American Behavioral Scientist 2012 Kaltenthaler 1718 34

7/28/2019 American Behavioral Scientist 2012 Kaltenthaler 1718 34

1/18

http://abs.sagepub.com/American Behavioral Scientist

http://abs.sagepub.com/content/56/12/1718The online version of this article can be found at:

DOI: 10.1177/0002764212463357

20122012 56: 1718 originally published online 19 OctoberAmerican Behavioral Scientist

Karl Kaltenthaler and William J. MillerHyper-Partisanship

The Polarized American : Views on Humanity and the Sources of

Published by:

http://www.sagepublications.com

can be found at:American Behavioral ScientistAdditional services and information for

http://abs.sagepub.com/cgi/alertsEmail Alerts:

http://abs.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://abs.sagepub.com/content/56/12/1718.refs.htmlCitations:

What is This?

- Oct 19, 2012OnlineFirst Version of Record

- Nov 5, 2012Version of Record>>

by guest on December 3, 2012abs.sagepub.comDownloaded from
http://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/content/56/12/1718http://abs.sagepub.com/content/56/12/1718http://www.sagepublications.com/http://abs.sagepub.com/cgi/alertshttp://abs.sagepub.com/cgi/alertshttp://abs.sagepub.com/subscriptionshttp://abs.sagepub.com/subscriptionshttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsReprints.navhttp://www.sagepub.com/journalsPermissions.navhttp://abs.sagepub.com/content/56/12/1718.refs.htmlhttp://abs.sagepub.com/content/56/12/1718.refs.htmlhttp://abs.sagepub.com/content/56/12/1718.refs.htmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://abs.sagepub.com/content/early/2012/10/17/0002764212463357.full.pdfhttp://abs.sagepub.com/content/early/2012/10/17/0002764212463357.full.pdfhttp://abs.sagepub.com/content/56/12/1718.full.pdfhttp://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/http://online.sagepub.com/site/sphelp/vorhelp.xhtmlhttp://abs.sagepub.com/content/early/2012/10/17/0002764212463357.full.pdfhttp://abs.sagepub.com/content/56/12/1718.full.pdfhttp://abs.sagepub.com/content/56/12/1718.refs.htmlhttp://www.sagepub.com/journalsPermissions.navhttp://www.sagepub.com/journalsReprints.navhttp://abs.sagepub.com/subscriptionshttp://abs.sagepub.com/cgi/alertshttp://www.sagepublications.com/http://abs.sagepub.com/content/56/12/1718http://abs.sagepub.com/


2/18

American Behavioral Scientist

56(12) 17181734

2012 SAGE PublicationsReprints and permission: http://www.

sagepub.com/journalsPermissions.nav

DOI: 10.1177/0002764212463357

http://abs.sagepub.com

ABS463357ABS561210.1177/0002764212463357American Behavioral ScientistKaltenthaler and Miller 2012SAGE Publications

Reprintsand permission: http://www.agepub.com/journalsPermissions.nav

1The University of Akron and Case Western Reserve University, Akron, OH, USA2Flagler College, St. Augustine, FL, USA

Corresponding Author:

Karl Kaltenthaler, Department of Political Science, The University of Akron, Akron, OH 44325-1904, USA

Email: [email protected]

The Polarized

American: Views on

Humanity and the Sourcesof Hyper-Partisanship

Karl Kaltenthaler1 and William J. Miller2

Abstract

In recent years, many possible explanations have been offered for why America hasbecome a more polarized country. In this article, we introduce a new variable to thediscussion. Rather than focusing on specific policy beliefs or political attitudes, weinstead suggest that social trust can be an important explanatory factor in predictingwhich Americans are most likely to be polarized. Through the analysis of ANES data,we find that individuals who are strong trusters of others are more likely to identifyas Strong Democrats while those less likely to do so are more likely to self-identify as

Strong Republicans. The article concludes by looking at how considering social trustas a predictor of polarization impacts our understanding of American politics.

Keywords

social trust, polarization, public opinion, citizen attitudes

One of the most debated issues in the study of American politics centers on the alleged

polarization of the electorate. Some scholars argue that the American electorate isbecoming increasingly polarized (e.g., Koch, 1994; McDonald & Popkin, 2001;

Olson, 2008). Others contend that American voters are not becoming more polarized

(Hetherington, 2009). Still others have maintained that it is not the electorate that has

become increasingly polarized but rather the political elite in the United States, par-

ticularly in the U.S. Congress (Bond & Fleisher, 2000; Carmines & Stimson, 1989;

Article

http://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/


3/18

Kaltenthaler and Miller 1719

Dixit & Weibull, 2007; Jones, 2001; Layman, 2001; Layman & Carsey, 2002a; Mann,

2006; McCarty, Poole, & Rosenthal, 2009; Poole & Rosenthal, 1984; Rohde, 1991).

This article does not seek to weigh in on the debate about whether the electorate or

the Congress has become more polarized over time; instead, it explores those indi-viduals who think of themselves at the polar extremes of the mainstream American

party spectrum. In other words, this article develops an argument about why some

Americans view themselves as strong partisans, either of Republican or Democratic

leanings. Thus, we hope to explain what individual-level factors make one think of

oneself as a strong Republican or a strong Democrat.

Our argument is that a crucial factor in determining whether one is a hyper-

partisan is how an individual views humanity. We contend that the more negative a

view of humanity a person has, the more likely she is to self-identify as a strong

Republican. Conversely, the more positive someone is toward the nature of humanity,the more likely one is to self-identify as a strong Democrat. When an individual

thinks of herself as a strong Democrat or Republican, she is thinking of herself as

being either a strong liberal Democrat or a strong conservative Republican. A strong

conservative Republican is likely to view the social world from a pessimistic basis,

believing that people are, generally, not trustworthy. A strong liberal Democrat is

likely to have a more positive view of humanity. Thus, it is the social psychology of

the individual that shapes a large part of the ideological predisposition that culminates

in a particular partisan identity.

This study will first explore what others have argued about the sources of polariza-tion among the American electorate. The section to follow develops an argument about

how views of humanity shape partisan worldviews. From this, a set of hypotheses are

derived that are then tested in a regression analysis employing data from the 2008

American National Election Study (ANES). The article concludes with what the find-

ings tell us about the nature of polarized partisanship in the United States.

Definitions of Political Polarization

What do we mean when we refer to political polarization within the electorate?Esteban and Ray (1991, 1994) have argued that polarization is the process whereby

individuals feel alienated from each other. They identify with a particular group, say

a political party or an ideological movement, and they believe themselves to be at

odds with the out-group. They call this an identification-alienation framework. In

a polarized situation, by identifying with one group, an individual feels alienated from

another group.

DiMaggio, Evans, and Bryson (1996) have created a definition of polarization that

is often used in the literature. They posit that polarization in the electorate is the

movement toward the poles of the Left-Right distribution of voters. Thus, if the elec-torate is polarized, voters are clustered on the ideological Left and Right extremes,

with the middle being relatively empty. What is important for purposes of this study

is who the people at the polar ends of the distribution of the electorate are and why

they are there.



4/18

1720 American Behavioral Scientist56(12)

Common Explanations of Elite and Voter Political Polarization

There has been a large amount of literature generated on the issue of political polariza-

tion in the United States in recent decades. The literature tends to focus on politicalelite polarization, voter polarization, or the nexus between the two.

Some have argued that the institutional changes of the past several decades in the

United States should have reduced political elite polarization because measures to

promote openness in government should logically have reduced distrust among elites

and thus reduced a sense of us versus them (see Stasavage, 2007). But Stasavage

(2007) has argued that the opposite has happened. He and others have pointed to the

political parties convention delegates, candidates, and members of Congress as all

having grown increasingly polarized on social welfare, racial, and cultural issues

(Bond & Fleisher, 2000; Carmines & Stimson, 1989; Layman, 2001; Layman &Carsey, 2002a; Poole & Rosenthal, 1984; Rohde, 1991; Stasavage, 2007; Stone,

Rapoport, & Abramowitz, 1990). Some have argued that when the leadership of the

parties polarizes, it can pull the whole party system in a centripetal pattern. (Layman

& Carsey, 2002b, p. 200). Some of the explanations offered for this elite polarization

are gerrymandering (Wolfson, 2006) and ones location in the electoral calendar

(Nicholson, 2003). This elite polarization may seem to make voters more polarized

because the parties have offered them polarized candidates. The voterspositions on

the issues may be moderate, even if theirchoices in the voting booth are not (McGhee

& Krimm, 2009, p. 348).Layman and Carsey (2002b, p. 227) argue that elite polarization is indeed having

an impact on voter polarization in a concrete way:

Our evidence confirms that this is part of the partisan change process. However,

we also clearly show that some individuals respond to increases in elite-level

polarization not by changing their party ties but by changing their issue attitudes

in response to those ties. When Democratic and Republican elites are clearly

differentiated on an issue and citizens recognize that difference, then a signifi-

cant number of party identifiers will bring their attitudes on the issue closer intoline with the stance of their partys leaders. When party elites are polarized on

more than one issue agenda, the process of party identifiers converting their

positions on issues toward the dominant stand of their parties means that issue

dimensions, such as social welfare and culture, that were once orthogonal to

each other will not remain so over time.

Layman and Carsey (2002b, p. 228) also argue that party politics exerts a strong

influence on political values and issue attitudes among the electorate. Those individu-

als who are most attached to a particular political party are likely to be most influencedby the cues that they get from the party leadership on issues (Carmines & Stimson,

1989; Claassen & Highton, 2009, p. 546; Kam, 2005; Nie, 1976; Zaller, 1992). But it

is not just partisan attachment; it is also awareness of politics that will play a role in

http://abs.sagepub.com/http://abs.sagepub.com/http://abs.sagepub.com/


5/18


shaping how elite behavior and cues affect the electorate. Thus, the most politically

sophisticated and aware people are most likely to be affected by relative positions of

political elites (see Sniderman, Brody, & Tetlock, 1991; Zaller, 1992).

The above arguments about the process of elite and voter polarization, by nomeans, represent a consensus view. Several scholars have argued that the data do not

bear out the argument that the American electorate in becoming more polarized.

Hetherington (2001, 2009) has shown that if one uses the definition of polarization

as a growing bi-modality of American voters at the extremes, then there is little evi-

dence of polarization. The distribution of American voters is not clearing out the

middle as the polarization hypothesis would suggest.

McGhee and Krimm (2009) are also prominent critics of the polarization thesis.

They point to the growth in independent registration among voters. This would indi-

cate less partisan attachment and not more. It might make the electorate appear morepolarized, however, because the voters who are left with partisan registrationswho

then vote in the primaries (particularly closed primaries)are the most ardent parti-

sans and most ideological voters, thus creating the appearance of a more polarized

electorate.

The Focus of This Study

This study does not seek to answer the question of whether there is indeed a growing

process of polarization in the American electorate or what has or has not driven thatprocess. This study is not interested in change in partisanship over time but rather

explores the extremes of the partisan distribution in the United States in a cross-

sectional snap-shot. This study focuses on explaining what individual-level factors

put voters at the extreme ends of the Democrat-Republican distribution of voters.

The studys dependent variable is self-identification as a strong Democrat or a

strong Republican. Thus, we seek to identify the individual-level factors that would

account for why an individual would think of himself or herself as a strong Democrat

or strong Republican.

It is common for models meant to predict partisanship to include ideology as anexplanatory variable (Abramowitz & Saunders, 1998, 2006; Alvarez & Garcia Bedolla,

2003; Carmines & Berkman, 1994; Carmines & Stanley, 1992; Franklin & Jackson,

1983). We have chosen to do the same despite acknowledging that they are so closely

linked in a persons mind that there may be a high degree of endogeneity in a model with

both variables included. This is because a person who views herself as a strong partisan

is also very likely to view herself as a strong ideologue, leaning toward one or the other

end of the ideological distribution. To control for this possibility, we will present results

demonstrating the inclusion and exclusion of ideology as an explanatory variable.

Views on Humanity and Ideological Self-Identification

We contend that a persons view on the basic nature of humanity will play a large role

in shaping her general ideological worldview. A persons basic view of human nature



6/18


refers to whether a person believes that other people can be thought of as being inher-

ently good or inherently bad. If an individual thinks of human nature as inherently

bad, that person will be less willing to trust other people in general. If, on the other

hand, a person views human nature as being inherently good, this person is morelikely to trust others in general than the person who has a negative view of basic

human nature.

This propensity to trust or not trust others has profound implications toward how

one views politics (Binning, 2007; Brewer & Steenbergen, 2002; Brewer, Gross, Aday,

& Wilnat, 2004; Citrin & Sides, 2008; Herreros & Criado, 2009; Inglehart, 1997;

Rothstein & Uslaner, 2005; Torgler, 2007; Uslaner, 2000-2001). While much research

has focused on how social trust affects a whole range of attitudes toward political

phenomena, there has been very little research on how social trust affects basic parti-

san self-identification. We contend that this is an important oversight as social trustforms a crucial part of the psychological basis for an individuals general ideological

worldview.

Let us first address what we mean by ideology. Ideology is a coherent set of ideas

meant to generally explain political phenomena that a person confronts as well as offer

a general set of remedies to political issues that arise. Scholars have looked at political

ideology as being a set of ideas that sit along two basic dimensions. The classical ideo-

logical dimension that has been referred to for a very long time is the traditional Left-

Right dimension (Lipset & Rokkan, 1967). This dimension was characterized by how

an individual viewed the proper role of the state in managing the economy and distrib-uting wealth in a society. People who were on the Left side of the ideological spectrum

were likely more in favor of the state playing an activist role in the economy and

redistributing wealth from the more fortunate (wealthy) to the less fortunate (the poor).

This view was premised on the argument that the wealthy achieved their wealth in a

system that was largely stacked in their favor. Thus, it was only fair that the state play

a redistributive role and manage the market in order to aid those who were losers

from the outcomes of the economic and political system.

The traditional Right, on the other hand, largely opposed a large redistributive and

economic management role for the state, as the wealthy were assumed to have achievedtheir economic gains through hard work and perseverance and therefore a large, redis-

tributive, and economic activist state was not fair to the wealthy. Thus, the Left and the

Right both couched their general arguments about the state in terms of fairness.

Political science viewed the political bases of support for these Left and Right posi-

tions as a function of a persons class background (Lipset & Rokkan, 1967). Thus, a

person who was born into the working class was likely to be more supportive of the

Left, because of her identification with those who looked to the state for redress of

social and economic inequality. Conversely, those who were born into the middle and

upper classes would tend to support the Right because they would want the state toplay a minimal role in redistributing wealth to those who had lower incomes as funds

for that redistribution would come from the wealthy through taxation and other means.

Therefore, a persons placement on the Left-Right ideological space was viewed as a

function of economic self-interest.



7/18


But there has been another ideological dimension that has been identified by politicalscientists more recently that is meaningful: the Right authoritarian-Left libertarian dimen-

sion (Kitschelt, 1995). This ideological dimension does not center on economic issues as

the traditional Left-Right dimension does but rather is concerned with the amount of

freedom or constraint the state should create in society for its citizens. A person who is

viewed as being more on the Right authoritarian end of the dimension would believe that

the state should play a larger law and order role in society as well as play some role in

preserving traditional moral values. An individual who is seen as more towards the Left

libertarian end of the dimension would believe that the state should try to maximize indi-

vidual freedom of expression and democratic participation and place less emphasis onlaw and order and the state playing a role in defending traditional moral values. The Right

authoritarian has a view of society as needing a Leviathan, and the Left libertarian view

is that man is best served by allowing people the maximum input into governance without

being threatened by a state that infringes on their personal rights.

Inglehart (1977, 1997) has argued that the Right authoritarianLeft libertarian

dimension of politics1

has grown more important in many advanced industrial

democracies as basic needs have been met by economic growth and the creation of

welfare states.2 Thus, political parties in such states have tended to add authoritarian-

libertarian issues to their platforms as these issues have resonated more with theirpublics. Right parties have tended to take authoritarian stances, arguing for the need

for law and order measures in society, whereas Left parties have tended to take more

libertarian stances in this New Politics.3

When the two dimensions of ideology are combined, we are presented with a

schema that looks like the following:

When we apply this schema to American politics, we find that because of the nature

of the electoral system and its tendency to punish small parties, the two main parties,

the Republicans and the Democrats, have absorbed New Politics issues into their

platforms alongside the traditional stances on economic distributional issues. TheRepublican Party has maintained its traditional stance as being the party less inclined

to support economic redistribution in the American party system. The Democratic

Party has maintained its traditional platform of being the major American political

party more inclined to support economic redistribution and an economically activist

state to aid the economically less advantaged.



8/18


In terms of the New Politics issues, the Republican Party has also taken on the

mantle of being the party home for Right authoritarian votersthat is, those who are

more in favor of law and order politics and the state playing a role in preserving tradi-

tional moral values. The Democratic Party, on the other hand, has become the majorparty home for Left libertarian voters who want the polity to increase inclusion in the

political process, protect minority rights, and keep the state from trying to protect con-

servative, traditional morality. Thus, most Republican and Democratic voters can be

thought of as situated along the diagonal axis that bisects traditional Left-Right politics

as well as the New Politics. Thus, voters for the two major parties are a mix of mate-

rialists and those who vote along New Politicsthat is, authoritarian-libertarian lines.

We posit that the degree to which one trusts others will play a very important role

in determining the leanings of the highly partisan American. We hypothesize that the

more social trust an individual says she has, the more likely she is to self-identify as astrong Democrat. Conversely, the less social trust an individual says he has, the more

likely that person is to self-identify as a strong Republican.

Social trust will condition attitudes toward both traditional Left-Right politics, as well

as New Politics. A person with a higher relative degree of social trust will have a more

positive attitude toward redistributive policies because that person will be more positive

toward other people, in general, and believe that the recipients of aid from the government

are deserving of such help. Thus, the social truster will be more likely to identify with the

Left than the Right. Such a person would more likely be a Democrat than a Republican.

Conversely, a person who is less trusting of other people, relative to others, will beless willing to have the state redistribute wealth to others because that person would

not be positively inclined toward others, in general, and would thus not favor such

policies. This person would identify more with the Right and thus would more likely

think of herself as a Republican.

When it comes to New Politics, a persons relative propensity to trust others will

also influence her identification with the Right and the Left. A social truster will likely

identify with the Left on New Politics issues because such a person would not look to

the state to constrain others through law and order policies because such a person would

feel less threatened by others. This identification with the Left on New Politics issueswould more likely make the person think of herself as a Democrat than as a Republican.

Those who are relatively less trusting of others would be more likely to identify

with the Right on New Politics issues. The person who is relatively less trusting of

others would be more likely to want the state to play a more constrained role in soci-

ety in terms of law and order because of the more generalized lack of trust in others.

Because of this identification with the Right on New Politics issues, the less trust-

ing individual is more likely to be a Republican in the American context.

Data, Measures, and Methods

We employ ordered logistic regression analyses given the dependent variable is bino-

mial and ordinal in nature. As mentioned earlier, we employ data from the 2008

American National Election Survey. The number of respondents in this sample was

2,203. Our final models include 518 respondents who responded to all questions



9/18


employed as variables in our study. Most respondents are excluded for not responding

to questions regarding their ideology or income. Based on diagnostics, there are no

clear patterns in the excluded cases.

The Dependent Variable

The dependent variable in this model is strong partisan beliefs. The ANES asks respon-

dents to identify the political party they most closely identify with and are able to break

down respondents into a 7-point scale ranging from strong Democrat to strong

Republican. In order to examine our question of what individual-level factors deter-

mine whether someone is a strong Democrat or strong Republican, we create a binary

variable for each. Respondents who identify as strong Democrats are coded 1, with all

other respondents as 0. Then, all respondents who identify as strong Republicans (7 onthe 7-point scale) are coded 1, with all other respondents as 0. By doing so, we end up

with two dependent variables.

A Picture of Polarization

The breakdown of the dependent variables find that 25.0% of respondents label them-

selves as strong Democrats, while 9.9% declare to be strong Republicans. While it is

unclear we there are so many fewer strong Republicans compared to strong

Democrats, it could be that many of the strong conservatives have come to label them-selves Independents rather than strong Republicans.

We can go further in a descriptive analysis by looking at demographic differences

between strong Democrats, strong Republicans, nonpolarized individuals, and all respon-

dents. By doing so, we can get a picture of the typical polarized Democrats and Republicans.

When we look at the demographics (as seen in Table 1), we find that there are many

splits that follow widely known partisan trends. Strong Democrats, for example,

appear to have relatively lower incomes, while strong Republicans report relatively

higher incomes. Likewise, strong Republicans are more likely to be married than the

average respondent to the ANES, while the typical strong Democrat is less likely to bemarried. The characteristic racial divide emergeswith strong Republicans more

likely to be White than the average respondent and strong Democrats more likely to be

members of a minority group. Lastly, we see a gender gap in terms of polarization with

women more likely to be strong Democrats and men strong Republicans than their

corresponding average respondents. None of these findings are particularly unex-

pected and confirm previous findings and beliefs.

Where we find more unique results are when looking at region and age. First, we

can see that the Northeast and Southern Census regions have clear polarization pat-

terns. Individuals living in the Northeast are considerably less likely to be polarized,while those in the South are more likely to be politically polarized than nonpolarized.

It is, by far, the most polarized region in the country. The South is a region where

strong Republicans outnumber strong Democrats, but not by a large margin. The North

Central region has less polarization and there is a slightly greater likelihood of strong



10/18


Democrats than Republicans and the West shows the same for strong Republicans

outnumbering strong Democrats. This suggests that there are some marked regional

considerations related to polarization in the United States. A pattern emerges with age

as well, with older respondents being more likely to be polarized (regardless of party)than their younger counterparts. As individuals age, they appear to move toward the

poles when considering their levels of partisanship.

The Independent Variables

Below are the questions that we use to test the predictive power of our argument about

the role of social trust contributing to polarization in the United States. We also test a

set of alternative hypotheses to determine relative predictive power of the most com-

mon arguments about what drives said polarization.We include a very common measure of relative social trust in the model to test our

primary hypothesis. The question asks, Can people be trusted? The respondent has the

choice of answering that eitheryou cant be too carefulormost people can be trusted.

We argue that those individuals who believe people can be trusted will be more likely

to be strong Democrats and that those who do not believe people can be trusted will be

more likely to be strong Republican identifiers.

Given the links often discussed between religious beliefs and religiosity and parti-

sanship, we include two religious measures (Abramowitz & Saunders, 2006; Carmines

& Stanley, 1992; Franklin & Jackson, 1983). First, the variable examining religiosityasks, Would you say you religion provides someguidance in your day-to-day living,

quite a bit of guidance, or a great deal of guidance in your day-to-day life? Respondents

are coded for not believing religion is important, religion provides some guidance, reli-

gion provides quite a bit of guidance, and religion provides a great deal of guidance.

The religious beliefs measure asks, Which of the these statements comes closest to

describing your feelings about the Bible? Respondents are asked to choose from the

following: 1 = The Bible is the actual word of God and is to be taken literally, word for

word; 2 = The Bible is the word of God but not everything in it should be taken literally,

word for word; and 3 = The Bible is a book written by men and is not the word of God.We use a 7-point scale of ideology to measure where respondents place them-

selves in terms of political beliefs (Abramowitz & Saunders, 1998, 2006; Alvarez

& Bedolla, 2003; Carmines & Berkman, 1994; Carmines & Stanley, 1992; Franklin

& Jackson, 1983). The scale runs from 1, indicating strong liberal, to 7, indicating

strong conservative. Strong liberals are predicted to be strong Democrats, and

strong conservatives are most likely to be strong Republicans. As touched on above,

there are concerns regarding endogeneity when utilizing ideology to predict parti-

sanship. As a result, we will present our models showing results both with and

without ideology included. We also include a measure of race. Our variable isdummy coded, with all non-White respondents coded as 1 and White respondents

coded 0. Previous literature has demonstrated that race needs to be considered

when examining partisanship and polarization (Avery, 2007; Hutchings & Valentino,

2004; McDaniel & Ellison, 2008). As with typical party identification, White



11/18


respondents are predicted to more likely be strong Republicans, and non-White

respondents are predicted to be more likely strong Democrats.

We also include a number of control variables for gender, age, education, and

income as these are very common controls in studies of public attitudes toward polar-

ization in the United States. We also include a measure of job security as a way to

control for the economic insecurity of the respondent. We include this measure because

it is possible that the respondents economic insecurity can affect the responses givenin the survey.

Results of the Analyses

The ordered logistic regression analysis results for our model are presented in Table 2.

Models 1 and 2 examine predictors of strong Democrats. Model 1 includes ide-

ology as an explanatory variable, while Model 2 removes ideology given possible

concerns of multicollinearity from having it explain party identification (a variable

it is closely linked with). We find relative consistency between both models interms of predictions. However, our variable of most interest theoreticallysocial

trustis not significant when ideology is included. When ideology is excluded,

however, we find that the more individuals trust others, the more likely they are to

Table 1. Demographic Breakdown

AllRespondents

StrongDemocrat

StrongRepublican Nonpolarized

Lowest 25% income 22.8 27.1 10.2 23.4

25%-50% income 26 30.1 19.7 25.6

50%-75% income 29.5 24.7 32.5 30.7

Highest 25% income 21.7 18.1 37.6 20.3

Married 42.2 36.7 61.4 41.5

Not married 57.8 63.3 38.6 58.5

White 51.4 27.4 82.1 55.9

Not White 48.6 72.6 17.9 44.1

Northeast 10.9 10.2 5.7 12North Central 17.1 17.4 13.5 17.5

South 47.3 50.3 55.7 44.9

West 24.7 22.1 25.2 25.6

Male 43 39 46.7 43.9

Female 57 61 53.3 56.1

Lowest 25% age 25.6 18.4 13.3 30.3

25%-50% age 24.2 22.4 24 24.9

50%-75% age 25.3 29.5 28 23.2

Highest 25% age 24.9 29.7 34.7 21.6



12/18


identify as strong Democrats (significant at the .05 level). Ideology, as would beexpected, is significant at the .01 level when included in the model, with strong

liberals showing as being more disposed to being strong Democrats. Being non-

White proves to be a strong predictor (.01 level) in both models in the assumed

direction. When ideology is excluded, age becomes the only other significant pre-

dictor (at the .05 level). Being older is linked with individuals being more likely to

be strong Democrats. No other variables produce significant relationships, includ-

ing religion or other demographics.

For the strong Republican modelsthree and fourwe find similar results.

When we account for ideology, it is significant at the .01 level and demonstrates thatstrong conservatives are more likely to be strong Republicans. In the same model,

social trust is significant at the .05 level. Unlike in the Democrat model, however,

the relationship is negative (as hypothesized) and suggests that social distrusters are

more likely to be strong Republicans. When ideology is removed from the model,

we again see congruity with the Democratic model. Social trust remains significant

at the .05 level in the negative direction. Non-White racial statusas it does in the

Democratic model without ideologyremains as a significant predictor (at the .01

level) and demonstrates that White respondents are more likely to be strong

Republicans. Unlike in the Democratic models, however, in our Republican modelthat excludes ideology, we find both religious variables to be significant predictors.

At the .05 level, individuals who believe the Bible to be the literal word of God and

who claim to use religion to guide their life are shown to be more likely to be strong

Table 2. Logistic Regression Model Results: Dependent VariableStrong Partisanship

Model 1: Strong

Democrat

Model 2: Strong

Democrat

Model 3: Strong

Republican

Model 3: Strong

Republican

Independent

variables Coef. SEFirst

Diff. Coef. SEFirst

Diff. Coef. SEFirst

Diff. Coef. SEFirst

Diff.

Social trust .289 .244 .505** .197 .085 -.721** .308 .053 -.669** .269 .051

Religion guide life -.206 .155 -.090 .123 .358 .196 .442** .170 .062

Biblical literalism -.047 .206 -.147 .164 .281 .272 .556** .236 .070

Ideology -.389*** .085 .375 .852*** .131 .454

Non-White 1.327*** .267 .202 1.534*** .220 .251 -1.109*** .335 .083 -1.596*** .296 .134

Age .019* .009 .184 .020** .008 .225 -.007 .012 .001 .010

Education .055 .051 .083 .045 -.064 .072 -.028 .065

Household income -.037 .124 -.111 .100 .167 .166 .261 .142

Gender .239 .184 .141 .148 -.137 .219 -.200 .170Worry lose job .114 .148 .118 .114 -.167 .212 -.272 .185

Pseudo R2 .1357 .1042 .2626 .1509

Log likelihood -235.52 -357.93 -169.05 -227.89

Log likelihood 2 73.94 83.26 120.38 81.01

N 516 713 516 713

Note: Figures are unstandardized coefficients shown alongside standard errors.

*p < .1. **p < .05. ***p < .01.



13/18


Republicans. This finding aligns with the belief that conservative Christians are

more likely to be strong Republicans.

When we consider the impact played by ideology in both models, we find that the

inclusion of ideology does far more in explaining strong Republican identificationcompared to strong Democrats. If we look at the pseudo-R

2values for the models, we

find that the inclusion of ideology leads to the strong Democrat model explaining

approximately 3.5% more of the variance in our dependent variable, whereas its inclu-

sion leads to an increase in model explanatory power of just over 10% when looking

at the strong Republican model.

Given that we use logit, the coefficients reported in Table 2 do not represent the

marginal effects of the independent variables on the dependent variables. As a result,

we report the first differences of our significant independent variables in Table 2 as

well. The first difference reported measures the probability of the dependent variable,signifying identification as a strong partisan, when the independent variable of interest

is moving from its minimum to its maximum value, while all remaining independent

variables are held at their appropriate measures of central tendency. This allows us to

consider the substantive significance of each independent variable to explaining varia-

tion in the dependent variable.

When looking at the strong Democrat models, we find that ideology has the largest

first difference value.375. This value tells us that taking an individual from strong

liberal to strong conservative makes them 37.5% more likely to be a strong Democrat

when all other variables are held constant at their central tendency. Non-White racialstatus is the second most substantive predictor (.202) followed by age (.184). With

ideology removed, however, we see social trust having a significant impact in the

model with a value of .085, yet it still has less substantive impact than non-White

racial status (.251) and age (.225).

Both Republican models present similar results. In Model 3, ideology has a first

difference of .454, while social trust is only at .053. Non-White racial status splits the

two with a first difference value of .083. When ideology is removed in Model 4, being

non-White has the largest first difference value (.134) while biblical literalism, reli-

gion guide life, and social trust and all increase the likelihood of someone being astrong Republican by between 5.1% and 7.0% when moved from their minimum to

maximum values.

In short, our results demonstrate that while social trust is an important predictor that

must be considered when discussing the increased polarization of the American elec-

torate, previous research on partisan identity has ignored this important social psycho-

logical factor.

Conclusion

This article has examined what factors lead Americans to identify as polarized parti-

sans. Moreover, we have attempted to test congruency in explanatory factors between

strong Democrats and strong Republicans. Most importantly, we introduced the



14/18


concept of social trust into Americas understanding of political polarization. We

argued that Americans general predispositions toward others will impact the likeli-

hood of their declaring to be strong partisans. For Democrats, we posited that being

social trusters would increase the chances of an individual identifying as a strongpartisan, while for Republicans we hypothesized that being a distruster would increase

the odds of identifying as a strong partisan.

Much of the current polarization research focuses on the role of varying ideological

and demographic characteristics, including ideology, age, income, and religious

beliefs and behavior. In this article, however, we demonstrate how social trust condi-

tions the psychological basis for an individuals general ideological worldview

including their propensity to be hyper-partisans. By including social trust, the

discussion about polarization begins to see a shift in interpretation. First, it moves the

decision away from any easily traced demographic basis. While variables such as race,income, or education can be gleaned from national census data, there is no similar

measure of individual-level social trust available. Without being able to fully account

for what has been shown to be a meaningful predictor, it makes the process of predict-

ing the future of polarization more difficult.

It is especially important to note this when we remember that social trust still proves

to be significant even when ideology is included within the model. It comes as little

surprise that strong liberals would be strong Democrats or strong conservatives would

be strong Republicans given the long observed relationship between ideology and

partisanship in the United States. This only goes further to explain why it is importantto consider the potential role played by social trust in shaping political attitudes.

If hyper-partisanship is impacted by individual social trust, it makes it significantly

more difficult to be optimistic regarding hope for more bipartisan political beliefs,

attitudes, and policies in the United States. When polarization is based on politics or

demographic characteristics, we can change policies or alter the discourse to bring

people back to the middle. If it is caused by social psychologythat is, general predis-

positions towards othershowever, it will require a more fundamental shift in citizen

attitudes, which would be far more difficult to achieve.

One of the key questions that arise from this research is, What does the relation-ship between social trust and polarized partisanship indicate for the future of partisan

relations in the United States? There is reason to expect that polarized partisanship is

likely to increase in the near to medium term. The reason for this is that the economic

crisis that befell the United States in 2008 has left in its wake a weakened economy

and bleak prospects for many younger Americans. As we know that an individuals

pessimism about his or her future tends to make a person less trusting of others, this

means that the proportion of people who have a dim view of humanity will grow in

this country. This will increase the ranks of the hyper-partisan Right, as will also

likely increase the ranks of the partisan Left who look to the government to protectand aid them from economic circumstances. Thus, this research does not offer much

room for optimism that the highly partisan nature of American politics will diminish

in the near future.



15/18


Declaration of Conflicting Interests

The authors declared no conflicts of interest with respect to the authorship and/or publication

of this article.

Funding

The authors received no financial support for the research and/or authorship of this article.

Notes

1. These issues are often referred to as New Politics issues.

2. Ingelhart referred to postmaterialist values as values that stressed helping others achieve a

voice in politics and society and placed emphasis on things like environmental problems.

He viewed this as a new dimension of the Left. They developed in societies where basicneeds were met and the more economically secure could focus on matters beyond their

own pocketbooks. In multiparty systems, postmaterialists would often join Green parties.

In party systems given to two-party dominance, the postmaterialists would likely join the

left-leaning party in the system. This would be the case in the United States.

3. This libertarianism should not be confused with the specific American brand of liber-

tarianism, which is more focused on economic freedom from state action, although it does

include an element of social freedom as well. American libertarianism is a more conser-

vative form of libertarianism, which focuses more on reducing the states role in society

overall, whereas the brand of libertarian politics referred to by scholars such as Inglehart(1977, 1997) envisions a positive and even activist role for the state to play in society by

bringing people into the political process and intervening in societal processes to ensure

that individuals are free from oppression from others.

References

Abramowitz, A., & Saunders, K. (1998). Ideological realignment and the U.S. electorate. The

Journal of Politics, 60, 634-652.

Abramowitz, A., & Saunders, K. (2006). Partisanship in the American electorate: Social identity

vs. ideology.Political Research Quarterly, 59, 175-187.Alvarez, M., & Bedolla, L. G. (2003). The foundations of Latino voter partisanship: Evidence

from the 2000 election.Journal of Politics, 65, 31-49.

Avery, J. M. (2007). Race, partisanship, and political trust followingBush versus Gore (2000).

Political Behavior, 29, 327-342.

Binning, K. (2007). Its us against the world: How distrust in Americans versus people-in-

general shapes competitive foreign policy preferences.Political Psychology, 28, 777-799.

Bond, J. R., & Fleisher, R. (2000).Polarized politics: Congress and the President in a partisan

era. Washington, DC: CQ Press.

Brewer, P., Gross, K., Aday, S., & Willnat, L. (2004). International trust and public opinionabout world affairs.American Journal of Political Science, 46, 93-109.

Brewer, P., & Steenbergen, M. (2002). All against all: How beliefs about human nature shape

foreign policy options.Political Psychology, 23, 39-58.



16/18


Citrin, J., & Sides, J. (2008). Immigration and the imagined community in Europe and the

United States.Political Studies, 56, 33-56.

Carmines, E., & Berkman, M. (1994). Ethos, ideology, and partisanship: Exploring the paradox

of conservative Democrats.Political Behavior, 16, 203-218.Carmines, E., & Stanley, H. (1992). The transformation of the New Deal party system: Social

groups, political ideology, and changing partisanship among Northern whites, 1972-1988.

Political Behavior, 14, 213-237.

Carmines, E., & Stimson, J. (1989).Issue evolution: Race and the transformation of American

politics. Princeton, NJ: Princeton University Press.

Claassen, R. L., & Highton, B. (2009). Policy polarization among party elites and the signifi-

cance of political awareness in the mass public.Political Research Quarterly, 62, 538-551.

DiMaggio, P., Evans, J., & Bryson, B. (1996). Have Americans social attitudes become more

polarized?American Journal of Sociology, 102(3), 690-755.Dixit, A. K., & Weibull, J. W. (2007). Political polarization.Proceedings of the National Acad-

emy of Sciences, 104, 7351.

Esteban, J., & Ray, D. (1991). On the measurement of polarization (Institute for Economic

Development Discussion Paper 18). Boston: Institute for Economic Development.

Esteban, J., & Ray, D. (1994). On the measurement of polarization.Econometrica, 62, 819-851.

Franklin, C., & Jackson, J. (1983). The dynamics of party identification. The American Political

Science Review, 77, 957-973.

Herreros, F., & Criado, H. (2009). Social trust, social capital, and perceptions of immigration.

Political Studies, 57, 337-355.Hetherington, M. J. (2001). Resurgent mass partisanship: The role of elite polarization.Ameri-

can Political Science Review, 95, 619-632.

Hetherington, M. J. (2009). Review article: Putting polarization in perspective.British Journal

of Political Science, 39, 413-448.

Hutchings, V. L., & Valentino, N. A. (2004). The centrality of race in American politics.Annual

Review of Political Science, 7, 383-408.

Inglehart, R. (1977). The silent revolution: Changing values and political styles. Princeton, NJ:

Princeton University Press.

Inglehart, R. (1997).Modernization and postmodernization: Cultural, economic, and politicalchange in 43 Societies. Princeton, NJ: Princeton University Press.

Jones, D. R. (2001). Party polarization and legislative gridlock. Political Research Quarterly,

54, 125-141.

Kam, C. D. (2005). Who toes the party line? Cues, values, and individual differences.Political

Behavior, 27, 163-182.

Kitschelt, H. (1995). The Radical right in Western Europe. Ann Arbor: University of Michigan

Press.

Koch, J. W. (1994). Group identification in political context.Political Psychology, 15, 687-698.

Layman, G. (2001). The great divide: Religious and cultural conflict in American party politics.New York: Columbia University Press.

Layman, G. C., & Carsey, T. M. (2002a). Party polarization and conflict extension in the

American electorate.American Journal of Political Science, 46, 786-802.



17/18


Layman, G. C., & Carsey, T. M. (2002b). Party polarization and party structuring of policy atti-

tudes: A comparison of three NES panel studies.Political Behavior, 24, 199-236.

Lipset, S. M., & Rokkan, S. (1967). Party systems and voter alignments: Cross-national per-

spectives (International Yearbook of Political Behavior Research, Volume 7). New York:The Free Press.

Mann, T. E. (2006). Polarizing the House of Representatives: How much does gerrymandering

matter?Red and Blue Nation, 263-283.

McCarty, N., Poole, K. T., & Rosenthal, H. (2009). Does gerrymandering cause polarization?

American Journal of Political Science, 53, 666-680.

McDaniel, E. L., & Ellison, C. G. (2008). Gods party? Race, religion, and partisanship over

time.Political Research Quarterly, 61, 180-191.

McDonald, M. P., & Popkin, S. L. (2001). The myth of the vanishing voter.American Political

Science Review, 95, 963-974.McGhee, E., & Krimm, D. (2009). Party registration and the geography of party polarization.

Polity, 41, 345-367.

Nicholson, S. P. (2003). The political environment and ballot proposition awareness.American

Journal of Political Science, 47, 403-410.

Nie, N. (1976). The changing American voter. Cambridge, MA: Harvard University Press.

Olson, J. (2008). Whiteness and the polarization of American politics.Political Research Quar-

terly, 61, 704-718.

Poole, K. T., & Rosenthal, H. (1984). The polarization of American politics. The Journal of

Politics, 46, 1061-1079.Rohde, D. (1991).Parties and leaders in the postreform House. Chicago: Chicago University

Press.

Rothstein, B., & Uslaner, E. (2005). All for all: Equality, corruption, and social trust. World

Politics, 58, 41-72.

Sniderman, P. M., Brody, R. A., & Tetlock, P. E. (1991).Reasoning and choice: Explorations in

political psychology. New York: Cambridge University Press.

Stasavage, D. (2007). Polarization and publicity: Rethinking the benefits of deliberative democ-

racy.Journal of Politics, 69, 59-72.

Stone, W., Rapoport, R., & Abramowitz, A. (1990). Sex and the caucus participant: The gendergap in 1984 and 1988.American Journal of Political Science, 34, 725-740.

Torgler, B. (2007). Trust in international organizations: An empirical investigation focusing on

the United Nations (Working Paper 2006-20). Basel, Switzerland: Center for Research in

Economics, Management, and the Arts.

Uslaner, E. (2000-2001). Producing and consuming trust. Political Science Quarterly, 115,

569-590.

Wolfson, A. (2006). Red and blue nation? Causes, consequences, and correction of Americas

polarized politics. Issues in Governance Studies, Number 1. Washington, DC: The Brook-

ings Institution.Zaller, J. (1992). The nature and origins of mass opinion. Cambridge, UK: Cambridge Univer-

sity Press.



18/18


Bios

Karl Kaltenthaler is the director of research at the Ray C. Bliss Institute of Applied Politics

and a professor in the Department of Political Science at the University of Akron. He is also onthe faculty in the Department of Political Science at Case Western Reserve University. He is the

author of several articles and books on public opinion, political economy, and political extrem-

ism. His current research centers on the role that social psychology plays in shaping attitudes

toward politics.

William J. Miller is an assistant professor of public administration at Flagler College. He is

the author of several articles and books on the Tea Party movement, campaign strategy, public

opinion, political extremism, and the pedagogy of political science and public administration.

His current research examines how social psychology (particularly social trust) impacts citizenattitudes toward various policies and functions of government.

Documents

American Behavioral Scientist 2012 Kaltenthaler 1718 34