University of Southern California Dissertations and Theses

Thawing rivalries and fading friendships: a multi-method approach to rapprochement and alienation

(USC Thesis Other)

Thawing rivalries and fading friendships: a multi-method approach to rapprochement and alienation

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content ! i!
!
!
!
!
!
!
!
Thawing Rivalries and Fading Friendships
A Multi-Method Approach to Rapprochement and
Alienation

by

Rod Albuyeh

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
POLITICAL SCIENCE AND INTERNATIONAL
RELATIONS

August 2016
! ii!

Dedication

For Katie

! iii!
Table of Contents

Acknowledgements iv

Chapter 1 1
Introduction

Chapter 2 18
Thawing Rivalries and Fading Friendships: An Experimental
Approach to Rapprochement and Alienation

Chapter 3 58
Enmity is What Leaders Make of It: The Microfoundations for
Perceptions of Adversaries

Chapter 4 101
Deterrence Without Threat or Force: Individual Differences in
Sensitivity to Past Actions

Chapter 5 121
Conclusion

! iv!
Acknowledgements

First and foremost, I extend my gratitude to my dissertation committee: Patrick James,
Brian Rathbun, and Jesse Graham. Pat has been an advisor, mentor, and friend to me throughout
this process. I am immensely grateful to him for taking me on as a mentee back when I only had
a vague idea of where my research was headed. He struck the right balance between pushing me
hard to improve my work while still providing encouragement and positive reinforcement when I
needed it most. I also offer my thanks to Brian for the thoughtful investment he put into helping
me develop my work as it has evolved (through how many iterations of revisions, I have lost
count) over the years, and for his believing in my ability to grow as a scholar. I offer my thanks
to Jesse for his invaluable guidance, especially in his helping me better emulate psychologists in
my methodology. I have also enjoyed participating in several sessions at his Values, Ideology,
and Morality Lab (VIMLab) and always found the discussions fascinating and the atmosphere
fun.
I would also like to acknowledge a number of other individuals who have contributed to
my graduate development. I offer my thanks to Ben Graham and Jonathan Markowitz for their
faith in me as a research assistant, and providing the opportunity to learn how to be more
organized, efficient, and tenacious as a scholar. I learned a great deal about professionalism by
being in their company and for that I am grateful. I also extend my thanks to Benjamin Cohen,
my master’s thesis advisor at UC Santa Barbara, for the example he sets as a scholar. He went
well above the call of duty in his support. I also acknowledge the support of other faculty
members such as Rand Wilcox for teaching me how to approach statistics more critically and
effectively, Steve Lamy for his commitment to the art of teaching, Garrett Glasgow for
introducing me to the wonders of R programming and quantitative methods, Eric RAN Smith for
showing me how Political Science has a heart, Saori Katada for her service as POIR director and
for always keeping the students’ best interests in mind, and to Jennifer Ramos and Carolyn
James for allowing me to learn through involvement in ISA-West. I also extend my thanks to
Veri Chavarin, Cathy Ballard, and Indira Persad for being great at what they do and for their
friendship. I offer a special thanks to all of my undergraduate professors at UC Irvine and De
Anza College for their inspiration.
I also extend my thanks to my colleagues for all the laughter and camaraderie. To my
future groomsmen Michael Albert and Julian Gottlieb, thanks for everything. To Marcus Arrajj,
thanks for being the un-graduate student. To Mark Paradis, Therese Anders, Youssef Chouhoud,
Nicolas DeZamaroczy, Thora Giallouri, Joey Huddleston, Tom Jamieson, Brian Knafou, Elli
Menounou, Simon Radford, Güez Salinas, Jihyun Shin, Ming-Min Yang, Ronald Osborn, Justin
Berry, Chin-Hao Huang, Fabian Borges-Herrero, Seanon Wong, Matthew Mendez, Scott
! v!
Englund,—you all have qualities that I admire and I am grateful for having had you as my
colleagues, you showed me that graduate school was a friendly place.
One of the benefits of being getting married within two months of defending my
dissertation is that I have a list of friends and family members easily at hand. Feri, Dara, Sanam,
and Ali, thank you for welcoming me into your home when I was at Irvine. Thank you Jarred for
being a reliable friend, thank you Roein and Henry for always being there. To the Ehsanis:
Faribourz, Sarita, Farzad, Silke, Ghazaleh, and Mimi, I appreciate and admire you in the example
you set as a family and the way you treat one another, and the way you raise your children. To
the Mitchell family: Betsy, Greg, Christie, Michael, Emma, and many others, thank you for
welcoming me so warmly. Mahtab and Mansour, I admire your relationship and appreciate your
kindness and senses of humor. Farzaneh, Iman, and Helen, you are great people. Kourosh and
Bijan, I am so grateful to have you as brothers. Ashley and Jibreel, you are welcome additions to
the family.
I extend my thanks to my two greatest inspirations, my father Farrokh and my
grandfather Kambiz. Thank you both for your support, pressure, encouragement, more pressure,
more support, more encouragement, etcetera etcetera. I can say for certain that I would not have
made it this far were it not for you. I would not have pursued my education to the extent that I
did. Thank you for everything.
Last but not least, I send my thanks to Moneer.
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! vi!
!
!
! 1!
Chapter 1

Introduction

This article-style dissertation consists of three studies tied together by the theme of belief
revision in international relations. This exploration of belief revision is somewhat timely given
that in the spring of 2015, the top foreign ministers of the P5+1 states—the United States, United
Kingdom, France, Russia, China plus Germany—as well as the European Union minister, stood
alongside the Iranian Minister of foreign affairs to announce the framework for a comprehensive
agreement on the Iranian nuclear program. This indicated a major shift from almost four decades
of tension between Iran and the West, especially the United States. A similar shift took place in
the summer of 2015 between the United States and Cuba, although recent relations between the
two countries had not been nearly as hostile as between the US and Iran. I am not aware of any
members of the scholarly community that would have anticipated these shifts a decade ago.
Throughout the processes of rapprochement in the Iran and Cuba cases, decision-makers
and members of the public disagreed about how to approach negotiations and interpret the
meaning of actions by the states under scrutiny. Even now that the Iran nuclear deal is complete
and they have complied with the sanctions thus far, it is not uncommon to hear politicians
discuss the return of sanctions and de-normalization of relations. Is this simply a matter of
hawks distrusting other states and doves being eager to give the benefit of the doubt? Perhaps.
But another question to consider is how the behavior of states is filtered through differing
individual cognitive schemata.
! 2!
In this introduction, I outline literature concerning how states and individuals assess the
intent of other states and, by extension, make predictions about their future behaviors. Since my
goal here is to explore the role of subjectivity in rapprochement and alienation, I also discuss
literature on simplified models of the political universe such as authoritarianism, social
dominance orientation, and ideology.

Perceived Intent in International Relations Theory: Circumstance and Subjectivity
Many scholars in the field of international relations have, to a greater or lesser degree,
maintained that the intent of other states is a key factor in explaining conflict, cooperation, and
state behavior more generally. For instance, Mearsheimer (2014) says that due to the confluence
of (1) the anarchic international system, (2) the offensive nature of great power military
capabilities, (3) a lack of certainty about the intentions of other states, (4) survival as states’
primary goal, and (5) the rational nature of state units, great powers assume the worst of the
intentions of rivals and maintain an offensive disposition towards other states, and judge threats
by way of capabilities rather than intentions. According to this view, intentions are seen as
constant, and therefore rival states are expected to discount behavior that is intended to signal
intent. Others, however, have emphasized the critical role of signals when they are associated
with material cost. According to this latter view, states try to manipulate the beliefs other actors
hold about them, usually by way of creating audience costs or sinking costs through military
expenditures (Fearon 1997). Some forms of communication are considered “cheap,” such as
private communication and diplomacy (Sartori 2002). However it has been argued that even
cheap talk can affect perceptions of intentions (Trager 2010).
! 3!
Another body of work in international relations scholarship, dealing with state reputations
for resolve and credibility, has emphasized the role of past behaviors as informing states’ beliefs
about other states’ intentions, with the predominant focus being on the intentions of rivals (which
they usually discuss in terms of apparent resolve to use force) in crisis situations (Huth 1997).
Apparent resolve is more important than actual resolve, since a challenger can never determine
how resolute a defender actually is. Challenger’s task is to use incomplete information to form
reputational inferences about the defender. Some have found that past actions do not inform
perceptions of intentions (Hopf 1994), arguing instead that current preferences and capabilities
are better indicators (M. Tomz 2007; Press 2007). Others have countered that past actions do
matter in informing the behavior of other states (Weisiger and Yarhi-Milo 2015). For instance
Downs and Jones (2002) find that inferences will be made the more broadly the action deviates
from the interests of the actor perceiving the actions. The debate over whether past actions
matter in actors’ assessments of other states’ intentions endures, and no consensus has emerged
in the field.
Political psychologists have also grappled with the question of how actors update their
beliefs about other states’ intentions, and have attempted to add a dose of subjectivity to the
explanation. For instance, Mercer (2010) finds that observers only update their assessments of
intentions when the signal conflicts with their interests. When it does not, observers attribute
behavior to circumstances. Image theorists such as Herrmann and Fischerkeller (1995) create a
typology of state types and attempt to predict and explain state behaviors based on their views of
other states. For image theorists, the key dimensions of interest are relative power (i.e.
capabilities), status, and goal compatibility, which feed directly into perceived intentions. In
keeping with the field’s focus on conflict and cooperation, the main focus is on the ally and
! 4!
adversary ideal-types, where allies are assumed to have benign intentions and adversaries are
expected to have hostile ones. For instance, Bilali (2010) finds that a state with equal
capabilities and cultural status elicits fear-based responses, but when the state signals non-
hostility actors will seek opportunities to cooperate.
One of the issues underlying much of the scholarship on state intentions is that
individuals are generally assumed to either not vary in the dynamics underlying their
assessments or other states, or that assessments are expected to converge relatively quickly. This
also holds true for most realist scholars, deterrence theorists, and even many political
psychologists. With only a few exceptions, almost none have considered the interplay between
material factors, circumstantial factors, and individual-level variation. Two scholars in particular
stand out in providing a synthesis of these factors.
One recent
1
example of this comprehensive approach is Murray (1997), who traces
dramatic changes in American opinion leaders’ views about the Soviet Union and containment
policy at the end of the Cold War. She notes that there was a broad consensus among American
leaders in terms of containment policy, yet American leaders embraced different stereotypes
about the Soviet adversary. There was a hard-line view that believed US military superiority and
vigilant containment provided the best means to security, and a soft-line perspective that
advocated more accommodating tactics such as arms control agreements. For the hard liners,
Gorbachev’s signaling was not received warmly due to wariness over soviet intentions, while the
soft-liners were more receptive to his attempts at rapprochement. In the end, end the collapse of
the Soviet Union was too strong of an indicator for even the staunchest hard-liners to ignore. Yet
while beliefs about the hostile intent of the Soviet Union were mitigated with its collapse,
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
1
Operational code scholars have also pursued this line of thought, however that research program tends to be more
concerned with foreign policy.
! 5!
decision-makers’ broader beliefs about the appropriate way to conduct foreign affairs remained
stable.
Murray’s key conclusions are that attitudes towards rivals can change while general
policy orientations do not, and that “domestic ideological orientations” are constrain these views.
This is a key intuition into the forces underlying belief change among international actors,
however her findings seem to suggest that the causal arrow begins with domestic ideological
orientations and ends with foreign policy orientations. It is more likely, however, that either the
domestic and foreign policy realms are falsely distinguished from each other, or that there is a
third underlying factor that underlies both.
The second, and perhaps best example of this approach is provided by Yarhi-Milo
(2014), in her account of how decision-makers update their beliefs about other states. She
focuses on adversary intentions in particular, and advances a framework centered on how
decision-makers filter and interpret the same signals. She compares conventional approaches to
her selective attention thesis, the latter of which posits that individual perceptual biases and
organizational interests influence which types of indicators observers will view as credible
signals of adversary intentions. Thus, decision-makers base their interpretations on their own
theories, expectations, and needs, sometimes ignoring costly signals and paying more attention to
information that is more emotionally or personally vivid to them. This is an excellent
contribution to comprehensive scholarship on inter-state belief change, as Yarhi-Milo correctly
identifies the processes and dynamics underlying common disagreements between policy-makers
over how to interpret signals from adversaries. However her assessment of subjectivity could
benefit from social psychology which discusses the factors underlying differences in belief-
revision.
! 6!
Yarhi-Milo identifies a factor underlying belief-revision that is often ignored by other
international relations scholars: that decision-makers interpret a signal as informative in a
manner consistent with their expectations about the relationship between behavior and a state’s
underlying characteristics. That is, they interpret evidence as informative according to their
theories about what kinds of signals are indicative of type. Unfortunately, her discussion of US
decision-makers focuses on personal idiosyncrasies, rather than on behavioral batterns that might
be more widely generalizable. She does, though, recognize (in similar fashion to Murray) that
the hawk-dove ideological dynamic seems to underlie personal theories about what kinds of
signals are indicative of an adversary’s intentions. But to be fair, Yarhi-Milo was not assessing
belief revision per say, but rather the conditions under which signals and indices would be
treated as informative by decision-makers.
While Murray and Yarhi-Milo’s scholarship considers structure and circumstance when
explaining beliefs about other states, it emphasizes the fact that the physical world is filtered
through a human mind which relies on simplified models of reality in updating—a subjectivist
perspective that can be traced back to William James in the 19
th
century. I seek to expand on this
view that privileges subjectivity in two ways: first, I intend to identify factors common to all
people that underlie beliefs about intentions of other states. Second, I intend to move beyond
scholarship that focuses on beliefs about adversaries alone (which of course is the product of the
Cold War context, the wake of which international relations scholarship does not yet seem to be
clear from) and expand the focus to include allies as an ideal type, as it may be the case that the
processes underlying rapprochement differ from those underlying alienation

! 7!
Authoritarianism and Social Dominance Orientation: Simplified Models of the Political
Universe
There certainly is something underlying the divide in foreign policy belief systems among
American decision-makers since the collapse of the Soviet Union, and perhaps that is the same
divide underlying the divide in beliefs about what types of signals credibly reveal the intentions
of other states (Yarhi-Milo 2014). We might be content to know that the divide tends to occur
along party lines in the US, where Republicans tend to advocate foreign policy vigilance and
discount overtures from other states, and Democrats tend to advocate a more cooperative posture
and be more receptive to overtures. But why is this the case? Is it simply a matter of party
politics, of ideologically driven hawkism and doveism?
Ideology is often thought of as being multidimensional, and no doubt political ideology is
associated with many factors. For instance, while social psychologists define political ideology
broadly as a value-neutral “set of beliefs about the proper order of society and how it can be
achieved,” and describe the polarity as involving the split between acceptance and resistance of
social change and inequality, a meta-analysis by Jost et al. (2009) finds that “the dimensional
structure and attitudinal contents of liberalism and conservatism…stem, at least in part, from
basic social psychological orientations concerning uncertainty and threat” (311). Jost et al. link
ideology to such factors as: death anxiety, beliefs about system instability, fear of threat and loss,
dogmatism, tolerance of ambiguity, personal needs for order and closure, cognitive complexity,
and self-esteem.
As for which of these factors might most directly pertain to the question of the processes
underlying belief revision between states, the answer must involve the notion of threat, and it
must be reconcilable with both domestic and foreign policy orientations, in particular regarding
! 8!
inter-group relations. Thus, I turn my attention to social psychological scholarship on prejudice,
which has been found to vary systematically among individuals (Duckitt 2001). The first major
assessment of individual-level prejudice was Adorno et al.'s (1950) theory of the authoritarian
personality, which held that prejudice, in-group glorification, political conservatism, and
profascist attitudes covaried powerfully to form an attitudinal schema. This research program
remained dormant for decades, but was revived by Robert Altemeyer who constructed the right-
wing authoritarian (RWA) scale to measure Adorno's authoritarian personality dimension
(Altemeyer 1981).
Duckitt links RWA to “authoritarianism, social conservatism, or traditionalism, at one
pole, versus openness, autonomy, liberalism, and personal freedom at the other pole.” The
prejudice that arises from this schema can come in the following qualitative type:
“[The RWA] kind of prejudice will be driven by fear and threat generating self-
protective, defensive motivational needs for social control and security. Here out-
groups will be disliked because they are seen as threatening and dangerous to
social or group cohesion, security, order, values, and stability. This would tend to
create a particular kind of readiness to categorize in a prejudiced fashion; that is,
categorizing the social world into “us,” who are good, normal, moral, decent
people and who are threatened by “them,” who are bad, disruptive, immoral, and
deviant.” (85)

The key takeaway here is that RWA is driven by fear of out-groups, who are seen as a threat to
order and stability, and thus individuals high in RWA will be more likely to be vigilant in their
dealings with out-groups. Although the group categorization is value driven, based on a
dichotomization of groups into good and evil, the political universe in general is seen as
threatening, and all out-groups are assumed to be hostile. This dynamic should generalize to
individuals’ views of other states. Individuals high in RWA might be assumed to see all other
states, allies and adversaries alike, as being more threatening. Additionally, these individuals
! 9!
might be more likely to discount information that disconfirms their prior beliefs that out-groups
are not to be trusted.
While Adorno initially named RWA the “authoritarian personality,” it came to be known
as right-wing authoritarianism, perhaps due to its correlation with measures of political
conservatism. However, there is no mention of how authoritarian personalities with a left-wing
bent might map onto this scale. For example, it’s hard to expect that Mao Zedong, Joseph Stalin,
or Pol Pot would compare not compare similarly to notable right-wing authoritarians like Benito
Mussolini and Adolf Hitler. Thus, I will henceforward refer to the scale with the same
terminology as its progenitor: Authoritarian Personality (AP).
We might expect individuals who are high in AP, who view the political universe as
dangerous and threatening, to be wary of adversaries and allies alike, since their fear based view
of other states leads them to assume that all states will take any advantage at the first
opportunity. By the same token individuals who are low in AP, who tend to view the political
universe as harmonious, might view both adversaries and allies more warmly, since their
optimistic view leads them to believe that even adversarial relations can be remedied, perhaps by
way of identification of common goals and interests.
A paradigm complementary to authoritarian personality theory is social dominance
theory, developed by Sidanius and Pratto. They proposed an attitudinal dimension called social
dominance orientation (SDO), referring to it as “a general attitudinal orientation toward
intergroup relations, reflecting whether one generally prefers such relations to be equal, versus
hierarchical” and the “extent to which one desires that one's in-group dominate and be superior
to out-groups” (Sidanius and Pratto 1993, 742).
! 10!
Duckitt finds that even though AP and SDO are distinct dimensions and predict similar
political phenomena, including prejudice and nationalism, they do not predict each other well.
He cites ten independent studies assessing the correlation between AP and SDO, with correlation
coefficients varying between .11 and .21. Thus, the two scales are very close to being orthogonal.
In disentangling the differences between AP and SDO, he ties SDO to “economic conservatism,
power, or belief in hierarchy or inequality at its one pole versus egalitarianism, humanitarianism,
social welfare, or concern at its other pole.” In effect, SDO refers to a Machiavellian, amoral
“competitive jungle” worldview, while AP is linked to values and traditions (Duckitt 2001).
According to Duckitt, SDO can be differentiated as follows:
“The [SDO] kind of prejudice will be driven by a view of the world as a
competitive jungle characterized by a ruthless and amoral struggle for resources
and power in which the fittest and most powerful succeed and the unfit and weak
fail. This activates self-enhancement motives for power, status, superiority, and
dominance. Here out-groups will be derogated because they are seen as inferior,
weak, inadequate, and failures. This would create a readiness for a different kind
of prejudiced social categorization, with the social world categorized into “us,”
who are superior, strong, competent, and dominant (or should rightfully be
dominant), versus “them,” who are inferior, incompetent, weak, and worthless.”
(85)

Political Ideology
A construct closely related to AP and SDO is political ideology. Jost et al. (2009)
provide the most comprehensive review to date of the comparative psychological
differences between liberals and conservatives. They employ Erikson’s (2003) definition
of political ideology as a value-neutral “set of beliefs about the proper order of society
and how it can be achieved”, and describe the notion of the left-right dimension as
involving the split between acceptance and resistance of social change and inequality.
They claim that this dichotomy is robust across cultures, and that its economic and social
! 11!
components are correlated to the point of near-redundancy. While ideology has a socially
constructed aspect, they claim “the dimensional structure and attitudinal contents of
liberalism and conservatism...stem, at least in part, from basic social psychological
orientations concerning uncertainty and threat” (311). They cite experimental studies
linking conservatism with death anxiety, beliefs about system instability, fear of threat
and loss, dogmatism, tolerance of ambiguity, and personal needs for order and closure,
and other studies linking liberalism with openness to new experiences, cognitive
complexity, tolerance of uncertainty, and self-esteem.
More importantly for the purposes of this study, political ideology has been linked
to AP and SDO. In other words, conservatives are more likely to be vigilant in their
dealings with out-groups, as the political universe in general is seen as threatening since
all other states are considered out-groups. Meanwhile liberals tend to see the world in
fundamentally the opposite manner, and accordingly believe that the political universe is
fundamentally harmonious and that conflicts generally arise due to a failure to properly
communicate.

Moving Forward
It is my view that the closely related constructs of AP, SDO, and political ideology
should be at the center of any assessment of inter-state belief revision. Subjectivity has a key
role in any examination of how states revise their beliefs about the intentions of other states.
Still, conventional wisdoms concerning structure must be taken into consideration, the following
chapters differ by design in the degree to which they emphasize it. Chapter 2 begins the
assessment by expanding the focus of belief revision to compare how its processes differ
! 12!
between the contexts of rapprochement and alienation. Due in large part to the backdrop of the
Cold War, the preponderance of scholarship on belief revision has focused on either deterrence
between adversaries, where states signal resolve to not back down when challenged (Hopf 1994;
Huth 1997; Downs and Jones 2002; Mercer 2010), or on rapprochement between hostile
adversaries (Murray 1997; Yarhi-Milo 2014). Using a novel survey experiment, Chapter 2
expands into fresher territory by examining the process of alienation, where relations between
allies become more hostile, alongside rapprochement, where relations between adversaries
become more harmonious. Between the two contexts, I find that the logic of costly signaling
differs fundamentally, where hostile signals are far more likely to lead to belief revision than
reassuring signals, and increasing cost in hostile signals does not lead to a greater magnitude in
belief revision regarding the threat posed by other states. More importantly, however, I find that
political belief systems are the most powerful determinant of how individuals perceive the intent
of other states.
Chapter 3 returns to more familiar territory by narrowing the focus to adversary relations
and examining belief revision in the context of the Cold War. While Chapter 2 had a relatively
short time horizon in its experimental prompt, Chapter 3 leverages a natural experiment over a
longer time horizon between 1985 and 1991, where the “treatment” is the rise of Gorbachev and
a series of reassuring signals that took place at the end of the Cold War immediately prior to the
Soviet Union’s collapse. Here I continue with the examination of the role political belief
systems, this time examining their effect on complexity of thought regarding adversaries.
Employing an automated content analytical technique to assess US legislators’ beliefs about the
Soviet Union, I find that when structural conditions are in line with decision-makers’ ideological
expectations, complexity of thought about the adversary increases. Conversely, when structural
! 13!
conditions do not accord with ideological expectations, decision-makers become more simplistic
in their views of the adversary.
Chapter 4 strips out the political context to assess in isolation individual differences in
sensitivity to information about past actions, by way of an incentivized market-entry bargaining
game. This chapter is motivated by policymakers’ common claim that war is justified when
national credibility and reputation for resolve is at stake. A debate in scholarship on reputation
has formed around this notion, where some claim that past actions do matter in deterring
challengers (Hopf 1994; Press 2007; Mercer 2010), and others claim that past actions do not
matter (Sartori 2002; Tomz and Weeks 2013; Trager and Vavreck 2011; Sartori 2013) Given
that chapters 2 and 3 examine the role of material factors such as costly signaling and political
belief systems, chapter 4 narrows the focus to individual differences in attributional tendencies in
a minimalist bargaining context. Holding past behavior constant, I find that individuals with a
situational attributional style are more likely to play the game’s optimal strategy as dictated by
the observed actor’s past behavior and the structure of payoffs given by the game.
Chapter 5, rather than recapitulating the findings and implications of the three empirical
chapters, critically assesses conventional wisdoms and parallel debates in various domains of
international scholarship concerned with beliefs and expectations about states, and relates these
ideas to the debate leading up to the Iran nuclear deal and foreign policy axioms espoused by
policymakers. I find that theory in international images, deterrence, and operational code each
centers on similar questions in international relations. The lack of cross-pollination between
these approaches, in my view, has led to missed opportunities for theory development and
refinement, as each emphasizes different aspects of structure and agency. I close the chapter
with a discussion of the practical implications of my empirical findings, a proposed synthesis
! 14!
between the approaches with an eye towards the inclusion of political psychology as a
foundation, and call for a return to the study of individuals in international relations research.
This return, it seems, is already well underway.

References
!
Adorno, T.W., Else Frenkel-Brunswik, Daniel J. Levinson, and R. Nevitt Sanford. 1950. The
Authoritarian Personality. Vol. xxxiii. Oxford, England: Harpers.
Altemeyer, Bob. 1981. Right-Wing Authoritarianism. University of Manitoba press.
Bilali, Rezarta. 2010. “Assessing the Internal Validity of Image Theory in the Context of
Turkey-U.S. Relations.” Political Psychology 31 (2): 275–303. doi:10.1111/j.1467-
9221.2009.00752.x.
Castano, Emanuele, Alain Bonacossa, and Peter Gries. 2015. “National Images as Integrated
Schemas: Subliminal Primes of Image Attributes Shape Foreign Policy Preferences.”
Political Psychology, March, n/a – n/a. doi:10.1111/pops.12259.
Downs, George W., and Michael A. Jones. 2002. “Reputation, Compliance, and International
Law.” The Journal of Legal Studies 31 (S1): S95–114. doi:10.1086/340405.
!
Duckitt, John. 2001. “A Dual-Process Cognitive-Motivational Theory of Ideology and
Prejudice.” Advances in Experimental Social Psychology 33: 41–114.
Fearon, James D. 1994. “Domestic Political Audiences and the Escalation of International
Disputes.” American Political Science Review 88 (03): 577–92. doi:10.2307/2944796.
! 15!
———. 1997. “Signaling Foreign Policy Interests Tying Hands versus Sinking Costs.” Journal
of Conflict Resolution 41 (1): 68–90. doi:10.1177/0022002797041001004.
Healy, Andrew, and Gabriel S. Lenz. 2014. “Substituting the End for the Whole: Why Voters
Respond Primarily to the Election-Year Economy.” American Journal of Political
Science 58 (1): 31–47. doi:10.1111/ajps.12053.
Herrmann, Richard K., and Michael P. Fischerkeller. 1995. “Beyond the Enemy Image and
Spiral Model: Cognitive–strategic Research after the Cold War.” International
Organization 49 (03): 415–50. doi:10.1017/S0020818300033336.
Hopf, Ted. 1994. Peripheral Visions: Deterrence Theory and American Foreign Policy in the
Third World, 1965-1990. University of Michigan Press.
Huth, Paul K. 1997. “Reputations and Deterrence: A Theoretical and Empirical Assessment.”
Security Studies 7 (1): 72–99. doi:10.1080/09636419708429334.
Johnson, Dominic D. P., Rose McDermott, Emily S. Barrett, Jonathan Cowden, Richard
Wrangham, Matthew H. McIntyre, and Stephen Peter Rosen. 2006. “Overconfidence in
Wargames: Experimental Evidence on Expectations, Aggression, Gender and
Testosterone.” Proceedings of the Royal Society of London B: Biological Sciences 273
(1600): 2513–20. doi:10.1098/rspb.2006.3606.
Jost, John T., Christopher M. Federico, and Jaime L. Napier. 2009. “Political Ideology: Its
Structure, Functions, and Elective Affinities.” Annual Review of Psychology 60 (1): 307–
37. doi:10.1146/annurev.psych.60.110707.163600.
Kertzer, Joshua D., Jonathan Renshon, and Keren Yarhi-Milo. 2015. “How Do Observers Assess
Resolve?” Working Paper. http://jonathanrenshon. com/Site/CurrentResearch
files/KertzerRenshonYarhi-Milo021415. pdf.
! 16!
Kriner, Douglas, and Francis Shen. 2014. “Responding to War on Capitol Hill: Battlefield
Casualties, Congressional Response, and Public Support for the War in Iraq.” American
Journal of Political Science 58 (1): 157–74. doi:10.1111/ajps.12055.
Mearsheimer, John. 2014. The Tragedy of Great Power Politics. Updated Edition. New York:
W. W. Norton & Company.
Mercer, Jonathan. 2010. Reputation And International Politics. First edition. Cornell University
Press.
Murray, Kathleen Shoon. 1997. Anchors against Change: American Opinion Leaders’ Beliefs
After the Cold War. Ann Arbor: University of Michigan Press.
Press, Daryl G. 2007. Calculating Credibility: How Leaders Assess Military Threats. 1st ed.
Cornell University Press.
Sartori, Anne E. 2002. “The Might of the Pen: A Reputational Theory of Communication in
International Disputes.” International Organization 56 (01): 121–49.
doi:10.1162/002081802753485151.
Schelling, Thomas C. 2008. Arms and Influence. New Haven, Conn.; London: Yale University
Press.
Sidanius,! Jim,! and! Felicia! Pratto.! 1993.! “The! Dynamics! of! Social! Dominance! and! the!
Inevitability!of!Oppression.”!Prejudice,*Politics,*and*Race*in*America*Today:!173–211.!
!
Tomz, Michael. 2007. Reputation and International Cooperation: Sovereign Debt across Three
Centuries. Princeton University Press.
Tomz, Michael R., and Jessica L. P. Weeks. 2013. “Public Opinion and the Democratic Peace.”
American Political Science Review 107 (04): 849–65. doi:10.1017/S0003055413000488.
! 17!
Trager, Robert F. 2010. “Diplomatic Calculus in Anarchy: How Communication Matters.”
American Political Science Review 104 (02): 347–68. doi:10.1017/S0003055410000158.
Wallace, Geoffrey P.R. 2013. “International Law and Public Attitudes Toward Torture: An
Experimental Study.” International Organization 67 (01): 105–40.
doi:10.1017/S0020818312000343.
Weisiger, Alex, and Keren Yarhi-Milo. 2015. “Revisiting Reputation: How Past Actions Matter
in International Politics.” International Organization 69 (02): 473–95.
doi:10.1017/S0020818314000393.
Yarhi-Milo, Keren. 2014. Knowing the Adversary: Leaders, Intelligence, and Assessment of
Intentions in International Relations. Princeton University Press.
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! 18!
Chapter 2

Thawing Rivalries and Fading Friendships: An
Experimental Approach to Rapprochement and Alienation

Although international relations theorists generally assume that actors update
their beliefs about the intentions of adversaries and allies based on structure,
costly signaling, and past actions, little is known about how the process of
rapprochement between adversaries differs from the process of alienation
between allies, particularly with respect to the nature and degree of costly
signaling. Furthermore, until recently the role of the individual in these
processes has been only been engaged by a small number of scholars, and fewer
still have integrated this perspective with conventional approaches to
rapprochement and alienation. Drawing from findings in social psychology, I
present results from an original survey experiment showing that (1) political
belief systems are a powerful determinant of how individuals perceive the
intentions of other states, more so than an observed state’s signaling behavior;
(2) there are diminishing returns in increasing the cost of a signal; and (3) hostile
signals are more effective in signaling intent than reassuring signals.

When states attempt to make costly signals of intent to change the beliefs of adversaries
or allies, what factors influence the degree to which observers update their prior beliefs about
intentions? International Relations theory places a great deal of focus on the effects of costly
signaling, the perceived hostile intent of adversaries, and to a lesser degree the perceived benign
intent of allies. Yet, even with recent developments in US-Iran and US-Cuba relations, we still
understand little about belief change as it pertains to the process of rapprochement between
adversaries, where intentions are viewed as becoming more benign—and the field is virtually
silent about the process of alienation between allies, where intentions are viewed as becoming
more hostile.
In attempting to assess how state actors form their beliefs about the intentions of other
states, many IR theorists focus on the role played by structure and circumstance. For instance,
! 19!
realists argue that states focus on relative capabilities when assessing intentions, thus the only
type of signal that would lead to a change in beliefs about intentions is a change in armaments
(Mearsheimer 2014). The logic of costly signaling follows from this, and stipulates that
incremental costly changes in armaments change ex-ante beliefs about intent, hostile or benign
(Yarhi-Milo 2014). In a similar vein, deterrence and reputation theorists emphasize the role that
past or current behaviors play in updating actors’ beliefs about the intentions of other states
(Tomz 2007; Press 2007). What these perspectives overlook is the fact that individuals do not
interpret the costliness of signals in the same way, even when these individuals have similar
priors and identical access to information (for instance, following the Iran nuclear deal in 2015
decision-makers remained divided on whether sanctions should be lifted). Other IR scholars
focus on the role of subjective or dispositional factors. Political psychologists have found that
there are some global human biases that color the way signals are perceived (Mercer 2010),
while others have pointed to differences in individuals’ personal theories about how the political
world functions that influences how they view the perceptions of other states in light of
“objective” material factors (Yarhi-Milo 2014).
I argue that the structural approaches, in focusing almost exclusively on relations among
adversaries, have ignored an important component of structure: pairwise variation in the ideal-
typical ally-adversary binary. As will be illustrated in this chapter by way of an original survey
experiment, implemented online via Amazon Mechanical Turk and the Qualtrics advanced
survey platform, the process of updating beliefs about the intentions of other states is determined
by whether the observed state is an ally signaling hostile intentions or an adversary signaling
benign intentions. I find that costly signaling is, in the short run, only effective when an ally is
signaling hostile intentions. Costly signaling has little to no effect when an adversary is
! 20!
signaling benign intentions. I also find, somewhat paradoxically, that where costly signaling is
effective, the degree of costliness generally does not matter beyond the minimal requirement of
including some change in capabilities.
An experimental methodology allows me to measure the role individual variation in
subjectivity plays. Moving beyond basic differences in political ideology, which is often
couched in terms of “dovishness” and “hawkishness,” I draw from scholarship in social
psychology and argue that Authoritarian Personality (AP) impacts how individuals view allies
and adversaries alike. When individuals assess the intentions of adversaries and allies, their view
of them is strongly influenced by whether they view the political universe as harmonious (low
authoritarianism) or conflictual (high authoritarianism). Put differently, individuals tend to view
individual states in a manner consistent with their political universe. Those who see the universe
as harmonious will hold more favorable views of allies and adversaries alike, while the opposite
holds for individuals who see the universe as conflictual. The paradoxical finding here is that
causal effect of this belief system is a stronger determinant than costly signaling itself of
individuals’ beliefs about the intentions of other states. I also find that while political ideology is
moderately related to AP, there is no causal mediation effect between these variables and thus
AP stands independently in its influence on beliefs about state intentions.
My argument proceeds in three parts. I first outline a set of hypotheses regarding belief
change that both tests the structural and circumstantial perspectives from IR while also making
allowances for variation in individual dispositions. I then discuss my experimental methodology
and present the results, which are derived from a robust nonparametric methodology, and
conclude with a discussion of the implications of my findings.

! 21!

Theoretical Expectations
In this section, I derive a series of hypotheses derived from structural and
dispositional approaches to beliefs about state intentions. The first step is to determine
some reference points in ideal-typical state types. Much of the literature dealing with
state intentions focuses on adversary intentions, however I argue that narrowing the focus
to adversaries might bias the process of updating, and it may be the case that the process
of updating differs between rapprochement and alienation between states. Thus, I offer
general reference points for allies and adversaries as follows:

I define an ally as a state that has:
1.! The ability to inflict significant material harm to an actor’s home
state
2.! Interests that an actor understands to be harmonious with the home
state
3.! A consistent record, within an actor’s memory, of publicly stated
friendliness toward the home state
4.! A consistent record, within an actor’s memory, of policies that are
understood to complement those of the home state

Conversely, I define an enemy as a state that has:
1.! The ability to inflict significant material harm to an actor’s home
state
! 22!
2.! Interests that an actor understands to conflict with the home state
3.! A consistent record, within an actor’s memory, of publicly stated
antagonism toward the home state
4.! A consistent record, within an actor’s memory, of policies that are
understood to undermine those of the home state

For both ideal types
2
, only the third and fourth points allow for a target state to send
signals that disconfirm its type. For example, an ally can break the pattern of friendly
rhetoric toward the home state with a hostile or critical statement, or an enemy can begin
to make reassuring rhetorical overtures. These might be thought to constitute “cheap
talk” or non-costly signals, however some actors may consider the reputational costs of
cheap talk and be more likely to update based on these signals. Troop movements, on the
other hand, might be thought to be costlier as a signal. An ally might engage in
alienation by enacting new policies that undermine a home state’s interests, or an enemy
could engage in rapprochement by enacting policies complementary to a home state.
Dispositional factors may influence the degree to which individuals update or
anchor to pre-existing beliefs about the intentions of other states in the face of belief-
disconfirming information. Individuals may vary in their proclivity to treat new signals
as informative versus attributing the signal to circumstances. For instance, individuals
high in AP who view the political universe as threatening will be more likely to view all
states, whether they be allies or adversaries, as more threatening. This is akin to the
attitude of “keep your friends close, and your enemies closer,” wherein no out-group can
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2
I exclude other ideal types that might include minor powers relative to the home state, such as something akin to
North Korea’s relationship to the US (a degenerate state according to Herrman and Fisherkeller) or Taiwan’s
relationship to China (which Herrmann and Fisherkeller might call a colony image).
! 23!
be fully trusted, since all states are viewed as fundamentally hostile, only acting on that
hostility when the benefits outweigh the costs. Thus my first hypothesis, Authoritarian
Worldview Confirmation Hypothesis, says:

Authoritarian Worldview Confirmation Hypothesis: For observed allies and
adversaries, the higher an individual’s AP, the more they will tend to believe the
observed state has a hostile intent towards them.

SDO is also a key factor to consider. It is difficult to anticipate if and how SDO might differ
from AP in its implications on the process of inter-state belief revision, thus the Dominance
Worldview Confirmation Hypothesis is similar to its authoritarian variant:

Dominance Worldview Confirmation Hypothesis: For observed allies and
adversaries, the higher an individual’s SDO, the more they will tend to believe the
observed state has hostile intent towards them.

I also employ political ideology as a baseline for comparison to AP and SDO. It is possible that a
simple self-reported measure of political ideology performs as well as the more elaborate factor
scores for AP and SDO. It is also possible that ideology’s effects on belief revision align more
strongly with either of AP or SDO on the rapprochement and alienation contexts. In any case, I
define the Ideological Worldview Confirmation Hypothesis as follows:

! 24!
Ideological Worldview Confirmation Hypothesis: For observed allies and
adversaries, the higher an individual’s conservatism, the more they will tend to
believe the observed state has hostile intent towards them.

The next hypothesis follows from literature on costly signaling, as it is often claimed that
sinking costs is a more effective signal than cheap talk. Less is said, however, about how much
of a cost must be sunk to be a credible signal. We should expect that “costless” cheap talk
(assuming no reputational or audience costs) is less effective than cheap talk paired with a
relatively low cost signal involving military expenditure, which in turn is less effective than
cheap talk paired with a relatively high cost signal involving military expenditure. Thus, the
Incremental Costly Signaling Hypothesis and Military Expenditure Signaling Hypothesis weigh
gradations of costly signals as follows:

Incremental Costly Signaling Hypothesis: Belief-disconfirming signals via public
statements result in less updating in beliefs about intentions than those paired
military expenditures.

Military Expenditure Signaling Hypothesis: The belief revision associated with
the Incremental Costly Signaling Hypothesis increases as the cost of military
expenditures increases.

As mentioned earlier, it may be the case that the process of rapprochement, where adversaries
attempt to credibly signal benign intent, differs from that of alienation, where allies attempt to
! 25!
credibly signal hostile intent. Following from Downs and Jones’ (2002) finding that a
reputational inference will be more extreme the broader the action deviates from the interests of
the actor, I derive the Threat Vigilance Hypothesis:

Threat Vigilance Hypothesis: Allies can more effectively signal hostile intent than
adversaries can signal benign intent.

This hypothesis is akin to an interaction effect, where a state’s type interacts with costly
signaling such that estimates of the effects of costly signaling differ between allies and
adversaries. Another interaction effect I anticipate concerns the interplay between the worldview
variables and costly signaling. Individuals who are high in either of the scales in the direction of
threat vigilance, i.e. high in AP, SDO, or conservatism, might be more sensitive to costly hostile
signals from allies. Thus I advance the Threat Interaction Hypothesis:

Threat Interaction Hypothesis: There will be a positive interaction effect between
the worldview measures (AP, SDO, and ideology) and costly signaling in the
alienation scenario.

Methods
I use an experimental methodology to test my hypotheses. It is now well known that
the use of experimental methods in international relations scholarship has grown
exponentially in recent years. The value in such an approach lies in its’ ability to identify
! 26!
causal effects. In the summer of August 2015, I surveyed 1002 American adults recruited
via Amazon Mechanical Turk, using the Qualtrics survey design platform. Participants
were paid $2.30 for their participation in a roughly 15-minute survey experiment.
Because my experimental design did not require a physical presence in a lab, the survey
experiment was a highly efficient use of resources. In keeping with the “best practices”
in the field, I limited participation in the study to workers located in the United States,
who had completed at least 50 human intelligence tasks (HITs) and whose approval rate
was greater than 95%. The survey included three simple attention checks, including
asking participants if humans need to eat food in order to survive, and one slightly more
demanding attention check determining if participants paid attention to and understood
the brief treatment they were presented with. The survey terminated if any of these
attention checks were failed and participants were informed in advance that they would
not receive compensation in this instance. The resulting sample size after dropping failed
attention checks was 862.
As has been noted elsewhere (Kertzer, Renshon, and Yarhi-Milo 2015), Amazon’s
Mechanical Turk has grown in popularity in experimental social science. Berinsky, Huber,
and Lenz (2012, 366) show that Mturk samples are “often more representative of the general
population and substantially less expensive to recruit” than other “convenience samples” often
used in political science. They also demonstrate the ability to replicate results from nationally
representative samples. Experiments using MTurk samples have been published in almost all
notable journals in the field, including APSR (Tomz and Weeks 2013), the American Journal of
Political Science (Healy and Lenz 2014), International Organization (Wallace 2013) and Journal
of Conflict Resolution (Kriner and Shen 2014).
! 27!
Fielding an experiment dealing with questions of international relations to the mass
public poses some issues, because often the goal of a survey is to aggregate up to the elite or
state level. My view is that assessing the views of the public is important. As Fearon (1994)
notes, crises in democracies are carried out in front of political audiences evaluating the skill and
performance of leaders who they have the ability to remove from office. Moreover, it is not
feasible to begin a study testing new hypotheses on an elite sample, and we can also assume that
many of the human dynamics prevailing among the public are to some degree generalizable to
leaders. A public sample also provides a better sense of the conditions under which views about
the intentions other states, which are often debated, might converge into a consensus.
The survey experiment employs a fully-crossed, quasi-factorial design, allowing subjects
to take on all possible combinations of situational factors. The observed state type and level of
costliness are randomly manipulated. The first situational factor has two levels: the target state is
either an ally or an adversary sending a signal that disconfirms its type, so an ally sends a hostile
signal while an adversary sends a reassuring signal; while the second situational factor has three
levels: the costliness of the signal, which can be low (public statements), medium (public
statements + small change in military funding and mobilization), or high (public statements +
large change in military funding and mobilization. Thus, this constitutes a 2x3 factorial design,
resulting in six possible combinations of factors. I refer to the design as quasi-factorial because
the ally and adversary factors are not perfectly comparable, as will become evident, thus the
design can also be seen as two separate experiments with costliness being a single three-level
factor in both. However, the two designs are deliberately similar to allow for comparison. For
the full set of situational prompts, see Appendix A1.
! 28!
Participants were presented at random with a dispositional questionnaire and a foreign
policy scenario with follow-up questions. In the scenario, participants were told that they would
be reading about a hypothetical state and asked several questions about that state, referred to only
as “the country.” Then, they were told that the US had maintained either hostile or harmonious
relations with that state for the last fifty years, that it had proven to be consistently
hostile/harmonious in public statements and statecraft. In all scenarios, the state was said to
possess nuclear weapons and a military that is half as strong as the US military, thus satisfying
the condition that the state must be able to impose material harm on the US. Participants were
then told that high-ranking officials in the country had been either publicly critical (in the
alienation scenario) or supportive (in the rapprochement scenario) of a new naval deployment by
the US in a resource-rich region in international waters, and that this type of statement was
unprecedented for the country over the past fifty years. This constitutes “cheap talk,” and
accordingly is a relatively low cost signal. Some participants were told nothing further about
that state, while others were told that the country had also announced a change in naval funding
(either an increase for the alienation scenario or a decrease for the rapprochement scenario) and
has either scheduled naval exercises in the contested area (for the alienation scenario) or
unscheduled naval exercises in that area (for the rapprochement scenario). For the case of the
medium cost signal, participants were told that the change in naval funding is 25%, while for the
high cost signal participants were told that the change was 75%
3
.
Following the scenario, participants were presented with a number of questions intended
to measure their beliefs about the intent of the observed state. They were asked, in random
order, a number of Likert scale questions: (1) to indicate how much of a threat the country
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
3
It may be the case that this is too subtle of a manipulation for costly signaling. In a future study I will increase cost
in a more pronounced fashion.
! 29!
represents to the US; (2) to what extent they would support or oppose the US agreeing to the
country’s request for economic assistance; (3) to what extent they would support or oppose the
US signing a new military cooperation agreement with the country; and (4) to what extent they
would support or oppose the US signing a new trade agreement with the state when it is not clear
who will benefit more from the agreement. The first question of threat is the most direct
assessment of perceived intent, as a state’s perceived level of threat is directly a reflection of its
perceived intent. The second, third, and fourth questions are an attempt to determine the policy
ramifications that should arise from perceived threat, and to find out which foreign policy
domains the perceived intent of states is tied to most closely.
To measure demographics and dispositions, participants were presented (randomly
either before or after the scenario) a questionnaire including ten items from Altemeyer’s
Right Wing Authoritarianism questionnaire (see Appendix A2), and a battery of
demographic controls such as age, gender, ideology, and socioeconomic status. Figure 1
indicates the overall design of the experiment.
Virtually all experimental designs in the international relations involve a variety of
tradeoffs. For instance, some ask individuals to take on the perspective of a decision-maker,
even if it has been shown that taking on the perspective of a foreign policy leader results in
individuals behaving more hawkishly than they otherwise would (Kertzer and Renshon 2014).
The hypothetical scenario itself might also be problematic, given its minimal nature, leading to a
tradeoff between a short scenario that is easily comprehensible and a more detailed scenario that
might lose the attention of participants. I opted for the former route,

! 30!
Figure 1. Signaling Experiment Structure

since Castano et al. (2015) find that even minimally invoking the images of other states with a 40
millisecond prime (consisting of a single word adjective) activates strong schemata of
adversaries and allies that shape attitudes and policy preferences towards those states. Thus, I
expect that even a short scenario outlining the last fifty years of relations with a hypothetical
adversary is sufficient to invoke the ideal-typical stereotype.

Data and Results
Before discussing the statistical model used to test my hypotheses, it is necessary to
provide a brief descriptive summary of the sample. Participants were 47% female, on average 35
! 31!
years old, predominantly white, and liberal in their political ideology, and the vast majority had
at least some college education. A score for authoritarian personality was derived for each
participant by way of an exploratory factor analysis of responses to the ten items from the RWA
and SDO questionnaires. A scree test strongly indicated the existence of only one underlying
factor in both instances, both according to the eigenvalue criteria and the scree plot elbow rule
(see Appendix A3). Using robust weighted least squares estimation and a promax rotation, a
factor analysis of the ten items indicates that each item possesses a standardized loading ranging
between 0.61 and 0.80 (Appendix A4). Factor scores were generated for each participant based
on these results, and the factor scores were compared to participants scores for authoritarian
personality based on a simple additive index, and the correlation between the crude additive
scale and the robust latent factor scores was 0.9986, indicating that the results of the factor
analysis are extremely reliable and that the overall quality of the data in terms of user
attentiveness is quite high.
The distribution of AP and SDO scores are non-normal as indicated by a Shapiro-Wilk
normality test, as are virtually all of the non-categorical demographic variables. Thus, a
parametric model based on the assumption of normality is likely not appropriate, since we cannot
know how large the sample size must be for the central limit theorem to apply. The measures
also possess a number of outliers, thus a model that is robust to outliers and leverage points is
also preferable. To deal with non-normality and outliers, I employ bootstrap sampling for all
analyses used herein, and rely on outlier-robust bootstrapped MM estimation for the main
analysis in this piece. I then compare the results to bootstrapped least squares and bootstrapped
ordered probit regression, because Likert scale dependent variables can be appropriately handled
by both. I find that the choice of approach to outliers and the scale of the dependent variable
! 32!
makes negligible difference in levels of significance, magnitude, and direction of effect (See
Appendix A5 and A6). These findings are robust to various approaches to estimation.
As for model specification, I include each of the worldview measures separately and the
costliness of the signal (either low, medium, or high) as explanatory variables, and include
controls such as gender, age, education, political interest, and an intergroup threat factor. Since
AP and political ideology are highly correlated (r = 0.61), it is problematic to include both in the
model, but I employ a follow-up causal mediation analysis to identify any potential mediation
effects between the two. Social Dominance and Authoritarianism are also too strongly correlated
to include in a model together without risking collinearity (r = 0.51), as are Social Dominance
and ideology (r = 0.45). Table 1 indicates the zero-order correlations between these variables.

Table 1. Zero-Order Correlations between AP, SDO, and Ideology

Table 2 indicates the results of the primary bootstrapped MM estimation results (while the OLS
and GLM estimates are presented in Appendices A5-A6). Table 3 directly compares the
coefficients for AP, SDO, and political ideology.

! 33!
Table 2. Bootstrap MM Estimates

What stands out immediately is that authoritarian personality is statistically significant on
all dependent variables in both the alienation and rapprochement scenarios, and moreover the
magnitude of effect is quite high. For instance, as authoritarian personality increases from its
minimum value to its maximum value, which has a range of roughly 4 (from -1.28 to 3.39), we
should expect an increase in perceived threat of almost a full unit in the perceived threat of allies
sending hostile signals, which accounts for a roughly 17% increase in threat assessment. The
magnitude of effect is roughly half for the rapprochement scenario, however the overall effect is
still more substantial than the other significant binary variables for that scenario, those being
gender and political interest.

! 34!
Table 3. Coefficients for Worldview Variables

A similar pattern holds for the policy choices, where higher AP scores are significantly
associated with a lower likelihood to offer economic aid, engage in a military agreement, or sign
a new trade agreement—in both scenarios. Due to the slight difference of scaling for these
variables, the magnitude of effect in some instances is even higher than for the threat assessment.
For instance, as we move from the lowest to highest extreme in AP, individuals become roughly
25% less likely to sign a military or trade agreement with an adversary, and between 12-18% less
likely to offer economic aid to any state, or sign a military or trade agreement in the alienation
scenario. In sum, there is strong support for the Authoritarian Worldview Confirmation
Hypothesis, which posits that the higher an individual’s AP, the more they will tend to believe
the observed state has a hostile intent and poses a threat. The results for SDO are more
inconsistent, however. This worldview construct lacks AP’s statistical significance in two
instances: for perceived threat in the rapprochement scenario and for support for a military
agreement in the alienation scenario. Moreover, even where SDO’s directionality and
significance accords with AP, the magnitude of effect is lower in almost all instances. Thus I
find support for the Dominance Worldview Confirmation Hypothesis although the results are
relatively less robust than the Authoritarian variant of the hypothesis. The findings are muddier
for ideology, which only has statistical significance in half of the analyses. Ideology
significantly affects support for granting economic aid in both the alienation and rapprochement
! 35!
scenarios, for support for a military agreement in the rapprochement scenario, and for support for
a trade agreement in the rapprochement scenario. When ideology is statistically significant, the
directionality of its relationship is in accordance with the coefficients for AP and SDO, although
the magnitude of effect is lower. The emergent pattern is that when it comes to specific policy
questions in the rapprochement scenarios, the results for political ideology match those of AP
and SDO. In other words, we might expect ideology to be a possible compromise predictor
when studying policy preferences of political elites, where assessing their levels of AP and SDO
is not possible. In any case, I find mixed support for the Ideological Worldview Confirmation
Hypothesis.
There is mixed support for the Incremental Costly Signaling Hypothesis, contingent on
whether the signaling is occurring in the rapprochement or alienation scenario, and no support
for the Military Expenditure Signaling Hypothesis. Levels of costly signaling only differ in their
effect at times in the ally alienation scenario, and even within that subset the effects are
inconsistent. For instance, both medium-cost and high-cost signals result in a significantly
higher assessment of the threat of allies who are signaling hostility when compared to low-cost,
however the overlapping confidence intervals between the two coefficients indicate there is no
significant difference between medium and high cost signaling. Figure 2 illustrates the overlap
in confidence intervals between medium cost and high cost signals for the perceived threat of
allies signaling hostility.
On the other hand, high cost signals are significantly different from low cost signals in
reducing the likelihood that economic or military aid will be offered in the alienation scenario,
while medium cost signals are not significantly different than low- or high-cost signals.

! 36!

Figure 2. Coefficient Plots – Perceived Threat in Alienation Scenario

Interestingly, there is no significant difference between low-, medium-, and high-cost signals in
their effect on the likelihood of signing a new trade agreement. This may be due to the wording
of the question, which is unique from the other questions in its emphasis that “it is not clear who
will benefit more from the agreement,” perhaps leading the participants to be less sensitive to the
degree of costliness of the hostile signal.
There is strong support for the Threat Vigilance Hypothesis, which says that Allies can
more effectively signal hostile intent than adversaries can signal benign intent. With only one
exception, virtually none of the coefficients for costliness of signal are significant for the
rapprochement scenario, which indicates that higher cost benign signals are no more effective in
! 37!
the short term than low cost signals. Meanwhile, as already discussed, increasing the costs of
hostile signals in the alienation scenario can be effective, although it is not clear whether
increases in the cost of the signal result in a greater change of beliefs about intentions in the mind
of observers.
It is important to note that the magnitude of effect of costly signaling, even where it had a
large effect—for instance for the question of offering economic aid in the face of hostile
signaling in the alienation scenario—was still lower in overall magnitude than the effect of
authoritarian personality. While a high-cost hostile signal resulted in a 7% decrease in
individuals’ favor for offering economic aid, moving along the continuum of AP among
respondents results in a 17% decrease.
Table 4. Interaction Effects Comparison

Table 4 provides results for interaction effects between the three worldview variables and
costly signaling. The results are unequivocal: there are no statistically significant interaction
effects and thus, there is no support for the threat interaction hypothesis, which indicates that
individuals who are sensitive to threat are not more sensitive to costly threatening signals than
they are to cheap threatening signals. For individuals high in AP, SDO, and political
conservatism alike, threat is threat.
! 38!
All but one of the control variables had a consistent significant effect: gender. Women
were more likely to view the target state as threatening in both the alienation and rapprochement
scenarios, and were less likely to support economic aid, as well as military and trade agreements
in the rapprochement scenario, which indicates that women are less receptive in general to
benign signals from adversaries than men. This is consistent with other experimental findings
which have indicated strong gender differences in foreign policy behavior (Johnson et al. 2006).
Before concluding, the relationship between the worldview constructs and political
ideology must be explored. Since AP has the most consistent effects and is more strongly
associated with political ideology than SDO, I will narrow my focus to the relationship between
AP and ideology. The respondents surveyed displayed a reasonably high association (r = 0.80)
between their ideological self-placement (via a 7pt Likert scale from extremely liberal to
extremely conservative) and their self-placement between the two main political parties in the
US (via a 7pt Likert scale from strong Republican to strong Democrat). Given the high
association between AP and ideology (r = 0.61) as illustrated in Figure 3, we must consider
whether AP is having a direct effect on perceived intentions of the hypothetical state, or if it is
being mediated through political ideology.
Rather than estimating causal mediation effects by way of structural equation modeling, I
employ Imai, Keele, and Tingley’s (2010) general method for nonlinear non-parametric
mediation analysis. The results of this analysis are in Table 2, which are calculated using
N=1000 simulations and 95% quasi-Bayesian confidence intervals, and show that authoritarian
personality’s effect on perceived threat in the alienation and rapprochement scenarios is not
mediated by political ideology, indicating that authoritarian personality is an independent
predictor of the perceived intent of other states.
! 39!
Table 5. Causal Mediation Analysis
Alienation Rapprochement
Average Causal Mediation Effect -0.05 0.02
Average Direct Effect 0.28* 0.10*
Total Effect 0.22* 0.13*
Proportion Mediated -0.24 0.18
Note: *p < .05; Nonparametric Bootstrap Confidence Intervals with the Percentile Method.
Figure 3. Authoritarian Personality and Political Ideology

! 40!

Conclusion
The concept of belief revision is at the core of many questions of international politics.
Most scholars assume that actors update prior beliefs about the intentions of other states when
presented with satisfactory evidence through costly signaling, but the findings here support
another view: that individuals tend to hold beliefs about state intentions that logically cohere
with their overall belief that the political universe is either harmonious or conflictual. This
applies most strongly to the AP construct, but a case can be made for SDO as well and to a lesser
degree political ideology. Given that individuals vary so much in belief systems, it should be no
surprise that observers with similar priors and identical access to information about other states
often disagree about whether signals are viewed as being indicative of changed intentions. This
disagreement over state intentions is likely to persist even when “objective” material conflicts of
interests are resolved; thus we may never see convergence in beliefs about states’ intentions.
The recent disagreement over whether or not Iran can be trusted with nuclear technology is only
one example among many in history.
Moreover, regardless of individual variation in political belief systems, hostile signals
from allies are more likely to be treated as credible than benign signals from adversaries. This
indicates that the process of rapprochement is fundamentally different than alienation, and that
the core of this difference lies in the nature of the signal being hostile or benign. Individuals
seem to find threatening signals as being more salient and informative than reassuring signals in
the short term. But just as individuals disagree over the perceived intent of other states, they also
disagree about how to interpret individual signals. In any case, these findings seem to provide
! 41!
some support to realists’ skepticism that politics can be less conflictual. Power politics may
indeed be the result of human sensitivities to threats.
Within the realist school of thought, my findings seem to suggest an answer to some
missing links in the causal mechanism underlying offensive realism, which some have claimed
has poorly specified micro-foundations. If states are not inherently aggressive, as offensive
realism claims, there should be no reason for great powers to fear rivals since their principal
motive is simply survival (James 2009). In other words, offensive realism’s aggressive zero sum
logic does not follow from the claims that states do not have a hard-wired “will-to-power” hard
wired into them and instead only seek survival. My findings suggest the missing link in this
instance: states may not have an inherent will to power, however even the mere perception that
rival states are predatory and hostile will spur them to aggressively maximize power by way of a
survivalist logic. Furthermore, offensive realism’s tragic characterization of the world as unable
to escape rivalry and suspicion among states is supported by my finding that hostile signals are
more likely to be taken seriously than benign signals. Interestingly, this finding was anticipated
in James’ critical assessment of offensive realism:
From an offensive realist point of view, decision makers should be more sensitive
in their probability calculations to negative rather than positive information about
rivals…Bad experience is weighted more heavily than good experience, which
produces vigilance about the possibility of war but also the ability to optimize
military preparedness based on updated beliefs rather than more static estimates
about capabilities. (Ibid, 54)

In any case, while I do find that individuals are more sensitive to negative than positive signals,
my findings do not support offensive realism’s claim that all actors make worst case assumptions
about rivals’ intentions. Instead, only some actors do, those that already view the political
universe as being fundamentally hostile. Future research should assess further the nuances of the
role of self-preservation in the interpretation of signals. I suspect self preservation is a strong
! 42!
motivational foundation, one rewarded by natural selection over the course of human evolution.
If one assumes the worst of others when reacting to threats, they will prepare accordingly and
should be more likely to survive when that threat is followed by aggression.
As for the question of how costly a signal must be to be informative to observers, we are
left with some questions answered, and others left unsolved. Generally speaking, costly
signaling is only more effective than cheap talk in the case of hostile signaling, while it seems to
have no effect in the case of benign signaling. But even in the alienation scenario, it seems to
only matter to observers that the target state incur some costs, but the difference between a
quarter change in military funding and triple that amount has no effect on the perceived
credibility of the signal. Thus, policymakers might not need to break the bank in attempting to
credibly signal hostile intent.
While this study is focused on identifying the dispositional and structural underpinnings
of inter-state belief change, a broader agenda is to provide a greater understanding of how
signaling may be received by others. It is common for policymakers to let concerns for
maintaining or changing certain beliefs of other states guide policy, including the decision to
engage in war. As Schelling said, “we lost thirty thousand dead in Korea to save face for the
United States and the United Nations, not to save South Korea for the South Koreans, and it was
undoubtedly worth it. Soviet expectations about the behavior of the United States are one of the
most valuable assets we possess in world affairs” (Schelling 2008). The findings of this study
indicate that short term hostile signaling is effective enough to lead individuals to discount a
harmonious past. However signaling has shown to have less of an effect on beliefs about
intentions than individuals’ own belief systems, which are the most powerful predictor of beliefs
! 43!
about the intentions of other states. A promising avenue to explore would be to assess further the
interaction between political worldviews and costly signaling.
If costly signaling seems to only effective when the signal is hostile, and if individuals
are more inclined to view other states according to their view of the political universe, are
prospects for rapprochement between states dim? Not necessarily. We have seen instances in the
past where signals and indices being evidenced by a state were too costly, too salient, and most
importantly too consistent over a long period of time to be ignored, such as was the case with
Détente with the Soviet Union.

References

Berinsky, Adam J., Gregory A. Huber, and Gabriel S. Lenz. 2012. “Evaluating Online Labor
Markets for Experimental Research: Amazon.com’s Mechanical Turk.” Political Analysis
20 (3): 351–68. doi:10.1093/pan/mpr057.
Duckitt, John. 2001. “A Dual-Process Cognitive-Motivational Theory of Ideology and
Prejudice.” Advances in Experimental Social Psychology 33: 41–114.
———. 1997. “Signaling Foreign Policy Interests Tying Hands versus Sinking Costs.” Journal
of Conflict Resolution 41 (1): 68–90. doi:10.1177/0022002797041001004.
Herrmann, Richard K., and Michael P. Fischerkeller. 1995. “Beyond the Enemy Image and
Spiral Model: Cognitive–strategic Research after the Cold War.” International
Organization 49 (03): 415–50. doi:10.1017/S0020818300033336.
Hopf, Ted. 1994. Peripheral Visions: Deterrence Theory and American Foreign Policy in the
Third World, 1965-1990. University of Michigan Press.
! 44!
Huth, Paul K. 1997. “Reputations and Deterrence: A Theoretical and Empirical Assessment.”
Security Studies 7 (1): 72–99. doi:10.1080/09636419708429334.
Imai, Kosuke, Luke Keele, and Dustin Tingley. 2010. “A General Approach to Causal Mediation
Analysis.” Psychological Methods 15 (4): 309.
James, Patrick. 2009. “Elaborating on Offensive Realism.” In Rethinking Realism in
International Relations: Between Tradition and Innovation, edited by Annette Freyberg-
Inan, Ewan Harrison, and Patrick James, 1 edition. Baltimore, Md: Johns Hopkins
University Press.
Johnson, Dominic D. P., Rose McDermott, Emily S. Barrett, Jonathan Cowden, Richard
Wrangham, Matthew H. McIntyre, and Stephen Peter Rosen. 2006. “Overconfidence in
Wargames: Experimental Evidence on Expectations, Aggression, Gender and
Testosterone.” Proceedings of the Royal Society of London B: Biological Sciences 273
(1600): 2513–20. doi:10.1098/rspb.2006.3606.
Kertzer, Joshua D., and Jonathan Renshon. 2014. “Putting Things in Perspective: Mental
Simulation in Experimental Political Science.” Unpublished Manuscript.
Kriner, Douglas, and Francis Shen. 2014. “Responding to War on Capitol Hill: Battlefield
Casualties, Congressional Response, and Public Support for the War in Iraq.” American
Journal of Political Science 58 (1): 157–74. doi:10.1111/ajps.12055.
Schelling, Thomas C. 2008. Arms and Influence. New Haven, Conn.; London: Yale University
Press.
Tomz, Michael R., and Jessica L. P. Weeks. 2013. “Public Opinion and the Democratic Peace.”
American Political Science Review 107 (04): 849–65. doi:10.1017/S0003055413000488.
! 45!
Trager, Robert F. 2010. “Diplomatic Calculus in Anarchy: How Communication Matters.”
American Political Science Review 104 (02): 347–68. doi:10.1017/S0003055410000158.

! 46!
Appendix A1. Experimental Prompts

Alienation – Low Cost Signal – Low Threat

Alienation – Low Cost Signal – High Threat

Alienation – Medium Cost Signal – Low Threat

! 47!

Alienation – Medium Cost Signal – High Threat

Alienation – High Cost Signal – Low Threat

Alienation – High Cost Signal – High Threat

! 48!

Rapprochement – Low Cost Signal – Low Threat

Rapprochement – Low Cost Signal – High Threat

Rapprochement – Medium Cost Signal – Low Threat

! 49!

Rapprochement – Medium Cost Signal – High Threat

Rapprochement – High Cost Signal – Low Threat

Rapprochement – High Cost Signal – High Threat

! 50!
Appendix A2. Authoritarian Personality and Social Dominance Questionnaire Items

Participants were given a prompt which began with: “Select the answer that best reflects
the extent to which you agree or disagree with each of the following statements:” followed by ten
items with a 5pt Likert scale, where they could indicate one of the following answers: strongly
agree, agree, neither agree nor disagree, disagree, or strongly disagree. Items were coded to
allow for consistency on the AUP scale while inverting the positive and negative orientation of
the questions. The ten items included the following, with a * indicating those that had inverse
coding:

AUP Item 1: The established authorities generally turn out to be right about things, while the
radicals and protestors are usually just “loud mouths” showing off their ignorance.

AUP Item 2: Our country desperately needs a mighty leader who will do what has to be done to
destroy the radical new ways and sinfulness that are ruining us.

*AUP Item 3: Gays and lesbians are just as healthy and moral as anybody else.

*AUP Item 4: Atheists and others who have rebelled against the established religions are no
doubt every bit as good and virtuous as those who attend church regularly.

AUP Item 5: The only way our country can get through the crisis ahead is to get back to our
traditional values, put some tough leaders in power, and silence the troublemakers spreading
bad ideas.
! 51!

*AUP Item 6: Our country needs thinkers who have the courage to defy traditional ways, even if
this upsets many people.

AUP Item 7: The “old-fashioned ways” and the “old-fashioned values” still show the best way
to live.

*AUP Item 8: Some of the best people in our country are those who are challenging our
government, criticizing religion, and ignoring the “normal way things are supposed to be done.”

*AUP Item 9: A “woman’s place” should be wherever she wants to be. The days when women
are submissive to their husbands and social conventions belong strictly in the past.

*AUP Item 10: There is no “one right way” to live life; everybody has to create their own way.

*SDO Item 1: Some groups of people are simply inferior to other groups.

*SDO Item 2: In getting what you want, it is sometimes necessary to use force against other
groups.

*SDO Item 3: It isokay if some groups have more of a chance in life than others.

*SDO Item 4: To get ahead in life, it is sometimes necessary to step on other groups.
! 52!

*SDO Item 5: If certain groups stayed in their place, we would have fewer problems.

*SDO Item 6: It is probably a good thing that certain groups are at the top and other groups are
at the bottom.

*SDO Item 7: Inferior groups should stay in their place.

*SDO Item 8: We should do what we can to equalize conditions for different groups.

*SDO Item 9: We would have fewer problems if we treated people more equally.

*SDO Item 10: We should strive to make incomes as equal as possible.

! 53!
Appendix A3. Scree Plots for Authoritarian Personality and SDO Factors

Authoritarian Personality Scree Plot

! 54!
Social Dominance Orientation Scree Plot

! 55!

Appendix A4. Factor Loadings for Authoritarian Personality and Social Dominance
Orientation Factors

Authoritarian Personality Factor Loadings
Questionnaire Item Factor Loading (Weighted Least Squares)
AP Item 1 0.62
AP Item 2 0.76
AP Item 3 0.79
AP Item 4 0.78
AP Item 5 0.80
AP Item 6 0.77
AP Item 7 0.79
AP Item 8 0.79
AP Item 9 0.61
AP Item 10 0.70

Social Dominance Orientation Factor Loadings
Questionnaire Item Factor Loading (Weighted Least Squares)
SDO Item 1 0.81
SDO Item 2 0.73
SDO Item 3 0.77
SDO Item 4 0.75
SDO Item 5 0.79
SDO Item 6 0.86
SDO Item 7 0.84
SDO Item 8 0.77
SDO Item 9 0.76
SDO Item 10 0.62

! 56!
Appendix A5. Bootstrapped Least-Squares Estimation

!
In this instance, parameter estimates are the same as OLS, but bootstrapping generally widens
confidence intervals. Still, effect sizes and hypothesis tests remain the same regardless of how
we conceptualize the dependent variable.

! 57!
Appendix A6. Bootstrapped Ordered Probit Estimation

In this instance, parameter estimates are the same as GLM, but bootstrapping generally widens
confidence intervals. Still, effect sizes and hypothesis tests remain the same regardless of how
we conceptualize the dependent variable.

!
!
!
!
!
!
!
!
!
!
!
!
!
!
! 58!
Chapter 3

Enmity is What Leaders Make of It: The
Microfoundations for Perceptions of Adversaries

Although a great deal of scholarship in international relations reflects on
beliefs about adversaries, little is understood about how individuals differ
in these views as the adversary’s behavior changes over time. I
hypothesize that an individual’s image of the political universe should
interact with structural factors to determine complexity of thought
regarding adversaries. To assess this hypothesis, I employ an automated
quantitative content analytical technique to assess US legislators’ beliefs
about the Soviet Union between 1985 and 1991. I find that when structural
conditions were in line with legislators’ ideological expectations,
complexity of thought about the Soviet Union increased. Conversely, when
structural conditions did not accord with ideological expectations,
legislators became more simplistic in their view of the Soviet Union.

Near the end of the Cold War, leaders in the United States began to seriously
reconsider how vigilant they should be in dealings with the Soviet adversary. Generally
speaking, the debate arose due to the Soviets' claims of a softened orientation toward the
West, which coincided with their making shifts toward economic and political
liberalization, drawing back missiles in Europe, withdrawing from Afghanistan, and
perhaps most importantly ending the Brezhnev Doctrine. While initially U.S.
policymakers were skeptical of the Soviet shifts resulting in lasting change, repeated
overtures by Mikhail Gorbachev combined with political instability within the Soviet
Union eventually began to convince some in the U.S. that an opportunity existed for more
peaceful bilateral relations.
! 59!
This opportunity seemed to be most apparent to Democrats, who (at first, quietly)
introduced legislation that represented an unprecedented shift in Cold War policy: direct
but conditional material aid to the Soviet Union. The Republicans chided their
colleagues for their naivety, questioning the sincerity of the Soviets' stated shift in
posture and claiming that the Soviets could not be trusted to use the aid in good faith.
This dynamic is encapsulated by a comment by then-Congressman Dick Cheney: “[The
Democrats] have seen some changes and apparently believe everything now is just fine.
It is as if they had decided to give away their overcoats on the first sunny day in January”
(Broder 1989).
! The dynamics underlying the formation and change of perceptions of other states,
particularly with regard to adversaries, have been engaged by various realist scholars and
political psychologists. However, save for a few exceptions, there has been little
systematic effort to assess the microfoundations of this type of belief updating. For
instance, both image theorists and deterrence theorists similarly assume that images and
reputations are intersubjectively shared, and that belief change and formation occurs
similarly for all individuals. Others assume that updating is driven by subjective beliefs
about what forms of signals are credible, and organizational motivations. One of my
goals in this paper is to advance a hypothesis that explains which actors are in general
more likely to update their beliefs about adversaries, with an eye towards maintaining
grounding in a cognitive framework.
To explain substantive individual-level variation in images, I refer to
psychological literature on cognitive complexity, which has been linked to political
ideology. Narrowing the focus to adversary images as an ideal type, I will explore the
! 60!
degree to which decision-makers images of the enemy are more-or-less complex. I will
suggest that greater complexity in an enemy image is indicative of the degree to which
the adversary’s behavior fits an individual’s worldview. I hypothesize that under
conditions where an adversary is sending image-inconsistent signals, conservative
decision-makers will hold less complex (in other words, more simplistic) enemy images
than their liberal counterparts, and that this difference in sensitivity to signaling can
account for variation over time. When the adversary sends image consistent signals, on
the other hand, conservative decision-makers will become more nuanced in their thinking
of adversaries.
To assess the current hypothesis, I employ a quantitative content analytical
technique to assess the statement complexity of U.S. Senators regarding the Soviet Union
in the winter of 1985 and the summer of 1991. I find that political ideology is a
significant determinant of the complexity of enemy images, and that there is an
interaction effect between political ideology and structure, such that when the adversary
is exhibiting hostility, the relationship between conservatism and complexity is positive,
and when the adversary is not exhibiting hostility, the relationship between conservatism
and complexity is negative. This suggests that when the adversary behaves as decision-
makers expect, they become more nuanced in their views about that adversary, while
when the adversary behaves in ways that are not expected, decision-makers become more
simplistic in their views of that adversary.
The ensuing sections are organized as follows. First, I open with an exposition of
realist and psychological approaches to beliefs about other states, highlighting a lack of
firm microfoundations and some diversity in the use of similar concepts. Next, I
! 61!
critically engage the cognitive and social psychological literature to develop and assess a
hypothesis involving systematic variation in enemy images. Finally, I close with a
discussion regarding how the sources and outcomes of this variation have implications
for academic theorizing in international relations and practical thought in foreign policy.

Images of the Political Universe

Given recent advances in psychology involving differences in how individuals
manage information, I seek to assess the microfoundations of images. More specifically,
I am concerned with whether some individuals are more or less inclined to hold extreme
images of other states, beginning in this study with the enemy image under conditions
where the putative adversary is demonstrating information inconsistent with its putative
ideal type. My task in this section is to review research on the psychology of political
ideology, and in particular how it is related to cognitive complexity. I will argue that
political ideology should be linked to the complexity of decision-makers' enemy images.
Jost et al. (2009) provide the most comprehensive review to date of the
comparative psychological differences between liberals and conservatives. They employ
Erikson’s (2003) definition of political ideology as a value-neutral “set of beliefs about
the proper order of society and how it can be achieved”, and describe the notion of the
left-right dimension as involving the split between acceptance and resistance of social
change and inequality. They claim that this dichotomy is robust across cultures, and that
its economic and social components are correlated to the point of near-redundancy.
While ideology has a socially constructed aspect, they claim “the dimensional structure
and attitudinal contents of liberalism and conservatism...stem, at least in part, from basic
! 62!
social psychological orientations concerning uncertainty and threat” (311). They cite
experimental studies linking conservatism with death anxiety, beliefs about system
instability, fear of threat and loss, dogmatism, tolerance of ambiguity, and personal needs
for order and closure, and other studies linking liberalism with openness to new
experiences, cognitive complexity, tolerance of uncertainty, and self-esteem.
More importantly for the purposes of this study, political ideology has been linked
to views about out-groups, such that conservatives tend to see out-groups as more
threatening and dangerous to social cohesion, order, values, and stability, and to
categorize the social world into “us,” who are good, normal, moral, decent people, versus
“them,” who are bad, disruptive, immoral, and deviant (Duckitt 2001). In other words,
conservatives are more likely to be vigilant in their dealings with out-groups, as the
political universe in general is seen as threatening since all other states are considered
out-groups. Meanwhile liberals tend to see the world in fundamentally the opposite
manner, and accordingly believe that the political universe is fundamentally harmonious
and that conflicts generally arise due to a failure to properly communicate.
There is an underlying tension in focusing on political ideology as the main
explanatory variable after having emphasized the role of schemas like AP and SDO in
Chapter 2. The goal in this chapter is not to establish political ideology as holding
equivalent explanatory power to AP and SDO, but rather to determine whether it is an
efficient way to understand and anticipate elites’ thought processes regarding signaling
behavior in a rapprochement context. The results in Chapter 2 indicate that while
ideology is indeed much weaker in explanatory power when compared to AP and SDO,
! 63!
however it does show to be an effective predictor in the context of rapprochement in the
domain of policy preferences.
In studying the beliefs of decision-makers, political psychologists often have to
make compromises in assessing elite beliefs and attitudes. I would be hard pressed to
generate a factor score for any Senator in 1985, for instance. But the benefit of
employing political ideology in this study is that it is that it is an easily measurable
component of a person's attitude or worldview. In the area of foreign policy, elites often
self-identify according to some ideology. If they do not self-identify, there are reliable
ways to measure their position on the left-right spectrum. The task is even easier with
members of congress, who have voting records that can be used to infer their political
ideology. American political action groups seem to believe that representatives’ voting
records can effectively be used to infer their ideological stances. For instance, the
Conservative Review uses representatives’ votes to give them a “Liberty Score” which is
an indication of their alignment with the conservative ideology.
The central argument here is as follows: when faced with an adversary does not
behave in accordance with its ideal type, i.e. an adversary attempting to signal
reassurance, conservatives will tend to exhibit lower complexity in thought than liberals.
When the adversary does behave in accordance with its ideal type, i.e. an adversary
behaving with hostility, there should be no significant differences between liberals and
conservatives.

I formalize the argument within the Cold War rapprochement context with the following
hypotheses:
! 64!

H1: If the Soviet Union is exhibiting hostility towards the US, there will be
no relationship between conservatism and complexity of views of the Soviet
Union.

H2: If the Soviet Union is not exhibiting hostility towards the US, there will
be a negative relationship between conservatism and complexity of views of
the Soviet Union

The primary explanatory variable in this study is political ideology, which will be
operationalized using Poole and Rosenthal's (2001) DW-NOMINATE (dynamic,
weighted, nominal three-step estimation) scores, which include a score on a liberal-
conservative scale based strictly on voting behavior with values ranging from -1 to 1 for
any given year of congress. The score is based on a probabilistic voting model wherein
representatives' “yea” or “nay” votes are used to calculate their distance from the center,
which is itself calculated based on “cutting lines” across parties (Everson, Valelly, and
Wiseman 2009). The benefit of this indicator is its reliability, although there may be
some limitations on its external validity.
To rule out competing causes, I will control for several variables. First, as
mentioned earlier, accountability in a given domain can affect situational conceptual
complexity. Bearing in mind Tetlock's (1985) definition of accountability as “pressures
to justify one's causal interpretations of behavior to others,” accountability in
international relations involves responsibility over foreign policy. In the context of the
! 65!
U.S. Senate, it is true that all members have some degree of foreign policy accountability,
but their relative levels of accountability are not equal. A senator's status as member of
either the armed services or foreign relations committees would qualify him or her as
having relatively high accountability (compared to non-members of those committees) in
foreign policy. These individuals know that the President and other top officials will pay
attention to their input, that what they say might actually make a difference in foreign
policy, and that the attentive public expects them to have a higher degree of expertise in
foreign policy than non-foreign policy members. Therefore, they should feel pressure to
make sound foreign policy decisions and exhibit nuanced cause-effect attributions
involving other states.
I treat accountability as a dichotomous variable, with values being either high or
low, and operationalize it based on membership on either the armed services or foreign
relations committee. Representatives who are not on either of these committees will be
considered low accountability individuals in foreign policy. Granted, the nature of
accountability will change depending on the issue area. For example, members on
budgetary committees may at times become held to a higher level of accountability in
foreign policy depending on the issue being debated. In general, however, members of
the foreign affairs and armed services committees can be considered high accountability
on most foreign policy issues.
I will also control for political party using a dummy variable where 0 = democrat
and 1 = republican, and for education using a three-point scale, such that 0 = high school
graduate; 1 = college graduate; and 2 = postgraduate degree. Age and length of service in
the Senate will also be controlled for, measured by years. Another control variable is a
! 66!
target state's own efforts to manipulate its image. States have the ability to deliberately
manipulate their image through signaling, manipulation of indices, exaggeration, or
misrepresentation (Jervis 2015), which may affect the collectivity of their images. By
selecting cases in involving the same target state during the same time period, this
variable is effectively controlled for since all decision-makers are making attributions
involving the same actor, given similar information.
Age is an important control variable since a decision-maker's historical context
informs analogical reasoning. Since decision-makers have been shown to draw on
historical analogies as a way of dealing with uncertain situations (Houghton 1996), I
expect that decision-makers' respective histories in dealing with a target state might
affect their images of it. While age is not a perfect operationalization, given that
individuals often have disparate personal experiences in the same time period, I expect
that a seventy-year-old's understanding of the Soviet Union in 1991 would systematically
differ significantly from that of a thirty-year-old. For instance, living as an adult through
the Cuban missile crisis might make an individual more vigilant in their dealings with the
Soviet Union.
The dependent variable in this study is the degree to which enemy images
resemble simplified schemata. The more complex and nuanced the enemy image, the less
it should resemble a parsimonious ideal-type. The task of measuring the complexity of
individually held images is formidable given that cognitive schema cannot be directly
measured, thus I operationalize complexity of images by assessing the complexity of
verbal statements made regarding a target state. The best available tool for this task with
the highest possible degree reliability is computer-automated qualitative content analysis.
! 67!
In 1999 Margaret Hermann, in collaboration with Social Science Automation, Inc.,
released a coding scheme for ProfilerPlus, a text coding and analytic platform. This
coding scheme, dubbed Leadership Trait Analysis (LTA), quantitatively assesses and
codes text for several cognitive components of leadership style, including: belief in
ability to control events, need for power and influence, conceptual complexity, openness
to contextual information, self-confidence, task orientation, distrust, and in-group
bias/distrust of others (Hermann 2005, 11).
The complexity variable at issue here is closely matched by the LTA system's
conceptual complexity variable. Hermann's LTA manual refers to conceptual complexity
as “the degree of differentiation which an individual shows in describing or discussing
other people, places, policies, ideas, or things. The more conceptually complex
individual can see varying reasons for a particular position...” (Ibid., 22). The LTA
system scans particular words to code for conceptual complexity, such as:
“approximately,” “possibility,” and “trend”; while other words decrease a statement's
complexity score, such as: “absolutely,” “without a doubt,” “certainly,” and
“irreversible” (Ibid., 30). It may be the case that, given the verbiage, the conceptual
complexity coding system is actually capturing an individual’s degree of certainty in
discussing a concept, and not the degree of complexity. But in the context of legislative
statements regarding foreign policy, I argue that certainty is equivalent to complexity.
Individuals with a high degree of knowledge, expertise, and nuance in the area of foreign
policy generally do not think in absolutes but rather probabilistically, and those who rely
on absolutes are also those who leverage simplified stereotypes. The world policymakers
seek to describe is always complex, and complexity can only be described uncertainly.
! 68!
I use the LTA coding schemes in ProfilerPlus to assess speeches given by
members of the U.S. Congress, for two reasons: first and most importantly, their images
of other states influence the direction of U.S. foreign policy. Not only are their collective
images of other states causal antecedents to U.S. foreign policy behavior given the
reactions they anticipate from other states, but also their discourse serves in part to frame
the general narrative involving how other states should be perceived. Second, members
of Congress spend time on the floor making lengthy and completely transcribed
statements regarding foreign policy, often centering on the anticipated behavior of other
states.
While the conventional wisdom says that the president dominates in foreign
affairs, Carter and Scott (2009) have argued that members of congress act as foreign
policy entrepreneurs who proactively engage foreign policy issues in order to shape
policy by means beyond writing their preferences into law, such as holding hearings,
engaging executive branch officials directly, and writing op-eds. The authors find that
members of congress who take on the role of foreign policy entrepreneurs are indeed
effective at exerting influence over many aspects of foreign policy, such as broader
questions of war and peace, to foreign aid, to the activities of foreign intelligence, and
even the activities of policy agencies within the executive branch.
I narrow my focus to the perceptions of U.S. Senators and members of the House
of Representatives involving the Soviet Union at the end of the Cold War, specifically
from 1985 to 1991. At this point in time, the Soviet Union satisfied Herrmann and
Fischerkeller's ideal-typical criteria for enemy images, but a number of image-
disconfirming changes had occurred that required U.S. policymakers to reconsider
! 69!
longstanding policy strategies for dealing the Soviet Union. The Soviets were not only
struggling with internal turmoil, but also had a new leadership that was calling for
increased government transparency and market reforms, in addition to calling for thawed
relations with the West (CQ Almanac 1991). In effect, US decision-makers had to decide
whether or not to respond to the Soviets' apparent relative losses and policy overtures
with a softened stance.
The sample of statements is drawn in two phases. The first sample is drawn from
the winter of 1985, between January 1 and March 1, immediately before the ascension
reform-minded Mikhail Gorbachev, and immediately following a phase of deterioration in
relations between the Soviet Union and the United States arising from the Soviet invasion
of Afghanistan five years earlier. The second sample is drawn from the summer of 1991,
between May 1
st
and August 31
st
, during which the Soviet Union was dissolving and
beginning to initiate democratic and market reforms. At the same time, congress began
to debate whether the Bush administration was handling the Soviet transition effectively.
Some members argued that the President was not doing enough to facilitate the transition,
while others argued that any support given to Soviet Union would bolster an inherently
hostile regime. Academics and some Soviet economists discussed the notion of a “grand
bargain,” where the US would provide economic aid in return for liberalizing Soviet
reforms. The idea was picked up by the national media and was subsequently discussed
by members of the House and Senate (CQ Almanac 1991).
The key question under discussion was: “what can we expect the Soviets to do
with direct economic aid from the United States?” which carried the implicit question of
whether the Soviet Union had decreased in its hostility toward the U.S. Put differently,
! 70!
central to this debate was whether the Soviets still represented the same extreme image of
a hostile enemy that they had for decades. If they did, it would not be reasonable to offer
direct economic aid, however stringent the conditions.
For this study, the Congressional Record was scanned to identify as many
statements as possible on the Senate and House floors where the subject was the Soviet
Union. 795 unique, paragraph-length (between 50 and 150 words) statements were
identified and individually processed by the ProfilerPlus LTA scheme. Of those
statements, 45% were from Republicans while 55% were from Democrats. Ideologically,
the two parties were polarized in terms of DW-NOMINATE scores, with only 20
members being more than a full standard deviation away from the party mean toward the
center.

Method of Analysis
Hypothesis testing with quantitative data requires selecting an appropriate model
with valid assumptions. It has been demonstrated that standard ordinary least squares
techniques often give inaccurate confidence intervals even where there are small
departures from normality (Wilcox and Keselman 2003). It is commonly taught that the
limit theorem obviates this concern, however in applied work we can never know how
large our sample sizes must be to validate the assumption of normality. Many political
scientists have dealt with this by employing nonparametric techniques such as Bayesian
modeling, but Bayesian inference for applied research is controversial even in the field of
statistics (Gelman 2008).
! 71!
A modern frequentist approach to nonparametric modeling is growing in
popularity, called robust statistics, and has been found to give accurate confidence
intervals under a variety of situations, including non-normality. Some robust techniques,
such as the increasingly popular bootstrapped confidence intervals in regression,
constitute a harder test than conventional methods. Other robust methods can increase
statistical power, which is understood as the probability of correctly rejecting the null
hypothesis when it is indeed false.
The first step in determining whether robust methods can be useful is exploratory
analysis of data. We must begin by determining whether data is normally distributed.
Figures 1 and 2 show a histogram of the Senate distribution of DW-NOMINATE scores
in 1985 and 1991. It is clear that these distributions are non-symmetrical, skewed, and
apparently bimodal. Since we do not know how large our sample size must be for the
central limit theorem to apply, employing conventional OLS and its associated parametric
assumptions does not seem ideal. Instead, I employ robust estimation by way of robust
standard errors clustered my members of congress, which constitutes a harder test than
traditional OLS alone.

! 72!
Figure 1. Senate Distribution of DW-Nominate Scores in 1985
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
!
! 73!
Figure 2. Senate Distribution of DW-Nominate Scores in 1991
!
!
!
! There are a number of issues in this methodology that need to be addressed. First
is the issue of whether the ProfilerPlus LTA value for conceptual complexity sufficiently
operationalizes my dependent variable: the complexity of enemy images. An argument
could be made that what this technique is actually capturing is an individual's overall
level of complexity that is evident in all statements, regardless of the topic. Indeed, the
psychology literature cited earlier supports the claim that individuals vary in their general
level of complexity, and Tetlock’s study involving the statements of U.S. Senators
operated at this level of generality. One of the factors that differentiates my design from
! 74!
his, in addition to content analytic techniques and choice of statistical model, is the
selection of a particular subcategory of statements that involves a single other state.
While there may be generalized differences that hold consistently across party lines, I
also expect that the gap in the complexity of images will vary over time, depending on
the stimuli provided by a target-state. As I will attempt to demonstrate in the next
section, there is provisional evidence that the complexity of images of the Soviet Union
held by members of the Senate were more similar in earlier years of the Cold War, and
began to diverge sharply in the period following Gorbachev's ascendance in Soviet
Politics.
Another potential issue is that staffers often “ghostwrite” Senators’ statements.
The ProfilerPlus manual in fact suggests prioritizing impromptu statements over prepared
ones. Unfortunately, the amount of unprepared statements from that period is limited
when compared to the abundance and accessibility of the Congressional Record.
However, I believe it is fair to assume that Senators have vetted their speechwriters and
are satisfied with the speeches that are written on their behalf. With that said, there may
still be issues with speechwriters' cognitive peculiarities influencing the complexity
evident in enemy images.
Some might dispute the construct validity of using ProfilerPlus to measure the
complexity of images, as it was not specifically designed for image complexity but rather
a more general conceptual complexity. However, the criterion the system uses, which
involves a bank of verbal indicators of nuanced versus absolutist thinking, is similar to
what one would use in creating a manual coding scheme for measuring enemy images,
although it likely misses some verbal nuance that can only be identified through the
! 75!
relationship of words, rather than their mere presence. In any case, I am concerned here
with relative variation, not absolute, thus the lack of contextual nuance likely has limited
implications.
A pitfall of selecting a dyad involving major adversaries is that the proclivity of
individuals on the left and right to have more or less complex enemy images may be
epiphenomenal to the ideational content of political ideology itself, or even to electoral
considerations. Typically the left is associated with foreign policy doveism while the
right is associated with hawkism. If I do find a difference between the left and right in
terms of the complexity of images, it may be due to these factors. Indeed it is difficult,
but not impossible, to disentangle underlying interests and images. It seems that
controlling for political party identification, alongside ideology, should mitigate this
issue substantially. Thus, we might expect updating dynamics for center-right versus far-
right Republicans to differ.

Results
Estimation for all results is conducted by way of robust linear regression with
clustered standard errors
4
. The results for the period of 1985 are presented in Model 1 within
Table 1. In this case, the coefficient for political ideology is positive and statistically
significant, and possesses a high magnitude such that moving from the left to right accounts
for a roughly 25% increase in complexity of statements about the Soviet adversary. Thus, in
this period where the Soviets were displaying hostility to the U.S., individuals who were
more conservative exhibited a greater degree of conceptual complexity in discussing the
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
4
The findings are cross-validated via other techniques like Maximum Likelihood Estimation with Fixed
Effects, bootstrapped confidence intervals, MM- and S-estimation in Appendix A, and elasticity tested in
Appendix B.
! 76!
Soviet Union. Thus I find no support for H1, the expected null finding for the period of
1985. This surprise directional finding where I expected a non-finding suggests a different
intuition which will discuss in the conclusion of this chapter. The effect of identification
with the Republican party has a statistically significant but weak negative effect on
complexity, as does years of education, while the number of statements has a significant but
weak positive effect on complexity.

Table 1. Conceptual Complexity of Enemy Images in 1985 and 1991

Variable
Model 1: 1985 Data Model 2: 1991 Data
Ideology 0.12 (0.04)* -0.15 (0.08)+
Party ID -0.07 (0.03)* 0.17 (0.06)*
Accountability 0.00 (0.03) -0.02 (0.03)
Tenure (logged) 0.02 (0.03) 0.05 (0.02)*
Age -0.00 (0.00) -0.00 (0.01)
Education -0.06 (0.03)* -0.07 (0.03)*
Statements (logged) 0.05 (0.02)* 0.02 (0.02)
Population (logged) -0.01 (0.01) -0.01 (0.01)
Senator -0.03 (0.03) 0.00 (0.00)
Constant 0.69 (0.29)* 0.86 (0.21)

R
2
= 0.10 R
2
= 0.03
Notes: Standard Errors in Parentheses; + p < .10, * p < .05

The results for 1991 are presented in Model 2 within Table 1. Here, the coefficient
for political ideology is negative and statistically significant, and possesses a high magnitude
such that moving from the left to right accounts for a roughly 30% decrease in complexity of
statements about the Soviet adversary. In other words, following the ascension of Gorbachev
and his attempts to thaw relations with the U.S., and following the initial signs of collapse of
! 77!
the Soviet Union, individuals who were more conservative exhibited a lesser degree of
conceptual complexity in discussing the Soviet Union. The coefficient for party is once
again significant and the sign has flipped, years of education remains significant and weakly
negative, and the effect of tenure has become statistically significant and positive in this
period.

Longitudinal Cross-Case Comparison: Assessing Belief Revision Over Time
To assess the quantitative findings’ validity by my own qualitative assessment, I
select two cases to ensure variation on the independent variable, DW-NOMINATE score,
while being matched for committee membership, age, tenure, and absolute value on the
ideological scale. To avoid selecting on the dependent variable, I have selected cases based
only upon the independent and matching variables, without regard for the members’ policy
positions regarding the Soviet Union. Senators John Warner, a Republican, and James Exon,
a Democrat, meet these criteria. In 1991, both were members of the Armed services
committee, had a tenure of twelve years in the senate, and were within six years of age.
Furthermore, they were both roughly equidistant from the center in terms of their DW-
NOMINATE score with Warner rated as .25 and Exon rated as -.20, an absolute difference of
.05, which constitutes a 2.5% difference on the -1 to 1 scale.
I proceed by taking two “snapshots” of the statements made by both Senators, one in
1985 and the other in 1991. In 1985, Gorbachev had recently assented to power and had not
yet had time to establish his “new thinking” in the eyes of US policymakers. At the outset of
his term, he and President Reagan began the first of a series of summits to discuss arms
control. At this time, congress had almost unanimously consented to continue funding for
! 78!
the MX missile, which was generally agreed to be a key bargaining chip in the talks. At this
point in time, liberals in congress avoided challenging Reagan on issues that involved
confrontation with the Soviets. By the end of the year, Gorbachev and Reagan agreed in
principle to a 50% cut in nuclear weapons, and developed a closer personal relationship than
had previous US and Soviet leaders. On the other hand, the US continued to deal with issues
of Soviet espionage, with a number of cases being brought up against military specialists for
selling secrets to the Soviet Union.
The apparent Soviet posture towards the US had changed dramatically by 1991. At
this point, Gorbachev had established himself as a different type of Soviet leader, had
enacted a number of policies which signaled, at least for the time being, a reorientation of
Soviet policy towards the United States, as well as internal policy in the direction of
openness, economic liberalization, and democratization. US decision-makers essentially had
an unprecedented situation with the Soviet Union, given essentially new stimuli from which
to draw inferences about them. In what follows, I will assess statements by Warner and
Exon side by side, comparing key statements from each. To collect statements, I have
scanned the Congressional Record during the MX missile debate in 1985 and the aid-for-
reform talks in 1991, scanning all statements made by each Senator for specific attributions
regarding the Soviet Union, specifically concerning the Soviet Union’s future behavior.
Figures 3 and 4 compare key statements for both senators, which I follow with some
interpretation.

! 79!
Figure 3. 1985 Debates on MX Missile Funding
Exon (D)

May 22, 1985 – National Defense
Authorization Act for Fiscal Year 1986

“The Soviet Union poses a serious challenge
to our national interests and security and,
accordingly, authorized and appropriated
large increases in defense spending...The
Threat from the Soviet Union and its allies
has grown; it exists; it must be countered.
This does not mean that we have to match
this threat missile for missile...The Soviets
may be 6 feet tall, but that does not mean
they are 10 feet tall. They have weaknesses
and disadvantages that our policies...can
exploit.”

“The last several decades have seen a
tremendous Soviet arms buildup across the
spectrum of military capabilities. This
strengthened power in conjunction with an
aggressive Soviet foreign policy is very
worrisome...We have bought greater measures
of protection and insurance, but the Soviets are
still charging ahead and the threat remains.
Although we would prefer a static situation of
equivalence, the Soviets are bent on
conventional and nuclear superiority.”
June 4, 1985 – National Defense Authorization
Act for Fiscal Year 1986
“SDI will also serve as a hedge for our country
against a possible Soviet breakout from current
or future arms control agreements. We must all
remember that the President has reported to
Congress that the Soviet Union has violated the
Helsinki Final Act, the Geneva protocol on
chemical weapons, the biological and toxin
weapons convention and two provisions of the
SALT II Treaty, specifically telemetry
encryption and ICBM modernization...We
cannot stand idly by and continue to tolerate
Soviet arms control treaty violations. The
Warner (R)

May 21-23, 1985 – National Defense
Authorization Act for Fiscal Year 1986

“This Nation will continue in the next decade
in the shadow of a significant imbalance in
strategic force capabilities that threatens
important elements of our retaliatory
capability while affording the Soviets a
virtual sanctuary for their own.”

“The Soviet leadership perceives that they
can engage in crisis bargaining – in the Third
World or elsewhere – with the expectation
that the United States will back down in the
face of significant asymmetries in important
strategic capabilities.”

“This asymmetry favoring the Soviet Union
would be a powerful disincentive for them to
negotiate in good faith with the United States
in Geneva.”
June 4, 1985 – National Defense Authorization
Act for Fiscal Year 1986
“The Soviets returned to the negotiating table
primarily as a result of the enhanced SDI
Program and ballistic missile defense work,
and at this time to gut the program would be
simply to jerk the chairs from beneath our
negotiators and, indeed, the Soviets could pack
up and go home having won their victory over
SDI.”

! 80!

Not surprisingly, given the broad bipartisan support for the MX missile program, both
Senators support providing it funding. The ways in which they describe the Soviet Union are
also quite similar. Both focus on maintaining a strategic military advantage over the Soviet
Union, which carries the implication that they believed deterrence on the part of the US was
the only way to keep the Soviet Union at bay. Both maintain the importance of the MX
program as a key bargaining chip, which implies that they believe that the Soviet Union
would be aggressive and strategic in negotiations over nuclear weapons at the Geneva
Summit. The key difference I find is that Warner takes more time to emphasize Soviet
defense spending, and touches on some of the specific weaknesses within the Soviet Union.
However Exon does allow for slightly more ambiguity in his assessment of the Soviet Union
than Warner, who focuses predominantly on Soviet hostility and aggressive bargaining
tactics. In both cases, the Senators agree that the United States must remain vigilant and
maintain military superiority over the Soviet Union in order to successfully negotiate over
arms reductions.

strategic defense initiative research and
development will further tell the Soviet Union
that they had better get into compliance and
negotiate seriously on future arms reductions.”
! 81!
Figure 4. 1991 Debates on Aid for Soviet Reform
Exon (D)

May 8, 1991 – “Observations on the Soviet
Union”

“The current political landscape of the Soviet
Union cannot be sketched in black and white
terms. The situation is complex with several
forces pulling at the political leadership of
the nation.”

“While the Gorbachev-Yeltsin controversy
occupied center stage, it was obvious that
there are several forces simultaneously at
work in the Soviet Union including
democratic forces, a restless labor movement,
nationalist and secessionist movements, and
the military industrial complex, the
Communist Party and the Government
bureaucracy all trying to define the new
Soviet Union.”

“Since 1919, the Communist Party and its
ideology has been the glue that has held the
vast Soviet Nation of diverse republics
together. That glue is rapidly disintegrating.
Communism has been discredited. The
Soviet people have lost faith in the party. It
presently appears that the Soviet Union lacks
a coherent belief system. Unlike the new
democracies of Eastern Europe where
economic reforms have been real and
democratic, the movement from a command
economy toward a market economy in the
Soviet Union, has been thus far insufficient
to engender the confidence, hope or support
of the people. During the early phases of
perestroika, the Soviet people were grateful
to Gorbachev for glasnost. Today, the painful
economic reforms have failed to convert the
economy to efficiency and have turned the
man on the street against Gorbachev. With
the freedoms of glasnost citizens are able to
voice their frustrations and they are.”
Warner (R)

July 31, 1991 – National Defense
Authorization Act

“The important promises made to the world
by the Soviet leaders have not yet been
translated...into concrete and irreversible
changes in their military industrial complex.
That remains the single most difficult
challenge of President Gorbachev and indeed
the several Presidents of the Soviet republics,
as they begin to grapple their economic
problems, to bring under control the
awesome strength and momentum of the
Soviet military- industrial complex.”

August 1, 1991 – National Defense
Authorization Act

“There must be confidence that the new
program meets the right threat efficiently and
does not undermine old-fashioned deterrence
or play into the hands of Soviet hard-liners.”

“I have stated many times, on the Senate
floor, in the press, and in the public, that I
welcome and applaud the historic changes in
the Soviet Union. We must continue to do
whatever we can to encourage the further
democratization of the Soviet Union. I have
also cautioned, however, that many of the
important promises made to the world by
Soviet leaders have not yet translated into
concrete and irreversible change. Mr.
President, in this context, we cannot forget
the significant and formidable nuclear
ballistic missile capability of the Soviet
Union. As the Chairman of the Joint Chiefs
of Staff, Gen. Colin Powell, stated in recent
testimony before the Senate Armed Services
Committee: 'the Soviet Union remains the
one country in the world with the means to
destroy the United States in a single
! 82!
“It is clear that President Gorbachev's
political future ebbs and flows on a rapidly
changing sea. While Gorbachev has been a
leader who the United States has been able to
work with, American policy with regard to
the Soviet Union should not be based solely
on the political survival of one personality.
Treaties, trade agreements, and exchange
programs are between nations, not individual
presidents. Our policy toward the Soviet
Union should attempt to look beyond the
horizon and beyond President Gorbachev.
Our policies should also be cognizant of the
political authority of the Soviet Republics. In
the future, it is clear that the individual
republics will come to wield more authority
than the past.”
“America must continue to exercise caution
in dealing with the Soviet Union and be
certain that Soviet Government actions match
the words of Soviet political leaders.
President Reagan appropriately adopted as a
motto for the United States-Soviet military
relations, “Trust but Verify.” Overall, I was
somewhat encouraged that from reformers
and hardliners alike, there was no evidence
of an interest in returning to the bad old days
of the cold war.”

June 6, 1991 – “Areas for United States-
Soviet Cooperation”

“The Soviet Union is undergoing rapid and
dramatic change. The economic, political,
and social underpinnings of Soviet society
are crumbling and that nation is scrambling
to rebuild and reshape its very foundation.
This is no easy task for a multiethnic nation
which covers one-sixth of the globe.”

“In light of the dramatic changes which have
occurred thus far in the Soviet Union there
are several areas where cautious but
constructive cooperation is appropriate.
These areas emphasize the mutual peaceful
aspirations of the United States, the Soviet
devastating attack.”

“Soviet strategic defenses continue to be
upgraded, despite budget reductions. The
Moscow ABM system is nearing full
operational capability, and strategic SAMs
and fighter interceptors are being upgraded.
The Soviets are continuing work on deep
underground facilities for leadership
protection and command continuity.”

! 83!
Union, and the world community.”

“I am...concerned that in spite of reductions
in other areas of military activity, the Soviet
Union continues to modernize their strategic
forces. As such, the United States has no
option but to continue our modernization
program.”

“Given the political tension I observed in the
Soviet Union, food shortages could unleash a
series of reactions and emotions within that
country which could further undermine any
movement toward reform.”

July 31, 1991 – National Defense
Authorization

“We judge the current Soviet leadership to
be less concerned with world domination, to
be more inward focused in an effort to rescue
the faltering Soviet economy. Because we
assess the Soviets as more inward focused we
ourselves have made several adjustments to
our own nuclear posture, reflecting the
changed assumptions about Soviet
intentions.”

“We must keep in place necessary strategic
forces, and infrastructure to provide
deterrence of even a hostile Soviet leader for
we may not have the lead time necessary to
expand our strategic forces should the Soviet
leadership and their positions suddenly
change.”

“But we can and should react to varying
assessments of the Soviet intentions, standing
down systems, and saving defense dollars
when we judge the Soviet intentions to be
benign but ready to rapidly increase our
deterrent capability should their intention
appear to shift.”

August 1, 1991 – National Defense
Authorization Act
! 84!

“I am not ready to say...that America is going
to ignore the continuing research and
development, and that we know going on
inside of the Soviet Union is some type of
SDI, whatever their version in the Soviet
Union is Called.”

The difference between the 1985 and 1991 snapshots is striking. The first thing to
note is that Senator Exon speaks in far greater detail about the Soviet Union than he did
during the MX missile talks. Although he and Warner were both on the Armed Services
committee, Exon had gone on a number of delegations to the Soviet Union to meet with
Soviet Officials. He cites the situation on the ground, referring to mid-level Soviet officials
by name, commenting on shortages in foodstuffs and issues with Soviet telecommunications.
He also directly acknowledges the complexity of the Soviet situation, citing multiple forces
pulling the Soviets in different directions, referring to domestic, bureaucratic, and military
interests asserting pressure on the state leadership. He comments on public opinion, the
laying bare of the communist ideology, and notes Soviet relations with other states such as
the chill in Soviet-Japanese relations. In general, he finds the Soviet leadership to be “less
concerned with world domination, to be more inward focused in an effort to rescue the
faltering Soviet economy.” In effect, he has updated his overall assessment of the Soviet
Union in a positive direction. That is not to say that he has cast caution aside in his dealings
with the Soviets. He comments that while Gorbachev is a different kind of leader, there is no
guarantee that all leaders that follow will be as amenable to cooperation with the United
States. Thus he recommends staying vigilant in defense affairs, while still rewarding a
softened stance by the Soviets with similar reorientation by the United States. In general,
! 85!
Exon has responded to the change in structural conditions by becoming more nuanced in his
thinking of the Soviet Union.
On the other hand, Senator Warner’s assessment of the Soviet Union has become
more simplistic and seems to anchor to his previous view of their hostility. He emphasizes
their growing military strength and the reversibility of recent Soviet reforms. In effect, he
still believes that at its core the Soviet Union is hostile to the United States. He also speaks
in generic terms of the existence of “Soviet hardliners,” although he does briefly recognize
and congratulate the Soviet Union for its recent democratization. Overall, Warner’s view of
the Soviet Union represents that of the simplistic hostile adversary stereotype.
While the two senators are somewhat difficult to differentiate in terms of the
complexity of their views of the Soviet Union in 1985, the divergence in their statements in
1991 indicate strong support for the view that when decision-makers are faced with
information that confirms their worldview, they become more nuanced in their thinking,
while when they are faced with information that disconfirms their worldview, they become
more stereotypical and simplistic in their thinking. It also seems that Senator Exon
dramatically updated his assessment of the Soviet Union, while Warner’s view remained
generally the same.

Conclusion
The results suggest that political ideology is a significant determinant of the
complexity of decision-makers' enemy images under conditions where they are faced with
image-inconsistent information, even when they all have access to similar information.
While I did find support for H2, which anticipated a negative relationship between
! 86!
conservatism and complexity of views regarding the Soviet Union in 1991, I did not find
support for H1 which anticipated no relationship between ideology and complexity of views
of the Soviet Union in 1985. Instead, I found that when the Soviets were displaying hostility
to the U.S., individuals who were more conservative exhibited a greater degree of conceptual
complexity in discussing the Soviet Union. I speculate that the underlying pattern is that
decision-makers tend to be more nuanced in their thinking of adversaries when those
adversaries behave in accordance with individual decision-makers’ preferred worldviews.
Thus if a decision-maker is inclined to believe that the political universe is hostile, a belief
that is associated with conservatism, his or her views of adversaries will be more nuanced the
more those adversaries behave with hostility. Conversely, if an individual is inclined to
believe the political universe is harmonious, a belief associated with liberalism, his or her
view of adversaries will be more nuanced the more those adversaries exhibit friendliness.
Put differently, decision-makers become more multivariate in their thinking as the political
reality is more reflective of their preconceived image of the political universe. When
circumstances do not conform with expectations, individuals should tend to fall back on
generalizations and stereotypes, which often do not reflect complexity and nuance.
An implication of this finding is that decision-makers will tend to ignore information
that does not fit with their preconceived worldview. The choice between paying attention to
and ignoring image-disconfirming information had clear policy implications in the case of
the U.S. Senate in the summer of 1991. The distinction amounted to the choice between the
continuation of a long-established aggressive posture toward the adversary, on the one hand;
and the unprecedented extension of an olive branch, on the other. If images are not the
immediate causal antecedents of foreign policy decision-making, they are very close to it.
! 87!
The solution to debates concerning when states might favor absolute or relative gains does
seem to be linked to views of the enemy, as image theorists contend. However, we might
trace the causal process a further causal step backward to assess the individual-level sources
of contested images. This might be less of an imperative when we are concerned with
autocratic states or crisis situations, but it is crucial to understanding relations involving
democratic states in non-crisis situations, which constitute a great deal of international
politics.
U.S. foreign policy has long been concerned with projecting an image of resolve to
deter enemies and reinforce links with allies, as it is often the justification for military
intervention. Indeed, the United States has suffered tragic losses in the name of protecting
its image. However if U.S. behavior is filtered through other state elites' idiosyncratic
cognitive schema, and if states cannot straightforwardly engineer their images, it might not
be worthwhile to make great commitments for the purpose of protecting an image. It is
simply too difficult to determine whether a bluff will be met with a raise or a fold, especially
under conditions of relative parity between powers as was the case during the Cold War.

! 88!
References
Bilali, Rezarta. 2010. “Assessing the Internal Validity of Image Theory in the Context of
Turkey-U.S. Relations.” Political Psychology 31 (2): 275–303. doi:10.1111/j.1467-
9221.2009.00752.x.
Broder, John. 1989. “Cheney Bitingly Assails House Defense Budget.” Los Angeles Times,
August 24. http://articles.latimes.com/1989-08-24/news/mn-1377_1_defense-budget.
Carter, Ralph G., and James M. Scott. 2009. Choosing to Lead: Understanding Congressional
Foreign Policy Entrepreneurs. Duke University Press.
CQ Almanac 1991. n.d. “500 Million Approved for Soviet Aid,” 47 edition.
Downs, George W., and Michael A. Jones. 2002. “Reputation, Compliance, and International
Law.” The Journal of Legal Studies 31 (S1): S95–114. doi:10.1086/340405.
Duckitt, John. 2001. “A Dual-Process Cognitive-Motivational Theory of Ideology and
Prejudice.” Advances in Experimental Social Psychology 33: 41–114.
Erikson, Robert S. 2010. American Public Opinion: Its Origins, Content, and Impact. 8 edition.
Boston: Pearson.
Everson, Phil, Rick Valelly, and Jim Wiseman. 2009. “NOMINATE and American Political
History: A Primer.” Hentet Fra Http://www. Politicalbubbles.
Net/nominate_and_political_history_primer. Pdf (tilgået 23-06-2014).
Fearon, James D. 1997. “Signaling Foreign Policy Interests Tying Hands versus Sinking Costs.”
Journal of Conflict Resolution 41 (1): 68–90. doi:10.1177/0022002797041001004.
Gelman, Andrew. 2008. “Objections to Bayesian Statistics.” Bayesian Analysis 3 (3): 445–49.
! 89!
Hermann, Margaret. 2005. “Assessing Leadership Style: A Trait Analysis.” In The Psychological
Assessment of Political Leaders: With Profiles of Saddam Hussein and Bill Clinton,
edited by Jerrold M. Post, 1 edition. Ann Arbor: University of Michigan Press.
Hermann, Margaret G. 2005. “Assessing Leadership Style: A Trait Analysis.” The Psychological
Assessment of Political Leaders, 178–212.
Herrmann, Richard K., and Michael P. Fischerkeller. 1995. “Beyond the Enemy Image and
Spiral Model: Cognitive–strategic Research after the Cold War.” International
Organization 49 (03): 415–50. doi:10.1017/S0020818300033336.
Hopf, Ted. 1994. Peripheral Visions: Deterrence Theory and American Foreign Policy in the
Third World, 1965-1990. University of Michigan Press.
Houghton, David Patrick. 1996. “The Role of Analogical Reasoning in Novel Foreign-Policy
Situations.” British Journal of Political Science 26 (04): 523–52.
Huth, Paul K. 1997. “Reputations and Deterrence: A Theoretical and Empirical Assessment.”
Security Studies 7 (1): 72–99. doi:10.1080/09636419708429334.
Jervis, Robert. 2015. Perception and Misperception in International Politics. Princeton
University Press.
Jost, John T., Christopher M. Federico, and Jaime L. Napier. 2009. “Political Ideology: Its
Structure, Functions, and Elective Affinities.” Annual Review of Psychology 60 (1): 307–
37. doi:10.1146/annurev.psych.60.110707.163600.
Mearsheimer, John. 2014. The Tragedy of Great Power Politics. Updated Edition. New York:
W. W. Norton & Company.
Mercer, Jonathan. 2010. Reputation And International Politics. First edition. Cornell University
Press.
! 90!
Murray, Kathleen Shoon. 1997. Anchors against Change: American Opinion Leaders’ Beliefs
After the Cold War. Ann Arbor: University of Michigan Press.
Poole, Keith T., and Howard Rosenthal. 2001. “D-Nominate after 10 Years: A Comparative
Update to Congress: A Political-Economic History of Roll-Call Voting.” Legislative
Studies Quarterly, 5–29.
Press, Daryl G. 2007. Calculating Credibility: How Leaders Assess Military Threats. 1st ed.
Cornell University Press.
Sartori, Anne E. 2002. “The Might of the Pen: A Reputational Theory of Communication in
International Disputes.” International Organization 56 (01): 121–49.
doi:10.1162/002081802753485151.
Tetlock, Philip E. 1985. “Accountability: A Social Check on the Fundamental Attribution Error.”
Social Psychology Quarterly, 227–36.
Tomz, Michael. 2007. Reputation and International Cooperation: Sovereign Debt across Three
Centuries. Princeton University Press.
Trager, Robert F. 2010. “Diplomatic Calculus in Anarchy: How Communication Matters.”
American Political Science Review 104 (02): 347–68. doi:10.1017/S0003055410000158.
Weisiger, Alex, and Keren Yarhi-Milo. 2015. “Revisiting Reputation: How Past Actions Matter
in International Politics.” International Organization 69 (02): 473–95.
doi:10.1017/S0020818314000393.
Wilcox, Rand R., and H. J. Keselman. 2003. “Modern Robust Data Analysis Methods: Measures
of Central Tendency.” Psychological Methods 8 (3): 254.
Yarhi-Milo, Keren. 2014. Knowing the Adversary: Leaders, Intelligence, and Assessment of
Intentions in International Relations. Princeton University Press.
!
91!
!
Appendix A

Table A1. Conceptual Complexity of Enemy Images
Variable
Model 1: Non-Robust Clustered Standard Errors Model 2: Maximum Likelihood Estimation, Fixed Effects
Ideology -0.19 (0.09)* -0.19 (0.08)*
Party ID 0.19 (0.06)* 0.20 (0.06)*
Accountability -0.02 (0.03) -0.03 (0.03)
Tenure (logged) 0.05 (0.02)* 0.05 (0.02)*
Age -0.01 (0.01)+ -0.03 (0.01)*
Education -0.10 (0.04)* -0.10 (0.04)*
Statements (logged) 0.01 (0.01) 0.01 (0.02)
Population (logged) -0.01 (0.01) -0.01 (0.01)
Constant 0.95 (0.22)* 0.95 (0.22)*

R
2
= 0.04 Prob > X
2
= 0.01
RMSE = .2755 Log likelihood = -63.58
Notes: Standard Errors in Parentheses
+ p < .10, * p < .05

!
92!
!
Table A2. Conceptual Complexity of Enemy Images
Variable
Model 3: Bootstrapped Confidence Intervals Model 4: Bootstrapped CIs + Clustered Robust SEs
Ideology -0.19 (0.09)* -0.19 (0.11)+
Party ID 0.19 (0.06)* 0.20 (0.07)*
Accountability -0.03 (0.03) -0.03 (0.04)
Tenure (logged) 0.05 (0.02)* 0.05 (0.03)+
Age -0.03 (0.01)* -0.03 (0.01)*
Education -0.10 (0.03)* -0.10 (0.04)*
Statements (logged) 0.01 (0.02) 0.01 (0.03)
Population (logged) -0.01 (0.01) -0.01 (0.02)
Constant 0.95 (0.22)* 0.95 (0.28)*

Replications = 599 Replications = 599
Adjusted R
2
= 0.02 Adjusted R
2
= 0.02
P
Notes: Standard Errors in Parentheses; + p < .10, * p < .05

!
93!
!
Table A3. Conceptual Complexity of Enemy Images
Variable
Model 5: M-Estimator Model 6: S-Estimator Model 7: MM-Estimator
Ideology -0.17 (0.09)+ -0.22 (0.17) -0.16 (0.10)+
Party ID 0.18 (0.07)* 0.22 (0.13)+ 0.17 (0.07)*
Accountability -0.02 (0.03) -0.06 (0.05) -0.02 (0.03)
Tenure (logged) 0.05 (0.02)* 0.07 (0.05) 0.05 (0.03)+
Age -0.03 (0.01)+ -0.04 (0.06) -0.03 (0.02)
Education -0.08 (0.04)* -0.01 (0.09) -0.07 (0.04)*
Statements (logged) 0.02 (0.02) 0.04 (0.03) 0.02 (0.02)
Population (logged) 0.01 (0.01) 0.02 (0.03) -0.01 (0.01)
Constant 0.88 (0.21)* 0.89 (0.44)* 0.86 (0.23)*

Huber k = 1.34 Bisquare k = 1.55 M-estimate: k = 3.44
Scale estimate = 0.28 Scale estimate = 0.28 Scale estimate = 0.28
Notes: 525 Standard Errors in Parentheses; + p < .10, **p < .05

!
94!
!
Appendix B. Elasticity Testing of Regression Models

Model 1: Clustered Standard Errors

Variable

Ideology -0.19 (0.09)* -0.18 (0.09)* -0.17 (0.08)* -0.16 (0.08)* -0.15 (0.08)+ -0.14 (0.08)+ -0.05 (0.06)
Party ID 0.20 (0.06)* 0.19 (0.06)* 0.18 (0.06)* 0.18 (0.06)* 0.16 (0.06)* 0.15 (0.06)* 0.09 (0.05)+
Education -0.10 (0.04)* -0.11 (0.04)* -0.10 (0.03)* -0.10 (0.03)* -0.07 (0.03)* -0.06 (0.03)+
Tenure
(logged)
0.05 (0.02)* 0.06 (0.02)* 0.06 (0.02)* -0.06 (0.02)* 0.03 (0.02)*
Age -0.003
(0.002)+
-0.003 (0.002) + -0.004 (0.002)* -0.004
(0.002)*

Accountabili
ty
-0.03 (0.03) -0.02 (0.03) -0.02 (0.03)
Population
(logged)
-0.01 (0.01) -0.01 (0.01)
Statements
(logged)
0.01 (0.02)
Constant 0.95 (0.22)* 0.99 (0.22)* 0.89 (0.44)* 0.79 (0.11)* 0.60 (0.06)* 0.67 (0.04)* 0.60 (0.03)*

Prob > F =
0.009
Prob > F = 0.005 Prob > F = 0.002 Prob > F =
0.002
Prob > F = 0.01 Prob > F = 0.03 Prob > F = 0.07
RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28
Notes: Standard Errors in Parentheses
+ p < .10, **p < .05

!
95!
!
Model 2: Maximum Likelihood Estimation, Fixed Effects

Variable

Ideology -0.19 (0.08)* -0.18 (0.09)* -0.17 (0.08)* -0.16 (0.08)* -0.15 (0.08)+ -0.14 (0.08)+ -0.05 (0.06)
Party ID 0.20 (0.06)* 0.19 (0.06)* 0.18 (0.06)* 0.18 (0.06)* 0.16 (0.06)* 0.15 (0.06)* 0.09 (0.05)+
Education -0.03 (0.03) -0.11 (0.04)* -0.10 (0.03)* -0.10 (0.03)* -0.07 (0.03)* -0.06 (0.03)+
Tenure
(logged)
0.05 (0.02)* 0.06 (0.02)* 0.06 (0.02)* -0.06 (0.02)* 0.03 (0.02)*
Age -0.03 (0.01)* -0.003 (0.002) + -0.004 (0.002)* -0.004
(0.002)*

Accountabili
ty
-0.10 (0.04)* -0.02 (0.03) -0.02 (0.03)
Population
(logged)
0.01 (0.02) -0.01 (0.01)
Statements
(logged)
-0.01 (0.01)
Constant 0.95 (0.22)* 0.99 (0.22)* 0.89 (0.44)* 0.79 (0.11)* 0.60 (0.06)* 0.67 (0.04)* 0.60 (0.03)*

Log likelihood
= -63.58
RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28 RMSE = 0.28
Notes: Standard Errors in Parentheses
+ p < .10, * p < .05

!
96!
!
Model 3: Bootstrapped Confidence Intervals

Variable

Ideology -0.19 (0.09)* -0.18 (0.09)* -0.17 (0.09)* -0.16 (0.08)* -0.15 (0.08)+ -0.14 (0.08)+ -0.05 (0.06)
Party ID 0.19 (0.06)* 0.19 (0.06)* 0.18 (0.06)* 0.18 (0.06)* 0.16 (0.06)* 0.15 (0.06)* 0.09 (0.05)+
Education -0.03 (0.03) -0.11 (0.03)* -0.10 (0.03)* -0.10 (0.03)* -0.07 (0.03)* -0.06 (0.03)+
Tenure
(logged)
0.05 (0.02)* 0.06 (0.02)* 0.06 (0.02)* -0.06 (0.02)* 0.03 (0.02)*
Age -0.03 (0.01)* -0.003 (0.002) + -0.004 (0.002)+ -0.004
(0.002)*

Accountabili
ty
-0.10 (0.03)* -0.02 (0.03) -0.02 (0.03)
Population
(logged)
0.01 (0.02) -0.01 (0.01)
Statements
(logged)
-0.01 (0.01)
Constant 0.95 (0.22)* 0.99 (0.22)* 0.80 (0.11)* 0.79 (0.11)* 0.60 (0.06)* 0.67 (0.04)* 0.60 (0.03)*

Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Prob > X
2
=
0.007
Prob > X
2
=
0.003
Prob > X
2
= 0.01 Prob > X
2
=
0.01
Prob > X
2
= 0.01 Prob > X
2
= 0.01 Prob > X
2
= 0.01
Notes: Standard Errors in Parentheses
+ p < .10, * p < .05

!
97!
!
Model 4: Bootstrapped Confidence Intervals + Clustered Robust Standard Errors

Variable

Ideology -0.19 (0.11)+ -0.18 (0.10)+ -0.17 (0.10)+ -0.16 (0.10)++ -0.15 (0.09)++ -0.14 (0.09)++ -0.05 (0.08)
Party ID 0.20 (0.07)* 0.19 (0.07)* 0.18 (0.07)* 0.18 (0.07)* 0.16 (0.06)* 0.15 (0.06)* 0.09 (0.05)+
Education -0.03 (0.04) -0.11 (0.04)* -0.10 (0.04)* -0.10 (0.04)* -0.07 (0.03)* -0.06 (0.03)*
Tenure
(logged)
0.05 (0.03)+ 0.06 (0.03)* 0.06 (0.03)* -0.06 (0.02)* 0.03 (0.02)*
Age -0.03 (0.01)* -0.003 (0.003) -0.004 (0.003)++ -0.004
(0.002)++

Accountabili
ty
-0.10 (0.04)* -0.02 (0.03) -0.02 (0.03)
Population
(logged)
0.01 (0.03) -0.01 (0.02)
Statements
(logged)
-0.01 (0.02)
Constant 0.95 (0.28)* 0.99 (0.27)* 0.80 (0.16)* 0.79 (0.11)* 0.60 (0.06)* 0.67 (0.05)* 0.60 (0.03)*

Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Replications =
599
Prob > X
2
=
0.02
Prob > X
2
=
0.003
Prob > X
2
= 0.01 Prob > X
2
=
0.01
Prob > X
2
= 0.01 Prob > X
2
= 0.01 Prob > X
2
= 0.01
Notes: Standard Errors in Parentheses
++ p < .15, + p < .10, * p < .05

!
98!
!
Model 5: M-Estimator

Variable

Ideology -0.17 (0.09)+ -0.16 (0.09)+ -0.15 (0.09)+ -0.14 (0.08)+ -0.14 (0.08)+ -0.13 (0.08)+ -0.05 (0.06)
Party ID 0.18 (0.07)* 0.17 (0.07)* 0.17 (0.07)* 0.16 (0.07)* 0.15 (0.06)* 0.14 (0.06)* 0.09 (0.05)+
Education -0.02 (0.03) -0.09 (0.04)* -0.09 (0.04)* -0.09 (0.03)* -0.06 (0.03)+ -0.05 (0.03)++
Tenure
(logged)
0.05 (0.02)* 0.06 (0.02)* 0.06 (0.02)* 0.06 (0.02)* 0.03 (0.02)+
Age -0.03 (0.01)+ -0.003 (0.002)+ -0.003 (0.002)* -0.003
(0.002)*

Accountabili
ty
-0.08 (0.04)* -0.01 (0.03) -0.01 (0.03)
Population
(logged)
0.02 (0.02) -0.01 (0.01)
Statements
(logged)
0.01 (0.01)
Constant 0.88 (0.21)* 0.92 (0.22)* 0.80 (0.11)* 0.79 (0.10)* 0.61 (0.06)* 0.68 (0.04)* 0.63 (0.03)*

Huber k = 1.34 Huber k = 1.34 Huber k = 1.34 Huber k =
1.34
Huber k = 1.34 Huber k = 1.34 Huber k = 1.34
Scale estimate
= 0.28
Scale estimate =
0.28
Scale estimate =
0.28
Scale estimate
= 0.28
Scale estimate =
0.28
Scale estimate =
0.25
Scale estimate =
0.25
Notes: Standard Errors in Parentheses
++ p < .15, + p < .10, * p < .05

!
99!
!
Model 6: S-Estimator

Variable

Ideology -0.22 (0.17) -0.20 (0.16) -0.21 (0.19) -0.20 (0.20) -0.18 (0.17) -0.11 (0.17) -0.12 (0.13)
Party ID 0.22 (0.13)+ 0.21 (0.12)+ 0.22 (0.15)* 0.21 (0.16) 0.18 (0.14) 0.13 (0.14) 0.14 (0.12)
Education -0.06 (0.05) -0.03 (0.07) -0.04 (0.10) -0.05 (0.14) 0.001 (0.06) 0.003 (0.08)
Tenure
(logged)
0.07 (0.05) 0.09 (0.05)+ 0.10 (0.05)* 0.11 (0.05)* 0.07 (0.04)+
Age -0.04 (0.06) -0.004 (0.006) -0.006 (0.007) -0.006 (0.008)
Accountabili
ty
-0.01 (0.09) -0.04 (0.05) -0.04 (0.06)
Population
(logged)
0.04 (0.03) -0.02 (0.03)
Statements
(logged)
0.02 (0.03)
Constant 0.89 (0.44)* 1.03 (0.41)* 0.77 (0.48)++ 0.80 (0.63) 0.42 (0.15)* 0.61 (0.12)* 0.61 (0.05)*

Bisquare k =
1.55
Bisquare k =
1.55
Bisquare k = 1.55 Bisquare k =
1.55
Bisquare k =
1.55
Bisquare k =
1.55
Bisquare k =
1.55
Scale estimate
= 0.28
Scale estimate =
0.28
Scale estimate =
0.28
Scale estimate
= 0.28
Scale estimate =
0.28
Scale estimate =
0.28
Scale estimate =
0.28
Notes: Standard Errors in Parentheses
++ p < .15, + p < .10, * p < .05

!
100!
!
Model 7: MM-Estimator

Variable

Ideology -0.16 (0.10)+ -0.15 (0.09)+ -0.14 (0.09)++ -0.14 (0.09)++ -0.13 (0.08)++ -0.12 (0.08) -0.06 (0.07)
Party ID 0.17 (0.07)* 0.16 (0.07)* 0.16 (0.07)* 0.15 (0.07)* 0.14 (0.07)* 0.13 (0.07)* 0.09 (0.05)++
Education -0.02 (0.03) -0.08 (0.04)* -0.08 (0.04)* -0.07 (0.04)* -0.05 (0.03)++ -0.04 (0.03)
Tenure
(logged)
0.05 (0.03)+ 0.06 (0.03)* 0.06 (0.02)* 0.06 (0.02)* 0.03 (0.02)+
Age -0.03 (0.02) -0.003 (0.002)++ -0.003 (0.002)+ -0.003
(0.002)+

Accountabili
ty
-0.07 (0.04)* -0.01 (0.03) -0.01 (0.03)
Population
(logged)
0.02 (0.02) -0.01 (0.01)
Statements
(logged)
-0.01 (0.01)
Constant 0.86 (0.23)* 0.89 (0.23)* 0.79 (0.12)* 0.78 (0.12)* 0.60 (0.06)* 0.67 (0.05)* 0.63 (0.03)*

M-estimate: k
= 3.44
M-estimate: k =
3.44
M-estimate: k =
3.44
M-estimate: k
= 3.44
M-estimate: k =
3.44
M-estimate: k =
3.44
M-estimate: k =
3.44
Scale estimate
= 0.28
Scale estimate =
0.28
Scale estimate =
0.27
Scale estimate
= .28
Scale estimate =
0.28
Scale estimate =
0.28
Scale estimate =
0.28
Notes: Standard Errors in Parentheses
++ p < .15, + p < .10, * p < .05

!
101!
!
!
Chapter 4

Deterrence Without Threat or Force: Individual
Differences in Sensitivity to Past Actions

Policymakers often justify war based on the need to protect national credibility
and reputation for resolve, and a debate has formed in deterrence scholarship
concerning whether or not observers draw reputational inferences based on past
behavior. This research has emphasized the role of material factors, perceived
interests, and/or cognitive biases in determining whether or not states’ efforts to
cultivate their reputation should be effective in changing the beliefs of observers.
Yet few have isolated the role individual differences in cognition play. It is quite
tenable to assume that some individuals might be more likely than others to
dismiss a target state’s past actions when weighing the decision to initiate
conflict. In this study, I leverage an incentivized market-entry bargaining
experiment to assess individuals’ general propensity to challenge an observed
actor. Borrowing from social psychology’s attributional style literature, I find
that individuals with a situational attributional style are more likely to play the
optimum strategy as dictated by both the observed actor’s behavior and the
structure of payoffs given by the game.

Decision-makers waging war out of concern for maintaining a reputation for resolve has
existed perhaps as long as the international system itself. Policymakers seem convinced that by
exhibiting “toughness” in international affairs, they can effectively influence the beliefs of other
international actors in the ways they expect. As modern international relations scholarship
developed over the course of the Cold War, it is no surprise that a substantial body of scholarship
developed to better understand the process of signaling to cultivate a reputation for resolve
(Schelling 1960; Russett 1963; Singer and Small 1966; George and Smoke 1974; Snyder and
Diesing 1977; Jervis 1979; Mearsheimer 1983; Huth 1988; Jervis, Lebow, and Stein 1989;
Powell 1990).
!
102!
!
Assessments of state resolve typically fall under the umbrella of deterrence theory, which
is concerned with states’ reputations for resolve, where reputation is defined as “a belief about a
trait or behavioral tendency of an actor, based on that actor’s past behavior; usually widely held”
(Dafoe, Renshon, and Huth 2014). While the focus has generally been on reputation for resolve,
others have assessed reputations for other tendencies such as risk acceptance (Crescenzi 1999),
repaying debt (Tomz 2007), or being a reliable ally (Miller 2012). While reputation is generally
thought of as an attribute of states, it is at its core a psychological and ideational concept (Mercer
2010), and a recent body of work has developed that approaches reputation from a cognitive or
social psychological standpoint (Larson and Shevchenko 2010; Dafoe, Renshon, and Huth 2014).
While scholars generally agree that decision-makers care about reputation (McMahon
1991; Kagan 1995; Tang 2005; O’Neill 2006; Lebow 2008), they have come to no agreement as
to whether observers actually update their assessments about other states based on how states
behave, and mainstream scholarship on reputation does not find compelling evidence that elites’
beliefs about current resolve are influenced by past behavior (Dafoe, Renshon, and Huth 2014).
Scholars like Hopf (1994) find that US military behavior did not influence Soviet inferences
about US credibility, which Mercer (2010) attributes to a confirmatory attributional pattern.
Similarly, Press (2007) finds that past actions do not inform elites’ assessments of the credibility
of threat, opting instead for a current calculus logic where present interests and power play a
more substantial role. Yet other scholars remain convinced that reputational inferences are
indeed drawn, such as Trager and Vavreck (2011) who find that threats induce reactions among
domestic audiences, Tomz (2007) who finds that states develop reputations among international
investors based on their past payments, or Sartori (2002) who finds that countries can affect their
!
103!
!
reputation through diplomatic signals since other states know they have a desire to maintain a
reputation for honesty.
One of the newer efforts among scholars has been to assess the conditions under which
reputation matters. In an assessment of the effects of time horizons on reputational inferences,
Tingley and Walter (2011) find that individuals invest in reputation regardless of whether the
game is played many or a few times. They also find that investing in a reputation for toughness
helps defenders deter entry, although its effect varies from the early stages of a game to the late
stages. They also find some individual variation in updating, citing limits in cognitive ability to
take into account the structure of the game for some individuals.
The study by Tingley and Walter effectively broaches the question of when reputations
are drawn, and also hints at who draws these inferences. This study will expand on the second of
these questions, drawing from an adapted version of Tingley and Walter’s market-entry
deterrence game. I begin in the first section by outlining the model and adapted version of this
game. I then introduce research on attributional style, which I argue impacts individuals’
propensity to make reputational inferences. In the following section I advance a series of
hypotheses concerning reputational inferences that are derived from research on attributional
style, and then describe the market-entry deterrence game employed to assess these hypotheses
in a laboratory experiment. Following that I assess the results of the laboratory experiment, and
conclude with a discussion of the overall findings and implications of this study.

!
104!
!
The Model and Market Entry-Deterrence Game
The game used in this piece is an adapted version
5
of that used by Tingley and Walter,
which is essentially designed to focus on one type of reputation, a reputation for resolve. Like
Tingley and Walter, I define a reputation for resolve as a “belief by others that a player who
fought a challenger in the past will continue to fight challengers in the future, given a sufficiently
similar situation.” (Ibid. 346). The authors adapted their version of the game from a game
commonly used in economics (Jung, Kagel, and Levin 1994), where a monopolist can keep firms
out of the marketplace by way of price attrition. Aggressively lowering prices signals to other
firms that entry will be costly, and maintaining this posture allows the monopolist to develop a
reputation for toughness. In this instance, the short-term cost of lowering prices is expected to be
offset by the long-term gains of deterring new market entrants.
In the game, defenders face a series of potential challengers who must decide whether or
not to challenge the defender. The defender must also decide in advance whether they will fight
entry or allow the challenger to enter. In the beginning of the game, defenders’ payoff structures
are randomly assigned to committed or uncommitted categories, such that committed defenders
should always prefer to fight entry based on payoffs and uncommitted defenders will prefer not
to defend. It is the entrant’s task to decide whether to challenge or not to challenge, knowing
that there is some probability of facing a committed defender. Figure 1 indicates the structure of
a single shot of the game.
In Tingley and Walter’s version of the game, the number of repetitions is varied, such
that defenders either face four or eight entrants. Once the defender makes its choice, if an
entrant decides to challenge, the defender’s choice to stand firm or back down is recorded and
displayed to new entrants in ensuing iterations of the game. Based on this information, entrants
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
5
In collaboration with Brian Rathbun, Joshua Kertzer, and Mark Paradis
!
105!
!
are expected to make inferences about whether the defender is committed or uncommitted. How
the defender behaves towards an early entrant should presumably indicate how the defender is
likely to behave towards later entrants.
In the adapted version of the game used here, the number of repetitions does not vary.
Instead, my focus is on how individual differences affect the decision to challenge a defender, a
decision which is an indicator of perceived resolve. I maintain the order of the moves in each
iteration, where nature randomly determines the defender’s type; the challenger uses this
information to decide their move, while the defender also determines their strategy.

Figure 1. Structure of a Single Iteration of the Game

From Tingley and Walter (2011)

Attributional Style
Attributional style refers to an individual's propensity to commit the “fundamental
attribution error” of overemphasizing the dispositions in making causal attributions for
the behavior of others. A dispositional attribution should be more closely linked to a
reputational inference, since the rationale for perceived behavior is given to the object
!
106!
!
under observation, rather than its environment. In any case, a number of studies have
demonstrated that individuals vary systematically with respect to how likely they are to
emphasize dispositions in making causal attributions. For example, Cochran, Boots, and
Chamlin (2006) find that conservatives' support for capital punishment can be explained
in part by their dispositional attributional style, which focuses more on the inherent
characteristics of the individual offender than the situational-systemic factors that may
have given rise to the criminal behavior in question. A similar pattern is found by
Weiner et al. (2010), who find that conservatives are more likely to exhibit anger toward
the poor because they blame poverty on individual dispositions, such as laziness or lack
of interest, as primary contributing causes to poverty. On the other hand, liberals are
more likely to be sympathetic to the poor since they attribute poverty to situational
factors such as a lack of resources or institutional support.
In a related study, Gomez and Wilson (2006) assess the relationship between
attributional complexity (which, they argue, is a component of political sophistication)
and attributional style regarding responsibility for the economy. They find that less
attributionally complex voters focus attributions on the President as the locus of
responsibility, rather than focusing on Congress as a structural constraint on Presidential
policy-making. Chirumbolo et al. (2004) find a relationship between the need for closure
and attributional style, such that individuals with a high need for closure tend to blame
individual dispositions for social problems rather than systemic factors.

!
107!
!
Hypothesis
Following from the literature on attributional style, I contend that individual-level factors
impact reputational inferences; thus, the absence of an impact of individual-level factors
constitutes the null hypothesis. While this is admittedly a “straw man” argument that is only
intended to provide a reference point, it does seem to be an ironically accurate statement on the
literature on image updating. Because the game was designed so that all defenders do not fight
until the last two rounds, all defenders in the game reveal an uncommitted type. Thus, the
structure of the game dictates that challenging under all circumstances is the optimum strategy.
I hypothesize that in the bargaining game context, given the same information, challengers with a
situational attributional style will be more likely to submit a challenge to defenders than
challengers with a dispositional attributional style, since they should exhibit a better
understanding of the structure of the game in weighing not only the costs and benefits of
challenging in individual rounds but also the incentives faced by defenders. The logic of this
hypothesis follows from the notion that once a defender’s past behavior is revealed, the
remainder of the decision to challenge depends on the structure of the game. Thus, those with a
situational attributional style will focus on the structure of the game, which dictates challenging
in all instances. This hypothesis may seem counterintuitive given that individuals with a
dispositional attributional style might be expected to focus on past actions more and thus be
inclined to challenge an uncommitted defender. However, in this context where the only obvious
information is past actions, only focusing on an individual defender’s past actions constitutes a
form of cognitive laziness which implies that an individual’s understanding of the overall
structure of the game may be compromised.

!
108!
!
Methods
The experimental subjects included undergraduates at a large private university on the
west coast. They were recruited via flyer and were paid $10 for attendance, plus an additional
performance-based sum (which averaged roughly $4). Participants were presented a
demographic and psychometric questionnaire in addition to experimental surveys, in random
order. The setting was a university lecture hall, with three researchers present to monitor
compliance and comprehension, as well as to explain the rules of the game.
Attributional style was measured using eight items (see Appendix A) from the
Attributional Style Assessment Test I (ASAT—I) where individuals were presented with
common situations of personal success and failure, with different possible explanations for the
outcomes in each situation. For instance, participants were told, “You have just succeeded at
completing the crossword puzzle in the daily paper,” and given five options that they could say
account for their performance in that instance. These are: a) “I used the right strategy to
complete the puzzle”; b) “I am good at crossword puzzles”; c) “I tried very hard to complete the
puzzle”; d) “I have the personality traits necessary for completing crossword puzzles”; e) “I was
in the right mood for a crossword puzzle”; f) “other circumstances (people, situations) produced
this outcome”. Each of these options indicates attributions for personal success or failure to
skill, personality, effort, or circumstances. Two factor analyses (using promax rotation) were
used to generate factor scores for dispositional and situational attributional style separately,
where loadings satisfied the eigenvalue and scree test (see Appendix B).
In total, 209 participants completed the study, with the following characteristics: 23 years
old on average, 54% female, predominantly Caucasian and Asian, and predominantly liberal. No
doubt, this is far from a representative sample of the general population, and even further from
!
109!
!
the population of elites one might seek to generalize to, however the baseline hypotheses should
hold just as well to this sample. The experiment used herein is part of a broader study that
examines both challenger and defender behavior, thus two separate survey instruments were used
with eight rounds each. Participants were told that they would be facing a new opponent in
every round, and were not informed of the defender’s type. Rather, they had to infer the type
based on the defender’s history. In each period, challengers receive information about the
defender’s history, and determine whether or not they indeed want to challenge the defender.
The history only shows the decisions for defender behavior for periods in which they were
challenged, thus the information might be scarce, particularly in early periods. In any case, since
the goal is only to identify individual differences, the commitment level of the defenders is held
constant for the first five rounds, such that defenders either do not fight or back down.

Results and Analysis
To assess the role of attributional style on the probability of fighting, I conduct a period-
by-period robust logit with the standardized difference between situational and dispositional
attributional style (scaled where higher scores indicate a greater leaning towards a situational
attributional style) as the predictor. Gender, age, race, participation in an economics course,
SAT score, political party identification, order of role, competitive and dangerous worldview
factors are included as controls. Since I am interested in the relative impact of attributional styles
as opposed to individual impact of attribution components, the difference between attributional
styles is the appropriate measure. Table 2 presents the results from the regressions. Participants
with an attributional style that places a relatively larger focus on the circumstance are
significantly more likely to fight in five of the eight periods, although the effects in two of the
!
110!
!
periods were significant at the less conventional 0.10 level. Given the small sample size, these
are important findings. The results are clearest, at least from a statistical significance point of
view, in rounds 2-4.
What these findings suggest is that individual differences in attributional style do matter
in the decision of how to deal with observed history in the behavior of another. Those with a
more situational attributional style will tend to challenge more, which given the low level of
fighting demonstrated by the defenders is indicative of a better understanding of the structure of
the game and a more optimal approach given the type of defenders faced (noting that the
defenders always exhibited a low level of commitment in this game).
!
111!
!
Table 2. Propensity to Fight by Round and Aggregates

Variable Round 1 Round 2 Round 3 Round 4 Round 5 Round 6 Round 7 Round 8
Fight
Ratio
Fight
Count
Gender -0.81* -0.63 -1.12** -0.84* -1.11** -0.93** -0.35 -0.70 -0.34** -0.83**

(-2.04) (-1.78) (-2.88) (-2.13) (-2.84) (-2.68) (-0.90) (-1.65) (-2.72) (-2.92)
Age 0.03 0.07 -0.11 -0.22 -0.19 -0.03 -0.03 0.05 -0.01 -0.04

(0.20) (0.53) (-0.89) (-1.60) (-1.31) (-0.27) (-0.18) (0.37) (-0.27) (-0.38)
Race 0.21 0.21 0.38** 0.43** 0.265 0.09 0.20 0.07 0.05 0.09

(1.54) (1.59) (3.02) (3.14) (1.92) (0.74) (1.48) (0.47) (1.35) (0.97)
Econ 0.11 -0.09 -0.22 0.05 -0.61 -0.10 -0.73 -0.45 0.03 0.22

(0.25) (-0.20) (-0.47) (0.11) (-1.35) (-0.22) (-1.52) (-0.90) (0.28) (0.69)
SAT 0.02 0.11 0.02 0.08 0.04 -0.03 -0.03 -0.04 -0.003 -0.01

(0.26) (1.30) (0.19) (0.91) (0.49) (-0.33) (-0.28) (-0.54) (-0.12) (-0.14)
Party ID 0.13 -0.05 0.01 0.02 0.05 0.18 -0.05 0.16 0.021 0.03

(0.87) (-0.37) (0.09) (0.11) (0.32) (1.41) (-0.32) (1.02) (0.47) (0.30)
Order 0.78* 0.27 -0.37 0.01 -0.10 0.06 0.20 0.65 0.06 0.02

(2.04) (0.77) (-1.02) (0.02) (-0.28) (0.18) (0.55) (1.55) (0.53) (0.06)
C-World -0.58* -0.23 -0.40 -0.57* 0.01 0.01 0.03 0.20 -0.06 -0.14

(-2.06) (-0.89) (-1.45) (-2.12) (0.05) (0.06) (0.11) (0.63) (-0.71) (-0.73)
D-World 0.25 0.13 0.08 0.03 0.01 0.07 0.12 0.09 -0.002 -0.04

(1.06) (0.58) (0.34) (0.14) (0.03) (0.30) (0.52) (0.37) (-0.03) (-0.25)
Att-
Style 0.13 0.46** 0.34* 0.39* 0.28# 0.15 0.28# 0.06 0.10* 0.27*

(0.83) (2.87) (2.19) (2.29) (1.74) (1.03) (1.84) (0.32) (2.00) (2.42)
_cons -3.05 -3.25 1.48 1.85 3.42 0.01 0.15 -2.24 0.26 N/A
(-0.97) (-1.06) (0.49) (0.59) (1.10) (0.00) (0.05) (-0.66) (0.27) (N/A)
Notes: Test statistics in parentheses. Logit with robust (Huber-White) standard errors for individual rounds, ordinal logit with robust
(Huber-White) standard errors for fight count, tobit regression for fight ratio. # p < 0.10, * p < 0.05, ** p < 0.01

!
112!
!

Conclusion
In this paper, I use a repeated-entry deterrence game to assess the role of dispositional
factors in the level of sensitivity to past actions in previous iterations of the game, where the
defender is exhibiting low levels of commitment. I hypothesized that a situational attributional
style should result in greater awareness of and behavior in accordance with the payoff structure
of the game, and found evidence to this effect with relatively few degrees of freedom. This
supports the claim that individual differences in attributional style matter in individuals’
sensitivity to information about past actions. It is likely the case that any claim or assessment of
reputation that does not take individual differences into account is missing a key component of
the causal picture.
In the reputation literature, a great deal of attention is given to the effects of signaling
behavior, and an insufficient amount to the behavior of those observing the signals. There are
only a few recent exceptions to this trend but it seems that there is a growing appreciation for
subjectivity’s role in signaling behavior. Given that policymakers often assert the importance of
establishing credibility, it is important to understand who is likely to find credibility in behavior
that is not necessarily self evident, or that can be explained away to circumstances. As Schelling
said, “we lost thirty thousand dead in Korea to save face for the United States and the United
Nations, not to save South Korea for the South Koreans, and it was undoubtedly worth it. Soviet
expectations about the behavior of the United States are one of the most valuable assets we
possess in world affairs.” However, if a state’s behavior and signaling is filtered through other
state elites’ idiosyncratic cognitive schema, states cannot engineer their images in a predictable
fashion. If that is the case, it might not be worthwhile to make great commitments for the
!
113!
!
purpose of protecting an image. It is simply too difficult to determine whether a bluff will be
met with a raise or a fold. In any case, it seems that our study has the potential to demonstrate
some of the behavioral conditions under which signaling might be effective.

!
114!
!
References
Chirumbolo, Antonio, Alessandra Areni, and Gilda Sensales. 2004. “Need for Cognitive Closure
and Politics: Voting, Political Attitudes and Attributional Style.” International Journal of
Psychology 39 (4): 245–53.
Cochran, John K., Denise Paquette Boots, and Mitchell B. Chamlin. 2006. “Political Identity and
Support for Capital Punishment: A Test of Attribution Theory.” Journal of Crime and
Justice 29 (1): 45–79.
Crescenzi, Mark JC. 1999. “Violence and Uncertainty in Transitions.” Journal of Conflict
Resolution 43 (2): 192–212.
Dafoe, Allan, Jonathan Renshon, and Paul Huth. 2014. “Reputation and Status as Motives for
War.” Annual Review of Political Science 17: 371–93.
George, Alexander L., and Richard Smoke. 1974. Deterrence in American Foreign Policy:
Theory and Practice. Columbia University Press.
Gomez, Brad T., and J. Matthew Wilson. 2006. “Rethinking Symbolic Racism: Evidence of
Attribution Bias.” Journal of Politics 68 (3): 611–25.
Hopf, Ted. 1994. Peripheral Visions: Deterrence Theory and American Foreign Policy in the
Third World, 1965-1990. University of Michigan Press.
Huth, Paul K. 1988. Extended Deterrence and the Prevention of War. Yale University Press.
Jervis, Robert. 1979. “Deterrence Theory Revisited.” World Politics: A Quarterly Journal of
International Relations, 289–324.
Jervis, Robert, Richard Ned Lebow, and Janice Gross Stein. 1989. Psychology and Deterrence.
JHU Press.
!
115!
!
Jung, Yun Joo, John H. Kagel, and Dan Levin. 1994. “On the Existence of Predatory Pricing: An
Experimental Study of Reputation and Entry Deterrence in the Chain-Store Game.” The
RAND Journal of Economics, 72–93.
Kagan, Donald. 1995. On the Origins of War and the Preservation of Peace. Doubleday New
York.
Larson, Deborah Welch, and Alexei Shevchenko. 2010. “Status Seekers: Chinese and Russian
Responses to US Primacy.” International Security 34 (4): 63–95.
Lebow, Richard Ned. 2008. A Cultural Theory of International Relations. Cambridge University
Press.
McMahon, Robert J. 1991. “Credibility and World Power: Exploring the Psychological
Dimension in Postwar American Diplomacy.” Diplomatic History 15 (4): 455–72.
Mearsheimer, John J. 1983. Conventional Deterrence. Cornell University Press.
Mercer, Jonathan. 2010. Reputation And International Politics. First edition. Cornell University
Press.
Miller, Gregory D. 2012. The Shadow of the Past: Reputation and Military Alliances before the
First World War. Cornell University Press.
O’Neill, Barry. 2006. “Nuclear Weapons and National Prestige.” Cowles Foundation for
Research in Economics, Yale University.
Powell, Robert. 1990. Nuclear Deterrence Theory: The Search for Credibility. Cambridge
University Press.
Press, Daryl G. 2007. Calculating Credibility: How Leaders Assess Military Threats. 1st ed.
Cornell University Press.
Russett, Bruce M. 1963. “The Calculus of Deterrence.” Journal of Conflict Resolution, 97–109.
!
116!
!
Sartori, Anne E. 2002. “The Might of the Pen: A Reputational Theory of Communication in
International Disputes.” International Organization 56 (01): 121–49.
doi:10.1162/002081802753485151.
Schelling, Thomas C. 1960. “The Strategy of Conflict.” Cambridge, Mass.
Singer, J. David, and Melvin Small. 1966. “Formal Alliances, 1815—1939 A Quantitative
Description.” Journal of Peace Research 3 (1): 1–31.
Snyder, Glenn Herald, and Paul Diesing. 1977. Conflict Among Nations: Bargaing, Decision
Making, and Systems Structure in International Crises. Princeton University Press.
Tang, Shiping. 2005. “Reputation, Cult of Reputation, and International Conflict.” Security
Studies 14 (1): 34–62.
Tingley, Dustin H., and Barbara F. Walter. 2011. “The Effect of Repeated Play on Reputation
Building: An Experimental Approach.” International Organization 65 (02): 343–65.
Tomz, Michael. 2007. Reputation and International Cooperation: Sovereign Debt across Three
Centuries. Princeton University Press.
Trager, Robert F., and Lynn Vavreck. 2011. “The Political Costs of Crisis Bargaining:
Presidential Rhetoric and the Role of Party.” American Journal of Political Science 55
(3): 526–45.
Weiner, Bernard, Danny Osborne, and Udo Rudolph. 2010. “An Attributional Analysis of
Reactions to Poverty: The Political Ideology of the Giver and the Perceived Morality of
the Receiver.” Personality and Social Psychology Review, 1088868310387615.

!
117!
!
Appendix A. ASAT-I Items

!
118!
!

!
119!
!

!
120!
!
Appendix B. Scree Plot of Eigenvalues for Attributional Style Items

Dispositional Factor

Situational Factor

-.5 0 .5 1 1.5 2
Eigenvalues
0 2 4 6 8
Number
Scree plot of eigenvalues after factor
!
121!
!
Chapter 5

Conclusion

This dissertation has been motivated in large part by the slow thawing of tensions
between Iran and the United States over the course of the Obama presidency. Upon being
inaugurated, President Obama broke with the Bush administration’s past hostility towards the
Muslim world in saying:
“To the Muslim world, we seek a new way forward, based on
mutual interest and mutual respect. To those leaders around the
globe who seek to sow conflict, or blame their society’s ills on the
West—know that your people will judge you on what you can
build, not what you destroy. To those who cling to power through
corruption and deceit and the silencing of dissent, know that you
are on the wrong side of history; but that we will extend a hand if
you are willing to unclench your fist.”
6

It was while sitting in an introductory international relations graduate seminar that I first asked
the question of how Obama’s rise to the presidency would affect America’s reputation abroad.
The professor leading the seminar responded to the effect of, “well, that would be a good
dissertation topic, wouldn’t it?” which resulted in every graduate student in the room scribbling
or typing the question down. It turns out that I am the only one of that group who remained
curious enough about the question of how beliefs about states are revised given new information.
In any event, the Iran nuclear deal and the events leading up to it were the motivation
behind the experiment in Chapter 2. Since Hassan Rowhani’s election to President, Iran began
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
6
From “TRANSCRIPT: Obama’s Inaugural Address.” ABC News. January 19.
http://abcnews.go.com/Politics/Inauguration/president-obama-inauguration-speech-
transcript/story?id=6689022.

!
122!
!
sending strong signals to the international community that it was willing to work with the West
on a nuclear agreement. Upon being elected, Rowhani called for a resolution to the issue within
“months, not years” and made it a major component of his agenda (Baker 2013). September
2013 saw the first (publicly acknowledged) direct talks between the US and Iran in 30 years.
While the talks were not successful, Iran immediately followed by reaching an agreement with
the IAEA to “resolve all outstanding issues” including managed access to nuclear sites for
inspectors. They also discontinued nuclear expansion, while the existing stock of 20% enriched
uranium was placed under IAEA monitoring, awaiting conversion to harmless oxide (Erlanger
2013). Furthermore, Iran resumed relations with Britain by appointing a new diplomatic head
(Gordon and Sengupta 2013), and enacted modest social liberalization which included
terminating “fashion police” arrests for violators of Islamic law involving attire (Gladstone
2013).
On the issue of a negotiated nuclear deal with Iran, a coalition initially formed between
the US, Britain, Germany, Russia, and China, while Israel and France took strong opposing
stances to any deal deemed overly permissive. What might explain this divide? The “usual
suspect” explanation would point to differing interests, but the agreement between the US,
Russia, and China might suggest that there are significant overlaps in geopolitical interests. The
key question is: what differentiated Israel and France? Beginning with Israeli interests at play,
there seem to be three key points to consider:

1. The Geneva talks initially included a provision that Iran would be allowed to continue work
on the Arak nuclear facility, where Iran’s 20% enriched uranium was held, so long as it promised
not to begin loading fuel into a new reactor. Experts argued that Arak could generate two bombs
worth of weapons grade plutonium per year. If Iran were to produce nuclear weapons at this site,
Israel’s option of a targeted strike would be far less sustainable, since striking the completed
weapons would spread radiation on a scale similar to the Chernobyl disaster. This would
contaminate roughly 28,000 square kilometers, which would include the capital Tehran, which
!
123!
!
has a population of roughly 14 million, and the Shi’a holy city of Qom, home to roughly one
million, among other cities (Vick 2014). Still, Iran’s agreement to halt activities at Arak did not
assuage Israel’s concerns.

2. Public opposition to a deal with Iran is strong in Israel, and in parts of the United States as
well. Lobbyists within and outside the US pushed hawks in the US congress to take a strong
stance on the issue, ostensibly with the goal that Iran relinquish all demands in full (Dreyfuss
2014). But it was clear that Iran would not capitulate completely on the issue, given the possible
loss of credibility suffered domestically and internationally. Thus, some middle ground was
necessary to a negotiated settlement.

3. Iran was within a few months of “critical capacity.” If negotiations failed, Israel would find
itself in a situation where Iran was a threshold nuclear power, while the US was clearly adverse
to military action (Herzog 2013). Thus, failed negotiations would potentially put Israel at a
greater disadvantage than a less-than-satisfactory negotiated settlement.

If these issues are framed in terms of generic international relations perspectives, they
constitute three factors that affected the Israeli position on the Iranian nuclear program:
international norms (in this case, against excessive destruction caused by force), public opinion,
and relative positioning. Neither of these perspectives indicates that Israel had a clear interest in
ignoring Iran’s overtures and stalling the negotiations. A case could be made that a negotiated
settlement would be to their advantage so long as the Arak provision is satisfactorily dealt with.
Since Iran’s status as a nuclear threshold state was a fait accompli, a deal to halt further growth
of Iran’s nuclear program would protect Israel’s relative advantage in the region.
The picture of France’s motivations on the nuclear talks is muddier. Initially, France
held up the Geneva talks, refusing to accept a “sucker’s deal” on the Iranian nuclear program
(Llana 2013). A number of explanations have been offered to explain French intransigence,
including but not limited to:

1. A desire to strengthen France’s position in the middle east, which involved a strategy of
bolstering relationships with the Arab states and Israel (Mitnick 2013).

!
124!
!
2. French companies had signed billions of dollars worth of energy and military contracts with
Saudi Arabia, and thus pressured the government in an attempt to maintain the need for Saudi
military spending (Seener 2013).

3. France may have been attempting to publicly assert their power in an attempt to make up for
the embarrassment following from the recent French retreat from Syria (Landler 2014).

If these factors are reframed in the same fashion as the previous case, they constitute three
factors: relative positioning, logrolling, and credibility maintenance. In any case, French
intransigence did not last long. In November of 2013, France moved in alignment with the rest
of the “P5+1.” They benefited by being able to argue that their posture resulted in a tougher
deal, including Iran’s agreement to halt activities at Arak. In the end, however, it seems that
France gave in to international momentum on the issue (Llana 2013), and thus maintains better
relations with the United States, United Kingdom, and Germany.
While the explanations accounting for French and Israeli positions on the Geneva talks
are not comprehensive, we might consider how these causal components fit together. One
approach to this is through the lens of systemism, which attempts to diagrammatically specify
causal linkages and interactions within and across different levels of analysis (James and James
Forthcoming). A systemist diagram illustrates causal pathways, in similar fashion to path
diagrams used by structural equation models. Two factors differentiate a systemist diagram from
a path diagram: first, it is more deliberate in specifying the composition of participants, the
environment that bounds the system, the structure of rules governing interactions among
participants, and the causal mechanisms that link the variables together. Second, it has slightly
different notation. While path diagrams employ different shapes for latent and observed
variables, systemist diagrams do not use shapes to distinguish the ontological status of different
variables, but use broken lines to indicate “interrogation points,” meaning new causal factors of
!
125!
!
interest, as well as new connections between existing variables. A (non-exhaustive) systemist
diagram of state policy on the Geneva talks, might look like Figure 1-1.

Figure 1-1. Systemist Model of Geneva Policy

In this illustration, directional arrows refer to one-way causal flows, while non-
directional lines indicate interactions and top-to-bottom orientation indicates levels of analysis.
The six broad factors combine to influence the outcome variable, which in this case is a state’s
position on the Geneva talks. More generally however, we might think of the outcome variable
to be a proxy for a state’s view of whether Iran can be expected to be friendly or hostile. In any
case, the diagram seems incomplete, especially given the unlikely coalition between rivals China
and Russia, on one end, and the US, UK, France, and Germany, on the other end. The two
camps may be subject to the same international norms, but they have different alliance
!
126!
!
commitments, different geopolitical interests, and presumably different types of domestic
commitments. If these factors do not explain Israel’s dismissal of Iranian overtures, what other
factors might be at play? More generally, why are some states more or less likely to update their
images of other states when those states are sending (sometimes costly) image-disconfirming
signals? Why to some states take a “tougher” stance than the strategic context dictates? Note
that Figure 1.1 makes no allowance for individual-level factors that deal with how Iranian
overtures would be interpreted, and furthermore indicates that geopolitical interests and relative
position is some objective, measurable factor. The default position among rationalists in
international relations is that international actors update their assessments of other states in
Bayesian fashion, based on generally agreed upon prior beliefs and similar assessment and
weighting of new information, and furthermore that geopolitics and relative standing can be
observed objectively. These views have been subject to major qualifications, particularly in the
subfield of political psychology, but there is still not a satisfactory answer to the question of
“who updates, and why?” Figure 1-2 illustrates this question with a systemist interrogation point
indicated in broken lines, which problematizes how image-disconfirming signals would be
interpreted in light of a subjective assessment of geopolitical interests.

!
127!
!

Figure 1-2. Interrogation Points in Geneva Policy Systemist Model

In several respects, this dissertation is an attempt to assess this interrogation point, with
each empirical chapter serving a different function. Chapter 2 disaggregated the process of
belief revision structurally by comparing the processes of rapprochement and alienation, finding
that (1) individual differences matter such that political belief systems are a powerful
determinant of how individuals perceive the intentions of other states; (2) the costliness of
signals only matter up to a point; and (3) hostile signals are more effective in signaling intent
than reassuring signals. Chapter 3 cross-validates some of the structural and idiographic findings
from Chapter 2 using content analysis of decision-makers public statements, and finds that when
structural conditions are in line with decision-makers ideologically driven expectations, nuance
in thought about adversaries increase. Conversely when conditions did not fit ideologically
!
128!
!
driven expectations, decision-makers became more simplistic in their view of the adversary. The
bargaining game in Chapter 4 strips down political context and focuses almost exclusively on the
role of individual differences in attributional style in belief revision in a competitive setting,
finding that individuals with a situational attributional style tend to be more aware of optimum
strategies dictated by structure than those with a dispositional attributional style. What ties the
empirical chapters together is that they all emphasize the key role of subjectivity in the process
of inter-state belief revision.
In the remainder of this concluding chapter, I comment on the state of the literature on
belief revision as a whole, as follows: first, I outline scholarship that has assessed the dynamics
underlying the formation and change of state images, with an eye toward three subfields in
particular: image theory, deterrence, and operational code. The first thing I find is that the
concept of an image overlaps considerably with deterrence theory’s concept of reputation, with
marginal theoretical and practical differences between the two. This diversity of concepts seems
to have resulted in missed opportunities for engagement and development. Furthermore, neither
approach has satisfactorily explored the micro-foundations of images, although other traditions
in the field of foreign policy analysis have.
Before proceeding I should note that the question of belief revision is wide in scope,
dealing with a variety of theoretical and methodological approaches within and between
international relations and social psychology, it does not possess the hard core of assumptions
that would constitute a Lakatosian research program (Lakatos 1976). Instead, this study has been
undertaken with the approach of analytic eclecticism, as advanced by Sil and Katzenstein (2010).
Eclectic scholarship, as they advance it, has three features: first, it encourages practical
engagement between putatively unresolvable metaphysical divides. Second, it is driven by
!
129!
!
problems that are wider in scope than those dealt with by international relations paradigms, and
accordingly confronts messier “real world” problems. Third, it offers causal accounts that
pragmatically recombine and reconceptualize theories and approaches embedded within and
between research traditions. This is not with the aim of a unified synthesis, but rather to generate
flexible frameworks organized around concrete problems (Sil and Katzenstein 2010). The theory
and methodology employed herein embodies this approach, particularly in regard to how
psychological theories and methods are adapted to assess questions concerning international
politics.
Accordingly, this dissertation may be vulnerable to questions about its theoretical and
methodological orientation toward a level of analysis. If the tripartite categorization advanced
by Kenneth Waltz (1959) is a reference point, components of the theories and methods used
herein involve components of all three: the individual, state, and system. Furthermore, I also
downplay questions about the generalizability of psychological theories and methods to
questions of foreign policy, which have been dealt with at length by other scholars (Druckman et
al. 2011). What I hope to have gained, however, is a more satisfying assessment of multicausal
phenomena within a coherent framework.

International Images and Reputation: Parallel Concepts, Missed Opportunities

Pluralism and eclecticism in theoretical approaches to the study of international politics
can be desirable and valuable. However when there is a diversity of concepts used across
approaches dealing with the same phenomenon without serious engagement between those
approaches, there might be missed opportunities for theory development and refinement. This
seems to be the case with the collective treatment of state images in the international relations
!
130!
!
literature. In particular, three sub-fields of international relations stand out in this regard: image
theory, deterrence theory, and operational code.
The popularity of image theory, which can be thought of as falling under the broad
subfield of political psychology, has grown as of late due to the failure of rationalist theories to
explain the outbreak and termination of conflict (Herrmann 2013). Within this literature, there
seem to be two applications of the image concept. In some cases, images have been referred to as
“mental representations we use to frame and organize the complicated world around us” (Voss
and Dorsey 1992). In this case, an image refers to one’s view of how the political universe
works. Another characterization of an image is provided by Herrmann and Fischerkeller: “a
subject's cognitive construction or mental representation of another actor in the political world”
(Herrmann and Fischerkeller 1995). Rather than being oriented towards the political universe,
this view is concerned with particular actors in international relations. It is this definition of an
image that seems to have the most conceptual overlap with other subfields of international
relations.
Image theory stands as a counter-point to systemic theories of international relations,
such as neorealism and neoliberalism, that are more concerned with system-level outcomes than
the specific content of any state's foreign policy. Image theory, instead, is concerned with
perceptions of the other as determinants of foreign policy vis-à-vis that state, and is particularly
focused on the images held by high-level decision-makers. Despite this difference in goals,
image theorists have attempted to engage questions fundamental to (perhaps, debatably) systemic
theories, such as questions involving the conditions under which states might prefer absolute or
relative gains.
!
131!
!
One of the earliest explorations of national images was conducted by Kenneth Boulding,
who argued that images varied along two dimensions: perceived hostility versus friendliness, and
strength versus weakness (Boulding 1959). Later scholars continued in assessing dimensional
aspects of images (Deutsch and Merritt 1965). As this research program developed in the
shadow of the Cold War, the concept of the enemy image generated the most interest. For
example, White contrasted stereotypical images of the self and other as the diabolical image of
the enemy coterminous with the moral and virile self-image (White 1970).
However the most famous exposition of the enemy image was provided by Ole Holsti's
study of John Foster Dulles' perception of the Soviet Union, which has been dubbed the
“inherent bad faith” model of information processing. Holsti found that Dulles saw the Soviet
Union as implacably hostile, and discounted any information that might falsify that view. His
enemy image of the Soviet Union was grounded in the view that they had similar capability and
status, and were led by a coherent, opportunistic leadership (Holsti 1962). Holsti's model was
later applied to analyze the operational codes of John F. Kennedy, Henry Kissinger, and others
(Stuart and Starr 1981).
Later scholars explored types of images beyond the enemy image. Herrmann and
Fischerkeller offer the most well known study of this kind, wherein they attempt to remedy
image theory’s narrow focus on conflictual pairings between states of similar capability. They
advance a broader taxonomy of images that varies along three dimensions: relative power,
perceived threat versus opportunity, and cultural differences. Leaving aside questions regarding
the origins of images, they advance five ideal-typical images—enemy, ally, imperialist,
degenerate, and colony—and explore how these views affect policy toward specific states. They
make the quite plausible claim that differing images have implications on state preferences for
!
132!
!
relative and absolute gains, as well as the specific directions of foreign policy regarding other
states. The authors treat elites as the drivers of prevailing imagery, although they do make some
allowances for domestic political constraints on images (Herrmann and Fischerkeller 1995).
Image theorists have also focused on the motivational foundations of images. For
example, Lebow has argued that when leaders are domestically weak, they are motivated to
construct an image of international system that overestimates their relative power (Lebow 1981).
Others have argued that conflicts of interests between states will lead to demonization of the
opposing state in such pairings (Reeder et al. 2005). This suggests that when a conflict of
interest is remedied, enemy images should be forgotten quite quickly, but the historical record
shows that this is not always the case. For example, the fall of the Soviet Union did not instantly
alter enemy images for decision-makers within the United States (Albuyeh Forthcoming).
Certainly, motivational factors can push the momentum of images in different directions, but
shifts do not seem to occur at similar rates for all actors, even when situated within the same
country.
It is important to note that Herrmann and Fischerkeller's typology identifies referent
points for the extreme version of each image. My goal in this paper is to identify who varies in
terms of holding more or less ideal-typical images, beginning by asking which decision-makers’
images most closely resemble the ideal-typical enemy construct, particularly under conditions
where there a state is observed to be sending image-disconfirming signals, such as an enemy
signaling friendliness or an ally signaling hostility.

!
133!
!
The Relationship between Images and Reputation

While image theory attempts to explain behavior based on motives and perceptions
within states, deterrence theory focuses on material constraints and incentives in explaining state
behavior. This paradigm is concerned with how defender states use threats of retaliation to
convince challengers that an attack has a low probability of success. The credibility of a deterrent
threat hinges on the defender's military power and apparent resolve to use force (Huth 1997).
Apparent resolve is more important than actual resolve, since a challenger can never determine
how resolute a defender actually is. The challenger's task is to make use of incomplete
information to form reputational inferences about the defender. It follows that the particular
aspect of reputation that deterrence theorists are typically concerned with are reputations for
resolve.
Reputation is generally defined in one of two ways. Most commonly, it is said to be a
judgment of an actor's disposition that is then used to predict future behavior (Mercer 2010),
while others have defined it as a belief about an actor's preferences and capabilities in a given
issue area (Tomz 2007). The first definition treats reputation as holding regardless of
circumstantial, case-specific factors, while the second definition can vary from case to case.
Deterrence theorists often employ rational choice models, with varying degrees of
formality, to analyze how states might act in light of their expectations of how other actors will
behave. Effective deterrence hinges on information processing, deterrence theorists typically
assume information processing occurs in Bayesian fashion. Furthermore, they contend that
commitments are interdependent, such that signaling in one instance can affect a state's
reputation in another (Mercer 2010).
!
134!
!
While deterrence theorists typically make, at best, limited allowances for limitations in
information processing, Jonathan Mercer applies insights gained from social psychology to
deterrence theory. Borrowing from scholarship on the “fundamental attribution error,” which
holds that individuals often discount situational factors and exaggerate dispositions in making
causal attributions about others' behavior, he finds that states are given situational attributions
when they behave desirably in the eyes of other states, and dispositional attributions when they
behave undesirably. Thus if an adversary stands firm, an observer attributes it to the adversary's
disposition. But if the adversary backs down, which is presumably desirable behavior, the
observer attributes the behavior to situational factors (Ibid.). Based on his assertion that only
dispositional attributions can contribute to reputation, what seems to follow from Mercer's
argument is that actors can only create a reputation based on their undesirable behavior. Thus, an
odd implication of Mercer's theory is an adversary can only gain a reputation for having resolve,
while an ally can only gain a reputation for lacking it. Interestingly, these findings, as they
involve discounting desirable adversary behavior, are quite consistent with Holsti's inherent bad
faith model.
Images and reputations seem to have significant overlap, both conceptually and in terms
of their implications. As mentioned earlier, Herrmann and Fischerkeller's definition of an image
is “a subject's cognitive construction or mental representation of another actor in the political
world,” which they then use to make inferences about behavior towards other states. Reputations,
on the other hand, are defined either as judgments of an actor's disposition that are used to
predict future behavior (Ibid.), or a belief about an actor's preferences and capabilities in a given
issue area (Tomz 2007). I will refer to the first definition as dispositional reputation and the
second as circumstantial reputation. The definition of images align quite closely with
!
135!
!
dispositional reputation, as both constitute schemata that are used to predict the policy of other
states, which in turn informs a state's own foreign policy. In this regard, the concept of image
seems quite consistent with the dispositional definition of reputation.
The situational definition of reputation also shares some conceptual overlap with
Herrmann and Fischerkeller's definition of image, who use three dimensions to build their
typology: (1) capabilities, (2) a value for whether an actor presents a threat, an opportunity to
exploit, or a chance for mutual gain, and (3) a cultural dimension. Herrmann and Fischerkeller's
capability dimension of image maps cleanly onto the first component of the situational
reputation. Their second dimension, in particular for the values of “presents a threat” and “a
chance for mutual gain,” while it seems to refer to the payoff structure between states, might also
be thought to refer to preferences in a given issue area. In this respect, their second dimension
might be thought to map onto situational reputation as well, only factor unique to image theory
being the cultural dimension.
Herrmann and Fischerkeller's definition of image is broad enough to encompass aspects
of both definitions of reputation. In either case, it seems difficult to conceptualize an image of
the other that does not operate in similar fashion to a reputation, although reputations are
typically used to refer to reputations for credibility. It seems, therefore, that images encompass a
broader set of issue areas than reputations, and thus refer to more general foreign policy
orientations toward other states, while reputations for credibility are often applied to narrower
issue areas such as nuclear deterrence—although effective nuclear deterrence involves
substantially more than simply possessing nuclear weapons.
To illustrate the implications of this conceptual overlap, consider Herrmann and
Fisherkeller's predicted policies arising from the ideal-typical enemy image. On the three
!
136!
!
dimensions of images, enemies are thought to be comparable in capability, have an apparent lack
of mutual interests and are therefore threats, and have comparable cultures in terms of status. The
strategy they argue that follows from this is containment, which involves deterrence, building
major alliance systems, protecting geopolitical assets, and protecting credibility as an “attractive
ally,” which presumably means being perceived resolute in military conflict. Thus, the policy
that follows from the enemy image is a generalized deterrence strategy to contain the enemy's
anticipated hostile behavior. It seems fair to conclude that image and reputations are
fundamentally similar concepts.

Ideographic Approaches to State Images

While approaches in image theory and deterrence theory have generally been interested
in creating nomothetic models of images that assume that all individuals construct images
similarly, scholarship in operational code has considered how individuals differ in terms of the
views they hold of other states, although this has primarily been concerned with general attitudes
concerning all other states. Pioneered in 1969 by Alexander George (George 1969), this
literature attempts to isolate cognitive and affective aspects of decision-makers mental constructs
and categorize them into a typology of political belief systems. George attempted to assess the
philosophical content of decision-makers' operational codes, asking questions about their
attitudes to questions such as: what is the essential nature of political life?; Is the political future
predictable?--and instrumental questions such as: what is the best approach for selecting goals or
objectives for political action?; how are the goals of action pursued most effectively (Walker
1990)?
!
137!
!
Building upon George's philosophical and instrumental questions, Ole Holsti created a
typology of six operational code belief systems, which is based on an individual's view of the
nature of the political universe and the fundamental sources of conflict within it, as indicated in
Table 1.

Table 1. Holsti's Typology of Operational Code Belief Systems
What is the Fundamental Nature of the Political Universe?
What are the Fundamental
Sources of Conflict?
Harmonious (conflict is
temporary)
Conflictual (conflict is
permanent)
Human nature A D
Attributes of nations B E
International system C F

Holsti characterizes operational code as: “a configuration of ideas and attitudes in which
the elements are bound together by some form of constraint or functional interdependence,” as
well as beliefs about opponent images (Holsti 1962). It seems that such a belief system should
have implications as to how images of other states might be formed or change, especially as it
involves questions about the fundamental nature of the political universe. If an individual
believes the political universe is primarily conflictual, for example, they might be more likely to
ignore overtures from adversaries and presume that any decrease in conflict is temporary, while
one who believes the universe is harmonious would be more receptive to such overtures.
The value in this approach is that it does not treat human nature as a constant. Individuals
are expected to vary in how they are oriented toward other states and how they interpret
information coming in. They also remedy image theorists' tendency to focus on judgments about
specific other actors without reference to particular individuals' worldviews, which should
inform their image of other states. However a lingering problem with this approach, even in
!
138!
!
current iterations (Walker and Schafer 2011), is that it has not engaged other approaches that
consider how individuals vary as information processes, namely cognitive and social psychology
which in recent years have made major advances in understanding how individuals vary in their
schematic orientation towards the world. As will be shown in the following section, how
individuals view the fundamental nature of the political universe has received serious analysis by
scholars in social psychology.

Moving Forward
The choice between paying attention to and ignoring image-disconfirming information
had clear policy implications both in the case of the Iranian nuclear deal and in the aid-for-
reform debates at the end of the Cold War. In both instances, the distinction amounted to the
choice between the continuation of a long-established aggressive posture toward the adversary,
on the one hand; and the unprecedented extension of an olive branch, on the other. If images are
not the immediate causal antecedents of foreign policy decision-making, they are very close to it.
The solution to debates concerning when states might favor absolute or relative gains does seem
to be linked to views of the enemy, as image theorists contend. However, we might trace the
causal process a further causal step backward to assess the individual-level sources of contested
images. This might be less of an imperative when we are concerned with autocratic states or
crisis situations, but it is crucial to understanding relations involving democratic states in non-
crisis situations, which constitute a great deal of international politics.
U.S. foreign policy has long been concerned with projecting an image of resolve to deter
enemies and reinforce links with allies, as it is often the justification for military intervention.
Indeed, the United States has suffered tragic losses in the name of protecting its image. As
!
139!
!
Thomas Schelling said, “we lost thirty thousand dead in Korea to save face for the United States
and the United Nations, not to save South Korea for the South Koreans, and it was undoubtedly
worth it. Soviet expectations about the behavior of the United States are one of the most valuable
assets we possess in world affairs”
1
. However if U.S. behavior is filtered through other state
elites' idiosyncratic cognitive schema, and if states cannot straightforwardly engineer their
images as I have suggested, it might not be worthwhile to make great commitments for the
purpose of protecting an image. It is simply too difficult to determine whether a bluff will be met
with a raise or a fold, especially under conditions of relative parity between powers as was the
case during the Cold War.
A compelling question is how the cognitive substructure underlying images interact with
intersubjectivity. If we consider a quasi-consensus involving the image of another state (for
example: North Korea's rogue image) as a social construction, linking cognitive foundations of
images with their discursive outcomes could provide one potential solution to the elusive link
between political psychology and constructivism. In this case, we would have to look at how
individual psychology, institutional structures, and power dynamics within and between states
interact in order to create some loose consensus regarding the reputation of other states. It seems
that the Republican and Democratic platforms regarding foreign policy could be reflective of the
psychology underlying political ideology, although causal winds likely travel in both directions
between these factors.
As mentioned earlier, one of the main weaknesses of operational code theory is that it is
based almost entirely on inductive theory and intuition. Inductive theory building has clear value,
especially in the areas of culture, norms, and ideas. However there is always room for
engagement with approaches that might have insights to offer. In the case of operational code,
!
140!
!
the greatest potential for cross-pollination seems to lie in social and cognitive psychology. Most
relevant to the questions assessed by operational code are psychological studies of political
ideology, which deal with one of the key questions of operational code: how people view the
fundamental nature of the political universe.
Recall Jost et al.'s (2009) analysis of political ideology, which includes a link between
orientations concerning threat and uncertainty, and in particular beliefs about system instability.
In the same vein, Duckitt (2001) cites conservatism as being linked with the view of the world as
being threatening and dangerous along one cognitive-social dimension, and as an amoral
competitive jungle along another. In other words, one's political orientation is linked to their
view of the political world as being conflictual or harmonious, which has implications for
behavior toward out-groups. We might consider how these ideas map onto Holsti's typology of
Operational Code Belief Systems. Like the psychological literature on political ideology,
operational code focuses on individual-level differences in information processing. Like the
concept of political ideology, operational code involves individuals' fundamental views about the
nature of the political universe and about the fundamental sources of conflict. Since political
ideology has also been linked to conceptual complexity in making causal attributions, it also has
implications for operational code's dimension that asks about an decision-maker's view of the
fundamental sources of conflict. We may be able to identify one's orientation on the
“fundamental nature of the political universe” dimension via political ideology, and one's
orientation along that axis might condition their position on the “fundamental sources of
conflict” axis, as follows:

!
141!
!
A Reformulated Typology of Belief Systems Regarding Adversary Images
What is the Fundamental Nature of the Political Universe?
What are the Fundamental
Sources of Conflict?
Harmonious (conflict is
temporary)
Conflictual (conflict is
permanent)
Human nature Liberal N/A (or conservative)
Attributes of nations Liberal Conservative (or N/A)
International system Liberal N/A
Regarding the question of the fundamental nature of the political universe, liberalism
should be associated with the belief that the universe is harmonious and that conflict is
temporary, whereas conservatism should be associated with the view that conflict is permanent
and that the universe is inherently conflictual. For the sources of conflict question, the literature
on cognitive complexity and ideology would place liberals in any of the three positions for
human nature, attributes of nations, and international system. Conservatives, on the other hand,
would be placed primarily in the human nature or attributes of nations position for the sources of
conflict axis.
Since it seems that the answer to the “nature of the universe” question should condition
responses to the “fundamental sources of conflict” question, a psychologically informed
operational code seems to provide an analytical refinement of the belief system typology. The re-
mapping also indicates that the operational code typology might be over-specified. As Stephen
Walker found in his review of the operational code literature, there is a great deal of hybrid type
individuals according to this typology, who overlap across more than one type
3
. It might be
preferable to operational code as a bipartite typology, given the broad and systematic set of
implications one's view of the political universe can have.
The fields of international relations and foreign policy analysis are ripe for the return of
scholarship in a similar vein as operational code. Given the recent boom in international relations
!
142!
!
research involving psychological approaches and experimental methodology, the field seems to
be on the precipice of a resurgence of individual-level research. Scholarship in international
relations could benefit from a program similar to operational code analysis, although a re-
branding would be helpful to create a distance from an outdated approach. “Operational
Psychology” might be an apt title. Regardless of the name, such a shift would be timely and
necessary for a field that has for so long neglected its roots in studying how individual
particularities can impact international politics.
!
!
!
!
References

“500 Million Approved for Soviet Aid.” 1992. CQ Almanac 1991 47th ed. Congressional
Quarterly: 463–70.
Adorno, T. W., Daniel J. Levinson, and Nevitt Sanford. 1950. “1. The Authoritarian
Personality.”
Albuyeh, Rod. Forthcoming. “Enmity Is in the Eye of the Beholder: The Microfoundations of
Images.”
———. Forthcoming. “The Ideology of Irredentism: Identity, Threat, and Realpolitik.”
Allison, Graham T. 2012. “Conceptual Models and the Cuban Missile Crisis.” American
Political Science Review 63 (03): 689–718. doi:10.2307/1954423.
Altemeyer, Bob. 1981. Right-Wing Authoritarianism. [Winnipeg]: University of Manitoba Press.
Baker, Peter. 2013. “Breaking Ground, Obama Converses With Iran Leader: [Foreign Desk].”
New York Times, Late Edition (East Coast), September 28, sec. A.
!
143!
!
Boulding, K. E. 1959. “National Images and International Systems.” Journal of Conflict
Resolution 3 (2): 120–31. doi:10.1177/002200275900300204.
Chirumbolo, Antonio, Alessandra Areni, and Gilda Sensales. 2004. “Need for Cognitive Closure
and Politics: Voting, Political Attitudes and Attributional Style.” International Journal of
Psychology 39 (4): 245–53. doi:10.1080/00207590444000005.
Cochran, John K., Denise Paquette Boots, and Mitchell B. Chamlin. 2006. “Political Identity and
Support for Capital Punishment: A Test of Attribution Theory.” Journal of Crime and
Justice 29 (1): 45–79. doi:10.1080/0735648X.2006.9721217.
Collier, David. 2011. “Understanding Process Tracing.” PS: Political Science & Politics 44 (04):
823–30. doi:10.1017/S1049096511001429.
Deutsch, K, and R Merritt. 1965. “Effects of Events on National and International Images.” In
International Behavior: A Social-Psychological Analysis, 130–87. New York: Holt,
Rinehart and Winston.
Dreyfuss, Bob. 2014. “Can Congress, Hawks and the Israel Lobby Wreck the US-Iran Talks?”
The Nation. Accessed February 22. http://www.thenation.com/blog/176893/can-
congress-hawks-israel-lobby-wreck-us-iran-talks.
Druckman, James N, Donald P Green, James H Kuklinski, and Arthur Lupia, ed. 2011.
Cambridge Handbook of Experimental Political Science. Cambridge; New York:
Cambridge University Press.
Duckitt, John. 2001. “A Dual-Process Cognitive-Motivational Theory of Ideology and
Prejudice.” Advances in Experimental Social Psychology 33: 41–114.
Erlanger, Steven. 2013. “Iran Would Eliminate Stock of Some of Its Enriched Uranium Under
Deal: [Foreign Desk].” New York Times, Late Edition (East Coast), November 23, sec. A.
!
144!
!
Fearon, James D. 1997. “Signaling Foreign Policy Interests Tying Hands versus Sinking Costs.”
Journal of Conflict Resolution 41 (1): 68–90. doi:10.1177/0022002797041001004.
Ford, Thomas E., Frank Gambino, Hanjoon Lee, Edward Mayo, and Mark A. Ferguson. 2004.
“THE ROLE OF ACCOUNTABILITY IN SUPPRESSING
MANAGERS’PREINTERVIEW BIAS AGAINST AFRICAN-AMERICAN SALES
JOB APPLICANTS.” Journal of Personal Selling and Sales Management 24 (2): 113–
24.
George, Alexander L, and Andrew Bennett. 2005. Case Studies and Theory Development in the
Social Sciences. Cambridge, Mass.: MIT Press.
George, Alexander L. 1969. “The ‘Operational Code’: A Neglected Approach to the Study of
Political Leaders and Decision-Making.” International Studies Quarterly 13 (2): 190.
doi:10.2307/3013944.
Gladstone, Rick. 2013. “Senate Bill Threatens Nuclear Deal, Analysts Say: [Foreign Desk].”
New York Times, Late Edition (East Coast), December 21, sec. A.
Gomez, Brad T., and J. Matthew Wilson. 2003. “Causal Attribution and Economic Voting in
American Congressional Elections.” Political Research Quarterly 56 (3): 271–82.
doi:10.1177/106591290305600303.
Gordon, Michael R., and Somini Sengupta. 2013. “France Sets Conditions for Including Iran, an
Assad Ally, in Syria Peace Talks: [Foreign Desk].” New York Times, Late Edition (East
Coast), September 24, sec. A.
Herrmann, Richard. 2013. “Perceptions and Image Theory in International Relations.” In The
Oxford Handbook of Political Psychology. Oxford: Oxford University Press.
!
145!
!
Herrmann, Richard K., and Michael P. Fischerkeller. 1995. “Beyond the Enemy Image and
Spiral Model: Cognitive–strategic Research after the Cold War.” International
Organization 49 (03): 415–50. doi:10.1017/S0020818300033336.
Herzog, Michael. 2013. “Israel’s New Focus on the Iran Nuclear Deal.” The Guardian,
December 1, sec. Comment is free.
http://www.theguardian.com/commentisfree/2013/dec/01/israel-new-focus-iran-nuclear-
endgame.
Holsti, Ole R. 1962. “The Belief System and National Images: A Case Study.” Journal of
Conflict Resolution 6 (3): 244–52. doi:10.1177/002200276200600306.
Houghton, David Patrick. 1996. “The Role of Analogical Reasoning in Novel Foreign-Policy
Situations.” British Journal of Political Science 26 (04): 523–52.
doi:10.1017/S0007123400007596.
Huth, Paul K. 1997. “Reputations and Deterrence: A Theoretical and Empirical Assessment.”
Security Studies 7 (1): 72–99. doi:10.1080/09636419708429334.
James, Carolyn, and Patrick James. Forthcoming. “Systemism and Foreign Policy Analysis.” In .
Jost, John T., Christopher M. Federico, and Jaime L. Napier. 2009. “Political Ideology: Its
Structure, Functions, and Elective Affinities.” Annual Review of Psychology 60 (1): 307–
37. doi:10.1146/annurev.psych.60.110707.163600.
Lakatos, Imre. 1976. “Falsification and the Methodology of Scientific Research Programmes.” In
Can Theories Be Refuted?, 205–59. Springer.
Landler, Mark. 2014. “Obama Greets France’s Leader, but Warns Against Doing Business With
Iran.” The New York Times, February 11.
!
146!
!
http://www.nytimes.com/2014/02/12/world/middleeast/obama-warns-companies-against-
trying-to-evade-iran-sanctions.html.
Lebow, Richard Ned. 1981. Between Peace and War: The Nature of International Crisis. Johns
Hopkins University Press Baltimore.
Llana, Sara Miller. 2013. “‘Spoiler’ Role Over, France Joins the Party Celebrating an Iran
Nuclear Deal.” Christian Science Monitor, November 24.
http://www.csmonitor.com/World/Security-Watch/2013/1124/Spoiler-role-over-France-
joins-the-party-celebrating-an-Iran-nuclear-deal.
Mercer, Jonathan, Cornell University Press. 2010. Reputation and International Politics. Ithaca;
London: Cornell University Press.
Mill, John Stuart. 1891. A System of Logic: Ratiocinative and Inductive!: Being a Connected
View of the Principles of Evidence and the Methods of Scientific Investigation. Hawaii:
University of the Pacific.
Mitnick, Joshua. 2013. “Israelis Applaud France for ‘Standing up’ as Iran Talks Resume in
Geneva.” Christian Science Monitor, November 19.
http://www.csmonitor.com/World/Middle-East/2013/1119/Israelis-applaud-France-for-
standing-up-as-Iran-talks-resume-in-Geneva.
Odell, John S. 2001. “Case Study Methods in International Political Economy.” International
Studies Perspectives 2 (2): 161–76.
Putnam, Robert D. 1988. “Diplomacy and Domestic Politics: The Logic of Two-Level Games.”
International Organization 42 (03): 427–60. doi:10.1017/S0020818300027697.
!
147!
!
Reeder, Glenn D., John B. Pryor, Michael JA Wohl, and Michael L. Griswell. 2005. “On
Attributing Negative Motives to Others Who Disagree with Our Opinions.” Personality
and Social Psychology Bulletin 31 (11): 1498–1510.
Schelling, Thomas C. 2008. Arms and Influence. New Haven, Conn.; London: Yale University
Press.
Seener, Barak. 2013. “Why Israel, Gulf States Are Wary of Iran Nuclear Talks.” CNN.
http://www.cnn.com/2013/11/19/opinion/iran-israel-seener-oped/index.html.
Sidanius, Jim, and Felicia Pratto. 1993. “The Dynamics of Social Dominance and the
Inevitability of Oppression.” Prejudice, Politics, and Race in America Today: 173–211.
Sil, Rudra, and Peter J. Katzenstein. 2010. “Analytic Eclecticism in the Study of World Politics:
Reconfiguring Problems and Mechanisms across Research Traditions.” Perspectives on
Politics 8 (02): 411–31. doi:10.1017/S1537592710001179.
Stuart, Douglas, and Harvey Starr. 1981. “The ‘Inherent Bad Faith Model’ Reconsidered: Dulles,
Kennedy, and Kissinger.” Political Psychology 3 (3-4): 1–33. doi:10.2307/3791139.
Tarrow, Sidney. 2010. “The Strategy of Paired Comparison: Toward a Theory of Practice.”
Comparative Political Studies 43 (2): 230–59. doi:10.1177/0010414009350044.
Tetlock, Philip E. 1983. “Accountability and Complexity of Thought.” Journal of Personality
and Social Psychology 45 (1): 74–83. doi:10.1037/0022-3514.45.1.74.
———. 1985. “Accountability: A Social Check on the Fundamental Attribution Error.” Social
Psychology Quarterly 48 (3): 227. doi:10.2307/3033683.
———. 1992. “The Impact of Accountability on Judgment and Choice: Toward a Social
Contingency Model.” Advances in Experimental Social Psychology 25 (3): 331–76.
!
148!
!
Tomz, Michael. 2007. Reputation and International Cooperation: Sovereign Debt across Three
Centuries. Princeton: Princeton University Press.
Vick, Karl. 2014. “If Iran Can Get This Reactor Online, Israel May Not Be Able to Bomb It.”
Time. Accessed February 22. http://world.time.com/2013/10/29/if-iran-can-get-this-
reactor-online-it-may-be-invulnerable-to-military-attack/.
Voss, James F., and Ellen Dorsey. 1992. “Perception and International Relations: An Overview.”
In Political Psychology and Foreign Policy, edited by E. Singer and V. Hudson, 3–30.
Boulder, CO, US: Westview Press.
Walker, Stephen G, Malici, and Mark Schafer. 2011. Rethinking Foreign Policy Analysis: States,
Leaders, and the Microfoundations of Behavioral International Relations. New York:
Routledge.
Walker, Stephen G. 1990. “The Evolution of Operational Code Analysis.” Political Psychology
11 (2): 403. doi:10.2307/3791696.
Waltz, Kenneth N. 1959. Man, the State, and War; a Theoretical Analysis. New York: Columbia
University Press.
Weiner, Bernard, Danny Osborne, and Udo Rudolph. 2011. “An Attributional Analysis of
Reactions to Poverty: The Political Ideology of the Giver and the Perceived Morality of
the Receiver.” Personality and Social Psychology Review 15 (2): 199–213.
doi:10.1177/1088868310387615.
White, Ralph K. 1970. Nobody Wanted War: Misperception in Vietnam and Other Wars. Garden
City, N.Y.: Doubleday.
Wilcox, Rand R. 1998. “How Many Discoveries Have Been Lost by Ignoring Modern Statistical
Methods?” American Psychologist 53 (3): 300.
!
149!
!
Wilcox, Rand R., and H. J. Keselman. 2003. “Modern Robust Data Analysis Methods: Measures
of Central Tendency.” Psychological Methods 8 (3): 254.

Abstract (if available)

Abstract This article-style dissertation consists of three studies tied together by the theme of belief revision in international relations. The first study expands the focus of belief revision to compare how its processes differ between the contexts of rapprochement and alienation, and employs a novel survey experiment to achieve this. The second study narrows the focus to adversary relations, leveraging a natural experiment to examine belief revision in the context of the Cold War. The third study strips out political context to assess in isolation individual differences in sensitivity to information about past actions, by way of an incentivized market-entry bargaining game. The overarching finding across these three studies is that individual subjectivity is the most powerful determinant of belief revision in international relations, particularly in the case of political belief systems.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

When passions run high: emotions and the communication of intentions in face-to-face diplomacy

PDF

The political psychology of democracy promotion, or how democracies democratically promote democracy promotion

PDF

Forced migration as a cause and consequence of conflict

PDF

Hard feelings: friendship, betrayal, and outrage in international relations

PDF

The paradox of personalism: internal political constraints on nuclear posture in authoritarian regimes

PDF

The IO contestation project: a multi-method approach towards understanding the ways states contest international organizations.

PDF

Anthropomorphic sociality theory: how connections to nonhumans connect us to humans

PDF

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

PDF

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

PDF

Perspectives on state capacity and the political geography of conflict

PDF

Are you as old as you feel? A quantitative and qualitative analysis of leader age and foreign policy decision-making

PDF

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

PDF

Sexual problems among midlife and older adults: individual differences in sexual distress, comorbidity patterns and risk factors, and dyadic pathways linking to sexual dissatisfaction

PDF

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

PDF

Offline social functioning and online communication: how social competence translates to an online context

PDF

Expressing values and group identity through behavior and language

PDF

When bullets are not enough: determinants of chemical weapons usage and abstention

PDF

Information, public opinion, and international relations

PDF

The social making of authoritarian environmentalism: protest-litigation nexus and policy changes in China

Asset Metadata

Creator Albuyeh, Rod (author)

Core Title Thawing rivalries and fading friendships: a multi-method approach to rapprochement and alienation

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Political Science and International Relations

Publication Date 06/30/2018

Defense Date 08/01/2016

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

Tag Foreign policy,images,International relations,OAI-PMH Harvest,political psychology,Reputation,updating

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor James, Patrick (committee chair), Graham, Jesse (committee member), Rathbun, Brian (committee member)

Creator Email albuyeh@gmail.com,albuyeh@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-258752

Unique identifier UC11279396

Identifier etd-AlbuyehRod-4492.pdf (filename),usctheses-c40-258752 (legacy record id)

Legacy Identifier etd-AlbuyehRod-4492.pdf

Dmrecord 258752

Document Type Dissertation

Format application/pdf (imt)

Rights Albuyeh, Rod

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