Partisan Hearts and Minds: Political Parties and the Social Identities of Voters / Edition 1

by Donald Green, Bradley Palmquist, Eric Schickler

Overview

In this authoritative study, three political scientists demonstrate that identification with political parties powerfully determines how citizens look at politics and cast their ballots.
"Partisan Hearts and Minds is a profound breakthrough in our understanding of partisan loyalties and makes a major contribution to the study of political attitudes and voting behavior."—Paul Abramson, Michigan State University
"This book will be influential the moment it appears. It will be the starting point for all further treatments of the topic."—Richard Johnston, University of British Columbia
"The grounding of partisanship in social identities is the most important theoretical contribution to the study of partisanship in the last two decades."—Morris P. Fiorina, Stanford University

Product Details

ISBN-13:	9780300101560
Publisher:	Yale University Press
Publication date:	04/10/2004
Series:	The Institution for Social and Policy Studies
Edition description:	New Edition
Pages:	288
Product dimensions:	6.12(w) x 9.25(h) x (d)

About the Author

Donald Green is A. Whitney Griswold Professor of Political Science and director of the Institution for Social and Policy Studies at Yale University. Bradley Palmquist has taught at Harvard University and Vanderbilt University. Eric Schickler is professor of government at Harvard University.

Read an Excerpt

Partisan Hearts and Minds

Political Parties and the Social Identities of Voters

---

By Donald Green

Yale University Press

Copyright © 2004 Donald Green
All right reserved.
ISBN: 0300101562

---

Chapter One

Introduction

In the context of public opinion research, the term partisanship is something of a double entendre, calling to mind both partisan cheering at sports events and affiliation with political parties. Both meanings, as it happens, comport with what those who study elections typically have in mind when discussing partisan attitudes. Ask a sample of ordinary citizens to assess the president's integrity, policy initiatives, or performance in office, and one finds sharp disagreement between Democrats and Republicans. Indeed, Democrats and Republicans offer contrasting views not only on party leaders and their programs but also on their family, friends, pets-anything that has become emblematic of a political party.

Of course, it is hardly news that Democrats and Republicans disagree about politics. Early survey researchers noted in 1936 that 83% of Republicans believed that Franklin Delano Roosevelt's policies were leading the country down the road to dictatorship, a view shared by only 9% of Democrats (Key 1961:246). What makes partisanship interesting, and what was not apparent to researchers until the 1950s, is the fact that voters who call themselves Republicans at age thirty-two will most likely continue to do so at age eighty-two. Recessions, wars, and dramatic swings in the political fortunes of the parties tend to leave a shallow imprint on the partisan affiliations of adults, just as doctrinal and organizational disputes within Christian sects typically have little effect on the religious affiliations of churchgoers. To be sure, one generation may be more enamored of the Republican Party or the Lutheran church than the last, but the pace at which adults change their group attachments tends to be slow.

This degree of persistence is surprising because identification with political parties is a minor part of the typical American's self-conception. Race, sex, ethnicity, religion, region, and social class come immediately to mind as core social identities; political party does not. The core identities suffuse nearly all of our day-to-day interactions with others; it is difficult to imagine a social gathering in which people fail to take notice of accents, skin color, or secondary sexual characteristics. Partisan stereotypes and self-images, by contrast, are called to mind sporadically, when one glances at the newspaper or discusses politics with friends. It is not unusual for people to be unaware of the party affiliations of their friends, even when these friends' other group identities are known in minute detail. The political self is for the most part eclipsed by other selves-cultural, economic, spiritual, sexual, familial, athletic, artistic. In those instances when our attention turns to politics, however, partisan attachments become highly influential, whereas more fundamental social identities, such as sex, religion, or social class, tend to have less predictive power.

A simple example drawn from the classic Youth-Parent Socialization Panel Study (Jennings, Markus, and Niemi 1991) illustrates the point. In this survey, a national sample of 728 parents of high school students were interviewed in and reinterviewed in 1982. In 1965, respondents were asked the canonical question about their political partisanship: "Generally speaking, do you usually think of yourself as a Republican, Democrat, Independent, or what?" On the basis of this question, respondents were classified as Democrats (45.5%), Republicans (30.0%), or Independents (24.0%) or as "apolitical" (0.5%). In 1982, these respondents reported which presidential candidate they had voted for in 1980. Ronald Reagan received votes from 89.5% of those who in 1965 had labeled themselves Republicans but from just 33.8% of those who had earlier labeled themselves Democrats. Independents fell in the middle, at 65.1%. It is remarkable to think that identities formed before the rise and fall of Richard Nixon, Gerald Ford, Jimmy Carter, bell-bottoms, disco, stagflation, and gasoline shortages could so powerfully shape presidential preferences seventeen years later.

Remarkable, too, is that these sharp partisan differences eclipse corresponding sex, class, or religion effects. For example, in this survey, Reagan received 57.8% of the male and 58.6% of the female parents' votes. He garnered the support of nearly identical proportions of Protestants and Catholics (58.2% and 60.7%, respectively). A mere sixteen-point gap separated Jews (41.9%) from Protestants. Even class differences pale compared with the predictive power of party identification: 63.9% of those identifying as middle class in 1982 voted for Reagan, whereas 47.1% of those identifying as working class voted for this conservative Republican. It is notable that sex, religion, and class were weaker predictors of the vote for Ronald Reagan than was party identification measured during the Johnson administration.

Of the seemingly "fundamental" social identities, only race is a powerful predictor of electoral choice. The relationship between race and voting, however, is largely traceable to the fact that since the early 1960s, a preponderance of African Americans have identified themselves as Democrats. Seldom have black voters since that time offered much electoral support to black Republicans or Independents running against white Democrats. Tellingly, groups with less sharply defined partisan proclivities, such as Chinese Americans, are less prone to vote in distinctive ways. Our point is not that race is unimportant but rather that its influence on electoral choice is mediated largely by partisan affiliation.

The persistence and motive power of partisan identities vis-à-vis other sorts of social identities are all the more remarkable given that political party organizations have almost no detectable presence in Americans' everyday lives. In many ways, political parties in the United States have declined since the 1950s, as urban political machines, patronage jobs, party levers in voting booths, and backroom nominating procedures have waned or disappeared. As campaigns have grown more reliant on direct mail, commercial phone banks, and television, the role of party activists has receded. As the percentage of adults who work for or with parties has dropped (Putnam 2000), the level of personal contact with party activists has diminished accordingly. To be sure, the parties have access to a great deal of money. But the so-called soft money that courses through the veins of the political parties, although justified politically as a means of "party building," tends to fund candidates' advertising campaigns that, ironically, seldom bother to mention political parties (Krasno and Seltz 2000). Given these changes in institutional structure and campaign style, it is tempting to assume that partisan identities have become a thing of the past and hold little sway over voting decisions in what has increasingly become a candidate-centered, rather than party-centered, political environment. It turns out, however, that the decline of party institutions has not rendered party attachments irrelevant, nor has it diminished the influence of party attachments on vote choice. Parties are not what they used to be, but partisan groups-Democrats and Republicans-remain important objects of social identification.

This book builds a case for the continuing theoretical and political significance of party identities. Because we do not view party attachments as a product of how voters evaluate party leaders or party platforms, our view of party identification differs from that of other leading interpretations. Currently in political science the prevailing way of thinking about partisan identities emphasizes the extent to which they are shaped by rational evaluations of party platforms and performance in office. Following in an intellectual tradition founded by Anthony Downs (1957), many scholars contend that partisanship reflects a citizen's level of policy agreement with the two parties. Voters with more conservative views gravitate toward the Republican Party; liberals, to the Democratic Party. Over time, the relative attractiveness of the two parties may change, as leaders in each party maneuver left and right in an effort to garner electoral support. Another prominent argument concerns the manner in which people adjust their party identification as they assess the incumbent administration's management of domestic and foreign affairs. Morris Fiorina (1981) argues that partisanship constitutes a "running tally" of performance evaluations, as citizens accumulate information about each party's capacity to govern. Whether people focus on policies or performance in office, the characterization of partisan identification is similar. Voters, like consumers in the marketplace, form loyalties based on their evaluations of what parties deliver.

Our view, which hearkens back to earlier social-psychological perspectives on partisanship, draws a parallel between party identification and religious identification. Partisan attachments form relatively early in adulthood. To be sure, party issue positions have something to do with the attractiveness of partisan labels to young adults, much as religious doctrines have something to do with the attractiveness of religious denominations. But causality also flows in the other direction: When people feel a sense of belonging to a given social group, they absorb the doctrinal positions that the group advocates. However party and religious identifications come about, once they take root in early adulthood, they often persist. Partisan identities are enduring features of citizens' self-conceptions. They do not merely come and go with election cycles and campaign ephemera. The public's interest in party politics climbs as elections draw near, but partisan self-conceptions remain intact during peaks and lulls in party competition.

These are strong empirical claims, and all of them have been challenged at one time or another in the enormous corpus of research on party identification. Our aim is therefore to reevaluate the evidence on partisan stability and change. Much of the research on partisanship suffers from a simple but nonetheless debilitating weakness: Scholars tend to look at only one type of data. Some researchers work exclusively with panel surveys; others, with aggregate data (for example, the percentage of Democrats recorded in successive surveys). Or they tend to focus attention on one country, one time period, or just one survey. This book pulls together a wide array of survey data-aggregate, individual, U.S., comparative-dating back to the 1950s. This approach enables us to develop a synoptic account of the nature of partisanship and the sources of partisan change.

Another distinctive feature of this book is its attentiveness to methodological nuance. Many previous analysts of party identification have overlooked measurement problems in the survey data they analyzed. The standard measures of party identification, although more reliable than most survey questions, are nonetheless fallible. Respondents, interviewers, and data coders make mistakes. As a result, quantitative ratings derived from surveys do not always reflect respondents' true levels of identification with political parties. Of course, differentiating mistakes from true changes in attitude is a complicated matter, and in the chapters that follow we devote a great deal of attention to this complex but important issue. Recognizing that some readers may be reluctant to wade through the arcane technical details, we intersperse our mathematical discussions with summaries of the substantive results. Suffice it to say that standard survey assessments of party identification contain enough measurement error to distort statistical findings, and unless one makes specific allowances for measurement error, one risks overstating the malleability of party attachments.

The most important contribution of this book, however, is theoretical. Few subjects in political science have received as much attention as party identification. Yet it is very difficult to find a single work that offers a clear explanation of what it is and why it behaves as it does. Those who return to The American Voter (Campbell et al. 1960) for guidance find its discussion of party identification terse and a bit vague. Party identification is characterized as a "psychological orientation," an "enduring attachment," but there is little discussion of self-conceptions, and one must read between the lines to work out the implied analogy between party identification and class identification. Subsequent works on partisanship have generally used The American Voter as a foil; as a result, the "traditional" view of party identification has not been fully articulated and defended empirically.

IDENTIFICATION VERSUS EVALUATION

To highlight our interpretation of partisan attachments, we return to the analogy between parties and religious denominations. Like members of political parties, members of a religion or religious denomination comprise an identifiable social group, cleave to distinctive underlying doctrines, and maintain (to varying degrees) an adversarial relationship toward other religions. And like party identification, religious identification is often acquired early in life as a product of one's family environment or early adult socialization. As a member of a religion, one is indoctrinated into that religion's precepts, much as partisans learn the slogans and nostrums of their party. To be sure, some adults choose their religion by canvassing various sects' theological positions, selecting the one that best fits their personal values. But these people seem to be a minority, for among adults over thirty years of age, religious affiliation, once established, tends to remain intact. A more common avenue for shifting religious affiliation is a changing small-group environment, in particular, marriage to a person of another faith. In such instances, people may commit themselves to their new faith for reasons that have little to do with its overall doctrinal attractiveness, or they may alter their perception of the new religion as they come to see it through their spouse's eyes. Parallel observations may be made about partisan identities, which also change as regional and occupational mobility put adults into contact with new friends and social groups, some of which may have different partisan coloration (Brown 1981; Gimpel 1999).

The analogy between partisan and religious identification has an implication that many political scientists find disconcerting: Partisan change among adults often has little to do with unfolding political and economic events. Scholars who study politics like to think that politics matters.

Continues...

Continues...

---

Excerpted from Partisan Hearts and Minds by Donald Green Copyright © 2004 by Donald Green. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---

Table of Contents

	Preface	vii
1	Introduction	1
2	Partisan Groups as Objects of Identification	24
3	A Closer Look at Partisan Stability	52
4	Partisan Stability: Evidence from Aggregate Data	85
5	Partisan Stability and Voter Learning	109
6	Party Realignment in the American South	140
7	Partisan Stability outside the United States	164
8	How Partisan Attachments Structure Politics	204
	Appendix	231
	Notes	235
	References	245
	Index	255