that might influence the nature of news coverage of
international affairs is the institutional environment. As a cornerstone of democratic
states, the news media can be a counterweight to elite rhetoric by fact checking,
questioning arguments, and monitoring political acts. As a result, in states with
media freedom, the news media can place constraints on foreign policymakers. As a
result, Van Belle (1997) argues that the mechanism driving the democratic peace is
press freedom, and that democracies without a free press would be as likely to fight
one another than if they were not democratic. Choi and James (2007) found that
media openness dampens conflict propensity as measured by militarized interstate
disputes. Baum and Potter (2015) found that states with a free, independent press
are less like to initiate conflict, reciprocate challenges, and join multi-state coalitions
to enter conflict. In short, a free press is essential as a constraint on the use of
military force. Without an open information environment, it is much easier for
leaders to avoid the domestic costs of foreign policy decisions in the absence of a free
press.
22
The Supply of Information about International Affairs in the Digital Age
While most scholarship examines traditional news media coverage of international
affairs, recent technological advances have augmented the way that people receive
mediated information about the world. Since the publication of The Nature and
Origins of Mass Opinion (Zaller 1992), the internet has expanded from a fledgling
information system used by universities to a global system used by approximately
3 billion people across the world (Farrell 2012; Golder and Macy 2014; Pierce, Red-
lawsk, and Cohen 2017). The rise of the internet has increased the availability of
political information, as people can read news websites from traditional news media
outlets at a fraction of the cost of a newspaper or cable subscription. Simultaneously,
the relativecosts of news production have decreased, contributing to the proliferation
of digital news providers.
Broadsheet newspapers have long been considered an important element of demo-
cratic societies, providing the mass public with authoritative coverage of political
events (Davis 2007, 2010). However, the proliferation of online news providers and
changing information consumption patterns have challenged the business models of
even the top newspapers. Some premium news organizations such as the Times
(London), New York Times, Washington Post, and the Wall Street Journal have set
up paywalls or limits on readership that require people to subscribe to view content
online. Other news organizations like the Guardian have membership subscriptions,
where supporters donate money to support the organization much like PBS and NPR
campaigns in the United States. The advent of the internet has forced traditional
quality newspapers to adapt. Thus far, the newspapers of record have managed to
23
survive these changes, and some have arguably grown stronger as demand for quality
reporting increases.
As online tools have developed, people also receive information about politics
from an increasing variety of sources. People may now receive news from the online
versions of traditional news providers, but also learn from peers in social networks
(Barberá, Jost, Nagler, Tucker, and Bonneau 2015; Barberá, Wang, Bonneau, Jost,
Nagler, Tucker, and González-Bailón 2015; Colleoni, Rozza, and Arvidsson 2014;
Conover et al. 2012; Eckles and Bakshy 2017; Kramer, Guillory, and Hancock 2014;
Ruths and Pfeffer 2014; Salganik 2017), partisan blogs (Adamic and Glance 2005),
or directly from politicians themselves (Barberá and Zeitzoff n.d.; Theocharis et al.
2016). This has arguably reduced the agenda setting power of traditional news media
(Djerf-Pierre and Shehata n.d.).
However, the plethora of online information presents new challenges to civil
democracy. Arguably, the internet has led to political polarization (Farrell 2012), as
people sort into ideological social networks and limit their exposure to contradictory
information (Adamic and Glance 2005; Barberá 2015; Colleoni, Rozza, and Arvids-
son 2014; Conover et al. 2012; Vaccari et al. 2016). The increasing prominence of
“fake news” (Allcott and Gentzkow 2017), bots that automatically generate politi-
cal messages (Shao et al. 2017), and misinformation on social media raises further
challenges to the public.
Information about international affairs has never been more accessible, but the
increased access comes with increased cognitive burdens associated with processing
this information. While it is important to understand the supply of information for
24
an understanding of public opinion about foreign policy, it only constitutes part of
the story about how people receive information about international affairs.
2.2.2 The Demand for Information about International Af-
fairs
Tounderstandhowpeoplereceiveinformationaboutinternationalaffairs, itiscritical
to examine how individuals’ demand for information shapes their attitudes about
international relations. To date, this aspect of public opinion and foreign policy has
been understudied.
For instance, the most comprehensive accounts of information, public opinion,
and international relations introduced the notion of the “information gap” between
leaders and the public (Baum and Potter 2008, 2015). Baum and Potter (2015)
argue that in areas of foreign policy, leaders have traditionally enjoyed informational
advantages such as internal briefings, acess to classified information and intelligence
reports to which the average member of the public does not have access. This gap
remains in place until the news media is able to access facts that challenge the
official version of events, which gradually occurs over time. Yet, their theory does
not examine how individuals’ demand for information can narrow the information
gap. However, Feldman, Huddy, and Marcus (2015) demonstrated that people are
capableofsearchingforinformationandformingforeignpolicyattitudesindependent
of the news media and political elites.
It is not just the accessibility of information that may cause public opinion to
diverge from elite leadership, but it is also people’s willingness to exert effort into
25
seeking information. In fact, “for many political issues, information is not scarce;
rather it is the cognitive resources that a person can use to process information are
scarce” (Lupia and McCubbins 1998, 6). Because people are limited in their ability
to make sense of a complex world, it is important to understand the conditions under
which people are motivated to pay attention to international relations.
Priorscholarshiphasnotfullyaddressedtheconditionsunderwhichpeoplesearch
for information about international affairs, and how they use this information to
learn about foreign policy and form political attitudes. Instead, the importance of
information acquisition has remained implicit or absent in models of public opinion
and foreign policy. I aim to address this gap in the understanding of public opinion
about foreign policy.
Information Search
To understand how people think about foreign policy, it is vital to understand the
conditions under which people search for information about international affairs. It
is impossible to understand how individuals form political attitudes without under-
standing why people search for information, when people search for information, and
whether people use information to confirm or challenge their prior beliefs (Lau and
Redlawsk 2006).
Previous research suggests that people typically seek out information to con-
firm preexisting beliefs and preferences (Jonas et al. 2001). People consciously seek
information that reinforces their beliefs even if available information provides con-
tradictory evidence (Frey 1986; Frey, Schulz-Hardt, and Stahlberg 2013; Johnston
26
1996; Pinkley, Griffith, and Northcraft 1995). This could lead to poor decisionmak-
ing as people forgo alternative perspectives, leading to unfortunate outcomes such as
“Groupthink” (Janis 1972, 1982; Nemeth and Rogers 1996) or “Polythink” (Mintz
and Wayne 2016a,b). However, there are reasons to believe that people may search
for information to help them when they believe there are incentives to make accurate
decisions.
In this respect, it is important to consider the context in which decisions are
made (Johns 2018). While people might generally rely on information shortcuts
when forming attitudes towards foreign policy, there are periods when people are
incentivized to exert extra attention to international relations.
In the context of international relations, scholars of public opinion and foreign
policy are most interested in people’s attitudes about responses to crises including
debates about entering military conflict, responses to terrorist attacks, responses to
macroeconomic conditions, and international agreements. All of these topics have
potential effects on the state of the country, and they could have direct effects on
people’s lives.
If learning is active (Lupia and McCubbins 1998), people might actively search
for information to form attitudes about policies where they perceive there are real
incentives for accurate decision making. During such periods, it is likely that people
search for additional information to become better informed about international
affairs. In fact, when people perceive their choices are consequential, they may even
search for information when it is not directly useful for their decision making (Eliaz
and Schotter 2007, 2010; Loewenstein 1994; Powdthavee and Riyanto 2015). As
an illustrative example, cable news consumption soared in response to crisis after
27
the 9/11 attacks in September 2001 as people actively pursued information about
international affairs (Althaus 2002).
Baum and Groeling (2010) argued that states cannot control the framing of real
events for perpetuity - instead, reality will eventually “assert itself” as the public be-
comes aware of the actual situation in terms of its success and the costs of conflict.
During crises where people are directly affected, they will likely additional informa-
tion about the foreign policy. As people consume this information, attitudes towards
domestic governments can rapidly change depending on their domestic public’s in-
terpretation of their response. During these periods of crisis, the information gap
between elites and the public narrows as people pay extra attention towards seeking
information about international relations. As a result, information search could have
profound implications for the study of public opinion about foreign policy.
Information Avoidance
However, demand for information is not unidirectional. Instead, people may actively
avoid information about international affairs, and individual responses to crisis may
vary. People may actively avoid information for a variety of reasons, which may com-
promisepeople’sabilitytomakeaccuratedecisions, althoughitmayactuallyimprove
decision making in some instances (Golman, Hagmann, and Loewenstein 2017). It is
important to examine the conditions under which people actively avoid information
so that scholars can better understand how people think about international affairs.
Previous research indicates that people may avoid information by evading infor-
mation they might find painful, such as personal evaluations (Eil and Rao 2011), or
information about medical conditions (Ganguly and Tasoff n.d.). Individuals might
28
also be inattentive to information that challenges what they want to hear (Brock
and Balloun 1967; Taylor and Brown 1988), or their prior beliefs (Benoît and Dubra
2011; Glaeser and Sunstein 2013; Lord, Ross, and Lepper 1979).
People might also forget pertinent information that may help them avoid reliving
painful events or experiences (Akerlof and Dickens 1982; Anderson and Huddleston
2012;BénabouandTirole2002;ShuandGino2012). Asaresult, peoplemayactively
avoid searching for information that has the potential to challenge their prior beliefs
or upset them.
In the context of international relations, it is important to consider how foreign
policy crises may induce information avoidance among individuals, and the implica-
tions of this for public opinion about foreign policy. Political attitudes about foreign
policy could be a function of individuals’ attitudes towards risk and loss (Kőszegi
2010; Kőszegi and Rabin 2009), where people avoid information to avoid disappoint-
ment (Köszegi 2006).
Anxiety might also lead people to avoid information about international affairs,
as individuals anticipate that new information might make them feel even worse
(Albertson and Gadarian 2015; Caplin and Leahy 2001; Huddy et al. 2005; Kőszegi
2003; Maslow 1963). For example, people living in Hawaii may not want to hear
about improvements in the range of North Korea’s nuclear missiles. In an uncer-
tain international environment during crisis, people may therefore avoid information
because it might make them feel less secure.
People also may avoid paying attention to information that may confirm bad
news (Falk and Zimmermann 2016; Miller 1987), and people might avoid new infor-
mation because it may have overwhelming effects on perceptions of utility (Golman
29
and Loewenstein 2015). People may also want to avoid information to avert feelings
of regret about previous decisions (Loomes and Sugden 1982); to maintain optimism
about outcomes (Brunnermeier and Parker 2005; Oster, Shoulson, and Dorsey 2013;
Sharot 2011); to avoid cognitive dissonance (Abelson et al. 1968; Chater and Loewen-
stein 2016; Festinger 1962; Scherer, Windschitl, and Smith 2013); or because they
have already invested in perpetuating beliefs about an issue (Bénabou and Tirole
2011; Golman et al. 2016). Patterns of information avoidance may have significant
effects on people’s attitudes towards conflict, especially if they anticipate bad news.
In sum, it is important to understand the conditions under which people avoid
information about international affairs. In spite of incentives to make correct de-
cisions, information avoidance might lead to the reinforcement of individuals’ prior
beliefs and negate the effect of new information on people’s attitudes towards for-
eign policy. As a result, it is important to consider how increases and decreases in
individuals’ demand for information affects public opinion about foreign policy.
2.3 Towards an Understanding of Information,
Public Opinion, and Foreign Policy
Beyond simply acknowledging the importance of individuals’ attention and their
demand for information, it is important to outline steps for scholars to incorporate
these ideas into their work. Measuring and testing the role of demand for information
involves some innovation above and beyond standard experimental and observational
approaches to the study of public opinion and foreign policy, but it is possible. This
research has the chance to reshape the landscape of public opinion scholarship in IR,
30
taking context into account, and improving the generalizability of research to the
real world.
2.3.1 Attention, Information, and Political Knowledge
For democratic states to function effectively, citizens should acquire information
and use that to make decisions to hold elected officials accountable. If citizens are
uninformed, they cannot make good decisions at the voting booth or be an effective
constraint on foreign policymakers. This is an important concern for all political
debates, but the stakes are particularly high in foreign policy.
Political science scholarship suggests reasons for skepticism about citizen com-
petence. The literature concludes that the public knows very little about politics
in general, and foreign policy in particular (Delli Carpini and Keeter 1996; Lupia
2015). Multiple studies suggest the public does not always fulfill their democratic
obligations as one might hope. For instance, during deliberations about foreign pol-
icy, people struggle to place the relevant countries on a map (Dropp, Kertzer, and
Zeitzoff 2014); while people routinely overestimate the number of immigrants in their
country (Sides and Citrin 2007). In short, the evidence suggests the public knows
very little about politics, and that they struggle to influence policymakers when they
do have informed opinions.
Arguably there are no more important decisions for domestic publics than deci-
sions related to international security. Conflict is costly, and the burdens of military
campaigns are suffered by domestic audiences as their taxes fund the military, their
security is under threat from opposing forces, and they provide the troops for the
war effort. Given the significant personal and financial stakes, the public should be
31
reticent about entering conflict unless it is necessary. During such periods, when it
matters most, are people capable of acquiring knowledge about international affairs
to make informed decisions about foreign policy?
2.3.2 Attention, Information, and Political Predictions
Another important element of public opinion about foreign policy concerns people’s
ability to make accurate predictions about international affairs. If people support
policies without some awareness of their implications, this presents important prob-
lems for policymakers who are responsible to their individual constituents and their
country.
Previous research suggests that it is extremely difficult to make predictions about
international affairs, even for experts in the area (Tetlock 1998, 1999, 2006). Tetlock
(1999, 351) found that area experts “were only slightly more accurate in making
predictions [about foreign affairs] than one would expect from chance.”
However, recent scholarship suggests that people are capable of making accurate
predictions about the world over a period of time (Mellers et al. 2015; Tetlock 1998,
1992; Tetlock and Gardner 2016; Ungar et al. 2012). Tetlock and Gardner (2016, 88)
found that “we now know a few hundred ordinary people and some simple math can
not only compete with professionals supported by a multibillion-dollar apparatus but
also beat them.”
Although these “superforecasters” are not representative of the broader popula-
tion, it begs the question, if members of the public are capable of outperforming
experts with predicting international affairs, can this type of behavior be induced?
32
Do incentives for accuracy lead people to search for information, and ultimately
improve their ability to make predictions about international affairs?
2.3.3 Attention, Information, and Foreign Policy Attitudes
The vast majority of scholarship in public opinion about foreign policy examines
how people form political attitudes. While previous scholarship has neglected to pay
attention to the demand for information, and how this influences political attitudes,
there are reasons to believe there could be information effects when people perceive
there are incentives for accuracy.
People might generally rely on information shortcuts when forming attitudes
towards foreign policy, but there are periods when people are incentivized to exert
extra attention to international relations, such as during when the state is on the
brink of war, following a terrorist attack, or during periods where there is debate
about foreign policy. In these situations, information could become particularly
valuable when people anticipate that there are important consequences of a given
foreign policy decision, so it is important to make the correct choice.
During crisis, people typically report that foreign affairs are the most important
issue facing the U.S. (Heffington, Park, and Williams 2017); and people could per-
ceive personal consequences of the crisis (Huddy et al. 2002). As such, there are
direct incentives for them to pay attention to debates about appropriate responses,
search for information about the crisis, and ultimately form attitudes about the for-
eign policy based on their improved political knowledge. During crisis, when there
are incentives for people to make accurate judgments, individuals are more likely to
search for information about foreign affairs.
33
If people search for information in a high information environment, it is possi-
ble that people update their preferences according to the information they acquire,
independent of political ideology or party ID. For instance, Feldman, Huddy, and
Marcus (2015) found that people who consumed high amounts of political informa-
tion and information about Iraq became much less supportive of the U.S.-led war in
Iraq. Similarly, people who consumed political information from newspapers (and
not television news) were resistant to the war (Feldman, Huddy, and Marcus 2015).
In short, how does information acquisition and improved political knowledge influ-
ence people’s attitudes about foreign policy?
2.4 Conclusion
Although the literature is vast, there remains significant scope for an improved un-
derstanding of public opinion about foreign policy. Until now, scholars have focused
on the supply of information to the mass public, but there are reasons to believe that
shifts in demand for information have significant effects on public opinion about for-
eign policy. Scholars should take a comprehensive approach to information to better
understand the conditions under which people learn about international affairs, how
this influences attitudes about foreign policy, and the conditions under which the
public constrains foreign policymakers.
The implications of this research agenda are significant for the field of interna-
tional relations. Until now, scholars have only developed a partial understanding
of the way people think about international affairs. If people’s political knowledge,
34
ability to make predictions, their attitudes, and their willingness to hold elected of-
ficials to account are contingent on their demand for information, there is a large
research agenda ahead to build a more complete picture of public opinion about for-
eign policy. Although there is a substantial amount of research in this subfield, the
research program introduced here promises to provide the basis for an exciting new
area of research in IR.
In sum, scholars should consider how individuals’ demand for information shapes
their political knowledge and their political attitudes about foreign policy. The pro-
cess of information acquisition and attitude formation is dynamic in an information-
rich world, and theories of public opinion and foreign policy would benefit from
taking that into account in future studies. This chapter marks an initial step in the
right direction towards an improved understanding of information, public opinion,
and international relations.
35
Chapter 3
A Theory of Incentives,
Information, and Public Opinion
about Foreign Policy
While people might generally rely on information shortcuts such as party ID to
form attitudes towards foreign policy, there are conditions under which people are
incentivized to exert extra attention to international affairs. Similarly, individuals
may have systematically different information consumption patterns depending on
their perceptions of incentives to be informed about foreign policy debates. In this
chapter, I outline a theory of incentives, information, and public opinion about for-
eign policy that explains variation within and between individuals in their political
knowledge, their ability to make accurate predictions, and ultimately their attitudes
about foreign policy.
36
The chapter is structured in six further sections. First, I outline assumptions
about the nature of public opinion about foreign policy. Second, I introduce a de-
cision theoretic model outlining the conditions under which people attention to in-
ternational affairs, and how this influences political knowledge, political predictions,
and ultimately political attitudes about foreign policy. Third, I discuss hypotheses
derived from the theory for the dependent variables of interest: information search,
political knowledge, political predictions, and political attitudes. The fourth section
defines the constructs of interest, before the fifth section discusses the operational-
ization of the constructs across different types of data. Finally, I conclude with a
brief summary of the theory and the contribution for the understanding of public
opinion about foreign policy.
3.1 Theoretical Assumptions
I make several assumptions about the nature of political behavior that allow me to
build a theory of incentives, information, and public opinion. These can be broadly
conceived as assumptions of instrumental rationality on the part of individuals in a
democratic public.
First, I assume that people are goal-oriented, and that they are able rank the
importance of each of these goals. In other words, people can identify what they
want to achieve, and they prioritize the pursuit of goals that are most important to
them.
Second, people seek to maximize their utility. In other words, everyone is moti-
vated by the pursuit of their objectives and to achieve what they consider to be the
37
best possible outcome. People vary in their attitudes towards risk and uncertainty,
their ability to calculate the probability of outcomes, and the order of their prefer-
ences. However, people are united in their pursuit to receive the greatest possible
benefit for the least amount of costs.
Third, I assume that people act to make accurate decisions and avoid mistakes
where possible. In the pursuit of their goals, people want to make decisions that give
them the best chance of achieving their objectives. As a result, people wish to avoid
making poor decisions that prevent them from achieving their goals.
Finally, I assume that attention is a scarce resource. In other words, people’s
attention to a given task comes with opportunity costs of alternative tasks people
could have devoted their attention to. For example, if someone decides to read the
newspaper, any attention to the newspaper comes at the expense of the universe of
all possible alternative tasks that person could have completed instead.
3.2 Incentives, Information, and Public Opinion
about Foreign Policy
Having laid out these assumptions, it is possible to consider the conditions under
which people search for information about politics, how information search influences
political knowledge, and how this information influences political attitudes about
foreign policy.
Lupia and McCubbins make two critical propositions that are particularly per-
tinent. First, that “people choose when and what to learn” (Lupia and McCubbins
1998, 20). They consider learning to be active and not automatic because it requires
38
effort, and because effort is a scarce resource for everyone, people have to decide
what to pay attention to (Lupia and McCubbins 1998, 22). Accordingly, attention
is necessary for learning.
Lupia and McCubbins (1998, 20) posit “information is valuable only when it im-
proves the accuracy of predictions about the consequences of choices.” They make
the case that “the purpose of paying attention is to make reasoned choices,” which
people need to “have beliefs that generate accurate predictions about the conse-
quences of their actions” (Lupia and McCubbins 1998, 24). As such, “the desired
consequence of paying attention to a stimulus is to acquire knowledge sufficient for
reasoned choice” (Lupia and McCubbins 1998, 24). Accordingly, if information does
not help people gain knowledge about a particular issue, it is not valuable enough to
demand their attention, which could be better spent elsewhere.
If people are goal-oriented utility maximizers seeking to use their scarce atten-
tion to make accurate decisions and avoid mistakes, it implies that learning about
international affairs is active. Accordingly, people search for information and learn
about politics only when it helps them achieve their goals (Lupia and McCubbins
1998).
If information does not help people gain knowledge about an issue, it is not
valuable enough to demand their attention, which could be better spent elsewhere.
This behavior could explain why people readily adopt the information shortcuts,
psychological biases or heuristics to make decisions about politics that have been
exhaustively catalogued in previous work (Gaines et al. 2007; Gigerenzer et al. 1999;
Kahneman 2013; Kahneman and Tversky 1979; Lupia and McCubbins 1998; Lupia,
McCubbins, and Popkin 2000; Simon 1978; Tversky and Kahneman 1981). However,
39
while people might generally rely on information shortcuts when forming attitudes
towards foreign policy, individuals may be incentivized to exert extra attention to in-
ternational affairs due to intrinsic personal characteristics, or in response to changes
in the context in which they make decisions about foreign policy. Under these con-
ditions, information could become particularly valuable when one anticipates that
could benefit from or they could bear costs from a given foreign policy decision, so
it is important to make the correct choice.
Correct decisions can be considered as choices made by individuals that help them
achieve their goals. In the context of foreign policy, we can think of individuals as
being motivated by the pursuit of security and prosperity.
1
People are motivated to pursue security and prosperity, but they are not willing
to pay any price for those goals. Instead, people want their governments to enact
foreign policies that make them feel more secure, or that provide financial benefits
for them, and to avoid policies that deliver neither of these benefits. As Page and
Bouton (2006, 38) found:
1
Of course, the notion of correct decisions has a long history in political science and decision
science,withaprominentexamplebeingCondorcet’sJurytheorem,whereagroupofvotersbecomes
more likely to make the correct decision as the number of voters increases, if the probability of
each individual voting for the correct decision is greater than 0.5 (Mueller 1993). Similarly, Lau
and Redlawsk (2006) provide an excellent overview of the literature on voting “correctly” in the
American political context, while providing outlining how voting correctly means that individuals
should vote according to their informed preferences about policy.
40
Several of the American public’s top-ranked foreign policy goals involve
physical security against attack by international terrorists or hostile for-
eign states. But nearly as important in the minds of ordinary Amer-
icans...are a cluster of goals related to security of domestic well-being:
economic, social, and cultural security at home.
2
For example, this could explain why the American public is often resistant to
foreign aid that does not serve some instrumental purpose (Büthe, Major, and Souza
2012; Milner and Tingley 2013, 2015). In short, people want a positive return from
their taxes. If the state pursues a military conflict that people anticipate will not
serve to make their lives more secure or more wealthy, we should expect individuals
to resist such policies. Similarly, if a state agrees a trade agreement that individuals
do not perceive will benefit them, we should expect that they should oppose it.
People are more likely to pay attention to foreign policy debates, search for infor-
mation to acquire knowledge about the options available to their state, and update
their attitudes from this information when they perceive there are incentives to make
correct decisions. The incentives for correct decisions could be the result of the value
that an individual places on being correct, or their sense of responsibility to make
2
Page and Bouton (2006, 38-44) cited a 2002 CCMR/GMF poll as supporting evidence, which
revealed what Americans thought should be very important goals for U.S. foreign policy. Partici-
pants were asked whether each policy goal was “very important,” a “somewhat important foreign
policy goal,” or “not an important goal at all”. Of policies reported as being very important, na-
tionalsecurityprioritiesthatcouldmakepeoplefeelmoresecurewereatthetopofthelist, including
combating international terrorism (91 percent), and preventing the spread of nuclear weapons (90
percent). Of policies reported as very important, policies that would provide or protect economic
interests were next important, including protecting the jobs of American workers (85 percent),
stopping the flow of illegal drugs into the United States (81 percent), securing adequate supplies
of energy (75 percent), and controlling and reducing illegal immigration (70 percent). Policies that
included justice and assisting foreign populations who required assistance were considered consid-
erably lower priorities than policies aimed at providing security and domestic economic prosperity.
41
informed judgments about politics. This could be the result of individuals placing
a high value on a specific issue at hand, or it could be that they are intrinsically
motivated to pursue correct decisions out of a sense of civic responsibility as voters.
Incentives foraccuracy couldalso beshapedbythe contextin which anindividual
is asked to make a judgment about a foreign policy. For instance, foreign policy crises
could constitute such occasions, as people are motivated to learn about possible
responses (Brecher and Wilkenfeld 1997; Brecher et al. 2017; Feldman, Huddy, and
Marcus 2015, 35-37).
3
During crisis, people typically report that foreign affairs are
the most important issue facing the U.S. (Heffington, Park, and Williams 2017); and
people could perceive personal consequences of the crisis (Huddy et al. 2002). As
such, individuals could perceive there are direct incentives for them to pay attention
to debates about appropriate responses, search for information about the crisis, and
ultimately form opinions about the foreign policy based on their improved political
knowledge. During crisis, when there are incentives for people to make accurate
judgments, individualsaremorelikelytosearchforinformationaboutforeignaffairs.
4
People are also more likely to devote attention to international affairs when they
perceivethattheiropinionmighthavesomeinfluence. Forinstance, ifaforeignpolicy
is not debated before a decision is made by the executive or legislative branches of
3
Brecher and Wilkenfeld (1997, 3) define a foreign policy crisis for a state as “a situation with
three necessary and sufficient conditions deriving from a change in a state’s internal or external
environment. All three are perceptions held by the highest-level decision makers of the state actor
concerned: a threat to one or more basic values, along with an awareness of finite time for response
to the value threat, and a heightened probability of involvement in military hostilities.”
4
I frame this calculation positively, in that people are incentivized to make correct decisions.
However, onecouldalsodescribethecalculationasonewherepeoplearemotivatedtoavoidmistakes
(Lupia and McCubbins 1998).
42
government, there is little expected benefit for an individual to pay any attention to
that particular issue.
However, if there is a public debate, and an individual’s decision may have some
influence on the government’s decision, there may be some expected benefit from
paying attention to an issue and making the correct decision. Of course, just as
there is an infinitesimal probability than any one individual casts a decisive vote
(Downs 1957); it is unlikely than any one individual’s opinion will convince any
given legislator or leader to adopt their preferred policy. However, the appearance
of public debate, even if it is not consequential for the policy, may convince an
individual that it is worth paying attention to a foreign policy debate.
5
This theory can be expressed formally in a simple decision theoretic model that
explains the conditions under which people devote attention to an issue, and when
they are motivated to search for information. This model can be expressed as:
E(Return) =E(Benefit)−(E(AttentionCosts)+E(CognitionCosts)),
(3.1)
where E(Return) is the expected return of paying attention to a given stimu-
lus, E(Benefit) is the expected benefit of paying attention to a given stimulus,
E(AttentionCosts) denotes the expected costs of paying attention to a stimulus,
and E(CognitionCosts) are expected costs of processing and reconciling informa-
tion with one’s prior beliefs.
5
This could be analogous to an individual voting because of perceived civic obligations; or
because they do not want to be perceived as neglecting their duty as citizens.
43
The expected benefit of attention is:
E(Benefit) =
cd
X
i=1
(P
CAUSE
∗(CorrectDecision
i
)−P
PREVENT
∗(CorrectDecision
i
))
×(P
DecisionConsequence
),
(3.2)
where
cd
P
i=1
denotes the sum of all correct decisions, P
CAUSE
∗ (CorrectDecision
i
)
refers to the probability that attention will cause a correct decision, andP
PREVENT
∗
(CorrectDecision
i
) describes the probability that attention to an issue will prevent
a correct decision.
P
DecisionConsequence
refers to the probability that the individual’s decision matters
for the eventual foreign policy. For instance, if a foreign policy issue is not debated
beforeadecisionismadebytheexecutiveorlegislativebranchesofgovernment, there
islittleexpectedbenefitforanindividualtopayanyattentiontothatparticularissue.
However, if there is some public debate, and an individual perceives their opinion
may have some influence on the government’s decision, there may be some expected
benefit from paying attention to an issue and making the correct decision.
Expected attention costs can be described as:
E(AttentionCosts) =OpportunityCosts +InformationCosts. (3.3)
Opportunity costs are the costs of paying attention to a given stimulus, given that it
comes at the expense of all other possibly things an individual could pay attention
to. Information costs refers to all other costs associated with attention, which could
44
involve financial costs of information, such as the costs of buying a newspaper, but
also the time costs associated with paying attention to this information.
Finally, expected cognition costs can be described as:
E(CognitionCosts) =ComplexityCosts +ReconciliationCosts, (3.4)
where complexity refers to the difficulty of processing new information, and reconcil-
iation costs describes the cognitive costs of reconciling new information against one’s
prior beliefs.
3.3 Implications of the Model
There are some important observational implications from this simple decision the-
oretic model.
3.3.1 Incentives and Information Search
First, whereE(Return)> 0, individuals should be more likely to search for informa-
tion about foreign policy. If there is a large expected benefit from paying attention
to a given issue, individuals should devote attention to it, and search for further
information about it.
People could perceive the incentives for correct decisions as being contingent
on the context. On such occasions, one could expect that individuals change their
minds about the incentives for accuracy based on their understanding of the stakes
of the decision. In the context of international relations, one would expect these
45
changes in perceptions of incentives for accuracy during periods of crises, when the
expected return from attention is particularly large. Because the stakes are high,
people are motivated to search for information, which would not be the case if the
foreignpolicyatstakewasconsideredunimportantbyagivenindividual. Asaresult,
within-individual differences should emerge during periods of crisis in international
relations.
Perception of incentives for accuracy could also be driven by individual-level
characteristics that predetermine some types of individuals to perceive there are in-
centives for accuracy, when other individuals might not. Individual-level differences
could be the result of beliefs about civic duty, about the importance of being an
informed member of society, and based on an individual’s understanding of the for-
eign policy debate. In such circumstances, we should expect to observe differences
between individuals’ political behavior as the result of their perceptions of incentives
for accuracy. Individuals should search for information if they perceive that there
are incentives for accuracy and if it will help them make correct decisions.
On the other hand, whereE(Return)< 0, people are less likely to pay attention
to international affairs, and to search for information to make correct decisions. For
mostperiodsofforeignpolicy, thisisthestatusquo, wheretherearefewincentivesfor
individual members of the public to make correct decisions about foreign policy. In
the absence of such incentives, there is little expected return from investing attention
in international affairs, and little reward from acquiring more information about the
foreign policy debate.
If an individual perceives that they have little potential effect on the outcome
of a policy, it is less likely they will devote attention to the policy debate or search
46
for information in the pursuit of the best possible outcome. As a result, we should
expect that people are more willing to invest attention into a policy discussion when
they consider there is some potential for their voice to be influential. If there is
public debate, we should expect that people are more willing to devote attention to
a policy decision than in its absence.
Furthermore, the costs of cognition factor into individuals’ willingness to acquire
information to make correct decisions. If an individual anticipates that the burden
of processing new information is too high, they may decide to avoid searching for
information about a given policy issue. As a result, one should expect to see that
information search is more likely when people anticipate that processing new infor-
mation will not be difficult. As an example, one could expect that acquiring new
information about a country’s location on a map will be relatively simple, requiring
little cognitive effort to absorb new information. On the other hand, acquiring new
information about a country’s geopolitical situation would be more difficult, and in-
dividuals might only search for information if they perceive the costs of processing
new information would be relatively low.
The costs of reconciliation also likely factor in to individuals’ decision to search
for information about a given policy issue. If one anticipates that their belief system
might be challenged, they might avoid searching for information, and instead avoid
the possibility of being forced to reconcile their prior beliefs with new evidence that
challenges them. As a result, people with lower levels of attachment to political
parties or their political ideology might be more willing to search for new informa-
tion. Conversely, people who harbor strong attachments to their political party or
47
their political ideology are less likely to acquire additional information when they
anticipate that it will challenge their priors.
Building on Lupia and McCubbins (1998), I expect that an individual will de-
vote attention to international affairs when the expected benefit of their attention is
greater than the expected costs of paying attention to a stimulus, and the expected
costs of processing and reconciling new information with their prior beliefs. The ex-
pected benefit from attention is their expected probability that attention to a foreign
policy debate might help them make the correct decision about the best response,
and if they act to persuade their elected representatives to pursue this action, it
could help them achieve their goals of security and prosperity.
There must be significant benefits from paying attention to a foreign policy de-
bate, because there are significant costs of attention to international affairs. First,
there are opportunity costs of all the other things an individual could be paying at-
tention to (Lupia and McCubbins 1998). Second, there may be costs of information.
These could range from small amounts of money such as the cost of a newspa-
per. However, there could also be a psychological toll whjere reading about security
threats could lead to increased anxiety about the state of the world (Albertson and
Gadarian 2015; Huddy and Feldman 2011; Huddy et al. 2005; Huddy, Feldman, and
Weber 2007).
Furthermore, there are cognitive costs of absorbing and processing information
about international affairs. New information could be complex, forcing people to
exertextraeffortintoprocessingtheinformationastheyformattitudesaboutforeign
policy. Newinformationcouldalsorequirethatindividualsreconcileinformationthat
48
contradicts their prior beliefs, which could be problematic for people with strong
priors.
Taking this into account, when people expect that the return from their attention
is greater than its costs, they should be more likely to search for information about
foreign policy. If there is a large expected benefit from paying attention to a given
issue, individuals should devote attention to it, and search for information about the
policy at stake.
On the other hand, where people perceive the expected return from attention is
less than the costs of paying attention to a given stimuli, people are less likely to
pay attention to international affairs, and to search for information to make correct
decisions. For most periods of foreign policy, this is the status quo, where there are
few incentives for individual members of the public to make correct decisions about
foreign policy. In the absence of such incentives, there is little expected return from
investing attention in international affairs, and little reward from acquiring more
information about the foreign policy debate. As a result, the first hypothesis is:
Hypothesis 1 Incentives and Information Hypothesis
Individuals given incentives for accuracy are more likely to search for information
about international affairs than individuals not given these incentives.
3.3.2 Incentives and Political Knowledge
Ifincentiveschangepeople’sinformationsearchbehavior, Iexpectinformationcauses
people to become more knowledgeable about the foreign policy debate. Previous
49
studies show that people respond to financial incentives and additional time to im-
prove political knowledge (Prior and Lupia 2008); and people update their beliefs
in response to incentives, even if they are slightly less responsive than the perfect
application of Bayes’ rule (Hill 2017).
6
Building on these studies, I expect to find similar results when people are given
incentives for accuracy, they improve their political knowledge. I expect that all
individuals given incentives to correctly answer questions are more likely to increase
their political knowledge, and this is especially likely if people search for information
about each policy. I expect that incentives for accuracy will overcome partisanship,
as the expected benefit of the information is greater than the utility gained from an
erroneous partisan interpretation of the evidence.
When given incentives to make accurate decisions, people should search for in-
formation and use that to improve their political knowledge. I expect there will be
direct effects of incentives on political knowledge, but also mediated effects through
information search. This leads to the second set of hypotheses:
Hypothesis 2 Incentives and Knowledge Hypothesis
Individuals given incentives for accuracy are more likely to correctly answer questions
about international affairs than individuals not given these incentives.
Hypothesis 2a Incentives and Knowledge Mediation Hypothesis
6
According to Bayes’ rule, an individual’s beliefs are the product of their prior beliefs and new
information, and people should update their beliefs as they are exposed to new evidence. Over
time, this should result in people having attitudes that are consistent with evidence, even if their
prior attitudes stood in opposition to this evidence.
50
Individuals given incentives for accuracy are more likely to search for information
about international affairs, which makes them more likely to correctly answer ques-
tions about international affairs than individuals not given these incentives.
3.3.3 Incentives and Political Predictions
If incentives change people’s information search behavior, it is reasonable to expect
that information leads to increased accuracy of their predictions about international
affairs.
I expected that all individuals given incentives to correctly answer questions be-
come more likely to increase the accuracy of their predictions, and this was especially
likely if people searched for information about each policy. Previous studies show
that people respond to financial incentives and additional time to improve political
knowledge (Prior and Lupia 2008); and people update their beliefs in response to
incentives, even if they are slightly less responsive than the perfect application of
Bayes’ rule (Hill 2017).
7
Building on these studies, I expected to find that when people are given incentives
for accuracy, they become more accurate in their predictions about international
affairs. When given incentives to make accurate decisions, people should search for
information and use that to improve the accuracy of their predictions. I expected
there will be direct effects of incentives on predictions, but also mediated effects
through information search.
7
According to Bayes’ rule, an individual’s beliefs are the product of their prior beliefs and new
information, and people should update their beliefs as they are exposed to new evidence. Over
time, this should result in people having attitudes that are consistent with evidence, even if their
prior attitudes stood in opposition to this evidence.
51
I expected that the effect of the incentives would gradually decrease as they be-
comelesscertain. Previousresearchshowsthatpeoplerespondwelltocertainpayoffs
in the form of cash, but lotteries are less powerful, and could even demotivate in-
dividuals in experiments (Pforr et al. 2015; Warriner et al. 1996; Zheng, Gong, and
Pavlou 2017). Time horizons also matter, as prepaid incentives are more powerful
than payments promised at a later date (Mercer et al. 2015; Pforr et al. 2015; Scher-
penzeel and Toepoel 2012). Although people respond more to immediate returns, if
participants perceive the payoffs as large enough, they can respond favorably even
to payments promised at a later date (Hsu et al. 2017).
I expected that people who receive more certain incentives in the form of payoffs
for every correct prediction will be more accurate in their predictions about interna-
tional affairs than the other treatment groups. As such, the bonus treatment should
be the most powerful treatment with positive effects on the accuracy of predictions.
Building on previous scholarship, I expected that the lottery treatment will have de-
motivating effects, leading to less accurate predictions than individuals given other
treatment conditions. This leads to the third set of hypotheses:
Hypothesis 3 Incentives and Predictions Hypothesis
Individuals given incentives for accuracy make more accurate predictions about in-
ternational affairs than individuals not given these incentives.
Hypothesis 3a Incentives and Predictions Mediation Hypothesis
Individuals given incentives for accuracy are more likely to search for information
about international affairs, which helps them make more accurate predictions about
international affairs than individuals not given these incentives.
52
Hypothesis 4 Bonus Incentive Hypothesis
Individuals given incentives for accuracy in the form of a bonus for every correct
prediction are more likely to make accurate predictions about international affairs
than individuals not given these incentives.
Hypothesis 5 Lottery Incentive Hypothesis
Individuals given incentives for accuracy in the form of a ticket for a lottery for every
correct prediction are less likely to make accurate predictions about international
affairs than individuals not given these incentives.
3.3.4 Incentives and Foreign Policy Attitudes
Previous research suggests reasons for pessimism about the influence of new informa-
tion on political behavior, as people generally adopt partisan cues such as party ID
in low information surveys and experiments (Abramowitz 2010; Bartels 2002; Brady
and Sniderman 1985; Druckman, Peterson, and Slothuus 2013; Fiorina and Abrams
2008; Zaller 1992). However, if people search for information in a high information
environment, people might update their preferences according to the information
they acquire, independent of political ideology or party ID. For instance, Feldman,
Huddy, and Marcus (2015) found that people who consumed high amounts of polit-
ical information about Iraq became much less supportive of the 2003 Iraq War.
While people pursue correct decisions, what they consider those decisions to be
is likely shaped by their prior beliefs and their world view. Although I expect that
people are responsive to changes in the expected benefits of attention to an issue, it
is likely that partisanship continues to play an important role in attitude formation.
53
It is important to determine to what extent partisan affiliations matter, and the
conditions under which people update their attitudes in response to new information
in spite of their pre-existing identities and beliefs.
Where political parties have clearly defined positions, it is likely that political
partisans are more likely to take the position of their own party than those who do
not identify with a political party. Previous research demonstrates that when given
competing frames, partisans double down on their previous beliefs, becoming more
steadfast in their attitudes about a given policy (Druckman and Jacobs 2015; Jacobs
and Shapiro 2000; Katz 2000). It is reasonable to expect that partisans will continue
to adopt party positions, all else being equal.
Similarly, the costs of reconciliation might also determine how people process in-
formation that contradicts their prior beliefs. Partisan individuals who have strong
attachments to their party or their political ideology are less likely to be moved by
new information, as the costs of reconciling the information may prove to be too dif-
ficult. This is particularly true in the contemporary information environment, where
people can indulge partisan biases through consuming media that reinforces their
prior beliefs, while avoiding information that might challenge pre-existing beliefs.
People who indicate a large depth of feeling about their attachment to their party
are therefore less likely to move as a result of new information. Similarly, people at
either end of the political spectrum, individuals who identify as either very liberal or
veryconservativearelesslikelytomovetheiropinionsthanmoremoderatepartisans.
However, I expect that all individuals who perceive incentives to correctly answer
questions are more likely to move their position about a given foreign policy than
those who do not perceive these incentives. This is especially likely if people search
54
for information about each policy because this information might contribute towards
them changing their position. Finally, I expect that partisanship may be a less
powerful predictor of attitudes than incentives. If people are incentivized to search
for information to make accurate choices about foreign policy, this information may
override political partisanship. If this is true, political partisanship may be overcome
through incentives to make accurate decisions about foreign policy. This could be
manifest in direct effects of incentives on attitudes, or through information search as
people acquire new information, and update according to this information.
Prior scholarship has identified how information search processes are biased,
where people search for confirmatory evidence instead of seeking out contrary evi-
dence (Druckman, Fein, and Leeper 2012; Leeper 2014). Leeper (2014) found that
individuals’ information processing behavior, rather than information search itself,
played the stronger role in driving mass polarization. However, it is possible that
incentivized information search has a stronger effect on attitudes than information
processing. In particular, it is possible that if one is motivated enough to search
for information when there are incentives to make correct decisions, they are more
responsive to this information than they would be otherwise.
If people are incentivized to search for information to make accurate choices
about foreign policy, this information may overcome political partisanship. This
could be manifest in direct effects of incentives on attitudes, through information
search as people acquire new information and update their attitudes, or through
political knowledge as people update their attitudes based on what they learn about
international affairs. This leads to the final set of hypotheses:
Hypothesis 6 Incentives and Attitudes Hypothesis
55
Individuals given incentives to correctly answer questions about a policy are more
likely to move their position about the policy than individuals not given these incen-
tives.
Hypothesis 6a Incentives, Information, and Attitudes Mediation Hypothesis
Individuals given incentives to correctly answer questions about a policy become more
likely to search for information about foreign policy, and they are more likely to move
their position about the policy than individuals not given these incentives.
Hypothesis 6b Incentives, Knowledge, and Attitudes Mediation Hypothesis
Individuals given incentives to correctly answer questions about a policy become more
likely to correctly answer questions about foreign policy, and they are more likely to
move their position about the policy than individuals not given these incentives.
3.4 Definition of Key Constructs
In this section, I introduce and define the key concepts for tests of the theory of
incentives, information, andpublicopinionaboutforeignpolicy. Itisimportanttobe
clear and specific, as although there are similarities between several of the concepts,
they have important differences that have implications for empirical tests conducted
later in the dissertation. Table 3.1 presents a summary of the key constructs and
their definitions to be used throughout the dissertation. This section expands on the
rationale behind the definitions, and what is meant by each term.
56
3.4.1 Incentives for Accuracy
I argue that to understand public opinion about foreign policy, it is vital to consider
the context in which decisions about international affairs are made. I discuss this in
terms of incentives for accuracy, where people are motivated to make correct decisions
about foreign policy.
It is important to acknowledge that incentives for accuracy are not necessarily
perceived similarly across the entire population. Instead, people may perceive in-
centives for accuracy differently based on their personality type, location, education,
income, or any number of other personal characteristics. As such, incentives for ac-
curacy are important only insofar as they influence an individual’s response - if there
are consequences for a decision but an individual does not perceive them as such,
they are unlikely to have much effect on individual behavior.
Incentives for accuracy are essentially analogous to perception of consequences,
where people perceive there are important consequences of the foreign policy judg-
ment, where they stand to receive either benefits from correct decisions or costs from
incorrect decisions (Lupia and McCubbins 1998).
57
Table 3.1: Definition of Key Constructs.
Construct Definition
Incentives for Accuracy motivation to make correct decisions
Information data or facts about international relations and/or foreign policy
Information Search the active pursuit of information
Attention selective concentration on one stimuli at the expense of other stimuli
Political Knowledge the ability to accurately answer factual questions about international relations and/or foreign policy
Political Predictions the ability to accurately forecast events in international relations
Political Attitudes Attitudes expressed views or judgments that captures an individual’s perspective about foreign policy
58
3.4.2 Information
Information is a fundamental component of public opinion about foreign policy, but
much scholarship has neglected to study the conditions under which people search
for information about international affairs. Downs (1957, 79) defined information as
“data about the current developments in and status of those variables which are the
objects of contextual knowledge.” Information is considered easier and less costly to
acquire than knowledge, yet one “can be knowledgeable without being informed, or
informed without being knowledgeable, but he cannot interpret information without
contextual knowledge” (Downs 1957, 79). This is useful for the study of information
in politics, where information is context-dependent.
Accordingly, I define information in the context of international relations as data
or facts about international relations and/or foreign policy. However, it is important
to consider information as distinct from knowledge. As Lupia and McCubbins (1998)
said:
The belief that more information is necessarily better is usually based on
the assumption that information and knowledge are equivalent. If this
assumption is true, then more information cannot make you worse off.
However, for most information, this assumption is clearly erroneous.
As such, I suggest that information cannot be interpreted without some prior
knowledge about the circumstances about politics or the political issue at hand.
59
3.4.3 Information Search and Information Avoidance
If information is data or facts about international relations and/or foreign policy,
information search refers to the process through which an individual actively pursues
data or facts because they perceive it might give them some utility.
8
Information avoidance refers to the reverse process of information search, where
individuals actively eschew information about international affairs (Golman, Hag-
mann, and Loewenstein 2017). This is conceptually different from the absence of
information search - information avoidance involves actively avoiding information
even if it is available and useful for decision making.
3.4.4 Attention
While access to information is critical, there is no guarantee that people will use
information about international relations even if they have access to it. There is no
shortage of information available to people in 21st century democracies. However,
people choose not to choose to read the newspaper; watch the news on television;
read transcripts of statements by their city council-member, district representative,
governor, senator, or president; or read Nate Silver’s latest predictions on FiveThir-
tyEight.com, for example.
8
This is distinct from information exposure, which refers to the consumption of information
where it is not actively pursued, but nevertheless people are exposed to information about inter-
national affairs. An example of this could include the consumption of “soft news,” where people
become more knowledgeable about foreign policy in spite of not having actively looked for infor-
mation about foreign policy debates (Baum 2002b, 2005; Prior 2003). As such, information may
not be actively sought, but people could still learn about international affairs from the information
they consume.
60
Inshort, attentionisscarce(Broadbent2013;LupiaandMcCubbins1998;Popkin
1994; Schneider and Shiffrin 1977; Simon 1971; Stroud 2017). As such, an individual
has to actively make choices about where to exert their scarce attention. I define
attention as an individual’s selective concentration on one stimuli at the expense of
other stimuli on which an individual could otherwise devote their focus.
3.4.5 Political Knowledge
Political knowledge is a contested term. Some authors provide a variety of different
definitions, while others avoid the conceptual minefield by avoiding defining the
concept altogether. Delli Carpini and Keeter (1996, 10) defined political knowledge
to be “the range of factual information about politics that is stored in long-term
memory.” Barabas et al. (2014, 841-842) built on this earlier work by developing
a helpful framework for the evaluation of political knowledge questions with the
temporal dimension as either static or surveillance, and the topical dimension as
either general or policy-specific.
I build on the work of Gilens (2001), who identified that policy-specific knowledge
was critical to understanding people’s political judgments. I am agnostic about
whether it comes from their long-term memory but it is important that knowledge be
useful in the evaluation of the policy debate at hand. As a result, for the purposes of
this study, political knowledge is policy-specific. This is manifest in an individual’s
ability to accurately answer factual questions about international relations and/or
foreign policy. Political knowledge is the product of information, where data about a
particular issue contributes towards a better understanding of the issue, to the point
61
where one can make a reasoned judgment about the potential effects of different
policy options.
3.4.6 Political Predictions
Where I consider political knowledge to refer to an individual’s ability to correctly
answer questions about international affairs, political predictions are the ability to
accurately forecast events in international relations. Going one beyond an under-
standing of political facts, it requires participants to be able to foresee the conse-
quences of policies and make predictions that tie their hands. As a result, political
predictions are closely aligned to the forecasting project developed by Phil Tetlock
and his collaborators at the Good Judgment Project (Tetlock 1992; Tetlock and
Gardner 2016; Ungar et al. 2012), and are closer to the Lupia and McCubbins (1998)
definition of political knowledge as the “the ability to predict the consequences of
actions.”
3.4.7 Political Attitudes
In this dissertation, I consider political attitudes to be an individual’s view or judg-
ment that captures their perspective towards a foreign policy. Attitudes are typically
more deeply rooted in an individual’s predispositions, values, and world view than
simplyreflectingtop-of-the-headresponsesofparticipants(ZallerandFeldman1992).
Most public opinion scholarship involves the use of nationally representative pub-
lic opinion polls, mass surveys, or experiments using convenience or probability sam-
ples.
62
3.5 Operationalization of Variables
Table 3.2 indicates how the variables discussed in this chapter could be operational-
izedandmeasuredacrossobservationalandexperimentalresearch. Intheparagraphs
to follow, I briefly outline some recommendations about the operationalization of
these variables in public opinion and foreign policy research.
Context can be taken into account through the incorporation of incentives for
accuracy, which could change individuals’ demand for information. Experimentally,
scholars could use monetary incentives to manipulate the decision making environ-
ment to replicate the effects of changes in the geopolitical situation in the real world.
Analogous observational data could include the onset of foreign policy crisis, peri-
ods of public debate where an individual perceives their opinion could have some
influence, and public referenda where individuals’ votes directly affect foreign policy.
Attention can be measured through an individual’s concentration on one stimuli
at the expense of other stimuli. In experimental research, this can be measured
throughparticipants’timespentondifferentsectionsofthesurvey, theirperformance
in attention check questions, and their eye movements. In public opinion polls, self-
reports of how closely participants followed a particular foreign policy debate could
capture attention in an observational setting.
Information search and avoidance can be captured through literal information
search through online portals. In an experiment, this could be measured through
online behavior in the laboratory, or self-reports about information search.
63
Table 3.2: Operationalization of Constructs across Experimental and Observational Studies.
Construct Experimental Variable Observational Variable
Incentives for Accuracy Financial Incentives Onset of Foreign Policy Crisis, Public Debate, Public Referendums
Attention Attention to Survey or Stimuli Following of Foreign Policy Debate
Information Search/Avoidance Information Search/Avoidance Google Searches, Information Consumption
Political Knowledge Political Knowledge Battery Political Knowledge Battery
Political Predictions Political Predictions Battery Political Prediction Battery
Foreign Policy Attitudes Foreign Policy Attitudes Foreign Policy Attitudes
64
Observational studies could measure information search and avoidance through
aggregate-level Google Trends and information consumption through news websites.
However, longitudinal individual-level tracking of online information search is prefer-
able, allowing for the analysis of dynamic trends in response to events.
Ultimately, scholars of public opinion are most concerned with how the public
learns about foreign policy, their ability to make predictions about the effects of
foreign policy, and their attitudes about foreign policy. There are well-established
literatures in how to construct batteries of questions to capture political knowledge,
predictions, and attitudes about foreign policy.
However, for the study of the effects of information search and avoidance on these
dependent variables, it is important to have batteries of questions that are policy-
specific and relate to the context of the survey (Gilens 2001). As such, off-the-shelf
batteries are less useful than custom batteries of questions developed for the study
at hand.
3.6 Conclusion
To understand how people think about foreign policy, it is vital to understand the
conditions under which they are asked to make decisions. While previous literature
correctly identifies that people use information shortcuts and heuristics to form at-
titudes about foreign policy using little information, this work fails to address how
people respond when asked to make decisions under duress, when there are incentives
to make correct decisions.
65
In this chapter, I introduced a model of public opinion that takes this into ac-
count. I argue that to understand public opinion about foreign policy, it is necessary
to consider how perceptions of incentives influence political behavior. When people
perceive there are incentives for accuracy, they are more likely to search for informa-
tion, and use this information to improve their political knowledge, the accuracy of
their predictions, and ultimately the way they think about foreign policy.
While I introduce a general model in this chapter, there are many different ap-
plications of the theory to study the effects of incentives and information search on
publicopinionaboutforeignpolicy. Insubsequentchapters, Iconductaseriesoftests
using experimental and observational data to illustrate the empirical plausibility of
the model.
66
Chapter 4
Motivated Learning and Public
Opinion about Foreign Policy
For democratic states to function effectively, citizens should acquire information and
use it to make informed judgments, and ultimately hold elected officials accountable.
If citizens are uninformed, they cannot make good decisions at the voting booth or
be an effective constraint on foreign policymakers. This is an important concern for
all political debates, but the stakes are particularly high in foreign policy.
Arguably there are no more important decisions for domestic publics than deci-
sions related to international security. Conflict is costly, and the burdens of military
campaigns are suffered by domestic audiences as their taxes fund the military, their
security is under threat from opposing forces, and they provide the troops for the
war effort. Given the significant personal and financial stakes, the public should be
reticent about entering conflict unless it is necessary.
67
However, information about international affairs is also more costly and difficult
to acquire than information about domestic politics. Baum and Potter (2008, 2015)
demonstrated that political elites enjoy information advantages over the general pub-
lic who do not have access to the same internal briefings, classified documents, and
staff members to help them make sense of complex foreign policy debates. It is diffi-
cult for people to make sense of international affairs given the difficulty of acquiring
and processing information to make sense of a world beyond their own experience.
However, during foreign policy debates, the public is required to overcome these
barriers and form judgments about foreign policies that could have significant con-
sequences. When it matters most, are people capable of searching for information
and improving their political knowledge to make informed decisions about foreign
policy?
Furthermore, it remains an open question whether people respond to incentives
when making predictions about the future. Previous research suggests that even
experts find it difficult to make predictions about international affairs (Tetlock 1998,
1999, 2006). However, recent scholarship suggests that people are capable of mak-
ing accurate predictions about the world over a period of time (Mellers et al. 2015;
Tetlock 1998, 1992; Tetlock and Gardner 2016; Ungar et al. 2012). This begs the
question, if members of the public are capable of outperforming experts in predict-
ing international affairs, can this behavior be induced? Do incentives for accuracy
improve people’s ability to make predictions about international affairs?
Finally, ifattentionisascarceresource(LupiaandMcCubbins1998;Popkin1994;
Schneider and Shiffrin 1977; Stroud 2017), when are people motivated to learn about
international affairs? When do people search for information about international
68
affairs? Finally, how does information acquisition and improved political knowledge
influence public opinion about foreign policy?
Inthischapter, Ibuildonpriorresearchtotestatheoryofincentives,information,
and public opinion about foreign policy. I exploit the advantages of experimental
researchtotesthowpeoplerespondtoincentivesforaccuracy. Iexaminetheeffectsof
financial incentives on individuals’ information search, political knowledge, political
predictions, and foreign policy attitudes.
The results provide the foundation for an improved understanding of the role of
information in public opinion and international relations, demonstrating that people
are capable of improving their political knowledge when it matters most, and that
they use this knowledge to update their opinions about foreign policy. This has
important implications for the understanding of public opinion about foreign policy,
suggesting that the public is capable of informed decision making when they perceive
the stakes as high.
4.1 Experimental Design
The controlled environment of experiments allow scholars the ability to control for al-
ternative explanations by holding confounding factors constant. Random assignment
to treatment, if successful, ensures that one can examine the effects of the treatments
on the dependent variables. Due to these benefits, I employ an incentivized online
survey experiment to test the effects of incentives on information search, political
knowledge, and foreign policy attitudes.
69
The experiment featured a 4-group posttest only design, where three groups were
exposed to a treatment, and the other group received the control condition. The
foreign policy debates address eight of the most significant issues that the United
States currently faces. Security-related foreign policy debates include responses to
the threat of terrorism, the Ukraine conflict, the threat presented by North Korea,
and the Iran nuclear deal. Foreign policy debates not directly related to security
include climate change, policy responses to unauthorized immigration, U.S.-Russia
relations, and international trade with China.
1
4.1.1 Sample
298 participants were recruited through MTurk. The sample was limited to U.S. cit-
izens, but it is not a probability sample of the entire population. However, previous
research suggests that the MTurk platform produces samples that are more diverse
than other commonly used convenience samples such as undergraduates (Berinsky,
Huber, and Lenz 2012; Huff and Tingley 2015). MTurk also produces similar experi-
mentalresultsfromparticipantstoothercommonlyusedsamples(Casler, Bickel, and
Hackett 2013; Mullinix et al. 2015); the population does not differ from population-
based samples in unmeasurable dimensions (Levay, Freese, and Druckman 2016);
and MTurk participants perform well on maintaining attention throughout surveys
(Hauser and Schwarz 2016).
As such, while it is important to acknowledge the limitations of the sample, it
provides a useful means for studying the effects of incentives on political behavior
1
Further details about expectations for foreign policy attitudes given the dynamics of U.S.
politics in August 2017 are reported in Appendix B.9.
70
regarding information search and attitude formation. All participants were compen-
sated $0.75 for participating in the survey experiment, and any earned bonuses were
paid on top of this amount.
4.1.2 Procedure
Figure 4.1 outlines the design of the experiment. Appendix A reports the full design
including the text of the treatments, questions, and coding of variables.
First, participants were presented with a battery of demographic questions. After
a common introduction, participants were randomly assigned to receive a treatment
or the control condition. The bonus treatment indicated that participants would
receive a bonus of $0.50 for every question they correctly answered. However, one
could also consider that there is no predetermined and uniform utility for correct
decisions about foreign policy in the real world. For this reason, a second treatment
condition involved a single bonus payment of $4.00, which was given to participantsif
they correctly answered a randomly selected question. The third treatment awarded
participants a lottery ticket for every correct answer, which gave them an entry
into a lottery for a $400 cash prize. Participants assigned to the control condition
proceeded to the next stage of the survey, and they received no bonus payment for
correct answers.
Participants were asked a series of questions about foreign policy that form the
variables of interest in the study. These can be divided into three different groups.
71
Figure 4.1: The Experimental Design.
Demographics 1
Common In tro duction
Exp osure to
T reatmen t/Con trol
P olitical
Kno wledge Battery
Predictions Battery
Information
Searc h Battery
P olitical
Kno wledge Answ ers
P olitical
A ttitudes Battery
Demographics 2
Demographics 3
End of Surv ey
Random Assignmen t to T reatmen t
Randomized Order
Randomized Order
72
The eight political knowledge questions have correct answers, measured by the total
number of questions participants correctly answered.
2
While all financial incentives could lead to improvements in the accuracy of pre-
dictions, I expected that the bonus treatment would be most powerful because of the
guaranteed reward for correctly answering each prediction question, with the random
bonus treatment less powerful because of the introduction of risk and uncertainty
about each individual question. Finally, I expected that the lottery treatment would
be least powerful out of the three treatments given the low probability that accurate
responses would lead to a return on their efforts. Participants assigned to the con-
trol condition simply proceeded to the next stage of the survey, and they received
no bonus payment for correct answers.
From there, participants were asked a series of questions asking them to make
predictions about international affairs, which were adapted from Tetlock (2006) and
his collaborators’ Good Judgment project. Participants were required to indicate
a response to each dichotomous question, where the correct prediction would be
revealed a month from the date of the survey. Participants’ accuracy was measured
by the total number of questions participants correctly answered out of eight, as well
as their accuracy for individual questions.
Afterwards, participants were asked a series of questions relating to information
search. I asked participants to indicate whether they searched for information to help
them answer the questions in the survey. This dichotomous variable is the primary
2
Apretestof200MTurkworkerswasconductedpriortofieldingthefullexperimenttodetermine
the degree of difficulty. In general, people’s political knowledge was low. The mean number of
correct answers was 2.25 out of 8 questions, and no participants correctly answered all 8 questions.
73
measure of information search. The battery also included questions about the tools
participants employed to search for information, and their trust in the sources.
Participants were randomly presented with the answers to the political knowledge
questions before or after completion of the attitudes battery. Finally, participants
answered questions about their foreign policy attitudes. Each of the questions ad-
dressed a contemporary foreign policy debate in the U.S. Participants were asked to
indicate whether they agreed or disagreed with each statement on a 7-point likert
scale, whether they: strongly disagree (-3); disagree (-2); somewhat disagree (-1);
neither agree nor disagree (0); somewhat agree (1); agree (2); or strongly agree (3).
Table 4.1 presents the exact question wording of each statement as presented to
participants.
Table 4.1: Topics and Statement Wording for Foreign Policy Attitude Questions.
Foreign Policy Foreign Policy Attitudes
Muslim Ban The U.S. should ban entry to travelers from Syria, Iran,
Libya, Somalia, Sudan and Yemen.
Russia Sanctions The U.S. should continue to impose trade sanctions
against Russia in response to the Ukraine conflict.
Strike N. Korea The U.S. should launch a pre-emptive military strike
against North Korea.
Iran Nuclear Deal The U.S. should have signed the nuclear deal with Iran
in 2015.
Paris Agreement The U.S. should have withdrawn from the Paris
Climate Agreement.
Russia Relations The U.S. should seek to improve diplomatic relations
with Russia.
Path to Citizenship The U.S. government should provide a path to citizenship
for unauthorized immigrants to the United States.
China Trade The U.S. should impose greater trade restrictions on China
to protect American companies and American workers.
74
The order of the final two batteries of demographic questions was randomized
in case question ordering affects participants’ responses to questions about their
political beliefs, affiliations, and levels of interest in politics.
3
Participants were also
asked to complete two comprehension questions to ensure that they read the survey
carefully and accurately.
Finally, after completing all questions in the survey, participants were provided
with a code for reimbursement through MTurk. Participants received compensation
for their time, and any earned bonuses for the political knowledge questions on
August 7, 2017. Payment for the predictions questions was completed on September
7, 2017.
4.2 Results
The experiment revealed support for most of the hypotheses introduced in Chap-
ter 3, providing support of the theory of incentives, information search, and public
opinion about foreign policy. The results are presented using OLS regressions for
ease of interpretation, but the results are robust to using logistic regressions for the
dichotomous variables in the study. All regressions are conducted using robust stan-
dard errors. Commonly used political covariates are also reported for illustrative
purposes, including Party ID, political ideology including whether the participant is
politically moderate, and the amount of policy agreement.
4
3
There were no effects of the order of responses in the survey (not reported).
4
Full results including alternative specification of the models and additional covariates are
presented in Appendix B.
75
Figure 4.2: Determinants of Information Search.
Coefficient plot of the determinants of information search during the experiment.
Point estimates indicate the effect of each variable on information search with 95%
confidence intervals. The plot was created using the coefplot package in Stata
(Jann 2014).
4.2.1 Incentives and Information Search
As expected, the treatments had strong positive effects on information search and
attention to the survey. Figure 4.2 reports the determinants of information search.
First, the bonus treatment had large positive effects on the probability of an in-
dividual searching for information in the survey to answer the political knowledge
76
questions. Substantively, the effect of an individual receiving the bonus treatment
increased the incidence of information search by 0.218, holding all other covariates
constant. To put this in context, when additional covariates are reported, the ef-
fect of the bonus treatment was more than ten times greater than political interest,
and over four times more powerful than news consumption.
5
These results provide
support the Incentives and Information Hypothesis, suggesting that incentives for
accuracy increase people’s willingness to search for information. There were no sta-
tistically significant effects of Party ID, political ideology, or policy agreement with
either party on information search.
Because the measure of information search relies on a self-reported answer, there
is a possibility that people may not truthfully answer the question. To demonstrate
the plausibility of the mechanism and to alleviate fears participants did not answer
the question truthfully, I tested whether information search had a positive effect
on the amount of time spent on the survey, as one would expect if the mechanism
worked as expected. The full results of these tests are reported in Appendix B, but
searchingforinformationledtopeoplespendingoversixminuteslongeronthesurvey
than those who did not search for information. As a result, it appears as though
the information search variable worked as expected, increasing our confidence that
participants answered the question honestly.
5
Full experimental results are reported in Appendix B.
77
4.2.2 Incentives and Political Knowledge
As anticipated, there were significant positive effects of the treatments on political
knowledge, as participants exposed to incentives to make correct decisions sought in-
formation to improve their likelihood of getting questions correct. Figure 4.3 reports
the determinants of political knowledge. As expected, searching for information had
a large positive relationship with political knowledge, and searching for information
led to an average increase of 2.75 correct answers. As a point of comparison, in
models with additional covariates, political interest and news consumption increased
the number of correct answers by 0.237 and 0.231 respectively.
6
Both the bonus and random bonus treatments had positive effects on partici-
pants’ political knowledge score, increasing it by an average of 1.31 and 1.09 correct
answers respectively. Collectively, these results support the hypotheses that indicate
that incentives for accuracy leading to improvements in political knowledge about
international affairs.
6
Reported in full models in Appendix B.
78
Figure 4.3: Determinants of Political Knowledge.
Dependent Variable: Number of Correct Answers for Political Knowledge Questions.
Coefficient plot of the determinants of political knowledge during the experiment.
Point estimates indicate the number of political knowledge questions correctly
answered by participants with 95% confidence intervals. The plot was created using
the coefplot package in Stata (Jann 2014).
79
Table 4.2: Direct Treatment Effects and Mediation Effects of Incentives on the Number of Correct Answers
for Individual Political Knowledge Questions.
Mediating Variable: Information Search.
ACME Direct Effect Total Effect % Total Effect Mediated
Bonus Treatment 0.517 0.858 1.375 0.378
(0.161, 0.884) (0.413, 1.291) (0.834, 1.933) (0.378, 0.619)
Random Bonus Treatment -0.063 0.567 0.504 -0.117
(-0.420, 0.299) (0.098, 1.020) (-0.100, 1.090) (-0.887, 0.628)
Lottery Treatment -0.182 -0.556 -0.738 0.243
(-0.541, 0.178) (-1.026, -0.100) (-1.337, -0.150) (0.133, 1.011)
The results were calculated using 1000 simulations, and 95% confidence intervals are reported in brackets.
The table was created using the mediation package in Stata (Hicks and Tingley 2011).
80
Figure 4.4: Determinants of Individual Political Knowledge Responses.
Dependent Variable: Correct Answers for Individual Political Knowledge Questions.
Coefficient plot of the determinants of correct answers for individual political
knowledge questions in the experiment. Point estimates indicate the effect of each
variable on the accuracy of the response with 95% confidence intervals. The plot
was created using the coefplot package in Stata (Jann 2014).
Table 4.2 presents the results of causal mediation analysis using the techniques
developed by Imai and his collaborators to determine the mediating effects of in-
formation search on political knowledge (Hicks and Tingley 2011; Imai, Jo, and
Stuart 2011; Imai, Tingley, and Yamamoto 2013; Imai et al. 2011; Imai, Keele, and
Yamamoto 2010). The results indicate that the effect of the bonus treatment was
81
strongly mediated by information search, constituting almost 38 percent of the aver-
age treatment effect. Similarly, information search mediated the lottery treatment,
although this had a negative effect on political knowledge. Information search damp-
ened the effects of the random bonus treatment on political knowledge, although the
average treatment effect remained positive. Collectively, the results present support
for the Incentives and Knowledge Mediation Hypothesis. Especially for the bonus
treatment, information search was a powerful mediating variable that helps explain
the treatment effects of that particular incentive.
Finally, Figure 4.4 presents results for individual political knowledge questions to
determine whether there are heterogenous treatment effects depending on the nature
of individual political knowledge questions. There are three important implications
of the figure. First, information search was a significant predictor of accuracy for
all individual political knowledge questions, irrespective of the issue area. There
was a considerable amount of variation of the magnitude of the relationship though,
ranging from improving accuracy by 0.145 regarding the countries that supported
the Iran nuclear deal, to increasing accuracy by 0.534 for the number of unauthorized
immigrants in the United States.
Second, the bonus treatment had positive effects on accuracy for individual po-
litical knowledge questions, but the effects were not statistically significant for the
questions on the Paris Agreement, the number of deaths in Ukraine, and the current
range of North Korean missiles. Similarly, the random bonus treatment had statis-
tically significant positive effects for a handful of questions, but not for the number
of deaths from Jihadist terrorism in the U.S., the number of deaths in Ukraine, the
82
name of the Special Counsel appointed to oversee Russian interference in the 2016
U.S. election, and the range of North Korean missiles.
Finally, it is important to note that alternative explanations for accuracy in po-
litical knowledge did not hold up. Party ID had no relationship with the accuracy
of responses, and there were few meaningful effects of policy agreement with either
major political party. Political ideology only had a positive relationship with accu-
racy about the U.S.-China trade deficit, with a one unit increase in conservatism
associated with an increase in accuracy by 0.065. The results indicate that moti-
vated reasoning or partisanship had little effect on political knowledge, despite the
experiment having been designed to elicit these biases if present.
In sum, the results present support for all three knowledge hypotheses. Incentives
for accuracy, particularly in the form of the bonus treatment, had positive effects on
political knowledge. However, these effects are heavily mediated by information
search, where incentives for accuracy lead people to search for information, and use
that information to improve their political knowledge.
4.2.3 Incentives and Political Predictions
The experiment revealed mixed support for the predictions hypotheses. Against ex-
pectations, there were no direct positive effects of financial incentives on the accuracy
of correct political predictions.
However, the lottery treatment had negative effects on correct predictions, sug-
gesting that it had a demotivating effect on participants in that treatment group.
Information search strongly mediated the effect of the bonus treatment on the num-
ber of accurate predictions, but this effect did not occur for other treatment groups.
83
Overall, the results indicate support for some elements of the theory, but they also
suggest several areas for further research.
Figure 4.5 reports the determinants of correct predictions. The figure indicates
that the bonus treatments had no statistically significant positive effects on the
accuracy of predictions, meaning that the null hypothesis of no positive effect of
incentives cannot be rejected.
However, the lottery treatment had statistically significant negative effects on the
number of correct predictions, suggesting that participants were demotivated by the
treatment. As mentioned earlier, this finding matches previous research that sug-
gests that lottery incentives lead to poor performance in surveys and experiments,
as participants respond negatively to the uncertainty of their efforts being rewarded.
Substantively, receiving the lottery treatment decreased the number of correct pre-
dictions by 0.45.
As expected, searching for information had large direct positive effects on the
accuracy of predictions, and the average treatment effect on participants’ scores was
an increase of 0.30 correct predictions. This supports the theory that searching for
information increases the accuracy of predictions.
To determine whether information search mediated the effects of the treatments
on the accuracy of participants’ predictions about international affairs, I conducted
causal mediation analysis using the techniques developed by Imai and his collab-
orators (Hicks and Tingley 2011; Imai, Jo, and Stuart 2011; Imai, Tingley, and
Yamamoto 2013; Imai et al. 2011; Imai, Keele, and Yamamoto 2010). Table 4.3
presents the results of this analysis.
84
Figure 4.5: Determinants of Correct Predictions.
Dependent Variable: Number of Correct Answers for Prediction Questions.
Coefficient plot of the determinants of political knowledge during the experiment.
Point estimates indicate the number of political knowledge questions correctly
answered by participants with 95% confidence intervals. The plot was created using
the coefplot package in Stata (Jann 2014).
85
Table 4.3: Direct Treatment Effects and Mediation Effects of Incentives on the Number of Correct Answers
for Prediction Questions.
Mediating Variable: Information Search.
ACME Direct Effect Total Effect % Total Effect Mediated
Bonus Treatment 0.058 0.080 0.066 0.241
(0.003, 0.145) (-0.332, 0.343) (-0.279, 0.408) (-5.458, 4.953)
Random Bonus Treatment -0.003 0.352 0.349 -0.008
(-0.057, 0.050) (0.014, 0.684) (0.006, 0.696) (-0.052, -0.003)
Lottery Treatment -0.019 -0.478 -0.497 0.037
(-0.079, 0.023) (-0.814, -0.147) (-0.835, -0.159) (0.022, 0.108)
The results were calculated using 1000 simulations, and 95% confidence intervals are reported in brackets.
The table was created using the mediation package in Stata (Hicks and Tingley 2011).
86
The table indicates that the effect of the bonus treatment was strongly mediated
by information search, constituting over 24 percent of the average treatment effect
on the number of correct predictions. However, information search did not mediate
the effects of the random bonus and lottery treatments, a result that is logically
consistent with the result that neither of these treatment conditions led to increased
information search relative to the control condition.
Taken together, the results indicate some support for the Incentives and Predic-
tions Mediation Hypothesis, but only when the incentives are large enough to prompt
information search. For the participants provided with the bonus treatment, infor-
mation search was a powerful mediating variable that helps explain the treatment
effects of that particular incentive on the number of correct predictions.
4.2.4 Incentives and Foreign Policy Attitudes
The results show some support for the hypotheses relating to information and foreign
policy attitudes. There are three important implications from the experiment.
First, political knowledge had significant effects on attitudes relating to security.
Increases in political knowledge led individuals to become less supportive of the
Muslim Ban and a pre-emptive strike against North Korea, but more supportive
of sanctions against Russia in response to the Ukraine conflict and the American
support for the Iran nuclear deal. Table 4.4 reports the results of the experiment
relating to this set of political attitudes.
However, political knowledge had little effect on questions not directly related to
security matters, illustrated in Table 4.5. This set of questions includes statements
about the U.S. withdrawal from the Paris Climate Agreement, a path to citizenship
87
for unauthorized immigrants, the state of the U.S.-Russia relationship, and trade
restrictions against China.
This was an unexpected pattern of results that emerged from the experiment, but
it supports previous research that argues that security is distinct from other politics,
and is especially powerful as a rhetorical tool (Balzacq 2010; Buzan, Waever, and
de Wilde 1997; Hayes 2015). While improvements in political knowledge appear to
inform attitudes about security, people do not appear to update attitudes about
other foreign policies based on this knowledge.
Second, the results challenge arguments that Party ID predicts attitudes about
foreign policy. However, there were small but consistent effects of partisan policy
agreement on attitudes towards both security and non-security related foreign poli-
cies. The effects were also in the expected direction for these policies. For example,
agreement with Republican policies had a positive effect on support for the Muslim
Ban, and a negative effect on support for the Iran Nuclear deal. Agreement with
Democratic Party policies had negative effects on attitudes towards the Muslim Ban,
and positive effects on support for Russian sanctions and the Iran Nuclear deal.
Third, people with moderate political ideology were much more likely to move
their foreign policy attitudes. Moderates supported the Muslim Ban and a strike
against North Korea, and opposed sanctions against Russia. However, moderates
were no more or less likely to shift their attitudes about issues that did not deal
directly with security. These results provide some preliminary evidence that moder-
ates were more likely to move than ideologues, but only where security matters were
involved.
88
Table 4.4: Attitudes about Foreign Policies related to Security.
(1) (2) (3) (4)
Muslim Ban Russia Sanctions Strike N.Korea Iran Deal
Bonus Treatment 0.324 -0.258 0.366 0.035
(0.301) (0.251) (0.255) (0.253)
Random Bonus Treatment 0.588
∗
-0.081 0.236 0.308
(0.291) (0.248) (0.217) (0.259)
Lottery Treatment 0.603
∗
0.196 0.668
∗∗
0.207
(0.266) (0.244) (0.252) (0.256)
Political Knowledge -0.149
∗∗
0.212
∗∗∗
-0.100
+
0.122
∗
(0.056) (0.047) (0.051) (0.048)
Information Search 0.222 -0.437
+
0.027 -0.099
(0.239) (0.235) (0.218) (0.218)
Information Exposure 0.017 -0.016 0.158 -0.043
(0.184) (0.170) (0.168) (0.173)
Democrat -0.166 0.047 0.085 -0.066
(0.260) (0.264) (0.218) (0.233)
Republican 0.376 0.148 0.867
∗
0.490
(0.388) (0.384) (0.407) (0.314)
Moderate 0.506
∗
-0.667
∗∗
0.426
∗
0.031
(0.216) (0.206) (0.196) (0.189)
Dem. Policy Agreement -0.018
∗∗∗
0.010
∗
-0.002 0.018
∗∗∗
(0.005) (0.004) (0.004) (0.004)
Rep. Policy Agreement 0.022
∗∗∗
-0.011
+
0.011
∗
-0.031
∗∗∗
(0.006) (0.006) (0.005) (0.005)
Constant -0.540 0.699
+
-2.238
∗∗∗
-0.275
(0.422) (0.390) (0.307) (0.359)
Observations 298 298 298 298
R
2
0.449 0.243 0.250 0.413
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
89
Table 4.5: Attitudes about Foreign Policies not related to Security.
(1) (2) (3) (4)
Paris Agreement Path to Citizenship Russia Relations China Trade
Bonus Treatment 0.048 0.014 -0.034 0.475
+
(0.241) (0.280) (0.280) (0.250)
Random Bonus Treatment 0.026 -0.254 0.245 0.134
(0.234) (0.274) (0.268) (0.258)
Lottery Treatment 0.197 0.076 0.220 0.392
(0.259) (0.279) (0.276) (0.259)
Political Knowledge -0.006 -0.014 0.037 -0.007
(0.044) (0.061) (0.053) (0.053)
Information Search 0.117 -0.089 -0.024 0.102
(0.188) (0.262) (0.251) (0.237)
Information Exposure 0.078 -0.181 0.150 0.270
(0.170) (0.187) (0.180) (0.171)
Democrat 0.175 -0.176 0.534
∗
0.127
(0.211) (0.258) (0.259) (0.239)
Republican 0.682 0.087 -0.215 0.402
(0.422) (0.391) (0.367) (0.277)
Moderate 0.213 -0.340 0.023 0.032
(0.219) (0.221) (0.197) (0.177)
Dem. Policy Agreement -0.029
∗∗∗
0.025
∗∗∗
-0.017
∗∗∗
-0.005
(0.004) (0.005) (0.005) (0.004)
Rep. Policy Agreement 0.021
∗∗∗
-0.027
∗∗∗
0.014
∗
0.011
∗
(0.006) (0.006) (0.005) (0.005)
Constant -0.908
∗
0.488 0.631
+
-0.375
(0.383) (0.422) (0.378) (0.370)
Observations 298 298 298 298
R
2
0.537 0.430 0.145 0.140
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
90
Table 4.6: Direct and Mediated Effects of the Bonus Treatment on Political Attitudes (Mediating Variable:
Political Knowledge).
Muslim Ban Russia Sanctions Strike N. Korea Iran Deal Paris Agreement Path to Citizenship Russia Relations China Trade
ACME -0.105 0.169 -0.083 0.114 -0.008 -0.026 0.039 -0.006
(-0.222, -0.017) (0.057, 0.304) (-0.184, -0.007) (0.022, 0.227) (-0.090, 0.075) (-0.122, 0.060) (-0.047, 0.137) (-0.088, 0.077)
Direct Effect -0.099 -0.288 0.061 -0.153 -0.025 0.09 -0.201 0.298
(-0.528, 0.318) (-0.683, 0.095) (-0.324, 0.437) (-0.550, 0.234) (-0.418, 0.358) (-0.341, 0.509) (-0.622, 0.208) (-0.096, 0.682)
Total Effect -0.204 -0.119 -0.021 -0.153 -0.032 0.064 -0.162 0.292
(-0.622, 0.206) (-0.486, 0.292) (-0.398, 0.345) (-0.413, 0.368) (-0.411, 0.340) (-0.352, 0.469) (-0.561, 0.251) (-0.087, 0.665)
% Total 0.355 -0.608 0.251 -0.338 0.024 -0.084 -0.143 -0.018
Effect Mediated (-4.213, 6.027) (-13.642, 10.657) (-6.240, 8.782) (-5.762, 11.616) (-0.609, 0.605) (-2.759, 2.154) (-2.923, 2.147) (-0.193, 0.161)
Table 4.7: Direct and Mediated Effects of the Random Bonus Treatment on Political Attitudes (Mediating
Variable: Political Knowledge).
Muslim Ban Russia Sanctions Strike N. Korea Iran Deal Paris Agreement Path to Citizenship Russia Relations China Trade
ACME -0.096 0.129 -0.062 0.081 -0.006 -0.010 0.019 0.010
(-0.208, -0.014) (0.030, 0.249) (-0.150, 0.000) (0.007, 0.178) (-0.075, 0.061) (-0.087, 0.061) (-0.050, 0.099) (-0.056, 0.082)
Direct Effect 0.273 -0.044 -0.116 0.229 -0.055 -0.284 0.188 -0.174
(-0.149, 0.682) (-0.434, 0.335) (-0.496, 0.254) (-0.163, 0.609) (-0.443, 0.322) (-0.708, 0.127) (-0.226, 0.592) (-0.563, 0.206)
Total Effect 0.177 0.084 -0.178 0.309 -0.061 -0.295 0.208 -0.163
(-0.241, 0.581) (-0.286, 0.488) (-0.555, 0.184) (-0.064, 0.711) (-0.437, 0.309) (-0.706, 0.108) (-0.192, 0.609) (-0.539, 0.210)
% Total -0.330 0.479 0.238 0.239 0.019 0.031 0.066 -0.040
Effect Mediated (-6.996, 4.520) (-12.566, 6.883) (-2.743, 3.830) (-1.559, 2.970) (-0.521, 0.628) (-0.314, 0.355) (-1.137, 0.732) (-0.703, 0.556)
Table 4.8: Direct and Mediated Effects of the Lottery Treatment on Political Attitudes (Mediating Variable:
Political Knowledge).
Muslim Ban Russia Sanctions Strike N. Korea Iran Deal Paris Agreement Path to Citizenship Russia Relations China Trade
ACME 0.071 -0.116 0.047 -0.077 0.003 0.013 -0.024 -0.008
(0.002, 0.171) (-0.247, -0.021) (-0.007, 0.132) (-0.180, -0.006) (-0.061, 0.069) (-0.054, 0.088) (-0.102, 0.037) (-0.076, 0.053)
Direct Effect 0.314 0.301 0.479 0.097 0.173 0.152 0.152 0.202
(-0.109, 0.726) (-0.089, 0.681) (0.101, 0.847) (-0.297, 0.480) (-0.216, 0.552) (-0.275, 0.567) (-0.265, 0.558) (-0.189, 0.583)
Total Effect 0.385 0.185 0.526 0.02 0.176 0.165 0.128 0.194
(-0.038, 0.805) (-0.207, 0.573) (0.150, 0.907) (-0.375, 0.410) (-0.211, 0.561) (-0.261, 0.595) (-0.284, 0.536) (-0.194, 0.576)
% Total 0.173 -0.413 0.088 -0.239 0.012 0.048 -0.092 -0.031
Effect Mediated (-0.688, 1.429) (-5.935, 5.365) (0.051, 0.301) (-6.937, 5.846) (-0.132, 0.174) (-0.660, 0.570) (-1.846, 1.146) (-0.396, 0.365)
The results were calculated using 1000 simulations, and 95% confidence intervals are reported in brackets.
The tables were created using the mediation package in Stata (Hicks and Tingley 2011).
91
The key result regarding incentives and foreign policy attitudes is that people
who answered questions correctly in the survey think differently about foreign policy
than other individuals. However, incentives to make correct decisions appear to have
few direct effects on foreign policy attitudes.
Mediation analysis demonstrates that political knowledge had an important me-
diating effect on attitudes about foreign policy. Tables 4.6, 4.7, and 4.8 present
the direct and mediated effects of the treatments on foreign policy attitudes, me-
diated by political knowledge as measured in the experiment. The results indicate
that political knowledge was a relatively powerful mediator for responses about the
security-related foreign policies across all treatment groups, but it explained little of
the variance regarding the other four foreign policy issues.
Specifically, politicalknowledgeexplainedbetween25and61percentoftheeffects
of the bonus treatment on attitudes relating to security. The effects of political
knowledge as a mediating variable for the random bonus treatment were also strong,
explaining between 23 and 48 percent of the effects of this treatment on security-
related foreign policies, while political knowledge also explained between 8 and 41
percent of the effects of the lottery treatment on these policies. However, these
mediating effects did not extend to foreign policies not directly related to security.
Instead, very little of the variance is explained by political knowledge.
Collectively, theresultsindicatethatpeople’sacquisitionofpoliticalknowledgein
response to financial incentives has effects on attitudes about security, where people
use their knowledge base to form opinions about foreign policy. However, where
security is not at stake, people do not use knowledge to inform their attitudes about
92
foreign policies. In such circumstances, commonly used shortcuts may trump the
influence of political knowledge on foreign policy attitudes.
4.3 Discussion
Although this chapter presents an intuitive theory of motivated learning and pub-
lic opinion about foreign policy, it challenges previous research. Political scientists
have generally questioned the ability of the public to acquire knowledge about pol-
itics to learn what they need to know to make informed decisions. The evidence
presented in this chapter provides compelling reasons to be more optimistic about
citizen competence when making important decisions about foreign policy.
First, the data suggest that people are capable of searching for information about
international affairs and using this information to improve their political knowledge
when they have incentives for accuracy.
Second, this chapter provides evidence that people can become more accurate
in their predictions about international affairs when they are given incentives for
accuracy. When provided with financial incentives with guaranteed rewards for ac-
curate predictions, participants became more likely to search for information, which
improved their accuracy. However, the lottery treatment had direct negative effects
on accuracy, demotivating participants assigned to the treatment.
Finally, the chapter demonstrates that motivated learning affects public opinion
about foreign policy, as people update their attitudes about security-related foreign
policy. People can fulfill their democratic obligations as an informed citizenry that
can produce informed decisions using available information when it matters most.
93
This is a significant contribution to the understanding of public opinion about
foreign policy, and it presents many further opportunities to build on the results of
this chapter. Future research should pay greater attention to the conditions under
which people have incentives for accuracy, how these influence information search
and individuals’ political knowledge, and ultimately how this process affects public
opinion about foreign policy.
94
Chapter 5
The Microlevel Foundations of
Attention to International Affairs
The previous chapter presented evidence from an incentivized online survey exper-
iment to demonstrate how people respond to incentives to search for information,
and use that to 1) improve their knowledge about international affairs; 2) make
predictions about international affairs; and 3) update their attitudes about foreign
policy. However, one could argue that people behave differently in the controlled
environment of an experiment than they do in real life. Furthermore, the use of
incentives in the experiment features a scenario where every participant received the
same incentives as other participants assigned to their treatment group. Outside of
the laboratory, people might have different perceptions of incentives to make correct
decisions when forming attitudes about foreign policy. To address this concern, the
following three chapters use observational data to complement the findings from the
incentivized online survey experiment.
95
In this chapter, I argue that an improved understanding of the dynamics of at-
tention is essential to better understand public opinion about foreign policy. While
scholars have developed a thorough understanding of how people form political at-
titudes in low-information environments, this scholarship has not considered when
and how people might seek information about international relations, and how this
influences people’s attitudes about foreign policy.
To address this significant shortcoming, I use multilevel regression to estimate
how being a registered voter influences individual attention to foreign policy debates
over U.S. military action in Syria. The chapter demonstrates that registered voters
are systematically more likely to search for information about international affairs
than other individuals. The results have important implications for the study of pub-
lic opinion and international relations, suggesting that voters are capable of informed
decision making, and that previous studies may have underestimated the willingness
of voters to fulfill their democratic obligations.
This chapter proceeds in four sections. First, I introduce my theory of voting,
attention, andinternationalaffairs. Second, Ioutlinethedataandmethodsemployed
in this study. The third section introduces results from multilevel regression of public
opinion polls taken during debates over U.S. military action in Syria. The chapter
concludes with a brief summary of the chapter’s contribution to the understanding
of public opinion and foreign policy.
96
5.1 The Conscientious Voter
Much has been written about the motivations of people to turn out and vote in
elections. For instance, despite the likelihood of people casting the deciding vote,
people still vote out of a sense of civic duty, or in response to social pressure (Green
and Gerber 2008). Yet, less is known about how an individual’s type as a voter or
non voter influences their ability to make informed judgments about foreign policy.
In this section, I introduce a theory of the conscientious voter, a type of person who
pays attention to international affairs out of intrinsic civic motivations.
Previous research has identified how people may vote out of a sense of civic duty
(Blais 2000; Downs 1957; Riker and Ordeshook 1973). Other work has identified
how intrinsic motivations may be complemented by other-regarding actions, where
people vote in the expectation that their turnout records can be made public (Cial-
dini and Goldstein 2004; Lerner and Tetlock 1999). Furthermore, field experiments
demonstrate that people respond to social pressure to vote (Green and Gerber 2015).
While this research illustrates the motivations of voters and what compels them to
exercise their right to vote in elections, questions remain about whether the same
types of people who vote behave differently to the rest of the population in their
attention to international affairs.
In particular, if there are intrinsic motivations to voting, one should expect that
similar motivations propel individuals to fulfill their perceived civic duty elsewhere.
For instance, if someone votes out of a sense of duty, they are also likely to pay
attention to policy debates to ensure they make an informed decision at the polls. In
97
short, there are reasons to believe that voters are conscientious across many different
aspects of civic life, not just at the ballot box.
I expect that an individual will devote attention to international affairs when
the expected benefit of their attention is greater than the expected costs of pay-
ing attention to a stimulus, and the expected costs of processing and reconciling
new information with their prior beliefs. I expect that voters have a systematically
greater perception of the expected benefit of attention to international affairs than
non voters. This should be reflected in the fact that holding everything else con-
stant, including the incentives to make correct decisions and party ID, voters are
more likely to pay attention to international affairs than non voters.
In contrast, people who do not vote are likely to perceive the expected benefit of
their attention to international affairs as low. As a result, they are less likely than
voterstopayattentiontoforeignpolicydebatessothattheycanmakeinformeddeci-
sions about foreign policy. Building on the hypotheses introduced earlier, I introduce
an additional hypothesis to test how voter type influences individuals’ attention to
international affairs:
Hypothesis 7 Attention Hypothesis
Registered voters are more likely to pay attention to international affairs than other
individuals.
98
5.2 Data and Methods
5.2.1 Case Selection: U.S. Public Opinion about Military
Action in Syria, 2013
American public opinion about Syria represents a good case for testing the theory
of the conscientious voter for several reasons. In particular, there is variation in
the incentives to make correct decisions, as there are two distinct periods for which
one can draw distinctions: 1) hypothetical responses to al-Assad’s use of chemical
weapons before August 2013; and 2) a period of intense public discussion and debate
after it was apparent that al-Assad had used chemical weapons, ranging from late
August to early September 2013.
In the first period, there were arguably few extrinsic incentives for the public to
make correct decisions because public opinion was inconsequential. In the second
period, during the public debate, there were arguably greater extrinsic incentives
for the public to make the correct decision as politicians contemplated a vote in
the U.S. House of Representatives, and there was time for people to lobby their
representatives.
This allows the researcher to evaluate the effects of voter type on attention to
international affairs during both a period where intervention in Syria was largely
hypothetical, and during a period of sustained political debate about the most ap-
propriate response to chemical weapons in August and September 2013. If voter
type explains attention to international affairs, this should predict attention to Syria
across all periods.
99
5.2.2 Data
Dependent Variable: Attention to Syria
Individual-level data for the independent variable is provided by three Pew Research
Centersurveys(PewResearchCenterandthePress2013a,b,c).
1
Participantsinthree
surveys were asked, “Did you follow the news story about possible U.S. airstrikes in
Syria and diplomatic efforts to have Syria give up control of its chemical weapons
very closely, fairly closely, not too closely or not at all closely?” Participants were
given a 4-point likert scale to indicate their response. I removed nonresponses and
recoded to range from 1) not at all closely; 2) not too closely; 3) fairly closely; to
4) very closely. In the absence of better measures, I use this variable to measure
people’s attention to the news story.
2
There was considerable movement in the different variables of interest, as people
responded to the changing foreign policy crisis by increasing their attention to the
crisis and updating their attitudes. Figure 5.1 illustrates that attention to the Syria
news story increased considerably from April to September, as people followed the
situation in Syria increasingly closely as the potential for conflict with Syria became
more likely.
1
Further information about the surveys is available to view in Appendix C.2.
2
Unfortunately, there is no data to measure information search at the individual level with
observational data, so this is not tested using public opinion poll data. The Pew surveys ask
participants to indicate how closely they have followed the Syrian news story in three different polls
during the course of 2013, a self-reported measure of attention. Behavioral measures of information
search and attention would be preferable, but in their absence I make use of this imperfect measure
of attention.
100
Figure 5.1: Attention to the Syria News Story Over Time.
Coefficient plot of attention to Syria in three different Pew surveys during 2013.
Point estimates indicate how closely participants followed the situation in Syria,
with responses ranging from 1 (not at all closely) to 4 (very closely). The
differences between each point estimate are statistically significant at a 95% level.
The plot was created using the coefplot package in Stata (Jann 2014).
101
Independent Variable: Voter Type
The independent variable is whether the voter was a registered voter. Participants
were asked, “Which of these statements best describes you?” Available responses
included: 1)Areyouabsolutelycertainthatyouareregisteredtovoteatyourcurrent
address? 2) Are you probably registered, but there is a chance your registration
has lapsed 3) Are you not registered to vote at your current address? or 4) Don’t
know/refused. I recoded the variable so that it became a dichotomous variable where
1 indicated the participant was certain they are registered to vote, and 0 indicated
all other responses. Data was provided by the three Pew Research Center surveys
(Pew Research Center and the Press 2013a,b,c)
Individual-level Control Variables
Some other individual-level variables could explain attention to the Syria crisis, so
they are included in the analysis in order to rule out alternative explanations. First,
support for the U.S. military action could predict attention to the crisis, as people
continue to pay attention to the crisis after they have formed an opinion about the
appropriate foreign policy. Unfortunately, the exact question wording varied over
time, but each question asked participants about their opinion about U.S. military
102
action in response to the use of chemical weapons.
3
All responses are recoded to
indicate 0 for oppose, 1 for favor, and nonresponses were not included in the analysis.
A host of control variables are included to control for confounding factors that
could have effects on individual attitudes about foreign policy. First, party ID is
included in order to control against partisan affiliations having effects on attention to
the Syria crisis. These are operationalized as dummy variables where 1 indicates the
participant is a member of a political party (Republican, Democrat, or Independent),
and 0 indicates they are not a member of that party.
Attentioncouldalsosystematicallyvaryaccordingtowhetheranindividualisem-
ployed or not (Employed), or their sex their sex (Male). Both variables are captured
as dummy variables. Finally, continuous measures of age, education, and income are
also included in the models. The data for all individual-level variables is provided
by the Pew Research Center (Pew Research Center and the Press 2013a,b,c).
State-level Control Variables
It is also likely that an individual’s political environment could influence their atten-
tion to international affairs, and attention could be explained by state-level environ-
mental factors that vary within a sample. As a result, I include a series of variables
to control for this possibility.
3
The survey from April 25-28 asked participants, “If it is confirmed that the Syrian government
has used chemical weapons against anti-government groups, would you favor or oppose the U.S.
and its allies taking military action against the Syrian government?” The survey from August 29-
September 1 asked participants, “Would you favor or oppose the U.S. conducting military airstrikes
against Syria in response to reports that the Syrian government used chemical weapons?” The
survey from September 12-15 asked participants, “If Syria does not give up control of its chemical
weapons, wouldyoufavororopposetheU.S.conductingmilitaryairstrikesagainstSyriainresponse
to reports that the Syrian government used chemical weapons?”
103
First, I include the number of active duty member of the armed forces within each
state. If an individual personally knows some members of the armed forces, this may
influence their attention to international affairs because they might look out for their
possible deployment, and for news about potential conflict. This is measured as a
continuous variable, measured at the state-year. The data is sourced from the U.S.
Department of Defence (Defence Manpower Data Center 2017), and the data was
correct as of September 30, 2013.
Second, the number of veterans of the armed forces in a state could also influence
people’s attention to international affairs, as people who have served in the mili-
tary and people who know veterans may also lead people to pay a greater amount
of attention to international affairs than they would have otherwise. This data is
provided by the U.S. Census (U. S. Census Bureau 2017).
It is also possible that the political environment in the state influences individual
attention to international affairs. To control for this possibility, the vote for each
majorpoliticalpartyinthe2012Presidentialelectionwasincludedinthemodel. This
data comes from the U.S. National Archives and Records Administration (National
Archives and Records Administration 2017).
Finally, information search is captured, because the amount of information search
about the crisis in a state could influence how closely people follow the news about
Syria. This is measured at the state level by disaggregating Google Trends data
for news searches for “Syria” in the week prior to the end of each wave of surveys
conducted by the Pew Research Center. The data comes from the Google Trends
platform.
104
DescriptivestatisticsforthevariablesusedinthisstudyareprovidedinAppendix
C.1. The tables indicate a large amount of variation in the independent and depen-
dent variables of interest. It is important to note that the number of observations in
each dataset for the dependent variable, attention to the debates over U.S. military
action in Syria, is slightly smaller than the number of observations for other variables
after nonresponses and “don’t know” responses were removed.
5.2.3 Model
The analysis is conducted using multilevel logistic regression to estimate the effect of
foreign policy crisis on individual attention to Syria, and ultimately support for U.S.
military action in Syria. This technique has the advantage of estimating the effect
of individual-level and state-level factors on attention to international affairs (Jäger
2017; Wang et al. 2015). Pooling respondents across units has benefits for estimating
attitudes from survey data, even where there are relatively few observations.
The determinants of attention to international affairs is estimated through a
mixed regression model with both fixed effects and random effects. This can be
expressed as:
y =Xβ +Zu+, (5.1)
wherey is the known vector of observations,X is an unknown vector of fixed effects,
u is an unknown vector of random effects, and is an unknown vector of random
errors.
105
5.2.4 Results
The results from the mixed regression model illustrate some support for the theory
of voters as systematically varying in their political behavior to non-voters. Figure
5.2 and Table 5.1 present results of the multilevel regressions. There are three main
results that are worth discussing further.
First, it appears from the results that if an individual is a registered voter, this
has considerable positive effects on their attention to international affairs. Across all
threewavesofthesurvey, beingaregisteredvoterhadstatisticallysignificantpositive
effects on individuals’ attention to the Syria crisis. In April, being a registered voter
led to increases in attention by 0.267, and this increased to the point where being
a registered voter led to increases in attention by 0.421 in the September survey.
This suggests that voters do systematically vary from non-voters in their attention
to international affairs, providing evidence in support for the attention hypothesis.
Second, political attitudes and partisanship had few effects on individual atten-
tion to the Syria crisis. In the April wave, there were statistically significant positive
effects of support for military action in Syria on attention to the crisis, but this dis-
appeared in later surveys. Partisanship had no effect on attention to Syria, which is
particularly interesting given there was a clear partisan divide in attitudes towards
the conflict in the September survey, where Democrats were much more likely to
support intervention than Republicans.
106
Figure 5.2: The Determinants of Attention to the Syria News Story.
Coefficient plot of attention to Syria in three different Pew surveys during 2013.
Point estimates indicate how closely participants followed the situation in Syria,
with responses ranging from 1 (not at all closely) to 4 (very closely). The plot was
created using the coefplot package in Stata (Jann 2014).
107
Table 5.1: Determinants of Attention to the Syria News Story in 2013.
(1) (2) (3)
April 25-28 August 29-September 1 September 12-15
Registered Voter 0.264
∗∗
0.238
∗∗
0.388
∗∗∗
(0.102) (0.092) (0.094)
Republican -0.167 0.047 0.136
(0.170) (0.140) (0.134)
Democrat -0.043 0.078 0.098
(0.165) (0.135) (0.130)
Independent -0.025 0.009 0.081
(0.164) (0.133) (0.126)
Employed -0.051 -0.101 -0.123
(0.085) (0.080) (0.080)
Male 0.044 0.257
∗∗∗
0.151
∗
(0.075) (0.069) (0.067)
Age 0.013
∗∗∗
0.016
∗∗∗
0.007
∗∗∗
(0.002) (0.002) (0.002)
Education 0.027 0.041
∗
0.037
+
(0.023) (0.020) (0.020)
Income 0.033
+
0.072
∗∗∗
0.051
∗∗∗
(0.018) (0.016) (0.015)
No. Active Duty Military 0.000 0.000 0.000
(0.000) (0.000) (0.000)
No. Veterans -0.035 -0.014 -0.037
(0.033) (0.030) (0.028)
Democratic Party Vote -0.000 -0.000
+
-0.000
(0.000) (0.000) (0.000)
Republican Party Vote 0.000 0.000
+
-0.000
(0.000) (0.000) (0.000)
Information Search (April) 0.001
(0.003)
Information Search (August) 0.011
∗
(0.005)
Information Search (September) -0.004
(0.004)
Constant 1.491
∗∗∗
1.197
∗∗
2.489
∗∗∗
(0.389) (0.378) (0.358)
lns1_1_1
Constant -20.843
∗∗∗
-3.083 -24.935
∗∗∗
(3.498) (2.612) (3.407)
lnsig_e
Constant 0.044
+
-0.073
∗∗
-0.108
∗∗∗
(0.025) (0.026) (0.025)
Observations 828 791 791
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
108
Finally, some of the control variables had some effects on attention to the crisis.
Individuals who identified as male, age and income all had positive effects on atten-
tion to the Syria policy debates in the August and September surveys, effects that
were not expected in the theory. However, information search at the state level had
no statistically significant effects at the 0.05 level, even though this was expected to
have a positive effect on individual attention to the foreign policy debates.
In sum, the analysis indicates support for the hypotheses introduced in the chap-
ter. However, the limited number of relevant questions in the survey, and the ob-
servational nature of the data mean it is difficult to establish a causal relationship
between the independent and dependent variables. While the data provide some pre-
liminary evidence in support of the theory of the conscientious voter, experimental
data can further build on the findings to evaluate how voters vary from non-voters
in their attention to international affairs.
5.3 Discussion
The chapter provides supporting evidence for the theory that voters behave system-
atically differently than the rest of the public. The first study demonstrates that
registered voters pay much greater attention to international affairs, demonstrating
this pattern of behavior both prior to a foreign policy crisis, and during a foreign
policy crisis where there was public debate about the appropriate response to Bashar
al-Assad’s use of chemical weapons against his own people.
The results suggest that scholars of public opinion about foreign policy should be
careful about the populations they study. The evidence suggests that voters behave
109
systematically differently to the rest of the population in their patterns of attention
to internanational affairs.
Many studies theorize that public opinion is important in democratic states be-
cause the public can constrain foreign policymakers at the ballot box. Yet, empirical
research relies on studies of the mass public, and not just voters which raises concerns
that we might be making misleading inferences about our populations of interest. It
is important to consider how variation in perceptions of incentives for accuracy in-
fluence attention to international affairs in the pursuit of an improved understanding
of public opinion about foreign policy.
110
Chapter 6
Perceptions of Impact and Public
Opinion about Foreign Policy
The previous chapter demonstrated that individuals’ voter type influences their at-
tention to international affairs. In this chapter, I argue that people’s foreign policy
attitudes are contingent on perceptions of impact and their attention to international
affairs. I test this theory through analysis of data from the British Election Survey
immediately prior to the referendum on the membership of the United Kingdom in
the European Union [E.U.] The results indicate that individuals who perceived the
impact of the vote as being high were significantly less likely to vote to leave the
E.U., holding potential confounding variables constant. The results have important
implications for the understanding of public opinion about foreign policy, showing
that scholars should consider how perceptions of impact influence political attitudes
and political behavior.
111
The chapter proceeds in four further sections. First, I introduce my theory of
perceptions of impact and public opinion about foreign policy. Second, I introduce
the case, the data, and the methods used to examine the relationship between per-
ceptions of impact and voting intentions in the European Union Referendum in the
United Kingdom. The third section introduces the results, showing a consistent neg-
ative relationship between perceptions of impact and the likelihood of voting to leave
the E.U.. The chapter concludes with a discussion of the implications of the results
and the understanding of public opinion about foreign policy.
6.1 Perceptions of Impact and Public Opinion
about Foreign Policy
In this chapter, I argue that people form attitudes about foreign policy on the basis
of their perceptions about the stakes involved in the foreign policy debate. People
are more likely to pay attention to foreign policy debates, search for information
to acquire knowledge about the options available to their state, and update their
attitudes when they perceive there are important consequences of the decision for
their country and when they perceive that their opinion might have some influence.
For instance, if a foreign policy is perceived to have little impact on the country,
there are few incentives for people to learn about the foreign policy debate. Instead,
it is rational for individuals to exert little effort into information acquisition and to
use information shortcuts to make an choice that is consistent with their preferences.
On the other hand, people might be motivated to learn about a foreign policy debate
if they perceive the stakes as high.
112
Similarly, if an individual perceives their opinion matters, they are more likely
to invest time and effort into information acquisition than otherwise, even if their
assumptionofpoliticalinfluenceisunfounded. Forinstance, ifapolicyisnotdebated
beforeadecisionismadebytheexecutiveorlegislativebranchesofgovernment, there
is little expected benefit for an individual to pay any attention to that particular
issue. However, if there is a public debate, and an individual perceives that their
decision may have some influence on the government’s decision, an individual may
perceive there is some expected benefit from paying attention to an issue and making
the correct decision.
Of course, just as there is a tiny probability than any one individual casts a
decisive vote (Downs 1957); it is unlikely than an individual’s opinion will convince
any given legislator or leader to adopt their preferred policy. However, people are
prone to overestimate their political influence. For example, a Pew Research Center
survey of American public opinion in 2015 revealed that 50 percent of participants
thought that “ordinary people can do a lot to influence government, if they make the
effort;” and58percentreportedthat“votinggivespeoplelikemesomesayabouthow
government runs things” (Doherty, Kiley, and Tyson 2015). Given the widespread
illusion of their influence over government decision making, periods of public debate
may further persuade individuals to pay attention to foreign policy, even if their
opinion actually has little effect on foreign policy.
1
Public referendums represent an example where an individual’s opinion has as
much influence in the policymaking process as they ever will. Even if an individual’s
1
Itisalsoimportanttoacknowledgethatpeoplemayalsopayattentiontoforeignpolicydebates
out of general political interest, for entertainment, or out of a sense of civic duty to be an informed
citizen.
113
vote is unlikely to change the outcome of the referendum, this example of direct
democracy gives the public the chance to make a decision that would otherwise be
made by their elected representatives. Although referendums on foreign policy are
relatively rare, there have been several high profile examples including Costa Rica
voting to approve the Dominican Republic - Central America Free Trade Agree-
ment [DR-CAFTA] in 2007; and France, the Netherlands, Luxembourg, and Spain
conducted referendums to ratify the proposed Constitution for Europe in 2005.
I expect that this is consistent with general attitudes towards risk in foreign
policy attitude formation. If people perceive the impact as high, they are less likely
to support foreign policies they perceive as risky. Because the outcomes of many
major foreign policy decisions are uncertain, individuals who perceive the stakes as
high could become risk averse given the possibility of significant costs to the country
if the state pursues a policy with a poor outcome (Kahneman 2013; Kahneman and
Tversky 1979; Tversky and Kahneman 1981).
In the case of Brexit, both campaigns emphasized the impact of leaving the Euro-
pean Union, but they took very different positions about what the impact would be.
The government leaflet explaining their preference to remain in the E.U. emphasized
that “it’s a big decision...one that will affect you, your family, and your children for
decades to come” (Government of the United Kingdom 2016, 2). The Government
of the United Kingdom (2016, 7-8) stated, “If the UK voted to leave the EU, the
resulting economic shock would risk higher prices on some household goods” and
“create years of uncertainty and potential economic disruption [that] would reduce
investment and cost jobs.” In short, much of the Remain campaign relied on negative
predictions about the high impact of Brexit on the country’s future.
114
In contrast, the Leave campaign emphasized how the impact of leaving the Eu-
ropean Union would dramatically improve the country’s future. However, they used
strikingly similar claims about the impact to the Remain campaign. For instance,
the official Vote Leave website said, “The referendum on 23 June is the most im-
portant vote for a generation - it’s about who makes our laws, who controls our
borders and whether our money is spent on our priorities” (Vote Leave 2016a). The
Vote Leave leaflet argued that staying in the E.U. would be costly, stating “we can’t
take the risk of voting to stay...Voting to stay in the EU is a threat to our jobs and
our security” (Vote Leave 2016b, 6). The official Leave campaign also stressed the
economic benefits of leaving to “spend money on our priorities like the NHS” instead
of sending £350 million to the E.U. every week (Vote Leave 2016b, 3-4).
One of the intriguing developments of the Leave campaign was the presence of
a rival campaign that also arguably contributed to Brexit, Leave.EU. Accordingly,
it is important to consider how they discussed the impact of leaving the E.U. The
campaign implied leaving the E.U. would lead to households having “£1,000 more
each year” as the result of “cheaper food bills, no membership fees, with the cost
of regulations lifted” (Leave.EU 2016b). Furthermore, the Leave.EU leaflet stated,
“OnlybyleavingtheEuropeanUnionwilltheUKhavetotalcontrolofitsbordersand
be able to deport dangerous foreign criminals;” that “leaving the EU will improve our
qualityoflife: withitsrighttostrikedealsrestored, theUKwillbeabletogainbetter
access to those markets that buy most of our goods and services;” and that “leaving
theEUwouldprovidea£15.3bnwindfall, enoughtopayfor35newhospitalsandstill
double current EU spending on Research and Development in the UK” (Leave.EU
115
2016a). In short, both the official Vote Leave and Leave.EU campaigns emphasized
how Brexit would have a large positive impact on the country.
I expect that individuals who perceive there are significant stakes of foreign policy
decisions are likely to oppose a policy change where outcomes are uncertain. This is
especially likely if people pay attention to the foreign policy debate for information
about the policy as this information might contribute towards them becoming more
risk averse as the weigh up the potential outcomes of the decision. Finally, I expect
that partisanship may be a less powerful predictor of attitudes than perception of
impact.
In this case of the Brexit referendum, I expect that as people’s perception of
the impact of leaving the E.U. increases, their willingness to vote Leave decreases.
Similarly, if they pay attention to the debate, I expect that this will decrease the
likelihood of voting to leave the European Union. Building on hypotheses introduced
earlier, I introduce two further hypothesis specific to the Brexit referendum:
Hypothesis 8 Impact Leave Hypothesis
If an individual perceives the impact of leaving the E.U. as high, they are less likely
to vote to leave the E.U.
Hypothesis 9 Attention Leave Hypothesis
If an individual pays attention to the E.U. Referendum, they are less likely to vote
to leave the E.U.
116
6.2 Data and Methods
6.2.1 Case Selection: U.K. Public Opinion about the E.U.
Referendum, 2016
British public opinion about the E.U. Referendum completed on 23 June 2016 is a
good case for testing the theory of perceptions of impact and foreign policy attitudes
for several reasons. First, there is a wide variety of individual perceptions of impact
of Britain leaving the European Union among participants in the British Election
Survey.
Second, the case is substantively important. British Prime Minister David
Cameron’s decision to let the public decide the future of British membership of
the European Union is an example of a consequential foreign policy decision made
directly by the public. As such, given the significance of the vote and its potential
consequences, it is a vital case to understand how people think about foreign policy
in the knowledge that they had a direct say in this foreign policy debate.
Furthermore, voters responded to the plebiscite by exercising their ability to
vote about continued membership of the European Union. Figure 6.1 compares the
turnout in the E.U. Referendum to General Elections for the U.K. Parliament, as
measured by the percentage of valid votes compared to the number of registered vot-
ers. The figure indicates that turnout was relatively high, representing the greatest
percentage of valid votes out of all registered voters in a national election since the
1992 General Election. Furthermore, turnout in the election exceeded the turnout in
both the 2015 and 2017 General Elections, suggesting that the public was motivated
to have their voice heard even more than in elections.
117
Figure 6.1: Voter Turnout in UK Parliament General Elections and the E.U. Refer-
endum, 1979-2017.
Source: The Electoral Commission (2018)
6.2.2 Data
The data for all variables is provided by questions from Wave 8 of the 2015 British
Election Survey, led by a team of researchers at the University of Manchester, the
University of Oxford, and the University of Nottingham (Evans et al. 2016). All
survey participants were interviewed between 6 May 2016 and 22 June 2016, meaning
118
the final interviews in the wave were completed immediately prior to the referendum.
In total, 33,502 participants completed the survey in the wave.
Dependent Variable: Voting Leave in the E.U. Referendum
It is important to note that the reasons why an individual might vote to leave the
European Union might be different to the reasons why they did not vote to remain
in the European Union, and vice versa. There is a possibility there are different de-
terminants of each vote, so I model the two substantive votes in the E.U. referendum
separately.
The dependent variables are derived from individuals’ voting intentions in the
E.U. Referendum. This is operationalized as the answer to the following question:
“If you do vote in the referendum on Britain’s membership of the European Union,
how do you think you will vote?” Participants were given three possible responses
to this question: 0) Remain in the E.U.; 1) Leave the E.U.; 2) Don’t know. I
transformed this into a dichotomous variable for the referendum removing “Don’t
know” responses and nonresponses from the analysis: Vote Leave.
2
Independent Variables: Perception of Impact and Attention
There are two independent variables in this study. First, perception of impact
is measured by participants’ estimates about the potential impact of leaving the
EuropeanUnionwouldhaveontheUnitedKingdom. Itisimportanttonotethatthis
2
A robustness check with a different coding scheme is included in Appendix D.5, where -1
indicates the intention to vote to remain in the E.U.; 0 indicates a “don’t know” or nonresponse;
and 1 indicates voting leave. This alternative coding of the dependent variable does not change the
results.
119
measure does not frame leaving the E.U. as either a positive or negative development
for the country - just that it would have an effect on the country. Perception of
impact is operationalized as participants’ response to the following question: “How
much impact do you think that Britain leaving the E.U. would have on the country
as a whole?” Responses ranged from: 1) no impact at all; 2) a small impact; 3) a
moderate impact; 4) a large impact; 5) a very large impact. Again, “Don’t know”
responses were dropped from the analysis so it became a dichotomous variable.
The second independent variable of interest is whether people paid attention to
the debate about the European Union. Prior to the debate, the British Government
sent a leaflet outlining the case for the country to remain in the United Kingdom.
Although the document advanced their preferred result, people who were attentive
to the debate would likely read the leaflet, even if they disagreed with the content
of the leaflet. As a result, attention is measured by whether participants read the
leaflet ahead of the referendum. Participants were asked, “Have you received and
read the UK Government’s ‘Why the Government believes that voting to remain in
the European Union is the best decision for the UK?’” Participants were asked to
choose one of the following responses: 1) Yes, I received it and read it; 2) Yes, I
received it but I haven’t read it; 3) No, I haven’t received it; or 9999) Don’t know.
Their response was recoded to indicate 0 if they had received it but had not read
the leaflet, 1 to indicate if they had received it and read the leaflet, and all other
responses were removed from the analysis.
120
Control Variables
I also include control variables that could conceivably influence the likelihood of an
individual’s voting intentions in the E.U. Referendum in the United Kingdom in
order to rule out alternative explanations for the effect of incentives and attention
on foreign policy attitudes.
Political Control Variables
First, I include their level of political interest in the E.U. Referendum. It
is possible that people’s interest is correlated with a position on the referendum, so it
is important to include this in the analysis. Participants’ level of interest in the ref-
erendum was measured by their response to the following question: “How interested
are you in the EU referendum that will be held on June 23rd?”. Responses could
include: 1) Not at all interested; 2) Not very interested; 3) Somewhat interested; or
4) Very interested. Again, nonresponses and “Don’t know” responses were excluded
from the analysis.
Second, I include measures of political knowledge about the E.U. Refer-
endum. Participants were given a series of five questions to answer about the rela-
tionship between the United Kingdom and the European Union. Political knowledge
was measured as the number of questions that each participant correctly answered
out of six in total. Further details about the questions and the answers are available
to view in Appendix D.1.
Third, in keeping with prior research, it is possible that there are systematic
differences in political behavior between voters and non-voters. It is possible that
voters have systematically different foreign policy attitudes than people who do not
vote. As a result, I include a dichotomous measure of voter type, as measured by
121
whether the participant voted in the 2015 General Election in the United Kingdom
(Voter).
I also include how each individual voted in the previous election. As party mem-
bership is not as common in the United Kingdom compared to other democracies
such as the U.S., previous voting behavior is a better measure of party ID than
membership. I include dichotomous measures of previous voting, where a 0 indi-
cates the individual did not vote for that party, and 1 indicates they did vote for
the party. I include these variables for previous towards the six major parties that
received more than one percent of the vote in the previous election: Conservative,
Labour, Liberal Democrat (Lib Dem), Scottish National Party (SNP),
United Kingdom Independence Party (UKIP), and the Green Party.
Risk and Uncertainty Control Variables
Itispossiblethatindividuals’votingintentionsinthereferendumwerecontingent
on their risk acceptance and their perception of certainty about the effects of leaving
the E.U. As I result, i include measures of risk and uncertainty to control for this
alternative explanation of voting leave and voting remain in the referendum.
First, I include a measure of an individual’s degree of certainty about the
effect of leaving the E.U. on the United Kingdom. While the dependent
variable measures their perception of impact, this measure captures how certain they
are about their perceptions of the effect of leaving the European Union. Participants
were asked, “How sure are you about what would happen to the UK if it left the
EU?” Responses could include: 1) Very unsure; 2) Quite unsure; 3) Quite sure; or
4) Very sure. Nonresponses and “Don’t know” responses were excluded from the
analysis.
122
I also include a measure of risk taking to determine if attitudes towards risk
influenced voting intentions in the referendum. Participants were asked about their
willingness to take risks in general, with responses ranging from 1) very unwilling to
take risks to 4) very willing to take risks.
Region Control Variables
It is also important to control for regions within the United Kingdom. The
E.U. Referendum exposed significant differences in political attitudes and behavior
between regions in the U.K., with London, Scotland, and Northern Ireland voting
to remain in the European Union, and many parts of the North and rural areas
supporting leaving the European Union. To control for regional effects, I introduce
dichotomous dummy variables for regions derived from the government office regions.
These include theNorth East, North West, Yorkshire and the Humber, East
Midlands, West Midlands, East of England, London, South East, South
West, Wales, Scotland, and Northern Ireland.
Demographic Control Variables
Finally, I include a number of demographic variables that could otherwise explain
variation in the dependent variable. These include a dichotomous measure of sex
(Male), a continuous measure of age, and a continuous measure of income. I also
include dichotomous measures of race and ethnicity for people who identify with one
of the three main demographic groups: White-British,White-Other, andBlack,
Asian, and Ethnic Minority (B.A.M.E.). The latter category is particularly
unfortunate as it groups together people from a variety of different cultural and
ethnic backgrounds as “other” to the majority-white state.
123
However,itisafrequently-usedcategoryintheUnitedKingdom,anditisauseful,
albeit blunt, instrument for measuring differences between people who identify with
an ethnic minority and people who identify as white British in particular. As a
result, despite the problems with the category, I use it in this paper. Further details
about all control variables, and the codebook for the variables included in models
in this paper are provided in Appendix D.1. Descriptive statistics for the variables
used in this study are provided in Appendix D.2.
6.2.3 Model
Because the dependent variables are dichotomous, analysis is conducted using simple
logistic regression with robust standard errors to estimate the effect of perceptions of
impact and attention to the E.U. Referendum on the likelihood of voting to leave and
remain in the European Union. Alternative specifications of the model are reported
in Appendix D.4 using a probit model, and the results are robust to the different
specification. The model for voting Leave can be expressed as the following:
p(x) =P(VoteLeave = 1|x) =P(β
1
Impact +β
2
Attention +u> 0|x) (6.1)
where the dependent variable is the probability p that an individual intends to vote
to leave the European Union (VoteLeave = 1).
124
6.3 Results
The results from indicate clear support for the hypotheses presented in this paper.
There was a clear negative relationship between individuals’ perceptions of the im-
pact of leaving the European Union and the likelihood of voting to leave the E.U.
Similarly, there was a negative relationship between attention as measured by read-
ing the leaflet about the referendum and the predicted probability of voting Leave.
In this section, I discuss these findings in more detail, and the implications of the
results for the understanding of public opinion about foreign policy.
Figure 6.2 present the results of the logistic regressions on voting Leave.
3
There
are several key implications of the results that are worth further discussion.
First, the results indicate the presence of a strong negative relationship between
perceptions of impact and the predicted probability of an individual voting to leave
the European Union. Substantively, in the model with all control variables included,
for every one unit increase in the perception of the impact of the U.K. leaving the
E.U. on the country, there is a 1.315 decrease in the log-odds of an individual voting
to leave the E.U.
The results indicate that we can reject the null hypothesis of no effect of per-
ceptions of incentives on voting for Brexit as perceptions of impact of leaving the
European Union were negatively correlated with voting leave in the Brexit referen-
dum.
3
Appendix D reports the results including coefficients for all variables and robustness checks.
125
Figure 6.2: Determinants of Voting to Leave the E.U.
Dependent Variable: The Predicted Probability of Voting to Leave the European
Union.
Point estimates indicate the predicted probability of voting Leave with 95%
confidence intervals. The plot was created using the coefplot package in Stata
(Jann 2014).
126
Second, thereissupportfortheattentionhypothesis. Attentionappearstohavea
negative relationship with the predicted probability of voting to leave the European
Union. In the full model, paying attention to the government’s leaflet appears to
decrease the log-odds of voting to leave the E.U. by 0.233. These results appear to
supporttheAttentionLeavehypotheses. Furtherdiscussionofmarginaleffectsbelow
helps to explain the relationship between variables on voting intentions in greater
depth.
Third, itisimportanttonotethatpoliticalinterestinthereferendumispositively
correlated with the predicted probability of voting to leave the E.U. For every one
unit increase in interest in the referendum, the log-odds of voting to leave the E.U.
increase by 0.582. At first glance this is surprising, given that political interest is
generally associated with high levels of political knowledge and attention to politics
(Lecheler and de Vreese 2017; Prior 2010; Strömbäck and Shehata 2010), which are
both negatively correlated with voting Leave. This begs the question, why would
political interest correlate with voting Leave?
One possible explanation is there could be an important difference between gen-
eralized political interest, and interest in a specific election or referendum. People
who scored low in political interest, as measured by interest in politics, voted in the
E.U. referendum in significantly greater numbers than they had in the 2015 general
election and the 2011 Alternative Vote referendum (NatCen Social Research 2017).
Similarly, NatCen Social Research (2017) found that “while 62% of those with not
very much or no interest at all in politics voted to leave, only 40% of those with
quite a lot or a great deal of interest did so.” As a result, it appears that people
who were relatively uninterested in politics were more likely to vote in the Brexit
127
referendum than other elections, and to vote Leave when they did get to the polls.
As a result, political interest in the Brexit referendum might function differently to
political interest writ large.
A reason why there is a difference between generalized and particularized polit-
ical interest is the enthusiasm gap between the proponents of leaving the E.U. and
advocates of remaining in the E.U. Despite having two rival campaigns, the Leave
campaign appears to have generated more enthusiasm than Remain’s arguments.
Focus groups conducted by Ipsos MORI and IPPR in Scotland provide some ev-
idence for this. Among participants, “When invited to spontaneously outline the
pros and cons of EU membership, people generally find it easier to be critical and
express those criticisms with feeling and passion,” whereas “there are perceived bene-
fits of EU membership but participants are generally less passionate and vocal about
these” (Diffley et al. 2016). People who are interested in the referendum might have
responded to criticism of the E.U. that ignited their willingness to vote Leave, while
the Remain campaign did not prompt similar feelings of enthusiasm for the status
quo. However, this is still an interesting finding that should be explored in further
research.
A fourth implication of Figure 6.2 is the small but statistically significant nega-
tive relationship between political knowledge and the predicted probability of voting
Leave. Substantively, for every one unit increase in political knowledge about the
E.U. increased, there was a 0.086 decrease in the log-odds of voting to leave the
European Union.
Fifth, votingintentionstrackedwithexpectationsaboutvotingbasedonpartyID.
As one might expect, individuals who voted for UKIP in the 2015 General Election
128
were significantly more likely to vote Leave than other parties. Having voted for
UKIP in 2015 increased the log-odds of voting to leave the E.U by 3.025. Voting for
the Conservative Party in the 2015 election also increased the likelihood of voting
Leave, but voting for all other parties had the opposite effect.
Sixth, it appears that perceptions of certainty and risk taking affected voting
intentions in the referendum. Certainty about the effects of leaving the E.U. and
willingness to take risks are positively correlated with voting Leave. Unsurprisingly,
those who are naturally risk averse were less likely to vote for a foreign policy change,
and uncertainty about the implications of leaving the E.U. was associated with a
reluctance to depart from the status quo in foreign policy.
There were also noticeable regional differences in voting intentions. In the full
model with all covariates, it appears voting in the E.U. referendum mirrored tradi-
tional regional differences in the United Kingdom between the North and the South,
and between urban centers like London and rural areas.
Finally, it is interesting to note that individuals who identified as white but not
of British origin were significantly less likely to vote for Brexit than British whites.
This could reflect the fact that many white migrants to the United Kingdom from the
E.U. stood to suffer the consequences of any potential backlash against immigrants
if the British left the E.U., so they opposed this possibility.
Figure 6.3 presents the marginal effects of perceptions of the interaction of impact
and attention on the predicted probability of voting to leave the European Union in
the referendum. There are two key implications from this figure.
129
Figure 6.3: The Marginal Effects of Perceptions of Impact and Attention on Voting
to Leave the E.U.
Dependent Variable: The Predicted Probability of Voting to Leave the European
Union.
Margins plot of the effect of perceptions of impact and attention on voting Leave.
Point estimates indicate the predicted probability of voting Leave with 95%
confidence intervals.
130
First, the figure illustrates the strong negative relationship between perceptions
of impact and voting Leave in the E.U. Referendum. Figure 6.3 illustrates how the
predicted probability of voting Leave in the referendum decreases markedly as the
individual perceives the impact of the U.K. leaving the E.U. on the country.
Second, the figure illustrate how attention has a negative effect on the predicted
probability of voting to leave the European Union, but that this effect is statistically
distinct from the effect of incentives alone when an individual perceives the stakes
as especially high. The plot shows that when an individual perceives the impact of
Brexit would be “large” or “very large,” paying attention to the debate decreases
the predicted probability of voting Leave compared to if an individual does not pay
attention.
In sum, the results indicate support for the hypotheses presented in this paper.
Individuals who perceive the stakes as high - that there is likely a large impact of
voting to leave the E.U. - were significantly less likely to vote to leave the European
Union. Similarly, attention is also negatively correlated with voting Leave, partic-
ularly when the individual perceived the impact as being high. The results suggest
that context matters, and that variation in perceptions of impact have important
implications for how people think about foreign policy.
6.4 Discussion
The chapter provides compelling evidence in support of the theory of perceptions of
impact and public opinion about foreign policy. Evidence from the E.U. Referendum
in the United Kingdom demonstrates that perceptions of impact matter for attitudes
131
about foreign policy - people were far less likely to vote for Brexit if they perceived
the stakes were high. Similarly, attention decreased the predicted probability of
an individual voting to leave the European Union, especially when the individual
perceived the stakes as high.
These findings have important implications for the study of public opinion about
foreign policy. Until now, there has been insufficient attention to the conditions
under which people have incentives to make accurate decisions about foreign policy,
and how this shapes their attitudes about foreign policy. The evidence presented
here suggests that there is an important relationship between perception of impact
and political attitudes, and it suggests there should be much more effort devoted to
understanding these dynamics in public opinion.
However, it is important to note that this chapter constitutes a tentative first
step, and there are important limitations of the chapter in its current form. Future
research should build on these results in the pursuit of a greater understanding of
public opinion about foreign policy. One important limitation is that causal inference
is problematic given the observational data. In particular, it is difficult to rule out
concerns of endogeneity and it is difficult to establish the direction of causality -
whether impact cause political attitudes, or vice versa. I do not offer a solution to
this concern in this chapter, but future research could use experiments to get leverage
oncausalrelationshipsbetweenvariables,andtheeffectsofimpactonattitudesabout
foreign policy.
A second limitation of the chapter is that it uses a rare event to make a broader
point about public opinion about foreign policy. It is rare that public referendums
areemployedinthecontextofforeignpolicy, eveniftherearesomenotableexamples.
132
However, while the case itself might be rare, it does not necessarily follow that these
dynamics are any less apparent during other foreign policy debates. While the public
may not always be able to directly influence policy outcomes, previous research has
demonstrated that the public does constrain foreign policymakers (Baum and Potter
2015). Future research can test the theory in a series of different foreign policy
debates and situations to build on the findings presented in this chapter.
A third limitation is that there was widespread disagreement about the stakes
in the context of Brexit, but this is not necessarily always the case. For instance if
a country is debating whether to engage in military conflict, to respond to terrorist
attacks, or to enter trade negotiations, one could argue that there would be much
more agreement about the impact. However this is not necessarily the case. For
instance, people struggle to make accurate predictions about the implications of
foreign policy decisions (Mellers et al. 2015; Tetlock 1998, 1999, 2006; Ungar et al.
2012). As a result, it is possible that much more heterogeneity between individuals
exists regarding perceived impact in the realm of foreign policy.
Future studies should build on the findings presented in this chapter to develop
a better understanding of the conditions under which individuals perceive impact,
and how this influences their attitudes about foreign policy. Hopefully, this chapter
marks the beginning of an improved appreciation of the context in which the public
forms attitudes about foreign policy. When people perceive the stakes as high, they
are liable to pay greater attention to the foreign policy debate, and ultimately update
their attitudes. Future research should build on this study to further develop the
understanding of public opinion and foreign policy.
133
Chapter 7
Crisis, Information Search, and
Public Opinion about Foreign
Policy
Previous research suggests that political attitudes about foreign policy are largely
explained through elite cues, where people follow the positions of leaders of their
political party. However, recent research suggests a more optimistic vision of public
opinion - that when people are motivated to search for information, they are capable
of using new information to make up their minds independent of elite opinion. In
this chapter, I examine how the onset of foreign policy crisis leads people to search
for information about international affairs, and to update their attitudes about for-
eign policy in response to new information. Using evidence from debates about U.S.
134
foreign policy towards Iraq, Syria, and ISIS, I demonstrate that dynamics of cri-
sis and information search explain variation in public opinion about foreign policy,
independent of elite cues.
The chapter proceeds in seven further sections. First, I outline a theory of infor-
mation search, public opinion, and international relations. I identify the conditions
under which I expect people to search for information, and how this should influence
political attitudes about foreign policy. Second, I outline the empirical strategy for
the chapter, including the use of data from public opinion polls, Google Trends data,
and the use of historical political documents. The next three sections of the chapter
then examine the three most important recent foreign policy crises faced by the U.S.
in chronological order: 1) Iraq; 2) Syria; and 3) ISIS. The chapter concludes with a
discussion of the implications of the findings of this chapter.
7.1 Crisis, Information Search, and Public Opin-
ion about Foreign Policy
As discussed in earlier chapters, I argue that people are more likely to pay atten-
tion to foreign policy debates, search for information to acquire knowledge about
the options available to their state, and update their attitudes when they perceive
incentives to make the correct decision. In previous chapters, the focus has been
on individual-level variation between subjects, where voters behave differently to
non-voters, and perceptions of the impact of the Brexit referendum determined in-
dividuals’ voting intentions. In this chapter, I examine variation in perceptions of
incentives for accuracy at the population level.
135
Context matters. When people are asked to make decisions about foreign policy
during periods of crisis and when there is a period of public debate, they are likely
to behave differently than during periods when they perceive the stakes as lower.
Rather than perceptions of incentives being solely based on individual perception,
the geopolitical situation can induce changes in individual behavior.
Inparticular, foreignpolicycrisesmayprovideincentivesforpeopletolearnabout
international affairs (Brecher and Wilkenfeld 1997; Brecher et al. 2017; Feldman,
Huddy, and Marcus 2015). When people perceive the existence of a threat to their
country, they become more likely to search for information about the foreign policy
debate, and to use that information to update their attitudes about the foreign
policy based on this information. During crisis, when there are incentives for people
to make accurate judgments, individuals are more likely to search for information
about foreign affairs.
Individuals are also more likely to search for information and to update their
attitudes about foreign policy when they perceive it is worth the effort. If a foreign
policyisnotdebatedbeforeadecisionismadebytheexecutiveorlegislativebranches
of government, there are few incentives for an individual to search for information
about the best response. However, if there is a period of extended public debate,
and there is a chance to lobby democratically elected representatives, people may
become more likely to devote effort into information search, and then update their
attitudes as a result of this information.
If this theory is plausible, there are several observable implications in public
opinion data. First, during foreign policy crisis and when people have a chance to
weigh in on foreign policy debates, people should increase their information search
136
about the foreign policy issue. In short, when people have situational incentives for
accuracy, they should search for information in the pursuit of knowledge about the
foreign policy issue.
Second, one should expect that during periods of crisis and when there is public
debate about the appropriate response, people update their attitudes about foreign
policy. It is more difficult to discern what the empirical implication of updating at-
titudes should be, given that it is impossible to observe the counterfactual attitudes,
and there are myriad confounding factors.
However, oneclearobservableimplicationofthepublicupdatingtheirattitudesis
that it should not simply reflect elite opinion. One of the most powerful determinants
of public opinion about foreign policy, and politics more generally, are party ID and
elite cues (Abramowitz 2010; Bartels 2002; Brady and Sniderman 1985; Druckman,
Peterson, and Slothuus 2013; Fiorina and Abrams 2008; Zaller 1992). If the public
updates their foreign policy attitudes according to new information, we should see
that these attitudes do not perfectly correspond to positions adopted by their co-
partisan political leaders, but instead be the result of independent evaluation of
policy during the crisis.
1
While people may still be influenced by party cues, if people consume information
to help them make accurate decisions, one should observe that the information has
independent effects on attitudes (Feldman, Huddy, and Marcus 2015). If people are
1
Again, it is useful to consider Bayes’ rule in evaluating how people respond to new information
and update their attitudes. According to Bayes, people should change their beliefs in response to
information. Ultimately, people should be responsive to information and update their attitudes
with new evidence, regardless of their priors.
137
incentivized to search for information to make accurate choices about foreign policy,
this information may overcome partisan affiliations.
7.2 Empirical Strategy of the Plausibility Probe
In this chapter, I evaluate the plausibility of my theory of crisis, information search,
and political attitudes about foreign policy using observational data of public opinion
about major U.S. foreign policy debates in the early twenty-first century.
There are significant limitations to the available data, including the fact that
the data is aggregated to the population level. Thus, there are unobserved variables
that could confound the relationship between variables. The absence of covariates
means I cannot control for possible confounds, making causal identification all but
impossible using the data introduced in the chapter.
As a result of these limitations, this chapter is best described as a plausibil-
ity probe. Time-series models such as an autoregressive integrated moving average
(ARIMA) are necessary to make causal arguments about trends and patterns over
time. Unfortunately, the absence of longitudinal time series data make this all but
impossible, so any patterns observed in the figures presented in this chapter cannot
be interpreted as causal. Instead, this chapter presents what data is available to
evaluate the plausibility of the theory introduced in this dissertation, even if one
cannot make inferences about its validity based on this data.
I attempt to illustrate the plausibility of the theory of crisis, information search,
and attitudes about foreign policy, albeit with these important caveats given the
138
incomplete nature of the data. Future research could build on the results presented
here and test hypotheses with improved data tracking individual behavior over time.
In this chapter, I aim to build on the results of the experimental analysis with
analysis of public opinion polls about U.S. foreign policy towards Iraq, Syria, and
ISIS from 2002-2017. The cases are all chosen as important foreign policy crises
facing the United States in the early twenty-first century, and they all illustrate the
process of information search, motivated learning, and attitude formation among the
U.S. public.
Table 7.1: Operationalization of Constructs.
Construct Observational Variable
Incentives for Accuracy Foreign Policy Crisis + Public Debate
Information Search Google Searches
Foreign Policy Attitudes Foreign Policy Attitudes
One important limitation is that variables of interest do not always align between
surveys for each crisis. Table 7.1 introduces the main variables of interest in the
public opinion polls.
23
The table provides a summary of the operationalization of
constructs in this study. More detail about the data for each study is provided in the
description of each case. The available data illustrate the plausibility of the theory
2
Further information about the operationalization and measurement of variables is presented
in supplementary files.
3
Google Trends provides data for Web searches from 2004, and News searches from 2008. The
two different types of search are highly correlated with one another, so they could almost be used
interchangeably. However, News searches have the benefit of being focused even more narrowly on
current foreign policy debates, whereas web searches could be more broad in the intent of the user’s
information search. As a result, for public opinion data from 2008 onwards (the cases of Syria
and ISIS), I use data from News searches as the best means to measure information search about
foreign policy. To allow for comparisons over time, I use Web searches for public opinion data that
predates 2004 (the case of Iraq).
139
using data from surveys relating to U.S. military action in response to foreign policy
crises with Iraq, Syria, and ISIS.
7.2.1 Observable Implications of Theory
There are two important observable implications that would suggest that the theory
isplausible. First, oneshouldobservethatpeoplesearchforinformationduringcrisis.
If the theory is correct, there should be increases in information search in response
to crisis, as people seek to learn about international affairs as they recognize the
importance of making the correct foreign policy decision. If there is no discernible
change in information search during such periods, this would challenge the theory.
Second, one should also observe changing attitudes about foreign policy at the
same time as information search, where people do not simply respond to crisis to
consume information, but use it to update their opinions about foreign policy. How-
ever, if people do not respond to crisis and information search by updating their
attitudes about foreign policy, this would challenge the theory.
7.3 U.S. Public Opinion about Military Action in
Iraq, 2002-2008
Thefirstforeignpolicydebateisthecrisisthatculminatedinthe2003US-ledinvasion
of Iraq. As the most prominent US military conflict in the last twenty years, it is an
important conflict that has had an enduring legacy.
140
Afterthe9/11terrorattacksonhomesoil, theUnitedStatesrespondedbylaunch-
ing a “War on Terror” against the perpetrators of the events, and the states who
harbored the terrorists. After the invasion of Afghanistan in late 2001, the Bush Ad-
ministration’s attention turned to Iraq, a long-term adversary of the United States.
Saddam Hussein had long been cagey about his programs developing weapons of
mass destruction (W.M.D.), and the U.S. had imposed a series of sanctions and lim-
ited attacks on Iraq in response to Hussein’s refusal to allow U.N. weapons inspectors
to inspect his facilities.
From early 2002, leading officials started to build a case for invading Iraq and
overthrowing Saddam Hussein. Bush’s State of the Union speech targeted Iraq as
one of the three powers of the “axis of evil,” and in February U.S. military forces had
begun preparations for a conflict in Iraq. On September 12, the foreign policy crisis
began with Bush’s speech to the U.N. General Assembly that outlined a threat for
multilateral military action against Iraq if they failed to meet a series of prescribed
conditions to destroy their weapons programs.
Iraq initially complied, welcoming U.N. weapons inspectors, and providing them
with a declaration of weapons capabilities. However, because there was little new
information in the document, the U.S. declared the regime to have breached U.N.
Security Council Resolution 1441. In early 2003, the U.S. engaged in diplomatic
efforts to gather support for multilateral action, and they pursued a Security Council
Resolution authorizing the invasion of Iraq. However, this effort was unsuccessful,
and the U.S. eventually invaded Iraq without the resolution on March 20, leading a
coalition including troops from the United Kingdom and Spain.
141
The coalition forces overwhelmed the Iraqi troops, and they seized Baghdad on
April 9, 2003. On May 1, 2003, George W. Bush announced the victory of the
coalition forces on the U.S.S. Lincoln. Saddam Hussein was eventually captured in
December 2003. However, efforts at installing a stable democratic regime in Iraq
have been more problematic, as the power vacuum after Hussein’s departure helped
exacerbate sectarian tensions between Sunni and Shia. Iraq subsequently lurched
into continuing violence as insurgents, and then rival sects used guerrilla tactics and
improvised devices to inflict violence.
In an effort to secure Iraq more effectively, Bush announced the “surge” of U.S.
troops to enter Iraq in early 2007. The surge was meant to help the Iraqi regime
become capable of maintaining their own security without the need for continued
U.S. involvement. This was met with significant resistance from politicians and the
public who had little appetite for continued involvement in Iraq as casualties from
the invasion and the occupation mounted.
Table 7.2: Crisis, Debate, and Events in the U.S. Military Action in Iraq.
Foreign Policy Crisis Public Debate U.S. President
January 1 – September 12, 2002 7 3 George W. Bush (R)
September 12, 2002 – May 1, 2003 3 3 George W. Bush (R)
May 1, 2003 – June 30, 2008 7 3 George W. Bush (R)
Table 7.2 presents the timeline of crisis, debate, and events in the U.S. military
action in Iraq. The table shows the three distinct phases of the conflict.
4
First,
4
The timeline corresponds to the coding of the foreign policy crisis by the ICB Data project
(Brecher and Wilkenfeld 1997; Brecher et al. 2017). The ICB data project has information for
476 international crises in total, spanning from 1918-2015, involving a total of 1052 actors. A
complementary project on international incidents that meet some but not all of the crisis criteria
is also in development (Brecher et al. 2017).
142
a period of debate occurred throughout 2002 as the public debated whether the
U.S. should enter conflict with Iraq. After Bush’s U.N. speech, there was a period
of foreign policy crisis that endured through the administrations threats to Iraq,
debates in the United Nations, weapons inspections, and finally the military conflict
in Iraq. Afterwards, the third period endured from the end of the war through to the
end of the data in mid-2008. While there was not a foreign policy crisis during this
period, U.S. forces continued to be involved in Iraq and bear the costs of conflict.
7.3.1 Case Selection
The war in Iraq constitutes a good case for the study of public opinion about Iraq
because it is arguably the United States’ most significant conflict since the end of
the Cold War. Over 15 years later, U.S. forces remain stationed in Iraq. Since the
invasion, much more has become known about the Iraqi weapons programs - namely
that the war was waged on the basis of W.M.D. that may never have existed. It
is also an interesting case for the study of the public’s attitudes towards conflict
because of the temporal variation from the lead-up to the war, the crisis itself, and
then the ongoing U.S. presence in Iraq.
7.3.2 Data
DataforthecaseofIraqisprovidedbytwosources. First, publicopiniondataisfrom
the Eshbaugh-Soha and Linebarger (2014) study of presidential leadership, media
coverage, and public opinion about Iraq. This monthly time series data primarily
consists of participants’ responses to a question asked in the USA Today/Gallup
143
from 2003 to 2008: “In view of the developments since we first sent our troops to
Iraq, do you think the United States made a mistake in sending troops to Iraq, or
not?” Where participants answered that the war was not a mistake, this was coded
as “support”. At other points in the data, alternative question wording was used,
but this was highly correlated with the Gallup measure.
5
Information search data consists of Google Web searches of "Iraq War" from
locations in the United States from 2004 to the end of June 2008. This data is
aggregated to the month level to match the unit of analysis in the public opinion
data.
7.3.3 Evidence from Public Opinion Polls
The data from the case of Iraq present some initial evidence in support of the theory
of crisis, information search, and political attitudes about foreign policy. Figure
7.1 illustrates support for U.S. military action in Iraq from 2002-2008 using the
Eshbaugh-Soha and Linebarger (2014) data. There are three main points to take
away from this figure.
First, there appears to have been an increase in support for the Iraq War dur-
ing the latter half of the foreign policy crisis. This period coincides with the war
itself after the U.S.-led coalition eventually invaded Iraq. This surge in support is
consistent with the notion of rally effects during war, where public opinion falls in
behind leaders in a wave of patriotic fervor during conflict (Baum 2002a; Brody 1992;
Gadarian 2010; James and Rioux 1998; Kam and Ramos 2008; Lian and Oneal 1993;
5
For more detail about the coding and selection of data, please refer to Eshbaugh-Soha and
Linebarger (2014).
144
Mueller 1973). Immediately after the end of the crisis, support for the war drops
even more dramatically, even though support remains well above fifty percent until
2004.
Unfortunately no information search data is available until 2004, so one is unable
to determine how information search relates to these dramatic changes in support
for the war at the start of the conflict. It is also important to note that even at the
beginning of the foreign policy crisis, support for the war was high, with well over
sixty percent of the public in support of war against Saddam Hussein’s regime.
Second, the figure indicates a general downward trend in support for the conflict
over time. At the same time, there appears to be a pronounced decrease in informa-
tion search about the conflict. While one is unable to draw any causal association
between the two variables due to limitations of the data, it is interesting that the
variables appear to be related. If there is a causal relationship between information
search and attitudes about conflict, this is the type of pattern one might expect. At
the very least, the pattern is not unexpected according to the theory of incentives,
information search, and public opinion about foreign policy.
145
Figure 7.1: U.S. Public Support for the Iraq War, 2002-2008
Plot of support for the Iraq War, 2002-2008. The X axis indicates time, and the Y
axis indicates overall support for the war in Iraq. Green dots indicate overall
support. Relative Google Web Searches for “Iraq War” are superimposed over the
plot from 2004 onwards, with relative searches on the right hand side of the figure
indicating the correlation between literal searches for information about the Iraq
War and attitudes towards the conflict.
146
Figure 7.2: U.S. Public Support for the Iraq War by Party, 2002-2008.
Plot of support for the Iraq War by Party, 2002-2008. The X axis indicates time,
and the Y axis indicates overall support for the war in Iraq. Blue dots indicate
overall support among Democrats, and red dots indicate overall support among
Republicans. Relative Google Web Searches for “Iraq War” are superimposed over
the plot from 2004 onwards, with relative searches on the right hand side of the
figure indicating the correlation between literal searches for information about the
Iraq War and attitudes towards the conflict.
147
Finally, it is important to note that George W. Bush’s announcement of the
“surge” appears to have had little effect on information search or on public opinion
about the war. This is consistent with the expectation that there were few incentives
for the public to pay attention to the debate, to search for information, and to update
their attitudes about the conflict given that the surge did not constitute a crisis.
Instead, the minimal public responsiveness to the announcement is what one should
expect given the theory of incentives, information search, and public opinion about
foreign policy.
Figure 7.2 illustrates support for the war in Iraq from 2002-2008 by party. There
are three main points from this figure worthy of further consideration.
First, there is a clear partisan divide over time, where Republicans generally
supported the war, while Democrats increasingly opposed the war. In particular, it
is important to note that the difference between the parties truly became a chasm
during the second half of the foreign policy crisis, when the U.S. led the invasion
of Iraq. Prior to the crisis, over sixty percent of Democrats supported the war. By
the time of the invasion in March, this had dipped under forty percent of Democrats
before recovering as the war was won.
Second, while partisans of both major parties decreased their support for the war
over time, there is a distinct difference in the rate of decline. Republican support for
the war remained relatively consistent, and support for the war remained well above
sixty percent in 2008. On the other hand, after the end of the crisis, Democratic
support for the war plummeted, eventually bottoming out with less than twenty
percent of Democrats supporting the war by 2008.
148
Finally, and perhaps most interesting, there appears to be a correlation between
information search and attitudes about the war among Democrats. While it might
appear at first glance that the figure demonstrates crude partisan differences, the
astute observer might see important nuances. It is striking that every spike in infor-
mation search appears to correlate with a small but distinct drop in support for the
war among Democrats. Similarly, drops in information search appear to correlate
with small but distinct increases in support for the war. This is an interesting obser-
vation which appears to indicate that as Democrats search for information about the
war, they became less supportive. However, when they do not search for information,
attitudes towards the war became more positive.
In sum, the patterns of crisis, information search, and public opinion about the
Iraq War demonstrate some tentative support for the theory, but data limitations
mean that causal links cannot be drawn between the variables of interest. However,
the consideration of additional cases can help illustrate the plausibility of the theory
across time and space.
7.4 U.S. Public Opinion about Military Action in
Syria, 2012-2017
In August 2012, Barack Obama threatened to use U.S. military forces to intervene if
Bashar al-Assad crossed this “red line,” a humanitarian norm that the U.S. and its
allies couldnot tolerate. For the mostpart, the American publicinitially backedhim.
The few public opinion polls available during this period indicated broad support for
149
hypothetical military action against Syria if Assad used chemical weapons against
civilians.
Table 7.3: Crisis, Debate, and U.S. Military Action in Syria.
Foreign Policy Crisis Public Debate U.S. President
January 1-August 21, 2013 7 7 Barack Obama (D)
August 21-September 21, 2013 3 3 Barack Obama (D)
April 2017 3 7 Donald Trump (R)
Table 7.3 presents an overview of the foreign policy crises with Syria. On August
21, 2013 Syrian government forces used chemical weapons to kill civilians in neigh-
borhoods near Damascus. UN investigators confirmed that sarin gas was deployed
using ground-to-ground missiles, implicating the Syrian regime who were known to
possess sarin gas. As this information emerged, the Obama administration consid-
ered whether to follow through on their previous threat or to back down.
Although Obama stated that he supported military action in Syria in retaliation
for the attacks, he sought a vote in Congress to authorize the strike. A public debate
followed over the following weeks, during which support for military intervention in
Syria rapidly dissipated. Ultimately, the vote never materialized. Instead, the U.S.,
Russia, and Syria negotiated a diplomatic settlement to the dispute where Syria
would surrender its chemical weapons and sign the Chemical Weapons Convention
in September 2013.
In April 2017, new images emerged of another government-perpetrated attack
on Syrian civilians, this time in Khan Sheikhoun, in the Idlib Province. After being
shownimagesofvictims, DonaldTrumpstated, “WhathappenedinSyriaistrulyone
of the egregious crimes, and it shouldn’t have happened, and it shouldn’t be allowed
150
to happen”(Shear and Gordon 2017). In response, the U.S. launched 59 Tomahawk
cruise missiles at the Al Shayrat airfield, the base from which the chemical attack was
launched. This response was rapid - there were only 63 hours between the gas attack
and the U.S. retaliation, and no public consultation. The few public opinion polls
addressing the U.S. strike afterwards reported positive public sentiment towards the
military action, although there were significant differences between partisans of the
two major parties in the U.S.
7.4.1 Case Selection
American public opinion about Syria represents a good case for evaluating the plau-
sibility of the theory for several reasons. First, two different U.S. Presidents from
different parties faced very similar situations after it became clear Assad had used
chemical weapons. This allows for the rare opportunity to compare public responses
to remarkably similar incidents while there is variation in the executive branch of
government. Second, information search data is available through Google Trends for
the entire duration of the crises, a luxury that makes it useful for illustrating the
theory using population-level data.
Third, there is variation in the incentives to make correct decisions, as there are
three distinct periods for which one can draw distinctions: 1) hypothetical responses
to Assad’s use of chemical weapons before August 2013; 2) a period of intense public
discussion and debate after it was apparent that Assad had used chemical weapons,
ranging from late August to early September 2013; 3) the public response after
Trump’sunilateralactiontostrikeSyriainresponsetoAssadusingchemicalweapons
in April 2017. In the first and third periods, there were few incentives for the public
151
to make correct decisions because public opinion was inconsequential. In the second
period, during the public debate, there were greater incentives for the public to
make the correct decision as politicians contemplated a vote in the U.S. House of
Representatives, and there was time for people to lobby their representatives.
Finally, the case is substantively important. The protracted civil conflict in Syria
has arguably produced the greatest humanitarian crisis in recent memory. Since
the onset of the Arab Spring in 2011, Syria has been ravaged by a violent, complex
civil conflict between and among Bashar al-Assad’s incumbent regime, an informal
coalition of rebel groups, and ISIS forces. Syria is also a venue for geopolitical
competition between great powers, as Assad’s regime is supported by Russia, while
the U.S., the U.K., and their allies prefer Assad’s overthrowal.
Civilians have been caught in the crossfire, with large numbers of noncombatant
casualties leading to the mass displacement of people. Hundreds of thousands of
refugeeshavefledtoEuropeandneighboringMiddleEasterncountries, oftenatgreat
personal risk. Exacerbating the mass destruction of property and killing of civilians,
Assad has used chemical weapons against his own people on several occasions.
7.4.2 Data
The data for the case of Syria is provided by all publicly available surveys of public
opinion where participants were asked questions about their attitudes towards U.S.
military action in Syria. The exact question wording of the questions varied between
survey companies and the surveys were conducted intermittently, meaning that the
data is unsuitable for time series analysis. However, the data are useful for visual
representation of attitudes towards conflict in Syria, even if statistical analysis is
152
inappropriate given the limitations of the data. Appendix E.1 provides further in-
formation about each public opinion poll, sampling methods, margins of error, and
exact question wording.
The data for information search is provided by Google News searches of “Syria”
from locations in the United States from December 2012 to May 2017. This data is
aggregated to the week level.
7.4.3 Evidence from Public Opinion Polls
Although data limitations limit the ability of population-level data to address all
aspects of the theory, the data that is available provide support for the argument
proposed in the chapter. The ICB project codes the foreign policy crisis as having
started on August 21, as video of the chemical weapons attack emerged, and the
crisis terminated on September 21, as Syria complied with the agreed framework to
give up its chemical weapons (Brecher and Wilkenfeld 1997; Brecher et al. 2017).
Figure 7.3 illustrates support and opposition to U.S. military action in Syria in
response to Assad’s use of chemical weapons against his own people from 2012-2017
using all publicly available public opinion poll data.
There are three key implications from this figure. First, the figure provides sup-
port for the argument that people increase their information search when they per-
ceive there are incentives for accuracy. A significant increase in information search
about Syria occurred during the period of public debate in late August and early
September 2013 as the Obama administration considered intervention in Syria. The
absence of similar spikes following the Trump administration’s action in April 2017
can be explained by the fact that the action was unilateral - there was no period
153
of public debate, and there were therefore few incentives for people to make correct
decisions about the conflict.
Second, Figure 7.3 demonstrates that foreign policy attitudes changed over time,
coinciding with information search patterns, suggesting that information search may
have had an important role in affecting attitudes towards conflict in Syria. It appears
that the U.S. public is broadly supportive of conflict in Syria in response to Assad’s
use of chemical weapons against civilians. However, as people searched for informa-
tion, they updated their attitudes to become less supportive of U.S. intervention. In
2017, people did not search for information about Syria, and levels of support for
conflict were similar to those in early 2013 before it became clear Assad had used
chemical weapons.
Third, the figure suggests support for the argument that incentives led people to
update their attitudes, although it is not clear whether this had a direct effect or
whether this effect was mediated through information search. It is important to note
that a dramatic shift in attitudes towards the conflict occurred during the period of
public debate in early September, 2013, when the incentives for people to make the
correct decision regarding possible intervention were at their highest. When there
was little debate about action in 2017, it appears as though people did not update,
and instead reverted back to using information shortcuts to inform their opinions.
In sum, Figure 7.3 suggests that if one considers incentives to make correct de-
cisions as analogous to the onset of foreign policy crises, the figure suggests that
people responded to these incentives by searching for information, and updating
their attitudes with new information.
154
Figure 7.3: Support for U.S. Military Action in Syria, 2012-2017.
Plot of support for U.S. military action in Syria in response to Assad’s use of
chemical weapons, 2012-2017. The X axis indicates time, and the Y axis indicates
support and opposition for U.S. military action in Syria in response to Assad’s use
of chemical weapons against civilians. Green dots indicate total support and red
dots indicate total opposition to military conflict in Syria. “Don’t know”,
“Unsure”, and non-responses are not reported in the figure. Relative Google News
Searches for “Syria” are superimposed over the plot, with relative searches on the
right hand side of the figure indicating the correlation between literal searches for
information about the crisis and attitudes towards U.S. military action.
155
Figure 7.4: Support for U.S. Military Action in Syria by Party, 2012-2017.
Plot of support for U.S. military action in Syria in response to Assad’s use of
chemical weapons, 2012-2017. The X axis indicates time, and the Y axis indicates
support for U.S. military action in Syria in response to Assad’s use of chemical
weapons against civilians. Blue dots indicate support among Democrats and red
dots indicate support among Republicans. “Don’t know”, “Unsure”, and
non-responses are not reported in the figure. Relative Google News Searches for
“Syria” are superimposed over the plot, with relative searches on the right hand
side of the figure indicating the correlation between literal searches for information
about the crisis and attitudes towards U.S. military action.
156
Figure 7.4 illustrates net support for U.S. military action by party. There are
three important implications from this figure. First, the figure indicates that parti-
sanship or party cues do not explain the complete picture. Among partisans of both
main parties in the U.S., there was net support for military action if Assad crossed
the “red line” prior to the chemical attacks in August 2013.
However, this support disappeared during the public deliberations over military
action in late August and early September 2013, and this was particularly pro-
nounced among Republicans. In contrast, Trump’s response to Assad’s chemical
attack on civilians in April 2017 was almost unanimously supported by Republicans,
but Democrats remained largely opposed to it, or neutral at best. These results
indicate that partisans tend to adopt party lines, but that party positions are not
deterministic.
Second, attitudes among partisans converged in early September during the peak
of public debate about military action in Syria, and when Syria was firmly on the
public agenda. In other words, as the incentives for individuals to make the correct
decision about whether the U.S. should intervene, partisans of both persuasions
became less supportive of conflict. In the absence of incentives to make correct
decisions, such as when intervention was in response to hypothetical use of chemical
weapons (pre-August 2013) or after Trump had already taken unilateral action (April
2017), people adopted party cues.
Finally, it is important to note that the attitude convergence occurred simulta-
neously with an increase in Google News searches for information about Syria. This
provides some evidence in support of the plausibility of the incentives, information
search, and attitudes mediation hypothesis.
157
Collectively, thesefiguressuggestthatpeoplerespondedtoperceivedincentivesto
makecorrectdecisionsbysearchingforinformationduringtheperiodofpublicdebate
in September 2013, and updating their attitudes to converge on broad opposition to
militaryactioninSyria, inspiteofpartisanaffiliations. Thepolarizedresponsetothe
unilateral action taken by the Trump administration in 2017 could be driven by the
fact that people did not perceive any incentives to make the correct decision because
public opinion was not consulted, meaning few incentives to search for information
andmakeaninformedjudgmentaboutthebestresponse. Asaresult, peopleadopted
party cues.
7.5 U.S. Public Opinion about Military Action
against ISIS, 2013-2016
The final case for the illustration of the theory is U.S. public opinion about airstrikes
and the provision of U.S. troops to fight against the Islamic State of Iraq and Greater
Syria [ISIS, also known as ISIL]. The case presents an illustration of the theory with
the U.S. fighting against a non-state actor, but there are similar patterns of crisis,
information search, and the formation of attitudes about foreign policy as when the
public makes decisions about interstate foreign policies.
Table 7.4 presents an overview of the foreign policy crises with ISIS. ISIS emerged
out of the wreckage of the Iraq War, with the group emerging from the leftover
remnants of Al-Qaeda in Iraq, which was renamed the Islamic State in Iraq [ISI] by
Abu Ayyub al-Masri in 2006, before ISI merged with the Nusra Front to become ISIS
in April 2013.
158
In early 2014, ISIS rose to prominence as a threat to regional stability and gov-
ernance in Iraq and Syria as they captured the cities of Fallujah, Raqqa, Mosul,
Tikrit, and the Mosul Dam in quick succession. Concurrently, ISIS engaged in ex-
treme violence, sexual violence, mass atrocities, the destruction of sites of historical
and cultural importance, and it stands accused of human rights abuses, war crimes,
and ethnic cleansing.
On June 29, ISIS declared itself as a caliphate, calling themselves the “Islamic
State.” Spokesman Abu Mohamed al-Adnani called for Muslims to “gather around
your caliph, so that you may return as you once were for ages, kings of the earth
and knights of war” (Al Jazeera and Agencies 2014). However, the announcement of
the caliphate did not necessarily attract much interest among the average member
of the public.
The U.S. began airstrikes against ISIS positions in early August 2014, and on
September 10 Barack Obama declared that the United States would “degrade and
ultimately destroy the terrorist group known as ISIL” (Obama 2014). Obama out-
lined a four-pronged strategy where the U.S. would: 1) conduct airstrikes; 2) increase
supporttootherforcesfightingontheground; 3)preventISISattacks; and4)provide
humanitarian assistance to victims of ISIS. At this stage, there was little discussion
of the provision of U.S. troops to combat ISIS, but this changed in response to events
in the region, and closer to home.
However, two major events in particular prompted reconsideration about the
deployment of troops to fight ISIS militants in Syria and Iraq. First, among other
acts of violence, the spectacle of the video of ISIS-affiliated militants beheading 21
159
Table 7.4: Crisis, Debate, and U.S. Military Action against ISIS.
Crisis Public Debate U.S. President
2003 – June 29, 2014 Rise of ISIS 7 Barack Obama (D)
June 29, 2014 Declaration of Caliphate 7 Barack Obama (D)
September 10, 2014 Obama’s Vow to Confront ISIS 3 Barack Obama (D)
February 15, 2015 Mass Beheading of Egyptian Christians 3 Barack Obama (D)
November 13, 2015 Bataclan Attack in Paris 3 Barack Obama (D)
Egyptian Christians on the shores of the Mediterranean prompted a lot of attention
from observers in February 2015.
The terrorist attacks in Paris centered on the Bataclan Theatre on November
13, 2015 constituted the second event. 130 people were killed by a coordinated
attack from suicide bombers, and gunmen, and the attack drew large amounts of
attention from around the world. ISIS claimed responsibility for the attacks shortly
afterwards, and the attackers appear to have been planned in Syria and orchestrated
from a terrorist cell in Belgium.
Even among a series of other attacks in Iraq, Syria, and abroad, these two events
wereparticularlypowerfulinattractingnewscoverageanddrawingpeople’sattention
to the conflict in the region.
7.5.1 Case Selection
U.S. public opinion about military action against ISIS represents a good case for
the illustration of the theory of crisis, information search, and public opinion about
foreign policy for several reasons.
First, as the protagonist of a number of terrorist attack around the world, ISIS
constitutes perhaps the most salient external threat to the security of the American
160
public in recent years. Given the diffuse nature of the network, they are arguably
the most difficult threats to counter too, especially compared to traditional security
threats from foreign military forces. This is particularly important given that the
public may have a different calculus about the value of information given the per-
ceived proximity of the threat and arguably added incentives to make the correct
decision to counter the threat.
Second, as a non-state actor, ISIS also adds a different dimension to the study of
public opinion about foreign policy. Debates about the response to ISIS demonstrate
the complexity of forming foreign policy against non-state terrorist organizations.
This is particularly consequential given the dynamics of conflict in the twenty-first
century, with the increased incidence of intrastate conflict and conflict against non-
state actors amid the relative decline of interstate conflict.
Third, the two distinct foreign policy tools debated among policymakers and the
public make the case an interesting one to examine attitudes towards conflict. This
allows for the examination of the conditions under which the public are willing to
escalate their involvement in a foreign conflict from airstrikes with a minimal risk to
U.S. armed forces, to putting troops in harm’s way by putting “boots on the ground.”
7.5.2 Data
The data for public opinion polls about U.S. military action against ISIS is provided
by all publicly available surveys of public opinion where participants were asked
questions about their attitudes towards 1) U.S. airstrikes against ISIS militants in
Iraq and Syria; and 2) the deployment of U.S. troops to fight ISIS militants in Iraq
and Syria.
161
Question wording of the questions varied between survey companies, and again
the data was collected sporadically, preventing meaningful statistical analysis. How-
ever, there are important trends one can observe from the visual representation of
the data over time, even if causal relationships between variables cannot be identi-
fied. Appendix E.2 and Appendix E.3 provide further information about each public
opinion poll, sampling methods, margins of error, and exact question wording for the
two dependent variables of interest here: public opinion about U.S. airstrikes against
ISIS, and the deployment of troops to fight ISIS on the ground.
The information search data is provided by Google News searches of "ISIS" from
locations in the United States from the beginning of 2013 to the end of 2016, aggre-
gated to the week level to allow for comparisons over time.
7.5.3 Evidence from Public Opinion Polls
Public Support for U.S. Airstrikes against ISIS
Figure 7.5 illustrates support for U.S. airstrikes in Iraq and Syria against the Islamic
militants of ISIS using all available public opinion polls from 2013-2016. There are
two critical things to note from this figure.
First, the figure illustrates the dynamics of information search about ISIS. Al-
though there are many attacks and atrocities carried out by ISIS, information search
spiked in response to Obama’s speech with his aggressive vow to confront ISIS, and
the release of the video of the mass beheading of Egyptian Christians. However,
searches for information about ISIS were most prominent in response to the Paris
attacks, with the greatest amount of searches during that period. In short, the figure
162
illustrates that people search for information in response to major events, illustrating
the plausibility of the theory of crisis, information search, and public opinion about
foreign policy.
Second, the figure illustrates how attitudes towards airstrikes changed over time,
withthepublicbecomingincreasinglysupportiveofairstrikesagainstISISinresponse
to events. When ISIS declared itself a caliphate, roughly equal amounts of the public
supported and opposed strikes against the organization. However, after Obama
explicitly vowed to “degrage and ultimately destroy ISIS” (Obama 2014), there was
a pronounced increase in support for airstrikes against ISIS, and this persisted over
time as people maintained their willingness to confront ISIS through airstrikes.
At first, support and opposition to airstrikes hovered above forty percent. How-
ever, once Obama announced his plan to confront ISIS, support for airstrikes rose to
above seventy percent, suggesting that people changed their attitudes in response to
the speech and the information they actively sought to help them learn about the
situation. This support for his plan appears to have persisted over time, although the
absence of polls with questions about airstrikes outside of the crisis means one can-
not know whether this is true or whether the combination of crisis and information
search produced increases in support at the time of polling.
It is important to note here that the shift in attitudes is consistent with theories
of elite cues, where people supported the President once he announced his opposition
to ISIS and his commitment to destroy them. However, diving into a different depen-
dent, more costly variable reveals insights about the process of attitude formation as
people make judgments about foreign policy towards ISIS.
163
Figure 7.5: Public Support for U.S. Airstrikes against ISIS, 2013-2016.
Plot of support for U.S. airstrikes against ISIS, 2013-2016. The X axis indicates
time, and the Y axis indicates support for U.S. airstrikes against Islamic Militants
in Iraq and Syria. Green dots indicate total support and red dots indicate total
opposition. “Don’t know”, “Unsure”, and non-responses are not reported in the
figure. Relative Google News Searches for “ISIS” are superimposed over the plot,
with relative searches on the right hand side of the figure indicating the correlation
between literal searches for information about the crisis and attitudes towards U.S.
airstrikes against ISIS.
164
Public Support for the Deployment of U.S. Troops
There is a significant difference between public support for airstrikes and their will-
ingness to support putting their own troops in harm’s way. Even if an individual
supported airstrikes, it is a vastly different proposition to put troops on the ground,
as there is a much greater chance for the loss of American life on the ground than in
the air, especially given the U.S. dominance of the air space in the region. As such, it
is important to consider support for the deployment of troops to fight against ISIS.
Figure 7.6 indicates U.S. public opinion towards the deployment of American
troops to fight against ISIS in Syria and Iraq. As noted before, information search
peaked during the Paris attacks, while there are notable increases in response to
Obama’s speech outlining the American strategy to confront ISIS, the beheading
of Egyptian Christians, and especially the Paris attacks centered on the Bataclan
Theatre.
The figure depicts gradual but sustained changes in attitudes about deployment
of U.S. troops to fight ISIS over time, reflecting significant changes in public attitudes
towards the conflict. While support for the deployment of troops increased after the
Obama speech, the plurality of voters continued to oppose putting boots on the
ground in the fight against ISIS.
However, a dramatic change occurred in response to the beheading of Coptic
Christians in February 2015, and over time support for the deployment of troops
continuedtoincreasegraduallyovertime. TheParisattacksbroughtanotherincrease
in support for the use of troops to counter ISIS, with a number of public opinion
polls placing support above fifty percent at the same time as information search
dramatically increased.
165
Figure 7.6: Public Support for the Deployment of U.S. Troops to Fight against ISIS,
2013-2016.
Plot of support for the Deployment of U.S. Troops against ISIS, 2013-2016. The X
axis indicates time, and the Y axis indicates support for the deployment of U.S.
troops to fight against Islamic Militants in Iraq and Syria. Green dots indicate
total support and red dots indicate total opposition. “Don’t know”, “Unsure”, and
non-responses are not reported in the figure. Relative Google News Searches for
“ISIS” are superimposed over the plot, with relative searches on the right hand side
of the figure indicating the correlation between literal searches for information
about the crisis and attitudes towards the deployment of U.S. troops to fight ISIS.
166
In sum, the figure suggests that the combination of events and information search
contributed towards increases in support of the deployment of troops to fight against
ISIS. The costly provision of troops was initially opposed, but over time as ISIS
atrocities mounted, it appears as though people searched for information in response
to these events, and they updated their attitudes as a result.
7.6 Discussion
The evidence from the three cases presented here does not demonstrate irrefutable
proof of the theory of crisis, information search, and public opinion about foreign
policy. However, there is sufficient evidence to warrant further investigation into
these phenomena using data that can test the theory empirically in the future. Ob-
servational data from public opinion polls of the three major foreign policy debates
for the U.S. in the last two decades suggests the plausibility of two key components
of the theory presented here.
First, people search for information during periods of crisis. Far from being an
irrational, uninformed public, the evidence presented in this chapter suggests people
searchforinformationduringperiodsofcrisis. Inshort, peoplesearchforinformation
when they perceive it is in their interests to make correct decisions, when the stakes
are at their highest.
Second, the chapter suggests the plausibility of the argument that abrupt changes
in foreign policy attitudes may be related to information search. The evidence in
this chapter suggests there is a relationship between crisis, information search, and
167
attitudes about conflict. It appears as though people search for information during
crisis, and then update their attitudes during this period, independent of elite cues.
The evidence suggests that people can fulfill their democratic obligations as an
informed citizenry that can produce informed decisions using available information
when it matters most. This is a significant contribution to the understanding of
public opinion about foreign policy, and it presents many further opportunities to
build on the insights of this chapter to gain insights into how people form attitudes
about international affairs. Future research should use longitudinal individual-level
data to illustrate that the theory of crisis, information search, and public opinion
about foreign policy holds up in the real world as well as the laboratory.
168
Chapter 8
Conclusion
To understand how people think about international affairs, it is important to under-
stand the conditions under which people make decisions about foreign policy. This
dissertation examines how incentives for accuracy influence people’s political knowl-
edge, their ability to make accurate predictions about international relations, and
their attitudes about foreign policy.
Evidence from experimental and observational data demonstrates that people
given incentives for accuracy update their attitudes about foreign policy. In short,
context matters. People search for information, improve their political knowledge,
and change their mind about foreign policy when there are incentives for accurate
decisions about foreign policy. When it matters most, people fulfill their democratic
obligations as an informed citizenry.
This concluding chapter proceeds in four sections. First, I summarize the argu-
ments put forward in the dissertation, and the evidence presented in support of the
theory of incentives, information, and public opinion about foreign policy. Second,
169
I discuss the policy implications of the dissertation. Third, I outline several areas
for future research to build on the dissertation in the pursuit of an improved un-
derstanding of public opinion about foreign policy. Finally, the chapter ends with a
brief conclusion.
8.1 Summary of Arguments and Evidence
The dissertation argued that incentives lead to information search, which in turn
leads to people improving their political knowledge, improving the accuracy of their
predictions about international affairs, and ultimately updating their attitudes about
foreign policy.
Chapter 2 introduced the motivation for the project. The literature review il-
lustrated how previous scholarship helped develop a good understanding of public
opinion about foreign policy, but significant gaps remain. In particular, I high-
lighted how previous scholarship fails to address the conditions under which people
shift their demand for information about international relations, and how this affects
public opinion about foreign policy.
In Chapter 3, I addressed this shortcoming in previous work. In the chapter
I introduced a general theory of incentives, information, and public opinion about
foreign policy. I outlined a decision theoretic model that illustrates the conditions
under which people are motivated to pay attention to international affairs. This
formed the basis of the arguments expanded upon throughout the dissertation, with
a variety of different applications.
170
Chapter 4 examined how people respond to financial incentives in an incentivized
online survey experiment. The chapter demonstrated that people given incentives
for accuracy search for information, and that they use this information to improve
their political knowledge about international affairs. The chapter also illustrated
that people given lottery incentives become less accurate in their predictions about
international affairs as they are demotivated by the uncertainty of receiving a reward
for their efforts.
Finally, the results showed that people update their attitudes about foreign pol-
icy as the result of receiving incentives for accuracy, searching for information, and
improving their political knowledge about international affairs. Collectively, these
results demonstrated that people respond to incentives for accuracy, changing their
patterns of information search, with implications for their political knowledge, po-
litical predictions, and attitudes about foreign policy. They suggested that public
opinion is contingent on the conditions under which people perceive incentives for
accuracy in foreign policy decision making.
Chapters 5, 6, and 7 built on the experimental findings by illustrating how people
respond to perceptions of incentives for accuracy in the real world. These chapters
examined how differences in perceptions of incentives for accuracy influenced public
opinion about foreign policy.
Chapter 5 used multilevel regression of public opinion data to demonstrate that
incentives for accuracy influence attention to international affairs. The results sug-
gested that registered voters systematically devoted more attention to the news sto-
ries about U.S. military action in Syria.
171
Chapter 6 analyzed data from the E.U. Referendum in the United Kingdom to
demonstrate that people’s voting intentions were contingent on their perceptions of
the impact of the U.K. leaving the E.U. People who perceived the stakes as being
the highest were those least likely to vote Leave and most likely to vote Remain,
demonstratingtheimportanceofperceptionsofincentivesforaccuracyduringforeign
policy debates.
Finally, Chapter 7 featured an extended plausibility probe of the theory, exam-
ining how crisis affects people’s information search behavior and their foreign policy
attitudes over time. Data limitations meant no causal relationship could be inferred
from the evidence presented in the chapter. However, people did appear to search for
information about foreign policy debates during periods of crisis, and this coincided
with shifts in attitudes towards different foreign policies.
Collectively, the dissertation provided a range of different evidence suggesting
that scholars must pay greater attention to the conditions under which people make
decisions about foreign policy. The evidence suggests that people respond to per-
ceived incentives for accuracy by searching for information, improving their knowl-
edge, and updating their attitudes when they consider the stakes of the foreign policy
debate are high.
8.2 Policy Implications
There are important policy implications of this research. Politicians and policymak-
ers often attempt to persuade voters to adopt their policy position, to encourage
likely voters to get to the polls, and to convince swing voters to cast their vote
172
for them during elections. Particularly in an age of polarized politics in the United
States, this means that there is a narrow group of voters for politicians to target their
messages towards, given that fellow partisans are already on board, and partisans of
the other major party are unlikely to be won over by direct attempts at persuasion.
The evidence from the dissertation suggests that direct attempts at persuasion
about foreign policy might not be as efficient as persuading voters that their opinion
matters, and that thestakes are high. Given the empirical resultsin this dissertation,
one should expect to see that if the public is convinced that there are incentives for
making correct decisions, they are likely to search for information, improve their
political knowledge, and update their attitudes about foreign policy.
If political actors attempt to persuade voters that the stakes are high and that
there are important incentives for accuracy, they could be liable to demand informa-
tion and ultimately change their attitudes. This might be a more effective way of
communicating with the public than simply trying to persuade people to adopt one’s
own position, especially when this is unlikely given the strong influence of motivated
reasoning and party ID in the absence of perceived incentives for accuracy.
8.3 Future Research
It is important to acknowledge that although this dissertation presents evidence in
support of the theory of incentives, information, and public opinion about foreign
policy, it has significant limitations that should be addressed in future work. I hope
173
this dissertation marks a first step in the right direction for an improved understand-
ing of public opinion about foreign policy, but there are exciting opportunities for
future research to further build on the dissertation.
8.3.1 Incentives for Accuracy
Future research could evaluate how the presentation of incentives influences people’s
political behavior. The incentives are framed positively in the experiment in Chapter
4,wherepeoplereceivearewardforcorrectanswerstobatteriesofpoliticalknowledge
and prediction questions. However, the incentives to make correct decisions could be
negative, and foreign policy crises can be considered as periods where people do not
receive rewards for making the right decision, but they are punished for taking the
wrong decision. For example, making the wrong decision about a potential conflict
might be considered as costly for individuals, but making the correct decision might
not be considered as a benefit.
Depending on how incentives are framed, there could be systematic differences in
individual behavior (Kahneman and Tversky 1979; Tversky and Kahneman 1981).
Furthermore, negatively framed incentives could induce anxiety on the part of par-
ticipants, altering their cognitive response to the treatment (Gadarian 2010). As a
result, future studies could alter the treatments so that people lose money for inac-
curacy, instead of earning money for accurate responses. This could lead to different
results, and they could potentially better replicate how people behave during foreign
policy crises.
174
8.3.2 Attention
Future research should examine how people respond to incentives for accuracy in
their attention to foreign policy debates. Although the results in Chapter 5 provide
support for the theory of the conscientious voter, it is important to note that some
important limitations remain.
First, theobservationaldataonlyreferstoasinglecaseoverthreedifferentpolling
periods.
1
ItcouldbethattheSyriancasewasananomaly, andthatpoliticalbehavior
among voters does not depart from the rest of the public so dramatically during other
periods. Futureresearchshouldincludeadditionalcasesforcomparison, todetermine
whether the evidence from the multilevel models presented here generalize to other
cases.
Second, the observational nature of the data mean there could be unobserved
variables that could confound the analysis. This is an especially important concern
given that there could be a relationship between voter type and political interest
or civic duty, or other factors that one could reasonably expect to influence atten-
tion to international affairs. Future research could isolate the effect of voter type in
experiments, where assigning the ability to vote for a foreign policy could affect peo-
ple’s attention and their attitudes about foreign policy. Random assignment would
eliminate the possibility of unobserved confounding variables, helping to establish a
1
It is important to note the benefits of attention to particular issues such as the Syrian case to
better understand political behavior of voters. Generalized political interest and news consumption
are less useful measures of attention than particularized attention to an issue because more specific
attention captures responses to a given foreign policy debate. If we think of attention as motivated
and active, measures of attention to a specific issue helps shed light on the conditions under which
peoplemightconstrainforeignpolicymakers, whenpeopledevoteattentiontoforeignpolicydebates.
175
causal relationship between perceptions of incentives as determined by voter type,
attention, and attitudes about foreign policy.
Third, future studies could measure attention directly. This data in Chapter
5 relies on participants’ self-reports for measures of attention to foreign affairs in
the polls conducted by Pew Research Center. Future research could unobtrusively
observepoliticalbehaviorusingcomputerswithsoftwaretrackingparticipants’search
patterns, extending the results of this chapter using behavioral measures.
8.3.3 Information Search
In the incentivized online survey experiment in Chapter 4, participants responded to
incentives for accuracy by searching for information. Future studies should triangu-
late these results in studies of information search using available observational data.
For example, panel surveys like those offered by YouGov Pulse provide unobtrusive
measures of panelists’ online information search behavior, which could to provide
empirical evidence of the same patterns of information search. Scholars could also
use aggregate public opinion polls and news consumption data from periodical mag-
azines and television news to build a greater understanding of the conditions under
which people search for information about international affairs.
Furthermore, future experiments could use different designs to better replicate
how information search occurs in the real world. First, an important limitation of
the experiment is that the experiment did not explicitly attach costs to information.
However, there is a cost of searching for information in the real world, as people’s
attention is scarce and people have to pay for many sources of political information.
People have an infinite number of alternative things they could attend to instead of
176
using search engines to look up answers to questions about political knowledge, and
make predictions about international affairs. To explicitly include the costs of infor-
mation search, future studies should introduce costs for searching for information,
manipulating bonuses to better replicate the costs of information acquisition in the
real world.
Second, future studies could use alternative control conditions. The experiment
featured a common introduction that encouraged all participants to search for an-
swers to the questions. Although the control group did not receive incentives to get
answers correct, this could mean that individuals exposed to the control condition
searched for information more than they would have done otherwise. This means
that the differences between the treatment groups represent a difficult test of the the-
ory, meaning that one can be confident about the effects of the incentives treatments
compared to the control group. However, the baseline amount of political knowledge
might be lower than the scores achieved by the control group in this experiment.
Finally, future studies should examine how emotion affects individuals’ informa-
tion search behavior. As discussed in the literature review, people may actively avoid
information if they anticipate that they will respond poorly to new information (Al-
bertson and Gadarian 2015; Caplin and Leahy 2001; Huddy et al. 2005; Kőszegi 2003;
Maslow 1963). Further research should examine how affect and emotion influence
information search in both experimental and observational settings.
8.3.4 Information and Social Media
Future research should also examine how social media shapes the information en-
vironment, and its implications for public opinion about foreign policy. Although
177
information has never been more accessible, the credibility of online information
is often difficult than ever to discern, especially when it is purveyed by unverified
sources and distributed on social media.
Future research should examine patterns of attention, information search, and
political attitudes about foreign policy from social media accounts to unobtrusively
measure how people respond to external conditions. There are distinct advantages
of this research agenda.
First, scholars can track individuals over time, allowing for individual-level anal-
ysis of how people respond to foreign policy crisis. Second, the data is accessible and
publicly available through APIs. Analysis of social media data should be a venue for
further tests of the theory introduced in this dissertation using unobtrusive measures
of political attitudes and behavior.
8.3.5 The Public as Democratic Constraint
It is beyond the scope of this project to examine how incentives for accuracy and
demand for information influence the public’s ability and willingness to constrain
foreign policymakers. However, it is reasonable to expect that the public’s constraint
is contingent on the conditions under which decisions are made.
If the public is attentive and engaged in a foreign policy debate, this is likely to
affect political actors’ ability to pursue foreign policies. In particular, it is likely that
when the public is motivated to search for information, and improve their political
knowledge, it will be more politically difficult for politicians to adopt policies that
differ from the public’s preferences. During such circumstances, it is likely that the
public may be more influential than what scholars commonly expect (Druckman and
178
Jacobs 2015; Holsti and Rosenau 1984; Jacobs and Shapiro 2000; Katz 2000; Mueller
1973). There is an an exciting research agenda ahead to examine the conditions
under which the public acts to constrain foreign policy.
8.3.6 Beyond Foreign Policy: Incentives, Information, and
Public Opinion
Finally, future research should broaden the scope of inquiry beyond foreign policy
to examine how the theory of incentives and information applies to other political
topics and areas of public debate. For many political debates, there are cycles of
attention where policy adoption might be contingent on the public’s willingness to
search for information, improve their political knowledge, and shift their attitudes
in line with new information.
There are many examples of policy areas that could benefit from research using
this theoretical framework. For instance, it is likely that people’s perceptions of
incentives for accuracy and their willingness to search for information affect their
attitudes about climate change, gun control, disaster risk reduction, financial regu-
lation, homeland security, refugees, and crime, to name but a few examples. Future
research should broaden the application of the theory to other policy areas.
8.4 Conclusion
In sum, I hope this dissertation represents a step in the right direction for the study
of public opinion about foreign policy, but there is more exciting work ahead to
continue building on the arguments and evidence presented within it.
179
The dissertation demonstrates that people respond to incentives for accuracy,
using them as a catalyst to search for information, improve their political knowledge,
make predictions about international affairs, and update their attitudes as a result
of this process.
I also examine how individual-level variation in voter type affects people’s at-
tention to foreign policy debates, and how perceptions of impact explains voting
intentions in the Brexit referendum in the United Kingdom. Finally, the disserta-
tion showed that people appear to search for information about international affairs
during periods of crisis, and using this information to update their attitudes about
foreign policy.
IR scholarship focuses on understanding the causes and effects of foreign policy
where the stakes are high. However, theoretical and methodological approaches
to the study of public opinion and foreign policy have diverged from this simple
premise, instead focusing on studies where the stakes could not be lower. This
dissertation attempts to address this significant shortcoming in the pursuit of a
better understanding of information, public opinion, and international relations.
180
Appendix A
The Full Experimental Design
181
A.1 Demographics 1
1. Please enter your age.
2. What is the highest level of school you have completed?
• No formal education (0)
• 1st, 2nd, 3rd, or 4th grade (1)
• 5th or 6th grade (2)
• 7th or 8th grade (3)
• 9th grade (4)
• 10th grade (5)
• 11th grade (6)
• 12th grade, no diploma (7)
• High School Graduate - High School Diploma or the equivalent (8)
• Some college, no degree (9)
• Associate degree (10)
• Bachelor’s degree (11)
• Master’s degree (12)
• Professional or Doctorate degree (13)
3. Please indicate what you consider your racial background to be. We greatly
appreciate your effort to describe your background using the standard cate-
gories provided. These race categories may not fully describe you, but they
do match those used by the Census bureau. It helps us compare our survey
respondents to the U.S. population.
Please check one or more categories below to indicate what race(s) you consider
yourself to be.
• White (1)
• Black or African American (2)
• Hispanic or Latino (3)
• American Indian or Alaska Native (4)
• Asian Indian (5)
182
• Chinese (6)
• Filipino (7)
• Japanese (8)
• Korean (9)
• Vietnamese (10)
• Other Asian (11)
• Native Hawaiian (12)
• Guamanian or Chamorro (13)
• Samoan (14)
• Other Pacific Islander (15)
• Some other race (16)
4. What is your current gender identity?
• Male (0)
• Female (1)
• Transgender (2)
• I do not identify as female, male, or transgender (3)
SQ1 Political scientists are often interested to know how much people know politi-
cal institutions. In a democracy, it is very important that people know enough
about politics to make an informed decision when they vote for political can-
didates. To show to us that you have read this far, please ignore the question
and select Four years as your answer to the question.
How long is the term of office for a U.S. Senator?
• Two years (1)
• Four years (2)
• Six years (3)
5. The next question is about the total income of YOUR HOUSEHOLD for the
PAST 12 MONTHS. Please include your income PLUS the income of all mem-
bers living in your household (including cohabiting partners and armed force
members living at home). Please count income BEFORE TAXES and from all
183
sources (such as wages, salaries, tips, net income from a business, interest, div-
idends, child support, alimony, and Social Security, public assistance, pensions,
or retirement benefits).
• Under $25,000 (1)
• Between $25,000 and $49,999 (2)
• Between $50,000 and $74,999 (3)
• Between $75,000 and $99,999 (4)
• Between $100,000 and $124,999 (5)
• Between $125,000 and $149,999 (6)
• Over $150,000 (7)
6. Are you now married, widowed, divorced, separated, never married, or living
with a partner?
• Married (1)
• Widowed (2)
• Divorced (3)
• Separated (4)
• Never married (5)
• Living with partner (6)
A.2 Experimental Treatments
A.2.1 Common introduction
In the next part of the survey, you will be asked some questions about politics
and international relations. Many people don’t know the answers to these ques-
tions, butitishelpfultousifyouanswer, evenifyou’renotsureofthecorrectanswer.
There is NO DECEPTION in this study.
I encourage you to search the internet for answers you may not know off the
top of your head, and to take a guess on every question. At the end of the survey,
you will see the correct answers to the questions, allowing you to see how many
184
questions you answered correctly.
Please note that some questions refer to your own opinion, where there is not a
’correct’answer. Theintroductiontothesequestionsindicatesinboldandunderlined
text that these questions relate only to your opinion.
A.2.2 Control Condition
The first set of questions involve assessments of political knowledge. The answers
will be revealed by the end of the survey.
The second set of questions involve assessments of your political predictions in
2017. These will be revealed on September 7, 2017.
A.2.3 Bonus Treatment
There are two sets of questions for which there are correct answers. I will pay you a
bonus for answering these questions correctly. You will earn $0.50 for every correct
answer you give.
So, if you answer 5 of the 8 questions correctly, you will earn $2.50 as a bonus. If
you answer all 8 of the questions correctly, you will earn $4.00 as a bonus. The more
questions you answer correctly; the more money you will earn in the form of a bonus.
The first set of questions involve assessments of political knowledge. The
answers will be revealed by the end of the survey, and you will be awarded any
bonusyouearnedforthesequestionslessthan24hoursafteryoucompletethesurvey.
The second set of questions involve assessments of your political predictions in
2017. These will be revealed on September 7, 2017, and you will be awarded any
bonus you earned for these questions on that date.
1
A.2.4 Random Bonus Treatment
There are two sets of questions for which there are correct answers. I will pay you a
bonus for answering one question correctly from each set of questions. The relevant
1
Adapted from Prior and Lupia (2008).
185
question will be selected randomly. You will earn $4 for answering this question
correctly.
So, if you answer 5 of the 8 questions correctly and the randomly selected
question is one of the 5 you answered correctly, you will earn $4.00 as a bonus. If the
randomly selected question is one of the 3 questions you answered incorrectly, you
will earn $0.00 as a bonus. The more questions you answer correctly; the greater
the likelihood that you will receive more money in the form of a bonus.
The first set of questions involve assessments of political knowledge. The
answers will be revealed by the end of the survey, and you will be awarded any
bonusyouearnedforthesequestionslessthan24hoursafteryoucompletethesurvey.
The second set of questions involve assessments of your political predictions in
2017. These will be revealed on September 7, 2017, and I will email you with a
report about how well you predicted future events. You will be awarded your bonus
for these questions on that date.
A.2.5 Lottery Treatment
There are two sets of questions for which there are correct answers. For every
question you answer correctly, you earn a ticket to a lottery for a $400 cash prize
for that block of questions.
So, if you answer 5 of the 8 questions correctly, you will earn 5 entries into the
lottery. If you answer all 8 of the questions correctly, you will earn 8 tickets to
the lottery as a bonus. The more questions you answer correctly; the greater the
likelihood that you will win the $400 cash prize for each set of questions.
The first set of questions involve assessments of political knowledge. The answers
will be revealed by the end of the survey. The lottery will be drawn and the $400
cash prize will be awarded to the winner less than 24 hours after you complete the
survey.
The second set of questions involve assessments of your political predictions in
2017. These will be revealed on September 7, 2017, and I will email you with a
report about how well you predicted future events. The lottery will be drawn and
the $400 cash prize will be awarded to the winner on that date.
186
A.3 Political Knowledge Questions
1. AccordingtoNewAmerica,anon-partisanthinktank,whatisthefigureCLOS-
EST to the total number of people killed in the United States from Jihadist
terrorist attacks since 9/11?
• 0 (0)
• 100 (1)
• 200 (2)
• 300 (3)
• 400 (4)
• 500 (5)
• 600 (6)
• 700 (7)
• 800 (8)
• 900 (9)
• Over 1,000 (10)
2. The 2015 Paris Agreement addressed climate change and imposed targets for
states to mitigate against climate change. 195 states signed the treaty. How
many states have since ratified the treaty, committing to the targets outlined
in the agreement?
• Between 0 and 19 (0)
• Between 20 and 39 (1)
• Between 40 and 59 (2)
• Between 60 and 79 (3)
• Between 80 and 99 (4)
• Between 100 and 119 (5)
• Between 120 and 139 (6)
• Between 140 and 159 (7)
• Between 160 and 179 (8)
• Between 180 and 199 (9)
187
• Over 200 (10)
3. The two major political parties in the United States have different positions
on what to do about unauthorized immigrants.
The 2016 Republican Party Platform said that “we oppose any form of
amnesty for those who, by breaking the law, have disadvantaged those who
have obeyed it.”
The 2016 Democratic Party Platform said that they would work “to end the
forced and prolonged expulsion from the country that these immigrants endure
when trying to adjust their status.”
According to the Pew Research Center, how many unauthorized immigrants
lived in the U.S. in 2014?
• 1.1 million (1)
• 3.1 million (2)
• 5.1 million (3)
• 7.1 million (4)
• 9.1 million (5)
• 11.1 million (6)
• 13.1 million (7)
• 15.1 million (8)
• 17.1 million (9)
• 19.1 million (10)
4. During the 2016 Presidential election, the candidates from the two major
political parties in the U.S. had different opinions of Russian President
Vladimir Putin.
Republican nominee Donald Trump said that Putin had “done a really great
job of outsmarting the U.S.” and that he is “highly respected within his own
country and beyond.”
188
Democratic nominee Hillary Clinton said that Putin “is at heart, a bully”
and that he is “a very cold-blooded, calculating former KGB agent” who is
determined to “enrich himself and his closest colleagues.”
In early 2014, separatist forces invaded the Ukrainian territory of Crimea, with
support from the Russian government. Russia subsequently annexed Crimea.
Along with several other countries, the U.S. imposed sanctions on Russia as a
result.
According to the Office of the United Nations High Commissioner for Human
Rights, as of November 2016, what is the figure CLOSEST to the total number
of people killed during the Ukrainian conflict?
• 0 (0)
• 1,000 (1)
• 2,000 (2)
• 3,000 (3)
• 4,000 (4)
• 5,000 (5)
• 6,000 (6)
• 7,000 (7)
• 8,000 (8)
• 9,000 (9)
• 10,000 (10)
5. Several branches of the U.S. Government are investigating the possibility that
Russian hacks of Democratic National Committee material may have affected
the electoral process.
Furthermore, members of Trump’s campaign and his Administration are being
investigated for possible collusion with Russian operatives.
Who is the Special Counsel for the Department of Justice investigation into
possible collusion between President Trump’s campaign and Russia?
189
• Robert Mueller (0)
• Jared Kushner (1)
• Carter Page (2)
• Scott Pruitt (3)
• Jeff Sessions (4)
• James Comey (5)
• Rod Rosenstein (6)
• Jason Chaffetz (7)
• Mike Pence (8)
• Jay Sekulow (9)
• Sean Spicer (10)
6. The United States helped negotiate a major nuclear deal framework with Iran
in 2015.
The reaction to the deal was mixed in the United States House of Congress.
Most Democrats in Congress supported the deal, but most Republicans
opposed it.
Please indicate whether each of the following countries generally supported or
opposed the nuclear deal:
• Germany (Supported or Opposed)
• Iraq (Supported or Opposed)
• Israel (Supported or Opposed)
• Russia (Supported or Opposed)
• Saudi Arabia (Supported or Opposed)
• Turkey (Supported or Opposed)
• United Kingdom (Supported or Opposed)
190
7. National Security experts are concerned with recent North Korean tests of
long range ballistic missiles.
According to the CSIS Missile Defense Project, what is the largest maximum
known range of North Korean missiles that are currently operational?
• Up to 250 km (0)
• Up to 1,000 km (1)
• Up to 3,000 km (2)
• Up to 5,000 km (3)
• Up to 8,000 km (4)
• Up to 10,000 km (5)
• Up to 11,500 km (6)
• Up to 13,000 km (7)
• Up to 15,000 km (8)
• Up to 17,000 km (9)
• Up to 20,000 km (10)
8. What was the U.S. trade deficit with China in 2016?
• There was no trade deficit with China (0)
• Between $100 billion and $200 billion (1)
• Between $200 billion and $300 billion (2)
• Between $300 billion and $400 billion (3)
• Between $400 billion and $500 billion (4)
• Between $500 billion and $600 billion (5)
• Between $600 billion and $700 billion (6)
• Between $700 billion and $800 billion (7)
• Between $800 billion and $900 billion (8)
• Between $900 billion and $1 trillion (9)
• Over $ 1 trillion dollars (10)
191
A.4 Prediction Questions
1. To prevent a terrorist attack, will the Department of Homeland Security ban
laptops on domestic flights in the United States before September 7, 2017?
• No (0)
• Yes (1)
2. Will the National Snow and Ice Data Center report the extent (area) of Arctic
Sea Ice on September 7, 2017 to be below or above the extent (area) of Arctic
Sea Ice on September 7, 2016?
• Below (0)
• Above (1)
3. Will construction on a border wall along the U.S.-Mexico start before Septem-
ber 7, 2017?
• No (0)
• Yes (1)
4. Will the U.S. lift sanctions on Russia before September 7, 2017?
• No (0)
• Yes (1)
5. Will the House of Representatives begin impeachment proceedings against
President Trump before September 7, 2017?
• No (0)
• Yes (1)
6. Will the U.S. impose economic sanctions on Iran for activities related to a
nuclear program before September 7, 2017?
• No (0)
• Yes (1)
7. WillalethalconfrontationinvolvingthenationalmilitaryforcesofNorthKorea
and the United States result in a fatality for either side before September 7,
2017?
192
• No (0)
• Yes (1)
8. Will President Trump impose trade sanctions on China to reduce the trade
deficit between the two countries before September 7, 2017?
• No (0)
• Yes (1)
A.5 Information Search Questions
1. During this survey, did you search for information so that you could answer
the questions correctly?
• No (0)
• Yes (1)
2. During this survey, did you rely more on your own knowledge or did you seek
information so that you could answer the questions correctly?
• I relied more on my own knowledge to answer the questions (0)
• I relied more on information I searched for to answer the questions (1)
3. Answer If During this survey, did you search for information...Yes Is Selected
If you used them, which internet search engines did you use? Please tick all
that apply.
• Baidu (1)
• Bing (2)
• Google (3)
• Yahoo! (4)
• Other (5)
4. Answer If During this survey, did you search for information...Yes Is Selected
What sources did you use to answer the questions? Please tick all that apply.
193
• Academic sources - e.g. Google Scholar, JSTOR, university libraries (1)
• Blogs - e.g. FiveThirtyEight, The Monkey Cage (2)
• Books (3)
• News sources - e.g. Fox News, MSNBC, New York Times, NPR (4)
• Official sources - e.g. Government Factsheets, Government Documents,
Acts of Congress (5)
• Results from Search Engines (6)
• Social media sources - e.g. Facebook, Snapchat, Twitter (7)
• Wikipedia (8)
• Other ? please specify (9)
5. Answer If During this survey, did you search for information...Yes Is Selected
On a scale from 0 (extremely unconfident) to 10 (extremely confident), what
is your overall confidence about the TRUTHFULNESS of the information ac-
quired from sources you used to answer the questions?
• 0 (0)
• 1 (1)
• 2 (2)
• 3 (3)
• 4 (4)
• 5 (5)
• 6 (6)
• 7 (7)
• 8 (8)
• 9 (9)
• 10 (10)
6. Answer If During this survey, did you search for information...Yes Is Selected
On a scale from 0 (extremely unconfident) to 10 (extremely confident), what is
your overall confidence about the VALIDITY of the information of the sources
you used to answer the questions?
194
• 0 (0)
• 1 (1)
• 2 (2)
• 3 (3)
• 4 (4)
• 5 (5)
• 6 (6)
• 7 (7)
• 8 (8)
• 9 (9)
• 10 (10)
7. Of the political knowledge questions with correct answers, how many questions
do you think that you successfully answered?
• 0 (0)
• 1 (1)
• 2 (2)
• 3 (3)
• 4 (4)
• 5 (5)
• 6 (6)
• 7 (7)
• 8 (8)
• 9 (9)
• 10 (10)
8. Of the political prediction questions with correct answers, how many questions
do you expect to have successfully answered?
• 0 (0)
• 1 (1)
195
• 2 (2)
• 3 (3)
• 4 (4)
• 5 (5)
• 6 (6)
• 7 (7)
• 8 (8)
• 9 (9)
• 10 (10)
A.6 Answers to Political Knowledge Questions
(Randomly presented to participants be-
fore/after political attitudes questions)
1. AccordingtoNewAmerica,anon-partisanthinktank,whatisthefigureCLOS-
EST to the total number of people killed in the United States from Jihadist
terrorist attacks since 9/11?
• 0 (0)
• 100 (1)
• 200 (2)
• 300 (3)
• 400 (4)
• 500 (5)
• 600 (6)
• 700 (7)
• 800 (8)
• 900 (9)
• Over 1,000 (10)
The correct answer was 100. From October 2001 to June 30, 2017,
Jihadist terrorist attacks have killed 95 people in the United States
in total.
196
2. The 2015 Paris Agreement addressed climate change and imposed targets for
states to mitigate against climate change. 195 states signed the treaty. How
many states have since ratified the treaty, committing to the targets outlined
in the agreement?
• Between 0 and 19 (0)
• Between 20 and 39 (1)
• Between 40 and 59 (2)
• Between 60 and 79 (3)
• Between 80 and 99 (4)
• Between 100 and 119 (5)
• Between 120 and 139 (6)
• Between 140 and 159 (7)
• Between 160 and 179 (8)
• Between 180 and 199 (9)
• Over 200 (10)
The correct answer was between 140 and 159. As of June 30, 2017,
148 states have ratified the Paris Agreement.
3. The two major political parties in the United States have different positions
on what to do about unauthorized immigrants.
The 2016 Republican Party Platform said that "we oppose any form of
amnesty for those who, by breaking the law, have disadvantaged those who
have obeyed it."
The 2016 Democratic Party Platform said that they would work "to end the
forced and prolonged expulsion from the country that these immigrants endure
when trying to adjust their status."
According to the Pew Research Center, how many unauthorized immigrants
lived in the U.S. in 2014?
• 1.1 million (1)
197
• 3.1 million (2)
• 5.1 million (3)
• 7.1 million (4)
• 9.1 million (5)
• 11.1 million (6)
• 13.1 million (7)
• 15.1 million (8)
• 17.1 million (9)
• 19.1 million (10)
The correct answer was that there were 11.1 million unauthorized
immigrants living in the United States in 2014, per the research
conducted by the Pew Research Center.
4. During the 2016 Presidential election, the candidates from the two major
political parties in the U.S. had different opinions of Russian President
Vladimir Putin.
Republican nominee Donald Trump said that Putin had “done a really great
job of outsmarting the U.S.” and that he is “highly respected within his own
country and beyond.”
Democratic nominee Hillary Clinton said that Putin “is at heart, a bully”
and that he is “a very cold-blooded, calculating former KGB agent” who is
determined to“enrich himself and his closest colleagues.”
In early 2014, separatist forces invaded the Ukrainian territory of Crimea, with
support from the Russian government. Russia subsequently annexed Crimea.
According to the Office of the United Nations High Commissioner for Human
Rights, as of November 2016, what is the figure CLOSEST to the total number
of people killed during the Ukrainian conflict?
• 0 (0)
• 1,000 (1)
198
• 2,000 (2)
• 3,000 (3)
• 4,000 (4)
• 5,000 (5)
• 6,000 (6)
• 7,000 (7)
• 8,000 (8)
• 9,000 (9)
• 10,000 (10)
The correct answer was 10,000. As of December 16, 2016, 9,733 peo-
ple have been killed during the conflict, including Ukrainian armed
forces, civilians, and members of the armed groups involved in the
conflict.
5. Several branches of the U.S. Government are investigating the possibility that
Russian hacks of Democratic National Committee material may have affected
the electoral process.
Furthermore, members of Trump’s campaign and his Administration are being
investigated for possible collusion with Russian operatives.
Who is the Special Counsel for the Department of Justice investigation into
possible collusion between President Trump’s campaign and Russia?
• Robert Mueller (0)
• Jared Kushner (1)
• Carter Page (2)
• Scott Pruitt (3)
• Jeff Sessions (4)
• James Comey (5)
• Rod Rosenstein (6)
• Jason Chaffetz (7)
199
• Mike Pence (8)
• Jay Sekulow (9)
• Sean Spicer (10)
The correct answer was Robert Mueller. Mueller was appointed as
the Special Counsel by Deputy Attorney General Rod Rosenstein on
May 17, 2017.
6. The United States helped to negotiate a major nuclear deal framework with
Iran in 2015.
The reaction to the deal was mixed in the United States House of Congress.
Most Democrats in Congress supported the deal, but most Republicans
opposed it.
Please indicate whether each of the following countries generally supported or
opposed the nuclear deal:
• Germany (Supported or Opposed)
• Iraq (Supported or Opposed)
• Israel (Supported or Opposed)
• Russia (Supported or Opposed)
• Saudi Arabia (Supported or Opposed)
• Turkey (Supported or Opposed)
• United Kingdom (Supported or Opposed)
The correct answers are that Germany, Iraq, Russia, Saudi Arabia
(with reservations), Turkey, and the United Kingdom all supported
the nuclear deal with Iran. Of the countries listed, only Israel op-
posed the nuclear deal with Iran.
7. National Security experts are concerned with recent North Korean tests of
long range ballistic missiles.
According to the CSIS Missile Defense Project, what is the largest maximum
known range of North Korean missiles that are currently operational?
200
• Up to 250 km (0)
• Up to 1,000 km (1)
• Up to 3,000 km (2)
• Up to 5,000 km (3)
• Up to 8,000 km (4)
• Up to 10,000 km (5)
• Up to 11,500 km (6)
• Up to 13,000 km (7)
• Up to 15,000 km (8)
• Up to 17,000 km (9)
• Up to 20,000 km (10)
The correct answer is up to 15,000 km. According to the CSIS, the
Taepodong-2 missile has the largest range of operational missiles,
with its range estimated to be between 4,000-15,000 km.
8. What was the U.S. trade deficit with China in 2016?
• There was no trade deficit with China (0)
• Between $100 billion and $200 billion (1)
• Between $200 billion and $300 billion (2)
• Between $300 billion and $400 billion (3)
• Between $400 billion and $500 billion (4)
• Between $500 billion and $600 billion (5)
• Between $600 billion and $700 billion (6)
• Between $700 billion and $800 billion (7)
• Between $800 billion and $900 billion (8)
• Between $900 billion and $1 trillion (9)
• Over $ 1 trillion dollars (10)
The correct answer was between $300 billion and $400 billion. Ac-
cording to the U.S. Census Bureau, the United States had a trade
deficit of $347.016 billion with China in 2016.
201
A.7 Political Attitudes Questions
There are no correct answers for these questions, I am just interested in your opinion.
Please indicate whether you agree or disagree with each statement.
1. The U.S. should ban entry to travelers from Syria, Iran, Libya, Somalia, Sudan
and Yemen.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
2. The U.S. should have withdrawn from the Paris Climate Agreement.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
3. The U.S. government should provide a path to citizenship for unauthorized
immigrations to the United States.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
202
• Disagree (-2)
• Strongly disagree (-3)
4. The U.S. should continue to impose trade sanctions against Russia in response
to the Ukraine conflict.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
5. The U.S. should seek to improve diplomatic relations with Russia.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
6. The U.S. should launch a pre-emptive military strike against North Korea.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
7. The U.S. should have signed the nuclear deal with Iran in 2015.
203
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
8. TheU.S.shouldimposegreatertraderestrictionsonChinatoprotectAmerican
companies and American workers.
• Strongly agree (3)
• Agree (2)
• Somewhat disagree (1)
• Neither agree nor disagree (0)
• Somewhat disagree (-1)
• Disagree (-2)
• Strongly disagree (-3)
A.8 Answers to Prediction Questions
These will be revealed on September 7, 2017. You will receive an email with a list
of correct answers on this date.
A.9 Demographics 2 (Randomly assigned to par-
ticipants before/after Demographics 3)
1. Which statement best describes your current employment status?
• Employed full time (1)
• Employed part time (2)
• Unemployed looking for work (3)
• Unemployed not looking for work (4)
204
• Retired (5)
• Student (6)
• Disabled (7)
• Other (8)
2. Generally speaking, do you think of yourself as a...
• Republican (1)
• Democrat (2)
• Independent (3)
• Another party, please specify: (4)
• No preference (5)
3. Answer If Generally speaking, do you think of yourself as a...Republican Is
Selected
Would you call yourself a...
• Strong Republican (1)
• Not very strong Republican (2)
4. Answer If Generally speaking, do you think of yourself as a...Democrat Is
Selected
Would you call yourself a...
• Strong democrat (1)
• Not very strong Democrat (2)
5. Answer If Generally speaking, do you think of yourself as a... Independent
Is Selected And Generally speaking, do you think of yourself as a... Another
party, please specify: Is Selected And Generally speaking, do you think of
yourself as a... No preference Is Selected
Do you think of yourself as closer to the...
• Republican Party (1)
• Democratic Party (2)
205
6. On a scale from 0 (never) to 100 (all the time), please indicate how much you
agree with Republican Party policy positions.
7. On a scale from 0 (never) to 100 (all the time), please indicate how much you
agree with Democratic Party policy positions.
8. In general, do you think of yourself as...
• Extremely liberal (1)
• Liberal (2)
• Slightly liberal (3)
• Moderate, middle of the road (4)
• Slightly conservative (5)
• Conservative (6)
• Extremely conservative (7)
A.10 Demographics3(Randomlyassignedtopar-
ticipants before/after treatments)
1. Some people seem to follow what’s going on in government and public affairs
most of the time, whether there’s an election going on or not. Others aren’t
that interested.
Would you say you follow what’s going on in government and public affairs
most of the time, some of the time, only now and then, or hardly at all?
• Hardly at All (0)
• Only Now and Then (1)
• Some of the Time (2)
• Most of the Time (3)
2. Some people seem to read, listen to, or watch the news most of the time,
whether there’s an election going on or not. Others aren’t that interested.
Would you say you read, listen to, or watch the news most of the time, some
of the time, only now and then, or hardly at all?
• Hardly at All (0)
206
• Only Now and Then (1)
• Some of the Time (2)
• Most of the Time (3)
3. Please think about your news consumption over the past week. In the last
week, did you read, listen to, or watch the news?
• No (0)
• Yes (1)
SQ2 One problem faced by democratic countries around the world is that there are
sometimes few incentives for people to spend time and effort going to the polls
to fulfill their democratic right to vote in free and fair elections. Accordingly,
political scientists are interested in voter turnout. To let us know that you
have read this much, please answer that you have never voted in a Presidential
election in the following question.
Please indicate the last time you voted in a Presidential election.
• 2012
• 2008
• 2004
• 2000
• 1996
• 1992
• 1988
• Earlier than 1988
• I have never voted in a Presidential election
4. In the 2016 Presidential election, who did you vote for?
• Hillary Clinton (Democratic Party) (1)
• Gary Johnson (Libertarian Party) (2)
• Jill Stein (Green Party) (3)
• Donald Trump (Republican Party) (4)
• Other candidate/write in (5)
207
• I did not vote in the Presidential election (6)
5. What is your religion?
• Baptist - any denomination (1)
• Protestant (e.g. Methodist, Lutheran, Presbyterian, Episcopal) (2)
• Catholic (3)
• Mormon (4)
• Jewish (5)
• Muslim (6)
• Hindu (7)
• Buddhist (8)
• Pentecostal (9)
• Eastern Orthodox (10)
• Other Christian (11)
• Other non-Christian (12)
• None (13)
6. How often do you attend religious services? [Reverse Coded)
• More than once a week (5)
• Once a week (4)
• Once or twice a month (3)
• A few times a year (2)
• Once a year or less (1)
• Never (0)
208
Appendix B
Appendix for Chapter 4
209
B.1 Descriptive Statistics
Table B.1: Descriptive Statistics.
Count Mean Std. Dev. Min. Max.
Information Search 298 .3590604 .4805319 0 1
Democrat 298 .4765101 .500288 0 1
Republican 298 .204698 .4041594 0 1
Political Ideology 298 3.342282 1.772486 1 7
Moderate 298 .4127517 .493157 0 1
Dem. Policy Agreement 298 53.40268 30.31905 0 100
Rep. Policy Agreement 298 30.51007 29.49579 0 100
Education 298 10.18792 1.362653 7 13
Income 298 2.795302 1.447759 1 7
Male 298 .6107383 .488403 0 1
Political Interest 298 2.42953 .9193398 0 4
News Consumption 298 2.399329 .956072 0 4
Religious Attendance 298 1 1.504203 0 5
Age 298 38.57383 12.81162 20 77
White 298 .7315436 .4439018 0 1
Black 298 .1073826 .3101198 0 1
Hispanic/Latino 298 .090604 .287528 0 1
210
Table B.2: Descriptive Statistics of Individual Political Knowledge Questions.
Count Mean Std. Dev. Min. Max.
Correct No. Deaths from Jihadist Terrorism 298 .4496644 .4982966 0 1
Correct No. Ratifications of Paris Agreement 298 .3020134 .4599033 0 1
Correct No. Unauthorized Immigrants 298 .4865772 .5006605 0 1
Correct No. Ukraine Deaths 298 .2852349 .4522858 0 1
Correct Special Counsel 298 .7449664 .4366133 0 1
Correct International Support for Iran Deal 298 .1174497 .3224969 0 1
Correct N. Korea Missile Range 298 .1510067 .3586577 0 1
Correct U.S.-China Trade Deficit 298 .466443 .4997118 0 1
211
Table B.3: Descriptive Statistics of Individual Prediction Questions.
Count Mean Std. Dev. Min. Max.
Correct Prediction of Laptop Ban on Domestic Flights 298 .7449664 .4366133 0 1
Correct Prediction of Arctic Sea Ice Extent (Area) vs. 2016 298 .3221477 .4680852 0 1
Correct Prediction of Border Wall Construction 298 .8557047 .35198 0 1
Correct Prediction of Lift Russia Sanctions 298 .8624161 .3450421 0 1
Correct Prediction of Impeachment of Trump 298 .8892617 .3143356 0 1
Correct Prediction of U.S. Sanctions on Iran for Nuclear Deal 298 .647651 .4785055 0 1
Correct Prediction of Lethal Confrontation with N. Korea 298 .8422819 .3650896 0 1
Correct Prediction of U.S. Sanctions on China for Trade 298 .7348993 .4421292 0 1
212
Table B.4: Descriptive Statistics of Individual Foreign Policy Attitudes.
Count Mean Std. Deviation Min. Max.
Political Knowledge 298 3.369128 2.236366 0 8
Information Search 298 .3590604 .4805319 0 1
Muslim Ban 298 -.647651 2.088935 -3 3
Paris Agreement 298 -1.362416 2.113655 -3 3
Path to Citizenship 298 .6073826 2.083334 -3 3
Russia Sanctions 298 1.177852 1.655085 -3 3
Russia Relations 298 .6744966 1.663053 -3 3
Strike N. Korea 298 -1.567114 1.607178 -3 3
Iran Deal 298 .2852349 1.893419 -3 3
China Trade 298 .2315436 1.549436 -3 3
213
B.2 Full Experimental Results on Political
Knowledge using OLS Regression
214
Table B.5: The Determinants of Information Search.
(1) (2) (3)
Bonus Treatment 0.220
∗∗
0.218
∗∗
0.227
∗∗
(0.078) (0.079) (0.079)
Random Bonus Treatment 0.077 0.068 0.049
(0.077) (0.077) (0.078)
Lottery Treatment 0.037 0.037 0.049
(0.075) (0.076) (0.076)
Democrat -0.007 -0.010
(0.082) (0.083)
Republican -0.150 -0.163
(0.106) (0.105)
Political Ideology 0.041 0.054
+
(0.032) (0.031)
Moderate -0.085 -0.031
(0.066) (0.065)
Dem. Policy Agreement 0.001 0.001
(0.001) (0.001)
Rep. Policy Agreement -0.000 -0.000
(0.002) (0.002)
Education -0.005
(0.022)
Income -0.043
∗
(0.020)
Male -0.075
(0.060)
Political interest 0.020
(0.051)
News Consumption 0.047
(0.051)
Religious Attendance -0.017
(0.019)
Age 0.006
∗
(0.002)
White 0.055
(0.077)
Black -0.143
(0.099)
Hispanic/Latino -0.063
(0.107)
Constant 0.274
∗∗∗
0.186 -0.043
(0.053) (0.146) (0.290)
Observations 298 298 298
R
2
0.031 0.042 0.132
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
215
Table B.6: The Determinants of Political Knowledge.
(1) (2) (3) (4)
Bonus Treatment 1.890
∗∗∗
1.280
∗∗∗
1.312
∗∗∗
1.224
∗∗∗
(0.342) (0.269) (0.274) (0.250)
Random Bonus Treatment 1.309
∗∗∗
1.094
∗∗∗
1.088
∗∗∗
1.090
∗∗∗
(0.334) (0.255) (0.258) (0.257)
Lottery Treatment 0.323 0.220 0.240 0.182
(0.318) (0.252) (0.250) (0.255)
Information Search 2.782
∗∗∗
2.749
∗∗∗
2.612
∗∗∗
(0.219) (0.223) (0.237)
Democrat -0.222 -0.313
(0.295) (0.302)
Republican -0.210 -0.296
(0.358) (0.365)
Political Ideology 0.139 0.202
+
(0.105) (0.107)
Moderate -0.405
+
-0.344
(0.220) (0.211)
Dem. Policy Agreement 0.001 0.004
(0.005) (0.005)
Rep. Policy Agreement -0.006 -0.004
(0.006) (0.006)
Education 0.049
(0.079)
Income -0.002
(0.074)
Male 0.056
(0.212)
Political interest 0.237
(0.169)
News Consumption 0.231
(0.174)
Religious Attendance -0.070
(0.077)
Age -0.003
(0.008)
White -0.529
+
(0.281)
Black -0.953
∗∗
(0.319)
Hispanic/Latino -0.524
(0.354)
Constant 2.110
∗∗∗
1.347
∗∗∗
1.309
∗∗
0.085
(0.216) (0.161) (0.428) (0.897)
Observations 298 298 298 298
R
2
0.117 0.466 0.474 0.527
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
216
Table B.7: Correct No. Deaths from Jihadist Terrorism.
(1) (2) (3) (4)
Bonus Treatment 0.215
∗∗
0.160
∗
0.168
∗
0.159
+
(0.080) (0.079) (0.080) (0.081)
Random Bonus Treatment 0.144
+
0.124 0.133
+
0.147
+
(0.082) (0.081) (0.080) (0.082)
Lottery Treatment -0.046 -0.055 -0.056 -0.061
(0.079) (0.076) (0.077) (0.077)
Information Search 0.249
∗∗∗
0.254
∗∗∗
0.217
∗∗∗
(0.060) (0.059) (0.062)
Democrat -0.081 -0.071
(0.078) (0.080)
Republican 0.061 0.070
(0.114) (0.118)
Political Ideology -0.023 -0.018
(0.030) (0.032)
Moderate -0.064 -0.025
(0.066) (0.066)
Dem. Policy Agreement 0.001 0.001
(0.001) (0.001)
Rep. Policy Agreement 0.002 0.002
(0.002) (0.002)
Education 0.001
(0.022)
Income -0.025
(0.020)
Male 0.075
(0.061)
Political interest 0.083
(0.055)
News Consumption -0.016
(0.057)
Religious Attendance -0.016
(0.020)
Age 0.002
(0.002)
White -0.039
(0.079)
Black -0.047
(0.111)
Hispanic/Latino -0.035
(0.104)
Constant 0.370
∗∗∗
0.302
∗∗∗
0.315
∗
0.100
(0.057) (0.057) (0.141) (0.280)
Observations 298 298 298 298
R
2
0.045 0.101 0.120 0.155
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
217
Table B.8: Correct No. Ratifications of Paris Agreement.
(1) (2) (3) (4)
Bonus Treatment 0.224
∗∗
0.120
+
0.117
+
0.115
+
(0.073) (0.062) (0.063) (0.062)
Random Bonus Treatment 0.200
∗∗
0.163
∗
0.166
∗∗
0.160
∗
(0.074) (0.063) (0.064) (0.064)
Lottery Treatment 0.011 -0.007 -0.009 -0.019
(0.066) (0.060) (0.060) (0.064)
Information Search 0.474
∗∗∗
0.479
∗∗∗
0.467
∗∗∗
(0.053) (0.054) (0.056)
Democrat 0.024 0.002
(0.072) (0.075)
Republican 0.039 0.017
(0.084) (0.087)
Political Ideology -0.026 -0.012
(0.023) (0.024)
Moderate 0.052 0.032
(0.053) (0.054)
Dem. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Rep. Policy Agreement -0.000 -0.001
(0.001) (0.001)
Education -0.020
(0.019)
Income -0.001
(0.016)
Male -0.077
(0.052)
Political interest -0.011
(0.042)
News Consumption 0.070
+
(0.041)
Religious Attendance 0.011
(0.019)
Age -0.003
(0.002)
White -0.072
(0.069)
Black -0.200
∗
(0.080)
Hispanic/Latino -0.038
(0.089)
Constant 0.192
∗∗∗
0.062
+
0.191
+
0.458
+
(0.046) (0.037) (0.112) (0.245)
Observations 298 298 298 298
R
2
0.051 0.289 0.295 0.324
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
218
Table B.9: Correct No. Unauthorized Immigrants.
(1) (2) (3) (4)
Bonus Treatment 0.361
∗∗∗
0.243
∗∗∗
0.243
∗∗∗
0.235
∗∗∗
(0.077) (0.065) (0.066) (0.065)
Random Bonus Treatment 0.239
∗∗
0.197
∗∗
0.194
∗∗
0.212
∗∗
(0.080) (0.064) (0.064) (0.065)
Lottery Treatment 0.131
+
0.111 0.115
+
0.109
(0.079) (0.069) (0.070) (0.071)
Information Search 0.539
∗∗∗
0.534
∗∗∗
0.513
∗∗∗
(0.050) (0.051) (0.054)
Democrat 0.010 0.010
(0.076) (0.075)
Republican -0.033 -0.046
(0.097) (0.099)
Political Ideology 0.036 0.045
(0.028) (0.029)
Moderate 0.004 0.014
(0.053) (0.054)
Dem. Policy Agreement -0.000 -0.000
(0.001) (0.001)
Rep. Policy Agreement -0.002 -0.002
(0.002) (0.002)
Education 0.005
(0.018)
Income -0.027
(0.017)
Male -0.065
(0.052)
Political interest 0.077
(0.049)
News Consumption 0.032
(0.045)
Religious Attendance -0.012
(0.018)
Age -0.004
∗
(0.002)
White -0.212
∗∗
(0.072)
Black -0.209
∗
(0.092)
Hispanic/Latino -0.204
∗
(0.101)
Constant 0.301
∗∗∗
0.154
∗∗∗
0.112 0.243
(0.054) (0.042) (0.122) (0.238)
Observations 298 298 298 298
R
2
0.071 0.330 0.338 0.399
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
219
Table B.10: Correct No. Ukraine Deaths.
(1) (2) (3) (4)
Bonus Treatment 0.198
∗∗
0.126
+
0.131
+
0.107
(0.073) (0.071) (0.072) (0.076)
Random Bonus Treatment 0.146
∗
0.121
+
0.117 0.110
(0.072) (0.071) (0.071) (0.074)
Lottery Treatment 0.024 0.012 0.015 0.001
(0.067) (0.063) (0.063) (0.069)
Information Search 0.325
∗∗∗
0.318
∗∗∗
0.310
∗∗∗
(0.056) (0.057) (0.060)
Democrat -0.014 -0.052
(0.074) (0.077)
Republican -0.077 -0.093
(0.094) (0.093)
Political Ideology 0.036 0.044
+
(0.027) (0.027)
Moderate -0.061 -0.079
(0.058) (0.058)
Dem. Policy Agreement 0.000 0.001
(0.001) (0.001)
Rep. Policy Agreement -0.001 -0.001
(0.002) (0.002)
Education -0.004
(0.020)
Income 0.020
(0.019)
Male -0.037
(0.055)
Political interest 0.022
(0.048)
News Consumption 0.046
(0.047)
Religious Attendance 0.012
(0.019)
Age -0.001
(0.002)
White -0.087
(0.067)
Black -0.185
∗
(0.088)
Hispanic/Latino 0.026
(0.092)
Constant 0.192
∗∗∗
0.103
∗
0.049 -0.000
(0.046) (0.048) (0.117) (0.240)
Observations 298 298 298 298
R
2
0.033 0.149 0.156 0.191
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
220
Table B.11: Correct Special Counsel.
(1) (2) (3) (4)
Bonus Treatment 0.226
∗∗
0.160
∗
0.167
∗∗
0.135
∗
(0.068) (0.064) (0.062) (0.058)
Random Bonus Treatment 0.113 0.089 0.083 0.064
(0.076) (0.069) (0.067) (0.062)
Lottery Treatment 0.059 0.048 0.054 0.057
(0.078) (0.074) (0.071) (0.062)
Information Search 0.304
∗∗∗
0.290
∗∗∗
0.246
∗∗∗
(0.041) (0.042) (0.041)
Democrat -0.045 -0.078
(0.061) (0.057)
Republican -0.086 -0.103
(0.104) (0.091)
Political Ideology 0.011 0.016
(0.031) (0.027)
Moderate -0.176
∗∗
-0.141
∗∗
(0.058) (0.050)
Dem. Policy Agreement -0.001 -0.000
(0.001) (0.001)
Rep. Policy Agreement -0.003
+
-0.002
(0.002) (0.002)
Education 0.059
∗∗∗
(0.016)
Income 0.026
+
(0.015)
Male 0.043
(0.046)
Political interest 0.031
(0.044)
News Consumption 0.059
(0.043)
Religious Attendance -0.030
+
(0.016)
Age 0.006
∗∗∗
(0.002)
White -0.093
(0.064)
Black -0.096
(0.089)
Hispanic/Latino -0.122
(0.092)
Constant 0.644
∗∗∗
0.561
∗∗∗
0.771
∗∗∗
-0.302
(0.056) (0.056) (0.130) (0.219)
Observations 298 298 298 298
R
2
0.037 0.146 0.214 0.381
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
221
Table B.12: Correct International Support for Iran Deal.
(1) (2) (3) (4)
Bonus Treatment 0.193
∗∗∗
0.162
∗∗∗
0.168
∗∗∗
0.149
∗∗
(0.051) (0.048) (0.048) (0.047)
Random Bonus Treatment 0.135
∗∗
0.124
∗∗
0.127
∗∗
0.117
∗
(0.047) (0.047) (0.047) (0.049)
Lottery Treatment 0.027 0.021 0.020 0.000
(0.033) (0.033) (0.034) (0.036)
Information Search 0.144
∗∗∗
0.145
∗∗∗
0.120
∗∗
(0.041) (0.042) (0.046)
Democrat -0.025 -0.027
(0.053) (0.054)
Republican -0.002 -0.009
(0.049) (0.057)
Political Ideology 0.004 0.013
(0.018) (0.020)
Moderate -0.026 -0.011
(0.040) (0.041)
Dem. Policy Agreement -0.001 -0.000
(0.001) (0.001)
Rep. Policy Agreement -0.000 -0.000
(0.001) (0.001)
Education 0.001
(0.015)
Income -0.016
(0.014)
Male 0.070
+
(0.041)
Political interest 0.009
(0.028)
News Consumption 0.036
(0.028)
Religious Attendance -0.002
(0.014)
Age 0.001
(0.002)
White 0.010
(0.058)
Black -0.097
∗
(0.047)
Hispanic/Latino 0.034
(0.071)
Constant 0.027 -0.012 0.034 -0.153
(0.019) (0.025) (0.108) (0.215)
Observations 298 298 298 298
R
2
0.060 0.105 0.114 0.157
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
222
Table B.13: Correct N. Korea Missile Range.
(1) (2) (3) (4)
Bonus Treatment 0.138
∗
0.087 0.084 0.071
(0.060) (0.059) (0.059) (0.059)
Random Bonus Treatment 0.066 0.048 0.043 0.049
(0.055) (0.052) (0.052) (0.053)
Lottery Treatment 0.012 0.004 0.000 -0.007
(0.050) (0.046) (0.046) (0.048)
Information Search 0.233
∗∗∗
0.226
∗∗∗
0.237
∗∗∗
(0.050) (0.050) (0.054)
Democrat -0.034 -0.029
(0.059) (0.064)
Republican -0.130
+
-0.114
(0.072) (0.075)
Political Ideology 0.035
+
0.042
+
(0.021) (0.022)
Moderate -0.027 -0.028
(0.047) (0.048)
Dem. Policy Agreement 0.001 0.001
(0.001) (0.001)
Rep. Policy Agreement 0.000 0.000
(0.001) (0.001)
Education 0.002
(0.016)
Income 0.009
(0.015)
Male 0.075
+
(0.044)
Political interest -0.020
(0.029)
News Consumption 0.038
(0.032)
Religious Attendance -0.017
(0.014)
Age -0.001
(0.002)
White -0.032
(0.052)
Black -0.010
(0.076)
Hispanic/Latino 0.024
(0.074)
Constant 0.096
∗∗
0.032 -0.074 -0.182
(0.035) (0.029) (0.097) (0.222)
Observations 298 298 298 298
R
2
0.023 0.118 0.133 0.159
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
223
Table B.14: Correct U.S.-China Deficit.
(1) (2) (3) (4)
Bonus Treatment 0.336
∗∗∗
0.223
∗∗
0.234
∗∗∗
0.253
∗∗∗
(0.077) (0.068) (0.069) (0.071)
Random Bonus Treatment 0.266
∗∗∗
0.227
∗∗
0.225
∗∗
0.231
∗∗
(0.079) (0.069) (0.069) (0.073)
Lottery Treatment 0.104 0.085 0.100 0.102
(0.078) (0.071) (0.070) (0.073)
Information Search 0.514
∗∗∗
0.504
∗∗∗
0.503
∗∗∗
(0.051) (0.052) (0.054)
Democrat -0.057 -0.068
(0.069) (0.074)
Republican 0.017 -0.019
(0.097) (0.099)
Political Ideology 0.065
∗
0.072
∗
(0.029) (0.029)
Moderate -0.107
+
-0.106
+
(0.058) (0.060)
Dem. Policy Agreement 0.002
+
0.003
∗
(0.001) (0.001)
Rep. Policy Agreement -0.002 -0.001
(0.002) (0.002)
Education 0.006
(0.021)
Income 0.012
(0.017)
Male -0.028
(0.056)
Political interest 0.046
(0.047)
News Consumption -0.034
(0.046)
Religious Attendance -0.017
(0.019)
Age -0.003
(0.002)
White -0.003
(0.071)
Black -0.109
(0.103)
Hispanic/Latino -0.209
∗
(0.088)
Constant 0.288
∗∗∗
0.147
∗∗
-0.088 -0.080
(0.053) (0.047) (0.118) (0.240)
Observations 298 298 298 298
R
2
0.070 0.307 0.329 0.355
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
224
B.3 Full Experimental Results on Political
Knowledge using Logistic Regression
225
Table B.15: The Determinants of Information Search.
(1) (2) (3)
Bonus Treatment 0.949
∗∗
0.957
∗∗
1.103
∗∗
(0.348) (0.352) (0.397)
Random Bonus Treatment 0.361 0.326 0.271
(0.359) (0.361) (0.397)
Lottery Treatment 0.178 0.181 0.262
(0.364) (0.367) (0.394)
Democrat -0.029 -0.036
(0.360) (0.406)
Republican -0.735 -0.883
(0.520) (0.587)
Political Ideology 0.196 0.275
+
(0.146) (0.160)
Moderate -0.413 -0.188
(0.313) (0.347)
Dem. Policy Agreement 0.003 0.004
(0.007) (0.007)
Rep. Policy Agreement -0.001 -0.002
(0.008) (0.008)
Education -0.023
(0.106)
Income -0.219
∗
(0.103)
Male -0.392
(0.290)
Political interest 0.131
(0.262)
News Consumption 0.204
(0.264)
Religious Attendance -0.091
(0.099)
Age 0.025
∗
(0.011)
White 0.268
(0.363)
Black -0.749
(0.535)
Hispanic/Latino -0.301
(0.533)
Constant -0.975
∗∗∗
-1.398
∗
-2.583
+
(0.263) (0.677) (1.419)
Observations 298 298 298
Pseudo R
2
0.023 0.033 0.107
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
226
Table B.16: Correct No. Deaths from Jihadist Terrorism.
(1) (2) (3) (4)
Bonus Treatment 0.874
∗∗
0.687
∗
0.741
∗
0.729
∗
(0.336) (0.346) (0.354) (0.370)
Random Bonus Treatment 0.587
+
0.537 0.591
+
0.678
+
(0.337) (0.351) (0.354) (0.376)
Lottery Treatment -0.201 -0.255 -0.266 -0.289
(0.348) (0.356) (0.362) (0.365)
Information Search 1.057
∗∗∗
1.104
∗∗∗
0.980
∗∗∗
(0.258) (0.261) (0.275)
Democrat -0.365 -0.333
(0.353) (0.373)
Republican 0.278 0.365
(0.503) (0.532)
Political Ideology -0.105 -0.095
(0.137) (0.153)
Moderate -0.295 -0.109
(0.294) (0.308)
Dem. Policy Agreement 0.005 0.006
(0.007) (0.007)
Rep. Policy Agreement 0.010 0.010
(0.008) (0.008)
Education 0.014
(0.104)
Income -0.116
(0.093)
Male 0.353
(0.282)
Political interest 0.403
(0.257)
News Consumption -0.080
(0.259)
Religious Attendance -0.087
(0.095)
Age 0.010
(0.011)
White -0.185
(0.365)
Black -0.220
(0.508)
Hispanic/Latino -0.178
(0.479)
Constant -0.533
∗
-0.844
∗∗
-0.833 -1.989
(0.243) (0.262) (0.662) (1.320)
Observations 298 298 298 298
Pseudo R
2
0.033 0.076 0.091 0.121
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
227
Table B.17: Correct No. Ratifications of Paris Agreement.
(1) (2) (3) (4)
Bonus Treatment 1.098
∗∗
0.788
+
0.756
+
0.826
+
(0.377) (0.411) (0.418) (0.427)
Random Bonus Treatment 0.999
∗∗
1.088
∗
1.116
∗∗
1.202
∗∗
(0.382) (0.427) (0.427) (0.448)
Lottery Treatment 0.069 -0.033 -0.096 -0.054
(0.415) (0.479) (0.494) (0.518)
Information Search 2.421
∗∗∗
2.491
∗∗∗
2.562
∗∗∗
(0.308) (0.310) (0.336)
Democrat 0.097 -0.086
(0.426) (0.449)
Republican 0.239 0.109
(0.576) (0.631)
Political Ideology -0.203 -0.095
(0.159) (0.163)
Moderate 0.397 0.277
(0.341) (0.358)
Dem. Policy Agreement -0.008 -0.006
(0.008) (0.008)
Rep. Policy Agreement -0.002 -0.005
(0.009) (0.010)
Education -0.132
(0.132)
Income -0.007
(0.115)
Male -0.639
+
(0.347)
Political interest -0.061
(0.291)
News Consumption 0.496
+
(0.298)
Religious Attendance 0.076
(0.125)
Age -0.021
(0.014)
White -0.581
(0.481)
Black -1.668
∗
(0.732)
Hispanic/Latino -0.273
(0.588)
Constant -1.438
∗∗∗
-2.457
∗∗∗
-1.551
∗
0.131
(0.298) (0.352) (0.717) (1.509)
Observations 298 298 298 298
Pseudo R
2
0.042 0.243 0.251 0.288
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
228
Table B.18: Correct No. Unauthorized Immigrants.
(1) (2) (3) (4)
Bonus Treatment 1.515
∗∗∗
1.435
∗∗∗
1.478
∗∗∗
1.625
∗∗∗
(0.351) (0.394) (0.401) (0.452)
Random Bonus Treatment 1.003
∗∗
1.155
∗∗
1.155
∗∗
1.398
∗∗
(0.346) (0.383) (0.385) (0.441)
Lottery Treatment 0.569 0.680 0.715
+
0.727
(0.347) (0.423) (0.425) (0.472)
Information Search 2.662
∗∗∗
2.671
∗∗∗
2.840
∗∗∗
(0.329) (0.330) (0.382)
Democrat 0.092 0.150
(0.460) (0.517)
Republican -0.199 -0.203
(0.559) (0.654)
Political Ideology 0.228 0.314
+
(0.168) (0.188)
Moderate 0.006 0.036
(0.327) (0.367)
Dem. Policy Agreement -0.002 -0.003
(0.008) (0.009)
Rep. Policy Agreement -0.012 -0.011
(0.010) (0.011)
Education 0.021
(0.123)
Income -0.181
+
(0.110)
Male -0.427
(0.336)
Political interest 0.464
(0.313)
News Consumption 0.265
(0.287)
Religious Attendance -0.070
(0.118)
Age -0.028
∗
(0.014)
White -1.262
∗∗
(0.462)
Black -1.208
∗
(0.596)
Hispanic/Latino -1.245
+
(0.644)
Constant -0.841
∗∗∗
-1.761
∗∗∗
-2.084
∗∗
-1.424
(0.256) (0.303) (0.751) (1.540)
Observations 298 298 298 298
Pseudo R
2
0.053 0.265 0.273 0.341
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
229
Table B.19: Correct No. Ukraine Deaths.
(1) (2) (3) (4)
Bonus Treatment 0.990
∗∗
0.722
+
0.766
+
0.706
(0.379) (0.413) (0.416) (0.456)
Random Bonus Treatment 0.766
∗
0.718
+
0.712
+
0.733
(0.386) (0.426) (0.430) (0.473)
Lottery Treatment 0.151 0.096 0.137 0.147
(0.411) (0.431) (0.430) (0.467)
Information Search 1.599
∗∗∗
1.565
∗∗∗
1.620
∗∗∗
(0.280) (0.281) (0.318)
Democrat -0.096 -0.316
(0.397) (0.417)
Republican -0.519 -0.571
(0.596) (0.616)
Political Ideology 0.218 0.264
+
(0.155) (0.159)
Moderate -0.368 -0.465
(0.339) (0.346)
Dem. Policy Agreement 0.001 0.004
(0.008) (0.008)
Rep. Policy Agreement -0.005 -0.006
(0.009) (0.010)
Education -0.023
(0.118)
Income 0.115
(0.112)
Male -0.257
(0.326)
Political interest 0.121
(0.334)
News Consumption 0.282
(0.316)
Religious Attendance 0.088
(0.108)
Age -0.006
(0.012)
White -0.559
(0.386)
Black -1.342
+
(0.700)
Hispanic/Latino 0.132
(0.492)
Constant -1.438
∗∗∗
-2.033
∗∗∗
-2.399
∗∗∗
-2.786
∗
(0.298) (0.363) (0.713) (1.375)
Observations 298 298 298 298
Pseudo R
2
0.028 0.124 0.131 0.169
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
230
Table B.20: Correct Special Counsel.
(1) (2) (3) (4)
Bonus Treatment 1.310
∗∗
1.057
∗
1.140
∗
1.253
∗
(0.419) (0.436) (0.471) (0.533)
Random Bonus Treatment 0.543 0.472 0.496 0.772
(0.366) (0.377) (0.412) (0.531)
Lottery Treatment 0.268 0.237 0.254 0.337
(0.353) (0.384) (0.399) (0.461)
Information Search 2.401
∗∗∗
2.497
∗∗∗
3.025
∗∗∗
(0.486) (0.498) (0.615)
Democrat -0.418 -1.325
∗
(0.467) (0.598)
Republican -0.345 -0.889
(0.595) (0.706)
Political Ideology 0.016 0.234
(0.197) (0.232)
Moderate -1.092
∗∗
-1.452
∗∗
(0.365) (0.447)
Dem. Policy Agreement -0.006 0.004
(0.009) (0.012)
Rep. Policy Agreement -0.018
+
-0.019
(0.010) (0.013)
Education 0.449
∗∗
(0.153)
Income 0.365
∗
(0.149)
Male 0.182
(0.383)
Political interest 0.412
(0.329)
News Consumption 0.521
+
(0.300)
Religious Attendance -0.247
∗
(0.114)
Age 0.058
∗∗
(0.018)
White -0.977
(0.679)
Black -0.731
(0.778)
Hispanic/Latino -1.349
+
(0.777)
Constant 0.592
∗
0.147 1.731
+
-7.504
∗∗∗
(0.245) (0.263) (0.936) (2.179)
Observations 298 298 298 298
Pseudo R
2
0.035 0.155 0.223 0.437
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
231
Table B.21: Correct International Support for Iran Deal.
(1) (2) (3) (4)
Bonus Treatment 2.308
∗∗
2.073
∗∗
2.166
∗∗
2.106
∗∗
(0.769) (0.781) (0.759) (0.788)
Random Bonus Treatment 1.927
∗
1.874
∗
1.918
∗
1.891
∗
(0.785) (0.815) (0.804) (0.880)
Lottery Treatment 0.707 0.661 0.683 0.510
(0.884) (0.902) (0.897) (0.933)
Information Search 1.404
∗∗∗
1.404
∗∗∗
1.213
∗∗
(0.393) (0.404) (0.465)
Democrat -0.389 -0.555
(0.529) (0.568)
Republican -0.064 -0.119
(0.520) (0.670)
Political Ideology 0.007 0.022
(0.209) (0.268)
Moderate -0.245 -0.107
(0.410) (0.467)
Dem. Policy Agreement -0.006 -0.003
(0.011) (0.013)
Rep. Policy Agreement -0.001 -0.001
(0.010) (0.012)
Education -0.001
(0.177)
Income -0.170
(0.203)
Male 0.750
(0.583)
Political interest 0.298
(0.395)
News Consumption 0.257
(0.390)
Religious Attendance 0.012
(0.144)
Age 0.016
(0.017)
White 0.009
(0.676)
Black 0.000
(.)
Hispanic/Latino 0.486
(0.703)
Constant -3.570
∗∗∗
-4.165
∗∗∗
-3.625
∗
-5.722
∗
(0.718) (0.844) (1.451) (2.702)
Observations 298 298 298 266
Pseudo R
2
0.089 0.151 0.163 0.218
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
232
Table B.22: Correct N. Korea Missile Range.
(1) (2) (3) (4)
Bonus Treatment 1.057
∗
0.731 0.751 0.591
(0.481) (0.512) (0.508) (0.513)
Random Bonus Treatment 0.602 0.494 0.486 0.502
(0.508) (0.523) (0.516) (0.522)
Lottery Treatment 0.134 0.063 0.049 0.037
(0.547) (0.544) (0.557) (0.573)
Information Search 1.812
∗∗∗
1.788
∗∗∗
1.973
∗∗∗
(0.379) (0.379) (0.443)
Democrat -0.329 -0.240
(0.455) (0.545)
Republican -1.303
+
-1.087
(0.696) (0.706)
Political Ideology 0.320
+
0.378
+
(0.178) (0.201)
Moderate -0.275 -0.165
(0.410) (0.420)
Dem. Policy Agreement 0.007 0.010
(0.008) (0.009)
Rep. Policy Agreement 0.003 0.003
(0.010) (0.011)
Education 0.000
(0.143)
Income 0.107
(0.135)
Male 0.771
+
(0.468)
Political interest -0.220
(0.329)
News Consumption 0.356
(0.334)
Religious Attendance -0.157
(0.157)
Age -0.004
(0.018)
White -0.376
(0.454)
Black -0.020
(0.793)
Hispanic/Latino 0.272
(0.609)
Constant -2.244
∗∗∗
-3.032
∗∗∗
-4.086
∗∗∗
-5.222
∗∗
(0.398) (0.421) (0.908) (2.013)
Observations 298 298 298 298
Pseudo R
2
0.027 0.135 0.156 0.192
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
233
Table B.23: Correct U.S.-China Deficit.
(1) (2) (3) (4)
Bonus Treatment 1.411
∗∗∗
1.270
∗∗
1.436
∗∗∗
1.592
∗∗∗
(0.350) (0.398) (0.426) (0.452)
Random Bonus Treatment 1.124
∗∗
1.277
∗∗
1.380
∗∗∗
1.466
∗∗
(0.349) (0.400) (0.417) (0.460)
Lottery Treatment 0.467 0.516 0.675 0.663
(0.352) (0.436) (0.442) (0.468)
Information Search 2.450
∗∗∗
2.516
∗∗∗
2.660
∗∗∗
(0.303) (0.307) (0.337)
Democrat -0.356 -0.398
(0.421) (0.453)
Republican 0.127 -0.084
(0.554) (0.606)
Political Ideology 0.419
∗
0.479
∗∗
(0.173) (0.179)
Moderate -0.685
∗
-0.731
∗
(0.340) (0.363)
Dem. Policy Agreement 0.016
+
0.017
+
(0.008) (0.009)
Rep. Policy Agreement -0.010 -0.008
(0.009) (0.010)
Education 0.057
(0.125)
Income 0.074
(0.106)
Male -0.134
(0.336)
Political interest 0.276
(0.287)
News Consumption -0.176
(0.282)
Religious Attendance -0.109
(0.116)
Age -0.019
(0.014)
White 0.012
(0.432)
Black -0.662
(0.631)
Hispanic/Latino -1.282
∗
(0.554)
Constant -0.907
∗∗∗
-1.769
∗∗∗
-3.425
∗∗∗
-3.689
∗
(0.259) (0.329) (0.825) (1.600)
Observations 298 298 298 298
Pseudo R
2
0.052 0.242 0.268 0.297
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
234
B.4 Full Experimental Results on Political Pre-
dictions using OLS Regression
235
Table B.24: Political Predictions.
(1) (2) (3) (4)
Bonus Treatment -0.082 -0.082 -0.032 -0.009
(0.207) (0.207) (0.205) (0.223)
Random Bonus Treatment 0.175 0.175 0.194 0.234
(0.203) (0.203) (0.204) (0.213)
Lottery Treatment -0.448
∗
-0.448
∗
-0.432
+
-0.352
(0.220) (0.220) (0.222) (0.232)
Information Search 0.298
+
0.298
+
0.299
+
0.272
+
(0.153) (0.153) (0.154) (0.155)
Democrat 0.087 0.095
(0.204) (0.203)
Republican 0.244 0.331
(0.334) (0.302)
Political Ideology -0.036 -0.042
(0.098) (0.091)
Moderate -0.206 -0.143
(0.183) (0.174)
Dem. Policy Agreement -0.007
+
-0.007
+
(0.004) (0.004)
Rep. Policy Agreement -0.006 -0.002
(0.004) (0.004)
Education 0.094
(0.058)
Income 0.079
(0.053)
Male 0.040
(0.157)
Political interest 0.062
(0.147)
News Consumption -0.000
(0.152)
Religious Attendance -0.141
∗∗
(0.053)
Age 0.010
(0.006)
White -0.441
∗
(0.186)
Black -0.092
(0.262)
Hispanic/Latino -0.417
(0.299)
Constant 5.233
∗∗∗
5.233
∗∗∗
5.870
∗∗∗
4.486
∗∗∗
(0.155) (0.155) (0.414) (0.781)
Observations 298 298 298 298
R
2
0.045 0.045 0.070 0.151
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
236
Table B.25: Correct Prediction of Laptop Ban on Domestic Flights.
(1) (2) (3) (4)
Bonus Treatment 0.027 0.004 0.027 0.039
(0.071) (0.074) (0.073) (0.079)
Random Bonus Treatment 0.031 0.023 0.025 0.007
(0.071) (0.073) (0.073) (0.078)
Lottery Treatment -0.037 -0.041 -0.040 -0.045
(0.074) (0.075) (0.075) (0.079)
Information Search 0.104
+
0.102
+
0.108
∗
(0.053) (0.052) (0.054)
Democrat 0.115 0.107
(0.071) (0.075)
Republican 0.005 -0.010
(0.098) (0.101)
Political Ideology 0.015 0.019
(0.029) (0.031)
Moderate -0.089
+
-0.102
+
(0.054) (0.056)
Dem. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Rep. Policy Agreement 0.002 0.002
(0.002) (0.002)
Education 0.010
(0.020)
Income 0.014
(0.018)
Male -0.035
(0.054)
Political interest -0.042
(0.050)
News Consumption 0.015
(0.051)
Religious Attendance -0.008
(0.019)
Age -0.001
(0.002)
White -0.049
(0.073)
Black -0.179
+
(0.106)
Hispanic/Latino -0.120
(0.107)
Constant 0.740
∗∗∗
0.711
∗∗∗
0.662
∗∗∗
0.714
∗∗
(0.052) (0.054) (0.135) (0.247)
Observations 298 298 298 298
R
2
0.004 0.016 0.052 0.077
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
237
Table B.26: Correct Prediction of Arctic Sea Ice Extent (Area) vs. 2016.
(1) (2) (3) (4)
Bonus Treatment -0.112 -0.106 -0.083 -0.052
(0.078) (0.079) (0.081) (0.084)
Random Bonus Treatment -0.100 -0.098 -0.087 -0.030
(0.079) (0.080) (0.080) (0.082)
Lottery Treatment -0.141
+
-0.140
+
-0.127 -0.082
(0.078) (0.078) (0.078) (0.080)
Information Search -0.028 -0.019 0.002
(0.056) (0.056) (0.060)
Democrat 0.126
+
0.141
∗
(0.071) (0.069)
Republican 0.231
∗
0.280
∗
(0.112) (0.114)
Political Ideology 0.009 -0.010
(0.032) (0.032)
Moderate 0.031 0.033
(0.065) (0.067)
Dem. Policy Agreement -0.001 -0.002
(0.001) (0.001)
Rep. Policy Agreement -0.001 -0.001
(0.002) (0.002)
Education -0.025
(0.023)
Income 0.017
(0.019)
Male -0.064
(0.059)
Political interest 0.108
∗
(0.053)
News Consumption -0.119
∗
(0.053)
Religious Attendance -0.009
(0.018)
Age 0.000
(0.002)
White -0.102
(0.074)
Black 0.181
(0.119)
Hispanic/Latino -0.032
(0.103)
Constant 0.411
∗∗∗
0.419
∗∗∗
0.340
∗
0.700
∗
(0.058) (0.060) (0.146) (0.290)
Observations 298 298 298 298
R
2
0.013 0.014 0.049 0.101
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
238
Table B.27: Correct Prediction of Border Wall Construction.
(1) (2) (3) (4)
Bonus Treatment -0.046 -0.055 -0.075 -0.087
(0.056) (0.056) (0.056) (0.061)
Random Bonus Treatment -0.012 -0.015 -0.021 -0.036
(0.053) (0.054) (0.054) (0.051)
Lottery Treatment -0.080 -0.081 -0.085 -0.092
(0.059) (0.059) (0.058) (0.059)
Information Search 0.040 0.038 0.044
(0.041) (0.040) (0.039)
Democrat -0.057 -0.056
(0.051) (0.051)
Republican -0.061 -0.064
(0.099) (0.092)
Political Ideology -0.029 -0.020
(0.027) (0.025)
Moderate 0.037 0.040
(0.058) (0.055)
Dem. Policy Agreement 0.000 0.001
(0.001) (0.001)
Rep. Policy Agreement -0.001 0.000
(0.001) (0.001)
Education 0.035
∗
(0.018)
Income 0.014
(0.013)
Male 0.059
(0.043)
Political interest -0.070
(0.051)
News Consumption 0.075
(0.049)
Religious Attendance -0.031
∗
(0.016)
Age -0.000
(0.002)
White -0.036
(0.044)
Black -0.040
(0.076)
Hispanic/Latino -0.086
(0.077)
Constant 0.890
∗∗∗
0.880
∗∗∗
1.014
∗∗∗
0.578
∗∗
(0.037) (0.038) (0.089) (0.221)
Observations 298 298 298 298
R
2
0.008 0.011 0.065 0.141
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
239
Table B.28: Correct Prediction of Lifting of Russia Sanctions.
(1) (2) (3) (4)
Bonus Treatment 0.006 -0.012 0.001 0.022
(0.051) (0.052) (0.053) (0.057)
Random Bonus Treatment -0.012 -0.018 -0.018 0.003
(0.053) (0.053) (0.050) (0.053)
Lottery Treatment -0.107
+
-0.110
+
-0.101
+
-0.074
(0.061) (0.061) (0.060) (0.063)
Information Search 0.079
∗
0.071
+
0.053
(0.039) (0.039) (0.040)
Democrat -0.044 -0.052
(0.058) (0.059)
Republican -0.018 -0.015
(0.088) (0.086)
Political Ideology 0.028 0.025
(0.025) (0.026)
Moderate -0.115
∗
-0.100
∗
(0.047) (0.047)
Dem. Policy Agreement -0.002 -0.002
(0.001) (0.001)
Rep. Policy Agreement -0.004
∗∗
-0.003
∗
(0.001) (0.001)
Education 0.006
(0.014)
Income 0.015
(0.015)
Male -0.058
(0.040)
Political interest 0.077
+
(0.044)
News Consumption -0.050
(0.043)
Religious Attendance -0.023
(0.017)
Age 0.002
(0.002)
White -0.075
(0.052)
Black -0.037
(0.073)
Hispanic/Latino -0.126
(0.082)
Constant 0.890
∗∗∗
0.869
∗∗∗
1.045
∗∗∗
0.930
∗∗∗
(0.037) (0.040) (0.113) (0.205)
Observations 298 298 298 298
R
2
0.018 0.029 0.093 0.139
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
240
Table B.29: Correct Prediction of Impeachment of Trump.
(1) (2) (3) (4)
Bonus Treatment 0.046 0.035 0.039 0.027
(0.052) (0.053) (0.053) (0.055)
Random Bonus Treatment 0.069 0.065 0.068 0.072
(0.050) (0.050) (0.050) (0.050)
Lottery Treatment -0.012 -0.014 -0.017 -0.006
(0.058) (0.058) (0.059) (0.061)
Information Search 0.051 0.055 0.025
(0.036) (0.039) (0.037)
Democrat 0.056 0.044
(0.053) (0.051)
Republican 0.027 0.043
(0.088) (0.077)
Political Ideology -0.023 -0.023
(0.031) (0.027)
Moderate 0.009 0.035
(0.049) (0.046)
Dem. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Rep. Policy Agreement 0.002 0.002
∗
(0.001) (0.001)
Education 0.036
∗
(0.015)
Income 0.001
(0.014)
Male -0.009
(0.042)
Political interest 0.050
(0.037)
News Consumption -0.004
(0.036)
Religious Attendance -0.032
∗
(0.015)
Age 0.004
∗
(0.002)
White -0.076
+
(0.040)
Black -0.062
(0.067)
Hispanic/Latino 0.042
(0.057)
Constant 0.863
∗∗∗
0.849
∗∗∗
0.879
∗∗∗
0.320
(0.041) (0.042) (0.128) (0.223)
Observations 298 298 298 298
R
2
0.011 0.017 0.031 0.134
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
241
Table B.30: Correct Prediction of U.S. Sanctions on Iran for Nuclear Deal.
(1) (2) (3) (4)
Bonus Treatment -0.060 -0.078 -0.087 -0.119
(0.079) (0.081) (0.082) (0.086)
Random Bonus Treatment 0.005 -0.002 -0.004 -0.019
(0.079) (0.079) (0.079) (0.082)
Lottery Treatment 0.018 0.015 0.014 -0.006
(0.078) (0.079) (0.079) (0.083)
Information Search 0.081 0.076 0.092
(0.058) (0.058) (0.061)
Democrat -0.142
+
-0.149
∗
(0.075) (0.075)
Republican -0.079 -0.057
(0.114) (0.114)
Political Ideology 0.005 -0.005
(0.033) (0.034)
Moderate -0.049 -0.049
(0.066) (0.066)
Dem. Policy Agreement 0.001 0.002
(0.001) (0.001)
Rep. Policy Agreement -0.001 -0.001
(0.002) (0.002)
Education 0.012
(0.022)
Income 0.027
(0.020)
Male 0.131
∗
(0.063)
Political interest 0.015
(0.054)
News Consumption -0.044
(0.054)
Religious Attendance 0.009
(0.020)
Age 0.003
(0.002)
White -0.027
(0.080)
Black -0.044
(0.110)
Hispanic/Latino 0.083
(0.110)
Constant 0.658
∗∗∗
0.635
∗∗∗
0.669
∗∗∗
0.379
(0.056) (0.058) (0.147) (0.301)
Observations 298 298 298 298
R
2
0.004 0.010 0.026 0.063
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
242
Table B.31: Correct Prediction of Lethal Confrontation with N. Korea.
(1) (2) (3) (4)
Bonus Treatment 0.075 0.081 0.083 0.077
(0.059) (0.059) (0.060) (0.065)
Random Bonus Treatment 0.097
+
0.099
+
0.098
+
0.073
(0.058) (0.058) (0.059) (0.063)
Lottery Treatment -0.038 -0.037 -0.038 -0.036
(0.068) (0.068) (0.068) (0.070)
Information Search -0.030 -0.034 -0.033
(0.044) (0.045) (0.046)
Democrat -0.008 -0.017
(0.061) (0.058)
Republican -0.053 -0.044
(0.103) (0.102)
Political Ideology -0.002 -0.002
(0.030) (0.029)
Moderate -0.041 -0.047
(0.058) (0.056)
Dem. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Rep. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Education 0.015
(0.017)
Income 0.032
∗
(0.014)
Male 0.038
(0.046)
Political interest -0.087
(0.053)
News Consumption 0.065
(0.054)
Religious Attendance -0.007
(0.017)
Age 0.004
∗
(0.002)
White -0.029
(0.055)
Black -0.032
(0.095)
Hispanic/Latino -0.020
(0.084)
Constant 0.808
∗∗∗
0.816
∗∗∗
0.977
∗∗∗
0.644
∗∗
(0.046) (0.048) (0.118) (0.228)
Observations 298 298 298 298
R
2
0.023 0.024 0.042 0.099
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
243
Table B.32: Correct Prediction of U.S. Sanctions on China for Trade.
(1) (2) (3) (4)
Bonus Treatment -0.011 -0.029 -0.023 -0.034
(0.075) (0.076) (0.077) (0.079)
Random Bonus Treatment 0.126
+
0.119
+
0.129
+
0.145
∗
(0.069) (0.069) (0.068) (0.073)
Lottery Treatment -0.023 -0.026 -0.022 -0.018
(0.076) (0.076) (0.076) (0.077)
Information Search 0.082 0.086 0.073
(0.052) (0.053) (0.053)
Democrat -0.100 -0.073
(0.071) (0.074)
Republican 0.113 0.142
(0.111) (0.104)
Political Ideology -0.033 -0.032
(0.032) (0.033)
Moderate -0.039 -0.003
(0.061) (0.061)
Dem. Policy Agreement -0.001 -0.001
(0.001) (0.001)
Rep. Policy Agreement -0.002 -0.001
(0.002) (0.002)
Education 0.016
(0.020)
Income -0.014
(0.021)
Male 0.109
+
(0.060)
Political interest 0.024
(0.053)
News Consumption 0.017
(0.051)
Religious Attendance -0.030
(0.019)
Age 0.002
(0.002)
White -0.074
(0.068)
Black 0.078
(0.100)
Hispanic/Latino -0.074
(0.096)
Constant 0.712
∗∗∗
0.690
∗∗∗
0.952
∗∗∗
0.601
∗
(0.053) (0.056) (0.143) (0.257)
Observations 298 298 298 298
R
2
0.018 0.026 0.054 0.106
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
244
B.5 Full Experimental Results on Political Pre-
dictions using Logistic Regression
245
Table B.33: Correct Prediction of Laptop Ban on Domestic Flights.
(1) (2) (3) (4)
Bonus Treatment 0.143 0.022 0.179 0.249
(0.380) (0.400) (0.403) (0.439)
Random Bonus Treatment 0.165 0.125 0.164 0.073
(0.385) (0.394) (0.411) (0.438)
Lottery Treatment -0.184 -0.207 -0.218 -0.248
(0.369) (0.379) (0.389) (0.410)
Information Search 0.578
+
0.614
∗
0.664
∗
(0.310) (0.307) (0.320)
Democrat 0.594
+
0.574
(0.357) (0.383)
Republican 0.036 -0.046
(0.555) (0.565)
Political Ideology 0.131 0.154
(0.170) (0.182)
Moderate -0.624
+
-0.693
∗
(0.342) (0.350)
Dem. Policy Agreement -0.008 -0.008
(0.007) (0.007)
Rep. Policy Agreement 0.009 0.010
(0.008) (0.009)
Education 0.052
(0.107)
Income 0.080
(0.104)
Male -0.223
(0.306)
Political interest -0.219
(0.262)
News Consumption 0.080
(0.272)
Religious Attendance -0.047
(0.102)
Age -0.006
(0.012)
White -0.265
(0.392)
Black -0.900
+
(0.489)
Hispanic/Latino -0.655
(0.520)
Constant 1.045
∗∗∗
0.902
∗∗∗
0.608 0.903
(0.267) (0.274) (0.707) (1.339)
Observations 298 298 298 298
Pseudo R
2
0.003 0.015 0.051 0.073
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
246
Table B.34: Correct Prediction of Arctic Sea Ice Extent (Area) vs. 2016.
(1) (2) (3) (4)
Bonus Treatment -0.493 -0.465 -0.374 -0.258
(0.345) (0.347) (0.365) (0.397)
Random Bonus Treatment -0.436 -0.427 -0.392 -0.127
(0.347) (0.347) (0.356) (0.385)
Lottery Treatment -0.633
+
-0.629
+
-0.597 -0.382
(0.354) (0.354) (0.364) (0.388)
Information Search -0.132 -0.096 0.021
(0.264) (0.274) (0.302)
Democrat 0.653
+
0.748
∗
(0.379) (0.375)
Republican 1.060
∗
1.342
∗
(0.522) (0.550)
Political Ideology 0.044 -0.055
(0.147) (0.153)
Moderate 0.152 0.158
(0.296) (0.317)
Dem. Policy Agreement -0.006 -0.010
(0.007) (0.007)
Rep. Policy Agreement -0.004 -0.003
(0.008) (0.009)
Education -0.119
(0.110)
Income 0.086
(0.095)
Male -0.317
(0.288)
Political interest 0.528
∗
(0.267)
News Consumption -0.588
∗
(0.264)
Religious Attendance -0.046
(0.093)
Age 0.001
(0.011)
White -0.501
(0.375)
Black 0.837
(0.547)
Hispanic/Latino -0.141
(0.523)
Constant -0.360 -0.324 -0.708 1.087
(0.238) (0.249) (0.680) (1.404)
Observations 298 298 298 298
Pseudo R
2
0.010 0.011 0.039 0.082
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
247
Table B.35: Correct Prediction of Border Wall Construction.
(1) (2) (3) (4)
Bonus Treatment -0.405 -0.477 -0.675 -0.887
(0.490) (0.497) (0.526) (0.624)
Random Bonus Treatment -0.118 -0.143 -0.209 -0.376
(0.517) (0.524) (0.544) (0.547)
Lottery Treatment -0.640 -0.653 -0.733 -0.926
+
(0.479) (0.483) (0.499) (0.539)
Information Search 0.333 0.352 0.343
(0.352) (0.369) (0.375)
Democrat -0.599 -0.634
(0.520) (0.533)
Republican -0.376 -0.527
(0.675) (0.685)
Political Ideology -0.236 -0.116
(0.199) (0.176)
Moderate 0.127 0.156
(0.402) (0.407)
Dem. Policy Agreement 0.003 0.011
(0.009) (0.010)
Rep. Policy Agreement -0.007 -0.000
(0.012) (0.012)
Education 0.333
∗
(0.162)
Income 0.134
(0.145)
Male 0.537
(0.376)
Political interest -0.485
(0.411)
News Consumption 0.561
(0.394)
Religious Attendance -0.215
∗
(0.109)
Age -0.005
(0.016)
White -0.378
(0.484)
Black -0.331
(0.792)
Hispanic/Latino -0.742
(0.548)
Constant 2.095
∗∗∗
2.012
∗∗∗
3.392
∗∗∗
-0.688
(0.375) (0.375) (0.906) (1.995)
Observations 298 298 298 298
Pseudo R
2
0.009 0.013 0.074 0.165
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
248
Table B.36: Correct Prediction of Lift Russia Sanctions.
(1) (2) (3) (4)
Bonus Treatment 0.060 -0.092 -0.144 0.162
(0.530) (0.537) (0.584) (0.663)
Random Bonus Treatment -0.118 -0.171 -0.227 0.073
(0.517) (0.519) (0.522) (0.541)
Lottery Treatment -0.807
+
-0.842
+
-0.947
+
-0.678
(0.470) (0.473) (0.524) (0.583)
Information Search 0.757
+
0.838
∗
0.760
+
(0.406) (0.426) (0.458)
Democrat -0.515 -0.665
(0.537) (0.646)
Republican -0.173 0.029
(0.623) (0.683)
Political Ideology 0.192 0.181
(0.215) (0.242)
Moderate -0.936
∗
-0.895
∗
(0.393) (0.413)
Dem. Policy Agreement -0.019
+
-0.019
+
(0.011) (0.011)
Rep. Policy Agreement -0.036
∗∗
-0.034
∗∗
(0.011) (0.013)
Education -0.034
(0.138)
Income 0.210
(0.171)
Male -0.613
(0.397)
Political interest 0.621
+
(0.342)
News Consumption -0.327
(0.350)
Religious Attendance -0.166
(0.136)
Age 0.020
(0.018)
White -0.776
(0.539)
Black -0.379
(0.746)
Hispanic/Latino -0.944
(0.607)
Constant 2.095
∗∗∗
1.929
∗∗∗
4.349
∗∗
3.780
+
(0.375) (0.385) (1.332) (2.205)
Observations 298 298 298 298
Pseudo R
2
0.020 0.036 0.122 0.178
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
249
Table B.37: Correct Prediction of Impeachment of Trump.
(1) (2) (3) (4)
Bonus Treatment 0.462 0.347 0.412 0.152
(0.523) (0.536) (0.528) (0.569)
Random Bonus Treatment 0.784 0.748 0.804 0.968
(0.576) (0.578) (0.569) (0.673)
Lottery Treatment -0.095 -0.115 -0.133 0.026
(0.473) (0.475) (0.483) (0.573)
Information Search 0.572 0.635 0.279
(0.438) (0.461) (0.510)
Democrat 0.593 0.547
(0.536) (0.542)
Republican 0.275 0.189
(0.930) (0.883)
Political Ideology -0.213 -0.136
(0.282) (0.223)
Moderate 0.098 0.144
(0.524) (0.457)
Dem. Policy Agreement -0.009 -0.008
(0.010) (0.011)
Rep. Policy Agreement 0.016 0.028
∗
(0.011) (0.012)
Education 0.367
∗
(0.164)
Income 0.025
(0.176)
Male -0.053
(0.481)
Political interest 0.618
+
(0.349)
News Consumption 0.052
(0.323)
Religious Attendance -0.364
∗
(0.143)
Age 0.047
+
(0.025)
White -1.407
∗
(0.616)
Black -0.950
(0.668)
Hispanic/Latino 0.754
(0.957)
Constant 1.841
∗∗∗
1.707
∗∗∗
2.017
+
-3.697
+
(0.341) (0.352) (1.192) (2.183)
Observations 298 298 298 298
Pseudo R
2
0.016 0.025 0.046 0.210
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
250
Table B.38: Correct Prediction of U.S. Sanctions on Iran for Nuclear Deal.
(1) (2) (3) (4)
Bonus Treatment -0.258 -0.339 -0.390 -0.564
(0.339) (0.351) (0.360) (0.391)
Random Bonus Treatment 0.021 -0.007 -0.021 -0.093
(0.349) (0.351) (0.354) (0.379)
Lottery Treatment 0.082 0.069 0.068 -0.039
(0.351) (0.353) (0.357) (0.389)
Information Search 0.363 0.353 0.440
(0.261) (0.267) (0.286)
Democrat -0.667
+
-0.743
∗
(0.358) (0.370)
Republican -0.353 -0.246
(0.500) (0.513)
Political Ideology 0.023 -0.033
(0.149) (0.151)
Moderate -0.224 -0.235
(0.287) (0.297)
Dem. Policy Agreement 0.007 0.008
(0.007) (0.007)
Rep. Policy Agreement -0.003 -0.003
(0.009) (0.009)
Education 0.054
(0.101)
Income 0.129
(0.095)
Male 0.600
∗
(0.279)
Political interest 0.075
(0.240)
News Consumption -0.210
(0.237)
Religious Attendance 0.045
(0.089)
Age 0.015
(0.011)
White -0.130
(0.375)
Black -0.199
(0.481)
Hispanic/Latino 0.388
(0.529)
Constant 0.652
∗∗
0.557
∗
0.727 -0.536
(0.247) (0.255) (0.673) (1.370)
Observations 298 298 298 298
Pseudo R
2
0.003 0.008 0.021 0.050
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
251
Table B.39: Correct Prediction of Lethal Confrontation with N. Korea.
(1) (2) (3) (4)
Bonus Treatment 0.584 0.636 0.631 0.634
(0.464) (0.456) (0.472) (0.537)
Random Bonus Treatment 0.820
+
0.840
+
0.835
+
0.653
(0.497) (0.501) (0.506) (0.554)
Lottery Treatment -0.229 -0.220 -0.253 -0.213
(0.407) (0.408) (0.424) (0.473)
Information Search -0.229 -0.252 -0.293
(0.328) (0.343) (0.353)
Democrat -0.080 -0.171
(0.482) (0.479)
Republican -0.338 -0.278
(0.677) (0.753)
Political Ideology -0.040 -0.033
(0.207) (0.190)
Moderate -0.256 -0.321
(0.390) (0.404)
Dem. Policy Agreement -0.012 -0.008
(0.009) (0.010)
Rep. Policy Agreement -0.011 -0.008
(0.010) (0.011)
Education 0.080
(0.130)
Income 0.307
∗
(0.148)
Male 0.276
(0.367)
Political interest -0.582
(0.371)
News Consumption 0.412
(0.378)
Religious Attendance -0.052
(0.130)
Age 0.029
+
(0.016)
White -0.252
(0.510)
Black -0.255
(0.752)
Hispanic/Latino -0.076
(0.683)
Constant 1.438
∗∗∗
1.504
∗∗∗
2.890
∗∗
0.577
(0.298) (0.317) (1.012) (1.800)
Observations 298 298 298 298
Pseudo R
2
0.026 0.028 0.048 0.114
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
252
Table B.40: Correct Prediction of U.S. Sanctions on China for Trade.
(1) (2) (3) (4)
Bonus Treatment -0.053 -0.150 -0.147 -0.206
(0.360) (0.368) (0.382) (0.411)
Random Bonus Treatment 0.736
+
0.707
+
0.770
+
0.990
∗
(0.408) (0.411) (0.413) (0.483)
Lottery Treatment -0.110 -0.127 -0.125 -0.117
(0.361) (0.363) (0.375) (0.392)
Information Search 0.446 0.504
+
0.434
(0.291) (0.306) (0.309)
Democrat -0.569 -0.407
(0.398) (0.432)
Republican 0.572 0.807
(0.575) (0.587)
Political Ideology -0.182 -0.183
(0.164) (0.181)
Moderate -0.206 -0.027
(0.310) (0.329)
Dem. Policy Agreement -0.006 -0.006
(0.008) (0.008)
Rep. Policy Agreement -0.011 -0.009
(0.009) (0.010)
Education 0.095
(0.112)
Income -0.074
(0.113)
Male 0.647
∗
(0.316)
Political interest 0.180
(0.270)
News Consumption 0.061
(0.261)
Religious Attendance -0.160
+
(0.096)
Age 0.011
(0.012)
White -0.421
(0.401)
Black 0.476
(0.626)
Hispanic/Latino -0.370
(0.478)
Constant 0.907
∗∗∗
0.793
∗∗
2.302
∗∗
0.354
(0.259) (0.268) (0.856) (1.465)
Observations 298 298 298 298
Pseudo R
2
0.017 0.024 0.050 0.098
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
253
B.6 The Determinants of Time Spent on the Sur-
vey
254
Table B.41: The Determinants of Time Spent on the Survey (Seconds).
(1) (2) (3) (4)
Bonus Treatment 300.162
∗∗
215.274
∗
207.027
∗
219.849
∗
(97.910) (96.322) (98.557) (98.432)
Random Bonus Treatment 253.353
∗
223.433
∗
223.866
∗
281.471
∗∗
(108.174) (103.956) (104.893) (108.108)
Lottery Treatment -14.728 -28.972 -44.191 -23.156
(94.738) (92.360) (93.338) (97.125)
Information Search 386.670
∗∗∗
386.506
∗∗∗
374.063
∗∗∗
(71.264) (70.763) (74.492)
Democrat -118.722 -74.319
(98.874) (96.585)
Republican -124.845 -73.812
(147.506) (157.493)
Political Ideology -41.747 -52.187
(37.829) (39.494)
Moderate -59.445 -18.175
(78.067) (82.340)
Dem. Policy Agreement 0.662 -0.087
(2.124) (2.179)
Rep. Policy Agreement 3.482
+
3.348
(2.081) (2.061)
Education -13.540
(26.140)
Income -50.177
∗
(23.160)
Male 31.049
(76.462)
Political interest 138.828
+
(71.112)
News Consumption -118.366
(73.259)
Religious Attendance -20.004
(22.824)
Age 0.591
(2.731)
White -4.507
(80.733)
Black 203.656
+
(123.377)
Hispanic/Latino 102.254
(135.491)
Constant 990.890
∗∗∗
884.953
∗∗∗
995.444
∗∗∗
1182.731
∗∗
(70.122) (74.250) (228.668) (407.711)
Observations 298 298 298 298
R
2
0.052 0.136 0.151 0.194
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
255
B.7 Randomization Tests
256
Table B.42: Randomization Check: F-test of Joint Orthogonality.
Control Bonus
Treatment
Random
Bonus
Treatment
Lottery
Treatment
P-value from
Joint
Orthogonality
Test of
Treatment
Arms
Democrat 0.466 0.481 0.527 0.432 0.715
(0.059) (0.057) (0.058) (0.058)
Republican 0.260 0.156 0.189 0.216 0.446
(0.052) (0.042) (0.046) (0.048)
Political Ideology 3.438 3.182 3.351 3.405 0.817
(0.215) (0.191) (0.214) (0.205)
Moderate 0.384 0.468 0.378 0.419 0.667
(0.057) (0.057) (0.057) (0.058)
Dem. Policy Agreement 51.795 57.117 55.189 49.338 0.401
(3.594) (3.366) (3.679) (3.414)
Rep. Policy Agreement 31.425 27.909 29.068 33.757 0.630
(3.590) (3.272) (3.533) (3.306)
Education 9.890 10.429 10.324 10.095 0.072
(0.184) (0.131) (0.147) (0.163)
Income 2.836 2.844 2.797 2.703 0.931
(0.182) (0.168) (0.161) (0.162)
Male 0.534 0.662 0.541 0.703 0.078
(0.059) (0.054) (0.058) (0.053)
Political interest 2.411 2.623 2.257 2.419 0.107
(0.095) (0.091) (0.114) (0.122)
News Consumption 2.315 2.571 2.338 2.365 0.327
(0.108) (0.097) (0.111) (0.126)
Religious Attendance 0.658 1.130 1.108 1.095 0.169
(0.145) (0.183) (0.175) (0.187)
Age 38.726 38.221 40.986 36.378 0.182
(1.550) (1.470) (1.530) (1.362)
White 0.726 0.662 0.784 0.757 0.371
(0.053) (0.054) (0.048) (0.050)
Black 0.151 0.143 0.068 0.068 0.182
(0.042) (0.040) (0.029) (0.029)
Hispanic/Latino 0.068 0.143 0.054 0.095 0.243
(0.030) (0.040) (0.026) (0.034)
Standard errors in parentheses
257
Table B.43: Randomization Check: T-tests of the Differences of Means between Variables.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Control Condition Bonus
Treatment
Random
Bonus
Treatment
Lottery
Treatment
(1) vs. (2),
p-value
(1) vs. (3),
p-value
(1) vs. (4),
p-value
(2) vs. (3),
p-value
(2) vs. (4),
p-value
(3) vs. (4),
p-value
P-value from
Joint
Orthogonality
Test of
Treatment
Arms
Democrat 0.466 0.481 0.527 0.432 0.858 0.461 0.687 0.571 0.556 0.252 0.715
Republican 0.260 0.156 0.189 0.216 0.116 0.305 0.534 0.590 0.343 0.685 0.446
Political Ideology 3.438 3.182 3.351 3.405 0.372 0.774 0.912 0.554 0.426 0.856 0.817
Moderate 0.384 0.468 0.378 0.419 0.302 0.949 0.665 0.271 0.551 0.617 0.667
Dem. Policy Agreement 51.795 57.117 55.189 49.338 0.281 0.510 0.621 0.699 0.107 0.246 0.401
Rep. Policy Agreement 31.425 27.909 29.068 33.757 0.470 0.640 0.633 0.810 0.211 0.334 0.630
Education 9.890 10.429 10.324 10.095 0.017 0.066 0.407 0.597 0.112 0.297 0.072
Income 2.836 2.844 2.797 2.703 0.973 0.875 0.586 0.841 0.546 0.679 0.931
Male 0.534 0.662 0.541 0.703 0.111 0.940 0.036 0.128 0.597 0.042 0.078
Political interest 2.411 2.623 2.257 2.419 0.108 0.301 0.959 0.012 0.178 0.332 0.107
News Consumption 2.315 2.571 2.338 2.365 0.079 0.884 0.765 0.114 0.193 0.872 0.327
Religious Attendance 0.658 1.130 1.108 1.095 0.047 0.049 0.067 0.932 0.893 0.958 0.169
Age 38.726 38.221 40.986 36.378 0.813 0.301 0.257 0.194 0.360 0.026 0.182
White 0.726 0.662 0.784 0.757 0.401 0.419 0.673 0.097 0.204 0.698 0.371
Black 0.151 0.143 0.068 0.068 0.893 0.107 0.107 0.135 0.135 1.000 0.182
Hispanic/Latino 0.068 0.143 0.054 0.095 0.142 0.717 0.566 0.069 0.364 0.351 0.243
N 73 77 74 74
258
B.8 Causal Mediation Sensitivity Analysis
Figure B.1: Average Causal Mediation Effect of the Bonus Treatment on Political
Knowledge as a Function of the Degree of Violation of Sequential Ignorability As-
sumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
259
FigureB.2: AverageCausalMediationEffectoftheRandomBonusTreatmentonPo-
litical Knowledge as a Function of the Degree of Violation of Sequential Ignorability
Assumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
260
Figure B.3: Average Causal Mediation Effect of the Lottery Treatment on Politi-
cal Knowledge as a Function of the Degree of Violation of Sequential Ignorability
Assumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
261
Figure B.4: Average Causal Mediation Effect of the Bonus Treatment on the Num-
ber of Correct Predictions as a Function of the Degree of Violation of Sequential
Ignorability Assumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
262
FigureB.5: AverageCausalMediationEffectoftheRandomBonusTreatmentonthe
Number of Correct Predictions as a Function of the Degree of Violation of Sequential
Ignorability Assumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
263
Figure B.6: Average Causal Mediation Effect of the Lottery Treatment on the Num-
ber of Correct Predictions as a Function of the Degree of Violation of Sequential
Ignorability Assumption.
The results were calculated using 1000 simulations with 95% confidence intervals.
The table was created using the mediation package in Stata (Hicks and Tingley
2011).
264
B.9 Expected Foreign Policy Attitudes in Exper-
iment
If people only take positions aligned to party positions, we should expect to see
that their policy positions will match the party positions outlined in Table B.44. If
this is the case, there should be significant differences between the distribution of
attitudes among Democrats and Republicans on any given policy, and especially on
policies that are polarized at the elite level.
Table B.44: Expected and Actual Foreign Policy Attitudes Based on Party ID.
Foreign Policy Democratic Position Republican Position
Muslim Ban Oppose Support
Withdrawal from Paris Agreement Oppose Support
Path to Citizenship Support Oppose
Russian Sanctions Support Oppose
Improve Relations with Russia Oppose Support
North Korea Pre-emptive Strike No Expectation No Expectation
Iran Nuclear Deal Support Oppose
China Trade Restrictions Oppose Support
265
Table B.45 presents expectations of policy attitudes among those with ideological
preferences. If there are differences between ideologues of liberal and conservative
persuasions in the way they perceive each issue, this should be reflected in their
attitudes about policy in the experiment.
Table B.45: Expected Foreign Policy Attitudes Based on Political Ideology.
Foreign Policy Liberal Position Conservative Position
Muslim Ban Oppose Oppose
Withdrawal from Paris Agreement Oppose Support
Path to Citizenship Support Oppose
Russian Sanctions Support Support
Improve Relations with Russia Support Oppose
North Korea Pre-emptive Strike Oppose Support
Iran Nuclear Deal Support Oppose
China Trade Restrictions Support Oppose
266
Appendix C
Appendix for Chapter 5
267
C.1 Descriptive Statistics.
Table C.1: Descriptive Statistics for Variables in the Pew Poll, April 25-28, 2013.
Count Mean Std. Dev. Min. Max.
Attention 990 2.365657 1.088423 1 4
Registered Voter 1001 .8101898 .3923469 0 1
Republican 1003 .2522433 .4345168 0 1
Democrat 1003 .3230309 .4678677 0 1
Independent 1003 .3609172 .4805062 0 1
Employed 1000 .528 .4994652 0 1
Male 1003 .4825523 .4999448 0 1
Age 965 52.38653 18.79914 18 97
Education 999 4.681682 1.837806 1 8
Income 869 5.009206 2.46401 1 9
No. Active Duty Military 1005 42049.33 50458.88 122 157860
No. Veterans 1005 8.327264 1.572058 5.2 11.9
Democratic Party Vote 1006 2493509 2809537 69286 6.54e+07
Republican Party Vote 1006 2151686 2312228 21381 6.06e+07
Information Search (April) 1005 32.23085 14.68102 0 100
268
Table C.2: Descriptive Statistics for Variables in the Pew Poll, August 28-September
1, 2013.
Count Mean Std. Dev. Min. Max.
Attention 999 3.072072 1.040662 1 4
Registered Voter 976 .8217213 .3829434 0 1
Republican 1000 .226 .4184484 0 1
Democrat 1000 .301 .4589222 0 1
Independent 1000 .352 .4778329 0 1
Employed 993 .4884189 .5001177 0 1
Male 1000 .462 .4988034 0 1
Education 983 4.839268 1.884281 1 8
Income 810 5.028395 2.387738 1 9
No. Active Duty Military 1002 41083.06 49501.63 122 157860
No. Veterans 1002 8.262375 1.553139 5.2 11.9
Democratic Party Vote 1003 2493019 2810758 69286 6.54e+07
Republican Party Vote 1003 2166670 2321919 21381 6.06e+07
Information Search (August) 1002 26.77445 7.219343 14 100
269
Table C.3: Descriptive Statistics for Variables in the Pew Poll, September 12-15,
2013.
Count Mean Std. Dev. Min. Max.
Attention 1000 3.217 .9522698 1 4
Registered Voter 984 .8384146 .3682572 0 1
Republican 1002 .2355289 .4245409 0 1
Democrat 1002 .2964072 .4569008 0 1
Independent 1002 .3542914 .4785369 0 1
Employed 993 .5679758 .4956073 0 1
Male 1002 .493014 .5002009 0 1
Age 960 52.36146 18.5807 18 97
Education 986 4.85497 1.782105 1 8
Income 810 5.21358 2.481589 1 9
No. Active Duty Military 1005 47877.06 53952.4 122 157860
No. Veterans 1005 8.280199 1.624549 5.2 11.9
Democratic Party Vote 1006 2672569 2938229 69286 6.54e+07
Republican Party Vote 1006 2261235 2338928 21381 6.06e+07
Information Search (September) 1005 38.75423 9.766716 19 100
270
C.2 Polls about U.S. Military Action in Syria
This section of the Appendix provides more information about the three publicly
available public opinion polls regarding U.S. military action in Syria in response to
the use of chemical weapons by Syrian President Bashar al-Assad that are used in
this chapter. The information about the polls is presented in chronological order,
ordered by the final date of data collection of each poll.
It is important to note that the question wording and the sampling methods
vary between polls, which makes it difficult to compare responses between polls. As
a result, this data is only analyzed individually. The data itself is available in a
Stata.dta file, which is available from the author upon request, and it will also be
made publicly available upon publication.
1. Pew Poll, April 25-28, 2013
Publication Date: April 29, 2013
Number of Observations: 1003
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
2. Pew Poll, August 29-September 1, 2013
Publication Date: September 3, 2013
Number of Observations: 1000
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
3. Pew Poll, September 12-15, 2013
Publication Date: September 16, 2013
Number of Observations: 1002
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
271
Appendix D
Appendix for Chapter 6
D.1 Codebook
This section provides full details of the question wording and the codebook for all
variables used in this analysis. Full details about Wave 8 of the 2015 British Election
Survey (Evans et al. 2016) are available to view here.
D.1.1 Dependent Variables: Vote Intention
If you do vote in the referendum on Britain’s membership of the European Union,
how do you think you will vote? [euRefVote]
• Remain in the EU (0)
• Leave the EU (1) [Vote Leave = 1]
• Don’t know (9999) [excluded from analysis]
D.1.2 Independent Variables
Perception of Impact
How much impact do you think that Britain leaving the EU would have on the
country as a whole? [leaveImpactBritain]
• No impact at all (1)
• A small impact (2)
272
• A moderate impact (3)
• A large impact (4)
• A very large impact (5)
• Don’t know (9999) [excluded from analysis]
Attention
Have you received and read the UK Government’s leaflet “Why the Government
believes that voting to remain in the European Union is the best decision for the
UK”? [euGovtLeaflet]
• Yes, I received and read it (1)
• Yes, I received it but I haven’t read it (0)
• No, I haven’t received it (8888) [excluded from analysis]
• Don’t know (9999) [excluded from analysis]
D.1.3 Political Control Variables
Political Interest
How interested are you in the EU referendum that will be held on June 23rd? [eu-
RefInterest]
• Very interested (4)
• Somewhat interested (3)
• Not very interested (2)
• Not at all interested (1)
• Don’t know (9999) [excluded from analysis]
273
Political Knowledge
Political knowledge is the total number of correct answers for the six True/False
questions asked of participants to measure their knowledge about the relationship
between the United Kingdom and the European Union. Correct answers are in-
dicated in bold. Participants scored 1 for every correct response, and 0 for every
incorrect response and “Don’t know” response.
Question 1. Each EU Member State elects the same number of representatives
to the European Parliament [euKnowGrid1]
• True (1)
• False (2)
• Don’t know (9999)
Question 2. Switzerland is a member of the EU [euKnowGrid2]
• True (1)
• False (2)
• Don’t know (9999)
Question 3. Croatia is a member of the EU [euKnowGrid3]
• True (1)
• False (2)
• Don’t know (9999)
Question 4. The EU spends more on agriculture than any other policy area
[euKnowGrid4]
• True (1)
• False (2)
• Don’t know (9999)
Question 5. The European Court of Human Rights only has jurisdiction over EU
members [euKnowGrid5]
274
• True (1)
• False (2)
• Don’t know (9999)
Question 6. The European Union is made up of 15 member states [euKnowGrid6]
• True (1)
• False (2)
• Don’t know (9999)
Voter
Thinking back to the last UK General Election on May 7th 2015, a lot of people
didn’t manage to vote. How about you - did you manage to vote in the General
Election in 2015? [voted2015]
• Yes, voted (1)
• No, did not vote (0)
• Don’t know (9999)
2015 Vote
Which party did you vote for? [recallVote15]
• Conservative (1) [Conservative Vote 2015 = 1]
• Labour (2) [Labour Vote 2015 = 1]
• Liberal Democrat (3) [Lib Dem Vote 2015 = 1]
• Scottish National Party (SNP) (4) [SNP Vote 2015 = 1]
• Plaid Cymru (5)
• United Kingdom Independence Party (UKIP) (6) [UKIP Vote 2015 = 1]
• Green Party (7) [Green Vote 2015 = 1]
275
• British National Party (BNP) (8)
• Other (9)
• Don’t know (9999)
D.1.4 Risk and Uncertainty Control Variables
Certainty Effect Leave
How sure are you about what would happen to the UK if it left the EU? [certain-
tyUKLeave]
• Very unsure (1)
• Quite unsure (2)
• Quite sure (3)
• Very sure (4)
• Don’t know (9999) [excluded from analysis]
Risk Taking
Generally speaking, how willing are you to take risks? [riskTaking]
• Very willing to take risks (4)
• Somewhat willing to take risks (3)
• Somewhat unwilling to take risks (2)
• Very unwilling to take risks (1)
D.1.5 Region Control Variables
Region
Which area of the UK do you live in? [gor]
• North East (1) [North East = 1]
276
• North West (2) [North West = 1]
• Yorkshire and the Humber (3) [Yorkshire and the Humber = 1]
• East Midlands (4) [East Midlands = 1]
• West Midlands (5) [West Midlands = 1]
• East of England (6) [East of England = 1]
• London (7) [London = 1]
• South East (8) [South East = 1]
• South West (9) [South West = 1]
• Wales (10) [Wales = 1]
• Scotland (11) [Scotland = 1]
• Northern Ireland (12) [Northern Ireland = 1]
• Non UK & Invalid (13) [Other = 1]
D.1.6 Demographic Control Variables
Male
Are you male or female? [gender]
• Male (1)
• Female (0)
Age
What is your age? [age]
277
Household Income
What is your gross household income? [profile-gross-household]
• under 5,000 per year (1)
• 5,000 to 9,999 per year (2)
• 10,000 to 14,999 per year (3)
• 15,000 to 19,999 per year (4)
• 20,000 to 24,999 per year (5)
• 25,000 to 29,999 per year (6)
• 30,000 to 34,999 per year (7)
• 35,000 to 39,999 per year (8)
• 40,000 to 44,999 per year (9)
• 45,000 to 49,999 per year (10)
• 50,000 to 59,999 per year (11)
• 60,000 to 69,999 per year (12)
• 70,000 to 99,999 per year (13)
• 100,000 to 149,999 per year (14)
• 150,000 and over (15)
• Don’t know (16) [excluded from analysis]
• Prefer not to answer (17) [excluded from analysis]
278
Race and Ethnicity
To which of these groups do you consider you belong? [profile-ethnicity]
• White British (1) [White-British = 1]
• Any other white background (2) [White-Other = 1]
• White and Black Caribbean (3) [B.A.M.E. = 1]
• White and Black African (4) [B.A.M.E. = 1]
• White and Asian (5) [B.A.M.E. = 1]
• Any other mixed background (6) [B.A.M.E. = 1]
• Indian (7) [B.A.M.E. = 1]
• Pakistani (8) [B.A.M.E. = 1]
• Bangladeshi (9) [B.A.M.E. = 1]
• Any other Asian background (10) [B.A.M.E. = 1]
• Black Caribbean (11) [B.A.M.E. = 1]
• Black African (12) [B.A.M.E. = 1]
• Any other black background (13) [B.A.M.E. = 1]
• Chinese (14) [B.A.M.E. = 1]
• Other ethnic group (15) [B.A.M.E. = 1]
• Refused (16) [excluded from analysis]
279
D.2 Descriptive Statistics
Table D1 indicates a large amount of variation in the independent and dependent
variables of interest. It is important to note that some observations are dropped
on two variables especially: 1) Attention and 2) Household Income. The attention
variable relied on the delivery of the Government leaflet, so many observations were
removed from the analysis as a result of the document not having been successfully
delivered to participants in the survey. Household income is a sensitive variable that
many people prefer not to answer, which helps explain why roughly one quarter of
participants failed to answer that question.
280
Table D.1: Descriptive Statistics.
Count Mean Std. Dev. Min. Max.
Vote Leave 33502 .4714047 .4991891 0 1
Perception of Impact 31020 3.633914 1.039668 1 5
Read Leaflet 26192 .6190822 .4856217 0 1
Political Interest 33085 3.540517 .7046861 1 4
Political Knowledge 33502 2.655155 1.66549 0 6
Voter 31212 .9057093 .2922373 0 1
Conservative Vote 2015 33502 .2802519 .4491289 0 1
Labour Vote 2015 33502 .2591189 .4381575 0 1
LIb Dem Vote 2015 33502 .0711599 .2570956 0 1
SNP Vote 2015 33502 .0510417 .2200862 0 1
UKIP Vote 2015 33502 .1093666 .3121033 0 1
Green Vote 2015 33502 .0448928 .2070719 0 1
Certainty Effect Leave 27795 2.668897 .7441664 1 4
Risk Taking 33502 2.555191 .7289928 1 4
North East 33502 .0323563 .1769471 0 1
North West 33502 .0759059 .2648515 0 1
Yorkshire and the Humber 33502 .0638469 .2444837 0 1
East Midlands 33502 .0538475 .2257199 0 1
West Midlands 33502 .0578174 .2334014 0 1
East of England 33502 .0705928 .2561473 0 1
London 33502 .09683 .2957306 0 1
South East 33502 .1022029 .3029194 0 1
South West 33502 .0665035 .2491639 0 1
Wales 33502 .069369 .2540844 0 1
Scotland 33502 .1188884 .323662 0 1
Male 33502 .489135 .4998894 0 1
Age 33502 52.3932 16.16792 18 96
Income 25166 8.824923 5.218967 1 17
White-British 33502 .729837 .4440505 0 1
White-Other 33502 .0310429 .1734363 0 1
B.A.M.E. 33502 .0360277 .1863619 0 1
281
D.3 Full Results using Logistic Regression
Table D.2: Logit Regression: The Determinants of Voting to Leave the European Union.
(1) (2) (3) (4) (5) (6)
Perception of Impact -1.079
∗∗∗
-1.150
∗∗∗
-1.233
∗∗∗
-1.323
∗∗∗
-1.322
∗∗∗
-1.315
∗∗∗
(0.015) (0.017) (0.021) (0.024) (0.024) (0.028)
Read Leaflet 0.041 -0.131
∗∗∗
-0.171
∗∗∗
-0.186
∗∗∗
-0.233
∗∗∗
(0.030) (0.036) (0.040) (0.040) (0.046)
Political Control Variables 3 3 3 3
Risk and Uncertainty Control Variables 3 3 3
Region Control Variables 3 3
Demographic Control Variables 3
Constant 3.864
∗∗∗
4.204
∗∗∗
2.202
∗∗∗
0.689
∗∗∗
0.644
∗∗∗
-0.149
(0.056) (0.067) (0.132) (0.167) (0.173) (0.300)
Observations 31020 24852 23437 20694 20694 15558
Pseudo R
2
0.168 0.184 0.330 0.377 0.379 0.390
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
282
Table D.3: Logit Regression: The Determinants of Voting to Leave the European
Union.
(1) (2) (3) (4) (5) (6)
Perception of Impact -1.079
∗∗∗
-1.150
∗∗∗
-1.233
∗∗∗
-1.323
∗∗∗
-1.322
∗∗∗
-1.315
∗∗∗
(0.015) (0.017) (0.021) (0.024) (0.024) (0.028)
Read Leaflet 0.041 -0.131
∗∗∗
-0.171
∗∗∗
-0.186
∗∗∗
-0.233
∗∗∗
(0.030) (0.036) (0.040) (0.040) (0.046)
Political Interest 0.692
∗∗∗
0.594
∗∗∗
0.598
∗∗∗
0.582
∗∗∗
(0.032) (0.037) (0.037) (0.043)
Political Knowledge -0.056
∗∗∗
-0.098
∗∗∗
-0.092
∗∗∗
-0.086
∗∗∗
(0.011) (0.012) (0.012) (0.015)
Voter -0.283
∗
-0.263
∗
-0.189 -0.274
+
(0.113) (0.127) (0.130) (0.157)
Conservative Vote 2015 0.771
∗∗∗
0.841
∗∗∗
0.768
∗∗∗
0.685
∗∗∗
(0.098) (0.109) (0.113) (0.135)
Labour Vote 2015 -0.342
∗∗∗
-0.402
∗∗∗
-0.472
∗∗∗
-0.553
∗∗∗
(0.098) (0.110) (0.113) (0.135)
Lib Dem Vote 2015 -0.622
∗∗∗
-0.586
∗∗∗
-0.639
∗∗∗
-0.760
∗∗∗
(0.110) (0.122) (0.125) (0.150)
SNP Vote 2015 -0.307
∗∗
-0.496
∗∗∗
-0.288
∗
-0.337
∗
(0.116) (0.130) (0.144) (0.169)
UKIP Vote 2015 3.348
∗∗∗
3.229
∗∗∗
3.141
∗∗∗
3.025
∗∗∗
(0.132) (0.144) (0.146) (0.171)
Green Vote 2015 -0.904
∗∗∗
-0.909
∗∗∗
-0.986
∗∗∗
-0.988
∗∗∗
(0.126) (0.141) (0.144) (0.172)
Certainty Effect Leave 0.758
∗∗∗
0.761
∗∗∗
0.736
∗∗∗
(0.027) (0.027) (0.033)
Risk Taking 0.129
∗∗∗
0.131
∗∗∗
0.207
∗∗∗
(0.028) (0.028) (0.033)
North East 0.375
∗∗
0.746
∗∗∗
(0.114) (0.131)
North West 0.164
∗
0.520
∗∗∗
(0.081) (0.103)
Yorkshire and the Humber 0.198
∗
0.549
∗∗∗
(0.083) (0.105)
East Midlands 0.070 0.415
∗∗∗
(0.089) (0.111)
West Midlands 0.077 0.409
∗∗∗
(0.088) (0.111)
East of England 0.116 0.506
∗∗∗
(0.082) (0.105)
London -0.181
∗
0.265
∗∗
(0.074) (0.100)
South East -0.022 0.334
∗∗∗
(0.070) (0.096)
South West 0.017 0.298
∗∗
(0.084) (0.106)
Wales -0.168
∗
0.073
(0.084) (0.106)
Scotland -0.308
∗∗∗
0.000
(0.078) (.)
Male -0.210
∗∗∗
(0.046)
Age 0.017
∗∗∗
(0.002)
Household Income -0.024
∗∗∗
(0.004)
White-British -0.048
(0.206)
White-Other -0.694
∗∗
(0.235)
B.A.M.E. -0.127
(0.237)
Constant 3.864
∗∗∗
4.204
∗∗∗
2.202
∗∗∗
0.689
∗∗∗
0.644
∗∗∗
-0.149
(0.056) (0.067) (0.132) (0.167) (0.173) (0.300)
Observations 31020 24852 23437 20694 20694 15558
Pseudo R
2
0.168 0.184 0.330 0.377 0.379 0.390
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
283
D.4 Alternative Specification of the Model using
Probit Regression
284
Table D.4: Probit Regression: The Determinants of Voting to Leave the European
Union.
(1) (2) (3) (4) (5) (6)
Perception of Impact -0.645
∗∗∗
-0.686
∗∗∗
-0.709
∗∗∗
-0.750
∗∗∗
-0.749
∗∗∗
-0.745
∗∗∗
(0.009) (0.010) (0.012) (0.013) (0.013) (0.015)
Read Leaflet 0.026 -0.076
∗∗∗
-0.098
∗∗∗
-0.106
∗∗∗
-0.132
∗∗∗
(0.018) (0.021) (0.023) (0.023) (0.027)
Political Interest 0.396
∗∗∗
0.341
∗∗∗
0.343
∗∗∗
0.334
∗∗∗
(0.019) (0.021) (0.021) (0.024)
Political Knowledge -0.034
∗∗∗
-0.057
∗∗∗
-0.053
∗∗∗
-0.050
∗∗∗
(0.006) (0.007) (0.007) (0.009)
Voter -0.168
∗
-0.154
∗
-0.112 -0.152
+
(0.066) (0.074) (0.075) (0.090)
Conservative Vote 2015 0.455
∗∗∗
0.487
∗∗∗
0.448
∗∗∗
0.399
∗∗∗
(0.057) (0.063) (0.065) (0.077)
Labour Vote 2015 -0.208
∗∗∗
-0.239
∗∗∗
-0.278
∗∗∗
-0.323
∗∗∗
(0.057) (0.064) (0.065) (0.077)
Lib Dem Vote 2015 -0.371
∗∗∗
-0.347
∗∗∗
-0.374
∗∗∗
-0.437
∗∗∗
(0.064) (0.071) (0.072) (0.086)
SNP Vote 2015 -0.182
∗∗
-0.291
∗∗∗
-0.172
∗
-0.200
∗
(0.068) (0.075) (0.083) (0.097)
UKIP Vote 2015 1.824
∗∗∗
1.753
∗∗∗
1.703
∗∗∗
1.633
∗∗∗
(0.073) (0.080) (0.081) (0.095)
Green Vote 2015 -0.522
∗∗∗
-0.521
∗∗∗
-0.563
∗∗∗
-0.567
∗∗∗
(0.072) (0.080) (0.082) (0.097)
Certainty Effect Leave 0.430
∗∗∗
0.431
∗∗∗
0.415
∗∗∗
(0.016) (0.016) (0.019)
Risk Taking 0.070
∗∗∗
0.071
∗∗∗
0.114
∗∗∗
(0.016) (0.016) (0.019)
North East 0.211
∗∗
0.425
∗∗∗
(0.066) (0.076)
North West 0.094
∗
0.303
∗∗∗
(0.046) (0.059)
Yorkshire and the Humber 0.111
∗
0.316
∗∗∗
(0.048) (0.061)
East Midlands 0.040 0.239
∗∗∗
(0.052) (0.064)
West Midlands 0.040 0.232
∗∗∗
(0.052) (0.065)
East of England 0.059 0.283
∗∗∗
(0.047) (0.061)
London -0.104
∗
0.154
∗∗
(0.043) (0.058)
South East -0.013 0.196
∗∗∗
(0.040) (0.056)
South West 0.002 0.168
∗∗
(0.048) (0.061)
Wales -0.099
∗
0.042
(0.048) (0.061)
Scotland -0.176
∗∗∗
0.000
(0.045) (.)
Male -0.119
∗∗∗
(0.026)
Age 0.010
∗∗∗
(0.001)
Household Income -0.014
∗∗∗
(0.003)
White-British -0.046
(0.116)
White-Other -0.400
∗∗
(0.133)
B.A.M.E. -0.084
(0.134)
Constant 2.309
∗∗∗
2.508
∗∗∗
1.282
∗∗∗
0.387
∗∗∗
0.365
∗∗∗
-0.082
(0.032) (0.038) (0.075) (0.095) (0.099) (0.171)
Observations 31020 24852 23437 20694 20694 15558
Pseudo R
2
0.167 0.183 0.327 0.373 0.376 0.386
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
285
D.5 Full Results with Alternative Coding of the
Dependent Variable
286
Table D.5: OLS Regression: The Determinants of Voting to Leave the European
Union, including Don’t Know and Nonresponses.
(1) (2) (3) (4) (5) (6)
Perception of Impact -0.449
∗∗∗
-0.464
∗∗∗
-0.396
∗∗∗
-0.388
∗∗∗
-0.386
∗∗∗
-0.379
∗∗∗
(0.004) (0.004) (0.005) (0.005) (0.005) (0.006)
Read Leaflet -0.005 -0.050
∗∗∗
-0.055
∗∗∗
-0.059
∗∗∗
-0.074
∗∗∗
(0.011) (0.011) (0.011) (0.011) (0.013)
Political Interest 0.176
∗∗∗
0.147
∗∗∗
0.148
∗∗∗
0.142
∗∗∗
(0.009) (0.010) (0.010) (0.011)
Political Knowledge -0.023
∗∗∗
-0.031
∗∗∗
-0.029
∗∗∗
-0.027
∗∗∗
(0.003) (0.004) (0.004) (0.004)
Voter -0.111
∗∗
-0.102
∗
-0.074
+
-0.115
∗
(0.039) (0.042) (0.042) (0.049)
Conservative Vote 2015 0.272
∗∗∗
0.283
∗∗∗
0.256
∗∗∗
0.237
∗∗∗
(0.033) (0.036) (0.036) (0.042)
Labour Vote 2015 -0.183
∗∗∗
-0.190
∗∗∗
-0.215
∗∗∗
-0.228
∗∗∗
(0.033) (0.036) (0.036) (0.041)
Lib Dem Vote 2015 -0.259
∗∗∗
-0.244
∗∗∗
-0.264
∗∗∗
-0.278
∗∗∗
(0.037) (0.039) (0.039) (0.045)
SNP Vote 2015 -0.163
∗∗∗
-0.204
∗∗∗
-0.159
∗∗∗
-0.156
∗∗
(0.039) (0.041) (0.045) (0.051)
UKIP Vote 2015 0.766
∗∗∗
0.704
∗∗∗
0.673
∗∗∗
0.647
∗∗∗
(0.034) (0.036) (0.036) (0.042)
Green Vote 2015 -0.346
∗∗∗
-0.335
∗∗∗
-0.361
∗∗∗
-0.353
∗∗∗
(0.039) (0.041) (0.041) (0.048)
Certainty Effect Leave 0.188
∗∗∗
0.188
∗∗∗
0.175
∗∗∗
(0.007) (0.007) (0.009)
Risk Taking 0.037
∗∗∗
0.037
∗∗∗
0.057
∗∗∗
(0.008) (0.008) (0.009)
North East 0.103
∗∗
0.096
∗
(0.032) (0.038)
North West 0.049
∗
0.036
(0.023) (0.031)
Yorkshire and the Humber 0.056
∗
0.042
(0.023) (0.031)
East Midlands 0.017 0.000
(0.025) (.)
West Midlands 0.023 0.001
(0.025) (0.032)
East of England 0.032 0.025
(0.023) (0.030)
London -0.050
∗
-0.041
(0.021) (0.029)
South East -0.010 -0.025
(0.020) (0.028)
South West 0.009 -0.028
(0.023) (0.031)
Wales -0.068
∗∗
-0.111
∗∗∗
(0.023) (0.031)
Scotland -0.079
∗∗∗
-0.113
∗∗∗
(0.023) (0.032)
Male -0.060
∗∗∗
(0.013)
Age 0.005
∗∗∗
(0.000)
Household Income -0.007
∗∗∗
(0.001)
White-British -0.037
(0.053)
White-Other -0.194
∗∗
(0.059)
B.A.M.E. -0.059
(0.062)
Constant 1.647
∗∗∗
1.734
∗∗∗
0.958
∗∗∗
0.451
∗∗∗
0.437
∗∗∗
0.336
∗∗∗
(0.015) (0.017) (0.039) (0.049) (0.051) (0.083)
Observations 31020 24852 23437 20694 20694 15558
R
2
0.230 0.244 0.376 0.411 0.413 0.424
Standard errors in parentheses
+
p< 0.1,
∗
p< 0.05,
∗∗
p< 0.01,
∗∗∗
p< 0.001
287
Appendix E
Appendix for Chapter 7
E.1 Polls about U.S. Military Action in Syria
Appendix E.1 provides more information about all available public opinion polls
conducted by different polling firms and agencies with questions regarding U.S. mil-
itary action in Syria in response to the use of chemical weapons by Syrian President
Bashar al-Assad. This information is presented in chronological order, ordered by
the final date of data collection of each poll.
Again, it is important to note that the question wording and the sampling meth-
ods vary between polling agencies, which makes it difficult to compare attitudes to
U.S. intervention in Syria between polls. As a result, this data is only presented
for illustrative purposes in this chapter. The data itself is available in a Stata.dta
file, which is available from the author upon request. This data will also be made
publicly available upon publication.
1. ABC News/Washington Post Poll, December 13-16, 2012
Publication Date: April 6, 2013
Question: What if the Syrian government uses chemical weapons against its
people? In that case would you support or oppose U.S. military involvement
in Syria? (Asked of half sample)
Number of Observations: 1002
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
2. Pew Poll, April 25-28, 2013
Publication Date: April 29, 2013
288
Question: If it is confirmed that the Syrian government has used chemical
weapons against anti-government groups, would you favor or oppose the U.S.
and its allies taking military action against the Syrian government?
Number of Observations: 1003
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
3. CNN/ORC Poll, May 17-18, 2013
Publication Date: May 18, 2013
Question: If the United States were able to present evidence that convinced
you that the Syrian government has chemical weapons and has used them to
kill civilians in that country, do you think the U.S. would or would not be
justified in using military action against the Syrian government?
Number of Observations: 923
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
4. Reuters/Ipsos Poll, May 17-21, 2013
Publication Date: May 21, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1498
Population: Adults
Sampling Mode: Internet
Margin of Error: 2.9
5. Reuters/Ipsos Poll, May 24-28, 2013
Publication Date: May 28, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1985
Population: Adults
289
Sampling Mode: Internet
Margin of Error: 2.5
6. Reuters/Ipsos Poll, July 5-9, 2013
Publication Date: July 11, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1437
Population: Adults
Sampling Mode: Internet
Margin of Error: 2.9
7. Reuters/Ipsos Poll, August 19-23, 2013
Publication Date: August 26, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1448
Population: Adults
Sampling Mode: Internet
Margin of Error: 2.9
8. NBC News/Wall Street Journal Poll, August 28-29 2013
Publication Date: August 30, 2013
Question: It has been reported that the Syrian government has used chemical
weaponsonitscitizens. DoyouthinktheUnitedStatesshouldtakemilitaryac-
tion against the Syrian government in response to the use of chemical weapons
or not?
Number of Observations: 700
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
9. Reuters/Ipsos Poll, August 26-30, 2013
Publication Date: August 30, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
290
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 708
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.2
10. ABC News/Washington Post Poll, August 28-September 1, 2013
Publication Date: September 4, 2013
Question: The United States says it has determined that the Syrian govern-
ment has used chemical weapons in the civil war there. Given this, do you
support or oppose the United States launching missile strikes against the Syr-
ian government?
Number of Observations: 1012
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
11. Pew Poll, August 29-September 1, 2013
Publication Date: September 3, 2013
Question: Would you favor or oppose the U.S. conducting military airstrikes
against Syria in response to reports that the Syrian government used chemical
weapons?
Number of Observations: 1000
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
12. Reuters/Ipsos Poll, August 30-September 3, 2013
Publication Date: September 3, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1195
Population: Adults
Sampling Mode: Internet
Margin of Error: 3.2
291
13. YouGov/Economist Poll, September 3-4, 2013
Publication Date: September 9, 2013
Question: Do you support or oppose the use of military force against Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: Not available
14. Gallup Poll, September 3-4, 2013
Publication Date: September 6, 2013
Question: Would you favor or oppose the U.S. taking military action against
Syria in order to reduce that country’s ability to use chemical weapons?
Number of Observations: 1021
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
15. Rasmussen Reports Poll, September 6-7, 2013
Publication Date: September 8, 2013
Question: PresidentObamaisseekingcongressionalapprovalforU.S.military
attacks on Syria in retaliation for that government’s use of chemical weapons
against its own people. Do you favor or oppose U.S. military action against
Syria?
Number of Observations: 1000
Population: Likely Voters
Sampling Mode: Live Phone
Margin of Error: 3.0
16. CNN/ORC Poll, September 6-8, 2013
Publication Date: September 9, 2013
Question: After reports that the Syrian government has used chemical
weapons to kill civilians in that country, Congress is considering a resolution to
authorize limited U.S. military action in Syria. The resolution only authorizes
military action for 60 to 90 days and it bars the use of U.S. troops in a combat
role in Syria. Do you think Congress should or should not pass this resolution?
Number of Observations: 1022
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
292
17. Pew/USA Today Poll, September 4-8, 2013
Publication Date: September 9, 2013
Question: Would you favor or oppose the U.S. conducting military airstrikes
against Syria in response to reports that the Syrian government used chemical
weapons?
Number of Observations: 1506
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 2.9
18. ABC News/Washington Post Poll, September 4-8, 2013
Publication Date: September 10, 2013
Question: The United States says it has determined that the Syrian govern-
ment has used chemical weapons in the civil war there. Given this, do you
support or oppose the United States launching missile strikes against the Syr-
ian government?
Number of Observations: 1020
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
19. CBS/New York Times Poll, September 6-8, 2013
Publication Date: September 9, 2013
Question: In response to the Syrian government’s use of chemical weapons,
do you favor or oppose the United States launching military airstrikes against
Syrian military targets?
Number of Observations: 1011
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
20. NBC News/Wall Street Journal Poll, September 5-8, 2013
Publication Date: September 9, 2013
Question: It has been reported that the Syrian government has used chemical
weaponsonitscitizens. DoyouthinktheUnitedStatesshouldtakemilitaryac-
tion against the Syrian government in response to the use of chemical weapons
or not?
Number of Observations: 1000
Population: Adults
293
Sampling Mode: Live Phone
Margin of Error: 3.1
21. McClatchy-Marist Poll, September 7-8, 2013
Publication Date: September 9, 2013
Question: Do you favor or oppose limited U.S. air strikes on military targets
in Syria in response to the Syrian government’s use of chemical weapons?
Number of Observations: 963
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.2
22. Fox News Poll, September 6-8, 2013
Publication Date: September 9, 2013
Question: Do you favor or oppose using U.S. military force to punish Syria
for using chemical weapons?
Number of Observations: 900
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 3.0
23. Associated Press/GfK Poll, September 6-8, 2013
Publication Date: September 9, 2013
Question: As you may know, Barack Obama has asked Congress to vote to
support a U.S. military strike in Syria in response to a chemical weapons attack
there on August 21. Do you think Congress should vote in favor of a military
strike or should Congress vote against a military strike?
Number of Observations: 1007
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
24. YouGov/Economist Poll, September 7-9, 2013
Publication Date: September 11, 2013
Question: Do you support or oppose the use of military force against Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.6
294
25. Reuters/Ipsos Poll, September 5-9, 2013
Publication Date: September 9, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1450
Population: Adults
Sampling Mode: Internet
Margin of Error: 2.9
26. YouGov/Economist Poll, September 9-10, 2013
Publication Date: September 11, 2013
Question: Do you think the United States military should or should not use
air strikes to aid rebels in Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.0
27. Reuters/Ipsos Poll, September 6-10, 2013
Publication Date: September 12, 2013
Question: There have recently been reports of the Syrian government using
chemical weapons such as Sarin Gas against protestors. In your opinion, if
the Syrian government is using chemical weapons against Syrians, should the
United States intervene in Syria, or not?
Number of Observations: 1412
Population: Adults
Sampling Mode: Internet
Margin of Error: 3.0
28. Gallup Poll, September 11-12, 2013
Publication Date: September 13, 2013
Question: Would you favor or oppose the U.S. taking military action against
Syria in order to reduce that country’s ability to use chemical weapons?
Number of Observations: 1038
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
295
29. YouGov/Economist Poll, September 11-12, 2013
Publication Date: September 13, 2013
Question: Do you support or oppose the use of military force against Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.0
30. YouGov/Economist Poll, September 13-14, 2013
Publication Date: September 16, 2013
Question: Do you think the United States military should or should not use
air strikes to aid rebels in Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.0
31. Pew Poll, September 12-15, 2013
Publication Date: September 16, 2013
Question: If Syria does not give up control of its chemical weapons, would
you favor or oppose the U.S. conducting military airstrikes against Syria in
response to reports that the Syrian government used chemical weapons?
Number of Observations: 1002
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
32. ABC News/Washington Post Poll, September 12-15, 2013
Publication Date: September 17, 2013
Question: The United States says it has determined that the Syrian govern-
ment has used chemical weapons in the civil war there. Given this, do you
support or oppose the United States launching missile strikes against the Syr-
ian government?
Number of Observations: 1004
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
33. YouGov/Economist Poll, September 14-16, 2013
Publication Date: September 18, 2013
296
Question: Do you support or oppose the use of military force against Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.5
34. YouGov/Economist Poll, September 21-23, 2013
Publication Date: September 25, 2013
Question: Do you support or oppose the use of military force against Syria?
Number of Observations: 1000
Population: Adults
Sampling Mode: Internet
Margin of Error: 4.9
35. Gallup Poll, April 7-8, 2017
Publication Date: April 10, 2017
Question: Next, we have a question about the missile airstrikes the United
States has launched against Syria. Do you approve or disapprove of this U.S.
military action against Syria?
Number of Observations: 1015
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
36. ABC News/Washington Post Poll, April 7-9, 2017
Publication Date: April 12, 2017
Question: Do you support or oppose President Trump’s decision to launch
a missile strike on a Syrian air base in retaliation for the Syrian government
using chemical weapons against civilians?
Number of Observations: 900
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
37. YouGov/Economist Poll, April 10-11, 2017
Publication Date: April 12, 2017
Question: Do you support or oppose President Trump’s decision to order
military airstrikes against a Syrian air force base in response to a chemical
weapons attack against civilians?
Number of Observations: 1500
297
Population: Adults
Sampling Mode: Internet
Margin of Error: 3.2
38. Pew Poll, April 7-11, 2017
Publication Date: April 12, 2017
Question: Do you approve or disapprove of the U.S. conducting missile strikes
against Syria in response to reports that the Syrian government used chemical
weapons?
Number of Observations: 1062
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
39. YouGov/Economist Poll, April 15-18, 2017
Publication Date: April 18, 2017
Question: Do you support or oppose President Trump’s decision to order
military airstrikes against a Syrian air force base in response to a chemical
weapons attack against civilians?
Number of Observations: 1500
Population: Adults
Sampling Mode: Internet
Margin of Error: 3.1
40. NBC News/Wall Street Journal Poll, April 17-20, 2017
Publication Date: April 20, 2017
Question: Now, recently the U.S. launched cruise missiles from a U.S. naval
ship to destroy military units and infrastructure that were used to carry out a
chemical weapon attack by the government of Syria against their own people.
Do you support or oppose the decision to take this military action in Syria?
Number of Observations: 900
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.7
41. Fox News Poll, April 23-25, 2017
Publication Date: April 27, 2017
Question: Do you approve or disapprove of the United States using military
airstrikes to punish Syria for using chemical weapons?
Number of Observations: 1009
298
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 3.0
299
E.2 Polls about U.S. Airstrikes against Islamic
Militants
Appendix E.2 provides more information about all available public opinion polls con-
ducted by different polling firms and agencies with questions regarding U.S. military
airstrikes against the Islamic Militants of the Islamic State in Syria and Iraq. This
information is presented in chronological order, ordered by the final date of data
collection of each poll.
It is important to note that the question wording and the sampling methods
vary between polling agencies, which makes it difficult to compare attitudes to U.S.
airstrikes against ISIS between polls. As a result, this data is only presented for
illustrative purposes in this chapter. The data itself is available in a Stata.dta file,
which is available from the author upon request. This data will also be made publicly
available upon publication.
1. ABC News/Washington Post Poll, June 18-22, 2014
Publication Date: June 23, 2014
Question: Overall, do you support or oppose U.S. airstrikes against the Sunni
insurgents in Iraq?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
2. CBS News/New York Times Poll, June 20-22, 2014
Publication Date: June 23, 2014
Question: In response to the recent violence in Iraq, do you favor or oppose
theUnitedStatesusingMANNEDaircrafttocarryouttargetedattacksagainst
militants in Iraq?
Number of Observations: 1009
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
3. Quinnipac University Poll, June 24-30, 2014
Publication Date: July 1, 2014
Question: Do you think the U.S. should launch airstrikes against Islamic
militants in Iraq using piloted aircraft, or launch airstrikes that do not put
U.S. lives at risk such as drones or cruise missiles, or a combination of both
300
piloted aircraft and drones or cruise missiles, or should the U.S. not launch
airstrikes against Islamic militants in Iraq?
Number of Observations: 1446
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.6
4. ABC News/Washington Post Poll, August 13-17, 2014
Publication Date: August 18, 2014
Question: Overall, do you support or oppose U.S. airstrikes against the Sunni
insurgents in Iraq?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
5. Pew Research Center/USA Today Poll, August 14-17, 2014
Publication Date: August 18, 2014
Question: Do you approve or disapprove of the U.S. airstrikes against mili-
tants in Iraq in response to violence against civilians?
Number of Observations: 1000
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.6
6. ABC News/Washington Post Poll, September 4-7, 2014
Publication Date: September 8, 2014
Question: Overall, do you support or oppose U.S. airstrikes against the Sunni
insurgents in Iraq?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
7. CNN/ORC Poll, September 5-7, 2014
Publication Date: September 8, 2014
Question: Here are some actions which the United States and its allies might
take in response to the current situation regarding ISIS forces in Iraq and Syria.
Please tell me whether you favor or oppose each of the following: Military
airstrikes against ISIS forces in Syria?
301
Number of Observations: 1014
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
8. NBC News/Wall Street Journal Poll, September 3-7, 2014
Publication Date: September 8, 2014
Question: Now, more specifically, should U.S. military action against ISIS
be limited to only airstrikes, should it include both airstrikes and American
combat troops on the ground, or should U.S. military action not be taken at
all? [support for airstrikes]
Number of Observations: 1000
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 3.1
9. Pew Research Center Poll, September 11-14, 2014
Publication Date: September 15, 2014
Question: As you may know, Barack Obama has announced a plan for a
military campaign against Islamic militants in Iraq and Syria, involving U.S.
airstrikes and U.S. military training for opposition groups. Overall, do you
approve or disapprove of this plan?
Number of Observations: 1003
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.6
10. CBS News Poll, September 12-15, 2014
Publication Date: September 16, 2014
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
IRAQ?
Number of Observations: 1260
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
11. CBS News Poll, September 12-15, 2014
Publication Date: September 16, 2014
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
SYRIA?
302
Number of Observations: 1260
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
12. ABC News/Washington Post Poll, September 24-28, 2014
Publication Date: September 29, 2014
Question: Do you support or oppose the U.S.-led airstrikes that are under
way against Islamic State insurgents and related targets in Syria?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
13. CNN/ORC Poll, September 25-28, 2014
Publication Date: September 29, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL or as the Islamic State. As you may know,
the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria, but
there are no U.S. combat troops fighting ISIS forces in either country.
"Based on what you have read or heard, do you favor or oppose the military
airstrikes conducted by the United States and its Western European and Arab
allies against the ISIS forces in Iraq and Syria?
Number of Observations: 1045
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
14. CBS News Poll, October 3-6, 2014
Publication Date: October 7, 2014
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
IRAQ?
Number of Observations: 1260
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
15. CBS News Poll, October 3-6, 2014
Publication Date: October 7, 2014
303
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
SYRIA?
Number of Observations: 1260
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
16. NBC News/Wall Street Journal Poll, October 8-12, 2014
Publication Date: October 13, 2014
Question: Now, more specifically, should U.S. military action against ISIS
be limited to only airstrikes, should it include both airstrikes and American
combat troops on the ground, or should U.S. military action not be taken at
all? [support for airstrikes]
Number of Observations: 1000
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 3.1
17. CNN/ORC Poll, October 24-26, 2014
Publication Date: October 27, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL or as the Islamic State. As you may know,
the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria, but
there are no U.S. combat troops fighting ISIS forces in either country.
"Based on what you have read or heard, do you favor or oppose the military
airstrikes conducted by the United States and its Western European and Arab
allies against the ISIS forces in Iraq and Syria?
Number of Observations: 1045
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
18. CBS News Poll, October 23-27, 2014
Publication Date: October 28, 2014
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
Iraq and Syria?
Number of Observations: 1205
Population: Adults
304
Sampling Mode: Live Phone
Margin of Error: 3.0
19. CNN/ORC Poll, November 21-23, 2014
Publication Date: November 25, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL or as the Islamic State. As you may know,
the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria, but
there are no U.S. combat troops fighting ISIS forces in either country.
"Based on what you have read or heard, do you favor or oppose the military
airstrikes conducted by the United States and its Western European and Arab
allies against the ISIS forces in Iraq and Syria?
Number of Observations: 1045
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
20. CBS News Poll, February 13-17, 2015
Publication Date: February 18, 2015
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
Iraq and Syria?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
21. ABC News/Washington Post Poll, November 16-19, 2015
Publication Date: November 20, 2015
Question: Do you support or oppose increased U.S. airstrikes against the
Islamic State in Iraq and Syria?
Number of Observations: 1004
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
22. CBS News Poll, November 19-22, 2015
Publication Date: November 23, 2015
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
305
Iraq and Syria?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
306
E.3 Polls about U.S. Troops to Fight Islamic Mil-
itants
Appendix E.3 provides more information about all available public opinion polls con-
ducted by different polling firms and agencies with questions regarding sending U.S.
military personnel to fight against the Islamic Militants of the Islamic State in Syria
and Iraq. This was colloquially known throughout the foreign policy debate as ques-
tions about “boots on the ground.” This information is presented in chronological
order, ordered by the final date of data collection of each poll.
It is important to note that the question wording and the sampling methods
vary between polling agencies, which makes it difficult to compare attitudes to the
provision of U.S. troops to fight against ISIS between polls. As a result, this data is
only presented for illustrative purposes in this chapter. The data itself is available
in a Stata.dta file, which is available from the author upon request. This data will
also be made publicly available upon publication.
1. Gallup Poll, June 20-21, 2014
Publication Date: June 22, 2014
Question: Do you favor or oppose the United States taking direct military
action in Iraq to assist the Iraqi government in fighting militants there?
Number of Observations: 1012
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
2. ABC News/Washington Post Poll, June 18-22, 2014
Publication Date: June 23, 2014
Question: Overall, do you support or oppose sending U.S. ground forces to
combat the Sunni insurgents in Iraq?
Number of Observations: 1009
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
3. CBS News/New York Times Poll, June 20-22, 2014
Publication Date: June 23, 2014
Question: In response to the recent violence in Iraq, do you favor or oppose
the United States sending ground troops into Iraq?
Number of Observations: 1009
307
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
4. Quinnipac University Poll, June 24-30, 2014
Publication Date: July 1, 2014
Question: Do you support or oppose the U.S. sending ground troops back
into Iraq to help the Iraqi government defeat Islamic militants?
Number of Observations: 1446
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.6
5. CNN/ORC Poll, September 5-7, 2014
Publication Date: September 8, 2014
Question: HHerearesomeactionswhichtheUnitedStatesanditsalliesmight
take in response to the current situation regarding ISIS forces in Iraq and Syria.
Please tell me whether you favor or oppose each of the following: Sending U.S.
ground troops to Iraq and Syria?
Number of Observations: 1014
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
6. CBS News Poll, September 12-15, 2014
Publication Date: September 16, 2014
Question: Do you favor or oppose U.S. airstrikes against ISIS militants in
IRAQ?
Number of Observations: 1260
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
7. CBS News Poll, September 12-15, 2014
Publication Date: September 16, 2014
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1260
Population: Adults
308
Sampling Mode: Live Phone
Margin of Error: 3.0
8. Gallup Poll, September 20-21, 2014
Publication Date: September 22, 2014
Question: WouldyoufavororopposetheUnitedStatessendinggroundtroops
to Iraq and Syria in order to assist groups in those countries that are fighting
the Islamic militants?
Number of Observations: 1013
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
9. NBC News/Wall Street Journal Poll, September 19-25, 2014
Publication Date: September 26, 2014
Question: Ifitweredeterminedbythemilitarycommandersthatthebestway
to defeat the ISIS army was to use American military troops on the ground,
would you be in favor of this decision or would you be in opposition to this
decision, or do you really not have an opinion one way or the other?
Number of Observations: 1283
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 3.2
10. ABC News/Washington Post Poll, September 24-28, 2014
Publication Date: September 29, 2014
Question: Do you support or oppose sending U.S. forces to Iraq to train Iraqi
government troops and coordinate airstrikes against Islamic State fighters?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
11. CNN/ORC Poll, September 25-28, 2014
Publication Date: September 29, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
309
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
12. CBS News Poll, October 3-6, 2014
Publication Date: October 7, 2014
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
13. Pew Research Center Poll, October 15-20, 2014
Publication Date: October 21, 2014
Question: Would you favor or oppose the U.S. sending ground troops to fight
Islamic militants in Iraq and Syria?
Number of Observations: 1504
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 2.9
14. CNN/ORC Poll, October 24-26, 2014
Publication Date: October 27, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
310
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
15. CBS News Poll, October 23-27, 2014
Publication Date: October 28, 2014
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
16. Quinnipac University Poll, November 18-23, 2014
Publication Date: November 24, 2014
Question: Do you think the United States military should have combat troops
on the ground in Iraq or not?
Number of Observations: 1623
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.4
17. CNN/ORC Poll, November 21-23, 2014
Publication Date: November 25, 2014
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
18. NBC News/Wall Street Journal Poll, February 11-12, 2015
Publication Date: February 13, 2015
311
Question: When it comes to combating ISIS, the Islamic State group, do
you support the United States sending a large number of U.S. ground forces,
sending a limited number of U.S. ground forces, or not sending U.S. ground
forces at all?
Number of Observations: 603
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.0
19. CNN/ORC Poll, February 12-15, 2015
Publication Date: February 16, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
20. CBS News Poll, February 13-17, 2015
Publication Date: February 18, 2015
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
21. Pew Research Center Poll, February 18-22, 2015
Publication Date: February 23, 2015
Question: Would you favor or oppose the U.S. sending ground troops to fight
Islamic militants in Iraq and Syria?
Number of Observations: 1504
312
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 2.9
22. McClatchy-Marist Poll, March 1-4, 2015
Publication Date: March 5, 2015
Question: When it comes to combating ISIS, the Islamic State group, do
you support the United States sending a large number of U.S. ground forces,
sending a limited number of U.S. ground forces, or not sending U.S. ground
forces at all?
Number of Observations: 1253
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 2.8
23. CBS News Poll, March 21, 2015
Publication Date: March 22, 2015
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
24. CNN/ORC Poll, May 29-31, 2015
Publication Date: June 1, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
313
25. NBC News/Wall Street Journal Poll, May 14-18, 2015
Publication Date: May 19, 2015
Question: Now, as you may know, a terrorist group called ISIS that is op-
erating in Syria and Iraq has declared an independent Islamic state. When it
comes to the United States re-committing our military forces to Iraq to fight
against ISIS, which of the following best describes your point of view? You
favor sending troops in for as long as it takes to get the job done. You favor
sending in troops for a short period to allow the Iraqi Army to take over. OR,
You do not favor sending any troops in at all.
Number of Observations: 500
Population: Adults
Sampling Mode: Live Phone
Margin of Error: N/A
26. CBS News Poll, July 29-August 2, 2015
Publication Date: August 3, 2015
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
27. CNN/ORC Poll, August 13-16, 2015
Publication Date: August 17, 2015
Question: Do you favor or oppose the United States sending ground troops
into combat operations against ISIS forces in Iraq or Syria?
Number of Observations: 1001
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
28. CNN/ORC Poll, August 13-16, 2015
Publication Date: August 17, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
314
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
29. Quinnipac University Poll, August 20-25, 2015
Publication Date: August 26, 2015
Question: Would you support or oppose the U.S. sending ground troops to
fight ISIS in Iraq and Syria?
Number of Observations: 1140
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.9
30. CNN/ORC Poll, October 14-17, 2015
Publication Date: October 18, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
31. Bloomberg Politics Poll, November 16-17, 2015
Publication Date: November 18, 2015
Question: Turning now to the conflict involving ISIS, the Islamic State. Do
you think the U.S. should or should not send American troops to Iraq or Syria
to fight the Islamic State?
315
Number of Observations: 628
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.9
32. ABC News/Washington Post Poll, November 16-19, 2015
Publication Date: November 20, 2015
Question: Do you support or oppose the increased use of U.S. ground forces
against the Islamic State in Iraq and Syria?
Number of Observations: 1004
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.5
33. CBS News Poll, November 19-22, 2015
Publication Date: November 23, 2015
Question: Do you favor or oppose the U.S. sending ground troops into Iraq
or Syria to fight ISIS militants?
Number of Observations: 1205
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 3.0
34. Quinnipac University Poll, November 23-30, 2015
Publication Date: December 1, 2015
Question: Would you support or oppose the U.S. sending ground troops to
fight ISIS in Iraq and Syria?
Number of Observations: 1140
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.9
35. CNN/ORC Poll, November 27-December 1, 2015
Publication Date: December 2, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
316
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
36. Quinnipac University Poll, December 16-20, 2015
Publication Date: December 21, 2015
Question: Would you support or oppose the U.S. sending ground troops to
fight ISIS in Iraq and Syria?
Number of Observations: 1140
Population: Registered Voters
Sampling Mode: Live Phone
Margin of Error: 2.9
37. CNN/ORC Poll, December 17-21, 2015
Publication Date: December 22, 2015
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
38. CNN/ORC Poll, April 28-May 1, 2016
Publication Date: May 2, 2016
Question: Now here are a few questions about the Islamic militant group
often referred to as ISIS that controls some areas of Iraq and Syria. The group
is also sometimes referred to as ISIL, or as the Islamic State. As you may
317
know, the U.S. has conducted airstrikes against ISIS forces in Iraq and in Syria
and has special forces troops on the ground in both countries to advise and
assist in fighting ISIS, but there are no U.S. combat troops fighting ISIS forces
in either country.
Do you favor or oppose the United States sending ground troops into combat
operations against ISIS forces in Iraq or Syria?
Number of Observations: 496
Population: Adults
Sampling Mode: Live Phone
Margin of Error: 4.5
318
Bibliography
Abelson, Robert P, Elliot Ed Aronson, William J McGuire, Theodore M Newcomb,
Milton J Rosenberg, and Percy H Tannenbaum. 1968. “Theories of cognitive con-
sistency: A sourcebook.”.
Abramowitz, Alan. 2010. The Disappearing Center: Engaged Citizens, Polarization,
and American Democracy. Yale University Press.
Adamic, Lada A., and Natalie Glance. 2005. “The Political Blogosphere and the
2004 U.S. Election: Divided They Blog.” In Proceedings of the 3rd International
Workshop on Link Discovery. LinkKDD ’05, New York, NY, USA pp. 36–43.
Aday, Sean, Steven Livingston, and Maeve Hebert. 2005. “Embedding the truth:
A cross-cultural analysis of objectivity and television coverage of the Iraq war.”
Harvard International Journal of Press/Politics 10 (1): 3–21.
Akerlof, George A, and William T Dickens. 1982. “The economic consequences of
cognitive dissonance.” The American economic review 72 (3): 307–319.
Al Jazeera, and Agencies. 2014. “Sunni rebels declare new ’Islamic caliphate’.”
Albertson, Bethany, and Shana Kushner Gadarian. 2015. Anxious Politics:
Democratic Citizenship in a Threatening World. New York: Cambridge University
Press.
Allcott, Hunt, and Matthew Gentzkow. 2017. Social Media and Fake News in the
2016 Election. Working Paper 23089 National Bureau of Economic Research. DOI:
10.3386/w23089.
Almond, Gabriel Abraham. 1960. The American people and foreign policy. Praeger.
Althaus, Scott L. 2002. “American news consumption during times of national crisis.”
PS: Political Science & Politics 35 (3): 517–521.
Anderson, Michael C, and Ean Huddleston. 2012. “Towards a cognitive and neuro-
biological model of motivated forgetting.” In True and false recovered memories.
Springer.
Bahador, B. 2007. The CNN Effect in Action: How the News Media Pushed the
West Toward War in Kosovo. New York: Palgrave Macmillan.
319
Balzacq, Thierry, ed. 2010. Securitization Theory: How Security Problems Emerge
and Dissolve. Milton Park, Abingdon, Oxon ; New York: Routledge.
Barabas, Jason, Jennifer Jerit, William Pollock, and Carlisle Rainey. 2014. “The
question (s) of political knowledge.” American Political Science Review 108 (4):
840–855.
Barberá, Pablo. 2015. “Birds of the Same Feather Tweet Together: Bayesian Ideal
Point Estimation Using Twitter Data.” Political Analysis 23 (January): 76–91.
Barberá, Pablo, John T. Jost, Jonathan Nagler, Joshua A. Tucker, and Richard
Bonneau. 2015. “Tweeting From Left to Right: Is Online Political Communication
More Than an Echo Chamber?” Psychological Science 26 (October): 1531–1542.
Barberá, Pablo, Ning Wang, Richard Bonneau, John T. Jost, Jonathan Nagler,
Joshua Tucker, and Sandra González-Bailón. 2015. “The Critical Periphery in the
Growth of Social Protests.” PLOS ONE 10 (November): e0143611.
Barberá, Pablo, and Thomas Zeitzoff. N.d. “The New Public Address System: Why
Do World Leaders Adopt Social Media?” International Studies Quarterly. Forth-
coming.
Bargh, J. A., M. Chen, and L. Burrows. 1996. “Automaticity of social behavior:
direct effects of trait construct and stereotype-activation on action.” Journal of
Personality and Social Psychology 71 (August): 230–244.
Bartels, Larry M. 2002. “Beyond the Running Tally: Partisan Bias in Political Per-
ceptions.” Political Behavior 24 (June): 117–150.
Baum, Matthew A. 2002a. “The Constituent Foundations of the Rally-Round-the-
Flag Phenomenon.” International Studies Quarterly 46 (June): 263–298.
Baum, Matthew A. 2002b. “Sex, Lies, and War: How Soft News Brings Foreign
Policy to the Inattentive Public.” American Political Science Review 96 (March):
91–109.
Baum, Matthew A. 2005. Soft News Goes to War: Public Opinion and American
Foreign Policy in the New Media Age. Princeton University Press.
Baum, Matthew A, and Angela S Jamison. 2006. “The Oprah effect: How soft news
helps inattentive citizens vote consistently.” The Journal of Politics 68 (4): 946–
959.
Baum, Matthew A., and Philip B. K. Potter. 2008. “The Relationships Between
MassMedia, PublicOpinion, andForeignPolicy: TowardaTheoreticalSynthesis.”
Annual Review of Political Science 11 (1): 39–65.
Baum, Matthew A., and Philip B. K. Potter. 2015. War and Democratic Constraint:
How the Public Influences Foreign Policy. Princeton, New Jersey: Princeton Uni-
versity Press.
320
Baum, Matthew A, and Tim J Groeling. 2010. War stories: The causes and
consequences of public views of war. Princeton University Press.
Bénabou, Roland, and Jean Tirole. 2002. “Self-confidence and personal motivation.”
The Quarterly Journal of Economics 117 (3): 871–915.
Bénabou, Roland, and Jean Tirole. 2011. “Identity, morals, and taboos: Beliefs as
assets.” The Quarterly Journal of Economics 126 (2): 805–855.
Bennett, W Lance. 1990. “Toward a theory of press-state relations in the United
States.” Journal of communication 40 (2): 103–127.
Bennett, W. Lance and David L. Paletz, eds. 1994. Taken by Storm: The Media,
Public Opinion, and U.S. Foreign Policy in the Gulf War. Chicago: University Of
Chicago Press.
Bennett, W. Lance, Regina G. Lawrence, and Steven Livingston. 2007. When the
Press Fails: Political Power and the News Media from Iraq to Katrina. University
Of Chicago Press.
Benoît, Jean-Pierre, and Juan Dubra. 2011. “Apparent overconfidence.”
Econometrica 79 (5): 1591–1625.
Berinsky, Adam J. 2007. “Assuming the Costs of War: Events, Elites, and American
Public Support for Military Conflict.” Journal of Politics 69 (November): 975–997.
Berinsky, Adam J. 2009. In Time of War: Understanding American Public Opinion
from World War II to Iraq. Chicago ; London: University Of Chicago Press.
Berinsky, Adam J., Gregory A. Huber, and Gabriel S. Lenz. 2012. “Evaluating On-
line Labor Markets for Experimental Research: Amazon.com’s Mechanical Turk.”
Political Analysis 20 (July): 351–368.
Biernatzki, William E. 2003. “War and media.” Communication Research Trends
22 (3).
Blais, André. 2000. To Vote Or Not to Vote?: The Merits and Limits of Rational
Choice Theory. University of Pittsburgh Pre.
Bloch, Yaeli, and Sam Lehman-Wilzig. 2002. “An exploratory model of media-
government relations in international crises: US involvement in Bosnia 1992-1995.”
Media and Conflict: Framing Issues Making Policy Shaping Opinions: 153–173.
Bolsen, Toby, and Thomas J. Leeper. 2013. “Self-Interest and Attention to News
Among Issue Publics.” Political Communication 30 (July): 329–348.
Brady, Henry E., and Paul M. Sniderman. 1985. “Attitude Attribution: A Group
Basis for Political Reasoning.” American Political Science Review 79 (December):
1061–1078.
Brecher, Michael, and Jonathan Wilkenfeld. 1997. “A Study of Crisis.”.
Brecher, Michael, Jonathan Wilkenfeld, Kyle Beardsley, Patrick James, and
David M. Quinn. 2017. International Crisis Behavior Data Codebook, Version
321
12.
Broadbent, D. E. 2013. Perception and Communication. Elsevier.
Brock, Timothy C, and Joe L Balloun. 1967. “Behavioral receptivity to dissonant
information.” Journal of personality and social psychology 6 (4p1): 413.
Brockus, Susan. 2009. “Coming to You ?Live? Exclusive Witnessing and the Battle-
field Reporter.” Journal of Communication Inquiry 33 (1): 27–42.
Brody, Richard. 1991. Assessing the president: The media, elite opinion, and public
support. Palo Alto, CA: Stanford University Press.
Brody, Richard A. 1992. Assessing the President: The Media, Elite Opinion, and
Public Support. Stanford, Calif: Stanford University Press.
Brunnermeier, Markus K, and Jonathan A Parker. 2005. “Optimal expectations.”
The American Economic Review 95 (4): 1092–1118.
Buzan, Barry, Ole Waever, and Jaap de Wilde. 1997. Security: A New Framework
for Analysis. Boulder, Colo: Lynne Rienner Pub.
Büthe, Tim, Solomon Major, and André de Mello e Souza. 2012. “The Politics
of Private Foreign Aid: Humanitarian Principles, Economic Development Objec-
tives, and Organizational Interests in NGO Private Aid Allocation.” International
Organization 66 (October): 571–607.
Caplin, Andrew, and John Leahy. 2001. “Psychological expected utility theory and
anticipatory feelings.” The Quarterly Journal of Economics 116 (1): 55–79.
Casler, Krista, Lydia Bickel, and Elizabeth Hackett. 2013. “Separate but equal? A
comparison of participants and data gathered via Amazon’s MTurk, social media,
and face-to-face behavioral testing.” Computers in Human Behavior 29 (Novem-
ber): 2156–2160.
Chater, Nick, andGeorgeLoewenstein.2016.“Theunder-appreciateddriveforsense-
making.” Journal of Economic Behavior & Organization 126: 137–154.
Choi, Seung-Whan, and Patrick James. 2007. “Media Openness, Democracy and
Militarized Interstate Disputes.” British Journal of Political Science 37 (1): 23?46.
Chong, Dennis, and James N. Druckman. 2007a. “Framing Public Opinion in Com-
petitive Democracies.” The American Political Science Review 101 (4): 637–655.
Chong, Dennis, and James N. Druckman. 2007b. “Framing Theory.” Annual Review
of Political Science 10 (1): 103–126.
Chong, Dennis, and James N. Druckman. 2007c. “A Theory of Framing and Opin-
ion Formation in Competitive Elite Environments.” Journal of Communication 57
(March): 99–118.
Cialdini, Robert B., and and Noah J. Goldstein. 2004. “Social Influence: Compliance
and Conformity.” Annual Review of Psychology 55 (1): 591–621.
322
Clifford, Scott. 2017. “Individual Differences in Group Loyalty Predict Partisan
Strength.” Political Behavior 39 (September): 531–552.
Cohen, Bernard C. 1963. The Press and Foreign Policy. Institute of Governmental
Studies, University of California.
Colleoni, Elanor, Alessandro Rozza, and Adam Arvidsson. 2014. “Echo Chamber
or Public Sphere? Predicting Political Orientation and Measuring Political Ho-
mophily in Twitter Using Big Data.” Journal of Communication 64 (April): 317–
332.
Conover, Michael D., Bruno Gonçalves, Alessandro Flammini, and Filippo Menczer.
2012. “Partisan asymmetries in online political activity.” EPJ Data Science 1 (De-
cember): 6.
Cortell, Andrew P, Robert M Eisinger, and Scott L Althaus. 2009. “Why Embed?
Explaining the Bush Administration’s Decision to Embed Reporters in the 2003
Invasion of Iraq.” American Behavioral Scientist 52 (5): 657–677.
Davis, Aeron. 2007. The mediation of power: a critical introduction. Communication
and society London ; New York: Routledge.
Davis, Aeron. 2010. Political communication and social theory. Communication and
society London ; New York: Routledge.
Defence Manpower Data Center. 2017. “DoD Personnel, Workforce Reports and
Publications.”
Delli Carpini, Michael X., and Scott Keeter. 1996. What Americans Know about
Politics and why it Matters. Yale University Press.
Der Derian, James. 2008. “The desert of the real and the simulacrum of war.”
International Affairs 84 (5): 931–948.
Diffley, Mark, Sara Davidson, Sanah Zubairi, Colin Hockaday, and Russell Gunson.
2016. “The ‘Tinman? referendum: the EU debate in Scotland is lacking heart.”
Djerf-Pierre, Monika, and Adam Shehata. N.d. “Still an Agenda Setter: Traditional
News Media and Public Opinion During the Transition From Low to High Choice
Media Environments.” Journal of Communication. Forthcoming.
Doherty, Carroll, Jocelyn Kiley, and Alec Tyson. 2015. Beyond Distrust: How
Americans View Their Government: Broad Criticism, But Positive Performance
Ratings in Many Areas. Pew Research Center.
Downs, Anthony. 1957. An Economic Theory of Democracy. Boston: Harper and
Row.
Dropp, Kyle, Joshua D. Kertzer, and Thomas Zeitzoff. 2014. “The less Americans
know about Ukraine’s location, the more they want U.S. to intervene.” Washington
Post (April).
323
Druckman, James N., Erik Peterson, and Rune Slothuus. 2013. “How Elite Partisan
PolarizationAffectsPublicOpinionFormation.”AmericanPoliticalScienceReview
107 (February): 57–79.
Druckman, James N., Jordan Fein, and Thomas J. Leeper. 2012. “A Source of Bias in
Public Opinion Stability.” American Political Science Review 106 (May): 430–454.
Druckman, James N., and Lawrence R. Jacobs. 2015. Who Governs?: Presidents,
Public Opinion, and Manipulation. Chicago, IL: University of Chicago Press.
Drury, A Cooper, Richard Stuart Olson, and Douglas A Van Belle. 2005. “The
politics of humanitarian aid: US foreign disaster assistance, 1964–1995.” Journal
of Politics 67 (2): 454–473.
Eckles, Dean, and Eytan Bakshy. 2017. “Bias and high-dimensional adjustment in
observational studies of peer effects.” arXiv:1706.04692 [cs, stat] (June). arXiv:
1706.04692.
Eichenberg, Richard C. 2005. “Victory Has Many Friends: U.S. Public Opinion and
the Use of Military Force, 1981–2005.” International Security 30 (July): 140–177.
Eil, David, and Justin M Rao. 2011. “The good news-bad news effect: asymmetric
processing of objective information about yourself.” American Economic Journal:
Microeconomics 3 (2): 114–138.
Eliaz, Kfir, andAndrewSchotter.2007.“Experimentaltestingofintrinsicpreferences
for nonlnstrumental information.” The American economic review 97 (2): 166–169.
Eliaz, Kfir, and Andrew Schotter. 2010. “Paying for confidence: An experimental
study of the demand for non-instrumental information.” Games and Economic
Behavior 70 (2): 304–324.
Eshbaugh-Soha, Matthew, and Christopher Linebarger. 2014. “Presidential and Me-
dia Leadership of Public Opinion on Iraq.” Foreign Policy Analysis 10 (October):
351–369.
Evans, G., E. Fieldhouse, J. Green, H. Schmitt, C. van der Eijk, J. Mellon, and C.
Prosser. 2016. The 2016 EU Referendum campaign wave of the 2014-2018 British
ElectionStudyInternetPanel.Version1.6.TechnicalreportBritishElectionStudy.
DOI: 10.15127/1.293723.
Falk, Armin, and Florian Zimmermann. 2016. “Beliefs and utility: Experimental
evidence on preferences for information.”.
Farrell, Henry. 2012. “The Consequences of the Internet for Politics.” Annual Review
of Political Science 15 (1): 35–52.
Feldman, Stanley, Leonie Huddy, and George E. Marcus. 2015. Going to War in Iraq:
When Citizens and the Press Matter. University of Chicago Press.
Festinger, Leon. 1962. A theory of cognitive dissonance. Vol. 2 Stanford university
press.
324
Fiorina, Morris P., and Samuel J. Abrams. 2008. “Political Polarization in the Amer-
ican Public.” Annual Review of Political Science 11 (1): 563–588.
Forgas, J. P. 1995. “Mood and judgment: the affect infusion model (AIM).”
Psychological Bulletin 117 (January): 39–66.
Fox, Julia R, and Byungho Park. 2006. “The" I" of embedded reporting: An analysis
of CNN coverage of the" Shock and Awe" campaign.” Journal of Broadcasting &
Electronic Media 50 (1): 36–51.
Frey, Dieter. 1986. “Recent research on selective exposure to information.” In
Advances in experimental social psychology. Vol. 19. Elsevier.
Frey, Dieter, StefanSchulz-Hardt, andDagmarStahlberg.2013.“Informationseeking
among individuals and groups and possible consequences for decision-making in
business and politics.” Understanding group behavior 2: 211–225.
Gadarian, Shana Kushner. 2010. “The politics of threat: How terrorism news shapes
foreign policy attitudes.” The Journal of Politics 72 (2): 469–483.
Gaines, Brian J., James H. Kuklinski, Paul J. Quirk, Buddy Peyton, and Jay
Verkuilen. 2007. “Same Facts, Different Interpretations: Partisan Motivation and
Opinion on Iraq.” The Journal of Politics 69 (November): 957–974.
Galtung, Johan, and Mari Holmboe Ruge. 1965. “The Structure of Foreign News The
Presentation of the Congo, Cuba and Cyprus Crises in Four Norwegian Newspa-
pers.” Journal of Peace Research 2 (March): 64–90.
Ganguly, Ananda, and Joshua Tasoff. N.d. “Fantasy and Dread: The Demand for
Information and the Consumption Utility of the Future.” Management Science.
Forthcoming.
Gartner, Scott Sigmund, and Gary M. Segura. 1998. “War, Casualties, and Public
Opinion.” Journal of Conflict Resolution 42 (June): 278–300.
Gartner, Scott Sigmund, and Gary M. Segura. 2000. “Race, Casualties, and Opinion
in the Vietnam War.” Journal of Politics 62 (February): 115–146.
Gelpi, Christopher, Peter D. Feaver, and Jason Reifler. 2006. “Success Matters: Ca-
sualty Sensitivity and the War in Iraq.” International Security 30 (January): 7–46.
Gelpi, Christopher, Peter D. Feaver, and Jason Reifler. 2009. Paying the human costs
of war: American public opinion and casualties in military conflicts. Princeton
University Press.
Gigerenzer,Gerd, PeterM.Todd, theABCResearchGroup,andothers.1999.Simple
heuristics that make us smart. Oxford University Press.
Gilboa, Eytan. 2005a. “The CNN effect: The search for a communication theory of
international relations.” Political communication 22 (1): 27–44.
Gilboa, Eytan. 2005b. “Global television news and foreign policy: Debating the CNN
effect.” International Studies Perspectives 6 (3): 325–341.
325
Gilens, Martin. 2001. “Political ignorance and collective policy preferences.”
American Political Science Review 95 (2): 379–396.
Glaeser, Edward L, and Cass R Sunstein. 2013. Why does balanced news produce
unbalanced views? Technical report National Bureau of Economic Research.
Golder, Scott A., and Michael W. Macy. 2014. “Digital Footprints: Opportunities
and Challenges for Online Social Research.” Annual Review of Sociology 40 (1):
129–152.
Golman, Russell, David Hagmann, and George Loewenstein. 2017. “Information
avoidance.” Journal of Economic Literature 55 (1): 96–135.
Golman, Russell, and George Loewenstein. 2015. The Demand For, and Avoidance
of, Information. Technical report working paper.
Golman, Russell, George Loewenstein, Karl Ove Moene, and Luca Zarri. 2016. “The
preference for belief consonance.” The Journal of Economic Perspectives 30 (3):
165–187.
Government of the United Kingdom. 2016. Why the government believes that voting
to remain in the EU is the best decision for the UK - with references - GOV.UK.
Green, Donald P., and Alan S. Gerber. 2008. Get Out the Vote: How to Increase
Voter Turnout. 2nd ed. Washington, D.C: Brookings Institution Press.
Green, Donald P., and Alan S. Gerber. 2015. Get Out the Vote: How to Increase
Voter Turnout. Brookings Institution Press.
Gries, Peter. 2014. The Politics of American Foreign Policy: How Ideology Divides
Liberals and Conservatives over Foreign Affairs. Stanford, California: Stanford
University Press.
Haigh, Michel M, Michael Pfau, Jamie Danesi, Robert Tallmon, Tracy Bunko, Shan-
non Nyberg, Bertha Thompson, Chance Babin, Sal Cardella, Michael Mink et al.
2006. “A comparison of embedded and nonembedded print coverage of the US
invasion and occupation of Iraq.” Harvard International Journal of Press/Politics
11 (2): 139–153.
Hallin, Daniel C. 1989. The Uncensored War: The Media and Vietnam. Berkeley:
University of California Press.
Hamilton, James T. 2006. All the News That’s Fit to Sell: How the Market
Transforms Information into News. Princeton, NJ: Princeton University Press.
Harcup, Tony, and Deirdre O’Neill. 2001. “What Is News? Galtung and Ruge revis-
ited.” Journalism Studies 2 (January): 261–280.
Hatemi, Peter K, and Rose McDermott. 2016. “Give me attitudes.” Annual Review
of Political Science 19: 331–350.
Hauser, David J., and Norbert Schwarz. 2016. “Attentive Turkers: MTurk partici-
pants perform better on online attention checks than do subject pool participants.”
326
Behavior Research Methods 48 (March): 400–407.
Hayes, Jarrod. 2015. Constructing National Security: U.S. Relations with India and
China. Cambridge University Press.
Heffington, Colton, Brandon Beomseob Park, and Laron K Williams. 2017. “The
“Most Important Problem” Dataset (MIPD): a new dataset on American issue im-
portance.” Conflict Management and Peace Science (March): 0738894217691463.
Hicks, Raymond, and Dustin Tingley. 2011. “Causal mediation analysis.” Stata
Journal 11 (4): 605–619.
Hill, SethJ.2017.“Learningtogetherslowly: Bayesianlearningaboutpoliticalfacts.”
Journal of Politics 79 (4).
Holsti, Ole R. 1992. “Public Opinion and Foreign Policy: Challenges to the Almond-
Lippmann Consensus.” International Studies Quarterly 36 (December): 439–466.
Holsti, Ole R., and James N. Rosenau. 1984. American Leadership in World Affairs:
Vietnam and the Breakdown of Consensus. Allen & Unwin.
Hsu,JoanneW.,MaximilianD.Schmeiser,CatherineHaggerty,andShannonNelson.
2017. “The Effect of Large Monetary Incentives on Survey CompletionEvidence
from a Randomized Experiment with the Survey of Consumer Finances.” Public
Opinion Quarterly 81 (September): 736–747.
Huddy, Leonie, and Stanley Feldman. 2011. “Americans Respond Politically to
9/11: Understanding the Impact of the Terrorist Attacks and Their Aftermath.”
American Psychologist 66 (September): 455–467.
Huddy, Leonie, Stanley Feldman, Charles Taber, and Gallya Lahav. 2005. “Threat,
Anxiety, and Support of Antiterrorism Policies.” American Journal of Political
Science 49 (July): 593–608.
Huddy, Leonie, Stanley Feldman, and Christopher Weber. 2007. “The Political Con-
sequences of Perceived Threat and Felt Insecurity.” The ANNALS of the American
Academy of Political and Social Science 614 (November): 131–153.
Huddy, Leonie, Stanley Feldman, Theresa Capelos, and Colin Provost. 2002. “The
Consequences of Terrorism: Disentangling the Effects of Personal and National
Threat.” Political Psychology 23 (September): 485–509.
Huff,Connor,andDustinTingley.2015.““Whoarethesepeople?”Evaluatingthede-
mographic characteristics and political preferences of MTurk survey respondents.”
Research & Politics 2 (August).
Hurwitz, Jon, andMarkPeffley.1987.“HowareForeignPolicyAttitudesStructured?
A Hierarchical Model.” American Political Science Review 81 (December): 1099–
1120.
Imai, Kosuke, Booil Jo, and Elizabeth A. Stuart. 2011. “Commentary: Using Poten-
tial Outcomes to Understand Causal Mediation Analysis.” Multivariate Behavioral
327
Research 46 (September): 861–873.
Imai, Kosuke, Dustin Tingley, and Teppei Yamamoto. 2013. “Experimental designs
for identifying causal mechanisms.” Journal of the Royal Statistical Society: Series
A (Statistics in Society) 176 (January): 5–51.
Imai, Kosuke, Luke Keele, Dustin Tingley, and Teppei Yamamoto. 2011. “Unpacking
the Black Box of Causality: Learning about Causal Mechanisms from Experimen-
tal and Observational Studies.” American Political Science Review 105 (Novem-
ber): 765–789.
Imai, Kosuke, Luke Keele, and Teppei Yamamoto. 2010. “Identification, Inference
and Sensitivity Analysis for Causal Mediation Effects.” Statistical Science 25
(February): 51–71.
Iyengar, Shanto, andAdamSimon.1993.“NewsCoverageoftheGulfCrisisandPub-
lic Opinion: A Study of Agenda-Setting, Priming, and Framing.” Communication
Research 20 (June): 365–383.
Iyengar, Shanto, and Donald R. Kinder. 2010. News That Matters: Television and
American Opinion. Updated ed. Chicago: University Of Chicago Press.
Jacobs, Lawrence R., and Robert Y. Shapiro. 2000. Politicians Don’t Pander:
Political Manipulation and the Loss of Democratic Responsiveness. Chicago, IL:
University of Chicago Press.
James, Patrick, and Jean Sebastien Rioux. 1998. “International Crises and Linkage
Politics: The Experiences of the United States, 1953-1994.” Political Research
Quarterly 51 (September): 781–812.
Janis, Irving L. 1972. “Victims of groupthink: a psychological study of foreign-policy
decisions and fiascoes.”.
Janis, Irving Lester. 1982. Groupthink: Psychological studies of policy decisions and
fiasces. Houghton Mifflin.
Jann, Ben. 2014. “Plotting Regression Coefficients and Other Estimates.” The Stata
Journal 14 (4): 708–737.
Jentleson, Bruce W. 1992. “The pretty prudent public: Post post-Vietnam American
opinion on the use of military force.” International studies quarterly 36 (1): 49–74.
Johns, Gary. 2018. “Advances in the Treatment of Context in Organizational Re-
search.”AnnualReviewofOrganizationalPsychologyandOrganizationalBehavior
5 (1): 21-46.
Johnson, Dominic D.P., and Dominic Tierney. 2011. “The Rubicon Theory of War:
How the Path to Conflict Reaches the Point of No Return.” International Security
36 (July): 7–40.
Johnston, Lucy. 1996. “Resisting change: information-seeking and stereotype
change.” European Journal of social psychology 26 (5): 799–825.
328
Jonas, Eva, Stefan Schulz-Hardt, Dieter Frey, and Norman Thelen. 2001. “Confirma-
tionbiasinsequentialinformationsearchafterpreliminarydecisions: anexpansion
of dissonance theoretical research on selective exposure to information.” Journal
of personality and social psychology 80 (4): 557.
Jäger, Kai. 2017. “The potential of online sampling for studying political activists
around the world and across time.” Political Analysis 25 (July): 329–343.
Kahneman, Daniel. 2013. Thinking, Fast and Slow. New York: Farrar, Straus and
Giroux.
Kahneman, Daniel, and Amos Tversky. 1979. “Prospect Theory: An Analysis of
Decision under Risk.” Econometrica 47 (2): 263–292.
Kam, Cindy D., and Jennifer M. Ramos. 2008. “Joining and Leaving the Rally
Understanding the Surge and Decline in Presidential Approval Following 9/11.”
Public Opinion Quarterly 72 (4): 619–650.
Katz, Andrew Z. 2000. “Public Opinion and the Contradictions of Jimmy Carter’s
Foreign Policy.” Presidential Studies Quarterly 30 (December): 662–687.
Kertzer, Joshua D. 2013. “Making sense of isolationism: foreign policy mood as a
multilevel phenomenon.” The Journal of Politics 75 (1): 225–240.
Kertzer, Joshua D., and Brian C. Rathbun. 2015. “Fair is Fair: social Preferences
and reciprocity in international Politics.” World Politics 67 (October): 613–655.
Kertzer, Joshua D., Kathleen E. Powers, Brian C. Rathbun, and Ravi Iyer. 2014.
“Moral Support: How Moral Values Shape Foreign Policy Attitudes.” The Journal
of Politics 76 (July): 825–840.
Kertzer, Joshua D., and Thomas Zeitzoff. 2017. “A Bottom-Up Theory of Public
Opinion about Foreign Policy.” American Journal of Political Science 61 (July):
543–558.
Khong, Yuen Foong. 1992. Analogies at War: Korea, Munich, Dien Bien Phu, and
the Vietnam Decisions of 1965. Princeton University Press.
Kim, HeeMin, Jenifer Whitten-Woodring, and Patrick James. 2015. “The Role of
Media in the Repression–Protest Nexus: A Game-theoretic Model.” Journal of
Conflict Resolution 59 (6): 1017–1042.
Koopman, Cheryl, Jack Snyder, and Robert Jervis. 1990. “Theory-Driven versus
Data-Driven Assessment in a Crisis: A Survey of International Security Readers.”
Journal of Conflict Resolution 34 (December): 694–722.
Kőszegi, Botond. 2003. “Health anxiety and patient behavior.” Journal of health
economics 22 (6): 1073–1084.
Köszegi, Botond. 2006. “Ego utility, overconfidence, and task choice.” Journal of the
European Economic Association 4 (4): 673–707.
329
Kőszegi, Botond. 2010. “Utility from anticipation and personal equilibrium.”
Economic Theory 44 (3): 415–444.
Kőszegi, Botond, and Matthew Rabin. 2009. “Reference-dependent consumption
plans.” American Economic Review 99 (3): 909–36.
Kovach, Bill, and Tom Rosenstiel. 2007. The Elements of Journalism: What
Newspeople Should Know and the Public Should Expect, Completely Updated
and Revised. New York: Three Rivers Press.
Kramer, Adam D. I., Jamie E. Guillory, and Jeffrey T. Hancock. 2014. “Experi-
mental evidence of massive-scale emotional contagion through social networks.”
Proceedings of the National Academy of Sciences 111 (June): 8788–8790.
Kupchan, Charles A., and Peter L. Trubowitz. 2007. “Grand Strategy for a Divided
America.” Foreign Affairs 86 (July).
Lang, Kurt, and Gladys Engel Lang. 1966. “The Mass Media and Voting.” In Reader
in Public Opinion and Communication. 2nd ed. New York, N.Y.: Free Press.
Lau, Richard R., and David P. Redlawsk. 2006. How Voters Decide: Information
Processing in Election Campaigns. Cambridge ; New York: Cambridge University
Press.
Lazarus, R. S. 1991. “Progress on a cognitive-motivational-relational theory of emo-
tion.” The American Psychologist 46 (August): 819–834.
Leave.EU. 2016a. “Leaflet Arguments.”
Leave.EU. 2016b. “Our Campaign - Leave.eu.”
Lecheler, Sophie, and Claes H. de Vreese. 2017. “News Media, Knowledge, and Po-
litical Interest: Evidence of a Dual Role From a Field Experiment.” Journal of
Communication 67 (Aug): 545?564.
Leeper, Thomas J. 2014. “The Informational Basis for Mass Polarization.” Public
Opinion Quarterly 78 (January): 27–46.
Lerner, Jennifer S, and Philip E Tetlock. 1999. “Accounting for the effects of ac-
countability.” Psychological bulletin 125 (2): 255.
Levay, Kevin E., Jeremy Freese, and James N. Druckman. 2016. “The Demographic
and Political Composition of Mechanical Turk Samples.” SAGE Open 6 (January).
Lian, Bradley, and John R. Oneal. 1993. “Presidents, the Use of Military Force, and
Public Opinion.” Journal of Conflict Resolution 37 (June): 277–300.
Lippmann, Walter. 1927. The Phantom Public. Transaction Publishers.
Livingston, Steven. 1997. “Clarifying the CNN effect: An examination of media
effects according to type of military intervention.”
Livingston, Steven. 2011. “The CNN effect reconsidered (again): problematizing ICT
and global governance in the CNN effect research agenda.” Media, War & Conflict
4 (1): 20–36.
330
Livingston, Steven, and Todd Eachus. 1995. “Humanitarian crises and US foreign
policy: Somalia and the CNN effect reconsidered.” Political communication 12 (4):
413–429.
Lodge, Milton, and Charles S. Taber. 2013. The Rationalizing Voter. Cambridge ;
New York: Cambridge University Press.
Loewenstein, George. 1994. “The psychology of curiosity: A review and reinterpre-
tation.” Psychological bulletin 116 (1): 75.
Loomes, Graham, and Robert Sugden. 1982. “Regret theory: An alternative theory
of rational choice under uncertainty.” The economic journal 92 (368): 805–824.
Lord, Charles G, Lee Ross, and Mark R Lepper. 1979. “Biased assimilation and
attitude polarization: The effects of prior theories on subsequently considered
evidence.” Journal of personality and social psychology 37 (11): 2098.
Lupia, Arthur.2015.Uninformed: WhyPeopleSeemtoKnowSoLittleaboutPolitics
and What We Can Do about It. New York: Oxford University Press.
Lupia, Arthur, and Mathew D. McCubbins. 1998. The Democratic Dilemma: Can
Citizens Learn What They Need to Know? Cambridge, U.K. ; New York: Cam-
bridge University Press.
Lupia, Arthur, Mathew D. McCubbins, and Samuel L. Popkin, eds. 2000. Elements
of Reason: Cognition, Choice, and the Bounds of Rationality. Cambridge ; New
York: Cambridge University Press.
Malek, Abbas. 1996. “New York Times editorial position and US foreign policy: the
case of Iran revisited.” News Media and Foreign Relations: 224–45.
Maoz, Zeev, and Bruce Russett. 1993. “Normative and Structural Causes of Demo-
cratic Peace, 1946–1986.” American Political Science Review 87 (September): 624–
638.
Marcus, George E., W. Russell Neuman, and Michael MacKuen. 2000. Affective
Intelligence and Political Judgment. University of Chicago Press.
Maslow, Abraham H. 1963. “The need to know and the fear of knowing.” The Journal
of general psychology 68 (1): 111–125.
McCombs, Maxwell E., and Donald L. Shaw. 1972. “The Agenda-Setting Function
of Mass Media.” Public Opinion Quarterly 36 (January): 176–187.
Mellers, Barbara, Eric Stone, Pavel Atanasov, Nick Rohrbaugh, S. Emlen Metz, Lyle
Ungar, MichaelM.Bishop, MichaelHorowitz, EdMerkle, andPhilipTetlock.2015.
“The psychology of intelligence analysis: drivers of prediction accuracy in world
politics.” Journal of Experimental Psychology. Applied 21 (March): 1–14.
Mercer, Andrew, Andrew Caporaso, David Cantor, and Reanne Townsend. 2015.
“How Much Gets You How Much? Monetary Incentives and Response Rates in
Household Surveys.” Public Opinion Quarterly 79 (January): 105–129.
331
Mesquita, Bruce Bueno de, James D. Morrow, Randolph M. Siverson, and Alastair
Smith. 1999. “An Institutional Explanation of the Democratic Peace.” American
Political Science Review 93 (December): 791–807.
Miller, Suzanne M. 1987. “Monitoring and blunting: validation of a questionnaire
to assess styles of information seeking under threat.” Journal of personality and
social psychology 52 (2): 345.
Milner, Helen V., and Benjamin Judkins. 2004. “Partisanship, Trade Policy, and
Globalization: Is There a Left–Right Divide on Trade Policy?” International
Studies Quarterly 48 (March): 95–120.
Milner, Helen V., and Dustin Tingley. 2013. “Public Opinion and Foreign Aid: A
Review Essay.” International Interactions 39 (July): 389–401.
Milner, Helen V., and Dustin Tingley. 2015. Sailing the Water’s Edge: The Domestic
Politics of American Foreign Policy. Princeton, New Jersey: Princeton University
Press.
Mintz, Alex, and Carly Wayne. 2016a. “The Polythink Syndrome and Elite Group
Decision-Making.” Political Psychology 37 (S1): 3–21.
Mintz, Alex, and Carly Wayne. 2016b. The polythink syndrome: US foreign policy
decisions on 9/11, Afghanistan, Iraq, Iran, Syria, and ISIS. Stanford University
Press.
Molotch, Harvey, and Marilyn Lester. 1974. “News as purposive behavior: On the
strategic use of routine events, accidents, and scandals.” American sociological
review: 101–112.
Morgenthau, Hans Joachim. 1948. Politics Among Nations: The Struggle for Power
and Peace. A. A. Knopf.
Mueller, Dennis C. 1993. Public Choice II : A Revised Edition of Public Choice.
Cambridge University Press.
Mueller, John E. 1973. War, Presidents and Public Opinion. New York: John Wiley
& Sons Inc.
Mullinix, KevinJ., ThomasJ.Leeper, JamesN.Druckman, andJeremyFreese.2015.
“The Generalizability of Survey Experiments.” Journal of Experimental Political
Science 2 (January): 109–138.
NatCen Social Research. 2017. “British Social Attitudes: Brexit.”
National Archives and Records Administration. 2017. “U. S. Electoral College.”
Nemeth, Charlan, and John Rogers. 1996. “Dissent and the search for information.”
British Journal of Social Psychology 35 (1): 67–76.
Neuman, W. Russell. 2007. The Affect Effect: Dynamics of Emotion in Political
Thinking and Behavior. University of Chicago Press.
Obama, Barack. 2014. “Statement by the President on ISIL.”
332
Oster, Emily, Ira Shoulson, and E Dorsey. 2013. “Optimal expectations and lim-
ited medical testing: evidence from Huntington disease.” The American Economic
Review 103 (2): 804–830.
Page, Benjamin I., and Marshall M. Bouton. 2006. The Foreign Policy Disconnect:
What Americans Want from Our Leaders but Don’t Get. Chicago: University Of
Chicago Press.
Page, Benjamin I., and Robert Y. Shapiro. 1992. The Rational Public: Fifty Years of
Trends in Americans’ Policy Preferences. Chicago: University Of Chicago Press.
Petty, Richard E., and Duane T. Wegener. 1998. “Matching Versus Mismatching
Attitude Functions: Implications for Scrutiny of Persuasive Messages.” Personality
and Social Psychology Bulletin 24 (March): 227–240.
Petty, Richard E., and John T. Cacioppo. 1986. “The Elaboration Likelihood Model
of Persuasion.” Advances in Experimental Social Psychology 19 (January): 123–
205.
Pew Research Center, for the People, and the Press. 2013a. “Modest Support for
Military Force if Syria Used Chemical Weapons.”
PewResearchCenter, forthePeople, andthePress.2013b.“PublicBacksDiplomatic
Approach in Syria, But Distrusts Syria and Russia.”
Pew Research Center, for the People, and the Press. 2013c. “Public Opinion Runs
Against Syrian Airstrikes.”
Pfau, Michael, Michel Haigh, Mitchell Gettle, Michael Donnelly, Gregory Scott,
Dana Warr, and Elaine Wittenberg. 2004. “Embedding journalists in military
combat units: Impact on newspaper story frames and tone.” Journalism & Mass
Communication Quarterly 81 (1): 74–88.
Pforr, Klaus, Michael Blohm, Annelies G. Blom, Barbara Erdel, Barbara Felderer,
Mathis Fräßdorf, Kristin Hajek, Susanne Helmschrott, Corinna Kleinert, Achim
Koch, Ulrich Krieger, Martin Kroh, Silke Martin, Denise Saßenroth, Claudia
Schmiedeberg, Eva-Maria Trüdinger, and Beatrice Rammstedt. 2015. “Are In-
centive Effects on Response Rates and Nonresponse Bias in Large-scale, Face-to-
face Surveys Generalizable to Germany? Evidence from Ten Experiments.” Public
Opinion Quarterly 79 (January): 740–768.
Pierce, Douglas R., David P. Redlawsk, and William W. Cohen. 2017. “Social Influ-
ences on Online Political Information Search and Evaluation.” Political Behavior
39 (September): 651–673.
Pinkley, Robin L, Terri L Griffith, and Gregory B Northcraft. 1995. “" Fixed Pie" a
la Mode: Information Availability, Information Processing, and the Negotiation of
Suboptimal Agreements.” Organizational Behavior and Human Decision Processes
62 (1): 101–112.
333
Popkin, Samuel L. 1994. The Reasoning Voter: Communication and Persuasion in
Presidential Campaigns. Chicago: University Of Chicago Press.
Potter, David M, and Douglas Van Belle. 2008. “News coverage and Japanese foreign
disaster aid: a comparative example of bureaucratic responsiveness to the news
media.” International Relations of the Asia-Pacific 9 (2): 295–315.
Powdthavee, Nattavudh, and Yohanes E Riyanto. 2015. “Would you pay for trans-
parently useless advice? A test of boundaries of beliefs in the folly of predictions.”
Review of Economics and Statistics 97 (2): 257–272.
Prior, Markus. 2003. “Any good news in soft news? The impact of soft news prefer-
ence on political knowledge.” Political communication 20 (2): 149–171.
Prior, Markus. 2010. “You?ve Either Got It or You Don?t? The Stability of Political
Interest over the Life Cycle.” The Journal of Politics 72 (Jul): 747?766.
Prior, Markus, and Arthur Lupia. 2008. “Money, Time, and Political Knowledge:
Distinguishing Quick Recall and Political Learning Skills.” American Journal of
Political Science 52 (January): 169–183.
Pérez-Peña, Richard. 2008. “The War Endures, but Where’s the Media?” The New
York Times (March).
Rapport, Aaron. 2012. “The Long and Short of It: Cognitive Constraints on Leaders’
Assessments of “Postwar” Iraq.” International Security 37 (December): 133–171.
Rathbun, Brian C. 2004. Partisan Interventions: European Party Politics and Peace
Enforcement in the Balkans. Ithaca: Cornell University Press.
Rathbun, Brian C. 2011. “The ‘Magnificent Fraud’: Trust, International Coopera-
tion, and the Hidden Domestic Politics of American Multilateralism after World
War II1.” International Studies Quarterly 55 (March): 1–21.
Rathbun, Brian C., Joshua D. Kertzer, Jason Reifler, Paul Goren, and Thomas J.
Scotto. 2016. “Taking Foreign Policy Personally: Personal Values and Foreign
Policy Attitudes.” International Studies Quarterly 60 (March): 124–137.
Riker, William H., and Peter C. Ordeshook. 1973. An Introduction to Positive
Political Theory. Prentice-Hall.
Robertson, Lori. 2004. “Images of war: this year the American news media have
displayed pictures of burned bodies in Fallujah, flag-draped coffins coming home
from Iraq and the abuse of Iraqi prisoners at Abu Ghraib. But were they too
squeamish when it came to showing the carnage of war during the invasion last
year?” American Journalism Review 26 (5): 44–52.
Robinson, Piers. 1999. “The CNN effect: can the news media drive foreign policy?”
Review of international studies 25 (2): 301–309.
Robinson, Piers.2000.“Thepolicy-mediainteractionmodel: Measuringmediapower
during humanitarian crisis.” Journal of Peace Research 37 (5): 613–633.
334
Robinson, Piers. 2001. “Theorizing the influence of media on world politics: Models
of media influence on foreign policy.” European Journal of Communication 16 (4):
523–544.
Russett, Bruce. 1994. Grasping the Democratic Peace: Principles for a Post-Cold
War World. Princeton University Press.
Ruths, Derek, and Jürgen Pfeffer. 2014. “Social media for large studies of behavior.”
Science 346 (November): 1063–1064.
Salganik, MatthewJ.2017. BitbyBit: Social Research intheDigital Age.Princeton,
NJ: Princeton University Press.
Scherer, Aaron M, Paul D Windschitl, and Andrew R Smith. 2013. “Hope to be
right: Biased information seeking following arbitrary and informed predictions.”
Journal of Experimental Social Psychology 49 (1): 106–112.
Scherpenzeel, Annette, and Vera Toepoel. 2012. “Recruiting a Probability Sample
for an Online PanelEffects of Contact Mode, Incentives, and Information.” Public
Opinion Quarterly 76 (January): 470–490.
Schneider, Walter, and Richard M. Shiffrin. 1977. “Controlled and automatic human
information processing.” I. Detection, search, and attention. Psychological Review:
1–66.
Sears, David O., and Jonathan L. Freedman. 1967. “Selective Exposure to Informa-
tion: A Critical Review.” Public Opinion Quarterly 31 (January): 194–213.
Shao, Chengcheng, Giovanni Luca Ciampaglia, Onur Varol, Alessandro Flam-
mini, and Filippo Menczer. 2017. “The spread of fake news by social bots.”
arXiv:1707.07592 [physics] (July).
Sharot, Tali. 2011. “The optimism bias.” Current biology 21 (23): R941–R945.
Shear, Michael D., and Michael R. Gordon. 2017. “63 Hours: From Chemical Attack
to Trump’s Strike in Syria.” The New York Times (April).
Shoemaker, Pamela J. 1996. Mediating the Message: Theories of Influences on Mass
Media Content. Longman.
Shu, Lisa L, and Francesca Gino. 2012. “Sweeping dishonesty under the rug: how
unethical actions lead to forgetting of moral rules.” Journal of personality and
social psychology 102 (6): 1164.
Sides, John, and Jack Citrin. 2007. “European Opinion about Immigration: The
Role of Identities, Interests and Information.” British Journal of Political Science
37 (3): 477–504.
Simon, Herbert. 1971. “Designing Organizations for an Information-Rich World.” In
Computers, Communications, and the Public Interest, ed. Martin Greenberger.
Baltimore, MD: Johns Hopkins University Press.
335
Simon, Herbert A. 1978. “Rationality as Process and as Product of Thought.” The
American Economic Review 68 (2): 1–16.
Strömbäck, Jesper, and Adam Shehata. 2010. “Media malaise or a virtuous circle?
Exploring the causal relationships between news media exposure, political news
attention and political interest.” European Journal of Political Research 49 (Aug):
575?597.
Stroud, Natalie Jomini. 2017. “Attention as a Valuable Resource.” Political
Communication 34 (July): 479–489.
Suhay, Elizabeth, and Cengiz Erisen. N.d. “The Role of Anger in the Biased Assim-
ilation of Political Information.” Political Psychology. Forthcoming.
Suhay, Elizabeth, James N. Druckman, Patrick W. Kraft, Milton Lodge, and
Charles S. Taber. 2015. “Why People “Don’t Trust the Evidence”: Motivated Rea-
soning and Scientific Beliefs.” The ANNALS of the American Academy of Political
and Social Science 658 (March): 121–133.
Taber, Charles S., and Milton Lodge. 2006. “Motivated Skepticism in the Evaluation
of Political Beliefs.” American Journal of Political Science 50 (July): 755–769.
Taylor, Shelley E, and Jonathon D Brown. 1988. “Illusion and well-being: a social
psychological perspective on mental health.” Psychological bulletin 103 (2): 193.
Tetlock, Philip.1998.“Close-CallCounterfactualsandBelief-SystemDefenses: IWas
Not Almost Wrong but I Was Almost Right.” Journal of Personality and Social
Psychology 75 (September): 639–652.
Tetlock, Philip E. 1992. “Good Judgment in International Politics: Three Psycho-
logical Perspectives.” Political Psychology 13 (3): 517–539.
Tetlock, Philip E. 1999. “Theory-Driven Reasoning About Plausible Pasts and
Probable Futures in World Politics: Are We Prisoners of Our Preconceptions?”
American Journal of Political Science 43 (2): 335–366.
Tetlock, Philip E. 2006. Expert Political Judgment: How Good Is It? How Can We
Know? Princeton, NJ: Princeton University Press.
Tetlock, Philip E, and Dan Gardner. 2016. Superforecasting: The art and science of
prediction. Random House.
The Electoral Commission. 2018. “Electoral Commission | Electoral data.”
Theocharis, Yannis, Pablo Barberá, Zoltán Fazekas, Sebastian Adrian Popa, and
Olivier Parnet. 2016. “A Bad Workman Blames His Tweets: The Consequences of
Citizens’ Uncivil Twitter Use When Interacting With Party Candidates.” Journal
of Communication 66 (December): 1007–1031.
Thussu, Daya Kishan. 2003. “Live TV and Bloodless Deaths: War, Infotainment and
24/7 News.” In War and the Media: Reporting Conflict 24/7, ed. Daya Kishan
Thussu and Des Freedman. London: Sage.
336
Tomz, Michael R., and Jessica L. P. Weeks. 2013. “Public Opinion and the Demo-
cratic Peace.” American Political Science Review 107 (November): 849–865.
Tormala, Zakary L., and Richard E. Petty. 2002. “What Doesn’t Kill Me Makes Me
Stronger: The Effects of Resisting Persuasion on Attitude Certainty.” Journal of
Personality and Social Psychology 83 (December): 1298–1313.
Tormala, Zakary L., and Richard E. Petty. 2004. “Resistance to Persuasion and
Attitude Certainty: The Moderating Role of Elaboration.” Personality and Social
Psychology Bulletin 30 (November): 1446–1457.
Tuchman, Gaye. 1980. Making News. Later printing edition ed. New York: Free
Press.
Tversky, A., and D. Kahneman. 1981. “The framing of decisions and the psychology
of choice.” Science (New York, N.Y.) 211 (January): 453–458.
U. S. Census Bureau. 2017. “American FactFinder - Results.”
Ungar, Lyle, Barb Mellors, Jon Baron, Philip Tetlock, Jaime Ramos, and Sam Swift.
2012. “The good judgment project: A large scale test of different methods of com-
bining expert predictions.” AAAI Fall Symposium - Technical Report (January):
37–42.
Vaccari, Cristian, Augusto Valeriani, Pablo Barberá, John T. Jost, Jonathan Nagler,
and Joshua A. Tucker. 2016. “Of Echo Chambers and Contrarian Clubs: Exposure
to Political Disagreement Among German and Italian Users of Twitter.” Social
Media + Society 2 (September): 2056305116664221.
Van Belle, Douglas A. 1997. “Press freedom and the democratic peace.” Journal of
Peace Research 34 (4): 405–414.
Vote Leave. 2016a. “Briefing Room.”
Vote Leave. 2016b. “Why Vote Leave.”
Waltz, KennethN.1959.Man, theState, andWar: ATheoreticalAnalysis.Columbia
University Press.
Waltz, Kenneth N. 1979. Theory of International Politics. Long Grove, Ill: Waveland
Press.
Wang, Wei, DavidRothschild, SharadGoel, andAndrewGelman.2015.“Forecasting
elections with non-representative polls.” International Journal of Forecasting 31
(July): 980–991.
Warriner, Keith, John Goyder, Heidi Gjertsen, Paula Hohner, and Kathleen Mc-
spurren. 1996. “Charities, No; Lotteries, No; Cash, Yes: Main Effects and In-
teractions in a Canadian Incentives Experiment.” Public Opinion Quarterly 60
(January): 542–562.
White, David Manning. 1950. “The “Gate Keeper”: A Case Study in the Selection
of News.” Journalism Bulletin 27 (September): 383–390.
337
Whitten-Woodring, Jenifer. 2009. “Watchdog or lapdog? Media freedom, regime
type, and government respect for human rights.” International Studies Quarterly
53 (3): 595–625.
Whitten-Woodring, Jenifer, and Patrick James. 2012. “Fourth estate or mouth-
piece? A formal model of media, protest, and government repression.” Political
Communication 29 (2): 113–136.
Zajonc, R. B. 1980. “Feeling and thinking: Preferences need no inferences.” American
Psychologist: 151–175.
Zaller, John. 1999. “A theory of media politics.” Manuscript, October 24 (1999):
61–101.
Zaller, John R. 1992. The Nature and Origins of Mass Opinion. Cambridge England
; New York, NY, USA: Cambridge University Press.
Zaller, John, and Stanley Feldman. 1992. “A Simple Theory of the Survey Response:
Answering Questions versus Revealing Preferences.” American Journal of Political
Science 36 (3): 579–616.
Zheng, Alvin, Jing Gong, and Paul Pavlou. 2017. On Using the Lottery in Crowd-
funding Platforms: ‘Crowding in’ the Masses But ‘Crowding out’ Success. SSRN
Scholarly Paper ID 2916807 Social Science Research Network Rochester, NY: .
https://papers.ssrn.com/abstract=2916807.
338

Abstract (if available)

Abstract This dissertation contributes towards a more complete understanding of information, public opinion, and international relations. I demonstrate that scholars must understand the conditions under which people perceive incentives for accuracy to understand how people learn, think, feel, and behave regarding international affairs. Through a combination of an incentivized online survey experiment and analysis of public opinion polls, I find that people's attitudes and behavior are contingent on their perception of incentives for accuracy. If people consider the stakes are high and it is important to make the correct decision, they are motivated to: 1) search for information to help them make the correct decision

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Legitimizing self-determination: advancing the sovereignty of separatist movements

PDF

Political participation, public opinion, and the law

PDF

Public opinion and international affairs: a multi-method approach to foreign policy attitudes

PDF

Code 'war' theorizing: information and communication technology's impact on international relations theorizing, negotiation, and cyber relations

PDF

The role of local actors in international development: how incentive types affect foreign aid effectiveness

PDF

Neither purse nor sword: strategic minority opinions and judicial policy-making

PDF

Habits and policy: the socio-cognitive foundations of foreign policy stability

PDF

The changing state of tolerance and tolerance in changing states

PDF

Private interests in American government institutions

PDF

The origins and evolution of the U.S. alliance network: how military allies transform and transact

PDF

Analysis and prediction of malicious users on online social networks: applications of machine learning and network analysis in social science

PDF

Authoritarian religion: explaining when and why authoritarian governments instrumentalize religion

PDF

Like father, like son? A succession-based explanation for conflict initiation by authoritarian regimes

PDF

International politics and domestic institutional change: the rise of executive war-making autonomy in the United States

PDF

Perspectives on state capacity and the political geography of conflict

PDF

Circuit breakers: how policy entrepreneurs interrupted the electric flow with Peru’s first renewable energy legislation for the grid

PDF

A two-level analysis of foreign policy decision making: an empirical investigation of the case of China-Taiwan

PDF

Selling virtue: how human rights NGOs and their donors work together to create a better world ... for themselves

PDF

The fabled fourth estate: media freedom, democracy and human rights

PDF

Essays on behavioral decision making and perceptions: the role of information, aspirations and reference groups on persuasion, risk seeking and life satisfaction

Asset Metadata

Creator Jamieson, Thomas (author)

Core Title Information, public opinion, and international relations

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Publication Date 07/25/2020

Defense Date 04/24/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Brexit,incentives,Information,Iraq,ISIS,OAI-PMH Harvest,political knowledge,predictions,public opinion, foreign policy,Syria

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor James, Patrick (committee chair), Chong, Dennis (committee member), Seib, Philip (committee member), Weller, Nicholas (committee member)

Creator Email thomas_jamieson23@hotmail.co.nz,tjamieso@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-28035

Unique identifier UC11672504

Identifier etd-JamiesonTh-6465.pdf (filename),usctheses-c89-28035 (legacy record id)

Legacy Identifier etd-JamiesonTh-6465.pdf

Dmrecord 28035

Document Type Dissertation

Format application/pdf (imt)

Rights Jamieson, Thomas

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA