Conference PaperPDF Available

Abstract

Abstract: Partisans in American politics are increasingly biased against their political opponents on a personal level. This partisan rancor has been found to rival even racial prejudice in its intensity (Iyengar and Westwood 2015), and has only increased in the era of the Trump candidacy and presidency. But to what extent do extra-partisan social attitudes shape these political animosities? Who were Trump supporters before Trump came along? Can we form a picture of the Trump voter based on social attributes and attitudes alone? Recently, Mason and Wronski (2018) demonstrated that individuals who feel closer to party-aligned racial, religious, and ideological groups are more strongly attached to their party. Yet, little is known about the direction of causality in these relationships. And Trump has attracted at least some voters who are not traditional Republicans. In this paper, we explore how individuals' social attributes and attitudes towards social out-groups predict future Trump support and Republican Party approval. Using the Voter Study Group panel data (2011-2017) we find that particular social identities and feelings toward racial groups do generally predict future support for Trump, but Trump approval is more powerfully motivated by outgroup animosity, while Republican Party approval is more linked to ingroup affection.
Ingroup Lovers or Outgroup Haters?
The Social Roots of Trump Support and Partisan Identity
Lilliana Mason
University of Maryland, College Park
Julie Wronski
University of Mississippi
John Kane
New York University
Abstract: Partisans in American politics are increasingly biased against their political
opponents on a personal level. This partisan rancor has been found to rival even racial
prejudice in its intensity (Iyengar and Westwood 2015), and has only increased in the era
of the Trump candidacy and presidency. But to what extent do extra-partisan social
attitudes shape these political animosities? Who were Trump supporters before Trump
came along? Can we form a picture of the Trump voter based on social attributes and
attitudes alone? Recently, Mason and Wronski (2018) demonstrated that individuals who
feel closer to party-aligned racial, religious, and ideological groups are more strongly
attached to their party. Yet, little is known about the direction of causality in these
relationships. And Trump has attracted at least some voters who are not traditional
Republicans. In this paper, we explore how individuals' social attributes and attitudes
towards social out-groups predict future Trump support and Republican Party approval.
Using the Voter Study Group panel data (2011-2017) we find that particular social
identities and feelings toward racial groups do generally predict future support for Trump,
but Trump approval is more powerfully motivated by outgroup animosity, while
Republican Party approval is more linked to ingroup affection.
****PLEASE NOTE THAT THIS IS A VERY PRELIMINARY DRAFT*****
1
The study of political behavior has long involved questions about how social
group identities influence our political lives. In the early stages, it was assumed that
social group identities helped to inform voters about which party to choose (Berelson,
Lazarsfeld, and McPhee 1948; Campbell, Converse, Miller, and Stokes 1960). In effect,
group identities such as race and religion were assumed to be primary, and to inform
partisan identities. The 2016 election, however, both highlighted and obscured the effects
of social identities on voter choice. First, Trump attracted the votes of many white voters
who had not previously voted Republican (Major, Blodorn, and Blascovich 2016),
clarifying the link between party and race in an unprecedented manner. On the other
hand, Noel (2016) points out that the Trump nomination itself reflected a deep divide
within the Republican Party that has yet to be reconciled. While party and race are
increasingly aligned (Mason 2018), the Trump campaign connected the two in a way that
made even Republican Party leaders uncomfortable.
Here, we examine the relationship between Trump support and social group
identity and attitudes in an era of heightened social and affective polarization. Do Trump
supporters vote for him because they hate the Democratic Party’s social makeup? Or are
Trump supporters merely out to change the status quo? What role do social identities and
social prejudices play in predicting support for Trump in 2016?
Social Sorting
In recent decades, it has become generally well-established that certain social
groups tend to be Democratic, while others tend to be Republican. Not only in public
polls and electoral demographics, but also in political science, as partisan identities have
converged with ideological (Levendusky 2009), religious (Fiorina, Abrams, and Pope
2008; Layman 2001), racial (Giles and Hertz 1994; Mangum 2013) and other political
identities (Campbell and Putnam, 2011). This is often referred to as sorting, usually
known as an increasing correlation between party and ideology (Levendusky, 2009),
though sorting can also be understood as a distinctly social phenomenon. Mason (2018)
uses the term “social sorting” to define this trend, and finds that this type of sorting is
significantly related to affective polarization, or the emotional and visceral dislike of
partisan opponents. While these party-group associations are well-known at the aggregate
2
level, it is not clear how these associations are understood by voters, and therefore how
social sorting psychologically transforms into affective polarization or vote choice.
Ingroup Love or Outgroup Hate?
Brewer (1999) has distinguished between the “ingroup love” felt by ingroup
members toward their own side and the “outgroup hate” felt by ingroup members toward
the outgroup. Brewer (1999) makes the essential point that preferential treatment of
ingroup members does not always coincide with hostility toward outgroup members. She
explains that attachment to a social group can produce outgroup hostility under a few
conditions: feelings of moral superiority, perceived conflict over resources, forced
collaboration, a shared standard of relative worth, or political competition for power.
Cross-cutting identities reduce the prospect of outgroup hostility.
In support of the latter argument, Roccas and Brewer (2002) find that cross-
cutting identities do tend to increase tolerance toward outgroups. Applying this theory to
politics, Mason (2018) has found that as religious, ideological, and racial identities move
into alignment with partisan identities, partisans grow more prejudiced against their
partisan opponents. Recently, Mason and Wronski (2018) demonstrated that individuals
who are closer to these partisan-aligned groups, and cognitively understand the
connection between their racial, religious, and ideological groups and the parties, are
more strongly attached to their party. But how do the various social identities associated
with the parties affect either ingroup affection or outgroup animosity? Is it possible for
partisans to be motivated by preference for victory separately from simple hatred of “the
other”? And, furthermore, would feelings toward Trump himself – often characterized by
distinctly hostile outgroup attitudes be any different than feelings toward the two
parties?
In this paper, we explore both how individual social identities and attitudes
towards party-linked out-groups structure support for Trump and the two parties. We
suggest that, as partisanship becomes increasingly socially-defined, feelings toward
parties and their associated social groups may simultaneously affect each other. However,
it is possible that Trump himself, through his rhetoric or other means, has made the socio-
partisan divide increasingly clear (Sides, Tesler, and Vavreck 2017). His supporters
3
should therefore be expected to be the most aware of or responsive to the social
divide between parties. Trump supporters may be more motivated by outgroup animosity
than the more traditional and institutionalized social groups – parties – who require some
level of positive commitment rather than simple hostility.
Data and Methods
We turned to the Democracy Fund’s Voter Study Group survey, which collected
multiple waves of data in partnership with the Cooperative Campaign Analysis Project
(CCAP) and YouGov.1 This data set includes several thousand respondents who were re-
interviewed (online) across three survey waves: 2011, 2016, and 2017. Importantly, the
first wave of respondents was not yet familiar with Trump as a political figure, and
therefore can be used as a baseline against which to compare later waves. In other words,
those who would become Trump supporters in later waves were (likely) not yet Trump
supporters in 2011. This allows us to examine the extent to which pre-Trump attributes
contribute to post-Trump approval.
Panel Data
We analyze panel survey data from the Democracy Fund’s Voter Study Group. Because
we have multiple observations for every individual, we are able to investigate whether a
given citizen’s change in affect toward politically-aligned social groups during earlier
points in time is predictive of the degree to which that same citizen approves of Trump at
a later point in time. Our interest is in detecting those voters who experienced a change
in their feelings toward a variety or racial/religious groups before Trump’s presidency –
and using those changes to predict those who approve the most of Trump in 2017. Our
secondary aim is to compare these predictions against predictors of 2017 GOP approval
1 More details regarding survey methodology can be found here:
https://www.voterstudygroup.org/publications/2017-voter-
survey/methodology-for-2017-voter-survey
4
(via Republican Party feeling thermometers). By examining the difference between
precursors to Trump approval versus general Republican Party approval, the unique
contribution of Trump support can be discerned.
Respondents were asked to indicate their feelings toward five Democratic-aligned
social groups in each of the three waves: African-Americans, Hispanics, Muslims,
Lesbian/Gay (LG) and Jewish people. Respondents also indicated their feelings toward
two Republican-aligned social groups in each wave: Whites and Christian people.
Our primary analysis thus utilizes an OLS regression model with the following
specification:
Trump Approval (or Party approval) i (2017) = b0 + b1∆Groupi (2016-2011)
+ b2Groupi (2011) + b3PIDi (2011) + b4∆PIDi (2016-2011)
+ b5Ideologyi (2011) + b6∆Ideologyi (2016-2011)
+ b7
δ
i (2011) +
ϵ
i
where Group captures the average feeling thermometer rating of the group
examined in each model (Democratic-linked groups are African-American, Hispanic,
Muslim, LG, and Jewish people; Republican-linked groups are White and Christian
people). These thermometers are recoded to range from 0 to 1 instead of 0 to 100 for ease
of interpretation. PID indicates respondent i’s party identification (seven-point scale,
ranging from “Strong Democrat” (0) to “Strong Republican” (1)); Ideology indicates
respondent i’s self-placement on the five-point ideology scale (ranging from “Very
Liberal” (0) to “Very Conservative” (1)) ; and
δ
includes a bevy of socio-demographic
variables measured in 2011 (i.e., political interest, race, religion, educational attainment,
gender, age, and income).
5
Our model therefore aims to isolate the effect of change (∆) in affect toward
groups aligned with each party on Trump Support, Republican Party Approval, and
Democratic Party Approval by accounting for baseline (i.e., 2011) demographic and
political differences between respondents, as well as differences in baseline affect toward
the politically-aligned groups (distributions of the change in affect toward each group can
be found in the appendix). Trump Support and GOP Approval are both coded to range
from 0 to 1 to account for the different scales of the two variables (1-5 for Trump
approval and 0-100 for Republican thermometer).
We present results of OLS regressions predicting 2017 Trump approval and GOP
approval for each of the seven social groups listed above.
Blacks
In Figure 1, we examine the effect of changes in affect toward Blacks on Trump
and GOP approval. The first variable in both models is change in reported feelings
toward Blacks between 2011 and 2016. For those whose feelings toward Blacks grew
warmer during this 5-year period, their feelings toward Trump in 2017 declined. This is
controlling for baseline 2011 affect toward Blacks, for which a similar result is notable.
Higher 2011 affect toward Blacks is related to lower Trump approval in 2017. So, when
feelings toward Blacks are cold and grow increasingly colder over time, approval for
Trump increases. Interestingly, the same effect cannot be seen in predicting feelings for
the GOP. Warm feelings for Blacks in 2011 has, if anything, a positive effect on feelings
toward the GOP in 2017, and changes in that affect over time do not change feelings for
the GOP.
6
Other interesting differences emerge from the two models. The effects of
partisanship and ideology have similar effects on both Trump approval and GOP affect.
Stronger Republican identity in 2011 leads to higher 2017 approval of Trump and warmer
feelings toward Republicans. Those whose ideological identification grows more
conservative between 2011 and 2016 also report similar Trump support and GOP affect in
2017. However, baseline conservatism in 2011 seems to be a stronger predictor of future
Trump support than of future GOP affect.
For both Trump and Republicans, black respondents express less approval in
2017. The same is true of non-religious respondents. The only other notable
difference in predictors of Trump and Republicans is political interest. Higher levels
of reported political interest in 2011 correspond to higher Trump approval in 2017,
but lower Republican approval in 2017.
7
Figure 1. Black Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
To briefly focus on the role of change in affect toward Blacks, predicted values of the
two dependent variables were calculated, holding all other variables at their means or modes.
Figure 2 presents these results. The difference between the two sub-figures is notable. While
rising feelings of warmth toward Blacks generate distinctly negative effects on Trump 2017
approval, the effect on Republican approval is minimal. Trump’s approval is distinctly related
to changing attitudes toward Black Americans. By examining these effects in the panel, we
can identify within-subject changes over time. Trump’s approval in 2017 is highest among
those who grew to dislike Blacks in the years between 2011 and 2016.
Figure 2. Predicted Trump and GOP Approval by Change in Affect Toward Blacks.
Those respondents whose feelings toward Blacks grew much colder between 2011 and 2016
reported approval of Trump in 2017 of around 50 percent. Those whose feelings toward
Blacks grew much warmer report Trump approval in 2017 of about 30 percent. This
significant change is all the more notable because it holds baseline 2011 affect toward Blacks
constant. The effect is only from changes in affect toward Blacks over time. In comparison,
these changes have no effect on feelings toward the Republican Party.
Hispanics
In Figure 3, identical models to those presented in Figure 1 are run, but affect toward
Blacks is replaced by affect toward Hispanics. All other variables are unchanged. In a similar
finding, affect toward Hispanics and change in affect toward Hispanics is related to Trump
support and Republican support in opposite directions. Those whose feelings toward
Hispanics were warm in 2011, and grew warmer between 2011 and 2016 expressed less
approval of Trump in 2017. The same feelings toward Hispanics led to warmer feelings
toward Republicans in 2017. Even when controlling for self-identified party and ideology,
affection toward Hispanics generates disapproval of Trump and approval of Republicans. In
these models, a post-graduate education and black race reduces future Trump support, but
only a post-graduate education reduces Republican affect. The same relationship with political
interest as found in Figure 1 is replicated here. Those with higher interest in politics in 2011
ended up being more supportive of Trump and less supportive of Republicans in 2017.
Figure 3. Hispanic Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
A clearer focus on the effects of feelings toward Hispanics can be demonstrated using
predicted values, holding all other variables at their means or modes. In Figure 4, a similar pattern
appears. Those who grow less warm toward Hispanics during the 2011-2016 period are
significantly more approving of Trump in 2017, and significantly less approving of the GOP in
2017, than those whose feelings of warmth toward Hispanics increase.
Figure 4. Predicted Trump and GOP Approval by Change in Affect Toward Hispanics.
Muslims
One of the more salient targets of both Trump and general Republican ire has been the
Muslim-American community. However, the panel data in Figure 5 demonstrate that feelings
toward Muslims are significantly more powerful in predicting Trump support than Republican
support.
Figure 5. Muslim Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
In Figure 5,
baseline 2011 affect toward Muslims has a large and significant effect on Trump support in
2017. The difference between holding the coldest and the warmest feelings toward Muslims in
2011 is associated with a 31 percentage point decrease in 2017 support for Trump. As feelings
change between 2011 and 2016, moving from the most negative to the most positive feelings
toward Muslims further decreases support for Trump by another 28 percentage points. These
effects are three times as large as the effects associated with affect toward Blacks and
Hispanics.
However, when predicting 2017 feelings toward Republicans, feelings toward Muslims
are far less consequential. Baseline feelings toward Muslims in 2011 do not significantly
predict feelings toward Republicans. Increasing feelings of warmth toward Muslims between
2011 and 2016 do predict a decrease in warmth toward Republicans, but to a much smaller
degree (6 percentage point decline) than is seen with regard to Trump approval.
To more directly observe the effect of these feelings toward Muslims, Figure 6 presents
the predicted values of Trump and Republican support in 2017. The differences are
dramatically visible.
Figure 6. Predicted Trump and GOP Approval by Change in Affect Toward Muslims.
For those whose feelings toward Muslims moved from very warm to very cold between 2011
and 2016, approval of Trump in 2017 is predicted to be around 70 percent. The same change
in feelings toward Muslims generates a predicted feeling of warmth toward Republicans of
only about 50 degrees (neutral). On the other end of the spectrum, for those whose feelings
toward Muslims move from very negative to very positive, approval of Trump in 2017 is
predicted to be around 15 percent, while feelings toward Republicans are predicted to be
around 35 degrees (holding 2011 feelings toward Muslims constant). Feelings toward
Muslims as a group seem to be uniquely related to evaluations of Trump, rather than the
Republican Party as a whole.
Gay and Lesbian People
In Figure 7, baseline and changing affect toward gay and lesbian people is used to
predict Trump approval and GOP affect in 2017. The difference between very cold and very
warm affect toward LG people in 2011 is about a 21 percentage point reduction in Trump
approval in 2017. So affect toward LG individuals before Trump became a political figure is a
significant predictor of future Trump approval. Furthermore, as attitudes toward gays and
lesbians grow warmer during the 5 year period from 2011 to 2016, approval of Trump in 2017
erodes further, with the maximum shift from coldest to warmest predicting a 13 percentage
point drop in approval.
However, the same effects are not visible in predicting feelings toward Republicans.
While warm 2011 feelings toward LG people slightly reduces feelings of warmth toward the
Republican Party in 2017, any change that occurs within individuals over time does nothing to
change attitudes toward Republicans.
Figure 7. LGBT Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
A closer look at the effect
of feelings toward gay and lesbian people on approval of Trump and the GOP can be found in
Figure 8, which presents predicted values of 2017 Trump and GOP approval.
Figure 8. Predicted Trump and GOP Approval by Change in Affect Toward Gay and
Lesbian People.
According to Figure 8, respondents whose feelings toward LG people grew much colder
between 2011 and 2016, approval of Trump in 2017 is about 55 percent. For those whose
feelings toward gays and lesbians grew much warmer, approval of Trump is around 30
percent. In comparison, there is no significant difference in GOP affect whether feelings
toward LG people grew warmer or colder (holding 2011 affect toward this group constant).
Jewish People
Affect toward Jewish people is an interesting and somewhat unclear indicator of
partisan preference. While the Trump campaign and presidency has not denounced white
supremacist (and therefore anti-Semitic) groups who support them, the Republican Party is
also very supportive of the existence of the state of Israel, drawing support from pro-Israel
American Jewish people. At the same time, Jewish voters have been predominantly
Democrats for decades. The panel data, then present a useful look into the role of affect
toward Jewish people and Trump or Republican support.
Figure 9 presents the results of the OLS model. Feelings of warmth toward Jewish
people in 2011 are marginally predictive of Trump approval, with warmer affect for Jews in
2011 predicting higher support for Trump in 2017. This effect is small, with a coefficient of
0.06. Baseline support for Jews is a better predictor of GOP affect, with a coefficient of 0.17,
nearly three times the size of the effect for Trump. Furthermore, those respondents whose
feelings toward Jews grow warmer between 2011 and 2016 appear to feel even higher levels
of warmth toward the Republican Party, but this change has no effect on levels of support for
Trump.
Figure 9. Jewish Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
These results suggest that
support for Jewish people is decidedly mixed in predicting support for Trump and
Republicans2.
In order to examine the effects of changing affect alone, Figure 10 presents the
predicted values of support for Trump and the GOP at varying levels of changing affect
toward Jewish People.
Figure 10. Predicted Trump and GOP Approval by Change in Affect Toward Jewish
People.
2 We also note that in analyses not presented here, baseline a&ect and
changing a&ect toward Jewish people signi(cantly predicts support for
Democrats as well. Warm and warming feelings toward Jewish people
predicts warmer a&ect toward the Democratic Party in 2017.
In this figure, holding constant 2011 affect toward Jewish people, the change in affect alone
between 2011 and 2016 has no effect on Trump approval in 2017. For Republicans, however,
even holding baseline affect constant, the effect of increasing feelings of warmth toward
Jewish people are to move Republican feelings from 30 degrees to 55 degrees at the most
extreme levels of change. Of course, most of the variance in the affective change variable
occurs between -0.5 and 0.5. At these levels, feelings toward the GOP would increase from 35
degrees to 47 degrees, still a significant change.
Whites
While the prior groups examined tend to be associated with the Democratic Party, the
final two groups are those that tend to be associated with the Republican Party. These are
Whites and Christians.
We begin by examining changes in attitudes toward Whites in Figure 11. Interestingly,
attitudes toward White people seems to be more powerful at predicting feelings toward
Republicans than toward Trump, although both are significant and positive effects. For
Trump, 2011 feelings toward Whites predict 17 percentage points higher approval in 2017.
The maximum increase in affect toward Whites between 2011 and 2016 predicts a further 14
percentage point increase in approval of Trump in 2017. However, for Republican Party affect
in 2017, the warmest (2011) feelings toward Whites increase affect toward Republicans in
2017 by about 37 degrees. As these feelings toward Whites grow warmer by 2016, feelings
toward Republicans grow even warmer by another 25 degrees3.
3 In models not shown here, feelings toward Whites do not have any
signi(cant e&ect on Democratic Party a&ect in 2017.
Figure 11. White Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity)
These effects of White affect are nearly twice as powerful in predicting feelings
toward Republicans than toward Trump himself.
To demonstrate the isolated effects of change in affect toward Whites, figure 12
presents predicted values of Trump and GOP support based on change in White affect.
Figure 12. Predicted Trump and GOP Approval by Change in Affect Toward Whites.
The results presented in Figure 12 make clear that the effect of changing feelings about
White people are more stark in regards to the Republican Party. Holding 2011 affect constant,
the most negative change in feelings toward Whites predicts Trump approval of about 30
percent, and Republican feelings of about 20 degrees. However, the most dramatic increase in
warm feelings toward Whites predicts Trump approval of about 55 percent, and feelings
toward Republicans of about 70 degrees. It appears, then, that Trump’s unique attraction is
animosity toward partisan racial outgroups, more than support for the partisan racial ingroup.
Christians
The final group examined here is Christians. To the extent that Christianity (at least among
Whites) is generally associated with the Republican Party, we should expect warmer feelings
toward Christians to be associated with more support of Trump and the Republican Party. In
Figure 13, this is what we find. The warmest baseline feelings toward Christians in 2011
predict an increase in Trump 2017 approval by 29 percentage points, and an increase in
warmth toward Republicans by 37 degrees. Similarly, the maximum increase in feelings of
warmth toward Christians between 2011 and 2016 predicts an increase in 2017 Trump
approval by 20 percentage points, and an increase in GOP affect by 26 degrees. Though the
results are slightly stronger for feelings toward the Republican Party, both Trump and
Republicans are assisted by warm and warming feelings toward Christians in the years before
2017.
Figure 13. Christian Affect and Trump vs. GOP Approval (zoom in on electronic version for clarity
In order to examine the effect of changing affect toward Christians, predicted values of Trump
and Republican support are presented in Figure 14, holding all other variables in the model
constant.
Figure 12. Predicted Trump and GOP Approval by Change in Affect Toward Christians
Even holding constant 2011 baseline affect toward Christians, the results in Figure 12
demonstrate that when affect toward Christians grows warmer, feelings toward Trump and
Republicans improve as well. Those whose feelings grew extremely negative toward Christians
were predicted to approve of Trump in 2017 at about 20 percent of the full range of approval,
while feelings toward the Republican Party were predicted to be around 20 degrees. However,
the most positive changes in affect toward Christians predict Trump approval above 60
percentage points and Republican affect around 70 degrees of warmth. As feelings toward
Christians grow warmer, respondents grow substantially more positively inclined toward both
Trump and the Republican Party.
Discussion
It is important to note that in models not shown here, feelings toward these groups
largely predict Democratic Party affect as well. These results are positive and significant for
changes in affect toward Blacks, Hispanics, Jewish people, Muslims, and lesbian and gay
people. However, changing feelings toward Whites and Christians do not predict Democratic
Party affect.
Overall, then, these analyses offer substantial evidence that changes in affect toward
politically-relevant social groups between 2011 and 2016 were consequential for Trump
Support in 2017, even after accounting for key inter-personal differences (e.g., in party
identification, ideology, and political interest) in 2011. Citizens who, for example, developed
greater warmth for groups that associate with the Democratic Party (e.g., Muslims and LG
individuals) tended to view Trump less favorably. However, the analyses also suggest that not
all groups are of equal consequence for shaping affective polarization—while change in affect
toward Christians was strongly associated with Trump support, for example, change in affect
toward Jewish people was not significantly predictive of Trump support. Nevertheless, these
data and results provide considerable support for the argument that affect toward social groups
shapes how people view each of the political parties in the United States, and in particular
Donald Trump, whose candidacy and presidency brought these intergroup conflicts into the
open.
Conclusion
“The desire for a strong leader who can identify domestic enemies
and who promises to do something about them without worrying over
much about legalities those germs, mutated to fit the particular local
subcultures, are latent in every democratic electorate, waiting for
sufficiently widespread human suffering to provide conditions for the
explosive spread.”
Achen and Bartels (2016)
In two sets of data, we have demonstrated that feelings toward social groups can affect
feelings about parties and candidates, and that these effects are not identical for Trump
himself and the Republican Party as a whole. We consider these effects of social affect to be
normatively worrisome for a few reasons.
First, the influence of social group affect on partisan and presidential affect, indicates a
type of partisan competition rooted in racial and religious differences. Achen and Bartels
(2016) have demonstrated that the “folk theory” of democracy, in which parties compete for
voters based on rational policy positions, is a far cry from reality. Indeed, they argue that
American democracy is strongly, if not mainly, influenced by social group loyalties. As racial
and religious animosities increasingly generate partisan and candidate animosity, we are
overlaying partisan contests with a national legacy of racial and religious conflict. These types
of social conflicts are not known for their sensible compromise and measured debate. The
more these passionate and unruly battles are allowed to characterize democratic governance,
the less democratic American government will be.
Second, as Achen and Bartels (2016) state above, when partisanship is tightly wound
with other identities, it becomes easier to identify “domestic enemies.” The spread of
animosity between social and partisan identities means that more social groups in society can
be seen as adversaries, rather than as fellow citizens. These divisions, once made and
reinforced, have the power to undermine the concept of one cohesive public, for which
democracy is supposed to work. As partisan affective polarization deepens, it has the potential
to take on a life of its own, dividing ever larger numbers of Americans against each other in
the name of partisan victory.
Finally, the differences between predictors of Trump support and Republican Party
approval are worrisome. Animosity toward Republican social outgroups seems to be more
powerful at driving support for Trump than support for the Republican Party itself. It appears
that while support for Republicans is powerfully driven by increased feelings of warmth for
partisan ingroup members, support for Trump is driven by increasingly cold feelings toward
partisan outgroup members. These results suggest that Trump himself is driving a more
conflict-oriented type of political interaction. Even before Trump was a political figure, we
can locate his future supporters among those who dislike non-White and non-Christian
Americans.
References:
Achen, C. H., and Larry M. Bartels. 2016. Democracy for realists: Why elections do not produce
responsive government. Princeton University Press.
Ahler, Douglas J., and Gaurav Sood. 2018. “The Parties in Our Heads: Misperceptions about Party
Composition and Their Consequences.” The Journal of Politics 80 (3): 964–81.
https://doi.org/10.1086/697253.
Berelson, Paul F. Lazarsfeld, William N. McPhee Bernard R. 1948. Voting. Chicago The University
of Chicago Press.
Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. 1960. The
American Voter. Chicago: University Of Chicago Press.
Campbell, David E., and Robert D. Putnam. 2011. “Crashing the Tea Party.” The New York Times,
August 16, 2011. http://www.nytimes.com/2011/08/17/opinion/crashing-the-tea-party.html.
Fiorina, Morris P., Samuel A. Abrams, and Jeremy C. Pope. 2008. “Polarization in the American
Public: Misconceptions and Misreadings.” The Journal of Politics 70 (02): 556–560.
https://doi.org/10.1017/S002238160808050X.
Giles, Micheal W., and Kaenan Hertz. 1994. “Racial Threat and Partisan Identification.” The
American Political Science Review 88 (2): 317–26. https://doi.org/10.2307/2944706.
Green, Donald, Professor Bradley Palmquist, and Professor Eric Schickler. 2004. Partisan Hearts
and Minds. New Haven, Conn.; London: Yale University Press.
Huddy, Leonie, Lilliana Mason, and Lene Aarøe. 2015. “Expressive Partisanship: Campaign
Involvement, Political Emotion, and Partisan Identity.” American Political Science Review
109 (01): 1–17. https://doi.org/10.1017/S0003055414000604.
Iyengar, Shanto, Gaurav Sood, and Yphtach Lelkes. 2012. “Affect, Not Ideology A Social Identity
Perspective on Polarization.” Public Opinion Quarterly 76 (3): 405–31.
https://doi.org/10.1093/poq/nfs038.
Iyengar, Shanto, and Sean J. Westwood. 2015. “Fear and Loathing across Party Lines: New
Evidence on Group Polarization.” American Journal of Political Science 59 (3): 690–707.
https://doi.org/10.1111/ajps.12152.
Klar, Samara, Yanna Krupnikov, and John Barry Ryan. 2018. “Affective Polarization or Partisan
Disdain? Untangling a Dislike for the Opposing Party from a Dislike of Partisanship.” Public
Opinion Quarterly 82 (2): 379–90. https://doi.org/10.1093/poq/nfy014.
Layman, Geoffrey. 2001. The Great Divide: Religious and Cultural Conflict in American Party
Politics. Columbia University Press.
Levendusky, Matthew. 2009. The Partisan Sort: How Liberals Became Democrats and
Conservatives Became Republicans. University of Chicago Press.
Luttig, Matthew D., Christopher M. Federico, and Howard Lavine. 2017. “Supporters and
Opponents of Donald Trump Respond Differently to Racial Cues: An Experimental Analysis.”
Research & Politics 4 (4): 2053168017737411. https://doi.org/10.1177/2053168017737411.
Mangum, Maruice. 2013. “The Racial Underpinnings of Party Identification and Political
Ideology.” Social Science Quarterly 94 (5): 1222–44. https://doi.org/10.1111/ssqu.12029.
Mason, Lilliana. 2018. Uncivil Agreement: How Politics Became Our Identity. University of
Chicago Press. http://www.press.uchicago.edu/ucp/books/book/chicago/U/bo27527354.html.
Mason, Lilliana. 2018b. “Losing Common Ground: Social Sorting and Polarization.” The Forum
16 (1): 47–66. https://doi.org/10.1515/for-2018-0004.
Mason, Lilliana, and Julie Wronski. 2018. “One Tribe to Bind Them All: How Our Social Group
Attachments Strengthen Partisanship.” Political Psychology 39 (February): 257–77.
https://doi.org/10.1111/pops.12485.
Miller, Arthur H., Christopher Wlezien, and Anne Hildreth. 1991. “A Reference Group Theory of
Partisan Coalitions.” The Journal of Politics 53 (4): 1134–49.
https://doi.org/10.2307/2131871.
Schaffner, Brian F., Matthew Macwilliams, and Tatishe Nteta. 2018. “Understanding White
Polarization in the 2016 Vote for President: The Sobering Role of Racism and Sexism.”
Political Science Quarterly 133 (1): 9–34. https://doi.org/10.1002/polq.12737.
Sides, John, Michael Tesler, and Lynn Vavreck. 2017. “The 2016 U.S. Election: How Trump Lost
and Won.” Journal of Democracy 28 (2): 34–44. https://doi.org/10.1353/jod.2017.0022.
Tajfel, Henri, and John Turner. 1979. “An Integrative Theory of Intergroup Conflict.” In Intergroup
Relations: Essential Readings, edited by M. A. Hogg and D. Abrams, 94–109. Key Readings
in Social Psychology. New York, NY, US: Psychology Press.
Appendix:
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.