Tea Party Supporters, Establishment Republicans, and the ... · the Obama Administration’s mortgage relief efforts (Williamson, Skocpol, and Coggin 2011). Santelli argued that Obama

Tea Party Supporters, Establishment Republicans, and the Role of Social Networking

Media in the 2012 General Election Campaign

Peter L. Francia, Professor

Department of Political Science

East Carolina University

Greenville, NC 27858

[email protected]

Jonathan S. Morris, Associate Professor

Department of Political Science

East Carolina University

Greenville, NC 27858

[email protected]

Paper prepared for presentation at the State of the Parties Conference, University of Akron,

Akron, OH, November 7-8, 2013.

In the early months of 2013, two prominent Republicans, Karl Rove and Sarah Palin,

took aim at one another in the aftermath of the Republican Party’s disappointing 2012 election.

In February 2013, Rove announced plans that his Super PAC, American Crossroads, was

backing a group, the Conservative Victory Project – a group formed to oust ultra-conservative

Tea Party candidates in Republican primary elections (Rayfield 2013). A month later, at the 2013

CPAC convention, Palin defended the Tea Party and criticized Rove, telling the crowd, “If these

‘experts’ who keep losing elections and keep getting rehired and getting millions – if they feel

that strong about who gets to run in this party, then they should buck-up or stay in the truck”

(Killough 2013). Rove promptly replied that if he did run for office, he would “serve out” his

term and “would not leave office midterm” – a shot at Palin’s decision to resign two years before

the completion of her term as Alaska’s governor (Killough 2013). This war of words between

Rove and Palin was more than personal; it underscored what many political observers suggest

are real fractures – even a “civil war” – between Tea Party supporters and what some call

“Establishment Republicans” (see e.g., Martin, Rutenberg, and Peters 2013).

A divided Republican Party, however, runs counter to accounts from previous decades in

the scholarly literature. Indeed, the Democratic Party has been the source of much derision in the

past for their divisions. As the humorist and social commentator, Will Rogers, remarked in 1935,

“I am not a member of any organized party—I am a Democrat.” A generation later in 1976,

Morris Udall said, “When Democrats assemble a firing squad, they always gather in a circle”

(quoted in Mayer 1996, 1). Indeed, until fairly recently, the findings of political scientists largely

backed these assessments. Defection rates in presidential general elections (i.e., the percentage of

self-identified Democrats or Republicans who voted across party lines), for example, favored the

Republican Party in eighty percent of the elections over a six decade period from 1936 to 1996

(Mayer 1996; Campbell 2008, 83).

Others have noted that the rules of the Democratic Party’s presidential nomination

process, which award delegates proportionately instead of the winner-take-all format that

Republicans use in many states, can prolong the primary season, which increases the chance that

hard feelings from the losing side may carry-over into the general election (Lengle 1980, 1981;

Southwell 1986; Kamarck 1990; Kamarck and Goldstein 1994; Mayer 1996). Some have argued

further that the culture of the Democratic Party has been one that is “open, loud, and

confrontational” whereas the Republican Party fosters a culture that is more “closed, quiet, and

consensual” (Freeman 1986, 338). The Democratic Party is also more demographically diverse

than the Republican Party (Teixeira 2009) and, throughout the 1990s, contended with an

ideological battle between its liberal wing and the rise of more centrist Democrats who called

themselves “New Democrats” (Hale 1995). Even as recently as 2003, former Vermont governor

and presidential candidate, Howard Dean, remarked that his campaign to win the Democratic

nomination for president would differ from his other Democratic opponents in that he would

represent the “Democratic wing of the Democratic Party” – a reference to the ideological split

between liberals and centrists in the party. To summarize in the words of Lengle, Owen and

Sonner (1995, 371):

The pluralistic base, cross-cutting cleavages, and wide ideological range that have

been characteristic of the Democratic Party since its transformation from minority

to majority party status in the 1930s sow the seeds of conflict among the party

elite as well as between the elite and the mass membership of the rules, role,

priorities, and direction of the party. The minority status, homogeneous

membership, and philosophical coherence of the GOP inoculate it to some degree

from the effects of divisiveness.

Yet, in more recent years, stories of Democratic divisions have given way to accounts of

mounting divisions within the Republican Party. These popular accounts have been supported in

several academic studies that show the Republican Party has become much more factionalized in

recent years, especially with the rise of the Tea Party, which became very politically active in

2009 (Abramowitz 2011, 2013; Rapoport, Dost, Lovell, and Stone 2013). While Tea Party

activists provided energy and grassroots support that many credit for the historic Republican

landslide in the 2010 mid-term elections, many political observers also blamed the Tea Party for

helping to nominate such ideologically extreme candidates that it cost the Republican Party

likely victories in U.S. Senate contests in Nevada, Delaware, Colorado, Indiana, Missouri in the

2010 and 2012 elections. These losses have stirred debate within the Republican Party, and

ultimately motivated prominent figures, such as Karl Rove, to push back against the influence of

the Tea Party within the Republican Party.

Yet, to appreciate how divisive looming battles may be in the Republican Party,

particularly in comparison to those in the Democratic Party, this paper analyzes data from the

2012 presidential election to determine just how divided the Republican electorate really is.

Next, we attempt to speculate about whether these divisions are likely to persist and perhaps

even widen by examining where Tea Party Republicans look for their political information

compared to Establishment Republicans. Polarization increases when members of like-minded

groups repeatedly hear others voice the same opinions and sets of ideas (Levendusky 2009). This

“echo chamber” can have the effect of reinforcing one’s beliefs, and even pushing those beliefs

further to the extreme (Jamieson and Cappella 2008). If Tea Party Republicans search out more

conservative media outlets for their political information than establishment Republicans, it

seems plausible to speculate that Republican divisions might persist into future elections. Finally,

for a more in-depth look at the thinking of Tea Party Republicans and establishment

Republicans, we examine comments made on Twitter in the late stages of the 2012 election. We

find that there are indeed deep divisions within the Republican electorate – both in their political

attitudes and in where they search for political information; however, our overall analysis also

reveals that Tea Party Republicans and establishment Republicans largely came together in the

2012 election and were unified behind Mitt Romney and against Barack Obama.

The Tea Party

The growth of the conservative activist base over recent decades coalesced with the 2008

election of a progressive African-American President and the proposal and passage of the

Affordable Care Act – developments that prompted the eventual establishment of the Tea Party

(Abramowitz 2011). While the Tea Party movement has been compared to grassroots political

movements of the past and present (Occupy Wall Street), there are unique features that set it

apart. First, the movement was not initially sparked from the grassroots. Instead, its origins trace

to a February 9, 2009, televised diatribe by CNBC financial reporter Rick Santelli, who criticized

the Obama Administration’s mortgage relief efforts (Williamson, Skocpol, and Coggin 2011).

Santelli argued that Obama was “rewarding bad behavior” and invited all “capitalists” to a

“Chicago Tea Party.” Second, the Tea Party Nation, formed in April of 2009, was established as

a for-profit business. Third—and most significantly for purposes of this study—is that the Tea

Party use media and technology in an unprecedented fashion to organize and propagandize,

possibly exacerbating divisions within the Republican Party.

Indeed, the modern media itself has fractionalized, creating an environment in which

there are more sources of political information than ever before, but also one in which a

significant number of news consumers choose to expose themselves only to select points of

view, making it less likely that they will come across counter-arguments (Prior 2007, Sunstein

2009). Niche news has become commonplace in American politics, where individuals tend to

select news sources that fit with their pre-existing point of view (Stroud 2008, 2010, 2011). The

result is the creation of “echo-chambers” in which like-minded media consumers receive

information that reinforces and solidifies their political perceptions. This trend has been

particularly noticeable in the conservative media establishment with the Fox News Channel and

conservative talk radio (Jamieson and Cappella 2008). These communication outlets did not exist

as effective mobilization tools until recently, and the conservative movement has taken

advantage of them in order to exert their influence on the Republican establishment.

Social networking websites have also helped the conservative media establishment’s

effort to organize and exert influence through the Tea Party (Williamson, Skocpol, and Coggin

2011). Some recent computer science research has examined the relationships between Facebook

and Twitter networks and found a good deal of cohesion between conservative media groups in

terms of language and network structures (Livne, et al. 2011, Mascaro, et al. 2012). This research

has shown that, despite the public perception that the Tea Party is a loosely associated grassroots

movement, the cohesiveness of the Tea Party on social networking websites rivals that of

Republicans and Democrats. We are, however, not yet sure what role the Tea Party presence on

social networking websites plays in enhancing or dividing cohesiveness for the Republican

Party. Twitter has been a major factor in Tea Party organization, but we do not fully understand

trends regarding what Tea Party members are posting on Twitter.

What were Tea Party members saying about the political parties and the candidates in the

2012 presidential election campaign? Was the sentiment toward the various political entities

positive or negative, and does the sentiment reflect a divide within the Republican Party between

establishment Republicans and Tea Party members? In order to address these questions, we first

look at the extent of the differences in public opinion between establishment Republicans and

Tea Party members in the 2012 general election. We then look at the differences in media usage

patterns between establishment Republicans and Tea Party members. Finally, we assess the

sentiment of Twitter posts from users who mentioned Barack Obama, Mitt Romney, and/or

either of the political parties and included right-of-center identifiers to the posts (specifically

#tcot [Top Conservatives on Twitter], #gop, and/or #teaparty). Did this sentiment reflect a divide

or unification? The implications are discussed.

The Divided Republicans?

We begin our analysis by examining data from the American National Election Study

(ANES) 2012 Time Series Study. The ANES 2012 Time Series Study includes 5,914

respondents in the pre-election survey (and 5,510 respondents in its post-election survey) – the

most of any previous ANES Time Series Study. This large sample provides sufficient cases to

analyze sub-populations such as Tea Party Republicans, establishment Republicans, and to make

comparisons of intra-party divisions in the Republican Party with intra-party divisions in the

Democratic Party.

Respondents completed interviews – either face-to-face or through the Internet – during

the two months preceding the November election and then again during the two months

following the election (for more information about the ANES 2012 Time Series Study, see

www.electionstudies.org). Consistent with its previous surveys, the ANES 2012 Time Series

Study asks respondents to identify their partisan affiliation (“Generally speaking, do you usually

think of yourself as a Democrat, a Republican, an independent, or what?”) and partisan leanings

if an independent (“Do you think of yourself as closer to the Republican Party or to the

Democratic Party?”). The ANES 2012 Time Series Study also asks respondents about their

feelings toward the Tea Party (“Do you support, oppose, or neither support nor oppose the Tea

Party movement?”).

These questions allow for the identification of Tea Party Republicans. Those classified as

Tea Party Republicans were respondents who answered that they identified as Republicans and

reported that they were supporters of the Tea Party. We also include independents who leaned

toward the Republican Party and were supporters of the Tea Party in this category because

independent “leaners” exhibit political behavior and attitudes consistent with self-identified

partisans (Keith et al. 1992; Petrocik 2009). Those who answered that they identified as

Republicans, or as independents who leaned toward the Republican Party, but were neutral

toward or not supporters of the Tea Party fall into the category of establishment Republicans.

As noted earlier, the Democratic Party also has had its own internal divisions, which in

previous years have been measurably greater than in the Republican Party (Mayer 1996). These

historic divisions have long pitted conservative to moderate Democrats, typically from the South

and dubbed “Yellow Democrats” and later “New Democrats,” against those in the more liberal

and northern wing of the Democratic Party (Rae 1994, McGlennon 1998, Reiter 2007). Although

many southern conservatives gradually left the Democratic Party, creating a more ideologically

cohesive Democratic Party than in the past (see e.g., Abramowitz and Saunders 1998, 2006;

Levendusky 2009), recent descriptions of the Democratic Party as “ideologically diverse” remain

(Hopkins 2009). To capture this ideological diversity, self-identified Democrats and Democratic

leaners who answered that they were “extremely liberal” or “liberal” comprise the Liberal

Democrat category. Democrats who answered that they were more conservative than “extremely

liberal” or “liberal” (i.e., “slightly liberal,” “moderate,” or any of the “conservative” categories)

fall into the Moderate Democrat category. Those remaining respondents who had no partisan

attachments or leanings fall into the Pure Independent category.

Seven-Point Issue Position Scales

Table 1 shows the extent of partisan divisions on a series of political issues that required

respondents to identify their positions on a 7-point scale. On the question of whether the

government should provide fewer services in areas such as health and education or more services

in these areas even if it means increased spending, Tea Party Republicans hold significantly

more extreme positions than establishment Republicans. Some 68 percent of Tea Party

Republicans are on the far end of the seven-point scale, in which a “1” and “2” indicate support

for “many fewer government services,” compared to 33 percent of establishment Republicans.

This 35-point gap is considerable when compared against the two blocs in the Democratic Party.

Support for “many more government services” at the “6” and “7” categories of the scale was

highest among Liberal Democrats at 33 percent compared to 15 percent for Moderate Democrats,

reflecting a smaller 17-point difference. In addition, both blocs of Democrats have an ample

number of respondents who answer in the middle categories, with 58 percent of Liberal

Democrats and 64 percent of Moderate Democrats falling in the “5” or “4” category. This stands

in contrast to the very small minority of Tea Party Republicans who indicated a moderate

response.

[Insert Table 1 about here]

On a similar 7-point scale, respondents were asked to indicate their feelings toward

whether the “government in Washington should make every effort to improve the social and

economic position of blacks” or whether “the government should not make any special effort to

help blacks because they should help themselves.” Tea Party Republicans are again the most

extreme in their responses with 77 percent answering “6” or “7” on the side of the scale

indicating that blacks should help themselves compared to 60 percent of establishment

Republicans. Similar but somewhat smaller differences are present among Democrats, with a

heavy concentration of answers from both Liberal and Moderate Democrats in the middle

categories of the scale. These results are particularly interesting given that issues tied to civil

rights have historically been most divisive in the Democratic Party. While these results show

strong opposition to government assistance to blacks among both Tea Party Republicans and

establishment Republicans, Tea Party opposition is especially strong with a majority (51%)

placing themselves in the “7” category compared to 36 percent of establishment Republicans, 27

percent of Pure Independents, 16 percent of Moderate Democrats, and 7 percent of Liberal

Democrats.

When asked whether the government should “see to it that every person has a job and a

good standard of living,” 71 percent of Tea Party Republicans fall into the “6” or “7” category

that “each person should get ahead on their own” compared to 44 percent of establishment

Republicans. Democratic divisions are noticeably smaller. Twenty-three percent of Moderate

Democrats and 29 percent of Liberal Democrats place themselves in the “1” or “2” categories,

and with healthy-sized majorities in both groups falling in the moderate categories of “3,” “4,” or

“5.”

Likewise, on the need for the federal government “to regulate business to protect the

environment,” 42 percent of Tea Party Republicans place themselves in the “6” or “7” category

of “no regulation because it will not work and will cost jobs.” This compares with 14 percent of

establishment Republicans, 9 percent of Pure Independents, 5 percent of Moderate Democrats,

and 2 percent of Liberal Democrats. When examining the other side of the scale where “1”

indicates support to “regulate business to protect the environment,” strong divisions are also

present among Democrats: 48 percent of Liberal Democrats fall into the “1” category compared

to 26 percent of Moderate Democrats, representing a sizable 22-point gap.

On perhaps the most controversial subject of all in recent years – health insurance – large

divisions are again present within the Republican Party. Twice the percentage of Tea Party

Republicans (40%) to establishment Republicans (20%) support for the most extreme position of

that “all medical expenses should be paid by individuals through private insurance plans like

Blue Cross or other company paid plans.” On the Democratic side, divisions were comparable

with 36 percent of Liberal Democrats, but only 16 percent of Moderate Democrats, favoring the

position that a government insurance plan should be provided to “cover all medical and hospital

expenses for everyone.” These patterns are similar for the “2010 healthcare law” or what is more

popularly known as “Obamacare.” The most intense opposition to Obamacare overwhelmingly

comes from Tea Party Republicans: 70 percent express a “great deal” of opposition to the law

compared to 35 percent of establishment Republicans. Democrats are likewise divided: 53

percent of Liberal Democrats favor the law a “great deal” compared to 25 percent of Moderate

Democrats. Taken together, health insurance, much like environmental protection, is an issue

with significant chasms within both parties.

It is only on the 7-point scale for defense spending that divisions within the Democratic

Party stand out more significantly than they do in the Republican Party. Thirty-nine percent of

Liberal Democrats fall into the 1 or 2 category of favoring great decreases in defense spending

compared to 18 percent of Moderate Democrats. On the other side of the scale, where 6 and 7

represent support for great increases in defense spending, 34 percent of Tea Party Republicans

support this position compared to 23 percent of establishment Republicans. While Democrats are

more divided on this issue than Republicans, the results taken together, show Republicans to be

equally, and on some issues, more divided than Democrats.

Job Performance of President Obama

The ANES 2012 Time Series Study also provides a series of questions that measure

respondents’ evaluations of President Obama’s job performance. The results in Table 2 show that

Tea Party Republicans are consistently the most negative. Roughly 92 percent of Tea Party

Republicans “disapprove strongly” of President Obama’s handling of his job as President. In

comparison, 64 percent of establishment Republicans fall into this category – a rather high

percentage, but still a considerably lower percentage than that reported for Tea Party

Republicans. Democrats are also divided somewhat on President Obama’s job performance,

although to a lesser extent than among Republicans. Seventy percent of Liberal Democrats report

strong approval of President Obama’s handling of his job as President compared to 53 percent of

Moderate Democrats.


Overwhelming majorities of Tea Party Republicans also disapprove strongly of President

Obama’s handling of the economy (93%), healthcare (90%), foreign relations (87%), and the war

in Afghanistan (63%). Establishment Republicans are also quite critical of President Obama, but

significantly smaller percentages fall in the “disapprove strongly” categories on his handling of

the economy (72%), health care (70%), foreign relations (55%), and the war in Afghanistan

(40%). Liberal Democrats and Moderate Democrats are divided by similar margins, with the

exception of President Obama’s handling of the war in Afghanistan – an issue in which similar

percentages of Moderate Democrats and Liberal Democrats express approval. In total,

assessments of President Obama’s job performance reveal roughly similar gaps in intensity of

approval and disapproval among Republicans as among Democrats.

Feeling Thermometers

Beyond the issues, political divisions often extend into intense feelings toward political

figures, political and societal institutions, and even groups of people. Table 3 shows the average

“feeling thermometer” responses (a 0 to 100 scale in which 0-49 equals a “cool rating”; 50

equals a “neutral rating”; and 51-100 equals a “warm rating”) for Tea Party Republicans,

establishment Republicans, Pure Independents, Moderate Democrats, and Liberal Democrats.

The results reveal that Tea Party Republicans harbor the most negative or coolest feelings toward

President Obama, with an average feeling thermometer score of just 16. By comparison,

establishment Republicans are mostly cool toward President Obama (average score of 35), but

considerably less so than Tea Party Republicans. Predictably, Pure Independents, Moderate

Democrats, and Liberal Democrats have more positive feelings toward President Obama, but the

gap separating Moderate Democrats (average score 78) and Liberal Democrats (average score of

87) of 9 points was noticeably smaller than the gap of 19 points separating Tea Party

Republicans and establishment Republicans. Similar patterns emerge for other prominent figures

in the Democratic Party, including Vice President Joe Biden, First Lady Michelle Obama, and

former First Lady and U.S. Senator Hillary Clinton. Indeed, Tea Party Republicans’ negative

feelings toward all of these prominent figures in the Democratic Party are considerable when

considered in comparison to establishment Republicans. Not a single Democratic figure averages

higher than 29 (Hillary Clinton), whereas among establishment Republicans, not a single

Democratic figure scores lower than a 34 (Joe Biden).


When examining prominent Republican figures, we see more unity between Tea Party

and establishment Republicans for the Republican presidential nominee Mitt Romney. Tea Party

Republicans are quite warm in their evaluations of Romney (average score of 77), but

establishment Republicans are not far behind (average score of 67). Among Democrats, the gap

is larger: Moderate Democrats are somewhat cool toward Romney (average score of 35), but

Liberal Democrats are very cool toward Romney (average score of 18). Interestingly, although

not surprising given his strong reputation as a small-government conservative, Paul Ryan, was

more popular than Romney with Tea Party Republicans (average score of 79), but not so with

establishment Republicans (average score of 64). This 15-point gap mirrors the divisions in the

Democratic Party where Moderate Democrats felt relatively cool toward Ryan (average score of

35) and Liberal Democrats were even cooler (average score of 20). Likewise, an identical 11-

point gap separates Tea Party Republicans from establishment Republicans, and Moderate

Democrats from Liberal Democrats on feelings toward Romney’s wife, Ann Romney. It is only

on feelings toward former President George W. Bush in which Republicans are slightly more

unified than Democrats. The gap that separates Tea Party Republicans from establishment

Republicans is 8 points, whereas Moderate Democrats are, on average, 15 points warmer toward

George W. Bush than Liberal Democrats.

The results for former President Bush combined with those for Romney suggest that Tea

Party Republicans and establishment Republicans generally rally around their party’s standard-

bearer with similar like and intensity. The more significant internal party differences come in the

intensity of dislike for the opposing party’s key figures. In other words, both Tea Party

Republicans and Establishment Republicans generally liked Romney and Bush with similar

intensity. It is the much stronger intensity of dislike for Democratic leaders, notably President

Obama, among Tea Party Republicans than among establishment Republicans that divide the

Republican Party. Likewise, for Democrats, dislike for Republican leaders, such as Romney and

Bush, is a more significant source of division among Moderate Democrats and Liberal

Democrats than each side’s warmness toward Democratic figures. We see this same pattern

emerge again on the feeling thermometer questions for the Democratic Party, Republican Party,

liberals, and conservatives.

Outside of political leaders and groups, most of the remaining the feeling thermometer

items all show only modest divisions within the factions of each party. Differences of 10 points

or more are present among Tea Party Republicans and establishment Republicans on feelings

toward labor unions (Tea Party Republicans are 15 points more negative than establishment

Republicans), and perhaps not surprisingly, the Tea Party itself (Tea Party Republicans are 25

points more positive than establishment Republicans). Among Democrats, differences of 10 or

more points exist on feminists (Liberal Democrats are 17 points more favorable than Moderate

Democrats), gay men and lesbians (Liberal Democrats are 17 points more favorable than

Moderate Democrats), illegal immigrants (Liberal Democrats are 12 points more favorable than

Moderate Democrats), and atheists (Liberal Democrats are 14 points more favorable than

Moderate Democrats). These latter divisions hint that many of the Democrats’ old fault lines

have not completely disappeared. The party’s liberal wing continues to embrace more strongly

the groups that champion equal rights and oppose conservative traditional values – a source of

division with a lengthy history in the Democratic Party. Nonetheless, while Democratic divisions

are hardly new, the large divisions within the Republican Party between the Tea Party wing and

the establishment Republican wing represent a clear shift from previous accounts of a unified

Republican Party.

Party Divisions and the Echo Chamber

As the results above demonstrated, divisions within the Republican Party now rival or

exceed those in the Democratic Party. But, are these divisions fleeting, or are they likely to

persist? A possible answer to that question lies in the diversity of ideas and positions that

individuals expose themselves to when seeking out political information. As noted earlier, living

in the so-called “echo chamber” can reinforce and push individuals to adopt even more extreme

positions over time.

With this in mind, we compare the sources of political information that individuals in

each party faction rely upon and use. Table 4 begins by examining viewership of several Fox

News programs, all of which have content that would reinforce strongly conservative viewpoints

and positions. The results indicate that Tea Party Republicans, by very large margins, expose

themselves more regularly to Fox News programs such as the Fox Report, The O’Reilly Factor,

Hannity, On the Record with Greta Van Susteren, and Huckabee than do establishment

Republicans. In comparison, establishment Republicans are more likely than Tea Party

Republicans to watch “hard news” programs more regularly, including 20/20, the ABC World

News, Nightline, 60 Minutes, the CBS Evening News, and the NBC Nightly News. Outside of Fox

News, only Face the Nation and Meet the Press were watched slightly more regularly by Tea

Party Republicans than by establishment Republicans. Among the Democrats, there are no such

differences. Viewing habits by Moderate Democrats and Liberal Democrats are generally

similar.


Differences in radio listenership are especially pronounced among Republicans. Table 5

reveals that a sizable percentage of Tea Party Republicans who listen regularly to the radio

choose the Rush Limbaugh Show (48%), the Sean Hannity Show (37%), and the Glen Beck Show

(26%) – all hosts known for their extremely conservative views. Establishment Republicans are

much less likely to listen regularly to these radio programs with only a small minority reporting

that they do so for the Rush Limbaugh Show (14%), the Sean Hannity Show (11%), and the Glen

Beck Show (9%). Other conservative radio programs also draw considerably more listeners from

Tea Party Republicans than from establishment Republicans, including the Mark Levin Show

(16% of Tea Party Republicans versus 2% of establishment Republicans), Savage Nation (11%

of Tea Party Republicans versus 4% of establishment Republicans), the Laura Ingram Show

(11% of Tea Party Republicans versus 3% of establishment Republicans), and the Neal Boortz

Show (10% of Tea Party Republicans versus 3% of establishment Republicans). In contrast,

establishment Republicans are more likely to listen to programs on National Public Radio,

including All Things Considered, Morning Edition, Fresh Air, and Talk of the Nation. Among

Democrats, listenership among Moderate Democrats and Liberal Democrats is almost

nonexistent for each of the conservative talk shows; however, Liberal Democrats are

significantly more likely than Moderate Democrats to listen to National Public Radio – a venue

that reports mostly objective “hard news” although with some critics suggesting a liberal bias.


Readership of print media also shows some divisions (see Table 6), although the most

telling differences are across parties rather within the parties. Notably, 14 percent of Liberal

Democrats report that they read the print edition of the New York Times regularly compared to

just 1 percent of Tea Party Republicans. Tea Party Republicans instead prefer to read their news

from the Fox News website (38%), the conservative Drudge Report (22%), or other web-based

services such as Yahoo News (30%) whereas establishment Republicans visit the Fox News

website (21%) and the Drudge Report website (5%) far less regularly (see Table 7). In fact

among establishment Republicans, the top websites are the largely neutral-content sites of Yahoo

News (36%), Google News (25%), and even CNN (22%). Differences are smaller among

Democrats, although Liberal Democrats are more likely to visit the New York Times website

(22%) than Moderate Democrats (11%).

[Insert Tables 6 and 7 about here]

The GOP and Tea Party on Twitter in the 2012 General Election

The previous discussion demonstrates a clear divide between establishment Republicans

and Tea Party members. In addition to differences on political issues and evaluations of political

figures, media usage also varies, which may contribute to further polarization. Beyond

television, radio, and other traditional sources of political information, minority factions of the

GOP, particularly the Tea Party, now have new outlets available to them that provide a platform

to be heard. This platform is not only used to criticize the liberal and Democratic opposition, but

also to raise objections at what they consider flaws within the Republican Party.

In order to examine whether or not this trend exists, we accessed historical Twitter data

from Gnip.com, an independent company that collects and archives tweets. We asked Gnip to

collect all Tweets from October 1 through October 24, 2012 that contained the most common

conservative and Republican-associated hashtags. Hashtags are keywords or topics in a tweet

created by users as a method of organizing Twitter messages into categories. These terms are

marked by the # symbol placed in front of them. The most common Republican and

conservative-associated hashtags are #gop, #tcot (Top Conservatives on Twitter), and #teaparty

(Livne et al. 2011). We search for all tweets conducted in the English language that contained

these hashtags from October 1st through the 24

th. This search produced 4.5 million Tweets, from

which we randomly sampled over 490,000. Once we removed empty Tweets and re-Tweets, we

finished with a sample of 253,507 Tweets. In total, 226,003 Tweets contained #tcot, 19,916

contained #gop, and 30,693 Tweets contained #teaparty. President Obama was mentioned in

104,320 Tweets, and Mitt Romney was mentioned in 76,630 Tweets. Republicans were

mentioned in 5,123 Tweets and Democrats were mentioned in 7,965 Tweets.

Our interest is in the sentiment of the Tweets. That is, what is the overall tone toward

different political entities mentioned in the Tweets? Based on the nature of the right-of center

political hashtags contained in the sample, we would expect the sentiment to be more positive

toward Republican entities and more negative toward Democratic entities. We would also expect

#teaparty Tweets to be more negative in tone toward Republican entities than #gop Tweets.

In order to analyze the sentiment of each tweet, we rely on automated content analysis of

our sample using a sentiment dictionary file. Specifically, we rely on the Lexicoder Sentiment

Dictionary (LSD) developed by Young and Soroka for analysis of political text (Young and

Soroka 2012). The purpose of LSD is to, “Develop a sentiment dictionary that is more broadly

applicable across a wide range of research foci in political communication” (2012, 206). Pulling

from several existing lexicon dictionaries, the LSD is comprised of 4,567 positive and negative

words. The content analysis software, Lexicoder 2.0, applies the LSD to each individual Tweet

and provides a count of the number of positive and negative statements per Tweet. Lexicoder 2.0

and LSD are freely available at www.Lexicoder.com. (See Young and Soroka 2012 for an

illustration of the reliability and validity of the LSD in comparison to other sentiment analysis

dictionaries and human coding.)

We applied the LSD content analysis to all Tweets in the sample that mentioned the

candidate in question, but not the other candidate. Tweets that mentioned both candidates were

not included in the content analysis. Thus, we could be confident that the positive or negative

statements are properly assigned to the proper candidate. In total 72,610 Tweets mentioned

Barack Obama but not Mitt Romney, and 44,920 Tweets mentioned Mitt Romney but not Barack

Obama. For Barack Obama, the average number of positive references per Tweet was .45, and

the average number of negative references per Tweet was .80, yielding a positive/negative ratio

of .57. For Mitt Romney, the average number of positive references per Tweet was .57, and the

average number of positive references per Tweet was .49, yielding a positive/negative ratio of

1.16.

Figure 1 shows the positive/negative sentiment ratio for Mitt Romney and Barack Obama

across each day in our sample (October 1-24). Some trends are evident in Figure 1. The first

trend is that negativity dominates when Barack Obama is mentioned. The ratio of positive to

negative Tweets is significantly lower than one, indicating more negativity than positivity, which

is not too surprising. Second, the positive to negative sentiment ratio for Mitt Romney is

consistently higher than that of President Obama, which makes sense given the hashtags we used

to identify our sample of Tweets. Third, it is clear that President Obama’s ratio stays fairly

consistent, reflecting that he was a known quantity. A marginal uptick in the sentiment ratio

occurs on October 5 and 6, when the government reported unemployment had fallen below 8

percent for the first time in his presidency, but other than that the ratio stays low. Lastly, it can be

seen that Mitt Romney’s sentiment ratio did not vary randomly. The swings are clearly

associated with events, particularly presidential debates. This highlights the validity of the

measure and confirms our expectation that sentiment toward Mitt Romney would be more

positive than that of President Obama.

[Insert Figure 1 about here]

Our larger question, however, was whether or not sentiment toward Republican entities

such as Mitt Romney and the Republican Party as a whole varied between Tea Party posters and

GOP posters. Again, because we cannot identify Tea Party affiliation based on Twitter handles,

we must rely on the hashtags #teaparty and #gop as the differentiating factor. The findings in

Table 8 show the difference in sentiment toward Barack Obama, Mitt Romney, and the two

major political parties from #teaparty and #gop. Two major findings come from this analysis of

Twitter sentiment. First, there is very little difference between #teaparty and #gop in sentiment

toward the candidates for president. The second finding reflects a noticeable difference when it

comes to sentiment toward the political parties. The amount of negative sentiment coming from

the #teaparty and #gop are similar, but #teaparty is much less positive toward Democrats and

Republicans than #gop. This is highlighted in the positive/negative ratio in Tweets that mention

Republicans from #gop (.71) versus the Tea Party (.56).


Conclusion

The results in this paper provide some empirical confirmation that there are rather large

and significant divisions within the Republican Party. The heated rhetoric of Karl Rove and

Sarah Palin runs much deeper than a personal disagreement among the two, and instead

illustrates much deeper divisions that exist within the rank-and-file of the Republican Party. On

several issues and evaluations of political figures, these divisions exceed or were on par with

those within the Democratic Party – a reversal from earlier accounts that portrayed Republicans

as the more unified party. Perhaps as significantly, media use was also shown to vary between

Tea Party Republicans and establishment Republicans, with Tea Party Republicans turning to

conservative sources with much greater frequency – a prospect that may continue to divide or

even increase divisions within the Republican Party.

Despite these many differences, however, the results also show relatively strong unity

behind Mitt Romney among both Tea Party Republicans and establishment Republicans. In fact,

during the final weeks of the 2012 election, Tea Party Republicans and establishment

Republicans were tweeting in much the same way about their support for Romney and their

dislike for President Obama. All of this suggests that while the Republican Party’s two major

factions are divided in several significant respects that now rival or exceed divisions within the

Democratic Party, the Republican Party remains unified at election time in November. This

could certainly change with increased and more divisive primary elections. However, at least for

now, the Republican Party remains united in its shared goal of defeating Democrats.

TABLE 1

Partisan Divisions on Selected Issues in the 2012 Election

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Government Services Scale

1= Provide many fewer services 31% 9% 8% 3% 1%

2 37% 24% 11% 5% 2%

3 18% 31% 19% 14% 6%

4 9% 22% 35% 36% 26%

5 3% 10% 14% 28% 32%

6 1% 2% 6% 8% 22%

7= Provide many more services 1% 2% 8% 7% 11%

N 822 1305 655 1562 651

Govt. Assistance to Blacks Scale

1= Govt. should help blacks 0% 1% 4% 6% 10%

2 1% 2% 4% 6% 14%

3 2% 4% 7% 12% 16%

4 12% 19% 27% 30% 34%

5 10% 15% 14% 15% 12%

6 26% 24% 19% 14% 9%

7= Blacks should help themselves 51% 36% 27% 16% 7%

N 805 1288 684 1499 624

Guaranteed Job/Standard of Living

Scale

1= Govt. provide jobs/std. living 2% 5% 11% 11% 13%

2 2% 6% 10% 12% 16%

3 3% 9% 12% 17% 22%

4 9% 19% 30% 30% 29%

5 14% 18% 17% 16% 11%

6 32% 27% 11% 10% 6%

7= Each person get ahead on own 39% 17% 10% 4% 3%

N 837 1,306 757 1,645 670

Environment/Jobs Tradeoff Scale

1= Regulate/protect environment 3% 8% 22% 26% 48%

2 5% 13% 20% 26% 27%

3 8% 19% 16% 21% 13%

4 17% 29% 25% 16% 8%

5 24% 17% 8% 6% 3%

6 24% 10% 5% 3% 1%

7= No regulation/will cost jobs 18% 4% 4% 2% 1%

N 798 1,191 650 1,460 644

* Table 1 continued on the next page.

TABLE 1 (continued)

Partisan Divisions on Selected Issues in the 2012 Election

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Medical Insurance Scale

1= Government insurance plan 3% 4% 14% 16% 36%

2 1% 4% 10% 17% 19%

3 3% 9% 12% 16% 18%

4 9% 15% 27% 24% 17%

5 17% 20% 14% 13% 5%

6 28% 27% 11% 9% 4%

7= Private insurance plan 40% 20% 13% 5% 2%

N 836 1308 735 1611 672

Support 2010 Healthcare Law

1= Favor a great deal 1% 3% 7% 25% 53%

2 2% 10% 13% 27% 26%

3 1% 3% 4% 6% 3%

4 7% 24% 40% 29% 13%

5 2% 4% 4% 2% 1%

6 16% 21% 12% 5% 2%

7=Oppose a great deal 70% 35% 19% 7% 2%

N 849 1,396 815 1,767 697

Defense Spending Scale

1= Greatly decrease def. spending 1% 1% 8% 6% 17%

2 3% 5% 9% 12% 22%

3 6% 12% 17% 20% 23%

4 26% 36% 36% 35% 24%

5 29% 23% 16% 18% 10%

6 22% 15% 9% 6% 2%

7= Greatly increase def. spending 12% 8% 6% 4% 3%

N 829 1260 654 1558 641

Source: American National Election Study, 2012

Notes: Some columns may not add to 100 percent because of rounding.

TABLE 2

Partisan Divisions on President Obama’s Job Performance in 2012

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Obama: Handling of Job

1= Approve Strongly 2% 5% 19% 53% 70%

2 1% 10% 26% 30% 26%

3 5% 21% 20% 10% 3%

4=Disapprove Strongly 92% 64% 35% 8% 2%

N 850 1,371 779 1,743 696

Obama: Handling Economy


2 1% 9% 20% 30% 33%

3 5% 16% 18% 13% 3%


N 846 1,383 774 1,739 691

Obama: Handling Healthcare


2 2% 7% 22% 23% 17%

3 6% 16% 15% 8% 3%


N 839 1,369 790 1,741 697

Obama: Handling For. Relations


2 4% 19% 28% 26% 20%

3 8% 19% 18% 9% 4%


N 838 1,348 760 1,733 685

Obama: Handling Afghan. War


2 12% 24% 27% 27% 32%

3 21% 23% 17% 10% 9%


N 821 1,347 759 1,730 688



TABLE 3

Partisan Divisions on Political Leaders, Groups, and Institutions in the 2012 Election

Feeling Thermometer Items

100 = Favorable; 0 =Unfavorable

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Political Leaders

Barack Obama 16 35 53 78 87

Joe Biden 18 34 49 69 80

Michelle Obama 28 45 57 80 89

Hillary Clinton 29 47 55 73 83

Mitt Romney 77 67 45 32 18

Paul Ryan 79 64 45 35 20

Ann Romney 77 66 49 43 32

George W. Bush 72 64 44 31 16

Political Groups

Democratic Party 20 34 45 70 77

Republican Party 69 65 41 30 16

Liberals 23 37 45 55 76

Conservatives 77 64 50 47 30

Political and Social Institutions

Federal government 27 36 39 49 52

Congress 34 41 38 45 41

U.S. Supreme Court 51 56 52 61 57

Military 88 84 75 77 70

Big business 61 53 44 45 37

Labor unions 26 41 50 57 66

Racial, Ethnic, and Other Groups

Asian-Americans 66 64 61 64 71

Blacks 60 61 60 68 72

Hispanics 61 60 59 65 71

Whites 74 74 68 71 73

Feminists 35 46 47 55 72

Gay men and lesbians 39 46 48 55 72

Illegal immigrants 25 31 37 42 54

Tea Party 75 50 42 33 17

Religious Groups

Atheists 30 35 39 40 54

Catholics 68 63 54 59 52

Christians 82 76 66 70 60

Christian fundamentalists 65 54 45 45 28

Mormons 61 54 47 47 43

Muslims 34 40 43 48 54


TABLE 3 (continued)

Partisan Divisions on Feelings Toward Political Leaders, Groups, and Institutions in the 2012 Election

Feeling Thermometer Items

100 = Favorable; 0 =Unfavorable

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Economic Groups/Class

Middle class people 76 75 71 77 80

Working class people 84 82 80 83 82

Poor people 66 65 66 70 63

People on welfare 40 43 47 52 59

Rich people 63 54 45 45 43


TABLE 4

Partisan Divisions on Viewership of Fox News Programs and Other Television

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

FOX News: Fox Report

Does not watch regularly 53% 75% 91% 89% 94%

Watches regularly 47% 25% 9% 11% 6%

N 777 1266 661 1617 637

FOX News: The O’Reilly Factor



N 777 1266 661 1618 638

FOX News: Hannity



N 777 1266 661 1617 637

FOX News: Greta Van Susteren



N 777 1266 661 1618 637

FOX News: Huckabee



N 777 1266 661 1618 637

ABC: 20/20



N 777 1267 662 1617 637

ABC: World News



N 777 1266 661 1617 637

ABC: Nightline



N 777 1267 662 1618 637


TABLE 4 (continued)

Partisan Divisions on Viewership of Fox News Programs and Other Television

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

CBS: 60 Minutes



N 777 1266 661 1617 637

CBS: Evening News



N 777 1267 661 1617 637

CBS: Face the Nation



N 777 1266 661 1618 637

NBC: Nightly News



N 777 1267 661 1618 637

NBC: Meet the Press



N 777 1266 661 1617 637



TABLE 5

Partisan Divisions on Listenership of Conservative Talk Radio and Other Programs

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

The Rush Limbaugh Show

Does not listen to regularly 52% 86% 92% 96% 98%

Listens to regularly 48% 14% 8% 4% 2%

N 577 706 298 809 381

The Sean Hannity Show



N 577 706 298 809 382

Glenn Beck



N 577 705 297 808 381

The Mark Levin Show



N 577 706 298 809 382

Savage Nation (Michael Savage)



N 577 706 298 809 382

The Laura Ingraham Show



N 576 706 298 809 381

The Neal Boortz Show



N 577 706 298 809 381

All Things Considered (NPR)



N 577 705 298 808 381


TABLE 5 (continued)

Partisan Divisions on Listenership of Conservative Talk Radio and Other Programs

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Morning Edition (NPR)



N 577 706 298 809 381

Fresh Air (NPR)



N 577 706 298 809 381

Talk of the Nation (NPR)



N 577 706 298 808 381



TABLE 6

Partisan Divisions on Readership of Print Newspapers

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

Wall Street Journal

Does not read regularly 92% 93% 97% 97% 94%

Reads regularly 8% 7% 3% 3% 6%

N 472 730 308 873 377

USA Today



N 473 730 308 872 378

Washington Post



N 473 730 308 872 378

New York Times



N 472 730 308 872 377



TABLE 7

Partisan Divisions on Usage of Website News Sources

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

FOX news (foxnews.com)

Does not visit website regularly 62% 79% 91% 89% 96%

Visits website regularly 38% 21% 10% 11% 4%

N 471 670 317 773 416

Drudge Report (drudgereport.com)



N 471 671 317 773 416

Yahoo News (news.yahoo.com)



N 472 670 317 772 417

Google News (news.google.com)



N 471 671 317 773 417

CNN (cnn.com)



N 471 671 317 772 417

MSNBC



N 472 671 317 772 417

USA Today (usatoday.com)



N 472 671 317 773 417

Washington Post (washingtonpost.com)



N 472 671 317 773 417


TABLE 7 (continued)

Partisan Divisions on Usage of Website News Sources

Tea Party

Republican

Establishment

Republican

Pure

Independent

Moderate

Democrat

Liberal

Democrat

ABC News (abcnews.com)



N 471 671 317 772 416

New York Times (nytimes.com)



N 472 671 317 773 416



TABLE 8

Sentiment Analysis of Twitter Posts (#gop versus #teaparty)

October 1-24, 2012

Entity Mentioned Average Number of Positive

Statements per Tweet

Average Number of Negative

Statements per Tweet

Pos/Neg Ratio

#gop #teaparty #gop #teaparty #gop #teaparty

Barack Obama .38 .38 .73 .71 .52 .53

Mitt Romney .51 .47 .43 .40 1.18 1.18

Republicans .51 .41 .72 .73 .71 .56

Democrats .36 .33 .80 .87 .45 .37

References

Abramowitz, Alan I., and Kyle L. Saunders. 2006. "Exploring the Bases of Partisanship in the

American Electorate: Social Identity Vs. Ideology.” Political Research Quarterly 59:

175-87.

———. 1998. "Ideological Realignment in the Us Electorate.” The Journal of Politics 60: 634-

52.

Campbell, James E. 2008. The American Campaign: U.S. Presidential Campaigns and the

National Vote, 2nd ed. College Station, TX: Texas A&M University Press.

Freeman, Jo. 1986. "The Political Culture of the Democratic and Republican Parties." Political

Science Quarterly 101: 327-56.

Hale, Jon F. 1995. "The Making of the New Democrats." Political Science Quarterly 110: 207-

232.

Hopkins, David A. 2009. "The 2008 Election and the Political Geography of the New

Democratic Majority.” Polity 41: 368-87.

Jamieson, Kathleen Hall, and Joseph N. Cappella. 2008. Echo Chamber: Rush Limbaugh and the

Conservative Media Establishment: Oxford University Press, USA.

Kamarck, Elaine Ciulla. 1990. "Structure as Strategy: Presidential Nominating Politics in the

Post-Reform Era.” The Parties Respond: Changes in the American Party System: 160-

86.

Kamarck, Elaine Ciulla, and Kenneth M. Goldstein. 1994. "The Rules Do Matter: Post-Reform

Presidential Nominating Politics.” The parties respond (2nd ed.). Boulder, CO:

Westview.

Keith, Bruce E. 1992. The Myth of the Independent Voter: University of California Press.

Killough, Ashley. 2013. "Karl Rove Snaps Back at Sarah Palin." CNN, March 17.

http://politicalticker.blogs.cnn.com/2013/03/17/karl-rove-snaps-back-at-sarah-palin.

Lengle, James I. 1980. "Divisive Presidential Primaries and Party Electoral Prospects, 1932-

1976.” American Politics Research 8: 261-77.

———. 1981. Representation and Presidential Primaries: The Democratic Party in the Post

Reform Era: Greenwood Press.

Levendusky, Matthew. 2009. The Partisan Sort: How Liberals Became Democrats and

Conservatives Became Republicans. Chicago: University of Chicago Press.

Livne, Avishay, Matthew P. Simmons, Eytan Adar, and Lada A. Adamic. 2011. "The Party Is

over Here: Structure and Content in the 2010 Election." 2011.

Martin, Jonathan, Jim Rutenberg, and Jeremy W. Peters. 2013. "Fiscal Crisis Sounds the Change

in G.O.P.'s 'Civil War.'" The New York Times, October 19.

Mascaro, Christopher M., Alison Novak, and Sean Goggins. 2012. "Shepherding and

Censorship: Discourse Management in the Tea Party Patriots Facebook Group." 2012.

Mayer, William G., and Nelson W. Polsby. 1996. The Divided Democrats: Ideological Unity,

Party Reform, and Presidential Elections: Westview Press Boulder, CO.

McGlennon, John. 1998. "Factions in the Politics of the New South.” Party Organization and

Activism in the American South.

Petrocik, John R. 2009. "Measuring Party Support: Leaners Are Not Independents." Electoral

Studies 28: 562-72.

Prior, Markus. 2007. Post-Broadcast Democracy: How Media Choice Increases Inequality in

Political Involvement and Polarizes Elections: Cambridge University Press.

Rae, Nicol C. 1994. Southern Democrats: Oxford University Press.

Rayfield, Jillian. 2013. "Tea Party Unhappy About Karl Rove's War on the Tea Party." Salon,

February 4. http://www.salon.com/2013/02/04/tea_party_unhappy_about

_karl_roves_war_on_the_tea_party/

Reiter, Howard L. 2007. "Party Factions in 2004.” The State of the Parties: The Changing Role

of Contemporary American Politics.

Southwell, Priscilla L. 1986. "The Politics of Disgruntlement: Nonvoting and Defection Among

Supporters and Nomination Losers, 1968-1984. Political Behavior 8: 81-95.

Stroud, Natalie Jomini. 2008. "Media Use and Political Predispositions: Revisiting the Concept

of Selective Exposure.” Political Behavior 30: 341-66.

———. 2011. Niche News: The Politics of News Choice: The Politics of News Choice: Oxford

University Press, USA.

———. 2010. "Polarization and Partisan Selective Exposure.” Journal of Communication 60:

556-76.

Sunstein, Cass R. 2009. Republic. Com 2.0: Princeton University Press.

Teixeira, Ruy. 2009. "The Evolving Democratic Coalition: Prospects and Problems." In What's

Left of the Left: Democrats and Social Democrats in Challenging Times, ed. James

Cronin, George Ross, and James Shoch. Durham, NC: Duke University Press, Pp. 162-

187.

Wattenberg, Martin P. 1991. "The Republican Presidential Advantage in the Age of Party

Disunity.” The politics of divided government: 39-56.

Young, Lori, and Stuart Soroka. 2012. "Affective News: The Automated Coding of Sentiment in

Political Texts.” Political Communication 29: 205-31.

Documents

Tea Party Supporters, Establishment Republicans, and the ... · the Obama Administration’s mortgage relief efforts (Williamson, Skocpol, and Coggin 2011). Santelli argued that Obama