Citation
The 2006 Colorado gubernatorial campaign

Material Information

Title:
The 2006 Colorado gubernatorial campaign the efficacy of the functional theory of political campaign discourse
Creator:
Sills, Sarah Lynn
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
x, 73 leaves : ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Public Administration)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public Administration
Committee Chair:
Burton, Lloyd
Committee Members:
Walkosz, Barbara
Eads, Marci

Subjects

Subjects / Keywords:
Governors -- Election -- Colorado ( lcsh )
Elections -- Colorado ( lcsh )
Political oratory -- Colorado ( lcsh )
Elections ( fast )
Governors -- Election ( fast )
Political oratory ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 69-73).
General Note:
School of Public Affairs
Statement of Responsibility:
by Sarah Lynn Sills.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
166394769 ( OCLC )
ocn166394769
Classification:
LD1193.P86 2007m S54 ( lcc )

Full Text
THE 2006 COLORADO GUBERNATORIAL CAMPAIGN:
THE EFFICACY OF THE FUNCTIONAL THEORY OF
POLITICAL CAMPAIGN DISCOURSE
by
Sarah Lynn Sills
B.S. Colorado State University, 1999
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center
in partial fulfillment
Of the requirements for the degree of
Master of Public Administration
2007


2007 by Sarah Lynn Sills
All rights reserved.


This thesis for the Master of Public Administration
Degree by
Sarah Lynn Sills
Marci Eads
Date


Sills, Sarah Lynn (Master of Public Administration)
The 2006 Colorado Gubernatorial Campaign: The Efficacy of the Functional Theory
of Political Campaign Discourse
Thesis directed by Professor Lloyd Burton
ABSTRACT
This study investigates newspaper coverage of the 2006 Colorado gubernatorial
campaign based on the functional theory of political campaign discourse a
technique that heretofore has been used in the study of presidential election
campaigns, but not gubernatorial elections. A content analysis of candidate-generated
news stories showed acclaims are more frequent than attacks, which are more
frequent than defenses. In addition, the study found that the Democratic candidate
discussed policy issues more frequently than the Republican candidate who discussed
character issues more frequently. The candidate who discussed policy issues most
frequently won the election. Finally, the study showed that there was not a
relationship between the candidate who was behind in the polling data and the
frequency that they attacked the leading candidates character and policies.
This abstract accurately represented the content of the candidates thesis.
recommend its publication.
urton


DEDICATION
I dedicate this thesis to my husband Michael, who always believed in me and
supported me. I also dedicate this thesis to my father, Steve who instilled the
important of an education in me.


ACKNOWLEDGEMENT
I want to thank Jenny Kilgroe and Cheryl Hyink for their help in coding articles. I
also want to thank my thesis committee Dr. Lloyd Burton, Dr. Brenda Walkosz, and
Dr. Marci Eads for their assistance in reviewing this paper.


CONTENTS
Tables.................................................... x
CHAPTER
1. INTRODUCTION TO THE STUDY.............................. 1
Background.......................................... 2
Changing Electoral Partisanship............... 5
Background of Colorado Politics............... 6
Professional Significance............................ 13
Overview of Methodology.............................. 14
2. LITERATURE REVIEW....................................... 17
New Approaches to Campaign Rhetorical Analysis....... 17
Functional Theory of Political Campaign Discourse.... 19
Themes......................................... 20
Function to Acclaim, Attack, and Defend...... 20
Issue Policy and Character................... 22
Empirical Research............................. 25
3. METHODOLOGY............................................. 29
Research Questions and Hypotheses.................... 29
General Perspective.................................. 30
Research Context..................................... 32
Description of Data and Collection................... 33
Data Analysis Methodology............................ 35
Identification of Themes....................... 35
Coding for Function, Issue, and Topic.......... 37
Coding Methodology...................... 40
4. THE RESULTS OF THE STUDY................................ 42
Data Set Descriptives................................ 42
Results.............................................. 45
Gubernatorial Campaign Trends........................ 45
5. SUMMARY AND DISCUSSION.................................. 55
Statement of the Problem............................. 55
Review of the Methodology............................ 55
Summary of the Results............................... 56
viii


Discussion of the Results...................... 57
Recommendations for Future Research............ 63
APPENDIX
A: KEY WORDS FOR POLICY TOPICS...................... 67
REFERENCES................................................ 69
IX


LIST OF TABLES
Table
1.1 Colorado population and party affiliation 1970 2006............. 8
1.2 Colorado presidential selections and Congressional seats 1950 2004 9
1.3 Control of the Colorado Governorship and Legislature 1950-2004... 10
4.1 Articles by week of the 2006 Colorado gubernatorial campaign..... 43
4.2 Themes attributed to the candidates by week of the 2006 Colorado
gubernatorial campaign.............................................. 44
4.3 Acclaims, attacks, and defenses by week of the 2006 Colorado
gubernatorial campaign.............................................. 46
4.4 Policy versus character issues by week of the 2006 Colorado
gubernatorial campaign.............................................. 47
4.5 Comparison of function and issue references in the 2006 Colorado
gubernatorial campaign.............................................. 48
4.6 Comparison of function and issue by candidate in the 2006 Colorado
gubernatorial campaign.............................................. 48
4.7 Policy and character utterances by candidate in the 2006 Colorado
gubernatorial campaign.............................................. 49
4.8 Policy issues by candidate in the 2006 Colorado gubernatorial
campaign............................................................ 50
4.9 Comparison of function and policy issue in the 2006 Colorado
gubernatorial campaign.............................................. 51
4.10 Attacks versus polling data in the 2006 Colorado gubernatorial
campaign............................................................ 53
x


CHAPTER 1
INTRODUCTION TO THE STUDY
Numerous statements and analogies have been used to describe American
political campaigns as reflections of the current and changing political, economic, and
social environments. Edwin Black (1972) stated, Our political experience is seasonal
and cyclical (p. 125). Other scholars have likened the American political campaign
to a game of rhetorical chess, and their attempts to explain voters decisions as the
strategy to win the chess game (Swanson, 1972). William Benoit, in his 1999 study,
described campaigns simply as ...inherently comparative: A voter chooses among
two or more candidates, and the candidate who appears most suitable (on whatever
criteria are most important to a given voter) will receive that persons vote (p. 16);
therefore, in the most simplistic sense, a political campaign is a rhetorical exercise.
The primary goal of rhetoric or persuasion in a political campaign is the
effective communication of a candidates ideas of government and governance to
persuade voters to vote for a specific candidate. As a result, candidates must be aware
of their messages, how their messages are presented, and how voters perceive them.
A significant amount of research has been conducted on political campaigns,
specifically presidential campaigns from 1948 through 2004, to analyze the efficacy
of candidates campaign messages that acclaim, attack, or defend their or their
opponents character and policies. However, the same studies have not been
conducted as extensively on state level elections, including gubernatorial campaigns.
Therefore, this study will evaluate candidate-generated campaign messages from the


2006 Colorado gubernatorial campaign to determine if its rhetorical trends are similar
to those of presidential campaigns described by the functional theory of political
campaign discourse.
This initial chapter includes a general background of the study, which
includes a discussion of changing electoral partisanship and the shifting political
landscape in Colorado. The chapter also presents the studys professional significance
including a brief problem statement and concludes with a brief overview of the
methodology.
Background
The need to study political campaigns and their impact on society was argued
by Gronbeck in 1978. To agree with the .assumption governments are constituted
by collectivists to provide protection and service, one must note that problems
necessarily justify governments reason-for-being and in order to form a collectivist
government, the population must select representatives by voting (Gronbeck, 1978, p.
276). Gronbeck (1978) further argued that by voting people give elected officials
opportunities to reflect and act on problems. It is through the reflections and actions
of our elected representatives that voters have biases for and against candidates based
on the candidates party affiliation. For example, Democrats are seen as .. .willing
to rely on government intervention, whether to regulate the economy and business or
to assist the economically disadvantaged (Jarvis, 2004, p. 3). In contrast, society
views Republicans as ...champions of low taxes and of bringing business-like
efficiency to government... [and] supporters of entrepreneurial activity (Jarvis,
2


2004, p. 3). These perceptions of political parties are based on the rhetoric of the
candidates which begins during primary and general campaigns.
Primary and general campaign messages from presidential elections from
1948 to 2004 have been extensively analyzed for trends in discourse. These analyses
have included numerous forms of media including newspapers, television
advertisements, debates, and acceptance speeches. One theory frequently used to
provide a framework for understanding how campaign messages are presented to
constituents is the functional theory of political campaign discourse (Benoit, 2004).
The theory has been used to explore presidential campaigns from 1948 to 2004;
however, based on an extensive literature review, the evaluation of statewide and
local campaigns using the functional theory of political campaign discourse has not
been completed. The 2006 Colorado gubernatorial campaign provided an opportunity
to begin this research. This study will use the functional theory of political campaign
discourse as the primary theoretical framework for the analysis of the discourse of
acclaims, attacks, and defenses based on character and policy issues.
Intellectual, research, and social concerns significantly influenced the
conceptualization and execution of this study. From the intellectual and research
perspectives, the field of political campaign research is small, but researchers have
made significant contributions in effort to explain the rhetorical choices made by
campaigning candidates. Researchers have shown Democratic presidential candidates
discuss policy more than Republicans, who discuss character more (Benoit, 2004).
And researchers have shown Democrats attack on policy, and Republicans attack on
3


character (Benoit, 2004). For example, a study conducted by Benoit (2003) analyzed
presidential campaign messages from 1948-2000. The study concluded that
presidential candidates who discuss policy issues more frequently than character
issues are more likely to win elections (Benoit, 2003). However, it is relatively
unknown if these trends are similar in elections at the state and local levels. The study
presented here seeks to begin to fill this gap in political campaign research by
analyzing partisan politics and candidate campaign rhetoric at a state level by testing
similar hypotheses.
From a research perspective, the selection of the general campaign as the
period of analysis was significant. Researchers have evaluated primary and general
elections and concluded primary elections tend to be more polar within the parties as
Republican candidates seek to differentiate themselves from the other Republican
candidates, the same is true of Democratic primary candidates (Hansen and Benoit,
2001). In contrast, general elections have become more centrist as Democratic and
Republican candidates balance their policies between core party ideals to retain the
support of core constituents and moderate ideals to appeal to the undecided voters.
According to Hart and Childers (2005), more recent elections have required that
candidates direct their appeals to Petrociks (1996) median voter, who is
concerned with collective goods and the resolution of problems. Based on Petrociks
(1996) definition of a median voter, candidates should focus on the problems they
can resolve not the policies they will pursue. It is through problem resolution that
Republican and Democratic candidates can differentiate themselves in a general
4


election. For example, a Republican candidate would propose to decrease taxes, while
a Democrat would propose to increase taxes in response to economic difficulties
(Jarvis, 2004). Therefore, in primary elections candidates tend to separate themselves
based on policy beliefs, and in general elections candidates differentiate themselves
based on policy solutions. For the purposes of this study, the general election of the
2006 Colorado gubernatorial campaign was selected because it represented the
balancing of policy ideas and solutions appealing to the core party constituents and
the median voter, and it was representative of a larger cultural shift in Colorado.
Changing Electoral Partisanship
The final component that impacted the conceptualization and execution of this
study is the social aspect of shifting electoral partisanship, both in the United States
and more specifically in Colorado. People traditionally join or associate with groups
with whom they have something in common, thus allowing the group to further a
common interest or goal. Political parties are one such example of group membership.
Constituents either seek to alter or protect the political, social, and/or economic status
quo depending on the beneficial aspect to the individual (Petrocik, 1996).
Understandably, voters concerns change between elections as their priorities change
based on larger political, social, and economic shifts (Benoit, 2004). Further, because
voters are changing their preferences, candidates have become more centrist in an
effort to remain appealing to the core party members, while trying to attract the
median voter. According to S.L. Popkin (1994), Today in an environment of
diminishing party loyalty, campaigns and candidates exert a greater influence on
5


voters than they did in the elections of 1940 and 1948 (p. 2), and as a result it has
become more difficult, yet important, to differentiate one candidate from another.
Currently there are approximately 215 million voters in the United States
(Holder, 2006) who identify as a Republican, Democrat, or an Independent. A recent
analysis of partisanship has shown a decreasing trend in the importance of political
parties. In 1952, 23% of voters identified themselves as Independent voters, as
compared to 38% in 1992 (Benoit & Harthcock, 1999). Further in 2000, 39% of the
electorate was registered as Independent voters, 34% were Democrats, and 25% were
Republicans (Benoit, Hansen, & Holbert, 2005). Finally in 2003, the Pew Research
Center for the People & the Press showed 33% of registered voters were Republican,
34% were Democrats, and 33% were registered as Independents or other (Kohut,
2003). Finally, Benoit and Harthcock (1999) showed that from 14-27% of voters
identify with one party but vote for a candidate from another party. While these
studies show a substantial number of American voters identify with one of the major
political parties, they also support the assertion that political campaigns must focus on
the Independent vote in order to be successful.
Background of Colorado Politics
The 2006 Colorado gubernatorial campaign was selected for this study for two
specific reasons. First, Governor Bill Owens (Republican, R) was term-limited, and
Lieutenant Governor Jane Norton did not run for the governors office. Secondly, and
most importantly, Colorado is representative of the political shift from Republican to
Democratic politics that is occurring in the western states, including: Montana,
6


Nevada, Arizona, New Mexico, and Wyoming. One of the most significant factors
influencing this shift of Colorados political identity is the growing population. In
2004 Congressman Bob Beauprez (R), the 2006 Republican gubernatorial candidate,
warned, Republicans definitely cant take Colorado for granted.. .It has a mobile,
fluid population (Cohen, 2004). Colorado has been reshaped economically and
politically, by its successive waves of new residents, according to the National
Journal (2006, p. 1). The rapidly growing population in Colorado, as illustrated in
Table 1.1, has contributed to the political struggle between the Republican western
slope and eastern plains and the Democratic Rocky Mountain region and eastern
foothills including Denver. The different political ideologies can be seen by
comparing the religiously conservative city of Colorado Springs, home to Focus on
the Family and the Air Force Academy, to Boulder, the liberal home of the University
of Colorado. This shifting political climate makes Colorado an ideal state for the
evaluation of gubernatorial campaigns and the functional theory of political campaign
discourse.
During the 1970s and 1980s, policies in Colorado focused on environmental
issues including slow growth initiatives and the designation of open space areas.
These policies were initiated by Democrats including: Governor Dick Lamm, Senator
Gary Hart, Congresswoman Patricia Schroeder, and Congressman Tim Wirth (Baron
and Cohen, 2006). This trend was reversed in the 1990s when the
telecommunications and high technology industries brought family-oriented cultural
conservatives {National Journal, 2006, p. N/A) to Colorado. The growing
7


Table 1.1 Colorado population and party affiliation 1970 2006
Gubernatorial election year Colorado population8 Percent change in population Number registered votersb Percent change in number of registered votersb Percentage registered Republicans11 Percentage registered Democratsb Percentage independents or unaffiliated minor partiesb
1970 2,209,596 - 968,892 - 23.7% 28.0% 34.4%
1974 2,541,406 15,0% 1,227,492 26.7% 25.7% 33.3% 35.3%
1978 2,766,748 8.9% 1,345,004 9.6% 25.6% 32.3% 37.2%
1982 3,061,564 10.7% 1,455,734 8.2% 25.4% 25.6% 36.6%
1986 3,237,450 5.7% 1,810,998 24.4% 31.2% 29.2% 35.9%
1990 3,307,618 2.2% 1,921,653 6.1% 32.1% 29.4% 34.9%
1994 3,724,168 12.6% 2,033.094 5.8% 27.0% 25.9% 32.9%
1998 4,116,639 10.5% 2,581,245 27.0% 35.6% 30.5% 33.9%
2002 4,498,407 9.3% 3,093,969 19.9% 36.3% 31.2% 32.5%
2006 4,756,915 5.7% 3,000,836 -3.0% 35.7% 30.2% 34.2%
8 Colorado Governors Office of State Planning and Budgeting (2007)
b Colorado Secretary of State Elections Division, 2007
population contributed to the rise of the importance of Colorado Springs in Colorado
politics. With the exception of 1992, when Colorado elected Democrats Bill Clinton
to the Presidency and Ben Nighthorse Campbell to the United States Senate, the state
elected Republicans to office including Governor Bill Owens in 1998, the first
Republican governor since 1970 (Colorado State Archives, 2006).
In 2000, Colorado elected Texas Governor George W. Bush (R) to the
Presidency over Vice President A1 Gore (Democrat, D) 51% to 42%, respectively
(Cohen, 2004). And the Republicans retained control of the Governors office in 2002
by reelecting Governor Owens. Also in 2002 Colorado reelected United States
Senator Wayne Allard (R). The reelection of Allard gave Colorado two Republican
Senators since Senator Campbell changed parties from Democrat to Republican in
1995, Table 1.2.
8


Table 1.2 Colorado presidential selections and Congressional seats 1950 2004
President US Senate+ US House+
US Colorado Democrats Republican Democrats Republicans
1948 Truman (D) Truman (D) 1 1 1 3
1952 Eisenhower Eisenhower (R) 1 1 2 2
1956 Eisenhower Eisenhower (R) - 2 2 2
1960 Kennedy(D) Kennedy (D) 1 1 3 1
1964 Johnson(D) Johnson (D) - 2 2 2
1968 Nixon (R) Nixon (R) - 2 3 1
1972 Nixon (R) Nixon (R) - 2 2b 3
1976 Carter (D) Carter (D) 2 - 3 2
1980 Reagan (R) Reagan (R) 1 1 3 2
1984 Reagan (R) Reagan(R) 1 1 4 2
1988 Bush (R) Bush (R) 1 1 5 2
1992 Clinton (D) Clinton (D) 1 1 3 3
1996 Clinton (D) Clinton (D) - 2 2 4
2000 Bush (R) Bush (R) - 2 2 4
2004 Bush (R) Bush (R) - 2 2d 5
a Barone and Cohen, 2005
b The fifth United States Congressional district was added in 1972.
c The sixth United States Congressional district was added in 1982.
d The seventh United States Congressional district was added in 2004.
However, in 2004 the Republican trend appeared to begin to shift as
Massachusetts Senator and Democratic presidential nominee John Kerry visited
Colorado three times by August. Republican Congressman Beauprez criticized the
Bush administration for its communication on Iraq, a major concern for Coloradoans
due to a large military population, [it] has not been as targeted or clear and forceful
as it could be (Cohen, 2004, p. 2572). In spite of Beauprezs concerns, the war in
Iraq turned out voters and President Bush won the state by a margin of 5% (Baron
and Cohen, 2006).
In 2004 Colorado also had a highly contested United States Senate race to
replace retiring Republican Senator Campbell. The race was between Colorado
9


Attorney General Ken Salazar (D) and Coors Brewing mogul Pete Coors (R).
Analysts speculated that for Salazar to win in the 2000 pro-Bush state he would have
to attract Democratic voters outside of Denver and conservative voters in rural
regions of the state as well (Crowley, 2004). Salazar beat Coors 50%-46% (Davidson,
2004). The shift of Colorado from a Republican stronghold to a state on the verge of
Democratic control was strengthened as other races in the United States house
districts were won by Democrats in 2004 (Otto, 2006). The shift to a State Legislature
controlled by the Democrats continued in 2004 as the Democrats took control of the
Colorado Senate and maintained control of the House for the legislative session that
began in 2005, Table 1.3 (Otto, 2006).
Table 1.3 Control of the Colorado Governorship and Legislature 1950 2004
Colorado Senate a Colorado House
Governorb Democrats Republicans Democrats Republicans
1951c Thornton (R) 15 20 18 47
1955 Johnson (D) 15 20 29 36
1959 McNicols (D) 22 13 44 21
1963 Love (R) 15 20 24 41
1967 Love (R) 15 20 27 38
1971 Love (R) 14 21 27 38
1975 Lamm (D) 16 19 39 26
1979 Lamm (D) 13 22 27 38
1983 Lamm (D) 14 21 23 42
1987 Lamm (D) 10 25 25 40
1991 Romer (D) 12 23 27 38
1995 Romer (D) 16 19 24 41
1999 Owens (R) 15 20 25 40
2003 Owens (R) 17 18 28 37
a Colorado State Archives
b Otto, 2006
cNote: The year represents the first year of a governors tern. For example Thornton was elected in
November 1950 and took office in January 1951.
10


The Democratic political shift continued in Colorado into 2006 as a decrease
in traditional farming and mining jobs gave rise to social needs including: illegal
immigration, health care, transportation, and education (Walsh, 2006). According to
many analysts, the two most significant issues causing the political shift were illegal
immigration and gay marriage. During the regular 2006 Colorado legislative session,
in response to illegal immigration concerns, legislation was passed to include state
felonies for human trafficking and smuggling, requirements for employer verification
of citizenship status, and increased fines for counterfeit identification documents
(Washington, 2006). However, these pieces of legislation did not go far enough
according to Governor Owens. Governor Owens convened a special legislative
session to consider major illegal immigration reform. The Colorado State Legislature
considered denying illegal immigrants access to tax-payer funded services and
requirements for proof of citizenship to vote (Washington, 2006). Former Governor
Lamm and former Senator Hart, both Democrats, joined Governor Owens in favor of
more stringent illegal immigration policies and laws, while former Denver Mayor
Federico Pena (D) opposed a proposed ballot measure that would deny public
services to illegal immigrants (Steers, 2006a; Washington, 2006). The special session
enacted House Bill 1023, effective upon Governor Owenss signature, eliminating
access to public housing, unemployment insurance, and denied other tax-payer
funded services, including Medicaid and food stamps, to illegal immigrants (Jones,
2006).
11


Gay marriage and family values also contributed to the shifting political
divisions in Colorado. A United States Constitutional Amendment to define marriage
as the union of a man and a woman, was proposed by Colorado Congresswoman
Marilyn Musgrave (R) in 2003, further uniting Republicans. Governor Owens, who
was also at the time the co-chair of the Republican Platform Committee, was a
proponent of the anti-gay marriage issue. However, his marital problems weakened
the Republican pro-family values platform in Colorado (Cohen, 2004). Colorado
Congresswoman Diana DeGette (D) said, the gay marriage issue turns off the many
moderates who have moved there [to Colorado] with a real Libertarian streak...They
are fiscal conservatives and social moderates who feel abandoned by this
administration (Cohen, 2004, p. 2527). The sentiment of a liberal Colorado was
echoed by Howard Dean, former Democratic presidential candidate and Democratic
National Committee Chairman, who stated,
The Republican Party has become the big-brother party, and this is a very libertarian
part of the country... [Westerners] dont believe its the governments place to tell
them what to do with their personal lives, and this is a government that specializes in
telling people what to do with their personal lives. (Walsh, 2006, p. 2)
In 2006 the gay marriage issue was on the ballot as Colorado Constitutional
Amendment 43 to ban gay marriage, and as Referendum I to allow domestic
partnerships for homosexual couples (Miller, 2006).
While numerous races for United States Congressional seats and Colorado
Legislative seats were highly contested, the 2006 ballot issues also represented a shift
from Republican stronghold to Democratic control. The ballot was composed of
seven Constitutional Amendments including citizen petitions, education expenditures,
12


judge term-limits, ethics in government, minimum wage, a ban on gay marriage, and
legal possession of marijuana. The seven ballot referenda included property taxes for
disabled veterans, recall timelines, obsolete constitutional provisions, domestic
partnerships, school expenditures, and two illegal immigration issues (Colorado
Legislative Council, 2006). In conclusion, a term-limited Republican Governor,
shifting political opinions in Colorado, and numerous ballot initiatives and
Constitutional Amendments make Colorados 2006 gubernatorial election an ideal
case study for the functional theory of political campaign discourse.
Professional Significance
Presidential campaigns from 1948 to 2004 have been analyzed for trends and
shifts in rhetoric using the functional theory of political campaign discourse. Similar
analyses have not been extensively completed on state level campaigns. The purpose
of this study is to apply the functional theory of political campaign discourse to
gubernatorial campaigns, specifically the 2006 Colorado gubernatorial campaign.
The functional theory of political campaign discourse predicts that acclaims
will be more frequent than attacks, which will be more frequent than defenses
(Benoit, 2004). Policy related utterances are predicted to be more frequent than
character utterances (Benoit, 2004). These predictions are predominantly based on the
analysis of presidential campaigns, and this study tested these hypotheses and others
in the context of a gubernatorial election, as described in Chapter 3, Methodology.
13


This study is important to the field of political rhetoric because it seeks to
determine if the functional theory of political campaign discourse can assist in
constructing and deconstructing political campaigns at levels of elected office beyond
the presidency. According to Hart and Childers (2005), the analysis of political
campaigns, specifically presidential campaigns, is significant not only because it
gives us insight into the candidates, but because they can shed light into the
institution of the presidency, the evolving nature of political campaigns, the
sensitivities or blindness of the electorate (and the mass media), the changing fabric
of political ideals in the United States and perhaps more important, the nature of the
American political culture (p. 184). Analysis of the 2006 Colorado gubernatorial
campaign can provide insight in a similar fashion. Finally, this study can shed light on
the phenomenon of the shifting Colorado political landscape as described in the
section titled Background of Colorado Politics.
Overview of Methodology
This study evaluated the efficacy of the functional theory of political
campaign discourse using candidate-generated newspaper articles from the general
campaign of the 2006 Colorado gubernatorial election. The basic structure of the
methodology allowed for the examination of the frequency of acclaims, attacks, and
defenses on character and policy issues.
The methodology for this study was based on a content analysis of candidate-
generated newspaper articles published during the 2006 general campaign cycle from
14


August ninth through November seventh, 2006. The data set, composed of candidate-
generated articles, was collected by conducting a daily search of Colorado
newspapers using the Lexis Nexus Academic Universe. While numerous other studies
of presidential campaigns have analyzed candidate speeches, debates, and television
advertisements, this study did not include these other forms of campaign rhetoric.
They were not included due to availability of transcripts and time and resource
constraints. In addition, since this is an initial study it was important to analyze the
printed media, the newspaper, which reaches numerous voters on a daily basis. It
should also be noted that a candidates ideas of government and governance are
presented by numerous special interest groups, including 527 organizations, during a
campaign. For the purposes of this study, campaign messages presented by special
interest groups were excluded from the data set, because candidates cannot control
the content of this type of campaign message and this studys analysis was limited to
candidate-generated discourse.
Prior to coding the candidate-generated articles, the articles were divided into
themes in which a candidate expressed a complete idea about government and
governance. In addition, prior to coding the coders were trained on the functions and
issues as described by the functional theory of political campaign discourse, provided
a list of policy topics and key terms, and given coded examples and the coding
procedure. Finally, the coders completed tests to evaluate intercoder reliability.
The themes were coded for source, edition, date of publication, day of the
week, and week of the campaign. The candidate themes were also coded for function
15


(acclaim, attack, or defense), issue (character or policy), and policy topic. The policy
topics were predetermined by reviewing previously tested policy topics which were
amended further to reflect state issues specific to Colorado based on a review of the
candidates campaign websites and articles discussing voter concerns. The policy
topics included crime/drugs, economy/jobs, education, energy, environment, fiscal
responsibility, health care, immigration, morals/family values, transportation, water,
and other. Policy topics and key words can be found in Appendix A. After the articles
were coded, the data was entered into an SPSS database for analysis. The analysis
provided descriptive data as provided in Chapter 4, The Results of the Study.
Chapter 2, Literature Review, includes a theoretical review of the functional
theory of political campaign discourse and an empirical review of the research
conducted based on the theory. The third chapter includes an extensive description of
the studys methodology. The findings and a brief summery of the results are
presented in Chapter 4, The Results of the Study. Finally, Chapter 5, Summary and
Discussion, includes a discussion of the results. This final chapter also includes
conclusions and recommendations for future research.
16


CHAPTER 2
LITERATURE REVIEW
A growing body of research exists on the evaluation of presidential campaigns
based on the functional theory of political campaign discourse. This chapter explores
the theoretical basis of the functional theory of political campaign discourse as it has
been used to analyze presidential campaigns from 1948-2004. Acclaims, attacks, and
defenses found in campaign rhetoric will be examined as part of the theoretical
review, which will also include the use of character and policy issues. The chapter
will conclude by examining empirical research relevant to development of the
hypotheses posed by this study of the 2006 Colorado gubernatorial campaign.
New Approaches to Campaign Rhetorical Analysis
According to Pomper (1975), Many voters change their partisan choice from
one election to the next, and these changes are most closely related to their positions
on issues and their assessment of the abilities of the candidates (p. 10). Efforts to
understand how voters assess a candidate and their campaign rhetoric are evolving as
political campaigns are influenced by changing political, social, and economic
environments. The primary goal of political discourse, to convince the electorate one
candidate is preferable to another, has not changed campaign strategies; however, the
expectations of political of campaigns have evolved. The new expectations of
political campaigns are based on the new politic or polis which is a civic moral
17


principle with five basic criteria. The five criteria used to evaluate campaigns include
civic service, popular judgment, collectivization, access, and political accountability
(Gronbeck, 2003). The new politic is based on increased citizen participation through
active and continuous reevaluation of the government (Gronbeck, 2003). In response,
the new politic emphasizes active reevaluation of the electorates concerns by the
elected official.
The evolution of this new politic began with the introduction of a theory
called the genesis of rhetorical action by Edwin Black in 1965. According to Benoit
(2000) the genesis of rhetorical action provides a richer, more complete analysis of
discourse and rhetoric than traditional genre theory (p. 182). In addition, rhetorical
action emphasizes message characteristics as significant influences of discourse
(Benoit, 2004). The genesis of rhetorical action is rooted in the behavioral sciences,
and uses grounded scientific theory and research to analyze the impact of mass media
on political campaigns (Swanson, 1972). The genesis of rhetorical action is further
divided into macroscopic and microscopic issues. Macroscopic issues relate to
campaign rhetoric as they govern all campaign strategies (Swanson, 1972). In
contrast, the microscopic issues concern the specific tasks of campaigns to obtain
success (Swanson, 1972). As a result of these views of political analysis, the genesis
of rhetorical action and Gronbecks new politic, the functional theory of political
campaign discourse has developed as a framework to explain the interrelated
rhetorical contexts of political campaigns, issues, and processes.
18


Functional Theory of Political Campaign Discourse
Blacks genesis of rhetorical action is the basic framework for the functional
theory of political campaign discourse. As explained by Swanson (1972) the
functional analysis of a message would recognize the complex integrative nature of a
sophisticated political campaign and begin its study by attempting to identify the
function this particular message served in the broader strategy of the total campaign
(p. 39). The functional theory of political campaign discourse has been used in the
analysis of nomination acceptance addresses, presidential television spots, nominating
convention key note speeches, presidential debates, and media reports of these events
including newspaper articles (Benoit, 2000).
The theoretical framework is used to analyze campaign rhetoric which assists
researchers in understanding election outcomes including voting behavior and how
voters select candidates (Benoit & Harthcock, 1999). Therefore, the basic hypothesis
of the theory is that a relationship exists between campaign discourse and election
outcome (Benoit, 2003). Functional theory assumes that a voter selects the preferable
candidate based on salient criteria of the voter (Benoit, Blaney, & Pier, 1998). Using
the functional theory of political campaign discourse, campaign discourse is analyzed
by dividing statements into themes which are coded based on function and issue.
According to the theory, functions are acclaims, attacks, and defenses, and issues are
classified as either character or policy references. Themes, functions, and issues are
further described in the next three sections.
19


Themes
The theme is a unit of analysis used to evaluate political discourse. Benoit
(2004) applied Berelsons 1952 definition of a theme as a single assertion about a
subject (p. 6) and the smallest unit of discourse that is capable of expressing a
complete idea (Benoit, 2000a, p. 280). The use of a theme as a coding unit allows for
the discussions of a single idea, which may differ in length, ranging from a sentence
to a paragraph, to be analyzed in their entirely and weighted equally. According to
Amie and Benoit (2005), political statements coded as themes more accurately
express a candidates idea or meaning. Themes can have multiple functions -
acclaims, attacks, and defenses as long as they are concerned with the same subject
to express a complete idea (Amie and Benoit, 2005).
Function to Acclaim, Attack, and Defend
The function of campaign discourse is a basic level of evaluation and allows
candidates to differentiate themselves from each other (Benoit, 2000; Benoit, 2003).
Candidates use acclaims, attacks, and even defenses in effort to appear more desirable
than the opposing candidate. An acclaim is a statement of self-praise, positive
characteristics, and policy positions (Benoit, 2000). In the 1988 presidential campaign
Vice President George H. Bush (R) acclaimed, Over the past six years, eighteen
million jobs were created; interest rates were cut in half. Today, inflation is down,
taxes are down and the economy is strong (Benoit, 2000, p. 183). Bush used this
statement to call attention to improvements in the economy during his tenure as Vice
20


President from 1980 to 1988. Benoit and Harthcock (1999) also argue acclaims are
used more frequently in the utterances of abstract topics, goals, and ideas. For
example, in the 2000 presidential election Governor Bush (R) acclaimed,
I believe we need to encourage personal responsibility so people are accountable for
their actions. And I believe in government that is responsible to the people. Theres a
big difference in philosophy between my opponent and me. He trusts government; I
trust you. 1 trust you to invest some of your own social security money for higher
returns. I trust local people to run their own schools. I trust you with some of the
budget surplus. We should help people live their lives, and not run them. Im asking
for your vote. (Benoit, 2004, p. 11)
Finally, acclaims are traditionally the most frequent form of utterance (Benoit &
Harthcock, 1999).
An attack specifically questions and criticizes an opponents character or
policies (Benoit & Harthcock, 1999). In addition, attacks are an effort to make
another candidate appear less preferable thus increasing the attacking candidates
appeal (Benoit, 2000). For example, in 1996, President Clinton (D) attacked Senator
Bob Doles (R) opposition to the creation of the Department of Education and
proposed 15% tax cut (Benoit & Harthcock, 1999). Typically, incumbents are
attacked more frequently based on the current problems of the country, and because
they have a record concerning issues that voters consider important on a national
level (Petrocik, 1996). In 1960, Nixon (R) was considered the incumbent as the
current Vice President and he was attacked 123 times by Senator John F. Kennedy
(D) (Benoit & Harthcock, 1999).
Finally, candidates usually defend themselves and their policies in response to
attacks in effort to restore their desirability after an attack (Benoit & Harthcock,
1999). There are five types of defense. They are denial, defeasibility, shifting blame,
21


transcendence, and differentiation; of these, denials are used most frequently (Benoit
& Harthcock, 1999). However, a significant problem exists with the defense strategy.
In order to effectively defend, the defender must first remind voters of the attack
(Benoit & Harthcock, 1999). Then he will discuss his opponents issues thus
refraining from discussing his campaign issues (Benoit & Harthcock, 1999; Benoit,
2000). As a result the candidate and his defense are perceived as more reactive than
proactive, thus unfit for elected office. An example of a defense by President Bush in
the 1992 campaign is as follows,
My opponents say I spend too much time on foreign policy, as if it didnt matter that
school children once hid under their desks in drills to prepare for nuclear war. I saw
the chance to rid our childrens dreams of the nuclear nightmare, and I did. (Benoit,
2000, p. 184)
In this defense, Bush reminds voters of the attack, in which he was accused of
focusing too much time on foreign policy, before discussing one of Governor
Clintons issues, schools. As a result Bush appears to be more reactive than proactive.
Finally, the defense, as well as the acclaim and attack, can relate to past, present, or
future actions and positions.
Issue Policy and Character
In addition to the three functions, the functional theory of political campaign
discourse subdivides campaigns into two primary issues of discourse. They are policy
issues and character issues (Benoit, 2003). It is based on these two areas of discourse
that the candidates attempt to differentiate themselves from each other because
functional theory predicts that voters decide their vote based on the issue, either
policy or character, that is most important to them (Benoit, 2003). Therefore, if policy
22


issues are most important to the voter, then the candidate who discusses policy issues
most frequently is more likely to be elected. The opposite is also true; if voters deem
character issues to be most important, then the candidate who speaks of character
issues more frequently should win the election. While this hypothesis is not true for
elections of all levels of political office, it is true for national elections, specifically
presidential elections (Benoit, 2003).
According to the functional theory of political campaign discourse, character
and policy utterances are separate. Character is an important issue according to
Aristotle because, We believe good men more fully and more readily than others
(Benoit, 2004, p. 1). Character utterances are described as those that concern the
personal traits or abilities of the candidate (Benoit, 2003; Benoit & Harthcock, 1999).
Character qualities describe the sincerity, empathy, morality, and drive of the
candidate (Benoit, 2004). For example, in the 2000 campaign between Vice President
Gore (D) and Governor Bush, Bush continuously attacked Gores character.
Examples of these character attacks include, A1 Gore is bending the truth again, and
Why does A1 Gore say one thing when the truth is another (Benoit, 2003 p. 4).
In contrast, policy utterances are described as concerning government action
or inaction either in the past, present, or future (Benoit & Harthcock, 1999). Policy
utterances include problems or issues that can be attributed to government action or
inaction and a candidates views of government and governance. For example,
Governor Bushs stance for private school tuition vouchers and Vice President Gores
targeted tax cuts in the 2000 campaign (Benoit et ah, 2005) are examples of policy
23


issues. An example of a policy utterance by Governor Bush from the 2000 campaign
is,
I want to take one half of the surplus and dedicate it to social security, one quarter of
the surplus for important projects. And I want to send one quarter of the surplus back
to the people who pay the bills. I want everybody who pays taxes to have their tax
rates cut, and that stands in contrast to my worthy opponents plan which will
increase the size of government dramatically... (Benoit, 2003 p. 3)
This campaign statement includes references to Bushs political philosophy of smaller
government and basic governmental responsibilities to the citizens. The comment
also provides insight into his policy on the personal investment of the budget surplus.
The discussion of character and policy issues highlights one of the most
significant areas of concern of the functional theory of political campaign discourse,
the difficulty in separating character issues from policy issues. Candidates tend to
shift discussions of policy issues to character issues and vice versa, especially during
political discussions and debates. For example, in the 1972 presidential election
between President Nixon and Senator George McGovern (D), a campaign
advertisement by the Nixon campaign stated the following,
In 1964, Senator McGovern voted for the Gulf of Tonkin Resolution which
supported escalation of the war in Vietnam. Now he says he is against the war and
always has been. In 1965, McGovern said we should wage war rather than surrender
South Vietnam to communism. Three years later he said it would not be fatal to the
United States to have a communist government in South Vietnam. In 1967, he was
against unilateral withdrawal. Throughout the year he has proposed unconditional
amnesty for draft dodgers. But his running mate claims he proposed no such thing.
For the war, against the war. For amnesty, against amnesty. The question is, where
will he stand next? (Benoit, 2003 p. 6)
In this utterance, Nixon discussed the political stances of McGovern, as related to
Vietnam, communism, draft dodgers, and so forth. He also discussed character issues
related to the trust and believability of McGovern and his vice presidential running
mate. While the functional theory of political campaign discourse seeks to separate
24


character and policy, analysis of political campaigns show the two are extremely
interwoven. This area of concern related to the functional theory of political
campaign discourse is one of the theorys greatest strengths because it reminds
researchers that the new rhetoric of campaigns emphasizes the evaluation of both
policy and character issues in the context of the entire campaign.
Empirical Research
In effort to understand the interactions of acclaims, attacks, and defenses on
policy and character issues, we must understand the frequency with which
presidential campaign discourse has been studied and which mediums have been
analyzed. The functional theory of political campaign discourse has been evaluated in
the context of various forms of media including television advertisements (Benoit,
2000; Benoit, 2000a; Amie & Benoit, 2005), newspaper advertisements and World
Wide Web pages (Benoit, 2000a), direct mail advertisements (Benoit, 2000a; Benoit
& Stein, 2005), speeches (Benoit, 2000), and debates (Benoit & Harthcock, 1999).
Based on these analyses across acceptance addresses, television and direct mail
advertisements, keynote speeches, debates, and World Wide Web pages acclaims are
more frequent than attacks which are more frequent than defenses (Benoit, 2000;
Benoit & Harthcock, 1999; Benoit & Stein, 2005; Amie & Benoit, 2005; Benoit,
2000a). The results of multiple studies on acclaims, attacks, and defenses used in
presidential campaigns include findings that acclaims are more frequent than attacks,
which are more frequent than defenses (Benoit, 2004). In addition, incumbents
acclaim and defend more frequently than challengers (Benoit, 2003). An incumbents
25


use of the defense is based on a high frequency of attacks from the challenger (Benoit
& Harthcock, 1999; Benoit, 2000). Further, incumbents tend to acclaim past actions
and positions due to their elected position, while challengers attack past actions or
inactions of the incumbent (Benoit, 2000; Benoit & Harthcock, 1999). Further
analysis of the studies show a trend that policy is discussed more frequently than
character (Benoit & Harthcock, 1999; Benoit & Stein, 2005; Amie & Benoit, 2005)
Another interesting finding is that Democrats attack more on policy and less on
character in all forms of campaign utterances with the exception of acceptance
addresses (Benoit, 2004).
For example, in the 1960 presidential campaign debates between then Vice
President Nixon and Senator Kennedy, overall policy was discussed 78% of the time,
and the remaining 22% of the utterances were related to character (Benoit &
Harthcock, 1999). In comparison, in 1996 the campaign between then President
Clinton and Senator Dole, policy and character accounted for 66% and 25% of the
utterances, respectively (Benoit & Harthcock, 1999). While these two elections
clearly place a superior amount of significance on the discussion of policy issues over
character issues, a study conducted by Hansen and Benoit (2001) found that between
1952 and 1996 policy utterances accounted for 52% and character utterances
accounted for the remaining 48%. This relative balance between policy and character
is a more recent trend supported by Benoits (2003) finding that since 1976 public
opinion polls show policy is the most important criterion for vote choice.
26


A content analysis of television presidential advertisements from 1952-1996
showed Democrats discussed policy more frequently than Republicans, 71% to 62%,
respectively (Kaid & Johnston, 2001). This result was confirmed in a 2004 study of
primary and general election debates, television advertisements, and acceptance
speeches from 1948 to 2000. Benoit (2004) showed 67% of policy utterances were
attributed to Democrats as compared to 60% for Republicans. In comparison,
Republicans discussed character more frequently than Democrats, 40% to 33%,
respectively (Benoit, 2004). An analysis of presidential campaign messages from
1948 to 2000 showed election winners discussed policy issues more frequently than
character issues and the study also showed losers discussed character issues more
frequently (Benoit 2003).
In conclusion, the evolution of the functional theory of political campaign
discourse allows researchers to evaluate the impact of campaign discourse on voter
choice and election outcome. By conducting content analyses of presidential
campaigns from 1948 to 2004, researchers have concluded the following trends are
consistent, with few exceptions, in those campaigns,
Acclaims are more frequent than attacks, which are more frequent than
defenses (Benoit, 2000; Benoit & Harthcock, 1999; Benoit & Stein, 2005;
Amie & Benoit, 2005; Benoit, 2000a),
Democratic candidates discuss policy issues more frequently than
Republican candidates who discuss character issues more frequently
27


(Benoit and Harthcock, 1999; Kaid and Johnston, 2001; Benoit, 2004;
Benoit & Stein, 2005),
The winning candidate tends to discuss policy issues more frequently than
the losing candidate, who discusses character issues more frequently
(Benoit, 2003; Benoit & Stein, 2005).
These three conclusions based on fifty years of presidential campaigns are the
foundations of this studys analysis of the 2006 Colorado gubernatorial campaign.
28


CHAPTER 3
METHODOLOGY
In order to explore the efficacy of the functional theory of political campaign
discourse, this study evaluated candidate-generated campaign related articles
published in Colorado newspapers during the 2006 Colorado gubernatorial election.
The structure of the methodology allowed for the evaluation of the frequency of
acclaims, attacks, and defenses on policy and character issues in comparison to the
rhetorical trends of presidential campaigns. This chapter will explain the
methodology used to conduct the study. The sections of the chapter include the
research question and hypotheses, general perspective, research context, description
of data and collection, and data analysis methodology. The data analysis methodology
will include the procedures used to code for themes, function, issue, and topic.
Research Question and Hypotheses
Based on a comprehensive literature review, campaign rhetoric has been
analyzed retrospectively post election; therefore, this analysis was completed in a
similar manner. This initial research on the 2006 Colorado gubernatorial campaign
will investigate one research question, based on the functional theory of political
campaign discourse analysis of presidential campaigns,
. Do gubernatorial campaigns follow the same trends as presidential
campaigns as described by the functional theory of political campaign
discourse?
29


The research question will be evaluated based on four hypotheses, which are based on
the trends seen in presidential campaigns. They are,
HI: Acclaims will be more frequent than attacks which will be more
frequent than defenses.
H2: The Democratic gubernatorial candidate will discuss policy issues
more frequently than the Republican candidate; the Republican
gubernatorial candidate will discuss character issues more frequently than
the Democratic candidate.
H3: The candidate who is behind in voter polling data will attack the
character and policies of the leading candidate more frequently.
H4: The candidate making the most frequent references to policy issues
relative to character issues in the 2006 Colorado gubernatorial election
will be most likely to win the election.
General Perspective
The research method used to complete this study was designed as non-
experimental and cross-sectional. A non-experimental design was appropriate
because the analysis did not require a treatment group. The choice not to use a
treatment group was appropriate for this study because this study explored the
rhetoric of the candidates, not the response of individual constituents. Non-
experimental studies are subject to internal threats to validity because social research
is conducted in a constantly changing environment that is affected by economic,
political, and social events. Threats to internal validity included history, maturation,
statistical regression, selection, experimental mortality, testing, instrumentation, and
design contamination (OSullivan, Rassel, and Verner, 2003). History and maturation
could not be controlled by the researcher in this study. Examples of history and
30


maturation include external factors such as local, state, national, and international
events that may have affected shifts in campaign rhetoric.
The researcher was able to control for the threats to internal validity which
included data selection and the methods used to analyze the data. The data collection
methodology outlined in this chapter allowed the researcher to control for selection
by initially including all newspaper articles relevant to the campaign in the study. The
news articles were reviewed for content to determine whether or not they were
appropriate for the data set based on the criteria outlined in the section titled,
Description of data and collection. The researcher controlled the coding of the data by
evaluating inter-coder reliability prior to the coding of actual data sets and using a
code sheet for the articles. Statistical regression, experimental morality, and design
contamination were not threats to internal validity since the study did not include a
treatment group.
A cross-sectional study design was appropriate to collect data on multiple
relevant variables at once. The cross-sectional design allowed for the collection and
analysis of numerous variables, including: news source and edition, date of
publication and day of the week, function, who the function was attributed to, the
subject of the function, issue references, and policy topic. The cross-sectional
research design was applied to analyze the frequency of acclaims, attacks, and
defenses for both the Republican and Democratic candidates.
31


Research Context
The data used for this study consisted of articles published in Colorado
newspapers from the 2006 Colorado general election, August ninth through
November seventh 2006. While candidates use numerous other forms of media -
television and direct mail advertisements, debates, speeches, and more recently World
Wide Web pages to communicate their concepts of government and governance,
this initial study only evaluated newspaper articles. Newspaper articles were selected
because they are representative of the dynamic nature of political campaigns.
Robinson and Davis (1990) comparison of newspaper articles and television news
reports showed newspapers to be more in-depth. And according to multiple studies
cited by Brians and Wattenberg (1996) newspaper readers are able to distinguish
differences between candidates. Similarly, newspaper articles provide voters with
significant candidate knowledge as described by Hollihan, for national political
news coverage, the most thorough, comprehensive, and substantive information
regarding political campaigns, political issues, and public policies is available to
readers of comprehensive large city daily papers (Benoit et al., 2005, p. 356).
Finally, according to Cohen (1963), the press may not be successful much of the
time in telling people what to think, but it is stunningly successful in telling its
readers what to think about (p. 13). In addition newspaper articles were readily
available while transcripts of debates were not. The selection of newspaper articles
was also appropriate based on limited time and resources to complete the study.
32


In recent years numerous stakeholder groups, including so-called "527"
organizations (so named for that section of the Internal Revenue Code which defines
this category of political advocacy organization), have increasingly become involved
in the political landscape by campaigning for or against candidates through the use of
attack advertisements on television. While these groups are required to be
independent of the candidates and the candidates typically denounce the negative
campaign tactics, the candidates can benefit from the negative campaigning.
However, since the candidates do not have control over these types of campaign
tactics, they were specifically excluded form this study of candidate-generated
campaign discourse. Therefore, this study was limited to candidate-generated
newspaper articles that were directly attributed to either Bob Beauprezs (R) or Bill
Ritters (D) campaign. This selection of appropriate data has been used extensively
by Petrocik (1996); Benoit & Harthcock (1999); Benoit (2000, 2000a, 2003, & 2004);
Petrocik, Benoit, & Hansen (2003/4); Benoit et al. (2005); and Amie & Benoit
(2005).
Description of Data and Collection
The data set analyzed in this study was composed of candidate-generated
stories. As previously stated the articles represent the general campaign cycle from
August ninth to November seventh, 2006. A daily search of Colorado newspapers
using the Lexis-Nexus Academic Universe was completed to collect the candidate-
generated articles. As described by Petrocik (1996), candidate-generated stories
report the positions of candidates, including issues, problems, and policies. These
33


articles also can report speeches and candidate positions in order to identify problems,
issues, or policies (Petrocik et al., 2003/4). Petrocik et al. (2003/4) explained that the
stories can be positive if proposals or assessments are made. If the failures of the
opposing party of candidate are addressed the story is viewed as negative (Petrocik et
al., 2003/4). The positions of candidates can be reported by the candidates themselves
or their proxies. Proxies in a national campaign include the vice presidential
candidate and national party officials (Petrocik, 1996). For this study, proxies
included the candidates for Lieutenant Governor, campaign spokespersons, and
national party officials when speaking for their respective campaign. Newspaper
articles that focused on the third party gubernatorial candidates and editorials or
commentaries were excluded from this study. In addition articles focused on other
stakeholder groups, including 527 organizations, were excluded, since the candidates
cannot control their content, as previously described.
Benoit, Stein, and Hansen (2005) noted that in addition to the address of
policy and character issues by news coverage, newspapers articles address four other
major functions of campaign discourse. The horse race is the analysis of the standings
amongst candidates, voter reactions are responses to candidate statements, and other
voting information includes voting dates, advertising costs, donations, etc (Benoit et
al., 2005). These three functions were excluded from the study. The fourth function,
scandals or allegations of immoral or illegal wrongdoings were included in the data
sample. For example, the following statement by Bill Ritter, the Democratic
gubernatorial candidate, was included in the data set, If you are going to talk about
34


accountability.. .but youre utilizing information that can only be obtained illegally,
we need to know what your source is (Paulson, 2006b, p. N/A). Ritters statement
was in response to an accusation that Congressman Bob Beauprez, the Republican
gubernatorial candidate, made against him based on illegally obtained information
from a federal crime database.
The data was coded by date of publication; however, for analysis the data was
grouped into sets encompassing one week. The first cycle began on August ninth, and
therefore thirteen weeks of data were analyzed. The use of the data by divided week
in the campaign also allowed for the data to be compared to polling data to determine
how voter polls impacted the candidates rhetoric, as posed by hypothesis three.
Data Analysis Methodology
Based on previous studies, candidate-generated news stories in the data set
were analyzed in four phases identification of themes, coding for the function of the
theme, coding for issue (character or policy), and if the theme was coded for policy
issue, it was then coded for policy topic. This section will include the methodology
used to identify themes, functions, issues, and topics.
Identification of Themes
Prior to coding for function, issue, and topic, the news stories were divided
into themes, which are a unit of analysis used in the functional theory of political
campaign discourse. The theme as previously described is used to allow candidate
discourse that expresses a single subject to be coded to more accurately reflect the
35


candidates meaning. For example, the following theme, attributed to Beauprez,
illustrates a combination of attacks, acclaims, and defenses on both policy and
character issues related to abortion:
Ritter went too far when he said he would allow exceptions for fetal anomalies,
which could include children with Down syndrome, a common birth defect. Fetal
anomalies too? You're still pro-life too? I don't have a foot in both buckets and I
don't try to straddle a fence...If Roe v. Wade were overturned at the federal level and
a bill was put on [my] desk that protected the life of the mother, [I] would sign it.
(Beauprez as quoted in Paulson, 2006a, p. N/A)
This theme was coded as an attack on Ritters character and policies on morals/family
values as related to abortion and an acclaim of Beauprezs morals/family values
policies which more accurately represents the messages content than coding it for an
acclaim of either Beauprezs policy or an attack of Ritters of character or an attack of
Ritters policy.
The theme below illustrates an acclaim of fiscal responsibility and was
attributed to Beauprez and his proxy, campaign spokesman John Marshall,
Beauprez unveiled a fiscal accountability plan, calling on the state to float bonds
from tobacco settlement money to pay off debts and fund a new round of tax cuts for
businesses...Beauprez also called for re-examining the way the state rents office
space and purchases goods and services. Its a common-sense plan that keeps faith
with the voters. It stands in stark contrast to Bill spend it all Ritter. I think
fundamentally Bill Ritter is not a business-friendly candidate because he has never
walked in their shoes. Bob Beauprez has an organic understanding of business.
Thats what hes done all his life. (Steers, 2006d, p. 7A)
The use of themes allowed the example above to be coded as an acclaim of policy
related to fiscal responsibility that was attributed to Beauprez. It is also an acclaim of
Beauprezs character, and an attack of Ritters character, both attributed to Beauprez.
And finally, it is an attack on Ritters policies related to fiscal responsibility. The use
36


of the theme unit allows this example to be coded and weighted equally to represent
the complete intended meaning of the Beauprez campaign.
Coding for Function, Issue, and Topic
After the themes in the articles were identified, the themes were coded for
function acclaim, attack, or defense then by policy and character issue based on
the functional theory of political campaign discourse. For the purposes of this study,
acclaims were defined as positive statements. Attacks were defined as negative
statements; and refutations of attacks were defined as defenses. The rules used for
coding were as follows,
Acclaims portray the sponsoring candidate favorably,
Attacks portray the opposing candidate unfavorably, and
Defenses explicitly respond to a prior attack on the sponsoring candidate (Amie
& Benoit, 2005, p. 479).
The themes were then coded as having references to either character or policy
issues or both. Character references as defined by Amie and Benoit (2005) are
references to a candidates properties, abilities or attributes (p. 480) or personal
traits or abilities of the candidate (Benoit & Harthcock, 1999). In contrast, a policy
issue was defined as a reference to governmental action and problems amenable to
such action (Amie & Benoit, 2005, p. 480). Benoit and Harthcock (1999) defined
policy issues as issues concerning government action either in the past, present, or
future.
Finally, the themes were coded for policy topics. The topics used were based
on the scheme developed by Bundge and Farlie in 1983. Their scheme was modified
37


by Petrocik in 1996 to reflect characteristics of the American political system that
were not originally evaluated by Bundge and Farlie. The scheme as developed by
Petrocik in his 1996 study, Issue ownership in presidential elections, with a 1980
case study (p. 848) is as follows,
Big government References to government taxing and spending (not tied to a
specific welfare policy) or regulation of the economy,
Civil liberties Rights of accused, freedom of speech, etc.,
Civil and social order Crime, life styles, traditional values, unconvertible
groups or movements, etc.,
Civil rights References to ethnic or racial matters or groups,
Defense spending and policy References to the defense budget, weapons
systems, or military policy. Does not include military treaties or aid,
Economy Any reference to the economy or its condition,
Farmers and agriculture References to activities on behalf of farmers,
agriculture, or rural groups,
Foreign relations References to relations with other countries including
military treaties and aid,
Government functioning Any reference to government corruption or failures,
not classified elsewhere by substitutive topic,
Organized labor Any references to unions or labor,
Social class and group relations References to the needs of interest of groups in
relation to other groups, e.g. middle class vs. upper class, big vs. small business,
old vs. young, etc.,
Social welfare and spending References to government social welfare activity
that is not group specific,
Women Any reference to women, womens rights, sex equality, and
All other mentions.
The policy issues developed by Bundge and Farlie (1983) and amended by Petrocik
(1996) are based on national level political campaigns and issues. Petrociks scheme
appears to have been further amended by Hansen and Benoit (2001). Their scheme
included crime/drugs, education, foreign affairs, health care, Medicare/prescription
38


drugs, morals/family values, social security, and taxes/tax cuts (Hansen and Benoit,
2001).
The policy scheme used in this study was based on the 2001 Hansen and
Benoit list that was further amended to include state-related policy topics. The
researcher had intended to use Colorado polling data related to voter concerns to
determine the policy topics; however, this data was not available. The researcher
evaluated the candidates websites and news stories that reflected voters concerns to
determine the policy topics prior to coding the data set. The list of policy topics used
in this study included crime/drugs, education, economy/jobs, energy, environment,
fiscal responsibility, health care, immigration, morals/family values, transportation,
water, and other. The list of terms used by Hansen and Benoit (2001) was also
amended to include terms specific to Colorado. For example, terms for education
were amended to include the Colorado Student Assessment Program or CSAP and
Amendment 23. The topic of fiscal responsibility included Referendum C & D and
TABOR (Taxpayers Bill of Rights). A list of key terms for each of these categories
of policy topics is included in Appendix A, Key Words for Policy Topics.
The three examples that follow illustrate how the functions, issues, and topics
were coded in this data set.
1. We need to augment the water storage and we are going to have to build storage
both at high elevations and lower elevations because we must. Im going to get the
job done (Harmon, 2006, p. N/A). This example was coded as an acclaim on policy
related to water and an acclaim of character. Both themes were attributed to
Beauprez.
39


2. The hypocrisy of his position supporting Jesus Apodaca four years ago and
completely flip-flopping around this issue for political gain is shameful...And
perhaps the most classic Both Ways Bob moment weve seen this entire campaign,
and weve seen a lot (Couch, 2006b, p. B-01). This example was coded as an attack
on Beauprezs character by Ritters proxy, spokesman Evan Dreyer.
3. 1 dont regret opposing it (Referendum C). 1 dont deny that they needed the money,
they being the state. I would have created the short-term cash to meet immediate
obligations but fix the systematic problems that are still waiting to be fixed. Im not
king, Id be governor. I think its got to be a collaborative (Paulson, 2006d, p. N/A).
This example was coded as a defense of policy related to fiscal responsibility. This
statement was also coded as an acclaim of character related to leadership. Both the
defense and the acclaim were attributed to Beauprez.
It is important to note that the use of themes allowed for the first and third examples
to be coded twice. The first was coded as a policy acclaim and character acclaim and
the third was coded once as a defense and once as an acclaim.
CODING METHODOLOGY
Prior to coding the actual data set, the coders were given instructions for
coding, coding samples, and a test to ensure inter-coder reliability and validity. The
practice samples were composed of candidate-generated news stories from the 2006
seventh congressional district race between Democrat Ed Perlmutter and Republican
Rick ODonnell. Intercoder reliability was calculated by comparing the number of
coding agreements and disagreements between the coders and the researcher. The
calculation used variable A to equal the number of agreements where the coder
records the same response as the researcher, and variable D to equal the number of
disagreements when the coder recorded a different response than the researcher
(OSullivan et al., 2003, p. 114). The calculation is as follows,
Reliability = [A/ (A+D)] 100%
40


The reliability for coder one was 93% and coder two was 90%; therefore, both coders
were determined to be reliable. Finally, an SPSS database was created to manage the
results of the coding. The SPSS database was also used to run the descriptive data
analysis for each hypothesis, which will be presented in the chapter to follow.
This chapter has explained the methodology used in the quantitative analysis
of the rhetorical trends of the 2006 Colorado gubernatorial campaign, including the
research questions and hypotheses, general perspective, and research context. Finally,
the methods for data collection, coding, and analysis were presented. The results
obtained using the described methodology will be presented in Chapter 4, The Results
of the Study.
41


CHAPTER 4
THE RESULTS OF THE STUDY
As stated in Chapter 1, this study examined the rhetorical trends of the 2006
Colorado gubernatorial campaign to determine if gubernatorial campaigns follow the
same rhetorical trends as presidential campaigns described by the functional theory of
political campaign discourse. The first section of this chapter presents descriptive data
for the set of data collected for analysis. The remainder of the chapter is organized in
terms of the research question and four hypotheses posed in Chapter 3. First the
frequencies of acclaims, attacks, and defenses are presented. Then the frequencies of
character and policy issues relative to the Democratic and Republican candidates are
examined. The weekly frequencies of attacks are compared to polling data. And
finally the frequencies of character and policy mentions by candidate are compared to
the election outcome.
Data Set Descriptives
The data set used in this study was composed of candidate-generated
newspaper stories from the Denver Post, Rocky Mountain News, Pueblo Chiefton,
Boulder Daily Camera, and the Associated Press. The news stories from the
Associated Press were included in the study because a number of Colorados smaller
newspapers, the Daily Sentinel of Grand Junction and Fort Collins Coloradoan for
example, as well as the Denver Post and Rocky Mountain News print the Associated
Press stories. In total 328 articles were examined and 150 were coded for analysis,
42


Table 4.1 Articles by yveek of the 2006 Colorado gubernatorial campaign
Campaign week Associated Press Boulder Daily Camera Denver Post Pueblo Chiefton Rocky Mountain News Total articles
8/09/06-8/15/06 4 0 2 0 6 12
8/16/06-8/22/06 1 1 4 0 4 10
8/23/06-8/29/06 4 0 6 2 2 14
8/30/06-9/05/06 1 0 5 1 1 8
9/06/06-9/12/06 2 1 2 0 4 9
9/13/06-9/19/06 1 0 4 1 4 10
9/20/06-9/26/06 0 0 1 0 2 3
9/27/06-10/3/06 4 0 3 0 4 II
10/04/06-10/10/06 1 0 11 0 4 16
10/11/06-10/17/06 4 0 3 2 6 15
10/18/06-10/24/06 3 0 9 4 6 22
10/25/06-10/31/06 5 0 2 1 3 11
11/01/06-11/07/06 1 0 3 1 4 9
Total 31 2 55 12 50 150
see Table 4.1. The reduction of articles was based on the data collection methodology
described in Chapter 3; therefore, articles that reported on special interest groups and
their endorsement of a specific candidate, election day information, and articles that
only mentioned the names of the gubernatorial candidates were excluded. In addition,
articles that discussed campaign fundraising were excluded. For example, Beauprez
was frequently mentioned in articles that reported on the seventh congressional
district race between Ed Perlmutter (D) and Rick ODonnell (R), because the winner
would replace Beauprez as the districts representative.
The candidate-generated articles were subdivided into themes, in which a
candidate expressed a complete idea about government or governance. In total 470
themes were coded, and 50.6% of the themes were attributed to Beauprez and 49.4%
were attributed to Ritter, see Table 4.2. The number of themes per week varied
substantially over the course of the campaign, from sixty-one in the first week of the
campaign, August ninth through August fifteenth, compared to only five during the
43


Table 4.2 Themes attributed to the candidates by week of the 2006 Colorado
Theme is attributed to
Campaign week Beauprez Ritter Total number of themes
8/09/06-8/15/06 29 32 61
8/16/06-8/22/06 24 23 47
8/23/06-8/29/06 28 20 48
8/30/06-9/05/06 9 6 15
9/06/06-9/12/06 15 18 33
9/13/06-9/19/06 14 23 37
9/20/06-9/26/06 3 2 5
9/27/06-10/3/06 16 12 28
10/04/06-10/10/06 25 26 51
10/11/06-10/17/06 31 26 57
10/18/06-10/24/06 16 22 38
10/25/06-10/31/06 17 15 32
11/01/06-11/07/06 11 7 18
Total 238 (50.6%) 232 (49.4%) 470
seventh week, September twentieth thought September twenty-sixth. The newspapers
from the seventh week of the campaign were reexamined based on this finding. The
election related coverage during that particular week focused on the ballot measures,
role of the 527 organizations, voting machine problems, and changes in campaign
staff, all of which were specifically excluded from this study.
Of the 470 themes that were evaluated forty-nine were attributed to the
candidates proxy, which was the campaign spokesperson in all cases. Twenty-six
themes were attributed to Ritter campaign spokesman, Evan Dreyer, and twenty-three
were attributed to Beauprez spokesman, John Marshall.
44


Results
The evaluation of the 2006 Colorado gubernatorial campaign was on based a
single research question. This section will present the results by the four hypotheses
that were posed to evaluate whether or not gubernatorial campaigns follow the same
trends as presidential campaigns as described by the functional theory of political
campaign discourse.
Gubernatorial Campaign Trends
The four hypotheses predicted that gubernatorial campaign trends are similar
to the rhetorical trends of presidential campaigns. The first hypothesis predicted that
acclaims would be more frequent than attacks, which would be more frequent than
defenses. Evaluation of the data showed this relationship did occur; acclaims
accounted for 60.0% of the utterances, attacks accounted for 28.5%, and defenses
accounted for 11.6%, see Table 4.3. It should be noted that the total number of
functions, 526, is greater than the total number of coded themes, 470. This is because
each theme can represent multiple functions, as described in the literature review and
methodology chapters.
For example, Ritter acclaimed his record on crime during his tenure as
Denvers district attorney, "We put over 12,000 people in prison...We saw the most
dramatic decrease in violent crime from 1993 to 2000 that the city ever had, down to
numbers not seen since the 1960s.. .I'm proud of my record and proud of the work of
the people in that office with me" (Lindsay, 2006, p. 14A). In comparison, the
45


Table 4.3 Acclaims, attacks, and defenses by week of the 2006 Colorado gubernatorial
campaign_____________________________________________________________________
Campaign week Acclaim Section function Attack Defense Total functions
8/09/06-8/15/06 50 15 5 70
8/16/06-8/22/06 34 13 3 50
8/23/06-8/29/06 27 15 13 55
8/30/06-9/05/06 8 3 6 17
9/06/06-9/12/06 26 10 2 38
9/13/06-9/19/06 31 11 0 42
9/20/06-9/26/06 4 1 0 5
9/27/06-10/3/06 16 9 7 32
10/04/06-10/10/06 28 23 5 56
10/11/06-10/17/06 43 15 7 65
10/18/06-10/24/06 16 21 5 42
10/25/06-10/31/06 23 5 5 33
11/01/06-11/07/06 9 9 3 21
Total 315 (60.0%) 150(28.5%) 61 (11.6%) 526
following attack was made by Beauprez on Ritters character and lack of
governmental leadership experience, I dont make light of the career Bills had, but
he went to law school, became a lawyer and has spent his entire career in the city and
county of Denver in the DAs office. Bills never run a business, how can he run the
business of the state (Glathright, 2006a, p. 4A).
Ritter attacked Beauprezs congressional immigration record after
proclaiming his policy to address the illegal immigration issue in Colorado stating,
I will find responsible, achievable and practical solutions to the illegal immigration
crisis gripping our country and our state, Washington has failed to secure our
boarders, prosecute companies that knowingly hire undocumented workers and
punish human smugglers and to live up to their responsibilities and reimburse states
for the enforcement and social-service tabs we are paying. (Glathright, 2006b, 18A).
While this statement by Ritter does not directly attack Beauprezs immigration voting
record, it does strongly imply that Beauprez, a member of Congress, has not
contributed to finding a solution for illegal immigration. In response, Beauprez
defended his voting record stating, "I don't think I would have changed a single vote,
46


but to suggest to anybody that I'm going to remember what every line, what every
section, every chapter and verse of the myriad pieces of legislation that comes across
any congressman's desk in a congressional session, frankly, the answer is no"
(Paulson, 2006c, p. N/A).
The second hypothesis predicted the Democratic candidate, Ritter, would discuss
policy issues more frequently than the Republican candidate, Beauprez; in addition,
the hypothesis predicted the Republican candidate would discuss character issues
more frequently than the Democratic candidate. Table 4.4 shows policy issues were
discussed in 76.0% of the utterances as compared to character utterances, 24.0%.
Policy issues were discussed more frequently in every week of the campaign with the
exception of the eleventh week of the campaign, October eighteenth through October
twenty-fourth. A closer review of the eleventh week revealed that much of the
coverage was related to Beauprezs use of illegally obtained information in an attack
advertisement against Ritter. In response, Ritter repeatedly questioned the character
Table 4.4 Policy versus character issues by week of the 2006 Colorado gubernatorial campaign
Campaign week Issue of reference Character Policy Total issues
8/09/06-8/15/06 12 53 65
8/16/06-8/22/06 8 43 51
8/23/06-8/29/06 14 43 57
8/30/06-9/05/06 6 12 18
9/06/06-9/12/06 7 28 35
9/13/06-9/19/06 10 33 43
9/20/06-9/26/06 0 5 5
9/27/06-10/3/06 10 20 30
10/04/06-10/10/06 11 45 56
10/11/06-10/17/06 9 50 59
10/18/06-10/24/06 25 17 42
10/25/06-10/31/06 2 30 32
11/01/06-11/07/06 10 12 22
Total 124 (24.0%) 391 (76.0%) 515
47


and accountability of Beauprez in light of the use of the illegally obtained information
for political gain. In return Beauprez attacked Ritters judgment and leadership as
Denvers district attorney related to his plea bargain with a suspected illegal
immigrant, who was accused of California sexual assault (Paulson, 2006c, p. N/A).
Table 4.5 compares the frequencies of acclaims, attacks, and defenses based
on character and policy issues. Overall policy acclaims were the most frequent form
of campaign utterance and denials of policy and character were the least frequent. The
same trend is seen in a comparison of utterances by candidate, as shown in Table 4.6.
Table 4.5 Comparison of function and issue references in the 2006 Colorado gubernatorial campaign
Acclaim Section function Attack Denial
Character 56 70 18
Policy 286 110 49
Table 4.6 Comparison of function and issue by candidate in the 2006 Colorado gubernatorial
campaign
Section function
Candidate Issue Acclaim Attack Denial
Beauprez Character 32 31 15
Policy 139 59 28
171 90 43
Ritter Character 24 39 3
Policy 147 51 21
171 90 24
48


Table 4.7 Policy and character utterances by candidate in the 2006 Colorado gubernatorial
campaign__________________________________________________________________________
Party Candidate Issue of reference Character Policy
Republican Beauprez 78 (26.5%) 216(73.5%)
Democrat Ritter 66 (23.2%) 219(76.8%)
Total 144 (25.0%) 435 (75.0%)
Closer analysis of the frequencies of policy and character issue utterances by
candidate showed that Ritter, the Democratic candidate, discussed policy issues more
frequently than Beauprez, 76.8% and 73.5%, respectively. In contrast, Beauprez
discussed character issues more frequently, 26.5%, compared to 23.2% of Ritters
character utterances, see Table 4.7. For example, Ritter stated his policy on creating a
twenty-first century workforce in Colorado by increasing education for skilled
technology workers through high speed internet access which would allow long-
distance learning, telemedicine, and telecommuting. In response, Beauprez discussed
his leadership skills in relationship to managing his familys dairy farm and bank
(Smith, 2006).
Further analysis of the policy topics discussed by both candidates showed
immigration, fiscal responsibility, economy/jobs, energy, and health care to be the
five most commonly discussed policy topics as shown in Table 4.8. It should be noted
that based on the methodology described in Chapter 3, each policy theme was
potentially coded for up to five policy issues; therefore the number of policy topics in
Table 4.8 is greater than the number of policy themes in Table 4.4.
49


Table 4.8 Policy issues by candidate in the 2006 Colorado gubernatorial campaign
Policy issue Beauprez Ritter Total Percentage
Crime/drugs 23 14 37 6.2 %
Economy/jobs 28 39 67 11.2 %
Education 16 23 39 6.5 %
Energy 32 34 66 11.0%
Environment 17 22 39 6.5 %
Fiscal responsibility 38 38 76 12.7%
Health care 26 40 66 11.0%
Immigration 44 40 84 14.0%
Morals/family values 20 23 43 7.2 %
Transportation 13 19 32 5.4 %
Water 18 17 35 5.9 %
Other 10 4 14 2.3 %
Total 285 (47.7 %) 313 (52.3%) 598
Many of the references to illegal immigration occurred as Beauprez accused
Ritter of being soft on illegal immigrants during his tenure as Denvers district
attorney. These accusations were returned by Ritter as he accused Beauprez if not
taking action in Congress to enact comprehensive immigration reform. In addition, as
shown in Table 4.9, the candidates most frequently proclaimed their records and
attacked the opposing candidates records in the discussion of immigration.
The second most frequent policy topic discussed was fiscal responsibility in
the form of acclaims and attacks, Table 4.9. In contrast to illegal immigration, Ritter
repeatedly attacked Beauprez for his opposition to Referendum C in 2004, which
allowed the state to keep several billion dollars worth of taxes to fund state services.
Ritter consistently reminded voters he was in favor or Referendum C to fund health
care and education. This forced Beauprez to defend his opposition by suggesting the
state could float bonds from the tobacco settlement money to pay off debts and fund
tax cuts for businesses (Steers, 2006d, p. 7A).
50


Table 4.9 Comparison offunction and policy issue in the 2006 Colorado gubernatorial
campaign_____________________________________________________________________
Policy Acclaim Section function Attack Denial Total
Crime/drugs 10 19 12 41
Economy/jobs 50 19 7 76
Education 36 6 0 42
Energy 51 16 6 73
Environment 30 7 4 41
Fiscal responsibility 54 26 11 91
Health care 61 12 3 76
Immigration 35 47 19 101
Morals/family values 33 10 6 49
Transportation 29 10 0 39
Water 34 3 0 37
Other 11 5 2 18
Total 434 180 70 684
The Colorado economy was the third most discussed issue during the
campaign. Both candidates discussed how they would work to revive the economy
through the various mechanisms of economic development. For example, during a
debate, both candidates discussed the minimum wage amendment on the ballot. Ritter
proclaimed his support of the constitutional amendment that would increase the
minimum wage from $5.15 per hour to $6.85 per hour with adjustments for inflation
(Couch 2006a). In comparison, Beauprez proclaimed he was against such a measure
that would allow the minimum wage to be on autopilot in the state constitution
(Couch, 2006a, p. B-02).
The fourth major policy issue in the campaign was energy and was the third
most proclaimed issue, Table 4.9. Before stating his policies on energy, Ritter
criticized Beauprezs voting record to pay royalty subsidies to oil-shale developers,
allowing Clean Water Act and the Safe Drinking Water Act exemptions for oil and
gas companies, as well as supporting the $20 million dollar budget cut from the
51


National Renewable Energy Lab in Golden, Colorado (Lofholm, 2006). Ritter argued
for progressive energy policies including wind power, green bundling, and biodiesel
(Lofholm, 2006).
Finally, both candidates argued for health care reform in 11.0% of the policy
utterances and the most frequently acclaimed issue, Table 4.9. The candidates offered
specific policy options for increasing the number of insured Coloradoans. When
asked how they would reform Medicaid, the candidates responded,
Ritter: Certainly the rate of increase in Medicaid spending has outstripped every
other category. So it is a big budgetary pressure. There are things you can do to
address Medicaid spending, to address prescription drug purchasing that Owens
decided not to do. But you cant do it piecemeal. Look at what Massachusetts did:
They went to the federal government, received a waiver, then they combined all their
state dollar spending Medicaid, state employees and money for businesses, and
they said, we're going to have a statewide health plan. They didnt do it piece by
piece by piece.
Beauprez: We need to make Medicaid consumer-driven. And South Carolina and
Florida provide some great models. It's what most of us, who have an employer-
based system, use: a defined contribution. They tell patients, "Here's your
contribution based on your needs." Then, suppose your defined contribution is
$1,000. You can use it to buy health insurance. But if you adopt a wellness program,
a preventive care program, or get into job training, that's worth some more. And if
you end up with some change in your pocket, we'll split it with you. Now we're
really changing the paradigm. (Brand, 2006, p. 4C).
The remaining 40% of the policy utterances were attributed to crime/drugs,
economy/jobs, environment, morals and family values, transportation, water, and
other miscellaneous issues. Miscellaneous issues included affordable housing,
government ethics and accountability, and two references to the war in Iraq.
The third hypothesis predicted the candidate behind in voter polling data
would attack the leading candidates character and policies more frequently. Overall,
policy acclaims were the most frequent utterance for each candidate, followed by
policy attacks, and character attacks as previously shown in Table 4.6. Table 4.10
52


Table 4.10 Attacks versus polling data in the 2006 Colorado gubernatorial campaign
Attack Rasmussen polla Survey USA pollb
Beauprez Ritter Beauprez Ritter Beauprez Ritter
8/09/06-8/15/06 6 9 39% 48%
8/16/06-8/22/06 7 6 40% 50%
8/23/06-8/29/06 8 7
8/30/06-9/05/06 3 0
9/06/06-9/12/06 3 7
9/13/06-9/19/06 2 9 34% 50%
9/20/06-9/26/06 0 1
9/27/06-10/3/06 5 4 38% 55%
10/04/06-10/10/06 13 10
10/11/06-10/17/06 6 9
10/18/06-10/24/06 9 12 39% 51% 38% 56%
10/25/06-10/31/06 4 1
11/01/06-11/07/06 7 2 40% 50%
a Rasmussen, 2006
b Survey USA, 2006
shows a comparison of attacks for each candidate by week of the campaign. The table
also illustrates two voter polls, with their respective polling data. Based on both the
Rasmussen and Survey USA polls, Ritter led throughout the general election. Ritter,
attacked more frequently in weeks one, six, and eleven. Further review of these three
weeks of candidate-generated themes shows Ritter attacked Beauprez on the issues of
education, the economy, environment, fiscal responsibility, immigration, and
transportation. Ritter also attacked Beauprezs character based on his selection of
Janet Rowland (R) as his running mate for Lieutenant Governor, after her comments
comparing homosexuality to bestiality (Steers, 2006b). During the same these same
three weeks, Beauprez attacked Ritters stance on immigration. He also attacked
Ritters pro-life abortion stance in which he believes there should be exceptions for
rape, incest, and fetal anomalies (Paulson, 2006a).
53


In comparison, in three of the four weeks in which the Survey USA polls were
available, Beauprez attacked Ritter more frequently. Review of October eighteenth
through October twenty-fourth during which Ritter attacked more frequently, showed
the focus of the attacks were on Beauprezs use of illegally obtained information from
a federal law enforcement database. The attacks by Ritter on Beauprez focused
Beauprezs character, If you are going to talk about accountability.. .but you are
utilizing information that can only be obtained illegally, we need to know what your
source is (Paulson, 2006b, p. N/A). In the last week of the campaign, November first
through November seventh, Beauprez attacked Ritters pro-life stance on abortion,
health care, transportation, global warming, fiscal responsibility, and character.
Beauprez, Ive become absolutely convinced that while Bill talks about his
Colorado Promise, its a plan-less promise. He keeps talking about invest, invest,
invest, and I wonder where all the money is going to come from to accomplish the
mission at hand (Glathright, 2006c, p. 27A).
Finally, hypothesis four predicted the candidate who made the most frequent
references to policy issues would win the election. Based on Table 4.7 Ritter, the
Democratic candidate, made 76.8% of his total utterances on policy issues as
compared to Beauprez, who had 73.5% of his utterances attributed to policy issues
(Table 4.7). Ritter won the election 57% to 40% (Davidson, 2006).
54


CHAPTER 5
SUMMARY AND DISCUSSION
This final chapter of The 2006 Colorado gubernatorial campaign: The efficacy
of the functional theory of political campaign discourse will serve as the studys
summation. The chapter will include a summary of the problem statement,
methodology, and results. A discussion of the results will be followed by
recommendations for future research.
Statement of the Problem
As explained in Chapter 1, numerous studies have been conducted on political
campaign discourse but they have focused on discourse at the presidential level of
elected office. This study of the 2006 Colorado gubernatorial campaign sought to
examine campaign discourse, to determine if gubernatorial campaigns have the same
rhetorical trends as presidential campaigns as described by the functional theory of
political campaign discourse.
Review of the Methodology
As a case study, this study evaluated the efficacy of the functional theory of
political campaign discourse using candidate-generated news stories from the 2006
Colorado general gubernatorial campaign. The basic structure of the methodology
allowed for the examination of the frequency of acclaims, attacks, and defenses on
55


character and policy issues. The methodology was based on a content analysis of
candidate-generated newspaper articles published during the 2006 general campaign
cycle from August ninth through November seventh, 2006. Prior to coding the
candidate-generated articles, the articles were divided into themes in which a
candidate expressed a complete idea about government and governance. The data set
was coded for source, edition, date of publication, and day of the week. The candidate
articles were also coded for function (acclaim, attack, or defense), issue (character or
policy), and policy topic. The policy topics used for the data set included crime/drugs,
education, economy/jobs, energy, environment, fiscal responsibility, health care,
immigration, morals/family values, transportation, water, and other. After the articles
were coded, the data was entered into a database for analysis in order to complete a
descriptive data analysis.
Summary of the Results
This study was centered on the exploration of a single research question. The
research question sought to determine if gubernatorial campaigns follow the same
rhetorical trends as presidential campaigns. Analysis of candidate-generated
newspaper articles showed the candidates made acclaims more frequently than
attacks, which were more frequent than defenses, Table 4.3. Second, the analysis
showed Ritter, the Democratic candidate, discussed policy issues more frequently
than Beauprez, the Republican candidate, Table 4.7. In addition, Beauprez was found
to discuss character issues more frequently than Ritter.
56


The third hypothesis was posed to determine if the candidate behind in polling
data attacked the leading candidates policies and character more frequently. Polling
data was available for six of the thirteen weeks of the general campaign and Ritter led
Beauprez in all of the polls; however, in three of the weeks Ritter attacked Beauprez
more frequently than Beauprez attacked Ritter. In three of the weeks, Beauprez
attacked more frequently. Based on the available data, hypothesis three was not
confirmed. Finally, the fourth hypothesis suggested the candidate who discussed
policy issues most frequently would win the election. Based on the results obtained
for hypothesis two and further analysis this hypothesis was proved to be correct, since
Ritter discussed policy issues more frequently than Beauprez, and Ritter won the
election.
Discussion of the Results
This study was completed to begin to fill the gap in research of the
functional theory of political campaign discourse by exploring the theoretical
frameworks applicability to gubernatorial campaigns. One of the initial
findings showed a balance of coverage between Beauprez and Ritter, which is
in contrast to the belief that newspapers and the media in general are biased or
slanted to either the Right or Left. Very few of the articles prior to identifying
themes reported on only Beauprez or Ritter. The balance of coverage became
more apparent as the themes were analyzed. This analysis of candidate-
generated news stories, showed 238 themes were attributed to Beauprez and
57


232 were attributed to Ritter, Table 4.2. In part this balance may be due to the
frequent comparisons of the candidates policy positions. In addition many of
the candidate themes reported on the gubernatorial debates and speeches
which allow for comparisons between candidates.
The results of this study show that the 2006 Colorado gubernatorial
campaign had similar rhetorical trends as compared to those seen in
presidential campaigns. The first finding that acclaims were more frequent
than attacks, which were more frequent than defenses, was expected, since the
use of attacks and defenses are viewed as negative campaigning and can cause
the candidates to look reactive, especially in the case of the defense. However,
the proportion of acclaims, attacks, and defenses was an interesting finding
that is also true of presidential campaigns. The low frequency of defenses seen
in Table 4.3 is lower than expected, which can be attributed to the rational of
the incumbency style.
Incumbency style as described by Trent and Friedenberg (2004) is, a
blend of both symbolic and pragmatic communication strategies designed to
make any candidate appear as both good enough for the office sought and
possessing the office (an assumed incumbency stance) (p. 81). Pragmatic
strategies are frequently used by candidates who do not hold the office of the
presidency (Trent & Friedenberg, 2004). Of the eleven pragmatic strategies
described by Trent and Friedenberg (2004) five were exhibited by Beauprez.
For example, Beauprez frequently acclaimed his Congressional record, which
58


according to Trent and Friedenberg (2004) creates pseudo events in which
increased media attention is given to the candidate as well as showing his
ability to manipulate important domestic concerns. Beauprez also used his
office as a Congressional Representative to gamer support from national
Republican Party leaders, including the President in addition to building his
credentials as a leader. Beauprez also emphasized his accomplishments by
proclaiming the success of his familys bank and his activities as a
Congressional Representative. Beauprez can be described as having exhibited
an incumbency style in the 2006 Colorado gubernatorial campaign based on
the use of these pragmatic strategies.
However, Trent and Friedenberg (2004) note there are four
disadvantages to exhibiting the incumbency style. They are the candidate must
run on his or her record, the public may blame them for problems, they must
campaign and do the job to which they were previously elected, and they
typically have higher expectations regarding being the front runner (Trent &
Friedenberg, 2004). Prior research on presidential campaigns shows that
incumbents acclaim past deeds and records, while challengers attack the
incumbents past deeds more frequently (Benoit, 1999). Based on Beauprezs
use of the incumbency style, Ritter was expected to attack Beauprez more
frequently based on Beauprezs Congressional record, which allowed the
public to blame Beauprez for policy problems. However this was not the case,
59


as shown in Table 4.6, both candidates attacked each others policies at the
same rate.
The policy topic of immigration is the most significant examples of the use of
the incumbent style. Beauprez should have had a slight advantage as a Congressional
Representative, especially on this kind of performance related policy issue. However,
this may have been to Beauprezs disadvantage based on Ritters use of Beauprezs
Congressional voting record related to immigration. While Ritter dealt with
immigration issues as Denvers district attorney, Beauprez has a Congressional voting
record of not addressing the issue, according to Ritter. Therefore on numerous
occasions Beauprez attacked Ritters plea bargain record, and instead of defending
himself, Ritter attacked Beauprezs lack of a policy solution while in Congress thus
providing the public with someone to blame for problems associated with illegal
immigration. This resulted in Beauprezs attempt to separate himself from
Washington politics and President Bush (R) and foregoing the incumbency style
strategy of gamering support from national party leaders and the President, by stating
he was frustrated by the inaction with the Bush administration and Congress (Steers,
2006c). Based on this series of events, additional research should examine whether or
not Beauprez, exhibiting the incumbency style, praised past deeds more frequently
and did Ritter, as the challenger, attack Beauprezs past deeds more frequently.
The rate with which the candidates attacked each other is also
interesting since previous research has shown newspapers to report negative
campaigning more frequently (Benoit et al., 2005). In addition, Graber found
60


in a review of newspaper campaign articles that the candidates character was
mentioned twice as frequently as their policies (Benoit et al., 2005). However,
this study showed policy issues were mentioned in 76.0% of the utterances as
compared to character utterances at 24%, Table 4.4. This finding can
substantiate the claim made by Hollihan that national papers are a
comprehensive source of learning for political campaigns (Benoit et al.,
2005). Another interesting finding is that a large majority of Ritters attacks
on Beauprez were related to Beauprezs policies until the latter weeks of the
campaign. Three weeks before the election Beauprez attacked Ritters record
as Denvers district attorney based on illegally acquired crime database
information. As a result Ritter attacked Beauprezs character more frequently.
Ritter used these character attacks to question Beauprezs honesty, leadership,
and fairness. This finding may be attributed to the rationale that character
attacks are considered to be of poor taste and negative campaigning, which is
not representative of the office for which the candidates were seeking.
However, as previously discussed Beauprezs use of the incumbency style
increased his vulnerability to policy attacks due to his frequent reference to his
Congressional record and his impact on domestic issues.
The functional theory of political campaign discourse proposes Democratic
candidates will discuss policy issues more frequently than the Republican candidate
who will discuss character issues more frequently. As shown in Table 4.7, Ritter
discussed policy issues more frequently than Beauprez. This may have been attributed
61


to Petrociks (1996) issue ownership theory and the idea that state and local issues
tend to be owned by Democrats. For example, education is owned or viewed as better
handled by Democrats and was discussed more frequently by Ritter, Table 4.8. The
issue of health care, which is traditionally a national issue and owned by Republicans,
was discussed more frequently by Ritter. The only national issue traditionally owned
by Republicans was immigration, and it was discussed more frequently by Beauprez,
Table 4.8. Immigration was also the issue the candidates most frequently attacked on,
Table 4.9.
One of the significant shifts in American politics is that the differences
between candidates are becoming less clear. For example, both Beauprez and
Ritter were pro-life, however Ritters policy on abortion has exceptions and as
a result this issue was discussed only briefly during the campaign. On a
broader scale the differentiation between candidates is causing shifts in
traditional issue ownership of policy issues. For policies that are typically
federal immigration and health care and owned by the Republicans, are
becoming local issues since many view the federal government as not
resolving the. Therefore these and other similar issues appear to be becoming
Democratic issues.
Since this study is specifically examining only the 2006 Colorado
gubernatorial campaign its generalizability is limited. The purpose of this study was
not to explore shifts in campaign rhetoric over multiple election cycles or in the larger
political context and therefore only the 2006 campaign was analyzed. The trends seen
62


in Colorados 2006 gubernatorial campaign may or may not be applicable to states
that are significantly more Democratic or Republican. This lack of generalizability
can be attributed to Colorados shifting political landscape. While the sample size for
this study is significant it only contains newspaper articles, which can be significantly
more negative than actual candidate statements. The coding of articles can possess
some reliability concerns; however this was limited by verifying intercoder reliability.
Recommendations for future research
The functional theory of political campaign discourse and the research results
reported, provide a solid basis for further research. Areas for future research include
the relationship between character and policy attacks, regional implications and
policy issue selection, issue ownership, the influence of the press and agenda setting,
and other forms of campaign media.
As previously discussed, one weakness of the functional theory of political
campaign discourse is the difficulty in separating utterances of policy and character.
However, this weakness is also one of the theorys greatest strengths, as a reminder
that campaigns are about character and policy. While the discussion of character
issues accounted for only 24.0% of the utterances as shown in Table 4.4, further
analysis should be conducted to determine if attacks on policy are accompanied by
attacks on character. Based on this research finding, it is suspected that the correlation
would be low.
63


While Colorado is representative of a large political shift, within itself
Colorado suggests a polarized political climate. In order to begin this evaluation,
analysis of other Colorado gubernatorial campaigns should be completed, such as
multiyear analyses similar to those conducted on presidential campaigns. In addition
multi-campaign year analyses of debates, speeches, and television, newspaper, and
direct mail advertisements should be conducted. In an effort to understand the larger
political shift, future research should include the presentation of policies based on the
regions of the state. For example, as described in the introductory chapter Colorados
politics are divided along the Rocky Mountains, eastern plains, and western slope
regions. Analysis should include candidate speeches and debates in these regions to
determine if candidates address policy issues differently in differing regions of the
state. For example, throughout the campaign the candidates discussed energy
options, in northern Colorado they talked about alternative forms of energy including
wind, on the eastern plains they discussed water, and on the western slope oil was a
major concern.
Further research on gubernatorial campaigns should include an exploration of
policy issues discussed relative to party ownership as described by Petrocik (1996) to
determine if the ownership of Republican issues crime, economy, fiscal
responsibility, health care, immigration, and moral responsibility differs on a state
and local level from the national level of ownership. For example, in this election the
traditionally Republican owned issues of the economy, health care, and moral
responsibilities were discussed more frequently by Ritter the Democratic candidate
64


than by Beauprez. In comparison the candidates discussed fiscal responsibility
proportionally, Table 4.8. Research should also be conducted to understand why
candidates tend to focus on issues owned by their respective party when issues are un-
owned. As a result researchers should investigate the impact on a candidate who does
not address their opponents issues.
This study examined candidate-generated newspaper stories which while
providing a significant amount of candidate information they can also be more
negative than actual candidate statements as well as covering the horse race, ballot
initiatives, and scandals (Benoit et al., 2005). Therefore additional forms of media
including speeches, debates, candidate websites, and television, direct mail, and news
paper advertisements should be analyzed for rhetorical trends. In addition, the role of
the press can directly influence campaign discourse and the perception of issue
ownership as related to presidential campaigns via its biases. While numerous studies
have examined the information published by the press, few if any have investigated
biases in the press. Another issue that arises is if the media focuses on specific issues
to how much of an impact does this have in influencing voters issue concerns and
priorities. In addition few if any studies investigated the accuracy with which
candidate messages are translated into print/news reports and the reliability of the
translations.
In conclusion this analysis of the 2006 Colorado gubernatorial campaign used
the functional theory of political campaign discourse to determine if the rhetorical
trends were similar to those of presidential campaigns. Acclaims were shown to be
65


more frequent than attacks, which were more frequent than defenses. Ritter, the
Democratic candidate, discussed policy more frequently than Beauprez, who
discussed character issues more frequently. And therefore, the analysis showed the
candidate who discussed policy issues most frequently, Ritter, won the election. As a
result of these findings, we can begin to understand the trends of gubernatorial
campaigns as described by the functional theory of political campaign discourse.
Further understanding of gubernatorial campaign trends can assist communicators in
addressing one of the most significant political problems of today described by Arthur
Hadley as reported by Ladd (1994),
Americas present problem...is an apathetic cross-pressured society with strong
feelings of political importance, where more and more people find their lives out of
control, believe in luck, and refrain from voting. These growing numbers of
refrainers hang over our democratic process like a bomb, ready to explode and
change the course of our history...For us, now an increase in voting is a sign of
political health (p. 3).
66


APPENDIX A: KEY WORDS FOR POLICY TOPICS
Crime/drugs Education Economy/jobs Energy Environment Fiscal responsibility
Addicts academic Aerospace Alternative Drought Anti-tax
Anti crime achievement Biomedicine energy Forests Budget
Assault gaps Businesses Amendment Global Earmark
Columbine after-school Economic 37 warming Federal funding
Convictions Amendment development Agricultural Green Fraud
Criminal 23 &35 Economic growth fuels standards Funding
justice classes Economic incentives Bio-fuels Hayman fire mechanisms
Enforcement college Employment Ethanol Mountains Grants
Felony college Entrepreneurs Fossil fuels Open spaces Investments
First financial aid Minimum wages Gas Outdoor Money
responders community Small business Hybrid recreation PERA
ID theft colleges T elecommunications vehicles Pine beetles Procurement
Illegal CSAP Oil Pollution Purchasing
Judges Early Oil shale Road less areas Ref. C & D
Police childhood Renewable Wildfire Sales tax
Prevention education energy prevention Severance tax
programs Prisons Public safety Security Sentencing Treatment Victims Graduation High-school dropout rate Higher education Kindergarten K-12 Learning Library Pay-for- performance Principals School districts Teachers Solar Traditional energy Utilities Wind Wildlife State budget TABOR Taxes Tax payer
67


APPENDIX A: KEY WORDS FOR POLICY TOPICS (continued)
Health care Immigration Morals/family values Transportation Water Other
CBMS Citizenship Abortion Automobiles Colorado Iraq war
CHP+ status Abuse Bridges river Social
Dental Deportation Adoption Buses Conservation security
Disease HB 1023 Amendment 43 Car pool Farm-to-city
management Human Anti-family C-DOT water
Drug smuggling Anti-same-sex Fastracks Ground
companies Human Baptist Highways water
Hospitals trafficking Bible 1-225 Platte river
Immunizations Illegal Catholic 1-25 Reservoirs
Low-birth immigration Children 1-70 corridor Reuse
weight Undocumented Clergy Light rail Snow melt
Managed care worker Euthanasia Roads Snow pack
Medicaid Faith RTD Storage
Medical schools Faith-based Tolls Water basins
Medicare Family Traffic Water
Mental health Gay congestion districts
Pharmaceuticals Gay marriage Transit Water sheds
Prenatal care God US 287 Water
Prevention Homosexuality US 36 storage
SCH1P Ref 1 Xeriscaping
Uninsured Religion
Wellness
68


REFERENCES
- (2006, October 29). Colorado governor: Ritter by 12. Rasmussen Reports. Retrieved
December 12, 2006 from http://rasmussenreports.com
- (2006, December 12). Vote for governor. Survey USA. Retrieved December 12,
2006 from http://surveyusa.com
- (2006, December 24). Colorado state profile. National Journal. Retrieved December
24, 2006 from http://election.nationaljoumal.com.
- (2007, January 6). List of United States Representatives from Colorado. Retrieved
from http://wikipedia.com
Amie, D. & Benoit, W.L. (2005). Political television advertising in campaign 2000.
Communication Quarterly, 53 (4), 473-492.
Barone, M. and Cohen, R.E. (2005). The almanac of American politics. New York:
National Journal Group.
Benoit, W.L. (1999). Seeing spots: A functional analysis of presidential television
advertisements from 1952-1996. New York: Praeger.
Benoit, W.L. (2000). Beyond genre theory: The genesis of rhetorical action.
Communication Monographs, 67, 178-192.
Benoit, W.L. (2000a). A functional analysis of political advertising across media,
1998. Communication Studies, 51, 274-295.
Benoit, W.L. (2003). Topic of presidential campaign discourse and election outcome.
Western Journal of Communication, 67 (1), 97-112.
Benoit, W.L. (2004). Political party affiliation and presidential campaign discourse.
Communication Quarterly, 52 (2), 81-98.
Benoit, W.L., Blaney, J.R., & Pier, P.M. (1998). Functions of campaign 96
Acclaiming, attacking, and defending discourse. Westport, CT: Prager.
69


Benoit, W. L., Hansen, G. J. & Holbert, R. L. (2005). Campaign information sources
in the 2000 presidential election. Presented at the National Communication
Association conference in New Orleans.
Benoit, W. L., & Harthcock, A. (1999). Functions of the Great Debates: Acclaims,
attacks, and defense in the 1960 presidential debates. Communication
Monographs, 66, 341-357.
Benoit, W.L., & Stein, K.A. (2005). A functional analysis of presidential direct mail
advertising. Communication Studies, 56 (3), p. 203-225.
Benoit, W.L., Stein, K.A., 8c Hansen, G.J. (2005). New York Times coverage of
presidential campaigns. Journalism & Mass Communication Quarterly, 82,
356-376.
Black, E. (1973). Electing time. Quarterly Journal of Speech, 59 (2), 125-129.
Brand, R. (2006, August 12). A healthy and critical debate. Rocky Mountain
News. 4C. Retrieved January 2, 2007 from http://lexis-nexus.com.
Brians, C.L. & Wattenberg M.P. (1996). Campaign issue knowledge and salience:
comparing reception from TV commercials, TV news, and newspapers.
American Journal of Political Science, 40 (1), 172-194.
Bundge, I. & Farlie, D.J.(1983). Explaining and predicting elections. London: Allen
and Unwin.
Cohen, B.C. (1963). The press and foreign policy. Princeton, NJ: Princeton
University Press.
Cohen, R.E. (2004, August 28). Colorado. National Journal, 36 (35), 2572.
Colorado State Archives. (2006, September 24). List of governors. Retrieved
September 24, 2006 from http://www.colorado.gov.
Couch, M.P. (2006a, August 16). Election 2006 States fate up for debate. The
Denver Post. B-02. Retrieved January 6, 2007 from http://lexis-nexus.com.
Couch, M.P. (2006b, October 10). Beauprez 02, 06 stands differ. The Denver Post.
B-01. Retrieved January 6, 2007 from http://lexis-nexus.com.
70


Crowley, M. (2004). Crossing over: Ken Salazar Democratic superstar. The New
Republic 231 (16), 13-15.
Davidson, D. (2004). Official publication of the abstract of voters cast for the 2003
coordinated, 2004 primary, and 2004 general. State of Colorado Secretary of
State.
Davidson, D. (2006). Colorado cumulative report. State of Colorado Secretary of
State. Retrieved December 19, 2006 from http://www.sos.state.co.us.
Davidson, D. (2007). Colorado cumulative report. State of Colorado Secretary of
State. Retrieved January 3, 2007 from http://www.sos.state.co.us.
Glathright, A. (2006a, August 9). Gubernatorial hopefuls both face challenges. Rocky
Mountain News. A4. Retrieved January 1, 2007 from http://lexis-nexus.com.
Glathright, A. (2006b, September 13). Beauprez accountability plan contracts with
Ritters. Rocky Mountain News. A18. Retrieved January 2, 2007 from
http ://lexi s-nexus. com.
Glathright, A. (2006c, November 4). Rivals agree Nov., 7 to be watershed. Rocky
Mountain News. 27A. Retrieved January 1, 2007 from http://lexis-nexus.com.
Gronbeck, B.E. (1978). Functions of presidential campaigning. Communication
Monographs, 45, 268-280.
Gronbeck, B.E. (2003). Citizens search of the World Wide Web during elections:
Rhetoric and dialogism inn electoral politics. (Key note address) College
station February 27, 2003. Retrieved 13 October, 2005 from
http://www.uiowa.edu/policult/papers/index.html.
Flansen, G.L. & Benoit, W. L. (2001). The role of significant policy issues in the
2000 presidential primaries. American Behavioral Scientist, 44, 2083-2101.
Harmon, T. (2006, August 23). Beauprez talks water in two region stops. The Pueblo
Chiefton. Retrieved January 2, 2007 from http://lexis-nexus.com.
Hart, R.P. & Childers, J.P. (2005). The Evolution of Candidate Bush: A Rhetorical
Analysis. American Behavioral Scientist, 49 (2), 80-197.
71


Holder, K. (2006, March). Voting and registration in the election of 2004. U.S.
Department of Commerce: US Census Bureau.
Jarvis, S. E. (2004). Partisan patterns in presidential campaign speeches, 1948-2000.
Communication Quarterly, 52, 403-419.
Jones, C. (2006, July 11). Colorado passes bill to bar illegal immigrants form state
benefits. USA Today. Retrieved January 15, 2007 from http://usatoday.com.
Kaid, L.L. & Johnston, A. (2002). Image ads and issues ads in US presidential
advertising: Using video style to explore stylistic differences in televised
political ads from 1952-2000. Journal of Communication, 52, 281-300.
Kohut, A. (2003). Evenly divided and increasingly polarized 2004 Political
landscape. The Pew Research Center for the People & the Press. Retrieved
February 25, 2007 from http://people-press.org.
Ladd, E.C. (1994). Of political parties great and strong. American Enterprise, 5 (4),
1047-3572.
Legislative Council for the General Assembly Research Publication No. 554. (2006,
September 14). Analysis of the 2006 ballot proposals.
Lindsay, S. (2006, October 12). Many agree: Ritter wasn't soft as DA. Rocky
Mountain News. 24A. Retrieved January 7, 2007 from http://lexis-nexus.com.
Lofholm, N. (2006, October 1). Governor candidates aboil over oil. Denver Post. C-
02. Retrieved January 7, 2007 from http://lexis-nexus.com.
Miller, J.J. (2006, December 4). The color purple. National Review, 58, (22), 28-30.
Office of State Planning and Budgeting (2007). December 2006 forecast. Retrieved
January 5, 2007 from http://www.state.co.us/
Otto, M. (2006) Control by two major political parties of governorship and general
assembly in Colorado from 1891 to present. Colorado Joint Legislative
Library. (Personal communication).
OSullivan, E., Rassel, G. R., & Berner, M. (2003). Research Methods for Public
Administrators, Fourth Edition. New York: Longman.
72


Steers, S. (2006b, August 15). Beauprez picks Rowland. Rocky Mountain News. 14A.
Retrieved January 6, 2007 from http://www.rockymountiannews.com.
Steers, S (2006c, August 18). Beauprez blasts Bush, Senate approach to illegals.
Rocky Mountain News. 18 A. Retrieved January 6, 2007 from
http://www.rockymountiannews.com.
Steers, S. (2006d, August 21). Some Republicans turn toward Ritter. Rocky Mountain
News. 7A. Retrieved January 6, 2007 from
http://www.rockymountiannews.com.
Swanson, D. L. (1972). The new politics meets the old rhetoric: New directions in
campaign communication research. The Quarterly Journal of Speech, 58, (1),
31-40.
Trent, J.S. & Friedenberg, R.V. (2004). Political Campaign Communications
Principles & Practices, Fourth Edition. Lanham: Rowman & Littlefield
Publishers, Inc.
United States Senate (2007, January 6). Senators of the United States 1789-2007.
Retrieved January 6, 2007 from
http://www.senate.gov/artandhistory/history/resources/pdf/chronlist.pdf
Walsh, K.T. (2006, October 15). A new map out west. U.S. News and World Report.
Retrieved January 6, 2006 from http://www.usnews.com
Washington, A.M. (2006, June 29). Owens calls em back. Rocky Mountain News.
Retrieved January 6, 2007 from http://www.rockymountiannews.com.
74


Paulson, S.K. (2006a, August 11). Beauprez, Ritter clash on abortion, tax surplus. The
Associated Press State and Local Wire. Retrieved January 6, 2007 from
http://lexis-nexus.com.
Paulson, S.K. (2006b, October 18) Ritter asks Beauprez to tell how he got
information for attack ad. The Associated Press State and Local Wire.
Retrieved January 6, 2007 from http://lexis-nexus.com.
Paulson, S.K. (2006c, October 23). Beauprez fends off new questions surrounding
campaign ad. The Associated Press State and Local Wire. Retrieved January
6, 2007 from http://lexis-nexus.com.
Paulson, S.K. (2006d, October 25). Election 06: Beauprez says he has experience for
gov. job. The Associated Press State and Local Wire. Retrieved January 7,
2007 from http://lexis-nexus.com.
Petrocik, J. R. (1996). Issue ownership in presidential elections, with a 1980 case
study. American Journal of Political Science, 40, 825-850.
Petrocik, J. R., Benoit, W. L., & Hansen, G. L. (2003/2004). Issue ownership and
presidential campaigning, 1952-2000. Political Science Quarterly 118, 599-
626.
Pomper, G. M. (1975) Voters choice: Varieties of American electoral behavior. New
York: Dodd, Mead, & Company.
Popkin, S.L. (1994). The reasoning vote: Communication and persuasion in
presidential campaigns. Chicago: University of Chicago Press.
Robinson, J.P. & Dennis, K.D. (1990). Television news and the informed public: An
information-processing approach. Journal of Communication, 40, (30), 106-
119.
Smith, J. (2006, September 9). Tech the great equalizer. Rocky Mountain News.
4C. Retrieved January 3, 2007 from http://lexis-nexus.com.
Steers, S. (2006a, June 29). Big names lead skirmish over immigration. Rocky
Mountain News. Retrieved January 6, 2007 from
http://www.rockymountiannews.com.
73