Democratizing the Debates | The Annenberg Public Policy Center of

DEBATES

THE

DEMOCRATIZING

A Report of the Annenberg Working Group on

Presidential Campaign Debate Reform

TABLE OF CONTENTS

INTRODUCTION: THE ISSUE 5

THE GOAL OF REFORM: DEMOCRATIZING THE DEBATE PROCESS 7

Expanding and Enriching Debate Content 8

Broadening the Accessibility of theDebates 18

Improving the Transparency and Accountability of the Debate Process 20

CONCLUSION: 23

APPENDIX ONE: PROCESS 25

Working Group Biographies 26

APPENDIX TWO: SPONSORSHIP 31

APPENDIX THREE: YOUNGEST GROUP LEAST LIKELY TO WATCH MOST OFDEBATE 35

APPENDIX FOUR: 19602012 HOUSEHOLD RATINGS TRENDS: PRESIDENTIAL DEBATES 36

APPENDIX FIVE: ELEMENTS IN MEMORANDA OF UNDERSTANDING 37

APPENDIX SIX: AUDIENCE REACTION STUDIES 43

NOTES 47

Democratizing the DebatesDemocratizing the Debates

The Annenberg Debate Reform Working Group (for

biographies of members see Appendix One) was created

by the Annenberg Public Policy Center of the University

of Pennsylvania to explore ways to increase the value

and viewership of presidential general election debates,

taking into account the ways in which the rise of early

voting, the advent of social media, establishment of

new media networks, changes in campaign ﬁnance, and

the increase in the number of independent voters have

altered the electoral environment.

It would be dicult

to overstate the signiﬁcance of these changes.

 When general election presidential debates emerged

in 1960, network television dominated the media

landscape. In an era where 90% of television viewers

watched three networks, and had few other choices,

the “roadblocked” debates commanded the available

air space.

 When the Republican and Democratic political parties

formed the Commission on Presidential Debates

in 1988 (for a history of debate sponsorship see

Appendix Two), there was no Internet or vote by mail,

and “early” voting was primarily absentee balloting

with signiﬁcant restrictions.

 In 1988, there was no Fox News, no MSNBC, no

Univision or money spent by the campaigns on cable

advertising (and no digital on which to spend).

 In 1988, there were no Super PACs and by today’s

standards the levels of campaign-related “issue

advertising” and “independent expenditures” were

small. In a number of past decades, campaign

acceptance of federal ﬁnancing imposed limits

onspending.

Since 1960, transformational shifts in television

viewing – the plethora of cable channels, Internet

streaming, and other methods of viewing video

content – have dramatically eroded the power of

the “roadblock.” Nielsen data show that the percent of

U.S. TV households viewing the debates has declined

from 60% in 1960 to about 38% in 2012.

Additionally

Hispanic media now attract substantial audiences. In

both the July sweeps of 2013 and 2014, the number one

network among both those 18-49 and those 18-34 was

Univision.

In today’s environment, traditional media are working to

adapt to a world in which digital content is increasingly

a primary source of “news” for many Americans.

Moreover, the fastest-growing block of voters in

the country considers itself non-aligned

and not

represented by the major political parties who originally

formed the Commission on Presidential Debates. The

new “news” is often delivered in 140 characters, and

a voter’s most trusted information source is often a

“friend” from Facebook. Voting is conducted earlier and

earlier, by mail and in-person. Super PACs and other

funding organizations play an increasing role.

Those who organize debates and those who participate

in and watch them generally share the view that

overall: these exchanges should be informative and

not “canned”; the topics should be wide-ranging

and relevant to voter choice and interest; and the

vice-presidential as well as the presidential candidates

should be heard. The Working Group agrees. But

while in recent campaigns the debate process has

Democratizing the Debates

INTRODUCTION: THE ISSUE

Democratizing the Debates

incorporated more open formats, and moved from

the traditional timed answers and strict structures, in

general it has changed very little since 1992, when the

single moderator and town hall formats were adopted.

Very few “television programs” have succeeded with a

format that is more than two decades old.

Yet, these quadrennial events remain important. Not

only do they continue to attract a larger viewing

audience than any other campaign event or message,

but also, consistent with past data, a 2014 survey

conducted by Peter D. Hart and TargetPoint Consulting

for the Working Group found that “Presidential

Debates” were a top source of information in helping

voters with their decisions and deemed the “most

helpful” by a plurality of respondents. However, the

proportion viewing debates is not as large as it once

was or as it could or should be. Although Nielsen data

reveal that the numbers viewing broadcast and cable

debates have increased in every election year since

1996, the proportion viewing debates on TV and cable

is down from the 1992 level.

Importantly, a Nielsen

study commissioned by the Annenberg Public Policy

Center reveals that although viewership numbers have

increased somewhat since 1996, the largest growth is

among those aged 50-64,

a cohort socialized in an

earlier media era.

At the same time, substantial numbers fail to watch an

entire debate or multiple ones. Across the 2004-2012

election years, a plurality of debate viewers watched

only a single debate

(See also Appendix Three). In

particular Nielsen analysis conﬁrms that:

In 2012, 20.2% of viewers 18 or older

watched at least some of one debate, 16.2%

watched at least some of two debates, 15.3%

watched at least some of three debates

and 14.9% watched at least some of all four

debates. Across all debates, those who

viewed at least some of one averaged 35.1

minutes of viewing time, those who watched

at least some of two averaged 47.8 minutes,

those who viewed at least some of three

averaged 66.1 minutes and those who viewed

at least some of four averaged 172.5 minutes

of viewing time.

There is no question that debates have a unique

capacity to generate interest in the campaign, help

voters understand their choices in the upcoming

election, forecast governance, and increase the

likelihood that voters will cast a vote for the preferred

candidate rather than primarily because of opposition

to the opponent. With needed reforms, presidential

general election debates can do a better job of meeting

these goals and can also increase the level and amount

of viewership; without change, the proportions viewing

debates may decline further and the levels of viewership

among two important constituencies – the young and

Hispanics - stagnate.

Hempstead, NY - October 15,

2008: Senator McCain and Senator

Obama participated in the third of

three presidential debates, which

focused on the economy and

domestic policy. The candidates

were seated at a table with

moderator Bob Schieer.

THE GOAL OF REFORM:

DEMOCRATIZING THE DEBATE PROCESS

Overall, the eect of the desired changes is a

democratization of the debate process. To this end,

and as discussed at length in this Report, the Working

Group recommends:

Expanding and Enriching Debate Content

 Increase direct candidate exchanges and otherwise

enhance the capacity of candidates to engage each

other and communicate views and positions;

 Reduce candidate “gaming” of time-limited answers

and create opportunities to clarify an exchange or

respond to an attack;

 Enlarge the pool of potential moderators to include

print journalists, university presidents, retired judges

and other experts;

 Use alternate formats for some of the debates,

including a chess clock model that gives each

candidate an equal amount of time to draw upon;

 Expand the role of diverse media outlets and the

public in submitting questions for the debates; and

 Increase the representativeness of audiences and

questioners at town hall debates.

Broadening the Accessibility of

theDebates

 Embrace social media platforms, which are the

primary source of political information for a growing

number of Americans, and facilitate creative use of

debate content by social media platforms as well as

by major networks such as Univision, Telemundo,

and BET, by providing unimpeded access to an

unedited feed from each of the cameras and a role

in framing topics and questions; and

 Revise the debate timetable to take into account the

rise of early voting.

Improving the Transparency and

Accountability of the Debate Process

 Eliminate on-site audiences for debates other than

the town hall and, in the process, reduce the need

for major ﬁnancial sponsors and audiences ﬁlled

with donors;

 Publicly release the Memorandum of Understanding

governing the debates as soon as it is signed;

 Require the moderators to be signatories to the

MOU to ensure compliance with the agreements

about rules and formats; and

 Clearly articulate the standards required of polls used

to determine eligibility for the debates.

Reform of the presidential debates should be accomplished by re-shaping formats, emphasizing

candidate accountability, better aligning debates with the changing attitudes of the electorate, and

modifying the debate process and timing.

Democratizing the Debates

Data on total viewership for debates

from the Commission on Presidential

Debates Archive: http://www.debates.

org/index.php?page=debate-history;

Total voting age population data from

the Federal Election Commission: http://

www.fec.gov/pages/htmlto5.htm

Expanding and Enriching Debate Content

The proportion of the electorate viewing debates

is substantially lower than it once was (see also

AppendixFour).

Moreover, debates are not giving voters as substantive

an understanding of the candidates as they might.

Candidates and their party representatives view them as

a hybrid of Sunday morning interviews and gladiatorial

clashes, and express frustration with the constraints the

joint press conference structure imposes on their ability

to communicate their positions, priorities and core

political messages, and clarify distinctions between or

among the candidates.

The Working Group believes that debate formatting

needs to be rethought. There are several contributing

reasons. Across the past half century of scholarship on

debates, scholars have noted how format limitations

“have made it dicult for audiences to see the ‘real

substance’ of the candidates’ positions and policies”.

These same formatting conventions “not only thwart

sustained discussion of serious issues, but also

encourage one-liners and canned mini-speeches”.

Figure 1. Average percent across all debates of voting age population that watched the

presidential and vice presidential debates on television.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

57.81

39.31

41.19

36.13

32.7

33.03

18.49

18.26

23.02

26.16

25.26

1960 1964

-1972

1976 1980 1984 1988 1992 1996 2000 2004 2008 2012

Looking back at the 1960 debates is a striking exercise

that oers a model for considering reforms in the

current structure for the Working Group. With eight-

minute opening statements and longer answer times,

both personalities and positions were clearer and more

compelling. The highly moderated format we have

today has produced shorter answers, but the result has

been less substance and more equivocation. At the

same time, strict time allotments treat all questions as

equally important and encourage candidates to use all

of the allocated time as well: if the candidates try to

respond at greater length, they appear to “ﬁlibuster;”

ifthey abbreviate their responses, they convey lack of

interest or knowledge. In any case, the short-answer

format rewards those who resort to clever quips and

sound bites.

Because it believes that the general format has calciﬁed

over 52 years, and especially the past 20 years, when

innovation has largely stopped, the Working Group

has adopted two fundamental goals for evaluating

alternatives: (1) The voters should learn more about

who the candidates are, what they stand for and

what they would do in oce, and (2) The candidates

themselves should be responsible – and therefore

accountable – for the quality of their performance. With

a candidate-centric format, success or failure rests on

the individual candidates’ shoulders.

Figure 2. Percent of voting age population that watched the most widely viewed debate on

television.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

60.83

45.76

48.97

38.57

36.82

36.88

23.46

22.64

28.2

30.23

27.89

1960 1964

-1972

1976 1980 1984 1988 1992 1996 2000 2004 2008 2012

Data on total viewership for debates

from the Commission on Presidential

Debates Archive: http://www.debates.

org/index.php?page=debate-history;

Total voting age population data from

the Federal Election Commission: http://

www.fec.gov/pages/htmlto5.htm

Democratizing the Debates

Rethinking Formats for Debates

Many of the Working Group’s recommended format

changes have been proposed before. In one major

study, respondents who had participated in “Debate

Watch,” a voter education program involving tens

of thousands that was originally associated with

the Commission on Presidential Debates, favored

“something closer to a Lincoln-Douglas debate with

less intrusion from the moderator. This would include

cross-examination by the candidates, opening and

closing statements…a limited number of topics per

debate, dierent topics in each debate, more ﬂexible

time limits that would allow for more depth of analysis

and clearer comparisons and contrasts between or

among positions while avoiding discussion of a topic

from a previous question during a subsequent topic,

rules that allow the moderator to keep the candidates

on the topic...”

That work noted as well that “many

participants expressed a belief that cross-examination

would improve the debates by making candidates

more spontaneous and by providing viewers with

better information and bases for comparison. One of

the major criticisms of the existing formats was that

they did not produce enough interaction between the

candidates…” Because the Annenberg Working Group

agrees with many of these sentiments, its members

recommend retention of the town hall, the addition

of two new formats—the chess clock model and the

reformed standard model—and a re-evaluation of the

roles of the press and moderator.

Alternative Formats

At the core of the Working Group chess clock format

is an idea that has been circulating for more than two

decades. The seventh recommendation of the 1986

Institute of Politics-Twentieth Century Fund Report

also known as the Minow report (see Appendix Two)

read: “The use of journalists as questioners should

be eliminated in favor of allowing the candidates the

opportunity of questioning each other.”

The “Chess Clock” Model. Under this model, each

candidate is allotted approximately 45 minutes of

speaking time. Eight topics with equal blocks of time

are provided. Anytime a candidate is speaking, that

candidate’s clock visibly counts down. To take control

of the ﬂoor, a candidate simply hits the chess clock. No

answer, rebuttal or question may exceed three minutes.

The hard time stops are agreed upon; when a candidate

runs out of total time, he or she has exhausted the right

to speak. Remaining time at the end of the moderator-

posed topics can be used for a closing statement.

St. Louis, MO - October2,

2008: Democratic

supporters watch

the televised debate

between Democratic

vice presidential

candidate Joseph Biden

and his Republican

counterpart Sarah Palin

at the Social Hollywood.

Vicepresidential nominees

Palin and Biden clashed

at their crucial vice

presidential debate.

In the chess clock model, the candidates, not the

moderators, would be responsible for follow-up. As

discussed below, reform of the standard model would

have the moderator enforce time limits and raise topics

culled from a variety of sources. The moderators would

not be tasked with asking follow-up questions; instead,

candidates would be expected to challenge incomplete

and nonresponsive statements by the opposing side.

The topics would be drawn from a pool of submissions

from broadcast, print and social media, and from

the general public, vetted by an editorial committee

formed perhaps of representatives of leaders of the

presidential libraries, presidents of major public and

private universities, or the heads of major public

libraries. The selection ﬁlter is the question, “What will

a President face while in oce – and how is he or she

likely torespond?”

Under this model, candidates can choose to go into

greater detail on matters of greater importance to them;

they are not compelled to pad time on others.

The Reformed Standard Model. Under the reformed

standard model, there would be no chess clock and the

time would be allocated between the candidates on

the more traditional understanding: e.g., one minute for

response, 30 seconds for rebuttal. The changes would

otherwise remain the same as in the chess clock model

in two respects: the role of the moderator and the

source of the questions. Two additional features, giving

candidates more ﬂexibility to rebut or clarify and the

opportunity to prepare statements on some topics in

advance, would be added to increase their opportunity

to engage the opponent and feature central points of

their agendas.

This ﬁrst, which involves allocating to each candidate

two “points of personal privilege” in congressional

terms, or “challenge ﬂags,” allows each candidate to

exercise two 90-second opportunities to deviate from

the format and make a statement. This allows each

to clarify a previous response or respond to an attack

when the formal format would by rule (if enforced)

preclude it.

The second additional feature would call for each

candidate to be given two dierent topics ahead of time.

Each may prepare a four-minute response; the other

candidate, also supplied with the topics in advance,

has equivalent time to oer a counterstatement, rebut,

and cross-question the ﬁrst candidate. Alternatively, in

advance of one of the debates, each candidate would

select two topics with two additional ones decided

by the moderator through the reformed process

recommended in this Report, and two by some form of

social media ballot.

Democratizing the Debates

Town Hall Debates

The town hall format is an important feature of

presidential debates. Although the ﬁrst presidential

debate of 2012 drew the largest number of viewers,

the town hall “had a higher rating and held viewers’

attention longer.”

The town hall would follow roughly

the same format as used in recent years, but with,

again, the adjustment in the role of the moderator,

who would enforce process requirements and time

limits, but would not have a role in asking follow-

up questions or supplement the roster of questions

posted by the citizen-participants. The town hall

debate should include no live audience other than the

citizens who are the participants in the town hall, and

who ask the questions. The questioners should be

selected from undecided voters in battleground states

and not, as has occasionally been the case in the past,

from communities in less competitive jurisdictions

that do not oer the same desirable diversity of views.

The set should be designed to minimize the physical

“trac” both between the candidates and between the

questioners and the candidates, allowing for an orderly

and clear progression of debate exchanges.

Re-evaluating the Role of the Press

andModerators

At the request of the Annenberg Working Group, in

March 2014, Peter Hart convened ﬁve focus groups

in Denver involving individuals who voted in 2012 and

reported having watched some or all of one or more

2012 presidential general election debates.

“The single

largest criticism of the debates centers on the inability

of moderators to do their job,” concluded Hart. “Some

participants perceive some moderators to be biased and

ruining the fairness of the debate. Others complain that

the moderators either do not have the skills to control

the candidates or to call them on ‘non-answers.’”

Table 1: 2012 Debate Moderators—Concerns About Complaints

By Age (Extremely/Very Concerned Combined)

18-34 (A) 35-49 (B) 50-64 (C) 65+ (D) Total

The moderators tend to play favorites, giving one of the candidates

theedge

32%

BCD

43%

44%

41%

The moderators tend to lose control of the debate and the candidates

interrupt one another and go over the time limits

29%

27%

36%

42%

33%

The moderators overstep their boundaries and inject themselves in the

debate process

23%

32% 33%

35%

30%

The questions moderators pose are not the right questions on the

important issues

26% 27% 30% 33% 29%

A,B,C,D=statistically signiﬁcant at

95% conﬁdence level (p<.05). For

a fuller explanation, see note on

AppendixThree.

“Please tell me whether you are

extremely concerned, very concerned,

just somewhat concerned, or not

at all concerned about each of the

following complaints. If you have not

noticed this complaint, just say so.”

The national survey conducted for the Working Group

by Hart reinforces concerns about the role of the

moderator.

The Working Group raised an additional set of concerns.

In recent years, the moderator has been a television

journalist, and more often than not a current or former

anchor of a major network news program. When a

network uses a debate as an opportunity to showcase

its broadcasting talent and enhance its brand, the

somewhat arbitrary selection of moderators from a

maximum of four networks in a world now populated

by many more than that creates a competitive

advantage for the selected outlet. One result of all of

this is that the debates can take on the appearance of

marketing opportunities for the network whose reporter

or anchor is moderating. Moreover there is pressure on

the campaigns to push for or accept certain moderators

for debates. Also of concern is the fact that moderator

selection is not a transparent process.

The format changes recommended by this group would

address many of these concerns by focusing the role of

the moderator on moderating.

A review of press coverage surrounding the 2012

general election conﬁrms that the moderators – their

performance, the reactions to their performances,

perceived tilts toward one candidate or another – have

become an integral part of the presidential debate

story. This attention to the role of the moderator stems

in part from two aspects of the role which raise the

question of whether these debates are best moderated

by individuals in their role as journalists or, moderated

by an individual whose sole responsibility is ensuring

that the debate process works well. The Working Group

favors the latter. The criticisms of debate formats as

joint news conferences or joint Sunday show-type

interviews reveal the inherent tension in the role of

journalists acting in their capacity as journalists while

also performing as moderators.

At present, in the debates other than the town hall,

the moderator decides which questions are asked, or

which topic areas are covered, with as much or as little

input from the public or other journalists as he or she

wishes. This can result in questions that advance the

news agenda more than public understanding of the

candidates, their plans and position on issues. Press

Richmond, VA - October 15, 1992:

Democratic candidate AR Gov. Bill

Clinton (foreground) making point as

rivals GOP Pres. Bush & Independent

candidate TX businessman Ross Perot

listen, in 2ndpresidentialdebate.

Democratizing the Debates

control of content and pursuit of “follow-up” can create

an interview or Sunday show dynamic in which the

candidate is engaged with the moderator, as opposed

to the other candidate. Candidates end up preparing to

debate the moderator as well as their opponent(s).

Consistent with our belief that the debates should

be a forum for the views of, and exchanges between,

candidates, the Working Group recommends the

following areas for reform:

Development of Questions: The debates should

employ a more formal process of soliciting topics

and questions both from the general public through

a variety of platforms as well as from a broad group

of knowledgeable experts that would include print as

well as broadcast journalists. The questions could be

curated by a group, potentially made up of directors

or members of boards of presidential libraries and

major public libraries, or public and private university

presidents, with the moderator responsible for framing

the questions. This reform would invite greater diversity,

and give both the public and a broader representation

of the press corps an opportunity to identify topics and

questions that the debate should cover.

We believe that a full third of the questions by the

moderator in the debate should be obtained from

non-news sources. On an individual level, this change

gives voters and politically interested Americans

greater opportunity to shape the debates. Involving

the audience, however, demands participation that

enhances the debate viewing experience. Audience

participation should be much more than a novelty

– it should contribute to the greater dialogue and

provide a meaningful way to participate. Moreover,

audience participation has the potential to help direct

conversation and reaction before and after the debate

and in the process increase interest. If well structured,

a high level of interaction between the public and

the people running for president furthers our goal of

helping Americans make the most informed choices.

Expanding Pool of Potential Moderators: Television

networks argue that live, televised events can be

eectively moderated only by experienced broadcast

journalists. The challenge of moderating the debate,

with the producer giving guidance in the moderator’s

ear, is real – trying to make sure time is allocated fairly,

that the order of responses is correct, and that the

candidates’ focus shifts as needed to dierent topic

areas. Moreover, there is a belief that journalists oer an

informed perspective that ensures that important topics

are covered and candidates answer the questionsasked.

However, as noted, a moderator’s control of content

and pursuit of “follow-up” can create an interview

or Sunday show dynamic in which the candidate is

St. Louis, MO - October11

1992: Democratic

presidential candidate

Bill Clinton (center), U.S.

President George Bush

(right) and presidential

candidate Ross Perot

(left) answer questions

at the athletic center at

Washington University

during the ﬁrst of four U.S.

presidential debates.

engaged with the moderator, as opposed to the other

candidate.

With a more limited role for the moderator, there is no

reason that the potential moderator pool could not

be broadened to once again include print journalists

as well as other persons of stature such as university

presidents, retired judges, historians, and others

with demonstrated credibility– as was proposed at

the birth of the televised debates in 1960. This also

would address the diversity issue. In the past, debate

moderators have not reﬂected the diverse makeup of

the country.

Adopting a More Inclusive, TransparentSelection

Process: The current process of choosing moderators

is not transparent. We recommend that a designated

group, again potentially made up of presidential library

heads and board members, develop an initial list and

that the campaigns select a moderator from that list.

While the system may produce one moderator who

would facilitate more than one debate, we would

strongly suggest using dierent moderators for dierent

debates to add variety and increase public interest.

The Question of the Criteria for the

Inclusion of Minor Party or Independent

Candidates

Over the course of the presidential debates, there has

been limited independent or minor party candidate

participation. In 1980, independent candidate

JohnAnderson was included in the ﬁrst debate

(although Democratic nominee and incumbent

President Jimmy Carter declined to participate). In 1992,

independent candidate Ross Perot participated in all

of the presidential debates, but was not invited to the

debates four years later. Those are exceptions to the

general pattern that presidential debates feature only

the two major party candidates.

The Commission on Presidential Debates administers

a two-part test for inclusion of candidates: (1) Any

candidate included must have the ability to be elected

in the general election by qualifying for ballot inclusion

in states adding up to at least 270 electoral votes, and

(2) Any candidate who passed the ballot test must reach

at least 15% in independent national surveys in the

period leading up to the debates.

Whether they identify themselves as independents or

non-aligned or just refuse to state a preference, one

of the signiﬁcant changes in American elections since

the Democratic and Republican parties formed the

Commission in 1987 is the growth of those who call

themselves non-aligned voters in this country.

As important is the fact that 50% of those in the

millennial generation, now ranging in ages from 18-33,

described themselves as political independents in

March 2014.

Given this reality, and the fact that the independent/

non-aligned candidates have succeeded in winning

statewide races over the past decade, the Working

Group discussed whether the time has come to revisit

the standard for including candidates. It heard views

on this topic from advocates of liberalized standards

for the inclusion of independent candidates, including

from those arguing for a guaranteed invitation for at

least one independent candidate regardless of the

person’s standing in the polls and electoral viability. It

has been argued that the rules should take account of

the possibility that through inclusion in the debates an

independent candidate could build the potential for

victory that he or she did not have at the outset.

Democratizing the Debates

There is support in both the focus groups and in the

survey for a lower entry threshold. Where 41 percent of

those surveyed favor the status quo, 47 percent oppose

limiting “the debates to the two major party candidates

unless a third-party candidate can exceed 15% in the

polls.” The survey asked which of two statements came

closer to the respondent’s view:

Statement A: The rules for a third-party

candidate inclusion should be relaxed so

that it is easier for them to be part of the

debate. Even if it is unlikely that they will

win the presidency, it would make the major

candidates respond to their ideas.

Statement B: The rules for a third-party

candidate inclusion should not be changed,

because the third-party candidates take away

from the central purpose of listening to and

watching the two people who are most likely

to become the president.

In response, 56 percent said the rules should be relaxed

while 28 percent said they should not be changed.

Fifteen percent oered no opinion or did not know.

Similarly, respondents in Hart’s focus groups favored

“making it easier to allow third-party candidates in on

the debates.”

Figure 3. Party Identiﬁcation, Yearly Averages, 1988-2013

Based on multiple day polls

conducted by telephone.

Jones, J. M. (2014,

January8). “Record-High

42% of Americans Identify

as Independents.” Retrieved

August 4, 2014 from

http://www.gallup.com/

poll/166763/record-

high-americans-identify-

independents.aspx

% Republican

% Independent

% Democrat

‘88 ‘90 ‘92 ‘94 ‘96 ‘98 ‘00 ‘02 ‘06‘04 ‘08 ‘10 ‘12

% Republican

% Independent

% Democrat

‘88 ‘90 ‘92 ‘94 ‘96 ‘98 ‘00 ‘02 ‘06‘04 ‘08 ‘10 ‘12

Indiscussing alternatives for the American election, the

Working Group examined the current 15% threshold

for inclusion, measured by public opinion polling,

coupled with qualiﬁcation for the ballot in enough

states to win a majority of the Electoral College. It

considered the balance to be struck between ensuring

a diversity of views, and giving voters the opportunity to

consider the views of the candidates—the major party

candidates—with the highest likelihood of being elected.

On this question, there was not a consensus on the

bestsolution.

The Working Group concluded that the debate must

remain principally an exchange of views for the beneﬁt

of voters who are faced with a choice of a potential

president, and that therefore the debaters should have

by a fair measure a realistic chance of winning the

election. But the Group could not arrive at a consensus

and has not therefore made a recommendation for a

modiﬁed standard.

However, a majority of the Working Group agreed that

the selection criteria now in place should be replaced

with a structure that, in addition to demonstrated

capacity to win a majority of electoral votes, would

involve; a) lowering the threshold for the ﬁrst

presidential debate to 10%, b) raising it to 15% for the

second; and then c) increasing it to 25% for the third

and ﬁnal debate. This process would facilitate inclusion

of third-party or independent voices at the outset

of the debate schedule, while requiring a showing

of expanded support as the campaign—and debate

period—continues. These proposed changes respond to

the argument that independent candidates have to clear

too high a hurdle in the ﬁrst instance, and if given a

greater chance at the outset to participate, may be able

to build support.

Because standing in polls plays a critical role in

determining eligibility for debate participation, the

Working Group does believe that it is important that the

standards required of polls used to determine eligibility

are clear; the number and identity of the polls on which

the decision will rely are announced in advance; the

survey question that will be used to assess eligibility

is disclosed in advance; and answers are provided

in advance to such questions as: What happens if

a candidate falls below the polled threshold but is

nonetheless within the margin of error?

Democratizing the Debates

Broadening the Accessibility of

theDebates

The Internet has produced increased democratization

in the political process. Almost overnight, information

became available from a variety of sources, not just

the major networks, and opened a platform to the

average person for voicing opinions and sharing news

andcoverage.

Debates have failed to keep up with the evolving digital

viewing habits of the American public. At the same

time, social media have altered the ways in which the

public consumes the debates. The May 2014 survey

conducted by Hart found that “When young people

do watch debates, they are signiﬁcantly more likely

to actively follow the debate through social media

platforms, such as Twitter or Facebook. More than

a quarter (28%) of 18-34 year olds and about a ﬁfth

(19%) of 35-49 year olds said they both watched and

actively followed the 2012 debates on social media

platforms. Signiﬁcantly fewer older adults reported such

activity (12% 50-64 year olds and 8% of 65 and older).”

Broadcast and cable networks play an essential role, but

the shifting media habits of the American electorate –

characterized by the expanding number of cord-cutters

and o-the-grid segments – are noteworthy, and by

2016, the shift to digital media consumption will be

even morepronounced.

We would recommend embracing this changing

media landscape by providing full access to debate

content on a ﬂat, universal feed developed according

to predeﬁned, public technical standards and shared

with media companies and individuals alike. We would

also complement a common feed with clean, succinct,

accessible data and analytics. In the Hart survey, 55%of

the public surveyed and 69% of those 18-34 support

streaming debates over the Internet on external outlets

simultaneously as the debate is going on. Doing so

frees campaigns and media providers to focus on

what matters: creating valuable debate experiences for

thevoters.

To facilitate these changes, technological infrastructure

should provide a level playing ﬁeld for competition and

innovation. Universal access is ideal for this purpose.

Flexible and adaptable, this foundation will allow the

market to develop media delivery models and oerings

that suit viewers of dierent ages and habits.

Digital content providers deserve a central role in

setting these universal requirements. Even though we

don’t know what those will look like in 2016, let alone

2020 or 2024, if debates are to stay relevant, they must

adapt to the variety of viewing habits and technologies

in use.

An election comes down to the millions of individual

decisions made by voters. The voters—not broadcasters,

media providers, or networks—ought to determine

how, where, and when debate content is used. Open,

universal access to debates, with meaningful content

presented in a relevant way, can help ensure the

continuing viability of the presidential debates well into

the future.

Without the ability to predict the media technology

that will shape the future, the path to increased

consumption of debates is clear: eliminate the points

of friction that keep voters from being able to access

the debate on their platform of choice.

Revising the debate timetable to take into account the

changes in voting behavior since 1988, particularly

early voting. Because the fall calendar of an election

year includes the Summer Olympics, Major League

Baseball’s post-season, and the Jewish high holidays,

the challenges confronting presidential debate

schedulers are signiﬁcant. When party conventions are

held in late August or early September, the schedule

is compressed even more, with the result that debates

often cluster into the month of October.

Importantly, the increase in early and no-excuse

absentee voting means that October debates occur

after balloting has actually begun. In 2012, the

percentage of voters who cast ballots before Election

Day was 31.6%, according to a Census Bureau study.

The aected proportion of the electorate is large. Thirty

two states currently allow early voting. Washington

and Oregon conduct all of their voting by mail before

Election Day. The earliest early voting currently

occurs 45 days before Election Day (in South Dakota

and Idaho). While the data indicate that only a small

percentage of voters voted prior to the ﬁrst debate

in 2012 (0.90%), a more signiﬁcant number cast their

ballots before the third debate (6.89%) (see chart below).

Figure 4 shows that the number of absentee and early

Figure 4. The percent of absentee and early voting by date from 8 battleground states in the 2012PresidentialElection

States included: Colorado, Florida, Iowa,

Michigan, Nevada, Ohio, Virginia, and

Wisconsin. Data come from Catalist

Absentee/Early Voting daily reports.

Democratizing the Debates

voters increases rapidly day by day in the last two weeks

before the election, which reduces the appeal of having

a debate closer to Election Day. Moreover, all of the

active duty military and their families living abroad vote

absentee well in advance of Election Day.

Since fundraising and university logistics dictate that

the Commission pick dates and sites over a year in

advance of the election, there is currently no ﬂexibility

built into the debate schedule. However, there is

precedent for increasing ﬂexibility in scheduling. In

1996, the four Clinton-Bush-Perot debates were held

outside of the Commission’s announced schedule over

a nine-dayperiod.

In response to increased early voting and the important

role the debates play in informing voters about the

candidates, the Working Group recommends that:

 The ﬁrst debate should be scheduled mid-September

to give military families and voters who participate

in early voting the opportunity to see at least one

debate before casting their votes. Sixty-three percent

of the respondents in the Hart survey favored moving

the ﬁrst debate to early September before early

voting begins.

 The “debate season” of three presidential debates

and one vice presidential debate should occur in

a window of 19-25 days. The decision on the ﬁnal

schedule for the debates should be set by July 1 of

the election year.

 On-site audiences should be dispensed with to

eliminate the need for booking sites far in advance

and provide greater ﬂexibility in timing and location.

Improving the Transparency and

Accountability of the Debate Process

The Memorandum of Understanding

Since the presidential debates are an event held for the

beneﬁt of the public, they should be structured to meet

basic standards of transparency and accountability.

The adjustments necessary to achieve these goals are

neither complicated nor controversial.

The Memorandum of Understanding (MOU) negotiated

by the candidates settles various terms and conditions

of the debate (for a synthesis of MOU content, see

Appendix Five). In the past, these MOUs have not

been made routinely available to the press and public.

Because it works out details speciﬁc to a given venue

and addresses concerns important to the campaigns

(e.g., a riser for shorter candidates, whether a candidate

can take or bring notes), the MOU is an agreement

important to the debate process.

Upon signing, the MOU negotiated by the candidates

should be made available to the press and public. The

MOU should be posted to the website that the Group

recommends be maintained to provide supplemental

information about debate topics and to facilitate

citizenengagement.

In addition, all the key participants in the debate,

including the moderator, should sign the MOU. It

makes little sense to have the candidates negotiate

understandings that must be enforced by a moderator

who is not a party to the agreement. In the past, the

moderators, as representatives of the media, have

declined to sign the MOU out of a desire to protect

their role as journalists who are expected to keep their

distance (and objectivity). The Working Group has

recommended reforms to address this conﬂict in role,

proposing changes in the role of moderator and in

the role of the press generally. In the proposed model,

journalists who serve as moderators are working as

moderators, not journalists, during the debate: they

have agreed to perform the task as a public service, not

as an act of reporting. While reporting is unquestionably

an act of public service in other contexts, in this one,

it presents the live and demonstrated risks discussed

in the testimony to the Working Group. While under

these proposals members of the press would remain

available for selection as moderators, the Working

Group concludes that the person who wishes to be a

moderator should only do so on the condition that he

or she signs the MOU.

Reducing the Debate “Spectacle”

and its Cost

The debates have become an extravaganza, elaborate

in design and costly. Hosting universities construct

temporary buildings or retroﬁt spaces not just to

create a debate venue, but also for spin alleys,

candidate holding rooms, surrogate viewing areas,

press ﬁling centers, sta work spaces, and ticket

distribution. Streets are closed and transportation

systems developed to accommodate the movement

of Secret Service motorcades and hundreds of

audience members to and from the debate site. A

television studio is built in the debate hall; technology

is installed to give media and the campaigns access

to communication systems. Thousands of media,

campaign sta, surrogates, and audience members

travel to the site. In the hours prior to the debate,

corporate sponsors provide food and entertainment for

an audience comprising policy makers and inﬂuentials.

In years past, the debate site supported not just the

audience in the hall but also the crowd that gathers for

the events surrounding the debate and the surrogates

representing the campaigns before and after the debate

and available to “spin” themedia.

Substantial private funding is needed for a spectacle

on this scale. With the private ﬁnancing comes product

promotion and the spending for corporate branding

and “goodwill.” How these arrangements are reached,

and on what terms, should be a matter of general

public interest, but little information is provided or

known. Although the overall cost of producing this

extravaganza is high, little of it is directly necessary for

the main event.

All that is necessary for a presidential debate is a stage

for the candidates and a mechanism to transmit their

words and images to the public. Other features of

current debates including the spin room, the audience,

the beer tents, and the locked-down university campus

St.Louis, MO - October17,

2000: Republican

presidential nominee

George W. Bush (center)

answers a question

during his town hall style

debate with Democratic

presidential nominee

AlGore (center, right) at

Washington University

Democratizing the Debates

contribute to a spectacle that distracts from the main

purpose of the event: the discussion of the major issues

among the candidates for president.

No Spin Alley. Other costs now typical of the debate

process have become less necessary and useful than

in the past. For example, with the rise of social media,

the value of “spin alley” has diminished as the senior

campaign voices are more likely to use email or Twitter

to engage the press both during and after the debate.

These changes substantially lessen the need for an elite

facility where chosen political spinners and credentialed

journalists gather in person to engage in a tired ritual.

No In-Person Audience. The presence of the in-

person audience not only raises questions about the

seemliness of its composition but its presence carries

risks as well. Once the debate begins, despite warnings

to remain silent, audience reaction can and has

aected the impressions of those viewing at home (See

Appendix Five). Laughter, cheers or jeers also magnify

moments and distract attention from the substance

of the statements made by the candidates. Although

it is sometimes said that these eruptions are primarily

a problem for primary debates and have been rare in

general election debates, there is no reason to assume

that this good fortune will last. After all there have been

audible audience responses in general election debates:

examples include audience reaction to President

Reagan’s answer to a question about his age and the

response to the exchange between Senators Lloyd

Bentsen and Dan Quayle over any comparison of the

latter with President John Kennedy. Even one such

episode is too many.

Some argue that the presence of a live audience

provides positive energy that can bring the best

performances out of the candidates. However, the

presidential debates that are routinely put forth as the

most consequential and substantive ones, the 1960

Kennedy-Nixon debates, were held in a TV studio with a

very small on-site audience.

There is ample precedent for using a television studio

for political debates. In addition to the 1960 debate, the

1976 Carter-Ford presidential debates and the 2010

United Kingdom general election debates were held

in studios. Television studio debates are routine for

gubernatorial and senatorial elections.

Moreover, as discussed earlier, the audience creates the

need for a large logistical footprint that increases the

cost of the overall production and raises transparency

and accountability issues.

Some point to the presence of students from the

host university in the audience as justiﬁcation.

However, the percentage of the audience made

up of students relative to donors is small. There are

more cost-eective ways to involve students in civic

education (generally) and debate education (speciﬁcally)

than staging a onetime, multimillion-dollar event

oncampus.

CONCLUSION:

The Annenberg Debate Reform Working Group has presented this analysis and related recommendations in the

service of Presidential debates that reﬂect major changes in our electorate, politics and media. More can and should

be done to enrich their content, enlarge their audience and improve upon their accessibility. The members of the

Working Group are conﬁdent that while views about the particulars will vary, there is likely near-unanimity about

the vital importance of Presidential debates and, therefore, of the need to ensure that they continue to answer

the needs of the voters who have watched and listened to them, and to draw into our political process those who

havenot.

Democratizing the Debates

APPENDIX ONE

PROCESS

Held on September 26-27, 2013 — the 53rd anniversary of the ﬁrst Kennedy-Nixon debate — the Working Group’s

ﬁrst meeting focused on debate topics, the format of the debates, moderators, timing and number of the debates,

negotiations with candidates, audiences and interactivity, and debate sponsorship. Its second meeting, on

November 8, 2013, concentrated on the role of media in future presidential debates and the logistics involved in

producing them. The meeting included discussion with social and legacy media representatives and with those who

have handled logistics for campaigns in recent debates. In the third session on December 16, 2013, the Working

Group heard from scholars who have studied third-party candidacies as well as from individuals urging alternative

criteria for inclusion of third-party candidates in debates. Convened in Cambridge, Md., in early February 2014, the

fourth meeting reviewed what had been learned and identiﬁed research required to address unanswered questions.

On March 27-28, 2014, the ﬁfth meeting, which like the ﬁrst, second and third was held at the Annenberg Public

Policy Center in Philadelphia, focused on the logistics involved in producing debates, the ways in which social

media might be harnessed to increase the value and viewership for debates and the role of non-legacy networks.

The group focused on ﬁnancing models and on understanding the costs involved in alternative models of debate

production. Shifting to the West Coast for its sixth session May 11-13, 2014, the group met with executives from

major social media organizations to explore ways to increase audience interaction with debates and debate content.

During its seventh session June 18-19, 2014, the Working Group discussed the drafting of this report.

The Working Group also:

* discussed the history and value of presidential debates with Newton Minow, who played an indispensable

role in the institutionalization of presidential debates.

* met with broadcast and cable network representatives as well as with the director of the Commission

on Presidential Debates (CPD), Janet Brown, and separately with CPD co-chairs Mike McCurry and

FrankFahrenkopf.

* convened a group of debate scholars to assist the Working Group in evaluating debate formats.

To support the work of the group, the APPC commissioned a Nielsen study of viewership patterns in and across

debates, a series of focus groups on debates conducted by Peter D. Hart, and a survey of public attitudes about

debates also conducted by Hart along with TargetPoint Consulting. Annenberg doctoral student Eunji Kim

synthesized the content of the debate memoranda. APPC senior researcher Bruce Hardy synthesized NAES data on

close following of debates and with Jamieson conducted a series of experiments on the eects of the immediate

audience’s cheers and laughter on the home audience’s perception of the candidates. Kim and Hardy analyzed fall-

o patterns in viewing within and across debates. Annenberg researchers Jennifer Isaacman and Deborah Stinnett

identiﬁed dierences in questions asked by those in town halls and journalists, and Jamieson synthesized the

scholarly literature on debates’ importance andeects.

Democratizing the Debates

Working Group Biographies

Robert Barnett

Robert B. Barnett is a senior partner at Williams & Connolly in Washington, D.C. He has worked on nine national

presidential campaigns, focusing mainly on debate preparation. In addition to playing a role on the Democratic

Party’s debate preparation teams in 1976, 1980, 1984, 1988, 1992, 2000, 2004, 2008, and 2012, he played the role

of George H.W. Bush in practice debates with Geraldine Ferraro in 1984, Michael Dukakis in 1988, and practice

debated Bill Clinton more than 20 times in 1992. In 2000, he played the role of Dick Cheney in practices with Joe

Lieberman and in 2004 with John Edwards. In 2000 and 2006, he assisted Hillary Rodham Clinton with her Senate

debate preparations and helped prepare her for 23 presidential primary debates in 2008. He assisted Barack Obama

in 2008 and 2012.

Robert Bauer

Bob Bauer represented President Obama’s re-election campaign in 2012 before the Commission on Presidential

Debates and in negotiation with the Romney campaign of the Memorandum of Understanding between the

Presidential nominees. As a specialist of politics law, Bauer has represented the Obama campaign and others on

legal requirements governing candidate debates, including the rules of the Federal Election Commission and the

Internal Revenue Service.

Joel Benenson

Joel Benenson, CEO of Benenson Strategy Group, has been President Obama’s pollster since his ﬁrst run for the

White House. Benenson is a former journalist whose career in polling has focused on the values and language of

working and middle class voters. He was a member of Obama’s debate prep team through all 23 primary debates in

2008 and all six general election debates President Obama participated in, as well as Vice President Biden’s team in

both elections. He has also been on prep teams for several Governor and U.S. Senators.

Charles Black

Charles R. Black is Chairman of Prime Policy Group. Black is best known as one of America’s leading Republican

political strategists. He served as senior advisor to both President Ronald Reagan and President George H.W.

Bush. In 1990, Black served as chief spokesman for the Republican National Committee and served as a principal

public spokesman for President Bush in the 1992 presidential campaign. He served on President George W. Bush’s

2000 and 2004 campaigns as a volunteer political advisor and surrogate spokesman. Most recently he served as

the senior political advisor to Senator John McCain’s 2008 presidential campaign. As part of his involvement in

nine presidential campaigns, Black has served as a debate negotiator and candidate preparer in four presidential

campaigns. He has provided advice to those playing those roles in four other presidential campaigns.

Rick Davis

Rick Davis has been involved in local, national and international political campaigns since 1976. His ﬁrst presidential

campaign experience was for then-former Governor Ronald Reagan in 1979. During President Reagan’s re-election

in 1984, Davis managed the delegate selection process during the primary and among other duties during the

general election he directed logistic support for the debate team during the presidential debates. In 1988 and

1992 as a part of the George Bush Presidential campaign Davis served in a number of roles including helping to

manage the support and logistics (spin room) for the presidential debates. In 1996, as Deputy Campaign Manager

for Senator Bob Dole’s Presidential campaign, Davis served as one of the debate negotiation team members

chaired at that time by Governor Carol Campbell. Davis’ duties as Deputy Campaign Manager included overseeing

all debate planning, candidate preparation and on-site management. In 2000 and 2008 Davis served as Senator

John McCain’s National Campaign Manager. In the later campaign he oversaw all debate negotiation, candidate

preparation, media and logistics.

Anita Dunn

As White House Communications Director and senior advisor to President Obama’s presidential campaign, Anita

Dunn directed conventional and new media communications strategies, as well as research, speechwriting,

television booking, presidential events, and cabinet aairs press. Dunn has worked for a wide range of elected

ocials and candidates across the United States, including working for the Democratic Senatorial Campaign

Committee as Communications Director from 1987-1990 under Senator John Kerry and then-Senator John Breaux;

leading Senator Bill Bradley’s political and communication teams from 1991-1993; serving as a consultant to the

Democratic Senate Caucus in 1995, and again in 1999, during the impeachment trial of President Clinton; and

working as then-Senate Majority Leader Senator Tom Daschle’s Communications Director in 2001. She has worked

with a wide range of Democratic Party ocials including Senator Evan Bayh, Democratic Leader Nancy Pelosi, and

Congressman John Dingell.

Ben Ginsberg

Benjamin Ginsberg, a partner at Jones Day, has represented the campaigns of Mitt Romney and George W. Bush

before the Commission on Presidential Debates and in negotiations with their general election opponents of

debate rules. He represents numerous political candidates and parties, members of Congress and state legislatures,

Governors, corporations, trade associations, vendors, donors and individuals participating in the political process. In

2012 and 2008, he served as national counsel to the Romney for President campaigns. In 2004 and 2000, Ginsberg

served as national counsel to the Bush-Cheney presidential campaigns; he played a central role in the 2000 Florida

recount. He advises on election law issues, particularly those involving federal and state campaign ﬁnance laws,

government investigations, ethics rules, Internal Revenue Service issues impacting the political process, redistricting,

communications law, and election recounts and contests.

Democratizing the Debates

Kathleen Hall Jamieson

Kathleen Hall Jamieson is the Elizabeth Ware Packard Professor of Communication at the Annenberg School

for Communication and Walter and Leonore Annenberg Director of the Annenberg Public Policy Center of

the University of Pennsylvania.Sheis a Fellow of the American Academy of Arts and Sciences, the American

Philosophical Society, the American Academy of Political and Social Science and the International Communication

Association.She is the author or co-author of 18 books including: Packaging the Presidency (1984), Presidential

Debates: The Challenge of Creating an Informed Electorate (1988); Spiral of Cynicism: The Press and the Public

Good (1997); and Presidents Creating the Presidency (University of Chicago Press, 2008). With Kate Kenski and

Bruce Hardy, Jamieson wrote The Obama Victory (Oxford, 2010), winner of an American Publishers Award for

Professional and Scholarly Excellence (PROSE Award) in government and politics and the ICA outstanding book

award. She was a member of the IOP-Twentieth Century Fund Taskforce on reform of debates convened by

Newton Minow in 1986 (see Appendix Two).

Ron Klain

Ron Klain has served as Assistant to the President and Chief of Sta to the Vice President. In this capacity, he directs

the sta in its various activities in support of the Vice President’s agenda. Klain also advises the Vice President on

a wide array of policy and political matters. He was appointed to this position in November 1995. He previously

served in the Clinton administration from January 1993 to February 1995. First, as Associate Counsel to the

President, he directed judicial selection eorts for the White House. In this capacity, he also led the conﬁrmation

teams for high-proﬁle nominees such as Justice Ruth Bader Ginsburg and Attorney General Janet Reno. Later, Klain

served as Counselor and Chief of Sta to Attorney General Janet Reno. As such, he advised Reno on numerous

legal and policy matters, and coordinated the administration-wide eort to craft and win passage of the President’s

Crime Bill, and the ban on semi-automatic assault weapons. Prior to joining the administration, Klain served in the

Clinton/Gore Campaign, as Washington Issues Director and as a Domestic Policy Specialist on the Campaign’s

Debate Preparation Team. After serving on the Debate Working Group from September 2013 to September 2014,

Klain resigned on October 23, 2014 upon being named Ebola response coordinator for the Obama Administration.

Zac Moatt

Zac Moatt was the Digital Director for Mitt Romney for President, where he managed a department of over 150

with a budget of over $100 million. The digital department was responsible for the campaign’s digital strategy:

online advertising, social media, email marketing and online fundraising. Before joining the Mitt Romney campaign,

Moatt and Michael Beach founded Targeted Victory, a full service interactive advertising agency. It has quickly

grown to serve over 170 federal and national clients including the Republican National Committee, Marco Rubio for

Senate and FedEx. Prior to founding Targeted Victory, Moatt served as the Deputy Director for Statewide eorts

at Freedom’s Watch, the Director of Political Education for the Republican National Committee and the Victory

Director for the Maryland Republican Party for Governor Robert Ehrlich and Senate candidate Michael Steele.

Beth Myers

Beth Myers has a long history of involvement in public issues and campaigns. Most recently, she was Senior Advisor

for Mitt Romney’s 2012 presidential campaign, and Campaign Manager for his 2008 presidential race. Before that

she served as Chief of Sta through all four years of the Romney governorship. She previously worked as a litigation

associate at Akin, Gump, Strauss, Hauer & Feld LLP in Dallas, Texas. Starting on the 1980 Reagan campaign, Myers

has worked for a slew of candidates. Working for Market Opinion Research, she developed and implemented GOTV

campaigns in California, Texas, Massachusetts, Louisiana and Missouri. In 2008 and 2012, she participated in debate

prep with Governor Mitt Romney for over 30 primary debates, and in 2012 she managed the debate preparations for

Governor Romney’s three general election debates.

Neil Newhouse

Neil Newhouse is a partner and co-founder of Public Opinion Strategies, a national political and public aairs

research ﬁrm, and was named Pollster of the Year by the trade publication Campaigns and Elections for its work in

the 2002 election cycle. Newhouse was chief pollster for the 2012 Romney for President Campaign and involved in

Governor Romney’s 2012 debate prep. He also worked on George W. Bush’s re-election campaign. Newhouse has

worked in public opinion research for more than 25 years, directed the research for thousands of individual projects

and has experience in every aspect of opinion research. During the 2008 election cycle, Newhouse was the

Republican partner for the NBC News/Wall Street Journal polls and he has twice been named as one of the Money

20 political consultants in the country who make a dierence. Newhouse has won praise from both sides of the

political aisle, having worked on numerous Gubernatorial, Senate and Congressional campaigns. He was described

by Pennsylvania Democratic Governor Ed Rendell as one of the most respected pollsters in the country and

recruited by Senator Joe Lieberman to provide polling and strategic guidance in his successful 2006 Independent

bid for U.S. Senate in Connecticut.

Jim Perry

Jim Perry served as an advisor to Governor Mitt Romney during preparations for his 2012 general election debates.

He has helped numerous candidates prepare for debates including former Mississippi Governor Haley Barbour, for

whom Perry also served as Deputy Chief of Sta and Policy Director. Perry currently is an investment banker at

Morgan Stanley.

Democratizing the Debates

Joe Rospars

Joe Rospars is the founding partner and creative director of Blue State Digital. For both the 2008 and 2012

campaigns, Rospars was President Barack Obama’s principal digital strategist and advisor, overseeing the digital

integration of the unprecedented fundraising, communications, and grassroots mobilization eort. The digital arm

of the campaign provided the backbone of design and branding both online and o and engaged a record-breaking

number of Americans through mobile, social, video, and web platforms. Prior to the Obama campaigns, Rospars

led Blue State Digital’s work with Governor Howard Dean, from the founding of Democracy for America to Dean’s

historic 50-state strategy and the 2006 election victories.

Michael Sheehan

Michael Sheehan has been a member of the Debate Prep team for every Democratic Presidential and Vice

Presidential Candidate since 1988. He has also prepared numerous Democrats for their Congressional state-wide

debates. For the Obama and Clinton administrations alike, he has coached Inaugural Addresses, States of the Union,

prime time addresses, and press conferences. One of America’s leading communications trainers and strategists,

his expertise embraces every format and every forum whether media interview, major speech or high-stakes Q&A.

His ability to help people communicate at the highest possible level was dubbed by New York Magazine as “the

Sheehan eect.”

Stuart Stevens

For 25 years, Stuart Stevens has been the lead strategist and media consultant for political campaigns such as those

of Senator Rob Portman, Senator Roy Blunt, Governor Haley Barbour, Governor Tom Ridge and President George

W. Bush in 2000 and 2004. Most recently, Stevens was lead strategist for Governor Mitt Romney’s 2012 primary and

general election presidential races. Beginning his political career in his native Mississippi, Stevens worked on Thad

Cochran’s campaigns.

APPENDIX TWO

SPONSORSHIP

1960 Debates:

In February 1959 in the Lars Daly case, the Federal Communications Commission (FCC) ruled that newscasts were

covered by the equal time provisions of Section 315 of the Communications Act of 1934. Congress responded by

amending Section 315 to exclude bona ﬁde newscasts, regularly scheduled interviews, programs, and on-the-spot

reporting of bona ﬁde news events. That initial piece of legislation did not include debates. On June 27, 1960,

Congress passed a joint resolution (S.J. 207) temporarily suspending the “equal time” provision for the 1960

presidential and vice presidential candidates. As CBS president Frank Stanton noted, “At midnight November seventh

we automatically reverted to the equal time provision of Section 315 of the Communications Act, as revised, which,

by requiring broadcasters to give equal time to all candidates of all parties including splinter groups and faddists, for

all practical purposes outlaws broadcasts of face-to-face meetings” (in Kraus, 1962, p. 66).

Shortly after being nominated, Richard Nixon and John F. Kennedy accepted the networks’ invitation to appear in

“The Great Debates.” Each saw an advantage in debating. For Kennedy, the debates were an opportunity to dispel

the notion that he was too young and inexperienced to be president; for Nixon, the debates seemed an opportunity

to showcase his skills both as a television performer and debater. NBC board chairman Robert W. Sarno

emphasized that the Kennedy-Nixon debates:

Do not represent a donation of free time to the candidates. They are rather an exercise of broadcast

journalism in which the candidates have agreed to appear within a framework calculated to stimulate a

digniﬁed, genuinely informative airing of their views, and to test those views against each other.

Details concerning the 1960 debates were “worked out in protracted negotiations—twelve meetings in all—

between representatives of CBS, NBC, ABC, and the Mutual Broadcasting System, and advisors to the candidates,”

note Minow and Sloan (1987, p 12). “The sponsorship of the individual debates was determined by lot…” (Minow and

Sloan, 13).

Democratizing the Debates

1976 Debates:

Behind in the polls, in his convention acceptance address incumbent Republican Gerald Ford challenged his

Democratic opponent, Jimmy Carter, to debate. The challenger accepted. As a result, for the ﬁrst time in history

an incumbent debated a challenger. A vice presidential debate was held as well. Because the FCC’s 1975 Aspen

decision held that debates not sponsored by a broadcaster were “bona ﬁde news events,” the 1976 presidential

debates were able to be sponsored by the League of Women Voters. Despite reservations, the League acceded to

candidate demands that there be no reaction shots of the audience. Reliance on a press panel to pose questions

was widely criticized by reporters and academics as was the fact that the League had permitted the candidates to

decide the rules under which the debates would occur.

1980 Debates:

The 1980 debates were also sponsored by the League, which proposed the same schedule as 1976: three

presidential and one vice presidential debate. With John Anderson running as an independent, the League set

criteria to determine eligibility to participate in debates that included passing a 15% threshold in public opinion polls

and sucient ballot eligibility to be elected. When the incumbent Democrat Jimmy Carter refused to participate

in a debate with Anderson, a League-sponsored debate between Anderson and Republican party nominee Ronald

Reagan was held on September 21. By contrast it was vice presidential nominee George H.W. Bush and not vice

president Mondale who refused to participate in a vice presidential debate with Anderson’s vice presidential

nominee Patrick Lucey. After Anderson fell below the League’s threshold for eligibility, a Carter-Reagan debate was

held on October 28, a week before the election. The ﬂedgling cable network CNN included Anderson in the debate

by giving him the opportunity to respond to the questions oered to Carter and Reagan. Again the League acceded

to the demands of the candidates. So for example, the panelists for the Carter-Reagan debate were selected by

joint agreement of the two campaigns.

1983: A Twentieth Century Fund Task Force concluded: “The public is not well served when debates are negotiated

in the heat of the fall campaign and when the candidates’ tactical advantages become more central to the

negotiations than the public interest” (In Minow and Sloan, 35).

1984 Debates:

Although the FCC had altered the Aspen rule to permit networks to sponsor debates, the League retained that

franchise in 1984, sponsoring two presidential and one vice presidential debates.

1985: On November 26, 1985, Frank Fahrenkopf and Paul Kirk, the chairs of the two major parties, issued a

“Memorandum of Agreement on Presidential Candidate Joint Appearances” which declared that “to better fulﬁll

our parties’ responsibilities for educating and informing the American public and to strengthen the role of political

parties in the electoral process, it is our conclusion that future joint appearances should be principally and jointly

sponsored and conducted by the Republican and Democratic National Committees.”

1985: In February 1985, the Center for Strategic and International Studies at Georgetown formed a 40 person

bipartisan Commission on National Elections co-chaired by Robert Strauss and Melvin Laird. Issued in April 1986,

the Commission report endorsed the Fahrenkopf and Kirk memorandum and recommended that the two major

political parties institutionalize debates by forming an organization responsible for running them and by securing

the commitment of their respective candidates to participate in them.

1986: In 1986, the Harvard Institute of Politics-Twentieth Century Fund formed a 30 person study group chaired by

Newton Minow.

In December 1986, Minow, who had played a key role in televised U.S. presidential debates throughout their history,

convened a conference to evaluate 10 of his recommendations to improve the functioning of presidential debates:

1. Quadrennial presidential debates should be institutionalized.

2. The Democratic and Republican parties should make ﬁrm commitments to future presidential debates as an

important contribution to the public interest.

3. The Democratic and Republican parties should establish a bipartisan Presidential Debates Organization now to

administer the 1988 debates.

4. The Presidential Debates Organization should have an Advisory Committee composed of a broad, diverse group

of public citizens.

5. The Presidential Debates Organization should set the time, number, and location of presidential debates for the

1988 campaign well in advance, preferably in 1987.

6. At least three presidential debates, and one vice presidential debate, should be scheduled. The debates should

begin immediately after Labor Day and should conclude by the third week of October.

7. The use of journalists as questioners should be eliminated in favor of allowing the candidates the opportunity of

questioning each other.

8. The question of third-party candidates should not undermine the goal of institutionalizing debates between

the Democratic and Republican party candidates. (That question can be considered, in all its complexity, in the

context of a guaranteed minimum of debates between the major party candidates.)

9. To insure third party access, other avenues, such as free television time for candidates, should be explored and

adopted.

10. As in 1960, Congress should suspend Section 315 of the Federal Communications Act for presidential and vice-

presidential candidates in the 1988 election.

Democratizing the Debates

1987: On February 18, 1987, the chairs of the two major political parties formally announced the existence of a 10

member Commission on Presidential Debates which they would co-chair. The Commission was established to

sponsor vice presidential and presidential debates.

1988 Debates: Both the League of Women Voters and the Commission on Presidential Debates announced

dates for fall 1988 debates. When the League rejected the terms of the Memorandum of Understanding that

the campaigns handed to both the Commission and the League, and the Commission agreed to its terms, the

Commission became the sponsor of the two presidential and one vice presidential debate of 1988. Of particular

concern to the League was the absence of the chance for follow-up questions.

APPENDIX THREE

YOUNGEST GROUP LEAST LIKELY TO WATCH MOST

OFDEBATE

2012 Debate Viewing—Length

By Age (Among Debate Viewers)

(N=892)

18-34

(A)

35-49

(B)

50-64

(C)

65+

(D)

Total

All/Most (collapsed) 75%

BCD

88%

85%

89%

84%

Whole debate 42%

41%

45%

59%

ABC

46%

Most of the debate 33%

47%

39%

30%

38%

Some/a little (collapsed) 25%

BCD

11%

15%

11%

16%

Just some of the debate 21%

BCD

12%

10%

13%

Just a little of the debate 4% 4%

3% 1%

A,B,C and D represent each age category, respectively. The letter placed beside a data point means that the referent age category is statistically dierent from the data point age

category. “A” refers to 18-34, “B” refers to 35-49, “C” refers to 50-64, and “D” refers to 65+. For example, if there is a letter “C” next to a number under the 18-34 category (A), then that

means that the dierence between the number in the C category and the A category is statistically signiﬁcant at the 95% conﬁdence level (p<.05).

When you watch the debate do you usually watch the whole debate from the beginning to end, most of the debate, just some of the debate, or just a little of the debate? Data from

survey conducted by Peter D. Hart and TargetPoint Consulting. Survey conducted May 15-May 21, 2014. Margin of Error +/-3.28%.

Democratizing the Debates

APPENDIX FOUR

19602012 HOUSEHOLD RATINGS TRENDS:

PRESIDENTIAL DEBATES

Prepared for: Annenberg Public Policy Center

Source: Nielsen Npower

Note: HH GRP%= The sum of HH ratings for all telecasts of the particular debate

9/1/12

9/3/12

9/5/12

9/7/12

9/9/12

9/11/12

9/13/12

9/15/12

9/17/12

9/19/12

9/21/12

9/23/12

9/25/12

9/27/12

9/29/12

10/1/12

10/3/12

10/5/12

10/7/12

10/9/12

10/11/12

10/13/12

10/15/12

10/17/12

10/19/12

10/21/12

10/23/12

10/25/12

10/27/12

10/29/12

10/31/12

11/2/12

11/4/12

11/6/12

Debate #1

Debate #2

Debate #3

Debate #4

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #1

Debate #2

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

1960 1976 1984 19881980 1992 1996 20042000 2008 2012

HH GRP%

APPENDIX FIVE

ELEMENTS IN MEMORANDA OF UNDERSTANDING

Focus of Debates

From 1988 to 2012, the debates concentrated on two broad issue areas: domestic policy/economic

policy and foreign policy/national security.

In 2012, the Committee announced speciﬁc topics under each issue (not mentioned in MOU) area prior

to the debate. For the ﬁrst presidential debate on domestic policy, topics included economy, health care,

the role of government, and governing. For the third presidential debate on foreign policy, they included

America’s role in the world, war in Afghanistan and Pakistan, Israel and Iran, Middle East and terrorism,

and rise of China.

Format of Debates

Opening/Closing:

General agreement

(1) No opening statement by candidate (Exception: 1996)

(2) A closing statement that does not exceed 2 minutes

(3) Order determined by a coin toss

Change over time

(1) The length of a closing statement: 2 minutes (1988-2004) vs 90 seconds (2008)

a. In 2012, whether closing statement for the third presidential debate would be 90 seconds or 2

minutes was resolved by a coin toss.

Props & Notes:

General agreement

(1) Neither candidate shall be permitted to carry in notes or any other materials

(2) Moderator must interrupt if a candidate uses a prop (2000, 2004)

(3) Neither candidate may reference or cite any speciﬁc individual sitting in a debate audience

(2008,2012)

Democratizing the Debates

Neither ﬁlm footage nor video footage in a debate may be used publicly (1988)

Exception

In 1988, the MOU did not categorically prohibit the use of excerpts. Instead it stipulated that excerpts not

be used out of context or in a false or deceptive manner.

Order of Questioning:

General agreement

(1) The same question is to be asked of each candidate and the order then reversed.

(2) The order of questioning is determined by a coin toss.

(3) The winner of a coin toss shall have the option to take the ﬁrst or second question. The order of

closing statement is determined by reversing the order.

Apportioning of Time:

A moderator is responsible for all time limits and shall interrupt when a candidate exceeds the permitted

time limit.

Speciﬁc time segments were agreed to in 2008 and 2012.

 — 2008: 9 nine-minute segments

(2 minutes for each candidate + 5 minutes for open discussion)

 — 2012: 6 ﬁfteen-minute segments

(2 minutes for each candidate )

Direct Address/Questioning:

In general direct candidate-to-candidate questioning has been banned. Exceptions are made for

rhetorical questions. In 2008, this agreement was in force only in the second presidential debate

(townhall).

Town Hall Audience Questions:

General agreement

(1) A moderator shall exercise full authority to select the questions from the audience and

expanddiscussion

Change over time

(1) Earlier debates allowed moderator to ask brief follow-up questions to clarify.

(2) Audience is asked to submit questions to the moderator prior to the town hall debate

(since 2000).

(3) Prior to the debate, campaigns will be told the method used to select the citizens in the town hall

(since 2000).

(4) Moderators are asked not to “coach” the questioners (since 2008).

Exception

In 2008, the moderator used questions submitted on the Internet, which constituted 1/3 of the

totalquestions.

Press Panelist Format (Only applicable to 1988 and 1992):

Besides having a moderator, the presidential debates in 1988 and 1992 had a press panelist format in

which selected panelists asked questions of candidates.

General agreement on panelist selection

Each side submits 6 to 10 panelists. When 2 or more possible panelists are agreed upon, these names will

be submitted to the Commission which will select one panelist from each list. For the third panelist, the

Commission submits a list of 10 possible panelists to each side and then picks one from mutually agreed

upon panelists. Each debate will have dierent panelists.

Panelists were drawn from following institutions.

1988 First Presidential Debate:

Atlanta Constitution, the Orlando Sentinel, ABC News

Second Presidential Debate:

ABC News, Newsweek magazine, NBC News

1992 First Presidential Debate:

The Boston Globe, ABC News, Freedom Forum

Third Presidential Debate:

Reuters, CNN, United Press International

The press panelist format was abandoned after the 1992 debate.

Moderators

Moderator selection rule:

In general each side submits 1-2 possible moderators. When each agrees upon at least 1 possible

moderator from the other’s list, the name will be submitted to the Commission which will then select

moderator for mutually accepted candidate (1988, 1992, 1996).

Candidates have agreed to follow the Commission’s recommendation (since 2000).

Democratizing the Debates

Number of moderators:

Dierent moderators for each debate (1988, 1992, 1996, 2004, 2008, 2012)

One moderator for three presidential debates (2000)

Moderator role:

General agreement

The moderator will open the program, monitor length of answers, identify each topic before the

questions are asked and ensure that questions are balanced.

Change over time

The 2008 and 2012 MOUs stipulate that the moderator shall not ask the candidates for a “show of hands”

or make similar calls for response.

Exception

(1) The moderator was expected to ask follow-up questions during the second presidential debate. (1992)

(2) The moderator shall ensure that candidates address at least 16 questions. (2004)

Timing and Debate Locations

The number of days between the debate and the election day (unit: days)

Presidential Debate 2

Presidential Debate 3

Presidential Debate Vice Presidential Debate

1988 44 26 N/A 34

1992 23 19 15 21

1996 30 20 N/A 27

2000 35 27 21 33

2004 33 25 20 28

2008 39 28 20 33

2012 34 21 15 26

Average 34 days 24 days 18 days 29 days

Debate Locations

Presidential Debates

Vice Presidential Debate

1988 Winston-Salem, NC Los Angeles, CA N/A Omaha, NE

1992 St. Louis, MO Richmond, VA East Lansing, MI Atlanta, GA

1996 Hartford, CT San Diego, CA N/A St. Petersburg, FL

2000 Boston, MA Winston-Salem, NC St. Louis, MO Danville, KY

2004 Coral Gables, FL St. Louis, MO Tempe, AZ Cleveland, OH

2008 Oxford, MS Nashville, TN Hempstead, NY St. Louis, MO

2012 Denver, CO Hempstead, NY Boca Raton, FL Danville, KY

Number of Debates

Except for 1960 when there were four presidential debates and 1988 and 1996 which had two presidential debates,

there have been three presidential debates and one vice presidential debate.

Audience and Interactivity

Audience question

Changes in managing audience questions

1992 No speciﬁc limitation on audience questions

1996 Moderator is permitted to ask brief follow-ups to clarify ambiguous questions from audience

Restrictions Added Since 2000

2000

(1) Follow-up questions from audience are prohibited

(2) Audience shall submit their questions to the moderator before the debate

2004

If any audience member poses a question or makes a statement that is dierent from the one that the

audience member earlier submitted to the moderator, the moderator will cut o the questioner

2008 The moderator shall not ask follow-up questions or comment on questions asked by audience

Democratizing the Debates

Audience instruction

Audience is asked not to applaud or participate by any means other than silent observation

except as provided by the agreed upon rules of the town hall debate

Audience selection in town hall debate

(1) A certain number of uncommitted voters would be selected by

An independent research ﬁrm (1992, 1996, 2000)

Gallup Organization (2004, 2008, 2012)

(2) The number of people in audience

Approximately 250 people (1992, 1996)

Approximately 100-150 people (2000, 2004, 2008)

APPENDIX SIX

AUDIENCE REACTION STUDIES

The only major published study of general election debate audience reactions showed that they are able to aect

viewers’ perceptions of a candidate. In four experiments testing the impact of audience reaction on viewers’ rating

of the candidates’ debate performance and traits, Fein, Goethals, and Kugler

found consistent eects. APPC

studies conducted for the Working Group conﬁrm these ﬁndings.

Overall Design of the Studies

Participants for the APPC study of the eects of audience reaction in debates were recruited from Amazon’s crowd

sourcing platform Mechanical Turk which is a web service run by Amazon.com designed to crowd-source Human

Intelligence Tasks (HIT). Participants were randomly assigned to condition (around 300 respondents per condition)

and either exposed to the actual debate clip with no editing, a clip with debate reactions edited out in a way that

looked and sounded natural, or a control/baseline condition with no debate material. To orient those in the control

condition to the post-test questions, after exposure all participants were shown a screen shot of both candidates

with their names superimposed on their pictures.

Although we randomized conditions, we note that samples collected from Mechanical Turk are not entirely

representative of the general U.S. population. So, for example, the demographic proﬁle of the sample (Table 1)

shows that the sample is younger, more male, less Black, more Asian, less Hispanic, more highly educated, and

more left leaning politically than the general U.S. population. This distribution is consistent across the samples of

the two studies reported here and other published studies evaluating Mechanical Turk.

Experiment 1

The ﬁrst experiment used the 1984 Reagan/Mondale debate clip noted earlier in which Reagan dismissed concerns

about his age.

Trewhitt: Mr. President, I want to raise an issue that I think has been lurking out there for 2 or 3 weeks

and cast it speciﬁcally in national security terms. You already are the oldest President in history. And

some of your sta say you were tired after your most recent encounter with Mr. Mondale. I recall yet that

President Kennedy had to go for days on end with very little sleep during the Cuban missile crisis. Is there

any doubt in your mind that you would be able to function in such circumstances?

Reagan: Not at all, Mr. Trewhitt, and I want you to know that also I will not make age an issue of this

campaign. I am not going to exploit, for political purposes, my opponent’s youth and inexperience.

Democratizing the Debates

Prolonged audience laughter and applause

Reagan: If I still have time, I might add, Mr.Trewhitt, I might add that it was Seneca or it was Cicero, I

don’t know which, that said, “If it was not for the elders correcting the mistakes of the young, there

would be nostate.”

Trewhitt: Mr. President, I’d like to head for the fence and try to catch that one before it goes over, but I’ll

go on to another question.

Participants randomized to the ﬁrst condition were shown the clip in its entirety including the audience laughter.

Those randomized to the second condition saw the same clip with the audience laughter edited out. (The video

faded out after Reagan said “youth and inexperience” and faded back in before he said “it was Seneca.”) Those

randomized into the third condition were not exposed to any of the debate clips.

After viewing the clips the respondents were asked to rate the favorability of the candidates on a one (very

unfavorable) to eleven (very favorable) scale. Figure 1 charts the results of audience reaction on the favorability of

the candidates. Exposure to the debate clip with the full audience reaction signiﬁcantly aected the favorability

5.0

5.5

6.0

6.5

7.0

7.5

8.0

Actual Debate Debate Minus Audience Reaction

Impact of Audience Reaction

Favorability of the Candidates on a 1 to 11 Point Scale

Baseline

Debate #1

Debate #2

Debate #3

Debate #4

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #1

Debate #2

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

Debate #1

Debate #2

Debate #3

Ronald Reagan Walter Mondale

Figure 1. The impact of debate audience reaction on the favorability of Ronald Reagan and Walter Mondale

3.000

3.625

4.250

4.875

5.500

6.125

6.750

7.375

8.000

Actual Debate Clip Bentsen’s Audience Reaction

Edited Out

Quayle’s Audience Reaction

Edited Out

Both Audience Reactions

Edited Out

Impact of Audience Reaction

Favorability of Vice Presidential Candidates on a 1 to 11 Point Scale

Baseline

Dan Quayle Lloyd Bentsen

Figure 2: The impact of debate audience reaction on the favorability of Dan Quayle and Lloyd Bentsen

of both candidates compared to baseline. As one would expect, the impact is much greater for Reagan than

forMondale.

Experiment 2

The stimulus for our second experiment is an exchange from the October 5, 1988 debate between Republican Vice

President Nominee Dan Quayle and Democrat Lloyd Bentsen:

Dan Quayle: I have far more experience than many others that sought the oce of vice president of this

country. I have as much experience in the Congress as Jack Kennedy did when he sought the presidency.

I will be prepared to deal with the people in the Bush administration, if that unfortunate event would

everoccur.

Judy Woodru: Senator Bentsen.

Democratizing the Debates

Bentsen: Senator, I served with Jack Kennedy,

I knew Jack Kennedy, Jack Kennedy was

a friend of mine. Senator, you are no Jack

Kennedy. (Prolonged audience shouts and

applause) What has to be done in a situation

like that is to call in the -

Woodru: (To audience) Please, please, once

again you are only taking time away from

your own candidate.

Quayle: That was really uncalled for, Senator.

(Audience shouts and applause)

Bentsen: You are the one that was making

the comparison, Senator - and I’m one who

knew him well. And frankly I think you are

so far apart in the objectives you choose

for your country that I did not think the

comparison was well-taken.

The design of this experiment is similar to the ﬁrst but

with ﬁve conditions instead of three: Condition 1 is

the actual clip with no edits, condition 2 edits out the

audience reaction to Bentsen, condition 3 edits out the

reaction elicited by Quayle, condition 4 edits out the

reactions to both candidates.

Audience reactions had a signiﬁcant impact on the

favorability and perceptions of the candidates (Figure

2). Those in the baseline condition rated Bentsen only

slightly more favorably than Quayle while the dierence

in ratings in the full clip condition is a little over 2 scale

points. When the reaction to Bentsen is edited out, his

favorability drops and Quayle’s increases. When the

reaction to Quayle’s is edited out in condition 3, his

ratings are lower than in condition 2 but higher than

condition 1 suggesting that this audience response

had a negative eect on reception of the Republican

vice presidential nominee. All dierences are

statisticallysigniﬁcant.

NOTES

1. Chaired by Anita Dunn and Beth Myers, the Annenberg Debate Reform Working

Group included: Robert Barnett, Bob Bauer, Joel Benenson, Charles Black, Rick Davis,

Benjamin Ginsberg, Kathleen Hall Jamieson, Ron Klain, Neil Newhouse, Zac Moatt, Jim

Perry, Joe Rospars, Michael Sheehan and Stuart Stevens (for bios and a description of the

meetings held by the Working Group, see Appendix One).

2. “General Election Debates: 1960-2012.” Presentation by Josephine Holz to the

Annenberg Debate Working Group – December 16. 2013 - slide 17

3. Nielsen, NPM, 7/3/14-7/30/14, Mon-Sat 8-11p/Sun 7-11p/A18-49 (000), A18-34 (000).

See http://tvbythenumbers.zap2it.com/2014/07/31/univision-is-the-1-network-for-the-

second-consecutive-july-sweep-among-both-adults-18-49-and-adults-18-34/289148/.

Retrieved August 4, 2014.

4. A 2014 American Press Institute study found that “Among smartphone owners, 78

percent report using their device to get news in the last week. Seventy-three percent

of tablet owners use their device to get news. And people with more devices tend to

enjoy following the news more….Fully 4 in 10 Americans say they got news in the last

week from social media, through platforms such as Twitter or Facebook.” http://www.

americanpressinstitute.org/publications/reports/survey-research/personal-news-cycle/.

Retrieved August 4, 2014. An October 2013 Pew survey reported that “Overall, about

half of adult Facebook users, 47%, “ever” get news there. That amounts to 30% of the

population.” http://www.journalism.org/2013/10/24/the-role-of-news-on-facebook/.

Retrieved August 4, 2014

5. Jones, J. M. (2014, January 8). “Record-High 42% of Americans Identify as

Independents.” http://www.gallup.com/poll/166763/record-high-americans-identify-

independents.aspx. Retrieved August 4, 2014.

6. Nielsen report commissioned by the Annenberg Public Policy Center, “General

Election Debates: 1996-2012 Dec 9, 2013, p.7.

7. Nielsen report commissioned by the Annenberg Public Policy Center, “General

Election Debates: 1996-2012 Dec 9, 2013, p.4.

8. Nielsen report commissioned by the Annenberg Public Policy Center, “General

Election Debates: 1996-2012 Dec 9, 2013, p.4.

9. For across time viewership data, see Appendix Three.

10. Edward Hinck, Enacting the Presidency: Political Argument, Presidential Debates, and

Presidential Character (Westport: Praeger, 1993).

11. David Zarefsky, “Spectator Politics and the Revival of Public Argument,”

Communication Monographs 59, no. 4 (1992): 412.

12. Diana Carlin, Tammy Vigil, Susan Buehler, and Kelly McDonald, The Third Agenda

in U.S. Presidential Debates: DebateWatch and Viewer Reactions, 1996-2004. Westport,

Conn: Praeger, 2009, 706. Kindle edition.

13. Nielsen report commissioned by the Annenberg Public Policy Center, “Watching

Presidential Debates Live VS. Time–Shifted” Addendum. 03.2014. p. 4

14. Two of the groups were conducted among whites (one younger group and one

older group); two with Hispanics (one younger voter session done in English with a

Spanish moderator and the other done with older voters in Spanish); the ﬁfth group was

conducted among a cross section of all age groups and mixed races.

15. Pew Research and Social Demographic Trends (2014 March 7) “Millennials in

Adulthood: Detached from Institutions, Networked with Friends.” Available at: http://www.

pewsocialtrends.org/2014/03/07/millennials-in-adulthood

16. Fein, S., Goethals, G. R., & Kugler, M. B. (2007). Social inﬂuence on political

judgments: The case of presidential debates. Political Psychology, 28(2), 165-192.

Democratizing the Debates