The Moral Diversity of Polarization: Examining Intra-Party Philosophical Differences in a Polarized Political System Daniel J. Coffey 1 Abstract The polarization between American political parties is often conceptualized as unidimensional. Indeed, the liberal-conservative spectrum has strong empirical support. Since the American party system is federal in nature and the United States is culturally diverse, it is likely, however, that the observation and measurement of polarization hides both inter and intra-party variance. Recent work by social psychologists suggests that there are multiple moral foundations that capture most human judgments about morality. Social scientists have studied this question from the micro-level, but so far there are few studies of how these moral foundations might be altered by a public setting. In an exploratory study, I analyze eight state party platforms (four from each party) and find that party philosophy is articulated (grounded) in easily recognizable moral claims. I also find considerable intra-party diversity as state parties draw on different moral claims in constructing their public platforms. Importantly, the platforms indicate that parties build consensus for position by invoking multiple foundations. Since moral beliefs are both innate and acquired, studying how parties and politicians both reflect and produce the moral frames citizens use in their political evaluations can greatly illuminate the role of morality in partisan polarization. Prepared for delivery at the Midwest Political Science Conference, April 16-19, 2015, Chicago, IL 1 Fellow, Ray C. Bliss Institute of Applied Politics; Associate Professor, Department of Political Science, University of Akron. [email protected]
21
Embed
The Moral Diversity of Polarization: Examining Intra-Party Philosophical Differences in a Polarized Political System
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
The Moral Diversity of Polarization: Examining Intra-Party
Philosophical Differences in a Polarized Political System
Daniel J. Coffey1
Abstract
The polarization between American political parties is often conceptualized as unidimensional. Indeed, the liberal-conservative spectrum has strong empirical support. Since the American party system is federal in nature and the United States is culturally diverse, it is likely, however, that the observation and measurement of polarization hides both inter and intra-party variance. Recent work by social psychologists suggests that there are multiple moral foundations that capture most human judgments about morality. Social scientists have studied this question from the micro-level, but so far there are few studies of how these moral foundations might be altered by a public setting. In an exploratory study, I analyze eight state party platforms (four from each party) and find that party philosophy is articulated (grounded) in easily recognizable moral claims. I also find considerable intra-party diversity as state parties draw on different moral claims in constructing their public platforms. Importantly, the platforms indicate that parties build consensus for position by invoking multiple foundations. Since moral beliefs are both innate and acquired, studying how parties and politicians both reflect and produce the moral frames citizens use in their political evaluations can greatly illuminate the role of morality in partisan polarization.
Prepared for delivery at the Midwest Political Science Conference, April 16-19, 2015, Chicago, IL
1 Fellow, Ray C. Bliss Institute of Applied Politics; Associate Professor, Department of Political Science, University of Akron. [email protected]
There is little disagreement that American political parties are ideologically polarized.
Political science, however, has tended to treat ideology as an exogenous variable. Most studies tend
to “measure” ideology, but the degree to which ideology is distinct from party unity is unclear
(Krehibel 1998; Cox and McCubbins 1993). In sum, however, political scientists have found that a
unidimensional, liberal-conservative conceptualization of ideology to be empirically useful
(Abramowtiz 2010; Poole, McCarty and Rosenthal 2006; Jost, Nosek and Gosling 2008).
There is considerable research in social psychology that the underlying reasons for the
attraction to these polar ends is rooted in both biological and psychological factors (Hibbing 2013).
These explanations are perhaps less philosophical and more empirically driven than previous claims
about the nature of human morality. Considerable research suggests that the liberal-conservative
polarization most likely masks a multidimensional foundation, whether this is rooted in genetics,
culture, personality or socio-economics, or some combination of these factors (Carney et al 2009;
Mondack; 2010; Hetherington and Weiler 2009).
In recent years, social psychologists have proposed a “moral foundations” theory to describe
fundamental human dimensions of morality (Graham, Haidt, and Nosek 2009). Haidt 2012) has
shown that differences in party ideology are partially rooted in moral intuitions. Empirical work
testing moral foundations theory provides compelling evidence that five “taste buds” make up the
foundations for nearly all moral beliefs. These foundations are 1) care/harm, 2) fairness/cheating 3)
loyalty/betrayal, 4) authority/subversion and 5) sanctity/degradation.2 Use of the moral foundations,
theory, therefore, could help detect different types of intra-party ideological variance, such as
libertarian vs. religiously conservative Republicans or moderate “Blue Dog” vs. progressive
Democrats.
2 Haidt provides some evidence that there may in fact be six foundations, including a liberty/freedom foundation.
Moral foundations theory has had considerable success in explaining differences between
Democrats and Republicans in their policy preferences. In general, extensive survey research has
consistently demonstrated that liberals generally tend to endorse only two dimensions, while
conservatives tend to weigh a broader spectrum of considerations in their moral judgments
(Graham, Haidt, and Nosek 2009, Haidt 2012; Clifford and Jerit. 2013).
There is some debate whether there is a separate sixth dimension of liberty. Initially, Haidt
and his colleagues identified five foundations. Recent work has shown that the second foundation,
equality and reciprocity, taps into different considerations for Republicans and Democrats (Iyer,
Koleva, Graham, Ditto, and Haidt 2012). Libertarianism tends to be orthogonal to the liberal-
conservative divide; libertarians are less empathetic, but much more sensitive to violations of
procedural fairness, and yet often have attitudes streaked with anti-authoritarianism. The Tea Party
movement tends to focus more on restrictions on the power and authority of government and
places value on individual freedom from government. Subsequent research by Haidt and his
colleagues provides empirical support there are probably more than five dimensions and the
strongest candidate for inclusion would be the “liberty/oppression” foundation (Haidt 2013).
None of this tells us much about the variance properties of moral foundations theory. In this
paper, I examine state party platforms in order to uncover the moral principles underpinning party
issue positions. In previous studies, I have demonstrated that state party platforms exhibit a high
degree of within and between-party variance in terms of ideology and issue positons (Coffey 2007;
2011; 2014). There is good reason, to believe then, that cultural or contextual factors will influence
party morality due to the federal nature of the American party system and cultural diversity in the
United States. The review of state party platforms demonstrates that nearly all party positions are
grounded in moral terms that can be categorized into the dimensions identified by moral
foundations theory. Republicans and Democrats rely on different considerations as predicted by the
theory, such that Republican platforms endorse the full range of moral foundations, while
Democrats tend to restrict morality to the care/harm and fairness/equality foundations.
This research also explores the public nature of the platforms. Platforms are often written by
committees and are intended to foster intra-party unity. As a result, they represent some degree of
“reflective equilibrium” in that parties, seeking to build consensus, craft positions to gain the
support of factions within the organization (Noel 2013). While sometimes state party platform
committees include planks to simply appease a particular group, often party platform planks and the
language articulating the policy position is designed to appeal to the moral beliefs of a diverse group
of supporters. As such, state party platforms often endorse multiple foundations within each issue or
policy position. Studying how moral language is used by parties attempting to appeal to multiple
groups within an electorate provides a new avenue for understanding the intersection of politics and
morality. This study provides new insight into how party moral beliefs are developed, how they
change over time and how parties win support among different groups within the electorate.
Morality by Committee
While moral foundations theory has solid empirical grounding, the research in this field has largely
been focused on distilling the separate nature of each foundation. Indeed, the research largely been
survey-driven. In this sense, moral foundations can be used to accurately describe the innate
morality produced through evolution, as well as the acceptance of moral principles as they are
affected by social or cultural context. In sum, moral foundation theory has so far sought to explain
individual variance in moral beliefs (Graham, Haidt, and Nosek 2009, Haidt 2012).
Parties, by contrast, are composed of groups of individuals who may share basic personality
dispositions and moral beliefs, but who often disagree about policy or the exact rationale for
pursuing different policies. Noel’s (2013) recent work on political parties and ideology demonstrates
that ideology and partisanship can overlap substantially, but ideology is a fundamentally different
concept than the party which pursues that ideology. As such, party positions are more dynamic and
philosophically inconsistent than much recent work on the psychological roots of ideology would
predict.
Political science has shown that individual attitudes are highly susceptible to elite discourse
(Zaller 1992; Druckman, Peterson and Slothuus 2013) and interpersonal discussion (Mutz 2006 ).
Studies of priming and framing consistently show that individual preferences are highly affected by
context (Iyengar and Kinder 1982). Priming and framing effects can be limited when citizens are
exposed to competing frames and different considerations;. (Chong and Druckman 2013).
As a result, we have only broad predictions about how moral principles will be shaped
through public debate. Partisan expressions of political morality may be different than what can be
detected in a survey because the process for generating the policy position consists of bargaining and
reasoning among factions within the party. Political argumentation is a form of public is reasoning
to both persuade others, but often to change what the argument is about (Noel 2014). Political
language requires making appeals to others that will be found convincing (Hart, Childers and Lind
2013). Indeed, recent evidence shows that the appeal of moral foundations shifts as respondents are
exposed to different moral frames (Day, Fiske, Downing, and Trail 2014).
Party Platforms and Party Morality
A unique way to test these claims is by focusing on state party platforms. These are ideal for
several reasons. First, they are “official” pronouncements of party policy and this elite nature .
Second, parties are able to articulate not just their policies but the moral justification of their policy
positions. Third, individual states craft platforms so that internal variance, especially regional
variance can be detected. e public nature of platforms and their committee approach to writing leads
to multiple issues being addressed.
Not all states write platforms. Of the 100 state parties, about two-thirds write platforms
every four years; some issue platforms every two years while others issue platforms sporadically. The
platforms are usually fairly lengthy; they are often comparable in length to gubernatorial state of the
state speeches ad so they tend to provide rich descriptions of party beliefs and policy positions.
For several reasons we should expect heterogeneity in party moral beliefs. Social
psychologists have noted that while many are universal, culture can powerfully shape basic beliefs
about right and wrong. In the American context, it has long been noted that the states are culturally
quite distinct, even if there remains disagreement about how this can be empirically measured. At
the least, compositional differences are substantial across states (Gimpel and Sckekek 2004 Gimpel
and Schuknecht 2003; Stonecash, Brewer, and Mariani 2003); Talhelm et al 2013). States in the U.S.
have differed dramatically in their political development. Daniel Elazar’s (1968) trichotmomy of
political cultures (moralist, individualist, and traditionalist), continues to predict a range of state
policy differences, while other scholars similar typologies Lieske (2012; Woodward (2012). In fact, a
recent study found that basic personality differences exist across states (Rentfrow et al 2013). In
sum, we should expect some degree of diversity in terms of the acceptance of moral foundations
across U.S. states, both between and within parties.
Data and Methods
To assess party issue agendas, I performed a manual content analysis of eight state Democratic and
Republican platforms, four from each party written in 2008. The platforms were coded into six
moral foundations (including one for liberty) and into 25 issue categories by sentence. 1331
sentences were coded. The states selected were Maine, West Virginia, Idaho and California. The
platforms were selected for regional diversity and because they were comparable in terms of length
and ideology (see appendix 1). The four states selected for regional diversity and to make sure that
the platforms were of roughly equal length, as platform content and format can often vary
substantially from state to state. The four state platforms contain nearly equal numbers of
sentences.3
The platforms were coded manually with the assistance for additional coders for reliability.
Computer-assisted-text-analysis (CATA) programs have numerous advantages over human coding
(Grimmer and Stewart 2013 Young and Soroka 2011; Grimmer 2010; Weinberg 2010; Hart, Childers
and Lind 2013). At the same time, for exploratory work, human coding can preserve semantic
validity while definitions and coding rules are established. These can later be used as a baseline for
CATA methods (Lowe and Beniot and 2013).4
Few sentences in the state party platforms are descriptive; almost all sentences contain some
moral or prescriptive content. For example, the statement in the 2008 Idaho Republican platform
“We recognize the increasing role of technology in society” is descriptive and was not coded into
any of the moral categories. Later, however, the party states, “We believe parents have the right to
educate their children at home , that court rulings that undermine these rights devalue the important
role of parents and family in a child’s education.”
Few sources really allow a researcher to put the state party on the psychological couch, so to
speak. In the platforms, state parties expend considerable space to articulate exactly what they
believe in detailed terms both in policy specifics and in terms of the moral underpinnings of their
beliefs. The platforms represent the formal positions of the state organizations, whether they are put
together by the state party central committee or written by activists at the precinct level. Certainly,
3 A larger analysis is currently being conducted for all platforms written between 2008 and 2012 with over 20,000 individual sentences were coded, providing a detailed window into the priorities of state party leaders and activists. 4 A dictionary of moral terms is currently being tested for all platforms written in 2008 and 2012.This dataset contains only 400,000 words and is ideal for using CATA.
state platforms cannot be said to represent the single view of an entire state party. Indeed, part of
the value of studying these platforms is that they are windows into not only the core values of the
parties, but the conflicts and tensions that exist within and across state parties.
The platforms address numerous issues. Of the 25 categories, every single one was
addressed. Consequently, the platforms are representative of the full range of debate in American
politics. I have previously shown that state party platforms written from 2000 to 2004 are highly
polarized (Coffey 2007; 2011; 2014). Evidence of this polarization, along with intra-party ideological
heterogeneity is provided in Appendix II and III.
Table 1: Primary Moral Foundation of State Party Platform Sentences