University of Southern California Dissertations and Theses

The structure of strategic communication: theory, measurement, and effects

(USC Thesis Other)

The structure of strategic communication: theory, measurement, and effects

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content i

THE STRUCTURE OF STRATEGIC COMMUNICATION:
THEORY, MEASUREMENT, AND EFFECTS

DEREK J. HARMON

University of Southern California
Department of Management and Organization
3670 Trousdale Parkway – BRI 306
Los Angeles, CA 90089-0808
djharmon@usc.edu
651-271-0320

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BUSINESS ADMINISTRATION)

August 9, 2016

ii

ABSTRACT

This dissertation advances a novel approach that I refer to as the structure of strategic
communication. Leveraging theory on how people naturally structure their arguments, this
approach contends that organizational actors deploy arguments to influence others at two
structurally distinct levels—within the rules of the game or about the rules of the game. This
dissertation’s primary claim is that talking more about the rules of the game, which exposes the
assumptions underlying our social institutions to direct examination, may have profound
implications. I build evidence for this claim in two ways. First, I develop a new measurement
called the argument structure ratio (ASR) that conceptually and empirically captures how
explicit a speaker makes these assumptions in their communication. I outline a three-step
methodology for measuring the ASR of any collection of written texts. Second, I theorize and
empirically demonstrate how the ASR impacts an audience’s reaction. Using all public speeches
made by the Chairperson of the United States Federal Reserve from 1998 to 2014, I show that
the more they expose the assumptions underlying the Federal Reserve System, the more their
speeches produce market uncertainty. I argue that these findings fundamentally change how we
think about the role of strategic communication in market contexts. More generally, this work
provides a new way to conceptualize and study strategic communication that extends well
beyond financial markets to a variety of different organizational contexts and across multiple
levels of analysis. Taken together, this dissertation provides a theoretical and methodological
foundation upon which to conduct research on the structure of strategic communication.

iii

TABLE OF CONTENTS

ABSTRACT...................................................................................................................................... ii
LIST OF TABLES ............................................................................................................................v
LIST OF FIGURES ......................................................................................................................... vi
ACKNOWLEDGEMENTS ............................................................................................................ vii
CHAPTER 1 – THE STRUCTURE OF STRATEGIC COMMUNICATION...................................1
EXIS TING APPROACH TO STRATEGIC COMMUNICATION........................................................................... 3
Theoretical Perspectives..................................................................................................................................................... 3
Empirical Methodology ...................................................................................................................................................... 6
Limitations to Existing Approach ................................................................................................................................... 6
INTRODUCING THE S TRUCTURE OF STRATEGIC COMMUNICATION ................................................... 8
The Structure of Strategic Communication ................................................................................................................. 8
Dissertation Outline........................................................................................................................................................... 10
CHAPTER 2 – ARGUMENT STRUCTURE RATIO: MEASURING THE EXPLICITNESS OF
OUR ASSUMPTIONS IN COMMUNCIATION ............................................................................ 11
ABSTRACT ............................................................................................................................................................................. 11
INTRODUCTION .................................................................................................................................................................. 12
ARGUMENT STRUCTURE............................................................................................................................................... 14
The Toulmin Model ........................................................................................................................................................... 14
Two Structural Levels of Communication .................................................................................................................. 15
ARGUMENT STRUCTURE RATIO ............................................................................................................................... 17
Step 1: Define the discursive space and corpus ......................................................................................................... 18
Step 2: Clarify the two structural levels of communication in your corpus ...................................................... 19
Step 3: Code your corpus based on these two structural levels ............................................................................. 20
ILLUS TRATION AND VALIDATION: FEDERAL RE S ERVE SPEECHE S ..................................................... 22
Research Context ............................................................................................................................................................... 22
Face Validity........................................................................................................................................................................ 26
Comparison with Other Measures ................................................................................................................................ 27
CONCLUS ION AND FUTURE DIRECTIONS ............................................................................................................ 30
CHAPTER 3 – THE EFFECT OF ARGUMENT STRUCTURE ON MARKET UNCERTAINTY:
FEDERAL RESERVE CHAIRPERSON SPEECHES FROM 1998 – 2014 .................................... 32
ABSTRACT ............................................................................................................................................................................. 32
INTRODUCTION .................................................................................................................................................................. 33
UNITED S TATE S FEDERAL RES ERVE...................................................................................................................... 37
Federal Reserve System.................................................................................................................................................... 37
Federal Reserve Communications ................................................................................................................................. 39
ARGUMENT STRUCTURE AND MARKET UNCERTAINTY............................................................................. 40
Argument Structure .......................................................................................................................................................... 40
Argument Structure Ratio............................................................................................................................................... 43
Argument Structure Ratio and Market Uncertainty ............................................................................................... 45
EMOTION IN MARKETS .................................................................................................................................................. 47
Speech Emotion .................................................................................................................................................................. 48
Audience Emotion.............................................................................................................................................................. 50
METHODS ............................................................................................................................................................................... 52
Sample ................................................................................................................................................................................... 52
Endogeneity ......................................................................................................................................................................... 53
Dependent Variable ........................................................................................................................................................... 54
Independent Variables...................................................................................................................................................... 55
Control Variables............................................................................................................................................................... 59
RE S ULTS ................................................................................................................................................................................. 64
Endogeneity Considerations............................................................................................................................................ 72
Robustness Checks............................................................................................................................................................. 77
DISCUSS ION AND CONCLUS IONS .............................................................................................................................. 80
Implications for Institutional Theory ........................................................................................................................... 83
iv

Implications for the Sociology of Financial Markets ............................................................................................... 85
Future Research Directions............................................................................................................................................. 87
CHAPTER 4 – CONCLUSION ...................................................................................................... 91
LEVERAGING THE ARGUMENT STRUCTURE RATIO ..................................................................................... 96
ASR Further Theorizing .................................................................................................................................................. 96
ASR in Other Contexts ..................................................................................................................................................... 97
ASR Automated Tool ........................................................................................................................................................ 99
EXPANDING THE CONCEPT OF STRUCTURE.................................................................................................... 101
Structural As pects of Communication ....................................................................................................................... 101
Grammatical Structure .................................................................................................................................................. 103
Other Types of Communication Structure ............................................................................................................... 104
CONCLUDING THOUGHTS .......................................................................................................................................... 105
REFERENCES ............................................................................................................................. 107

v

LIST OF TABLES

TABLE 1 – Two Structural Levels of Fed Communication............................................................. 24
TABLE 2 – Sample Fed Speech Titles and Associated ASR ............................................................ 26
TABLE 3 – Comparing ASR with Other Measures ........................................................................ 29
TABLE 4 – Descriptive Statistics and Pearson Correlation Statistics ............................................. 65
TABLE 5 – Regression Models Predicting Market Uncertainty (t -1 to t0) ........................................ 68

vi

LIST OF FIGURES

FIGURE 1 – The Toulmin Model.................................................................................................... 14
FIGURE 2 – Average ASR of Fed Chairperson Speeches from 1998 to 2014 .................................. 27
FIGURE 3 – Fed Chairperson Speeches from 1998 - 2014 .............................................................. 53
FIGURE 4 – Main Effect of ASR on Market Uncertainty ............................................................... 70
FIGURE 5 – Interaction between ASR and Speech Positive Tone................................................... 71
FIGURE 6 – Interaction between ASR and Audience Fear............................................................. 72
FIGURE 7 – Ref erence Distribution of Impact for Covariates........................................................ 75

vii

ACKNOWLEDGEMENTS
My dissertation, along with my broader thinking and approach to scholarship, has been
shaped by many people over the last six years. I would like to take the time here to acknowledge
these individuals, as my accomplishments could not have been achieved without them. Nor could
I have finished this dissertation without their encouragement.
First and foremost, I would like to thank my advisor, dissertation committee chair, and
good friend, Peer Fiss. Peer has been a constant source of intellectual and emotional support for
me throughout my time in the program. His mentorship enabled me to find my confidence when
it waned, and my voice when I thought I had none. Knowing that Peer was always 100 percent
behind me, believing in my passions, my intuitions, and my decisions, was the primary reason I
am where I am today.
I would also like to thank the rest of my dissertation committee. Kyle Mayer helped me
manage important relationships over my tenure in the program, and provided insightful feedback
to better contextualize my work for different audiences. Scott Wiltermuth, often while beating
me in a friendly game of tennis, was always kind enough to offer his honest and straightforward
advice on my work and life in the profession. Jerry Hoberg was invaluable to my committee,
bringing to light both the technical complexities of my work as well as the larger story my
research was telling about the financial markets.
This dissertation, and the ideas within, could not have been developed if it were not for
my coauthors and good friends, Sandy Green and Tom Goodnight. Sandy was the person who
taught me how to identify my passion, as well as how to actually conduct research on it. I am
forever grateful for Sandy’s generosity in time and energy, teaching me everything he could to
prepare me for this profession. Tom was the person who constantly stimulated novelty in my
viii

thinking. Every time we talked, I walked away with a dozen research topics, all of which were
entirely plausible and similarly exciting. Sandy and Tom have taught me how this profession can
indeed be a manifestation of one’s passions. This dissertation is evidence of that lesson.
I would also like to thank my many friends at USC who helped me in countless ways.
Vern Glaser, Kari Olsen, and Nan Jia were my pillars of support, acting as sounding boards for
ideas and questions that ranged from being perfectly reasonable to borderline delusional. And I
am grateful that they never judged me for that. Mariam Krikorian, Roshni Raveendhran, Adele
Xing, and Jordian Rahimian provided constant encouragement and support, particularly during
the most trying times. Sarah Bonner was my most trusted mentor during my time at USC. I owe
a lot to her advice and wisdom over the years, as well as her unending kindness.
I would also like to thank several important people who helped me on Chapter 3 of this
dissertation in particular. Clare Chang, Sandhya Nadadur, and Heather Rietzfield were my
wonderfully hardworking and incredibly talented research assistants. They worked tirelessly with
me to understand the Federal Reserve context and code literally hundreds of speeches. I could
not have completed this dissertation without them. I would also like to thank Richard Peterson
and his teams at MarketPsych and Thomson Reuters for allowing me access to his wonderfully
rich market sentiment data.
I owe a great deal to those who have provided me with the emotional support I needed
during several trying years in this program. I would first like to thank my family, and in
particular, my mom and dad, who believed in me and stood by me even during my lowest points.
I know I was not always the most pleasant person to be around, and having their unending
support was necessary. I also would like to thank Bruce Howard for his years of helping me
understand myself and my relationships better. I have told this to Bruce before, but it is worth
ix

saying again. While this program wasn’t the most intellectually challenging thing I’ve ever done,
it was certainly the most emotionally challenging. Bruce gave me the skills and insight to not
only take on this challenge but come out the other side a more mature, integrated, better, and
happier person.
Finally, I would like to thank a number of friends who heard versions of this dissertation
over time and provided invaluable feedback to help nudge these ideas in the right direction and
mold it into what it is today. I would like to thank Eero Vaara, David Tan, Emilio Marti, Linda
Putnam, and Charlene Zietzsma. I also would like to thank Tim Pollock and the INFORMS
Dissertation Proposal Competition judges, my friends at the 2015 Medici Summer School
workshop, the attendees at the PhD workshop at the 2015 Alberta Institutions Conference, and
the Organizations and Strategy group at USC. Finally, I want to thank the incredibly insightful
feedback I received when presenting this work at the University of Georgia, Aalto University,
Cambridge University, HEC Paris, Michigan State University, Texas A&M, INSEAD, Arizona
State University, University of Michigan, UCLA, and the University of Alberta.

1

CHAPTER 1 – THE STRUCTURE OF STRATEGIC COMMUNICATION

CHAPTER 1

THE STRUCTURE OF STRATEGIC COMMUNICATION

Language plays a dual role in society. The words we use can embed us further within as
well as free us from our historical and culturally contingent surroundings. This two-fold
understanding of language is featured prominently in many foundational texts across the
humanities and social sciences. Drawing on the work of Heidegger (1927) and Wittgenstein
(1953), linguists and philosophers like Burke (1969), Rorty (1981), and Habermas (1984) argue
that language not only can dominate, control, and coerce human action, but it can also enable us
to recognize the contingencies of our social world and push back against the status quo.
Anthropologists like Geertz (1973) and sociologists like Berger and Luckmann (1966), Gramsci
(1971), and Goffman (1978) echo these same themes by articulating how language is
simultaneously a source of cultural hegemony and social change.
This dual role of language recently found its way into organizational and institutional
analysis in the management field as a possible explanation for how actors simultaneously
maintain but also change the very institutions in which they live. Organizational theorists often
call this the paradox of embedded agency. The apparent conundrum is that if actors’ beliefs,
language, and actions are all conditioned by the institutions in which they live, how can these
actors also change these very same institutions (Battilana & D’Aunno, 2009; Holm, 1995)?
Language, with its dual nature, was one explanation. Indeed, scholars had already demonstrated
in separate studies that language seems to explain institutional maintenance (Green, 2004; Green,
Li, & Nohria, 2009; Phillips, Lawrence, & Hardy, 2004) as well as changes to our institutions
2

(Lawrence, Hardy, & Phillips, 2002; Seo & Creed, 2002; Suddaby & Greenwood, 2005). Despite
this initial evidence, this idea that language might serve as a way out of this paradox still
remained too abstract. In particular, the precise aspects of communication that actually lead to
maintenance as opposed to change remained vague and elusive—until just recently.
In particular, Harmon, Green, and Goodnight (2015) proposed that the way we structure
our arguments may be one approach to identifying the precise inflection point in our everyday
communication that explains these two distinct outcomes. Harmon and colleagues draw on
Toulmin’s (1958) model of argument structure to propose that actors can communicate at two
structurally distinct levels. At one level, actors can argue within the rules of the game, or what
Harmon and colleagues call “intrafield rhetoric,” leaving implicit their deeper institutionalized
assumptions. At another level, actors can argue about the rules of the game themselves, what
Harmon and colleagues call “interfield rhetoric,” exposing and talking explicitly about these
fundamental institutionalized assumptions. These researchers theorized that arguing within the
rules of the game relates more to institutional maintenance because communication at this level
reproduces and reinforces the prevailing consensus surrounding the very assumptions that were
left implicit. In contrast, arguing about the rules of the game relates more to institutional change
because discussing these assumptions directly exposes the contingencies of our taken-for-granted
way of doing things and puts them at risk for alteration. While this basic conceptual distinction
holds substantial promise, its theoretical and pragmatic implications have yet to be explored.
The aim of this dissertation is to leverage this conceptual distinction—between
arguing within the rules of the game and arguing about the rules of the game—to develop a
theoretically and empirically novel approach to the study of strategic communication. In
doing so, this dissertation significantly advances Harmon, Green, and Goodnight’s (2015)
3

original work in several important ways. In particular, the current depiction of two structural
levels of communication remains too abstract and descriptive to be pragmatically useful for
actors in organizations. Like other social theorists recognizing the dual role of language in
society, Harmon and colleagues point to these two structural levels of communication as if
they were pure or abstract forms and describe them as separate and distinct processes.
However, everyday language usage is rarely so neatly distinguishable. As a result, we need a
way to conceptualize how organizational actors might flexibly and strategically exploit these
two structurally distinct levels of communication in everyday public discourse. Moreover,
since organizational actors regularly use communication to influence specific audiences, we
also need additional theorization surrounding the potential impact of communicating at these
different structural levels. By extending Harmon, Green, and Goodnight’s (2015) original
ideas to situations where organizational actors use language deliberately to influence others,
this dissertation aims to uncover an entirely new way of studying strategic communication.
EXISTING APPROACH TO STRATEGIC COMMUNICATION
Theoretical Perspectives
Strategic communication refers to the language used by actors with the aim to influence
others. In this sense, strategic communication is often intentional, reflecting a deliberate choice
by actors in regard to what and how they talk. However, strategic communication need not
always be intentional or deliberate. For instance, actors may desire to influence others, but they
may be uncertain about how an audience will react to a particular argument. As a result, their
choice of words may result in unintended consequences which nevertheless stemmed from a
strategic attempt aimed to influence others. Over the last 35 years, four theoretical
perspectives—impression management, rhetorical theory, framing, and storytelling—have
4

developed and offer different perspectives on the study of strategic communication. Each
perspective tackles a slightly different question and therefore considers different aspects of
language to be the primary driving force for influencing others.
Scholars adopting an impression management perspective emphasize how actors can use
language strategically to manage their own personal image or responsibility when negative
events threaten how others perceive them. In particular, when one’s image or reputation gets
threatened, actors have a choice to either acknowledge or deny responsibility for what has
transpired. Researchers have used this simple choice to explore the conditions under which
acknowledging and denying responsibility is most effective (Bettman & Weitz, 1983; Elsbach,
1994; Lamin & Zaheer, 2012; Salancik & Meindl, 1984; Schlenker, 1980; Staw, McKechnie, &
Puffer, 1983; Sutton & Callahan, 1987; Wade, Porac, & Pollock, 1997). For instance, investors
tend to react more negatively when an organization accepts responsibility for an accident
because an admission of guilt opens up the possibility of a lawsuit. In contrast, when an event is
a scandal, investors will react more negatively if they deny their involvement (Marcus &
Goodman, 1991). This simple decision by organizational actors to accept or deny responsibility
has continued to sustain this line of research until the present day.
Scholars drawing on rhetorical theory step back from this decision to accept or deny
responsibility and instead focus on the generalized forms of communication that are maximally
persuasive when influencing others. While the impression management perspective is older than
the rhetorical perspective in organizational analysis, rhetorical theory pre-dates the former by
several thousand years. In particular, scholars adopting a rhetorical understanding of strategic
communication draw primarily upon Aristotle (1991) and his argument that there exists three
fundamental components of the human condition—emotion (i.e., pathos), logic (i.e., logos), and
5

character (i.e., ethos)—that speakers can appeal to in order to persuade audiences. Organization
theorists since have examined how appealing these three components can help actors
institutionalize practices (Green, 2004) or legitimate actions (Erkama & Vaara, 2010), define
new organizational forms (Suddaby & Greenwood, 2005), and make sense of catastrophes
(Cornelissen, Mantere, & Vaara, 2014). Scholars therefore draw upon rhetorical theory to be able
to identify the universal rules for how communication resonates with an audience.
Researchers adopting a framing perspective have a similar goal as rhetoricians in that
they seek to explore how communication resonates with an audience. However, instead of
assuming that there are universal rules of persuasion that enabled resonance across all audiences
similarly, framing theorists focus on how actors might frame a particular situation, event, or
action in specific ways so as to resonate with a target audience. Drawing on the work of Goffman
(1974) as well as Snow and Benford (2000), organizational theorists have explored how actors
can frame important strategic decisions (Fiss & Zajac, 2006), everyday behavior (Gorgi &
Weber, 2015), and potentially controversial actions in a way that resonates with specified
audiences (Rhee & Fiss, 2014). By recognizing the fact that actors can use language to highlight
some interpretations while hiding others, framing scholars have pointed to a more contingent
understanding of how communication resonates with and influences others.
Finally, a smaller group of these scholars who explore how actors’ communication
resonates with target audiences have focused in on storytelling as one particularly influential
tactic. In this way, storytelling might be thought of as a subset of framing theory in that stories
are simply one of the many ways to highlight a preferred interpretation and hide a less appealing
one. Even still, storytelling has a long history of its own apart from the framing literature that
emphasizes how managers and entrepreneurs often need to be great storytellers to connect with
6

their desired audience (Bartel & Garud, 2009; Garud, Schildt, & Lant, 2014; Lounsbury &
Glynn, 2001; O’Connor, 2004). Researchers within this perspective have shown that stories or
narratives can connect to unique contextual features of the prevailing context, draw forth an
emotional resonance, and highlight future opportunities for audience members (Martens,
Jennings, & Jennings, 2007).
Empirical Methodology
Despite the differences between these four theoretical perspectives, all of them
empirically approach the study of strategic communication in a similar fashion. In particular, the
standard empirical approach is to collect public messages (e.g., press releases, IPO prospectuses,
quarterly earnings transcripts, etc.) communicated by organizational actors and extract specific
words or phrases embedded within the broader text. Extracting these words or phrases is an
attempt to capture the theoretical elements of interest within the message. For example, if the
organization apologized (e.g., Lamin & Zaheer, 2012), did they use the words “I’m sorry”
anywhere in the text? If the organization framed their adoption of a practice in a particular way
(Rhee & Fiss, 2014), which words did they include and exclude when they mentioned that
practice? As a result, the existing empirical approach searches for and codes just a few words or
phrases and ignores the rest of the broader text within which these words or phrases are
embedded.
Limitations to Existing Approach
This existing approach to the study of strategic communication however has several
important limitations. The primary and most substantial limitation is that this existing approach
focuses almost entirely on how strategic communication operates within a given system of
meaning. In particular, research across all four theoretical perspectives explores how strategic
7

communication strategies—such as apologies or denials (Lamin & Zaheer, 2012), justifications
(Wade et al., 1997), rhetorical appeals (Green, 2004), framing tactics (Rhee & Fiss, 2014), or
stories (Martens et al., 2007)—fit within an audience’s prevailing institutionalized assumptions.
Fiss and Zajac (2006: 1179) summarize the prevailing finding across this body of work by
stating that audiences “will respond more positively if a firm’s [communication] is in line with
the institutional context.”
However, this current focus overlooks the possibility that organizational actors not only
talk within the rules of a given meaning system but also about those very rules or assumptions
(Harmon et al., 2015; Toulmin, 1958). For instance, CEOs regularly talk about fundamental
strategic assumptions in order to clarify a business position (Cook, 2016; Drucker, 1994) or take
employees in a radically new direction (Furr & Dyer, 2014), leaders commonly discuss the
assumptions that ground their professions (Suddaby & Greenwood, 2005), and politicians often
debate ideological assumptions instead of how they might actually execute their policies
(Simons, 1994). Despite the fact that actors often communicate directly about these assumptions,
existing approaches to strategic communication contain no theoretical understanding of how
exposing these assumptions influences others and the broader institution within which these
actors reside. This omission is particularly troubling because the decision to either expose or hide
our assumptions when communicating can impact the persuasiveness of certain strategies over
others (Bitzer, 1959; Jackson & Jacobs, 1980), signal the depth of thinking engaged in by the
speaker (Werder, 1999), and even impact the stability of our existing institutions (Bitektine &
Haack, 2015; Harmon et al., 2015).
The second limitation concerns the way in which this existing approach conceptualizes
what strategic communication actually looks like in public messages. In particular, this approach
8

both conceptually and empirically presumes that certain words or phrases (e.g., apologies,
justifications, framing, stories, etc.) are the primary linguistic components that persuade
audiences. However, by isolating and examining the impact of only these small portions of an
overall text, scholars are overlooking the possibility that the persuasive force of communication
comes not from these several words or phrases but from the holistic impression formed from the
entire message that happens to also contain these words or phrases. This omission seems even
more problematic in light of the fact that linguists (Rorty, 1981; Wittgenstein, 1953) and
communication scholars (Burke, 1969; Toulmin, 1958) have long argued that the persuasive
force of words and phrases cannot be understood by audiences in isolation from the broader
context within which they occur.
INTRODUCING THE STRUCTURE OF STRATEGIC COMMUNICATION
This dissertation seeks to address these limitations by advancing a novel approach that I
refer to as the structure of strategic communication. Structure of course can refer to a variety of
things in communication. For example, linguists often talk about structure as grammar
(Wittgenstein, 1953), while others refer to structure as a conversational style (Grice, 1975;
Schwarz, 1996). In this dissertation, when I discuss structure I am referring specifically to
argument structure and, more precisely, to the fact that strategic communication used by actors
can occur at two structurally distinct levels of argumentation.
The Structure of Strategic Communication
The approach I put forth here leverages work by Toulmin (1958) and Harmon, Green,
and Goodnight (2015) to conceptualize a novel way to study strategic communication. This
approach ignores the specific communication strategies that actors might use (e.g., apologies,
denials, justifications, framing, etc.) and instead focuses on the degree to which these actors
9

make explicit their institutionalized assumptions when communicating. To achieve this, I use
aforementioned conceptual distinction in argument structure between arguing within the rules of
the game and arguing about the rules of the game. I propose that arguing more within the rules,
which leaves implicit these assumptions, tends to reinforce and reproduce the legitimacy of those
very assumptions. In contrast, I propose that arguing more about the rules of the game exposes
the contingencies of the institution and places the legitimacy of its taken-for-granted assumptions
at risk. This dissertation’s primary claim therefore is that talking more about the rules of the
game in a given public message may profoundly impact audiences and the overall stability of our
social institutions.
This new approach to the study of strategic communication also provides one way to
address the two limitations noted earlier. In particular, this approach not only acknowledges the
possibility that organizational actors can talk about the rules or assumptions that underlie our
institutions, but it also conceptualizes a way to capture the degree to which actors do so. This
provides a nice opportunity to develop new theory as well as a new empirical approach for
measuring how strategic communication influences others. Moreover, this approach also
conceptualizes communication more holistically by acknowledging that every aspect of a given
message would be categorized into one of two structural levels of talk (i.e., arguments within or
about the rules of the game). This provides the opportunity to conceptualize and examine the
impact of an entire message on others rather than the influence of just isolated words or phrases.
Taken together, this dissertation aims to build upon this basic conceptual distinction in argument
structure to reformulate how scholars might study strategic communication.

10

Dissertation Outline
I advance this approach in two parts. First, in Chapter 2 I develop a novel construct called
the argument structure ratio (ASR) that conceptually and empirically captures how explicit a
speaker makes their assumptions in communication. Second, in Chapter 3 I put to use the ASR
measure by theorizing and empirically demonstrating the profound implications of exposing
these assumptions to an audience. Using all public speeches made by the Chairperson of the
United States Federal Reserve from 1998 to 2014, I show that the more they expose the
assumptions underlying the Federal Reserve System, the more their speeches produce market
uncertainty. Chapter 4 steps back to reflect upon the theory, measurement, and effects of the
ASR and draws together several broader conclusions regarding the future study of the structure
of strategic communication.

11

CHAPTER 2 – ARGUMENT STRUCTURE RATIO: MEASURING THE
EXPLICITNESS OF OUR ASSUMPTIONS IN COMMUNCIATION

CHAPTER 2

ARGUMENT STRUCTURE RATIO:
MEASURING THE EXPLICITNESS OF OUR ASSUMPTIONS IN COMMUNCIATION

ABSTRACT

Existing approaches to conceptualizing and measuring the communication strategies actors use
to influence others has overlooked the fact that the actors involved maintain certain assumptions
that ground what it is they are doing. This work presumes that organizational actors take these
assumptions for granted and therefore leave them implicit in persuasive communication. This
however is not the case. Actors regularly make explicit these assumptions for a variety of
reasons. This paper develops a new construct called the argument structure ratio (ASR) that
conceptually and empirically captures how explicit a speaker makes these assumptions in
communication. I propose a three-step methodology for measuring the ASR of communication
that can be used on any collection of written texts obtained in archival or experimental settings. I
then illustrate the validity of this measure by examining the ASR of all public speeches made by
the Chairperson of the United States Federal Reserve from 1998 to 2014. Future directions for
using the ASR construct in a variety of organizational contexts and across multiple levels of
analysis are discussed.

Words: 173

Keywords: argument structure, strategic communication, persuasion, Federal Reserve

12

INTRODUCTION
Communication is a critical tool that actors regularly use to persuade others. Managers
for example use communication strategically to justify performance outcomes (Staw et al.,
1983), narrate their approach to business (Martens et al., 2007), frame controversial practices
(Rhee & Fiss, 2014), and explain strategic decisions to their employees (Rousseau & Tijoriwala,
1999) and investors (Fiss & Zajac, 2006). Top governmental officials similarly engage in
rhetorical strategies to shape how voters and market participants interpret their policies and
decisions (Abolafia, 2004; Emrich, Brower, Feldman, & Garland, 2001), and social activists
employ a variety of framing tactics to garner support from different constituencies (Benford &
Snow, 2000; King & Soule, 2007). This still growing body of work that stretches across a
number of literatures has thus sought to identify and empirically isolate a wide variety of
communication strategies that actors use to inform and influence others.
This existing approach of conceptualizing and measuring communication strategies
however has been limited. Specifically, this research has overlooked the fact that when trying to
persuade others, the actors involved maintain certain assumptions that ground what it is they are
doing (Harmon et al., 2015). For instance, when a CEO justifies the acquisition of another
company to investors, the assumption underlying this communication is that “profitability is
what matters here.” Contrast this with when an environmental social activist publicly decries this
same acquisition. The assumption underlying this communication is quite different, and likely
resembles the idea that “social well-being is what matters here.” These assumptions thus reflect
the “rules of the game” that actors appear to be collectively playing by in a given situation
(Berger & Luckmann, 1966; Friedland & Alford, 1991; Thornton, Ocasio, & Lounsbury, 2012).
Since these rules at times are left unstated and taken as givens when communicating with others
13

(Zucker, 1977), it is perhaps unsurprising that existing research has overlooked them and failed
to examine them directly.
The problem arises however when we recognize that these assumptions are often exposed
and stated explicitly in our communications. CEOs regularly discuss their assumptions in order
to clarify a strategic decision (Cook, 2016; Drucker, 1994) or explain a radically new business
direction (Furr & Dyer, 2014), and political party leaders seem to debate their ideological
assumptions sometimes more than actually discussing how they would execute their policies
(Simons, 1994). And while some case study research has recently highlighted the possibility that
actors can engage in discourse about these typically taken-for-granted assumptions (Green,
Babb, & Alpaslan, 2008; Suddaby & Greenwood, 2005), there is no existing work that addresses
how we might empirically measure the degree to which we make explicit our assumptions in
communication. This omission is particularly troubling because the decision to either expose or
hide our assumptions when communicating can impact the persuasiveness of certain strategies
over others (Bitzer, 1959; Jackson & Jacobs, 1980), signal the depth of thinking engaged in by
the speaker (Werder, 1999), and even provide an indication about the stability of our social
institutions (Bitektine & Haack, 2015; Harmon et al., 2015).
This paper addresses this by developing a new construct called the argument structure
ratio (ASR) that conceptually and empirically captures how explicit a speaker makes these
assumptions in communication. I begin by briefly explaining that Toulmin’s (1958) model of
argument structure provides the ideal theoretical scaffolding upon which to build this measure. I
then propose a three-step methodology for measuring the ASR of communication that can be
used on any collection of written texts obtained in archival or experimental settings. To illustrate
the validity of the ASR construct, I examine all public speeches made by the Chairperson of the
14

United States Federal Reserve from 1998 to 2014, and compare the ASR construct to other
potentially related measures of communication. Finally, future directions for using the ASR in
other organizational contexts and across different levels of analysis are discussed.
ARGUMENT STRUCTURE
The Toulmin Model
Argument structure concerns the way we naturally organize our reasoning in order to
make our communication with others legitimate and persuasive. The Toulmin Model is widely
regarded as one of the simplest and most powerful ways to conceptualize the structure
underlying everyday arguments (see Figure 1). Toulmin (1958) contends that all arguments start
with three basic structural components: an argument moves from data (i.e., the evidence actors
draw upon) to claim (i.e., the conclusion actors seek to establish) by virtue of a warrant (i.e., the
reason explains why the data support this claim). Importantly, attached to every data—warrant—
claim combination is a fourth component Toulmin called backing, which form the assumptions
that ground the overall reasonability of the argument. According to the Toulmin Model, the
backing thus reflects the underlying “rules of the game” so to speak. When the backing is
defined or collectively assumed in a given context, this encourages actors’ to play by these very
rules when choosing their data and warrants to support their claims.

FIGURE 1 – The Toulmin Model

15

For example, consider a situation where analysts are asking the CEO of a company
questions about whether their merger with another company remains a sound business decision.
The CEO might try to convince these analysts that the merger is still a sound decision (claim),
justifying this assertion with financial projections that look quite promising over the next two
years (data). Now the reason why this data support this claim is because mergers that continue to
make money are typically sound business decisions (warrant). While the data and warrant in this
case indeed work to support the claim made by the CEO, the reason why these components
function properly is because the parties involved in this context all generally hold an assumption
that “profitability” is what we are doing here (backing). That is, the backing provides the
presumptive basis for believing that using data related to financial performance would be a valid
and legitimate way to justify the claim that the merger is remains a sound business decision.
Toulmin contends that while every argument contains all four structural components, not
all components need to be made explicit in the argument itself. Indeed, the example above shows
how the backing in some arguments can be implicit and taken for granted by the actors involved.
This situation in fact resembles most existing research on strategic communication, which takes
the backing as a given and examines how actors use data or warrants to justify and persuade
others of their desired claim (Elsbach, 1994; Fiss & Zajac, 2006; Lamin & Zaheer, 2012;
Martens et al., 2007; Rhee & Fiss, 2014; Staw et al., 1983). While actors of course leave the
backing implicit in many everyday arguments, they can also make the backing explicit and
discuss those very assumptions directly (Green et al., 2008; Suddaby & Greenwood, 2005).
Two Structural Levels of Communication
Harmon, Green, and Goodnight (2015) were one of the first to recognize this and
proposed that actors can thus communicate at two structurally distinct levels. At one level, actors
16

can communicate “within the rules of the game.” This is where actors use data or warrants to
argue for or against a particular claim, while leaving implicit the backing. At another level,
actors can communicate “about the rules of the game.” This is where actors make explicit the
backing and discuss its nature and appropriateness in grounding the given context.
Consider the game of baseball as a simple example. Baseball has many rules, such as the
number of outs in an inning, when the substitution of players can occur, how and when to use
instant replay, the size of the strike zone, and even the etiquette its players should observe. For
much of the time, the validity of these rules is taken as a given by players, coaches, fans, and
sportscasters. However, when listening to post-game interviews after a controversial call
occurred during the course of the game, you often hear an impressive variation in the structural
level of talk. For instance, some interviewees continue to take as a given the validity of the rule
undergirding the controversial call and instead discuss how they could have played differently
within the prevailing rules to perhaps achieve a different outcome (i.e., arguing within the rules
of the game). Other interviewees however may take issue with the legitimacy of the rule
affecting the controversial call, talking about the validity of the rule itself (i.e., arguing about the
rules of the game).
Importantly, within in the same interview an actor can seamlessly shift between these two
structurally distinct levels. For instance, the interviewee might start by leaving the backing
implicit, then transition to talking explicitly about the backing, and finally shift back to again
leaving the backing implicit. The possibility of variation across these two structural levels in a
single interview or speech act provides the basis for developing a new measure based on the
degree to which actors make explicit their assumptions in any public message.

17

ARGUMENT STRUCTURE RATIO
The argument structure ratio (ASR) conceptually captures the variation between
communication that occurs within the rules of the game (i.e., engages the structural components
of data, warrant, and claim) and about the rules of the game (i.e., engages the structural
components of backing). The ASR therefore measures how explicit a speaker makes these
backing-related assumptions in communication. Any communication that seeks to influence
others, whether obtained in archival or experimental settings, has an ASR associated with it. The
ASR is empirically calculated in the following manner:

ASR = (number of arguments that expose the backing / total number of arguments)

ASR scores therefore range from 0 to 1. Communication with higher ASRs contains more
backing-related talk and, thus, reflects the fact that these messages are increasingly making
explicit the assumptions underlying the given context. Two considerations are worth mentioning.
First, consistent with prior work that suggests that people regularly cluster their arguments into
paragraph form (Green et al., 2009), each paragraph of a given message should be coded as one
argument. As a result, every paragraph is coded as either exposing the backing or not exposing
the backing, and the ratio above is calculated.
Second, a ratio is used instead of the raw number of arguments that expose the backing.
Keep in mind that the ASR is a measure of the overall impression of a speech act, specifically in
regard to the degree to which these assumptions are made explicit. As a result, a ratio captures
the relative weight that a speech act exposes the backing, whereas a raw count measure does not.
For instance, compare a message that has five arguments that expose the backing and 10
18

arguments in total, with a message that also has five arguments that expose the backing but 50
arguments in total. Counting the raw number of arguments that expose the backing treats these
two messages identically, whereas a ratio recognizes that these same five backing-related
paragraphs is likely going to be interpreted differently given the substantial difference in the
overall message length.
In the remainder of this section, I propose a three-step methodology for measuring the
ASR of communication. Following these three steps will enable a researcher to code the ASR of
any communication that seeks to influence others. An illustration executing these three steps is
provided later.
Step 1: Define the discursive space and corpus
A discursive space is a bounded site of persuasion in which actors use communication—
verbal speech or written text—to impose their meanings on particular ideas, activities, or
situations (Hardy & Maguire, 2010). The bounded aspect of this definition refers to a site where
identifiable actors are interacting or in dialogue with each other. This bounded relationship can
be bidirectional where multiple parties are communicating at conferences or events (Ansari,
Wijen, & Gray, 2013; Suddaby & Greenwood, 2005) or unidirectional where actors give
speeches or issue public messages while everyone else listens and evaluates (Martens et al.,
2007; Rhee & Fiss, 2014). Boundedness is an important criterion because it helps to concretely
identify the backing or assumptions underlying arguments in that particular context. By
persuasion I mean any situation wherein actors use communication to influence others’ beliefs or
interpretations. Most public communications meet this criterion, from company press releases
(Lamin & Zaheer, 2012) or annual reports (Staw et al., 1983) to employee memos (Rousseau &
Tijoriwala, 1999) to testimony (Suddaby & Greenwood, 2005). Finally, I refer to communication
19

as verbal speech or written text to highlight that the researcher, after defining the discursive
space, then needs to identify clearly a corpus, or collection of documents, that he or she wants to
code for ASR.
Step 2: Clarify the two structural levels of communication in your corpus
Once the discursive space is defined and the corpus one wants to examine empirically is
identified, the next step is to clarify what the two structurally distinct levels of communication
look like. It is important to note that this clarification is done at the corpus level, not at the level
of each individual speech act. There are two reasons for this. First, communication with others
does not occur in isolation but rather always operates within a broader context and metanarrative.
As such, a researcher needs to understand this broader context and the potential factors that
might influence the communication (Fiss & Zajac, 2006; Lamin & Zaheer, 2012). Second, one of
the primary aims of this measure is to be able to compare the ASR of multiple speech acts with
each other within a given corpus. Clarifying these two structural levels at the corpus level
enables the researcher to develop a standardized “ideal type” argument within that space,
allowing one to compare each individual speech act against this predefined model.
To do this, the researcher should first read sample speech acts within the identified
corpus as well as additional literature (e.g., research articles, popular press books, newspaper
articles, etc.) so as to immerse oneself in and become familiar with the discursive space. Once
the researcher has reached a point of informational saturation, I recommend documenting the
different types of words and phrases that typically encapsulate standard claims, data, warrants,
and backings within that corpus. Identifying these “content flags” for each of the argument
structure components enables the researcher to understand clearly and in a reproducible manner
the coding of paragraphs as either containing the backing or not. While the presence of backing-
20

related content flags in particular do not of course automatically result in coding a paragraph as
exposing the backing, they do help to signal an increased likelihood that the argument should be
coded as such.
At this point, researchers should keep in mind two additional considerations. First, the
ASR may not have much variation in situations where the communication within that discursive
space is highly institutionalized (Zucker, 1977). In these situations, the backing may be so
entirely taken-for-granted that actors would not even think to talk about them (Bitektine &
Haack, 2015). Even still, small changes in these types of situations can trigger mental alarms
(Tost, 2011) and prompt actors to start talking about their assumptions (Zilber, 2002).
Leveraging these sorts of changes in one’s corpus time period can be useful research design tool.
Second, the backing underlying a given corpus can sometimes change or evolve over extended
periods of time. As such, researchers should use their judgment about the likelihood that the
backing within their corpus has shifted during the examined time period. If it has, then
identifying different content flags for different theoretically-derived time periods may be useful.
Step 3: Code your corpus based on these two structural levels
Based on the clarification between these two structural levels in the corpus, one is ready
to code the ASR of each speech act. Remember that the speech act itself is the unit of analysis,
while the paragraph is the unit of data collection. The researcher or trained assistants should
systematically read each speech act, coding each paragraph as either exposing the backing or not
exposing the backing. Upon completion of coding each paragraph, one calculates the ASR by
dividing the number of paragraphs that expose the backing by the total number of paragraphs in
the overall speech act.
21

There are several considerations to keep in mind during this coding process. First, while
many paragraphs will contain only data, warrants, and claims or only backing, some paragraphs
may have a combination of these two structural levels. As a general rule, whenever the backing
is exposed at all in the paragraph, that paragraph should be coded as exposing the backing.
Second, some paragraphs in a speech act may not appear to fit into one of the four structural
components of the Toulmin Model. For example, sometimes speeches begin with formalities like
thanking the audience for attending and press releases often have standard legal information at
the end. The researcher should use his or her judgment to determine whether to exclude these
paragraphs or always code them as arguments that do not expose the backing. The latter option
of course reduces the ASR of a given speech act. However, this does not pose an issue if most
speech acts in the corpus have similar types of paragraphs since then this reduction of the ASR
will be systematically applied throughout the sample.
Finally, researchers should also follow steps recommended by experts in content analysis
(Krippendorff, 2003; Neuendorf, 2001) to ensure the ASR coding is done in a reliable manner.
First, ensure that you have at least one additional coder and that this coder becomes familiar with
the discursive space using the same process as noted above. Second, engage in pilot coding,
which entails everyone coding a handful of the same speech acts and discussing their coding
decisions until a consensus is reached on each decision. Third, engage in reliability coding,
which entails everyone coding a predefined number of the same speech acts without discussing
their decisions (Krippendorff, 2003: 240) and checking the interrater reliability (Hayes &
Krippendorff, 2007). Fourth, engage in independent coding, which entails everyone
independently coding different speech acts. Fifth, ensure that interrater reliability remained
22

robust throughout your independent coding by having everyone code a handful of the same
speech acts at the end of the sample and again calculate interrater reliability.
ILLUSTRATION AND VALIDATION: FEDERAL RESERVE SPEECHES
Research Context
To illustrate and validate the development of the ASR measure, I apply this three-step
methodology to public speeches given by the Chairperson of the United States Federal Reserve
(Fed). The Fed is the central banking system of the United States and was established in 1913 as
a way to protect investors during financial panics by guaranteeing liquidity and acting as the
lender of last resort. With its central base of operations in Washington, D.C., the presidentially
appointed seven-member Board of Governors (with one member appointed as the Chairperson)
oversees the twelve regional Federal Reserve Banks and the broader Federal Reserve System.
The Fed’s primary objective is to maintain confidence and market stability (Bernanke, 2015),
making them one of the most important and powerful institutions in the world (Abolafia, 2004;
Cruikshank & Sicilia, 1999; Holmes, 2013). This context is useful to illustrate the ASR because
one of the primary ways the Fed tries to achieve this objective is through communicating with
the public (Bernanke, 2013; Yellen, 2013).
Step 1: Define the discursive space and corpus. The discursive space in this context is a
bounded set of discussions about United States monetary policymaking. This discussion revolves
primarily around the Fed and occurs in both bidirectional interactions (e.g., at academic
conferences, in academic institutions, popular culture books, etc.) and unidirectional speech acts
(e.g., Chairperson speeches given to the public, press releases issued in the market, testimony to
Congress, etc.). These unidirectional speech acts by the Fed Chairperson in particular function as
acts of persuasion directed towards the United States economy and financial markets (Bernanke,
23

2015; Cruikshank & Sicilia, 1999; Holmes, 2013; Yellen, 2013). For this illustration, I define the
corpus of interest as all public speeches given by the Fed Chairperson from 1998 to 2014 (N =
339). Since the Fed Chairperson is perhaps the most prominent actor in the United States
economy (Holmes, 2013) and his or her speeches are uniquely a reflection of their own opinions
rather than the Fed as a whole (Bernanke, 2015), this is a reasonable choice.
Step 2: Clarify the two structural levels of communication in your corpus. I began this
step by reading a random sample of Chairperson speeches during my defined time period. I
supplemented this reading with academic articles (e.g., Abolafia, 2004, 2012; Chen & Clements,
2007; Fligstein, Brundage, & Schultz, 2014) and popular press books about the Fed and central
banking in the United States (e.g., Bernanke, 2015; Cruikshank & Sicilia, 1999; Holmes, 2013;
Paul, 2009; Shiller, 2000; Steil, 2013). Based on this background knowledge, I documented the
different types of words and phrases that typically encapsulate standard Fed Chairperson claims,
data, warrants, and backings within that corpus (see Table 1). Most important in this coding
scheme is the backing. In this context, the backing reflects a stable set of assumptions underlying
the objectives, nature, and boundaries of United States monetary policymaking. The backing is
therefore defined clearly by the Fed’s dual Congressional mandate (i.e., to maximize
employment and maintain price stability) and the set of conventional tools (e.g., open market
operations, discount rate, reserve requirements) used to conduct monetary policy. These
objectives and tools thus form the “rules of the game” that are regularly taken as givens by most
market participants.

24

TABLE 1 – Two Structural Levels of Fed Communication

Step 3: Code your corpus based on these two structural levels. Based on these two
predefined structural levels of communication within the Fed speech corpus, three research
assistants and I coded every paragraph within all 339 Fed Chairperson speeches as either
exposing the backing or not. Paragraphs that do not do expose the backing use data and
sometimes warrants to draw conclusions about a claim, although they need not engage in all
three components. For instance, in a speech on September 26, 2005 to the American Bankers
Association, Chairperson Greenspan makes an initial claim, supports this claim with data, and
then reasserts the claim.

This enormous increase in housing values and mortgage debt has been spurred by the decline in mortgage
interest rates, which remain historically low (claim). Indeed, the thirty-year fixed-rate mortgage, currently
around 5 3/4 percent, is about 1/2 percentage point below its level of late spring 2004, just before the
Federal Open Market Committee (FOMC) embarked on the current cycle of policy tightening (data). This
decline in mortgage rates and other long-term interest rates in the context of a concurrent rise in the federal
funds rate is without precedent in recent U.S. experience (claim) (Greenspan, 2005).

Similarly, in a speech on January 3, 2014 to the American Economic Association, Chairperson
Bernanke makes an initial claim and then supports this claim with a variety of data.

Structural Level Component Basic Argument and Related Content Flags
Claim Conclusions about the state of the US economy .
"the economy is…[stable/recovering/slowing…]"
Data Economic-based evidence .
"inflation rate" "unemployment rate" "GDP" "production investment" "asset
values" "economic indicators"
Warrant Explaining why evidence supports these conclusion .
"inflation rate is slowing economic recovery" "economic indicators show
evidence for a slowing economy"
Arguing "about the
rules of the game"
Backing Objectives, nature, or boundaries of US monetary policymaking .
"central banking" "framework" "maximum employment" "price stability"
"conventional tools" "macroprudential supervision"
Arguing "within the
rules of the game"
25

The economy has made considerable progress since the recovery officially began some four and a half
years ago (claim). Payroll employment has risen by 7-1/2 million jobs from its trough (data). Real GDP has
grown in 16 of 17 quarters (data), and the level of real GDP in the third quarter of 2013 was 5-1/2 percent
above its pre-recession peak (data). The unemployment rate has fallen from 10 percent in the fall of 2009
to 7 percent recently (data). Industrial production and equipment investment have matched or exceeded
pre-recession peaks (data) (Bernanke, 2014).

In contrast, when the Chairperson exposes the backing, he or she begins discussing the
nature and boundaries of monetary policy itself. For example, in a speech in given at the Federal
Reserve Bank of St. Louis on October 11, 2001, Chairperson Greenspan lays bare the very
assumptions underlying the Federal Reserve System.

We at the Federal Reserve are given two mandates that are not often spelled out explicitly. First, to
implement an effective monetary policy to meet our legislated objectives (backing). Second, to do so in a
most open and transparent manner in recognition that we, as unelected officials, are accountable both to the
Congress from which we derive our monetary policy mission and to the American people (backing)
(Greenspan, 2001).

Similarly, in a speech on October 18, 2011 at the Federal Reserve Bank of Boston’s 56th
Economic Conference, Chairperson Bernanke reaffirms the Fed’s dual Congressional mandate,
followed by a discussion about how inflation targeting fits into their monetary policy framework.

The Federal Reserve is accountable to the Congress for two objectives —maximum employment and price
stability, on an equal footing—and it does not have a formal, numerical inflation target. But, as a practical
matter, the Federal Reserve's policy framework has many of the elements of flexible inflation targeting. In
particular, like flexible inflation targeters, the Federal Open Market Committee (FOMC) is committed to
stabilizing inflation over the medium run while retaining the flexibility to help offset cyclical fluctuations
in economic activity and employment (backing) (Bernanke, 2011).

For each speech, I then calculated the ASR by taking the ratio of the paragraphs that
exposed the backing divided by the total paragraphs in the speech. I ensured reliability of this
measure by engaging in reliability coding for the first 60 speeches (Krippendorff’s alpha = .88)
and validated the robustness of this measure by again conducting reliability coding for the last 10
speeches (Krippendorff’s alpha = .84).
26

Face Validity
I conducted two tests to assess the face validity of the ASR measure. First, if ASR is
indeed capturing the overall impression of the speech act that relates specifically to the degree to
which the backing is made explicit, then one might expect the titles of these Fed speeches to
somehow also reflect that overall impression. That is, since titles typically reflect the overarching
emphasis of the speech, we should expect low ASR speeches to have titles that signal a
discussion about the stability of the economy. In contrast, we should expect high ASR speeches
to have titles that ignore the economy but instead signal a discussion about the very nature of
United States monetary policymaking itself. A random selection of four speeches in the bottom
and top 10 percent of ASRs in the overall sample confirms this expectation (see Table 2),
providing some initial face validity to this construct.

TABLE 2 – Sample Fed Speech Titles and Associated ASR

Second, we also should be able to theorize the conditions under which one would expect
ASR to be higher and validate this expectation empirically. Identifying such conditions is not as
simple as it may appear, since the use of argument structure is likely to be a strategic decision on
the part of the Fed to some degree. However, there is one major environmental jolt that occurred
during the period from 1998 to 2014 that should theoretically have forced the Fed Chairperson to
Date Speaker Speech Title ASR
11/2/1999 Greenspan Mortgage Markets and Economic Activity 0.05
6/28/2001 Greenspan Impact of Energy on the Economy 0.03
6/9/2010 Bernanke Fostering Workforce Development 0.00
6/7/2011 Bernanke The U.S. Economic Outlook 0.07
10/19/2000 Greenspan Challenges for Monetary Policymakers 0.29
8/26/2005 Greenspan Reflections on Central Banking 0.34
1/5/2007 Bernanke Central Banking and Bank Supervision in the United States 0.64
10/19/2007 Bernanke Monetary Policy under Uncertainty 0.63
27

expose more of its backing. Specifically, the financial crisis starting in late 2007 shook the very
foundations of the United States economy (Lounsbury & Hirsch, 2010) and central banking
(Abolafia, 2012). Indeed, by late 2008 the federal funds rate—the Fed’s primary monetary policy
tool—was essentially zero, forcing the Chairperson and the broader central banking community
to reexamine their fundamental assumptions and monetary policy framework in order to figure
out what to do. This rate remained zero through 2014, leading central bankers to continue
discussing the very nature of monetary policymaking in the United States. Using the financial
crisis as a natural experiment, we would expect the ASR to steadily increase during this
timeframe and remain high throughout the latter years in my sample. Figure 2 graphs the average
ASR of Fed Chairperson speeches during this time period and strongly confirms this expectation.

FIGURE 2 – Average ASR of Fed Chairperson Speeches from 1998 to 2014

Comparison with Other Measures
To the best of my knowledge, there is no comparable measure to the ASR. Despite the
fact that no other measures claim to directly capture the explicitness of assumptions in
28

communication, there are at least three constructs that may indirectly tap into this fundamental
conceptual distinction. First, words differ in their level of abstraction. Some words are more
concrete in that they refer to tangible objects, persons, or actions that one can often point to and
observe in reality. Other words refer to more abstract concepts, hypothetical constructs, or
generalized ideas that are more easily objectified and harder to point to and observe in reality.
One might hypothesize that talk about the backing is thus inherently more abstract than talk
about data, warrants, and claims, since the backing reflects the generalized, objectified
assumptions underlying what it is we are doing here. To examine this possibility, I calculated the
level of abstractness or concreteness in each speech using three distinct measures of this
concept: Mergenthaler’s (1996) word dictionary of abstract nouns (e.g., nouns often ending in –
ism, -age, -ness, etc.), Friendly, Franklin, Hoffman, and Rubin’s (1982) measure of concreteness
imagery that is also known as the Toronto Word Pool, and Brysbaert, Warriner, and Kuperman
(2014) measure of concreteness words and two-word expressions.
Second, based on this same line of thinking, one might also propose that speeches with
higher exposed backing should also exhibit more vagueness because talking about objectified,
abstract assumptions will likely lack verbal clarity. I examine this possibility by using Hiller,
Fisher, and Kaess’s (1969) word dictionary (e.g., maybe, various, perhaps, probably, etc.) that
assesses the level of vagueness in each speech. Third, a final potentially related measure to the
ASR is the level speech act complexity. Similar to the argument for vagueness, one might
propose that talking more about the backing is simply going to be a more complex endeavor than
discussing everyday normal data and claims. To explore this, I use the Flesch-Kincaid reading
grade level (Kincaid, Fishburne Jr, Rogers, & Chissom, 1975), which is used extensively in the
field of education as well as in writing insurance and legal documents (McClure, 1987). The
29

score is based on a formula that considers the total number of words, sentences, and syllables in
a text, and produces the U.S. grade level required to understand that text.
Table 3 reports the correlations between the ASR and these other measures. The only
measure to correlate significantly with the ASR is speech complexity. This is perhaps
unsurprising when considering that fact that it is simply harder to articulate and communicate
about one’s fundamental assumptions. As such, it makes perfect sense that when trying to do so,
the Fed Chairperson uses more words and sentences to get his or her argument across.
Nevertheless, the correlation is 0.42, suggesting a meaningful relationship with the ASR but not
a substitute by any means. It may seem puzzling at first that the ASR is not correlated with
abstractness, concreteness, and vagueness. However, this finding makes sense when considering
the fact that these measures attempt to capture the abstractness, concreteness, and vagueness of
the object of the sentence, while failing to differentiate between whether that object relates to a
claim (i.e., the economy) or the backing (i.e., the framework for monetary policy). In contrast,
the ASR ignores how abstract, concrete, or vague one describes an object, and instead focuses on
capturing the variation in the argument’s object. Given these considerations, the ASR appears to
uniquely capture a novel conceptual distinction in our communication.

TABLE 3 – Comparing ASR with Other Measures

1 2 3 4 5
1. ASR
2. Abstractness (Mergenthaler 1996) 0.03
3. Concreteness (Friendly et al. 1982) -0.06 -0.06
4. Concreteness (Brysbaert et al. 2014) -0.01 -0.25** 0.49**
5. Vagueness (Hiller et al. 1969) -0.09 0.08 -0.04 -0.04
6. Complexity (Kincaid et al. 1975) 0.42** -0.06 0.04 0.05 -0.05
30

CONCLUSION AND FUTURE DIRECTIONS
This paper developed a new measure of how actors communicate called the argument
structure ratio (ASR), which conceptually and empirically captures how explicit a speaker makes
their assumptions in communication. Existing research until now has generally ignored or
overlooked these assumptions underlying how we communicate, presuming that they are taken-
for-granted in a particular context (Elsbach, 1994; Fiss & Zajac, 2006; Lamin & Zaheer, 2012;
Rhee & Fiss, 2014). By identifying these assumptions and when they are made explicit in
communication, the ASR measure adds an altogether new dimension of communication to this
still growing body of research. As a result, the ASR opens up a number of avenues for future
research in different contexts and across multiple levels of analysis.
At the individual and group level, managerial or leader communication could be
examined to explore the role of ASR in consensus-building. For instance, communicating major
changes within an organization or team can be challenging (Rousseau & Tijoriwala, 1999), and
making explicit one’s assumptions could potentially build consensus in some situations but also
alienate employees in others. The ASR might also be useful in examining conflict management
situations (Dewulf et al., 2009; Putnam & Holmer, 1992), where one group of people may be
exploiting these assumptions to force change while another group is strategically avoiding
discussion of these assumptions because they want to avoid risking changes to their existing
framework. Researchers studying employee voice (Burris, 2011; Detert, Burris, Harrison, &
Martin, 2013) may find the ASR construct useful as well when examining the broader
communicative conditions under which individual employees are likely to speak up and question
the assumptions underlying their organization.
31

We might also examine how the ASR of CEO or top management communication at the
organizational level influences different stakeholders. For example, organizational founders are
required to communicate with prospective investors with their prospectuses when going public
(Martens et al., 2007). Scholars could explore how the ASR of these communications impact the
firm’s their ability to raise capital. Similarly, given the importance of quarterly earnings calls
with analysts (Lee, 2015) one might explore how the ASR of this communication influences
analysts’ ratings after the meeting. More generally, organizations communicate through press
releases and other public communications all the time, defending their actions (Lamin & Zaheer,
2012) and framing their decisions (Rhee & Fiss, 2014). Researchers could examine how the ASR
of these communications impact how the stock market and media react.
Researchers might also find examining the ASR at the institutional and societal level
fruitful. For example, since different institutional systems and cultures maintain very different
assumptions (Griswold, 2012), it would be interesting to examine how revealing one’s
assumptions might produce different effects across nations. One might also consider examining
the ASR over time at the discursive space or societal level to gain insight about collective beliefs
of actors in that space. Consider again Figure 2 that graphs the average ASR from 1998 to 2014.
This graph suggests that the ASR, when tracked over time, might shed light on the underlying
collective distress emerging within a given system. With that in mind, it may be possible to
explore the ASR as either a leading or lagging indicator of market bubbles and busts.
The ASR thus provides scholars studying a variety of different communication-related
phenomena a novel and powerful tool. Given its distinctiveness from the way most research
across these different research areas conceptualize communication, I believe that the ASR will
enable researchers to gather interesting and altogether new insights in the future.
32

CHAPTER 3 – THE EFFECT OF ARGUMENT STRUCTURE ON MARKET
UNCERTAINTY: FEDERAL RESERVE CHAIRPERSON SPEECHES FROM 1998 –
2014

CHAPTER 3

THE EFFECT OF ARGUMENT STRUCTURE ON MARKET UNCERTAINTY:
FEDERAL RESERVE CHAIRPERSON SPEECHES FROM 1998 – 2014

ABSTRACT

This study investigates how communication that exposes an institution’s assumptions to direct
examination creates uncertainty and instability in the market. Leveraging theory on how we
naturally structure our arguments, I develop a novel measurement called the argument structure
ratio (ASR) that conceptually and empirically captures how explicit a speaker makes these
assumptions. Using all public speeches made by the Chairperson of the United States Federal
Reserve from 1998 to 2014, I demonstrate that the more they expose the assumptions underlying
the Federal Reserve System, the more their speeches produce market uncertainty, as measured by
market volatility (i.e., the VIX Index). This however creates a practical tension for the Federal
Reserve, since they are simultaneously responsible for being transparent about their monetary
policy operations but also for reducing market uncertainty. To resolve this tension, I build on
theory related to the role of emotion in markets and identify two conditions under which the
Chairperson can still discuss these assumptions without creating uncertainty. This study
contributes a new perspective on the role of strategic communication in market contexts, offering
new insights for institutional theory, the sociology of financial markets, and the study of
argument structure in other organizational contexts.

Words: 197

Keywords: strategic communication, argument structure, emotions, institutions, markets

33

INTRODUCTION
In recent years, an increasing number of people are paying attention to how prominent
actors’ communications influence markets. Popular business press for example often expresses
concern over how corporate leaders manage investors’ expectations (Hutton, 2001; Lev, 2011;
Schweitzer, Brooks, and Galinsky, 2015) and how just a few words from top governmental
officials can move entire markets (Leubsdorf, 2015; Hilsenrath, 2015). These concerns have
been reinforced by several decades of academic research that explores how prominent actors’
public messages influence the way market participants interpret their activities and future
behavior (Staw, McKechnie, and Puffer, 1983; Wade, Porac, and Pollock, 1997; Westphal and
Zajac, 1998; Emrich et al., 2001). This line of research has shown that carefully crafted messages
can help to manage important organizational issues, such as the justification of potentially
controversial practices (Elsbach, 1994; Graffin, Carpenter, and Boivie, 2011; Lamin and Zaheer,
2012; Zavyalova et al., 2012; Rhee and Fiss, 2014) or the decision to engage in broader strategic
changes (Elsbach, Sutton, and Principe, 1998; Arndt and Bigelow, 2000; Fiss and Zajac, 2006;
Kennedy and Fiss, 2009). The continued growth of this research highlights the importance of and
sustained interest in understanding how strategic communication affects the direction and
stability of markets.
In this study, I take a social constructionist view of strategic communication in markets.
This view sits in contrast to the financial economics understanding of communication as
information that reduces asymmetries (Akerlof, 1970; Spence, 1973), thereby reducing
uncertainty at both the individual and market levels (Van Buskirk, 2012). A social
constructionist view holds that communication is not just raw information but instead a mode of
meaning-making that can shape people’s beliefs and interpretations (Berger and Luckmann,
34

1966; Goffman, 1974). This view has led scholars to identify a number of different strategies that
actors might use, such as apologies and denials (Marcus and Goodman, 1991), justifications
(Wade, Porac, and Pollock, 1997), stories (Lounsbury and Glynn, 2001), framing tactics (Gorgi
and Weber, 2015), and narratives (Martens, Jennings, and Jennings, 2007). By focusing on how
communication operates within specific market contexts, this work has demonstrated that using
strategies that fit within those institutionalized rules and assumptions tend to produce more
favorable market reactions. For example, Lamin and Zaheer (2012) find that the effectiveness of
a firm’s strategies to defend their sweatshop labor practices differed on Main Street vs. Wall
Street because “these worlds operate by separate moralities.” Rhee and Fiss (2014) similarly
show that the market reacts more positively when a firm frames their reasons for adopting the
poison pill in a manner that aligns with the dominant institutional logic. As Fiss and Zajac (2006:
1179) note, “markets will respond more positively if a firm’s [communication] is in line with the
institutional context.”
However, this tendency to focus on how strategic communication operates within a given
system of meaning overlooks a substantial portion of communication regularly used by
prominent actors in market contexts. Anecdotal evidence and case studies suggest that prominent
actors not only talk within the rules of a given meaning system but also about those very rules.
Just as sportscasters can shift from talking about the implementation of a rule on the field to
talking about the validity of the rule itself, it is often the responsibility of prominent actors to
provide commentary about the very institution in which they are an authority. Indeed, CEOs
regularly talk about fundamental strategic assumptions in order to clarify a business position
(Drucker, 1994; Cook, 2016) or take employees in a radically new direction (Furr and Dyer,
2014), leaders commonly discuss the assumptions that ground their professions (Suddaby and
35

Greenwood, 2005), and politicians often debate ideological assumptions instead of how they
might actually execute their policies (Simons, 1994). Yet despite the clear prevalence of actors
communicating directly about these collectively held assumptions, there is little theoretical
understanding of how exposing these assumptions influences others and the broader institution
within which these actors reside. This is particularly troubling because several scholars have
speculated that opening up the foundations of an institution to direct examination may have
destabilizing effects for the system itself (Harmon, Green, and Goodnight, 2015; Bitektine and
Haack, 2015).
This study advances a theoretical approach to explore how the financial market responds
to communication from prominent actors when they expose an institution’s assumptions to direct
examination. To do so, I draw on Toulmin’s (1958) model of argument structure—which
includes the structural components of data, warrant, claim, and backing—to propose that actors
can communicate at two structurally distinct levels in the same message. At one level, actors can
argue within the rules of the game, which is the type of talk much of the existing research
concerns. This is where actors use data or warrants to argue for a particular claim, while leaving
implicit the assumptions—or backing—that ground the prevailing institutionalized context. By
leaving implicit these assumptions, I propose that this structural level of talk tends to reinforce
and reproduce the legitimacy of those very assumptions. At another level, actors can argue about
the rules of the game themselves. This is where actors talk explicitly about the backing, which I
argue exposes the contingencies of the institution, placing the legitimacy of its taken-for-granted
assumptions at risk. I use this distinction to develop a novel measurement—called the argument
structure ratio (ASR)—which captures the variation between these two structurally distinct levels
in communication. Strategic messages with high ASRs contain a large proportion of backing-
36

related talk and, therefore, reflect the fact that these messages are making explicit the
assumptions underlying a given meaning system. My primary argument in this paper is that
strategic messages with higher ASRs point to the contingencies and therefore potential instability
of those very assumptions that undergird a given institutional arrangement and, as a result, will
increase uncertainty.
I contend that examining the argument structure of public messages alters how scholars
and practitioners understand the role of strategic communication in markets and institutions more
generally. First, by exploring the effects of opening up an institution’s assumptions to direct
examination, this study extends our existing social constructionist view of strategic
communication in markets. Organization theorists widely agree that assumptions form the basis
of our institutions (Berger and Luckmann, 1966), providing taken-for-granted guidelines for how
we organize ourselves (Suchman, 1995; Thornton, Ocasio, and Lounsbury, 2012). While we
often leave these assumptions implicit in daily life (Zucker, 1977; Green, Li, and Nohria, 2009),
I show that making them explicit may have important and potentially destabilizing effects on
institutions and markets. Examining the implications of the ASR of strategic communication thus
reveals a new way of thinking about the role of social assumptions and consensus undergirding a
community of actors. Second, this study also suggests a substantial qualification to the
information asymmetry reduction argument held by financial economists. While this perspective
assumes that more communication should reduce overall uncertainty in the market (Spence,
1973), this study suggests that this may not always be true. In particular, I show that more
communication can actually increase uncertainty if it exposes the fundamental assumptions
underlying the prevailing institution to direct examination.
37

I test these arguments in an empirical context where the prominent actors expressly use
strategic communication to reduce market uncertainty: public speeches made by the Chairperson
of the United States Federal Reserve (Fed). Specifically, I examine how the ASR of Fed
speeches from 1998 to 2014 affects market uncertainty, as measured by market volatility (i.e.,
the VIX Index). My findings show that in general more communication from the Fed reduces
uncertainty in the market as one might expect, but that over and above this effect, the more the
Fed makes explicit their assumptions (i.e., by talking more about the backing), the more this
actually increases market uncertainty. I then theorize how this market reaction is likely an
emotional response and, if so, how a deeper understanding of the role of emotions in markets can
offer a theoretically-driven but pragmatic solution to when the Fed can talk about the backing
without creating these undesirable effects on the market. First, I theorize how the emotional
positivity of their speeches can mask the sensitive nature of the backing. Second, using propriety
daily market sentiment data from MarketPsych and Thomson Reuters, I explore how the level of
fear in the business news media leading up to the speech also can mitigate negative reactions to
the discussion of the assumptions underlying the Federal Reserve System.
UNITED STATES FEDERAL RESERVE
Federal Reserve System
The Federal Reserve is the central banking system of the United States. The Fed was
established in 1913 to protect investors during financial panics by guaranteeing liquidity and
acting as the lender of last resort. Based in Washington, D.C., the presidentially appointed seven-
member Board of Governors (with one member appointed as the Chairperson) oversees the
twelve regional Federal Reserve Banks and the broader Federal Reserve System. The structure of
the Fed is unique in that they are “independent within the government” rather than “independent
38

of government” (Federal Reserve Board, 2016). This structure, along with their staggered 14-
year terms for Governors, provides the Fed with unusual but arguably necessary independence
from political pressure and involvement. As a result, the Fed’s basic framework changes
infrequently, with the last major modification coming with the Federal Reserve Reform Act of
1977.
Since this last reform the fundamental assumptions that ground the basic framework of
the Federal Reserve System have remained reasonably stable. These assumptions concern the
nature and boundaries of United States monetary policymaking, which includes the Fed’s
fundamental objectives and the monetary policymaking tools they use to achieve those
objectives. The Fed’s objectives, which are also known as the Fed’s dual Congressional mandate,
are to maximize employment and maintain price stability. To achieve these objectives, the Fed
uses a variety of conventional tools—like engaging in open market operations, setting the
discount rate, or changing member bank reserve requirements—to conduct monetary policy.
These objectives and tools form the “rules of the game” underlying the institution of United
States monetary policymaking. These fundamental assumptions thus form the taken-for-granted
foundation of a bounded, specialized discourse about United States monetary policy involving
highly educated and sophisticated participants (e.g., the Fed, economists, option traders, etc.).
The presumptive nature of these objectives and operations are elemental to the stability of
the United States economy and entire financial system, making the Fed one of the most
important and powerful institutions in the world (Cruikshank and Sicilia, 1999; Abolafia, 2004;
Holmes, 2013). Because of this, the Fed also retains a broader responsibility to maintain
confidence and market stability (Bernanke, 2015). Prior to the 1990s, the Fed’s method of
achieving this was to “never explain, but behave predictably” (Yellen, 2013). So long as the Fed
39

did not act in unexpected or surprising ways, the idea was that market participants would find
certainty and comfort through simply observing such predictability. But by the early 1990s, the
Fed began to realize that their effectiveness in influencing the market also depended on their
ability to shape people’s expectations of the future, “specifically by helping the public
understand how it intends to conduct policy over time, and what the likely implications of those
actions will be for economic conditions” (Yellen, 2013). The Fed recognized that the most direct
means of achieving this was through communicating with the public (Bernanke, 2013).
Federal Reserve Communications
Since 1996, the Fed has used communication to manage the public’s expectations and
maintain market stability. One of the most influential communication channels has been
speeches given by the Fed Chairperson. These speeches are formal, planned presentations given
all over the world (e.g., Washington, D.C., regional Federal Reserve Banks, market competition
conferences, academic institutions, etc.). Although speeches are physically presented to local
audiences, the transcripts for these speeches are made available to the broader public on the
Federal Reserve website when the speech begins. Speeches cover a wide range of topics entirely
of the Chairperson’s own choosing (Bernanke, 2015). Importantly, these speeches are expressly
not the opinion of the Board of Governors but instead reflect the opinions of the Chairperson him
or herself. In this way, speeches differ in meaningful ways from the two other major forms of
Fed communication—press releases and testimony—both of which are more routinized,
constrained, and therefore repetitive in their wording than are speeches. Speeches thus offer the
Chairperson an important and unique opportunity to personally influence the markets.
The last two decades have demonstrated just how critical Fed communication has become
in trying to maintain overall market stability and confidence (Bernanke, 2013, 2015; Holmes,
40

2013). Indeed, for much of this time, either conventional monetary tools failed to influence the
economy in the way economists would predict or these tools reached their practical limits (e.g.,
the federal funds rate reached zero), leaving many to feel as though the Fed was left “using
communication—mere words—as its primary monetary policy tool” (Yellen, 2013). The old
adage—when the Fed speaks, the world listens—became increasingly more apparent during this
time period. But it had also become surprisingly unclear exactly what the world was listening to.
The traditional answer from financial economics of course was that the market was listening to
these communications to simply gather new information that should reduce the overall level of
uncertainty about the future direction of the economy. However, anecdotal evidence continually
pointed to occasions when Fed communications actually created uncertainty instead of reducing
it (Holmes, 2013). I therefore propose that one reason for the ambiguous direction of the effect of
the Fed’s communication is that the market may also be listening to the structural level at which
the Fed is talking—whether they are leaving their assumptions implicit or exposing them to
public scrutiny. To examine this possibility, I examine the underlying argument structure of
strategic communications from the Fed and its potential effect on the market. To do so, I first
review what argument structure is and develop theory as to why it matters in the context of the
Fed.
ARGUMENT STRUCTURE AND MARKET UNCERTAINTY
Argument Structure
Argument structure concerns the way we naturally organize our reasoning in order to
make our communication with others legitimate and persuasive. Aristotle (1991) developed one
of the first approaches to conceptualizing argument structure. Aristotle suggested that the most
persuasive arguments were naturally arranged into syllogistic forms. Syllogisms traditionally
41

contain three structural components: a major premise (e.g., if the acquisition of a company
improves financial performance, then it is a sound business decision) and a minor premise (e.g.,
the acquisition of Company Alpha improves financial performance) that necessarily leads to a
conclusion (e.g., therefore, the acquisition of Company Alpha is a sound business decision).
Since Aristotle’s time, a large amount of research in communication and rhetorical studies has
explored how different syllogistic structures impact the persuasiveness of arguments (Bitzer,
1959; Jackson and Jacobs, 1980; Conley, 1984). Several organization theorists have used the
Aristotelian syllogism in their research (Heracleous and Barrett, 2001; Green, Li, and Nohria,
2009); however, this model of argument structure has gained little traction in organizational
analysis.
This is in part due to two limitations of the syllogistic approach when studying strategic
communication in organizations and institutions. The first is that syllogisms fundamentally are
deductive proofs, suggesting that clearly stated premises should necessarily persuade an audience
of a desired conclusion. This however leaves little room for the political contestation many
researchers tell us occurs regularly in institutions (Holm, 1995; Fligstein, 1997; Seo and Creed,
2002; Covaleski, Dirsmith, and Rittenberg, 2003). This leads to the second and perhaps most
important limitation, which is that syllogisms have no way to conceptualize the source of the
socially shared assumptions implicitly underlying our arguments. For instance, the previously
stated major premise (i.e., actions that improve financial performance are sound business
decisions) is grounded by the assumption that “profitability” is the generally accepted purpose of
business operations. This may not be true however because organizing one’s actions around the
profitability assumption is not the only way to conduct business. Nevertheless, the syllogistic
approach does not account for the existence or potential contestation of these assumptions.
42

These limitations led British philosopher Stephen Toulmin to develop an alternative
approach to argument structure that I argue resonates better with our understanding of strategic
communication and meaning structures underlying our institutions. Like Aristotle, Toulmin
(1958) begins with the same three structural components: actors he argues provide data (i.e.,
minor premise) to support a claim (i.e., conclusion), which is further supported by virtue of a
warrant (i.e., major premise, explaining why the data support the claim). But unlike Aristotle,
Toulmin observes that everyday arguments rarely start with abstract assertions that deductively
lead to conclusions. Instead, arguments tend to function in the opposite direction, beginning with
claims and only contain data and warrants to the extent that the audience demands further
justification. This picture of argumentation is not only far more realistic for how organizational
actors actually justify their decisions (Pfeffer, 1981; Staw, McKechnie, and Puffer, 1983), but it
also provides room for political contestation (Toulmin, Rieke, and Janik, 1984; Bouwmeester,
2013).
Crucially, Toulmin also differs from Aristotle in that he adds to his model an important
fourth structural component called the backing (see Figure 1). The backing is the often implicit
assumptions that form the basis for the most appropriate way to ground the argument (Stephen
Toulmin, 1958; Goodnight, 1993). In the example above, the implied backing was
“profitability,” which provides the presumptive basis for believing that using data related to
financial performance would be a valid and legitimate way to justify the claim that acquiring
Company Alpha was a sound business decision. With the conceptual addition of the backing, I
argue that the Toulmin Model provides a theoretical foundation that resonates strongly with a
wide group of organization theorists who conceptualize our social world as being grounded on
43

often implicit, socially constructed but nevertheless shared assumptions (Berger and Luckmann,
1966; Friedland and Alford, 1991; Townley, 2008; Thornton, Ocasio, and Lounsbury, 2012).
In this study, the concept of the backing is critical because it provides a conceptual and
empirical way to point to an observable second structural level of communication (Harmon,
Green, and Goodnight, 2015). Indeed, much of our daily communication takes place within the
rules of the game at the data-warrant-claim level, leaving implicit the backing. This is the level
of communication that most of the existing research on strategic communication examines
(Green et al., 2009; see also Elsbach, 1994; Lamin & Zaheer, 2012; Rhee & Fiss, 2014).
However, prominent actors also regularly talk about the backing, discussing these assumptions
directly and talking “about the rules of the game” underlying the existing institutionalized
arrangement (Bitektine and Haack, 2015). Several scholars have recently conceptualized
communication occurring at these two structurally distinct levels as entirely separate processes
(Harmon, Green, and Goodnight, 2015), but we can also imagine that every day public
communication simultaneously contains both of these structural levels in an attempt to make a
maximally persuasive argument.
Argument Structure Ratio
To explore this possibility, I develop a new theoretical construct that I call the argument
structure ratio (ASR). Conceptually, the ASR captures the variation between communication that
occurs within the rules of the game (i.e., engages the structural components of data, warrant, and
claim) and about the rules of the game (i.e., engages the structural components of backing).
Theoretically, the ASR assesses the degree to which a message makes explicit the assumptions
underlying the existing institutionalized arrangement. Thus, messages with high ASRs contain a
large proportion of backing-related talk.
44

Consider the game of baseball. Baseball has many rules, such as the number of outs in an
inning, when the substitution of players can occur, how and when to use instant replay, the size
of the strike zone, and even the etiquette its players should observe. For much of the time, the
validity of these rules is taken as a given by players, coaches, fans, and sportscasters. However,
when listening to post-game interviews after a controversial call occurred during the game, you
often hear an impressive variation in the structural level of talk. For instance, some interviewees
continue to take as a given the validity of the rule undergirding the controversial call and instead
discuss how they could have played differently within the prevailing rules to perhaps achieve a
different outcome (e.g., low ASR messages). In contrast, other interviewees may take issue with
the appropriateness of the rule affecting the controversial call, talking about the validity of the
rule itself (e.g., high ASR messages).
Now consider the context of the Fed, where the game is not baseball but United States
monetary policymaking. The rules underlying this game, as described earlier, are related to the
objectives and conventional tools that form the basis of the Federal Reserve System. And just as
in post-game baseball interviews, Fed Chairperson speeches can exhibit the same variation in
argument structure. For instance, in speeches where the Chairperson is talking primarily within
the rules of the game, he or she typically is making claims about the state of the US economy and
providing economic data to support these claims. For example, in a speech to the Economic Club
of Washington, D.C. on December 7, 2009, Chairperson Bernanke asserts an initial claim about
the recent recovery of the economy and then provides three pieces of data to justify this claim:

A number of factors support the view that the recovery will continue next year (claim). Importantly,
corporations are having relatively little difficulty raising funds in the bond and stock markets (data), stock
prices and other asset values have recovered significantly from their lows (data), and a variety of indicators
suggest that fears of systemic collapse have receded substantially (data) (Bernanke, 2009).

45

In contrast, in speeches where the Chairperson is primarily talking about the rules of the
game, he or she articulates the backing explicitly by discussing the nature or boundaries of
monetary policy. For example, in a speech given at the Federal Reserve Bank of St. Louis on
October 11, 2001, Chairperson Greenspan lays bare the assumptions underlying the Federal
Reserve System:

We at the Federal Reserve are given two mandates that are not often spelled out explicitly. First, to
implement an effective monetary policy to meet our legislated objectives (backing). Second, to do so in a
most open and transparent manner in recognition that we, as unelected officials, are accountable both to the
Congress from which we derive our monetary policy mission and to the American people (backing)
(Greenspan, 2001).

I argue that the variation in the amount of backing-related talk in a Fed Chairperson’s
speech will lead to very different inferences about the stability of the Federal Reserve System as
an institution and, by extension, the United States economy.
Argument Structure Ratio and Market Uncertainty
The more the Fed Chairperson leaves the backing implicit in their public communications
(i.e., speeches with low ASRs), the more it reinforces the collective comfort over and legitimacy
of those assumptions underlying the United States Federal Reserve System (Harmon, Green, and
Goodnight, 2015). This is consistent with the notion in institutional theory that completely
legitimate ideas, practices, or organizations essentially “go without saying” (Meyer and Scott,
1983; Suchman, 1995; Green, Li, and Nohria, 2009; Tost, 2011). Indeed, if the backing is
legitimate, then naturally it too should “go without saying.” Moreover, if the Fed Chairperson is
talking and decided not to discuss these assumptions, then this provides even more authoritative
proof that these assumptions are indeed legitimate (Green, 2004; Rhee and Fiss, 2014; Bitektine
and Haack, 2015: 51). By reinforcing the legitimacy of the Federal Reserve System, the Fed
Chairperson is implying that we need not even question the stability of the institution that sits at
46

the epicenter of the United States market economy. This sense of collective comfort allows
market participants to feel more at ease deferring to this shared feeling of stability instead of
cultivating their individual judgment, leading to a “greater conformity and isomorphism” in
judgments and decisions openly expressed in the market (Bitektine and Haack, 2015: 53–54;
Meyer and Rowan, 1977; DiMaggio and Powell, 1983). As a result, low ASR speeches narrow
the overall range of expected directions the market could go in the future, thereby reducing
market uncertainty.
In contrast, the more the Fed Chairperson makes the backing explicit in their public
communications (i.e., speeches with high ASRs), the more it creates anxiety by putting at risk
the legitimacy of those assumptions underlying the United States Federal Reserve System
(Harmon, Green, and Goodnight, 2015). Since the backing typically goes without saying and is
presumed legitimate, the Fed Chairperson pointing to these assumptions exposes the
contingencies of the institution and signals to market participants the possibility that there might
be a problem with the legitimacy of the United States monetary policy framework (Bitektine and
Haack, 2015: 58). Indeed, if the Fed was not reexamining those assumptions themselves
(Bernanke, 2015) or these assumptions were not being questioned by external parties thereby
forcing the Fed to address them publicly (Paul, 2009), then the Chairperson would have had little
reason to state them. This is especially true when it comes to Chairperson, who is in charge of
maintaining the stability of that very system (Holmes, 2013). By signaling a risk to the
legitimacy of the Federal Reserve System, the Fed Chairperson is implying that this institution
sitting at the epicenter of the United States market economy may be questioned (Meyer and
Scott, 1983; Suchman, 1995; Green, Li, and Nohria, 2009; Tost, 2011). I argue that this will
prompt market participants to feel a reduced sense of collective comfort and thus increased
47

anxiety about deferring to this now seemingly unstable institutional foundation (Tost, 2011;
Voronov and Vince, 2012; Haack, Pfarrer, and Scherer, 2014). This in turn I propose will lead
them to rely more on their individual judgments of what is going on in the market, producing
greater heterogeneity in overall market judgments and decisions. As a result, high ASR speeches
expand the overall range of expected directions the market could go in the future, thereby
increasing market uncertainty.
Hypothesis 1. The ASR of Fed Chairperson speeches will be positively associated with
market uncertainty.

If this prediction is true, it creates a critical tension for the Fed. On the one hand, the
Fed’s broader responsibility is to maintain market stability by reducing uncertainty, not
increasing it (Abolafia, 2004; Bernanke, 2015). On the other hand, the Fed at times must discuss
their fundamental assumptions if they are to maintain transparency regarding their monetary
policy operations as promised (Bernanke, 2013). Under what conditions then can the Fed discuss
these assumptions without creating market uncertainty? Conceptually, the answer to this question
may rest with the role emotions plays in my Hypothesis 1 prediction. In particular, I argued that
the Fed Chairperson exposing their assumptions will dislodge the feelings of collective comfort
and security that ground the Federal Reserve System, thereby producing emotional anxiety that
creates uncertainty within the market about the future state of the US economy. If this is the case,
then a deeper understanding of how other emotion-based considerations might counteract this
effect could offer a theoretically-driven solution.
EMOTION IN MARKETS
Emotion has long played a prominent role in markets. Early work painted a negative
picture of human emotion, likening them to our underlying “animal spirits” that produced
48

unpredictability in markets (Keynes, 1936). This notion was popularly captured in Nobel Prize-
winner Robert Shiller’s (2000) book, Irrational Exuberance, echoing the same phrase Alan
Greenspan uttered in his December 5, 1996 speech that sent shockwaves through capital markets
around the world (Greenspan, 1996; Wessel, 1996). More recently however scholars have started
to demonstrate that emotion is far from unpredictable (Pollock and Rindova, 2003; Tetlock,
2007; Tetlock, Saar-Tsechansky, and Macskassy, 2008; Pfarrer, Pollock, and Rindova, 2010) and
can operate in market contexts in two distinct yet reciprocal ways: 1) emotional communication
can directly affect people’s market actions and 2) people’s existing emotional state can influence
how they react to financial market events. I explore both of these relationships in order to
theorize the conditions under which the Fed Chairperson should be able to expose their
assumptions without creating market uncertainty.
Speech Emotion
The idea that emotional language can influence people’s beliefs and actions is not new.
Aristotle (1991) was one of the first to argue that emotion is an important aspect of the human
condition that speakers can appeal to in order to persuade audiences. Organization theorists since
have examined how appeals to emotion can help actors institutionalize practices (Green, 2004) or
legitimate actions (Erkama and Vaara, 2010), define new organizational forms (Suddaby and
Greenwood, 2005), and make sense of catastrophes (Cornelissen, Mantere, and Vaara, 2014). To
explore the potential impact of a Fed speech’s emotion, we first need a way to conceptualize the
different types of emotional content in communication.
Communication valence, or talking in a positive versus negative tone, is arguably the
most basic distinction between emotions found in communication (Tversky and Kahneman,
1981; Pennebaker, Mehl, and Niederhoffer, 2003). Further supporting the usefulness of this
49

distinction is a number of studies that demonstrate the organizational benefits of communicating
in a positive rather than negative tone. For instance, information conveyed in a positive light
generally leads audiences to rate people’s performances as better (Levin, 1987), perceive
management control systems as stronger (Schneider, Holstrum, and Marden, 1993), support
organizational practices more (Davis and Bobko, 1986), and generally evaluate issues more
favorably (Levin, Schneider, and Gaeth, 1998). These beneficial effects of communicating in a
positive tone hold in financial markets as well. Davis, Piger, and Sedor (2012) show that using a
positive tone in earnings press releases produces a better short term stock market reaction, even
after controlling for firm characteristics related to fundamentals (Huang, Teoh, and Zhang,
2013). These researchers argue that strategically conveying information in a positive tone signals
an optimistic outlook (even if untrue), thus creating confidence and certainty in the future.
I propose that the Fed Chairperson may engage in this very same linguistic strategy. In
particular, since one of the Fed’s responsibilities is to build market confidence, the Fed
Chairperson may at times seek to narrate or give their speech in a more positive tone in order to
signal optimism about the future. I would expect that doing this would produce a main effect of
reducing market uncertainty, consistent with the aforementioned research that shows positive
tone produces favorable market reactions. More important, I also argue that this would have a
moderating effect on Hypothesis 1. Specifically, giving a speech in a positive tone enables the
Fed Chairperson to encase his or her talk about the backing in an overarching positive-sounding
story. Doing so increases the likelihood that market participants will interpret the Fed’s
statements about their assumptions as a positive signal that their message is reasonable and well-
supported. As a result, the explicit discussion of one’s fundamental assumptions will be less
likely to come across as worrisome or anxiety-producing and, thus, will not signal as great of a
50

risk to the legitimacy of the Federal Reserve System. The Fed Chairperson therefore should be
able to discuss the backing directly without producing market uncertainty so long as the speech
is conveyed in an overtly positive tone.
Hypothesis 2. The positive tone of the Fed Chairperson’s speech will weaken the positive
effect of the ASR on market uncertainty.

Audience Emotion
Markets are also already full of emotion, which can impact how people react (Shiller,
2000; Kahneman, 2011). In this sense, emotions operate as part of people’s baseline and color
their interpretations of and reactions to market events (Pollock and Rindova, 2003). Mountains
of anecdotal evidence highlight how emotions influence our decisions made every day (Gino,
2015), from choosing short-term investment decisions (Statman, 2013) to understanding long-
term macroeconomic issues (Carrns, 2013). Academic research has recognized this role of
emotions in markets as well, demonstrating that the presence of certain emotions can amplify an
audience’s reaction to events (Barberis, Shleifer, and Vishny, 1998; Baker and Wurgler, 2006).
Arguably the most powerful amplifying emotion that arises regularly in financial markets is fear
(Krugman, 2001), which is also cited as the emotional driver of market bubbles as well as the
very financial panics the Fed seeks to avoid (Shiller, 1988, 2000; Holmes, 2013; Bernanke,
2015).
Fear produces two key features that likely influence how people will react to events in
market contexts. First, fear creates an arousal in people that heightens their attention. Indeed,
psychologists (Niedenthal and Kitayama, 2013) and economists (Kahneman, 2011) alike argue
that arousal from emotions like fear modifies “the allocation of attentional resources and
heightens sensitivity to environmental cues” (Lane, Chua, and Dolan, 1999: 889). Second, fear
51

also creates pessimism about the future. In particular, psychologists have found that being in a
state of fear predisposes individuals to interpret events and their environment with increased
pessimism and negativity (Lerner and Keltner, 2001; Lerner, Small, and Loewenstein, 2004).
Taken together, markets that contain elements of fear are thus more likely to lead market
participants to pay closer attention and negatively overreact to events than participants in
markets containing little to no fear.
Based on these considerations, it seems likely that at least some amount of fear would
need to be present already in the market for my original Hypothesis 1 to hold. In particular,
market participants would need to be aroused enough to actually pay attention to the Fed
speeches, and this heightened attention may in turn lead to greater scrutiny of backing-related
talk and more pessimism about the presumably unstable Federal Reserve System. If this is
theoretically the case, then the removal of this fear from the market should diminish this
predicted effect. Specifically, when there are low levels of fear leading up to a Fed speech, I
argue that market participants will likely be paying less attention to the speech, and that the
attention they do allocate will involve less scrutiny of any included backing-related talk. Indeed,
the Fed stating their assumptions in an environment that contains low levels of fear will increase
the likelihood that market participants will interpret the Chairperson’s overall message in a more
positive and less pessimistic way. In other words, this lack of heightened emotion in people’s
baseline will thus provide a safer and less reactive set of conditions in which to expose these
assumptions. The Fed Chairperson therefore should be able to discuss the backing directly
without producing market uncertainty so long as there are low levels of fear prior to the speech.
Hypothesis 3. Low levels of fear prior to the Fed Chairperson’s speech will weaken the
positive effect of the ASR on market uncertainty.
52

METHODS
Sample
My initial sample consisted of all Fed Chairperson speeches given between January 1,
1998 and December 31, 2014, totaling 344. Five speeches were removed because, based on an
outlier analysis, their studentized residuals exceeded plus or minus three. While the findings
remain consistent with outliers included, I removed them because the average spike in market
uncertainty on these days was 10 times that of all other days. To examine why and also to
validate further the reasonability of removing these outliers, I examined what happened in the
markets on these particular days. Unsurprisingly, I found that extreme market events co-occurred
with the Fed speeches on these days that both significantly increased uncertainty (i.e., Greece’s
sovereign credit rating downgraded to junk, 2010 Flash Crash, and Italy’s borrowing costs hit
all-time record high) as well as decreased uncertainty (i.e., Eurozone approved Greece loan, and
US Supreme Court announces long-awaited ruling). The final sample thus consisted of 339
speeches. Of these, 159 were given by Alan Greenspan, 166 by Ben Bernanke, and 14 by Janet
Yellen (see Figure 3).

53

FIGURE 3 – Fed Chairperson Speeches from 1998 - 2014

Endogeneity
In the ideal experiment, I would be able to randomly assign the ASR of Fed speeches and
observe the change in the level of market uncertainty on the day of the speech. Since I am unable
to perform this ideal experiment, I face a situation in which market factors that influence market
uncertainty may also influence how the Fed talks. I acknowledge that Fed speeches can be
influenced by some market factors, since the Fed’s job is to pay attention and respond to market
conditions. However, we also know that the Fed’s communication, over and above these existing
factors, can independently influence market uncertainty (Cruikshank and Sicilia, 1999; Nikkinen
and Sahlström, 2004; E.-T. J. Chen and Clements, 2007). Therefore, my empirical design tries to
isolate this independent effect specifically with respect to the ASR. I also report the results from
a number of supplementary tests to provide further evidence of the robustness of my main
findings.

54

Dependent Variable
I measure market uncertainty using the VIX volatility index. The VIX is a daily index
calculated by the Chicago Board Options Exchange “by averaging the weighted prices of S&P
500 puts and calls over a wide range of strike prices” (Chicago Board Options Exchange, 2003:
2). The VIX thus represents option traders’ estimates of the direction of the S&P 500 over the
next month by providing an aggregate measure of the dispersion or variance of option prices on
any given day. The higher the variance across option traders’ beliefs about where the market will
go in the next month, the more uncertainty there is in the market. The VIX is an appropriate
measure for market uncertainty for three reasons. First, the VIX is the world’s premier barometer
for measuring expected market volatility (Chicago Board Options Exchange, 2003). And since
scholars and market participants interpret expected volatility as the market’s expectation of the
average volatility over the remaining life of the option contract (Merton, 1973), it is reasonable
that the uncertainty is reflected in this measure (Nikkinen and Sahlström, 2004). Second, the
VIX is the standard approach finance scholars use to measure market uncertainty (Connolly,
Stivers, and Sun, 2005; Ang et al., 2006; Andersson, Krylova, and Vähämaa, 2008; Bia’lkowski,
Gottschalk, and Wisniewski, 2008). Third, option traders represent a highly educated and
sophisticated audience and are important participants within this highly specialized discourse
about United States monetary policymaking.
To remain consistent with existing finance research that examines the impact of Fed
communication on the VIX (Nikkinen and Sahlström, 2004; E.-T. J. Chen and Clements, 2007;
Nikkinen, Sahlström, and Äijö, 2007; Jamali, 2009; Vähämaa and Äijö, 2011), I use a two-day
event window (t-1 to t0) to measure the change in market uncertainty produced by a Fed speech.
1

1
A two-day event window is more appropriate than shorter event windows, like 15-minutes (Ederington and Lee,
1993) or 45-minutes (C. R. Chen, Mohan, and Steiner, 1999), because of the type of information I am examining.
55

That is, I measured the change in the VIX from the market close on the day before the speech to
the market close on the day of the speech. The speech date was determined by the date and time
stamp on the speech itself, and corroborated with the date and time information I received from
my Freedom of Information Act Request No. G-2015-00191. I employ the following regression
model:
ln(VIXit/VIXit-1) = α + βASRit + ηControlsit + εit

Independent Variables
Argument structure ratio. Speeches were coded for their ASR by the author and three
business school undergraduate students with familiarity on macroeconomics and monetary
policy. The coding process followed five steps, consistent with recommendations for content
analysis (Neuendorf, 2001; Krippendorff, 2003). First, I wrote a detailed manual for how to code
for the ASR, which was used to train the three undergraduate coders. Second, we engaged in
pilot coding. To do so, all four coders coded 20 speeches (five at a time) independently and we
subsequently discussed our coding decisions until we reached a consensus on each speech. Third,
we engaged in reliability coding. To do so, all four coders coded 60 speeches, or 18 percent of
the final sample, which is an appropriate number in this context (Krippendorff, 2003: 240).
Interrater reliability was acceptable (Krippendorff’s alpha = 0.88) (Hayes and Krippendorff,
2007), suggesting consistency across the four coders. Fourth, we engaged in independent coding.

These shorter event windows studies examine how the market reacts to a single information point released at the
same time of the day (e.g., unemployment rate). The information I am examining is not a single number or decision,
but instead a speech that takes some time to be digested by the market. A two-day event window is also more
appropriate than longer event windows, like 15 or 30 days, because the VIX is a sensitive measure that absorbs
information quickly. This makes longer event windows more empirically challenging when trying to control for
alternative influences. Because of this, scholars that examine VIX over periods of time longer than a few days are
typically exploring only correlational relationships between the VIX and other macroeconomic factors (e.g.,
Connolly, Stivers, and Sun, 2005; Andersson, Krylova, and Vähämaa, 2008), and not looking at changes in the VIX
during an event window as I am here.
56

To do so, the remaining speeches were randomly assigned to one of the four coders. Fifth, to
validate ongoing interrater reliability, I randomly selected ten speeches towards the end of our
independent coding that we again all coded, noting no substantial deviations from original
reliability testing (Krippendorff’s alpha = .84).
The unit of analysis of the ASR is the speech, calculated by the following formula:
ASR = (number of arguments that expose the backing / total number of arguments)

There are two considerations to keep in mind. First, a ratio is used instead of the raw
number of arguments that expose the backing because the relative influence of a single backing-
related argument will differ substantially depending on the length of the overall speech. Second,
following prior work (e.g., Green et al., 2009), we coded each paragraph as one argument.
Conceptually, the paragraph was the appropriate unit of data collection because people regularly
cluster their arguments into paragraph form, which is especially true in formal speeches when
they are trying to enhance transitional clarity for their audience. Pragmatically, coding at a finer-
grained level (e.g., each individual sentence) would be challenging. Since on average there were
25 paragraphs per speech, and 6 sentences per paragraph, we would have gone from manually
coding over 8,000 paragraphs to coding over 50,000 sentences. Moreover, since the majority of
paragraphs remained clearly at one of these two levels, this further validated the reasonableness
of the paragraph as an appropriate unit of data collection. For the handful of paragraphs that were
less clear cut, all coders followed the same general rule: If the Fed Chairperson exposed the
backing at all, the paragraph was coded as arguing about the rules of the game.
For a paragraph to be coded as arguing within the rules of the game (i.e., the backing was
not exposed), the Chairperson must engage only in the structural components of data or warrants
to draw conclusions about a claim. However, they need not engage in all three components. Most
57

frequently, the Chairperson provided some sort of evidence about actions that they have taken or
economic-related metrics they have collected in order to make some sort of a claim about the
state of the economy. For instance, in a speech on September 26, 2005 to the American Bankers
Association, Chairperson Greenspan makes an initial claim, supports this claim with data, and
then reasserts the claim.

This enormous increase in housing values and mortgage debt has been spurred by the decline in mortgage
interest rates, which remain historically low (claim). Indeed, the thirty-year fixed-rate mortgage, currently
around 5 3/4 percent, is about 1/2 percentage point below its level of late spring 2004, just before the
Federal Open Market Committee (FOMC) embarked on the current cycle of policy tightening (data). This
decline in mortgage rates and other long-term interest rates in the context of a concurrent rise in the federal
funds rate is without precedent in recent U.S. experience (claim) (Greenspan, 2005).

Similarly, in a speech on January 3, 2014 to the American Economic Association, Chairperson
Bernanke makes an initial claim and then supports this claim with a variety of data.

The economy has made considerable progress since the recovery officially began some four and a half
years ago (claim). Payroll employment has risen by 7-1/2 million jobs from its trough (data). Real GDP has
grown in 16 of 17 quarters (data), and the level of real GDP in the third quarter of 2013 was 5-1/2 percent
above its pre-recession peak (data). The unemployment rate has fallen from 10 percent in the fall of 2009
to 7 percent recently (data). Industrial production and equipment investment have matched or exceeded
pre-recession peaks (data) (Bernanke, 2014).

For a paragraph to be coded as arguing about the rules of the game (i.e., the backing was
exposed), the Chairperson at some point must explicitly and directly engage the backing. This
most frequently occurs when the Chairperson reflects on the nature and boundaries of monetary
policy objectives and conventional tools. For instance, in a speech on October 18, 2011 at the
Federal Reserve Bank of Boston’s 56th Economic Conference, Chairperson Bernanke reaffirms
the Fed’s dual Congressional mandate, followed by a discussion about how inflation targeting
fits into their monetary policy framework.

The Federal Reserve is accountable to the Congress for two objectives —maximum employment and price
stability, on an equal footing—and it does not have a formal, numerical inflation target. But, as a practical
58

matter, the Federal Reserve's policy framework has many of the elements of flexible inflation targeting. In
particular, like flexible inflation targeters, the FOMC is committed to stabilizing inflation over the medium
run while retaining the flexibility to help offset cyclical fluctuations in economic activity and employment
(backing) (Bernanke, 2011).

Similarly, in a speech on April 16, 2014 to the Economic Club of New York, Chairperson Yellen
discusses the challenges of monetary policy when the primary conventional tool—the federal
funds rate—is pinned near zero. Yellen discusses how the Fed was forced to expand the
boundaries of their monetary policy toolkit and consider alternative, perhaps less conventional
tools.

The idea that monetary policy should react in a systematic manner in order to blunt the effects of shocks
has remained central in the FOMC's policymaking during this recovery. However, the application of this
idea has been more challenging. With the federal funds rate pinned near zero, the FOMC has been forced to
rely on two less familiar policy tools the first one being forward guidance regarding the future setting of the
federal funds rate and the second being largescale asset purchases. There are no time tested guidelines for
how these tools should be adjusted in response to changes in the outlook. As the episodes recounted earlier
illustrate, the FOMC has continued to try to adjust its policy tools in a systematic manner in response to
new information about the economy. But because both the tools and the economic conditions have been
unfamiliar, it has also been critical that the FOMC communicate how it expects to deploy its tools in
response to material changes in the outlook (backing) (Yellen, 2014).

Speech positive tone. Following existing work (Pfarrer, Pollock, and Rindova, 2010;
Rhee and Fiss, 2014), I used text analysis software Linguistic Inquiry and Word Count (LIWC)
to create an index that captures the relative amount of positive emotional words in relation to all
the words in the speech. Positive emotional content is captured using a dictionary approach,
whereby words psychometrically related to positive emotion (e.g., happy, good, nice, positive,
great, favorable, etc.) are compared against all the words in each Fed speech. This positive
emotion word dictionary was compiled and validated by Pennebaker and his colleagues (2007;
2007; 2010).
Audience Fear. I was granted full access to the Thomson Reuters Market Psych Indices
database. This proprietary database contains daily indices on 48 different emotions present in the
59

United States media from January 1, 1998 to present. Every five minutes, their algorithms scrape
business news media sources using word dictionaries associated with these 48 emotions. At the
end of the day, an average is taken to create a daily index for each market emotion. Their
business news sources include Reuters and top business print newspapers, as well as Internet
business news sources starting in 2005. For the purposes of this paper, I am using the daily fear
index (e.g., containing words like worrisome, concerning, anxious, fearful, panicky, etc.). While
it would have been ideal to have a direct measure of option traders’ fear, no such measurement
exists. However, option traders and other market participants pay careful attention to the
business news media (Pollock and Rindova, 2003), making this a reasonable proxy for audience
fear. I constructed an audience fear variable by taking the average of the fear index over a three-
day window (t-3 to t-1) before the day of the speech, thereby capturing the market’s emotional
baseline from which they would hear the speech.
2

Control Variables
To remain as orthodox in my research design as possible, I controlled for key variables in
organization theory research on strategic communication as well as in financial economics
research on the Fed and the VIX. I also added control variables unique to my specific research
question that may play a conceptually meaningful role. I grouped these control variables into
three categories based on distinct theoretically-driven concerns underlying my research design.
Existing market conditions. First, I controlled for conditions present in the market before
each Fed speech that could simultaneously influence the ASR of these speeches and produce
changes in market uncertainty directly. For instance, I controlled for existing market uncertainty

2
To provide additional comfort surrounding the accuracy and reasonableness of this audience fear variable, I also
validated that this measure correlated with other Thomson Reuters Market Psych emotional indices in the way that
one would expect during the same three-day window. As expected, fear positively correlated with gloom (r = 0.77)
and stress (r = 0.81) and negatively correlated with sentiment (r = -0.46) and optimism (r = -0.54).
60

prior to the speech because greater preexisting uncertainty could encourage the Fed to talk more
about their backing and, at the same time, increase future variation in the VIX. I calculated the
VIX raw by taking its 30-day average prior to each speech. I also controlled for existing market
conditions directly linked to Fed policies. Specifically, I controlled for the unemployment rate
and inflation rate because they are market indicators of the Fed’s operations and performance
with regard to their dual mandate of maximizing employment and maintaining price stability,
respectively. I gathered this data from the Bureau of Labor Statistics. Since using average
monthly rates prior to each speech created issues of multicollinearity but did not change my
findings, I decided instead to use the slope of the previous six months unemployment and
inflation data. While not ideal, this control variable has the added benefit of capturing the
direction the economy appears to be moving rather than the static values of unemployment and
inflation.
I also controlled for whether the prevailing monetary policy conditions reflected an
expansionary or contractionary time period. If the most recent FOMC meeting resulted in
lowering the federal funds rate (e.g., expansionary monetary policy), the observation was coded
“1,” otherwise it was coded “0.” If the most recent FOMC meeting resulted raising the federal
funds rate (e.g., contractionary monetary policy), the observation was coded “1,” otherwise it
was coded “0.” Moreover, since 2008 there have been increasing levels of dissent within these
FOMC meetings (Plosser, 2015; Hilsenrath, 2016). The level of dissent across the 12 voting
members in these meetings could provide an important signal about the legitimacy of the
monetary policy framework, prompt the Chairperson to discuss their backing in future speeches,
and create more uncertainty in the market. Besides the Chairperson, the voting members include
the six other Governors from the Board in Washington, D.C. and five rotating Presidents from
61

regional banks (with the New York Fed president always voting). Since Governors are part of the
Board in D.C. but regional Presidents are not, the meaning of a dissent from these two parties
may indicate different things to the market. As such, I controlled for dissent governor and dissent
president separately by counting the number of dissenting votes from each group at the last
FOMC meeting prior to each speech.
I also controlled for several conditions surrounding the Fed speech itself. For instance, I
controlled for the speech location because speeches given in Washington, D.C. may contain
qualitatively different messages than speeches given elsewhere around the world. Indeed,
Washington, D.C. has been the headquarters of the Federal Reserve System since its inception in
1913 and is also the meeting location of the FOMC. For these reasons, Washington, D.C. is
considered both domestically and internationally the epicenter of central banking and monetary
policymaking. I coded speeches located in Washington, D.C. as “1,” otherwise “0.” Finally,
consistent with work in finance that suggests that releasing new informatio n into the market
earlier in the week produces more volatility (Chang, Pinegar, and Ravichandran, 1993; Dubois
and Louvet, 1996; Choudhry, 2000; Connolly, Stivers, and Sun, 2005), I controlled for the
weekday on which the speech took place. Speeches given on Mondays were coded as “1,”
Tuesdays as “2,” and so on.
Contemporaneous information. Second, I controlled for new information released into
the market around the same time as when the Fed speech is given because such information
could hypothetically be correlated with the ASR of the speech and more importantly directly
impact market uncertainty rather than the speech. Prior work suggests that three scheduled
macroeconomic news releases have a major impact on the financial markets (Ederington and
Lee, 1993; Nikkinen and Sahlström, 2004): the Consumer Price Index report, the Producer Price
62

Index report, and the Unemployment report. I compiled the dates of these major announcements
from the Bureau of Labor Statistics. A dummy variable is created for each report, coded “1” if
the report was released on the same day as the speech, “0” if not. I also control for additional
information the Fed itself introduced into the market. I control for BOG speeches if a Governor
also gave a speech on the same day as the Chairperson, testimony if another member of the Fed
testified to Congress on that same day, and press releases if the public relations department at
the Fed issued any type of press release on that same day. For each of these items, a dummy
variable was coded “1” if the information was released on the same day as the speech, “0” if not.
Speech characteristics. Third, I also controlled for additional characteristics of the Fed
speech itself that may be the source of change in market uncertainty rather than my ASR
measure. Since longer speeches may contain more information for market participants to
interpret, thereby reducing information asymmetry and uncertainty (Spence, 1973; Van Buskirk,
2012), I controlled for the speech word count. I also controlled for the actual level of speech
uncertainty to ensure that it is not simply the Chairperson sounding more uncertain that is
creating market uncertainty. I used the Financial Sentiments Dictionary created by Loughran and
McDonaold (2011) to measure the percentage of words in each speech related uncertainty (e.g.,
uncertain, variability, depend, contingency, fluctuate, indefinite, etc.).
One prominent characteristic of Fed Chairperson communication that supposedly
influences the market is called “Fedspeak,” which according to former Fed Vice Chairperson
Alan Blinder (2001) is complex, abstract, or vague language used by Fed Chairpersons to
obfuscate sensitive subjects so as to avoid unnecessary market uncertainty. In an interview on
September 16, 2007, Alan Greenspan referred to Fedspeak as an intentional form of “syntax
destruction, which sounded as though [he] were answering the question but in fact had not” (60
63

minutes, 2007). I controlled for the three primary aspects of Fedspeak. I controlled for speech
complexity using the Flesch-Kincaid reading grade level (Kincaid et al., 1975), which is used
extensively in the field of education as well as in writing insurance and legal documents
(McClure, 1987). The score is based on a formula that considers the total number of words,
sentences, and syllables in a text, and produces the U.S. grade level required to understand that
text. This measure comes as close as possible to Greenspan’s notion of “syntax destruction.” I
controlled for speech abstractness using a word dictionary compiled and validated by
Mergenthaler (1996) that contains a comprehensive list of abstract nouns (e.g., nouns often
ending in –ism, -age, -ness, etc.). I controlled for speech vagueness using a partial dictionary
compiled and validated by Hiller and colleagues (1969) that contains words that signal a lack of
verbal clarity (e.g., maybe, various, perhaps, probably, etc.).
Researchers have suggested that the positive tone of communication cannot be
understood in isolation from its negative tone (Pfarrer, Pollock, and Rindova, 2010). Thus, to
control for how negative emotion contained in the speech may counteract the positive emotion
used, I controlled for speech net emotion by using the LIWC index called “affect,” which nets all
positive emotional content (i.e., the variable noted above) with all negative emotional content.
Finally, I controlled for the topic of the speech. Since topics are inherently context-
dependent, no a prior categories were available. To address this, I inductively derived topics by
reading a randomized selection of 40 speeches, continually writing down possible topics as I
read. After 40 speeches, I had a list of eight topics. I then read another randomized selection of
40 speeches to validate these topics, identifying a ninth topic. The nine topics were: the state of
the economy, the financial crisis, financial literacy, central banking, the banking system,
globalization, economic history, commencement addresses, and remarks based on special
64

occasions. I discussed these nine topics with my three research assistants who were intimately
familiar with the context, noting no issues and no additional topics. One of the research assistants
and I then coded all speeches using a forced-choice methodology with these nine topics, with
interrater agreement of 99 percent. I controlled for these nine topics using dummy variables.
RESULTS
To conduct this event study analysis, I used OLS regression with topic- and year-fixed
effects to estimate the effect of Fed speeches’ ASR on market uncertainty.
3
Descriptive statistics
and correlations for all variables are shown in Table 4. As the table shows, market uncertainty
for the two-day window (t-1 to t0) is negative. Comparing days with Fed speeches (M = -0.0051,
SD = .06) to days without Fed speeches (M = 0.0000, SD = .07) shows a marginally significant
difference, t(4294) = 1.677, p < .10, indicating that Fed speeches in general reduce the VIX
Index by 0.5 percent. This provides preliminary evidence consistent with the financial
economists’ view that more communication reduces information asymmetries and thus reduces
market uncertainty.

3
Including speaker-fixed effects to this specification produces the same results . However, I remove them for my
primary analyses because speaker-fixed effects correlate with year-fixed effects.
65

TABLE 4 – Descriptive Statistics and Pearson Correlation Statistics
Mean SD 1 2 3 4 5 6 7 8 9 10 11 12
1. Market uncertainty -0.005068 0.06
2. ASR 0.23 0.22 0.09
3. Speech positive tone 3.42 1.01 -0.05 -0.08
4. Audience fear 0.0084 0.00 0.05 -0.22 -0.16
5. VIX raw 21.84 8.83 -0.01 0.05 0.05 0.19
6. Unemployment rate 0.0087 0.12 0.03 0.05 0.07 0.09 0.63
7. Inflation rate -0.0174 0.26 0.02 -0.11 -0.07 0.15 -0.29 -0.40
8. Expansionary monetary policy 0.16 0.37 0.04 0.01 0.05 0.15 0.35 0.36 -0.09
9. Contractionary monetary policy 0.18 0.39 0.14 -0.17 -0.12 0.17 -0.28 -0.18 0.17 -0.20
10. Dissent governor 0.04 0.19 0.00 -0.10 0.00 0.13 0.06 -0.08 0.10 -0.08 0.16
11. Dissent president 0.47 0.64 -0.17 0.16 0.18 -0.20 -0.01 -0.12 -0.06 0.13 -0.28 0.01
12. Speech location 0.29 0.46 0.07 -0.10 0.18 -0.02 0.02 0.08 0.01 0.02 -0.01 -0.05 -0.03
13. Weekday 3.50 1.50 -0.16 0.00 -0.05 -0.04 -0.01 0.05 -0.01 -0.02 -0.02 -0.11 0.03 -0.16
14. Consumer Price Index report 0.07 0.25 0.06 0.08 0.03 0.03 -0.03 -0.09 0.04 -0.02 -0.04 -0.05 0.02 0.06
15. Producer Price Index report 0.06 0.24 0.14 0.05 -0.10 0.01 0.01 0.06 0.00 0.06 0.04 0.02 -0.04 0.08
16. Unemployment report 0.05 0.23 -0.09 0.05 -0.02 -0.07 0.04 0.08 -0.03 0.01 0.06 -0.05 -0.05 -0.09
17. BOG speeches 0.23 0.42 -0.01 -0.03 -0.11 0.11 -0.01 -0.01 0.02 0.05 0.08 0.01 -0.11 -0.01
18. Testimony 0.08 0.28 0.05 -0.02 0.12 0.05 0.06 0.01 -0.04 -0.01 0.00 0.06 0.01 0.11
19. Press releases 0.51 0.50 -0.06 0.06 0.07 -0.02 0.07 0.07 0.01 0.04 -0.05 0.06 0.13 0.16
20. Speech word count 2667.45 1210.28 -0.06 0.23 -0.41 -0.01 0.06 0.09 -0.09 0.03 -0.08 -0.09 -0.03 -0.30
21. Speech uncertainty 1.24 0.68 -0.01 0.08 -0.05 0.03 -0.08 -0.09 0.04 -0.05 0.05 0.02 0.02 -0.03
22. Speech complexity 15.95 1.80 0.05 0.42 -0.27 -0.11 0.03 0.01 -0.07 0.00 -0.06 -0.05 0.09 -0.04
23. Speech abstractness 10.18 1.62 -0.16 0.03 0.04 0.02 -0.07 -0.08 -0.04 -0.08 -0.04 -0.04 -0.02 -0.12
24. Speech vagueness 1.12 0.37 -0.06 -0.08 -0.14 -0.08 -0.08 -0.07 0.00 0.05 -0.06 -0.05 0.04 -0.15
25. Speech net emotion 5.22 1.13 0.02 0.13 0.70 -0.13 0.09 0.11 -0.06 0.08 -0.10 0.01 0.11 0.03

66

Mean SD 13 14 15 16 17 18 19 20 21 22 23 24
14. Consumer Price Index report 0.07 0.25 0.03
15. Producer Price Index report 0.06 0.24 0.01 -0.07
16. Unemployment report 0.05 0.23 0.22 -0.06 -0.06
17. BOG speeches 0.23 0.42 0.10 0.02 0.01 -0.07
18. Testimony 0.08 0.28 -0.05 -0.04 0.10 -0.02 -0.06
19. Press releases 0.51 0.50 -0.23 0.03 0.03 -0.08 -0.03 0.06
20. Speech word count 2667.45 1210.28 0.02 0.01 0.03 0.00 0.00 -0.11 -0.03
21. Speech uncertainty 1.24 0.68 0.14 0.12 0.05 -0.08 0.07 0.05 -0.04 0.11
22. Speech complexity 15.95 1.80 0.01 0.04 -0.02 0.00 -0.08 0.02 -0.05 0.36 -0.01
23. Speech abstractness 10.18 1.62 0.05 -0.05 0.07 -0.01 -0.01 -0.03 -0.03 0.09 0.12 -0.06
24. Speech vagueness 1.12 0.37 -0.02 0.03 -0.04 -0.06 -0.07 -0.06 0.02 0.28 0.02 -0.05 0.08
25. Speech net emotion 5.22 1.13 -0.06 0.10 -0.01 -0.04 -0.05 0.03 0.03 -0.07 0.01 -0.14 0.09 -0.04
Note: 339 speeches from 1/1/1998 - 12/31/2014. Correlations above absolute value of 0.11 (0.14) are significant at 0.05 (0.01) for two-tailed test.

67

I also find that the mean for ASR is 0.23, as seen on Table 4, which indicates that
speeches tend not to expose the backing most of the time (since the ASR ranges from 0 to 1).
This is reasonable given most of our everyday language does not discuss directly our
fundamental taken-for-granted assumptions. Greenspan’s ASR (0.16) did however differ from
Bernanke’s (0.29) and Yellen’s (0.27). Figure 2, which recall charts the average ASR for each
year during the sample period, suggests that major external influences like the Financial Crisis of
2008 may have encouraged Bernanke and Yellen to discuss the assumptions underlying
monetary policy more often. Yet this figure also demonstrates a degree of construct validity for
the ASR because institutional theorists have suggested that discussion of institutional
assumptions may increase as the cognitive legitimacy of a given system declines (Green, Li, and
Nohria, 2009; Harmon, Green, and Goodnight, 2015), which is what occurred during the
Financial Crisis of 2008.
Table 5 reports the results of the OLS regression models. Model 1 represents the baseline
model, which includes only control variables. Examining patterns within these variables
validates several expectations we should have about trends in the data. For instance, consistent
with the financial economics view of communication (Van Buskirk, 2012), I find that longer
speeches reduce market uncertainty. Speeches given earlier in the week also produce more
uncertainty than speeches given later in the week, a finding consistent with prior finance research
(Connolly, Stivers, and Sun, 2005). In line with the anticipated aims of engaging in Fedspeak (60
minutes, 2007), more abstract speeches reduce market uncertainty. Furthermore, signals related
to potential disruptions in the economy—time periods of contractionary monetary policy—also
relate to increases in market uncertainty as one might expect. Interestingly, increased dissent
from regional presidents in the last FOMC meeting leads to reductions of uncertainty on Fed
68

speech days. This may be because dissent from the regional president creates uncertainty about
the specific situation inside the Fed that may dissipate once the Chairperson gives a speech and
reassures the market.

TABLE 5 – Regression Models Predicting Market Uncertainty (t
-1
to t
0
)

Model 1 Model 2 Model 3 Model 4 Model 5

ASR 0.040*** 0.030** 0.046*** 0.035**
(0.017) (0.017) (0.018) (0.018)
Speech positive tone -0.012** -0.011**
(0.006) (0.006)
ASR * Speech positive tone -0.044*** -0.040**
(0.018) (0.018)
Audience fear -2.247 -2.829
(2.776) (2.697)
ASR * Audience fear 28.448*** 23.698**
(11.356) (10.874)
VIX raw 0.000 0.000 0.000 0.000 0.001
(0.001) (0.001) (0.001) (0.001) (0.001)
Unemployment rate 0.021 0.024 0.019 0.025 0.020
(0.059) (0.059) (0.060) (0.059) (0.060)
Inflation rate 0.013 0.016 0.015 0.016 0.016
(0.015) (0.015) (0.016) (0.015) (0.016)
Expansionary monetary policy -0.000 0.002 0.003 -0.002 -0.001
(0.021) (0.021) (0.021) (0.021) (0.021)
Contractionary monetary policy 0.023* 0.025* 0.025* 0.025* 0.025**
(0.013) (0.013) (0.013) (0.013) (0.013)
Dissent governor 0.008 0.009 0.007 0.014 0.011
(0.019) (0.019) (0.019) (0.019) (0.018)
Dissent president -0.021*** -0.021*** -0.019*** -0.018*** -0.017**
(0.007) (0.007) (0.007) (0.007) (0.007)
Speech location -0.003 -0.001 0.000 -0.002 -0.001
(0.007) (0.007) (0.007) (0.007) (0.008)
Weekday -0.006** -0.005** -0.006** -0.005** -0.005**
(0.002) (0.002) (0.002) (0.002) (0.002)
Consumer Price Index report 0.017 0.014 0.015 0.018 0.018
(0.016) (0.016) (0.015) (0.015) (0.014)
Producer Price Index report 0.030* 0.027 0.025 0.026 0.024
(0.018) (0.018) (0.017) (0.017) (0.017)
Unemployment report -0.016 -0.019 -0.015 -0.018 -0.015
69

(0.015) (0.014) (0.014) (0.014) (0.014)
BOG speeches -0.001 -0.002 -0.003 -0.000 -0.002
(0.007) (0.007) (0.007) (0.007) (0.007)
Testimony 0.005 0.006 0.013 0.012 0.017
(0.011) (0.012) (0.011) (0.011) (0.011)
Press releases -0.010 -0.010 -0.010 -0.008 -0.008
(0.007) (0.007) (0.007) (0.007) (0.007)
Speech word count -0.000** -0.000** -0.000*** -0.000*** -0.000***
(0.000) (0.000) (0.000) (0.000) (0.000)
Speech uncertainty 0.002 0.001 0.000 -0.000 -0.001
(0.004) (0.004) (0.004) (0.004) (0.004)
Speech complexity 0.002 0.001 0.000 0.001 0.001
(0.002) (0.002) (0.002) (0.002) (0.002)
Speech abstraction -0.005*** -0.006*** -0.006*** -0.006*** -0.006***
(0.002) (0.002) (0.002) (0.002) (0.002)
Speech vagueness -0.002 -0.000 -0.002 -0.002 -0.003
(0.009) (0.009) (0.009) (0.009) (0.009)
Speech net emotion 0.000 -0.000 0.005 -0.001 0.004
(0.003) (0.003) (0.005) (0.003) (0.005)
Constant 0.036 0.055 0.047 0.057 0.041
(0.049) (0.049) (0.050) (0.050) (0.050)

Observations 339 339 339 339 339
R-squared 0.244 0.256 0.279 0.277 0.296
Adjusted R-squared 0.127 0.139 0.160 0.158 0.173
df 45 46 48 48 50
Average model VIF 2.38 2.38 2.48 2.40 2.50
Note: Results show robust regressions with robust standard errors in parentheses. Significance tests
are one-tailed for directional hypotheses, two-tailed for control variables. All models include year-
and topic-fixed effects.
*** p<0.01, ** p<0.05, * p<0.1

Model 2 adds the focal independent variable to test Hypothesis 1, which predicted that
the ASR of Fed Chairperson speeches will be positively associated with market uncertainty.
Consistent with this hypothesis, I find support across all models for the idea that the more the
Chairperson exposes the institution’s fundamental assumptions to direct examination, the more
market participants become uncertain (see Figure 4). What this means is that a one standard
deviation increase in ASR (i.e., 0.22) will increase market uncertainty (i.e., the VIX Index) on
the day of the speech by 0.90 percent.
70

FIGURE 4 – Main Effect of ASR on Market Uncertainty

Model 3 adds the first moderating variable and interaction term to test Hypothesis 2,
which predicted that the positive tone of the Fed Chairperson’s speech will weaken the positive
effect of ASR on market uncertainty. I find that the positive tone of the speech had an
independent negative effect on market uncertainty, suggesting that more positive speeches tend
to decrease market uncertainty, a finding that we would expect based upon prior literature. I also
find a significant interaction (see Figure 5), which indicates that as the positive tone of the
speech increases, the positive effect of ASR on market uncertainty decreases, thus supporting
Hypothesis 2. In addition, the economic significance of the ASR’s impact increases when there is
a low amount of positive tone in a speech. Specifically, when a speech’s positive tone is one
71

standard deviation below its mean, a one standard deviation increase in ASR (i.e., 0.22) will
increase market uncertainty (i.e., the VIX Index) on the day of the speech by 1.72 percent.

FIGURE 5 – Interaction between ASR and Speech Positive Tone

Model 4 adds the second moderating variable and interaction term to test Hypothesis 3,
which predicted that low levels of fear prior to the Fed Chairperson’s speech will weaken the
positive effect of ASR on market uncertainty. I find that this hypothesis is also supported by
evidence of a significant interaction (see Figure 6). This suggests that when the audience’s fear
prior to the speech is high, the positive association between ASR and market uncertainty holds.
However, when the audience’s fear is low, this relationship is mitigated, providing support for
Hypothesis 3. Moreover, the economic significance of the ASR’s impact also increases when
there is a high degree of audience fear before a speech. Specifically, when the audience’s fear is
72

one standard deviation above its mean, a one standard deviation increase in ASR (i.e., 0.22) will
increase market uncertainty (i.e., the VIX Index) on the day of the speech by 2.03 percent.

FIGURE 6 – Interaction between ASR and Audience Fear

Endogeneity Considerations
There are two general endogeneity concerns with regard to this research design. The first
relates to reverse causality, where changes in market uncertainty on the day of the speech
influence the ASR of the Fed speech given that day. Conceptually, this concern faces a temporal
disconnect between when the speeches are written and presented. Speeches are typically written
several weeks in advance (Bernanke, 2015) and are not released publicly until the time the
speech begins. As such, changes in market uncertainty on the day of a speech are unlikely to
influence the writing of that speech weeks beforehand. Empirically, my research design also
73

controlled for market uncertainty prior to each speech, providing additional comfort surrounding
this concern for reverse causality.
The second and perhaps more pertinent concern relates to the presence of a confounding
variable, where something simultaneously influences the ASR of Fed speeches and market
uncertainty in the exact manner as predicted. I address this concern conceptually, empirically,
and with a lab experiment. Even though I of course cannot completely remove this concern, my
hope is that these tests when considered together provide reasonable comfort regarding the
predicted and observed relationship between the ASR of Fed speeches and market uncertainty.
Conceptual tests. In order for the concern regarding a confounding variable to hold, we
should theoretically expect that market conditions that are likely to create uncertainty will also
lead the Fed to talk more about the backing. However, I find that the Fed actually exhibits the
opposite tendency. As Table 4 indicates, the economy being in a time of contractionary monetary
policy relates to increases in market uncertainty (i.e., positively correlates at 0.14) but to
decreases in the ASR of Fed speeches (i.e., negatively correlates at 0.17). In other words, the Fed
actually seems to avoid talking about the backing under the precise conditions that are most
likely to produce the most uncertainty. This is further validated by the fact that ASR increases
when fear is diminishing (i.e., negatively correlates at 0.22). The tendency to behave in this way
is suggestive of the fact that the ASR could be a strategic choice on the Fed’s part, since it
somewhat makes sense to stop talking about sensitive assumptions if the market is already
anxious. This observation however does make the existence of a confounding variable less
likely.
Furthermore, if the ASR of Fed speeches is just a reflection of the existing conditions
present in the market rather than a strategic choice, then we should expect the ASRs of speeches
74

given temporally near one another to be similar. To examine this expectation, I took all 339
speeches and calculated the number of calendar days between each of them. I selected speech
pairs that were within three days of one another. There were 55 pairs meeting this condition, for
a total of 110 speeches. When comparing these proximal speeches, I found that the average
difference in ASR between each pair was 0.21 (or roughly one standard deviation), and that the
second speech in 25 of the 55 pairs had a higher ASR. These observations violate the expectation
that market conditions are the primary determinant of the ASR. To further validate this point, I
reproduced support for Hypothesis 1 (B = .067, p = .032) despite the fact that the sample was
only 110 speeches. This provides additional support that the ASR of Fed speeches is indeed
influencing market uncertainty independently of existing market conditions.
Empirical tests. There are also several standard empirical methods that further mitigate
concerns related to a confounding variable. As noted earlier, I controlled for all major variables
in organization theory research on strategic communication, in financial economics research on
the Federal Reserve and the VIX. Furthermore, even if one is unable to successfully identify,
measure, and thus control for a potentially confounding variable, scholars have developed an
empirical method to calculate how large this hypothetical confounding variable would have to be
to invalidate the inferences made from his or her predictions (Frank, 2000). Following this
approach, I first calculated the Impact Threshold for a Confound Variable (ITCV), which is the
statistical size a confounding variable would have to be to invalidate the inference made in
Hypothesis 1 by diminishing its effect to be below significance. In my study, a confounding
variable would have to be positively correlated with both the ASR and market uncertainty at
0.18, making my ITCV equal to .033. Next, I compare the ITCV with the impact of all other
covariates in my model (see Figure 7). When doing so, I find that the size of the confounding
75

variable needed to invalidate my results is between 3 – 10 times larger than every other standard
control variable used in existing literature. This provides some additional comfort that
identifying a variable this large, which would invalidate my inference, is somewhat less likely.

FIGURE 7 – Reference Distribution of Impact for Covariates

Lab experiment. I also conducted a controlled lab experiment that executed as close of
the “ideal experiment” as possible, whereby I randomly assigned the exposure of backing to one
condition and not the others. Since gaining access to the same audience as in the above analysis
(i.e., option traders) is unlikely and unrealistic, I decided to conduct this test with everyday
working adults. Using this audience however required that I change the argument that
participants read from Fed speeches, which requires a reasonable level of economics and
monetary policy familiarity, to a simple business-related statement about the profitability when
76

acquiring a company, which requires very little specialized knowledge. The clear downside to
this design is that it does not test the same Fed speeches or same audience’s reaction. However,
in the context of the other analyses conducted here, the benefit of this experiment is that it tightly
controls for alternative influences that could be at play in a non-experimental setting and
importantly demonstrates a causal relationship between my independent and dependent variable.
I paid 120 Mechanical Turk workers $1 each to participate in an online study. These
participants were 50 percent female, with an average age of 37 (SD = 10.98) and 14 years (SD =
10.93) of full-time work experience. Participants were randomly assigned into one of three
conditions, each of which read a simple statement from a company about a recent acquisition
decision and then answered the following question: “How uncertain are you that the acquisition
of Company Alpha is a sound business decision?” Answers could range from 0 (not at all
uncertain) to 100 (completely uncertain). The first condition was the baseline condition and
contained a claim and one piece of data (note that the Toulmin labels were not seen by the
participants): “We are acquiring Company Alpha (claim). This decision improves our financial
performance (data).” The second condition contained the same claim and data, but also added a
second piece of data: “We are acquiring Company Alpha (claim). This decision improves our
financial performance (data). This will also allow us to reinvest our excess cash to benefit our
shareholders directly (data).” Finally, the third condition contained the original claim and data,
but instead added a direct statement about the backing: “We are acquiring Company Alpha
(claim). This decision improves our financial performance (data). The basis for this decision is
grounded in the assumption that profitability is what matters here (backing).” If the financial
economics view is correct, then the second and third conditions should both reduce uncertainty
because they contain more information than the baseline condition. However, if the theory I
77

developed here is correct, then the second condition should reduce uncertainty but the third
condition should increase it.
The statistical test of interest here is not a comparison between condition means, but a
comparison of their variances. While the mean in each of the three conditions represents the
average level of individual level uncertainty, the variance in each condition represents the
aggregate level of uncertainty of the audience as a representative collective (which is what the
VIX Index is assessing). The appropriate statistical test to examine differences in variances is
Levene’s test (Levene, 1960). I control for gender, age, and number of years worked. As
expected, the second condition (M = 32.18, σ
2
= 413.74) had a significantly lower variance than
the first baseline condition (M = 43.47, σ
2
= 734.41), F(1, 79) = 4.77, p = .032. This comparison
produced a significant difference in the means as well (F(1, 79) = 4.626, p = .035), suggesting
that more data reduces the overall aggregate level of uncertainty as well as the average level of
an individual’s uncertainty. This supports the financial economics view. However, also as
expected, the third condition had a significantly higher variance (M = 44.00, σ
2
= 1175.90) than
the first baseline condition, F(1, 79) = 6.22, p = .015. This provides causal evidence from a
tightly controlled experiment that exposing the backing can produce a greater variance across
participants’ reactions.
Robustness Checks
I conducted a number of robustness checks based on concerns that may arise given the
time period of my data or the research design itself. For instance, since my data contains one of
the largest financial crises of the last century, one might wonder if my findings are primarily
driven by this unique time period rather than being a more stable effect. Including year fixed-
effects should empirically control for this temporal variation in my dependent variable.
78

Nevertheless, to explore this even further, I first removed the year-dummy variables from my
regressions and replaced them with a financial crisis period dummy. I explored period dummies
of two years (i.e., 2007 – 2008), three years (i.e., 2007 – 2009), four years (2007 – 2010), and
from 2007 on. In all cases, all three of my hypotheses were again strongly supported. Moreover,
none of these dummy variables interacted with my ASR variable to predict market uncertainty.
Furthermore, if I also add to these financial crisis period dummies a year-continuous dummy
variable (i.e., 1998 = 1, 1999 = 2, etc.) to capture the trending effects of the economy getting
worse over time, I again find consistent results. This indicates that my results are not driven by
the financial crisis period and instead appear to be a stable mechanism across these years.
Yet other factors during the financial crisis may be influencing my results. Specifically,
starting in late 2008 the Fed started to engage in unconventional monetary policy with something
called quantitative easing. This represented a relatively unusual practice whereby the Fed
purchased securities directly from the market in order to lower interest rates and increase the
money supply. These purchases in turn were reflected on the Fed’s balance sheet as assets,
increasing nearly 500 percent between 2008 and 2014, from about $900 billion to over $4.5
trillion. It is fair to say that quantitative easing has thus influenced how the Fed views the
boundaries of their own monetary policy framework and has also had a large impact on the
broader market (Yang and Zhou, 2016). I tried to control for quantitative easing by taking the
average total assets on the Fed’s balance sheet for the month prior to each speech, but this
produces issues of multicollinearity and thus I left it out of my primary analyses. Nevertheless,
whether I include this variable with or without the year fixed-effects, I still find consistent
support for all my predictions.
79

One might also question whether it matters how the Chairperson talks about the backing
for this effect to hold. For example, while most of the time the Chairperson makes reflections
about the existing backing in a very descriptive manner, at times he or she talks about changing
the backing or what the backing may look like in the future. This sort of change- and future-
oriented language is potentially different from descriptive statements about the existing system
and may be what is primarily driving my effects. To examine this possibility, myself and one
research assistant re-coded every single paragraph that exposed the backing to determine if it
referred in any way to change or if it was future-oriented. Of all the paragraphs that exposed the
backing, 9.3 percent related to change and 10.3 percent were future-oriented. I then calculated
the percentage of backing-related paragraphs in each speech that referred to these two factors.
Controlling for both of these considerations again produced consistent results, suggesting that it
does not matter how the Chairperson talks about the backing per se, just that he or she exposes it.
One might also propose that just talking about the backing itself is unexpected and, thus,
it is the unexpectedness and not the ASR of the speech that is driving my results. To explore this
potential alternative mechanism, I created an ASRChange variable by subtracting a speech’s
ASR by the chronologically previous speech’s ASR. If it is really the unexpectedness of talking
about one’s assumptions that is driving my results, then this newly created variable should
predict market uncertainty, since a greater change in ASR from the last speech should be
increasingly unexpected. However, running Model 2 with this new predictor variable produces a
nonsignificant result (p = .291). I also re-ran Model 2 with this new variable as a control instead
and it does not change my original results. This provides some evidence that it is not just the
unexpectedness of talking about the backing that is driving my results, but again just the
exposure of the backing itself.
80

Finally, I also validated that my results were robust in a number of additional ways. In
particular, I would expect that my results should not hold for the two-day event window before
the day of the speech (t-2 to t-1) since the speech is not made public until the date and time
denoted on the transcript. Consistent with this expectation, ASR did not predict market
uncertainty (B = .01, p = .631) for this time period. Similarly, consistent with existing research
that demonstrates that the influence of new information on the VIX Index does not typically last
longer than one day (Nikkinen and Sahlström, 2004), I found that that the effect of the ASR
dissipates in the two-day event window after the day of the speech (t0 to t+1) (B = -.01, p = .636).
Lastly, I examined alternative specifications for a number of control variables to ensure that my
results were robust to these changes. I tried using the actual federal funds rate as well as the
change in this rate instead of dummy variables for expansionary and contractionary monetary
policy, a two-day (t-2 to t-1) and four-day (t-4 to t-1) window for my audience fear variable instead
of a three-day window, and several different windows (e.g., 15-day and 180-day) for calculating
the average for VIX raw instead of a 30-day average. My findings remained consistent across
these specifications.
DISCUSSION AND CONCLUSIONS
This study set out to examine how the market responds to communication from
prominent actors when they expose an institution’s assumptions to direct examination. Using all
public speeches made by the Chairperson of the United States Federal Reserve from 1998 to
2014, I demonstrated that the more they make explicit the assumptions underlying the Federal
Reserve System, the more their speeches produced uncertainty in the market. Moreover, since
the Fed generally wants to avoid increasing uncertainty but at times must discuss these
assumptions, I theorized how they might exploit the way emotion works in markets to achieve
81

both of these objectives simultaneously. Specifically, I showed how increasing the emotional
positivity of their speeches and discussing these assumptions when there are low levels of fear in
the market are two ways to avoid these undesirable market effects. I suggest that this study
contributes a new perspective on the role of strategic communication in market contexts.
In particular, this work builds on but also substantially extends the social constructionist
view of communication in markets. This still growing body of research tends to focus on
identifying different communication strategies that operate within a set of institutionalized rules
and assumptions that are taken as givens (Fiss and Zajac, 2006; Martens, Jennings, and Jennings,
2007; Lamin and Zaheer, 2012; Rhee and Fiss, 2014). It is true that these rules and assumptions
underlying our social institutions are often left implicit and taken for granted in everyday life.
But this of course is not always the case. What happens when prominent actors in that
community start to expose these assumptions to public scrutiny? This is not a question scholars
in this space have yet asked, perhaps in part because we academics also take certain assumptions
as givens (e.g., like how strategic communication supposedly works). By pulling back our
collective vantage point, this study seeks to expand our understanding of the possibilities for how
communication interacts with our socially constructed institutions.
This study also provides an important qualification to the financial economics view of
communication. This still dominant perspective contends that more communication should
reduce information asymmetries between parties (Akerlof, 1970; Spence, 1973) and therefore
reduce the overall level of uncertainty in the market (Van Buskirk, 2012). Although reversing
this finding has seemed unlikely, there have been ways to mitigate this effect. For instance,
communicating too much can produce a tapered-off effect because people cannot process all that
new information (Graffin, Carpenter, and Boivie, 2011), and more communication about
82

something people already know may have no effect at all (Tetlock, 2007). However, this study
points to a fundamental distinction in the communication itself—its structural level—that
challenges this prevailing finding. Specifically, I show that more communication can actually
increase uncertainty if it exposes the assumptions underlying the existing institution to direct
examination. By pulling apart these two structurally distinct levels of communication, this study
provides an important caveat to a dominant assumption for how information operates in markets.
The arguments made in this paper also further our understanding of how emotions
operate in financial markets. While scholars have become increasingly interested in the role of
emotion in these contexts (Tetlock, 2007; Pfarrer, Pollock, and Rindova, 2010; Rhee and Fiss,
2014), existing work has primarily emphasized how emotions amplify market reactions
(Barberis, Shleifer, and Vishny, 1998; Baker and Wurgler, 2006). This study demonstrates how
actors might leverage emotion to not only amplify but also suppress these reactions. In particular,
while low amounts of positive tone and high levels of audience fear amplify market reactions
when discussing the backing, I also find that increasing positive tone and talking when there are
low levels of fear are ways to suppress this same reaction. By better understanding emotion—
both in one’s own communication and in the market itself—this study points to several ways
actors might better control how an audience is likely to react.
More broadly, this work points to an important market-level outcome—uncertainty—that
organizations may not always consider directly but could be important to success. Indeed,
organizations tend to focus on the direction of their audience’s evaluations, or their stock price,
and how their actions or communications influence the change in this level of favorability
towards their activities. Perhaps unsurprisingly, management theorists also tend to focus on stock
price as their primary dependent variable when examining the symbolic impact of organizational
83

activities on the market (Fiss and Zajac, 2006; Martens, Jennings, and Jennings, 2007; Rhee and
Fiss, 2014). However, organizations and scholars are also interested in an audience’s aggregate
level of uncertainty (Sanders and Boivie, 2004), which reflects the degree of social consensus
underlying the audience’s evaluation. By exploring the market uncertainty as an outcome, this
study seeks to renew our theoretical and practical interest in the importance of social consensus
underlying our institutions (Mead, 1962; also see Bitektine & Haack, 2015) and how
organizations might directly impact their audience’s overall level of agreement in their
evaluations of their activities.
Implications for Institutional Theory
This study has several important implications for institutional theory. In particular, this
work contributes to how we theorize and empirically examine the microfoundations of
institutions. Institutional theorists have conceptualized the micro-level plumbing of institutions
in a variety of different ways (Powell and Colyvas, 2008) by using frameworks based on emotion
(Voronov and Vince, 2012; Creed et al., 2014), social psychological evaluations (Bitektine,
2011; Tost, 2011), interaction rituals (Gray, Purdy, and Ansari, 2015), and practices and
performances (Smets, Morris, and Greenwood, 2012; Lok and De Rond, 2013). Several scholars
have argued that communication may also be a promising approach to understand the
microfoundations of institutions. For instance, Loewenstein, Ocasio, and Jones (2012) suggest
that vocabularies—or a related collections of words—may be one possible way to understand the
how meaning coheres within our institutions (e.g., Suddaby and Greenwood, 2005; Dunn and
Jones, 2010).
I propose an alternative approach that conceptualizes arguments as the micro-level
plumbing that maintains and changes our institutions. While vocabularies reflect the words
84

actors typically use in different institutionalized contexts, arguments inject agency into how
actors use communication strategically to stabilize or destabilize institutions. In this sense, these
approaches may actually be quite complementary in that arguments are precisely how actors use
vocabularies to accomplish their goals. However, unlike the vocabularies approach, an argument
approach importantly mirrors the structural depth and taken-for-granted nature inherent in our
nested institutions (Friedland and Alford, 1991; Thornton, Ocasio, and Lounsbury, 2012).
Specifically, argument structure reflects the underlying implicit meaning structure of our
institutions, enabling scholars to actually observe whether or not actors make explicit these
assumptions or not. This study thus provides a conceptual as well as empirical approach to link
the taken-for-granted components of our institutions with the communicative aspects of everyday
symbolic action.
This in turn may generate important insights for our understanding of cognitive
legitimacy. While substantive forms of legitimacy (e.g., pragmatic and moral) are relatively
simple to observe and measure because they reflect the presence of people’s evaluation about an
organization or action (Suchman, 1995; Bitektine, 2011; Tost, 2011), cognitive legitimacy
reflects how much institutional actors take for granted the assumptions and thus actually
represents the increasing absence of an evaluation itself (Green, Li, and Nohria, 2009). Indeed,
things that are entirely cognitively legitimate are not questioned, considered, or thought of,
representing a uniquely powerful form of legitimacy (Suchman, 1995). How then are we as
scholars supposed to study or examine changes in cognitive legitimacy directly? This study
proposes that the ASR is a conceptual and empirically observable component in our strategic
communication that has a direct relationship with people’s cognitive legitimacy and, by
extension, the stability of our institutions. For instance, low ASR communication strengthens the
85

presumed cognitive legitimacy, reinforcing the taken-for-grantedness of the backing (Aldrich
and Fiol, 1994: 652), and further stabilizing the very foundations of the institutions (Holm, 1995;
Tost, 2011). In contrast, high ASR communication disrupts cognitive legitimacy because it
makes explicit the very assumptions that were previously taken-for-granted, triggering mental
alarms (Tost, 2011) or existential crises (Voronov and Vince, 2012), and increasing the risk for
changes to the foundation of the institution (Holm, 1995; Suddaby and Greenwood, 2005).
Implications for the Sociology of Financial Markets
This study also makes an important contribution to a new but fast-growing literature on
the sociology of financial markets (Lounsbury and Hirsch, 2010; Carruthers and Kim, 2011;
Knorr Cetina and Preda, 2012). This body of work argues that financial economists failed to
understand and prevent the recent collapse of the global financial system and looks to
sociological insights to improve traditional theories of market behavior. One of the primary
insights coming out of this work is that market behavior does not function the way we might
think because prominent actors that oversee our system (e.g., top political officials, the Federal
Reserve, regulatory committees, etc.) have become conditioned by the historically situated
assumptions underlying our institutions (Rubtsova et al., 2010). Fligstein, Brundage, and Schultz
(2014) adopt this line of thinking when they examined transcripts from the Fed’s FOMC
meetings, arguing that the Fed’s “main analytic framework for making sense of the economy,
macroeconomic theory, made it difficult for them to connect the disparate events that comprised
the financial crisis into a coherent whole” (p. 2). Abolafia (2012) echoes this same thinking by
arguing this historically conditioned thinking can translate into Fed communication being a
source of technocratic control. Drawing on sociological notions of how discourse can be a
conduit for hegemony (Bakhtin, 1982; Steinberg, 1999), Abolafia elicits the following warning:
86

The danger is that proficient masters of spin become so confident in their technical discourse that the
restraints of uncertainty and legitimacy are no longer sufficient to encourage prudent questioning of the
current operating models. The result is that discourse dominates the dialectic of technical rationality and the
pragmatism of expert judgment is neglected…The threat is that the domination of any technical discourse
inhibits the power of judgment to the point where the discourse is performing the policy rather the experts
operating a discourse (Abolafia, 2012: 112).

My findings suggest a potential problem with this overall line of thinking. In particular, I
find that the Fed not only recognizes but also discusses the nature and boundaries of their
framework or operating model (i.e., the backing) directly. It seems reasonable to expect that if
the Fed was blinded entirely by their own assumptions as Fligstein and Abolafia suggest, then
the Fed would not be discussing their backing at all and we should not observe any variation in
the ASR. My finding may in part be due to the fact that I examined the Chairperson’s personally-
developed speeches rather than the FOMC meeting transcripts. Indeed, FOMC meetings are held
primarily to discuss the operations and implementation of monetary policy and, thus,
unsurprisingly tend to contain mostly conversations within the rules of the game (i.e., little direct
discussion of the backing). In contrast, speeches are occasions where the Chairperson reflects
more broadly about his or her beliefs regarding monetary policy operations and about the
framework itself. My empirical results thus suggest that the Fed may have been more aware of
their own framework’s limitations than prior work has assumed (see Bernanke, 2015).
Nevertheless, despite the fact that Greenspan and others’ have openly acknowledged flaws in
their monetary policy framework (PBS News Hour, 2008), being aware of these limitations is
different than fixing them.
With that being said, my findings also point to the possibility that the Fed, whether
intentionally or unintentionally, could still be using the ASR of their communication as a tool for
market control. While my results primarily emphasize the effects of opening up the Fed’s
87

assumptions to direct examination (i.e., high ASR speeches), not exposing the backing (i.e., low
ASR speeches) may be thought of as an equally strategic decision to keep markets from raising
questions, particularly in times of turmoil. Consistent with the themes in Abolafia’s work, I
argued that low ASR speeches imply the legitimacy of and stability within the Federal Reserve
System and, by extension, the broader market economy. Importantly, this ignores the fact that the
legitimacy and stability of this very system perhaps should have been questioned but that this
questioning has been disallowed. The theoretical framework proposed here thus highlights that
while prominent market actors’ thinking may not be as conditioned by their own assumptions as
we might have thought, their strategic use of argument structure may still have controlled the
prevailing market discourse.
Future Research Directions
This study also generates a number of opportunities for future research. In particular,
while this study focused on the VIX Index as the market’s reaction, sociological approaches to
financial markets have convincingly argued that the architecture of markets is not universalistic
but instead socially constructed and culturally contingent (Fligstein, 2001; Fligstein and Calder,
2001; Lounsbury and Hirsch, 2010). Thus, examining only the VIX Index prioritizes a highly
sophisticated audience that is embedded within the capital markets and overlooks other
potentially important audiences. Although using the VIX Index in this case was a reasonable
decision given my stated theoretical aims and that it is widely accepted as the premier barometer
for measuring market uncertainty, it would be useful and important to also examine how the
ASR impacts other market audiences. For instance, examining how bond or foreign currency
markets respond to the ASR of Fed speeches may reveal deeper concerns underlying the stability
of interest rates. Exploring how audiences outside the capital markets may be interesting as well,
88

such as how regulatory authorities alter their policy decisions or how the media amplifies or
spins the Fed’s original message. Such considerations could shed light on potential cultural
contingencies within markets and demonstrate how the same message might be refracted in
different ways.
Future research might also considering exploring the ASR of strategic communication
from a wider variety of actors. The primary reason for emphasizing prominent actors in this
study was that the speaker likely needs to be an authority within the social institution in which
they are talking for people to listen or take their discussion of the backing seriously. Indeed, we
will typically ignore the person standing on the corner shouting about the end of the world
through a megaphone, but we may listen to that same message if conveyed to us by our priest. It
is possible however that non-authorities could garner the same amount of attention in certain
public forums. For example, some have argued that social media has democratized the notion of
authority, enabling just about anyone to voice their opinions and create a platform upon which
the world will listen (Edgerly et al., 2009). Moreover, minority actors can sometimes force
authority figures to address or at least recognize their concerns publicly. This dynamic, which is
a basic assumption underlying the social movements literature (King and Soule, 2007; King,
2008), occurs regularly with politicians, organizations, or sports icons who are constantly
scrutinized in the public light. Thus, while this study only examined prominent actors, there are a
number of other forces that likely play a role in this story and are certainly worthy of future
consideration.
This also raises the question about the extent to which these findings would extend to
other contexts. The Fed Chairperson is certainly in a unique position, being perhaps the most
influential speaker when it comes to communicating within the market (Cruikshank and Sicilia,
89

1999). This uniqueness was required because my dependent variable of interest was at the
market level uncertainty of the entire S&P 500. My contention is that these findings are likely to
generalize to situations where the speaker’s prominence and a community of actors interested in
what the speaker has to say are similarly commensurate. For instance, managers often attempt to
influence their employees’ opinions during organizational change efforts (Meyer, Brooks, and
Goes, 1990). Since uncertainty and social consensus of employees regarding the future direction
of their organization is critical in this situation (Rousseau and Tijoriwala, 1999), one might
consider looking at how the ASR of organizational change communication influences employee
morale. Corporate leaders similarly use strategic communication to manage the expectations of a
variety of external audiences. For instance, organizational founders are required to communicate
with prospective investors with their prospectuses when going public (Martens, Jennings, and
Jennings, 2007). Given the critical role uncertainty plays in a firms’ success when going through
an IPO (Sanders and Boivie, 2004; Pollock and Gulati, 2007; Pollock, Rindova, and Maggitti,
2008), one might predict that talking more about their business plan assumptions would diminish
their ability to raise capital. Moreover, CEOs and CFOs also communicate with analysts on a
regular basis (Lee, 2015), and the ASR of their quarterly earnings calls may influence the
divergence of analysts’ ratings after the meeting. Further research might seek to extend the
findings found in this study to a variety of other important organizational contexts.
One of the classic sociological observations made by Berger and Luckmann (1966: 64) is
that our “institutions are built upon language.” Since that time, a number of theorists have
referenced this theme (Friedland and Alford, 1991) and have tried to explain better what exactly
they meant (Suddaby and Greenwood, 2005; Loewenstein, Ocasio, and Jones, 2012). This study
seeks to provide additional clarity to Berger and Luckmann’s abstract assertion by pointing
90

concretely to our communicative assumptions—or what Toulmin calls the backing—as what
specifically our institutions are built upon. In this sense, I view the development of the ASR
construct as both a theoretical and empirical entry point into the further study of how language
operates within—and at the boundaries of—our social institutions.
91

CHAPTER 4 – CONCLUSION

CHAPTER 4

CONCLUSION

This dissertation advances a novel approach that I refer to as the structure of strategic
communication. Drawing on the idea that actors can deploy arguments at two structurally distinct
levels—within the rules of the game or about the rules of the game—I argue that talking more
about these rules, which exposes the assumptions underlying our social institutions to direct
examination, has profound implications. To show evidence for this claim, I first developed a
construct called the argument structure ratio (ASR) that measures how explicit a speaker makes
these assumptions in communication. I then theorized the impact of the ASR on the market,
demonstrating that the more the Fed Chairperson makes explicit their assumptions, the more
their communication produces uncertainty in the United States financial market. Taken together,
this dissertation presents a unified approach to begin studying the structure of strategic
communication.
I argue that this approach has the potential to revitalize a large body of research on
strategic communication that has grown increasingly stale. In particular, existing work in this
space seems to be somewhat stuck asking the same question and arriving at the same answer.
Specifically, scholars for the last 30 years or so have explored the question of how to use
communication strategically to receive a favorable reaction from an audience (Elsbach, 1994;
Erkama & Vaara, 2010; Green, 2004; Lamin & Zaheer, 2012; Marcus & Goodman, 1991;
Martens et al., 2007; Pollock & Rindova, 2003; Rhee & Fiss, 2014; Vaara, Tienari, & Laurila,
92

2006). This in turn led to the cataloguing of different strategies actors might use (e.g., apologies,
denials, justifications, framing, stories, etc.) and the testing of when different strategies worked
and when they did not. Across a variety of studies and contexts, the same recurring answer to
this question was that a strategy was most effective when it met the expectations of the target
audience (Fiss & Zajac, 2006). However, since there are countless ways in which this fit or
resonance with an audience might occur, researchers in this space seem to be stuck cataloguing
examples of the same mechanism with no end in sight.
Interestingly, these researchers might gain some insight by reflecting upon how social
psychologists studying persuasion handled an almost identical impasse over 30 years ago. Early
persuasion researchers during the 1950s and 1960s, like organizational scholars, drew upon
Aristotle and others to catalogue all the factors that influenced the persuasiveness of a message.
Their lists of potential influences grew so large and so fine-grained that findings from seemingly
identical persuasion studies started to contradict each other, suggesting a frightening conclusion
that the number of factors needed to be considered to predict how persuasion works might be
unrealistic. While this led some researchers to give up on this topic of study, others sought to
reformulate the way persuasion should be studied. Specifically, these scholars—like Petty and
Cacioppo (1986) among others—suggested that we should ignore the countless strategies of
persuasion and instead refocus our attention on the fact that these strategies can be processed in
two fundamental ways and that this difference in processing is what we should study. By shifting
the focus of study to a less contingent and more stable aspect of how communication works to
influence others, social psychologists were able to revive and reenergize the study of persuasion
for decades thereafter.
93

This dissertation aims to shift the focus of strategic communication in organizational
analysis in an analogous way. Like persuasion researchers 30 years ago, I am similarly
suggesting that we should place less of an emphasis on the countless strategies of persuasion,
since we already know that there are any number of things actors can say and the effectiveness of
such strategies will depend upon a variety of factors. Instead, I advocate that we focus on the fact
that these strategies can be communicated across two structurally distinct levels, which reflects
the degree to which actors make explicit their assumptions, and that this difference in the
explicitness of assumptions is what we should study. By shifting our focus from the specific
content of communication to its underlying argument structure, my aim is to reformulate the
way strategic communication might be studied in the future.
More specifically, I believe that a focus on argument structure could help to revitalize the
study of strategic communication in three significant ways. First, conceptualizing
communication as argument structure suggests an entirely new mechanism through which
strategic communication impacts audiences. Recall that the prevailing mechanism for how
communication works is based on the notion of fit, or what some refer to as resonance. That is,
the more communication fits or resonates with an audience’s expectations, the more effective it
is. The reason this mechanism works however is because these scholars are focused on the
content or strategies of communication. Specifically, if the content is something the audience can
relate to, then it is more appealing or attractive. This mechanism may even reflect a deeper
evolutionary mechanism based on the idea of homophily, which was hinted at by Aristotle
(1991) when he pointed out that similarity and conceptual closeness in rhetoric can enhance the
credibility of the speaker.
94

However, by shifting the focus from the content to argument structure, this dissertation
points instead to a mechanism related to an emotional reaction to the exposure of something that
should not be exposed. In particular, an audience reacts with increased uncertainty when the
assumptions underlying institutions are exposed not because doing so is unexpected but rather
because doing so highlights the contingencies of deeply held assumptions that we feel
uncomfortable considering directly. Talking about that which should not be talked about thus
produces existential anxiety that has to do with our fundamental understanding of the social
world now being at risk. This mechanism thus takes seriously the historical basis of our social
institutions. Cultural and institutional knowledge gets passed on from generation to generation,
and the standard ways of doing things in everyday situations goes unquestioned for decades
(Zucker, 1977). While existing theories of communication of course acknowledge these
historically embedded assumptions, the approach proposed here points directly at these typically
taken-for-granted assumptions and examines precisely what happens when they are unearthed.
Second, this new conceptualization of how strategic communication works naturally
opens up an opportunity for the development of novel independent variables that prior work has
overlooked. In particular, this dissertation seeks to make a substantial contribution by developing
an entirely a new methodological approach to measure variation in argument structure—what I
called the ASR. The ASR stands in stark contrast to existing communication strategies. For
instance, while existing strategies capture the content of what is said, the ASR captures the depth
to which the speaker goes to convey this point. Moreover, while existing strategies capture only
words or phrases as the aspects of communication influencing others, the ASR demonstrates how
the overall impression of an entire message can impact an audience. The ASR thus is unlike any
communication strategy existing scholars have considered. As I elaborate in greater detail in the
95

following section, I believe that the ASR provides a valid, reproducible, and powerful tool for
researchers to leverage across a variety of contexts and levels of analysis in future research.
Third, this new conceptualization also reveals different ways strategic communication
impacts audiences. Existing work has generally examined how communication strategies
produce more favorable reactions from audiences. This led to the aforementioned focus on how
communication content can produce shifts in people’s beliefs (e.g., changes in stock price,
evaluations of legitimacy, etc.). However, by changing our focus from communication content to
argument structure, this dissertation highlights how the ASR of communication impacts an
equally important outcome to audience favorability—audience uncertainty. Despite not being
studied consistently as an outcome in organizational analyses, the overall level of uncertainty or
social consensus of an audience is an important outcome for organizations. Indeed, uncertainty
impacts investing patterns, commitment to action, and the ability for organizations to raise
capital (Sanders & Boivie, 2004). Moreover, while organizations may often want to reduce
audience uncertainty, they also may seek to increase uncertainty so as to obfuscate sensitive
issues and distract investors (Graffin et al., 2011).
Taken together, the aim of this dissertation is to revitalize the study of strategic
communication by reformulating how communication influences others. While the theory,
measurement, and effects of the ASR discussed in this analysis provides one example of how to
explore these ideas, there may be other ways to extend these basic principles. In particular, I
believe that there are many ways to leverage the ASR as a theoretical and empirical construct to
explore interesting questions in a variety of other contexts. At a more general level, I argue that
this dissertation might also be thought of as a theoretical starting point to explore other types of
structure underlying strategic communication. I therefore close this dissertation with an
96

exploration of these considerations with the aim to encourage future research in this new and
exciting space.
LEVERAGING THE ARGUMENT STRUCTURE RATIO
ASR Further Theorizing
Researchers interested in leveraging the ASR to explore its impact in other contexts
might consider examining the boundary conditions of the theory proposed here. Indeed, my
primary argument was that strategic messages with higher ASRs point to the contingencies and
therefore potential instability of those very assumptions that undergird a given institutional
arrangement and, as a result, will increase uncertainty. However, this argument makes two
important assumptions. First, I assumed that the backing or assumptions in the research context
were relatively stable. Second, I assumed that the audience evaluating the ASR was deeply
embedded within the institution whose assumptions were being exposed. Further theorizing is
warranted to explore the impact of ASR when either of these assumptions are violated.
In particular, the backing underlying an institution is not always stable. For example,
consider the Internet IPO boom during the late 1990s. During this time, a number of companies
were going public and claiming an association with the Internet to ride the wave of investor
hype. However, the Internet at this point was still relatively new and the rules of the game so to
speak were still being formulated. The stability of the assumptions in this context stands in stark
contrast to the stability of the assumptions underlying Federal Reserve System, the latter of
which had been left unchanged for decades. When the backing is not yet stable and these
assumptions are still being formulated, exposing these assumptions may not increase uncertainty.
Indeed, the mechanism that I argued creates uncertainty—questioning that which should not be
questioned—does not exist when the backing is not yet stable. As a result, it is possible that
97

exposing assumptions that are still being formulated will have no effect or may even decrease
uncertainty. More work is needed to theorize this important boundary condition.
Scholars might also benefit from further theorization related to how audiences not deeply
embedded within the institution being questioned might react to the ASR. I argued here that
options traders react with uncertainty when the Fed exposes the assumptions underlying United
States monetary policymaking to direct examination. However, these option traders are highly
sophisticated and already embedded within the economics and central banking institutionalized
frameworks. Indeed, they already take for granted the objectives and tools used by the Fed. As a
result, when the Fed points to the assumptions underlying these considerations, the option traders
react as I predicted. The same may not be said for audience members that are not embedded in
the same way. For instance, your average everyday citizen likely does not have the same level of
sophistication or education regarding United States monetary policymaking. As a result, when
the Fed exposes these assumptions, this revealing moment may not be nearly as emotionally
jarring. This observation suggests that more theorization could be useful to better understand the
broader impact of the ASR.
ASR in Other Contexts
The ASR also presents an opportunity for scholars to explore potentially novel effects
within their research contexts of interest. In particular, I contend that the ASR might be used to
revisit some existing questions in the literature from an entirely new perspective. Specifically,
the large body of research on strategic communication discussed at length in this dissertation has
examined a number of important contexts where organizational actors use language to influence
others. Exploring the impact of the ASR in these contexts may elicit new insights. For instance,
entrepreneurs are required to communicate with prospective investors through a prospectus when
98

going public. While prior work has extensively explored how storytelling influences these firms’
ability to raise capital (Martens et al., 2007), researchers might also look at how the ASR of these
communications impact investor uncertainty as another important mechanism of IPO success
(Sanders & Boivie, 2004). Moreover, organizations also regularly communicate with a variety of
stakeholders through press releases. While researchers have shown that defending or
acknowledging their actions influences their legitimacy (Lamin & Zaheer, 2012) and that their
framing can portray their decisions in a more desirable light (Rhee & Fiss, 2014), future work
could examine how the ASR of these communications may create or reduce uncertainty
surrounding these very same reactions.
Scholars might also leverage the ASR to explore entirely new and exciting questions. In
particular, while this dissertation focused on the effects of the ASR on audiences’ reactions,
researchers might instead want to understand the conditions under which actors expose these
institutionalized assumptions. Indeed, this idea of voicing dissent, speaking up, or pointing to
things that are traditionally taken for granted has been a theme across several literatures. For
instance, researchers studying employee voice (Burris, 2011; Detert et al., 2013) are interested in
understanding when and why individual employees are likely to speak up and question the
assumptions underlying their organization. Moreover, institutional theorists studying legitimacy
and institutional change (Battilana & D’Aunno, 2009; Bitektine & Haack, 2015; Harmon et al.,
2015) are interested in similar questions regarding when and why actors recognize and seek to
change the very assumptions that they typically take for granted. Researchers might therefore
begin to explore the specific conditions under which the ASR of a community or discursive
space changes and actors begin to expose these assumptions directly.
99

This notion of focusing on the constraints on communication in turn points to a
potentially novel avenue of study. In particular, scholars across the social sciences have long
recognized the constraints on actors’ cognition, which has generated decades of fruitful research
on the concept of bounded rationality (Kahneman, 2011; Simon, 1982). I contend that the ASR
points researchers to a related but entirely distinct concept that I refer to as bounded
communicative rationality. That is, just as our cognition or thinking is constrained in predictable
ways, so too is our communication. I see the concept of bounded communicative rationality to be
just as powerful as bounded rationality, with applications at the individual level (e.g., when
dissent is voiced to friends or family), organizational level (e.g., when managerial authority is
challenged), institutional level (e.g., when change to institutions is initiated), and even the
political or societal level (e.g., when citizens speak out against oppressive or authoritarian
regimes). Furthermore, given the complicated relationship between our thoughts and words
(Schwarz, 1996), I also see exciting research to be done on how bounded rationality and bounded
communicative rationality relate and influence one another. In this way, I believe that the ASR is
capable of generating entirely new and exciting ways to think about the role of communication in
social life.
ASR Automated Tool
Leveraging the ASR in other contexts holds substantial promise. However, an obvious
limitation is that coding communication for ASR is resource-intensive. Indeed, while the
methodology outlined in Chapter 2 is valid and reproducible, it is admittedly not easy. Obtaining
an ASR for your desired communications requires extensive training, reliability coding, and the
time and resources to code the entire sample. One way to improve the efficiency of this
methodology as well as its potential impact is to automate its coding. I am currently in the
100

process of finalizing an automated tool for public use. However, I would like to highlight for
researchers interested in using the ASR how this tool might work in the near future.
The ASR automated tool is a dictionary approach that works by searching for words
associated with the backing in your corpus and providing a ratio of those words to the overall
words in each speech act. The researcher first must identify a small number of words (e.g.,
between 5 and 10) that refer to or are closely associated with the backing in their context. These
words are then automatically “trained” against a database containing millions of news articles
over the last decade or so. By training, I mean that the ASR automated tool is searching the
semantic space surrounding each of these words to create a larger, customized dictionary of
semantically similar words related to the backing. This produces a dictionary that is more closely
related to the backing that could be done manually. The researcher then uploads the corpus and
the ASR automated tool produces a number for each speech act that represents the ratio of
backing-related words to overall words. I piloted the use of this tool using my Fed Chairperson
speech corpus. By inputting seven backing-related words in the Fed context ["central",
"institutional", "policy", "system", "regulation", "risk", "supervisory"], the automated measure
correlated with my reliability-tested speeches at 0.80, and reproduced my primary prediction that
higher ASR creates more uncertainty (p < .001).
Keep in mind however that because the backing or assumptions are context-dependent,
complete automation is not possible. As a result, researchers will still be required to identify
small number of backing-related words that reflect the core assumptions present in the corpus.
What this means is that researchers should still follow Step 1 (Define the discursive space and
corpus) and Step 2 (Clarify the two structural levels of communication in your corpus) detailed
in Chapter 2 to identify these words. I would also highly recommend researchers perform part of
101

Step 3 (Code your corpus based on these two structural levels) by coding a small sample of the
corpus. This will help validate that the words inputted into the tool are indeed picking up
references to the backing. Nevertheless, even though the ASR cannot be fully automated, this
tool should enable researchers to code large corpuses without the substantial associated costs. By
developing this tool, I thus hope to broaden the impact of the ASR.
EXPANDING THE CONCEPT OF STRUCTURE
This dissertation advanced a novel approach that I referred to as the structure of strategic
communication. The type of structure I was of course referring to was argument structure, which
turns out to be one promising direction for research. However, because this is just one way to
conceptualize structure, this dissertation admittedly represents just a single case of a broader
direction I would like to see scholars studying communication take. It is for this reason I
intentionally labeled my approach the structure of strategic communication so as to highlight and
encourage researchers to consider more seriously not just argument structure per se but the
structural aspects of communication more generally.
Structural Aspects of Communication
Structural aspects of communication are the latent patterns that underlie how
communication is conveyed. It is helpful to understand communication structure by relating it to
communication content. While communication content refers to what actors say, communication
structure refers to how what they say fits together to make it intelligible. That is, irrespective of
motives, goals, age, gender, topic, audience, location, etc., if a speaker wants to convey an
intelligible message to others, he or she will follow specific structural rules for communicating.
As a result, unlike communication content, the structural aspects of communication tend to
remain more stable across contexts and actors. Because of this stability and invariance across
102

contexts, I argue that this is why structure may be a generative and fruitful foundation upon
which to examine strategic communication in future research.
Structural aspects of communication reside at an implicit third order level within
everyday communication. I have already discussed at length the first and second order levels.
The first order level refers to communication occurs within the rules of the game. The second
order level refers to the rules of the game itself, and this second order can remain implicit or be
discussed directly in communication. With regard to these first two levels, the game referred to is
the backing, or the collective assumptions regarding “what it is we are doing here.” However,
these first two levels always occur within an even deeper set of assumptions that I refer to as the
third order level that grounds even the discussion about the backing. These deeper assumptions
reflect the rules of a very different sort of game—the game of communicating with others. These
rules are the structural aspects of communication, forming the grounds upon which human
beings can interact intelligibly with one another.
What are the structural aspects of communication? What are the rules of this deeper level
game of communicating with others? Argument structure is of course one example. Indeed,
whenever a speaker wants to make an argument, it is nearly universally accepted that the speaker
will provide some sort or data to support their desired claim. This is the case regardless of the
location of the argument, the speaker, the desired claim, or the type of data elicited. Indeed,
imagining an argument without a claim (even an implicit one) is nearly impossible. This is
because societies have defined collectively what it means to make an argument and, by
definition, arguments have to make a claim. Beyond argument structure, there are other systems
of rules that ground how humans communicate with one another and may provide other potential
bases upon which to study the structure of strategic communication. I discuss in detail one
103

particularly prominent structural aspect of communication—grammatical structure—followed by
several other possibilities for future research.
Grammatical Structure
Grammar refers to the set of structural rules governing the composition of words and
phrases in any natural language (Wittgenstein, 1953). Grammar functions as perhaps the most
basic structural aspect of communication, since it forms the rudimentary mechanics of how the
“atoms” of our symbolic universe (i.e., words) configure together to create meaning. Like
argument structure, grammar is incredibly stable across speakers, audiences, and even cultures,
and contains standard structural components. Actors typically express a complete unit of thought
by including standard grammatical components like subjects and verbs. However, they may also
express additional things through the inclusion of additional grammatical components like
objects, adjectives, and so on. By exploring these structural components of grammar in different
ways, just as I did with the structural elements of arguments, scholars may uncover similarly
unique insights.
In particular, one might consider examining the degree to which certain grammatical
components are included or excluded in a given speech act. The increased use of adjectives, for
example, may reflect more conceptual thinking because adjectives work grammatically to
modify in greater detail existing points already conveyed in the sentence. Moreover, the
decreased use of verbs may reflect that the speaker is less ready to take immediate action because
he or she is avoiding the use of the structural component that signals movement in time or space.
Relatedly, one might also consider the sequencing of these grammatical components. Indeed,
English speakers typically place the subject first, followed by either the verb or the object.
Variation in this sequencing may signal the level of responsibility the speaker is willing to take
104

for the topic discussed because the sequencing signals an implicit priority with regard to how the
speaker relates to that topic.
Another way to examine grammatical structure might be to explore how all of the
possible grammatical components configure together in increasingly complex structures.
Sentence structure for instance can get more complex through the use of more grammatical
components (e.g., including multiple verbs and adjectives) as well as longer grammatical
components (e.g., adjectives with eight instead of four letters). While such complexity may be
desirable when discussing certain topics, it may also be used strategically to enhance audience
confusion or increase uncertainty (60 minutes, 2007). Such considerations may be particularly
relevant in public communications between organizations and stakeholders like analysts or
investors (Graffin et al., 2011). Taken together, examining grammar as a second structural aspect
of strategic communication may yield important insights in future research.
Other Types of Communication Structure
There are also other types of communication structure that may reveal novel insights.
Indeed, the rules by which humans communicate in order to remain intelligible to each other
extend well beyond argument and grammatical structure. Although the aim of this dissertation is
not to outline each of these other structural aspects in great detail, I briefly outline two additional
possibilities here so as to potentially spark some interest and future research.
First, just as we following grammatical sequencing rules, we also abide by certain
sequencing rules in a number of other situations. For example, scholars have hinted at the fact
that narratives or stories tend to follow different structural sequences (Martens et al., 2007;
O’Connor, 2004). Scholars might explore how variations in narrative structure lead to different
reactions from audiences. As with narratives or stories, there also may be an underlying structure
105

to arguments. While I did not explore this idea in this dissertation, it is possible that the
sequencing by which actors argue within and about the rules of the game could have differential
effects. Indeed, a speech beginning and ending with arguments that occur within the rules of the
game may produce a different audience reaction than speeches beginning and ending with
arguments about the rules of the game. Considering these sequencing possibilities may yield
interesting insights.
Second, actors also tend to follow what some have called a conversational structure
(Grice, 1968; Schwarz, 1996). Conversational structure refers to how people talk and cooperate
with one another in everyday conversation. The overarching rule that Grice argues people follow
when conversing is: “Make your contribution such as it is required, at the stage at which it
occurs, by the accepted purpose or direction of the talk exchange in which you are engaged”
(Grice, 1975). Based on this overarching rule, Grice developed four maxims that referred to the
principles of quantity (e.g., where one tries to be as informative as one possibly can, and gives as
much information as is needed, and no more), quality (e.g., where one tries to be truthful, and
does not give information that is false or that is not supported by evidence), relation (e.g., where
one tries to be relevant, and says things that are pertinent to the discussion), and manner (e.g.,
when one tries to be as clear, as brief, and as orderly as one can in what one says, and where one
avoids obscurity and ambiguity). Scholars might examine these structural aspects of conversation
to explore how variation across as well as violations of such considerations might influence
different audiences.
CONCLUDING THOUGHTS
This dissertation sought to develop a new approach to the study of strategic
communication. By leveraging how people naturally structure their arguments, I advanced not
106

only new theory but also a novel methodology to explore these ideas. In doing so, I believe that
this dissertation also makes a broader claim about the dual role language plays in society that I
discussed in the introduction. In particular, social theorists in philosophy (Rorty, 1981),
linguistics (Habermas, 1984), anthropology (Geertz, 1973), sociology (Berger & Luckmann,
1966), and management (Harmon et al., 2015) all have argued that the words we use can
simultaneously maintain as well as bring about change to the social institutions in which we live.
However, these theories tend to conceptualize the maintenance and change of institutions as two
distinct processes, thereby separating this linguistic duality into analytically separate
components. While separating social processes that are inherently recursive can be useful for
analytical purposes (Archer, 1982), some have argued that we should instead sacrifice our ability
to examine these processes empirically so as to be truer to the circular nature of social life
(Giddens, 1984). This dissertation proposes a middle ground. Specifically, the ASR is a singular
construct that captures the precise inflection point in our everyday argument structure that
produces these diverging effects of maintenance and change. This inflection point—which is the
keystone of this dissertation—may be the theoretically and empirically observable location to
begin studying more rigorously the circularity of social life.

107

REFERENCES

60 minutes. 2007, September 16. Alan Greenspan’s use of “fedspeak.”
https://www.youtube.com/watch?v=0liegrixknA.
Abolafia, M. Y. 2004. Framing moves: interpretive politics at the Federal Reserve. Journal of
Public Administration Research and Theory, 14(3): 349–370.
Abolafia, M. Y. 2012. Central banking and the triumph of technical rationality. Handbook of the
Sociology of Finance, 94–114.
Akerlof, G. A. 1970. The Market for “Lemons”: Quality Uncertainty and the Market Mechanism.
The Quarterly Journal of Economics, 84(3): 488–500.
Aldrich, H. E., & Fiol, C. M. 1994. Fools rush in? The institutional context of industry creation.
Academy of Management Review, 19(4): 645–670.
Andersson, M., Krylova, E., & Vähämaa, S. 2008. Why does the correlation between stock and
bond returns vary over time? Applied Financial Economics, 18(2): 139–151.
Ang, A., Hodrick, R. J., Xing, Y., & Zhang, X. 2006. The cross-section of volatility and
expected returns. The Journal of Finance, 61(1): 259–299.
Ansari, S., Wijen, F., & Gray, B. 2013. Constructing a climate change logic: An institutional
perspective on the “tragedy of the commons.” Organization Science, 24(4): 1014–1040.
Archer, M. S. 1982. Morphogenesis versus structuration: on combining structure and action.
British Journal of Sociology, 33(4): 455–483.
Aristotle. 1991. The art of rhetoric. New York: Penguin Books.
Arndt, M., & Bigelow, B. 2000. Presenting structural innovation in an institutional environment:
Hospitals’ use of impression management. Administrative Science Quarterly, 45(3): 494–
522.
108

Baker, M., & Wurgler, J. 2006. Investor sentiment and the cross-section of stock returns. The
Journal of Finance, 61(4): 1645–1680.
Bakhtin, M. M. 1982. The dialogic imagination: Four essays, vol. 1. Univ of Texas Pr.
Barberis, N., Shleifer, A., & Vishny, R. 1998. A model of investor sentiment. Journal of
Financial Economics, 49(3): 307–343.
Bartel, C. A., & Garud, R. 2009. The role of narratives in sustaining organizational innovation.
Organization Science, 20(1): 107–117.
Battilana, J., & D’Aunno, T. 2009. Institutional work and the paradox of embedded agency.
Institutional Work: Actors and Agency in Institutional Studies of Organizations, 31–58.
Benford, R. D., & Snow, D. A. 2000. Framing processes and social movements: An overview
and assessment. Annual Review of Sociology, 26(1): 611–639.
Berger, P. L., & Luckmann, T. 1966. The social construction of reality. New York: Anchor.
Bernanke, B. S. 2009, December 7. Frequently Asked Questions. At the Economic Club of
Washington D.C., Washington D.C.
Bernanke, B. S. 2011, October 18. The Effects of the Great Recession on Central Bank Doctrine
and Practice. At the Federal Reserve Bank of Boston 56th Economic Conference,
Boston, Massachusetts.
Bernanke, B. S. 2013. Communication and Monetary Policy. At the National Economists Club
Annual Dinner, Herbert Stein Memorial Lecture, Washington, D.C., on November 19,
2013.
Bernanke, B. S. 2014, January 3. The Federal Reserve: Looking Back, Looking Forward. Annual
Meeting of the American Economic Association, Philadelphia, Pennsylvania.
109

Bernanke, B. S. 2015. The Courage to Act: A Memoir of a Crisis and Its Aftermath. New York,
NY: W. W. Norton & Company.
Bettman, J. R., & Weitz, B. A. 1983. Attributions in the board room: Causal reasoning in
corporate annual reports. Administrative Science Quarterly, 165–183.
Bia’lkowski, J., Gottschalk, K., & Wisniewski, T. P. 2008. Stock market volatility around
national elections. Journal of Banking & Finance, 32(9): 1941–1953.
Bitektine, A. 2011. Toward a theory of social judgments of organizations: The case of
legitimacy, reputation, and status. Academy of Management Review, 36(1): 151–179.
Bitektine, A., & Haack, P. 2015. The “Macro” and the “Micro” of Legitimacy: Toward a
Multilevel Theory of the Legitimacy Process. Academy of Management Review, 40(1):
49–75.
Bitzer, L. F. 1959. Aristotle’s enthymeme revisited. Quarterly Journal of Speech, 45(4): 399–
408.
Blinder, A. S. 2001. How Do Central Banks Talk?. Centre for Economic Policy Research.
Bouwmeester, O. 2013. Field dependency of argumentation rationality in decision-making
debates. Journal of Management Inquiry, 22(4): 415–433.
Brysbaert, M., Warriner, A. B., & Kuperman, V. 2014. Concreteness ratings for 40 thousand
generally known English word lemmas. Behavior Research Methods, 46(3): 904–911.
Burke, K. 1969. A grammar of motives. Berkeley: University of California Press.
Burris, E. 2011. The risks and rewards of speaking up: Managerial responses to employee voice.
Academy of Management Journal.
Carrns, A. 2013, March 25. Expense and Emotions Affect Decisions About Long-Term Care.
The New York Times.
110

Carruthers, B. G., & Kim, J.-C. 2011. The sociology of finance. Annual Review of Sociology, 37:
239–259.
Chang, E. C., Pinegar, J. M., & Ravichandran, R. 1993. International evidence on the robustness
of the day-of-the-week effect. Journal of Financial and Quantitative Analysis, 28(04):
497–513.
Chen, C. R., Mohan, N. J., & Steiner, T. L. 1999. Discount rate changes, stock market returns,
volatility, and trading volume: Evidence from intraday data and implications for market
efficiency. Journal of Banking & Finance, 23(6): 897–924.
Chen, E.-T. J., & Clements, A. 2007. S&P 500 implied volatility and monetary policy
announcements. Finance Research Letters, 4(4): 227–232.
Chicago Board Options Exchange. 2003. VIX CBOE volatility index. White Paper, Www. Cboe.
Com.
Choudhry, T. 2000. Day of the week effect in emerging Asian stock markets: evidence from the
GARCH model. Applied Financial Economics, 10(3): 235–242.
Conley, T. M. 1984. The enthymeme in perspective. Quarterly Journal of Speech, 70(2): 168–
187.
Connolly, R., Stivers, C., & Sun, L. 2005. Stock market uncertainty and the stock-bond return
relation. Journal of Financial and Quantitative Analysis, 40(01): 161–194.
Cook, T. 2016, February 16. Customer Letter.
Cornelissen, J. P., Mantere, S., & Vaara, E. 2014. The contraction of meaning: The combined
effect of communication, emotions, and materiality on sensemaking in the Stockwell
shooting. Journal of Management Studies, 51(5): 699–736.
111

Covaleski, M. A., Dirsmith, M. W., & Rittenberg, L. 2003. Jurisdictional disputes over
professional work: the institutionalization of the global knowledge expert. Accounting,
Organizations and Society, 28(4): 323–355.
Creed, W. D., Hudson, B. A., Okhuysen, G. A., & Smith-Crowe, K. 2014. Swimming in a sea of
shame: Incorporating emotion into explanations of institutional reproduction and change.
Academy of Management Review, 39(3): 275–301.
Cruikshank, J. L., & Sicilia, D. B. 1999. The Greenspan Effect: Words That Move the World’s
Markets. New York: McGraw-Hill Companies.
Davis, A. K., Piger, J. M., & Sedor, L. M. 2012. Beyond the Numbers: Measuring the
Information Content of Earnings Press Release Language. Contemporary Accounting
Research, 29(3): 845–868.
Davis, M. A., & Bobko, P. 1986. Contextual effects on escalation processes in public sector
decision making. Organizational Behavior and Human Decision Processes, 37(1): 121–
138.
Detert, J. R., Burris, E. R., Harrison, D. A., & Martin, S. R. 2013. Voice Flows to and around
Leaders Understanding When Units Are Helped or Hurt by Employee Voice.
Administrative Science Quarterly, 58(4): 624–668.
Dewulf, A., Gray, B., Putnam, L., Lewicki, R., Aarts, N., et al. 2009. Disentangling approaches
to framing in conflict and negotiation research: A meta-paradigmatic perspective. Human
Relations, 62(2): 155–193.
DiMaggio, P. J., & Powell, W. W. 1983. The iron cage revisited: Institutional isomorphism and
collective rationality in organizational fields. American Sociological Review, 147–160.
Drucker, P. F. 1994. The theory of the business. Harvard Business Review, 72(5): 95–104.
112

Dubois, M., & Louvet, P. 1996. The day-of-the-week effect: The international evidence. Journal
of Banking & Finance, 20(9): 1463–1484.
Dunn, M. B., & Jones, C. 2010. Institutional logics and institutional pluralism: The contestation
of care and science logics in medical education, 1967–2005. Administrative Science
Quarterly, 55(1): 114–149.
Ederington, L. H., & Lee, J. H. 1993. How markets process information: News releases and
volatility. The Journal of Finance, 48(4): 1161–1191.
Edgerly, S., Vraga, E., Fung, T., Moon, T. J., Yoo, W. H., et al. 2009. YouTube as a public
sphere: The proposition 8 debate. Association of Internet Researchers Conference,
Milwaukee, WI.
Elsbach, K. D. 1994. Managing organizational legitimacy in the California cattle industry: The
construction and effectiveness of verbal accounts. Administrative Science Quarterly,
39(1): 57–88.
Elsbach, K. D., Sutton, R. I., & Principe, K. E. 1998. Averting expected challenges through
anticipatory impression management: A study of hospital billing. Organization Science,
9(1): 68–86.
Emrich, C. G., Brower, H. H., Feldman, J. M., & Garland, H. 2001. Images in words:
Presidential rhetoric, charisma, and greatness. Administrative Science Quarterly, 46(3):
527–557.
Erkama, N., & Vaara, E. 2010. Struggles over legitimacy in global organizational restructuring:
A rhetorical perspective on legitimation strategies and dynamics in a shutdown case.
Organization Studies, 31(7): 813–839.
113

Federal Reserve Board. 2016. FRB: Who owns the Federal Reserve?.
http://www.federalreserve.gov/faqs/about_14986.htm.
Fiss, P. C., & Zajac, E. J. 2006. The symbolic management of strategic change: Sensegiving via
framing and decoupling. Academy of Management Journal, 49(6): 1173.
Fligstein, N. 1997. Social skill and institutional theory. American Behavioral Scientist, 40(4):
397–405.
Fligstein, N. 2001. Social skill and the theory of fields. Sociological Theory, 19(2): 105–125.
Fligstein, N., Brundage, J. S., & Schultz, M. 2014. Why the Federal Reserve Failed to See the
Financial Crisis of 2008: The Role of “Macroeconomics” as a Sense making and
Cultural Frame.
Fligstein, N., & Calder, R. 2001. Architecture of Markets. Emerging Trends in the Social and
Behavioral Sciences: An Interdisciplinary, Searchable, and Linkable Resource.
Frank, K. 2000. The Impact of a Confounding Variable on a Regression Coefficient.
Sociological Methods and Research, 29(2): 147–194.
Friedland, R., & Alford, R. R. 1991. Bringing society back in: Symbols, practices and
institutional contradictions. The new institutionalism in organizational analysis (Walter
W. Powell and PaulJ. DiMaggio (eds.)): 232–263.
Friendly, M., Franklin, P. E., Hoffman, D., & Rubin, D. C. 1982. The Toronto Word Pool:
Norms for imagery, concreteness, orthographic variables, and grammatical usage for
1,080 words. Behavior Research Methods & Instrumentation, 14(4): 375–399.
Furr, N., & Dyer, J. H. 2014. Leading your team into the unknown. Harvard Business Review,
92(12): 80–88.
114

Garud, R., Schildt, H. A., & Lant, T. K. 2014. Entrepreneurial storytelling, future expectations,
and the paradox of legitimacy. Organization Science, 25(5): 1479–1492.
Geertz, C. 1973. The interpretation of cultures. New York: Basic, 412–453.
Giddens, A. 1984. The constitution of society: Outline of the theory of structuration. Berkeley:
University of California Press.
Gino, F. 2015. Don’t Let Emotions Screw Up Your Decisions. Harvard Business Review, 93(5):
4–10.
Goffman, E. 1974. Frame analysis: An essay on the organization of experience. Harvard
University Press.
Goffman, E. 1978. The presentation of self in everyday life. Harmondsworth.
Goodnight, G. T. 1993. Legitimation Inferences: An Additional Component for the Toulmin
Model. Informal Logic, 15(1): 41–52.
Gorgi, S., & Weber, K. 2015. Marks of Distinction Framing and Audience Appreciation in the
Context of Investment Advice. Administrative Science Quarterly.
Graffin, S. D., Carpenter, M. A., & Boivie, S. 2011. What’s all that (strategic) noise?
anticipatory impression management in CEO succession. Strategic Management Journal,
32(7): 748–770.
Gramsci, A. 1971. Selections from the Prison Notebooks, edited and translated by Quintin Hoare
and Geoffrey Nowell Smith. New York: International.
Gray, B., Purdy, J. M., & Ansari, S. (Shaz). 2015. From Interactions to Institutions:
Microprocesses of Framing and Mechanisms for the Structuring of Institutional Fields.
Academy of Management Review, 40(1): 115–143.
115

Green, S. E. J. 2004. A rhetorical theory of diffusion. Academy of Management Review, 29(4):
653–669.
Green, S. E. J., Babb, M., & Alpaslan, C. M. 2008. Institutional Field Dynamics and the
Competition Between Institutional Logics The Role of Rhetoric in the Evolving Control
of the Modern Corporation. Management Communication Quarterly, 22(1): 40–73.
Green, S. E. J., Li, Y., & Nohria, N. 2009. Suspended in self-spun webs of significance: A
rhetorical model of institutionalization and institutionally embedded agency. Academy of
Management Journal, 52(1): 11–36.
Greenspan, A. 1996, December 5. Central banking in a democratic society. At the Annual
Dinner and Francis Boyer Lecture of The American Enterprise Institute for Public Policy
Research, Washington, D.C.
Greenspan, A. 2001, October 11. Transparency in monetary policy. At the Federal Reserve Bank
of St. Louis, Economic Policy Conference, St. Louis, Missouri.
Greenspan, A. 2005, September 26. Mortgage banking. American Bankers Association Annual
Convention, Palm Desert, California.
Grice, H. P. 1968. Utterer’s meaning, sentence-meaning, and word-meaning. Foundations of
Language, 4(3): 225–242.
Grice, H. P. 1975. Logic and conversation, in Syntax and semantics, vol 3, Speech acts. Ed.
Peter Cole and Jerry Morgan, 41–58.
Griswold, W. 2012. Cultures and societies in a changing world. Sage.
Haack, P., Pfarrer, M. D., & Scherer, A. G. 2014. Legitimacy-as-Feeling: How Affect Leads to
Vertical Legitimacy Spillovers in Transnational Governance. Journal of Management
Studies, 51(4): 634–666.
116

Habermas, J. 1984. The theory of communicative action: Reason and the rationalization of
society, vol. 1. Boston: Beacon Press.
Hardy, C., & Maguire, S. 2010. Discourse, field-configuring events, and change in organizations
and institutional fields: Narratives of DDT and the Stockholm Convention. Academy of
Management Journal, 53(6): 1365–1392.
Harmon, D., Green, S., & Goodnight, G. 2015. A Model of Rhetorical Legitimation: The
Structure of Communication and Cognition Underlying Institutional Maintenance and
Change. Academy of Management Review, 40(1): 76–95.
Hayes, A. F., & Krippendorff, K. 2007. Answering the call for a standard reliability measure for
coding data. Communication Methods and Measures, 1(1): 77–89.
Heidegger, M. 1927. Being and time. New York: Harper & Row.
Heracleous, L., & Barrett, M. 2001. Organizational change as discourse: Communicative actions
and deep structures in the context of information technology implementation. Academy of
Management Journal, 44(4): 755–778.
Hiller, J. H., Fisher, G. A., & Kaess, W. 1969. A computer investigation of verbal characteristics
of effective classroom lecturing. American Educational Research Journal, 661–675.
Hilsenrath, J. 2015, October 26. The Fed Strives for a Clear Signal on Interest Rates. Wall Street
Journal.
Hilsenrath, J. 2016, March 24. The Decline of Dissent at the Fed. Wall Street Journal.
Holmes, D. R. 2013. Economy of Words: Communicative Imperatives in Central Banks.
Chicago ; London: University Of Chicago Press.
Holm, P. 1995. The dynamics of institutionalization: Transformation processes in Norwegian
fisheries. Administrative Science Quarterly, 40(3): 398–422.
117

Huang, X., Teoh, S. H., & Zhang, Y. 2013. Tone management. The Accounting Review, 89(3):
1083–1113.
Hutton, A. 2001. Four rules for taking your message to Wall Street. Harvard Business Review,
79(5): 125–32.
Jackson, S., & Jacobs, S. 1980. Structure of conversational argument: Pragmatic bases for the
enthymeme. Quarterly Journal of Speech, 66(3): 251–265.
Jamali, I. 2009. Stock market volatility and monetary policy. Concordia University.
Kahneman, D. 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.
Kennedy, M. T., & Fiss, P. C. 2009. Institutionalization, framing, and diffusion: The logic of
TQM adoption and implementation decisions among US hospitals. Academy of
Management Journal, 52(5): 897–918.
Keynes, J. M. 1936. The general theory of interest, employment and money. London: Macmillan.
Kincaid, J. P., Fishburne Jr, R. P., Rogers, R. L., & Chissom, B. S. 1975. Derivation of new
readability formulas (automated readability index, fog count and flesch reading ease
formula) for navy enlisted personnel. DTIC Document.
King, B. G. 2008. A political mediation model of corporate response to social movement
activism. Administrative Science Quarterly, 53(3): 395–421.
King, B. G., & Soule, S. A. 2007. Social movements as extra-institutional entrepreneurs: The
effect of protests on stock price returns. Administrative Science Quarterly, 52(3): 413–
442.
Knorr Cetina, K., & Preda, A. 2012. The Oxford Handbook of the Sociology of Finance. OUP
Oxford.
118

Krippendorff, K. H. 2003. Content Analysis: An Introduction to Its Methodology (2nd edition).
Thousand Oaks, Calif: Sage Publications, Inc.
Krugman, P. 2001, September 30. Fear Itself. The New York Times.
Lamin, A., & Zaheer, S. 2012. Wall Street vs. Main Street: Firm strategies for defending
legitimacy and their impact on different stakeholders. Organization Science, 23(1): 47–
66.
Lane, R. D., Chua, P. M., & Dolan, R. J. 1999. Common effects of emotional valence, arousal
and attention on neural activation during visual processing of pictures. Neuropsychologia,
37(9): 989–997.
Lawrence, T. B., Hardy, C., & Phillips, N. 2002. Institutional effects of interorganizational
collaboration: The emergence of proto-institutions. Academy of Management Journal,
45(1): 281–290.
Lee, J. 2015. Can Investors Detect Managers’ Lack of Spontaneity? Adherence to Predetermined
Scripts during Earnings Conference Calls. The Accounting Review, 91(1): 229–250.
Lerner, J. S., & Keltner, D. 2001. Fear, anger, and risk. Journal of Personality and Social
Psychology, 81(1): 146.
Lerner, J. S., Small, D. A., & Loewenstein, G. 2004. Heart strings and purse strings carryover
effects of emotions on economic decisions. Psychological Science, 15(5): 337–341.
Leubsdorf, B. 2015, January 4. Communications Key as Fed Prepares to Raise Rates. Wall Street
Journal.
Lev, B. 2011. How to Win Investors Over. Harvard Business Review, 89(11): 4–10.
Levene, H. 1960. Robust tests for equality of variances. Contributions to Probability and
Statistics: Essays in Honor of Harold Hotelling, 2: 278–292.
119

Levin, I. P. 1987. Associative effects of information framing on human judgments. Annual
meeting of the Midwestern Psychological Association, May, Chicago, IL.
Levin, I. P., Schneider, S. L., & Gaeth, G. J. 1998. All frames are not created equal: A typology
and critical analysis of framing effects. Organizational Behavior and Human Decision
Processes, 76(2): 149–188.
Loewenstein, J., Ocasio, W., & Jones, C. 2012. Vocabularies and vocabulary structure: A new
approach linking categories, practices, and institutions. The Academy of Management
Annals, 6(1): 41–86.
Lok, J., & De Rond, M. 2013. On the plasticity of institutions: Containing and restoring practice
breakdowns at the Cambridge University Boat Club. Academy of Management Journal,
56(1): 185–207.
Loughran, T., & McDonaold, B. 2011. When Is a Liability Not a Liability? Textual Analysis,
Dictionaries, and 10-Ks. The Journal of Finance, 66(1): 35–65.
Lounsbury, M., & Glynn, M. A. 2001. Cultural entrepreneurship: Stories, legitimacy, and the
acquisition of resources. Strategic Management Journal, 22(6-7): 545–564.
Lounsbury, M., & Hirsch, P. M. 2010. Markets on trial: The economic sociology of the US
financial crisis. Emerald Group Publishing.
Marcus, A. A., & Goodman, R. S. 1991. Victims and shareholders: The dilemmas of presenting
corporate policy during a crisis. Academy of Management Journal, 34(2): 281–305.
Martens, M. L., Jennings, J. E., & Jennings, P. D. 2007. Do the stories they tell get them the
money they need? The role of entrepreneurial narratives in resource acquisition. Academy
of Management Journal, 50(5): 1107–1132.
120

McClure, G. M. 1987. Readability formulas: Useful or useless. IEEE Transactions on
Professional Communications, 30: 12–15.
Mead, G. H. 1962. Mind, Self and Society: From the Standpoint of a Social Behaviorist.
Chicago: Chicago University Press.
Mergenthaler, E. 1996. Emotion–abstraction patterns in verbatim protocols: A new way of
describing psychotherapeutic processes. Journal of Consulting and Clinical Psychology,
64(6): 1306.
Merton, R. C. 1973. Theory of rational option pricing. The Bell Journal of Economics and
Management Science, 141–183.
Meyer, A. D., Brooks, G. R., & Goes, J. B. 1990. Environmental jolts and industry revolutions:
Organizational responses to discontinuous change. Strategic Management Journal, 11(5):
93–110.
Meyer, J. W., & Rowan, B. 1977. Institutionalized organizations: Formal structure as myth and
ceremony. American Journal of Sociology, 83(2): 340–363.
Meyer, J. W., & Scott, W. R. 1983. Centralization and the legitimacy problems of local
government. Organizational Environments: Ritual and Rationality, 199–215.
Neuendorf, K. A. 2001. The Content Analysis Guidebook (1st edition). Thousand Oaks, Calif:
SAGE Publications, Inc.
Niedenthal, P. M., & Kitayama, S. 2013. The heart’s eye: Emotional influences in perception
and attention. Academic Press.
Nikkinen, J., & Sahlström, P. 2004. Impact of the federal open market committee’s meetings and
scheduled macroeconomic news on stock market uncertainty. International Review of
Financial Analysis, 13(1): 1–12.
121

Nikkinen, J., Sahlström, P., & Äijö, J. 2007. Turn-of-the-month and intramonth effects:
Explanation from the important macroeconomic news announcements. Journal of
Futures Markets, 27(2): 105–126.
O’Connor, E. 2004. Storytelling to be real: narrative, legitimacy building and venturing.
Narrative and Discursive Approaches in Entrepreneurship: A Second Movements in
Entrepreneurship Book, 105–124.
Paul, R. 2009. End the Fed. Grand Central Publishing.
PBS News Hour. 2008, October 23. Greenspan Admits “Flaw” to Congress, Predicts More
Economic Problems.
Pennebaker, J. W., Booth, R. J., & Francis, M. E. 2007. Linguistic inquiry and word count:
LIWC [Computer software]. Austin, TX: Liwc. Net.
Pennebaker, J. W., Chung, C. K., Ireland, M., Gonzales, A., & Booth, R. J. 2007. The
development and psychometric properties of LIWC2007.
Pennebaker, J. W., Mehl, M. R., & Niederhoffer, K. G. 2003. Psychological aspects of natural
language use: Our words, our selves. Annual Review of Psychology, 54(1): 547–577.
Petty, R. E., & Cacioppo, J. T. 1986. The elaboration likelihood model of persuasion. Springer.
Pfarrer, M. D., Pollock, T. G., & Rindova, V. P. 2010. A tale of two assets: The effects of firm
reputation and celebrity on earnings surprises and investors’ reactions. Academy of
Management Journal, 53(5): 1131–1152.
Pfeffer, J. 1981. Management as symbolic action: The creation and maintenance of
organizational paradigms. Graduate School of Business, Stanford University.
Phillips, N., Lawrence, T. B., & Hardy, C. 2004. Discourse and institutions. Academy of
Management Review, 29(4): 635–652.
122

Plosser, C. I. 2015, December 9. The Veneer of Consensus at the Fed. Wall Street Journal.
Pollock, T. G., & Gulati, R. 2007. Standing out from the crowd: the visibility-enhancing effects
of IPO-related signals on alliance formation by entrepreneurial firms. Strategic
Organization, 5(4): 339–372.
Pollock, T. G., & Rindova, V. P. 2003. Media legitimation effects in the market for initial public
offerings. Academy of Management Journal, 46(5): 631–642.
Pollock, T. G., Rindova, V. P., & Maggitti, P. G. 2008. Market watch: Information and
availability cascades among the media and investors in the US IPO market. Academy of
Management Journal, 51(2): 335–358.
Powell, W. W., & Colyvas, J. A. 2008. Microfoundations of institutional theory. The Sage
Handbook of Organizational Institutionalism, 840.
Putnam, L. L., & Holmer, M. 1992. Framing, reframing, and issue development. Communication
and Negotiation, 128: 155.
Rhee, E., & Fiss, P. 2014. Framing Controversial Practices: Frame salience, source credibility,
and stock market reaction. Academy of Management Journal.
Rorty, R. 1981. Philosophy and the Mirror of Nature (1ST edition). Princeton, N.J.: Princeton
University Press.
Rousseau, D. M., & Tijoriwala, S. A. 1999. What’s a good reason to change? Motivated
reasoning and social accounts in promoting organizational change. Journal of Applied
Psychology, 84(4): 514.
Rubtsova, A., DeJordy, R., Glynn, M. A., & Zald, M. 2010. The social construction of causality:
The effects of institutional myths on financial regulation. Markets on Trial: The
Economic Sociology of the US Financial Crisis, 201–244.
123

Salancik, G. R., & Meindl, J. R. 1984. Corporate attributions as strategic illusions of
management control. Administrative Science Quarterly, 238–254.
Sanders, W. G., & Boivie, S. 2004. Sorting things out: Valuation of new firms in uncertain
markets. Strategic Management Journal, 25(2): 167–186.
Schlenker, B. R. 1980. Impression management: The self-concept, social identity, and
interpersonal relations. Monterey, CA: Brooks/Cole.
Schneider, S. L., Holstrum, G. L., & Marden, R. E. 1993. Framing effects and audit decision
making: Control environment evaluation. Poster presented at the annual meeting of the
Society for Judgment and Decision Making, November, Washington, DC.
Schwarz, N. 1996. Cognition and communication: Judgmental biases, research methods, and the
logic of conversation. Ann Arbor, 1001: 48106–1248.
Schweitzer, M. E., Brooks, A. W., & Galinsky, A. D. 2015. The organizational apology.
Harvard Business Review, 94(9): 44–52.
Seo, M. G., & Creed, W. D. 2002. Institutional contradictions, praxis, and institutional change: A
dialectical perspective. Academy of Management Review, 27(2): 222–247.
Shiller, R. J. 1988. Causes of changing financial market volatility. Proceedings, Federal Reserve
Bank of Kansas City, 1–32.
Shiller, R. J. 2000. Irrational Exuberance (Second edition). Princeton, N.J: Princeton University
Press.
Simon, H. A. 1982. Models of bounded rationality: Behavioral economics and business
organization. Cambridge, MA: MIT Press.
Simons, H. W. 1994. “Going meta”: Definition and political applications. Quarterly Journal of
Speech, 80(4): 468–481.
124

Smets, M., Morris, T. I. M., & Greenwood, R. 2012. From practice to field: A multilevel model
of practice-driven institutional change. Academy of Management Journal, 55(4): 877–
904.
Spence, M. 1973. Job market signaling. The Quarterly Journal of Economics, 355–374.
Statman, M. 2013, April 21. Investor, Know Yourself. Wall Street Journal.
Staw, B. M., McKechnie, P. I., & Puffer, S. M. 1983. The justification of organizational
performance. Administrative Science Quarterly, 582–600.
Steil, B. 2013. The battle of Bretton Woods: John Maynard Keynes, Harry Dexter White, and the
making of a new world order. Princeton University Press.
Steinberg, M. W. 1999. The Talk and Back Talk of Collective Action: A Dialogic Analysis of
Repertoires of Discourse among Nineteenth-Century English Cotton Spinners 1.
American Journal of Sociology, 105(3): 736–780.
Suchman, M. C. 1995. Managing legitimacy: Strategic and institutional approaches. Academy of
Management Review, 20(3): 571–610.
Suddaby, R., & Greenwood, R. 2005. Rhetorical strategies of legitimacy. Administrative Science
Quarterly, 50(1): 35–67.
Sutton, R. I., & Callahan, A. L. 1987. The stigma of bankruptcy: Spoiled organizational image
and its management. Academy of Management Journal, 30(3): 405–436.
Tausczik, Y. R., & Pennebaker, J. W. 2010. The psychological meaning of words: LIWC and
computerized text analysis methods. Journal of Language and Social Psychology, 29(1):
24–54.
Tetlock, P. C. 2007. Giving content to investor sentiment: The role of media in the stock market.
The Journal of Finance, 62(3): 1139–1168.
125

Tetlock, P. C., Saar-Tsechansky, M., & Macskassy, S. 2008. More than words: Quantifying
language to measure firms’ fundamentals. The Journal of Finance, 63(3): 1437–1467.
Thornton, P. H., Ocasio, W., & Lounsbury, M. 2012. The institutional logics perspective: A new
approach to culture, structure, and process. Oxford: Oxford University Press.
Tost, L. P. 2011. An integrative model of legitimacy judgments. Academy of Management
Review, 36(4): 686–710.
Toulmin, S. 1958. The uses of argument. Cambridge: Cambridge University Press.
Toulmin, S., Rieke, R. D., & Janik, A. 1984. An introduction to reasoning. New York:
Macmillan.
Townley, B. 2008. Reason’s Neglect: rationality and organizing. Oxford: Oxford University
Press.
Tversky, A., & Kahneman, D. 1981. The framing of decisions and the psychology of choice.
Science, 211(4481): 453.
Vaara, E., Tienari, J., & Laurila, J. 2006. Pulp and paper fiction: On the discursive legitimation
of global industrial restructuring. Organization Studies, 27(6): 789–813.
Vähämaa, S., & Äijö, J. 2011. The Fed’s policy decisions and implied volatility. Journal of
Futures Markets, 31(10): 995–1010.
Van Buskirk, A. 2012. Disclosure frequency and information asymmetry. Review of Quantitative
Finance and Accounting, 38(4): 411–440.
Voronov, M., & Vince, R. 2012. Integrating emotions into the analysis of institutional work.
Academy of Management Review, 37(1): 58–81.
Wade, J. B., Porac, J. F., & Pollock, T. G. 1997. Worth, words, and the justification of executive
pay. Journal of Organizational Behavior, 18(s 1): 641–664.
126

Werder, A. v. 1999. Argumentation rationality of management decisions. Organization Science,
10(5): 672–690.
Wessel, D. 1996, December 6. Fed Chairman Pops the Big Question. Wall Street Journal.
Westphal, J. D., & Zajac, E. J. 1998. The symbolic management of stockholders: Corporate
governance reforms and shareholder reactions. Administrative Science Quarterly, 127–
153.
Wittgenstein, L. 1953. Philosophical Investigations. Oxford: Basil & Blackwell.
Yang, Z., & Zhou, Y. 2016. Quantitative Easing and Volatility Spillovers Across Countries and
Asset Classes. Management Science.
Yellen, J. 2014, April 16. Monetary Policy and the Economic Recovery. At the Economic Club
of New York, New York, New York.
Yellen, J. L. 2013. Communication in Monetary Policy, At the Society of American Business
Editors and Writers 50th Anniversary Conference, Washington, D.C., on April 4, 2013.
Zavyalova, A., Pfarrer, M. D., Reger, R. K., & Shapiro, D. L. 2012. Managing the message: The
effects of firm actions and industry spillovers on media coverage following wrongdoing.
Academy of Management Journal, 55(5): 1079–1101.
Zilber, T. B. 2002. Institutionalization as an interplay between actions, meanings, and actors:
The case of a rape crisis center in Israel. Academy of Management Journal, 45(1): 234–
254.
Zucker, L. G. 1977. The role of institutionalization in cultural persistence. American
Sociological Review, 42(5): 726–743.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Essays on strategic categorization

PDF

Enchanted algorithms: the quantification of organizational decision-making

PDF

Better now than later: the cost of victims' delayed accusations

PDF

User lobbying amongst high-growth tech startups: the next evolution of constituency building

PDF

Three essays on young entrepreneurial firms

PDF

Strategic audience partitioning: antecedents and consequences

PDF

Empirical essays on relationships between alliance experience and firm capability development

PDF

For the love of the game? ownership and control in the NBA

PDF

Technology, behavior tracking, and the future of work

PDF

Who's the physician in charge? Generalist and specialist jurisdictions in professional practice

PDF

Automated contracts and the lawyers who don't review them: adoption and use of machine learning technology

PDF

Empirical essays on alliances and innovation in the biopharmaceutical industry

PDF

A question of identity: the rhetoric and argument of conspiracy theories

PDF

The convergence of investor relations and public relations: fitting investor relations into the overall corporate communications strategy

PDF

Strategic communications in dietary supplement marketing — a case study of the marketing of Coenzyme Q10 supplements

PDF

Whole foods market -- a communication plan to engage the German market

PDF

The curse of loyalty: interdependent self construal and support for corruption

PDF

Essays on understanding consumer contribution behaviors in the context of crowdfunding

PDF

Essays on information and financial economics

PDF

Expectation dynamics and stock returns

Asset Metadata

Creator Harmon, Derek J. (author)

Core Title The structure of strategic communication: theory, measurement, and effects

School Marshall School of Business

Degree Doctor of Philosophy

Degree Program Business Administration

Publication Date 07/22/2016

Defense Date 05/10/2016

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

Tag argument structure,Emotions,institutions,Markets,OAI-PMH Harvest,strategic communication

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Fiss, Peer (committee chair), Hoberg, Gerard (committee member), Mayer, Kyle (committee member), Wiltermuth, Scott (committee member)

Creator Email derek.j.harmon@gmail.com,djharmon@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-273631

Unique identifier UC11281360

Identifier etd-HarmonDere-4575.pdf (filename),usctheses-c40-273631 (legacy record id)

Legacy Identifier etd-HarmonDere-4575.pdf

Dmrecord 273631

Document Type Dissertation

Format application/pdf (imt)

Rights Harmon, Derek J.

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