W ’ I V

advertisement
WHAT’S IN A VOICE?
AN EXAMINATION OF THE GENDER DIFFERENCES IN POLITICAL ADVERTISING VOICEOVERS
WILLIAM A. CUSEY
COMM 629: CAPSTONE IN POLITICAL COMMUNICATION
APRIL 28, 2014
ABSTRACT
Fifty years ago, the American public heard the first ever female
voiceover in a political advertisement. Now, fifty years later, the
political voiceover landscape remains dominated by male voiceovers:
male voiceovers outnumber female voiceovers by a 3 to 1 ratio.
However, what is not clear is exactly why this phenomenon exists. The
fundamental question this paper considers is: why is there still such a
dramatic disparity between the use of female and male voiceover
talents in political ads? This paper analyzes data attained from the
coding of 1,431 political ads (including 734 voiceover ads) that aired
during the 2008 and 2012 Presidential elections. Additionally, several
first person interviews were conducted with political ad creators,
voiceover talents, producers, and researchers to provide context,
insight, and credibility to the findings uncovered during the coding
analysis.
1
INTRODUCTION
The date is September 12, 1964. Lyndon Baines Johnson is President of the United States
and he is running for reelection. Less than one year before, popular Democratic President John
Fitzgerald Kennedy was brutally assassinated. Tensions and fears are running high. Ninety percent
of Americans believe that nuclear war is possible and 38 percent believe that nuclear conflict is
likely.1 Enter conservative Senator Barry Goldwater, the Republican Presidential nominee: “If we
are not prepared, under any circumstances, ever to fight a nuclear war, we might just as well do as
the pacifists and the collaborationists propose—dump our entire arsenal into the ocean.”2
Warmonger, trigger-happy, loose cannon, unpredictable—this is Goldwater’s reputation. It is the
genesis of Johnson’s political advertising strategy and the principal reason why a particular TV spot
is airing this night. This particular black-and-white spot opens to show an innocent little girl eating
an ice cream cone outside. The sun is shining radiantly down on her light brown hair—it looks like
a beautiful summer day. But, the voiceover in the background is dripping with fear and insecurity:
“Do you know what people used to do? They used to explode atomic bombs in the
air. Now children should have lots of vitamin A and calcium, but they shouldn't have
any strontium 90 or cesium 137. These things come from atomic bombs, and they
are radioactive. They can make you die. Do you know what people finally did? They
got together and signed a nuclear test ban treaty. And then the radioactive poison
started to go away. But now there's a man who wants to be President of the United
States, and he doesn't like this treaty. He fought against it. He even voted against it.
He wants to go on testing more bombs. His name is Barry Goldwater, and if he is
elected they might start testing all over again.”3
This ad, called “Little Girl, Ice Cream Cone,” was the Johnson campaign’s sequel to the
infamous “Peace, Little Girl” (a.k.a. “Daisy”) spot. It continued the fear-mongering theme of the
Robert Mann, Daisy Petals and Mushroom Clouds: LBJ, Barry Goldwater, and the Ad that
Changed American Politics (Baton Rouge: Louisiana State University Press, 2011), 9.
2 Barry Goldwater, Why Not Victory? A Fresh Look at American Foreign Policy (New York:
MacFadden, 1962), 25.
3 “Little Girl, Ice Cream Cone,” The Living Room Candidate: Presidential Campaign
Commercials 1952-2012, accessed March 16, 2014,
http://www.livingroomcandidate.org/commercials/1964.
1
2
“Daisy” spot: if Goldwater is elected President he will start a nuclear war, which will put our
children’s lives in danger. But, “Little Girl, Ice Cream Cone,” is not only significant for its theme and
message; it is significant because it featured the first ever female voiceover talent in a political ad.4
Now, fifty years later, what does the political voiceover landscape look like? The available
evidence depicts an environment that is dominated by male voiceover talents.5 However, what is
not clear from the evidence is exactly why this is the case. Thus, the fundamental question that this
paper will consider is: why is there such a huge disparity between the use of male and female
voiceovers in political advertising?
In order to offer a possible explanation to this query, this paper examines both “the what”
and “the why” of political voiceovers. Concerning “the what,” this paper conducts an in-depth
analysis of political ads featuring traditional voiceovers6 that aired during the primary and general
elections of the 2008 and 2012 Presidential campaigns.7 A review of the available secondary
literature reveals that political scientists have seldom studied political advertising voiceovers.
There has been little research published on the subject, particularly concerning gender.8 To date,
there has only been one significant piece of work examining the patterns and trends of political
voiceovers with respect to gender: “In a Different Voice: Explaining the Use of Men and Women
Mann, Daisy Petals and Mushroom Clouds, 70.
Patricia Strach, Katherine Zuber, Erika Franklin Fowler, Travis N. Ridout, and Kathleen
Searles, “In a Different Voice? Explaining the Use of Men and Women Voiceovers in Political
Advertising” (presentation, Annual Meeting of the American Political Science Association, Seattle,
WA, September 1-4, 2011), 3.
6 Throughout this paper, when the term “traditional voiceover” is used, it means an
anonymous voice that narrates an ad in English (not Spanish), where the individual narrating (i.e.,
the voiceover talent) is not physically represented in the ad, nor is the identity of the voiceover
talent expressly known or stated at any point during the ad.
7 While this paper will analyze ads that feature traditional voiceovers, it does not consider
ads that feature the candidate, a well-known or celebrity individual, or a physically represented onair talent as the principal spokesperson. The thought process here is to conduct an analysis that
only focuses on the voice presented in the ad, and not on other characteristics, such as who the
person is or what the person looks like. Analyzing political ads that feature voices beyond
traditional voiceovers would have necessitated the consideration of these other characteristics and
would have disrupted the focus of the paper.
8 Strach et al., “In a Different Voice?,” 3.
4
5
3
Voiceovers in Political Advertising,” by Strach, Zuber, Fowler, Ridout, and Searles.9 In the paper,
Strach et al. “examine under what circumstances men’s and women’s voices are used in campaign
advertisements to create continuous and discontinuous messages and the effect that voice and
(dis)continuity have on interpretation of campaign messages.”10 As the basis for much of their
findings, Strach et al. coded 4,887 unique ads that aired during the 2010 midterm elections in
House, Senate, and gubernatorial campaigns. This paper, on the other hand, examines 1431
Presidential ads aired in 2008 and 2012—of which 734 featured traditional voiceover talents.11 The
data accumulated for this paper will allow for a meaningful comparison to be made between this
work and the work of Strach et al. Further, the specific data collected for this study is important
because no one has yet published a study of the gender differences of advertising voiceovers in
connection with Presidential campaigns.
Although this paper may at first glance appear to mimic Strach et al., the methodology used
to analyze the political ads is more nuanced in certain ways. The vast majority of political scientists
and political practitioners, including Strach et al., have long categorized political ads in terms of
negative vs. positive ads. A few researchers in the field have discussed the problematic nature of
this simplistic frame.12 So, in an effort to move beyond the overly simplified view of positive vs.
negative political ads, this study defines and codes the identified ads along a varied emotional
spectrum. The hope is that through this emotional analysis, a more meaningful and useful set of
Ibid.
Ibid., 3.
11 This is not an exhaustive list of the ads that aired during the 2008 and 2012 Presidential
cycles. Partly because some ads from the primaries are no longer available. But, the author is
confident that the ads reviewed in this study represent the overwhelming majority of the total.
12 Ted Brader, Campaigning for Hearts and Minds: How Emotional Appeals in Political Ads
Work (Chicago, IL: The University of Chicago Press, 2006); William A. Cusey, “Are Negative Political
Ads Effective? The Disconnect between Academics and Professionals” (Communication Theory
literature review, American University, 2012); Kim L. Fridkin and Patrick J. Kenney, “The
Dimensions of Negative Messages,” American Politics Research 36, no. 5 (2008): 694-723; Travis N.
Ridout and Kathleen Searles, “It’s My Campaign I’ll Cry if I Want to: How and When Campaigns Use
Emotional Appeals,” Political Psychology 32, (2011): 439-458.
9
10
4
conclusions can be drawn from the data. For example, is there a higher proportion of female
voiceovers in anger toned ads, or of male voiceovers in fear toned ads? Or, are there more women
used to narrate empathetic or compassionate ads, and more men used to narrate heroic or
protection driven ads? These are the types of questions that the traditional political advertising
frame of positive vs. negative cannot begin to answer.
Moving beyond “the what” to “the why” this paper seeks to discover why and how political
ad makers use male and female voiceover talents in constructing a political spot. What is the
process for selecting a voice for a particular ad? What type of voice is the ad creator seeking: male
vs. female, soft vs. harsh, heavy vs. light, etc.? What is the perceived gain on behalf of the candidate
by using either a male or female voice in a political ad? Why might an ad maker pursue a male voice
over a female voice, or vice versa? Through a series of first person interviews with political ad
creators, voiceover talents, producers, researchers and others involved in the business of making
and/or studying political ads, this paper attempts to explain the questions posed above.
The political ad is arguably the most important element in defining a candidate's image
during a campaign. Campaigns spend countless hours developing these spots in the hopes of
winning the race. Thus, understanding how, why and when campaign staff use the gender of voice
to their advantage will help shed light on a crucial aspect of not only the political advertising
process, but of the political campaign process as well.
Ultimately, this paper synthesizes the raw data obtained from the coding of the Presidential
campaign ads with the information gleaned from the various first person interviews to try and
make sense of the fundamental question of this paper: why, fifty years after “Little Girl, Ice Cream
Cone,” is there still such a dramatic disparity between the use of female and male voiceover talents
in political ads?
DATA COLLECTION AND RESEARCH METHODOLOGY
Before the questions posed above can be answered, it is important to discuss the research
5
methodology. The research collection process focused primarily on the accumulation of both
qualitative and quantitative primary source research: first-person interviews and the coding of
political ads. The relative absence of existing data on political advertising voiceovers necessitated
the formulation of new data. The first-person interviews give additional context and understanding
to the data collected through the coding process. Interviews were conducted from February
through April, 2014. Interviewees were asked a range of questions designed to elicit responses that
would shed light on the political voiceover experience. Ad makers were asked how and why they
chose certain voiceover talents. Voiceover talents were asked what made their voice special and
how a voice could affect an ad’s credibility. Professors with specialties in political advertising were
asked to provide a unique outside perspective on the nature of political voiceovers. For interviews
conducted after the conclusion of the coding analysis, interviewees were presented with
preliminary data and asked to analyze the findings. Overall, a total of 7 interviews were conducted.
While the interviews provided helpful insight and analysis, the main thrust of the research
involved the coding of 1431 primary and general election political ads that aired during both the
2008 and 2012 Presidential campaigns. Of the total number of ads viewed, 734 ads featured
traditional voiceovers. The vast majority of these ads were viewed on the Stanford Political
Communication Lab (Stanford PCL) website, which proved to be invaluable in the collection of this
data. Stanford PCL has collected an archive of nearly all of the Presidential political ads (both
primary and general) going back to the 2000 cycle. Additional advertising material for the study,
not included in the Stanford PCL archive, was found on YouTube.
In order to extract and analyze meaningful data sets, the author constructed a coding matrix
in Microsoft Excel that identified 17 different characteristics (codes) for each ad: (1) Ad Name, (2)
Air Date, (3) Voiceover Gender, (4) Candidate Gender, (5) Ad Sponsor, (6) Type of Ad Sponsor, (7)
Party ID, (8) Ad Medium, (9) Election Year, (10) Election Type, (11) Emotional Tone of Ad, (12)
Voiceover Tone, (13) Target Gender, (14) Target Age, (15) Target Voting Bloc, (16) Issues, and (17)
6
Character Frames. These characteristics are explained in further detail below. Only ads that
featured a traditional voiceover were coded for all 17 traits. Ads that did not feature a traditional
voiceover (e.g., candidate voiced ads), were only coded for traits 1-10. Because this paper’s sole
focus is the examination of political advertising voiceovers, coding the substantive, emotional traits
of ads outside of this category would have disrupted the focus of this paper.
The first 10 numbered codes of the matrix represent the vitals of the ad—who made it,
when it aired, what type of election, the gender of both the candidate and the voiceover, etc. This
diverse set of vitals was chosen to increase the likelihood of finding substantive patterns of how the
voiceovers were used in the ads, especially with regard to gender. Code 3, Voiceover Gender, was
broken out into four categories: (1) Female, (2) Male, (3) Both Female and Male, and (4) No
Voiceover. This is the same coding scheme that Strach et al. used in their study and was utilized
here to provide one-to-one comparison opportunities between the two studies. For code 4,
Candidate Gender, ads that were not directly sponsored by the candidates (i.e., ads sponsored by
the party or an outside group) received the code of N/A. Code 5, Ad Sponsor, lists the names of the
ad sponsors. In order to somewhat contain the size of the coding list (which stands at 28), outside
groups were not individually named, but instead were grouped together and labeled either Pro
Dem Outside Group or Pro GOP Outside Group. Related to code 5, code 6 was devised to encourage
voiceover comparisons depending on whether the ads were sponsored by a candidate, party
committee, or outside group. Comparing how Republicans and Democrats utilized voiceovers via
code 7 was a crucial part of this analysis too. Code 8, Ad Medium, was developed as another way to
analyze the data and look for patterns. This coding category was made possible because the
Stanford PCL archive provided TV, radio, and Web ads in their collection. Furthermore, Election
Year (code 9) and Election Type (code 10) were coded separately to again increase the probability
of finding interesting and substantive findings.
The second half of the matrix (codes 11-17) includes more substantive traits dealing with
7
the emotional tone of the ad, the target audience, the main issues presented, and the character
frames associated with the ad. Code 11, Emotional Tone of Ad, coded for 7 distinct groups of
emotions: (1) Fear/Worry/Insecurity, (2) Anger/Resentment/Defiance, (3) Enthusiasm/Hope/
Change, (4) Security/Trust/Protection, (5) Compassion/Empathy, (6) Nostalgia/Reminiscence, (7)
Heroism/Determination. The primary goal in coding for the emotional theme of the ad was to move
beyond the more simplistic negative vs. positive ad frame and add sophistication to the study.
Code 12, Voiceover Tone, was broken out into four separate categories: (1) Positive Tone,
(2) Negative Tone, (3) Tonal Shift from Positive to Negative, and (4) Tonal Shift from Negative to
Positive. The idea here was to listen for how the voiceover talent manipulated his or her voice
during the ad. Codes 13, 14, and 15 all looked at the target audience. The target audience of an ad
can have a dramatic impact on how the ad is made and who the voiceover is, so multiple target
audience codes were included to provide greater depth to the analysis.
Code 16 categorized the ad’s issues into 19 separate categories: (1) Healthcare, (2)
Jobs/Economy, (3) Taxes, (4) Debt/Deficit, (5) Wars/Terrorism/Veterans, (6) Immigration, (7)
Energy, (8) Environment, (9) Role of Government, (10) Education, (11) Social Security, (12)
Poverty/Welfare, (13) Ethics/Lobbying Influence, (14) Crime/Violence, (15) Social Issues, (16)
Character Driven, (17) Other, (18) Multiple Issues, (19) N/A. Category 16, Character Driven, was
used in situations when the ad either engaged in ad hominem style attacks of a candidate, or
positive personality promotions of a candidate.
Lastly code 17, Character Frames, was broken out into 19 distinct character frames, of
which 1-10 were largely positive and 11-19 were largely negative: (1) Patriot/Pride in America, (2)
Principled/Conviction/Integrity, (3)One of us/Relatable, (4) Cares about Family, (5) American
Dream/Overcome Adversity, (6) City on a Hill/Nostalgia, (7) Doer/Gets things done, (8) Protect
from Evil, (9) Force for Good, (10) Unifier, (11) Flip-flopper, (12) Too Extreme, (13) Wastes
Money/Resources, (14) Out of Touch, (15) Hypocrite, (16) Abuse of Power, (17) Inexperienced,
8
(18) Playing Politics/Avoiding the Issues, (19) Threat to "Our America.”
The analysis of the coded data largely juxtaposed code 3, Voiceover Gender, against the
other codes of the matrix. The main purpose of the data analysis was to look for trends and
patterns associated with the use of traditional political voiceovers with respect to gender. For
example, did one party use female voiceovers more than male voiceovers? Did women mostly
narrate emotionally negative or emotionally positive ads? Did female candidates use female
voiceovers? What types of emotional codes were associated with which types of narrators? Was
there a gender difference between how voiceovers were used in 2008 vs. 2012? How did Strach et
al.’s analysis of 2010 midterm ads compare to this study’s analysis of 2008 and 2012 Presidential
ads? In order to draw out the answers to these questions, the author utilized the pivot table
function in Microsoft Excel. This proved to be an effective tool in analyzing the collected data.
LIMITATIONS
While every effort was made to produce a comprehensive set of codes that adequately and
accurately detailed the aspects of each political ad considered for this project, there are, however,
some aspects to the coding matrix that limit the findings of the study. With respect to the vitals half
of the coding matrix, there was one limitation that became apparent following the conclusion of the
coding process relating to code 3, Voiceover Gender. Although the four part coding scheme used for
this study did facilitate the collection of valuable and useful data, it could have included a more
detailed subset of codes for the category No Voiceover: candidate voiceover/appearance, celebrity
voiceover/appearance, Spanish language voiceover, average Joe/Jane voiceover/appearance, and
actor appearance. This more intricate level of coding would have produced a more sophisticated set
of results—and a future study on this topic may merit the inclusion of this expanded code set.
Concerning the second half of the coding matrix, one limitation for code 11, Emotional Tone
of Ad, was discovered part way through the data collection phase: the lack of a distinct category for
disappointment. Although there were not a very high number of disappointment ads, there were
9
enough to have warranted a separate category. These ads were largely coded as anger ads because
many of the ads featuring disappointment also featured anger themes. Further, the data for code 13,
Target Gender, and code 15, Target Voting Bloc, each produced fairly lopsided numbers, with most
ads categorized as Gender Neutral and Base. This may be an indication that the ads were incorrectly
coded for these two categories, thereby signifying a limitation to the findings. Lastly, one minor
limitation to code 16, Issues, is the absence of a true foreign policy code. Although fairly uncommon,
pure foreign policy ads did appear during the coding process; these ads were coded under Other.
FINDINGS AND ANALYSIS
An analysis of the data collected for this paper revealed five significant findings: (I) female
candidates did not utilize any female voiceovers in their political ads; (II) candidates were less
likely to use female voiceovers than parties and outside groups; (III) Republicans were more likely
to use female voiceovers than Democrats; (IV) female voiceovers appeared in more feminine centric
ads than masculine themed ads; (V) female voiceover talents narrated far more emotionally
negative ads than emotionally positive ads. But, before delving too deeply into this paper’s principal
findings, it is important to provide some contextual findings concerning the subsets for each code.
This will provide the foundation for the more
complex analyses discussed later in the paper by
specifying the frequency of each coded subset.
The most crucial of these foundational
Code 3: Voiceover Gender
All Ads
VO Ads
Female
12%
23%
Male
38%
74%
Both M & F
1%
3%
No Voiceover
49%
N/A
N=1431
N=734
numbers is within code 3, Voiceover Gender, because the gender of the voiceover talents and the
breakdown of those numbers lies at the heart of this paper’s central question: why are there so few
female voiceovers? This key question along with code 3’s findings will be discussed further in the
discussion section of this paper.
Data from codes 4 through 7 provide detailed information about the sponsor of the ad.
10
Code 4: Candidate Gender
Female
Male
N/A (Party/Outside Ads)
All Ads
6%
66%
27%
N=1431
VO Ads
6%
64%
30%
N=734
These tables show the frequency of ads
containing the specific coded characteristic, both
as a percentage of all ads viewed and as a
percentage of all traditional voiceover ads viewed.
Data for code 4 shows that roughly two-thirds of the
ads reviewed were sponsored by male candidates,
Code 6: Type of Ad
Sponsor
Candidate
Party Committee
Outside Group
Grand Total
All Ads
VO Ads
70%
12%
18%
N=1431
69%
7%
24%
N=734
whereas only 6% were sponsored by female candidates. The table for code 5 demonstrates that
Code 5: Ad Sponsor
Barack Obama
Hillary Clinton
John McCain
Mitt Romney
Ron Paul
Mike Huckabee
Rudy Giuliani
Bill Richardson
Chris Dodd
John Edwards
Tim Pawlenty
Michele Bachmann
Herman Cain
Rick Perry
Rick Santorum
Newt Gingrich
Jon Huntsman
Bob Barr
Mike Gravel
Fred Thompson
Joe Biden
Tom Tancredo
Dennis Kucinich
Buddy Roemer
DNC/other Party Cmte
RNC/other Party Cmte
Pro Dem Outside Group
Pro GOP Outside Group
All Ads
22%
6%
9%
16%
2%
0.8%
0.8%
1%
0.4%
2%
0.4%
0.6%
0.4%
2%
1%
2%
0.6%
0.1%
0.2%
0.6%
0.1%
0.3%
0.8%
0.1%
3%
8%
4%
14%
N=1431
VO Ads
23%
6%
13%
13%
3.4%
0.1%
0.8%
1%
0.4%
0.7%
0.4%
0.1%
0.4%
1%
2%
0.8%
0%
0.1%
0.3%
0.3%
0.3%
0.3%
0.8%
0.3%
2%
5%
5%
19%
N=734
Barack Obama, Pro GOP Outside Groups, John
McCain, the RNC, Mitt Romney, and Hillary Clinton
aired the most ads during 2008 and 2012.
Additionally, 2008 Republican Presidential nominee
John McCain used a traditional voiceover talent in
72% of all campaign ads in 2008. However, 2012
Republican Presidential nominee Mitt Romney only
used traditional voiceover’s 43% of the time.13 The
data for code 6 shows that party committees were
less likely to feature traditional voiceover talents in
their ads. But, outside groups, such as PACs and
SuperPACs, favored traditional voiceover style ads
more often. Lastly, the data for code 7 shows that
Democratic sponsored ads were outmatched 60% to
40% by Republican sponsored ads. This difference is
This number includes ads Mitt Romney ran during the 2008 Republican Presidential
primary campaign, as well as all of his 2012 campaign ads.
13
11
All Ads
VO Ads
potentially attributable to the large amount of outside
Democrat
40%
40%
spending by Republican SuperPACs in 2012.14 Also, the
Republican
60%
60%
N=1431
N=734
Code 7: Party ID
Grand Total
fact that there was not a Democratic primary in 2012
also contributed to a lower percentage for Democratic sponsored ads.
The data for code 8, Ad Medium, is fascinating
TV
76%
VO
Ads
84%
Radio
4%
6%
Web
21%
9%
N=1431
N=734
Code 8: Ad Medium
because Web-based voiceover ads represented a
significantly smaller percentage than total Web-based
ads run in 2008 and 2012. Subjectively speaking, these
All Ads
Web-based ads featured more movie-trailer style or documentary style ads—ads with quick cuts,
news footage, and on-the-street interviews—ads that did not utilize traditional voiceovers.
Code 9: Election Year
All Ads
VO Ads
2008
48%
53%
2012
52%
47%
N=1431
N=734
Code 10: Election Type
All Ads
VO Ads
Primary
56%
47%
General
44%
53%
N=1431
N=734
The Election Year (code 9) and Election Type (code 10) tables above show that there were
more traditional voiceover style ads aired during the 2008 Presidential election and during both
general elections than in the 2012 Presidential election and the primary campaigns. The tables for
codes 11 and 12 illustrate that the vast majority of
traditional voiceover ads viewed for this study were
negative. Fear and anger ads taken together equaled
60% of voiceover ads. Additionally, 53% of
voiceover ads featured a negative voiceover tone
and 71% of voiceover ads presented at least some
Code 11: Emotional Tone of Ad
VO Ads
Fear/Worry/Insecurity
35%
Anger/Resentment/Defiance
25%
Enthusiasm/Hope/Change
14%
Security/Trust/Protection
5%
Compassion/Empathy
3%
Nostalgia/Reminiscence
2%
Heroism/Determination
16%
N=734
negative tonal elements. This lopsided ratio between negative and positive political ads stems from
“Election 2012: Independent Spending Totals,” NYTimes.com, last modified April 18, 2014,
http://elections.nytimes.com/2012/campaign-finance/independent-expenditures/totals.
14
12
Code 12: Voiceover Tone
VO Ads
the widely held belief among political practitioners that
Positive Tone
30%
Negative Tone
53%
in order to win a campaign, the campaign must go
Tonal Shift P to N
3%
Tonal Shift N to P
15%
negative.15 However, whether the decision to go
N=734
negative will actually benefit a campaign and lead to
victory is unclear from the available academic data.16 Further targeted research is needed to draw
out those conclusions.
Data from codes 13, 14 and 15 below demonstrate the frequency of each respective
targeted demographic. The Target Gender table (code 13) shows that an overwhelming majority of
traditional voiceover ads were targeted towards a gender neutral audience. Data for code 14
illustrates that traditional voiceover ads were predominantly targeted either towards a generalaged audience or a middle-aged audience. Lastly, the data found that traditional voiceover ads
target base voters in very high proportions.
Code 13:
Target Gender
Female
VO
Ads
9%
Code 14: Target Age
Young Adult (18-29)
VO
Ads
10%
Male
4%
Middle Aged (30-59)
34%
Gender Neutral
87%
Seniors (60+)
5%
General
50%
N=734
N=734
Code 15: Target Voting
Bloc
Base
VO
Ads
90%
Independent/Undecided
10%
Not Discernable
0.1%
N=734
With respect to the issues identified in each voiceover ad, not surprisingly, jobs and the
economy, taxes, healthcare, ethics, and wars and terrorism were all top issues considered. However,
somewhat unexpectedly, roughly a quarter of all voiceover ads were character driven ads. This
figure includes both ads that attacked an opponent for “reprehensible” behavior or promoted a
candidate for “redeeming” characteristics. Although not a one-to-one comparison, this finding is
Richard R. Lau, Lee Sigelman, and Ivy Brown Rovner, “The Effects of Negative Political
Campaigns: A Meta-Analytic Reassessment,” The Journal of Politics 69, no. 4 (2007): 1176-1209;
Fridkin and Kenney, “The Dimensions of Negative Messages,” (2008).
16 Lau et al., “The Effects of Negative Political Campaigns,” (2007).
15
13
further supported by Darrell West’s analysis that, of late, the messaging of an ad is increasingly
related to a candidate’s “personal qualities” and not a “specific policy.”17 At the top of the Character
Frames list (code 17) were voiceover ads that featured a negative frame. In fact, three out of the top
five character frames were negative: Too Extreme (15%), Out of Touch (13%), and Threat to “Our
America” (8%). Also in the top five were the positive character frames Doer/Gets Things Done
(15%), and Principled/Conviction/Integrity (8%).
Code 16: Issues
VO Ads
Code 17: Character Frames
VO Ads
Healthcare
5%
Patriot/Pride in America
4%
Jobs/Economy
21%
Principled/Conviction/Integrity
8%
Taxes
7%
One of us/Relatable
2%
Debt/Deficit
2%
Cares about Family
1%
Wars/Terrorism/Veterans
7%
American Dream/Overcome Adversity
1%
Immigration
1%
City on a Hill/Nostalgia
0.4%
Energy
4%
Doer/Gets things Done
15%
Environment
1%
Protect from Evil
3%
Role of Government
4%
Force for Good
5%
Education
1%
Unifier
2%
0.4%
Flip-flopper
2%
Poverty/Welfare
1%
Too Extreme
15%
Ethics/Lobbying Influence
6%
Wastes Money/Resources
4%
Out of Touch
13%
Social Security
Crime/Violence
0.5%
Social Issues
3%
Hypocrite
3%
Character Driven
26%
Abuse of Power
6%
Other
3%
Inexperienced
3%
Multiple Issues
6%
Playing Politics/Avoiding the Issues
4%
Threat to "Our America"
8%
N/A
0.3%
N=734
N=734
Darrell M. West, Air Wars: Television Advertising in Election Campaigns, 1952-2008,
(Washington, D.C.: CQ Press, 2010), 50-52.
17
14
FINDING I: FEMALE VOICEOVERS ABSENT FROM FEMALE CANDIDATE SPOTS
As Table 1.1 shows, female Presidential candidates Hillary Clinton (2008) and Michele
Bachmann (2012) did not use any female voiceovers in their campaign ads.18 In 2008, Hillary
Clinton ran 84 political ads, 43 of which employed a traditional voiceover. All 43 voiceover ads
aired by the Clinton campaign used a male voiceover talent. Likewise, Michele Bachmann aired 9
political ads in 2012—most of which she voiced herself—and the one voiceover ad she did air
featured a male actor. Moreover, Strach et al. similarly found that female Congressional and
gubernatorial candidates running in 2010 shunned female voiceover talents: only 10.5% of ads
aired by these candidates exclusively featured a female narrator.19
TABLE 1.1
Women Presidential
Candidates
Michele Bachmann
Hillary Clinton
Voiceover Gender
Female
Male
Both M & F
No Voiceover
0%
11%
0%
89%
0%
51%
0%
49%
N=1431
These are fascinating statistics, though not altogether surprising. There is a generally held
belief among ad makers that in situations where a candidate has less name recognition, such as in
smaller market races, it is important to use a voiceover of the opposite sex to avoid audience
confusion between the voiceover talent and the candidate.20 This theory could explain why Michele
Bachman eschewed female voiceovers, because she had fairly low name recognition and support
when she ran for President in 2012.21
Hillary Clinton did run an ad in 2008 that featured Maya Angelou’s endorsement of her. It
was the closest the Clinton campaign came to running an ad with a female voiceover without
actually doing so. Here is the ad: https://www.youtube.com/watch?v=JtYwQvFd_bg.
19 Strach et al., “In a Different Voice?,” 29.
20 Strach et al., “In a Different Voice?,” 19; Daniel Stone, “How Voice Actors Are Chosen for
Political Ads,” Newsweek, last modified March 13, 2010, http://www.newsweek.com/how-voiceactors-are-chosen-political-ads-89351. This belief is not entirely supported by the data when
considering male candidates in down-ballot races. Strach et al. found that male candidates still used
male voiceovers in 44% of all ads and only used a female voiceover 18% of the time.
21 “2012 Republican Presidential Nomination Polling Data,” Real Clear Politics,
http://realclearpolitics.com/epolls/2012/president/us/republican_presidential_nomination1452.html.
18
15
However, that belief does not adequately explain why Hillary Clinton’s campaign exclusively
used male actors in her voiceover spots, because as the former First Lady and Senator of New York,
Clinton had enormously high name recognition. When asked why Clinton did not use any female
voiceovers, Kathryn Klvana, a well-established and
experienced Democratic voiceover talent, offered one
interesting explanation: “In 2008, Hillary was attempting
to run away from the fact that she was a woman. She
wanted people to think she was a man.”22 This idea is
reinforced by the high level of newsworthiness attached
TABLE 1.2
Hillary Clinton: Emotional Tone of Ads
Fear/Worry/Insecurity
9%
Anger/Resentment/Defiance
7%
Enthusiasm/Hope/Change
12%
Security/Trust/Protection
7%
Compassion/Empathy
7%
Nostalgia/Reminiscence
0%
Heroism/Determination
58%
to Hillary’s “emotional moment” prior to the 2008 New
N=734
Hampshire primary.23 Furthermore, this explanation is supported by the data in Table 1.2, which
shows that a majority of Hillary Clinton’s voiceover ads portrayed traditionally masculine24
emotions and themes, such as heroism, determination, security, and strength.
Moreover, it appears that Hillary Clinton’s masculine political image may have harmed her
ability to win the 2008 Democratic primary. Professor Susan Douglas wrote a piece in 2007 entitled
“Why Women Hate Hillary,” describing the detrimental effect this masculine persona had among
potential female supporters: “Hillary wants to be more like a man in her demeanor and politics,
leaving some basic tenets of feminism in the dust. She is like patriarchy in sheep’s clothing.”25
Essentially, Professor Douglas explained that progressive women—a cohort that in her opinion
should have overwhelmingly supported Hillary’s nomination—did not support her and felt
Kathryn Klvana, Interview by William A. Cusey, March 28, 2014.
Karen Breslau, “Hillary Clinton’s Emotional Moment, Newsweek, last modified March 13,
2010, http://www.newsweek.com/hillary-clintons-emotional-moment-87141.
24 “Gender & Gender Identity,” Planned Parenthood,
http://www.plannedparenthood.org/health-topics/sexual-orientation-gender/gender-genderidentity-26530.htm.
25 Susan Douglas, “Why Women Hate Hillary,” In These Times, April 26, 2007,
http://inthesetimes.com/article/3129/why_women_hate_hillary.
22
23
16
betrayed by her lack of femininity. Kathryn Klvana echoed Professor Douglas’s comments and said
that “Hillary Clinton’s overly mannish portrait hurt her image and likely doomed her 2008
campaign.”26 In fact, Hillary Clinton barely won 50% of female voters in 2008 against Democratic
opponent Barack Obama. She also did not win the support of all female elected leaders, such as
Senators Amy Klobuchar and Claire McCaskill, who supported Obama’s nomination instead.27 Even
a few more percentage points of support among women could have made the difference and given
her the nomination. In 2016, if Hillary Clinton decides to run, it will be interesting to see whether
she will continue to rely on male voiceovers and masculine themes or pivot towards more feminine
characteristics by employing more female voiceovers in her political ads.
FINDING II: CANDIDATES LESS LIKELY TO USE FEMALE VOICEOVERS THAN PARTIES AND OUTSIDE GROUPS
As Tables 2.1 and 2.2 demonstrate, non-candidate sponsored ads were much more likely to
feature female voiceovers than candidate sponsored ads. But, before considering this finding in
more depth, it is important to understand how often these different entities used or did not use
voiceovers in their political spots. Table 2.1 provides the context for this later discussion and
further provides the foundation
for the numbers represented in
Tables 2.2 and 2.3. Table 2.1
shows that 70% of ads sponsored
by the major party committees
TABLE 2.1
Type of Ad
Sponsor
Candidate
Party
Committee
Outside Group
Voiceover Gender
Female
Male
Both M & F
8%
41%
1%
No
Voiceover
49%
15%
13%
2%
70%
24%
42%
2%
32%
N=1431
did not employ a traditional voiceover talent. Conversely, approximately 70% of outside group
sponsored ads did feature voiceovers. This means that the sample size in Table 2.2 for Outside
Group ads (175) is much larger than the sample size for Party Committee ads (50). The sample size
Klvana Interview.
Penny Lee, “No banking on the Women’s Vote,” U.S. News and World Report Online, April 4,
2013, http://www.usnews.com/opinion/blogs/penny-lee/2013/04/04/hillary-clinton-shouldntcount-on-the-womens-vote-in-2016.
26
27
17
for candidate sponsored ads, however, is 1005 for all ads considered and 509 for traditional
voiceover ads.
Keeping this in mind, Table 2.2 shows a fairly wide disparity in the use of female voiceovers
between party committees and outside groups
on the one hand, and candidates on the other.
Party committees, mainly the RNC and DNC,
used a female voice talent in 56% of their
TABLE 2.2
Type of Ad
Sponsor
Candidate
Party
Committee
Outside Group
traditional voiceover ads. Likewise, outside
Voiceover Gender
Female
Male
Both M & F
16%
82%
2%
50%
44%
6%
35%
62%
3%
N=734
groups, such as Crossroads and Priorities USA, used a female voiceover in 38% of traditional
voiceover spots. However, Presidential candidates running in 2008 and 2012 only used a female
voice talent in 18% of ads featuring a traditional voiceover, or a mere 9% in ads aired overall.
So, why the large disparity? Why are non-candidate sponsored ads utilizing female
voiceovers more often than candidate sponsored ads? One possible explanation relates to the
emotionality of the ads aired by the different types of ad sponsors: the vast majority of traditional
voiceover ads aired by both party committees and outside groups were either fear or anger toned
ads. Table 2.3 shows that 90% of party sponsored voiceover ads and 77% of outside group
TABLE 2.3
Type of Ad Sponsor
Emotional Tone of Ad
Candidate
sponsored ads featured
Fear/Worry/Insecurity
29%
Party
Committee
62%
Outside
Group
43%
Anger/Resentment/Defiance
22%
28%
34%
Enthusiasm/Hope/Change
16%
2%
9%
Security/Trust/Protection
7%
2%
1%
Compassion/Empathy
4%
2%
2%
Nostalgia/Reminiscence
2%
0%
2%
Heroism/Determination
20%
4%
9%
N=734
negative emotional themes.
These statistics regarding
emotionally negative ads are
important because women are
far more likely to narrate ads
with negative emotions than
ads with positive emotions. Therefore, because party committees and outside groups each aired a
high percentage of negative ads, and because women narrated negative ads at a higher rate, it
18
follows that party committees and outside groups each featured a higher percentage of female
voiceovers in their ads. This major finding—that women voice more negative ads than positive
ads—will be discussed more fully in Finding V, but it is important to highlight here as an
explanation for the voiceover disparity among different types of ad sponsors.
One further explanation for this finding could simply be that the sample size is not large
enough to make any statistically significant findings. Strach et al., who coded nearly 3.5 times the
number of ads this study coded, found fairly different results with regard to this section’s specific
finding.28 For example, they found that 96% of all party sponsored ads used either a male, female,
or combination voiceover and, of those ads, 76% used a male voice.29 Strach et al. also found that
candidates used female voiceovers in 37% of all ads featuring voiceovers, whereas party
committees only used a female voice talent in 24% of voiceover ads30—essentially the opposite
finding of this section. However, it is also possible that the differences between this study and
Strach et al.’s study could be unrelated to sample size and instead could be related to the sample
itself. As previously mentioned, Strach et al. coded ads that aired during the 2010 midterms,
whereas this study analyzed Presidential ads. Perhaps different types of races have different
patterns of how and when voiceovers are used in political ads. Since the research on this topic is so
sparse, future study is needed to determine whether this finding is applicable.
FINDING III: REPUBLICANS MORE LIKELY TO USE FEMALE VOICEOVERS THAN DEMOCRATS
In political campaigning, every action is taken for a specific reason with the intention of
getting a candidate elected. Whether those reasons are based on factual evidence, or merely based
on the gut feelings and reactions of the campaign operative, the actions these individuals take are
28 Strach et al. did not code for ads from outside groups, so that figure could not be
compared here.
29 Strach et al., “In a Different Voice?,” 30.
30 Ibid.
19
purposeful. So, considering that elemental fact about politics, it is intriguing to note that during the
2008 and 2012 Presidential campaigns, Republicans were nearly twice as likely as Democrats to
use a female voiceover in their political ads. Table 3.1 illustrates that in traditional voiceover style
ads, Republicans employed a female voice
TABLE 3.1
actor 32% of the time compared with 17% for
Democrats. Similarly, Strach et al.’s study of
Party ID
Voiceover Gender
Female
Male
Both M & F
Democrat
16%
83%
1%
Republican
28%
69%
4%
N=734
the 2010 midterms also found that Republicans utilized more women voiceovers than Democrats:
39% of Republican voiceover ads used a female narrator, whereas Democrats used a female
voiceover 34% of the time.31 Additionally, Table 3.2 sheds light on this disparity between
Democrats and Republicans: Republican Presidential candidates John McCain and Mitt Romney
utilized far more female voiceovers (28% each) than Democratic Presidential Candidates Hillary
Clinton and Barack Obama (0% and 18%). Further, the RNC used more than double the number of
female voice talents in their voiceover ads than the DNC did.
Thus, assuming that every
political action is taken for a reason,
why did Republicans all across the
country, in three different election
cycles, use more female voiceovers in
their political ads than Democrats?
American University Professor
TABLE 3.2
Voiceover Gender
Ad Sponsor
Female
Male
Both M & F
Barack Obama
17%
82%
1%
Hillary Clinton
0%
100%
0%
John McCain
25%
72%
3%
Mitt Romney
24%
71%
4%
Ron Paul
8%
84%
8%
DNC/other Party Cmte)
29%
64%
7%
RNC/other Party Cmte)
62%
32%
6%
Pro Dem Outside Group
38%
60%
3%
Pro GOP Outside Group
35%
63%
3%
Andrew Babb, who spent the vast
N=734
majority of his career as a professional ad man, offered this explanation: “Republicans didn't want
to come across as tough guys, so they went with more women—to mitigate their war on women.”32
31
32
Strach et al., “In a Different Voice?,” 30.
Professor Andrew Babb, Interview by William A. Cusey, April 15, 2014.
20
This idea—that Republicans wanted to shore up support among women by using more female
voiceovers—was further echoed by Professor Leonard Steinhorn33 and longtime voiceover actor
John Doe.34
The so-called “war on women” refers to both past and current Republican sponsored
legislative proposals designed to limit women’s access to various health related services, namely
contraception and abortion rights. The term became much more widely used during the 2012
election cycle after Missouri Senatorial Candidate Todd Akin issued his infamous “legitimate rape”
comments.35 The “war on women” concept became such a commonly used and researched phrase
that it spawned the creation of a Wikipedia page.36 Moreover, even before the start of their so-called
“war on women,” Republicans struggled to attract female voters: in 2008, Barack Obama won 56%
of female voters against Republican nominee John McCain.37 That trend continued in 2012, when
President Obama won 55% of women.38 Furthermore, several media stories reported that the
Republicans’ “war on women” cost them the Presidency and control of the Senate in 2012.39
With all of this evidence concerning low support among female voters, Republicans had to
figure out how to, as Professor Babb said, “mitigate” this weakness and show the voters that this
“war on women” was a myth. Thus, it is likely that one of the ways Republicans sought to increase
Leonard Steinhorn, personal e-mail communication, April 7, 2014.
John Doe, Interview by William A. Cusey, April 10, 2014. John Doe is a pseudonym. He
requested that his true name be kept confidential.
35 Lori Moore, “Rep. Todd Akin: The Statement and the Reaction,” The New York Times,
August 20, 2012, http://www.nytimes.com/2012/08/21/us/politics/rep-todd-akin-legitimaterape-statement-and-reaction.html?_r=0.
36 “War on Women,” Wikipedia, http://en.wikipedia.org/wiki/War_on_Women.
37 “Exit Polls,” CNN.com, November 5, 2008,
http://www.cnn.com/ELECTION/2008/results/polls.main/.
38 “President: Full Results,” CNN.com, last modified December 10, 2012,
http://www.cnn.com/election/2012/results/race/president.
39 Monika Bauerlein and Clara Jeffery, “WTF, GOP?,” Mother Jones, May/June, 2012,
http://www.motherjones.com/politics/2012/04/gop-war-on-women; Frank Rich, “Stag Party,”
New York Magazine, March 25, 2012, http://nymag.com/news/frank-rich/gop-women-problem2012-4/; Nancy Cohen, “Women Fought the GOP’s 2012 ‘War on Women’ and Won,” The Guardian,
November 7, 2012, http://www.theguardian.com/commentisfree/2012/nov/07/women-foughtgop-2012-waronwomen.
33
34
21
support among women was through the increased use of female voiceovers. Whether this
mitigation strategy proved effective at boosting female support is unclear. Many Republican
strategists believe that the “war on women” is still a significant problem and are actively working to
re-engineer the conversation away from a “war on women” to a “war for women”: entities such as
the RNC and several Republican leaning PR and consulting firms are leading this fight.40 However,
going forward, it will likely take more than a few extra female-voiced ads and other clever
messaging techniques to mitigate the damage done from the ongoing “war on women” storyline.
FINDING IV: MORE FEMALE VOICEOVERS IN FEMININE CENTRIC ADS THAN MALE THEMED ADS
This finding is neither surprising nor groundbreaking. It does not cause the reader to
experience cognitive dissonance. In fact, it is a finding that was entirely expected upon entering into
this study. But, if this finding is so banal, what makes it significant enough to be considered a
principal finding? It is significant because it lays the statistical foundation for the primary
discussion section of this paper, which offers answers as to why there so few female voiceovers in
political advertising. The research analysis for this section uncovered four specific sub-findings
related to the frequency of female and male voiceovers: (1) the gender targeting of ads, (2) the
emotional tone of the ads, (3) the issues represented in the ads, and (4) the positive character
frames exhibited by the ads.
Concerning gender targeted ads, Table 4.1
shows that within the universe of traditional
voiceover ads, women narrated 80% of female
TABLE 4.1
Voiceover Gender
Target Gender
Female
Male
Female
Male
77%
0%
20%
100%
Both
M&F
3%
0%
Gender Neutral
18%
79%
3%
N=734
targeted ads. Likewise, male targeted ads were dominated by male voiceover talents. Somewhat
surprisingly though, women only voiced 21% of ads geared toward a gender neutral audience. This
sub-finding speaks to the broader point about the general lack of female voiceovers in political ads.
Aviva Shen, “Inside The GOP’s Attempt to Change Its ‘War on Women’ Image,”
ThinkProgress, March 29, 2014, http://thinkprogress.org/election/2014/03/29/3419593/insidethe-gops-attempt-to-change-its-war-on-women-image/.
40
22
Table 4.2 demonstrates that, when juxtaposed against male voiceovers, women narrated
compassion and empathy toned ads, as well as fear and insecurity toned ads, 35% of the time.
As Table 4.2 further shows, of all of the
TABLE 4.2
Voiceover Gender
emotional tones coded for in this study,
Emotional Tone of Ad
Female
Male
Fear/Worry/Insecurity
32%
65%
Both
M&F
3%
Anger/Resentment/Defiance
22%
75%
3%
Enthusiasm/Hope/Change
16%
82%
2%
Security/Trust/Protection
15%
79%
5%
Compassion/Empathy
35%
65%
0%
Nostalgia/Reminiscence
17%
83%
0%
Heroism/Determination
13%
85%
3%
women voiced these groups of ads at a
rate 10 points above the next highest set
of emotions. These emotional
categories—compassion/empathy and
fear/insecurity—are more commonly
N=734
associated with feminine characteristics than the other emotions represented in this study.41
TABLE 4.3
Voiceover Gender
Issues
Female
Male
Healthcare
46%
Jobs/Economy
Voiceover ads that were analyzed for
51%
Both
M&F
3%
different issue patterns revealed that
20%
78%
3%
women overwhelmingly narrated ads
Taxes
38%
58%
4%
Debt/Deficit
33%
50%
17%
Wars/Terrorism/Veterans
6%
94%
0%
Immigration
22%
78%
0%
Energy
30%
70%
0%
Environment
43%
57%
0%
Role of Government
19%
78%
4%
Education
11%
89%
0%
largely refer to abortion, contraception, and
Social Security
33%
33%
33%
Poverty/Welfare
25%
75%
0%
marriage equality issues—issues more
Ethics/Lobbying Influence
22%
78%
0%
Crime/Violence
25%
75%
0%
Social Issues
Character Driven
79%
19%
21%
79%
0%
3%
N=734
featuring social issues. Table 4.3 depicts
that female voiceovers were featured in
79% of all social issue related voiceover
ads. In the realm of politics, social issue ads
often associated with women than men.42
Additionally, Table 4.3 shows that
voiceover ads discussing social security
“Gender & Gender Identity,” Planned Parenthood.
Laurie Kellman, “Social Issues Retake U.S. Politics, 2012 Elections,” Huffington Post:
HuffPost Politics via AP Wire, February 10, 2012,
41
42
23
(66%), the debt or deficit (50%), healthcare (49%), environment (43%), and taxes (42%) featured
female voiceovers at a rate well above the 26% average for this study. Table 4.3 further illustrates
that men narrated the vast majority of ads relating to war, national defense, and violence: wars,
terrorism, and veterans (94%), immigration (78%), and crime/violence (75%). On the topic of jobs
and the economy—the number one issue for American voters in both 200843 and 201244—male
voice actors again dominated
TABLE 4.4
Voiceover Gender
by voicing 78% of ads in that
Positive Character Frames
Female
Male
Patriot/Pride in America
6%
91%
Both
M&F
3%
Principled/Conviction/Integrity
13%
83%
3%
One of us/Relatable
20%
80%
0%
Cares about Family
33%
67%
0%
American Dream/Overcome Adversity
25%
75%
0%
City on a Hill/Nostalgia
33%
67%
0%
Doer/Gets things done
18%
80%
2%
Protect from Evil
4%
96%
0%
Force for Good
8%
92%
0%
Unifier
0%
83%
17%
category.
The fourth sub-finding
of this section depicts the
frequency of male and female
voiceovers with respect to
positive character frames.
Table 4.4 illustrates that
N=734
female voiceovers were most common in nostalgic and family-centered frames, appearing 33% of
the time in each type of ad. Male voiceovers, on the other hand, overwhelmingly appeared in
patriotic and leadership oriented ads: “protect from evil” (96%), “force for good” (92%),
“patriot/pride in America” (91%), and “unifier” (83%). It is fascinating that some of the most
fundamental character frames for any President are so heavily dominated by male voiceover
talents. As discussed above, the finding that female narrators voiced more feminine centric ads and
http://www.huffingtonpost.com/2012/02/10/social-issues-politics-2012elections_n_1267779.html.
43 Lydia Saad, “Economy Reigns Supreme for Voters: More than Half Rate it ‘Extremely
Important’ to their Vote for President,” Gallup.com, October 29, 2008,
http://www.gallup.com/poll/111586/economy-reigns-supreme-voters.aspx.
44 Lydia Saad, “Economy Is Dominant Issue for Americans as Election Nears: Total Mentions
of Economic Issues are as High Today as in October 2008,” Gallup.com, October 22, 2012,
http://www.gallup.com/poll/158267/economy-dominant-issue-americans-election-nears.aspx.
24
male narrators voiced more masculine centric ads was not exactly a surprise. However, the degree
to which female voiceovers were excluded from some of the traditionally masculine emotions,
issues, and characteristics was definitely surprising.
FINDING V: WOMEN NARRATED MORE EMOTIONALLY NEGATIVE ADS THAN EMOTIONALLY POSITIVE ADS
“Seventy Percent? That surprises me—it really does.”45 That was Professor Babb’s reaction
when asked what he thought about the fact that 70% of political voiceovers narrated by women
TABLE 5.1
Voiceover Gender
Voiceover Tone
Female
Male
Positive Tone
18%
33%
Both
M&F
26%
Negative Tone
70%
48%
47%
Tonal Shift P to N
2%
3%
11%
Tonal Shift N to P
10%
16%
16%
conveyed a negative tone. Table 5.1 depicts that
N=734
statistic and also shows that 82% of voiceovers
narrated by women contained some negative tonal
elements. Voiceovers featuring either a male talent
or both a male and female talent were still more
negative than positive, but less so than when only female voiceover talents were employed.
Table 5.2 further supports this
TABLE 5.2
Voiceover Gender
Fear/Worry/Insecurity
48%
30%
Both
M&F
37%
female narrations bore either
Anger/Resentment/Defiance
24%
25%
26%
fear or anger related emotions.
Enthusiasm/Hope/Change
9%
15%
11%
Security/Trust/Protection
4%
6%
11%
Lastly, Table 5.3 (below)
Compassion/Empathy
5%
3%
0%
Nostalgia/Reminiscence
1%
2%
0%
Heroism/Determination
9%
19%
16%
finding, showing that 72% of
depicts the frequency of male
and female voiceovers with
Emotional Tone of Ad
Female
Male
N=734
respect to each negative character frames. It illustrates that female voiceovers appeared more often
than the 26% average for this study in several categories of negatively framed ads: flip-flopper
(53%), too extreme (42%), threat to “our America” (38%), wastes money or resources (35%), out
of touch (31%), and inexperienced (29%). The main takeaway from this data is that women
voiceover talents simply do not narrate that many positively themed ads—they are mostly confined
45
Babb Interview.
25
to voicing emotionally negative ads.
TABLE 5.3
Negative Character Frames
Voiceover Gender
Female
Male
Both M & F
Flip-flopper
53%
47%
0%
Too Extreme
38%
59%
4%
Wastes Money/Resources
23%
65%
12%
Out of Touch
27%
69%
4%
Hypocrite
9%
91%
0%
Abuse of Power
22%
78%
0%
Inexperienced
29%
71%
0%
Playing Politics/Avoiding the Issues
14%
83%
3%
Threat to "Our America"
38%
62%
0%
N=734
So, why is this happening? Longtime political voiceover talent Sheldon Smith offered this
explanation: “The last several election cycles there’s been kind of a philosophy, which has been
widely held by political media consultants, that if you do a negative commercial you really need to
use a female voice. Because it somehow softens the message.”46 Essentially, Sheldon Smith made the
claim that political ad makers are hiring women to narrate negatively themed political ads because
they believe using a female voice will improve the credibility and efficacy of the ad by softening the
impact with the viewers. Kathryn Klvana verified this claim: “This is absolutely true; I have heard
producers tell me that’s why women are hired sometimes—to soften negative attack ads.”47
Professor Babb further confirmed the existence of this belief and agreed with the general consensus
of the producers that female voiceovers have a beneficial impact when employed in negative ads: “I
don't know if [a female voiceover] softens the attack, but it makes it less strident and over the top.”
Professor Robert Mann, author of Daisy Petals and Mushroom Clouds, substantiated the
existence of this philosophy regarding women voice talents and negative ads, but questioned its
accuracy: “There is a theory that if you want to attack your opponent, you should use a woman to
46
47
Sheldon Smith, Interview by William A. Cusey and Jesse Strauss, February 19, 2014.
Klvana Interview.
26
narrate your ads. The idea is that a woman’s voice would lighten the impact of the ad. However, I
don’t know how true that is, it seems to me that I’ve heard plenty of men voicing negative ads
too.”48 And, he is right, at least about the last part. Although not shown in one of the tables, men
narrated 67% of all negative toned voiceover ads considered in this study. So, even though
approximately three-quarters of the ads voiced by women conveyed a negative tone, men still
narrated the vast majority of all emotionally negative political ads. This is a fascinating statistic
given the fact that there is so much attention placed on whether to employ a female voiceover talent
to narrate a negatively themed political ad.
DISCUSSION
Fifty years ago, the Johnson campaign introduced America to the first female political
voiceover with its “Little Girl, Ice Cream Cone” spot. This ground breaking moment occurred at a
time when women were relatively non-existent in the political process: only 14 women—two in the
Senate and 12 in the House—served in Congress in 1964,49 and of the fifty state governors, not one
was a woman.50 Likewise, female voiceovers were relatively absent from political advertising
campaigns until the early 1990s.51 Now, in 2014, women have made progress in breaking into
politics, but are still vastly underrepresented.52 There are currently only five female governors, 20
Robert Mann, Interview by William A. Cusey, March 18, 2014.
U.S. House of Representatives: Office of the Historian, Office of Art & Archives, Office of
the Clerk, “Mapping Congress,” U.S. House of Representative: History, Art & Archives, accessed April
21, 2014, http://history.house.gov/Map/Mapping-Congress/.
50 Center for American Women and Politics, Eagleton Institute of Politics, Rutgers, The State
University of New Jersey, “Statewide Elective Executive Women 2014,” last modified March, 2014,
http://www.cawp.rutgers.edu/fast_facts/levels_of_office/documents/stwide.pdf.
51 Klvana Interview; Babb Interview.
52 Professors Jennifer Lawless and Richard Fox have thoroughly studied why this gender
gap exists: “the fundamental reason for women’s under-representation is that women do not run
for office. There is a substantial and persistent gender gap in political ambition; men tend to have it,
and women don’t.” Jennifer L. Lawless and Richard L. Fox, “Girls Just Wanna Not Run: The Gender
Gap in Young Americans’ Political Ambition,” (Women & Politics Institute, Washington, D.C., March
2013): 1, http://www.american.edu/spa/wpi/upload/girls-just-wanna-not-run_policy-report.pdf.
48
49
27
female Senators, and 79 female House Members.53 Further, this study found that female voiceover
talents are also greatly underrepresented in the political process today: as discussed previously,
only 26% of voiceover ads (13% of all ads) reviewed for this paper featured at least a partial female
voiceover.54 Similarly, Strach et al. found that only 35% of political voiceover ads (24% of all ads)
featured a female narrator.55 The following section offers several possible explanations as to why
this imbalance exists.
First, it has been proposed by some people in the field that the reason for the large
discrepancy between male and female voiceovers is connected to research performed by the
campaign. Professor Mann posited that “campaigns today are so data driven. So, the disparity might
exist because the data these advertisers are finding shows that male voiceovers elicit the right
emotional response from the target audience.”56 Moreover, voiceover artist John Doe suggested that
“a lot of these ads are tested and they’re always looking for the advantage. So, it might be that the
research has shown women to be effective in these situations and not in others.”57 With respect to
the broader trend toward more data-driven advertising campaigns, they are both correct.58 In 2008
Center for American Women and Politics, Eagleton Institute of Politics, Rutgers, The State
University of New Jersey, “Current Numbers of Women Officeholders,” accessed April 21, 2014,
http://www.cawp.rutgers.edu/fast_facts/levels_of_office/Current_Numbers.php.
54 Please see Table “Code 3: Voiceover Gender” on page 9.
55 This study’s numbers are likely lower than Strach et al.’s numbers because the 2008
primary featured an extremely low percentage of female voiceover style ads (probably because of
Hillary Clinton’s ads): just 9% compared to the 2008 general election with 36%, the 2012 primary
election with 29% and the 2012 general election with 28%. Excluding the 2008 primary ads from
the total calculation of female voiceovers (where N=552 rather than N=734) changes the
percentage of female voiceovers from 26% to 31%, which is much closer to Strach et al.’s topline
number of 35%. So, it is possible, that the 2008 primary represents an anomalous outlier to the
overall trend.
56 Mann Interview.
57 Doe Interview.
58 Colin Delany, “Data-Driven Politics,” Campaign Insider, May 17, 2012,
http://www.campaignsandelections.com/campaign-insider/319937/datadriven-politics.thtml;
Author Unknown, “The Data Driven Campaign,” The American Prospect, accessed April 21, 2014,
http://prospect.org/article/data-driven-campaign; Brian Fung, “Microsoft is Using Your Data to
Target Political Ads on Xbox Live,” The Washington Post, March 7, 2014,
http://www.washingtonpost.com/blogs/the-switch/wp/2014/03/07/microsoft-is-using-yourdata-to-target-political-ads-on-xbox-live/.
53
28
and even more so in 2012, campaigns have engaged in micro-targeting data analyses in an effort to
make the art of voter persuasion more effective and successful.59 However, Sheldon Smith lamented
what he sees as a general lack of these types of detailed data analyses being utilized during the
voiceover selection process for political ads:
“A lot of research goes into the ad in terms of content…I don’t think there’s a parallel
amount of research that goes into what kind of [voice] actor are we going to hire to
do this. Is it going to be a man? Is it going to be a women? Is it going to be someone
young or someone older? I think that kind of happens more often than not by
happenstance.”60
As someone with in-depth knowledge of the voiceover selection process, Sheldon Smith
offered a different explanation for how voiceover talents are chosen: “Usually, the decision of how
to select a voice I think is asked based on who do you know that perhaps you’ve worked with
before. And in almost every case, you have.”61 So, what Sheldon Smith touched on here, and what
others in the field of political advertising are saying, is that voiceovers are chosen based on
relationships.62 In a town dominated by networking and connections, this concept is not surprising.
It is also not surprising because producers want to ensure that they are going to create a great ad
for the candidate; from their perspective, the best way to accomplish that is by going with a
voiceover talent they know and trust, and have worked with many times before. As an ad man,
Professor Babb followed this procedure for hiring voice actors to narrate his spots: “You can’t
underestimate people's sort of inertia…lots of people just keep going back to the actors they know,
Michael Scherer, “Inside the Secret World of the Data Crunchers Who Helped Obama
Win,” Time.com: Swampland, November 7, 2012, http://swampland.time.com/2012/11/07/insidethe-secret-world-of-quants-and-data-crunchers-who-helped-obama-win/; Christina Macholan and
Evan Rowe, “Obama for America Uses Google Analytics to Democratize Rapid, Data-Driven Decision
Making,” Google Analytics Blog, August 6, 2013, http://analytics.blogspot.com/2013/08/obama-foramerica-uses-google-analytics.html.
60 Smith Interview.
61 Smith Interview.
62 Terrance Green, Interview by Jesse Strauss, February 17, 2014; Paul Wilson, Interview by
Jesse Strauss, February 26, 2014; Smith Interview, Klvana Interview, Doe Interview, Babb
Interview.
59
29
[because] you want someone reliable that you know does a great job.”63 Well-established
Republican ad maker Paul Wilson further confirmed this behavior: “We have a number of
announcers that we’ve worked with over the years. And, the one on that commercial is very famous.
His name’s Shel Smith. And, he’s made literally thousands of commercials…”64
However, this practice of choosing voiceover talents based on pre-existing relationships
could be negatively impacting the number of opportunities for female voiceover talents. Kathryn
Klvana explained why: “Who becomes the voiceover is related to relationships—who the agency
trusts. It’s hard to break in if you’re a new voice. Political voiceovers are very fast paced and the
producers want to be able to count on the voices they hire. There’s a kind of loyalty in the
industry.”65 Given the difficulty faced by new voiceover talents trying to break into the field,
coupled with the paucity of existing established female voiceover talents, it is unlikely that the 3 to
1 ratio of male to female voiceovers is unlikely to change under this current paradigmatic system.
At least with respect to political voiceovers, ad makers are still operating under an outdated
schema that prefers men over women for no other reason than that it is the way they have always
done it. So, these ad makers—of which all of the well-known ones appear to be men: Terrance
Green, Mark Putnam, Fred Davis, Paul Wilson—are very focused on preexisting relationships and
established precedents. Many of them started working in the business before female voiceover
artists became more common. Therefore, they are probably going to continue to select more male
voiceover talents because that is who they have worked with and that is who they trust. So, there
seems to be a gender lag on the political voiceover business—a lag that is affecting the number of
opportunities available to female voice talents. Until the old guard political ad men retire, or until
there is a paradigm shift, there will likely continue to be significantly fewer female voiceovers in
political ads than male voiceovers.
Babb Interview.
Wilson Interview.
65 Klvana Interview (emphasis added).
63
64
30
The broader point here is that under this current paradigm, female voice talents are mostly
limited to a set of niche ads—ads targeted towards specific cohorts and not towards larger, more
general audiences. Findings IV and V both discuss this phenomenon. Finding IV detailed a political
voiceover universe that mostly limits female narrators to feminine centric ads, while men largely
narrate not only masculine themed ads, but also more widely targeted gender neutral ads. Further,
finding V discussed the widely held belief among ad makers that you should choose a woman to
narrate a negative political ad. However, Sheldon Smith strongly disagreed with that reasoning:
“I don’t think that’s necessarily true [that women should narrate negative ads]. I
think you pick the gender of the talent you hire based on how you want to convey the
message. And what the content of the message is. If it’s a message that is specifically
aimed at female voters and you need a female voice to convey that message then you
absolutely should hire a female. On the other hand, if it’s a more generic message it
could be done by either a male or female voice. So, I don’t happen to subscribe to the
theory that there is any standard rule of thumb that applies to who does negative
political commercials—or positive political commercials for that matter.”66
Essentially, Sheldon Smith argued that the voiceover selection process should be made on a caseby-case basis depending on the content and the message of each ad, and not by any overarching
rules dictating the use of differently gendered voices in certain specific situations.
But, this is a situation where the old adage, “it’s easier said than done” applies. As this paper
describes, persistent barriers remain between the current role that female voice actors play in the
political advertising realm—that of the niche voiceover artist—and a place where female and male
voiceover talents are utilized the way Sheldon Smith recommends. Kathryn Klvana makes the point
that, right now,
“We’re not at a point where there would be a woman who would be the main voice
for a campaign…Women tend to voice ads later in the campaign. This is when
campaigns start to mix up their ads, to start airing more diversely targeted ads. This
is when I get a lot of work. Men, such as Sheldon Smith, start much earlier in the
campaign with the more generally targeted, biographical spots. As a man, someone
like Sheldon is much more likely to be the ‘voice of the campaign.’”67
66
67
Smith Interview.
Klvana Interview.
31
Ironically, Sheldon Smith, who argued for a more balanced system of selecting voiceovers,
unintentionally represents one of the obstacles to the higher inclusion of female voiceovers in
political ads. He’s helping to perpetuate, through no fault of his own, a system that incentivizes the
selection of male voiceovers for ‘voice of the campaign’ style roles, and simultaneously excludes
women from such opportunities.
CONCLUSION
This research has shown that there is an evident gender bias in the selection of voiceover
talents for political ads. This bias does not, however, appear to be intentional on the part of the ad
makers. On the contrary, it appears to be merely a continuation of a method for choosing
voiceovers that adheres to an outdated paradigm based on relationships and intuition. But, there
are possibilities for change—possibilities that could shift the current paradigm and allow for
greater inclusion of female voiceover talents.
Despite the current underrepresentation by women in the field of political voiceover work,
female voiceovers are on an upward trend—at least in Presidential campaigns. In 2008, 23% of
political voiceover ads featured a female narrator, whereas in 2012, that number increased to 29%.
The author of this work is optimistic that these numbers will continue to increase. However, many
obstacles remain. Until a female narrator becomes ‘the voice’ of a major Presidential campaign,
female voiceovers will most likely continue to represent the minority of overall political voiceovers.
That is the next big step, and it will probably need to come from a female Presidential candidate. In
2016, perhaps Hillary Clinton, Elizabeth Warren, Kirsten Gillibrand, or Susana Martinez—if they
elect to run68—will decide to take a chance on a female voiceover, and show America that you do
not need to be a man, nor do you need to project an image of masculinity, to become President of
the United States.
Larry J. Sabato, “2016 Presidential Watch,” Sabato’s Crystal Ball, March 20, 2014,
http://www.centerforpolitics.org/crystalball/articles/category/2016-president/.
68
32
BIBLIOGRAPHY
“2012 Republican Presidential Nomination Polling Data.” Real Clear Politics.
http://realclearpolitics.com/epolls/2012/president/us/republican_presidential_nominatio
n-1452.html.
Bauerlein, Monika, and Clara Jeffery. “WTF, GOP?.” Mother Jones. May/June, 2012.
http://www.motherjones.com/politics/2012/04/gop-war-on-women.
Brader, Ted. Campaigning for Hearts and Minds: How Emotional Appeals in Political Ads Work.
Chicago, IL: The University of Chicago Press, 2006.
Breslau, Karen. “Hillary Clinton’s Emotional Moment. Newsweek. Last modified March 13, 2010.
http://www.newsweek.com/hillary-clintons-emotional-moment-87141.
Center for American Women and Politics, Eagleton Institute of Politics, Rutgers, The State University
of New Jersey. “Current Numbers of Women Officeholders.” Accessed April 21, 2014.
http://www.cawp.rutgers.edu/fast_facts/levels_of_office/Current_Numbers.php.
—. “Statewide Elective Executive Women 2014.” Last modified March, 2014.
http://www.cawp.rutgers.edu/fast_facts/levels_of_office/documents/stwide.pdf.
“Clinton’s Angelou Ad.” YouTube. April 30, 2008. https://www.youtube.com/watch?v=JtYwQvFd_bg.
Cohen, Nancy. “Women Fought the GOP’s 2012 ‘War on Women’ and Won.” The Guardian.
November 7, 2012. http://www.theguardian.com/commentisfree/2012/nov/07/womenfought-gop-2012-waronwomen.
Cusey, William A. “Are Negative Political Ads Effective? The Disconnect between Academics and
Professionals.” Communication Theory literature review, American University, 2012.
Delany, Colin. “Data-Driven Politics.” Campaign Insider. May 17, 2012.
http://www.campaignsandelections.com/campaign-insider/319937/datadrivenpolitics.thtml
33
Douglas, Susan. “Why Women Hate Hillary.” In These Times. April 26, 2007.
http://inthesetimes.com/article/3129/why_women_hate_hillary.
“Election 2012: Independent Spending Totals.” NYTimes.com. Last modified April 18, 2014.
http://elections.nytimes.com/2012/campaign-finance/independent-expenditures/totals.
“Exit Polls.” CNN.com. November 5, 2008.
http://www.cnn.com/ELECTION/2008/results/polls.main/.
Fridkin, Kim L. and Patrick J. Kenney. “The Dimensions of Negative Messages.” American Politics
Research 36, no. 5 (2008): 694-723.
Fung, Brian. “Microsoft is Using Your Data to Target Political Ads on Xbox Live.” The Washington
Post. March 7, 2014. http://www.washingtonpost.com/blogs/theswitch/wp/2014/03/07/microsoft-is-using-your-data-to-target-political-ads-on-xboxlive/.
“Gender & Gender Identity.” Planned Parenthood. http://www.plannedparenthood.org/healthtopics/sexual-orientation-gender/gender-gender-identity-26530.htm.
Goldwater, Barry. Why Not Victory? A Fresh Look at American Foreign Policy. New York: MacFadden,
1962.
Kellman, Laurie. “Social Issues Retake U.S. Politics, 2012 Elections.” Huffington Post: HuffPost
Politics via AP Wire. February 10, 2012.
http://www.huffingtonpost.com/2012/02/10/social-issues-politics-2012elections_n_1267779.html.
Lau, Richard R., Lee Sigelman, and Ivy Brown Rovner. “The Effects of Negative Political Campaigns:
A Meta-Analytic Reassessment.” The Journal of Politics 69, no. 4 (2007): 1176-1209.
Lawless, Jennifer L. and Richard L. Fox, “Girls Just Wanna Not Run: The Gender Gap in Young
Americans’ Political Ambition.” Women & Politics Institute, Washington, D.C., March 2013.
http://www.american.edu/spa/wpi/upload/girls-just-wanna-not-run_policy-report.pdf.
34
Lee, Penny. “No banking on the Women’s Vote.” U.S. News and World Report Online. April 4, 2013.
http://www.usnews.com/opinion/blogs/penny-lee/2013/04/04/hillary-clinton-shouldntcount-on-the-womens-vote-in-2016.
“Little Girl, Ice Cream Cone.” The Living Room Candidate: Presidential Campaign Commercials 19522012. Accessed March 16, 2014. http://www.livingroomcandidate.org/commercials/1964.
Macholan, Christina, and Evan Rowe. “Obama for America Uses Google Analytics to Democratize
Rapid, Data-Driven Decision Making.” Google Analytics Blog. August 6, 2013.
http://analytics.blogspot.com/2013/08/obama-for-america-uses-google-analytics.html.
Mann, Robert. Daisy Petals and Mushroom Clouds: LBJ, Barry Goldwater, and the Ad that Changed
American Politics. Baton Rouge: Louisiana State University Press, 2011.
Moore, Lori. “Rep. Todd Akin: The Statement and the Reaction.” The New York Times. August 20,
2012. http://www.nytimes.com/2012/08/21/us/politics/rep-todd-akin-legitimate-rapestatement-and-reaction.html?_r=0.
“President: Full Results.” CNN.com. last modified December 10, 2012.
http://www.cnn.com/election/2012/results/race/president.
Rich, Frank. “Stag Party.” New York Magazine. March 25, 2012. http://nymag.com/news/frankrich/gop-women-problem-2012-4/.
Ridout, Travis N. and Kathleen Searles. “It’s My Campaign I’ll Cry if I Want to: How and When
Campaigns Use Emotional Appeals.” Political Psychology 32, (2011): 439-458.
Saad, Lydia. “Economy Reigns Supreme for Voters: More than Half Rate it ‘Extremely Important’ to
their Vote for President.” Gallup.com. October 29, 2008.
http://www.gallup.com/poll/111586/economy-reigns-supreme-voters.aspx.
—. “Economy Is Dominant Issue for Americans as Election Nears: Total Mentions of Economic
Issues are as High Today as in October 2008.” Gallup.com. October 22, 2012.
35
http://www.gallup.com/poll/158267/economy-dominant-issue-americans-electionnears.aspx.
Sabato, Larry J. “2016 Presidential Watch.” Sabato’s Crystal Ball. March 20, 2014.
http://www.centerforpolitics.org/crystalball/articles/category/2016-president/.
Scherer, Michael. “Inside the Secret World of the Data Crunchers Who Helped Obama Win.”
Time.com: Swampland. November 7, 2012.
http://swampland.time.com/2012/11/07/inside-the-secret-world-of-quants-and-datacrunchers-who-helped-obama-win/
Shen, Aviva. “Inside The GOP’s Attempt to Change Its ‘War on Women’ Image.” ThinkProgress.
March 29, 2014. http://thinkprogress.org/election/2014/03/29/3419593/inside-thegops-attempt-to-change-its-war-on-women-image/.
Stone, Daniel. “How Voice Actors Are Chosen for Political Ads.” Newsweek. Last modified March 13,
2010. http://www.newsweek.com/how-voice-actors-are-chosen-political-ads-89351.
Strach, Patricia, Katherine Zuber, Erika Franklin Fowler, Travis N. Ridout, and Kathleen Searles. “In
a Different Voice? Explaining the Use of Men and Women Voiceovers in Political
Advertising.” Presentation at the Annual Meeting of the American Political Science
Association, Seattle, WA, September 1-4, 2011.
“The Data Driven Campaign.” The American Prospect. Accessed April 21, 2014.
http://prospect.org/article/data-driven-campaign
U.S. House of Representatives: Office of the Historian, Office of Art & Archives, Office of the Clerk.
“Mapping Congress.” U.S. House of Representative: History, Art & Archives. Accessed April 21,
2014. http://history.house.gov/Map/Mapping-Congress/.
“War on Women.” Wikipedia. http://en.wikipedia.org/wiki/War_on_Women.
West, Darrell M. Air Wars: Television Advertising in Election Campaigns, 1952-2008. Washington,
D.C.: CQ Press, 2010.
36
Download