Racial Stasis: The Millennial Generation and the Stagnation of Racial Attitudes in American Politics 9780226643762

Racial progress in the United States has hit a wall, and the rise of white nationalism is but one manifestation of this.

237 94 5MB

English Pages 304 [336] Year 2020

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Racial Stasis: The Millennial Generation and the Stagnation of Racial Attitudes in American Politics
 9780226643762

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Racial Stasis

Racial Stasis The Millennial Generation and the Stagnation of Racial Attitudes in American Politics C h r i st o p h e r D . D e S a n t e a n d C a n d i s Watts S m i t h

The University of Chicago Press Chicago and London

The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London

© 2020 by The University of Chicago All rights reserved. No part of this book may be used or reproduced in any manner whatsoever without written permission, except in the case of brief quotations in critical articles and reviews. For more information, contact the University of Chicago Press, 1427 E. 60th St., Chicago, IL 60637. Published 2020 Printed in the United States of America 29 28 27 26 25 24 23 22 21 20   1 2 3 4 5 ISBN-­13: 978-­0-­226-­64359-­5 (cloth) ISBN-­13: 978-­0-­226-­64362-­5 (paper) ISBN-­13: 978-­0-­226-­64376-­2 (e-­book) DOI: https://doi.org/10.7208/chicago/9780226643762.001.0001 Library of Congress Cataloging-in-Publication Data Names: DeSante, Christopher D., author. | Smith, Candis Watts, author. Title: Racial stasis : the millennial generation and the stagnation of racial attitudes in American politics / Christopher D. DeSante and Candis Watts Smith. Description: Chicago : University of Chicago Press, 2020. | Includes bibliographical references and index. Identifiers: LCCN 2019025423 | ISBN 9780226643595 (cloth) | ISBN 9780226643625 (paperback) | ISBN 9780226643762 (ebook) Subjects: LCSH: Racism—United States. | Discrimination—United States. | United States—Ethnic relations. Classification: LCC E185.615 .D475 2019 | DDC 305.800973—dc23 LC record available at https://lccn.loc.gov/2019025423 ♾ This paper meets the requirements of ANSI/NISO Z39.48-­1992 (Permanence of Paper).

For Melissa Spas and Terrell Smith

C o n t e n ts

List of Tables / ix List of Figures / xi Preface / xiii

Not All Change Is Progress: An Introduction / 1 P a r t I . L ay o f t h e L a n d

One

/ Nature of the Game: The Racial Stasis Hypothesis / 21 T wo

Three

/ Is Race Special? / 39

/ New Attitudes or Old Measures? / 71

P a r t I I . C o u n t e r v a ili n g F o r c e s

f o ur

five

/ Millennials on Racism / 101

/ Racialized Policy Preferences / 143

P a r t I I I . A H o li s t ic M e a s u r e

six

/ New Attitudes, New Measures / 177

seven

/ The Structure, Nature, and Role of  Twenty-­First-­Century Racial Attitudes / 195 eight

/ The FIRE This Time / 225

C o n clus i o n :

Is Resuscitation Possible? / 247

Appendix A. Everything You Need to Know about the APC Intrinsic Estimator / 261 Appendix B. A Brief Note on Factor Analysis / 269 Appendix C. Interview Schedule and Respondent Demographics / 275 Appendix D. Supplemental APC-­IE Tables / 285 Notes / 293 Works Cited / 297 Index / 315

Tabl e s

Table 2.1.

APC-­IE: White-­Black SES Gap Is Due to “Inborn Disability” / 52

Table 3.1.

Cross-­Generational Differences in Racial Resentment / 79

Table 3.2. Intrinsic Estimates of Age-­Period-­Cohort Effects on Racial Resentment / 81 Table 3.3.

Test of Measurement Invariance across Age Groups (2012 ANES) / 85

Table 3.4. Racial Resentment, Generational Status, and Stereotypes of Blacks as Violent / 90 Table 3.5.

Racial Resentment, Generational Status, and Stereotypes of Blacks as Lazy / 90

Table 3.6. Racial Resentment, Generational Status, and Attitudes toward Racialized Targets / 93 Table 4.1.

Racial Background of Friend Groups / 127

Table 5.1.

Relationship between Policy Interpretation and Level of Support / 163

Table 6.1.

CoBRAS Racial Privilege / 188

Table 6.2.

CoBRAS Institutional Discrimination / 188

Table 6.3.

CoBRAS Blatant Racism / 190

Table 6.4.

PCRW Racial Empathy / 191

Table 6.5.

PCRW White Guilt / 192

Table 6.6.

PCRW Fear of Other Groups / 192

Table 6.7. Explicit Racial Resentment across Generations / 193 Table 7.1.

Predicted Subscales for CoBRAS and PCRW / 197

Table 7.2.

Confirmatory Factor Analysis of CoBRAS / 197

Table 7.3.

Confirmatory Factor Analysis of PCRW / 199

Table 7.4.

Confirmatory Factor Analysis of Racial Resentment / 200

Table 7.5.

Confirmatory Factor Analysis of Conservative Ideology / 201

x / Tables Table 7.6.

Predicting Old-­Fashioned Racism: Whites as More Intelligent / 205

Table 7.7.

Predicting Old-­Fashioned Racism: Whites as Harder Workers / 205

Table 7.8.

Predicting Old-­Fashioned Racism: Blacks as More Violent / 206

Table 7.9.

Predicting Old-­Fashioned Racism: Whites as More Trustworthy / 206

Table 7.10. Predicting Support for Racialized Targets / 208 Table 7.11. Correlations between Independently Estimated Dimensions / 212 Table 7.12. Higher Order Factor Loadings / 213 Table 7.13. Old-­Fashioned Racism Predicted by Two Dimensions of Racial Attitudes / 217 Table 7.14. Old-­Fashioned Racism Predicted by Two Dimensions, by Generational Status / 218 Table 7.15. Predicting Nonracialized Policy Positions by Generational Status / 219 Table 7.16. Predicting Racialized Policy Positions by Generational Status / 220 Table 8.1.

FIRE by Generational Status / 229

Table 8.2. Predicting Opposition to a Close Family Member Marrying Someone Who Is Not White / 234 Table A.1. APC Table of Biological Racism (GSS) / 265 Table B.1.

Miles between Various American Cities (Truncated) / 270

Table C.1. Interview Respondent Characteristics / 280 Table D.1. APC-­IE: Feelings regarding the Women’s Liberation Movement / 285 Table D.2. APC-­IE: Affective Ratings of Gays and Lesbians / 286 Table D.3. APC-­IE: Pro-­White Affect / 286 Table D.4. APC-­IE: Equalitarianism / 287 Table D.5. APC-­IE: Support for Men and Women Having Equal Roles (0/1) / 287 Table D.6. APC-­IE: Support Workplace Protections for Gays and Lesbians (0/1) / 288 Table D.7. APC-­IE: Support Workplace Protections for Blacks (0/1) / 288 Table D.8. APC-­IE: Support Women in the Workplace (0/1) / 289 Table D.9. APC-­IE: Support Affirmative Action for Blacks (0/1) / 289 Table D.10. APC-­IE: Support Affirmative Action for Blacks (0/1) / 290 Table D.11. APC-­IE: Support Affirmative Action for Women (0/1) / 290 Table D.12. APC-­IE: Support Ban on Interracial Marriage (0/1) / 291 Table D.13. APC-­IE: Oppose Relative Marrying Member of Another Race (0/1) / 291 Table D.14. APC-­IE: Agree Government Should Help Blacks (0/1) / 292

F i gur e s

I.1.

Racial Resentment (1986–­2016) / 3

I.2.

Racial and Ethnic Demographics of  Four American Generations, 2014 / 5

1.1. The Evolution of Racial Attitudes in the United States / 30 2.1. How Millennials and Older Whites Prioritize Racial Issues / 41 2.2. Proportion of  White Respondents Who Believe Racial Disparities Can Be Attributed to Black “Inborn Disability” / 51 2.3. Illustrated Age, Period, and Cohort Effects from Table 2.1 / 53 2.4. Age Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women / 56 2.5. Period Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women / 57 2.6. Cohort Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women / 59 2.7. Age, Period, and Cohort Effects on Attitudes toward Abstract Egalitarianism / 60 2.8. Age, Period, and Cohort Effects on Attitudes toward Specific Applications of Egalitarian Principles / 62 2.9. Age Effects on “Easy” and “Hard” Applications of Egalitarianism / 64 2.10. Period Effects on “Easy” and “Hard” Applications of Egalitarianism / 66 2.11. Cohort Effects on “Easy” and “Hard” Applications of Egalitarianism / 67 2.12. Differences in Issue Importance across Generations / 68 3.1. Trends in Racial Resentment / 80 3.2. Graphical Display of Age, Period, and Cohort Effects from Table 3.2 / 82 3.3. Measurement Model for Racism / 85 3.4. Visualization of Regression Estimates from Tables 3.4 and 3.5 / 92

xii / Figures

3.5. Visualization of Regression Estimates from Table 3.6 / 94 3.6. White Americans’ Perceptions of Racial Discrimination, by Generation / 97 4.1. Millennials Living in “Diverse” Neighborhoods / 119 4.2. What Comes to Mind When You Think of Diversity? / 121 7.1. Factor Effects on Measures of “Old-­Fashioned” Racism / 203 7.2. A Single-­Order Latent Factor Model / 210 7.3. A Second-­Order Latent Factor Model / 211 7.4. The Two Dimensions of  Whites’ Racial Attitudes / 214 7.5. The Two Dimensions of Racial Attitudes by Generation / 215 7.6. Predicting Support for Amnesty / 221 8.1. FIRE and Its Correlates / 231 8.2. Racial Attitudes and Primary Voting in 2016 / 236 8.3. Racial Attitudes, Generation, and Primary Voting in 2016 / 238 8.4. Racial Attitudes, Generation, and the 2016 General Election / 239 8.5. FIRE in the 2016 Presidential Election / 241 8.6. FIRE and Trump’s Popularity among White Obama Voters / 242 C.1. Countervailing Forces on Bumper Stickers / 248 A.1. Two Examples of Systems of Equations / 262 A.2. Trust in Government 1958–­2012 (ANES) / 266 A.3. Trust in Government by Birth Cohort (ANES) / 267 B.1. Unrotated Factor Solution / 271 B.2. Rotated Factor Solution / 272

P r e fac e

In many ways, the authors of this book represent Dr. Martin Luther King Jr.’s dream. One of the authors of this book is a White man, the other a Black woman. We went to graduate school together at Duke University, where this project first began to take flight. We each have a doctorate. We both live comfortable, upper-­middle-­class lives as college professors. Ostensibly, we are equals in an American society that promises benefits for all who work hard. As we were both born in the early 1980s, we are also millennials, members of a generation that is markedly different from those that have come before us. Millennials, ourselves included, have lived their entire existence in an America where two people like us can work together and be friends with one another without any pushback from our peers or our families. Most people in our generation were socialized to believe that all people are equal despite race or gender, to value diversity, and to appreciate multiculturalism. We were inculcated to believe that America’s deep racial divisions had been healed. We belong to a generation that was lectured about the history of American racism where everything was presented in the past tense. Indeed, what makes millennials different is that most of our experiences of overt racism primarily come secondhand, usually in the form of history lessons taught during the month of February. We have been tasked by our predecessors to put the final nail in racism’s coffin, to be color-­blind, and to help America finally reach its post-­racial goal. Without a doubt, America has made great strides in terms of racial prog­ ress. Those who seek to provide evidence that race is becoming a less divisive issue in American politics can point to the two of us, or people like us in their own lives. They can point to Oprah, the Carters (Beyoncé and Jay-­Z), and LeBron James as successful, über-­rich Black Americans. They can point to Sonia Sotomayor, Marco Rubio, and Julián Castro as signs that Latinx

xiv / Preface

people are politically incorporated. Americans today can point to Lisa Ling, Mindy Kaling, and Jackie Chan as household names while being shocked at the poor taste Hollywood executives showed when casting Whites to portray characters of color only a few decades ago. In the political realm, many Americans point to Barack Obama’s 2008 election to the presidency as the ushering in of a “post-­racial” America. Not only was there a high turnout of African Americans and Latinx people, the overwhelming majority of whom voted for Obama, younger Whites also turned out in record numbers and also gave the majority of their votes to Obama. White Americans under the age of thirty preferred Obama to Senator John McCain 54 to 44 percent—­the reverse of the aggregate White population, which gave 54 percent of its vote to McCain while only giving 44 percent of its vote to Obama (Dahl 2008; Keeter, Horowitz, and Tyson 2008). For some, this fact served and still serves as prima facie evidence not only of a significant decline in anti-­Black racial attitudes among White Americans but also that America has begun realizing its post-­racial dream (Nagourney 2008; Thernstrom and Thernstrom 2008; Tolson 2008). So why are so many Americans, especially the two of us, still talking about race? For every story that fits the post-­racial narrative, there is another that shows there is still progress to be made. Again, consider the two of us. Research shows that 75 percent of White Americans do not have a non-­White friend. In a scenario where a White American has a hundred friends, ninety-­ one of them would be White. In the same scenario for Blacks, eighty-­three of those friends would be Black (Ingraham 2014). Furthermore, of all of the doctorates awarded in the United States, only 7 percent of them were earned by Black Americans, and among the 10,595 members of the American Political Science Association, 338 of them are Black, and only 173 of those are Black women. The chances of us interacting, no less being friends, were really quite low. What’s more, Christopher began studying race in American politics because he was constantly receiving mixed messages regarding racial progress from his friends, his family, and the media. Growing up in a working-­class household in Pennsylvania, he was often around folks who were not hesitant to express what scholars now refer to as “old-­fashioned” racial animus. Friends and family alike would unapologetically characterize Blacks as lazy, unintelligent, dishonest, and prone to criminal behavior and would use the N-­word when talking about African Americans. These comments were relayed as matters of fact, rarely examined by many adults in his early life. In graduate school, when he began working on racial attitudes, many of his

Preface / xv

White classmates would ask him a question that Candis never got: “Why do you work on race?” We acknowledge that we are just a small sample, an anecdote. Looking through a wider lens, we find many more signs of inequality. Of the 540 Americans who have acquired the status of billionaire, there are only 3 who are Black: Oprah Winfrey, Robert Smith, and, most recently, Michael Jordan. While there are some very wealthy Blacks in the United States, White Americans own about ten times more wealth than Black families, and optimists calculate that if the average Black family accumulates wealth at the same rate Black families have in the previous three decades, it would take that family about 228 years to amass the same amount of wealth that the average White family has today (Asante-­Muhammad et al. 2016). Meanwhile, about 6.4 percent of non-­Hispanic White families are living in poverty, in contrast to the 20.2 percent of Black families who are in the same condition (US Census Bureau 2016). We should add that many (conservative White) Americans also rely on the “model minority” myth to deflect attention from the ongoing discrimination that Asian Americans face, and movie executives still cast White actors to play roles written for Asian and Asian American characters. We can go on and on with statistics about the disparities—­many of them growing—­between Whites and people of color, particularly Blacks and Latinx people, but, ultimately, we wrote this book because of what we noticed in our interactions with our millennial peers and also with our college-­ aged students, many of whom belong to our generational cohort. We make two central claims in this book that are controversial, and we provide a surfeit of evidence to support both. The first is that racial progress in the United States has hit a wall. When Howard Schuman and his colleagues were writing their seminal text Racial Attitudes in America (originally published in 1970), there was a great deal of optimism regarding the trajectory of racial attitudes in the United States. In contrast, we argue that progress has, at best, slowed. In some cases, we find evidence that progress has completely stalled. While overtly racist attitudes have certainly declined over the past several decades, symbolic racial attitudes have essentially flatlined during the past thirty years. Furthermore, we show that anti-­Black stereotypes, anti-­ Black affect, and racial resentment are still very prevalent among the White American population, and, building on the work of other scholars, we show that anti-­Black sentiment exerts a stronger influence on some Americans’ partisanship, policy preferences, and vote choice than we have seen in the past couple of decades (Yadon and Piston 2018).

xvi / Preface

It is said that you are either part of the problem or part of the solution. The second main claim we make in this book is that millennials are not part of the solution, and therefore they are likely part of the problem. This is not merely a jab at millennials. As scholars of political science, and racial and ethnic politics more specifically, we are steeped in an understanding of the history of racial attitudes in America, but we are also insiders of the millennial generation. In a way, we are bilingual: fluent in the language of race that millennials are speaking but conversant in the racial language of our parents and grandparents. This represents an important departure from extant political science work on race and racial attitudes. Previous work has focused on racial attitudes in the language and socialization of past generations and, as a result, continues to use the same measures to estimate “racial progress.” Ours is a different approach. We show that White millennials are so removed from Jim Crow and the civil rights era that they have little understanding of the structural nature of racial inequalities in the United States and therefore lack the contextual knowledge to be actively anti-­racist. So while White millennials may be open to the idea of interracial marriage or living next to a Latinx family, they do not understand why policies like affirmative action still need to exist. As a result, and like their predecessors, they are wary of supporting these kinds of policies. What we show is that even though White millennials’ language and rationale around race, racism, and racial inequalities are different from that of previous generations, the end result is the same. We demonstrate that we are in a state of racial stasis. We offer this argument in the hope that our readers, our peers, and our students will be persuaded that we can all be part of the solution.

Acknowledgments During the summer of 2010, the two of us began talking to each other about many of our White colleagues and peers who claimed to be anti-­racist but weren’t actually working toward dismantling a system marked by White supremacy. In fact, we could provide several examples to the contrary; one that still stands out involves the use of racial humor wrapped up in a cloak of “I’m so not racist that I can make racist jokes.” We thought about producing a project to out these kinds of shenanigans, paint a more accurate portrait of the so-­called progressive millennial, and pin down a measurement to get at the aspects of their racial attitudes that the traditional political science literature has heretofore failed to capture. What all of this means, in the context of giving thanks to those who have supported us through the process

Preface / xvii

of writing this book, is that we have been working on this project for the better part of a decade, and a whole lot of people have shared insights, feedback, words of wisdom, and encouragement along the way. Consequently, we have a whole lot of thanks to share. If we have failed to mention you by name, please charge it to our exhausted heads and not our hearts. In the years that we’ve worked on this project, we’ve presented at a half dozen conferences or more and had posts at a total of five institutions: Duke, Oberlin College, Williams College, University of North Carolina at Chapel Hill, and Indiana University. At each of these places, our colleagues (often turned friends) have been tremendously helpful. We’d like to say thank you to John Aldrich, Antoine Banks, Frank Baumgartner, Andrea Benjamin, Bill Bianco, Eduardo Bonilla-­Silva, Ted Carmines, Cassandra Davis, Vivian Ferrillo, Colin Fisk, VaNatta Ford, Matthew Fowler, Bernard Fraga, Matthew Hayes, Kerry Haynie, Margie Hershey, Vince Hutchings, Jeff Israel, Ashley Jardina, Christopher Johnston, Cindy Kam, Aaron King, Rebecca Kreitzer, Michael MacDonald, James Manigault-­Bryant, Rhon Manigault-­Bryant, Natalie Masuoka, Paula McClain, Nicole Mellow, Jennifer Merolla, Jacob Montgomery, Ngoni Munemo, Brendan Nyhan, Diana O’Brien, Efrén Pérez, Rene Rocha, Debbie Schildkraut, Eric Schmidt, Brigitte Seim, Regina Smyth, Michael Tesler, Isaac Unah, and Nick Valentino. Our gracious colleagues at Washington University in St. Louis, Penn State University, the American Politics Research Group at the University of North Carolina Department of Political Science, and the American Politics Workshop at Indiana University invited us to present our work and provided productive and constructive criticism. We appreciate you. We must also give a huge thanks to Amber Ellis, Gabby Malina, and Corey Michon—­our three wonderful undergraduate research assistants. This book has been through several iterations, and we hope this one is the best. We are so pleased that Chuck Myers supported this project. While we are honored to join the ranks of University of Chicago Press authors, we must also thank the anonymous reviewer at Cambridge who provided comments that allowed us to get the book ready for prime time, the wealth of feedback from all of the reviewers at Chicago and at Oxford University Press, and also David McBride. We appreciate the glorious editorial help of Serene Yang and the indexing talents of laurie prendergast. We would be remiss to not thank Stephen Ansolabehere and Brian Schaffner for taking a chance on our FIRE measure on the Cooperative Congressional Election Study. Some of the original data collection was funded by the National Science Foundation (NSF award 1122624) as well as a Collaborative Research and Creative Activity Fund grant from the Office of the Vice Provost for

xviii / Preface

Research at Indiana University. Most figures were made using Hadley Wickham’s (2016) ggplot2 package for R. Those near and dear to us have been willing to hear all about this project and give head nods (which go a long way), high fives, and shout-­outs. Candis would like to thank Mr. Cheeks alongside Kimberly Bickham, Rose Buckelew, David Cortez, Kim Yi Dionne, Chryl Laird, Deb LoBiondo, Tehama Lopez Bunyasi, Sarah Mayorga-­Gallo, Julian Wamble, and the amazing scholars who uplift others through @WomenAlsoKnow and @POCalsoknow. Christopher acknowledges Andy Bell, Betsy Bell, Sam Bestvater, Scott Clifford, Scott de Marchi, Kent Freeze, Melanie Freeze, Thomas Gift IV, Marc Hetherington, Sunshine Hillygus, Lin-­Manuel Miranda, Frank Orlando, Jason Reifler, Dave Rohde, Melissa Sands, David Sparks, Nathan Teetor, and Dane Wendell. Special thanks go to Christopher’s brothers Peter and Paul, and to Ashley Jardina—his best of friends in the worst of times. One thing that is “interesting” about academia is that those who go into it are people who sometimes let their work consume them. Academia can be brutal on one’s mental health and one’s relationships. Needless to say, we have to thank those who have been closest to us during our best and worst times: Melissa Spas and Terrell Smith, the two millennials whom we still love most. We dedicate this book to them. Christopher adds: Unlike Candis, this is my very first book. As a result, I have a few more things to say. I have a number of people to thank for the support and encouragement that led to this book’s completion. My educational journey took me to Allegheny College, where I first met our fantastic editor Chuck Myers, and later to Vanderbilt University, where I began my PhD as an aspiring political theorist. My first year in Nashville, I had the very good fortune of taking two courses that forever set me off on a different path: Quantitative Methods with Suzanne Globetti and Public Opinion with Marc Hetherington. Seeing how much I enjoyed computational statistics and methods, both Marc and Suzanne encouraged me to shift my focus and supported my applications to graduate programs as an Americanist. In 2007, I moved down the road (I-­40) to Durham and began in the PhD program at Duke University. There I worked with John Aldrich and alongside Dave Rohde in Rohde’s Political Institutions and Public Choice center. With the continued support of Marc, and the generosity of both John and Dave, I finished at Duke in 2012. Now that I have finally completed this manuscript, I will turn back to working on Schwartz values and partisanship in America. For those readers who have heard either Candis or I talk about “our book” over the last six or seven years, here it is. It took far longer than either

Preface / xix

of us would have liked, but we are optimistic that you will find our arguments interesting and the evidence compelling. In terms of one final acknowledgment for this project, the person I would actually like to thank more than anyone else is Candis. In an occupation that puts tremendous pressure on junior scholars to publish and to do so quickly, Candis is a coauthor who demonstrated incredible patience with me while we finished this book. As someone who struggles with treatment-­ resistant major depression (which is exactly as much fun as it sounds), at times I did not want to work on this (or any) project. During these times, Candis always emphasized how important it was, to her, that I was taking care of myself. In the spring of 2015, when this book was first completely drafted, I mentioned to Candis that one of the reasons I was dragging my feet on the project was because “owing her work” was something that helped me get through each day. Understanding my meaning, Candis responded with an open heart, telling me that “if writing a book with me is going to keep you around, I think we should scrap this book and start all over from the beginning; you are more important to me than our book.” I am so glad to have her as a collaborator and friend . . . even if she really disliked me when we first met.

Not All Change Is Progress: An Introduction

It is to our credit that the efforts at reconciliation have been unceasing, and while our progress toward the achievement of greater universal human rights has moved in zig-­zag fashion and has often been blocked or delayed, in the main, the trend has been upward. —Pauli Murray, 1953 Yes, our progress has been uneven. The work of democracy has always been hard, contentious and sometimes bloody. For every two steps forward, it often feels we take one step back. But the long sweep of America has been defined by forward motion, a constant widening of our founding creed to embrace all, and not just some. —Barack Obama, 2017

Americans, as Pauli Murray suggests in the epigraph, have typically expected unceasing and continuous progress toward racial egalitarianism, even after incorporating setbacks in their prediction-­making calculus. Over a half century after Murray made this observation, the nation’s first self-­identified African American president left office with a note of optimism, asserting that the arc of US history bends toward inclusion and egalitarianism. Indeed, between the 1940s and the late twentieth century, the United States saw momentous and positive change in White Americans’ racial attitudes. In that time, biological racism, overtly racist attitudes, and a tolerance for racial inequality declined significantly and rapidly. For example, in 1944 Americans were asked, “Do you think Negroes should have as good of a chance as white people to get any kind of job or do you think white people should have the first chance at any job?” Over half (55 percent) of White Americans agreed that members of their racial group should have preferential treatment. By 1972, in a single generation, this number dropped to 3 percent.

2 / Introduction

Similarly, in 1942 nearly 70 percent of White Americans believed that White and Black schoolchildren should attend separate schools. By 1995, only 4 percent of  Whites agreed that separate schools for children of different racial groups was good education policy (Schuman et al. 1997). When it comes to measures of overt, biological racism, a vanishingly small proportion of Americans foster anti-­Black attitudes. But, something has gone wrong in recent years. Rather than the expected upward trend of two steps forward, one step back, we witness stagnation. Racial progress once seemed inevitable, but it no longer does. Even though we have seen significant changes in what most Americans clearly understand as racist attitudes, we have not seen the same changes in more symbolic matters. Figure I.1 shows the remarkable stability of the four items that make up the racial resentment scale, an important measure of symbolic racial attitudes, between 1986 and 2016. We see very little change over the past three decades on measures of symbolic racism, and it seems that anti-­ Black stereotypes and affect seem to have hit a wall as well. How can it be that 53 percent of Whites score Whites as harder working than Blacks, and 44 percent score Blacks as less intelligent than Whites (Grant 2014)? Stale and harmful stereotypes persist, and in some cases, we see an increase in anti-­Black attitudes. In 2008, 49 percent of Americans held implicit anti-­Black attitudes; four years later, this bumped up to 56 per­ cent. In that same time, the proportion of Americans willing to express explicit anti-­Black attitudes also increased from 48 percent to 51 percent (AP 2012). By 2016, approximately sixty-­three million Americans were willing to throw their support behind a presidential candidate who claimed that most Mexican immigrants are drug traffickers and rapists; proclaimed that Black Americans are at war with themselves and have nothing to lose; called for a ban on Muslims entering the country; and has equivocated on his stance on whether the internment of Japanese citizens during World War II was the right thing to do—­all of this is separate and apart from any of his homophobic, transphobic, and sexist proclamations. While many have suggested that most people voted for the forty-­fifth president because of his populist rhetoric and their economic hardships, a growing body of research provides evidence for the central role racial antipathy played in a large sector of the electorate’s voting calculus (Pettigrew 2017; Schaffner, MacWilliams, and Nteta 2018; Sides, Tesler, and Vavreck 2017). Not only do we provide evidence that racial progress has flatlined, we also offer several explanations for why. There is probably a plethora of reasons why we see the trend illustrated in figure I.1, but a large piece of the puzzle can be explained by the central claim of this book: Younger Whites,

I.1.  Racial Resentment (1986–­2016)

4 / Introduction

namely members of the millennial generation, are not doing the work that young peo­­ ple in the past have done to make significant and positive changes in aggregate ra­­ cial attitudes or in policies that aim to ameliorate racial disparities. Instead, there is a series of countervailing forces that prevents the positive characteristics of  White mil­­ lennials (e.g., values of egalitarianism and diversity, a recognition of White privilege) from coming to fruition. We focus on young people, and millennials in particular, for several reasons. History shows that major changes in attitudes about minority groups tend to come along due to cohort replacement or generational change. Consider, for example, the fact that young people, Black and White alike, were integral to the civil rights movement. As we all know, this movement had major policy implications. Similar to the Student Nonviolent Coordinating Committee, which played a role during the fight for civil rights, Students for a Democratic Society (SDS) was another group led by baby boomers that was actively engaged in the anti-­Vietnam movement that fostered and thrived on college campuses around the country. We focus on millennials because they constitute a particularly special group in the American political landscape, perhaps having the greatest potential to make positive change in matters concerning racial egalitarianism. Why? Sheer numbers alone get us pretty far. Millennials represent about one out of three American adults. What’s more, they tend to have more progressive attitudes than their predecessors. Over two-­thirds of millennials approve of same-­sex marriages, about the same proportion that favors the decriminalization of marijuana. Moreover, millennials are relatively progressive on some racialized issues; over half of all millennials believe in a pathway to American citizenship for undocumented immigrants. Furthermore, this group is the most educated living generation. For example, in the 2016 Cooperative Congressional Election Study, approximately half of millennials ages twenty-­six to thirty-­two had already graduated from college (Ansolabehere and Shaffner 2017). Research shows that with more education, people tend to become more racially liberal and less racially resentful. Furthermore, since these people are attending college, they are more likely to interact with people who are not like them, which ideally would lead to a greater acceptance and appreciation for personal differences. Given how different millennials are from their predecessors, if any single generation has the potential to usher in a more progressive era, it’s them. These kinds of statistics and facts may lead people to believe that millennials are, politically, an overwhelmingly liberal group. Indeed, many pundits, journalists, and scholars have praised the millennial generation

Not All Change Is Progress  /  5 Racial Group

Percent

100

Other

75

Asian

50

Black Hispanic (any race)

25

White 0 Silent Generation (b. 1928−45)

Baby Boomers (b. 1946−64)

Gen Xers (b. 1965−80)

Millennials (b. 1981−96)

Generation Source: Pew Research Center

I.2.  Racial and Ethnic Demographics of Four American Generations, 2014

for being the most tolerant and open living generation (Rutten 2008; Taylor and Pew Research Center 2014; Zogby 2009), but these summary statistics do not tell the whole story. The shape of millennials’ collective attitudes is, in part, likely to be a result of the millennial generation being the most diverse generation in American history: over 40 percent of the millennial generation is non-­White. In 2014, 57 percent of millennial adults identified as White in comparison to 61 percent of Gen Xers (born between 1965 and 1980), 72 percent of boomers (1946–­1964), and 78 percent of members of the silent generation (1928–­1945) (Pew Research Center 2015). Figure I.2 illustrates the racial and ethnic diversity of the millennial generation in comparison to their predecessors. Millennials of color tend to be more liberal on a range of issues in comparison to their White counterparts, especially on racial issues. For example, about 55 percent of non-­White millennials identified racial cleavages as the largest source of division in America; only 35 percent of their White counterparts agreed. Fifty-­three percent of all millennials favor a larger government that provides more services, but that support is being driven by the 71 percent of non-­White millennials who favor such a policy. In contrast, fewer than four in ten White millennials believe we should expand government spending (GenForward Project 2016). Additionally, more and more studies reveal that White Americans, including millennials, are actually quite anxious about an increasingly racially diverse demographic profile (Craig and Richeson 2014a; Jimenez 2017; Smith and Mayorga-­Gallo 2017). We are not alone in our focus on millennials. As Winograd and Hais note, “to have a clear sense of where America is headed in the future requires a thorough understanding of the behaviors and attitudes of the Millennial

6 / Introduction

Generation” (2011, 1). Millennials have been decisive in recent elections and will continue to be for decades into the future. The way they understand both the causes of and solutions to problems will influence the policy changes that we are likely to see moving forward. The way they understand race and racism will shape our collective racial grammar for generations to come. Millennials are the children and grandchildren of generations that pushed for policies that nudged America toward its better angels—­ egalitarianism and opportunity for success—­but young White Americans are not necessarily doing the same thing, or at least not to the same degree as their parents and grandparents. So, how could it be that the millennial generation, composing an increasingly larger portion of the American electorate, contributes to a stagnation in racial attitudes? How can we make such a claim in the face of data that show that young White Americans are more likely to marry interracially? To believe that all people should be treated equally regardless of race? To have given the overwhelming majority of their votes to Obama in 2008? We have one overarching objective: to explain why there has not been sustained, significant, positive change when it comes to racial attitudes in the hearts and minds of White America. We explore the two phenomena we believe to be responsible for what we see empirically. The first concerns racial ideology. We rely on Stuart Hall’s definition of ideology: “those systems of meanings, concepts, categories, and representations which make sense of the world” (1996, 48). Racial ideology, to be more specific, can best be understood as “the racially based frameworks used by actors to explain and justify (dominant race) or challenge (subordinate race or races) the racial status quo” (Bonilla-­Silva 2014, 9). Relatedly, “the central component of any dominant racial ideology is its frames or set paths for interpreting information” (74, emphasis in the original). Race frames are the “lenses through which individuals understand the role of race in society,” and they “impact the interpretation of social phenomena by making certain aspects prominent and obscuring others” (Warikoo and de Novais 2014, 860). There is a pair of things to keep in mind about racial ideology. First, there are often multiple ideologies that exist simultaneously, but generally speaking, the ones held by dominant groups are ones that allow for the maintenance of existing hierarchies. Second, ideologies evolve over time. Racism, in particular, has been referred to as a “scavenger ideology,” because over time it discards the parts that make it too obvious, transparent, or abhorrent and incorporates new components to keep it both relevant and

Not All Change Is Progress  /  7

elusive, usually by appropriating the values of (middle-­class) White Americans (Mosse 1995, 164). By examining the racial ideology and frames that White millennial Americans rely on most, we are able to accurately describe the shape of today’s dominant racial ideologies and to do so in a nuanced way. Through grounded theory, we uncover a series of countervailing forces—­ factors that work simultaneously but in opposition—­to help explain what we see empirically and to predict what we should expect for the structure, nature, and effect of racial attitudes in years to come. Our theory of countervailing forces helps to explain the apparent stagnation in racial progress. Through our interviews with young people, we show that White millennials do have racially inclusive attitudes, do value diversity, and do not want to be racist (or even appear racist) themselves. But, we also show that while they have gained the ability to see racism as a structural and systemic problem, they are not more emotionally invested in fixing these problems. Baby boomers, in contrast, are more likely to have had direct experiences with overt racism over the course of their lives and thus have greater familiarity with systemic racism than the average White millennial (whether most boomers are still willing to admit the role of structural racism on contemporary racial disparities is a separate matter we address in part 3). Consequently, White millennial Americans tend to have a commitment to color-­blind racial ideology and rely on diversity logic. Their emerging racial vocabulary is quite different from how political scientists, in particular, tend to discuss, describe, and measure the racial attitudes and sensibilities of the average non-­Hispanic White American. These contemporary racial frames are ostensibly less hostile toward Blacks in comparison to how scholars typically depict Whites’ racial attitudes. Ultimately, however, contemporary sentiments lead to the same lack of support for policies aimed at decreasing racial disparities and inequalities that we have seen among Generation Xers, boomers, and even the silent generation in some cases. The data we provide help us to see that one of the reasons why racially progressive attitudes have ceased their “steady march” is because today’s dominant racial ideologies provide no room for structural racism as an explanation for persistent racial disparities and inegalitarianism. The second issue at hand, then, is methodological. As mentioned, great minds have suggested that racial progress in the United States comes in fits and starts. It waxes and wanes. It goes forward two steps and back one. It may be best characterized as “zigzag” or serpentine (Murray 1953; Omi and Winant 1994; Smith 1991). Generally, Americans have expected racial attitudes to continue to liberalize due to the ongoing changes in social

8 / Introduction

norms as well as the replacement of older, less tolerant individuals with younger, more tolerant ones. However, there has not been a systematic assessment of these changes over the past two decades or so. Additionally, not only does there exist a large body of scholarly evidence suggesting that the configuration of racial attitudes has changed within the past twenty years, but also some research shows that there are both cognitive and affective components of racial animus (Forman and Lewis 2015; Banks and Valentino 2012; Sears 1993); the research on the latter point has been undervalued and largely overlooked in the literature aiming to quantitatively measure racial animus. Our findings speak to, and bolster, these bodies of literature. From this we take that as the level, structure, and nature of racial attitudes evolve, so must the measures we use to ascertain the extent to which these attitudes exist and influence politics. We are challenged then, and required as responsible social scientists, to think more creatively and more holistically about how to measure Americans’ racial attitudes in the twenty-­first century. Political science still largely relies on a measure of racial attitudes developed well before any millennial was born, and while tried-­and-­true measures like the racial resentment scale remain helpful in understanding Americans’ racial attitudes, resentment is only one ingredient in the racial attitudes soup du jour. What’s more, in the political science literature, “racism” or “racial attitudes” are primarily conceived of as “prejudice”—­an individual’s dislike of another person due to their race. We rely not only on our training as political scientists but also on the tools provided by critical race scholars, sociologists, and psychologists who have a wider perspective on how “racial attitudes” might be conceptualized beyond prejudice alone. This broader view also takes into account whether one acknowledges racial privilege (or disadvantage), whether one is willing to admit to or is aware of ongoing structural inequalities, the extent to which one has an understanding of one’s place in the racial hierarchy, one’s fear of other racial groups, and one’s empathy and level of concern about the racial status quo. All of these facets come together to form what many refer to as, simply, “racial attitudes.” In the following pages, we argue that when scholars rely on older measures, we miss out on the new ways of seeing race that are hiding in plain sight. A major contribution of Racial Stasis is the development of a new measure of racial attitudes, one that is theoretically and empirically derived over the course of this book. Our measure, which we call FIRE—­Fear, acknowledgment of Institutional Racism, and Empathy—­serves as a more

Not All Change Is Progress  /  9

nuanced, holistic, and multidimensional conceptualization of contemporary racial attitudes.

Key Terms There are a few key terms that we rely on throughout this book. The first is generation, which we sometimes use interchangeably with cohort. We focus on generational groups in the way that Karl Mannheim thinks of them, as a group of people who are close in years of birth and therefore “have a common location in the historical dimension of the social process” (1952, 290). Similarly, scholars like Norman Ryder (1965) theorized that cohorts tend to share a certain set of attitudes or values because they may also share distinctive experiences at specific points in their lives. For instance, the authors of this book were about seventeen or eighteen years old on Septem­ ber 11, 2001, when the terrorist attacks on the Twin Towers occurred. We have a very different understanding of 9/11 and the events that occurred after it than many of our current students; our oldest undergraduate students were three or four at the time and have lived their entire lives in a country at war. As such, our generation and that of our students may have different attitudes about American foreign policy, war, and terrorism. We may also have different constructions of Muslims, Arabs, and Islam, given the shifting media depictions and narratives of these groups over time. But it should be mentioned that children aren’t socialized in a vacuum (Hyman 1959); some core ideas are likely to be passed down from one generation to the next, though ideas, values, and norms shift over time and get interpreted through different generational lenses. A closely related term is millennial or millennial generation. People often use millennial in a pejorative way to discuss all young people, especially those with a sense of undeserved entitlement, but the eldest millennials at the time of this writing are now approaching forty (ourselves included), and the youngest are able to legally vote, drink, and purchase a handgun. Millennials are those people who were born between the early 1980s and the mid-­1990s. The Pew Research Center has declared that those born between 1981 and 1996 should be considered millennials, and for our analyses we adopt the Pew standard.1 To be sure, there is nothing especially different between someone born on December 31, 1996, and someone born on January 1, 1997. The boundaries around generations are not totally arbitrary, but the hard cutoffs are a handy tool for the kinds of analysis used in research like ours (Dimock 2018).

10 / Introduction

The third term that we want to highlight is structural racism. Typically, when most (White) Americans think of racism, they think of individual attitudes rooted in prejudice, or interpersonal discrimination. In fact, we provide a great deal of evidence for this in part 2. While individual bigotry is an important aspect of racism to consider, we encourage our readers not only to think about overt behaviors (e.g., lynching) and attitudes (e.g., White nationalism) but also to consider and incorporate structural racism into their understanding of how race and racism function in American society. Structural racism is a feature of society whereby patterns of public policy, institutions, dominant ideologies, and popular representations serve to perpetuate social, political, and economic inequities along racial lines. By having a broader understanding of racism—­as both interpersonal and systemic—­we realize that we have to think more creatively about how to describe and measure the role of racism in a society like the United States of America.

Data Sources We rely on numerous data sources and methodological strategies throughout this book. We use both qualitative and quantitative data. Each of our analyses has both strengths and weaknesses, and we discuss them when we use different methodological tools. By using a range of data and methods, we can triangulate and corroborate our findings. More importantly, we can use what we believe are the most appropriate methods to test our hypotheses. Whether using confirmatory factor analysis, multivariate regression, or age-­period-­cohort analyses, we strive to explain why we have chosen that method as the most useful in answering the question at hand; in appendixes A and B we try our best to outline, in simple terms, both how to use and how to understand these methods. Our qualitative data derive from a set of semi-­structured, face-­to-­face interviews with forty-­three White millennials conducted by three fantastic research assistants—­all young, White women—­during the summer of 2014, about a year after George Zimmerman was acquitted for the murder of  Trayvon Martin and just before Michael Brown, an unarmed teenager, was fatally shot by a White police officer in Ferguson, Missouri. These interviews were completed at three sites—­one in the Northeast, one in the Midwest, and another in the South—­but the respondents themselves hailed from a wider geographic area. The goal of the interviews was to ascertain how young Whites understand the roles of race and racism in American society. Scholars show that Whites tend to provide different, often more liberal, answers

Not All Change Is Progress  /  11

in their survey responses on racial matters than they do when they are asked to elaborate on their rationale (Bonilla-­Silva 2014). By analyzing interview data, we can gain a better understanding of the racial logic, racial grammar, and racial ideologies that young people rely on in their explanations of why the world works the way it does. Again, we focus on White millennials because they give us insight into the boundaries of the race talk that will dominate American racial ideology moving forward. To be clear, it was not our intent to select a random group of individuals for these interviews. We focused on non-­Hispanic White millennials. Those we interviewed were highly educated—­nearly all of them were either current college students or college graduates at the time of the interview. We focused on this group because these individuals are likely to think critically about race matters, be aware of racial privilege, and perhaps have been exposed to notions of structural, institutional, or systemic racism. Therefore, if conventional wisdom is correct, any bias that we see in this group should be toward more racially progressive attitudes: racial tolerance, appreciation for diversity and inclusion, critical thinking, and liberal policy preferences. Though the names of our respondents have been changed, their demographic information and the interview schedule can be found in appendix C; some data, such as college names, have been removed to ensure anonymity. Typically, the quotations we rely on are reported verbatim, but some have been edited for clarity—­for example, by removing fillers such as “like” in order to allow for an easier read. Our interviewers, all of whom were also White, asked respondents a series of questions about topics ranging from their understanding of the American dream to their attitudes about affirmative action. The interviews generally occurred in a public place and lasted between 23 and 107 minutes. After the interviews were completed, they were transcribed by a research assistant and coded by the two of us. We rely on grounded theory, or an inductive process of theory building. That is to say, we analyzed the data and took note of the patterns that arose across the responses of our participants in order to build a theory of contemporary racial attitudes and racial stagnation. We build on the findings of our qualitative data analysis with quantitative, large-­n, nationally representative data sources. To begin, we examine changes in racial attitudes over several decades and, just as above, rely on social scientists’ “gold standard” data sets: the American National Election Studies (ANES), the General Social Survey (GSS), and the Cooperative Congressional Election Study (CCES), and also US Census data. These data are most prominently employed in chapters 1 and 2. As we turn to examine

12 / Introduction

individuals’ racial attitudes, particularly in chapters 6 through 8, we rely on the 2014 and 2016 Cooperative Congressional Election Studies. For 2014, we have a subset of data with over seven hundred self-­identified White respondents. For 2016, we have over forty thousand White respondents.

Plan of the Book and a Preview of Our Findings As previously mentioned, racism is structural, not just a series or aggregate of attitudes (Bonilla-­Silva 1997). But we focus on attitudes here because (a) we can measure attitudes (for the most part, though not perfectly) and (b) we know that attitudes influence political outcomes and, therefore, af­ fect structure (Omi and Winant 1994). While it is clear to us that race in­ fluences Americans’ life chances and that racial attitudes and political attitudes are linked, there still exists a debate around these relationships. As a re­sult, the first three chapters—­which make up part 1—­are concerned with des­cribing the United States’ existing racial landscape. In chapter 1, we briefly discuss this debate, since this book seeks to both participate in and expand on it. Then we shift our attention to the changing nature of racial attitudes and what these changes mean for social scientists aiming to measure, describe, explain, and predict these attitudes. It is here that we outline our central hypothesis: The United States is experiencing what we call racial stasis, and White millennials are not helping to usher in more progressive racial attitudes at the rate we might expect. Chapter 2 begins with the question, Is race special? Here, we are concerned with (a) whether Whites’ attitudes toward African Americans have really improved over time and (b) the extent to which measurable changes over time originate from generational change or if they are due to some other mechanism of social change. Throughout the book, we use a variety of words like “improved” synonymously with words like “progressed” or “liberalized,” largely because we believe that it would be a normatively good thing for racial animus to decline, for the reliance on racial stereotypes to diminish, and for racial empathy to increase. In that same spirit, and because change is relative, we compare trends in attitudes about African Americans with trends in attitudes about women and members of lesbian and gay communities. In these analyses, we provide evidence for a foundational claim we have already made: compared to attitudes on many other issues and across several similar domains, progress in racial attitudes has stagnated. This is a hard pill to swallow given the body of prima facie evidence that almost any American can cite to show that things have changed over time: Obama’s election and reelection, the inclusion of people of color

Not All Change Is Progress  /  13

at various levels of government, the racial and ethnic diversity on America’s elite college campuses, and so on. Our results, however, reveal that racial attitudes, especially toward Black Americans, are in fact the exception rather than the rule. Race, as an issue in mass public opinion, is “special” insofar as attitudes on racialized issues have not changed during periods when attitudes toward other historically marginalized groups shifted quite dramatically. Further, we show that White millennials are not very different from their predecessors, and on some racialized issues they are more conservative than some older Whites, though we take the findings in chapter 2 with caution. In chapter 3, we use quantitative data to determine whether the measure that political scientists rely on most to assess levels of symbolic racism in the United States, the racial resentment scale (RRS), accurately assesses the level, shape, and nature of millennials’ racial attitudes. We do this because, in a sense, the findings in chapter 2 suggest that millennials are not really so different from their predecessors, and that does not comport with conventional wisdom. We have to keep in mind that the RRS was developed to measure attitudes in a particular historical context—­an America that had just experienced the civil rights movement, where opportunities had been opened to Blacks for the first time. Since then, there have been shifts in social norms, immigration patterns, and the overall racial landscape, so we test whether the questions posed by the RRS will bring the same things to mind for a generation that is far removed from such a tumultuous racial era as they do for older White Americans. Here, we assess whether the measures we most often use are appropriately capturing racial attitudes, measuring the same attitudes in the same way across birth cohorts. We examine how expressions of racism have changed over time and conclude that the measures of racial animus that most scholars use might not be capturing the same things across the generational gap. We find that not only are White millennials answering the questions to the racial resentment scale in systematically different ways than their older peers, but also their attitudes as measured by the RRS are more strongly connected with measures of “old-­fashioned,” biological racism and uncorrelated with the ideological components of the measure. Why might this be the case? As we discuss in chapter 3, we believe White millennials have a different relationship with racial matters in the United States than previous generations did. As scholars of racial and ethnic politics, we find it clear that today’s young Whites are introducing a different conceptualization and language to frame and discuss issues of race and racism in contemporary America. We turn to this in part 2. With the help of three White, college-­aged, millennial

14 / Introduction

women, we interviewed over forty White millennials about their attitudes about racism, diversity, and racial privilege. Our respondents were asked questions like, What, according to you, is racism? Do you value diversity; why? To what extent are Whites and people of color advantaged or disadvantaged because of their race? What do you think about affirmative action and stop-­and-­frisk? In chapters 4 and 5, through our analysis of millennials’ responses to these questions, we are able to take a giant step toward gaining a better understanding of why today’s young Whites are not leading the charge in racial progress with regard to the aggregate American public opinion on issues surrounding race and racism. Through our interviews, we are able to uncover four sets of countervailing forces that help structure their attitudes about race and racialized policies. To preview, we find that White millennials feel disconnected from an era of overt racism. Raised in an environment that has preached color-­ blindness, they often claim to not notice race, and many believe it is impolite to even talk about race. This mindset makes it difficult for millennials to see the connections between race and inequality in America. This color-­ blindness brings with it two beliefs that work to hinder racial progress, as we will show. First, rooted in a somewhat egalitarian ethos, millennials believe that race should no longer matter in America. Obviously, in a perfect world no one would (need to) be treated differently based on the color of their skin, but a problem arises when the belief that race should not matter transforms into a belief that race does not matter. As a result, talking about racial issues (a necessary mechanism for racial progress) causes cognitive dissonance for the color-­blind millennial; mentioning race and racial differences is an affront to their conception of contemporary America and their place in it. We find that young people’s racial attitudes are constrained in systematically different ways than their predecessors’ attitudes have been described as being (Kinder and Sanders 1996; Kinder and Sears 1981). Young people are self-­conscious when interviewed and talking openly about race; they do not think about racism as systemic, generally speaking. We show that being color-­blind is different from, though not fully divorced from, being racially resentful. Using the insights from our millennial interviewees, we are able to derive expectations for how to build a new measure of racial attitudes, which we lay out in part 3. In the final set of chapters, we use these findings to interrogate newer measures of racial attitudes that appear to be more in line with an emerging racial grammar (Bonilla-­Silva 2012). These measures attempt to capture notions of knowledge about racial privilege and institutional racism as well as empathy to other racial groups, White guilt, and fear. In chapters 6 and 7,

Not All Change Is Progress  /  15

we are able to develop a multidimensional framework for mapping Whites’ racial attitudes, ultimately deriving a measure that maps onto our findings from part 2. Our final substantive chapter, chapter 8, focuses on the political and policy implications of contemporary racial attitudes, particularly as they relate to the 2016 presidential election. We conclude by reflecting on what we have taken away from this book and on the broader implications of our results for political science, for the role of racial attitudes in the United States, and for American democracy. We would be remiss to not consider the fact that we are also living in both a post-­Ferguson America and a post-­Charlottesville America,2 and thus we take the opportunity to briefly address our expectations for Generation Z, or the post-­millennial generation, before we close the book.

Before We Move On: Why White People? Given the demographic shift we previously mentioned, one might ask why we focus primarily on White Americans in this ever-­changing environment. The main reason is because we recognize that demographic changes will not necessarily lead to proportional changes in power dynamics. In fact, this new level of diversity may heighten Whites’ racial anxiety and only further entrench America’s so-­called racial divide. Scholars have already begun to provide evidence that many White Americans “experience the impending ‘majority-­minority’ shift as a threat to their dominant (social, economic, political, and cultural) status”; this impending sense of doom is manifested through more conservative political attitudes and policy preferences, including those closely related to matters of racial equity (e.g., criminal justice, diversity initiatives, and immigration policy) (Craig, Rucker, and Richeson 2018, 206, 210). And, perhaps, this phenomenon is best illustrated by the rise of the Tea Party (Parker and Barreto 2013), Donald J. Trump’s election to the American presidency (Sides, Tesler, and Vavreck 2017), and the increase in hate groups and hate crimes against people of color (Chen 2017). Nonetheless, critics may point out that Asian Americans are the fastest-­ growing racial group in the United States, that Latinx people make up a significant portion of the population, and that Black American voters were overrepresented in two of the last three presidential elections, such that their record voter turnout was decisive in Obama’s elections. Here’s the thing: Asians are the fastest-­growing racial group because of immigration, and, similarly, Latinx and, to some extent, Black populations are also immigrant-­ replenished groups. We have to consider not only the time it takes for immigrants to become citizens but also the actual number of individuals who

16 / Introduction

will be eligible for citizenship to begin with. We must also keep in mind that there are citizens, disproportionately Black and Latinx, who are not eligible to vote for other reasons, such as felon-­disenfranchisement laws (not to mention “successful” vote-­dilution tactics). Secondly, we should not expect a one-­to-­one relationship in the change in population and a change in the distribution of power. We know that a group’s proportion of the population may not tell us much about its members’ political and economic resources and access. Consider, for example, South Africa during apartheid (and afterward, too). Consider the antebellum American South. Consider almost any of the European colonies. Compared to Whites in each of these cases, racial and ethnic minorities made up a significant, if not an overwhelming, majority of the population, yet they were completely disenfranchised. The winners make the rules, and often these rules are crafted to keep the winners in power. Third, we are well aware that Whites’ racial attitudes alone do not dictate policy outcomes. To be sure, there is and has always been a dialectic push and pull among all racial groups in determining the shape of the racial landscape in the United States. For instance, Black freedom struggles have been at the forefront of shifting the United States toward egalitarianism (Taylor 2016; Lebron 2017; Lopez Bunyasi and Smith 2019b). Further, Whites’ attitudes do not exist in a vacuum—­indeed, Whites’ responses to racial progress often lead to regressive policy outcomes for all racial groups, including poor White Americans (López 2015; Omi and Winant 1994). But White racial attitudes do set the outer limits of what is possible for racial progress, given the group’s political dominance and prowess. We know that of all the things that tend to lead people to participate in politics, Whites have more of each of them in comparison to Blacks and Latinx people, and sometimes Asian Americans too (Junn 2006; Verba et al. 1993). Whites have more wealth and lower rates of poverty and incarceration in comparison to Blacks and Latinx people. They have higher returns on their education than most minority groups. They have higher wages, better health outcomes, higher rates of educational attainment, and higher rates of employment than the two largest minority groups. These things will not change dramatically or rapidly due to demographic shifts. Ultimately, we focus on Whites because they do not appear to be losing their privileged position anytime soon, despite perceptions otherwise. Finally, to be absolutely clear, we acknowledge and recognize that Whites are not a monolithic group. Each individual is different: some are liberal, others conservative; some incorporate their racial identity in their political decision-­making calculus, others do not even consider their racial identity;

Not All Change Is Progress  /  17

some are actively racist, some are actively anti-­racist; some enjoy Hamilton, others just haven’t heard it; and so on. Even though individuality is important in day-­to-­day life, as social scientists we focus primarily on averages and aggregates, highlight means and medians, and care little about extreme outliers. “Generally speaking” is a phrase that we could put in front of nearly all of our findings, and generalities are what we are focused on in Racial Stasis.

One

Nature of the Game: The Racial Stasis Hypothesis

Racism, we are not cured of it. And it’s not just a matter of it not being polite to say “nigger” in public. That’s not the measure of whether racism exists or not. —Barack Obama, 2015

We begin with two basic and (what we believe to be) reasonable assumptions. First, race matters in the United States. The America one experiences is largely a function of one’s racial-­group membership. In every domain of life that has the potential to produce more dignified living conditions or create more opportunities for success—­such as housing (Massey and Denton 1993; Massey et al. 2016; Rothstein 2017; Katznelson 2005; Yinger 1995; Ross and Turner 2005), education (Hannah-­Jones 2014; Camera 2016; Bright, Duefield, and Stone 1998; Ross and Turner 2005; Cottom 2017), healthcare (Olshansky et al. 2012), employment (Pager and Shepherd 2008), wealth (Oliver and Shapiro 1995; Darity 2011); criminal justice (Alexander 2010; Baumgartner, Grigg, and Mastro 2015), and public safety (Baumgartner, Epp, and Shoub 2018)—­people of color, especially Black Americans, are much less well off than White Americans. As points of fact, the Black unemployment rate has historically been twice as high as that of Whites (Bureau of Labor Statistics 2014), a White ex-­felon is more likely to get a job than a Black man without a criminal record (Pager 2007), and even though White Americans are more likely to use illegal drugs, Blacks are more likely to be imprisoned for drug offenses (Alexander 2010). And educational and monetary gains do not necessarily mean that Blacks will automatically be better off. For example, the median well-­educated Black family with a post-­college education has significantly less wealth (about $84,000) compared to their similarly educated White

22 / Chapter One

counterparts ($293,100), and this is still less than the $86,200 that the median White family with only some college education has, largely due to intergenerational transmission of assets (Hamilton et al. 2015). While income helps to pay one’s recurring bills, wealth provides families with a safety net to fall back on if they are met with hard times or to rely on for future investments, such as college tuition or additional homes. The average “American experience” for Whites is substantially different than that of other racial groups, especially Blacks. The second basic premise we rely on is the notion that Whites’ attitudes toward race have markedly changed over the last sixty years. This change is multidimensional. On one dimension, we see that what White Americans deem as socially acceptable expressions of racism has transformed over time. For instance, President Obama received an incredible amount of (negative) attention from the news and the nation when he used the “N-­word,” as presented in the epigraph of this chapter. Using the N-­word (in public) is no longer allowed, according to most Americans.1 In 2014, Donald Sterling, the longest-­tenured owner of an NBA team, was banned from the NBA for life for making racist comments about his team’s Black players and Magic Johnson. Four years later, John Schnatter, the founder of Papa John’s Pizza, came under fire for using similar language in a conference call, leading him to step down from his position and apologize for his language. This kind of behavior is considered a clear marker of a person with a racist heart and is thus unacceptable to most Americans today, even though it is well known that many White Americans use this language in private or in all-­White spaces. On another dimension, there has been a shift in racial ideology—­the frames of reference we use to explain racial phenomena, such as racial disparities. It used to be the case that Whites’ explanation of racial inequality was rooted in the idea that Whites are genetically or inherently superior to non-­Whites (Schuman et al. 1997). Currently and in contrast, White Americans are more likely to rely on a logic of color-­blindness to explain away racial disparities (Bonilla-­Silva 2014). Considering that we have seen significant shifts in racial attitudes and a decline in overt racism, it is tempting to believe that we can turn our attention away from race and racism and focus instead on class, culture, or individual behaviors to produce more racially equitable outcomes. But, we think it is important to discern whether what we are seeing today is a sign that racism is on the decline or whether, perhaps, today’s expressions of racism are more difficult to see, even if they are still very harmful. Therefore, we ask a series of questions that are important to scholars of both American

Nature of the Game  /  23

public opinion and racial and ethnic politics: Has there been a continuing decline of racial animus among Americans in the United States over the past few decades? Does cohort replacement always lead to higher levels of racial tolerance? Should we expect millennials to be even more racially tolerant than members of previous generations? Have levels of racial animus genuinely changed (declined), or has the manner in which racism is expressed in today’s society changed, leading us to introduce new types of error into widely used “modern” measures of racial attitudes? We use this chapter to begin to answer some of these questions and lay down the groundwork for our central argument and theory concerning trends in Americans’ racial attitudes. First, we briefly describe three pivotal bodies of literature that each attempt to explain the role of Whites’ racial attitudes in shaping their policy preferences, especially regarding those policies that may exacerbate racial inequity; we put these theories in conversation with one another to illuminate the debate that ensues around discussion of Whites’ racial attitudes. Then, we shift our focus to examining how the nature of  Whites’ racial attitudes has shifted over time. Against this backdrop, we outline our central hypothesis—­the racial stasis hypothesis—­and discuss the implications of our predictions for social scientists who seek to accurately describe and measure an ever-­evolving set of attitudes.

Oh, What a Tangled Web We Weave: Racial Attitudes and Politics The question of whether race matters in American politics is one that can be answered succinctly—­the answer is “yes”—­but that would be neither a satisfying nor a convincing answer. We would have to consider and elaborate on the history, founding, and ideological bedrock of these United States. We would have to consider the role of race in political ideology, partisanship, and party realignments as well as in candidate evaluations, voting behavior, the extent to which the Supreme Court protects minorities from the tyranny of the majority, congressional representation and behavior, and presidential agendas to begin to answer that question fully. Instead, we will take up a far more manageable question here: Are racial attitudes still a central component of Americans’ policy preferences and issue opinions, and have the ideological nature, structure, and role of these attitudes changed in the last sixty years? As Vincent Hutchings and Nicholas Valentino explain, “the ensuing debate over the impact of racial attitudes on policy preferences has been among the most contentious in all of public opinion research” (2004, 389), but one thing most social scientists can agree on is that there is a large and

24 / Chapter One

persistent gap between the political opinions of Whites and the political opinions of ethno-­racial minorities. The largest gap is often between Blacks and Whites; sometimes Latinx people’s and Asian Americans’ opinions mirror Blacks’, while at other times their attitudes more closely resemble Whites’ (Dawson 2012; Hutchings and Valentino 2004; Masuoka and Junn 2013), but one’s race is often a strong predictor of political attitudes. For example, in 2007, 52 percent of  Whites favored affirmative-­action programs to help Blacks get better jobs and education, while 89 percent of Blacks and 77 percent of Latinx people agreed (Pew Research Center 2009). In response to the aftermath of 2005’s Hurricane Katrina, 66 percent of Blacks believed that the government would have responded more quickly if most victims had been White; 77 percent of Whites believed that the response time would have been the same. Seventy-­one percent of Blacks felt that the events that occurred around Hurricane Katrina revealed that racial inequality is still a problem in the United States, compared to 32 percent of  Whites who agreed (Pew Research Center 2005). We see this gap in the reaction to more recent events as well. According to a Washington Post/ABC News poll, 87 percent of Blacks believed that the shooting of Trayvon Martin in February 2012 was unjustified, while only 33 percent of Whites agreed; Latinx people were in between, “with 16 percent saying the shooting was justified, 34 percent saying it was unjustified and 50 percent reporting they couldn’t say whether the shooting was or wasn’t justified” (Levinson 2013). Similarly, after the shooting of Michael Brown, an unarmed Black teen in Ferguson, Missouri, a Pew Research Center poll revealed that 80 percent of Blacks believed that the shooting and killing of Brown raised important issues about race; 37 percent of Whites agreed, while 47 percent of  Whites reported that they felt the topic was receiving too much attention (Pew Research Center 2014). People’s perceptions of reality shape their beliefs about what are appropriate policy responses. For example, if people believe that the fatal police shootings of Black people are isolated incidents, then we need only to get rid of a few “bad apples.” If people believe that these shootings represent a systemic pattern of behavior and are thus structural or deeply rooted in the institution, then the policy recommendations that flow from this perspective may include the replacement of leadership, better training for all police, or even the removal of firearms from society altogether. How we see the world informs how we think we ought to change it. At the most basic level, “race” matters insofar as people across different racial groups view the link between identity, politics, and policy quite differently.

Nature of the Game  /  25

The Principle-­Policy Gap Although there are glaring differences between racial groups’ attitudes, social scientists cannot agree on why this gap persists. What’s more, social scientists cannot agree on why Whites are inclined to support egalitarian values at similar levels as people of color but are not supportive of the policies that could ameliorate racial disparities. Many schools of thought have developed intuitions and theories to answer these questions (e.g., social dominance theory, system justification theory, social identity theory, realistic group conflict theory, authoritarian personality theory, ethnocentrism), but we will highlight what we believe are three of the major groupings. The first school of thought is composed of theories that emphasize so­ cial structure and group interests. Here, one would find theories like realis­ tic group conflict, “sense of group position,” and social dominance orientation. These theories primarily suggest that “individuals identify with their own racial and ethnic groups, that the competing interests of those groups generate intergroup conflict, and that dominant groups develop ideologies to justify and legitimize their hegemony” (Sears et al. 2000, 22). For scholars in this camp, most notably Jim Sidanius, Whites’ attitudes about racially liberal policies stem from the perception that these policies work against their group interest in a zero-­sum game for resources such as jobs, promotions, and college admission. The game is perceived as “zero-­sum” in the sense that, due to a limited pool of resources, if one group is bound to benefit or “win,” the other necessarily loses. The second group of theories that try to explain Whites’ opposition to racially liberal policies can be characterized by sociopsychological models of racial prejudice. Theories like symbolic racism, modern racism, racial resentment, the authoritarian personality, and covert racism can be bundled together. While each of these theories has its own set of nuances, they all emphasize early-­life socialization of prejudice and social values (Sears et al. 2000). These theories suggest that racial attitudes matter in public opinion but racism has taken on a new form so as to conform to what is deemed an acceptable expression of racial animus. Supporters of symbolic racism and racial resentment as theories to help explain Whites’ attitudes toward minorities and racialized policies suggest that (a) overt bigotry does not predict attitudes now as well as it did in the past because these attitudes are condemned; (b) Whites have been socialized to have negative stereotypes about Blacks, “leaving a reservoir of racial antipathy decoupled from racial beliefs”; (c) Whites perceive Blacks

26 / Chapter One

as not holding up their end of the deal as far as working hard to reach the American dream; and (d) Whites believe that racial discrimination is not a constraint that Blacks face in a society where all are equals under the law (Sears et al. 2000, 17). In contrast to the socio-­structural theories, this group of theories minimizes the idea that Whites do not support policies aimed at increasing opportunities for minorities out of a sense of personal threat, and instead places the emphasis on Whites’ prejudices being rooted in a belief that Blacks are violating particular American norms. The last group of theories focuses on what some call “race-­neutral” aspects of White Americans’ policy preferences. The position of researchers in this camp is that the only relevant component of racial resentment is one’s commitment to the values and norms Donald Kinder and David Sears discuss: equality, fairness, hard work, and individualism. Though not without their critics (Federico 2006; Tesler and Sears 2010; DeSante 2013; Bertrand, Mullainathan, and Shafir 2004; Butler and Broockman 2011), scholars like Paul Sniderman and his coauthors (Sniderman and Piazza 1993; Sniderman and Tetlock 1986; Sniderman and Brody 1977; Sniderman, Brody, and Kuklinski 1984) suggest racial antipathy is not at the center of the great ma­ j­ority of Whites’ attitudes because “there is no longer the issue of race, no simple, emotional, gut-­level response to any racial issue, driven by Blacks themselves” (Sears et al. 2000, 28). Instead, this group of scholars asserts that Whites do not support policies like affirmative action not because of racial antipathy but rather because Whites tend to be politically conservative and therefore have principled, rather than racial, reasons to object to racially liberal public policies. Some people believe that all Americans should work hard, take care of themselves, and not look to others (or the government) for help. Here, having this principled commitment to some core set of American values, disjointed from any overt racial prejudice, could lead someone to answer the racial resentment battery in ways that may appear to be resentful, but those responses could be arrived at mostly by way of a race-­neutral and conservative ideology. We would like to note that we do believe it is possible for some Whites to be opposed to policies aimed at reducing racial disparities, like affirma­ tive action, for completely race-­neutral reasons. It could be the case that they believe that institutional discrimination is no longer a problem for Blacks in America. As a corollary belief, since Whites and Blacks are equally likely to face the same sorts of obstacles in gaining admission into a university or being hired for a job, a program that treats people unequally is ultimately unfair. This is a perfectly consistent and internally valid argument; its con­ clusion is supported by the sum of the premises. However, given the evidence

Nature of the Game  /  27

we have presented thus far, we believe that for a person to reach this conclusion, they are likely to be either unfamiliar with or deliberately ignorant of large swaths of evidence that indicate major problems with their minor premises.2 As much as we wish it to be untrue, Whites and Blacks do not have the same opportunities in America. As mentioned, on every major predictor of “success” that most people think of, the disparities between Whites and Blacks are deeply rooted in America’s history of structural racism, seen in discriminatory housing practices and policies, unequal educational systems, disparate health outcomes, and the distribution of wealth. Overall, while it may certainly be the case that some Whites oppose racialized policies on the basis of a completely race-­neutral ideology, many scholars have recently shown that, on average, racial animus still plays an important role in how Whites make decisions in a variety of settings. Although these three sets of theoretical models differ vastly in their explanation of Whites’ racialized policy preferences, they do have a couple things in common. One is that they all agree, to some extent, that racial attitudes matter, even if among a small group of Americans. More importantly, underlying all of these theories is the notion that racial attitudes and their role in American public opinion will change over time—­either in the extent to which they matter or in how, exactly, racism rears its head. All of these scholars, like us, agree that both the level and the structure of racial animus have changed, and will continue to change, over time. The Changing Nature of Racial Attitudes When scholars talk about attitudes, generally, they are referring to an affective orientation toward some specific object (e.g., a group or a policy). An individual’s true attitudes, we know, are impossible to measure perfectly, and this difficulty is exacerbated when one tries to study attitudes around sensitive topics like race. It is challenging to determine whether and the extent to which racial considerations are an important component of Americans’ political attitudes not simply because expressions of racial animus that Whites find acceptable to display in public have radically changed over time but also because the ideological nature of racial attitudes evolves. As mentioned, there was a time when Whites not only sincerely believed that they were inherently and biologically superior to Blacks (nature) but also were willing to admit this openly to their family, their friends, and even researchers who asked (expression). The overwhelming majority of social scientists who focus on racial attitudes would agree that while old-­fashioned “Jim Crow” racism is out of

28 / Chapter One

vogue and the tenets of biological racism are eschewed by most Americans, racial considerations still influence White Americans’ attitudes, especially toward policies that prime such considerations (such as initiatives concerning affirmative action, criminal justice, school desegregation, social welfare allocation, or problems in “the inner city”), but there is still plenty of plausible deniability. For instance, instead of the suggestion that Blacks remain behind Whites on important socioeconomic indicators because Blacks are inherently inferior or do not have the intellect to compete with Whites, a more acceptable explanation of the Black-­White socioeconomic gap is that Blacks do not work hard enough to attain a better life because “Black culture” prevents them from aspiring to do well without the help of the government. The latter set of explanations does not belie any animus, but the implied policy proposals to address individual behavior or culture are likely to range from paternalistic and controlling (Soss, Fording, and Schram 2008; Soss, Fording, and Schram 2011) to benignly neglectful and neoliberal (Taylor 2016; Cohen 2010); this range of policy outputs is likely to maintain or deepen inequality. We mention all of this to say that scholars of racial attitudes must always keep in mind that racial attitudes change not only in level but also in their ideological nature and expression. After the civil rights movement, scholars noted that Whites were no longer willing to rely on an old-­fashioned “Jim Crow” style of racism and opted for newer versions. The racial resentment scale, which we discuss in great detail in chapter 3, is the most frequently employed measure of racial attitudes in political science. When the racial resentment scale was developed, its questions were viewed as subtle. Agreeing to the statement “If Blacks would only try harder, they could be just as well off as Whites” allowed Whites to express animus toward Blacks without relying on the outmoded and taboo rationale of biological inferiority. Our work suggests that, today, even this restrained expression is becoming less acceptable for Whites to express openly. A more delicate approach to racial expressions is becoming the racial lingua franca among Americans. Racial attitudes are transforming yet again, whereby more Americans are most likely to express racial attitudes that are best characterized by “color-­ blind” racial ideology. Ruth Frankenberg explains that a color-­blind perspective is a “mode of thinking about race organized around an effort not to ‘see’ or at any rate not to acknowledge race differences,” as this is the “‘polite’ language of race” (1993, 142). An emerging consensus among scholars, especially sociologists and critical race theorists, is that “color-­blindness” is currently the basis for contemporary American racial ideology and charac­

Nature of the Game  /  29

terizes the environment in which young Whites have been socialized (Bonilla-­Silva 2014; Bonilla-­Silva, Lewis, and Embrick 2004; Carr 1997; Forman 2004; Frankenberg 1993). Eduardo Bonilla-­Silva and Amanda Lewis (1999, 56) outline the elements that make up a color-­blind racial structure: “(1) The increasingly covert nature of racial discourse and practices; (2) the avoidance of racial terminology and the ever growing claim by Whites that they experience ‘reverse racism’; (3) the elaboration of a racial agenda over political matters that eschews direct racial references; (4) the invisibility of most mechanisms that reproduce racial inequality; and finally, (5) the rearticulation of a number of racial practices characteristic of the Jim Crow period of race relations.” At the most basic level, this means that people rely on so-­called politically correct modes of expression, where the words “race” and “racism” are completely avoided, deleted from vocabularies as well as from the list of potential factors that explain inequities between racial groups. Further, color-­blind ideology relies on abstract liberalism, which allows people to rationalize unfairness in the name of equal opportunity. In other words, it allows Whites to suggest that they support equal opportunity without also having a concern for the inequalities that exist between Whites and Blacks (Bonilla-­Silva 2014). The next step in this logic is that when policies like race-­conscious admissions are employed to reduce inequalities, these policies are then cast as “reverse discrimination.” Finally, this logic allows people to explain away persistent racial inequalities—­such as wealth and income inequalities; disparate poverty, incarceration, and infant-­mortality rates; and other barriers that constrain opportunities from being distributed more evenly across racial groups—­as the by-­products of something other than structural racism. Instead, people choose to rely on ideas about “class” differences or Black “culture” or the notion that some of the racial phenomena we see are just “natural” (e.g., residential segregation occurs because, as the adage goes, birds of a feather flock together) in order to explain ongoing racial inequity. This represents a different kind of logic than what most political scientists are attuned to or what previous scholars have tried to capture in survey measures. Taken together, the extant literature suggests that not only do the expressions of racial attitudes change over time but also the ideological nature of contemporary racial attitudes is probably best understood as a mixture of old and new. Old attitudes like Jim Crow racism die slow deaths, but during their decline, new attitudes are mounted on top, layered like an onion, as represented in figure 1.1. This evolution occurs in response to major shifts in society, such as massive shocks to the national economy, implementation

This symbolic racial attitude developed in response to Blacks’ ongoing demands for racial equality in the mid- to late-1960s, sometimes through riots and unapologetically Black freedom movements. Racial Resentment is a blend of anti-black animus and a sense that Black Americans are not living up to the opportunities provided to them by the policies that were developed in response to the Civil Rights Movement. Levels of racial resentment have been quite stable.

Post-Civil Rights Era through Present Day

Racial Resentment

Colorblind racial ideology requires individuals to ignore the structural basis of racial disparities, rely on non-racial explanations to explain racial phenomena, and deny the role of racial privilege in shaping Whites’ opportunity structure. It represents a shift in language and logic of previous dominant racial ideologies and expressions of racism, largely by avoiding talk of race and racism altogether. The absence of a structural explanation of racial inequality leads some to rely on logic that mimics Jim Crown Racism, such as Black “cultural” inferiority.

Late 20th Century through Present Day

Colorblind Racial Ideology

1.1.  The Evolution of Racial Attitudes in the United States

Jim Crow, or Old Fashioned Racism, is rooted in the notion that Whites are (culturally, intellectually, and genetically) superior to Blacks, the desire to segregate racial groups, and a preference for racially discriminatory policies. This bundle of ideas developed during the era of slavery in the United States, and though declining in prominence over time, it is still detectable in some individuals’ attitudes, and serves as the foundation of contemporary racial attitudes in the U.S.

Antebellum Era through Present Day

Jim Crow Racism

Nature of the Game  /  31

of landmark legislation and Supreme Court rulings, shifts in demographics, and potential generational replacement—­all of which we have seen in the past few decades.

Why Racial Stasis Now? Thus far, we have focused on change. There is always the potential for change and progress, but neither are inevitable. There are two major mechanisms of social change, both of which we consider and test for throughout this book: period effects and cohort effects. Yang et al. explain: “Period and cohort effects reflect the influence of social forces. Period variations often result from shifts in social, historical, and cultural environments. Cohort variations are conceived as the essence of social change and may reflect the effects of early life exposure to socioeconomic, behavioral, and environmental factors that act persistently over time to produce differences in life course outcomes for specific cohorts” (2008, 1698, emphasis added). Or in other words, period effects reflect the shifts in the ways in which the population, as a whole, manages particular attitudes and behaviors over time. For example, since World War II, the American population, on average, seems to hold increasingly liberal attitudes toward all sorts of groups—­racial and ethnic minorities, women, and gay and lesbian Americans (Brooks 2000; Schuman et al. 1997). An explanation solely based on “period effects” would point to some of the macro-­level changes that have influenced White Americans to shift their attitudes toward minorities over time, such as changing social norms, transformations in the economy and the law, or the success of massive social movements. Specific examples are seen in the facts that overt expressions of racial animus have become less socially acceptable; marriage equality was deemed constitutional by the Supreme Court; more Whites are well-­represented by minority representatives; women began to publicly fight back against sexual harassment in the workplace; public and private institutions celebrate (and even rely on) racial diversity; people watched or directly experienced terrorist attacks on the World Trade Center; the American electorate selected its first non-­White president in 2008; and a series of police shootings of unarmed Black people circulated via social media and mainstream American news outlets. These are large-­scale events that most Americans are well aware of, and thus, we might expect to see aggregate-­ level shifts in attitudes as a result. On the other hand, cohort effects describe the changes that occur due to the replacement of older Americans, who were socialized to value and believe certain things, by new generations, who have their own shared and

32 / Chapter One

unique set of experiences, attitudes, and worldview. For instance, it would be unsurprising for many Americans to hear a member of the silent generation or an older baby boomer express a sentiment like Cliven Bundy’s, the cattle rancher who opposes the federal government and came to fame when armed protesters answered his rally cry for the “sovereign citizen” after a twenty-­year dispute with the federal government came to a head. Bundy was using public land for his cattle to graze; he was later “pardoned” by President Donald J. Trump. In April 2014, Bundy said, “They [‘Negroes’] abort their young children, they put their young men in jail, because they never learned how to pick cotton. And I’ve often wondered, are they better off as slaves, picking cotton and having a family life and doing things, or are they better off under government subsidy? They didn’t get no more freedom. They got less freedom” (Nagourney 2014). While most of us would characterize a comment like this as racist, it is likely that the speaker would disagree. Indeed, even though President Trump has on a number of occasions castigated whole groups of people through rhetoric and policy, he has claimed that he is not racist. In early 2018, after he referred to Haiti, El Salvador, and many African nations as “shithole” countries, the White House released a statement noting that he is “one of the least racist people” anyone in the press had ever interviewed (Scott 2018). Given the cognitive gymnastics required to reach their conclusions, it may be easy to believe that these folks are delusional, but a more charitable reading is one that considers the fact that these individuals were socialized during a time when White supremacy was, for the most part, an unquestioned norm. In contrast, many people found themselves dumbfounded when photos of young White men chanting racist, anti-­Semitic, and homophobic slurs and slogans at the Unite the Right rally were released out of Charlottesville, Virginia, in August of 2017. The rally was organized by a millennial, Jason Kessler, and his Gen X counterpart, Richard Spencer, in an effort to gather White nationalists, bring them out of the shadows of the internet, and pro­­ test the removal of a Confederate statue. It led to the injury of dozens and to the murder of counterprotester Heather Heyer by White millennial and neo-­Nazi James Alex Fields Jr. How could young people do such terrible things, and why do there seem to be so many more young White nationalists than we thought? Cohort or generational replacement certainly has the potential to shift the average American’s opinion toward racial progressivism, but would that alone explain what Americans have recently been witnessing? Scholarship on cohort replacement tends to imply that cohort replace­ ment will continually produce more liberal racial attitudes among new co­

Nature of the Game  /  33

horts of White Americans as time goes on, but there are several key factors that motivate us to hypothesize that the United States is in a state of racial stasis and to further argue that White millennials are not helping to produce real racial progress, particularly at the structural level. First, we have to consider today’s racial landscape. Karl Mannheim’s (1952) theory of generations suggests that specific historical events can potentially shape the character of a generation. Those who came of age during and directly after the civil rights movement have drastically more favorable racial attitudes than their predecessors, in part because of the specific historical moment in which they grew up. The civil rights movement shed light on the horrors of unrestrained White supremacy, racial terrorism, and the physical and psychological violence of Jim Crow. There was a significant shift toward more equitable racial norms in White Americans’ values, due to the defeat of de jure Jim Crow policies (which deemed discrimination not just morally abhorrent but also illegal) and the rejection of biological racism. Though this shift in values did not necessarily lead Whites to more aggressively support and pursue policies that would bring racial equity to fruition, it is undeniable that there was a major transformation in America’s racial landscape. Nonetheless, the overt, hard-­hitting factors that jolted boomers to be more aware of racial inequity (and potentially rally against inegalitarianism) have dissipated. One of the outcomes of this shift in norms was a turn toward more subtle forms and expressions of racial animus. For instance, both John Ehrlichman, who served as domestic policy chief for President Richard Nixon, and Lee Atwater, a strategist for the Republican Party, explained that by the late 1960s, any successful politician had to speak a new language of race. First, Ehrlichman: The Nixon campaign in 1968, and the Nixon White House after that, had two enemies: the antiwar left and black people. You understand what I’m saying? We knew we couldn’t make it illegal to be either against the war or black, but by getting the public to associate the hippies with marijuana and blacks with heroin, and then criminalizing both heavily, we could disrupt those communities. We could arrest their leaders, raid their homes, break up their meetings, and vilify them night after night on the evening news. Did we know we were lying about the drugs? Of course we did. (Quoted by Baum 2016)

In a similar vein, Atwater explained the rhetoric of the “Southern strategy”—­ the effort to convert Southern, White Democrats to the Republican Party—­ in this way:

34 / Chapter One You start out in 1954 by saying, “Nigger, nigger, nigger.” By 1968 you can’t say “nigger”—­that hurts you, backfires. So you say stuff like, uh, forced busing, states’ rights, and all that stuff, and you’re getting so abstract. Now, you’re talking about cutting taxes, and all these things you’re talking about are totally economic things and a byproduct of them is, blacks get hurt worse than whites. . . . “We want to cut this,” is much more abstract than even the busing thing, uh, and a hell of a lot more abstract than “Nigger, nigger.” (Perlstein 2012)

The language of racism shifted to a subtle and covert form after the civil rights movement, perhaps making it more insidious. How so? Injustice thrives when the illusion of justice is perfected (Chiasson 2014). Outright racial discrimination was deemed illegal, and the US Supreme Court clearly stated that de jure segregation was unconstitutional, but Black Americans’ ability to attain the American dream remained highly constrained; they had more difficulty in fighting against de facto discrimination and structural racism, and White Americans’ racial ideology shifted from old-­fashioned racism to a more symbolic racism (Kinder and Sears 1981). Perhaps conse­ quently, those who came of age during the Reagan presidency developed more conservative racial attitudes (Wilson 1996). Later, during the Clinton years, Democrats lost their sense of racial progressivism, best illustrated by the 1994 crime bill and the 1996 welfare-­reform legislation, both of which put Blacks and other people of color (as well as poor Whites) at an even greater disadvantage (Taylor 2016). Later, in 2008, the election of Barack Obama was largely viewed through a color-­blind lens and as evidence that Americans now hold abstract anti-­racist, progressive values. It appears that over time the urgency to rid America of its racial disparities has declined precipitously, in part because it is difficult to fight what you cannot easily see, or what you do not want to see. To be sure, we are not making any causal claims, but the abovementioned timeline aligns closely with the evidence provided by a growing literature that there are signs that racial progress due to generational cohort replacement is slowing down and may even have come to a halt. For example, Wilson (1996) found that cohorts born after World War II were less prejudiced than those born prior to the war, but those born between 1961 and 1972 were not significantly different from those born between 1946 and 1960. Additionally, Andolina and Mayer (2003) found that Generation Xers born between 1967 and 1974 were more liberal than their predecessors in their support for school integration, but they were just as conservative in their attitudes concerning affirmative action and equal employment opportunities as those in previous generations. In an effort to update the

Nature of the Game  /  35

literature, Scott Blinder (2007) found that White Americans born between 1975 and 1986 were not significantly more progressive in their racialized policy preferences, and their policy attitudes were influenced by negative racial stereotypes. What’s more, scholars like Tyrone Forman (2004) find that young Whites are increasingly indifferent and apathetic to racial inequality. The evidence we produce in the remaining chapters of Racial Stasis will extend this literature. The contemporary period presents at least two possibilities for the millennial generation. The first centers on the notion that since members of the millennial generation have grown up in an America that celebrates various forms of racial and ethnic pluralism—­integration, diversity, and multiculturalism—­ and preaches anti-­racism, the youngest Americans could be the necessary catalyst to reenergize America’s liberalizing trends in terms of racial attitudes. A second possibility is that the contemporary environment could influ­ ence younger Americans to be just as conservative as, or perhaps even more conservative than, previous generations. Howard Schuman and his colleagues explain that the youngest Americans in their sample “have no firsthand knowledge of the events that occurred in the 1950s and early 1960s” and thus “can hardly feel the outrage and sense of urgency that swept through the entire nation” (Schuman et al. 1997, 202). At the time, they were writing about Generation X; this is an understatement for the millennial generation. A majority of young Americans do not learn about America’s racial history (Dillon 2011; Southern Poverty Law Center 2011). As a result, it may be the case that millennials do not believe or even have the language to begin to consider that racism plays a structural role in society. New expressions of racism like “racial resentment” and “symbolic racism” are, in part, based on the idea that the civil rights movement and the legislation that followed leveled the playing field (Henry and Sears 2002), so the youngest cohort group may be more likely to believe that a level playing field actually exists, as its members are farther distanced from a time of overt racial acrimony in comparison to previous generations. The more that people perceive equality, the less likely they are to support policies aimed toward maintaining egalitarianism. Valentino and Brader (2011), for example, reveal that after Barack Obama’s first election, people believed that there was less discrimination in the United States. Additionally, many of the members of the millennial generation voted for Obama, who is an excellent representation of the Horatio Alger myth. If President Obama could be successful despite systemic adversity, why can’t all Black people do it? In comparison to previous generations, millennials may experience a greater

36 / Chapter One

disconnect between what they see empirically and their understanding of the covert, structural reasons for disparities across racial groups. Additionally, millennials are the most diverse generation in American history. Millennials have been taught to celebrate diversity, though largely in an abstract, at-­arm’s-­length kind of way (Smith and Mayorga-­Gallo 2017). While White millennials have been trained to value racial diversity, they are also keenly aware that the imminent shift in demographics means that they will compete with minorities in ways that their predecessors never did, thereby potentially increasing feelings of racial threat. Historically, scholarship has revealed that Whites feel a greater sense of threat to Whites’ relative societal status as the number of people of color increases (Blaumer 1972; Blalock 1956), and still today, several scholars predict that “the increasing diversity of the nation may actually yield more intergroup hostility” (Craig and Richeson 2014a, 750). As White Americans consider the idea that the United States is on a trajectory to become a majority-­minority country, they tend to feel a greater sense of racial animosity and racial anxiety and express more conservative policy preferences, especially around those policies that may reduce racial inequity (Craig, Rucker, and Richeson 2018). Additionally, the racial socialization of White Americans is one that leaves much to be desired. Blinder (2007) theorizes that young Whites have or will experience a “two-­tracked socialization” process, where parents discuss abstract anti-­racist principles but do not directly address the onslaught of negative images of Blacks or the patterns of racially disparate outcomes in terms of systemic racism. Sociologists Amanda Lewis (2001) and Megan Underhill (2018) find that White parents consider themselves to be “good” parents if they keep their White children protected from information about racism, often avoiding the conversations that Black parents must have with their children to keep them safe. Needless to say, there may be an intergenerational transfer of a new, color-­blind racial ideology. Overall, the literature bends toward the notion that liberalizing trends in racial attitudes among Americans due to cohort replacement are unlikely. The extant scholarship on cohort replacement and racial attitudes leads us to expect White millennials, those born between 1981 and 1996, to be just as, if not more, racially conservative as their predecessors. But, it also suggests that we should not necessarily expect millennials to use the same racial grammar or rely on the same racial ideologies as their predecessors, or at least not to the same extent. Though we have seen millennials help move the bar in other domains of American life—­technology, LGBTQ equality, greater awareness of sexual assault and misconduct, and even gun policy—­it would not surprise us to find that millennials’ reliance on today’s dominant

Nature of the Game  /  37

racial ideology, color-­blindness, has slowed progress and contributed to racial stagnation.

Studying racial attitudes over time (and across generations) is a difficult task due to how the level, nature, and structure of racial attitudes change over time. In our quest to pin down a more accurate depiction of contemporary trends in racial attitudes and policy preferences, we take the time to update the literature by analyzing whether there have been recent changes in Americans’ racial attitudes, positive or otherwise. Our racial stasis hypothesis leads us to expect young White Americans to push aggregate opinions to the left in some domains of American life and policy but not necessarily on matters of race; chapter 2 provides the first of many tests of this hypothesis.

Two

Is Race Special?

There is little doubt in the minds of most Americans that over the past five or six decades, there has been a significant increase in tolerance for underrepresented groups. In addition to the transformation in the nature and expression of racial attitudes, attitudes toward gay and lesbian community members have drifted toward the political left (Loftus 2001; Yang 1997). What’s more, attitudes toward women’s rights, particularly as they relate to abortion and gender equality, have become more liberal since the 1960s as well (Cutler et al. 1980; Granberg and Granberg 1980; Thornton and Young-DeMarco 2004). In this chapter, we ascertain whether Americans have consistently held more progressive attitudes toward various types of minorities. Additionally, we examine underlying mechanisms of social change, with a special focus on the extent to which cohort replacement is doing the lion’s share of work to push the United States toward its better angels. We are aware that “liberalization in one area of public policy may coexist with equally striking instances of opinion stability” in other policy domains (Yang 1997, 477), and our findings in this chapter provide empirical evidence for this. Our results reveal that although Whites’ attitudes are liberalizing toward gays, lesbians, and women, trends toward increasing acceptance are exclusive to some minority groups while leaving others behind. Despite prima facie signs of racial progress, there exists a great deal of evidence that negative racial animus still lingers in American society. A robust body of scholarship shows that racial animus in various forms still influences factors that range from individuals’ employment prospects (Pager 2007; Bertrand and Mullainathan 2004) to criminal-­justice outcomes (Alexander 2010; Baumgartner, Epp, and Shoub 2018) to the ways in which social-­welfare funds are distributed (Soss, Fording, and Schram 2008;

40 / Chapter Two

DeSante 2013). Whether politicians will help an unknown constituent register to vote is also influenced by the race of the constituent (Butler and Broockman 2011). Research developed in the wake of Barack Obama’s presidential elections illuminates the critical role anti-­Black sentiment played in those elections (Andersen and Junn 2009; Block Jr. 2011; Hutchings 2009; Tesler and Sears 2010). Still more research reveals that Donald Trump’s election elicited virulent (and sometimes overt) racial animus and benefited from the racial antipathy or racial apathy of millions of White voters (Pettigrew 2017; Sides, Tesler, and Vavreck 2017; DeSante and Smith 2017). Over time, a “theme of increasing individualization, freedom, and tolerance of diversity of ideas and behaviors” has been important in transforming attitudes toward various minority groups and policies aimed to protect all sorts of groups (Thornton and Young-DeMarco 2004, 1010), but we are curious to know whether these ideas have been equally employed when Americans consider their attitudes toward Black Americans, gays and lesbians, and women. There are conflicting indicators concerning whether liberalizing trends toward various minorities, especially toward African Americans, are similarly even, consistent, and persistent. Figure 2.1 elucidates the issue of uneven progress quite clearly and helps further illustrate our argument that America has entered into a period of racial stasis. Responses to twenty policy questions are organized across four rows. The top row focuses on women, the second on gays and lesbians, the third on general spending preferences, and the bottom row depicts trends on policies that are either implicitly or explicitly racialized. The lines show the proportion of survey respondents who gave a liberal response (e.g., support affirmative action, support gays and lesbians, and favor the spending items). The feeling thermometer measures represent the difference in the scores that respondents gave to Whites and to Blacks and is rescaled to run from 0 to 1. What is made evident is that White Americans’ contemporary policy attitudes have largely trended to meet this society’s egalitarian principles as they concern gender roles, workplace protection for women and for lesbian and gay Americans, aid to the poor, education, the environment, healthcare, and social security. But, these trends toward equity do not extend to matters concerning Black Americans nor to policies that are most closely related to Blacks in the White American imagination, namely “welfare” (Gilens 1999). There is a stark contrast between the steady march toward liberalized policy positions on a wide range of policy domains and the plateaued trends and relatively conservative attitudes concerning Black Americans. While liberalism is a difficult concept to succinctly describe, particularly because “liberalism is a chimera that has changed its emphasis and even

Proportion of responses which are liberal

0.0

0.5

1.0

0.0

0.5

1.0

0.0

0.5

1.0

0.0

0.5

1.0

1976 1984 1992 2000 2008 2016

Government Ensure Fair Employment for Blacks

Spending on Education

Homosexual Sex is Always Acceptable

Support Equal Roles for Men and Women

Year of Survey

1976 1984 1992 2000 2008 2016

Government should help Blacks

Spending on the Environment

Support Gays and Lesbians Adopting Children

Women Belong in Politics

1976 1984 1992 2000 2008 2016

Spending on Welfare

Spending on Healthcare

Support Gays and Lesbians Serving in the Military

Women Belong in the Workforce

2.1. How Millennials and Older Whites Prioritize Racial Issues

Sources: American National Election Studies Cumulative Data Set (1970−2016) General Social Survey Cumulative Data Set (1972−2016) White Respondents Only

1976 1984 1992 2000 2008 2016

Support Affirmative Action for Blacks

Spending on Aid to Poor

Feeling Thermometer: Gays and Lesbians

Support Affirmative Action for Women

How the 'Liberalization' of Racial Attitudes Lags Behind Other Domains

1976 1984 1992 2000 2008 2016

Feeling Thermometer Difference: White−Black

Spending on Social Security

Support Workplace Protections for Gays and Lesbians

Abortion Should Always Be an Option

42 / Chapter Two

some key tenants over time,” (Smith 1990, 481), we follow Tom W. Smith, who describes liberalism as (1) reformist, opting for change and generally opposed to the status quo; (2) democratic, favoring a full extension of electoral rights; (3) libertarian, supporting civil liberties such as free speech and the right to protest; (4) reg­ ulatory and interventionist, backing the management of business and the economy by the government; (5) centralist, using the federal government to set and enforce national standards and regulate state and local governments; (6) humanitarian, favoring a social welfare system for the care and protection of society in general and the lower class in particular; (7) egalitarian, advocating equal treatment for all and perhaps equal conditions for all; and (8) permissive, tolerating and often approving nontraditional life styles and practices. (1990, 481)

Simply put, we use the term “liberalization” to mean drifting to the political left or as a synonym for progress, specifically in relation to attitudes about the groups of concern in this chapter. In light of the conflicting trends we’ve seen thus far, we take a deeper dive into a wide array of data in order to address the following questions: To what extent have attitudes toward Blacks, members of lesbian and gay communities, and women changed over time? Are attitudes toward some minority groups warming while others are not? Have Whites’ racial attitudes reached a point at which they are unlikely to further improve? Finally, to what extent are young people leading the charge in changing the contours of Americans’ attitudes about various minority groups?

The Potential of Young People Cohort replacement is a central component in many scholars’ explanations of the increasing prevalence of liberal attitudes in American society (Dowden and Robinson 1993; Firebaugh and Davis 1988; Hochschild, Weaver, and Burch 2012; Ryder 1965; Steeh and Schuman 1992; Wilson 1996). In the case of racial attitudes, scholars have suggested that cohort replacement leads to a process of “constant liberalization that occurs as older, less tolerant generations are replaced by new generations that have been socialized to be less hostile than their parents to racial integration” (Steeh and Schuman 1992, 341). The basis for such an argument centers on the idea that “individuals who belong to the same generation, who share the same year of birth, are endowed, to that extent, with a common location in the historical

Is Race Special?  /  43

dimension of the social process” (Mannheim 1952, 290). Each generational cohort experiences unique and distinct socialization processes. This may lead us to expect cohort-­based changes in worldview, attitudes, and behaviors (Schwadel 2011; Schwadel and Stout 2012), as based on “the potency of historically linked socialization experiences that produce different orientations as a result of cohort membership” (Duane and Scott 1996, 82). Research that emphasizes cohort replacement’s role in declining racial prejudice is convincing, but works that home in on cohort replacement have traditionally been unable to disentangle cohort effects from period and age effects (Dowden and Robinson 1993; Hochschild, Weaver, and Burch 2012; Steeh and Schuman 1992). Similarly, scholarship that focuses on macro changes across time fails to discern period effects from cohort and age effects (Hunt 2007; Kluegel 1990; Schuman et al. 1997). To be clear, age effects are defined by “variation associated with different age groups” (Yang et al. 2008, 1697). It is often assumed that as individuals get older, they are more likely to become more conservative, and there are some psychological reasons for this. As Cutler et al. note, there are “psychologically-­based, age-­related changes in the direction of greater rigidity, cautiousness, and growing resistance to change,” all of which lead to “a greater stake in the maintenance of the status quo” (1980, 115). Period effects refer to “contemporary circumstances and events that press upon the public consciousness in a manner that changes opinion for every­ one, not just for a particular demographic or political group” (Tate 2010, 31). In other words, period effects describe variation over time periods that affects the entire population simultaneously; signs of period effects appear when a society is affected by some phenomenon that is shared by all, shaping the collective memory. Major events such as a war, a transformative or unexpected presidential election, a shock to the economy, or the implementation of landmark legislation are experienced by all members of society and thus may shift the average person’s response to closely related issues. For instance, though young people and older people may have interpreted the meaning of the attacks of September 11, 2001, in the United States differently, the images of the falling Twin Towers and the loss of thousands of Americans’ lives washed over everyone; this was a shared experience that shaped all Americans’ collective memory and imagination. Cohort effects, in contrast, reflect “changes across groups of individuals who experience an initial event such as birth in the same year or years” (Yang et al. 2008, 1697). Perhaps the most dominant example of cohort ef­ fects involves the use (and abuse) of technology and social media. Millen­ nials are essentially “digital natives” and thus have a unique relationship

44 / Chapter Two

with their mobile devices, for instance, which simultaneously act as their phone, television, camera, mode of social connection, sleep companion, and coping mechanism. Post-­millennials are even more entrenched in their use of technology, while boomers may still get phone calls to landlines. Societal changes in attitudes toward various minority groups and in policy attitudes are often explained—­implicitly or explicitly—­by cohort effects (Andolina and Mayer 2003; Astin 1998; Cutler et al. 1980; Firebaugh and Davis 1988; Loftus 2001; Steeh and Schuman 1992; Wilson 1996). However, until recently, it has been difficult for researchers to disentangle these three effects—­age, period, and cohort—­from one another due to limitations in methodological tools and strategies; “age, period and cohort measures cannot simultaneously be included in a standard regression model due to linear dependency,” and it should be noted that analysis of “cohort effects are unreliable without including age in the model” (Schwadel and Stout 2012, 238). In addition to answering the questions we have presented, this chapter also addresses these methodological problems by utilizing an improved measurement strategy, the age-­period-­cohort intrinsic estimator (APC-­IE); we thoroughly explain this method in appendix A.

Trends in Attitudes Research suggests that Americans have become increasingly liberal, especially as their ideas relate to their opinions on civil liberties (Brooks 2000), Black Americans (Firebaugh and Davis 1988; Schuman et al. 1997), women (Thornton and Young-DeMarco 2004), and those who identify as lesbian or gay (Hicks and Lee 2006; Loftus 2001; Yang 1997; Kreitzer, Hamilton, and Tolbert 2014), but these attitudinal trends have not necessarily been mono­ tonically improving. Opinions toward these groups appear to wax and wane; progressive attitudes toward various minorities have fits and starts, and attitudes liberalize at faster rates toward some minorities than they do toward other groups. Attitudes toward Women, Gender Equality, and Abortion Arland Thornton and Linda Young-­DeMarco note, “A strong social movement from the improvement of women’s rights emerged in the last half of the 19th century and continued through the 1920s, became dormant after the achievement of women’s suffrage, and then reappeared with increased energy in the second half of the 20th century” (2004, 1010). American survey respondents have been asked on several occasions whether they believe

Is Race Special?  /  45

“the activities of married women are best confined to the home and family.” Between 1967 and 1997, the proportion of American college students—­ both men and women—­who answered in the affirmative declined precipitously; the number of people with this strongly held value was cut in half in thirty years’ time (Astin 1998, 120). Additionally, scholars show that approval of legal abortion increased alongside support for women’s rights (Granberg and Granberg 1980). Between 1965 and 1972, there was a major increase in favorable attitudes toward abortion rights; then, after the Supreme Court’s decision in Roe v. Wade was publicized in January 1973, attitudes toward abortion became considerably more liberal until 1974, when attitudes began to plateau (Cutler et al. 1980; Granberg and Granberg 1980; Saad 2009). More recently, studies show that over the past two decades there have been more Americans identifying as “pro-­life” compared to “pro-­choice,” but the proportion of the population that supports some level of abortion rights (i.e., abortion made legal under any circumstance or legal under certain circumstances) outnumbers those who say abortion should be illegal (Saad 2009). Overall, with each passing year, Americans have increasingly favored abortion rights. Concerning age, period, and cohort effects, scholars find, in general, attitudes toward abortion have become more favorable among all cohorts; that is to say, there has been a societal shift toward approving of abortions in a variety of circumstances. Stephen Cutler and his colleagues found that all birth cohorts, including the very oldest cohort, have become more accepting of abortion since the 1950s, and that older individuals are just as likely to change their attitudes in a liberal direction as young people are, though at a slower rate (Cutler et al. 1980). Attitudes toward Gays and Lesbians In 1965, 70 percent of the population believed that gays and lesbians are “more harmful than helpful to American life” (Hicks and Lee 2006, 59). During the late 1980s, “discrimination on the basis of sexual orientation in employment, housing, and public accommodations remain[ed] legal throughout the United States, except in Wisconsin and a few dozen municipalities that enacted protective legislation” (Herek 1988, 452). Since then, there has been significant improvement in public opinion toward gays and lesbians, and there has been more acceptance of members of lesbian and gay communities, perhaps evidenced by the increasing number of elected officials who are openly gay. Additionally, there have been important but limited policy changes toward gay and lesbian citizens, including the 2011

46 / Chapter Two

repeal of the military’s 1993 “Don’t Ask, Don’t Tell” policy, and states that enacted marriage-­equality laws despite the federal government’s stance on the matter (exemplified by the Defense of Marriage Act). These changes were followed up relatively quickly by the 2015 Supreme Court ruling in Obergefell v. Hodges, which deemed marriage as a fundamental right guaranteed to same-­sex couples.1 Between 1977 and 1990, public opinion became more liberal toward gays and lesbians, but according to some scholars, these attitudes plateaued after 1990 (Hicks and Lee 2006). More liberal attitudes have developed toward this group because of such factors as “people have become less religious, more people believe that homosexuality is innate rather than a choice, the advancement of civil rights and liberties, more gays and lesbians are out of the closet, and people have become more liberal in general” (Hicks and Lee 2006, 61). Further, the number of people reporting ties to someone who is gay or lesbian (e.g., an acquaintance, close friend, or family member) has increased over time, and research shows that personal contact with gays and lesbians improves attitudes toward these groups (Hicks and Lee 2006; Yang 1997). Those with higher levels of education also tend to be more liberal in their attitudes toward members of this minority group (Astin 1998; Hicks and Lee 2006; Loftus 2001). Nevertheless, there seem to be two faces of tolerance; that is to say, Americans’ support for the rights and protections of gays and lesbians have increased considerably even though many individuals still dislike members of this minority group. Jeni Loftus (2001) asserts that Americans distinguish between their attitudes toward the morality of homosexuality and the civil liberties of gays and lesbians. She finds that “Americans’ attitudes regarding the morality of homosexuality became slightly more liberal from 1973 to 1976, became increasingly conservative through 1990, and have become more liberal since 1990. Over the same 25-­year period, willingness to restrict the civil liberties of homosexuals declined steadily, the only departure being a brief increase in negative attitudes” between 1985 and 1987 (Loftus 2001, 778). Overall, Americans have been willing to cede more civil-­rights protections for gays and lesbians, but even this is limited to certain types of policies (Yang 1997; Lax and Phillips 2009). Finally, we see that tolerance toward this group has increased over time, but what this ultimately has meant is that even though attitudes toward protecting members of this minority group have become strikingly more liberal, affective attitudes toward members of this group may not be improving at the same rate (Brooks 2000; Loftus 2001; Yang 1997). Yang (1997)

Is Race Special?  /  47

shows that while “feeling thermometer” ratings toward gays and lesbians have improved over time, the reported attitudes toward members of lesbian and gay communities are still among the lowest among most social groups. Additionally, Brooks (2000) reports that there has been little change in beliefs about the acceptability of homosexuality, but there has been an increase in support for gays and lesbians to be treated fairly in the realms of housing and employment. Some scholars offer cohort replacement as a potential causal mechanism for attitudinal change, as scholars show that younger people tend to be more liberal on issues of gay rights (Astin 1998; Loftus 2001; Yang 1997). Period effects are also a potential contender, as it seems that the American population’s attitudes, in general, appear to be evolving on issues of civil liberties for gays and lesbians. Attitudes toward Black Americans and Racialized Policies While scholars have shown that racial prejudice and anti-­Black affect continue to influence the political attitudes and behaviors of Whites in the United States, one key finding is that, in many ways, America is in a better position today than it was several decades ago, because these negative attitudes have been on the decline since the second half of the twentieth century (Schuman et al. 1997). According to a number of scholars, one of the major factors of this decline is cohort replacement (Andolina and Mayer 2003; Firebaugh and Davis 1988; Schuman et al. 1997; Steeh and Schuman 1992). Firebaugh and Davis (1988) find some evidence that a decline in prejudice stems from individuals changing their attitudes over time, but they conclude that the most important source of the decrease in negative racial attitudes, on an aggregate level, stems from cohort replacement. The link between cohort replacement and a decline in prejudice comes from the idea that older, more prejudiced birth cohorts are replaced by younger, less prejudiced cohorts. Anti-­Black attitudes and the desire for more social distance between Whites and Blacks are highly associated with year of birth. White Americans born in an earlier era are more likely to report more prejudiced attitudes than members of younger cohorts (Firebaugh and Davis 1988; Schuman et al. 1997; Tuch 1984; Wilson 1996). For the most part, scholars have been quite optimistic about the power and role of cohort replacement in liberalizing Whites’ racial attitudes. Firebaugh and Davis (1988), for example, found that even during times when Americans’ political attitudes were becoming more conservative on many social issues, racial attitudes were still becoming more liberal. Steeh and Schuman (1992,

48 / Chapter Two

340, 341) conclude, “there is no indication of decreasing tolerance among cohorts coming of age in the 1980s,” and “differences among cohorts contribute positively to the development of racial liberalism in the last half of the 20th century.” But these data are outdated by at least two decades, largely assessing the sentiments of people born before 1975. As we outlined in the previous chapter, there are a number of reasons to hypothesize that the introduction of White millennials’ attitudes into aggregate calculations of racial attitudes will not produce progressivism. Instead, due to the “now you see it, now you don’t,” characteristic of contemporary structural racism, the dominance of color-­blind racial ideology, and the fact that millennials are being socialized to disengage with issues of racial inequity, we hypothesize that in the domain of race and Black Americans, we will see that racial issues are uniquely in stasis.

Data and Methodological Strategy We employ data from the General Social Survey (GSS) between 1972 and 2012 and the American National Election Studies (ANES) between 1964 and 2012 to analyze age, period, and cohort effects on American attitudes toward various groups (Blacks, gays and lesbians, and women), policies that are targeted to help these groups, and, more generally, affective attitudes toward these groups. The GSS is a nationally representative survey of American citizens and has been conducted annually or biannually since 1972. Its samples include noninstitutionalized individuals who are over the age of eighteen; all of the surveys between 1972 and 2012 have been merged by the National Opinion Research Center into a single data file (Smith, Marsden, and Hout 2011). The ANES surveys a cross-­sectional equal-­probability sample of American adults. We use the ANES cumulative data file, which includes all questions that have been asked in three or more of the ANES time-­series studies. Like the GSS, the ANES has been conducted regularly, with the most common span between two time-­series studies being two years; the exception to this is the 2006 study, which was not a traditional time-­series study, and thus there is a gap in some of our data between 2004 and 2008. Since this chapter is concerned with whether attitudes toward gays, lesbians, African Americans, and women have improved over time, and the extent to which expected changes originate from age, period, or cohort (APC) effects, we use APC intrinsic estimator (IE) models (Yang, Fu, and Land 2004; Yang et al. 2008).

Is Race Special?  /  49

Measurement Strategy We examine Whites’ attitudes toward Blacks, members of lesbian and gay communities, and women’s rights over time. Specifically, we analyze three distinct types of attitudes for women and gays and lesbians, and four types of attitudes toward Black Americans.2 Affective Orientation. First, utilizing feeling-­thermometer scores, we track the general affective orientation using standard measures on how warmly a respondent feels toward each of our groups: gays and lesbians, women, and African Americans. Concerns for Equal Treatment. Second, we analyze global attitudes toward equality of opportunity and whether the APC trends for an overall measure of egalitarianism translate into increased support for policies that focus on equal treatment. Specifically, we look at support for laws that would protect gays and lesbians in the workplace and affirmative-­action programs for women and racial minorities. Group-­Based Policies. Third, we examine opinions on policies that are specifically aimed at helping one particular group: whether gays should be allowed to serve in the military or adopt children, whether abortion should be permitted, and whether the government should increase its effort to provide Blacks with aid. Personal Involvement. Finally, we inspect opinions on policies that would involve more than an “easy” commitment to racial egalitarian principles and ideals (e.g., attitudes toward laws banning interracial marriage) and instead scrutinize the extent to which Whites are willing to personally interact with Blacks (i.e., whether they would be willing to live in a predominantly Black neighborhood as well as whether they would oppose a member of their family marrying a Black person). We chose these policies and dimensions because they allow us to paint a richer picture about the nature of change in Whites’ policy attitudes. If the traditional explanation of cohort replacement is true, then we should see cohort effects for all policy types, across issues and targets. If, however, race is somehow categorically different from other policy domains, we will be able to see if racial progress has stagnated.

Results Given the nature of the APC-­IE analysis, the results in tabular form are difficult to interpret. For illustrative purposes, we provide an example of results in tabular form and graphic form for one of the GSS questions we

50 / Chapter Two

analyzed. Thereafter, we provide only figures of the predicted values/probabilities in the main text, with tables appearing in appendix D. Throughout this chapter (and in other chapters where we use this statistical method), we will primarily focus on those differences in APC trends that are highly unlikely to be the product of random chance (i.e., those results that are statistically significant at conventional levels). The GSS queried, “On the average (Negroes/Blacks/African-­Americans) have worse jobs, income, and housing than White people. Do you think these differences are because most (Negroes/Blacks/African-­Americans) have less in-­born ability to learn?” This is a question that serves to mea­ sure “old-­fashioned,” biological racism. This question was on the GSS from 1977 until 2016, and of the nearly twenty thousand Whites who answered it over thirty-­five years, about 15 percent agreed. Figure 2.2 illustrates a downward trend in affirmative responses to this question. We would like to note, though, that it took about fifteen years for the proportion of people who agree with this statement to be cut in half. If we were conservative in our calculation of a half-­life of this attitude, then we’d expect affirmative responses to have gotten to a low of about 5–­6 percent by 2007. Figure 2.2 shows that this level has yet to be reached as of 2016. Nonetheless, we want to discern the effects of period and cohort effects on this downward trend. We estimated a response function to this question using the APC-­IE. The resulting coefficient estimates for the logistic regression (where agreement was coded as 1 and disagreement as 0) are shown in table 2.1. Table 2.1 presents coefficients and standard errors for a total of forty-­ four different parameters. Since our dependent variable is either a 0 or 1, these coefficients can be interpreted similarly to standard logistic regression coefficients. A positively signed coefficient means that the probability of a respondent in that age/period/cohort group agreeing with the statement in question increases compared to the overall rate for all of the nineteen thousand White respondents in the model. Thus, the coefficients for each type of variable represent how far the average member in that group deviates from the overall sample mean. Negative coefficients represent a decrease in the predicted probability. However, given the dichotomous variable, the number of parameter estimates, and what amounts to the simultaneous control for the other types of effects, we illustrate our main findings graphically. Figure 2.3 is a panel of three figures showing the marginal effects and confidence intervals for each of the different coefficient types, broken down by age, period, and co­ hort. In each of the panels, the y-­axis represents the predicted probability of agreeing that differences between Whites and Blacks are due to an “inborn

2.2.  Proportion of  White Respondents Who Believe Racial Disparities Can Be Attributed to Black “Inborn Disability”

Fig 2.2 - DeSante and Smith

52 / Chapter Two Table 2.1.  APC-­IE: White-­Black SES Gap Is Due to “Inborn Disability” Age

Period

Cohort

Group

Coef. (SE)

Year of Survey

Coef. (SE)

Year of Birth

Coef. (SE)

Year of Birth

Coef. (SE)

18–­24

−0.485* (0.103) −0.616* (0.0970) −0.543* (0.0905) −0.430* (0.0843) −0.280* (0.0805) −0.292* (0.0818) −0.101 (0.0792) 0.0452 (0.0795) 0.184* (0.0777) 0.289* (0.0786) 0.417* (0.0825) 0.563* (0.0922) 0.560* (0.110) 0.689* (0.126)

1976–­1980

0.735* (0.0683) 0.497* (0.0701) 0.432* (0.0476) −0.0242 (0.0538) −0.230* (0.0522) −0.324* (0.0851) −0.473* (0.0635) −0.613* (0.119)

1895–­1899

0.517 (0.657) 0.156 (0.345) 0.0620 (0.200) 0.335* (0.162) 0.0972 (0.139) 0.437* (0.123) 0.263* (0.112) 0.212* (0.108) 0.111 (0.107) 0.0557 (0.102) −0.0406 (0.0973)

1950–­1954

−0.367* (0.0946) −0.443* (0.0925) −0.289* (0.0836) −0.200* (0.0873) −0.287* (0.0965) −0.311* (0.121) −0.347* (0.148) −0.609* (0.221) 0.186 (0.194) 0.462 (0.384)

25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­89 80–­84 85+

1981–­1985 1986–­1990 1991–­1995 1996–­2000 2001–­2005 2006–­2010 2011–­2014

1900–­1904 1905–­1910 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949

1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995+

Constant

−2.65* (0.05)

N

19,828

Data: General Social Survey (GSS) cumulative data set (1972–­2012), White respondents only *p < 0.05

disability.” The x-­axis in each figure represents the coefficient whose effect we are showing. The solid black line represents the point estimates, while the dashed lines, above and below, represent the 95 percent confidence intervals around the estimates. As you can see, the smallest cohort groups (the eldest and the youngest) have the smallest number of members represented in the survey, and thus they have the widest confidence intervals. The first panel of figure 2.3 illustrates the age effects shown in table 2.1. The oldest Whites, controlling for period and cohort, are about twice as likely to express a belief in old-­fashioned racism as the youngest Whites in the data are; that is to say, as Whites get older, we see them becoming more likely to express “old-­fashioned” racial animus, even when controlling for the year in which the question was asked (the period) and which

Is Race Special?  /  53

generational cohort they belong to. There is also evidence for period effects, which are shown in the second panel. Over time, the likelihood that any one individual answers this question in the affirmative decreases from 0.30 in the period from 1976 to 1980, to 0.10 during the period ending in 2016. This suggests that there is a marked shift within the population over time toward more liberal responses to this question, though the proportion of

2.3. Illustrated Age, Period, and Cohort Effects from Table 2.1

54 / Chapter Two

Whites who agree with the statement regarding “inborn disability” is still not as low as some might hope. Cohort effects perform slightly differently, however. One thing that is immediately apparent, as we noted above, is that the standard errors and resulting confidence intervals for the earliest and latest cohorts are much larger than the others; this is due to those cohorts having a smaller sample size. For example, while we have over 19,000 respondents in this model, only 434 of them fall into the cohorts containing birth years of 1985 or later (cohorts 1990 and 1995). Of these 434 respondents, 8 percent agreed with the GSS question, a percentage that is twice as high as the preceding cohort and at about the same level as Whites born in the late 1960s and early 1970s. The third panel in figure 2.3 shows what, at first glance, appears to be a cohort effect in the direction that is directly at odds with the traditional ex­ planation of liberalization of attitudes through cohort replacement; there is an upswing in overtly racist sentiment beginning with millennials in the 1985 cohort. As table 2.1 reveals, while the 1976–­1980 cohort is significantly less likely than the 1970–­1974 cohort to express support for a biological explanation of racial differences, the 1980–­1984 cohort is not significantly different from its previous cohort; similarly, the 1985–­1989 cohort is not significantly different from its preceding cohort. The results suggest that we see stasis in “biological” racism, starting with members of the millennial generation; or in other words, members of the millennial generation do not necessarily support the idea that Blacks are innately inferior, but they are not rejecting this belief at a higher rate than their predecessors. Overall, the data reveal that the change in attitudes surrounding biological racism is almost exclusively due to age and period effects. As White Americans age and move through the life cycle, they are more likely to express this kind of attitude, though over the time span we have analyzed, all Whites are less likely to agree with the question over time. We should note, however, that millennials are conspicuously absent here, while boomers and older Gen Xers are much less likely to agree with the notion of biological racism. In fact, we can see that the most recent cohorts have the same predicted probability of agreeing with that statement (that Blacks have an “inborn disability”) as Whites born in the early twentieth century. Affective Orientations Next, we examine affective orientation toward Blacks, gays and lesbians, and the “women’s liberation movement,” and we do so through the use of

Is Race Special?  /  55

feeling-­thermometer items from the American National Election Studies (ANES). Feeling thermometers range from 0 to 100, where 0 to 49 is considered “cold” or “cool,” while scores of 50 indicate neutrality, and those above 50 are considered warm. The affective-­orientation measure related to Blacks is actually a comparative evaluation of Blacks and Whites; we label this “pro-­White affect.” That is to say, we subtract the respondents’ ratings of Blacks from their ratings of Whites to get a measure for the difference in affective orientation between in-­and out-­groups; a positive score means that the respondents feel more warmly toward Whites than they do toward Blacks. We begin by examining age effects, illustrated in figure 2.4. As expected, for all three measures, when respondents are older, they give less progressive responses, though there is hardly any difference across age in attitudes toward the women’s liberation movement. The middle line in figure 2.4 illustrates attitudes as they concern lesbians and gays. When White Americans are young, they tend to have a higher overall affective rating for gays and lesbians (about 43 degrees) compared to when they get older (averaging just under 35 degrees). The line at the bottom of figure 2.4 reveals the difference in affect between respondents’ in-­group (Whites) and their out-­group (Blacks). Overall, Whites across age groups feel more warmly toward their in-­group; social identity theory would lead us to expect this (Tajfel and Turner 1986), but we see that there are differences in the degree to which people feel more warmly toward their in-­group as they age. While those aged eighteen to twenty-­four, on average, rate Whites about 11 degrees higher than Blacks, older Whites have a difference closer to 18 degrees; again, these are the independent effects of age when controlling for the effects of both period and cohort. The question about women’s liberation was only asked between 1970 and 2000. Meanwhile, the question about feelings toward gays and lesbians first appeared on the 1984 ANES, and the feeling-­thermometer ratings of Whites and Blacks have appeared on nearly every ANES from 1964 until 2012. Figure 2.5 depicts period effects, or the extent to which White Americans’ aggregate attitudes transformed over those distinct survey periods. Figure 2.5 reveals a general progression of Whites’ attitudes over time. These results show that Americans are embracing more liberal attitudes for two out of the three groups, as the ratings of gays and lesbians warmed between 1990 and 2010 and the ratings for women’s groups warmed from 1970 to 2000. Meanwhile, the difference in feeling-­thermometer scores between Whites and Blacks declined between 1960 and 2006, but affective bias

Affect

0

25

50

75

100

20−24

25−29

30−34

35−39

45−49

50−54

Age Group

40−44

55−59

60−64

65−69

70−74

75−89

Pro−White Affect

Gays and Lesbians

Women's Movement

Target

2.4.  Age Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women

Source: American National Election Studies Cumulative Data Set (1948−2016)

15−19

Age Effects on General Affect Towards Groups

Affect

0

25

50

75

100

1970−1974

1975−1979

1980−1984

1985−1989

Period

1990−1994

1995−1999

2000−2004

2005−2009

2010−2015

2016

2.5.  Period Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women

Source: American National Election Studies Cumulative Data Set (1948−2016)

1965−1969

Period Effects on General Affect Towards Groups

Pro−White Affect

Gays and Lesbians

Women's Movement

Target

58 / Chapter Two

toward Whites began to increase starting in 2006, with a constant increase through 2016. Figure 2.5 helps us to see that at the beginning of that series, the average difference in ratings between Whites and Blacks was about 25 de­­ grees, and it reached a low of just 3.1 degrees in the 2002 study. However, isolating the period effects shows us that recent periods have been met with a slight increase in in-­group favoritism among White Americans. While in 2002 the difference in sentiments between Whites and Blacks was only a few degrees, it has only been increasing: 4.3 degrees in 2004, 6.5 degrees in 2008, and just over 9 degrees in 2012, a level not seen since approximately 1992. That is to say, we are seeing a reverse in trends as they concern racial animus. Again, it is interesting to note that during periods in which most White Americans are warming to other groups (women, gays and lesbians), they are becoming cooler toward Blacks relative to Whites. The cohort effects, shown in figure 2.6, reveal that more recent cohorts are willing to express more liberal attitudes toward gays and lesbians and women’s groups in comparison to previous cohorts, but they are not significantly different from older cohorts in terms of their racial attitudes. These results suggest that while cohort replacement serves as a plausible explanation for mass attitude change concerning women and lesbian and gay Americans, there is little evidence that White millennials are offsetting their older counterparts’ more racially conservative attitudes. For example, the cohorts born around 1990, on average, rated Whites about 8.2 degrees warmer than Blacks; in order to match this difference closely in terms of magnitude, one would have to look all the way back to the 1955 or 1960 cohorts. To be sure, even those earlier cohorts are just slightly more racially progressive (differences of 7.9 and 7.8 degrees, respectively) than those born around 1990. Egalitarianism: Broad and Specific We turn to examine trends in concerns for equal treatment. Between 1984 and 2008, the American National Election Studies asked respondents whether they agree or disagree that “our society should do whatever is necessary to make sure that everyone has an equal opportunity to succeed.” This question is about the broad principle of equal opportunity, an idea that Americans tend to embrace. We treat this question as a dichotomous variable: respondents who agreed were coded as 1, and those who were either on the fence (neither agreed nor disagreed) or disagreed were coded as 0. Figure 2.7 depicts the APC-­IE results, which show that, without exception, support for this broad form of egalitarianism never dips below 75 percent, and it is generally closer to 90 percent. Americans’ commitment to a thin

Affect

Cohort Effects on General Affect Towards Groups

Pro−White Affect

Gays and Lesbians

Women's Movement

Target

2.6.  Cohort Effects on General Affect toward Blacks, Gay & Lesbian Americans, and Women

Source: American National Election Studies Cumulative Data Set (1948−2016)

Cohort

9 4 9 4 9 4 9 4 9 4 9 4 9 4 9 4 9 4 9 8 89 90 90 91 91 92 92 93 93 94 94 95 95 96 96 97 97 98 98 99 −1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5−1 0−1 5 9 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 18 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19

0

25

50

75

100

60 / Chapter Two

Probability of Support

Age Effects on Egalitarianism 1.00 0.75 0.50 0.25 0.00 15−19

20−24

25−29

30−34

35−39

40−44

45−49

50−54

55−59

60−64

65−69

70−74

75−89

Age Group Source: American National Election Studies Cumulative Data Set (1948−2016)

Probability of Support

Period Effects on Egalitarianism 1.00 0.75 0.50 0.25 0.00 1995−1999

2000−2004

2005−2009

2010−2015

2016

Period Source: American National Election Studies Cumulative Data Set (1948−2016)

Probability of Support

Cohort Effects on Egalitarianism 1.00 0.75 0.50 0.25 0.00 1915−1919

1925−1929

1935−1939

1945−1949

1955−1959

1965−1969

1975−1979

1985−1989

Cohort Source: American National Election Studies Cumulative Data Set (1948−2016)

2.7.  Age, Period, and Cohort Effects on Attitudes toward Abstract Egalitarianism

conception of egalitarianism has not changed much these three and a half decades, people don’t respond differently as they age, and each generation seems to throw its support behind this sentiment about equally. It is easy to laud a broad principle of equal opportunity. However, while Americans have been socialized to glorify principles of equality, scholars

Is Race Special?  /  61

have shown that there is often discord between White Americans’ value of egalitarianism and their willingness to support public policies that put those principles into action (Kinder and Sanders 1996). Needless to say, Whites are likely to be less enthusiastic about equality when specific targets are mentioned. Figure 2.8 illustrates the age, period, and cohort effects on the probability that a respondent would support the application of an egalitarian value to Blacks, women, or gays and lesbians. Specifically, respon­ dents were asked about their desire for equal roles in the workplace for men and women, whether the government should see to it that Blacks get fair treatment in jobs, and whether the respondent believes the federal government should make sure gays and lesbians are treated fairly in the workplace. Figure 2.8 shows that as White Americans age, they generally become more conservative on issues regarding the protection of women and the protection of members of lesbian and gay communities. But one thing that stands out is the U-­shape we see in the top panel of figure 2.8. When White Americans are at the beginning and end of their (political) lives, they are more likely than at other life stages to support the government seeing to it that Blacks are treated fairly in their jobs; perhaps this effect is due to not being (or no longer being) in competition with Blacks in the workforce. The second panel of figure 2.8 depicts period effects. For policies related to women and gays and lesbians, we see considerable progress. In fact, it seems that ever since the percentage of respondents who approved of women in the workplace increased from 64 percent in 1972 to over 80 percent in 1998, the ANES ceased asking the question. Again, this shift of 16 points might not seem like a radical progression, but given that this large of a change in support occurred in only twenty-­five years, it shows that Americans’ attitudes regarding applied egalitarianism can (and do) change. The largest of these shifts occurs regarding attitudes toward gays and lesbians being protected from discrimination in the workplace. In 1988, the first time the question was asked by the ANES, just about 53 percent of Whites supported laws that would “protect homosexuals from job discrimination.” In 2016, that percentage had jumped to over 80 percent, a gain of 30 percent in just about twenty-­five years. Again, Whites seem to be warming toward gays and lesbians in a way that resembles how they warmed to policies and opinions regarding women. However, when we examine a similar question that asked respondents if the government should ensure fair treatment for Blacks, there is essentially no movement in Whites’ attitudes over time. In the 1986 ANES, support for ensuring fair treatment for Blacks was at just 48 percent, where it stayed until 2012, falling slightly in the 2016 ANES.

Age Effects on Attitudes Regarding Protecting Various Groups Predicted Probability

1.00

0.75

Group Equal Roles for Men and Women

0.50

Workplace Protections for Gays and Lesbians Workplace Protections for Racial Minorities

0.25

0.00 15−19 20−24 25−29 30−34 35−39 40−44 45−49 50−54 55−59 60−64 65−69 70−74 75−89

Age Group Source: American National Election Studies Cumulative Data Set (1948−2016)

Group

0.75

Equal Roles for Men and Women

0.50

Workplace Protections for Gays and Lesbians Workplace Protections for Racial Minorities

0.25

16 20

5 01

9

−2 20

10

00

4

−2 20

05

00

9

−2 20

00

99

4

−1 19

95

99

9

−1 19

90

98

4

−1 19

85

98

9

−1

97

80 19

−1

97

75 19

−1

96 −1

70 19

65 19

4

0.00 9

Predicted Probability

Period Effects on Attitudes Regarding Protecting Various Groups 1.00

Period Source: American National Election Studies Cumulative Data Set (1948−2016)

Group

0.75

Equal Roles for Men and Women

0.50

Workplace Protections for Gays and Lesbians Workplace Protections for Racial Minorities

0.25 0.00 18 95 − 19 18 00 99 − 19 19 05 04 − 19 19 10 09 − 19 19 15 14 − 19 19 20 19 − 19 19 25 24 − 19 19 30 29 − 19 19 35 34 − 19 19 40 39 − 19 19 45 44 − 19 19 50 49 − 19 19 55 54 − 19 19 60 59 − 19 19 65 64 − 19 19 70 69 − 19 19 75 74 − 19 19 80 79 − 19 19 85 84 − 19 19 90 89 −1 99 8

Predicted Probability

Cohort Effects on Attitudes Regarding Protecting Various Groups 1.00

Cohort Source: American National Election Studies Cumulative Data Set (1948−2016)

2.8.  Age, Period, and Cohort Effects on Attitudes toward Specific Applications of Egalitarian Principles

Is Race Special?  /  63

Never once does Whites’ affirmation for implementing policies aimed to protect Blacks from discrimination reach even the lowest levels of Whites’ support for protecting gays and lesbians, despite the fact that during this same time period, many Whites acknowledged that Blacks are discriminated against in terms of finding work. According to the 1990 GSS, only 11 percent of Whites thought Blacks experience no discrimination at all, while 72 percent said Blacks experience either “a lot” or “some” discrimination that keeps them from getting well-­paying jobs. This point is one worth emphasizing and is representative of what some might call “cognitive gymnastics”; not only do Whites acknowledge that they are committed to equality for all and that Blacks are discriminated against, they also still prefer nothing be done about it. The last panel in figure 2.8 depicts cohort effects. The results reveal a clear upward trend for support for protecting gays and lesbians in the workplace, with younger cohorts being much more supportive. As far as support for laws that protect Blacks from discrimination, the line only crosses the 50-­percent mark for those generations that would have come of age during the civil rights era. But again, this change is hardly as large as those we saw regarding policies aiming to help and protect other groups. Given the United States’ history, attitudes toward protecting minorities’ rights might seem like a relatively low bar for a commitment to equality for all persons. For a more nuanced test of  Whites’ commitment to racial egalitarianism, we turn to look at a set of seven policies that we think can be classified as either “easy” or “hard” in terms of the level of commitment to racial fairness it would take to support each one. The easy questions include whether an individual believes it is fair for a homeowner to consider race when deciding to sell a home, whether an individual would support a ban on interracial marriage, and two questions that capture respondents’ attitudes toward women in the workplace. The first “easy” question concerning women, repeated from the GSS, asks whether the respondent supports women in the workplace. The second question, from the ANES, asks respondents if a woman’s place is “in the home” or if “women and men should have an equal role.”3 The “hard” questions involve issues that come closer to home. On issues of race, we examine whether an individual would support a close relative or family member marrying a Black person. We also analyze respondents’ support for affirmative-­action policies to remedy past discrimination for both Blacks and women. Again, we present these results as a series of figures illus­ trating the age, period, and cohort effects. Figure 2.9 illustrates the age effects on these seven policy issues. As we saw with other issues, the results show that as Whites get older, they become

Predicted Probability

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

4 −2 20

−2

25

9

4 −3

30

−3

35

Source: ANES and GSS Cumulative Files White Respondents Only

19

− 15

9

4 −4

40

Age Effects on Various Issues

9

4 −5

50

9

−5

55

9

−6

65

4

−7

70

9

−1

15

4 −2

20

9 −2 25

4 −3 30

E) Support Ban on Interracial Marriage F) Agree Government should help Blacks

Age Group

9

−8

75

C) Agree Can't Discriminate in Housing Sale D) Oppose Relative Marrying Member of Different Race

4

−6

60

4 −4 40

9 −4 45

Hard

4 −5 50

H) Affirmative Action (ANES)

G) Affirmative Action (GSS)

9 −3 35

2.9.  Age Effects on “Easy” and “Hard” Applications of Egalitarianism

A) Support Equal Gender Roles B) Approve of Women in Workplace

−4

45

Easy

9 −5 55

60

4 −6

65

9 −6

70

4 −7

75

9 −8

Women Blacks

Is Race Special?  /  65

less likely to endorse racially liberal positions. When Whites are older, they are more likely to oppose a relative marrying someone who is Black, more likely to support a ban on interracial marriage, and more likely to believe homeowners should be able to discriminate on the basis of race when selling their homes. As Whites age, not only are they also less likely to support either of the easier women’s issues, but also they become less likely to support affirmative action for both Blacks and women. Period effects are shown in figure 2.10. There appears to be a clear progression of attitudes on nearly all measures, with the distinct exception of affirmative action for Blacks. White Americans, on the whole, are now less likely to believe that interracial marriage should be banned and are less likely to oppose a close relative marrying an African American; they are also more likely to agree that women belong in the workplace and should have more equal roles with men. These are all results that satisfy our deep intuition about Americans becoming increasingly progressive over time. However, support for affirmative action—­for women or for Blacks—­has always been fairly low. Nonetheless, the support for affirmative action for women has increased from about one in five White Americans in 1996 to 30 percent in the 2016 GSS. Meanwhile, support for affirmative action for Blacks has remained approximately half as popular, moving from having the support of 11 percent of the White American population in 1994 to having 18 percent support in 2016. Finally, the cohort effects allow us to test whether more recent cohorts, including millennials, are ushering Whites’ attitudes in the direction of racial progress. The results in figure 2.11 illuminate two trends: stagnation across generational cohorts and notable declines in progressive attitudes. First, we should point out that millennials’ attitudes on affirmative action for women or for Blacks are no different than their predecessors’. This generation is less likely to support affirmative action for women and appears to be less comfortable with a family member or close relative marrying an African American. This last set of results is counterintuitive to a commonsense notion of millennial progressivism but is in line with our argument that race is a special issue domain in American politics. What, though, does this say about White millennials? Do they prioritize race matters, as many suggest they do? If they are more progressive than previous generations, we would expect racial issues, or race relations, to be more important to them than to their parents and grandparents. In 2016, the Cooperative Congressional Election Study asked respondents how important fifteen different issues are to them. We estimated how important each issue would be among White respondents and then conducted

Predicted Probability

9 97

1

1

9

98

−1

5 98

−1 90 9 1

4 99 1

2

9

00

−2

5 00 2

5

01

−2

0 01

16

20 1

1

4

98

−1

0 98 1

9 98

−1 5 98

1

4 99 −1 0 99

1

9 99 −1 5 99

E) Support Ban on Interracial Marriage F) Agree Government should help Blacks

9

97

−1

5 97

Period C) Agree Can't Discriminate in Housing Sale D) Oppose Relative Marrying Member of Different Race

04

20

− 00 20

2

9 00 −2 5 00

2

−2 0 01

01

5

20

H) Affirmative Action (ANES)

G) Affirmative Action (GSS)

2

4 00 −2 0 00

Hard

2.10.  Period Effects on “Easy” and “Hard” Applications of Egalitarianism

9

99

−1

5 99

A) Support Equal Gender Roles B) Approve of Women in Workplace

4

98

−1

0 98

Easy

Source: ANES and GSS Cumulative Files White Respondents Only

−1 75 9 1

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

Period Effects on Various Issues

16

Women Blacks

Predicted Probability

18

Easy

Hard

C) Agree Can't Discriminate in Housing Sale D) Oppose Relative Marrying Member of Different Race

E) Support Ban on Interracial Marriage F) Agree Government should help Blacks

H) Affirmative Action (ANES)

G) Affirmative Action (GSS)

2.11.  Cohort Effects on “Easy” and “Hard” Applications of Egalitarianism

Source: ANES and GSS Cumulative Files White Respondents Only

A) Support Equal Gender Roles B) Approve of Women in Workplace

Cohort

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 89 90 91 92 93 94 95 96 97 98 93 94 95 96 97 98 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 −1 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 9 0 1 2 3 4 5 6 7 8 3 4 5 6 7 8 18 19 19 19 19 19 19 19 19 19 19 19 19 19 19 19

29 19

5− 92

1

19 19

− 15 19

9 90

1 5− 90

1

9

9 18

− 95

0.00

0.25

0.50

0.75

1.00

0.00

0.25

0.50

0.75

1.00

Cohort Effects on Various Issues

Women Blacks

68 / Chapter Two Marriage Equality Environment Abortion Gun Control Race Relations Jobs Health Care Taxes Crime Govt. Corruption Deficit Immigration Defense Spending National Security Social Security More Important to Older Whites

More Important to Millennials

Difference in Issue Importance Rescaled to Run (0−1) Within Groups Source: 2016 CCES Common Content White Respondents Only N>40,000; ~33,000 Older Whites, ~13,000 Millennials (born after 1979) Each Issue Asked of ~13,000 respondents.

2.12.  Differences in Issue Importance across Generations

a difference-­of-­means test across those groups. Given that each issue was asked of about thirteen thousand White respondents, a standard statistical test will be sensitive to minute differences on the 0–­1 scale. The results are shown in figure 2.12. In fact, the issue that had the smallest significant difference we found was abortion, which had a difference of just 0.03 across the groups (t = −3.98, p < 0.01). For all but two issues, we found significant differences: compared to older Whites, millennial Whites think that marriage equality, taking care of the environment, and abortion are more pressing issues facing America. Other issues, like social security, are more important to older Whites. The only issues where there is no statistical difference across the groups, which have been bolded on the vertical axes, are “gun control” and “race relations.” Again we see that White millennials do not seem to be prioritizing race relations; while both groups think “health care” is the most important

Is Race Special?  /  69

issue, neither group ranks race relations in the top half of all fifteen issues facing the country.

The United States was founded on principles of egalitarianism and liberty, and “liberal idealism reflects the fact that, in our historical tradition and government roots, America is a liberal nation” (Smith 1990, 500). But America has also been challenged with a dilemma: the United States has not always lived up to these founding ideals in practice. Nonetheless, American citizens and denizens have worked toward positive change, broader inclusion, and greater equality. In his second inaugural address, Barack Obama noted that We, the people, declare today that the most evident of truths—­that all of us are created equal—­is the star that guides us still, just as it guided our forebears through Seneca Falls, and Selma, and Stonewall, just as it guided all those men and women, sung and unsung, who left footprints along this great Mall, to hear a preacher say that we cannot walk alone, to hear a King proclaim that our individual freedom is inextricably bound to the freedom of every soul on Earth. (2013)

In referencing Seneca Falls, Selma, and Stonewall, Obama specifically called attention to the rights movements of three historically marginalized groups: women, African Americans, and gay and lesbian Americans. Over the past five decades, the citizens who supported and participated in these movements, the Supreme Court, and the presidents who have been willing to en­ force laws that require fair treatment have been some of the most critical ac­ tors in leading the United States to close the gap between liberal ideals and American reality. While we no longer are likely to see candidates like George Wallace (1963) touting “segregation now, segregation tomorrow, segregation forever,” we are seeing some evidence, marked by the levels of racial conservatism among young Whites, that racial progress in the United States seems to be stagnating, at best, or on the brink of making a U-­turn, at worst. In this chapter, we examined a number of trends in Whites’ attitudes toward three groups that are typically underrepresented among the politically and economically powerful or the socially well-­connected. Overall, the results in this chapter highlight the fact that race has a special place in White Americans’ hearts and minds, and further, these results make clear why scholars must persist in talking about race and racism. There has been an increase in tolerance toward various minority groups and an increase in support for policies that protect these groups, but the boundaries of this

70 / Chapter Two

support are exclusive. For the most part, Americans have become more lib­ eral toward women and members of lesbian and gay communities, but prog­ ress has stagnated (and at times reversed) toward Black Americans. White Americans, who make up the majority of the population and are the dominant group in America’s social, political, and economic hierarchies, are increasingly willing to stand up for the rights of women and gay and lesbian Americans, but their support does not extend to Blacks. What’s more, in stark contrast to many recent declarations of millennial progressivism, we find little evidence that younger Whites are the source of systemic change toward more liberal attitudes. Instead, we have provided evidence for the central claim of this book: racial attitudes have stagnated, and America’s youngest cohorts are not doing the work of pushing racial progressivism in the way that their predecessors have done so historically. The questions we analyzed in this chapter largely focus on “old school” measures of racial animus: social distance, approval of interracial marriage, and differences in in-­group and out-­group evaluations. We examined policy attitudes without also analyzing whether these preferences are in anyway undergirded by racial animus. Some might suggest that these results are not really that bad—­people do tend to prefer their group to others, and again, they may oppose affirmative action for nonracial reasons. As such, we take this chapter’s findings with caution. In the next chapter, we delve into the more symbolic aspects of racial attitudes, with a special focus on the measures of contemporary racial attitudes that most political scientists rely on. There, we can do a more sophisticated analysis that allows us to test whether young Whites’ racial attitudes are structured differently than those of older White Americans. It could be the case that young people provide responses to questions about race, racism, and racial inequity that are similar to their predecessors’, but have a different set of ideas that they are rooting their answers in. If millennials think the same way that older people do, we’re in trouble. However, if they have a new or different way of thinking about racial issues, then we should probably come up with a different way to measure their attitudes. Gaining a better understanding of the racial logic and racial grammar of young people could ultimately help to provide tools to directly address the way that they are (perhaps unintentionally) participating in the reproduction of racial inequality.

Three

New Attitudes or Old Measures?

Over ten years ago, in 2008, 54 percent of Americans under thirty years of age gave their vote to Barack Obama. In turn, many journalists, political commentators, and researchers declared that the racial barriers of the United States had fallen. Further, they suggested that millennials have helped the United States to realize its post-­racial dream. For example, public-­opinion pollster John Zogby labeled Americans born between 1979 and 1991 as the “First Globals.” Zogby notes that this group “acknowledges race but avoids dwelling on it” (Zogby 2009). Similarly, journalist Tim Rutten wrote of his appreciation of the generational gap between older people who are polarized by race and religion and young people who are not; he explains, “So far, Obama appears to be the principal beneficiary of this trend, not simply because ‘change’ is a mantra that resonates with the young but because he personifies and articulates the post-­racial America in which most of our young people live.” Rutten goes on to assert that “skin color is no longer a physical marker for most of them [young people]. By and large, our sons and daughters describe their friends as tall or short, funny or serious, as good students or poor athletes, but seldom—­as earlier generations would have done—­as a ‘Black guy’ or a ‘White girl’” (Rutten 2008). Even the Pew Research Center (2010) has asserted that the millennial generation is the most racially tolerant generation in history; Pew based this claim on the fact that millennials are more likely than any other group to approve of interracial marriage and interracial dating. These proclamations and findings fit a tidy “post-­racial” narrative—­that race and racism are no longer barriers in this country, thanks in large part to the young people whom we have finally socialized to overlook race and overcome racial prejudice. But we believe that these commentaries and surveys say more about a fundamental misunderstanding of race and racism in

72 / Chapter Three

the United States today than they do about young Whites’ attitudes. Those in the media are subject to a different standard of scrutiny with regard to measuring anything, really. Social scientists, on the other hand, are expected to develop and employ more nuanced measures of everything. In the case of explicit racial attitudes, the most frequently cited measure is the racial resentment scale, which actually tells a different story about Americans’ racial attitudes. Centering a well-­tested, frequently employed measure of racial attitudes, we use this chapter to answer the following question: To what degree have Americans’ racial attitudes—­based on the racial resentment scale—­changed over time? We determine that, overall, racial resentment is not decreasing over time but, contrary to our previous findings in some respects, young White people report lower levels of racial resentment than older White Americans. We then have to ask another set of questions: Do we find that young White people have lower levels of racial animus because they are more likely to be liberal (in the 2016 CCES, 38 percent of White millennials identify as liberals, compared to just 25 percent of older Whites), or are they simply different than their predecessors? Difference can arise in two ways. Millennials can foster less racial animosity and resentment than their predecessors, or their attitudes can be structured in a wholly different way than older people’s. In other words, they can have a different orientation toward the questions that political scientists have traditionally asked, thus introducing heterogeneity into the ideas that pop into people’s heads as they are answering the questions on the racial resentment scale. This chapter aims to adjudicate between these two possible explanations.

Whites’ Racial Attitudes: Measurement and Trends While it is well known that certain types of racial attitudes have declined significantly among the White American population since World War II, it is debated whether negative racial animus toward Blacks has actually declined. The best answer concerning the question of whether racism has decreased is likely to be “it depends on who is asking and what is being asked” (McConahay, Hardee, and Batts 1981, 563). A nuanced response is required be­ cause the meaning, configuration, and measurement of racial attitudes have changed over time; as social norms change, expressions, levels, and the struc­ ture of racial attitudes also change.1 Prior to the 1980s, racial-­attitude survey items primarily consisted of three types of questions. The first type queried respondents about matters of racial principles in the form of questions concerning whether society should

New Attitudes or Old Measures?  /  73

be racially integrated and whether people across racial groups should be treated equally. The second type of question focused on social distance and gauged respondents’ willingness to be in contact with out-­group members as neighbors, friends, potential in-­laws, and the like. The third type of question focused on the extent to which respondents supported the implementation of policies that would bring about integration or racial equality. Schuman and his colleagues (1997) found that overt bigotry, demands for racial segregation, the condoning of racial discrimination, and the belief that Blacks are biologically inferior to Whites all declined significantly in the era after World War II. In fact, by 1972 only 3 percent of Whites thought they should receive preference over African Americans, and by 1985 only 7 percent of White Americans favored racially segregated schools (Bobo and Smith 1998). Eventually, social-­distance questions were removed from national election surveys because there was so little variance in the responses. There is strong evidence that suggests a significant and positive transformation in Whites’ attitudes toward Blacks in the long arc of history, but with one major caveat: scholars found that even though attitudes toward Blacks had become more positive and Whites believed in racial equality, Whites’ attitudes toward implementing policies aimed at ameliorating racial disparities—­examined in the third set of questions—­did not significantly change over that same period (Schuman et al. 1997; Sears et al. 2000). The gap between the belief in racial equality and the support necessary to move racially equitable policies forward inspired researchers to investigate an apparent contradiction in White Americans’ attitudes. Sociopsychological research theorized that the gap between Whites’ principles and policy preferences could be well explained by a more subtle and covert form of racism that had developed among the American population. There have been many labels for this form of covert racism: aversive racism (Dovidio 2001), symbolic racism (Kinder and Sears 1981), racial resentment (Kinder and Sanders 1996), laissez-­faire racism (Bobo, Kluegel, and Smith 1997), modern racism (McConohay 1986), and subtle racism (Pettigrew and Meertens 1995), to name a few. While each of these concepts is rooted in its own nuanced theory, they all share two things in common: that most Whites are resistant to implementing policies that would lead to greater racial equality because, as the logic goes, the Civil Rights Act successfully “leveled the playing field,” and that the reason why Blacks lag behind Whites has to do with Blacks’ personal failings and general lack of motivation rather than structural inequalities. Racial resentment, in particular, has garnered substantial attention in political science. Scholars have noted that prior to the civil rights movement,

74 / Chapter Three

the “race problem” in America was marked by unequal opportunity, discrimination, and lack of enforcement of constitutional rights. After the late 1960s, the “race problem” centered on the threat that Blacks posed to the social order. More specifically, Kinder and Sanders (1996) explain, after the 1965 Watts riots, and others like it, Whites were increasingly likely to suggest that Blacks lag behind Whites on important socioeconomic indicators because of Blacks’ failure to live up to core American values. This reinterpretation of the American “race problem” is the foundation of racial resentment. Kinder and Sears explain that this modern form of racial attitudes is “A blend of affect and the kind of traditional American values embodied in the Protestant Ethic. . . . [This attitude] represents a form of resistance to change in the racial status quo based on moral feelings that Blacks violate such traditional American values as individualism and self-­reliance, the work ethic, obedience, and discipline” (1981, 416). There has been significant criticism of the racial resentment scale (Kluegel and Bobo 1993; Sniderman, Crosby, and Howell 2000; Sniderman and Piazza 1993; Sniderman and Tetlock 1986; Stoker 1998; Wilson and Davis 2011), but the principal creators of symbolic racism have revised the measurement over time in efforts to improve its validity and have shown that the construct is a coherent belief system that is independent of “old-­fashioned” racism and political conservatism (Henry and Sears 2002). Additionally, a plethora of studies have shown that racial resentment influences Whites’ attitudes on the death penalty (Unnever and Cullen 2007), social-­welfare policies (Kinder and Mendelberg 2000), affirmative action (Bobo 2000), vote choice (Kinder and Sears 1981; Tesler and Sears 2010), who is deserving of governmental help (DeSante 2013), and more.2 We do not question the value of racial resentment as a theory; instead, we are concerned with the cross-­generational validity of the measure. The move from old-­fashioned Jim Crow racism to modern forms of racism, like symbolic racism and racial resentment, came about because of major shifts in the American economic, political, and social landscape. Demographic changes due to the “Great Migration” of African Americans from the South to other parts of the country, the urbanization of Blacks, the development of minority organizations, economic factors such as industrialization, the decline of Southern planter elites’ power and Jim Crow social structures, and minority mobilization (in the civil rights movement as well as in race riots) are all factors that led to the development of what is now known as “new” racism (Bobo and Smith 1998; Bonilla-­Silva, Lewis, and Embrick 2004; Pendergrass 2013). Some of these historical phenomena are explicitly referenced in the racial resentment items. For example, one question asks how strongly a respon­ dent agrees with the statement “The Irish, Italians, Jews, and other minorities

New Attitudes or Old Measures?  /  75

overcame prejudice and worked their way up. Blacks should do the same without any special favors.” For members of the silent generation (those born between 1925 and 1945) or the baby boomers (those born between 1946 and 1964), the narratives of these specific immigrant groups might be more salient, if only due to the raw number of immigrants coming to America from, for example, Italy and Ireland. Between 1900 and 1949, the United States granted permanent-­residence status to over 3.8 million Italians and over 750,000 Irish. From 1950 until 2000, those numbers drastically declined to about 671,000 and 184,000, respectively. Over those same periods, the number of African immigrants increased nearly twentyfold, from 31,000 to nearly 600,000, and the number of people coming from Asia who were granted permanent-­resident status increased from 750,000 during the period from 1900 until 1949 to approximately 7.1 million in the second half of the previous century. Given their particular historical context, millennials may be more likely to think of different groups (e.g., those from China, Mexico, or India) than their predecessors do when they think of immigrants overcoming prejudice. Additionally, they may rely on an entirely different set of myths and stereotypes in their considerations of these groups. For instance, Jane Junn notes that over the past century, “the dominant trope for Asian Americans has shifted dramatically from coolie to model minority” (2007, 356, emphasis in the original). Though both stereotypes represent, in some ways (as we will show in chapter 4), a threat to White Americans’ social and economic standing, the former trope is overtly denigrating while the latter allows individuals to overlook ongoing discrimination against a group in light of that group’s ostensible success. The racial resentment scale was developed during a time of heated and clearly racialized battles over public policies. Michael Tesler explains: “the political debate over racial policies also shifted in the post-­civil rights era from OFR[old-­fashioned racism]-­related concerns about desegregation to an equality of outcomes agenda that evoked racially resentful anxieties about black deservingness and work ethic” (2013, 111). We believe that when older Whites are asked whether Blacks have gotten less than they deserve, they are likely to think about policies that aimed to equalize outcomes, such as busing and racial quotas—­issues that are not likely to be salient in the mind of someone who came of age long after racial quotas were banned by the US Supreme Court (Regents of the University of California v. Bakke, 1978) and forced busing and mandatory assignment had largely been replaced by magnet schools, where Whites can voluntarily integrate schools (Goldring and Smrekar 2000; Hannah-­Jones 2014; Wilson 2018).

76 / Chapter Three

Considering this, we may expect that these survey item questions may not be as strongly related to millennials’ racial attitudes as they would be for older Whites. While younger Whites may appear to harbor less racial animus than previous generations, the majority of the landmark events referred to by the survey items (e.g., the civil rights movement, the Great Migration) fall outside younger Whites’ historical memory. Consequently, we argue that the measures that incorporate these types of era-­specific references may fail to accurately describe and capture the attitudes of those who have a different set of references, experiences, and worldviews.

Theoretical Expectations As mentioned, there is no shortage of evidence that people might provide to bolster an argument that White millennials are more racially liberal than previous generational cohorts; indeed, our own data in this chapter reveal that Whites born after 1980 report lower levels of racial animus than older Whites, even when we employ what political scientists consider a gold stan­ dard of racial-­attitudes measurement. Nonetheless, the devil is in the details.3 There are two potential theoretical explanations for these seemingly high levels of racial liberalism (or seemingly lower levels of racist responses), particularly among White millennials. One relies on the continuous socialization model, and the other is rooted in a theory of symbolic politics. A theory of change based on the continuous socialization model predicts “an ever-­improving spiral” of racial attitudes over time due to the replacement of older, more racially conservative and prejudiced individuals by younger, more racially conscientious and less prejudiced Americans (Schuman et al. 1997, 199). Based on this theoretical model of attitudinal change, we would predict that young people will be more racially liberal than their predecessors and thus will lead the way toward racial progressivism. This theory would lead us to prioritize the idea that younger Whites are systematically different from previous generations and simply have less racial animosity toward Blacks, in an apples-­to-­apples comparison. The second explanation, which has been advanced by scholars before, is that current measures are no longer sensitive to contemporary expressions of racism, especially among younger Whites. This idea is rooted in a theory of symbolic politics. David Sears explains: “This theory holds that people acquire stable affective responses to particular symbols through a process of classical conditioning, which occurs most crucially at a relatively early age. These learned dispositions may or may not persist through adult life, but the strongest—­called ‘symbolic predispositions’—­do. The most important

New Attitudes or Old Measures?  /  77

of these predispositions in American politics include party identification, political ideology, and racial prejudice” (1993, 120). This theory centers socialization and upbringing and thus generational cohort status. It primes us to be cognizant of the notion that, given how different the racial landscape in which millennials have been socialized is from that of previous generations, today’s younger Whites are likely to have qualitatively different racial attitudes than their predecessors. Just as previous transformations in the American economic, political, and social environments have changed the levels and structure of racial attitudes (Bobo and Smith 1998; Bonilla-­ Silva, Lewis, and Embrick 2004), we may be in the midst of yet another reconfiguration, or even a step back toward older forms, of racial attitudes (Tesler 2013).

Empirical Expectations The extant literature provides a clear motivation for our central hypothesis regarding cross-­generational differences in how racial prejudice relates to the racial resentment scale. In keeping with the continuous socialization model, we expect that younger Whites will report more racially liberal or racially tolerant attitudes toward Black Americans than do older Whites. However, we also know that an individual’s attitudes can change over time and that people tend to become more conservative with age. Therefore, one might suggest that young people are racially liberal because they are young. We expect to find that people will report higher levels of racial resentment in their later years more so than when they are young, due to age or life-­cycle effects, but strong evidence for the continuous socialization model will be found if we see that newer birth cohorts cause attitudes at the aggregate level to shift toward more progressive attitudes. Put simply, we have the following cross-­generational hypotheses: H1: Younger Whites (millennials) will report lower levels of racial resentment than Whites born before 1980. H2A: Americans’ racial attitudes will become more racially conservative as they get older, even when we control for cohort and period effects; this is the “age effect” hypothesis. H2B: Given the change in America’s racial environment, we should see White Americans’ racial attitudes become more liberal over time, even after controlling for cohort and age effects; this is the “period effect” hypothesis.

78 / Chapter Three H2C: We will see a monotonic decline in racial animus across cohorts, controlling for period and age effects. Ultimately, millennial Whites will be less likely to express racially conservative opinions than their predecessors, if all else is equal; this is the “cohort effect” hypothesis.

The symbolic politics theory, however, requires us to test not only the quantity or levels of racial animus but also the qualitative differences. Millennials have lived their entire lives in an America that ostensibly values diversity and eschews overt racial animus. Ultimately, we expect White millennials to report lower levels of racial resentment largely because of the environment in which they were raised. As a result, we also expect that these environmental differences will produce different structures in racial attitudes. That is to say, the expected difference in means (H1) will not necessarily be due to the fact that millennials are less racially prejudiced; instead, it will be because younger Whites are not answering the racial resentment scale in the same way as older Whites. For the myriad reasons we discuss above, it is our expectation that younger Whites’ racial attitudes will be structured differently from those of older White Americans. In order to ascertain this, we test for measurement invariance; here, we expect to see that the survey items we use to measure latent prejudice—­via the racial resentment questions—­are not related to racial prejudice in the same way for older and younger Whites. It is likely the case that the link between latent racial resentment and the survey items will vary across generations. In other words, H3: The structural relationship between racial prejudice and the racial resentment scale will vary across generations.

Data, Methods, and Results Generational Differences: Racial Resentment Scale We turn to the American National Election Studies (ANES) surveys to test our first hypothesis. Both the 2012 and 2016 ANES surveys consist of both face-­to-­face interviews and an internet sample. In order to maintain comparisons across previous years’ surveys, we limit our analysis to non-­Hispanic White respondents who answered the questions face-­to-­face. While the number of respondents varies from question to question, we are left with approximately 1,550 complete cases. The four racial-­resentment questions ask respondents how strongly they agree with statements on a five-­point

New Attitudes or Old Measures?  /  79 Table 3.1.  Cross-­Generational Differences in Racial Resentment Older Whites

Millennials

T-­Statistic

Item 1: Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class.

Mean Std. Dev. N

2.18 1.41 1166

2.04 1.99 408

1.69*

Item 2: The Irish, Italians, Jews, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.

Mean Std. Dev. N

2.84 1.28 1165

2.57 1.29 407

3.79**

Item 3: It’s really a matter of some people not trying hard enough; if Blacks would only try harder, they could be just as well off as Whites.

Mean Std. Dev. N

2.30 1.37 1164

2.22 1.34 405

0.92

Item 4: Over the past few years, Blacks have gotten less than they deserve.

Mean Std. Dev. N

2.64 1.25 1160

2.41 1.24 406

3.26**

Mean Std. Dev. N

0.624 0.26 1149

0.579 0.26 402

2.91**

Four Item Additive Battery (0–­1)

Data: 2012 and 2016 American National Election Studies (ANES) Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.10, **p < 0.05

categorical scale; each item was recoded to run from 0 to 4, such that larger numbers correspond to a higher level of racial resentment. A value of 2.0 would indicate that the group was neither racially liberal nor conservative, but as we can see, all of the group means fall on the more resentful side of 2.0. In order to test our first hypothesis, which asserts that younger Whites will report lower levels of racial animus, we begin with difference-­of-­means tests for each of the four items. Table 3.1 shows the means of each item by generational status. The results show that younger Whites systematically report lower levels of racial resentment. With the exception of the third question, White millennials provide less racially resentful responses to each of the four items. Overall, we have support for our first hypothesis—­White millennials report lower levels of racial resentment than their predecessors. But again, these differences may arise due to the fact that people tend to be more liberal when they are young. Again, using the age-­period-­cohort intrinsic estimator, we can parse out whether these differences are due to age, period, and/or cohort effects to test our next set of hypotheses.

80 / Chapter Three

Generational Differences: Estimating Age, Period, and Cohort Effects For this particular estimation procedure, our data come from the 1986–­ 2016 ANES data sets, with a total of nearly ten thousand White respondents who answered all four racial-­resentment items. Figure 3.1 presents the average score for non-­Hispanic Whites on the additive racial resentment scale, for every year that all four items were asked on the ANES survey. Figure 3.1 shows us that there has been hardly any change in the level of racial resentment at the aggregate level over a thirty-­year period. The level of racial resentment in the population of White adults has held relatively stable, yet another example of racial stasis. Next, we must determine whether the trends are due to age effects, cohort effects, and/or period effects in order to test H2A, H2B, and H2C. Because we are interested in the cross-­generational change and explanatory power of the racial resentment scale, we add the four items and rescale the variable to range from 0 (less racial resentment) to 1 (most racially resentful). Following standard practices, age, period, and cohorts are grouped into five-­year categories.4 Consequently, we estimate coefficients for five periods, fourteen age groups, and eighteen cohorts. The results of the APC-­IE model are presented in table 3.2. The coefficients are estimated variations from the overall mean for each age, period, or birth cohort, controlling for the other predictors. The intercept is the overall mean averaged across periods, birth cohorts, and ages. The coefficients, which are similar to ordinary least squares regression coefficients, are easier

Racial Resentment (0−1)

Average Racial Resentment Over Time 1.00 0.75 0.50 0.25 0.00 1988

1992

1996

2000

2004

2008

Year Source: American National Election Studies Cumulative Data Set (1986−2016) Non−Hispanic White Respondents Only

3.1.  Trends in Racial Resentment

2012

2016

New Attitudes or Old Measures?  /  81 Table 3.2.  Intrinsic Estimates of Age-­Period-­Cohort Effects on Racial Resentment Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­89 80–­84

−0.036* −0.024* −0.006 −0.004 −0.032* −0.007 0.011 0.024* 0.029* 0.026* 0.017† −0.002 0.006

0.009 0.008 0.008 0.008 0.008 0.009 0.009 0.009 0.009 0.01 0.011 0.011 0.014

1985–­1989 1990–­1994 2000–­2004 2005–­2009 2006–­2010 2011–­2016

−0.014* −0.002 0.003 0.003 0.021* −0.012*

0.005 0.005 0.006 0.008 0.007 0.006

1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1998

0.047† 0.015 0.019 −0.003 −0.006 0.004 −0.008 −0.018† −0.018* −0.021* −0.008 0.012 0.027* 0.026* −0.031* −0.019 0.003 −0.022

0.028 0.016 0.014 0.013 0.011 0.011 0.011 0.01 0.009 0.009 0.008 0.008 0.008 0.01 0.012 0.014 0.016 0.021

Constant N = 10,209

0.631*

0.004

Data: 1986–­2012 American National Election Studies (ANES) Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test; †p < 0.05, one-­ tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

to interpret in graphical form, so we present the effects in the three panels of figure 3.2. As figure 3.2 shows, age is strongly related to feelings of racial resentment (RR), as expected by H2A. Independent of the period the survey was conducted in and of the birth cohort of a respondent, Whites tend to become marginally more racially resentful as they enter their late forties and early fifties, a trend that continues for most of their lives. When Whites are younger, aged thirty years or less, they tend to express less racial resentment than those older than them. We have some support for our age hypothesis, though the differences (which are statistically different from zero) are not as large as some scholars may expect. The largest difference between any two age groups is about 0.08 on the one-­point scale, or just 8 percent of the total possible change. The middle panel in figure 3.2 presents our estimated period effects. These data serve as evidence that racial resentment was significantly increased

82 / Chapter Three

3.2.  Graphical Display of Age, Period, and Cohort Effects from Table 3.2

among Whites during the Obama years, a finding that runs counter to what we expected (H2B). Despite one popular narrative that the election of President Obama ushered in a new, “post-­racial” America—­a premise Tesler and Sears (2010) and many others clearly refute (Hutchings 2009; Piston 2010; Tesler 2012; Valentino and Brader 2011)—­racial resentment among Whites has actually increased over the last twenty years. Certainly, the racialization

New Attitudes or Old Measures?  /  83

of the 2008 and 2012 presidential elections plays some part in this, but heretofore scholars could not precisely estimate just how great an effect these periods had compared to life-­cycle and cohort effects. Next, we turn to our hypothesis concerning cohort effects. The results for H2C are illustrated in the bottom panel of figure 3.2. First, we notice that older birth cohorts (those born in the early twentieth century) exhibit higher levels of racial resentment than do those born in the late 1940s, the 1950s, and the 1960s. However, beginning with those born in 1970, the level of racial resentment seems to increase slightly before beginning to decrease monotonically for those cohorts that make up the millennial generation. When examined in a single figure, the age-­period-­cohort effects all seem to be very small. While about half of the estimated coefficients reach conventional levels of significance, the largest effect is less than 0.05 on a one-­point scale. So, while we have some weak evidence for the continuous socialization model, racial resentment in the aggregate is remarkably stable. This leaves us wondering: Are these measures capturing the same constructs across generations? Racial Resentment and Measurement Invariance As our analyses heretofore have indicated, younger Whites appear to be less racially prejudiced than their older counterparts. Not only have we seen that millennials answer items on the racial resentment scale with lower average scores, but the APC-­IE analysis also indicates that the cohorts that form the millennial generation attenuate some of the aggregate levels of racial resentment. The question we now turn to is whether the relationship between racial prejudice, an unobservable (latent) trait, and these items that seek to measure that trait is approximately equal across generations. In order to figure this out, we perform a series of measurement invariance tests rooted in multiple-­groups confirmatory factor analysis using the face-­to-­face portion of the 2012 ANES. In order for each of the four items to be equivalent measures of racial animus across generations, they must display measurement invariance (Millsap 2012); that is to say, they must link the same underlying trait, on the same scale, in the same way for both older and younger Whites. If the mea­ sures fail to display measurement invariance, comparing the group means on scales, as scholars often do (and as we did earlier in this book), may be inappropriate, as the questions might be being answered in systematically different ways conditional on group membership (Brown 2012; Cheung and Rensvold 2002; Davidov 2009; Meredith 1993). What this means for the

84 / Chapter Three

central issue at hand is that while White millennials appear to exhibit lower levels of racial resentment—­simply evidenced by the means of each of the scale’s items—­these questions might not correlate with racial prejudice in the same way. Previously, political scientists have used this factor-­analytic technique to determine whether group differences exist across racial groups (Pérez and Hetherington 2014), nations in cross-­national surveys (Stegmueller 2011), and even the language in which the survey questions were asked (Pérez 2009). As Eldad Davidov succinctly describes, testing for measurement invariance can help check whether differences in groups’ means on survey items arise from “a different understanding of the question items and not due to ‘true’ differences across groups” (2009, 68); without testing for invariance, scholars are only able to make superficial comparisons. The model we estimate using confirmatory factor analysis is shown below: (1) ym = λmξ + δm. In equation (1), each item in the racial resentment scale (ym) is a function of latent racial animus (λmξ) as well as a random error term (δm). However, this model specification assumes that the observed items, which we have coded so that higher item scores correspond to higher levels of racial resentment, are related to an underlying response variable (y*m) that has a continuous distribution such that y m = y*m (Brown 2012; Pérez and Hetherington 2014). Alas, because the racial resentment items are ordinal scales, we violate this assumption. However, since the response variable, y*m , is   theoretically continuous, we can estimate a relationship between the latent factor (ξ) and our observed responses (ym). We do this by estimating four thresholds (τ) for each of the response items.5 The model we estimate is depicted in figure 3.3. First, we estimate the factor loadings6 and thresholds for each of the two groups (millennials and older Americans) separately using the robust weighted least squares procedure in Mplus. In our unrestricted model, factor loadings and thresholds are freely estimated across groups. We also estimate the error covariance be­­ tween items 2 and 3, as both items relate to work ethic at the individual, as opposed to the systemic, level. If this model specification in its unrestricted form has significant factor loadings (λ) and thresholds (τ) in each group and has a good fit to the data, we can then proceed to place restrictions on the model and constrain factor loadings and thresholds to be equal for both. The first two columns in table 3.3 show the standardized factor loadings, thresholds, and model-­fit

3.3.  Measurement Model for Racism Table 3.3.  Test of Measurement Invariance across Age Groups (2012 ANES) Freely Estimated

Constrained

Millennials

Full Sample

Older Whites

Item 1: Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class.

τ11 τ12 τ13 τ14

0.763* −1.21 −0.18 0.21 0.909

0.763* −1.272 −0.247 n.s. 0.611

1 −1.223 −0.222 0.067 0.675

Item 2: The Irish, Italians, Jews, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.

τ21 τ22 τ23 τ24

0.431 −1.489 −0.912 −0.244 0.508

0.618 −1.605 −1.021 −0.605 0.104

0.604 −1.642 −1.034 −0.53 0.205

τ31 τ32 τ33 τ34

0.541 −1.208 −0.759 n.s. 0.73

0.588 −1.223 −0.606 −0.094 0.558

0.578 −1.191 −0.635 −0.088 0.593

τ41 τ42 τ43 τ44

0.887 −1.729 −0.883 n.s. 0.679

0.807 −1.666 −0.938 −0.353 −0.403

0.827 −1.646 −0.899 −0.259 0.468

0.378 338 1.00 1.00 0.00 2.45 3 p ≈ 0.48

0.422 873

0.434 1211 0.987 0.98 0.071 73.66 18 p < 0.001

Item 3: It’s really a matter of some people not trying hard enough; if Blacks would only try harder, they could be just as well off as Whites.

Item 4: Over the past few years, Blacks have gotten less than they deserve.

Item 2 with Item 3 N TLI CFI RMSEA χ2 χ2 df

Data: 2012 American National Election Study (ANES)

86 / Chapter Three

statistics for our unrestricted estimates. Overall, this unrestricted model has a very good fit to the data; both the Tucker-­Lewis Index (TLI) and the confirmatory fit index (CFI) are near 1.0, the root mean squared error of approximation (RMSEA) falls below 0.10, and our χ2 statistic is small and insignificant.7 For interpretation purposes, what the (standardized) loading on item 2 for millennials indicates is that a standard deviation shift in latent prejudice would cause a 0.431 standard deviation shift in the response variable underlying that item; however, for older Whites the same change in latent racism is accompanied by an increase of a 0.618 standard deviation in the response variable for item 2. Clearly, some of the loadings appear to be different across the groups; the latent variable seems to be more strongly related to item 4 (a measure that asks about whether Blacks have recently gotten less than they deserve) among White millennials than among older Whites. Conversely, latent racism seems to be more strongly related to items 2 and 3 among older Whites. Substantively, and speaking to our hypothesis about the changing historical context, this particular difference may indicate that the “other minorities” item is not as strongly related to racial prejudice for younger Whites. On the other hand, the question about Blacks getting “less than they deserve” is more strongly related to racial prejudice among White millennials than it is to prejudice among older Whites. The feeling that Blacks have advantages over Whites, or are getting more than they deserve, is a sentiment common among those who espouse color-­blind racial ideology (Neville et al. 2000). However, whether these measures are functioning differently across the groups cannot be tested without comparing a constrained model to the freely estimated one; this model is presented in the rightmost column of table 3.3. At first glance it appears as if this constrained model fits our data relatively well; both fit indices (TLI and CFI) are at acceptable levels and the RMSEA is larger than before but still below 0.10. One notable change between the models is that the χ2 statistic is now larger in the constrained model, indicating that these constraints decrease the quality of the model’s fit. The appropriate test for this change is a χ2 difference test (Δχ2 = 68.93, Δdf = 16 , p < 0.001), which reveals that the constrained model fits the data significantly worse.8 Consequently, these results provide support for our third hypothesis: the structural relationship between racial prejudice and the racial resentment scale varies across generations. The immediate implication for this finding is that while we see that White millennials report lower levels of racial resentment, it is unclear whether they genuinely have lower levels of racial prejudice. If asked whether White millennials foster less racial antipathy than their predecessors, we would

New Attitudes or Old Measures?  /  87

have to say that we simply can’t tell, because we are not necessarily measuring the same relationship between the survey items and the latent variable across generational groups. The differences that are reported are not “true” differences and instead may arise because there is a different understanding of the questions across generations. When we rely on the racial resentment scale for millennials and others alike, we are essentially comparing apples and oranges. Just because the links between the racial resentment items and the latent variable are not exactly the same does not mean that the mea­ sure is not capturing something for younger Whites. As we lay out below, we believe it may be capturing something that is more akin to anti-­Black affect and less about core ideological principles.

Less Is More: How Latent Racism Operates in Younger Whites What we have uncovered are two types of generational differences—­one his­­ torical and the other psychosocial—­and these observations present us with an interesting opportunity to test several additional hypotheses. We have typically steered away from some of the debates surrounding racial resentment, such as debates about the degree to which the scale actually measures what its creators’ theorize, but our experiences talking to millennials about race, as well as the findings we have presented so far, inspire us to take part in this conversation at some level. The racial resentment scale is an important measure, but it has received a great deal of criticism. Some suggest that racial resentment is really no different than “old-­fashioned racism,” a claim that Sears and Henry refute (Henry and Sears 2002; Sears and Henry 2003). Additionally, as time has passed, some critics have suggested that this measure not only captures racial attitudes but also in fact might be more strongly related to political ideology (e.g., Sniderman and Tetlock 1986; Carmines, Sniderman, and Easter 2011). In other words, those in this “principled politics” camp suggest that Whites are not supportive of certain racialized policies because Whites are principled, political conservatives, rather than because they are anti-­Black. If the principled politics theory is correct, then we would expect to see that political ideology is strongly related to measures of racial resentment, and ideology should be the principal force shaping White Americans’ political attitudes. However, if the larger structure of racial attitudes is changing, we might expect that younger Whites’ racial attitudes and ideological principles might be only weakly correlated with one another. That is to say, the relative weights of the two constructs that scholars believe racial resentment measures—­principled conservative values and racial animus—­may not be

88 / Chapter Three

equal across generations. It may be the case that racial resentment among millennial Whites is measuring racial prejudice divorced from ideological principles. Furthermore, when we consider the psychosocial generational differences, compared to older Whites, millennial Whites’ racial resentment may more closely resemble old-­fashioned prejudice. We believe it is more difficult, psychologically, for younger Whites to signal racial prejudice given their upbringing; therefore, when they do, it might have a greater effect on related political attitudes. In short, whatever the threshold is for a White respondent to reveal racial resentment on a survey, we believe it is higher for millennials than for older Whites. For a millennial to express anything that resembles an anti-­Black attitude, they must really want to express it. As a result, we hypothesize that H4: The relationships between racial resentment, “old-­fashioned racism,” and political conservatism will differ across generational cohorts. Empirically, this will be shown in the ways that latent racial prejudice is related to old-­fashioned racism and political attitudes in different ways across generations.

What’s more, we have mentioned that young people were raised to be “color-­blind.” On one hand, this means that young people will work harder to show that they are not racist, but “color-­blindness” also means that they do not have a clear understanding of the structural and institutional character of contemporary racism. Their generational predecessors have taught them to avoid conversations about race, racial discrimination, and racial stereotypes (Lewis 2001; Hagerman 2018). As a consequence of being socialized in this racial milieu, “any continued inequality can easily be viewed as resulting from the failures of individuals to achieve rather than to racial discrimination by whites” (Schuman et al. 1997, 220). That is to say, since parents are not teaching their children about historically based structural racism, millennials may be inclined to rely on blaming individuals for ongoing patterns of racial inequality. What’s more, Tesler (2013, 2016) shows that in 2008, Obama heightened the link between Black Americans and the Democratic Party, and a large proportion of millennials came of age and voted for the first time during this political context. As such, we expect that those who are low in racial animus may overcompensate by providing a great deal of support to liberal causes and policy issues, while those who are high in racial animus may feel more negatively toward Obama compared to older White Americans. Put simply, our final hypothesis:

New Attitudes or Old Measures?  /  89 H5: The cross-­generational difference in structure will have empirical implications for the way latent racial animus influences Whites’ policy preferences. Racial animus will have a greater effect on the millennial generation’s racialized policy preferences than it will for older White Americans.

In order to test this, we look at data from the most recent American National Election Studies, which provide a number of variables with which we can test the hypothesis that racial resentment is more closely related to old-­fashioned racism for younger Whites than for older Whites. The 2016 ANES collected data from 3,038 non-­Hispanic White respondents, only 796 of whom were interviewed face-­to-­face. For our main independent variable, we use the four-­question racial resentment battery and scale it to run from 0 (least racially resentful) to 1 (most racially resentful). For dependent variables, we use the differences between how a respon­ dent placed Whites and Blacks on two “old-­fashioned” stereotype scales: rating Whites and Blacks as either hardworking or lazy and as either peaceful or violent. Each item asks the respondent to place the group on this scale from 1 to 7, which we rescaled to run from 0 to 6, where higher values represent a more positive rating (either more hardworking or more peaceful). We then subtract the rating of Blacks from the placement of Whites to get a difference score; a positive number indicates that the respondent believes Whites are either more hardworking or more peaceful. Thus, when we regress this variable on the racial resentment scale, the coefficient for racial resentment represents the magnitude of racial stereotyping between the least and most racially resentful respondents. We present these ordinary least squares regression models first as bivariate regressions for each group: millennials (n = 827) and older Whites (n = 2,211). We then present a third model that includes the interaction term between a dichotomous (0/1) variable indicating a millennial (someone born after 1979) and the racial resentment scale. The estimates for these statistical models are presented in tables 3.4 and 3.5. As we can see from the regression estimates presented in these tables, the effect of racial resentment (RR) on these stereotypes is signed in the expected direction: those who are more racially resentful hold more negative stereotypes of Blacks than of Whites. The model in the first column, for older Whites, predicts that older Whites at the lowest end of the racial resentment scale are likely to rate Whites as 0.57 points more violent than Blacks. However, the most racially resentful older Whites are predicted to rate Blacks as 2.14 (2.71 − 0.57) points more violent than Whites. More

90 / Chapter Three Table 3.4.  Racial Resentment, Generational Status, and Stereotypes of Blacks as Violent

Racial Resentment

Older Whites

White Millennials

Full Sample

2.71* (0.12)

3.49* (0.20)

−0.57* (0.08) 1,856 0.225

−0.99* (0.12) 665 0.291

2.71* (0.13) −0.42* (0.13) 0.79* (0.21) −0.57* (0.08) 2,521 0.254

Millennial RR × Millennial Intercept N Adjusted R2

Data: 2016 American National Election Study (ANES) Notes: *p < 0.05, standard errors in parentheses

Table 3.5.  Racial Resentment, Generational Status, and Stereotypes of Blacks as Lazy

Racial Resentment

Older Whites

Millennial Whites

Full Sample

2.20* (0.11)

2.74* (0.20)

−0.41* (0.07) 1,862 0.169

−0.67* (0.12) 666 0.209

2.20* (0.12) −0.26* (0.12) 0.54* (0.21) −0.41* (0.08) 2,528 0.186

Millennial RR × Millennial Intercept N Adjusted R2

Data: 2016 American National Election Study (ANES) Notes: *p < 0.05, standard errors in parentheses

interestingly, the coefficient for racial resentment is larger in magnitude for millennials (predicting a net change of 3.50 points), a finding significant in both the statistical and substantive senses. The third column of estimates in table 3.4 indicates that, among older Whites, the estimated difference between the most racially liberal person and the most racially conservative person is 2.14 (2.71 − 0.57), meaning that we would expect the most racially resentful person to rate Blacks as 2.14 points more violent than the least racially resentful person in that age group would. The interactive term between the indicator for a millennial respondent and the racial resentment scale tells us that the difference between the least racially resentful millennial and the most racially resentful

New Attitudes or Old Measures?  /  91

millennial is about 30 percent larger: 2.93 (2.14 + 0.79) points. The same can be shown in the models in table 3.4, which predict the old-­fashioned stereotype of Blacks as lazy. Here again, the effect for older Whites is sig­ nificantly smaller in magnitude (1.79) compared to the total effect for millennials (2.07). These effects are illustrated in figure 3.4. For our final set of analyses, we examine the differential effects of racial animus across generations on three racialized (policy) issues. Specifically, we examine one item that has been on the ANES for several decades and two items that were new in 2016. The first question asks respondents to place themselves on a seven-­point scale regarding how much the government should assist Blacks, which we recode to run from 0 (Blacks should help themselves) to 6 (the government should help Blacks). The second variable is made by taking the difference between how much discrimination the respondent believes Whites face in the United States and how much discrimination the respondent believes Black Americans face. Each of the composite variables asks the respondent to indicate, on a five-­point scale, how much discrimination each group faces, from 0 (“none at all”) to 4 (“a great deal”). By subtracting how much discrimination the respondent believes Blacks face from how much discrimination the respondent believes Whites face, we can estimate the degree to which a respondent believes Whites face more discrimination than Blacks. Thirty-­five percent of Whites believe that members of their racial group are discriminated the same amount or more than Blacks, with the majority of those believing that the groups are discriminated against equally. Using this measure allows us to test the notion that younger Whites may be adopting a color-­blind racial ideology, which leads them to believe in the notion of “reverse racism.” Finally, the third column in table 3.6 examines how racial resentment shapes affective ratings of Black Lives Matter, a social movement and global network that campaigns against anti-­Black racism (Taylor 2016; Lopez Bunyasi and Smith 2019b). This variable is a simple “feeling thermometer” score and ranges from 0 to 100; higher scores indicate greater positive affect. If White millennials are more racially progressive than previous generations, we should expect them to have higher affective ratings of the group. Specifically, if millennials are more racially progressive, we should see positive coefficients for both the millennial indicator variable and the interaction term. Again, each of the results supports our hypothesis. Not only does racial resentment have a strong correlation with thinking that “Blacks should help themselves” but also the effect is again significantly larger for White

3.4. Visualization of Regression Estimates from Tables 3.4 and 3.5

New Attitudes or Old Measures?  /  93 Table 3.6.  Racial Resentment, Generational Status, and Attitudes toward Racialized Targets

Racial Resentment Millennial RR × Millennial Intercept

N adj. R2

Aid to Blacks

Discrimination

Black Lives Matter

−3.99* (0.12) 0.58* (0.13) −0.66* (0.20) 4.51* (0.08)

2.53* (0.11) −0.29* (0.11) 0.52* (0.18) −2.69* (0.07)

−64.90* (2.12) 1.19 (2.25) −8.49* (3.59) 80.86* (1.40)

2,302 0.459

2,471 0.289

2,530 0.378

Data: 2016 American National Election Study (ANES) Notes: *p < 0.05, standard errors in parentheses

millennials. Despite the most racially liberal millennials having a higher level of support for aid to Blacks, the most racially resentful millennials are statistically indistinguishable from older respondents who score just as highly on the scale. The same is true for perceptions of discrimination: higher levels of racial resentment are associated with a respondent perceiving little difference in the levels of systemic discrimination faced by Blacks and Whites in the United States. Again, while the most racially liberal millennials acknowledge that Blacks face higher levels of discrimination, as evidenced by the negative and statistically significant coefficient for the millennial term (−0.29), this generational difference is ultimately washed out among those who are the most racially resentful. Finally, with regard to Black Lives Matter, White millennials are no warmer toward the group than older Whites are. The estimated feeling-­ thermometer score for the least racially resentful older Whites is about 81 degrees, while the most racially resentful older Whites would be estimated to rate Black Lives Matter at just under 16 degrees. The same effect for millennials, however, takes them from 82 degrees to about 9 degrees. This can be seen graphically in figure 3.5. Across each of the last set of tests, we see that among millennials, the marginal effect of latent racism has a stronger effect on their racialized policy preferences and attitudes. Looking at the interaction term in each model, we see that the slope for each group is significantly steeper. Ultimately, we find that the virulence of racial animus is heightened among millennials.

0.00

0.25

0.50

0.75

1.00

0.25

0.50

0.75

1.00 0.00

0.50

0.75

Racial Resentment

0.25

Support Black Lives Matter

1.00 0.00

0.25

0.50

0.75

Support Government Helping Blacks

3.5. Visualization of Regression Estimates from Table 3.6

Source: American National Election Study (2016), Non−Hispanic White Respondents Only *difference in slopes is statistically different from 0; p < 0.01

0.00

Believe Blacks are Discriminated More than Whites

Effects from Table 3.6 (All Variables Recoded to Run from 0−1)

1.00

Older Whites

Millennials

Group

New Attitudes or Old Measures?  /  95

In some ways, this perhaps speaks to the events of Charlottesville and to groups like the Proud Boys, a White nationalist organization.

What Is Going on Here? The extant literature reveals that it is at the intersection of cohort replacement and changes in social norms that we are likely to see differences in the level, structure, and role of racial attitudes. Major demographic, economic, political, and social changes led scholars like David Sears and Donald Kinder to develop new measures of racial attitudes. In the decades since the racial resentment scale was first developed, there has been another set of significant changes in America that would likely alter the nature and structure of racial attitudes: demographic shifts that have brought the United States closer to becoming a minority-­majority country, major increases in Black and Latinx political representation, the decline of manufacturing jobs in urban centers, the increase in numbers of members of racial minorities attaining higher education, and the election of the first non-­White US president are only a few examples. With these phenomena in mind, we may need to consider the fact that the structure of Whites’ racial attitudes may also be changing. At the time the racial resentment scale was developed (in the late 1970s and early 1980s), the survey items used were read and understood as subtler ways of tapping racial animus than the traditional biological-­racism or social-­distance questions. Today, in 2020, we are in a new era of racial politics, where the language of racial attitudes has a different structure from the one developed in the period immediately after the civil rights era. Researchers, pundits, and journalists have noted that there has been a shift in the acceptable expressions of racial attitudes over the past two de­ cades or so. This shift is largely marked by an increasing desire for political correctness and avoidance of race (Jackson 2008). The avoidance of discussing race largely stems from the perception that talking about race these days is “fraught with the risks of misunderstanding and social sanction”; the social norms of contemporary society guide Whites’ motivation to “avoid the appearance of prejudice” (Apfelbaum, Sommers, and Norton 2008, 918). Today, race and racial attitudes are more likely to be discussed in a “color-­ blind” way, if they are discussed at all. Frankenberg explains that a color-­ blind perspective is a “mode of thinking about race organized around an effort not to ‘see’ or at any rate not to acknowledge race differences,” as this is the “‘polite’ language of race” (1993, 142). An emerging consensus among scholars is that “color-­blindness” is currently the basis for contemporary

96 / Chapter Three

American racial ideology and characterizes the environment in which young Whites have been socialized (Bonilla-­Silva 2014; Bonilla-­Silva, Lewis, and Embrick 2004; Carr 1997; Forman 2004; Frankenberg 1993). Scott Blinder has developed a theory of “two-­tracked socialization,” which explains what we are likely to see empirically among young people (Blinder 2007). On the one hand, they have been socialized to avoid race, but they may face some cognitive dissonance because they do not have the language to explain racial inequalities. As such, we may see that young people are likely to be incredibly supportive of symbols that demonstrate that they are anti-­racist, but we may also see that they are less likely to pursue policies that ameliorate racial disparities because they do not have a firm grasp on the role of institutional, systemic, and structural factors that lead to racial disparities. For example, when asked how much discrimination various racial groups face, millennials and older Whites all recognize that certain groups are discriminated more than others. The 2016 ANES asked respondents, “How much discrimination is there in the United States today against each of the following groups?” Of the four groups mentioned, Whites were seen to be discriminated against least, followed by Asians, and then Hispanics. Blacks were perceived to experience the most discrimination in America. In addition to being asked about those groups, respondents were also asked whether they had experienced any discrimination on the basis of their race or ethnicity. Here again, millennials differentiate themselves from older Whites. Among all non-­Hispanic White respondents, millennials were more likely to report having been discriminated against on the basis of race. This is shown in figure 3.6 and speaks exactly to the cross-­pressures that millennials experience; not only are they more aware that Blacks and Hispanics are discriminated against, they also believe that they have experienced racial discrimination. Taken together, this leads us to believe that the measures of racial attitudes we often use in political science may fail to capture the sentiments that arise from being socialized in an era of color-­blindness. In addition to the fact that the racial resentment items reference specific historical events and racialized policies that lie outside of White millennials’ memory or knowledge (Dillon 2011), it is important to note that asking about race, and about “Blacks” specifically, raises social-­desirability red flags for younger respondents in a way that they do not for older Americans. When the racial resentment scale was developed, “compared with most efforts to measure racial animosity, these questions . . . [appeared] rather subtle” (Kinder and Sanders 1996, 106), but considering today’s environment, these questions

New Attitudes or Old Measures?  /  97

Perceived Discrimination

Whites' Perceptions of Discrimination in America (2016) A Great Deal

None

Yourself*

Whites

Asians

Hispanics*

Blacks*

Mill. Older Whites Whites

Mill. Older Whites Whites

Mill. Older Whites Whites

Mill. Older Whites Whites

Mill. Older Whites Whites

Source: American National Election Study (2016), Non−Hispanic White Respondents Only *difference in means is statistically different from 0; p < 0.01

3.6.  White Americans’ Perceptions of Racial Discrimination, by Generation

may miss the mark, as the mere mention of race is unacceptable, especially for younger Americans. Anecdotally, after the 2014 midterm election, voters were reportedly taken aback by the questions that compose the racial resentment battery; some felt that the asking of the questions implied that the interviewees were racists (Jordan 2014). Psychologists have also recorded this tendency to avoid talking about or even mentioning race. For example, Evan Apfelbaum and his colleagues (Apfelbaum et al. 2008; Apfelbaum, Sommers, and Norton 2008) show that Whites adopt “strategic color-­blindness,” or an effort to completely avoid mentioning race even in a task where pointing out someone’s race is actually helpful. They also found that younger children (aged eight to nine years) are able to outperform older children in a basic categorization task, where acknowledgment of race facilitates performance. Their research reveals that children as young as ten years old recognize the potential social sanction of mentioning race and avoid doing so even at the expense of successfully completing a task. These scholars suggest that we may be seeing a “critical transition in human social development” around the issues of race and racism (Apfelbaum et al. 2008, 1513).

There is a great deal of evidence that suggests that the racial resentment battery is, indeed, a coherent belief system that is different from old-­fashioned racism and predicts White Americans’ candidate evaluations and attitudes across a range of policy domains, but our results suggest that the nature of the racial resentment battery as a measure of racial attitudes may be changing. In the face of increasing heterogeneity brought forward by a new cohort

98 / Chapter Three

of Americans, namely millennials who have been socialized in a very different era, this measure may not tell the whole story about racial attitudes in the twenty-­first century. White Americans are now being socialized to avoid talking about race (Apfelbaum et al. 2008; Apfelbaum, Sommers, and Norton 2008; Lewis 2001), are more likely to feel that Whites are discriminated against (Dietrich 2015), and are less likely to understand the history of racial inequality in the United States (Dillon 2011). Our findings illustrate the implications of such socialization by revealing that the structure of young Whites’ racial attitudes is quite different than that of older Whites. Young people are likely to have a different set of ideas and references that come to mind when they are asked the questions in the racial resentment scale, and likely have a different set of social norms about what is acceptable to talk about as far as issues of race are concerned. The evidence we have presented here suggests that the heterogeneity that millennials introduce will reduce the reliability of the racial resentment scale over time. What this ultimately means is that we may need to think about developing a new measure of racial attitudes. Psychologists and sociologists have already begun building a foundation to measure “color-­ blind” racial attitudes (Neville et al. 2000; Bonilla-­Silva 2014) as well as racial attitudes that may arise as Whites become more cognizant of their racial identity due to the vast demographic changes that have occurred over the past three decades and are projected to occur henceforth (Spanierman and Heppner 2004; Spanierman et al. 2006). We build on this foundation in part 2. There, we make a concerted effort to ascertain how millennials think about racial matters. We do this by simply asking young White Americans about their conceptualizations about race, racism, and racial disparities; through the words of White millennials, we are better able to understand why they are neither as racially liberal as they believe they are nor as progressive as many have characterized them to be. Further, we are able to generate themes that help us to build a more accurate measure of contemporary racial attitudes.

F o ur

Millennials on Racism

We spend a great deal of time with millennials. Many of our colleagues and (former) students are members of the millennial generation. Over the course of our interactions with them, we have found a number of inconsistencies between the average American’s expectations about the millennial generation’s racial politics, and reality. This gap is often illustrated in very ugly ways. For example, every year around Halloween, we see a new set of images from college campuses across the country of White students in blackface and in costumes mocking entire racial and ethnic minority groups. From time to time, we learn of nooses hanging in public spaces on campuses like Duke University or University of California, Berkeley. There are even examples of more overt expressions of old-­fashioned racism, like in March 2015, when members of an all-­White University of Oklahoma frater­ nity, Sigma Alpha Epsilon, were filmed chanting, to the tune of “If You’re Happy and You Know It,” “There will never be a nigger SAE. There will never be a nigger SAE. You can hang ’em from a tree, but it will never start with me. There will never be a nigger SAE.” On a seemingly regular basis, young White people’s overtly racist remarks are moved out of the shadows of their “Finstas” (fake Instagram accounts) and into the “public” domain to be judged by all. In addition, Americans learned that the police officer who fatally shot Michael Brown was only twenty-­eight years old, himself a millennial. Similarly, the police officer who killed twelve-­year-­old Tamir Rice (who was holding a BB gun) was twenty-­six years old, another millennial. Dylann Roof, who killed nine African Americans in a Charleston, South Carolina, church, is also a millennial. Jason Kessler, who organized the Unite the Right rally in Charlottesville, Virginia, is a millennial, as is James Alex Fields Jr., the man who murdered Heather Heyer during the rally. A year later, women

102 / Chapter Four

like Avialae Horton, who is on the cusp between the millennial generation and Generation Z, helped to organize a second Unite the Right rally in Washington, DC (Pitofsky 2018). To be clear, we are not equating racist chants with racially motivated hate crimes, but we do want to point out the wide range of nonneutral, racist sentiments and behaviors among young White millennials that are often left out or ignored in narratives of unfettered racial progress, which is often attributed to that generation. In all, these incidents reveal that young Whites are not exempt from perpetuating racial stereotypes, hierarchies, or inequalities. Looking retrospectively, it is clear that the United States has become a more racially progressive place in many ways, but there is also a great deal of room for improvement. How do we explain this tug-­of-­war between the value of racial egalitarianism and the persistent manifestations of racism? In this chapter and the next, we test our central hypothesis: the United States is in a state of racial stasis because though, superficially, White millennials appear to be more racially progressive than their predecessors, they are not pulling aggregate White racial attitudes in a more progressive direction. We test this by analyzing the way that millennials express racial sentiments and provide explanations of ongoing inequality. We find that White millennials are not relying on “older” versions of racial animus or explanations of persistent racial disparities but instead employ a new set of racial logics—­a mix of color-­blind racial ideology and diversity ideology. Both allow millennials to appear to be more racially progressive, because they can claim to “not see race” while also appreciating racial diversity. Nonetheless, these ostensible humanist views are mitigated by what we see as a set of countervailing forces. For every harbinger of progress expressed in White millennials’ attitudes, there is something that signals peril for that progress. We can liken these dynamics to push-­and-­pull factors of aggregate racial progressivism. We were able to uncover four of these push-­pull dyads in our analysis of forty-­three interviews with White millennials. We refer to these pairs as the moving walkway of racism, the diversity dilemma, the empty knapsack, and the paradox of generations. We provide examples of the first three in this chapter, and the fourth rings more prominently in the next. Together, they help us to gain a better understanding of why we see stagnation in racial progressivism among the most recent generations of White Americans.

The Moving Walkway of Racism Being called a racist in the twenty-­first century is not simply an insult; many White Americans view it as a moral condemnation (Mayorga-­Gallo 2019;

Millennials on Racism  /  103

Picca and Feagin 2007). One could argue that the desire to not be viewed as a racist is a good thing. Members of the millennial generation have worked quite hard to not be viewed as racist; in polls they are the generation that has the highest support for notions of egalitarianism (Pew Research Center 2010). Here, we have an important push factor toward racial progressivism: a value of racial equality. The pull factor lies in the notion that young people, especially White millennials, have neither a firm grasp on what racism is nor a wide enough scope to see all of the ways that racism works in American society. This dyad is well illustrated by Beverly Daniel Tatum’s analogy of racism as a moving walkway: I sometimes visualize the ongoing cycle of racism as a moving walkway at the airport. Active racist behavior is equivalent to walking fast on the conveyor belt. The person engaged in active racist behavior has identified with the ideology of White supremacy and is moving with it. Passive racist behavior is equivalent to standing still on the walkway. No overt effort is being made, but the conveyor belt moves the bystanders along to the same destination as those who are actively walking. Some of the bystanders may feel the motion of the conveyor belt, see the active racists ahead of them, and choose to turn around, unwilling to go to the same destination as the White supremacists. But unless they are walking actively in the opposite direction at a speed faster than the conveyor belt—­unless they are actively antiracist—­they will find themselves car­­ ried along with the others. (Tatum 2003, 11–12)

White millennials—­and most White Americans—­are relatively passive bystanders. In many ways, we could say they are structural or complicit racists (Lopez Bunyasi and Smith 2019b). They are not actively pursuing racist outcomes, but they may be supporting policies and candidates that further embed racial inequity in our laws and institutions; or, at the very least, they are not doing anything to prevent or mitigate racism from influencing outcomes. So why are they sitting on the sidelines? The way that people make meaning of race, racism, and racial inequality will influence what they think should or could be done about the negative outcomes that stem from racism (Kluegel 1990; Kluegel and Bobo 1993; Kluegel and Smith 1981; Kluegel and Smith 1986). For example, if individuals believe that racism is simply a set of overt, negative, biased behaviors or that racism is a matter of the hearts and minds of individual Americans, then people may also believe that there isn’t much that can be done about racial inequality except wait for old people to die. Alternatively, if people

104 / Chapter Four

feel that racial inequality stems from racial minorities’ failures to live up to the opportunities presented to them by legislation enacted after the civil rights movement, then we might expect them to rally around neoliberal policies, or policies aimed at correcting behavior and “culture.” Still another possibility is that if most Americans adopt a structural interpretation of racism and view the United States as a racialized social system, or a society “in which economic, political, social, and ideological levels are partially structured by the placement of actors in racial categories or races” (Bonilla-­Silva 1997, 469), then we might also see that they will demand a set of policies aimed at undoing the political, economic, social, and perhaps even psychological inequalities that undergird the way American society is structured. We asked respondents, “What is racism?” or “What comes to mind when you think about racism?” In addition to answering these questions, re­ spondents often preemptively volunteered stories about what they view as racist acts or attitudes; they shared their views about what is not racist or racism and discussed how relevant they believe racism is in American society. We found that twenty-­nine (67 percent) of our forty-­three (White millennial) respondents understand racism in America as a set of ideas, beliefs, or overtly biased actions (mostly toward African Americans). There were ten young people (23 percent) who felt that racism is a thing of the past, so they did not have much to say about the shape of contemporary racism, and there were just four (9 percent) who understood racism as structural, institutional, or systemic. One thing that is important to keep in mind is that these responses are essentially “top of the head” responses, to borrow Zaller and Feldman’s term (1992). So, it’s not that millennials would never acknowledge the structural components of racism, but they are likely to do so only when they are prodded to consider these components. For example, on the 2016 Cooperative Congressional Election Study, some respondents were asked a similar question but provided with various possible answers. There, 75 percent of millennial Whites agreed that racism can be understood as “structural inequalities that advantage Whites”; two-­thirds of older Whites responded similarly. These proportions might seem high but are not when compared to the nearly 90 percent of White millennials (and 83 percent of older Whites) who believe that racism is best understood as “individuals acting in a way that disadvantages racial minorities.” Needless to say, when given the opportunity to provide a more nuanced response, millennials tend to do so; however, when they are not prompted to consider the structural aspects of inequality, they are remiss to consider them. We care about this because conceptualizing racism primarily as individ-

Millennials on Racism  /  105

ual prejudice or ignorance leads to a cascading set of beliefs, perceptions, and ideas about the role of race in American society. Our data reveal that (1) some millennials believe that racism is not that important of an issue or that if it is an important issue, it is exclusive to older Whites or to the southern region of the United States; (2) many millennials believe that racism is relevant, but it is really just a ghost—­a lingering issue from the past—­that haunts us today; and (3) most millennials view or explain racial phenomena (e.g., racial segregation, racial disparities in education) through the lenses of socioeconomic status and class. These three trains of logic minimize the perceived effects of race and racism in the minds of millennials, which is likely to lead them to be passive bystanders on the moving walkway of racism. What Is Racism? Although our respondents recognized that the way racism manifests itself over time changes, many still relied on relatively outmoded conceptualizations of racism. In 1945, anthropologist Ruth Benedict defined racism as “the dogma that one ethnic group is condemned by nature to congenital inferiority and another group is destined to congenital superiority” (1945, 87). This is a fairly outdated definition of racism, and while some of our respondents made comments about the perceived natural athletic superiority of African Americans, few relied on this biological basis of racism. For example, Easton, a twenty-­six-­year-­old from Baltimore, noted during his interview, “I just kind of realized that that type of racism, the type where Black people are stupider, and they’re not able to run the country, is gone.” Similarly, when we asked Dylan, a twenty-­four-­year-­old man from Cleveland, what comes to mind when he hears the word “racism,” he explained, “I don’t think we’re at this dictionary of races, family of mankind of—­I think we’ve broken out of the nineteenth century, ‘Let’s rank the races’ thing, but maybe not all the way. It wouldn’t surprise me if there was a state representative from Oklahoma who is compiling a list as we speak, but I do think that there are different racial tensions around different ethnic groups.” Dylan revealed a couple of things here that came up for many of his generational peers. First, he realized that there has been a change in the way that racism is expressed over time. Second, he described racism as “tensions around differ­ ent ethnic groups.” Finally, he suggested that if there are some old-­school biological racists in this country, they are likely to be older people from cer­ tain parts of the country—­in states like Oklahoma, of all places. Since Ruth Benedict’s writing, we have seen many updated renditions and definitions of racism appear. For example, Pierre van den Berghe explains in

106 / Chapter Four

his 1967 book Race and Racism: A Comparative Perspective that racism is “any set of beliefs that organic, genetically transmitted differences (whether real or imagined) between human groups are intrinsically associated with the presence or the absence of certain socially relevant abilities or characteristics, hence that such differences are a legitimate basis of invidious distinctions between groups socially defined as races” (11, emphasis added). In 1990, Richard Schaefer explained that racism is “a doctrine of racial supremacy, that one race is superior” (16, emphasis added). These definitions were developed in different eras, but what they share is the idea that racism is a set of beliefs or attitudes about particular groups, and consequently, these beliefs lead people to develop “negative attitudes towards an entire group of people” (Schaefer 1990, 53). Here, racism is rooted in the hearts and minds of some Americans, and expressions of racism are revealed in the form of unfair bias, discrimination, and negative stereotypes. However, more recently, scholars, journalists, and activists have tried to show that racism is not just about individual-­level attitudes. There has been a concerted effort to educate people about how racism is also embedded in social systems, institutions, laws, and public policies as well as in day-­to-­day interactions and individuals’ feelings about other groups (Bonilla-­Silva 1997, 1999, 2014; Feagin 2013; Katznelson 2005; Oliver and Shapiro 1995; Omi and Winant 1994; Kendi 2016; Taylor 2016). Despite this set of revelations, the narrower, older conceptualization of racism was most dominant among our millennial respondents. Here are some examples: Adeline (20, Lexington, MA): I would say, let’s see, racism, people who, I guess people who stop others from getting the same things that they want that are of different race or ethnicity. Callie (20, Riverside, IL): Hate speech, prejudice, segregation, all sorts of fallacies of thought, stereotypes. Charles (23, Fishers, IN): Like lynching, like old pictures of old lynchings, and things like that or racism. I’m a big sports fan. Like in Europe seeing soccer players get bananas thrown at them or racially abused because they’re African American [sic]. That’s generally what I think of. Overall abuse of someone just because of the race that they are. Mollie (20, North Carolina): Black people and jokes and the N-­word. Just a separation between the two [races, Black and White].

Millennials on Racism  /  107

In each of these cases, racism is seen as a matter of individuals’ attitudes and behaviors, and stereotyping is also a prevalent way to conceptualize racism. Reliance on this narrow racial logic allowed two things to happen. First, respondents were willing to call their own actions racist while not seeing their part in reproducing inequality as a terrible thing. For instance, Mackenzie, a twenty-­year-­old woman from Lexington, Massachusetts, noted, “I think [racism] definitely influences people. I think individuals, and like I could even see myself like still like applying stereotypes and stuff.” Secondly, and relatedly, since stereotyping and categorizing are things that our brains do to cope with the complexity of the world around us, racism in the form of stereotyping was not perceived to be that bad in the eyes of some of the respondents. This sentiment is best illustrated by Michael and Kaitlyn, both from suburbs of Boston. First, we have twenty-­year-­old Michael’s statement: So, I think it’s more like I said, it’s part of society because of how deep it is within, you know, human society and how humans are genetically or—­we’re social animals, right? Racism is meant to be in our DNA, genetics, whatever. It’s meant to be in human society because we’re made—­our brains force us to stereotype. It’s human nature. It’s just the way humans are, as much as you can control it, you know, I think racism is just a part of it. This is my answer is when I think of racism. I just think of society. It’s part of us. It’s not really that much to talk about in some sense.

Kaitlyn mirrored this sentiment: “And I feel if you’re, if people are still gonna be racist, it’s something deep down inside. Like some personal issue that you can’t solve the whole world, obviously.” For this group, racism is reduced to a matter of individual psychology and, for some, perhaps even genetics. Nearly all of the respondents were college-­educated or on their way to college, so they were likely to have all been exposed to the concept of stereotypes. In a way, they had just enough information to do quite a bit of damage. For instance, in the case of George Zimmerman (also a millennial), his defense team persuaded jurors that Trayvon Martin, a Black teenager in a hoodie, fulfilled a stereotype about who is dangerous, and therefore Zimmerman had simply reacted in a way that any “reasonable” adult who wants to protect his neighborhood and family might have. Zimmerman’s legal defense team, Michael, and Kaitlyn all employed the same rationale: we cannot help but be biased against members of out-­groups; therefore, racism is just an inevitable part of the way the world works, and there is very

108 / Chapter Four

little we can do to fix it. This is the logic of those passive passengers on the moving walkway. It’s Black. It’s White. In addition to holding the notion that racism is a series of individual beliefs or behaviors, White millennials tended to narrow their conceptualization even further by suggesting that racism is generally an issue that Black Americans face, and that if there were a culprit or two, Whites (especially men) and sometimes Blacks were the prime suspects dispensing “racism.” When we asked Claire, from Marlborough, Massachusetts, what comes to mind when she thinks about racism, she said, “I guess I think of the White man coming down on the Black man at the simplest term.” Similarly, Ella, a nineteen-­year-­old from a Boston suburb, explained, “Like a stereotypical White man is the epitome of racism to me.” Harper, a twenty-­one-­year-­old woman from Columbus, Indiana, also explained, “But I think America as a whole is more racist towards African Americans. I don’t know why—­we had the whole Pearl Harbor thing with Japan so, I don’t know why we’re not racist towards Japanese. Maybe we are, and I’ve just not seen it? I don’t know, but we have a serious issue with Black people so when I think of racism, I think of African Americans.” There was a pervasive notion that the world works as Black versus non-­Black, whereby all other racial and ethnic minority groups are relieved from the inequalities that originate from a racialized social system. Moreover, a number of individuals in the Boston area and in Indiana made comments suggesting that Asian immigrants and Asian Americans had “out-­Whited” White people, since people of Asian descent are moving into upscale, wealthy neighborhoods and now take up a disproportionate number of spaces in high school advanced-­placement classes and in elite colleges, domains previously dominated by Whites. As such, there was a sense that Asian Americans had overcome what White millennials defined as racism. Even fewer individuals discussed Latinx communities, which form the largest ethno-­racial minority group in the country. There was also a number of people who grappled with whether Whites are the only people who can be racist. Ava, a twenty-­one-­year-­old woman from Watertown, Massachusetts, explained, “I think of racism in the US between Caucasians and African Americans. But then I also think about how it can also go the other way. Like African American people can be racist against White people. I don’t know.” Likewise, Caden, who was born in Alabama but raised in North Carolina, said, “But I mean, I’ve seen people think rac­­ ism is like Whites hating on Blacks, but I’ve seen like Blacks hating on

Millennials on Racism  /  109

Whites just as bad like in Mississippi, where my mom grew up. It’s like that. And they will not talk to a White person, like they don’t associate with them.” Again, what these responses reveal is that young White people have a very narrow understanding of what racism is. For the great majority of them, they see racism as overt, negative forms of bias or discrimination perpetrated by individuals. They tend to see it as a problem that only Blacks face. Harper suggested that Asian Americans are above the fray, and there was no mention of any other racial or ethnic minority groups in the respondents’ answers to the questions posed about what they understand racism to be. This myopic view of racism excludes a serious consideration of the ways in which Latinx and Asian Americans (and other ethno-­racial minority groups, such as American Indians) are doled out disadvantages disproportionately in a racialized social system, even if those disadvantages are different from those of Black Americans. Since young White people are focused on blaming Blacks and Whites for their parts in perpetuating racism or are focused on the progress that Blacks have attained, they fail to see the patterns of racism that exist among other underrepresented minority groups. Nothing in My Lifetime Because most White millennials believe that racism is a phenomenon derived from the ignorant biases and stereotypes of a small proportion of White Americans, it is clear why so many of them do not “see” racism. Approximately one-­fourth of our respondents expressed a belief that racism is a thing of the past, and the majority of our respondents indicated that the only way one could ever “see,” experience, and thereby come to understand what racism means would be to grow up in the South or be born during a bygone era. When Alaina, twenty-­two, from Alexandria, Indiana, was asked what comes to mind when she thinks of racism, she simply said, “The civil rights movement, probably.” Matthew mimicked this quite closely: “Civil rights movement of the ’60s. Yeah, civil rights movement.” Joshua, a twenty-­five-­ year-­old from Shelbyville, Indiana, said, “Old White people and Black people. Old images. Nothing of my lifetime.” Maria, a nineteen-­year-­old woman from Zionsville, Indiana, said, “I think most about African Americans just because that is a large part of our history as far as slavery and things, so that would probably be what comes to mind most.” Similarly, Caden explained, “First thing off is, KKK, you know, 1950s. That’s the first thing that jumps into my head.” Caden made a historical reference, but then he also noted that “I mean yeah. It’s like, I’m from Alabama. I mean I know several people

110 / Chapter Four

down there that are still like in the KKK and stuff.” So, even though Caden said he knows people who are in the Ku Klux Klan, he still viewed racism as a relic of the past. As a final example, Kaitlyn noted that “I honestly, like I just think of like Black people and like discrim—­their discrimination. Like you know, like that whole, like, worlds apart. . . . Because, yeah, because like I don’t see it like, like today, like when you say, like I can’t, like I don’t have an image of like today’s [racism].” Our respondents anchored their ideas about racism in the sea of images that most American schoolchildren are exposed to during Black History Month or on Dr. Martin Luther King Jr.’s birthday: police with fire hoses and attack dogs, lynching, and groups of White adults yelling at Black children as they try to desegregate Southern schools. Another way that White millennials minimized the ubiquity of racism was to suggest that if racism, of any kind, is still alive and well, it is likely to remain where racism was supposedly born and has always resided: in the South. For example, Brody from Williamstown, Massachusetts, the oldest millennial we interviewed, explained, “I also think you’re not as exposed to it in this region.” For him, racism is not something that you’re likely to see in New England. Scarlett from Westchester County, New York, made mention of this regional aspect of racism in her answer to a question about the extent to which she believes White Americans have privileges because of their race: “I’m not sure in today’s society if [Whites having racial privilege is] entirely true, but definitely before there was so much diversity and awareness just of racism in general. Also, I’m only speaking from a northeastern perspective. I’ve only been to the South once, and I was little. I don’t really know what it’s like in other parts of the country.” Relatedly, Dylan, a twenty-­ four-­year-­old college graduate, told us this story: Oh, I know a couple of years ago that there was an incident at Oberlin where someone carved a swastika or drawn a swastika on the door of the [inaudible] House or one of those with a Jewish-­life houses. It prompted a relatively school-­wide discussion about race. [The place] where I went to school, had a couple of issues around Somali students who wore head scarves, feeling like they were being kind of essentialized or not taken seriously. In a lot of the sexual assault on college campus discussions I feel like there is an undercurrent of race, especially in cases that I’ve heard about where a White assailant, a White student and a Black student, or a White student and a, any sort of interracial potential sexual assault happens that tension tends to be really prevalent and that happened at [my college] while I was there.

Millennials on Racism  /  111

Then, after he told us about the role of race in the everyday interactions of women and individuals of color at his college, a liberal-­arts school in the Midwest, he explained, “Yeah, it’s definitely I didn’t, I didn’t grow up in the Jim Crow South so [I’ve had] a sheltered, upper-­middle-­class kid’s view of what American racism looks like.” Dylan managed to negate these overt instances of racism because they occurred outside of the South and, further, he suggested that because he didn’t grow up in “the Jim Crow South,” he has not been exposed to racism. Maria provided a similar sentiment when she noted, “Okay. I think it’s better than it was. I think that racism as a nation, as a whole, has pretty much gone away. I mean we might be having a very different conversation if we still—­if we were sitting in the South or if we were in an area that was very like racially segregated or discriminated against. But I think as a whole there have been a lot of things positively done for the Black community.” In a similar vein, Reagan, a twenty-­one-­year-­old woman from Elk Grove Village, Illinois, suggested, “I think in some areas that’s true. Definitely in the South, maybe more—­more, so. The views haven’t completely changed as much as in other places.” And Riley, a twenty-­one-­year-­ old woman from Connecticut, shared this story: “Yeah. I think—­I guess some of the—­some advantages are more visible than others. And a lot of this is just things I’ve read, so—­I mean, I think—­so I was—­I took this class on stereotypes last semester, and we watched a video. I think it was 2011, which is really depressing, but there was a segregated class somewhere in the South. I was like, really?” This attitude wasn’t just among those who lived in the Northeast or the Midwest. Even Mia, a twenty-­year-­old North Carolinian, participated in this conversation: “[Racism is] definitely still relevant, and more particularly in places, in small, predominantly White, Southern towns, where people, once again, are living in the past and haven’t accepted that things have changed, and still have prejudice towards diversity.” What is fascinating about these responses is the notion that racism in the form of racial prejudice, racial disparities and disadvantages, and racial segregation is perceived as particular to one part of the United States. When we look at the history of the United States, we can see that Black and Brown people throughout the country have faced severe interpersonal discrimination and the brunt of structural racism. If we look more closely today, we would note that many of the protests concerning racial inequalities, especially as they concern police brutality, have occurred on the West Coast (e.g., in response to the murder of Oscar Grant), in the Midwest (e.g., Philando Castile), and in the Northeast (e.g., Eric Garner). And in recent cases of police shootings of unarmed Blacks, those in the South have resulted in more

112 / Chapter Four

vigorous prosecution of White police officers than those in other parts of the country (Wilkerson 2015). What’s more, most of the country’s most segregated schools and highest levels of racial disparities in incarceration are not in the South (Reardon and Owens 2014; Camera 2016; Hannah-­Jones 2014; The Sentencing Project 2017). The geography of racial animus has transformed over time, whereby states in the Midwest and even the Plains states are now challenging the South’s dominant reputation for fostering the most negative racial sentiments (Kreitzer, Smith, and Suo 2018). Additionally, researchers examined what they considered racist posts on Twitter. What they found was that tweets that include overtly negative racial language and slurs—­very similar to our respondents’ conceptualizations of racism—­were prevalent across the entire eastern United States, not just the South (Edwards 2013). Needless to say, there is plenty of evidence that racism in all its forms is prevalent across the fifty states. It’s Still Important, But . . . Because younger White Americans have a constrained understanding of what racism is (and where it is likely to occur), it becomes apparent why many millennials are likely to disassociate the prevalent and ongoing racial disparities in the United States with contemporary systemic features like state actors and existing political institutions and laws. In other words, White millennials believe that racism is relevant, per se, but since racism is mostly something from the past, our job now is to manage its lingering effects. We asked respondents, “There is still a lot of talk about slavery and racism in the US. Do you think these issues are still relevant, or are they things of the past?” About 20 percent of the millennial respondents suggested, in one way or another, that these issues were not relevant. For example, Brooke, a twenty-­two-­year-­old hospice nurse from North Carolina, responded to our question this way: No, I feel like that they [Blacks] make them relevant, and they still carry a chip on their shoulder for something that I had nothing to do with. If you’re going to walk by me and be mad that I’m White, why? Because I did not have a cotton field that I hired you to work in. Like, that’s something that happened a long time ago, and I’m sure that people up north probably killed my ancestors, but I don’t go up north and be like, “Listen, you Union people,” you know? I mean, you have to get real, life goes on and stuff happens, but it’s not the present day fault, and you need to quit trying to get a free ride off of it.

Millennials on Racism  /  113

Thereafter, Lily, who interviewed alongside Brooke, succinctly stated, “I com­ pletely agree.” Charles also expounded on this issue: As far as slavery goes, I really, I don’t think slavery should be an issue anymore. It’s hard to tell people who were put through that or whose ancestors were that’s the way it is but I don’t think—­my family wasn’t even here when that went on. The African Americans that are alive today have never met anyone in their entire family who has ever been through racism itself or not racism but slavery itself. I don’t think you can really continue to talk about slavery as a major current issue when it happened 250 or 300 years ago. It’s just not well, over 100 years ago. There’s no one alive today that’s ever experienced it. So I don’t know that’s relevant as a topic as people make it out to be.

A number of respondents noted that sex slavery is probably a bigger issue to consider than historic chattel slavery because racial slavery is completely illegal. It is true that almost nobody alive right now in the United States has experienced chattel slavery, but the logic here is that since nobody currently living in this country has experienced slavery, then there isn’t anyone who has either benefited from or been disadvantaged by a four-­centuries-­ long era of slavery—­a logical and factual fallacy (Wilder 2013) expressed by only a few of our respondents. The majority of our respondents believed that race and racism, and to some extent the legacy of slavery, are still relevant today. Some answered our question in these terms: Jayce (22, St. Louis, MO): Well, I mean there’s absolutely still issues of slavery and racism, so they’re not a thing of the past. Hunter (21, Chicago, IL): Like even today we have a Black president [Barack Obama] and you still have people believing he is a Muslim from Kenya. That’s not because the wrong facts are out there, that’s because they’re racist and they don’t like—­like that’s part of it. I think we have incredibly virulent racism. I am stunned even at [my elite, Northeast liberal-­arts college] sometimes at how just like ignorant people are. Scarlett (19, Westchester County, NY): Slavery, more of the past. But racism definitely still exists, whether it’s embedded in people, whether they realize it or not. Like I know my grandma is inherently racist towards Asians just because she doesn’t even realize she is.

114 / Chapter Four Elena (21, Indianapolis, IN): And I think it’s easy to say in my position, “Oh, racism doesn’t exist anymore,” but it sure does. I don’t know how blatantly racist people are anymore, but there’s definitely an underlying hum of racism, and that’s cracked recently.

Here, we see that some White millennials do believe that racism and the effects of slavery influence the way the United States works today, but we also see that they tend to see racism in the form of individual acts and behav­ iors: Granny is racist, some people are “blatantly racist,” or ignorance leads to biased ideas about non-­Whites. What we tend to find is that largely because of this group’s narrow understanding of what these concepts—­race and racism—­are, White millennials are more likely to suggest that we need to keep these things in mind so that we do not repeat the past, and many of them advised that there is a fine line we must walk: yes, we must keep history in mind, but we must not let retrospection hold us back from potential progress. For example, in response to our question as to whether issues like slavery and racism are still relevant, Ella answered, “I think it’s important to talk about and to educate people about—­younger people—­but I don’t think it’s something that is helpful to hold on to.” Alaina from Alexandria, Indiana, responded, I think that they’re slightly a thing of the past because it’s not like—­at least other than [sex] trafficking which I don’t think that has to do with that—­but I think that [chattel] slavery is more like something in our history, and I think it’s important to remember your history and where you came from. But I think at the same time holding onto that is kind of holding back people from moving forward, if that makes sense? I don’t know. It’s something you can’t forget because it happened, but something that you kind of need to move past to better yourself.

As a final example, Brody, from Williamstown, responded to our question about historical injustices this way: That’s a really tough question. I feel that’s a really tough question. I’ve asked myself this a lot and I argue with myself because I think it’s important that we recognize what happened so we never do it again, but I also feel like by discussing it, it still brings awareness to the fact that it exists and I don’t know how to answer that question because I argue with it myself. When you teach people what slavery and racism is, you bring an awareness to it and I just feel like there’s a part of it that wishes you could just forget it and erase it, but it’s happened and it’s important and it needs to be recognized but I don’t know.

Millennials on Racism  /  115

Two things stand out here. First is that Alaina suggested that society should remember historical indignities and tragedies, but ultimately society is better off by not persistently bringing awareness to the issue. This sentiment mirrors large polls that show that many White Americans now believe that the best way to ameliorate racial disparities is by not talking about race and racism (MTV Strategic Insights and David Binder Research 2014). This lack of engagement with tough issues is an illustration of how color-­blind racial ideology can work to perpetuate a racialized social system. The second thing that we should highlight is that Brody wasn’t the only person who, at some point during the interview, got frustrated with the series of questions that we asked about race. Robin DiAngelo (2011, 2018) calls this stress “White fragility.” DiAngelo notes that White Americans tend to live in social environments—­which we will further elaborate on in the next section—­that protect them from the very kind of conversation that we have initiated here about racism, racial privilege, and racial inequalities. DiAngelo suggests that this insulated environment “builds white expectations for racial comfort while at the same time lowering the ability to tolerate racial stress,” thereby leading to a “state in which even a minimum amount of racial stress becomes intolerable, triggering a range of defensive moves” (DiAngelo 2011, 54). One of the defensive moves we saw most frequently occurred when several respondents would notify the interviewer that they are not racist themselves. Overall, the answers to the questions concerning how people understand racism and the extent to which they believe that racism is prevalent in society illustrate both color-­blind racial ideology and the moving walkway of racism. In this case, it is possible for young White people to recognize the existence of racism but simultaneously describe it as a relic of history, because they have such a narrow idea of what racism is. Rather than see racism as structural, institutional, or systemic—­whereby political, econo­mic, social, and psychological resources are allocated to the benefit of some racial groups and to the disadvantage of others—­they can cast racism off to a small segment of the American population (an old, and blatantly racist, grandparent), a time in history (seventeenth century to the 1960s), or a particular geographic region (the American South), and therefore be passive in the face of ongoing structural inequality. “It’s Like a Class Thing and Like a Money Thing” The last theme we outline that evokes the image of a passive bystander on the moving walkway of racism was demonstrated when our respondents

116 / Chapter Four

primarily viewed racism through the lenses of socioeconomic status and class. This idea is not new, but our conversations with White millennials reveal how prevalent this perspective has become in shaping Americans’ racial attitudes. A class paradigm of race suggests that “social divisions which assume a distinctive racial or ethnic character can be attributed or explained principally by reference to economic structures and processes” (Hall 1996, 17). Why does this matter? What we find is that White millennials tend to explain racial phenomena with ostensibly race-­neutral rationale. If racism is not perceived as a core mechanism of racial disparities, then people likely will not seek to address structural racism in the realm of politics and policy. Nearly half of the respondents primarily relied on class distinctions as a major explanation for racial phenomena. One common refrain is well represented by Mackenzie, who noted, “So I think like money, social class, defines you maybe more than race does now.” A large proportion of White millennials, like Mackenzie, felt race was just a cover-­up for the real underlying issue of class inequality. Skylar, a young woman from a wealthy, mostly White Boston suburb, provided another rendition of this claim: “I think there is a lot of class distinction which in the urban areas gets kind of covered up as racial issues, but it really is, ultimately I think, it’s all economic.” We found that, in general, young Whites tended to employ a class-­ stratification approach, which focuses on the social distribution of resources; this approach assumes that “individuals receiving roughly equal incomes, or partaking of equal quantities of wealth, are deemed to have similar ‘life chances’ and assigned to groups in a ‘status order’ or ranked hierarchy of ‘classes’” (Omi and Winant 1994, 27). For example, Easton, from Baltimore, suggested, “So like if a person from Canada comes here and becomes wildly successful, it’s not because they’re White but probably because they came from a middle-­class family in Canada.” Comparably, Daniella explained, Do I think that White people have certain advantages because of their skin color? I don’t believe it’s as simple as that. I think—­I really think it’s like a class thing and like a money thing. I think that I have more advantages because I grew up in an upper-­middle-­class home with two parents. But if I had grown up in the projects with a single mom and with a drug addict dad or a dad who’s in jail or whatever the circumstances were, I think that I would be way more limited with my opportunities. So I don’t know if I can attribute that to just race.

Race does not fully dictate the life chances of an individual or a group, but what both Easton and Daniella suggested is that if all things remain equal,

Millennials on Racism  /  117

we should expect equal outcomes among people of similar social classes, despite racial differences, though it is also worth noting how Daniella made the leap from growing up poor (“in the projects”) to having a parent who has some moral failings (“drug addict dad”). There is little empirical support for this logic. Research shows that even when we control for income, Whites have more wealth (Oliver and Shapiro 1995; Hamilton et al. 2015). Poor Whites do not live in hypersegregated neighborhoods predominantly comprising other poor Whites in the way that poor Black and Latinx Americans do (Massey and Denton 1993; Massey et al. 2016; Rugh, Albright, and Massey 2015). Research even shows that Black college graduates receive a lower return on their education and are less likely than their White peers to be gainfully employed (Cohen 2014; Badger et al. 2018; Wilson and Rod­ gers III 2016). So, even when we do control for class, inequalities rooted spe­ cifically in race still remain.

The Diversity Dilemma The United States has seen a cultural shift that has led to an increased appreciation for diversity and multiculturalism. Diversity is a common mantra in many aspects of American life: in Fortune 1000 companies (Embrick 2011), in the brochures and viewbooks that are dispatched from college admissions offices (Hartley and Morphew 2008), and in the majority opinions of the US Supreme Court (e.g., Grutter v. Bollinger, 2003). The second countervailing force that arose in our interviews concerns the notion that Americans’ propensity to value diversity is directly at odds with the fact that most Whites do not live, or seek to live, diverse lives; that is to say, while White millennials, in particular, value the idea of racial diversity and acceptance, they do not themselves embody that sentiment in the way they live their daily lives, even when they are in multiethnic spaces (Mayorga-­Gallo 2014, 2018; Smith and Mayorga-­Gallo 2017). As such, the value of diversity meets several critical challenges. One challenge is that even though people value diversity, diversity does not mean the same thing for everyone. Scholars find that people equate in value idiosyncratic characteristics (e.g., movie genre preferences) with structurally contingent characteristics (e.g., race, gender) (Smith and Mayorga-­ Gallo 2017; Embrick 2011). This means that a group of all-­White men with different personalities is viewed as equally diverse as a multiracial group of people with varied life experiences. It should also be noted that even though Americans value diversity and multiculturalism, they are not living diverse lives. Critics might point to

118 / Chapter Four

survey data that suggest otherwise. Indeed, figure 4.1 shows that in 2016, just over 50 percent of White millennials reported to the General Social Survey (GSS) that “in the last few years, [someone in their] family brought a friend who was [Black] home for dinner.” The figure also reveals that Whites are increasingly likely to report having a member of a different racial group as a neighbor, and White millennials even more so; nearly 80 percent of White millennials reported having neighbors of another race, compared to 70 percent of their older counterparts. Still, just because White millennials may live alongside people of color, this does not mean they are actually engaging in racially progressive behavior. Indeed, scholars like Sarah Mayorga-­ Gallo (2014, 2018) show that even many Whites who conscientiously move to racially diverse neighborhoods do not necessarily interact (in positive ways) with their Black and Latinx neighbors. Racial segregation in some areas of American life is actually increasing. Racial integration in American schools increased between the 1950s and the 1980s; since the early 1990s, schools have been in a process of continuous resegregation, largely due to continual White flight, the process of school secession, or the creation of predominantly White charter schools (Hannah-­ Jones 2014; Wilson 2018). Almost 75 percent of Black and Latinx schoolchildren attend schools that are predominantly minority (Hochschild, Weaver, and Burch 2012; Kozol 2005). Goldsmith finds that “segregation in schools and colleges contributes to the intergenerational transmission of residential locations in terms of racial composition” (2010, 1603); that is to say, the racial composition of teenagers’ neighborhoods and schools is highly associated with the racial composition of their neighborhoods when they are adults. One might argue that this isn’t the fault of young Americans; they live with their parents, who decide where to live and work. What we show below, however, is that even when these individuals are placed in more racially diverse settings—­either due to the busing of students of color to their high schools or due to the move to college—­they still live racially homophilous lives. Research shows that Whites develop a “White habitus” in racially homogenous spaces. White habitus limits Whites’ interactions with racial minorities, and “as a result of this conditioning, whites’ racialized attitudes and prejudice towards Blacks are continuously recycled and legitimated” (Bonilla-­Silva, Goar, and Embrick 2006, 247). In efforts to grapple with this set of countervailing forces—­the love of diversity versus White habitus—­we begin by examining how young White people talk about diversity, and then we shift to examine their friend groups and how they explain their racially homogenous neighborhoods and social networks. In all, it becomes clear how valuing “diversity” may not lead

Proportion

1995

2000

2005

2010

2015

4.1. Millennials Living in “Diverse” Neighborhoods

Source: General Social Survey Cumulative Data Set (1977−2016), White Respondents Only

Year

0.00 1990

0.00 1985

0.25

0.25

1970

0.50

0.50

1980

0.75

0.75

1975

1.00

1.00

Proportion of White Respondents Reporting Contact with Members of Another Race

Millennials

Older Whites

Have Over for Dinner (Last Asked in 2006)

Millennials

Older Whites

Have As Neighbors

120 / Chapter Four

Whites to have more (positive) interactions with racial minorities, reduce stereotypes about people of color, or increase awareness about how racial hierarchies are systematically perpetuated. Instead, we find that these two forces counteract one another, leading to a stagnation in racial attitudes across time and across recent generations. “I Think It’s a Beautiful Thing. . . . Diversity Is Awesome” We asked our respondents two questions about diversity. The first was “When you think of the word ‘diversity’ what comes to mind?” The word cloud in figure 4.2 presents the words, ideas, and concepts this group most closely relates to diversity. For them, diversity is about difference in background, and they tend to incorporate race, ethnicity, and gender as major components in their notions of diversity. A number of the respondents also included socioeconomic status or class, sexuality, and whether one has a disability. Respondents also drew up the notion of people being together, in an American melting pot. In fact, there were very few people who didn’t incorporate at least one of these identities or categories into their response to our first question; those who did not were very self-­conscious and purposeful about not doing so. For example, Jackson explained, “I would say in any setting, no matter where you’re at, there’s diversity. It’s a family business, and it’s three relatives who run the business, I’d say there’s very much diversity in that. Because everybody has a different personality. They have different things that make them laugh, different things that make them cry, different things that they like, different things they don’t like, hobbies, et cetera, et cetera.” But even for those like Jackson who wanted to show that they were above the fray and not bogged down by primarily thinking about race and ethnicity in the context of diversity, they still believed that diversity has value; it is something to attain, whether it is through gathering a group of individuals who have “different backgrounds” or have similar backgrounds but different personalities and interests. Nonetheless, for the most part, people described diversity more like Charles, a twenty-­three-­year-­old from Indiana, who stated that for him diversity is “Just people of all genders and races mixed together cooperating and doing things to achieve a better society.” Jacob, twenty-­one-­year-­old from Maryland, took it one step further: “Diversity is basically heterogeneity with—­in the context of political realm, it’s not only heterogeneity but also an acceptance of heterogeneity of whatever trait or dimension you’re going off

diversity

experiences

innate

groups dead class

doesn’t society

acceptance

racial

cultures

ses

reliance

cooperating differences

4.2. What Comes to Mind When You Think of Diversity?

sexuality

language

122 / Chapter Four

of and the desire for the same.” There is a sense that diversity is not just the gathering of individuals from different backgrounds but also an acceptance of these differences. Bethany, from western North Carolina, stated, “I think it’s a beautiful thing, I think diversity is awesome.” Others described diversity as “good,” “great,” “cool,” “important,” “positive,” and “practical.” You Need at Least Two Women in the Room The second set of questions we posed to respondents was “Do you think it’s important to have diversity in colleges? In workplaces? In government?” Three themes concerning the importance of diversity appeared. First, nearly all of the respondents relied on the diversity frame, which “views race as a positive cultural identity. The diversity frame also assumes that race shapes individuals’ world views and cultural practices, and that interaction across racial lines is positive and important” (Warikoo and de Novais 2014, 861). Secondly, there was a premium value placed on diversity in college settings more so than other settings, especially the workplace. And finally, there was a sense that diversity is a valued commodity; following sociologist Sarah Mayorga-­Gallo, we define commodification “as the treatment of Asian, Black, Latinx, and Native peoples as objects rather than humans for the benefit and satisfaction of others, namely White people” (2019, 10). To begin, most of the respondents felt that diversity in various settings is important. When Jayce, a twenty-­two-­year-­old from St. Louis, was asked if diversity was important, he stated, “Oh sure. I mean different perspectives can only help move humanity forward.” This notion of “different perspectives” was key for White millennials, in general. For example, Isabella, a twenty-­year-­ old from Western North Carolina, mentioned, “Oh yeah, because it brings more to the table. You know if you have the same group of people that have all the same ideas, you’re not gonna get anywhere, and you’re not gonna build off anything.” And Lila, a twenty-­eight-­year-­old from Philadelphia, provided an example of how having different perspectives just makes practical sense: Yes. I think that if you don’t have representation, you miss out on important things that you should consider. So, I think it most readily comes to mind with gender representations, things like advertising campaigns where like you see something on TV and you’re like, if there had been a single goddamned woman in that room, you would not have your PR disaster because someone would’ve said like, “This is incredibly offensive.” And actually, I really think you need at least two women in the room because if you’ve got one, she’s not going to speak up. But if you’ve got two, they’re going to look at each other,

Millennials on Racism  /  123 raise their eyebrows, and then look at the CEO and say like, “This is going to be a cluster-­fuck. You cannot run this ad.”

Lila noted that when you have a diverse setting, you are more likely to get ideas and perspectives about what is appropriate and desirable as well as get a better understanding of why certain things may be offensive to underrepresented groups. She added that you have to have some critical mass of minorities—­at least two, who can work together and support one another. In general, there was a positive reaction toward the idea of diversity, but there was also a special focus on the role of diversity in college. Respondents noted that college is a special moment in a person’s life to meet different types of people, learn new things, and gain a new perspective on life. For example, despite the fact that Easton, from Maryland, felt that diversity is “another great word that doesn’t mean anything,” he later stated, In that case actually I think differently about college because I owe it to like a college where diversity really wasn’t talked about at home, then I got here and it was talked about a lot more, and I think often it is ridiculous here but also sometimes helpful. But I actually think that the discussions I have had about diversity, and they’re mostly sexual diversity rather than racial diversity because my class is all White, but we have sexually identified [ourselves]. It’s very different things, have been really helpful, but they’re mostly private conversations that are happening. And so I was thinking that wouldn’t it be great if college could have a forum where those private conversations could happen.

Easton, as well as many of his generational cohort, suggested that college is the perfect time and place to interact with people who are not like you. He, like most of the other respondents, lived in primarily White neighborhoods, so the shift to college was one of the first times where he approached discussions of difference. Similarly, and put more simply, Harper, from Columbus, Indiana, explained, “I kind of think [that diversity is important] because it brings in a different cultural perspective, and I think that makes us more well-­rounded.” As mentioned, there was a premium placed on diversity on college campuses. Ella’s comments on whether job and college applications should inquire about the race of the applicant provide an example: “But it’s also important to create a diverse community because a lot of times it doesn’t really happen naturally. So I think in terms of college and stuff I think it is important to kind of handle that and try and force a diverse community. Or not force, but create a diverse community more. But I think on job

124 / Chapter Four

applications and everything, that’s really unnecessary.” Ella noted that it is common for people to live in homogenous communities; colleges, however, tend to actively seek out people from different kinds of communities and backgrounds, and she noted that colleges can encourage interracial interactions, which is good, but this kind of thing need not extend beyond the four-­year college experience. Caden, who was attending college in North Carolina, echoed this sentiment: Definitely for college because college is the time when you learn. Like, K[indergarten] through twelve, you learn about history and all that stuff, but I think it doesn’t really apply or click until you get to college, move off, and you see all these different kinds of people who have been brought up different, they have different views. Yeah, I mean definitely important in college because like me, I love to learn about different cultures and stuff. Now like when you move into the workplace—­I mean I think diversity is important everywhere but when it starts coming down to the ratios, going back. I don’t really know on that. Like at all.

Both Caden and Ella noted that college is an important time in the development of people’s ideas; it’s important to be exposed to different people in this very specific educational realm, but once you move on from college, being exposed to and interacting with individuals of “different cultures and stuff” comes at too much of a cost, particularly if you end up relying on “ratios” and affirmative action in order to develop or maintain a racially diverse environment. The words and rationale that people used to describe and explain the necessity of “diversity” are important to note and analyze, as we’ve done thus far. But it is also important to note the words and the rationale that were not used. None of our respondents suggested that diversity is about “power” or “equity”; instead, what we find is that for this group diversity is about individual gain and not a means to thwart White supremacy or racial inequalities. This is best illustrated in Caden’s and Ella’s sentiment that diversity is good in college, a time when they feel like they will individually benefit from non-­ Whites being nearby, but not when they feel that “diversity” is in direct competition with their personal opportunity to be gainfully employed. Diversity Is Fun! In addition to the notion that “diversity is awesome,” there was a sense that diversity is valued as a commodity. That is to say, having non-­White people

Millennials on Racism  /  125

around is seen as attractive because it will benefit Whites. Here are three examples: Hunter (21, Chicago, IL): Like I like the idea that they let people in to have more diversity because it ends up, even though I’m like a straight White guy, it lets me see other people. Like even though I grew up in a much more diverse background than certain people here, probably a lot of people here, it’s not like I’m done learning about other people. Like, “Oh, finished that chapter let me go live in a suburb right now.” Like it’s a constant process. Daniella (19, Lexington, MA): I think that it’s important for places—­or like especially universities, also the workplace—­to have a lot of diversity because, again, when you meet people of different cultural and ethnic and racial backgrounds, you get to expand your horizons because they’ve had different experiences, and they can teach you something. Kaitlyn (21, Lexington, MA): I think that like you do kinda need it in a way to make sure like—­because it’s like, it’s no fun—­well, like I guess that sounds so mean. But like it’s kind of like more interesting when you like meet peo­ ple from, like, different backgrounds and different eth—­Because you like learn things, like along the way.

Hunter explained that for him diversity is important because it benefits him. It exposes him to different people that he might not have interacted with otherwise. Similarly, Daniella, who is from a wealthy suburb in Massachusetts, explained that diversity is important because she gets to expand her horizons—­because “[people of color] have had different experiences, and they can teach you something.” What Daniella’s and Hunter’s responses are illustrating is the notion that Whites rarely see themselves as having race or racialized experiences. Hunter is a “straight White guy,” and Daniella’s life is seen as normal, whereas non-­Whites’ lives are something to examine and be exposed to. Kaitlyn used the word “fun,” which is perhaps the most recognizable example of commodification. If there aren’t non-­Whites around, “it’s no fun.” She even caught herself when she said this, suggesting that what she’s saying “sounds so mean.” For her, non-­Whites provide a source of entertainment in her life; they color (pun intended) her own normal, White existence. Altogether, we are able to see a pattern whereby White millennials value diversity because they view people of color as objects that will enhance their lives and do the work of teaching them about non-­White American culture and experiences.

126 / Chapter Four

The thing about diversity is that “diversity initiatives were originally meant to address systemic racial and gender inequality in corporate environments,” but what is illustrated here and in our everyday lives is that “their contemporary amorphous conceptualizations have reduced their usefulness for structural change” (Mayorga-­Gallo 2014, 23). Young people value diversity and multiculturalism not because diversity may be a signal that institutions are working toward more equitable outcomes across politically relevant identity groups, but because diverse settings represent a personal benefit for them. We see this most clearly in how respondents endorse having diversity in college settings but less so in workplace settings. To be fair, college brochures often “depict students as a diverse, young, fun-­loving crowd,” sell to students the idea that at their college “it’s all about me,” and rarely define what “diversity” means to that college apart from geography or interests (Hartley and Morphew 2008, 678). Hartley and Morphew’s work reveals that colleges themselves have (publicly) shifted from a mission of diversity as a way to address inequality toward an amorphous conceptualization and commodification of diversity in efforts to recruit the “best” students they can get. So perhaps it follows that students have incorporated this highly curated picture of what diversity is and what it means to be in a diverse environment into their own definitions of these ideas. Which Birds Flock Together? White Americans value diversity, but they do not live diverse lives. They tend to live in very racially segregated neighborhoods; thirty-­six of the forty-­ three respondents (84 percent) described their neighborhood as White or predominantly White. Most Americans do live in racially segregated neighborhoods and attend racially segregated schools (Hochschild, Weaver, and Burch 2012; Kozol 2005; Massey and Denton 1993). A large proportion of our respondents were college students or college graduates and thus were living in or had recently moved from a place where they were likely to believe that experiences with a diverse group of people are necessary, appropriate, and personally fulfilling. Thus, we would be safe to expect our respon­ dents’ friend groups to be more diverse than the average White American’s. We asked respondents to name their closest three friends, then we asked them to describe the kinds of things they do with their friends, and, finally, we asked for the demographic characteristics of those friends. Table 4.1 illustrates the racial makeup of our forty-­three White millennial respondents’ friend groups. The racial makeup of our respondents’ friend groups mirrors that of a national sample. Christopher Ingraham found that almost

Millennials on Racism  /  127 Table 4.1.  Racial Background of Friend Groups

Friend 1 Friend 2 Friend 3 Total Percent

White (N)

Asian (N)

Black (N)

Latino (N)

Other (N)

39 39 34 86.80%

0 1 3 3.10%

2 1 0 2.30%

1 0 3 3.10%

1 2 3 4.70%

Data: Face-­to-­face interview respondents

three-­quarters of Whites had no non-­White friends, and that if the average White person had a hundred friends, ninety-­one of them would be White; there would be one each Black, Latinx, Asian, mixed-­race, and other; and three would be of unknown race (2014). Thirty-­two (74 percent) of our respondents only named White friends in their closest circle of friends, six (14 percent) had one non-­White friend, three (7 percent) had two non-­ White friends, and just two (4.6 percent) did not name any White friends. To be fair, almost half of the respondents noted that the three individuals they had named were not truly representative of their “real” friend group or their day-­to-­day interactions with non-­White people. For example, Brooke, who had named three White friends and was perhaps conscious of the social pressures associated with doing so, made the following public-­service announcement: “But I just want everybody that’s going to be listening to know that I have an African American first cousin who was adopted, love him, half Cuban cousin, a half Black cousin, a half Latino cousin, and I would be willing to adopt any ethnicity child if I need to. It has nothing to do with racism. I just want to set that out there for the rest of the interview.” Caden also noted, “I’m not a racist. I have Black family.” This connects to what we mentioned earlier. This group of individuals has a very narrow idea of what racism is; in this case, they feel that the mere existence of people of color in their family relieves them of the possibility of being racist. They have little to no control over the composition of their extended family, and the love of one individual from an out-­group is not a sufficient proxy for other preferences and beliefs that may produce inequities for whole groups of people. In a similar but less extreme vein, most of the other respondents noted that even though they described an all-­White friend group, they are in some way connected to non-­Whites or have, at some point in time, had friends who are persons of color. For instance, Adeline explained, “I feel though like a lot of our friends, we have a lot of diverse friendships. I know a lot of our friends were always friends with the METCO kids. We always got along

128 / Chapter Four

well with them. That was never really a problem.” METCO is a voluntary desegregation program that buses students from Boston and Springfield, Massachusetts, into neighborhoods and schools that agree to accept the bused students in efforts to reduce racial isolation and expand educational opportunities for those involved. Adeline explained that while she named only White friends, her friends were always friends with the students of color who were bused into their wealthy, White suburb. She and her friends got along well with her friends’ friends. Chloe, from Lincoln, Massachusetts, vaguely explained, “Yeah, and I mean, I guess a lot of my friends are White, but I do have some like friends of different races. There’s kind of like—­there are definitely differences, cultural differences, but you can definitely still be friends with someone just because they are of a different race.” Similarly, Dylan noted, “I remember having a lot of friends from other ethnic groups.” Kaitlyn’s response was also instructive, as she explained to the interviewer, kaitlyn: But like, I have friends that are of different ones [races and ethnicities]. But yeah, I think it should just be left up to the individual. interviewer: Okay. And what are the names of your friends who aren’t White? kaitlyn: Oh my God, I’m trying to think. Well, one’s named [Anna], but she’s Asian. interviewer: Okay. kaitlyn: I don’t know if that matters. interviewer: No. kaitlyn: Then there’s like [George] and [Luke] and, like, they’re Black. interviewer: Okay. kaitlyn: I don’t know, they, they have like, they don’t have like unique—­Like different names, like to the background. But like yeah. interviewer: Okay. kaitlyn: They’re both like more like, I don’t want to say Americanized, but like they are.

Here, after Kaitlyn was asked who her friends are, what kinds of things they do together, and what their races are, she brought up the fact that she does have friends who are not White, but she had difficulty in naming the people of color she claimed to be close to, which suggests that she is not actually friends with the people that she was able to eke out of her memory. Additionally, we see another example of how White millennials do not necessarily see Asian Americans and underrepresented racial groups; Kaitlyn mentioned Anna and noted, “but she’s Asian.” As a consequence of this,

Millennials on Racism  /  129

Kaitlyn needed to raise her credibility. She did this by naming two Black friends, but the ones she supposedly has—­George and Luke—­didn’t necessarily help her because they don’t have clearly identifiable or “unique” names that are typically associated with Black people, like Jamal or DeShawn (Butler and Broockman 2011). In turn, she told us their names and then assured the interviewer that they are Black, but that they are assimilated or “Americanized” Blacks, to explain why they have the names that they do. This interaction not only tells us that Kaitlyn doesn’t have any non-­White friends but also illustrates the effects of  White habitus: Kaitlyn has a series of ideas about what non-­Whiteness and Blackness are and typically look like, and this vision is constrained and limited to a series of stereotypes. In all, what we see here is that most White millennials do not have diverse friend groups, despite their admiration for diversity or their desire to prove that they are not racist and have some connection with non-­Whites over time, through other friends, or tangentially to people whose names they have dif­ficulty remembering. How do White millennials explain why they tend to live in racially homogenous neighborhoods and have homophilous social networks? Our respondents relied on the naturalization frame of color-­blind racial ideology; this frame “allows Whites to explain away racial phenomena by suggesting they are natural occurrences” (Bonilla-­Silva 2014, 76). Like class and socioeconomic status, naturalization provides a race-­neutral explanation for racial phenomena. We tend to see the naturalization frame through words like “natural,” “comfort,” “interests,” and “coincidence.” White millennials expressed a sense that there is something about human genetics that leads to living racially segregated lives. For instance, Michael cited his sociology teacher, noting that, “So, one way my support sociology professor put it very simply is how you’re—­races love to stick with races, and that’s how we’ll always be, and usually if you’re of a different race you’re most likely to live through those neighborhoods.” Peter, from Indianapolis, explained, “I think that people of the same race and same background tend to live together because they identify with each other. That’s how we are as humans. We run in packs. I don’t see anything wrong with that.” There is a notion that it is natural to group people by race—­which is well known to be a socially constructed category—­because like wolves and birds, humans run in packs and flock together, respectively. But one thing that none of our respondents explained is why we group around race rather than some other identity or characteristic: gender, major, dietary preferences, height, et cetera. “Comfort” was also a regularly provided explanation of why people live racially homogenous lives even when they value diversity. For example,

130 / Chapter Four

thirty-­two-­year-­old Claire explained, “You gravitate to people that are like you. You feel more comfortable I think sometimes with people that are like you. So it would be nice to learn about other cultures and people that come from different walks of life, but unfortunately that’s just not always a reality.” Similarly, Heather from Carmel, Indiana, noted, “A lot of times you’ll look to see who you’re comfortable with and it’s usually the same race. I feel like we’re not ignorant, like thinking people are worse or better, but we still kind of keep to each other, our races.” Claire, Heather, and many others noted that there is something natural and comforting about being with people who share your racial group; nature leads you to “gravitate to people that are like you,” and that is where you feel more comfortable. Daniella, who noted, “I mean I like living in a diverse place, like broadens your mind,” told this story: Like I went to this one talk on [diversity], and they were like, we just feel like—­someone, an African American girl, was speaking for herself, and she said, you know, “If I’m in a class, if I get there first to the class and I sit down, like all the White students, none of them want to sit next to me.” And I mean for me, as a White individual, that is like really hard to believe. Like I don’t think I would ever go into a class and see a Black girl and be like, “Wow, I’m not going to sit next to her because she’s Black.” But at the same time, it’s like, I think there are subconscious things about it. Like if I walk into a room and I’m like, “Who should I sit next to? Who looks nice?” I’m not saying that African American people are not nice—­that’s a terrible thing to say, and I don’t believe that at all. But like maybe I’d sit next to someone who looked like me, and I would be like, “Wow, I can relate to her because she looks like me.” And so that’s why I’m going to sit next to her.

Daniella explains that she has a hard time believing that people would avoid sitting next to a Black person because of their race, but when she thinks back on how she makes her own seat selection, she tends to find someone who looks like her, who shares her racial identity; it’s not that she avoids sitting next to the Black woman because she’s Black, but she instead finds comfort in sitting next to a White woman, as White habitus would have conditioned her to feel. Here, Daniella makes a distinction without a difference. In the end, Daniella does not want to sit next to the Black woman even in her own hypothetical situation. Even though Daniella has the same information about the Black woman and the White woman in this thought experiment, she still presumes that she will be able to “relate to” the woman who looks “nice,” and in this case, that person is a White woman.

Millennials on Racism  /  131

Another reason that respondents relied on to explain why they have racially homogenous friend groups is “interests.” For example, Ava noted that she used to have friends from all sorts of ethnic groups, but “then as high school went on, it actually became more segregated ethnically, I think.” She explained that “Yeah, I mean my friend group kinda broke up. And then some of them regrouped and I don’t know—­I think it was more interest based than race based.” Heather also noted that interests are what dictates who she is friends with: “The reason I’m friends with those other people is because we went to—­we were in the same clubs; I ran cross country, so it wasn’t anything to do about [race].” There’s a notion that people of different racial groups have vastly different interests, and perhaps they do, but it is interesting how many respondents felt that racial segregation by interests was not a racial phenomenon. A third reason that White millennials employed was that of “coincidence.” Thirty-­four-­year-­old Brody provided a clear example of this: “I choose my friends because of who they are, not race or ethnicity or whatever that thing. . . . I mean I’m not, do you, I’ve had friends of all kinds so it’s, you know, I guess kind of coincidental that I have three White male friends but that doesn’t mean I chose them because they’re three White male friends, you know.” Also, Kaitlyn explained, “I feel like it’s just like a coincidence that I’m friends, like my three best friends are of the same ethnicity background. But like, I have friends that are of different ones.” Alaina suggested that perhaps we are all thinking about this too much, stating, “I don’t think [living racially segregated lives] is an actual race issue, I just think it’s where people populate. I think people—­if it becomes a race issue, it’s because people are overthinking it.” A quick Google search tells us that a coincidence is a “remarkable concurrence of events or circumstances without apparent causal connection.” And yet, none of what we see is remarkable or without causal connection; it is common for most Whites to have predominantly White social networks, live in predominantly White neighborhoods, and go to predominantly White educational institutions for their entire lives, a series of patterns that misses the mark of “coincidence.” Each of these things is closely related to one another and, more importantly, is related to the distribution of (political, economic, and social) power and (material) resources by race, where Whites, on average, have more of both than non-­Whites do. Actually, attrib­ uting all of these kinds of things to coincidence serves as a means to perpetuate this asymmetry of power and resources. The last way that we see young White people explain away the racial component of their racially segregated lives is through the use of class or

132 / Chapter Four

socioeconomic differences. For instance, Daniella noted, “I think that more along the lines of people like from different socioeconomic statuses should like be more integrated. I think that like, you know, the upper middle class doesn’t necessarily deal with like people who are of lower socioeconomic standing. I think like that’s more important than like making it a race thing.” Dylan also explained, “So in theory I would have loved to spend more time around people who came from different backgrounds, spoke different languages at home, any number of factors of diversity both economic, ethnic, and even ideological. But I also loved my sort of high-­cost, bougie, Montessori education and those two things unfortunately didn’t work coincident [sic].” Again, we see the notion of “coincidence”: it is a coincidence that Dylan’s life is racially homogenous; it is a “coincidence” of class that all of the wealthy people who could afford his “high-­cost, bougie, Montessori education” just happen to be White. Daniella and Dylan provide examples of how people rely on the idea that ostensible racial phenomena are best understood as class phenomena in disguise. White millennials appear to be incredibly liberal because they truly do value diversity; they have good intentions of interacting with people who they believe can teach them something. But when we dig deeper, we find that they value the idea of diversity, not the actuality of it. Sarah Mayorga-­ Gallo’s work helps us to explain this gap in principles and preferences. She reveals a hidden side of the diversity frame, noting that this frame “helps individuals who live within an increasingly multicultural environment reconcile a national emphasis on egalitarianism with pervasive racial inequality,” and she goes on to explain, “As a part of this reconciliation, diversity ideology dictates that intentions, as opposed to outcomes, are what truly matter” (Mayorga-­Gallo 2014, 23). So, while young people can express an appreciation for differences, they do not actually have enough engagement with people of different racial groups on a day-­to-­day basis to be considered racially liberal. Diversity is embraced and celebrated, but interacting outside of one’s racial group is seen as unnatural, uncomfortable, and inconvenient. The thing is, this is not a contradiction, because actions aren’t required to be considered a racial liberal for members of this group.

The Empty Knapsack Peggy McIntosh wrote an insightful and influential article titled “White Privilege: Unpacking the Invisible Knapsack.” She, like many Whiteness and critical race scholars, notes that Whites are socialized not to recognize the privileges that come with their racial-­group membership (DiAngelo 2018;

Millennials on Racism  /  133

Wise 2011; Hartmann, Gerteis, and Croll 2009; Moore and Bell 2011; Pu­ gliese 2002; Lewis 2004; Jensen 2011). McIntosh develops the analogy of the invisible knapsack to help us understand this phenomenon. She explains: “I have come to see white privilege as an invisible package of unearned assets that I can count on cashing in everyday, but about which I was ‘meant’ to remain oblivious. White privilege is like an invisible weightless knapsack of special provisions, maps, passports, codebooks, visas, clothes, tools, and blank checks” (McIntosh 1992, 30). Sociologist Douglas Hartmann and his colleagues note, “By hiding the structural relations of race, this ideology of neutrality and fairness is believed to obscure the source of both white difference and advantage”; this ideology allows for an “attribution error,” whereby “Whites may be able to see and understand the ways that blacks and others have been disadvantaged by the racial system, but they tend instead to attribute their own success to individual effort and hard work” (Hartmann, Gerteis, and Croll 2009, 408). This lack of recognition of Whiteness and White privilege is well cited in the literature, but as we have mentioned before, White millennials have bucked the trend. The last dyad of countervailing forces that we focus on in this chapter concerns the push and pull of the recognition of White identity, Whiteness, and White racial privilege. We have mentioned before that the millennial generation is special for a number of reasons, but the one that is most significant here is that this is the most racially diverse generation of Americans in history. The demographics of the United States are changing so rapidly that the US Census Bureau predicts that by 2043 there will be no racial group that composes a demographic majority. California, Hawaii, New Mexico, Texas, and the District of Columbia as well as several of America’s (major) cities can already be characterized as majority-­minority. In turn, “the shrinking size of the white population as well as the increased presence of nonwhites in prominent positions has rendered whiteness more visible rather than as an implicit synonym for American” (McDermott and Samson 2005, 248). Young Whites’ recognition of Whiteness and White privilege and their strong sense of egalitarianism mean that they have the potential to walk against the moving walkway of racism that we discussed previously. But, as the trend goes, we find that there is a countervailing force that may prevent this recognition of White privilege from leading to more racially progressive attitudes on the part of White millennials. As we saw at the end of chapter 3, millennials’ cognizance of White privilege is coupled with the increased perception among White Americans of “reverse discrimination.” Just as the growth in diversity might make Whites feel more aware of their racial group, a growth in this awareness might be accompanied by a

134 / Chapter Four

sense of loss among many White Americans (Craig and Richeson 2014a, b). CNN asked, “Are Whites racially oppressed?” (Blake 2011), to which many replied a resounding “Yes!” This kind of sentiment is best illustrated by White millennials at institutions of higher education like Texas A&M University and University of California, Berkeley, who have hosted “affirmative action bake sales” in efforts to show that underrepresented racial groups are gaining unfair advantages over White students in the college admissions process (Dietrich 2015). Similarly, millennial Abigail Fisher took her case of “reverse discrimination” against the University of Texas at Austin to the Supreme Court. Relatedly, White supremacist Jason Kessler, organizer of the deadly Unite the Right protests, argued that he is not a White supremacist or “even a white nationalist” but rather a “civil and human rights advocate focusing on the under-­represented Caucasian demographic” (National Public Radio 2018). What these examples reveal on the part of these young White people and their supporters is both an awareness of one’s racial-­group membership and identity as White and a lack of understanding about the structural nature of racism in American society. McIntosh declares, “Describing white privilege makes one newly accountable” (1992, 31). Indeed, we are able to show that White millennials are more likely to acknowledge that they have privileges, but our respon­ dents also seem no more likely to care about racial injustice. This dyad of countervailing forces is the very definition of a paradox, because we have two things that are in contradiction simultaneously existing in the minds of many individuals in the millennial generation. Here we find that young White people are exploring the contents of their invisible knapsacks and even feeling the weight of them on their backs but, at the same time, believe that people of color have similar knapsacks with the same contents in them. These two forces cancel one another out, leaving us at a standstill. Below, we provide examples of how this plays out. “We Have All of the Advantages for the Most Part” We asked respondents whether they believe that Whites have advantages due to their race. The overwhelming response was “Yes.” Twenty-­eight respon­ dents (65 percent) agreed that Whites do see advantages in today’s society because of their skin color. When we asked respondents to elaborate or provide an example, most people pointed to their observations that a larger proportion of Whites tends to have more money or be more affluent in comparison to other groups, there are more positive stereotypes of Whites, Whites are more likely to get job opportunities or be hired (mostly due to subconscious

Millennials on Racism  /  135

bias), and Whites are more likely to have positive outcomes with law enforcement and the criminal justice system; some respondents also mentioned that there are some positive psychological advantages that Whites receive. Most people named one or two of these factors, but Lila, a twenty-­ eight-­year-­old woman from Philadelphia, summed up these ideas nicely: So, I can be driving my car and be reasonably certain that if I’m pulled over it’s because I was doing something wrong while driving my car, and not because of who I am. And I’m interacting with the police; I’m generally not concerned about my personal safety. When I have kids, you know, if they come out looking like me, I can be reasonably confident in their safety in a really broad range of contexts. I am not looked at to be exemplar of my race. In most rooms, I’m not going to be the only person of my race, and if I am, it’s not going to be notable or commented on. I’m not self-­conscious about my racial background. I don’t have to feel like I have to make excuses for it. I don’t have explain it if I am eating something for lunch that I grew up with. Nobody’s going to, like, ask questions about it and be like, “Oh my God, what is that? Did you learn how to make that from your grandmother?” Like, “Oh that smells weird,” if I have people over for dinner and make a food that I grew up learning how to make, nobody’s going to exoticize it, and nobody is going to be like, “Oh I don’t eat weird ethnic food.“ Like my background is the norm and the default, and I don’t ever have to explain it or assume that I’m going to have to explain it.

Lila named a great number of the advantages Peggy McIntosh identifies in her knapsack; here, we see that this knapsack is not invisible to Lila. Reagan, a young woman from Elk Grove Village, Illinois, also suggested, “Yeah, and I—­and people can deny it and say, no, it’s based on our merit and blah-­ blah-­blah, it’s not based on privilege, or yada-­yada, but I’m sitting here and I’m going to say, yes, that definitely I—­sometimes get more—­have more opportunity because I’m a White girl.” Reagan implied that White privilege is associated with a myth of meritocracy, and she also noted that she is presented with some opportunities because of racial-­group membership. In a similar vein, Sebastian, from Chicago, provided an example of the psychological wages of Whiteness (Du Bois 2007): I think there’s the issue of access. I think if you’re White, you can feel I guess more confident just emotionally going to a lot of places that might make you better off like a job interview or—­because if you’re White, you can go to a job interview and be like, okay, I don’t need to worry about my race. No one is going

136 / Chapter Four to be racist against me. I can just prepare for the job interview. And I’m not Black. But imagine that if you are Black, and this is an issue that people have to deal with racism, that’s going to be in your head that they might be prejudiced against me because I’m Black. And I think these little things like that definitely make life harder if you’re not White just because if you’re White, you have an absence of concern about your skin color, which lets you focus on other things.

The great majority of our respondents believed Whites receive unearned privileges of all sorts. They noted Whites are more likely to be hired by other Whites because, as Mackenzie from Massachusetts said, “I think it’s just a subconscious, like what we’re more comfortable with. What’s more traditional, what we’re used to.” Some respondents even mentioned that there is also a gender component that has to be accounted for as well. Charles, for example, said, “The male thing, too, I think that being male helps you out in terms of getting jobs and keeping jobs and people tend to look, it’s proven, pay more and get more jobs than women do for the same level of achievements.” As a final example, Hunter explained, Like I’m never gonna—­I’m never gonna make somebody cross the street when they walk in front of me or if they’re walking towards me. I’ve never heard a car click shut. I’ve never been addressed as “Boy.” Yeah I have so much White privilege, so much male privilege. It rocks. Like at the end of the day I’m aware of it, and I’m a little guilty over it. I certainly think a lot about it. And I try to check it. But at the end of the day I have huge benefits.

Hunter is well aware of the contents of his knapsack and thoroughly appreciates the resources that are conferred to him in large part due to this racial-­group membership; later he noted, “That said, would I trade that, like give up my White privilege? Absolutely not. . . . And that actually sounds like moderately racist, I would not want to be Black.” In contrast to what the extant literature on Whiteness suggests, many of our respondents were sure Whites have more positive experiences and noticed that Whites tend to be seen in a better light than racial minorities, especially by employers and law enforcement; empirically, they are correct. “I Would Just as Equally Say There’s an Advantage of Being African American” While the overwhelming majority of White millennials we interviewed believed that Whites receive unearned privileges, our respondents were also

Millennials on Racism  /  137

likely to do three things that could potentially counter their awareness. First, there was still a substantial group of individuals who could not see White privilege at all. Second, many could not see how they personally benefited from White privilege, even if they recognized that White privilege systematically advantages Whites in general. Finally, there were just as many responses suggesting that Whites face systematic disadvantages or that minorities receive advantages at the expense of Whites as there were responses affirming White privilege. We found that people were able to see systematic White privilege while also noting systematic White racial disadvantages. About a quarter of the respondents suggested either that Whites do not have any advantages because of race or that other racial groups have just as many advantages as Whites do. For example, Samantha suggested, “I’m not sure in today’s society if that’s entirely true [that Whites have structural advantages], but definitely before there was so much diversity and awareness just of racism in general.” From Samantha’s and others’ perspectives, Whites may have had advantages historically, but America has evolved as a society and thus produces few benefits that are primarily afforded to Whites. Other respondents tended to suggest either that they don’t receive any personal advantages or that any advantages they have are a direct result of their parents’ class status or the way they were raised rather than of race. It was common to hear a sentiment like Brody’s: “Advantages, I don’t believe that that’s true because I haven’t been exposed or even made aware of certain circumstances that it’s occurred. I just don’t see it. I don’t know. It’s not, I’ve never really looked to be honest with you either. It’s not like I just go around and think about these things on a day-­to-­day basis.” Here, we see that Brody doesn’t believe that White privilege exists, and we also get a glimpse into what Whiteness scholars have been trying to illuminate for years: the notion that most Whites tend to not recognize their Whiteness. Similarly, Harper responded in this way to our question about whether she has noticed the way White privilege works in her life: “I’ve never thought about it. Yeah, I’ve never thought about it. I don’t think it comes to my race, I think it comes down to my parents’ jobs but that could come down to race. I don’t know. For me, personally, I’m just lucky enough to have parents that have always been willing to pay for everything.” Again, we see a lack of awareness of one’s personal White privilege and a reliance on class status and coincidence (this time, the luck of being born to a certain family). White millennials were nearly as likely to note that Whites are at a disadvantage in some way. Twenty-­three respondents (53 percent) provided examples of ways in which Whites are at a disadvantage due to their racial-­group membership, and nineteen (44 percent) identified systematic

138 / Chapter Four

advantages that minorities have due to their race or ethnicity. Since most of the respondents had, at some point, gone through the college application process, it is not surprising that they relied most commonly on affirmative action and diversity initiatives as examples of White disadvantage. Whites perceived that they were excluded from applying to various scholarships or that their chances to get into college were severely limited. Here are some examples: Adeline (20, Lexington, MA): I remember when I was applying to college; I was kind of torn on that issue. I kind of felt like in some ways it wasn’t fair. I remember my, one of my friends, she’s Jewish and Black and I remember when she’s applying to college, she was getting into really good schools, and I was kind of like it’s not fair. She’s a minority. She’s getting into all of these great schools. Caden (19, Alabama): Um, yeah, actually like when I was applying to college, and I was looking at scholarships, I get on a scholarship website and it’s like, looking at the eligibility stuff. It’s like African American, Latino, all this stuff. And literally, there’s a scholarship for everything except for middle-­class White kids. I found one, and it said only for middle-­class White males. Jackson (23, Sarasota, FL): I’d say I’ve had a disadvantage that I’m White because I haven’t been available for lots of scholarships, and that African Americans and Latin Americans have been available for.

There was a sense that people of color were getting into colleges that they did not deserve to get into, because of their race, and this was a personal disadvantage for the respondents. Adeline, for example, implied that her Jewish friend was “getting into all of these great schools” because she was Black rather than because she deserved entry into college. Relatedly, two people noted that Asians and Asian Americans put Whites at a disadvantage because they are “too smart.” Caden and Jackson represent those who felt that they were getting the short end of the stick in the scholarship game because of their race. A number of individuals noted that there were many scholarships to which they could not apply because the scholarships were exclusive to non-­Whites. Empirically, this observation doesn’t hold up. Anti-­racist activist Tim Wise explains, It is simply false that scholarships for people of color crowd out monies for white students. According to a national study by the General Accounting

Millennials on Racism  /  139 Office, less than four percent of scholarship money in the U.S. is represented by awards that consider race as a factor at all, while only 0.25 percent (one quarter of one percent) of all undergrad scholarship dollars come from awards that are restricted to persons of color alone. In other words, whites are fully capable of competing for and receiving any of the other monies—­ roughly 99.75 percent of all scholarship funds out there for college. (2011)

Despite the facts, these respondents felt there were way too many scholarships for non-­Whites, and the 0.25 percent of scholarships that are available for non-­Whites exclusively put them at a unique advantage. In addition to holding the sentiment that affirmative action and minority-­targeted grants create benefits from which Whites are excluded, respondents also felt that Whites are disadvantaged because Whites cannot talk openly about race and because people of color harbor hostility toward Whites, especially in spaces where Whites are the minority. For instance, Ella explained, There is a lot of hostility towards White people. Not that it’s necessarily undeserved, but sometimes it is. Just on an individual basis sometimes it is. Second, sometimes there’s discrimination against White people in that way. Like all White people are a certain way. Like “it’s such a White thing to do.” Everyone’s seen like Twitter handles and stuff that are “White girl problems” and what not. That seems discriminatory. Sometimes I felt personally offended or something like that; I don’t do that.

This is a common refrain among White millennials: Whites get stereotyped too. Respondents relayed a sense that people of color harbor resentment toward Whites, which Ella did not think is always undeserved. However, even the examples Ella pointed out—­“It’s such a White thing to do,” or her reference to #WhiteGirlProblems, a Twitter hashtag that trended briefly—­are false equivalents to the types and effects of stereotypes that people of color face. The things that White people are associated with doing and enjoying (i.e., food and clothing preferences) do not denigrate the group. Furthermore, there is such a range of popularly available White female images that White women do not actually face a set of negative tropes, unlike Black women and Latinas (Harris-­Perry 2011). (Though this may be changing due to the onslaught of White women, such as “BBQ Becky” and “Permit Patty,” who have been caught calling the police on Black people who are minding their own business.) Ours is not an effort to play in the “Oppression Olympics” (Hancock 2011); instead, it is important to recognize that the

140 / Chapter Four

disadvantages that White millennials perceive Whites face are not actually systematic, structural, or institutional disadvantages at all.

Racial attitudes, racial ideologies, and the role of racism transform over time. Here, we examined the racial ideology of young White people and the frames that they employ most, to gain a better understanding of why they seem more liberal than their predecessors but are not moving the racial-­ attitudes needle in a progressive direction. Again, we focused on young people because social scientists have predicted that cohort change along with changes in norms and values would help the United States evolve into a more equitable society. What’s more, the millennial generation is the largest living generation, makes up a big proportion of the electorate, and will shape our attitudes about race and racism for decades to come. Mary Jackman states, “Ideology is a political instrument, not an exercise in personal logic” (1994, 69). This is helpful because it explains why we see so many examples of cognitive gymnastics, leaps in logic, and contradictions in our informants’ responses. This chapter illuminated these contradictions, thus providing evidence for our racial stasis hypothesis. Specifically, a series of countervailing forces came to the fore. We found that young people are motivated to avoid the “racist” label, they value diversity, and they are cognizant of White privilege. All of these things are important values and ideals to possess in the march toward racial progress. But each of these significant steps is counteracted and contradicted. To be sure, what we see is not old biological racism or searing racial antagonism but instead a softer, ostensibly benign neglect and (perhaps purposeful) ignorance of institutional racism. Some of the sentiments our respondents relayed in this chapter overlap with what scholars often refer to as racial resentment (Kinder and Sanders 1996), but they also introduced a newer set of ideas that could only come to the fore in the twenty-­first century. For instance, demographic changes are triggering a sense of racial loss and “reverse discrimination.” Or relatedly, the increased prominence or salience of White identity is rather new (Schildkraut 2017; Cramer 2016; Jardina 2019). An awareness of Whiteness ought to produce greater awareness of racial inequality (McIntosh 1992) and, perhaps, racial guilt, but our research and others’ suggest that racial fear or apathy may be more likely to arise (Forman 2004; Craig and Richeson 2014a, b). Taken together, our respondents’ answers help us to better understand the stag­ nation in racial progressivism, and reveal new elements of contemporary ra­

Millennials on Racism  /  141

cism that have yet to be captured in the measures relied on most by po­lit­ ical scientists. Racial attitudes often influence policy preferences. In the next chapter, in our efforts to better understand the composition and configuration of Whites’ contemporary racial attitudes, we examine how White millennials interpret two racialized policies.

Five

Racialized Policy Preferences

Police brutality is nothing new, but the ability of average citizens all over the country to view these incidents through the lens of another person’s camera phone—­sometimes in real time—­is rather new. The fatal (and illegal) choke hold of Eric Garner in New York, the police shooting of Walter Scott in the back, and the body slamming of a petite, bikini-­clad Black teenager in McKinney, Texas, by a police officer because the teen had appeared out of place in a predominantly White neighborhood were all filmed by bystanders on their mobile phones and then repeatedly broadcast on national news and social media. National and state law-­enforcement agencies rarely collect data around incidents like these, but a steady stream of videos provides evidence of police brutality to a large swath of Americans who are typically distant from or even skeptical of this reality. Indeed, about a year after #BlackLivesMatter went viral, a Washington Post poll found that 60 percent of Americans believed that more changes are needed to give Blacks equal rights. This was a fourteen-­point increase from the previous year. The Post found that “majority opinion flipped among whites, with 53 percent saying more changes are needed, compared with 39 percent in 2014” (Clement 2015). But what kinds of policies are people willing to implement? In this chapter, we explore the extent to which White American millennials are willing to throw their support behind initiatives that concern racial disparities, the exclusion of underrepresented minority groups from various opportunities, and/or negative racial stereotypes. Specifically, we asked White millennials what they think of two racialized policies: affirmative action and a policy commonly known as stop-­and-­frisk. Affirmative action is a divisive policy; some view it as aiming to reduce racial disparities, while others see it as directly (and unfairly) benefiting

144 / Chapter Five

Blacks and other minorities to the detriment of Whites. We asked questions like “When you hear ‘affirmative action,’ what do you think of? What comes to mind?” and “What do you think about affirmative action in the college admissions process? What about in the workplace?” We should note that the topic of affirmative action often came up in our interviews long before we prompted our respondents to think about it. That is to say, when respondents were asked questions about the American dream or questions about whether people should identify as just American rather than as a “hyphenated” American, opinions about affirmative action were voiced. The use of stop-­and-­frisk, or Terry stops, has also been at the forefront of American minds due to the conversations in the media that were sparked by this constant stream of civilian-­filmed footage of some police officers at their worst. These conversations have been renewed in response to President Trump’s “tough on crime” campaign promises and his “urban” agenda. Some see Terry stops as an effort to prevent crime, but others see it as a pol­ icy that ultimately allows the police to racially profile, target, and harass members of Black and Latinx communities, especially men. We posed the following to our interview respondents: “There is a program in New York City that allows police with reasonable suspicion of criminal activity to stop individuals and search them. This program is commonly known as ‘stop-­ and-­frisk.’ What you do you think about this policy?”1 A striking difference between the way our informants responded to the topic of stop-­and-­frisk and the way they responded to affirmative action came to the forefront. White millennials were by and large against stop-­ and-­frisk, as many of them recognized the racial ramifications of the way the policy has been implemented. On the other hand, of the forty-­two respon­ dents who had something to say about affirmative action, only about a quarter of them (eleven) were unequivocally supportive of the policy. In part due to the fourth countervailing force—­the paradox of generations—­our respondents’ conservative policy preferences were not very different from what was already on record; they relied on intergenerationally passed-­down and outdated interpretations of affirmative action as well as stale (racial) stereotypes to inform their policy preferences. Additionally, we find that through an analysis of the language and meaning-­making our respondents employed around these two racialized policies, it becomes clearer that we need a more holistic measure of White Americans’ racial attitudes—­one that captures both what people know about the existence and role of racism and how they feel about it.

Racialized Policy Preferences  /  145

When Is a Policy Racist? Stop-­and-­frisk reportedly leads police officers to consider race when selecting which individuals to stop. Meanwhile, affirmative-­action policies also require an institution or organization to consider the race of a candidate among a myriad of other characteristics. As we have shown, those who ascribe to a color-­blind ideology believe that we should not consider race in our interpersonal interactions and certainly not to enforce state-­sanctioned policies. Are both of these policies racist? How can we know when it is racist to consider race and when it is not? Michael Omi and Howard Winant help us to understand and answer this conundrum. They develop a theory of racial formation—­“the sociohistorical process by which racial categories are created, inhabited, transformed, and destroyed”—­in an effort to help understand what racism is and how it influences American society (Omi and Winant 1994, 55). They emphasize the idea of “racial projects,” which link racial structure and representation; to be clearer, they explain, “A racial project is simultaneously an interpretation, representation, or explanation of racial dynamics, and an effort to reorganize and redistribute resources along particular racial lines. Racial projects connect what race means in a particular discursive practice and the ways in which both social structures and everyday experiences are racially organized, based upon that meaning” (56, emphasis in the original). Both stop-­and-­frisk and affirmative action are racial projects because they are macro-­level social processes that treat people differently due to race (even if not initially intended to do so). But, as Omi and Winant elaborate, “a racial project can be defined as racist if and only if it creates or reproduces structures of domination based on essentialist categories of race” (71, emphasis in the original). Affirmative action or race-­conscious admissions policies “employ racial criteria in assessing eligibility,” but “they do not generally essentialize race, because they seek to overcome specific socially and historically constructed inequalities” (73). It is likely that stop-­and-­frisk was not created with the expressed intention to racially profile and treat Blacks and Latinx people very differently from Whites in an effort to reduce crime, but statistics indicate that police officers (across racial groups) do essentialize race in their implementation of this policy. For example, the New York Civil Liberties Union (NYCLU) found that in 2011, only 11 percent of stops “were based on a description of a violent crime suspect. On the other hand, from 2002 to 2011, black and Latino residents made up close to 90 percent of people stopped”; of these stops, 88 percent were of innocent civilians. Further, the NYCLU revealed

146 / Chapter Five

that “even in neighborhoods that are predominantly white, black and Latino New Yorkers face the disproportionate brunt. For example, in 2011, Black and Latino New Yorkers made up 24 percent of the population in Park Slope, but 79 percent of stops” (New York Civil Liberties Union 2011). It becomes clear that race is a significant factor in police officers’ perception of whom they view with “reasonable suspicion.” Nonetheless, those who subscribe to a color-­blind ideology disregard this difference. So, while color-­blindness is different from older racial ideologies, such as racial resentment, its logic ultimately leads us to the same outcome: a lack of support for policies aimed at reducing socially and historically structured racial disparities. Our White millennial respondents are not supportive of a policy like stop-­and-­frisk that targets Blacks and perpetuates racial inequalities, but they are not in support of policies like affirmative action that aim to reduce racial inequality either.

The Paradox of Generations Before we move on to our respondents’ interpretation and evaluation of these two racialized policies, we should point out the last set of countervailing forces. An increasing number of White Americans, especially younger Whites, have adopted new racial ideologies like color-­blindness, but they have not simply cast off old racial ideologies altogether. Scholars have asserted that cohort replacement has the potential to be a process of “constant liberalization that occurs as older, less tolerant generations are replaced by new generations that have been socialized to be less hostile than their parents to racial integration” (Steeh and Schuman 1992, 341). This may seem intuitive; as societal norms change, people who are socialized within a particular historical moment are typically more likely to hold more progressive values than their predecessors, even if just on the edges. And the process repeats over the course of generations. On the other hand, new generational cohorts do not make themselves, and they do not raise themselves. We must also consider the intergenerational transmission of values, ideas, and stereotypes. That is to say, while young people may have a particular set of attitudes due to their unique ex­ periences, they are also likely to have been socialized by people with less progressive attitudes. Scholars like Schuman and his colleagues (1997) show that there have been positive changes in attitudes around social distancing (e.g., interracial marriage and dating, segregation vs. integration in schools, willingness to have a racial minority as a neighbor) due to cohort replacement. But psychologists have shown that certain attitudes that favor

Racialized Policy Preferences  /  147

one’s in-­group are developed at a very young age, when minds are impressionable and open to accepting the (old-­school) values of parents (Jennings and Niemi 1968). What’s more, Americans of all racial and ethnic groups are socialized in an environment where negative stereotypes about people of color are nearly ubiquitous, and members of all racial groups are similarly inundated by images of Blacks as criminals, Latinx people as “illegal aliens,” and Asian Americans as math geniuses (Gilens 1999; Masuoka and Junn 2013). Donald Kinder, David Sears, and their colleagues (Kinder 1986; Kinder and Mendelberg 2000; Kinder and Sanders 1996; Kinder and Sears 1981; Sears 1988; Sears and Henry 2003; Sears et al. 1980) reveal that there has been a significant shift in the ways in which racism is expressed, but ultimately these new expressions are an evolved form of anti-­Black animus. As a result, young White Americans are not likely to espouse overt, Jim-­Crow racist sentiments, but they are still influenced by the stereotypical ideas and images that pervade America’s racial landscape. The effect of (unintentionally) intercepting these stereotypes may be exacerbated by the fact that most White parents do not talk about race with their children, even though the parents seek to be anti-­racist and hope to raise anti-­racist children (Lewis 2001; Underhill 2018). As we have mentioned before, Scott Blinder (2007) refers to this process as “two-­tracked socialization,” whereby young Whites have been socialized to avoid discussing race, which causes cognitive dissonance because they do not have the language to explain racial inequalities. What’s more, social norms are often transformed in a more progressive direction due to changes in policies, behavior, and values. But others are often reconfigured—­especially those that are positively valenced and deeply ingrained in the American psyche (e.g., meritocracy, individualism, equal opportunity of the American dream). What this means is that these values may be passed on from one generation to the next but are at times unchecked for the negative implications they may pose for realizing meaningful racial equality. Baby boomers and members of Generation X, like many generations before, may pass these values onto the millennial generation, but this time around, these values are likely conduits for newer expressions of racism, such as color-­blindness. For instance, most people value meritocracy and teach their children to appreciate it as well, but we often fail to clarify that meritocracy is not a helpful explanation of patterns of social, political, and economic racial hierarchies (Guinier 2015). Below, several examples of how this set of countervailing forces shapes our respondents’ policy preferences will surface.

148 / Chapter Five

Stop and Frisk We asked our respondents about stop-­and-­frisk because at the time the interviews were conducted, during the summer of 2014, this policy was a sa­ lient topic in the news along with an increasing attention to police brutality. Of the thirty-­nine respondents who were asked and answered the question about stop-­and-­frisk, twenty-­six (two-­thirds) of them were unequivocally op­ posed. Twelve of the respondents supported the policy. The role of racism was brought up among both those who supported stop-­and-­frisk and those who opposed it, but the major mechanisms that sent people into these two different groups were the respondents’ individual levels of tolerance toward racism and empathy toward minorities. For instance, Hunter from Chicago mentioned, I think it [the policy] got the job done [reducing crime]. That said, I think it is racist. Ideologically I think it worked out fine. Or I think it’s a good idea being able to stop and frisk people. That said the way it manifested itself in practicality was really racist and terrible. And I would much rather take a little more crime. And here’s me saying like somebody who doesn’t really have to deal with again a super scary neighborhood, yeah. I think it’s racist. I don’t like stop and frisk. I don’t like stop and frisk how it’s actually practiced. I think certain policies like that could be beneficial sometimes but as they played out they clearly aren’t and therefore I don’t think we should have them.

Hunter stated a belief that even though the stop-­and-­frisk policy has been ef­ fective in reducing crime, he doesn’t like the racial ramifications of the law; he’d rather have more crime than implement what he sees as a racist policy. In contrast, a number of those who were in support of stop-­and-­frisk were aware of the racial ramifications but felt that the potential benefits outweighed the cost of racially biased outcomes. Kaitlyn provided an example of the counterpoint: I mean, yeah, it is. I mean, they are being racist in a way, when they like look at you and they’re like “Oh.” Because if you see like a White man walking, like you’re not really gonna like, like think that much about it. But if you see like a Muslim or a Black person, like you’re probably gonna be like, “What are they up to?” But like at the same time, I’d rather them do that than, like, have like something serious and bad happen.

In a similar vein, Harper from Columbus, Indiana, noted that it’s hard to separate the benefits from the costs:

Racialized Policy Preferences  /  149 I see both points. I think that it can stop crime because some people just aren’t good at hiding when they’re doing something illegal. Some people just have that guilty look. But then I also am brought back to the Trayvon Williams [sic] case where he wasn’t doing anything wrong but he was stopped because he was Black and then he was shot in cold blood for no good reason. So I think it’s both, and there’s no way to really separate which is the better outcome.

Respondents like Kaitlyn and Harper suggested that even though some members of underrepresented groups are more likely to be unnecessarily stopped (and sometimes killed) because of their race, in the long run, the policy makes everyone, or at least people like the respondents, safer.2 These respondents seem to have a higher tolerance for racial bias and its negative effects because they value their personal security over the civil rights and liberties of racial minorities. Relatedly, empathy also influenced the policy preferences of our respondents. For instance, Reagan noted, “So I don’t know if it actually is causing less crime, but I feel like overall it’s really not going to make a big enough difference—­to just stop [minorities disproportionately]—­I would be pissed off if someone just came up to me and stopped me and searched me down, especially if I was a minority. I’m sure that’s—­that that’s just asking for trouble, I think.” When Reagan put herself in the shoes of a person who may be stopped largely because of her race, it infuriated her even though she realized she is not likely to be a person who is directly affected by this policy. Conversely, Elena explained, “I think I would have a different opinion if I was of a different race, but it’s kind of the same in airports. Like, I want all the precautions taken for safety. I’m fine with—­at the airport this last time, I had to get my hands fogged and stuff and random search. I was fine with that. But I think I would get annoyed with it if it happened all the time. So yes to stop-­and-­frisk, but do White people sometimes too.” Elena realized that people of color are likely to be treated differently than Whites, but she saw this as a minor annoyance rather than a systematic problem; for her, the problem can be solved by annoying more White people. Similar to what we noted in the previous chapter, we see that even though both groups—­those for and against stop-­and-­frisk—­are cognizant of racial disparities, an underwhelming level of racial empathy influences some people’s policy preferences. A third major difference between the two groups is that those who preferred stop-­and-­frisk felt that stereotypes are derived from statistics, which are, from their perspectives, always neutral and factual. Daniella provided an example of this sentiment:

150 / Chapter Five Yeah, we also talked about this in Women and Gender Studies. Well, specifi­ cally for stop and frisk, I think that it’s unfortunate that they feel like they have to stop African American people more, and they are just naturally more suspicious of, especially African American males. But at the same time, I think that the crime rate in those areas, the usual perpetrators are African American males. That’s kind of harsh, and that’s also a generalization, and I don’t know any statistics, but it’s like they have reason to believe that these people are dangerous. And so it’s just like a tough balancing act because it’s unfair to assume something about someone just because of their race. But it’s also for the safety of the people. For example, in airports, if they stop Middle Eastern people, it’s not necessarily right to just assume that because you’re Middle Eastern, you’re going to be a terrorist, but who has been the terrorists? And is it likely to be me? Or is it—­like a nineteen-­year-­old White girl? Or is it likely to be a middle-­aged Middle Eastern man?

To begin, Daniella suggested that the statistics that indicate racial discrimination in stop-­and-­frisk are perfectly understandable and even reasonable because police are and should be “naturally more suspicious” of African American men, who are “the usual perpetrators” of crime, writ large. Here, Daniella used the “naturalization frame” of color-­blind racism. Eduardo Bonilla-­Silva explains: “Naturalization is a frame that allows whites to explain away racial phenomena by suggesting they are natural occurrences” (2014, 76). Daniella implied that the feelings police have about Black men are “natural.” Additionally, for Daniella, and many others, it is reasonable to stop and search random Black men and middle-­aged Middle Eastern men because we all “have reason to believe these people are dangerous” (emphasis added). Daniella admitted she doesn’t “know any statistics,” so it is clear that her evaluation of Black and Middle Eastern men and her opinion on this policy are based on preconceived and negative racial stereotypes. She asked rhetorically, “Who has been the terrorists?” but the truth of the matter is, most of the terror attacks in the United States are carried out by right-­wing extremists, including White supremacists, militias, and “sovereign citizens” (Morlin 2018; Ruiz-­Grossman 2017). Daniella did not appear to realize that a larger racialized social system accrues positive stereotypes to some groups and pernicious stereotypes to other groups (Schneider and Ingram 1993). In the end, her attitudes about the groups she mentioned here and her policy preferences actually serve to perpetuate racial inequalities. Jackson from Sarasota, Florida, had a similar feeling and rationale:

Racialized Policy Preferences  /  151 And I don’t have a problem with a—­if there’s a suspicious driver, pulling them over, because they’re African American. The person can whine and scream about, “You just pulled me over because I’m an African American.” Well, that is a fact. You are an African American, and I did pull you over. Those are two undisputable facts. But you were doing some suspicious activity. And so I don’t think that’s beyond the realm of reasoning. And if there is a—­if you look at particular prisons and you look at particular races, and you see that one is more prone to a certain area of crime. If there is one particular area of Miami where the Cubans are involved in narcotics deals, if you see a Cuban driving a huge Escalade, Ferrari, or whatever, and he’s driving it all over the road, this guy might be involved in narcotics, so you pull him over. But if you were to say, you can’t pull him over because he was Cuban—­if you say that, you’re ignoring the fact that there is a huge Cuban narcotic problem, and just because it involves Cubans. I think if you do that, you’re ignoring valuable information.

There are two issues that we want to point out here. The first point is that this kind of response hinges on what Cathy Cohen calls partial truths, “those familiar images and narratives of [underrepresented minorities] engaged in seemingly deviant behavior that are accepted as truth” and “do not need irrefutable evidence to be effective” (Cohen 2010, 23). Michelle Alexander, for example, shows that “about 90 percent of those sentenced to prison for a drug offense in Illinois are African American”; as such, you would find more Black men in prison for crimes around drugs, but what Jackson did not seem to realize or mention is the fact that “white drug offenders are rarely arrested, and when they are, they are treated more favorably at every stage of the criminal justice process, including plea bargaining and sentencing” (Alexander 2010, 184). Indeed, studies show that White professionals may be the most likely of any group to engage in illegal drug activity over their lifetimes, but they are the “least likely to face the criminal justice system for doing so” (Alexander 2010, 192). Accordingly, it is partially true that “if you look at particular prisons and you look at particular races, . . . you see that one is more prone to a certain area of crime,” as Jackson said, but Jackson missed out on the larger, systematic, discriminatory factors that lead us to see prisons that have overrepresented Black and Latinx populations. Jackson’s sentiments raise another red flag. Jackson reasonably justified that if a person is driving “all over the road,” you should probably pull the driver over, but Jackson reinforced his preference for racially biased policies in his next sentence. Jackson suggested that, rather than pulling this person

152 / Chapter Five

over just for dangerous driving, we should also consider the race or ethnicity of the person when doing so since this information may allow us to find another, unrelated reason to pull this person over anyway. In the example that Jackson provided, we shouldn’t pull the person over because they are driving dangerously, we should do so because it is a Cuban person driving dangerously and therefore this may indicate that the driver is involved in narcotics—­rather than the more common and obvious crime of drunk driving.3 These are two representative responses of those who supported stop-­ and-­frisk. To summarize, people in this group believed that some people just look more suspicious, and often those are people of color. Secondly, they believed that the people who are most likely to be stopped by the police are those who fit the profile of people who are more likely to commit crimes. This line of reasoning suggests that because the majority of people who are convicted of certain crimes are of a particular group, all members of that group are suspicious. On its face, this logic seems reasonable. Indeed, Miles helped us think through this rationale: “You can’t blame the police for pegging African Americans as more likely to commit a crime than Whites because if you went in a town and I were to say 90 percent of the crimes, not 90 percent, a larger percent of the crimes are committed with people with pinky rings and there was a crime committed. Who would you look for first, the people with the pinky ring or the people without a pinky ring?” Let’s think about this in another way and use another example. Between 1982 and 2012, there were sixty-­two mass shootings in the United States. Sixty-­four percent of them were committed by White people—­almost exclusively White men. Should we be fearful of all White men who walk into an elementary school, a Black church, a movie theater, or an Army base because forty White men, or 0.0000003 percent of the White male population, have committed such an atrocity? Should we stop and frisk more White men in the face of an increasing number of mass shootings in the United States? Even though most mass shootings are executed by White men, most White men—­the overwhelming majority—­do not commit mass shootings, and therefore, we would see it as completely ridiculous to assert that because members of some groups are more likely to commit (and be convicted of ) a certain crime, we should be willing to allow all members of that group to be subjected to racially discriminatory policies. Those who subscribed to this line of thinking were a minority—­although not a small proportion (31 percent). Most of our respondents did not support stop-­and-­frisk and believed that the policy had too many racially pernicious discriminatory effects to justify its continuation. Claire, for example,

Racialized Policy Preferences  /  153

represented this group fairly well when she stated, “I think that’s a pretty dangerous one in that it sounds like it—­the potential for profiling and stereotyping is pretty high. And it sounds like a pretty dicey piece of legislation.” Those who opposed the policy suggested that even though it is intended to protect citizens, it might be enacted by individuals who have racially biased views, thereby leading to racially biased outcomes. For instance, Alaina, from Alexandria, Indiana, mentioned, I think it [the policy] leads to unfair racial profiling a lot, actually. I have a friend whose father is a police officer, and he’s a very, very good man. But I know there are people on his squad who—­in the city—­who are more likely to stop African Americans or Latinos than they are White people who are just as likely to commit a crime. I mean, I understand the policy might have been set in order to protect people, but I don’t think that’s how it’s being used at all sometimes.

Alaina, like a number of others, separated intent from outcomes and recognized the room for discrimination in the execution of the policy. People also mentioned that they learned about these outcomes through various sources—­classes, YouTube, or the news—­and that police officers, whether they wanted to or not, were more likely to target Blacks, either due to career incentives or unconscious biases. Ava, from Watertown, Massachusetts, explained: Yeah. I’ve watched—­I’ve seen a couple of videos on YouTube of people talking about stop and frisk. I have to say I really don’t like it because yes, I think they have caught more people with illegal hand weapons or with drugs or committing or planning to commit criminal activities. Also I think that’s just because they keep upping the number of people that each policeman has to stop and frisk. I forget if it’s by the end of the week or the month or whenever, but they keep upping the number. So, yes, if you constantly up in numbers of times that you have to stop people, then yes, statistically, you would probably find more stuff. But then also it encourages racial profiling so much.

Similarly, Madelyn noted, Absolutely ridiculous because they—­I’ve seen videos where like some people are—­and they’re usually Black males or Hispanic males—­they’re stopped on the street multiple times a day, for walking down the street. And it’s because these policemen have quotas that they have to reach, and a certain amount

154 / Chapter Five of tickets they have to write and this and that. I’ve seen like interviews with these policemen that are just—­they hate their jobs because they have to do this, and they don’t want to have to do this. They know these people aren’t doing anything wrong. They’re just doing it because they have to, to meet certain standards. Absolutely ridiculous. It should not be happening. Because it’s discriminatory against groups of people, and it’s not benefiting anybody.

These kinds of sentiments are not born out of paranoia. In a federal class-­ action lawsuit, Floyd v. City of New York, a police officer testified that he was mandated to log at least five stop-­and-­frisks, make one arrest, and write twenty tickets each month (Carver 2013). Even New York Police Department commissioner William Bratton admitted that there were arrest quotas when he promised to “focus on the quality of policy actions, with less emphasis on our numbers” (Morales and Tracy 2014), but nearly a year into his tenure, quota-­based policing was still part and parcel of police tactics in New York City (Mathias 2014). Moreover, several of our respondents felt that, aside from having incentives to stop certain people, police were prone to stop Black and Latinx people more often due to the (often unintended) influence of the stereotypes that we all know, but may or may not subscribe to, about members of these groups. For instance, Riley, from Connecticut, explained why she could not support stop-­and-­frisk: Yeah. I mean, I think—­I feel like if you’re gonna do it, I don’t think it’s necessarily—­I’m really not—­I’m really against guns, so getting guns off the streets, that’s a great thing. But I think definitely if they are gonna have a policy like that, it needs to be very clear what is suspicious activity. And then officers also need to be aware of—­not necessarily—­that they might subconsciously be stopping more Black people than White people, or whatever.

Here, Riley referred to unconscious stereotypes and implicit racial biases that police may have, just as we all do. An individual holding and relying on negative stereotypes serves as a salient example of modern-­day racism for White Americans. Meanwhile, other respondents were simply aware that some police have explicit biases and would use their power to discriminate against racial minorities, which our respondents viewed as an egregious behavior. Jordan and Sebastian provided examples of this sentiment. First, Jordan: “But—­ yeah, I think just most people judge with their eyes so White people can’t be trusted to not just single out minorities more than White people. Stop

Racialized Policy Preferences  /  155

and frisk—­has crime gone down since this policy started? I don’t know. If it had significantly then sure, but I doubt it has. I don’t like it.” Jordan suggested that Whites are simply more likely to react negatively toward racial minorities than they are to Whites. Studies show that even in a controlled lab setting, police are more likely to shoot an unarmed Black suspect than an armed White suspect (Plant and Peruche 2005).4 In the same vein, Sebastian suggested that the policy simply will not be implemented fairly because there are police who are racist: “So when you see that Blacks actually have less in terms of legal possessions when they’re stopped and frisked, I think we have to realize that it’s a racist policy then. And I think there’s a difference between a policy itself being that racist and an execution of policy being racist because if this policy was enacted by perfect police officers who weren’t racist, maybe things would work out.” All in all, those who were against stop-­and-­ frisk were concerned about the policy primarily because of its negative racial ramifications—­intentional or otherwise. As shown in chapter 4, White millennials tend to understand racism in terms of stereotypes and overt discrimination. It’s easy for them to see how this plays out in YouTube videos and statistics that illustrate and provide evidence for the argument that police are more likely to harass Black and Latinx men than Whites. For this group of respondents, it’s pretty cut and dry: racial discrimination is wrong. It’s wrong to work out your racist attitudes on people of color. It’s wrong to work out your implicit biases on people of color. It’s wrong to use minorities as easy targets in efforts to increase the number of tickets police have to write. The policy is just plain wrong, even if the intention of the policy was something wholly different, because of the clearly racially discriminatory effects. Arguably, these attitudes are laudable, but in these responses, we do not see many explanations that focus on structural racism. Stop-­and-­frisk policies tend to be enacted in racially segregated neighborhoods; these neighborhoods tend to see high numbers of certain crimes and also high poverty. Stop-­and-­frisk is a policy that is aimed at stamping out a symptom rather than the causes of a larger, structural problem. Furthermore, research shows that “even when these stops yield arrests, almost all are low-­level, many resulting directly from citizens questioning the rights of the police to stop them in the first place. While most of these arrests don’t result in criminal convictions, they often trigger severe consequences—­including job loss, eviction, and even deportation of permanent residents who are not citizens” (Communities United for Police Reform 2012). Young White people have focused on the notion that police are likely to stop people because of race, but they have not taken the next step to recognize the larger, more racially

156 / Chapter Five

pernicious effects that these policies have on communities of color, because their understanding of “racism” is quite limited. Nonetheless, White millennials’ attitudes about stop-­and-­frisk reveal a sign of what many may identify as racially progressive policy preferences.

Affirmative Action Most of our forty-­three White millennial respondents had something to say about affirmative action, whether or not we prompted them to talk about it. While most of our respondents were opposed to stop-­and-­frisk because they believed it was racially discriminatory, we found that they were also very much opposed to affirmative action. Affirmative action and racially conscious admissions policies were originally developed to ameliorate institutionalized and covert forms of race (and gender) discrimination, but today their use and justification have evolved to focus on issues of diversity. So, while our millennial respondents were highly opposed to a policy that is racially discriminatory, they were also largely opposed to a policy or set of policies that aims to close gaps between racial groups. Their responses, ideas, and considerations about affirmative action were much more nuanced than their ideas about stop-­and-­frisk, and there were a number of themes that arose as our respondents talked about affirmative action, especially as it related to the college admissions process. To begin, we found that there were three major ways that people thought about affirmative action: as quotas, as a means to address diversity, or as a way to make up for past discrimination and provide equal opportunities to minorities. Secondly, we found that the way that people understood race-­conscious admissions and hiring processes influenced their level of support. Finally, we found that even though there were some who were willing to make clear statements for or against these kinds of programs, most of our respondents were against them but used a particular semantic strategy to suggest that they are liberal or in support of the policy before ultimately making a statement against it: “Yes, but. . . .” We explore these themes around affirmative action in the remainder of this chapter. Quotas: You Do It on Percentages The most prevalent way that our respondents understood affirmative action and race-­conscious admissions policies was via the notion of “quotas,” even though the US Supreme Court outlawed quotas several decades ago. The Supreme Court’s decision in Regents of the University of California v. Bakke

Racialized Policy Preferences  /  157

(1978) upheld affirmative-­action policies, allowing race to be considered among a number of other characteristics in college admissions policies and processes, but ultimately the court ruled that quotas or a specific number or proportion of set-­aside seats was unconstitutional. This case occurred long before anyone in the millennial generation was born, and furthermore, it occurred when many of those in Generation X were teenagers; in all, two generations have lived in an America where quotas in college admissions are banned, but we see that the plurality of our respondents expressed a belief that affirmative-­action policies nonetheless enable a quota system. About 40 percent (sixteen) of our respondents believed that these kinds of policies were to, in Chloe’s words, “reserve spots for certain people.” As Grayson explained, “I’d say kneejerk, I think of quotas.” Even though millennials have been socialized in a different social and political milieu than their predecessors, they’ve gained insight about the way the world works from members of previous generations and use those insights to inform their policy preferences. Rather than seeing affirmative action as a compensatory policy that provides greater opportunity for future success, 40 percent of our respondents viewed affirmative action as a policy that gives unfair and unnecessary preferential treatment to some groups over others. Ophelia provided an example of this sentiment. When we asked her what comes to mind when she thinks of affirmative action, she said, “I think of quotas and like letting a certain number of different races into a school.” She explained how a system like this works: “Like you know if you set a number, maybe you have this like five people from Alabama. And then only five people from Alabama apply, you know? That’s just like, that’s an example like—­that seems silly then that you would have to admit all those kids if you have a couple kids from New Mexico who are more qualified.” Actually, the Supreme Court’s most recent landmark decisions on race-­conscious admissions policies, Grutter v. Bollinger (2003) and Gratz v. Bollinger (2003), also known as the University of Michigan cases, reaffirmed that universities may take race into consider­ ation among many other factors, noting that the hallmark of the University of Michigan Law School’s admissions policy “is its focus on academic ability coupled with a flexible assessment of applicants’ talents, experiences, and potential ‘to contribute to the learning of those around them’” (Grutter v. Bollinger, 2003). In this case, the court reiterated that quotas are “patently unconstitutional,” and ultimately determined that there is a compelling state interest to attain a diverse class. Justice Sandra Day O’Connor, in the majority opinion for Grutter, explained: “In order to cultivate a set of leaders with legitimacy in the eyes of the citizenry, it is necessary that the path

158 / Chapter Five

to leadership be visibly open to talented and qualified individuals of every race and ethnicity. All members of our heterogeneous society must have confidence in the openness and integrity of the educational institutions that provide this training. As we have recognized, law schools ‘cannot be effective in isolation from the individuals and institutions with which the law interacts.’ ” The court supported the University of Michigan Law School’s goals of attaining a diverse student body not to make up for past discrimination (which was allowed in Bakke) but instead as a means to create an environment for their entire student body to thrive. We should expect the vast majority of millennials to believe that colleges adopted and continue to adopt race-­conscious admissions policies for educational reasons, because unlike members of the boomer generation or even Generation X, the eldest millennial was born two years after the Bakke case and entered college nearly two decades after the case was settled. To boot, many members of the millennial generation did not even begin the college admissions process until after the Michigan cases were decided. Despite these chronological facts, we see that millennials are still most likely to adopt the understanding of affirmative action and race-­conscious admissions policies as “quotas,” a system that was deemed illegal many de­ cades ago. The Diversity Dilemma Resurfaces About one in six individuals viewed affirmative-­action policies mainly as a means for colleges to promote diversity. Even though White Americans place a high value on diversity, especially for their own learning and growth (Embrick 2011; Berrey 2015a; Edelman, Fuller, and Mara-Drita 2001; Berrey 2015b; Andersen 1999), our respondents tended to be skeptical about in­stitutions of higher education that value diversity. For instance, Easton, from Baltimore, mentioned, “Well I don’t know if affirmative action is about fairness. I think schools do it because they want more people of color in their community.” Similarly, when we asked Reagan from Elk Grove Village what comes to mind when she thinks of affirmative action, she replied, “I think of accepting races just to say, ‘Oh look, I’m not being racist.’ And maybe giving a scholarship or a job to someone of a minority race over, I don’t know, a Caucasian just because of their race so they can look like they’re not being racist. That’s what I think of.” This kind of skepticism was also marked in an exchange that Mackenzie from Lexington had with one of our interviewers:

Racialized Policy Preferences  /  159 mackenzie: Yeah, I don’t think, I think if you’re applying for like a school you should get it on merit, not on, on race. But like I understand, because then the school gets shit for not having enough variety, and then we’re like, “We should have variety.” So like, I guess I’m kind of being hypocritical, but like—­ interviewer: So do you think that, like, the purpose of affirmative action is not, is more to aid the school in, like, obtaining a diverse population, or do you think it’s like—­ mackenzie: To help the kids. interviewer: Yeah. mackenzie: I think it’s—­me, this is probably me just being negative. I think it’s the school. interviewer: Yeah. mackenzie: Because I’m like, “Fuck you, institution.” interviewer: Yeah. mackenzie: Because really, like they’re worried about funding, and funding depends on numbers, and so they’re trying to get numbers. interviewer: Yeah. mackenzie: That they need for their funding. I think when it really comes down to it, it’s all about money. It’s not actually about helping people, I’d say.

There were seven people in this group, and nearly all of them focused on the affirmative-­action policies in higher education, especially admissions. Many of them suggested that colleges and universities implement these policies to increase diversity, but as illustrated by Reagan and Mackenzie, there were underlying feelings of skepticism and resentment toward institutions for trying to attain a more diverse student population. Reagan, specifically, suggested that institutions accept people of color and provide scholarships in large part to show that “they’re not being racist.” Mackenzie, in particular, brought a sense of reflectiveness on the part of the respondents. On one hand, she suggested that most people feel that “we should have variety,” and she realized that because diversity is valued in society, “schools get shit for not having enough variety.” On the other hand, she derided the institution for actually implementing policies that aim to accomplish that very goal and create a more diverse environment; implicitly, she ridiculed these institutions of higher education because of her belief that meritocracy is sacrificed and because “it’s all about money. It’s not actually about helping people.” Irony and hypocrisy abound here, and Mackenzie was quite cognizant of that. Here, we see that colleges and universities get berated for

160 / Chapter Five

(a) doing what society thinks they should do and (b) doing what’s in their self-­interest. Recall in the previous chapter that millennials demonstrated that their value for diversity derives largely from self-­interest (i.e., desire for a better learning environment and increased chances to learn about other cultures). While nearly all of our respondents valued diversity, they did not necessarily also believe that diversity should be a priority for institutions of higher education. Here, we see how the “diversity dilemma” plays out in the policy preferences of White Americans. Perhaps, though, there is not a contradiction. As mentioned, diversity ideology only requires intention, not action (Mayorga-­Gallo 2019), and here, when we see action through affirmative action or race-­conscious admissions policies, many White Americans are none too pleased. To Ensure That Historical Inequalities Don’t Calcify Thirteen out of forty-­two respondents (31 percent) viewed affirmative action and race-­conscious admissions policies as efforts to “level the playing field,” given a legacy of racial discrimination in the United States. For example, when we asked Claire, a thirty-­two-­year-­old admissions counselor at a community college in Massachusetts, what comes to mind when she thinks of affirmative action, she said, “I think about a population that has historically had some challenges and there being steps and processes and accommodations made so that those groups can have a more equitable stake.” Similarly, Dylan, from Cleveland, explained, I think it’s a really important thing because I think that affirmative action existed for a long time before it was ever called affirmative action in its antithetical form when universities had quotas on the number of students that could be admitted. My grandfather got into Cornell by applying through the School of Forestry whose Jewish quota was not quite as full at the time as the School of Earth and Sciences, and then transferred to the School of Earth and Sciences. And I think affirmative action has always existed. We have now given it a name, and tried to equalize it. And in trying to equalize it, it has meant—­sorry, we have given it a name, and tried to make it fairer and in making it fairer have made it, in some people’s understanding unequal to the opposite side. I think the notion that affirmative action is giving undeserving minorities a break is completely nonsensical, but I think it’s a really important way to balance the scales of who we get in to higher education.

Racialized Policy Preferences  /  161

Dylan took an approach similar to Ira Katznelson’s in When Affirmative Action Was White (2005). Katznelson explains that historically America has had a number of “color-­blind” or race-­neutral public policies, like the G. I. Bill, that ultimately served to benefit White Americans almost exclusively. He calls these policies affirmative action for Whites. In the same vein, Dylan discussed the idea that there have always been race-­or ethnicity-­based policies in college admissions processes, similar to what Katznelson details in his book, but Dylan argued that over time there has been a shift from primarily benefiting White Anglo-­Saxon Protestants to making these policies “fairer.” Dylan noted that some people have come to believe that these policies are now “unequal to the opposite side” or, in common parlance, “reverse discrimination.” In all, however, when he thinks of affirmative action, he sees this kind of policy as a way to level the playing field or, in his words, a “way to balance the scales.” In addition to considering the legacy of racial inequality, some also noted ongoing discrimination as a factor that makes affirmative action necessary today. Meanwhile, others noted that there is also a gender component to affirmative action. These two sentiments are best represented by Daniella and Madelyn’s responses. First, Daniella, explained, “So affirmative action is providing opportunities for people of like minority groups, and also women. I think it’s a great concept that the government was providing opportunities and making sure that not all White men got everything.” Second, Madelyn stated, I think that it’s a really difficult—­I think it’s something that is very complex, and I think that systems like that do need to be in place, only because if they’re not, then people will be discriminated against, which they already are, but they’ll be even more discriminated against than they already are. The reason why affirmative action exists is just to give equal opportunities to people of color and to women. We forget that women are part of that. And actually, statistically, if you look at the statistics, the group that affirmative action has helped the most is White women. So I am helped.

These two respondents recognized that affirmative-­action policies are aimed to consider the role that race and gender play in the way that opportunities are distributed. In all, most respondents in this group saw affirmative action similarly to the way that Savannah from Ridgewood, New Jersey, saw it when she explained, “And so that’s what affirmative action tries to do: basically, to ensure that historical inequalities don’t calcify or don’t continue on into the future.” That is to say, they saw it as a means to compensate for past

162 / Chapter Five

injustices. Others in this group saw the policy as a compensatory policy, or a policy aimed to allow people of color and women of all races to compete in the market. Attitudes about Affirmative Action: Yes, No, and Heck No The way that people perceive and interpret a policy will influence their support for that policy. For example, in 1983, James Kluegel and Eliot Smith noted that “programs seen as promoting preferential treatment (e.g., hiring fixed numerical quotas of minorities) are overwhelmingly opposed by approximately 90 percent of whites”; meanwhile, programs believed to simply help Blacks gain access to jobs or higher education are supported by about 70 percent of Whites (1983, 797). We find evidence for this claim as well. Table 5.1 examines the relationship between the way our interview respondents perceived or understood “affirmative action” and their level of support for it. The three major ways that people interpreted affirmative action and race-­conscious admissions policies were as quotas, as a means to increase diversity, or as an effort to make up for a legacy of racial discrimination and inequality. When we examined our interview respondents’ attitudes toward and levels of support for affirmative action, we found that they gave three kinds of responses: unequivocal support (yes), unequivocal opposition (no), or a third category where the respondent gave reasons why they might think about supporting affirmative action but ultimately decided that there were more costs than benefits (yes, but . . .). Even though the “yes, but” responses are ultimately “no,” Eduardo Bonilla-­Silva reveals that this kind of answer is a semantic move typical of color-­blind racism (2014, 108). We will provide examples of each of these three categories in the following section. With a simple χ2 test of independence, we found that there is a statistically significant relationship between the way people interpret affirmative action and their level of support. Table 5.1 shows that not a single one of our respondents who understood affirmative action as a system of quotas supported the policy. About 28 percent of those who believed that affirmative-­ action policies are largely a way for institutions of higher education to create a diverse student body supported affirmative action. Conversely, we found that the majority of people who viewed the policy as a means to disrupt a legacy of racism and provide opportunities for minorities to access jobs and higher education supported affirmative action (54 percent). On the one hand, 54 percent is almost twice as high as the support given by those who viewed affirmative action as a diversity initiative and fifty-­four points higher

Racialized Policy Preferences  /  163 Table 5.1.  Relationship between Policy Interpretation and Level of Support

Interpretation

Level of Support

Quota Diversity Past / Equality Other Total

Yes

Yes, but . . . no

No

Total

0 2 7 2 11

9 2 5 2 18

7 3 1 2 13

16 7 13 6 42

χ2 = 12.76, df = 6;   p < 0.05 Data: Face-­to-­face interview respondents

than respondents who expressed a belief that affirmative action is a system of quotas, but it is much lower than the proportion of White respondents who shared a similar rationale in Kluegel and Smith’s 1983 study (70 percent in support). If That’s True, Then Yes Of the forty-­two respondents who engaged in this line of questioning, eleven of them (26 percent) gave their support to affirmative action and/ or race-­conscious admissions policies. As table 5.1 shows, most of these individuals expressed a belief that these policies were created and implemented in order to make up for racial discrimination or to provide people of color and women with equal access to opportunities. For example, Brody, from a small town in the Berkshires of Massachusetts, explained that even though “I don’t know that if it’s true, if they’re [minorities] not getting a fair chance,” he still supports affirmative-­action policies, since he sees them as a way of providing greater opportunities for those who are not getting a fair deal. He elaborated, If that’s true [that minorities continue to be discriminated against] then, yes, they deserve a fair chance and there’s rules that need to be put in place to ensure that they get a fair chance. I’m in support of it. It’s unfortunate to think that you would still need a fair chance in this day and age and that’s what I don’t know. That would be what I want to base my decision on is whether or not they’re given a fair chance, and if they’re not then I’d wanna know why first of all and because yeah, they should be given a fair chance and if you have to make a rule, it’s pretty sad but yeah, they should be given a fair chance.

164 / Chapter Five

Despite Brody’s naivete about the ongoing role of racial discrimination and structural racism, he was quite earnest, and his logic mirrors that of someone whom we would consider racially progressive: if racism constrains opportunities for people of color, then we should have policies that remove or lessen that constraint. One interesting thing is that even though some people believed that affirmative action shifts opportunities from Whites to others, they were still supportive of the policy because they felt that in the end they either (a) have lots of other opportunities available to them or (b) would benefit from affirmative action because they saw it as a diversity initiative. For example, Savannah noted, “I know that it’s [seeing fewer women of color] not because these girls are less smart than me or less talented or less anything really. It’s just they’ve been given less opportunities. And so why not try to even that out a little bit with more opportunities? I feel like I have a lot of stuff going for me.” Savannah expressed support for affirmative action because she realized that talent is equally distributed among racial groups but opportunity is not; she was supportive of shifting some opportunities to people of color because, in the end, she has other talents, privileges, and points of access to opportunities for success. To the second point, we found that some individuals who saw affirmative action as a means to increase diversity supported affirmative action because ultimately it would benefit them. For instance, Hunter, a twenty-­one-­year-­ old man from Chicago, explained, “I like the idea that they let people in to have more diversity because it ends up, even though I’m like a straight White guy, it lets me see other people.” And recall that Madelyn noted, “The reason why affirmative action exists is just to give equal opportunities to people of color and to women. . . . So I am helped.” Both of these cases represent the idea that affirmative action is good because it is good for everyone, even Whites, who may get to interact with non-­Whites, and White women, who may get access to various opportunities previously denied to them because of historical and ongoing systems of paternalism and patriarchy. Minorities Aren’t Minorities Anymore On the other end of the spectrum, there were thirteen respondents who were crystal clear in their opposition to affirmative action. These individuals largely fell into the category of people who believe that affirmative action and race-­conscious admissions policies are synonymous with quotas. Two major themes arose among those who outright rejected these kinds of policies, and they closely align with frames of color-­blind racial ideology:

Racialized Policy Preferences  /  165

abstract liberalism and minimization of racism. Abstract liberalism “involves using ideas associated with political liberalism (e.g., ‘equal opportunity,’ the idea that force should not be used to achieve social policy) and economic liberalism (e.g., choice, individualism) in an abstract manner to explain racial matters” (Bonilla-­Silva 2014, 76). In this case, respondents called on notions of meritocracy, noting that quotas are unfair and lead to unqualified individuals attaining admissions or jobs they do not deserve. “Minimization of racism” is a frame that suggests that racism is no longer a central factor affecting racial and ethnic minorities’ life chances. Here, we show that many of those who opposed affirmative action did so because they believed that racism no longer influences the opportunity structure for racial minorities. Jacob, from Maryland, provided an excellent example of abstract liberalism, and his response illustrates the link between one’s interpretation of a policy and one’s policy preference. Jacob explained, “Affirmative action to me is actually almost laudable in the fact that it is so explicit. To say we are going to have quotas or heavy unwarranted—­or heavy a priori biases towards accumulating individuals for these meritocratic positions on nonmeritocratic bases. So it is nice to have an enemy to fight against openly and explicitly.” Jacob saw affirmative action as both a quota and an enemy against meritocracy. He is erroneous on both counts, however. First, we have already discussed that affirmative-­action policies are not a quota system, nor could they be in this day and age. Second, Jacob assumed that meritocracy is the crucible of college admissions. We imagine that Jacob, like many other Whites, believes that meritocracy, and therefore university admissions, should largely be a function of quantitatively measurable outcomes like the SAT, the LSAT, and one’s GPA. Both of the Michigan cases and, more recently, Fisher v. University of Texas at Austin (2013) arose because the women who filed the cases believed that their test scores and GPAs were more meritorious than those of some of the racial minorities who gained admission. However, time and again, research shows that test scores are very highly correlated with applicants’ parents’ incomes and level of education (Goldfarb 2014; Rampell 2009; Schmidt 2007). Regarding the Fisher case, more specifically, Nikole Hannah-­Jones (2013) reports, “It’s true that the university, for whatever reason, offered provisional admission to some students with lower test scores and grades than Fisher. Five of those students were black or Latino. Forty-­two were white.” Furthermore, there were “168 black and Latino students with grades as good as or better than Fisher’s who were also denied entry into the university that year” (Hannah-­Jones 2013). Jacob, like many others, viewed meritocracy as a means to combat

166 / Chapter Five

unfairness, but when you look more closely at the details, nuance, and history of standardized tests, it becomes clear that meritocracy, particularly in the way that Jacob saw it, reflects unearned privileges and an imbalanced distribution of access to opportunity (see Guinier 2015). The second theme that arose among those who opposed affirmative action is well illustrated by Maria and Kaitlyn’s responses. Maria, who opposed affirmative action, shared this sentiment: “Okay. I understand the reasoning behind affirmative action like in college admissions and in hiring, and I understand that a company would want to have a racial diversity and to be known as that, but I think we’re getting far enough away from like the civil rights movement, and people my age shouldn’t be able to claim that like they had all these hardships; we weren’t born.” Sharing a similar explanation, Kaitlyn discussed this matter with one of our interviewers: interviewer: So do you think that we still need policies that guarantee minorities a fair shake in hiring and college admissions? kaitlyn: I don’t think so just because it’s like, we’re so far beyond it. interviewer: Yeah. kaitlyn: And like I think minorities, like what was considered a minority, like the race that were minorities, aren’t minorities anymore. interviewer: Yeah. kaitlyn: Like you know what I, like I feel like because there’s like, there are like, they’re not like—­I feel so racist. interviewer: No, it’s fine. Yeah. kaitlyn: But there like, there are like so many more Asians now. interviewer: Yeah. kaitlyn: And like more Blacks, and like I feel like they all like have the same rights, and they all were, like they all are really successful, that like there, there’s no need for that anymore. interviewer: Yeah. So do you think that it’s like at this point, it’s like pretty much an even playing field and like we just like don’t need policies like this anymore? kaitlyn: Yeah. Yeah.

Both Kaitlyn and Maria showed a belief that racism and discrimination are things of the past. Maria suggested that millennials or “people my age,” in particular, do not have a legitimate claim to being victims of racism or discrimination, since we have moved far enough away from the civil rights movement. Kaitlyn suggested that policies aimed at creating a more even playing field are unnecessary because “we’re so far beyond it” (beyond being a racist or discriminatory society). Kaitlyn also suggested that since she sees

Racialized Policy Preferences  /  167

more Asian Americans, that must indicate that this group must have more power or prominence, which is another reason to intimate that we are no longer in need of policies aimed at preventing discrimination or assuring fair access to jobs and higher education. Relatedly, she stated a belief that Blacks “have the same rights, and they were all . . . really successful”; she used this belief as evidence that racial and ethnic minorities are no longer minorities in the “numerical” sense (especially Asian Americans) or in the “underrepresented” sense of the word. Instead, Kaitlyn—­who later explained, “I think like Caucasians are like becoming more like a minority now than, like, other races, especially like in Lexington”—­felt that racial minorities have accrued the power, prominence, population, and rights to do well without policies that guarantee fairness in hiring and in college admissions. We demonstrated in chapter 4 that a sizable proportion of  White millennials believe racism is something that happened in the past, and therefore they are unable to see it in its current iteration or they believe that it no longer exists. Here, we see how this aspect of color-­blind racial ideology—­ the minimization of racism—­has a major influence on policy preferences. Our respondents’ answers also show how abstract liberalism, with a special focus on “meritocracy,” allows individuals to suggest that affirmative action and race-­conscious admissions policies are what is actually racist nowadays. Yes, but . . . The most frequent sentiment expressed by our respondents about their preference for affirmative action was “yes, but.” Upon first glance, the plurality (42 percent) of our respondents appear to be ambivalent about a set of policies that are controversial and generate a lot of debate. Ambivalence in survey responses may best be understood as “a problem of reconciling the multiple values, beliefs and principles simultaneously present in the political culture” (Feldman and Zaller 1992, 270). For example, Jennifer Hochschild interviewed twenty-­eight people from New Haven and found that when they tried to apply norms of economic inequality to modern problems of social welfare, ambivalence arose because they recognized the underlying conflicts between their own experiences and egalitarian beliefs; she found that “they find it easier to live with, and try to ignore, even distressing normative tensions than to undertake the enormous effort needed to resolve them” (1981, 258). Hochschild’s respondents have a value conflict and are thus ambivalent; the political implication “is to leave people in a state of political paralysis because they cannot think their way through their mixed beliefs to a definite perspective from which to act” (Hochschild 1981, 258).

168 / Chapter Five

In contrast, our respondents appeared to be ambivalent, but they were not distressed or paralyzed in their policy preferences on this issue. Harper, from Columbus, Indiana, provided an example of the appearance of ambivalence without the actuality of it: I want to say yes, but at the same time, I want to say no. Because I want to say no because, again, if they’re qualified, great, awesome, fantastic. They should be able to get the job, they should be able to go to the school, get into the college they want to get into. But at the same time, they’re so looked down upon that I feel like they still need help. At the same time that race question may not help them. Maybe the admissions committee is against Black people so all African American applications are going to get thrown in the trash. They could have some personal vendetta against some race and then it hurts them in the end. So, I’m kind of in the middle on that one, I guess. (Emphasis added.)

Even though Harper said that she is in the middle, she is not. She wants to say yes, but all of the rationale she provided supports her desire to “say no.” She suggested that if people are qualified, they would not need affirmative action or race-­conscious admissions policies. She used the phrase “at the same time,” implying that she is going to weigh another side of the debate, but she does not. She explained that asking for an individual’s race on admissions applications may ultimately do some minorities harm because of a hypothetical racist admissions officer who may have it out for Black people. Even though she ended with the idea that she is neither for nor against race-­conscious admissions policies, her underlying sentiment is opposition. We find that, in addition to using this semantic strategy, our respondents did appear to be engaged with various sides of the debate and, further, did appear to have a value conflict, of sorts. It came in the form of one of the countervailing forces: the diversity dilemma. Our respondents often began with the fact that they value diversity and realize that diversity on college campuses is very important, but when they weighed diversity, on the one hand, against fairness, merit, equal opportunity, and/or rigor, on the other hand, diversity always lost and so did support for affirmative action and race-­conscious admissions policies. Alaina from Indiana provided an example of the semantic strategy in combination with the diversity dilemma: I mean I understand the issue with having predominantly White males in a college and how that would look bad for a university because obviously that’s saying that they’re not diversified and they don’t acknowledge that is an important thing. Bloomington is a very liberal college so I can understand

Racialized Policy Preferences  /  169 the issue if tomorrow we stopped doing affirmative action—­or I don’t even know if we do. If we stopped and we only had White males applying or White females, I can see how that’d be an issue, but at the same time—­the fact that there are some students who get turned down from colleges or get turned down from jobs just because they’re not Black, or White.

Alaina appeared to weigh the pros and cons of affirmative action and race-­ conscious admissions policies. She “understood” why it would be problematic to have a largely homogenous campus, and that it would “look bad” for a university that does not acknowledge that diversity is important. Alaina also pointed out that if her college did not consider race or gender in its admissions process, the applicant pool might also change. These are all real, pressing concerns, and when we look at states and schools that have banned race-­ conscious admissions policies, their student bodies do become less diverse and more homogenous, sometimes even less White, in the case of some very elite schools. But ultimately, Alaina decided that “the fact that there are some students who get turned down” because of their race makes the policy unfair, and therefore she is against the policy. Another example comes from Peter, a twenty-­three-­year-­old from Concord, Massachusetts: I think that to some extent that it is a good thing, but as far as on a large scale like it is used right now, I believe that either way it can be racist. It is a racist thing because you’re excluding one group based on the color of their skin. You’re excluding Whites based on the color of their skin to get more African Americans or people of Latino descent in to admissions. I think that in order to propel diversity in a school that affirmative action can be used responsibly on a smaller scale.

Peter began with the sentiment that affirmative action is a good thing, and even suggested that when used “on a smaller scale,” it would be helpful to “propel diversity,” but in all, he considered the program to be racist because “you’re excluding one group based on the color of their skin.” This logic appears to be reasonable on its face, but there are a number of contradictions, or, at least, there are some dots that have not been well connected here. Peter suggested that diversity is important and racial exclusion is bad, but he failed to realize that diversity initiatives, race-­conscious admissions policies, and affirmative action are all aimed at addressing both of the issues he is concerned about. These dots could not be connected for Peter because he saw diversity as synonymous with White exclusion. He ended by saying

170 / Chapter Five

that he believes affirmative action can be used on a smaller scale, but it is unclear how much smaller of a scale you can have than a college campus or a professional school. As revealed in table 5.1, the largest proportion of people in this “yes, but” category is made up of those who see affirmative action as “quotas.” Here, we find that people weigh the value of diversity with the unfairness of quotas. Callie, a twenty-­year-­old woman from Riverside, Illinois, explained, “I think when affirmative action came out, it was a good idea, but I think that also it incentivizes people to hire the wrong people just to fill a quota. And so it has that certain unintended results because then you have people who aren’t qualified getting jobs.” Similarly, Elena, a twenty-­one-­year-­old from Indianapolis, shared her ideas on this issue: I understand college applications, why they ask for your race because of affirmative action and stuff. I think that’s a double-­edged sword though. It’s—­I don’t think universities should have quotas. I think that that’s—­it may be—­it leaves—­if you’re looking for quota of X amount of Black people, but you’re not looking for X amount of Asian people, or Mexican people, or women, or White—­if you’re trying to fill these boxes with a certain amount of people, you’re putting them into a box, where—­you want to have people all different races and ethnicities bringing these different cultural backgrounds to your school. And that’s a positive thing. But I think trying to fill a certain number of people in one area is putting other people at a disadvantage.

Both Callie and Elena saw the good and the bad of these kinds of policies; they are a “double-­edged sword.” But both of these responses represent the fact that a sizable proportion of White millennials believes that race-­ conscious admissions policies amount to a simple matter of putting warm, unqualified bodies in reserved spots. Our respondents suggested that they are ambivalent, but when it came down to it, they believed that these policies produce perverse incentives and unfair disadvantages for Whites.

In summary, the responses we presented in this chapter illustrate how a person’s perceptions of a policy shape their preference for it. By contrasting stop-­and-­frisk with affirmative action and race-­conscious admissions policies, we see that White millennials have a very low tolerance for openly discriminatory public policies. In the case of stop-­and-­frisk, many of our respondents suggested that even though a policy may not be intentionally racist, if the outcome of the policy is perniciously discriminatory, then

Racialized Policy Preferences  /  171

they are not willing to support it. This is a step in the right direction as far as race politics goes. On the other hand, we also found that a majority of our respondents were not willing to support a set of policies aimed at eliminating racial disparities or increasing diversity, the latter of which they claimed to highly value. For our respondents, diversity is largely something that benefits them as Whites (see also Berrey 2015a; Smith and Mayorga-­ Gallo 2017). When they believe that diversity will lead to a disadvantage for them or members of their group, they throw diversity by the wayside. In the end they see meritocracy, fairness, rigor, and qualification as mutually exclusive from diversity, and when they weigh the first four things against the latter, diversity always loses and so do the policies that seek to create diverse educational and work settings. Overall, we see the ways in which the countervailing forces are entangled and reinforce one another, leading to racial stasis. We found that most of our respondents viewed race-­conscious admissions policies and affirmative action through the outdated, outmoded, and illegal frame of a racial quota system, despite the fact that all of our respon­ dents were born well after the landmark Supreme Court decision that declared quotas unconstitutional. This shows that even though birth cohorts are likely to have their own ideas, standards, norms, and values that are shaped by a specific temporal context, they are still influenced by previous birth cohorts and generations. Millennials are largely raised by boomers, who did live during a time when racial quotas were legal and endorsed by various levels of government in the effort to equitably dispense opportunities. In a way, what we find “makes sense,” but it is disturbing that the erroneous notion that affirmative action is a quota system has yet to be corrected. What’s more, we see that those who were willing to support stop-­ and-­frisk and racial profiling in policing, more generally, relied on inaccurate but incredibly prevalent stereotypes of Black criminality as well as tropes that undergird Islamophobia. Taken together, a few things stand out to us that will help us move forward in more accurately depicting the shape of White Americans’ racial attitudes in the twenty-­first century. First, young White people are not necessarily using the same language of racial animus or logic that structures older White Americans’ attitudes. Racial resentment, both an important theory and a measure of contemporary racial attitudes, is based on a reservoir of negative racial stereotypes and the belief that Blacks aren’t holding up their end of the deal (Henry and Sears 2002; Kinder and Sanders 1996; Sears, Sidanius, and Bobo 2000), but that isn’t the language we heard among our respondents. They did not bring up the notion that Blacks have not

172 / Chapter Five

held up their end of the deal to attain the American dream or suggest that Blacks refuse to work hard like other (historical) minority groups. They did not even suggest that Blacks aren’t deserving of opportunities presented by affirmative-­action policies. Instead, it seems that they believe that we have moved past needing to ensure racial fairness. To be sure, this sentiment is a direct descendant of racial resentment, but here we see a more contemporary expression of racial attitudes, rooted in color-­blind ideology. Research in sociology and psychology show that an increasing number of Whites’ racial attitudes can best be understood as color-­blind (Apfelbaum, Sommers, and Norton 2008; Bonilla-­Silva 2014; Carr 1997; Mueller 2017; Neville et al. 2000). These scholars demonstrate that there is either a lack of awareness or a lack of acknowledgment of structural racism (Mills 2007). White Americans know negative racial stereotypes, but most deplore them (even though they use them). Young Whites, especially, do not seem to foster animosity toward Blacks. Instead, White millennials’ racial attitudes are based on the notion that race and racism no longer uniquely structure Americans’ life chances. The second issue that surfaced here concerns how people respond or react, emotionally, to what they see around them. Although there is a great deal of heterogeneity on this matter, we found that a number of our respondents had low levels of empathy for racial and ethnic minorities; this was displayed most in this chapter, where people recognized that stop-­ and-­frisk may lead to racial profiling and increased harassment for people in specific groups (i.e., Black men, Muslims), but they valued their own safety over the civil rights of others. Our results corroborate extant literature, which reveals that Whites, especially young Whites, may be characterized as racially apathetic, or simply indifferent to existing racial inequities (Forman 2004). Finally, we would like to note that traditionally, political science research on racial attitudes focuses on Whites’ attitudes about Blacks, as the prototypical minority group. Considering that a great deal of this research was developed when African Americans were the largest racial minority group in the United States, this makes perfect sense. But, here we see that while Black Americans do still come to the forefront of White Americans’ racial imag­ ination, there have been significant demographic shifts in the American population over the past two decades, not to mention a change in global affairs, that have made a different set of groups more salient in the minds of Americans, across the racial spectrum. Here, we find that our respon­ dents had specific considerations of Asian Americans, people from the Middle East, Muslims, and Latinx communities in addition to Black Americans

Racialized Policy Preferences  /  173

when making calculations about their racialized policy preferences. We take that as a sign that we should begin considering a wider array of ethno-­racial groups when assessing White Americans’ racial attitudes. In all, it is the dueling set of values, principles, and policy preferences that have surfaced thus far in this book that lead us to argue that we are more likely to see racial stasis in the United States than to see increasingly progressive racial attitudes. As White Americans’ racial grammar evolves, it also becomes clearer to us that we need to develop a new measure. Perhaps this measure includes aspects of racial resentment (that attitude is likely to be around for quite a while), but we also need it to be more nuanced, holistic, and able to capture both knowledge around issues of racial inequity and emotional responses to our racial reality. Specifically, we ought to capture whether people are aware of institutional racism and racial privilege, the extent to which people feel empathy for racial minorities and racial injustice, and, perhaps, feelings of White guilt. Lastly, we may want to think about how to tap notions of racial politics that extend beyond how Whites feel about Black Americans only. Using the findings that have surfaced thus far as a backdrop, we tackle the challenge of developing a new measure of White Americans’ contemporary racial attitudes in part 3.

Six

New Attitudes, New Measures

Researchers have been concerned with how to measure Whites’ racial attitudes for quite some time. Over time, scholars change the questions they ask of the American population for a myriad of reasons, but ultimately they throw out old measures and develop new ones as acceptable expressions of racism evolve. As we mentioned, scholars used to ask questions about social distance and biological inferiority, but at least two things happened: First, a growing number of people learned that admitting a belief that Blacks are inherently inferior was out of vogue and taboo; their publicly displayed attitudes “changed” due to an increasing pressure of social desirability. Second, members from both older and newer generational cohorts began to rely on new language and a new dominant racial ideology to explain ongoing racial disparities. When the door marked “Jim Crow Racism” closed, the “symbolic racism” window opened, and scholars followed up with new measures. Those who have put forth the theories and measures to tap into symbolic racism, including the oft-­employed racial resentment scale, posit that Whites’ racial attitudes in the era after the civil rights movement are best understood as a blend of traditional American values and anti-­Black affect (Kinder and Sanders 1996; Sears and Henry 2003; Henry and Sears 2002; Sears and Kinder 1970; Kinder and Sears 1981). By the late 1970s, a new set of attitudes had surfaced, and they were based on the premise that Blacks have not held up their end of the deal presented by a series of landmark acts of Congress and Supreme Court decisions, including the two Brown v. Board of Education decisions, the Voting Rights Act of 1965, and a loose set of affirmative-­action policies. Now, over thirty years since the first studies of  Whites’ post–­civil rights ra­ cial attitudes, we question whether Kinder and Sears’s (1981, 416) original

178 / Chapter Six

assertion that “symbolic racism may be, politically, the most potent vehicle for racial prejudice today” still holds. Some scholars might argue that we are premature in looking for different ways of asking White Americans how and what they think about race and racism in contemporary American politics. If any of our readers fear that we might be attempting to “reinvent the wheel,” we would say that we are in good company. As millennials ourselves, we recognize that there is a disconnect between some of the measures of racial prejudice most frequently used in political science scholarship and how our families, our colleagues, and especially our students talk (or don’t talk) about race these days. To be sure, the preceding chapters have already provided evidence that the nature and structure of contemporary racial attitudes have yet again evolved. With that in mind, how can we best conceptualize and more accurately capture contemporary expressions of racial attitudes in the twenty-­first century with quantitative measures? We begin to answer that question in this chapter. There are some things that we must keep in mind. First, White Americans are far more likely to use color-­blind language in discussing race (Bonilla-­ Silva and Lewis 1999; Bonilla-­Silva 2014; Forman and Lewis 2015; Carr 1997; Mueller 2017; Apfelbaum, Sommers, and Norton 2008). Second, most folks are sensitive to the boundaries of acceptable language in public discourse, and today’s standards of social desirability call for Americans to not mention, notice, or discuss race. The questions posed by the racial resentment scale, in some ways, represent an affront to what is deemed appropriate talk around issues of race (Apfelbaum, Sommers, and Norton 2008). The racial grammar of Americans’ race talk is evolving; if we aim to understand White Americans’ attitudinal structure, we should utilize measures that more adroitly navigate the language of race that people are likely to employ. Additionally, we are moving farther away from the different historical points of reference captured in the scale that political scientists use so frequently, and considerations of groups other than Black Americans are shaping Whites’ racial sentiments; we need to build a set of questions that are abreast of these changes. What’s more, it has always been the case that racism has both affective and cognitive components, though scholars have tended to focus primarily on the latter; if we are going to capture the dynamics of all of the countervailing forces, we will need to capture people’s knowledge about and emotional responses to contemporary racial dynamics. Fortunately for us, we don’t have to fully reinvent the wheel, in part because scholars from across an array of disciplines have been thinking about this same question for quite some time. New scales have been developed

New Attitudes, New Measures  /  179

to address dimensions of racial attitudes that more closely mimic White Americans’ racial language and logic, but these measures, to the best of our knowledge, have yet to be analyzed and compared to older measures or to each other. We examine three new scales that, together, speak to the issues of concern that we laid out above. We assess White Americans’ responses to the Color-­Blind Racial Attitudes Scale (CoBRAS), the Psychosocial Costs of Racism to Whites (PCRW) scale, and the Explicit Racial Resentment (EXR) scale. Because we use nationally representative data, these data also serve to amplify the sentiments we gathered in the previous two chapters based on interviews with millennials; or in other words, the large sample that we examine, which includes Whites across generational cohorts, underscores the fact that observing millennials is important because they serve as a bellwether, helping us to understand how more and more White Americans, generally speaking, are thinking and talking about race and racism. Ultimately, we show that there are several different (contradictory and countervailing) components to Whites’ contemporary racial attitudes that ought to be captured by any accurate, updated, nuanced, and holistic measure.

Color-­Blind Racial Attitudes Scale While the racial resentment scale requires respondents to recall a history of ethnic European immigrants and blame Blacks themselves for the position they are in, research, including ours, reveals that White Americans are less likely to rely on historical depictions of “good” immigrants or even use race or racism as explanations for disparities between racial groups. Instead, White Americans focus on the present and are more likely to use class or socioeconomic status, interests, or “comfort” to explain ongoing racial phenomena such as economic disparities between groups and racial segregation in schools and housing. We note, as many others do, an increasing desire among Whites for political correctness and for avoidance of discussions surrounding racial issues (Jackson 2008). This avoidance largely stems from the perception that talking about race these days is “fraught with the risks of misunderstanding and social sanction”; the social norms of contemporary society guide Whites’ motivation to “avoid the appearance of prejudice” (Apfelbaum, Sommers, and Norton 2008, 918). With this in mind, an increasing amount of scholarship suggests that the most prevalent way White Americans understand and express their racial attitudes is through a color-­blind lens. Our evidence thus far encourages us to join the emerging consensus among scholars studying American racial ideology, who assert that color-­blindness is today’s “culturally sensitive approach to intergroup

180 / Chapter Six

contact” (Apfelbaum, Sommers, and Norton 2008, 930; see also Bonilla-­ Silva 2014; Bonilla-­Silva, Lewis, and Embrick 2004; Carr 1997; Forman 2004; Frankenberg 1993). Scholars have identified three manifestations of color-­blind racial attitudes: “(a) viewing race as an invisible characteristic (e.g., refusing to notice racial group membership for fear of appearing prejudiced); (b) viewing race as a taboo topic (e.g., adhering to a perceived norm that talking about or referring to racial designators is impolite); and (c) viewing social life as a nexus of individual relations (e.g., individual circumstance, and not intergroup relations, mostly account for one’s social life)” (Neville et al. 2004, 60; Schofield 1986). Sociologist Bonilla-­Silva (2014), with the use of large-­n qualitative data, determined that the main semantic frames of color-­blind ideology are a denial of the centrality of racial discrimination, abstract liberalism (e.g., “I believe in equal opportunity, and that’s why I oppose affirmative action”), naturalization of racial matters (e.g., “Residential segregation exists because, naturally, minorities prefer to live together”), and employment of a cultural explanation for minorities’ standings (e.g., “Blacks are poor because they are not in the habit of showing up for work”). We saw each of these at play in the previous chapters. In all, color-­blind racial ideology suggests that all groups are working on a level playing field and racial phenomena can best be explained by race-­ neutral factors. A related set of ideas comes in the form of diversity ide­ ology. Diversity ideology requires people to only warmly accept superficial differences between groups rather than actually do anything to assure that different kinds of people have access to political and social privileges that have historically been skewed toward those who are identified as White (Mayorga-­Gallo 2019; Warikoo and de Novais 2014). These two ideas—­ color-­blind ideology and diversity ideology—­speak directly to two of the countervailing forces: the moving walkway of racism and the diversity dilemma, respectively. When we examine the themes that arise in theories of color-­blind ideology and in related conceptualizations like diversity ideology, it becomes clear that frequently used measures of racial attitudes may not accurately gauge how White people think about race in twenty-­first-­century America. In an attempt to accurately quantify these sentiments, educational psychologist Helen Neville and her colleagues (2000) developed the Color-­Blind Racial Attitudes Scale (CoBRAS). This scale, which contains twenty items, assesses the cognitive aspects of color-­blind racial attitudes. Neville et al. (2000) determined that these twenty items could be best described by three

New Attitudes, New Measures  /  181

latent factors or dimensions. Neville and her colleagues label the first dimension “Racial Privilege.” We measure this construct using the following questions:1 ·· White people in the US have certain advantages because of the color of their skin. ·· Race is very important in determining who is successful and who is not. ·· Race plays an important role in who gets sent to prison. ·· Race plays a major role in the type of social services (such as type of health care or day care) that people receive in the US. ·· Racial and ethnic minorities do not have the same opportunities as White people in the US. ·· Everyone who works hard, no matter what race they are, has an equal chance to become rich.

These measures reveal respondents’ level of awareness regarding White privilege. While the survey asks about “White people,” it should be noted that other racial groups are not specifically mentioned—­particularly “Blacks” or “African Americans.” Here, respondents are directly asked about the extent to which they feel one’s racial-­group membership may influence one’s access to certain privileges or one’s chances of being disadvantaged by major political institutions in the United States. Considering what we learned in chapters 4 and 5, we expect respondents to disagree with the great majority of these questions, given respondents’ reliance on other, race-­neutral explanations of racial disparities. The second dimension of the CoBRAS is called “Institutional Discrimination”; this dimension measures the extent to which respondents are aware of the implications of institutional forms of racial discrimination. Questions used to measure this aspect of the CoBRAS include ·· Social policies, such as affirmative action, discriminate unfairly against White people. ·· White people in the US are discriminated against because of the color of their skin. ·· English should be the only official language in the US. ·· Due to racial discrimination, programs such as affirmative action are necessary to help create equality. ·· Racial and ethnic minorities in the US have certain advantages because of the color of their skin.

182 / Chapter Six

This measure mimics Bonilla-­Silva’s abstract liberalism racial frame, which allows respondents to frame racial matters in terms of political and economic liberalism. As a result, respondents can appear to be taking the moral high ground when they oppose practical remedies for racial disparities. Again, based on our conversations in the previous chapters, we suspect that respondents will report that policies such as affirmative action are unnecessary and that, further, they will suggest that Whites are at a disadvantage. What’s more, research shows that Whites are increasingly focused on the implications that race has in their own lives, where they may see their race as a disadvantage and believe that increased equality for racial minorities is directly linked to anti-­White bias (Eibach and Ehrlinger 2006; Norton and Sommers 2011). The third dimension of the Color-­Blind Racial Attitudes Scale is called “Blatant Racial Issues.” This dimension is measured using questions that ask about respondents’ acknowledgment of racism, generally speaking. Specifically, people are asked questions such as ·· Racial problems in the US are rare, isolated situations. ·· Talking about racial issues causes unnecessary tension. ·· Racism is a major problem in the US.

Respondents’ agreement with these questions provides an indication of how aware they are of general, pervasive racial discrimination. We feel safe in pre­­ dicting overwhelming agreement on the first two questions and less agree­­ ment on the question of whether racism is a major problem, particularly given that our interview respondents revealed that they have a narrow understanding of racism as being overt, racist behaviors and attitudes. Altogether, each of the three dimensions of the CoBRAS taps into the sentiments expressed by our interview respondents in part 2.

The Psychosocial Costs of Racism to Whites The second scale we examine is the Psychosocial Costs of Racism to Whites (PCRW) (Poteat and Spanierman 2008; Spanierman and Heppner 2004; Spanierman et al. 2006). In any racialized social system, some citizens and denizens will gain benefits, rights, and privileges at the expense of others. While no single volume could ever offer a full moral accounting of the costs and benefits of racism in the United States, there is a mountain of evidence that suggests that Whites, generally speaking, receive large structural benefits at the expense of members of other racial groups. As we have previously

New Attitudes, New Measures  /  183

shown, in comparison to racial minorities, White Americans have greater access to most of society’s resources. Furthermore, although White racial privilege normally goes unacknowledged by most Whites, some are certainly able to recognize the benefits bestowed on them by a society that describes their lives, tastes, interests, physical attributes, values, and the like as the norm. However, some might argue that while racism has many benefits for Whites, it may also come with costs, though they are qualitatively different from the costs borne by people of color. Psychologists Lisa Spanierman and Mary Heppner (2004) begin with the premise that racism can also negatively affect some members of the dominant group, and argue that there could be a variety of ways in which racism may affect Whites. For example, imagine three different people, all of whom are White. The first is someone who is actively anti-­racist: they believe that systemic inequalities and structural racism serve to benefit Whites at the expense of other groups (in particular, Blacks). The second is someone who believes that America has entered a post-­racial era. For this second person, racial identity is a thing of the past and discussing race relations only serves to divide some groups against one another. Finally, the third individual could be best characterized as a rugged individualist. For this person, minorities need to stop using racism as an excuse and simply need to work harder if they want to get ahead. All of these hypothetical individuals are well represented by our forty-­three interview respondents, of which many looked most like the second hypothetical person. One could imagine how each of these people might react if they were presented with evidence that reveals that Black and White Americans experience very different Americas when it comes to matters of education, housing, health, criminal justice, wealth, the quality of the environment, and the like. The first person, whom we label an “anti-­racist,” may be negatively affected, both cognitively and emotionally. Not only might this person come to see that they are complicit in perpetuating racial inequalities, this realization might also lead them to feel guilt and shame in relation to their racial identity, ultimately motivating them to work to combat racism. Similarly, our second person might begin to question whether “color-­blindness” is mechanism to help things change for the better or a symptom of a larger problem. Finally, the rugged individualist might experience some cognitive dissonance and become angry over the way many astute scholars describe the state of racial inequity. Thus, Whites might react to racism in a number of ways, and the costs of encountering racial facts could be emotional, cognitive, or even behavioral. No matter what precisely the costs are, we can borrow the language of previous scholars and describe them as psychosocial;

184 / Chapter Six

they predominantly affect how Whites think of themselves vis-­à-­vis other racial groups. These psychosocial attitudes are central to belief systems and may have profound effects on Whites’ political attitudes, their political behavior, and their preferences on policies that aim to ameliorate (or exacerbate) racial inequalities. Spanierman and her colleagues (Poteat and Spanierman 2008; Spanierman and Heppner 2004; Spanierman et al. 2006) identify three key ways Whites experience the negative effects of racism. These “psychosocial costs of racism to Whites” (PCRW) are measured using a set of sixteen Likert-­type survey items, which ultimately tap into two dimensions, the first of which has two subscales. One of the subscales of the first dimension is measured by agreement with the following statements: ·· I am angry that racism exists. ·· I become sad when I think about racial injustice. ·· It disturbs me when people express racist views. ·· When I hear about acts of racial violence, I become angry or depressed. ·· Racism is dehumanizing to people of all races, including Whites. ·· I feel helpless about not being able to eliminate racism.

This subscale is called and captures “White Empathetic Reactions toward Racism.” We have dueling expectations here. On the one hand, given that White Americans perceive being called a racist as not just an insult but also a moral condemnation, we expect White Americans to record fairly high responses to these questions. On the other hand, we found that, at times, our interview respondents expressed little racial empathy; this finding mimics data from the American Freshman study, which revealed that members of the millennial generation are less concerned than older Whites with being helpful in promoting racial understanding (Twenge, Campbell, and Freeman 2012). Needless to say, the data we rely on in this chapter should provide concrete evidence for one hypothesis or the other (or both). A second subscale, which taps “White Guilt,” includes the following statements: ·· Being White makes me feel personally responsible for racism. ·· I never feel ashamed about being White. ·· Sometimes I feel guilty about being White. ·· I am afraid that I abuse my power and privilege as a White person. ·· I feel good about being White.

New Attitudes, New Measures  /  185

This subscale is valuable for our purposes for a number of reasons. First, an increasing amount of research shows that Whites are aware of their racial identity and Whiteness (Wong and Cho 2005; McDermott and Samson 2005; Jardina 2019), and these survey items allow us to measure the extent to which people are aware of their racial identity. Second, these questions also allow us to assess respondents’ recognition of White racial privilege and their reaction to it. The final dimension, “White Fear of Others,” measures respondents’ comfort around people of other races. This subscale includes the following statements: ·· I often find myself fearful of people of other races. ·· I am distrustful of people of other races. ·· I have very few friends of other races. ·· I feel safe in most neighborhoods, regardless of the racial composition.

We suspect that most people will respond in socially desirable ways to these questions, on average, but any variation on these questions will be fruitful given the high price of social desirability presented by answering in the affirmative to the first three questions in this group.

(Explicit) Racial Resentment The statements below compose the four-­item battery of the racial resentment scale: ·· The Irish, Italians, Jews, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors. ·· Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class. ·· Over the past few years, Blacks have gotten less than they deserve. ·· It’s really a matter of some people not trying hard enough; if Blacks would only try harder, they could be just as well off as Whites.

There has been a great deal of debate about what these questions actually measure; principally, some scholars have misgivings about whether the battery simply and primarily taps into political ideology (Carmines, Sniderman, and Easter 2011). Given how White respondents’ answers to these four questions predict their attitudes on racial as well as nonracial policies,

186 / Chapter Six

Wilson and Davis (2011) set out to develop a revised form of the racial resentment scale (RRS), which they call Explicit Racial Resentment (EXR). This measure was developed in order to better separate the conservative values, most notably a sense of “rugged individualism,” from the anti-­Black component of racial resentment. In developing their scale, Wilson and Davis note that their questions “mainly differ from past resentment measures in their explicit connection between the source of the resentful feelings and the targeted racial group” (2011, 121). Instead of asking respondents about whether they believe “Blacks have gotten less than they deserve,” the Explicit Racial Resentment scale lives up to its name—­explicitly asking respondents how they feel about Blacks: ·· I resent any special considerations that African Americans receive because it’s unfair to other Americans. ·· I don’t understand why race is any different from what other people have to deal with. ·· For African Americans to succeed they need to stop using racism and slavery as excuses. ·· Special considerations for African Americans place me at an unfair disadvantage because I have done nothing to harm them. ·· African Americans bring up race only when they need to make an excuse for their failure.

This measure is designed to assess whether a respondent believes Blacks are receiving undue attention for their group’s problems. The first item explicitly asks about a respondent’s resentment: whether they believe Blacks receive undue attention. Other questions measure both anti-­Black affect and beliefs in individualism but do so directly. For example, the third question measures a combination of two beliefs: first, that the respondent believes that Blacks use racism as an excuse, and, second, that the respondent does not believe in structural racism in the United States.

Measuring CoBRAS, PCRW, and EXR We expect these measures to provide some additional leverage in accurately gauging racial attitudes in the twenty-­first century because they introduce a wider array of cognitive and affective sentiments than the most commonly used measures, they mirror the racial language and logic that Whites are employing more and more, and they could help us to more precisely pinpoint attitudinal areas of progressivism and stagnation.

New Attitudes, New Measures  /  187

The descriptive analyses we present below rely on data from the 2014 Cooperative Congressional Election Study, an online survey conducted by YouGov during the 2014 midterm election. Our sample of 1,000 respon­ dents contained 743 who identified as White, 154 of whom were categorized as millennials (being born in or since 1980). All questions used a six-­point agree/disagree Likert scale that did not have a “neither agree nor disagree” option; thus, the respondents were forced to indicate how they felt one way or another. We collapsed the responses to create dichotomous variables—­ thus separating all of those who agreed to any degree from those who dis­ agreed to any degree—­in order to do a comparative assessment of White millennials and their predecessors on nearly three dozen racial-­attitudes questions. CoBRAS We begin with the scale that aims to mimic the major themes of color-­blind racial ideology. First, we turn to the component of CoBRAS that measures the extent to which people are willing and able to acknowledge White racial privilege, shown in table 6.1. There appear to be some tensions between older Whites’ and millennial Whites’ racial belief systems. For instance, fewer than 50 percent of our White millennial respondents believed that “racial and ethnic minorities do not have the same opportunities” as Whites, but they were more likely to agree to this than their elder counterparts. This provides some promising evidence that perhaps millennials are more progressive on some of the basic racial issues. However, millennial Whites are less likely than older Whites to believe that “race is very important in determining who is successful and who is not” or that “race plays a major role in the type of social services that people receive in the US.” White millennials are no more likely than their older counterparts to acknowledge that being White is associated with receiving advantages, and they are just as likely to believe that race is not an important factor in shaping certain kinds of opportunities for minorities. If one were looking toward White millennials with the hope of finding a groundswell of anti-­racist activists, these responses would cause great concern. White millennials are no more committed to basic anti-­racist principles than their predecessors, and our respondents’ answers suggest that White Americans across generational cohorts lack critical knowledge with regard to basic facts: upward of three-­ quarters of each group expressed the belief that “everyone who works hard, no matter what race they are, has an equal chance to become rich.” As we have pointed out before, the racial gap in wealth is not explained by racial

188 / Chapter Six Table 6.1.  CoBRAS Racial Privilege Millennials (% Agree) White people in the US have certain advantages because of the color of their skin. Race is very important in determining who is successful and who is not. Race plays an important role in who gets sent to prison. Race plays a major role in the type of social services that people receive in the US. Racial and ethnic minorities do not have the same opportunities as White people in the US. Everyone who works hard, no matter what race they are, has an equal chance to become rich. White people are more to blame for racial discrimination than racial and ethnic minorities. N

Older Whites (% Agree)

54.55

48.62

18.3

24.44

39.87 42.86

37.74 56.43

37.91

29.91

73.38

74.61

27.92

27.76

155

385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.85; bold text indicates proportions significantly different across groups (p < 0.05)

Table 6.2.  CoBRAS Institutional Discrimination Millennials (% Agree) Social policies, such as affirmative action, discriminate unfairly against White people. White people in the US are discriminated against because of the color of their skin. English should be the only official language in the US. Due to racial discrimination, programs such as affirmative action are necessary to help create equality. Racial and ethnic minorities in the US have certain advantages because of the color of their skin. It is important that people begin to think of themselves as American and not African American, Mexican American, or Italian American. N

Older Whites (% Agree)

69.48

75.09

42.21

56.42

75.82 40.91

86.43 35.36

60.13

64.3

64.71

86.72

155

385

Data: 2014 Cooperative Congressional Election Study (CCES) Note: Cronbach’s α = 0.85. Bold text indicates proportions significantly different across groups (p < 0.05).

differences in culture or in orientation to hard work, education, or income (Darity 2011; Hamilton et al. 2015). Responses to the second dimension of the CoBRAS scale are illustrated in table 6.2; these charts show the degree to which respondents recognized and acknowledged institutional discrimination. Here, millennial Whites are

New Attitudes, New Measures  /  189

mostly identical to the older respondents. However, older Whites seem to be more supportive of some of the more nationally focused questions. For example, not only are older White respondents more likely to believe that English should be the official language of the United States, they also think that it is better if people start thinking of themselves as “American” as opposed to any “hyphenated” alternatives (e.g., “African American”). In many ways the latter results mirror our qualitative data. During the face-­to-­face interviews, we asked respondents questions similar to these, and nearly half of our White millennial respondents suggested that identifying as a “hyphenated American” is divisive because America is a “melting pot,” and even those who did not support the idea that people should start thinking of themselves as solely “American” still hoped for a time when people would feel that they could identify as “just American.” Those in the latter group suggested that when everyone identified as “just American,” that would signal the kind of societal progress for which they hoped. Our survey results also reveal that 43 percent of White millennials and 56 percent of older Whites believe that Whites “are discriminated against because of the color of their skin,” a difference that is unlikely to be due to chance (z = 3.14, p < 0.05). This finding reveals something new and important. Our results on millennials corroborate a recent MTV poll that showed that 48 percent of those between the ages of fourteen and twenty-­four believe that discrimination against Whites is as big a problem as discrimination against racial minorities (MTV Strategic Insights and David Binder Research 2014). A 2012 poll from the Public Religion Research Institute found that 58 percent of their White millennial respondents believed the same thing (Jones and Cox 2012). Our results are on the lower end, but what is surprising is the proportion of older Whites who believe this to be true, particularly since a large number of them lived during a time when overt racial discrimination was common and, well, overt. Here, we see that the language and logic of contemporary racial attitudes are pervasive among a large group of White Americans, but political science scholarship has fallen behind on capturing this shift. Finally, table 6.3 reveals that White millennials and older Whites are equally as likely to agree with each of the three questions we used to mea­ sure their awareness of the existence of racism. Nearly 69 percent of millennial Whites and 62 percent of older Whites believe that “racism is a major problem in the US,” and only about one in four believes that racial problems are “rare, isolated situations.” The two groups have statistically identical levels of agreement. These findings are a perfect illustration of countervailing forces. Even though nearly 70 percent of Whites believe that “racism

190 / Chapter Six Table 6.3.  CoBRAS Blatant Racism Millennials (% Agree) Racial problems in the US are rare, isolated situations. Talking about racial issues causes unnecessary tension. Racism is a major problem in the US. N

28.76 62.09 68.63 155

Older Whites (% Agree) 25.17 59.55 61.86 385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.60. All proportions are statistically indistinguishable across groups.

is a major problem in the US,” they also feel that talking about race causes unnecessary tension. It’s difficult to begin to solve a problem without talking about it. An examination of the CoBRAS scale underscores some of the ostensible contradictions we have outlined through the notion of countervailing forces. Specifically, an analysis of these questions reveals that even though the overwhelming majority of White Americans believe that racism is a problem, and nearly half recognize White privilege, less than half believe that race influences the opportunity structure for underrepresented minorities. Relatedly, even though respondents recognize the problem that racism presents to American society, they fail to see why some policies might be necessary to promote equality; our findings here mimic what we saw in the previous chapter. Needless to say, holding these kinds of attitudes simultaneously is the reason why we do not see much change in racial attitudes on the whole. To be sure, the simple (yet parsimonious) RRS does not allow us to see these incongruencies. Psychosocial Costs of Racism to Whites While the CoBRAS items help us to examine cognitive awareness of racial inequality and privilege, the PCRW scale tells us more about affective responses to the United States’ racial reality. Table 6.4 shows marked similarities between our two subsets of Whites regarding “Racial Empathy.” Recall that these questions concern whether Whites feel bad about racism and racial injustice. The only item in which there is a statistical difference between the two groups is the “racism is dehumanizing to all people, including Whites” item; younger Whites appear to be less likely to agree with this statement, though a substantial proportion of each group agrees (95 percent and 90 percent).

New Attitudes, New Measures  /  191

It should be noted that while an overwhelming majority of Whites report emotional reactions of sadness and anger when confronted with certain aspects of racism (acts of racial violence, racist views, etc.), there is still a substantial number of people who do not: 20 percent of respondents expressed that they become neither angry nor sad when they hear about racial violence, and nearly a quarter of White adults are not saddened by racial injustice. This is the variance that makes these newer scales helpful; they allow us to see if there are people who are cognitively blind and/or emotionally numb when it comes to issues of racial justice. The second dimension of the PCRW scale, “White Guilt”, has a number of items where White millennials expressed higher levels of remorse than their older peers. On two of the five items, millennials are more likely to agree. They are approximately twice as likely to agree with the statement “Sometimes I feel guilty about being White” (19 percent compared to 9 per­ cent), and over twice as likely to agree that “I am afraid that I abuse my power and privilege as a White person” (9.8 to 3.7 percent). Here, it seems that White millennials may be reporting the “guilt” that they may experience if they are conscious of their racial privileges, something that barely half of the millennials admitted to having (above in the CoBRAS scale). In terms of racial progress or identifying anti-­racist White millennials, these responses do not indicate that their cognitive components and their emotional reactions are necessarily in sync. These results are shown in table 6.5. The last component of the PCRW scale is presented in table 6.6. Here, we see similar patterns: younger Whites are less likely to agree that they find themselves “fearful of people of other races,” but are otherwise statistically

Table 6.4.  PCRW Racial Empathy Millennials (% Agree) I am angry that racism exists. I become sad when I think about racial injustice. It disturbs me when people express racist views. When I hear about acts of racial violence, I become angry or depressed. Racism is dehumanizing to people of all races, including Whites. I feel helpless about not being able to eliminate racism. N

Older Whites (% Agree)

88.31 75.32 82.47 77.12

88.72 71.68 85 78.62

90.2

94.83

54.25 155

53.47 385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.78; bold text indicates proportions significantly different across groups (p < 0.05)

192 / Chapter Six Table 6.5.  PCRW White Guilt Millennials (% Agree) I feel helpless about not being able to eliminate racism. Being White makes me feel personally responsible for racism. I never feel ashamed about being White. Sometimes I feel guilty about being White. I am afraid that I abuse my power and privilege as a White person. I feel good about being White. N

Older Whites (% Agree)

54.25 10.46 81.17 19.08 9.8

53.47 8.28 84.93 9.14 3.77

87.5 155

91.6 385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.73; bold text indicates proportions are significantly different (p < 0.05)

Table 6.6.  PCRW Fear of Other Groups Millennials (% Agree) I often find myself fearful of people of other races. I am distrustful of people of other races. I have very few friends of other races. I feel safe in most neighborhoods, regardless of the racial composition. N

Older Whites (% Agree)

19.61 17.11 29.61 59.74

26.68 19.69 34.93 58.15

155

385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.70; bold text indicates proportions are significantly different across groups (p < 0.05)

identical to older Whites. Above and beyond anything else, we find it critical to highlight the many similarities that arose between generational groups on so many questions. For those readers who were expecting large differences across our groups for these types of measures, our results must certainly be surprising, but again, as we have argued, values and ideas do get passed down from one generation to another. Explicit Racial Resentment Table 6.7 shows the percentage of each group that provided an affirmative answer for each of the statements in the Explicit Racial Resentment battery. In contrast to our findings for several of the other sets of questions we have examined thus far, younger Whites are less likely to agree with several premises posed by the EXR battery. They are less likely to express resentment for

New Attitudes, New Measures  /  193

special considerations Blacks may receive, for instance. Or, as another example, while two-­thirds of older Whites believe Black Americans only bring up race to excuse their failings, only about half of millennial Whites agree (a difference that is statistically different from zero). These differences are presumably a step in the right direction. On their face, these responses make younger Whites appear less resentful and, perhaps, more understanding of race in contemporary politics. Upon closer examination, however, the group differences on the other items seem to tell a different story. For example, White millennials are just as likely as their older counterparts to agree that race might be no different than “what other people have to deal with.” So, while younger Whites are less likely to explicitly say they resent African Americans for receiving undue benefits, there is evidence that indicates that some of the attitudes underlying their resentment might be missed. As an additional example, an overwhelming majority (76 percent) of White millennials agreed that “for African Americans to succeed, they need to stop using racism and slavery as excuses.” Recall that nearly the same proportion of Whites agreed that racism is a major problem.

Rather than rely on a measure that was developed well before the first millennial was born, we consider a wider set of measures and scales that mirror the way that an increasing number of White people are talking about race and racism. By putting almost three dozen questions into conversation with

Table 6.7.  Explicit Racial Resentment across Generations Millennials (% Agree) I don’t understand why race is any different from what other people have to deal with. I resent any special considerations that African Americans receive because it’s unfair to other Americans. For African Americans to succeed they need to stop using racism and slavery as excuses. Special considerations for African Americans place me at an unfair disadvantage because I have done nothing to harm them. African Americans bring up race only when they need to make an excuse for their failure. N

Older Whites (% Agree)

66.9

69.1

58.5

69.4

76

83.2

68.5

76

55.9

69.9

155

385

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Cronbach’s α = 0.86; bold text indicates proportions are significantly different (p < 0.05)

194 / Chapter Six

one another, we uncovered several important themes that will help us to build an updated, accurate, and nuanced measure of racial attitudes. First, our results reveal that color-­blindness is not unique to the millennial generation; color-­blind racial ideology is the dominant mode of racial thinking among White Americans. Second, our results show that it is important to ascertain both what people know or are willing to acknowledge about race and racism and how they feel about it. Third, we see that Whites’ racial attitudes are not “logical” in the commonsense use of the word. For instance, they are willing to acknowledge that racism is a problem but feel that Blacks use racism as an excuse, and assert that talking about racism is divisive. Nonetheless, any critical race scholar would suggest that it is this logic that serves to perpetuate White supremacy. We posit that the juxtaposition of these attitudes produces racial stasis. From these preliminary analyses, we see that the newer measures provide valuable insight into the structure of how White Americans, young and old, think about race and race relations in the United States. In the next chapter, we turn to a series of more rigorous quantitative analyses to estimate how exactly all of these items are related, assess the strength of their predictive values, and discern their role in shaping White Americans’ attitudes toward racialized policies and people.

Seven

The Structure, Nature, and Role of Twenty-­First-­Century Racial Attitudes

In the previous chapter, we outlined a series of newly developed, theory-­ driven measures aimed to capture contemporary expressions of racial animus and attitudes; rather than simply focus on notions of what Blacks have failed to do, these measures engage respondents’ attitudes and knowledge regarding institutional and structural racism, racial privilege, and Whiteness as well as feelings of racial guilt and empathy. These are some of the major components we noticed percolating in the way that Whites are talking about race matters these days. Though these newer scales and batteries were created independently, we believe that we can gain more traction in understanding contemporary racial attitudes by putting all of them—­new and old—­in conversation with one another. We do three sets of analyses in this chapter to gain a better understanding of the structure, nature, and role of Whites’ racial attitudes in the twenty-­first century, employing a sample of 743 Whites from the 2014 Cooperative Congressional Election Study. We should note that the measures we examined in chapter 6 were developed at different times (but much more recently than traditionally employed measures of racial attitudes) and by experts across an array of social-­science disciplines. To our knowledge, we are the first to examine and employ all of these measures simultaneously. As such, our first step is to determine whether the items we asked of our survey respondents actually fit the structure that Neville et al. (2000) hypothesized for the Color-­Blind Racial Attitudes Scale (CoBRAS) and Spanierman and Heppner (2004) determined for the Psychosocial Costs of Racism to Whites (PCRW) scale. We also assess the structure of the Explicit Racial Resentment (EXR) scale and the racial resentment scale (RRS) here. Our second step is another quality check of sorts. We examine the convergent validity and predictive value of these measures. Do these measures covary with related attitudes in expected

196 / Chapter Seven

directions? What exactly do these new measures offer us that previous mea­ sures do not? The final step we take is to examine these measures’ relationships with one another to determine whether we can find a more accurate model and representation of Whites’ racial attitudes as we see them play out today. We determine whether our adopted measures are tapping the same underlying construct. We find that the myriad of measures we use actually uncovers multiple dimensions of Whites’ racial attitudes, with each dimension providing greater detail of the structure of racial attitudes in the mass public.

Step 1: Replication We begin by determining whether the items we asked of our survey respondents actually fit the structure Neville et al. (2000) hypothesized for the Color-­Blind Racial Attitudes Scale and Spanierman and Heppner (2004) determined for the Psychosocial Costs of Racism to Whites scale; both sets of scholars asserted that the CoBRAS inventory and the PCRW scale are each supposed to measure three dimensions of racial attitudes, which are summarized in table 7.1. We asked all of our respondents nearly all of the questions that constitute the CoBRAS, the PCRW scale, the EXR scale, and the RRS. All questions used a six-­point agree/disagree Likert scale that did not have a “neither agree nor disagree” option; thus, the respondents were forced to indicate how they felt one way or another. All of the measurement models we present will be shown in tables with the factor loadings and several goodness-­ of-­fit statistics also presented. Given that we have a large sample size from the Cooperative Congressional Election Study, the traditional chi-­squared goodness-­of-­fit measure will be biased upward.1 As such, we present other goodness-­of-­fit measures that are less sensitive to sample size. All estimations were carried out in Mplus version 7.3, and suggested modifications to the basic model were made to improve model fit. Any additional pa­ rameters, including estimated error covariance terms, were left out of the tables for simplicity. In terms of predicting a respondent’s location on each of the dimensions, elaborated on below, the factor analysis automatically scales the latent variables to have means close to zero and variances close to one. The results from the confirmatory factor analysis (CFA) for the CoBRAS are presented in table 7.2. The factor loadings can be interpreted similarly to basic regression coefficients. For example, a one-­unit change in the latent variable “Unawareness of Racial Privilege” would lead us to expect the

Table 7.1.  Predicted Subscales for CoBRAS and PCRW CoBRAS

PCRW

1. Acknowledge Racial Privilege 2. Acknowledge Institutional Discrimination 3. Understanding of Blatant Racial Issues

1. Empathetic Reactions toward Racism 2. White Guilt 3. White Fear of Others

Table 7.2.  Confirmatory Factor Analysis of CoBRAS Unawareness of Racial Privilege White people in the US have certain advantages because of the color of their skin. Race is very important in determining who is successful and who is not. Race plays an important role in who gets sent to prison. Race plays a major role in the type of social services (such as type of health care or day care) that people receive in the US. Racial and ethnic minorities do not have the same opportunities as White people in the US. Racial and ethnic minorities in the US have certain advantages because of the color of their skin. Everyone who works hard, no matter what race they are, has an equal chance to become rich. White people are more to blame for racial discrimination than racial and ethnic minorities. Unawareness of Blatant Racial Issues Racial problems in the US are rare, isolated situations. Talking about racial issues causes unnecessary tension. Racism is a major problem in the US. Unawareness of Institutional Discrimination Social policies, such as affirmative action, discriminate unfairly against White people. White people in the US are discriminated against because of the color of their skin. English should be the only official language in the US. Due to racial discrimination, programs such as affirmative action are necessary to help create equality. It is important that people begin to think of themselves as American and not African American, Mexican American, or Italian American. Goodness of Fit Statistics CFI TLI RMSEA χ2 df P-­value N Data: 2014 Cooperative Congressional Election Study (CCES) Notes: * denotes loadings fixed to 1.00

Loading 1.00* −1.19 −1.00 −1.14 −0.45 0.92 −1.11 0.94

Loading 1.00* −0.14 −1.27 Loading 1.00* 1.36 0.84 −0.19 0.76 Estimate 0.98 0.97 0.04 214.5 82 < 0.05 743

198 / Chapter Seven

agreement with the statement “Race plays an important role in who gets sent to prison” would decrease by one point. As indicated by the confirmatory fit index (CFI), the Tucker-­Lewis Index (TLI), and the root mean squared error of approximation (RMSEA), the data fit the three-­factor model very well. All of the factor loadings are statistically different from zero at conventional levels (p < 0.05). In fact, even the smallest factor loading—­the relationship between whether one agrees with “Talking about racial issues causes unnecessary tension” and the latent measure of “Unawareness of Blatant Racial Issues”—­would have been observed by chance, if the true value were zero, just four times out of a hundred (p < 0.04). Similar goodness-­of-­fit statistics appear for our analysis of the PCRW items, presented in table 7.3. Here, we can see that the data fit very well to the hypothesized three-­factor solution; both the CFI and TLI are above 0.95, and our RMSEA is close to 0.05. As before, the first factor loading for each subscale is fixed to 1.00 to allow for identification of the model; again, all factor loadings are statistically different from zero (p < 0.01 in all cases). Looking at the factor loadings for the “empathy” dimension, a one-­unit change in the latent variable would be accompanied by a higher level of agreement with each of the observed variables, most strongly associated with the item that asks whether the respondent becomes sad when thinking about racial injustice: a one-­unit increase on the latent dimension of empathy would be met with a 1.35-­point increase in the respondent’s level of agreement. It is less strongly related with the other items, though these factor loadings are still quite large by any standard. Even the smallest loading on empathy estimates a 0.72-­point change in the survey response based on a one-­unit shift in a respondent’s position on the latent empathy dimension. In terms of the factor loadings for “White guilt,” the three items that deal explicitly with personal responsibility are positively related to the latent dimension: a one-­unit increase in White guilt is met with a one-­point change in feeling “personally responsible for racism” as well as feeling “guilty about being White”; a smaller but similar effect of 0.93 manifests in the item where a respondent is afraid that they abuse their power and White privilege. As expected, the two items that focus on the opposite of  White guilt, pride and general contentment with being White, are negatively related to the latent dimension. A one-­unit change in White guilt sees agreement with those two questions drop by about one-­half of one point on a six-­point scale. In addition to replicating CoBRAS and the PCRW, we examine the relationship between racial resentment and explicit racial resentment. As mentioned in chapter 3, Wilson and Davis suggest that the original set of racial

Twenty-First-Century Racial Attitudes  /  199 Table 7.3.  Confirmatory Factor Analysis of PCRW White Empathetic Reactions toward Racism I am angry that racism exists. I become sad when I think about racial injustice. It disturbs me when people express racist views. When I hear about acts of racial violence, I become angry or depressed. Racism is dehumanizing to people of all races, including Whites. I feel helpless about not being able to eliminate racism. White Fear of Others I often find myself fearful of people of other races. I am distrustful of people of other races. I have very few friends of other races. I feel safe in most neighborhoods, regardless of the racial composition.

Loading 1.00* 1.35 0.89 1.11 0.72 1.25 Loading 1.00* 0.98 0.61 −0.80

White Guilt

Loading

Being White makes me feel personally responsible for racism. I never feel ashamed about being White. Sometimes I feel guilty about being White. I am afraid that I abuse my power and privilege as a White person. I feel good about being White. Goodness of Fit Statistics CFI TLI RMSEA χ2 df P-­value N

1.00* −0.48 0.99 0.76 −0.35 Estimate 0.97 0.96 0.03 125 75 < 0.05 743

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: * denotes loadings fixed to 1.00

resentment questions only “allege that the underlying source of the [respondents’ answers] is resentment”; they developed a set of questions that aimed to remove any ambiguity of resentment (2011, 119; emphasis in the original). However, we hypothesized that these new items, while helpful, would not necessarily elicit a wholly different set of attitudes. In terms of using this measure, our hypothesis is that all four traditional racial-­resentment items (Kinder and Sanders 1996) and five items taken from the Explicit Racial Resentment battery will load on a single dimension; this is illustrated in table 7.4. The variables load very strongly on a single dimension. This hypothesis is confirmed, and the single factor we retained explains over two-­ thirds of the variance in all nine items.2 As the goodness-­of-­fit statistics,

200 / Chapter Seven Table 7.4.  Confirmatory Factor Analysis of Racial Resentment Explicit Racial Resentment Items I don’t understand why race is any different from what other people have to deal with. I resent any special considerations that African Americans receive because it’s unfair to other Americans. For African Americans to succeed they need to stop using racism and slavery as excuses. Special considerations for African Americans place me at an unfair disadvantage because I have done nothing to harm them. African Americans bring up race only when they need to make an excuse for their failure.

1.00* 1.17 1.16 1.13 1.21

Standard Racial Resentment Items Generations of slavery and discrimination have created conditions that make it difficult for Blacks to work their way out of the lower class. The Irish, Italians, Jews, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors. It’s really a matter of some people not trying hard enough; if Blacks would only try harder, they could be just as well off as Whites. Over the past few years, Blacks have gotten less than they deserve. Goodness of Fit Statistics CFI TLI RMSEA χ2 df P-­value N

−1.02 1.07 1.15 −0.85 Estimate 1.00 1.00 0.00 20.68 21 0.48 743

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: * denotes loadings fixed to 1.00

including the insignificant χ2 statistic, indicate, the data fit this model very well. Henceforth, we call this dimension “Racial Resentment” (RR). We have one final preliminary analysis to perform. One thing that we have alluded to and that became quite explicit in chapter 3 is that the relationship between racial attitudes—­measured by the original RRS—­and political ideology differs between young White people and older White Americans. A debate concerning the relationship between racial prejudice and political ideology is heated and ongoing. Since this book takes part in both contributing to and expanding that debate, we took the necessary step to include a test that helps determine whether various aspects of political ideology (i.e., fiscal values, cultural values, partisanship, conservatism/liberal­ ism) load on a single dimension.

Twenty-First-Century Racial Attitudes  /  201

As such, we estimate “Political Conservativism” to be a function of five variables: self-­reported conservative ideology (the traditional seven-­point measure, from very liberal to very conservative), affective ratings of the “Tea Party,” a dichotomous variable indicating the respondent’s support for gay marriage, and two measures regarding fiscal priorities. The first of these measures is a hundred-­point scale that elicits preferences for cutting government spending; the second asks respondents to choose which they would prefer as a solution to the government deficit, raising sales taxes (a regressive policy) or raising income taxes (usually a more progressive policy). These results are shown in table 7.5. The items load on a single dimension, with higher scores being associated with being more ideologically conservative. Thus far we have conducted several independent confirmatory factor analyses, predicting a total of eight latent factors: ·· A three-­factor solution to the Color-­Blind Racial Attitudes Scale (CoBRAS); ·· A three-­factor solution for Psychosocial Costs of Racism to Whites (PCRW); ·· A one-­factor solution to the racial resentment batteries, the traditional four-­ item measures, and measures of explicit racial resentment; and ·· A one-­factor solution that measures political ideology.

Each of these analyses shows good fit to the data, and we recover the number of hypothesized factors predicted by the developers of the various scales. The next set of analyses we conduct shows the relationship between these measures and other variables of interest.

Table 7.5.  Confirmatory Factor Analysis of Conservative Ideology Ideology -­7 point Scale Tea Party Favorability Support for Gay Marriage Cut Government Spending Increase Sales Tax

1.00* 0.82 −0.20 0.11 0.05 Goodness of Fit Statistics CFI TLI RMSEA χ2 df P-­value N

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: * denotes loadings fixed to 1.00

0.99 0.99 0.03 8.54 5 0.13 743

202 / Chapter Seven

Step 2: New Racism vs. Old Racism As an intermediary step, we examine the relationship between these relatively new measures and some measures that we have more familiarity with: old-­fashioned racism (OFR) and stereotypes. Given what we know about the underlying theories of each of the new racial-­attitudes measures, we expect that different dimensions will be correlated with different types of stereotypes and policy preferences. For example, we might expect those who are cognitively aware of racial injustices to be open to policies that aim to ameliorate racial disparities, or we might expect a person who has a fear of non-­Whites to also foster negative stereotypes about out-­group members. We test the convergent validity and predicted value of these new measures on Whites’ stereotypes and their racialized policy preferences. Old-­Fashioned Racism We begin with an examination of the relationship between the eight latent variables we uncovered in the previous section and several measures of old-­fashioned racial/racist stereotypes. First, we recoded each of the latent variables to have a minimum value of 0 and a maximum value of 1, which will allow for easy comparison throughout the remainder of the chapter. Our four measures of old-­fashioned prejudice are defined by the difference between a respondent’s evaluation of Whites and that respondent’s evaluation of Blacks on four different dimensions: Intelligent/Unintelligent, Hardworking/Lazy, Violent/Nonviolent, and Trustworthy/Untrustworthy. These variables have been coded so that a positive score indicates that the respondent rated Whites more positively on the dimension and a negative score indicates that the respondent rated Blacks more positively. Each dependent variable has a theoretical minimum of −1 (bias favoring Blacks) and maximum of +1 (bias favoring Whites), with a score of 0 representing a respondent rating both groups equally. The average scores on each of the four measures are 0.13 (violence), 0.11 (work ethic), 0.08 (trustworthiness), and 0.06 (intelligence). The mean of each of these measures is statistically different from zero, with a minimum t-­statistic of 5.25 (intelligence). Interestingly, with regard to cohorts, there is no statistical difference between White millennials and older Whites on any of the items. Figure 7.1 shows a “heat map,” which illustrates both the strength and direction of each of the eight dimensions’ relationships with the four sets of stereotypes. Those cells that are shaded white are most strongly negatively

Twenty-First-Century Racial Attitudes  /  203

Racial Empathy White Guilt Unaware: Blatant Fear of Others Ideology Unaware: Privilege Unaware: Institutional

en llig te

In

k

Et

hi

ce

c

t Tr us

W or

Vi o

le

nc

e

Racial Resentment

Data: 2014 CCES, N = 743 White Respondents

7.1.  Factor Effects on Measures of “Old-­Fashioned” Racism

correlated with the variable of interest; those that are darker are more strongly positively correlated. For example, racial empathy and differences in ratings on work ethic are strongly negatively correlated (r = −0.21), while racial resentment is very highly positively correlated with perceptions of Blacks as more violent than Whites (r = 0.40). Those areas on the heat map that are shaded gray are where the correlation between the two variables is close to zero, like the correlation between our measure of latent ideology and the differences in ratings between Whites and Blacks on intelligence (r = 0.09). On the whole, the relationships between the new measures and the OFR measures are in the expected directions. A heat map provides just one way of conceptualizing and visualizing the relationships that these new measures have with older measures. We can also rely on basic bivariate regression. An added benefit of regression is that we can also learn about the predictive power of these new variables through R2. R2 values are equivalent to squaring the bivariate correlations presented

204 / Chapter Seven

in the heat map, and they are estimates of the proportion of variance that we can explain in dependent variables by using the variance of each independent variable. Tables 7.6, 7.7, 7.8, and 7.9 provide another evaluation of the relationships between the eight dimensions of racial attitudes and our four stereotype measures. Table 7.6 reveals that for some of the more traditional “old-­fashioned” measures of racial prejudice, the variables are independently predictive and in the expected direction. Those who are racially resentful, fearful of other groups, unaware of institutional racism, and ideologically conservative are more likely to rate Blacks as less intelligent than Whites. Those who feel empathy, on the contrary, are more likely to move toward rating the groups more equally. Furthermore, table 7.6 shows that while resentment explains approximately 2 percent of variance in OFR, empathy and fear each explain over 4 percent. Table 7.7 presents the estimates for differences in the work-­ethic stereotypes. Recall that the developers of the traditional racial resentment scale claim that this set of items measures both anti-­Black affect and the belief that Blacks are violating the Protestant work ethic. The full resentment factor, which includes both RRS and EXR, accounts for about 8 percent of the variance in the ratings on work ethic and has the coefficient with the largest absolute magnitude (0.45). When we examine the intercept for the first model, one would predict that someone who scored the lowest on the resentment factor would rate Blacks approximately 0.17 points higher than Whites in terms of their work ethic, on a scale from 0 to 1. For those Whites who are the most resentful, this model predicts that they would rate Blacks 0.28 points lower than Whites. Again, we see that both empathy and guilt are negatively signed, while the CoBRAS subscales and our ideology measure are all positively signed and significant in their expected direction. Table 7.8 presents the bivariate regression analysis of each dimension being regressed on stereotypes of violence, stereotypes that many scholars might think are particularly antiquated. Here, the resentment factor explains about 16 percent of the variance, while the other factors each explain between 5 percent and 13 percent. Again, conservative ideology is positively correlated with measures of old-­fashioned prejudice but consistently has a relatively small coefficient when compared to the other measures. This is important because it is evidence that the other measures of racial attitudes are operating differently than they would if they were just representing ideology.

0.168* (0.043) −0.051* (0.026) 693 0.022

−0.254* (0.045) 0.188* (0.026) 693 0.044

−0.028 (0.042) 0.060* (0.020) 693 0.001

0.453* (0.058) −0.169* (0.035) 688 0.082

−0.387* (0.063) 0.310* (0.036) 688 0.052

−0.210* (0.058) 0.189* (0.027) 688 0.019

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: *p < 0.05, standard errors in parentheses

N R2

Intercept

Coefficient

Guilt 0.361* (0.061) −0.080* (0.032) 688 0.048

Fear

0.304* (0.061) −0.069* (0.035) 688 0.035

Racial Privilege

Empathy

Resentment

0.059 (0.045) 0.016 (0.026) 693 0.003

CoBRAS

0.256* (0.044) −0.078* (0.023) 693 0.046

PCRW

Table 7.7.  Predicting Old-­Fashioned Racism: Whites as Harder Workers

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: *p < 0.05; standard errors in parentheses

N R2

Intercept

Coefficient

Fear

Racial Privilege

Guilt

Resentment

Empathy

CoBRAS

PCRW

Table 7.6.  Predicting Old-­Fashioned Racism: Whites as More Intelligent

0.396* (0.058) −0.127* (0.035) 688 0.063

Institutional Discrimination

0.124* (0.043) −0.022 (0.026) 693 0.012

Institutional Discrimination

0.206* (0.062) −0.012 (0.035) 688 0.016

Blatant Racism

0.057 (0.045) 0.018 (0.025) 693 0.002

Blatant Racism

0.181* (0.041) −0.004 (0.025) 688 0.027

Ideology

0.062* (0.030) 0.013 (0.018) 693 0.006

Ideology

0.587* (0.052) −0.227* (0.032) 692 0.158

−0.424* (0.058) 0.352* (0.033) 692 0.072

−0.340* (0.053) 0.267* (0.025) 692 0.056

0.390* (0.057) −0.072* (0.029) 692 0.064

0.324* (0.050) −0.124* (0.031) 677 0.059

−0.325* (0.054) 0.246* (0.031) 677 0.052

−0.129* (0.050) 0.123* (0.023) 677 0.010

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: *p < 0.05, standard errors in parentheses

N R2

Intercept

Coefficient

0.309* (0.052) −0.084* (0.027) 677 0.049

Fear

0.186* (0.053) −0.034 (0.030) 677 0.018

Racial Privilege

Guilt

Resentment

Empathy

CoBRAS

0.477* (0.055) −0.141* (0.032) 692 0.097

PCRW

Table 7.9.  Predicting Old-­Fashioned Racism: Whites as More Trustworthy

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: *p < 0.05, standard errors in parentheses

N R2

Intercept

Coefficient

Fear

Racial Privilege

Guilt

Resentment

Empathy

CoBRAS

PCRW

Table 7.8.  Predicting Old-­Fashioned Racism: Blacks as More Violent

0.293* (0.051) −0.099* (0.030) 677 0.047

Institutional Discrimination

0.535* (0.053) −0.184* (0.031) 692 0.130

Institutional Discrimination

0.123* (0.053) 0.002 (0.029) 677 0.008

Blatant Racism

0.366* (0.057) −0.074* (0.031) 692 0.057

Blatant Racism

0.112* (0.035) 0.005 (0.022) 677 0.015

Ideology

0.294* (0.037) −0.045* (0.023) 692 0.082

Ideology

Twenty-First-Century Racial Attitudes  /  207

Similarly, table 7.9 reveals that each of the variables is associated with differences in ratings on trustworthiness, but the relative effects are not as large as we saw previously. Most of the factors each explain about 5 percent of the variance in the dependent variable, with resentment and empathy having the strongest relationships in terms of coefficient size. In summation, these analyses present evidence that each of the dimensions under consideration is independently related to other measures of racial animus. More often than not, our latent measures of empathy and White guilt are negatively correlated with having anti-­Black attitudes, while resentment, fear, conservative ideology, and all three subscales from the PCRW items are strongly related to old-­fashioned racist stereotypes. Taken together, when we examine these bivariate relationships, we see there are times when the resentment factor provides a healthy amount of predictive power, but only by examining other dimensions of racial attitudes are we able to move our discipline further on the matter of measuring racial sentiments. In other words, racial resentment provides some explanation for stereotypes related to work ethic and propensity toward violence, but by taking a step back and incorporating insights from other disciplines, we are able to develop a more nuanced approach to contemporary racial attitudes. In particular, we can now more clearly see that people’s awareness of institutional racism gives us just about the same amount of explanatory power—­and we can pinpoint the exact underlying construct that might be driving the gap between people’s attitudes toward Blacks and people’s attitudes toward Whites. Racialized Policy Preferences Next, we present similar analyses for three racialized policy and political issues: affirmative action, welfare, and presidential approval. Specifically, we examine whether the respondent supports affirmative action for Blacks, whether they believe spending on welfare programs should be cut, and whether they approve of President Barack Obama. The binary nature of each of the three dependent variables leads us to model each of these relationships using logistic regression. Since coefficients from a logistic regression are not immediately interpretable, we present the change in predicted probability associated with a change in each of the eight independent variables that is equivalent to moving along the interquartile range for each variable, shifting from the twenty-­fifth percentile on the scale to the seventy-­fifth percentile. We present these predicted probabilities, as well as the odds ratios (the ratio of the two predicted probabilities), in table 7.10.

208 / Chapter Seven Table 7.10.  Predicting Support for Racialized Targets

PCRW

CoBRAS

Resentment Empathy Guilt Fear Racial Privilege Institutional Discrimination Blatant Racism Ideology

Approve of Obama

Support Affirmative Action

Support Cutting Welfare

25th

75th

Ratio

25th

75th

Ratio

25th

75th

Ratio

0.50 0.25 0.18 0.37 0.53 0.47

0.13 0.42 0.49 0.32 0.11 0.67

3.85 0.60 0.37 1.16 4.82 0.70

0.36 0.17 0.12 0.28 0.39 0.37

0.07 0.33 0.37 0.24 0.06 0.06

5.14 0.52 0.32 1.17 6.50 6.17

0.30 0.49 0.56 0.43 0.26 0.28

0.58 0.37 0.31 0.44 0.59 0.57

0.52 1.32 1.81 0.98 0.44 0.49

0.50 0.67

0.17 0.04

2.94 16.75

0.37 0.43

0.10 0.07

3.70 6.14

0.30 0.19

0.56 0.66

0.54 0.29

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: All effects, except that of fear on welfare, are significant at conventional levels (p < 0.05). Ratios cal­ culated by dividing the predicted probability in the first column by that in the second column (e.g., 0.50/ 0.13 = 3.85).

With the exception of the effect of the racialized fear factor on respon­ dents’ desire to cut welfare spending (which is italicized), all coefficients have z-­statistics that have an absolute value larger than 2.0 and are statistically significant at conventional levels (p < 0.05). The ratios help to interpret the marginal effects in terms of the relative like­ lihood of each outcome, where a ratio greater than 1 indicates diminishing support and a ratio less than 1 denotes increasing predicted support. For example, the ratio of 3.85 in the first row and third column of table 7.10 indicates that someone at the twenty-­fifth percentile on our measure of latent resentment is 3.85 times as likely to have voted for Obama than someone who is at the seventy-­fifth percentile on that same measure; 3.85 is the ratio of 0.50 to 0.13. Relatedly, table 7.10 reveals that those who are less racially resentful are about five times more likely to support affirmative action (with a ratio of 5.14) and half as likely to prefer to cut welfare spending (with a ratio of 0.52). The ratios for ideology are even larger and in the same direction. Consistent with the theory of color-­blind racial attitudes, support of racialized policies like affirmative action seems to drop precipitously as the “unawareness” or lack of acknowledgment of institutional racism is more prevalent. For instance, those who are in the twenty-­fifth percentile of un­ awareness of racial privilege (and thus are aware of White privilege) are 6.5 times more likely to support affirmative action than those who are far less aware of the advantages that Whites receive due to racial privilege.

Twenty-First-Century Racial Attitudes  /  209

Similar effects are seen in awareness of institutional racism (6.17) as well as ignorance of general racial inequalities (3.70). It is perhaps unsurprising that the less aware an individual is of systematic or institutionalized racism, the more likely that individual is to want to cut welfare spending. Finally, the Psychosocial Costs of Racism to Whites subscales show that higher levels of empathy and White guilt are both related to a greater predicted probability of approving Obama and supporting affirmative action, but a lower predicted probability for wanting to cut welfare. These analyses provide an initial step to verify that each scale has some predictive validity. While we know that each dimension is related to anti-­ Black stereotypes, these estimates show us that the “new” scales we have in­ corporated into our research also translate into meaningful predictors of policy attitudes.

Step 3: A Second-­Order Model of Racial Attitudes These initial steps have allowed us to get some familiarity with each of the eight latent variables. We suspect that by putting the variables in conversation with one another—­rather than exploring them as totally separate, independent constructs—­we might be able to develop a more fully specified model of contemporary racial attitudes. In this final step, we examine our data in a way that looks at the relationship between the previously uncovered latent variables and find that they group nicely on two dimensions. Generally, the RRS, CoBRAS, PCRW, and EXR are modeled as a series of items aimed to measure one underlying, latent variable. They tend to be conceptualized as something like what is shown in figure 7.2: a unidimensional latent construct being represented by some number of observed variables. However, because we have asked over seven hundred people questions from each of these batteries aimed to measure various aspects of contemporary racial attitudes, we can leverage the data in a special way. Instead of simply looking at each of the scales and their subscales separately, we can test whether a higher order factor (or factors) can better explain the variance within and between the observed data. This allows us to see if items developed for one scale have a relationship with some of the items in another. Theoretically, we could find a relationship between and among items of the four constructs (resentment, PCRW, CoBRAS, ideology) as well as two second-­order factors. This set of relationships is illustrated in figure 7.3. Here, two second-­order factors explain the variance in the latent first-­order factors (here generically labeled as “construct” and not accurately reflecting some of the subscales), and the first-­order factors explain the variance in

210 / Chapter Seven

7.2.  A Single-­Order Latent Factor Model

the items used to measure each of our eight dimensions: racial resentment, ideology, three subscales from the CoBRAS items, and the three subscales from the PCRW battery. We begin this analysis by looking at the correlations between our previously estimated factor scores. These correlations are presented in table 7.11; this table reveals that some dimensions are very strongly related to one another while others have relatively small correlations and seem to be unrelated (e.g., guilt and empathy). This initial analysis suggests that perhaps multiple higher-­order dimensions might be able to better explain the variance between these different subscales. In fact, the correlation matrix has multiple eigenvectors with values greater than 1.0, which hints at the fact that a superstructure could explain how each of these subscales is related to one another and to a “higher” dimension. Next, we determine whether a higher-­order factor structure can account for the covariance between the latent dimensions we have already recovered. We do this, first, by looking at the eigenvalues, which provide a cursory indication of how many dimensions might be underlying each of the eight latent constructs. The three largest eigenvalues in our matrix are 4.6, 1.5, and 0.31. Ultimately, we are only concerned with the eigenvalues that are greater than 1.0. The fact that we have two eigenvectors with values greater than 1.0 indicates that a two-­factor second-­order model is likely to fit the data, which our preliminary analysis confirms. Finally, we fit a second-­order confirmatory-­factor analytic model using the raw survey data. This involves simultaneously estimating factor loadings for each of the eight first-­order factors and estimating how those factors load on two additional higher-­order latent variables. As this model estimates eighty-­three different parameters (factor loadings, thresholds, covariances,

e6

X6

e7

X7

e8

X8

Construct 3

7.3.  A Second-­Order Latent Factor Model

e5

e3

e2

e1

e4

X5

X3

X2

X1

X4

Construct 2

Construct 1

Second Order Factor 1

e9

X9

e10

X10

Second Order Factor 2

e11

X11

Construct 4

e12

X12

1 −0.443 −0.647 0.18 0.902 0.96 0.626 0.806 1 0.046 −0.569 −0.318 −0.388 −0.385 −0.403

Empathy

Data: 2014 Cooperative Congressional Election Study (CCES) Note: N = 743, non-­Hispanic Whites only

Resentment Empathy White Guilt Fear of Others Racial Privilege In. Discrim. Blatant Racism Political Ideology

Resentment

1 0.434 −0.752 −0.61 −0.402 −0.497

Guilt

1 −0.131 0.167 −0.009 0.156

Fear

Table 7.11.  Correlations between Independently Estimated Dimensions

1 0.846 0.728 0.776

Unaware of White Privilege

1 0.605 0.797

Unaware of Institutional Racism

1 0.696

Unaware of Blatant Racial Issues

1

Ideology

Twenty-First-Century Racial Attitudes  /  213 Table 7.12.  Higher Order Factor Loadings Second-­Order Factor 1 Racial Resentment Racial Empathy White Guilt Fear of Other Racial Groups Racial Privilege Institutional Discrimination Blatant Racism Conservative Ideology Correlation (F1, F2) = 0.20*

Second-­Order Factor 2

1.00α 0.81* −0.91* 0.00α 0.93* 1.14* 0.45* 1.11*

−1.63*

Data: 2014 Cooperative Congressional Election Study (CCES); N = 743 White respondents Notes: *p < 0.05; α parameter value fixed

and error covariances), we refrain from a tabular presentation of our results. In terms of the traditional measures of model fit, the second-­order factor analysis fits the data very well: the ratio of chi-­squared to degrees of freedom is below 2 (1.64), the CFI is 0.97, TLI is 0.96, and the RMSEA is below 0.03. The second-­order factor loadings are shown in table 7.12; all factor loadings are significantly different from zero at conventional levels.3 The first column in table 7.12 lists the eight latent variables that make up the first-­order factors (these were labeled “construct” in figures 7.2 and 7.3). The second and third columns provide information about the higher, second-­order factors. A graphical presentation of these factor loadings is shown in figure 7.4, with the factor labels for the first dimension being “fanned out” so that they can be read; the Cartesian coordinates for this plot are simply those given in table 7.12, with a zero being given where no loading was specified. The items that load onto the first of the two higher factors include resentment, White guilt, the three subscales of the CoBRAS, and political ideology. The factor loadings are illustrated in figure 7.4 in such a way that higher scores relate to more positive attitudes toward Blacks and other minorities, a higher level of empathy for racial minorities, and an increased awareness regarding racial issues. More specifically, respondents who score higher on factor 1 are more aware of various examples and types of racism, less ra­­ cially resentful, and less ideologically conservative. They are also likely to have some level of “White guilt” from the PCRW battery. All of these subfactors consist of items that relate to respondents’ understanding and awareness of race, racism, and racial privilege in America. This second-­order factor may indicate how knowledgeable an individual is regarding the racial

214 / Chapter Seven

7.4.  The Two Dimensions of Whites’ Racial Attitudes

status quo. Consequently, we label this more cognitively oriented dimension “recognition of institutional racism.” In contrast, the second dimension, factor 2, loads most strongly on the PCRW subscales related to fear and empathy. Respondents who score more highly on this factor have greater empathy toward racial minorities and are less fearful of other racial groups. As such, we label this more emotionally oriented factor the “fear-­empathy” dimension.

Gaining Traction . . . The results presented thus far provide several major contributions to the literature on racial attitudes in American politics. To begin, all of these various measurements are typically employed separately and independently, but here we show that these measures have relationships with one another. By putting these measures in conversation with one another, we gain insight into how various aspects of racial attitudes—­which go beyond prejudice—­ are related. Second, when we use these measurements independently to predict attitudes and behaviors of concern, we are capturing either how people think, understand, and perceive the way race works in the United States or how people feel about out-­group members. From our second-­order analysis, however, we see that there is both a cognitive dimension of racial attitudes

Twenty-First-Century Racial Attitudes  /  215

and an emotional dimension, and further, we find that these two dimensions are not very strongly correlated to one another (r = 0.20). Here, we can gain traction on whether feelings/emotion and knowledge/awareness have different, opposing, or additive effects on racialized policy preferences and attitudes toward various groups. Furthermore, these findings speak to the ongoing debate about the relationship between racial attitudes and principled, ideologically based rejections of racialized social policies. Instead of an either/or perspective, the second-­ order factor analysis allows us to conclude that both aspects are particularly important with regard to how Whites’ racial attitudes are organized. More importantly, by imposing constraints on the factor solutions so that the resulting estimated second-­order factors are minimally correlated, we can be sure that the component that does not load on the ideology items is hardly related to any measures of principled conservativism. Consequently, while our first factor, the cognitive dimension (which includes a nuanced measure of political ideology), is very strongly correlated with the traditional seven-­point measure of political ideology (r = −0.68),4 the correlation between ideology and the second dimension is estimated to be 0.15, which, although statistically distinguishable from zero, is relatively small. So where does this leave us with comparing older White generations to White millennials? Figure 7.5 provides an illustration of the distribution of attitudes for White millennials and older White Americans. Millennials score significantly higher on the cognitive dimension (t = 2.30, p < 0.05) in

Millennials

Empathetic Dimension Factor Loadings

Older Whites

Cognitive Dimension Factor Loadings 7.5.  The Two Dimensions of Racial Attitudes by Generation

216 / Chapter Seven

comparison to their predecessors. They score lower on the empathetic dimension than older Whites, though this difference is not statistically different from zero. This, in many ways, harkens back to what we saw in previous chapters when we interviewed millennial Whites about race and racism in America: younger Whites seem to be cognizant of racial privilege, but that knowledge may not necessarily translate into caring about racial injustice. In the final section of this chapter, we examine the role of these two dimensions on Whites’ attitudes toward Blacks and toward racialized and nonracialized policy preferences.

Examining Racial Attitudes in Two Dimensions We have spent the great majority of this chapter focused on the nitty-­gritty aspects of uncovering a model that helps us to more accurately conceptualize and measure contemporary racial attitudes. In this last section, we do a first-­pass set of analyses to examine the two latent dimensions we uncovered. Our goals are to (1) examine their convergent validity and predictive value, (2) explore how each of the dimensions relates to variables of interest, and (3) leverage our data and new measures to determine whether either dimension performs differently across generations. Specifically, we begin by looking at how the cognitive and empathetic dimensions of Whites’ racial attitudes predict the difference in scores that respondents give to Whites in comparisons to Blacks on four stereotypes: intelligence, trustworthiness, work ethic, and violence. Here, we estimate a series of linear models predicting these differences as a function of respon­ dents’ scores on each of the underlying latent dimensions. Again, increasing values on the cognitive dimension mean that people have greater awareness and acknowledgment of structural inequalities. Meanwhile, higher values on the empathetic dimension suggest that people are more empathetic and less fearful of people of other races. Furthermore, since the latent factors are estimated to have a mean of zero and a standard deviation of one, the regression coefficients for the cognitive and empathetic dimensions can be interpreted similarly to standardized regression coefficients. We begin by analyzing all of the respondents without regard to generational status. The results are presented in table 7.13. As with all regression models, we can interpret the intercept to be the predicted value when both independent variables equal zero. Since the cognitive and empathetic dimensions both have a mean of zero, we can say that a respondent who is average on both dimensions would be predicted to rate Blacks as 0.13 points

Twenty-First-Century Racial Attitudes  /  217 Table 7.13.  Old-­Fashioned Racism Predicted by Two Dimensions of Racial Attitudes Violence Intercept Cognitive Dimension Empathetic Dimension N R2

0.13* (0.01) −0.08* (0.02) −0.09* (0.02) 692 0.18

Intelligence 0.06* (0.01) −0.01 (0.01) −0.08* (0.02) 693 0.10

Trustworthy 0.08* (0.01) −0.03* (0.01) −0.08* (0.02) 677 0.09

Hardworking 0.11* (0.01) −0.05* (0.01) −0.10* (0.02) 688 0.12

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Standard errors in parentheses; *p < 0.05, two-­tailed

more violent and to see Whites as 0.06 points more intelligent, 0.08 points more trustworthy, and 0.11 points more hardworking. Table 7.13 shows that both the cognitive and empathetic dimensions are strongly related to old-­fashioned racial stereotypes. As a respondent becomes more cognitively aware of systemic racial bias, however, the differences for where they place Whites and Blacks on three of the four traits get significantly smaller. These effects, however, are not as large as those from the second (empathetic) dimension. As a respondent becomes more racially empathetic (and less fearful), we would predict that respondent to move from rating Whites as more favorable on all four dimensions to rating Whites and Blacks roughly equally. In other words, the extent to which Whites empathize with (or fear) non-­Whites influences the extent to which they stereotype Blacks, an effect seen by both the size and significance of all of the regression coefficients. We now look at how these measures relate to old-­fashioned stereotypes among White millennials separate from older Whites. The results are presented in table 7.14. From our perspective, these results are quite interesting. What we consistently see is that while both dimensions of racial attitudes influence the extent to which older Whites foster racial stereotypes, only the affective dimension influences young people’s stereotypes. It is feelings of fear of other groups and empathy for minorities that shift young Whites’ characterization of Blacks. We should note that the difference in R2-­values across these simple models shows that these two dimensions, especially the empathetic dimension, predict a greater proportion of the variation in White millennials’ attitudes than in the attitudes of older Whites. It may be the case that old-­fashioned racism—­the belief that Blacks

0.14* (0.01) −0.09* (0.01) −0.07* (0.02) 547 0.18

0.10* (0.02) −0.03 (0.02) −0.14* (0.06)

145 0.24

Older Whites

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Standard errors in parentheses; *p < 0.05, two-­tailed

N R2

Empathetic Dimension

Cognitive Dimension

Intercept

Millennials

Violence

146 0.17

0.07* (0.03) −0.01 (0.03) −0.13* (0.06)

Millennials

Intelligence

547 0.08

0.06* (0.01) −0.02* (0.01) −0.06* (0.01)

Older Whites

Table 7.14.  Old-­Fashioned Racism Predicted by Two Dimensions, by Generational Status

145 0.21

0.07* (0.03) −0.00 (0.03) −0.15* (0.06)

Millennials

Trust

532 0.07

0.08* (0.01) −0.04* (0.01) −0.06* (0.02)

Older Whites

143 0.21

0.08* (0.03) −0.01 (0.03) −0.16* (0.06)

Millennials

Work Ethic

545 0.11

0.11* (0.01) −0.07* (0.01) −0.08* (0.02)

Older Whites

Twenty-First-Century Racial Attitudes  /  219 Table 7.15.  Predicting Nonracialized Policy Positions by Generational Status

Intercept Cognitive Dimension Affective Dimension N R2

Support Assault Weapons Ban

Support EPA

Older Whites

Older Whites

0.72* (0.13) 1.16* (0.17) 0.12 (0.12) 586 0.16

Millennials 0.09 (0.22) 0.72* (0.26) 0.36 (0.23) 153 0.08

−0.26* (0.12) 1.50* (0.16) 0.18 (0.13) 583 0.24

Millennials −0.11 (0.24) 0.60* (0.29) −0.16 (0.21) 150 0.05

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Standard errors in parentheses; *p < 0.05, two-­tailed

are inferior to Whites on a number of desirable qualities—­is more deeply rooted in fear for younger Whites than it is for older Whites due to increased racial isolation and White habitus (Bonilla-­Silva, Goar, and Embrick 2006; Lewis 2001), or perhaps due to what Blinder (2007) calls “two-­track” socialization. Millennials are more distant from the era of overt racist practices that is at the core of today’s persistent racial inequalities, and considering that they are largely raised by people who avoid discussing matters of racial inequality head on (Hagerman 2018; Underhill 2018), young White people may rely on persistent stereotypes, which are often rooted in fear, to explain what they see. Next, we examine how and the extent to which the two dimensions predict policy preferences—­both racialized policies (i.e., amnesty for undocumented immigrants, support for affirmative action and welfare) and ostensibly race-­neutral policies (i.e., support for an assault weapons ban; attitudes concerning the Environmental Protection Agency, or EPA). We examine a wide range of policies in order to execute an initial test of discriminate validity. Ideally, a good measure of Whites’ racial attitudes only predicts policy preferences that are racialized, and not anything outside of that range. To be sure, the measure of racial attitudes that we have uncovered has two dimensions, the first of which includes political ideology; as such, we expect that the cognitive dimension may help to shape nonracialized policy attitudes, but the role of the second dimension—­which only incorporates aspects of  Whites’ feelings toward people of other races—­should be constrained to racialized policy preferences only. We begin with an examination of nonracialized policy preferences in table 7.15.

220 / Chapter Seven Table 7.16.  Predicting Racialized Policy Positions by Generational Status

Intercept Cognitive Dimension Affective Dimension N R2

Support Amnesty

Increase Welfare

Affirmative Action

Older Whites

Older Whites

Older Whites

−0.55* (0.126) 1.55* (0.15) 0.23* (0.12) 589 0.25

Millennials −0.07 (0.234) 0.932* (0.278) 0.63* (0.22) 154 0.14

−1.51* (0.152) 1.07* (0.15) 0.03 (0.12) 538 0.16

Millennials −1.44* (0.45) 0.89 (0.54) −0.18 (0.37) 134 0.08

−1.66* (0.17) 2.16* (0.24) 0.05 (0.15) 588 0.37

Millennials −1.23* (0.39) 1.26* (0.54) −0.08 (0.32) 152 0.15

Data: 2014 Cooperative Congressional Election Study (CCES) Notes: Standard errors in parentheses; *p < 0.05, two-­tailed

The results in table 7.15 reveal that nonracialized policy preferences are best predicted by the first dimension of respondents’ attitudes; this dimension includes measures of political ideology. We see that this dimension is doing the work we expected. Higher scores on the cognitive dimension are associated with more ideological liberalism; thus, as scores increase, we see that respondents are more in favor of a ban on assault weapons and express greater support of the EPA. The affective component of racial attitudes does not influence respondents’ preferences on these two nonracialized policies, as expected.5 These findings hold across the generational divide. Table 7.16 reveals the effects of the two dimensions on three racialized policies: amnesty, welfare, and affirmative action. These results show that the cognitive dimension of  Whites’ racial attitudes is a very strong predictor of the positions taken by Whites of both groups. The results also reveal that there are some instances where the effect of racial empathy is also a powerful predictor of racialized policy preferences. For instance, respondents were asked whether they agree that the government should “grant legal status to all illegal immigrants who have held jobs and paid taxes for at least 3 years, and not been convicted of any felony crimes.” Policies like this are sometimes referred to as “amnesty.” Figure 7.6 illustrates the marginal effects of the second dimension—­the empathetic factor—­on the probability that a respondent supports an amnesty program. These results are quite striking. As White millennials become more racially empathetic, they begin to favor this type of policy in overwhelming numbers. A significant, though smaller, effect is also seen for older Whites.

Twenty-First-Century Racial Attitudes  /  221

We highlight this finding, in particular, because it helps us to show that racialized policy attitudes cannot simply be explained by political ideology or racial prejudice; instead, we see that a second, unique dimension—­rooted in empathy for and fear of other groups—­that underlies Whites’ racial attitudes also influences their racialized policy preferences. In addition to the factors that are significant, the lack of significance in some places also piques our interest. For example, neither the cognitive nor affective dimension of racial attitudes influences White millennials’ attitudes on the issue of welfare. This could suggest that White millennials may view this policy as not necessarily a racialized issue, as their predecessors were socialized to do (Gilens 1999), but rather a class issue. Millennials may have unique opinions on welfare because a great deal of them entered into the labor market during the United States’ worst economic downturn in decades. For a time, they had high chances of being unemployed and were likely to have to return to live at home with their parents despite the fact that they are members of the most highly educated living generation in the United States (Luhby 2018; Rouse and Ross 2018). Consequently, they may have a different way of thinking about state-­sponsored social and economic safety nets. Their sentiments may be well illustrated by their support of 2016 presidential contender Bernie Sanders, a self-­identified Democratic socialist, and their marked lack of enthusiasm for capitalism (Newport 2018). Additionally, the results suggest that neither older Whites’ nor millennial Whites’ attitudes toward affirmative action are influenced by their affect, only by their ideology and cognition. There are two things we should note here. First, even if we exclude the specific political ideology from the first

Predicted Probability of Supporting Amnesty

Cognitive Dimension

Empathetic Dimension

1.00 0.75

Group Millennials

0.50

Older Whites

0.25 0.00 −2

−1

0

1

2

−2

Factor Score

−1

0

7.6.  Predicting Support for Amnesty

1

2

222 / Chapter Seven

dimension, we still maintain these results. Secondly, the question about affirmative action was worded in a way that requires people to consider historical and structural inequalities to answer the question, which would likely very strongly tap into the first dimension, a measure of knowledge about racial privilege and institutional racism.

For more than thirty years, scholars have primarily relied on one measure to capture variance in Whites’ racial policy preferences and political attitudes, but we argue that political science’s traditional measure only speaks to one aspect of contemporary racial attitudes. Our findings in this chapter support that assertion. Our analyses in this chapter show that when we ask a wider array of questions designed to assess various components of Whites’ racial attitudes, a single first-­order factor does not sufficiently explain the variance we observe in this richer set of survey items; instead, we see that eight first-­order factors fit the data quite well. And when these factors are put into conversation with one another, two second-­order dimensions emerge to explain the variance between the individual scales. Our results indicate that Whites’ racial attitudes can be best characterized as sitting on two (approximately) orthogonal dimensions. The first dimension, which maps onto aspects of cognitive thinking, accounts for Whites’ perceptions about the state of racism in America. This dimension assesses whether and the extent to which Whites recognize racism as an ongoing problem in the United States, captures respondents’ explanations of racial inequality (e.g., structure, White privilege, blatant racism), and includes aspects of what others have termed “White guilt,” a potential reaction to an awareness of the privilege that is bestowed on them by virtue of having been born and recognized by others and society as White. The second dimension, in contrast, captures something that political scientists have yet to leverage in a systematic way to understand Whites’ attitudes. Measures of the empathetic reactions to racism and of  Whites’ fear of other groups help anchor the two poles of the second latent dimension that our analyses uncovered. This dimension is particularly important as social psychologists and political psychologists have shown that Whites routinely dehumanize Blacks (Goff et al. 2008; Goff et al. 2014; Jardina and Piston 2016). Meanwhile, sociologists have shown that racial apathy is increasingly common among White Americans (Forman 2004). Thus, to completely understand the structure of Whites’ racial attitudes, we must take into account not only how Whites think about race and racism in America but also how they feel in the empathetic or fearful sense.

Twenty-First-Century Racial Attitudes  /  223

We have mentioned before that we intuited that White millennials’ racial attitudes differ in some shape, way, or form from those of their predecessors. Our analyses here provide greater insight into this line of inquiry. Not only do we find that Whites’ contemporary racial attitudes are well explained by two dimensions, but also we can leverage these independent components to undergo more rigorous analysis of how racial attitudes operate across generational cohorts. For instance, our analyses revealed that younger Whites seem to be more cognizant of their racial privilege than their predecessors are, but they lack the empathetic reactions to systemic racism that some might hope they would have. These findings illustrate two of the countervailing forces that we’ve presented—­the “empty knapsack” and the “moving walkway of racism”—­ and show how these forces work in tandem: millennial Whites are more likely to acknowledge their racial privilege, but they are no more likely to be saddened or angered by racial injustice. This is akin to those who are standing still on the moving walkway. However, here we see that they know precisely what the walkway affords them. Again, we gain greater discernment into America’s racial stasis. Despite all of these new insights, we still have a problem. We used nearly fifty questions to ascertain our findings here—­that’s about ten times as many questions as any social scientist at a large, public research university can afford to ask in a nationally representative sample. Needless to say, we need to balance our appreciation for depicting the complexity of racial attitudes with a goal of parsimony in measuring these attitudes. In the next (and last substantive) chapter, we devise a four-­question battery, which we have nicknamed FIRE, that taps into each of the components of racial attitudes that we have highlighted here.

Eight

The FIRE This Time

Donald Kinder and David Sears begin their influential article “Prejudice and Politics: Symbolic Racism versus Racial Threats to the Good Life” this way: “Although theories of prejudice have been extensively catalogued, empirical confrontations between competing theories are surprisingly rare” (1981, 414). This book represents one of those rare “empirical confrontations.” In the early 1980s, Kinder and Sears sought to measure and explain the configuration of White Americans’ racial attitudes that had evolved in the era after the civil rights movement; they tested their theory of “symbolic racism” against competing hypotheses related to “racial threats to whites’ private lives” and ultimately showed that symbolic racism could explain anti-­Black political attitudes and behavior (1981, 414). We set out on a similar intellectual endeavor in a new era of American political life, and our results thus far provide evidence that we ought to think of racial attitudes, generally speaking, as encompassing more than just prejudice by Whites against Blacks and, further, consider how contemporary racial attitudes may be composed of sentiments beyond “abstract, moralist resentments of blacks” (Kinder and Sears 1981, 414). Specifically, we have shown that two distinct higher-­order dimensions best describe the structure of Whites’ contemporary racial attitudes. The first dimension involves Whites’ awareness and recognition of the systemic racial bias against non-­Whites in America. The second is more affective in nature, having to do with whether Whites fear racial minorities and whether Whites are angry about racial inequity. To be sure, scholars have long noted that racial attitudes are likely to have both cognitive and affective components (Banks and Valentino 2012; Forman 2004; Pettigrew 2000), but the measures that are most often employed in large-­n, nationally representative surveys could not exploit this nuance. We hope Racial Stasis bridges this gap.

226 / Chapter Eight

In this last substantive chapter, we introduce a more holistic, nuanced, and useful measure of White Americans’ contemporary racial attitudes that taps into these two dimensions. We call this measure FIRE because it mea­ sures a combination of Fear that Whites may have toward people of color, an awareness of Institutional Racism, and racial Empathy. Until now, we have relied on nearly fifty questions to map White Americans’ contemporary racial attitudes. But asking that many questions is neither practical nor necessary. The first goal of this chapter is to outline a short-­form, parsimonious battery that taps into the two dimensions of  Whites’ racial atti­ tudes. Along with advocating for parsimony, we argue that any measure that aims to capture modern racial attitudes would need to meet four tests for validity (Carmines and Zeller 1979; DeVellis 2016). First, it would have face validity: we should know what we are measuring by the questions that are being asked. Second, it should have a level of convergent validity: items that at­ tempt to measure similar constructs ought to covary together in predictable ways. Third, it should have predictive validity. Fourth and finally, the measure should have discriminant validity: not only should the measure predict what it sets out to, but it should also be unrelated to measures to which it is not theoretically linked. After we describe the questions that compose the FIRE battery, we discuss how the measures should be used, execute several tests in efforts to ascertain whether this new measure meets these four criteria, and illustrate the battery’s added value with the help of data from the 2016 Cooperative Congressional Election Study (CCES) Common Content.1 The CCES surveyed over 60,000 Americans during the 2016 presidential election cycle; these data include 46,289 White Americans, of which 28.7 percent (13,273) are members of the millennial generation. With these data and the FIRE battery, we are also able to speak to one of the most pressing questions of the 2016 presidential election: What was the role of Americans’ racial attitudes in producing the election of Donald Trump?

Derivation of the FIRE Battery In the preceding chapters, we relied on questions from four batteries of racial attitudes and a scale that captured respondents’ political ideology in order to uncover two distinct higher-­order dimensions of contemporary racial attitudes. Needless to say, that is just too many questions for the average social scientist to ask on a survey. As such, we provide researchers with a short-­form battery that captures aspects of both dimensions. In order to do so, we conducted a series of computational tests to see which smaller subset of items would explain the most variance in our two

The FIRE This Time  /  227

second-­order dimensions. Specifically, we wrote a computational program to find and record the combinations of four or five of the fifty items that would simultaneously maximize the amount of variance (across the two dimensions). Computationally, this involved independently testing well over a hundred thousand unique combinations of items. In the circumstances where two combinations could explain approximately equal amounts of variance, we examined which subscales were being tapped. Any subsample that had multiple questions from a single subscale was passed over in favor of questions that capture each of the different theorized dimensions.2 Ideally, a truncated measure would include measures from both the emotional (fear/empathy) dimension and the cognitive dimension and the matrix of resulting correlations would be able to be decomposed into two dimensions (having two eigenvalues greater than 1.0). Out of this endeavor, we were able to meet these ideal circumstances. We landed on the following four questions: ·· I am fearful of people of other races. ·· White people in the US have certain advantages because of the color of their skin. ·· Racial problems in the US are rare, isolated situations. ·· I am angry that racism exists.

The first and fourth questions come from the Psychosocial Costs of Racism to Whites (PCRW) scale and tap into the emotional or affective dimension of contemporary racial attitudes. The remaining two questions are borrowed from the Color-­Blind Racial Attitudes Scale (CoBRAS), and both speak to the cognitive dimension that we uncovered in the previous chapter. Taken to­ gether, the four-­item battery has two eigenvalues above 1.0 (1.84 and 1.08) and explains approximately 82 percent of our first dimension and approximately 79 percent of the second dimension. To be sure, we call these four items a battery rather than a scale because the four questions should not, under any circumstances, be combined into a single additive scale. There are two reasons for this. First, the items are derived from a multidimensional structure, so combining them into a single dimension is antithetical to their derivation. Second, as empathy and fear are distinct emotions, these components of racial attitudes are likely to be theoretically linked to different outcome variables. Consequently, combining the two items into a single measure makes it impossible to isolate which of the two emotive poles is driving the outcome of interest. While we are not the first scholars to ask any one of these questions, we

228 / Chapter Eight

think that this particular set of four questions raises the bar on how we mea­ sure racial attitudes, in several ways. First, although Henry and Sears assert that it has been important to include specific mentions of Blacks in the development of modern-­day racial-­attitudes scales because “white prejudice against blacks has long been assumed to represent a significant obstacle to full racial equality” (2002, 258), our measures do not do this. This is an im­ portant shift. Given the demographic makeup of the United States, where Latinx people are the largest minority group and Asian Americans are the fastest-­growing group, it is necessary to broaden the questions that we pose to dominant members of society. Theoretically, the FIRE battery will have a longer life span, given that no particular racial or ethnic group is named. Second, we should note that none of the original racial resentment questions rose to the top of the most well-­performing items, suggesting that we ought to be taking a broader view on what constitutes racial attitudes. Finally, though some may find it strange that we included a question that explicitly asks whether people are fearful of members of other races, a question that may seem out of place in an era marked by color-­blindness, we believe this item will prove to be quite helpful. Fear has historically been elicited by political candidates in the United States, and arguably we still see that today—­be it in the form of implicit or explicit discussions of Muslims in a post-­9/11 world and of “invasions” of Central American asylum seekers, or in the call for a political agenda that polices Black and Latinx communities, who are, according to Trump, “living in hell. You walk down the street and you get shot” (Desjardins 2017; Joseph, D’Harlingue, and Wong 2008). Again, given that racial attitudes encompass more sentiments than prejudice, we ought to cast a broad, but theoretically driven, net. Distribution of Attitudes How do these ideas percolate among White Americans? Table 8.1 outlines the distribution of contemporary racial attitudes, highlighting the means and standard deviations for each of the four items, broken down by generational status. The results indicate that these items appear to have cross-­ generational comparability. Table 8.1 reveals that few White Americans report that they are fearful of racial minorities. White millennials are slightly less fearful than old Whites, but only very slightly—­on a five-­point scale, the difference amounts to just 0.04. While the difference-­in-­means statistic is statistically significant (t = 2.99, p < 0.01), it is not as if younger Whites are substantively less fearful of other groups.

37.39

37.56

31,052

24.47

25.4

40,015

23.58

22.47

3.83 1.11

12.69

12.31

3.84 1.12

1.67

2.05

Older Whites

8,963

3.87 1.14

38.15

28.60

18.64

10.99

3.36

Millennials

40,059

2.87 1.45

20.54

14.77

17.95

24.05

22.58

All Whites

31,080

2.98 1.45

22.60

15.90

18.34

22.75

20.31

Older Whites

8,979

2.48 1.37

13.43

10.87

16.60

28.54

30.47

Millennials

Recognize Whites Have Advantages

Data: 2016 Cooperative Congressional Election Study (CCES), White respondents only

Mean Standard Deviation N

1. Strongly Agree (%) 2. Somewhat Agree 3. Neither Agree nor Disagree 4. Somewhat Disagree 5. Strongly Disagree

All Whites

Fearful of Other Races

Table 8.1.  FIRE by Generational Status

40,006

3.65 1.22

29.98

31.26

17.25

15.91

5.37

All Whites

31,043

3.64 1.22

29.60

31.52

16.85

16.43

5.37

Older Whites

Racial Problems Are Rare

8,963

3.68 1.20

31.27

30.36

18.63

14.10

5.39

Millennials

40,056

1.73 0.97

2.06

2.9

16.18

24.05

54.7

All Whites

31,079

1.78 0.99

2.25

3.18

17.62

23.80

53.04

Older Whites

Angry Racism Exists

8,977

1.59 0.87

1.41

1.90

11.22

24.91

60.46

Millennials

230 / Chapter Eight

A similar finding holds true for acknowledging racism in America. When asked if they think that racial problems in the United States are rare, isolated incidents, White millennials are more likely to disagree than their older counterparts, but again by only four-­hundredths of a point on a five-­point scale. While this difference is statistically significant (t = 3.09, p < 0.01), it is largely due to having over forty thousand respondents; it is not necessarily substantively significant. It is not as if White millennials are in complete disagreement and older Whites are in complete agreement about the issue. We uncover larger differences on the other two items. White millennials report being angrier that racism exists than older Whites do by approximately one-­fifth of a point. And the largest difference, fully one-­half of a point, is seen in the question that asks for a respondent’s level of agreement with the statement “White people in the US have certain advantages because of the color of their skin.” On the five-­point scale, younger Whites average about 2.48 (between “somewhat agree” and “neither agree nor disagree”), while older Whites average about 2.98 (closer to being indifferent). To put these differences in perspective, White millennials and older White Americans differed by only about 0.04 on the other two questions; the differences here are between five and twelve times larger, respectively. Extant literature has suggested that White Americans’ level of education, partisanship, and political ideology are also likely to be related to the level and shape of their racial attitudes. As such, we examine the extent to which these three key demographic variables are correlated to the FIRE items. We depict these relationships in figure 8.1. The figure summarizes the average response on each of the four FIRE items, by generational status, across each of the demographic characteristics of concern. The solid line represents older Whites, and the dashed line represents White millennials. The first row of panels illustrates the relationship between the FIRE items across six levels of education.3 From left to right we can see that as respon­ dents gain more education, they tend to be slightly less fearful of other races and slightly more likely to recognize White privilege. The third and fourth panels indicate that education does not have a strong correlation with recognizing racial problems or with levels of racial empathy. What’s more, we see that there is little difference in these patterns across generations. Finally, the R2 terms for each item reveal how strongly education is related to each of the variables across the two groups. As far as the four FIRE items go, we can see that none of the measures is strongly correlated with education. The second row allows us to look at the relationship between responses to the FIRE battery and partisanship. The figure moves across the traditional seven-­point partisanship scale, from “strong Democrat” to “strong

4 3 2 1 0

4 3 2 1 0

4 3 2 1 0

Strong Liberal

Source: 2016 Cooperative Congressional Election Study Non−Hispanic White Respondents Only; N = 46,289

Strong Conservative

Recognize Racial Problems

Recognize Racial Problems

Strong Conservative

Mill. R2 = 0.19 ; Others R2 = 0.18

Strong Liberal

Post Grad

Strong Republican

Mill. R2 = 0.14 ; Others R2 = 0.09

Strong Democrat

No H.S.

2

Mill. R = 0.00 ; Others R = 0.00

2

Recognize Racial Problems

8.1.  FIRE and Its Correlates

Strong Conservative

Recognize Whites' Advantage

Strong Liberal

Post Grad

Strong Republican

Mill. R2 = 0.26 ; Others R2 = 0.31

Strong Democrat

Are Fearful of Other Races

Strong Republican

Mill. R2 = 0.02 ; Others R2 = 0.02

Ideology

Strong Democrat

Recognize Whites' Advantage Mill. R2 = 0.26 ; Others R2 = 0.25

No H.S.

Are Fearful of Other Races

Post Grad

2

Mill. R = 0.04 ; Others R = 0.05

2

Recognize Whites' Advantage

Mill. R2 = 0.02 ; Others R2 = 0.00

Partisanship

No H.S.

2

Mill. R = 0.00 ; Others R = 0.02

2

Are Fearful of Other Races

Education

Are Angry about Racism

Are Angry about Racism

Strong Conservative

Mill. R2 = 0.09 ; Others R2 = 0.13

Strong Liberal

Post Grad

Strong Republican

Mill. R2 = 0.11 ; Others R2 = 0.09

Strong Democrat

No H.S.

Mill. R2 = 0.00 ; Others R2 = 0.00

Are Angry about Racism

4 3 2 1 0

4 3 2 1 0

4 3 2 1 0

Millennials

Older Whites

Group

Millennials

Older Whites

Group

Millennials

Older Whites

Group

232 / Chapter Eight

Republican.” Here, we see that as Whites move out of the Democratic Party, they are also less likely to acknowledge persistent racial problems; across the two extremes is a difference of about 1.15 points (p < 0.001). The most strident Republicans are significantly less angry about racism than their Democratic counterparts and express more fear of other racial groups. Additionally, Democrats score over two full points higher—­that’s about half the scale!—­on recognizing White privilege, compared to strong Republicans.4 The third row in figure 8.1 tells a similar story when we contrast those who identified as strong liberals against those who identified as strong conservatives. Conservatives tend to be about 0.60 points more fearful, one full point less angry about racism, 1.50 points less likely to recognize racism is a problem, and over 2.25 points less likely to agree with the statement con­cerning racial privilege. In terms of absolute magnitude, the smallest t-­statistic of these four differences is well above any threshold for statistical significance (t = −23.03, df = 7,630, p < 0.001). While partisanship, political ideology, and racial attitudes are linked, based on the R2 terms, only about one-­quarter of the variance in any of these items can be explained by either partisanship or ideology. To be more specific, the correlation between ideology and “fearfulness of other races” is 0.15. Meanwhile, the correlation between partisanship and racial fear is even lower: 0.12. The measures of racial empathy, while more highly correlated, are still what most would consider “small” correlations (−0.33 for partisanship and −0.36 for ideology). This observation is of particular import because it shows that what we are measuring is not as enmeshed with partisanship or ideology as previous measures are. For instance, the correlations between racial resentment and the demographic characteristics we analyzed here are nearly twice as high.

Is FIRE a Good Measure? As mentioned, a good measure ought to meet a number of important criteria, including face, discriminant, convergent, and predictive validity. Additionally, we should be able to show the added value of this new measure. We have harped on the notion that racial attitudes are multidimensional and argued that a measure that mimics this characterization would provide more nuance into our understanding of the mechanisms that shape Americans’ responses to issues of race, racism, and racial inequity. In the remaining portions of this chapter, we execute several tests to determine the validity of the FIRE battery and demonstrate one of the comparative advantages that this measure has over other, unidimensional measures of

The FIRE This Time  /  233

racial attitudes. The analyses below show that not only do the FIRE items provide greater predictive power on issues of concern, but also, because we have individual measures for the two dimensions of  Whites’ racial attitudes, we can see which dimension triggers or influences other related attitudes, policy preferences, candidate evaluations, and the like. Discriminant Validity: Interracial Marriage While many “old-­fashioned racism” and social-­distance questions have been done away with, one that still remains poignant relates to how comfortable an individual would be with a member of their family marrying someone of a different race. Scholars have shown that one’s attitude about interracial marriage correlates with racial resentment, but the theoretical paradigm that birthed racial resentment does not help us to fully flesh out the link between the two. Theoretically, racial resentment arises where anti-­Black animus intersects with a belief that Blacks are violating traditional American norms and values. But what is it about one’s beliefs about the “Irish, Italians, and Jews overcoming prejudice and working their way up” or the idea that “if Blacks tried harder, they could be just as well off as Whites” that would shape one’s ideas about interracial marriage? The second challenge that traditional measure presents is that because this measure itself is a blend of attitudes, it is very difficult to parse out the effects of its component parts. One ought to expect that someone’s opposition to interracial marriage might be rooted in fear rather than in abstract, moralist resentments of Blacks; the FIRE measure allows us to test that directly. The FIRE items allow us to determine whether the same variable that was highly related to old-­fashioned stereotypes of violence, most notably the fear of racial minorities, would be of particular importance in understanding how Whites think about interracial marriage. To test this hypothesis, we estimated four statistical models that examine the underlying determinants of agreement with the statement “It would make me feel uneasy if a close relative of mine was planning to marry someone who is a member of another race.” This de­ pendent variable is coded to run from 0 (strongly disagree) to 5 (strongly agree). The results are presented in table 8.2. The first two columns in table 8.2 report the effects that the racial resentment scale and the four FIRE items have on opposition to interracial marriage; for ease of comparison, all of the independent variables have been coded to run from 0 to 1. Model 1 shows that racial resentment is statistically significantly related to the attitude. Model 2 reveals that all of our FIRE items are statistically related to this attitude, and the item measuring

234 / Chapter Eight Table 8.2.  Predicting Opposition to a Close Family Member Marrying Someone Who Is Not White Model 1 Racial Resentment

Model 2

1.66** (0.21)

F: Fear of Other Races (IR) Acknowledges Whites’ Advantages (IR) Recognizes Racism Is a Problem E: Angry about Racism

Model 3

1.89** (0.17) −0.52** (0.15) −0.55** (0.19) −0.80** (0.22)

0.40 (0.28) 1.86*** (0.17) −0.40* (0.20) −0.41** (0.19) −0.80*** (0.23)

1.97*** −0.19 633 0.25

1.60*** −0.33 617 0.26

Conservative Ideology Intercept N adj. R2

0.37*** −0.12 689 0.08

Model 4 0.30 (0.29) 1.92*** (0.19) −0.42* (0.21) −0.45* (0.20) −0.59** (0.23) 0.25 −0.22 1.38*** −0.37 593 0.27

Data from 2016 CCES. All independent variables are scaled to go from 0 to 1; the dependent variable ranges from 0 (strongly disagree) to 5 (strongly agree) with the statement that “it would make me feel uneasy if a close relative of mine was planning to marry someone of a different race.” Notes: Weighted OLS estimates for non-Hispanic White respondents only. Standard errors in parentheses. *p < 0.05, **p < 0.01, ***p < 0.001

fear of other races has the largest coefficient. Model 3 includes all four of the FIRE items as well as the four-­item racial resentment scale. Here, we see that racial resentment does not maintain statistical significance, but all of the FIRE items do, with the “fear” component of FIRE having the largest effect. Finally, model 4 is a fully specified model. Here we see that even when we include other measures such as political ideology, not only does racial resentment lose its statistical significance, but also the fear factor of the FIRE battery reigns supreme, as we hypothesized. This set of models provides evidence for both discriminant and predictive validity. What we like about the FIRE battery is that it allows us to isolate particular domain-­ specific mechanisms of influence. What’s more, it provides greater predictive value than traditionally relied-­on measures of racial attitudes. Racial resentment, alone, explains about 8 percent of the variance (R2) in Whites’ sentiments about interracial marriage. The FIRE battery, in contrast, explains over three times the variance that racial resentment does, even though both are composed of the same number of questions; the fear item alone can explain twice the variance that the racial resentment scale can. What’s more,

The FIRE This Time  /  235

we see that when racial resentment is accounted for alongside FIRE, it does not provide any added predictive benefit. Together, these models demonstrate that expanding our measures to include multiple components of racial thinking allows us to gain a better understanding of which aspects influence White Americans’ racialized policy preferences. Put simply, if a researcher were to only estimate model 1, they might conclude that racial resentment is the primary force driving opposition to interracial marriage; this is known as a type I error. By parsing out the dimensions, we are able to home in on the specific attitudinal mechanisms that can explain various political attitudes and behaviors and may also work to reproduce inequality. Predictive Validity: FIRE in the White House As you may know, we are political scientists. We began this project because we recognized that racial attitudes play a major role in Americans’ political attitudes, policy preferences, and candidate evaluations. We simply wanted to develop a measure that more accurately reflects America’s contemporary, dominant racial ideology and racial grammar. Now that we’ve developed a more holistic measure that measures what we think it measures and doesn’t measure what it shouldn’t (see above), we take the time to show its predictive power as it relates to presidential vote choice. There has been a great deal of debate about whether racial attitudes, economic woes, or just simply feeling “left out” drove so many White Americans to support a candidate who had no political experience, filed for bankruptcy on multiple occasions, and clearly spouted racist, homophobic, Islamophobic, and sexist rhetoric. Our concern here is not to adjudicate between all the possible factors that ignited people to support Donald Trump’s candidacy, but to discern the extent to which various components of people’s racial attitudes might have influenced their voting decision. Nevertheless, given that we can examine whether and how racial fear, empathy, and acknowledgment of structural racism separately influence Whites’ political behavior, we actually are able to provide some new insights into Americans’ choices in the 2016 primary and general elections. Adding Fuel to the FIRE: Descriptive Statistics We begin with how Whites voted in the 2016 primaries, given their importance. Figure 8.2 shows the point estimates for each of our four items across

Mean

0

1

2

3

4

Other Voters

Trump Voters

Democratic Voters

Trump Voters

Democratic Voters

Other Voters

Trump Voters

Recognize Racial Problems

8.2.  Racial Attitudes and Primary Voting in 2016

Other Voters

Recognize Whites' Advantage

'Other Voters' are those who reported voting for either Ted Cruz, John Kasich, Marco Rubio, Another Republican, or someone else besides Hillary Clinton, Bernie Sanders, or another Democratic candidate in the 2016 primaries.

Source: 2016 CCES Common Content N = 28,365 Non−Hispanic White Primary Voters

Democratic Voters

Are Fearful of Other Races

Democratic Voters

Other Voters

Are Angry about Racism

Trump Voters

The FIRE This Time  /  237

three different groups: Whites who voted for Donald Trump (“Trump voters”), Whites who voted for another Republican primary candidate (“other voters”), and Whites who voted in a primary but not for a Republican (“Democratic voters”). As figure 8.2 clearly shows, Trump voters scored significantly higher than other GOP voters on three of the four items. Not only were they less empathetic (less angered by racism), they were more likely to deny that Whites have racial privilege and to express far more fear of other racial groups. Democratic primary voters, on the other hand, scored far better on every single item: lower on fear, higher on recognizing institutional racism, and higher on racial empathy. How do White millennials compare to other primary voters? While about 52 percent of  White millennials reported participating in the primaries, and 58 percent of these respondents reported voting for a Democrat in the primaries, we can see that those White millennials who did not vote for a Democrat are significantly lower in terms of their recognition of institutional racism and their racial empathy. Figure 8.3 shows the point estimates and confidence intervals (which are essentially contained within the size of the points) for each of our four measures across the three types of voters: those who voted for a Democrat, those who voted for Donald Trump, and those who voted for another candidate. What is interesting to see in this figure is that those millennial Whites who voted for Trump in the primaries closely resemble older Whites who voted for another Republican candidate. As before, those who voted for Democrats are less fearful, more likely to recognize institutional racism, and generally angrier about racism. Among Democratic voters, White millennials are statistically indistinguishable from older Whites. Now we turn to how these items relate to voting behavior in the general election. As we did before, we can look at the averages of our four items across the generational divide. In figure 8.4, we can see that there are stark differences between those Whites who voted for Hillary Clinton and those who helped elect Donald Trump. While Trump’s White millennial supporters are more likely to recognize Whites’ advantages and are significantly angrier about racism than his older voters, they still lag far behind Clinton voters (regardless of age). What is most striking about our White millennial voters is that, compared to older Whites, White millennials are simulta­neously more likely to recognize that Whites have advantages and more likely to think racial problems are rare. So, while both Clinton’s and Trump’s youngest White supporters recognize institutional racism exists, they are less likely to think racial problems happen with any regularity.

Mean

0

1

2

3

4

Trump Voters

Democratic Republican Voters Voters

Democratic Republican Voters Voters

Trump Voters

Recognize Racial Problems

8.3.  Racial Attitudes, Generation, and Primary Voting in 2016

Trump Voters

Recognize Whites' Advantage

'Other Voters' are those who reported voting for either Ted Cruz, John Kasich, Marco Rubio, Another Republican, or someone else besides Hillary Clinton, Bernie Sanders, or another Democratic candidate in the 2016 primaries.

Source: 2016 CCES Common Content N = 28,365 Non−Hispanic White Primary Voters

Democratic Republican Voters Voters

Are Fearful of Other Races

Democratic Republican Voters Voters

Trump Voters

Are Angry about Racism

Millennials

Older Whites

Group

Mean

0

1

2

3

4

Trump Voters

Source: 2016 CCES Common Content N = 31,290 Non−Hispanic Whites.

Clinton Voters

Are Fearful of Other Races

Trump Voters

Clinton Voters

Trump Voters

Recognize Racial Problems

Clinton Voters

Trump Voters

Are Angry about Racism

8.4.  Racial Attitudes, Generation, and the 2016 General Election

Clinton Voters

Recognize Whites' Advantage

Millennials

Older Whites

Group

240 / Chapter Eight

Predictive Validity: Trump vs. Clinton As our final set of analyses, we test the predictive validity of the FIRE items on one of the most consequential US elections in recent history. Here, we estimate statistical models predicting whether a respondent would vote for Trump or Clinton in the general election. Over half (54 percent) of the 31,600 non-­Hispanic White respondents who reported voting for one of the major party candidates selected Trump. We collapse vote choice into a dichotomous dependent variable where a 0 means a vote for Clinton and a 1 a vote for Trump. We then use logistic regression to predict a Trump vote while including standard demographic controls: partisanship, ideology, age, sex of respondent, household income, level of education, and perceptions of the economy. Instead of presenting tables of logistic regression estimates, we present the marginal effect each item has on the predicted probability of voting for the Republican candidate, with the other explanatory variables set at their central tendencies.5 The panels in figure 8.5 show visually weighted regression estimates, where a wider line indicates less certainty about our estimates. As we can see, White voters who were high on fear and low on the other three items were predicted to vote for Donald Trump. Even those who had the lowest level of racial fear would be predicted to vote for him (because most White Americans voted for Trump). However, as we can see in the three rightmost panels, as acknowledgment of  Whites’ advantages, recognition of racial problems, and racial empathy all increase, the predicted probability of voting for Trump falls precipitously. In order to get a sense of the magnitude of these effects, we can hold all the independent variables at their central tendencies and then vary our FIRE items along their interquartile range. Moving from the first quartile to the third quartile on racialized fear changes the predicted probability that a respondent votes for Donald Trump from 0.43 to 0.68. For the other three variables, an increase in each variable leads to a significant decrease in the predicted probability. This shift goes from 0.61 to 0.23 for acknowledging Whites’ advantages and from 0.64 to 0.35 for recognizing racial problems, and as racial empathy moves along its interquartile range, the predicted probability of a Trump vote falls from 0.61 to 0.40. Finally, we examine how Trump’s unique campaign might have been manipulating the racial attitudes of Democrats to motivate them to vote for him as opposed to Hillary Clinton. We estimated the same model as above but restricted the sample to Whites who had reported voting for President Obama in 2012. This brought our total sample size to just over fourteen

8.5.  FIRE in the 2016 Presidential Election

8.6.  FIRE and Trump’s Popularity among White Obama Voters

The FIRE This Time  /  243

thousand respondents, 12 percent of whom reported voting for both Obama and Trump. Again we present the marginal effects of the FIRE items, holding demographics and perceptions of the economy at their central tendencies. Figure 8.6 reveals a large and significant effect for each of the four variables. With the exception of racialized fear, moving from the low end to the high end of each variable significantly changes the predicted probabilities. For those Whites who voted for Obama, an increase either in acknowledging institutional racism or in racial empathy makes it far more likely that they would vote for Clinton. In the three rightmost panels, we can see that moving from low to high scores on the three FIRE items brings the predicted probability of voting for Trump from well above 0.50 to at or below the 0.50 mark. Thus, we would predict that those White Obama voters who were particularly aware of racial problems in the United States and were angry about them were the least likely to defect.6 These findings put a great deal of what we have seen in the run-­up to and aftermath of  Trump’s election into perspective. Again, because we do not add together the responses of the four-­item battery, we are able to pinpoint the individual effects of each component of racial attitudes. As such, we can say with confidence, for example, that a fear of racial others played a critical role in Trump primary voters’ political views. In the aftermath of the election, there was a discussion about how Whites felt left out and disadvantaged; we are able to provide empirical data for this idea. Both White Democrats and White Republicans who were not willing to acknowledge aspects of  White privilege were much more likely to support Donald Trump in the primaries and in the general election. Similarly, we see that each component of the FIRE battery provided additional information as to why many White voters made such a dramatic shift in vote preferences—­from Obama to Trump. Hopefully, with this multidimensional measure we can develop more accurate predictions for future elections, more of which are likely to include emboldened candidates like the forty-­fifth president as well as candidates of color across the ideological spectrum.

Earlier, we mentioned that one question that has haunted scholars of American politics is the one concerning the extent to which racial attitudes continue to play a role in American politics. It is well documented that racial animus has had a central role in the shape of White Americans’ political attitudes, but it turns out racial animus was only one part of the problem. White Americans’ contemporary racial attitudes have a wider set of components than we typically think about. Yes, racial prejudice matters, but White

244 / Chapter Eight

Americans’ understandings of White racial privilege and structural racism and their feelings about these issues and out-­group members are important as well. The measure that we put forth here allows us to see that more clearly. What’s more, the items of the FIRE battery allow us to parse the effects of each of these components. Several findings in this chapter alone tell us a great deal about how White Americans’ political attitudes are shaped by their perspectives on race, racism, and underrepresented groups—­both broadly speaking and specific to an incredibly consequential presidential election. Generally speaking, our results in this chapter also highlight the issue of racial stasis, as signs of the countervailing forces remain visible. For instance, White millennials do present more progressive attitudes. Compared to their predecessors, they are more likely to express anger about racism and more likely to acknowledge their privilege. But we also found that nearly one in five White millennials (20 percent) simultaneously feels angry that racism exists and does not believe Whites have advantages because of the color of their skin. Even with this short battery of measures, we are able to see not only that a large chunk of young Whites abhor racism but also that they have a constrained understanding of what racism is. Ultimately, we have a large number of walking contradictions in American society that are help­ ing to produce and perpetuate ongoing racial inequities through their political stances and policy preferences. A great deal of ink has been spilled concerning whether Trump voters are racist. Without getting bogged down in a debate about who exactly is “racist,”7 our data clearly show that racial attitudes played an important role in both the primary election and the general election in shaping the White electorate’s response to a Trump presidency. Because we do not add together the responses of the four-­item battery, we are able to pinpoint the individual effects of each component of racial attitudes. As such, we can say with confidence, for example, that a fear of racial others played a yuge role shaping Trump’s supporters’ preferences in the primaries. What’s more, in the aftermath of the election, pundits conjectured that Whites, especially those who left the Obama coalition, were inclined to vote for a so-­called populist because they felt left out, left behind, and disadvantaged due to economic decline (in many working-­class areas of the country) and a demographic shift that they perceived would threaten their culture and stance in society (Norris 2018). We are able to speak to this notion, in some limited way. We found that both White Democrats and White Republicans who are in denial about how awesome it is to be White were much, much more likely to support Donald Trump in the primaries and in the

The FIRE This Time  /  245

general election. Relatedly, we found that racial attitudes likely played a role in turning “blue” stronghold states “red,” as our results reveal each of the four components of racial attitudes we highlight led many Democratic voters to switch to Trump in 2016. Many people have relied on “I voted for Obama” as a shorthand for “I’m not racist,” but our data present clear limitations for this declaration.

C o n clus i o n

Is Resuscitation Possible?

When there were exactly seventeen people attempting to clinch the GOP presidential nomination, Candis, in the self-­identified liberal bastion of Chapel Hill, North Carolina, noticed that students as well as average citizens seemed to be “Feeling the Bern” or were “with Her.” Meanwhile, Christopher noticed that a surprising number of young people in Bloomington, Indiana, leaned more to the right. One day, while he was minding his own business, he came upon this truck, and photographed it. Not only had the owner of this pickup placed a very large Confederate flag decal in the rear window, but the toolbox in the bed had another Confederate flag and a “Ben Carson ’16” bumper sticker (figure C.1). We doubt that the placement of these stickers was done with some sense of irony. The person who placed these decals on their vehicle, without a sense of cognitive dissonance, could somehow both support a Black man’s effort to become the president of the United States and memorialize the Confederacy’s attempt to maintain an institution that reduced that same man’s ancestors to property. Ben Carson, although a Black man, does not represent a threat to the racial status quo. During his campaign, Dr. Carson made no appeals to reduce racial inequality. Carson is a man who believes that Harriet Tubman should not replace Andrew Jackson on the twenty-­ dollar bill because Jackson was a “tremendous president” (as Carson said in a 2016 interview), and as Housing and Urban Development secretary, Carson has proposed raising rents for low-­income Americans as the wealthy get a tax cut. At first glance, the juxtaposition of these political symbols and ideas seems contradictory, but it is not. As the number of Republican primary candidates began to dwindle, pollsters predicted that in a hypothetical Clinton versus Trump contest, Clinton would win out, in part because millennials—­forming almost one-­third of

248 / Conclusion

C.1.  Countervailing Forces on Bumper Stickers

the voting-­eligible population—­would “choose Clinton by a crushing 52%–­ 19%, a preference that crossed all demographic lines. Among whites, she’d be backed by nearly 2–­1, 45%–­29%” (Page and Ung 2016). Trump ran an explicitly racist (and sexist and Islamophobic) campaign—­referring to undocumented immigrants as “rapists and murderers,” asserting that Muslims should be banned from entering the United States, suggesting that Japanese internment was the right thing to do, promising to pay the legal fees of one of his White supporters who had punched a Black Lives Matter protester, equivocating on whether he would enjoy the support of former Ku Klux Klan grand wizard David Duke, and so on. People assumed that Americans would reject this kind of rhetoric, with millennials leading the pack.

Is Resuscitation Possible?  /  249

In the end, millennials did give the majority of their votes to Clinton. According to a CNN exit poll, 56 percent of voters between the ages of eigh­ teen and twenty-­four voted for Clinton, in comparison to 34 percent for Trump; among those ages twenty-­five to twenty-­nine, the vote split was 54–­ 38 percent in Clinton’s favor, and those between thirty and thirty-­nine gave 51 percent of their votes to Clinton and 39 percent to Trump (CNN Pol­itics 2016). But, these statistics are for all millennials, across all demographic groups. Among Whites aged eighteen to twenty-­nine, 47 percent gave their votes to Trump, while 43 percent reported voting for Clinton (in comparison, 85 percent of Blacks in that age range voted for Clinton). This was unexpected by the media and polls, but perhaps it should not have been. Months after the 2016 presidential election, several hundred White supremacists, who felt emboldened by the new president’s overtly racist rhetoric, moved out of the shadows of the internet and marched unmasked on the campus of the University of Virginia, in Charlottesville. The group had a large contingent of White millennial men, who chanted racist and anti-­ Semitic slogans and brandished Nazi symbols and Confederate flags. Their goal was twofold—­bring together various iterations of White supremacists and protest the removal of a Confederate statue. There was a great deal of outrage well after the Unite the Right march ended, in large part because a White millennial woman named Heather Heyer was murdered by a millennial White supremacist. She was protesting hate; he was spewing it. A march like this, which was initiated and organized in large part by a White millennial, was unexpected by most Americans, but perhaps it should not have been. These kinds of on-­the-­surface contradictions and the bombardment of observations of both blatant and complicit racism that we have noticed since the two of us met in graduate school motivated us to take a deeper look into the racial attitudes of  White people, with a special focus on White millennials. Our data—­both qualitative and quantitative—­and a multimethodological analytical strategy allowed us to develop a theory of countervailing forces, test a hypothesis of racial stasis, and create a measurement that provides a more holistic and nuanced depiction of White America’s contemporary racial attitudes. In this concluding chapter, we briefly review the major findings in this book and discuss their implications for American politics and American society. Before closing, we briefly touch on three matters of import that we did not discuss here but deserve their own 320-­page books: millennials of color, implicit attitudes, and Generation Z—­the post-­ millennial generation.

250 / Conclusion

Major Takeaways: The More Things Change . . . We began this book by suggesting that America is in a state of racial stagnation. While attitudes toward members of lesbian and gay communities and women and the preferences concerning policies that affect and protect members of these groups seem to be chugging along in a positive direction, the same cannot be said for Whites’ attitudes toward racial equity and about Blacks in particular. Although previous research shows that young people have historically worked to move the United States toward more progressive attitudes concerning marginalized and underrepresented minority groups, our research reveals that the same cannot be said for White millennials—­or at least not to the same degree. We developed a theory of countervailing forces to help us better understand this phenomenon. This theory asserts that even though there are factors that help White millennials shift society’s racial attitudes in a progressive direction, there are just as many factors that actively work against such a shift, thus producing what we call “racial stasis.” Through our interviews, we found that while White millennials are invested in bringing to fruition an egalitarian society—­one where race does not influence an individual’s life chances—­they do not necessarily know how to do this, because they primarily see racism as overt displays of bigotry perpetrated by individuals at the micro-­level. That is to say, more and more Whites tend to understand racism as a set of attitudes and individual behaviors marked by racial bias and prejudice, rather than as a larger, systemic, and structural set of systems, policies, laws, and ways of being (e.g., White habitus, White fragility), many of which do not need any particular person to keep them going (Bonilla-­Silva 1997; Omi and Winant 1994; Bonilla-­ Silva, Goar, and Embrick 2006; DiAngelo 2018; Feagin 2013). Thus, our research adds to the body of literature that shows that when (systemic) racial inequalities are pointed out to Whites, they tend to use color-­blind, race-­ neutral, or class-­based explanations to rationalize the existence of persistent disparities. This is especially true when Whites focus on those inequalities that concern facts regarding their own lives (i.e., the friends they have chosen, the racial segregation within their neighborhoods and schools). Here, we found that because White millennials are not able to conceptualize the role that race plays in American society more broadly, they become passive travelers on the moving walkway of racism, rather than actively walking in the opposite, anti-­racist direction. Additionally, this book contributes to the growing critical diversity literature (Bell and Hartmann 2007; Berrey 2015a, b; Mayorga-­Gallo 2019; Warikoo and de Novais 2014). We found that young White people place a

Is Resuscitation Possible?  /  251

high value on diversity, but they (a) fail to live diverse lives (or in other words, they maintain racially homogenous social networks), and (b) primarily value diversity as a benefit and commodity for themselves rather than as a signal of a more equitable institution or inclusive social space. These are serious consequences for what scholars are calling diversity ideology. To reiterate, research shows that racial isolation serves to perpetuate (negative) stereotypes about groups with which people do not normally come in contact (Bonilla-­Silva, Goar, and Embrick 2006; Mayorga-­Gallo 2014). Part of the reason why we see that Americans have increasingly positive attitudes toward members of lesbian and gay communities and are also more supportive of policies like gay-­marriage and anti-­discrimination laws is because people across various demographics have interpersonal contact with lesbian and gay individuals (Becker 2012; Herek and Glunt 1993). We do not see as much interpersonal contact between members of different racial groups,1 but when we do, there is potential for a great deal of change. Political scientists Tatishe Nteta and Jill Greenlee found that among young Americans, the effect of having interpersonal contact with African Americans leads to a seventeen-­point (on a hundred-­point scale) decrease in support for the belief that African Americans are unintelligent (2013, 888). What’s more, when we get results like the ones we did in chapter 3, that show that young people’s racial sentiments are highly related to old­fas­hioned bigotry, it becomes clear that racial isolation has major, negative consequences for the shape and effect of White Americans’ racial attitudes. It is also discouraging (as we found in chapter 7) to see that, in terms of the two dimensions of racial attitudes, younger Whites’ lack of empathy for racial minorities is a stronger predictor of old-­fashioned racist attitudes than is the cognitive component meant to reflect an understanding of racial inequality. Our findings are depressing, but even more unfortunate is that they are not anomalous. We might read Nteta and Greenlee’s (2013) finding another way: people who do not come into contact with Black Americans are more likely to believe that Blacks are unintelligent. Michael Tesler’s (2013) work shows that old-­fashioned racial attitudes became increasingly connected with partisanship during the Obama presidency. What’s more, Ashley Jardina and Spencer Piston’s work (2016) reveals that Whites see Black Americans as less evolved than Whites. None of this mounting data bodes well for what we, or most reasonable people, would characterize as racial progress. In addition, while Whites have generally not “noticed” their own race or been aware of their racial privilege, we see a new racial consciousness grow­ing. Many Whites now are willing to admit to having White privilege. Presumably,

252 / Conclusion

if Whites are more likely to recognize that they have privilege, then they might also realize this unearned privilege affords them access to opportunities (e.g., better-­off social networks, benefit of the doubt in interactions with police or at a job interview) that people of color do not get. It would not be a far stretch to come to the conclusion that racial equity, then, requires Whites’ support of policies that provide people of color greater access to these privileges. However, what we presume to be a logical set of conclusions doesn’t manifest itself here. Instead, we find in both our face-­to-­face interviews and large-­n survey data that a great majority of White Americans believe that people of color also have racial privilege. In the 2016 CCES, we asked individuals if they support affirmative action for Blacks and Latinx people. A little less than 50 percent of  White millennials responded that they do support such a policy.2 On their face, these data are actually quite uplifting, given that only about a third of older Whites agreed. However, about the same percentage of young Whites believed that Whites should also be beneficiaries of affirmative action! We see that there is a shift in the way that people talk about race and racism, but ultimately, we also see more of the same outcomes—­a value for diversity and egalitarianism is met with a lack of support for policies that bring those values to fruition.

Major Takeaways: Shifting Measurement Strategy Currently, most political scientists rely on the racial resentment scale, which when developed nearly forty years ago allowed Whites to subtly call Blacks lazy and entitled without literally using those words. Our goal has been not only to describe the shifting nature and shape of racial attitudes but also to more accurately measure them and explore their political ramifications, and our findings illuminate a few things about depicting and measuring con­ temporary racial attitudes. First, we argue that White Americans have a new way of expressing their racial attitudes. This is in agreement with many other social scientists, especially those working in sociology and psychology. What we see as the rising dominant racial ideology in the United States—­ well characterized as color-­blind—­relies on a different racial grammar and logic than the concept of racial resentment does. Racial attitudes evolve, so current attitudes are not fully divorced from racial resentment; but, racial attitudes are also more nuanced and multidimensional than traditionally considered. Second, when we view contemporary racial attitudes through a wider lens, we are able to see that the phrase “racial attitudes” is not synonymous with “racial prejudice,” as old-­fashioned racial bigotry is only one kind of racial attitude (Chudy 2017). By taking time to discern the

Is Resuscitation Possible?  /  253

components of the construct commonly referred to as “racial attitudes,” we see that this construct is multidimensional. A more holistic measurement, then, is required to fully appreciate the complexity of White Americans’ racial attitudes. As we uncovered in chapters 6 and 7, there is more to Whites’ racial attitudes than simply resentment. In those chapters, through a series of painstaking analyses, we uncovered that Whites’ racial attitudes might best be described as two-­dimensional. The first dimension corresponds to Whites’ cognitive awareness of racial privilege and systematic racial inequality. The second dimension corresponds to Whites’ empathetic reactions to the role of racism in society and to their level of fear of racial out-­groups. While scholars have conjectured that racial attitudes have various components, including cognitive and affective parts, we believe we are the first to develop a measure that not only taps into multiple dimensions of racial attitudes but also does so in a way that meets the four requirements of a good measure. The FIRE measure succinctly describes and measures contemporary racial attitudes and displays face, convergent, predictive, and discriminant validity. We know that people are drawn to using the racial resentment scale. Not only has it been shown to tap into a coherent belief system, the measure also provides predictive power over a number of policy preferences, political attitudes, and behaviors. Aside from that, it is an appealing measure to use, because we have plenty of data over time that relies on it (including the data we relied on in part 1). We all love a measure that we can track over time, but here’s the problem: this measure was developed almost forty years ago. Kinder and Sears, the creators of the racial resentment battery, went on a quest to find better, more accurate measures of Whites’ racial sentiments because they noticed that people were talking about race matters in a different way around the 1970s; Kinder and Sears didn’t rely on measures developed during Jim Crow. Why should we rely on measures developed in the era immediately after the civil rights movement, given the vast changes that have occurred in the shape of the economy, in the demographic profile of the United States, and in what we collectively deem as socially acceptable norms? What’s more, we have shown that our measure of Fear, acknowledgment of Institutional Racism, and racial Empathy (FIRE) has just as much, if not more, predictive power as racial resentment. FIRE does not predict things that it should not, and because it is not a scale, we can see how each component of White Americans’ racial attitudes individually influences their policy preferences, political attitudes, and behaviors. For White Americans, sometimes just acknowledging institutional racism shapes one’s attitudes; other times, an attitude might be rooted in one’s

254 / Conclusion

fear of other races. The two-­dimensional structure is important to acknowledge also because it was derived to be practically (and statistically) independent of partisan and ideological leanings. In the 2016 CCES, our mea­ sure of racial fear among Whites is correlated with ideology at 0.15 and with partisanship at 0.10. The racial resentment questions are correlated with those same measures at 0.55 and 0.50, respectively. To be sure, we embark on this task with a sense of intellectual humility. We know that there will be plenty of debates around whether FIRE is the most valid or accurate measure out there. In fact, as scholars, we are encouraged by that. Prompting a debate and inspiring our discipline to move toward developing a more appropriate, accurate, and nuanced measure of racial attitudes are the underlying goals of our endeavor.3 We simply say that it is time to adopt a new measure of racial attitudes; we offer a proposed starting point with a surfeit of evidence to justify it.

Some Caveats There are three caveats that we want to consider prior to closing this book. The first caveat is that we have focused on explicit racial attitudes. Explicitly communicated attitudes are just the tip of the mental iceberg. Psychologists show that much of our thinking happens at an unconscious level, and although political scientists rarely use measures of implicit attitudes and bias, Efrén Pérez (2016) reveals that a great deal of our political thinking is influenced by unconscious thought. Pérez’s research shows that biased attitudes are not equally distributed about the population; instead, there are differences in levels across racial groups and differences in effects across levels of education. For example, Pérez found that educated White Americans are affected the most strongly by their negative, implicit biases on the issue of immigration. While Racial Stasis focused on finding measures that accurately capture what people think about race and politics, we would be remiss to suggest that our proposal for a two-­dimensional measurement strategy is the only way to measure racial attitudes, when so many of our preferences and our behaviors are dictated by our unconscious mind. It is important for political scientists to update measures when ways of thinking change and when expressions of various attitudes go in and out of vogue; but it will also be important for us to use multiple measures of racial attitudes, tapping both implicit and explicit sentiments. The second and third caveats are very closely related: millennials of color and Generation Z are two groups that we do not focus on in our analysis here. In a very early draft of this book, we received a criticism that our focus

Is Resuscitation Possible?  /  255

on young Whites is misguided in the face of the demographic transformation that the United States is experiencing right now. Early on, when we asked our readers to embark on this intellectual journey with us, we explained that we focus on Whites because we do not expect the demographic shift to result in a one-­to-­one shift in economic or political resources. It is certainly not the case that as non-­Whites become the numerical majority in the United States, Whites will automatically cede their wealth and resources to them. In a later draft of this book, we were asked, What about Generation Z? They seem to be doing something right in this post-­Ferguson world we’re living in. We take these points in turn. To the first matter at hand, we argue that instead of implying that “demographics is destiny” or suggesting that people of color’s attitudes would mute those of Whites as the population becomes less White, a more productive critique would be one that notes that White attitudes do not exist in a vacuum. Further, it would point out that it is the dialectic push and pull between Americans across racial and ethnic groups that shapes our politics and policies. To be sure, we do not mean to imply that Whites’ attitudes are the only attitudes worthy of study, but given the balance of power in America’s racialized social system, White attitudes ought to be carefully tracked and measured, which is what this book aims to do. Nevertheless, it is certainly important to consider the role people of color have in shaping America’s racial politics. Research on Black politics reveals that Blacks’ policy preferences have actually become more conservative over time (Tate 2010). What’s more, Blacks have become less likely to rely on structural inequality and racial discrimination as explanations of the Black-­White socioeconomic gap (Smith 2014). Further, scholars show that people of color, including Latinx and Asian Americans, rely on a color-­blind racial ideology to explain away racial inequalities just as Whites do, although not to the same extent (Bonilla-­ Silva 2014; O’Brien 2008; Vega 2014). We would like to note that people of color are socialized in the same milieu as Whites. Dozens of  YouTube clips, for example, reveal that young Black and Brown children still do not like the doll that looks like them more than they like the White doll. Children of color are washed over by the same media images, messages, and ideas that provide a foundation for America’s racialized social system as Whites are. When asked if the lives of Black women, lesbian and gay Blacks, transgender Blacks, and formerly incarcerated or undocumented Black immigrants matter equally, Black people are reluctant to equally throw their support behind all of these groups (Lopez Bunyasi and Smith 2019a). In all, what these kinds of results suggest is that an increasing number of Black, Latinx,

256 / Conclusion

and Asian Americans in the American population may potentially but not necessarily lead to an aggregate move to the left on racial issues or on policies aimed to ameliorate every racial inequality. Nonetheless, it must be noted that young people of color have taken a strong stance against ongoing structural inequalities in the forms of the Black Lives Matter movement (Lebron 2017; Ransby 2018; Taylor 2016), ac­ tivism for the DREAM Act, and activism against anti-­immigration policies (Zepeda-­Millán 2017; Zepeda-­Millán and Wallace 2013), as well as in the form of protests and sit-­ins focused on more local issues, such as racial inequality on elite college campuses. With the help of social media, young people of color have forced all Americans to grapple with these issues. Black Lives Matter protesters have essentially made it a requirement for Democratic presidential candidates to take a stance on issues of mass incarceration, police brutality, and even reparations. College students at the University of North Carolina, for example, have forced their administration to strip the name of a Ku Klux Klan member from the moniker of a building on campus. Students at Duke University worked to raise the hourly minimum wage of mostly Black and Latinx workers to twelve dollars an hour by mounting a sit-­in in the president’s office for nearly a week. Black students at the University of Missouri protested in response to issues of workplace benefits, faculty diversity, and the well-­being of marginalized students on campus, and the Mizzou football team threatened to refuse to play; their protest would have influenced a multimillion-­dollar, televised game. In response, both the president of the University of Missouri system and the chancellor resigned. Taken together, it appears that millennials of color may have their own set of dueling attitudes and racial ideologies that should be seriously considered. We do live in a post-­Ferguson world. This is a world in which Black millennials are not only leading a social movement aimed to produce more equitable outcomes for Black people who face various types of challenges due to interlocking systems of oppression, but also inspiring other movements and massive protests, including the Women’s March on Washington, the largest day of protest in American history; the #MeToo movement, which aims to highlight the prevalence of rape culture and reduce the frequency of sexual harassment, misconduct, and assault; and #NeverAgain, a student-­ led movement aimed at increasing gun regulations that came to the fore after an armed young man killed seventeen people at a high school in Parkland, Florida. But we also live in a post-­Charlottesville world. This is a world in which some White millennials are leading a countermovement to vie for White

Is Resuscitation Possible?  /  257

supremacy in the face of the changing demographics of the United States and have the unmitigated gall to host a second rally one year after the first ended someone’s life. This is also a world in which people, including folks like Oprah Winfrey, laud the activism of the White (and White-­passing) students of Parkland but criticize Black millennials for taking part in the leaderful contemporary movement for Black lives—­or worse, ignore the same demands for more effective gun regulation that are made by young Black people across the country. So, what are our expectations for Generation Z, those who are now being socialized in this cacophony of dueling movements? Let’s put it this way: if we find in some years’ time that Generation Z fails to get the racial-­equality needle moving at a faster clip, the responsibility will be ours. We fail our children, the next generation, when we do not guide them to better understand the mechanisms and shape of racial inequality. When people believe that it is better to not talk accurately and in an informed way about racism because they believe these conversations are divisive, we are in trouble. We see this in our classrooms all of the time, and we see how misconstrued facts have real implications for policies aimed at reducing racial inequality. People believe that intentional avoidance of race talk will lead to better outcomes, but informal and formal bans on talking about race actually serve to perpetuate racial inequalities (Joseph 2015; Marx 1998; Sawyer 2005). For instance, as we have shown, when people believe that affirmative action is synonymous with “quotas,” which it is not, they also believe that this policy amounts to “reverse racism,” and thus they are unwilling to support those kinds of policies. Or as another example, too many Whites still believe that “Black culture” is the main reason why Blacks lag behind Whites on important socioeconomic indicators. Data show that Blacks do not have different values around education, hard work, and family. Instead, being poor and Black just has different and much direr consequences than being poor and White (Darity 2011). An open, intentional, informed conversation starting with an explanation of structural inequality would allow us to address (or even debate) these issues, but research reveals that the average White parent does not talk to their children about race or racial inequality. Instead, White parents and teachers opt to teach White children to be “color-­blind” or not focus on racial differences at all (Hagerman 2018; Underhill 2018; Lewis 2001). What happens in the void is that negative stereotypes and racial misinformation go unchecked in the great majority of White households. Because people are reluctant to point to the ways in which racism shapes our society, young people have to find some other reason to explain racial phenomena

258 / Conclusion

such as residential segregation in housing and schools, mass incarceration, food deserts, or disparate health outcomes. What typically happens is that (a) people’s stereotypes about the groups that they do not interact with become hardened, and (b) people will look for nonracial explanations (e.g., class) or lean on neoliberal policy proposals (which focus on individual behaviors and not larger structures) to “solve” the problem. This is not working out for us now, and we have no reason to believe that this strategy will work for Generation Z. It is argued that education is the key; perhaps so, but we have to be intentional about what we teach this next generation. We cannot leave it up to young people to (mis)inform themselves. At Howard University’s 2016 commencement, President Obama noted that all the graduates had “a computer in [their] pocket that gives [them] the world at the touch of a button.” With this ability to constantly and instantly access information about the racial status quo, how is it that some Whites still believe that non-­Whites have systematic racial privilege? Here’s something to consider: On March 23, 2016, Microsoft introduced Tay, an AI (artificial intelligence) chatbot, to Twitter. They shut her down within twenty-­four hours because she became an anti-­Semitic, racist, sexist, homophobic teenager. Microsoft claims that Tay had been “attacked” by trolls, which is true. Within just a few minutes of Tay being online, people began to suggest to Tay that Jews were responsible for 9/11 and the like. Some of the time Tay just repeated what was said to her, but in addition to that, these trolls triggered Tay to search the Internet for source material to include in her replies. Journalist Paul Mason explains, “Some of  Tay’s most coherent hate-­speech had simply been copied and adapted from the vast store of antisemitic abuse that had been previously tweeted” (Mason 2016). What does this have to do with anything? This story reveals what we are up against. If any of us search the internet for “professional hair,” “CEO,” “welfare family,” “drug dealer,” “beautiful person,” or a host of other things, the images and information that are attached to the positive words are generally of White people, while the negative words will reveal Blacks (Noble 2018). The internet, where all members of Generation Z have citizenship, reflects not the reality of the world but what society believes is true. We have called on young people to help push us toward a more racially egalitarian society, but we have to be incredibly vigilant about the tools used and the path taken, as choosing the wrong ones, even with good intentions, may steep us into a greater deal of trouble. Between the end of the Obama presidency (when we first drafted this book) and two years into the Trump presidency, some things may be chang-

Is Resuscitation Possible?  /  259

ing. We alluded to the fact that, primarily due to the prevalence of publicly available videos of police fatally shooting unarmed (or legally armed) Black people, more and more White Americans are becoming increasingly grounded in the notion that racism exists and needs to be addressed. Additionally, in reaction to Donald Trump’s campaign, more people are likely to feel as Jenée Desmond-­Harris (2016, SR2) does: that, “for once, nobody is pretending that racism is at a frequency so high they can’t make it out. Racism is no longer being treated as a feeling, an allegation, a matter of opinion, or something that can be negated by the announcement of a black friend.” Somehow we have entered into a time when racism is rearing its head in incredibly overt ways. Despite the downside that comes with an onslaught of racist notions publicly spewed by no small number of Americans, the potential upside is that this makes it impossible to ignore racism or explain it away. It has been the case that when racism is at its ugliest and comes directly into our living rooms (via television, though now it’s also through your phone, computer, tablet, and even your watch), young people have taken to the streets and demanded change and redress, and sometimes have been successful. Maybe the Trump era, then, is the impetus that is needed to get the needle moving again. As we quoted earlier, President Obama described this in his farewell address: “for every two steps forward, it often feels we take one step back.” If Trump and the attitudes surrounding his administration are our “step back,” we are eager to begin taking the next two steps forward.

A pp e n d i x A

Everything You Need to Know about the APC Intrinsic Estimator

What’s the Problem? At some point in one of your math courses, you likely came across what are referred to as systems of equations. Usually, these are a set of two or three equations for which one equation solves for two or three variables. Sometimes a system has only a single solution, sometimes it has more than one, sometimes it has an infinite number of solutions, and sometimes it has no solution. Take, for example, this system of equations: X + 2Y = 9 and X + Y = 6. (System 1) This system can be solved by subtracting the second equation from the first and solving for Y; in this case, Y equals 3. Once you have identified the value of Y, you can then use it to identify the value for X; in this case, X also equals 3. Things get more complicated, however, if you have a system that does not have a unique solution. Take, for example, this system of equations: X + 2Y = 4 and 3X + 6Y = 15. (System 2) This system cannot be solved, as three times the first equation (3 (X + 2Y = 4) ) results in 3X + 6Y = 12, which is contradicted by the second equation, where 3X + 6Y = 15. This system is said to have no real solution. But what if we just slightly change the second equation? X + 2Y = 4 and 3X + 6Y = 12 (System 3) Unlike the previous system, this one can be solved. By multiplying the first equation by 3, we can see that the two equations we have are 3(X + 2Y = 4) and 3X + 6Y = 12, or 3X + 6Y = 12 and 3X + 6Y = 12.

262 / Appendix A System with An Infinite Number of Solutions

System with A Unique Solution (X = 3, Y=3) 10

10

8

8 6

6 4 2

2

0

0

−2

−2

Y

Y

4

X+2Y = 9

X+Y = 6

−4

Y = 2 − X/2

−4 −6

−6 −8

−8

−10

−10

−12

−12

1

2

3

4

X

5

6

7

8

9

10

1

2

3

4

X

5

6

7

8

A.1.  Two Examples of Systems of Equations

This system now has an infinite number of solutions: for any value one arbitrarily chooses for X (or Y), one can find a solution for the other. If 3X + 6Y = 12, and that is all the information we can get from both equations, possible solutions include (X = 4, Y = 0), (X = 0, Y = 2), and so on. In statistics, we usually call a system of equations like this collinear, because they are literally co-­linear: different expressions of the same single line. In this case, that line is Y = 2 − (X / 2). System 1 and system 3 are depicted in figure A.1.

So, How Does This Relate to APC Analyses? Well, the same problem of collinearity arises when we have individual respondents who are in the data occupying specific ages and periods, because, by definition, these terms are collinear. One’s birth cohort (C) is always related to the period (P) in which one is taking the survey and to one’s age (A) in this way: P = (A + C).1 This makes the model we are trying to estimate impossible to identify, and we are left with a situation similar to the right-­hand panel of figure A.1: we can either arbitrarily set either X or Y to some number and then solve for the other, or simply say that there is an infinite number of solutions. For demographers, a traditional approach has been to consider two adjacent age groups to be “equal,” giving them the information necessary to identify the other parameters in the system of equations. Because these constraints are arbitrary, however, this type of model specification might get different results based on who is making

9

10

The APC Intrinsic Estimator  /  263

those constraints on the data. For example, if Christopher and Candis each decided to solve for Y in the system of equations depicted in the right panel of figure A.1, and Christopher arbitrarily set X to 0 and Candis set X to 4, they would get different answers. Christopher would be convinced that Y = 2, and Candis would be sure that Y = 0.

How Does the APC-­IE Get around This? Through some mathematical concepts that are beyond the scope of this text, the intrinsic estimator places a different kind of constraint on the equations so that they can be solved. For example, if we have to estimate the coefficients for five different age groups—­let’s call these coefficients A1, A2, A3, A4, and A5—­the intrinsic estimator uses a constraint such that A1 + A2 + A3 + A4 + A5 = 0. A major advantage of the age, period, and cohort analysis we employ is that it allows us to independently estimate effects from each of the three factors of interest. Looking at age, period, and cohort simultaneously gives us the chance to say whether Whites become more racially conservative as they age (an age effect), whether younger Whites today are more progressive than younger Whites were ten or even fifty years ago (a cohort effect), or, ceteris paribus, whether everyone is becoming more progressive as time marches on (a period effect). Here the ceteris paribus qualification is particularly important, as we will present estimates of the separate effects for each of the three variables (at all their levels) through an estimation procedure that, to our knowledge, has yet to be used to analyze Whites’ racial attitudes.

How Do You Interpret APC-­IE Model Estimates in Tabular Form? APC-­IE estimates can be read just as simply as ordinary least squares regression (or logistic regression models) where the independent variables are all dichotomous. For instance, in table 2.1 we present logistic regression estimates for whether a respondent believes that Blacks have an inborn disability. If we want to estimate the probability that someone holds this belief based on their membership in the age group eighteen to twenty-­four, we can simply set the value of that variable to 1, set the values for all other  . In coefficients to 0, and calculate our prediction in the usual way: Yi = xi B this example, it only requires adding the intercept (our constant term of

264 / Appendix A

−2.65) to the coefficient of that group (−0.485) to get a predicted Ŷ of about −3.135. Because this is a logistic regression, in order to transform that into a predicted probability, we plug that value into this equation: Predicted Probability =

ey 0.0435 = = 0.041. y (1 + e ) 1.0435

Thus, all else equal, we would predict that the probability that someone in that age group (independent of period and of birth cohort) agrees with the statement is about 0.04. If you do the same procedure using the coefficient for the eighty to eighty-­four age group, you’ll see the predicted probability change to approximately 0.11. These end results (predicted values) are what we show in the graphics in this book. Because the mathematics behind the APC-­IE is neither the focus of this book nor the area of our expertise, all we can say is this: we believe that this methodology is appropriate and helpful given how it gets around the aforementioned “identification problem.” This is the linear dependency of the three factors that we care about: period = age + cohort. In turn, “age, period and cohort measures cannot simultaneously be included in a standard regression model due to linear dependency, and cohort effects are unreliable without including age in the model” (Schwadel and Stout 2012, 238).

What Does APC Data Look Like? Table A.1 helps illustrate the identification problem that researchers have traditionally encountered with this type of data. The table is a standard age-­by-­period data table, aggregated from the General Social Survey, showing the proportion of Whites in each of the age categories who agreed with the statement that the differences between Whites and Blacks in terms of socioeconomic status and housing could be attributed to “Blacks hav[ing] less inborn ability to learn.” The entry in the top right of table A.1 can be read as follows: in the 2010 and 2012 GSS, 173 White people aged eighteen to twenty-­four years answered the question on whether differences between Whites and Blacks could be explained by biological differences; of those, about 10 percent of them responded, “Yes.” What is interesting about this table is that it provides evidence that is inconsistent with the idea that younger Whites are far less racist than their predecessors. In fact, with the exception of a few of the oldest age categories (aged seventy and above), those youngest Whites agreed with the “old-­fashioned” biological explanation at a higher rate than any other age group in that period. Of course, these are just summary statistics and cannot tease out whether this difference is simply due to variability

The APC Intrinsic Estimator  /  265 Table A.1.  APC Table of Biological Racism (GSS) Period

18–­24 N 25–­29 30–­34 35–­39 40–­44

Age

45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84 85+

1975–­ 1979

1985–­ 1989

1990–­ 1994

1995–­ 1999

2000–­ 2004

2005–­ 2009

2010–­ 2012

0.15 159 0.09 145 0.12 130 0.17 119 0.25 111 0.28 102 0.29 111 0.27 113 0.48 84 0.53 68 0.51 65 0.47 36 0.33 12 0.57 7

0.12 409 0.12 523 0.13 481 0.14 470 0.12 384 0.2 282 0.25 238 0.28 229 0.28 272 0.36 277 0.39 202 0.35 160 0.42 89 0.38 58

0.09 386 0.09 430 0.08 514 0.08 531 0.09 470 0.09 398 0.16 290 0.2 232 0.23 245 0.26 239 0.32 253 0.32 195 0.28 113 0.39 71

0.05 244 0.06 288 0.04 339 0.07 365 0.09 331 0.07 303 0.1 250 0.12 185 0.12 152 0.16 135 0.23 137 0.22 92 0.34 65 0.19 52

0.07 254 0.04 253 0.08 261 0.07 281 0.1 313 0.07 288 0.06 266 0.15 202 0.18 160 0.14 148 0.18 122 0.3 106 0.29 78 0.24 55

0.07 196 0.03 185 0.06 202 0.04 214 0.09 209 0.08 265 0.05 244 0.06 212 0.11 196 0.14 143 0.14 132 0.17 93 0.2 61 0.27 45

0.1 173 0.06 150 0.04 171 0.05 172 0.04 190 0.03 173 0.09 186 0.08 168 0.09 173 0.1 157 0.13 101 0.27 77 0.16 62 0.26 50

Data: General Social Survey (GSS) cumulative data set (1972–­2016)

in random sampling. Fortunately, the age-­period-­cohort intrinsic estimator allows us to do this.

How Are the Data Grouped into Age and Period Categories? Age, period, and cohort are grouped into five-­year categories, which is typical among demographers and for the APC-­IE modeling technique employed here. Given the fact that only adults participated in these surveys, those respondents aged eighteen were included in the first age group. We have fifteen age groups ranging from eighteen to twenty-­four to eighty-­nine and older and twenty-­two cohorts, which are nine-­year periods and are labeled by the

Proportion

0.00

0.25

0.50

0.75

1.00

1964

1968

Source: American National Election Studies Cumulative Data Set (1948−2016)

1958

1976

1980

1984

1988

Year of Survey

1992

A.2.  Trust in Government 1958–­2012 (ANES)

1972

Proportion Trusting Government Always or Most of the Time

1996

2000

2004

2008

2012

Proportion

1

1

9

90

1 5− 90

1

4

91

−1

0 91

1

9

91

−1

5 91 1

4

92

−1

0 92

Source: American National Election Studies Cumulative Data Set (1948−2016)

4

90

−1

0 90

0.0

0.2

0.4

0.6

1

4

93

−1

0 93 1

9

93

−1

5 93 1

1

9

94

−1

5 94 1

4

95

1 0− 95

Birth Cohort

4

94

−1

0 94 1

9 95 −1

5 95

1

4 96 −1 0 96

A.3.  Trust in Government by Birth Cohort (ANES)

1

9

92

−1

5 92

Proportion Trusting Government Always or Most of the Time

1

9 96 −1 5 96

1

4 97 −1 0 97

1

9 97 −1 5 97

1

4 98 −1 0 98

1

5 98

−1

98

9

268 / Appendix A

middle year in each range (for example, the 1905 cohort contains individuals born between 1901 and 1909). While birth years might appear in multiple cohorts, no individual appears in multiple cohorts; again, this is the generally accepted way to analyze APC data (see Yang and Land 2013). For the models using data from the GSS, there are eight five-­year time periods; for the ANES, there are twelve (ranging from 1950–­1954 to 2010–­2015). For example, we know Americans’ trust in government has been declining since the 1950s (Hetherington 1998, 2005); this is illustrated in figure A.2. One may wonder whether (a) those citizens who were politically socialized during the Watergate scandal, and thus came of age in a scandal-­riddled political era, were left less trusting of government and (b) those individuals led the charge in Americans’ changing attitudes toward the government. We can examine whether those who were new to the political system (between ages eighteen and twenty-­five) at the time of the Watergate scandal introduced a new set of attitudes to the larger pool of American respondents; see figure A.3. This pivotal cohort is labeled “1940–­1944” in figure A.3, and we can see that—­all else equal—­Americans born in those five years were the first to begin the decline of trust in the aggregate. This illustrates the power of cohort replacement and the potential of young people to influence the delineations of American political attitudes.

A pp e n d i x B

A Brief Note on Factor Analysis

As this book is not a methodological text, some readers may benefit from a brief introduction to some of the statistical techniques we employ. Here, we take a moment to discuss the basic concepts underlying factor analysis, a statistical method used to reduce a large number of observed variables into a smaller number of latent (unobserved) variables. Latent variables are, in practice, often referred to as “dimensions.” The two quantities that will affect how many dimensions a factor analysis will produce are (1) the number of observed variables and (2) how strongly those variables are related to one another. These types of statistical models enable researchers to generate factor scores that place cases on latent dimensions in a meaningful way. For example, imagine a researcher has a set of data representing driving distances between a sample of twenty US cities; table B.1 is a subset. We can think of the cities as twenty independent variables, each of which has twenty different observations (the distance from that city to the others in the sample). Keep in mind that some entries will be identical to others (e.g., the distance between Boston and Atlanta is the same as that between Atlanta and Boston). More importantly, some of these independent variables will correlate with others in predictable ways. Since the distance from New York to Philadelphia is only about ninety miles, we expect that the distances between these cities and other cities will be strongly positively correlated (they correlate at approximately 0.99), while cities that are far apart, like Baltimore and Los Angeles, will have distances that are strongly negatively correlated (−0.94). What factor analysis allows us to do is find the unobserved patterns in a set of data that can explain the larger pattern of covariation we do observe. Factor analysis does this by looking for characteristic dimensions that can

. . . 1881.6 2129.3 2171.9

0 575 933.2 585.5 717.6

575 0 358.3 603.4 1207.3 . . . 2284.9 2444 2321.3

Baltimore 933.2 358.3 0 847.3 1544.4 . . . 2570.9 2684.8 2478.2

Boston

Adapted from http://www.mapcrow.info/united_states.html.

Atlanta Baltimore Boston Chicago Dallas . . . San Diego San Francisco Seattle

Atlanta

. . . 1724.8 1848.3 1727.7

585.5 603.4 847.3 0 799.9

Chicago

Table B.1.  Miles between Various American Cities (Truncated)

717.6 1207.3 1544.4 799.9 0 . . . 1177.6 1476.1 1674.4

Dallas . . . . . . . . . . .

. . . . . . . . . . . .

. . . . . . . . . . . .

. 1881.6 2284.9 2570.9 1724.8 1177.6 . . . 0 457.2 1060.8

2129.3 2444 2684.8 1848.3 1476.1 . . . 457.2 0 677.3

San Diego San Francisco

2171.9 2321.3 2478.2 1727.7 1674.4 . . . 1060.8 677.3 0

Seattle

B.1.  Unrotated Factor Solution

B.2.  Rotated Factor Solution

A Brief Note on Factor Analysis  /  273

explain large portions of the variance noticed in the sample. A set of strongly correlated measures, which could be unrelated to other measures, may represent some latent dimension. Ideally, factor analysis would allow the researchers to reduce many variables into a smaller number of latent factors, or dimensions, that are interpretable in a meaningful way. Ultimately, the number of factors that arise is determined by the magnitude of the eigenvalues associated with the factors. The proportion of variance that a factor can explain is found when its eigenvalue is divided by the sum of all the eigenvalues produced from the factor analysis. The factor analysis conducted on our sample data of distances between cities indicates that a two-­factor solution fits the data quite well. With eigenvalues of 12.5 and 5.7, these first two factors explain a total of about 99 percent of all variance in the twenty items. While both factors have a mean of 0.0, the first factor runs from −5.0 to 6.6, and the second takes values from −2.4 to 2.6. Still, what do these numbers mean? If we plot the variable labels against the first and second factors, we may recognize a pattern. Figure B.1 shows the names of the cities plotted along the two recovered factors. By simply transforming the factors to reflect how we often view these cities on a standard projection map, figure B.2 helps us to see that these two factors translate our observed distances between cities into a (relatively) accurate map; cities that we know are close together appear proximate in the figure, even though the layout is not exactly perfect. Higher scores on factor 1 are associated with cities in the American West; meanwhile, higher scores on factor 2 are associated with cities in the American South. Just as we “mapped” some cities in the continental United States, we mapped the landscape of Whites’ racial attitudes. In summary, what factor analysis allows us to do is reduce a large number of variables into a smaller number of dimensions. These dimensions are then made meaningful by showing which variables vary together in predictable ways. In this example, having prior knowledge about which cities are in what part of the country helps us to name the dimensions in a meaningful way (i.e., east/west and north/south). Similarly, using the content of various survey questions, we can hypothesize that Whites’ racial attitudes will fall on particular dimensions.

A pp e n d i x C

Interview Schedule and Respondent Demographics

Respondent Background ·· Tell me about yourself and your family. ·· Where are you from? ·· What kind of work do you do / do your parents do? ·· How would you describe your class background? (If needed: For example, working class, low income, middle class, upper class.) ·· What kind of neighborhood did you grow up in? What kind of people lived there? ·· What about your elementary and high schools? What kinds of kids went to school with you? Was your school diverse, or did most people come from the same background?

Friends ·· What are the first names of your three closest friends? ·· What kinds of things do you do together? ·· Tell me a little bit about them: Age? Race / ethnicity? Where are they from? ·· Most people have friends who are of the same race, live in neighborhoods that mostly have one racial group that lives there, and go to schools where most people are of the same race. What do you think about this? (If needed: Do you think something should be done to see that different people of different racial groups interact, or should this be left up to individuals?)

Politics ·· Would you describe yourself as politically engaged? Do you keep up with current events and pay attention to what your representatives are up to? Do you take the time to do things like vote, campaign for candidates, or try to persuade others about political issues?

276 / Appendix C ·· Concerning economic issues like federal government spending and social safety net policies, would you describe yourself as conservative or liberal? ·· Concerning social issues like abortion and gay marriage, would you describe yourself as conservative or liberal? ·· Do you believe that there is one political party that represents your interests and ideas best? If so, which one? If not, why?

American Identity ·· The American dream is based on the idea that if you work hard, you can be successful or even rich. Do you think this dream is still attainable for most Americans? ·· Do you think some people have it harder or easier attaining the American dream than others? Who has it easier? Who has it harder? ·· To what extent do you believe that immigration is negatively or positively influencing America as a whole—­its values and identity? ·· Some people think it’s important that people begin to think of themselves as American and not African American or Mexican American or Italian American. What do you think about that? ·· We have Black History Month and Latino History Month and the like. Do you think we should continue to teach schoolchildren about this kind of history?

State of Racial Issues in America ·· How would you describe the racial dynamics (e.g., how people of different races interact, conflict between racial groups) in your community? ·· How would you describe the racial dynamics in the US, on the whole? ·· Do you think that Barack Obama’s elections signal a significant shift in racial dynamics in the US? ·· There is still a lot of talk about slavery and racism in the US. Do you think these issues are still relevant or are they things of the past? ·· Some people think that by talking about race and racism or by asking people their race or ethnicity on job or college applications, we make things worse and cause unnecessary tension. What do you think?

Stereotypes and Intergenerational Transmission of Attitudes ·· As you were growing up, did your parents talk much about race or other racial groups? ·· Did you hear any stereotypes about other groups of people? Or tell jokes about people from different races? If so, can you provide some examples? ·· Do you subscribe to any of these attitudes? Do you see any truth in them?

Interview Schedule and Respondent Demographics  /  277

Policies ·· When you hear “affirmative action,” what do you think of? What comes to mind? ·· What do you think about affirmative action in the college admissions process? What about in the workplace? ·· Potential probe: Some people see providing extra points to students who have alumni parents or who are athletes as a different kind of affirmative action. What do you think? ·· Do we still need policies that guarantee minorities a fair shake in hiring or college admissions? ·· When there is unfairness due to race, what do you think should be done about it? ·· Do you think politicians and political leaders should talk about race and racism to help solve society’s problems?

Diversity ·· When you think of the word “diversity,” what comes to mind? ·· Do you think it’s important to have diversity in colleges? In workplaces? In government? ·· Do you think it’s important to make sure that people receive equal opportunities to succeed? ·· Some people think we should look at outcomes to understand inequality. For example, if we see that minorities are attaining worse outcomes than Whites, then that may mean that they are not getting the same oppor­tunities (or other disadvantages may exist). What do you think about that? ·· Do you think people receive equal opportunities in the United States?

“Reverse” Discrimination ·· Do you think that White people, in general, in the US have certain advantages because of their skin color? ·· What about yourself? Do you think you have any privileges or disadvantages because of your race? ·· How concerned are you that special privileges for Blacks and other minorities place you at an unfair disadvantage? ·· Do you think that social policies like affirmative action discriminate unfairly against White people? ·· Are White people in the US discriminated against because of their skin color? ·· Have you ever seen that happen? Or know someone who has been discriminated against because they are White?

278 / Appendix C

Minorities ·· Do you think racial minorities like Blacks and Latinos have certain advantages because of their skin color? Can you provide an example? ·· Blacks tend to fall behind Whites on a lot of important indicators like education and income. Why do you think this is the case? ·· Some people think that African Americans use racism as an excuse for not succeeding. What do you think?

Racial Resentment ·· What comes to mind when I say the following?

o  Generations

of slavery and discrimination have created conditions that

make it difficult for Blacks to work their way out of the lower class.

o  The

Irish, Italians, Jews, and many other minorities overcame prejudice

and worked their way up. Blacks should do the same without any special favors.

o  It’s



o

really a matter of some people not trying hard enough; if Blacks would

only try harder, they could be just as well off as Whites. Over the past few years, Blacks have gotten less than they deserve.

Policy Preferences and Attitudes ·· There is a program in New York City that allows police with reasonable suspicion of criminal activity to stop individuals and search them. This program is commonly known as “stop-­and-­frisk.” What do you think about this policy? (If needed: Explain that some people think this program is good because it can potentially stop crimes. Meanwhile others say that this policy rarely prevents crime and see this policy as a way to unfairly racially profile Blacks and Latinos. What do you think?) ·· Recently, a policy that is commonly known as “stand your ground” became the center of attention because of the killing of  Trayvon Martin by George Zimmerman. This law allows people to use self-­defense without an obligation of retreat—­usually in their home or in their vehicle. What do you think about this law?

Privilege and Success ·· Let’s think about your success. (For students:) Perhaps this means getting into and going to college. (For nonstudents:) Perhaps this means having a good job or providing for your family. Everyone has their own definition of success. Now let’s think about why you are successful. Some of this comes from things you can’t help; let’s call these “endowments.”

Interview Schedule and Respondent Demographics  /  279 For example, your parents’ income or education level, your gender, or your race. Some of your success comes from your hard work. If you could break these two down into two percentages that add up to 100 percent, how much would be attributed to the things you can control, and how much of your success comes from these “endowments”?

Age

20

22

21

34

22

19

20

23

18

32

19

Name

Adeline

Alaina

Ava

Brody

Brooke

Caden

Callie

Charles

Chloe

Claire

Daniella

Female

Female

Female

Male

Female

Male

Female

Male

Female

Female

Female

Gender

Marlborough, MA Lexington, MA

Lincoln, MA

Fishers, IN

Alabama and North Carolina Riverside, IL

Williamstown, MA North Carolina

Watertown, MA

Alexandria, IN

Lexington, MA

Place Where Respondent Resides / Was Born / Grew Up

Table C.1.  Interview Respondent Characteristics

Recent high school graduate Marketing / college recruiter Student, college

Student, college

Student, college

Student, college

Business banker Hospice nurse

Student, college

Student, college

Student, college

Occupation

Mostly White, middle to upper class Wealthy, upper middle class White, Asian, wealthy and upper middle class Predominantly White, middle class White, upper middle class, increasingly Asian

Suburban, sheltered, some diversity, upper middle class Mostly White, middle class White, Armenian, mixed incomes bet­ween lower middle and upper middle class Predominantly White, middle class Rural, safe, White, increase of Latinos in past few years Rural, White

Neighborhood Composition / Characteristics

Upper middle class

Upper middle class Upper middle class Upper middle class Middle class

Middle class

Upper lower / middle class

Middle class

Upper middle class

Lower class

Upper middle class

Family’s Class Status

Conservative

Moderate

Liberal

Conservative

Conservative

Conservative

Conservative

Liberal

Liberal

Conservative

Conservative

Ideology: Federal Spending / Economy

Liberal

Liberal

Liberal

Liberal

Liberal

Conservative

Liberal / Moderate Conservative

Liberal

Liberal

Conservative

Ideology: Social / Abortion / Gay Marriage

Leans Democrat Independent

Democratic

Republican

No party

Republican

Republican

No party

Democratic

No party

Democratic* (not an error)

Party Affiliation

Male

Female Male

21

19

21 21

20

21

20 23

21

22

25

21 28

Elena

Ella

Grayson Harper

Heather

Hunter

Isabella Jackson

Jacob

Jayce

Joshua

Kaitlyn Lila

Female Female

Male

Male

Male

Female

Male Female

Female

Female

Male

26

Easton

Male

24

Dylan

Lexington, MA Philadelphia, PA

Shelbyville, IN

Harford County, MD St. Louis, MO

North Carolina Sarasota, FL

Home Place / Carmel, IN Chicago, IL

Suburbs of Boston, MA Indiana Columbus, IN

Indianapolis, IN

Baltimore, MD

Cleveland, OH

Student, college Clinical psychologist

Student, college

Student, college

—­ Student, law school Student, college

Student, college

Student, college

Student, college Student, college

Student, college

Student, graduate Student, college

Student, graduate

Privileged, mostly White White and Hispanic, apartment complex Predominantly White —­

Rural, White Diverse in FL, not diverse in NC Rural, all White

Racially diverse

Professionals and businesspeople, mostly White White, upper middle class Predominantly White Middle / upper middle class White, upper class

Wealthy

Diverse college town, White suburb

Upper middle class Lower middle class Middle class Lower class

Middle Upper middle class Middle class

Middle class

Middle class

Upper middle class Poor Middle class

Lower / middle class; then upper middle class Upper middle class Upper middle class

Liberal Liberal

Liberal

Liberal / Socialist

Anarcho-­capitalist

N/A Conservative

Liberal

Conservative

Liberal Conservative

Liberal

Conservative

Liberal / Socialist

Liberal

Liberal Liberal

Liberal

Radically liberal Liberal

Moderate Conservative

Liberal

Conservative

Liberal Moderate

Liberal

Liberal

Liberal

Liberal

Democratic Democrat

Democratic

Democratic

Libertarian

No party / left of Democrats Independent Neither

Democrat Independent / no party Republican

Democratic

Leans Democrat Republican

No party / Democratic voter

Male Male

Female

20

19

19

22

20

20 23

20

23

Mackenzie

Madelyn

Maria

Mason

Mia

Michael Miles

Ophelia

Peter

Male

Female

Male

Female

Female

Female

Female

23

Lily

Gender

Age

Name

Table C.1.  (continued )

Indianapolis, IN

Concord, MA

Lexington, MA Valparaiso, IN

Georgia / North Carolina North Carolina

Zionsville, IN

Springfield, MA

Lexington, MA

North Carolina

Place Where Respondent Resides / Was Born / Grew Up

Student, college

Student, college

Student, college Recent college graduate

Student, college

Pastor

Student, college

Student, college

Student, college

Cosmetologist

Occupation

White, upper class

Predominantly White, upper middle class Diverse in GA, not in NC Rural, safe, White, increase of Latinos in past few years —­ Suburban, middle and upper class, predominantly White Mostly White, upper middle class

Rural, safe, White, increase of Latinos in past few years Affluent, predominantly White, Armenian, increasingly Asian High poverty, diverse

Neighborhood Composition / Characteristics

Upper class / wealthy

Upper middle class

Upper class Middle class

Lower middle class / poor Upper middle class Lower middle class Middle class

Middle class

Upper lower / middle class

Family’s Class Status

Conservative / Libertarian

Conservative

Liberal Liberal / indifferent

Moderate / liberal

Conservative

Conservative

Moderate

Conservative

Conservative

Ideology: Federal Spending / Economy

Conservative / Libertarian

Liberal

Liberal Liberal / indifferent

Liberal

Conservative

Moderate

Liberal

Liberal

Conservative

Ideology: Social / Abortion / Gay Marriage

Independent / leans Republican Libertarian

No party Democratic

Neither

Neither

Lean Democrat Republican

No party

Republican

Party Affiliation

21

21

24

21

19

21

19

Reagan

Riley

Sadie

Savannah

Scarlett

Sebastian

Skylar

Female

Male

Female

Female

Female

Female

Female

Westchester County, NY Chicago area, IL Lexington, MA

Ridgewood, NJ

Student, college

New Canaan, CT (currently living there; born in Japan) Los Angeles, CA

Student, college

Student, college

Student, college

Barista and book editor Student, college

Student, college

Elk Grove Village, IL

Predominantly White, suburban Predominantly White, increasingly Asian, few Jewish families

Upper middle class Upper middle class

Extremely racially diverse, mixed-­income Predominantly White, upper middle class / middle class suburb Wealthy, Jewish Upper middle class Upper middle class Upper middle class

Upper class

Middle class

Wealthy, predominantly White, conservative

Suburban, White, middle class

Liberal

Conservative

No opinion

Liberal

Liberal

Liberal

Conservative

Liberal

Liberal

Liberal

Liberal

Liberal

Liberal

Liberal

Democratic

Libertarian

Democrat

No party / leans Democrat Independent / leans Democrat

Independent/ leans Democratic Democrat

A pp e n d i x D

Supplemental APC-­IE Tables

Table D.1.  APC-­IE: Feelings regarding the Women’s Liberation Movement Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

2.213* 1.559* 1.289* −0.254 −0.243 −1.573* −0.828 −0.0173 0.0369 −1.363 −0.547 −0.0218 −0.25

0.762 0.676 0.643 0.623 0.631 0.665 0.695 0.728 0.744 0.786 0.865 0.985 1.208

1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999 2000–­2005

−20.68* −4.857* −0.75 3.70* 8.873* 6.092* 7.618*

0.625 0.454 0.458 0.563 0.425 0.446 0.541

1890–­1894 1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 Constant N = 16,742

−5.306 −1.929 −0.122 −3.469* −1.336 1.466 0.711 −0.501 −0.526 −0.0495 −1.771* −0.86 1.729* 0.794 −0.147 −0.468 1.572 3.236* 6.977* 52.06*

4.818 2.387 1.782 1.54 1.235 1.075 0.981 0.913 0.857 0.843 0.804 0.703 0.618 0.579 0.626 0.706 0.828 1.154 2.183 0.404  

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.2.  APC-­IE: Affective Ratings of Gays and Lesbians Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

2.12* 3.74* 3.28* 1.83* 2.63* 2.81* 1.49* −1.66* −1.43 −2.97* −2.30* −3.62* −5.93*

0.99 0.86 0.77 0.70 0.71 0.72 0.72 0.74 0.79 0.82 0.94 1.04 1.24

1985–­1989 1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

−11.97* −13.05* −5.29* 2.24* 6.58* 7.89* 13.61*

0.66 0.69 0.49 0.47 0.60 0.77 0.48

Constant N = 17,278

40.55*

0.45

1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

−8.77 2.20 −1.97 −0.03 0.53 −0.96 −3.51* −2.80* −1.48 0.21 1.09 0.20 −0.89 −0.45 −0.42 6.64* 3.00* 4.84* 2.56

5.35 2.75 1.92 1.59 1.37 1.27 1.13 1.06 0.95 0.84 0.77 0.71 0.67 0.75 0.81 0.93 1.12 1.29 1.61

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.3.  APC-­IE: Pro-­White Affect Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

−1.29* −1.60* −1.28* −1.11* −1.47* −1.25* −0.54 0.64 0.66 1.12* 1.76* 1.83* 2.52*

0.51 0.46 0.44 0.41 0.41 0.42 0.42 0.43 0.45 0.48 0.55 0.62 0.76

1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

10.42* 6.73* 3.67* 3.25* −1.15* 3.21* −2.48* −6.09* −8.68* −5.63* −3.26*

0.54 0.39 0.40 0.40 0.40 0.54 0.38 0.38 0.49 0.63 0.39

Constant N = 29,101

13.77*

0.28

1890–­1894 1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

0.16 4.28* 3.23* 2.71* 1.86* 1.98* 2.07* 0.79 −0.42 0.73 −0.86 −0.22 −1.72* −2.19* −2.05* −1.95* −1.38* −0.20 −2.41* −1.31 −2.36* −0.70

1.96 1.39 1.12 1.00 0.88 0.78 0.72 0.65 0.61 0.57 0.53 0.48 0.43 0.43 0.44 0.47 0.56 0.62 0.71 0.84 0.96 1.15

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.4.  APC-­IE: Equalitarianism Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

0.26 −0.16 −0.41* −0.18 −0.14 −0.42* 0.05 0.06 −0.19 0.32* 0.37* −0.04 0.49*

0.18 0.14 0.12 0.11 0.12 0.11 0.12 0.13 0.13 0.14 0.17 0.18 0.23

1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

0.22* −0.37* 0.34* 0.31* −0.50*

0.07 0.07 0.13 0.11 0.06

1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

−0.02 0.14 0.79* −0.03 −0.31 0.31 −0.26 −0.30* 0.09 0.01 −0.11 0.04 −0.10 −0.09 0.35* −0.10 −0.45

0.48 0.30 0.31 0.23 0.17 0.18 0.15 0.13 0.13 0.12 0.11 0.11 0.12 0.13 0.17 0.18 0.24

2.48*

0.06

Constant N = 11,603

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.5.  APC-­IE: Support for Men and Women Having Equal Roles (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84       Constant N = 16,567    

0.20* 0.18* 0.26* 0.19* 0.26* 0.28* 0.25* 0.18* −0.07 −0.11 −0.21* −0.43* −1.00*

0.07 0.07 0.07 0.07 0.07 0.08 0.08 0.08 0.07 0.08 0.09 0.09 0.11

1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014

−1.07* −0.70* −0.45* −0.13 0.21* 0.40* 0.74* 1.01*

0.05 0.05 0.06 0.07 0.06 0.07 0.12 0.18

1.60*

0.05

1890–­1894 1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989

0.31 −0.09 0.01 −0.43* −0.26* −0.37* −0.45* −0.37* −0.40* −0.24* −0.07 0.26* 0.36* 0.22* 0.21 0.36* 0.20 0.25 0.18 0.33

0.26 0.16 0.13 0.11 0.10 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.09 0.11 0.14 0.17 0.21 0.37 0.71

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.6.  APC-­IE: Support Workplace Protections for Gays and Lesbians (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84       Constant N = 11,386

0.22* 0.21* 0.15 0.06 0.11 0.03 0.06 −0.19* −0.11 −0.05 −0.02 −0.15 −0.32*

0.10 0.08 0.08 0.07 0.07 0.07 0.08 0.08 0.08 0.08 0.09 0.10 0.12

1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

−0.65* −0.41* −0.14* 0.26* 0.34* 0.61*

0.05 0.05 0.04 0.07 0.06 0.04

0.76*

0.04

1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

−0.33 0.04 0.10 −0.01 −0.07 −0.08 −0.24* 0.03 −0.08 0.17* −0.03 −0.08 −0.11 −0.07 0.23* 0.55* 0.09 −0.10

0.35 0.19 0.15 0.13 0.12 0.10 0.10 0.09 0.08 0.08 0.08 0.07 0.07 0.08 0.10 0.13 0.13 0.17

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.7.  APC-­IE: Support Workplace Protections for Blacks (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

0.40* 0.14* −0.04 0.01 0.04 0.05 0.05 −0.16* −0.04 −0.14 −0.17 −0.10 −0.04

0.08 0.07 0.07 0.06 0.06 0.06 0.06 0.07 0.07 0.08 0.09 0.10 0.12

1965–­1969 1970–­1974 1975–­1979 1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

−0.04 −0.13* 0.13* 0.11* 0.19* 0.04 0.03 −0.03 −0.31*

0.06 0.06 0.05 0.05 0.06 0.05 0.09 0.08 0.05

Constant N = 11,783

−0.27*

0.04

1885–­1889 1890–­1894 1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989

0.06 −0.57 0.04 0.23 −0.08 −0.05 0.01 −0.03 −0.16 −0.07 0.14 0.16* 0.16* 0.21* −0.03 0.07 −0.06 −0.06 0.20 −0.10 −0.07

0.47 0.31 0.20 0.16 0.13 0.12 0.11 0.10 0.09 0.09 0.08 0.07 0.06 0.06 0.06 0.08 0.09 0.11 0.13 0.14 0.16

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.8.  APC-­IE: Support Women in the Workplace (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84       Constant N = 20,198

0.55* 0.54* 0.45* 0.30* 0.24* 0.14* 0.10 −0.03 −0.17* −0.33* −0.45* −0.58* −0.78*

0.07 0.06 0.06 0.06 0.06 0.07 0.07 0.07 0.07 0.07 0.07 0.08 0.10

1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999 2000–­2004

−0.46* −0.40* 0.15* 0.13* 0.24* 0.34*

0.04 0.04 0.04 0.04 0.04 0.05

1.05*

0.03

1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984

−0.20 −0.77* −0.28* −0.29* −0.20* −0.18* 0.03 0.12 0.23* 0.19* 0.25* 0.36* 0.31* 0.17* 0.15* 0.04 0.15 −0.08

0.23 0.16 0.11 0.09 0.08 0.07 0.07 0.07 0.08 0.08 0.07 0.06 0.06 0.06 0.07 0.09 0.12 0.19

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.9.  APC-­IE: Support Affirmative Action for Blacks (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84       Constant N = 14,948

0.25* 0.07 −0.14 −0.09 0.13 0.17 0.12 0.11 −0.01 −0.09 −0.36* 0.04 −0.22

0.11 0.10 0.10 0.09 0.09 0.09 0.10 0.10 0.10 0.11 0.13 0.13 0.16

1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

0.31* −0.01 −0.15* −0.06 −0.23* 0.14*

0.07 0.06 0.06 0.10 0.10 0.05

−1.96*

0.05

1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

0.49 −0.23 0.19 0.09 −0.24 −0.14 −0.35* −0.04 −0.13 0.01 −0.11 −0.39* −0.21* −0.19 0.30* 0.38* 0.27 0.33*

0.38 0.24 0.18 0.16 0.16 0.14 0.14 0.12 0.11 0.10 0.09 0.09 0.09 0.10 0.10 0.13 0.14 0.16

Source: 1952–­2016 American National Election Studies Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.10.  APC-­IE: Support Affirmative Action for Blacks (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84       Constant N = 13,274

0.44* 0.19 0.10 −0.24* 0.18 0.04 −0.13 −0.04 −0.23 −0.27* 0.04 0.15 −0.23

0.14 0.12 0.11 0.11 0.10 0.10 0.10 0.11 0.12 0.13 0.14 0.15 0.19

1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

−0.17 −0.04 −0.18* 0.05 0.33*

0.09 0.05 0.08 0.05 0.06

−2.09*

0.07

1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

−0.81 −0.27 −0.02 0.21 0.13 0.06 0.20 0.28* 0.24 −0.03 0.21* −0.12 −0.06 −0.02 0.01 −0.18 0.17

0.85 0.33 0.25 0.21 0.18 0.17 0.15 0.14 0.13 0.12 0.10 0.10 0.09 0.09 0.11 0.14 0.19  

Source: 1972–­2016 General Social Survey Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.11.  APC-­IE: Support Affirmative Action for Women (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

0.01 −0.15 −0.20 −0.23 −0.16 −0.01 −0.19 0.16 0.33* −0.15 0.29* 0.40* −0.09

0.15 0.12 0.12 0.12 0.12 0.12 0.13 0.13 0.13 0.14 0.14 0.15 0.20

2000–­2004 2005–­2009 2010–­2014 2015–­2016

−0.07 −0.12 0.08 0.11

0.06 0.08 0.05 0.06

1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999

0.63 0.56* −0.14 0.15 −0.26 −0.29 −0.42* −0.17 −0.26 −0.06 −0.19 −0.08 −0.05 0.15 0.05 0.38

0.35 0.26 0.19 0.15 0.15 0.15 0.15 0.14 0.14 0.13 0.13 0.12 0.11 0.13 0.15 0.22

Constant N = 5,383

−0.98*

0.05

Source: 1972–­2016 General Social Survey Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.12.  APC-­IE: Support Ban on Interracial Marriage (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

−0.61* −0.70* −0.45* −0.48* −0.38* −0.22* −0.13* 0.07 0.35* 0.31* 0.67* 0.78* 0.77*

0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.06 0.07 0.09

1975–­1979 1980–­1984 1985–­1989 1990–­1994 1995–­1999 2000–­2004 2005–­2009

0.74* 0.45* 0.44* 0.14* −0.25* −0.64* −0.87*

0.04 0.04 0.04 0.04 0.05 0.05 0.11

Constant N = 24,346

−1.21*

0.06

1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989

0.41 0.37* 0.37* 0.44* 0.32* 0.31* 0.31* 0.26* 0.21* 0.24* −0.02 −0.29* −0.36* −0.23* −0.38* −0.14 0.02 −0.39 −1.44

0.23 0.15 0.11 0.10 0.09 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.08 0.09 0.10 0.11 0.13 0.20 0.96

Source: 1972–­2016 General Social Survey Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.13.  APC-­IE: Oppose Relative Marrying Member of Another Race (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

−0.63* −0.62* −0.59* −0.40* −0.31* −0.19* 0.01 0.17* 0.25* 0.31* 0.61* 0.59* 0.78*

0.11 0.10 0.09 0.08 0.08 0.08 0.08 0.08 0.08 0.09 0.10 0.11 0.13

1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014

1.38* 0.23* −0.15* −0.45* −1.01*

0.06 0.04 0.05 0.04 0.05

Constant N = 13,346

−0.48*

0.05

1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994

0.89 0.27 0.33 0.03 0.25* 0.36* 0.06 −0.03 −0.09 −0.12 −0.11 −0.13 −0.24* −0.34* −0.48* −0.30* −0.36

0.61 0.21 0.18 0.14 0.12 0.12 0.11 0.10 0.09 0.09 0.08 0.08 0.08 0.09 0.12 0.13 0.24

Source: 1972–­2016 General Social Survey Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Table D.14.  APC-­IE: Agree Government Should Help Blacks (0/1) Age

Period

Cohort

Group

Coef.

S.E.

Year of Survey

Coef.

S.E.

Year of Birth

Coef.

S.E.

18–­24 25–­29 30–­34 35–­39 40–­44 45–­49 50–­54 55–­59 60–­64 65–­69 70–­74 75–­79 80–­84

0.34* 0.02 −0.10 −0.07 −0.02 −0.14* −0.09 −0.06 −0.02 0.09 −0.02 −0.03 0.10

0.08 0.07 0.07 0.06 0.06 0.07 0.07 0.08 0.08 0.08 0.09 0.10 0.12

1975–­1979 1985–­1989 1990–­1994 1995–­1999 2000–­2004 2005–­2009 2010–­2014 2015–­2016

0.36* −0.03 0.03 −0.11* −0.04 −0.19* −0.13* 0.11

0.07 0.06 0.05 0.05 0.05 0.08 0.06 0.06

Constant N = 17,395

−1.40*

0.04

1895–­1899 1900–­1904 1905–­1909 1910–­1914 1915–­1919 1920–­1924 1925–­1929 1930–­1934 1935–­1939 1940–­1944 1945–­1949 1950–­1954 1955–­1959 1960–­1964 1965–­1969 1970–­1974 1975–­1979 1980–­1984 1985–­1989 1990–­1994

−0.01 0.20 0.19 0.12 −0.07 −0.16 −0.30* −0.22* −0.17 0.02 0.12 0.06 −0.08 0.02 0.06 −0.02 −0.16 0.15 0.03 0.22

0.46 0.26 0.18 0.14 0.13 0.12 0.12 0.11 0.10 0.09 0.07 0.07 0.06 0.06 0.07 0.08 0.09 0.11 0.12 0.18

Source: 1972–­2016 General Social Survey Cumulative Data File Notes: Non-­Hispanic White, face-­to-­face respondents only; *p < 0.05, two-­tailed test Estimates and standard errors calculated by the APC-­IE estimation procedure in Stata 13.

Notes

I n tr o duct i o n

1.

2.

Pew defines Generation X as those born between 1965 and 1980, boomers were born between 1946 and 1964, and the silent generation’s birth years are between 1928 and 1945. “Ferguson” refers to both the fatal shooting of Michael Brown, an unarmed black teenager, by a white police officer in August 2014, and the catapulting of the Black Lives Matter movement into the average American’s knowledge base. “Charlottesville” is a reference to the two-­day rally at Charlottesville, Virginia, in August 2017, just three years later, where neo-­Nazis, the so-­called alt-­right, Klansmen, and the like gathered in part to unite white nationalists across many stripes and to protest the potential removal of a Confederate statue. Dozens of people were injured, and three people died. C h apt e r 1

1. There is considerable debate on this issue, nonetheless, especially concerning whether people can use the N-­word as a term of endearment or friendliness and whether by “banning” the word, it is actually given more power. 2. Philosopher Charles Mills and sociologist Jennifer Mueller argue that the willful, aggressive ignorance of racial inequality and the efforts to mystify practical solutions are best understood within the framework they call the “epistemology of ignorance” (Mills 2007; Mueller 2017). C h apt e r 2

1.

2.

3.

Despite the landmark Obergefell decision that deems marriage equality constitutional, a wide array of anti-­LGBT policies exists, particularly at the state level. In some states, people are still allowed to discriminate on the basis of sexuality in hiring and housing. Generally speaking, though, the shape of policies has not yet met up with public opinion in this domain of American life (Lax and Phillips 2009). As one will see, the fourth domain of attitudes is specific to African Americans, and there are no questions that are equivalent for gays and lesbians, or for women, in the ANES or GSS. This question was asked on a seven-­point Likert scale that was rescaled to run from 0 to 1. Those who responded, “A woman’s place is in the home,” were coded 0, while those who responded, “Women and men should have an equal role,” were coded 1.

294  / Notes to Pages 72–155 C h apt e r 3

1.

For example, in the 1977 GSS, 327 of 1,267 whites (25.8 percent) agreed that differences between whites and blacks were based on blacks having an “inborn disability”; in the 2012 GSS, only 75 out of 960 whites expressed the same belief (7.8 percent). 2. The Kinder and Sears (1981) article has been cited well over a thousand times. 3. One of the most frequently cited points of evidence is millennials’ support for Obama. To be sure, 54 percent of whites between the ages of eighteen and twenty-­ nine voted for Barack Obama in 2008, but this declined to 44 percent in 2012. And according to a December 2014 Gallup poll, at that time Obama’s support among “millennial whites” was at just 34 percent, its lowest point since Obama took office. 4. This is the traditional method for demographers and for the APC-­IE method used here; for more information, please see appendix A, which discusses the method in further detail. 5. Given an item with K response categories, exactly K − 1 thresholds must be estimated; this concept is similar to the notion of “cut points” in ordered logistic regression. In our tables we denote cut point j for item i as τij. 6. The factor loading for the first item is set to 1.00 across groups for identification purposes. 7. Unlike other tests of hypotheses, typically in structural equation modeling an in­ significant χ2 statistic is an indicator of good model fit, whereas a large χ2 indicates the model is not supported by the data. For a detailed discussion of the goodness-­ of-­fit measures, see Bentler and Bonett (1980), Bentler (1990), or Bollen and Long (1993). 8. This test was conducted in Mplus version 6.11, using the DIFFTEST option for nested models using categorical variables (WLSMV). Additional invariance tests were carried out in an increasingly restrictive, stepwise fashion, but even less restrictive models (e.g., constraining factor loadings and freeing thresholds) were still accompanied by significantly larger χ2 statistics (Δχ2 = 10.3, p < 0.005). The substantive findings for these invariance tests are not unique to the ANES and have been replicated in several other nationally representative data sets; these additional analyses are available from the authors upon request. C h apt e r 5

1. If our interviewers felt that respondents needed more information or hesitated in answering the question altogether, they explained, “Some people think this program is good because it can potentially stop crimes. Meanwhile others say that this policy rarely prevents crime and see this policy as a way to unfairly racially profile blacks and Latinos. What do you think?” 2. Research shows that stop-­and-­frisk is not the reason for a decline in crime in New York City. Mayor Giuliani implemented this policy during a declining trend in crime in the city. Further, research shows that this policy is not proven to get guns off the street or prevent murders (New York Civil Liberties Union 2011). 3. Michelle Alexander (2010, 201) points out that drunk drivers are predominantly white and male. In 1990, 78 percent of arrests for drunk driving were of white men. 4. It should also be noted that after repeated simulations and extensive exposure to the computer program, racial biases were significantly reduced in this lab setting.

Notes to Pages 181–243  /  295 C h apt e r 6

1. In the full version of the CoBRAS, there are seven questions in the first dimension, seven in the second, and six in the third. We chose the questions that loaded highest on each dimension in order to develop a shortened version of the scale. C h apt e r 7

1.

For a full discussion of goodness-­of-­fit indices in structural equations, we encourage the reader to see Bentler (1990), Bollen and Long (1993), and Cheung and Rensvold (2002). 2. The largest eigenvalues of the correlation matrix were 6.0 and 0.64, consistent with a one-­factor solution. 3. The smallest test statistic, in absolute terms, for any of the factor loadings is t = 6.88, which corresponds to a p-­value far below 0.001. 4. Although a traditional seven-­point measure of political ideology is highly correlated with the first dimension of racial attitudes, we argue that this dimension is still very different from ideology. To begin, it captures white Americans’ knowledge and awareness of racial attitudes. Additionally, even if we remove all of the ideology-­related measures from this dimension, our results remain quite robust. 5. One could argue that under certain conditions, these policies could be racialized, but the CCES did not prime racial attitudes, as scholars do purposely in experiments to show the effects of racial priming on whites’ policy preferences. C h apt e r 8

1. It was really serendipitous to be able to place these four items on the Common Content of the 2016 Cooperative Congressional Election Study. While it certainly took some convincing, we are indebted to Brian Schaffner and Stephen Ansolabehere (Co-­PIs of the CCES) for selecting our items for inclusion on the large survey. 2. Technically speaking, we began by randomly selecting a smaller subsample of the available questions, selecting either four or five items. Second, we independently regressed each of the second-­order factor scores on that collection of items and recorded the multiple r2 term from the regression output. We then stored the selected questions and these statistics and looked for which combination of questions maximized the sum of the two r-­squared terms. 3. Those levels are no high school, high-­school graduate, some college, two-­year-­college graduate, four-­year-­college graduate, and post-­graduate. 4. The t-­statistic for this difference is −111.2, on 14,948 degrees of freedom. The probability that we’d observe data like this if the two groups were equally cognizant of whites’ advantages is practically zero. Certainly, this difference is statistically signifi­ cant at any level; if we wanted to write out the p-­value, it would be less than a decimal with over one thousand leading zeros (p < 10(−1000)). 5. We calculated the predicted probabilities as if each respondent were a female high-­ school graduate who thought the economy had worsened and was average in terms of her age, her income, her partisanship, and her ideology. 6. While we do not report the full model in the text, we should note that when each of the following is coded to run from 0 to 1, the coefficients with the largest magnitude are partisanship (3.50), ideology (2.16), acknowledging whites’ advantage (−1.86), and economy is worse (1.64), followed by the other three FIRE items: racial empathy

296  / Notes to Pages 244–262

7.

(−1.36), recognition of racial problems (−1.32), and racial fear (0.82). Of these, all are significant at the p < 0.001 level, with z-­statistics with an absolute value greater than 4.80. Ask us in person or read Candis’s other book, inspired by the findings in this one. C o n clus i o n

1. Even though we see more interracial marriage, whites are most likely to marry Asians and Asian Americans, then Latinx people, with interracial marriage with blacks lagging much further behind. 2. In our interviews, we found that many fewer respondents support affirmative-­action policies than we saw in the CCES. However, research shows that people tend to provide more liberal answers to quantitative surveys than in qualitative interviews (Bonilla-­Silva, Lewis, and Embrick 2004; Bonilla-­Silva 2014). It is only in qualitative interviews that people can say things like “Yes, but . . .” and explain their rationale; in a survey, they may simply answer, “Yes.” 3. If anyone has ever heard Christopher talk about it, he often credits the racial resentment measure as the impetus for his switching from being a political theorist to being a quantitative scholar of American politics. The particular question that stuck out to him was the question about “Irish, Italians, Jews and other minorities” working their way up without any favors; maybe our items can do the same for some other, younger scholars. A pp e n d i x A

1.

The general problem that the intrinsic estimator (hereafter IE) solves is that the standard design matrix, X, is not full rank, meaning that at least one column can be constructed by a linear combination of other columns. The mathematical implication for this is that the usual generalized linear model (see equation (1) below) cannot be calculated, because the matrix (X TX) cannot be inverted. (1)

β = (X T X)– X T Y

As we reviewed above, one way to solve this problem is to constrain some parameter to be equal and estimate the equation as normal with two parameters constrained to equality or a single parameter left out of the model. The IE avoids doing this by decomposing the matrix (X T X) into two parts: its nonnull and null column subspaces. Computationally, this involves finding the eigenvector of (X T X) that has an eigenvalue exactly equal to zero and using a principal components regression (where the nonzero eigenvectors are the independent variables) to remove the influence of the null-­space of (X T X). Again, the proofs of the relevant statistical properties of the IE are well beyond the scope of this work; for details on the estimator’s bias, efficiency, and consistency see Yang, Fu, and Land (2004). Previous methods required researchers to have and make strong assumptions about the data, and the estimates that the models put out are sensitive to this arbitrary choice of the identifying constraint (Yang, Fu, and Land 2004). In contrast, the APC-­IE does not depend on these atheoretical constraints, and it yields robust estimates of trends by age, period, and cohort. “Put in the simplest possible terms, the basic idea of the IE is to remove the influence of the design matrix on coefficient estimates” (Yang et al. 2008, 1707). Since the dependent variables are dichotomous, a logit link function adjusts the model accordingly.

W o rks C i t e d

Alexander, Michelle. 2010. The New Jim Crow: Mass Incarceration in the Age of Colorblindness. New York: New Press. Andersen, David J., and Jane Junn. 2009. “Deracializing Obama: White Voters and the 2004 Illinois U.S. Senate Race.” American Politics Research 38, no. 3: 443–­70. Andersen, Margaret L. 1999. “The Fiction of ‘Diversity without Oppression’: Race, Ethnicity, Identity, and Power.” In Critical Ethnicity: Countering the Waves of Identity Politics, edited by Robert H. Tai and Mary L. Kenyatta, 5–­20. Lanham, MD: Rowman & Littlefield. Andolina, Molly W., and Jeremy D. Mayer. 2003. “Demographic Shifts and Racial Attitudes: How Tolerant Are Whites in the Most Diverse Generation?” Social Science Journal 40, no. 1: 19–­31. Ansolabehere, Stephen, and Brian F. Shaffner. 2017. CCES Common Content, 2016. Cambridge, MA: Harvard Dataverse, V4. AP. 2012. “AP Poll: U.S. Majority Have Prejudice against Blacks.” USA Today, October 27, 2012. http://www.usatoday.com/story/news/politics/2012/10/27/poll-black-prejudice -america/1662067/. Apfelbaum, Evan P., Kristin Pauker, Nalini Ambady, Samuel R. Sommers, and Michael I. Norton. 2008. “Learning (Not) to Talk about Race: When Older Children Underperform in Social Categorization.” Developmental Psychology 44, no 5: 1513–­18. Apfelbaum, Evan P., Samuel R. Sommers, and Michael I. Norton. 2008. “Seeing Race and Seeming Racist? Evaluating Strategic Colorblindness in Social Interaction.” Journal of Personality and Social Psychology 95, no. 4: 918–­32. Asante-­Muhammad, Dedrick, Chuck Collins, Josh Hoxie, and Emanuel Nieves. 2016. The Ever-­Growing Gap: Without Change, African-­American and Latino Families Won’t Match White Wealth for Centuries. Washington, DC: Institute for Policy Studies. https://ips-dc .org/wp-content/uploads/2016/08/The-Ever-Growing-Gap-CFED_IPS-Final-2.pdf. Astin, Alexander W. 1998. “The Changing American College Student: Thirty-­Year Trends, 1966–­1996.” Review of Higher Education 21, no. 2: 115–­35. Badger, Emily, Claire Cain Miller, Adam Pearce, and Kevin Quealy. 2018. “Extensive Data Shows Punishing Reach of Racism for Black Boys.” New York Times, March 19, 2018. https://www.nytimes.com/interactive/2018/03/19/upshot/race-class-white-and-black -men.html. Banks, Antoine J., and Nicholas A. Valentino. 2012. “Emotional Substrates of  White Racial Attitudes.” American Journal of Political Science 56, no. 2: 286–­97.

298 / Works Cited Baum, Dan. 2016. “Legalize It All: How to Win the War on Drugs.” Harper’s, April 2016. https://harpers.org/archive/2016/04/legalize-it-all/. Baumgartner, Frank R., Derek A. Epp, and Kelsey Shoub. 2018. Suspect Citizens: What 20 Mil­lion Traffic Stops Tell Us about Policing and Race. New York: Cambridge University Press. Baumgartner, Frank R., Amanda J. Grigg, and Alisa Mastro. 2015. “# BlackLivesDon’tMatter: Race-­of-­Victim Effects in US Executions, 1976–­2013.” Politics, Groups, and Identities 3, no. 2: 209–­21. Becker, Amy B. 2012. “Determinants of Public Support for Same-­Sex Marriage: Generational Cohorts, Social Contact, and Shifting Attitudes.” International Journal of Public Opinion Research 24, no. 4: 524–­33. Bell, Joyce M., and Douglas Hartmann. 2007. “Diversity in Everyday Discourse: The Cultural Ambiguities and Consequences of ‘Happy Talk.’” American Sociological Review 72, no. 6: 895–­914. Benedict, Ruth. 1945. Race and Racism. London: Routledge and Kegan Paul. Bentler, Peter M. 1990. “Comparative Fit Indexes in Structural Models.” Psychological Bulletin 107, no. 2: 238–­46. Bentler, Peter M., and Douglas G. Bonett. 1980. “Significance Tests and Goodness of Fit in the Analysis of Covariance Structures.” Psychological Bulletin 88, no. 3: 588–­606. Berrey, Ellen. 2015a. “Diversity Is for White People: The Big Lie behind a Well-­Intended Word.” Salon, October 26, 2015. http://www.salon.com/2015/10/26/diversity_is_for _white_people_the_big_lie_behind_a_well_intended_word/. Berrey, Ellen. 2015b. The Enigma of Diversity: The Language of Race and the Limits of Racial Justice. Chicago: University of Chicago Press. Bertrand, Marianne, and Sendhil Mullainathan. 2004. “Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.” American Economic Review 94, no. 4: 991–­1013. Bertrand, Marianne, Sendhil Mullainathan, and Eldar Shafir. 2004. “A Behavioral-­ Economics View of Poverty.” American Economic Review 94, no. 2: 419–­23. Blake, John. 2011. “Are Whites Racially Oppressed?” CNN, March 4, 2011. http://www.cnn .com/2010/US/12/21/white.persecution/index.html. Blalock, Hubert M. 1956. “Economic Discrimination and Negro Increase.” American Sociological Review 21, no. 5: 584–­88. Blaumer, Bob. 1972. Racial Oppression in America. New York: Harper & Row. Blinder, Scott B. 2007. “Dissonance Persists: Reproduction of Racial Attitudes among Post-­Civil Rights Cohorts of White Americans.” American Politics Research 35, no. 3: 299–­335. Block, Ray, Jr. 2011. “Backing Barack Because He’s Black: Racially Motivated Voting in the 2008 Election.” Social Science Quarterly 92, no. 2: 423–­46. Bobo, Lawrence. 2000. “Race and Beliefs about Affirmative Action: Assessing the Effects of Interests, Group Threat, Ideology, and Racism.” In Racialized Politics: The Debate about Racism in America, edited by David O. Sears, Jim Sidanius, and Lawrence Bobo, 137–­64. Chicago: University of Chicago Press. Bobo, Lawrence, James R. Kluegel, and Ryan A. Smith. 1997. “Laissez-­Faire Racism: The Crystallization of a Kinder, Gentler, Antiblack Ideology.” In Racial Attitudes in the 1990s: Continuity and Change, edited by Steven A. Tuch and Jack K. Martin, 16–­42. Westport, CT: Prager. Bobo, Lawrence D., and Ryan A. Smith. 1998. “From Jim Crow Racism to Laissez Faire Racism: The Transformation of Racial Attitudes.” In Beyond Pluralism: The Conception of

Works Cited / 299 Groups and Group Identities in America, edited by Wendy E. Katkin, Ned Landsman, and Andrea Tyree, 182–­220. Chicago: University of Illinois Press. Bollen, Kenneth A., and J. Scott Long, eds. 1993. Testing Structural Equation Models. Newbury Park, CA: Sage Publications. Bonilla-­Silva, Eduardo. 1997. “Rethinking Racism: Toward a Structural Interpretation.” American Sociological Review 62, no. 3: 465–­80. Bonilla-­Silva, Eduardo. 2012. “The Invisible Weight of Whiteness: The Racial Grammar of Everyday Life in Contemporary America.” Ethnic and Racial Studies 35, no. 2: 173–­94. Bonilla-­Silva, Eduardo. 2014. Racism without Racists: Color-­blind Racism and the Persistence of Racial Inequality in the United States. 4th ed. Lanham, MD: Rowman & Littlefield. Bonilla-­Silva, Eduardo, Carla Goar, and David G. Embrick. 2006. “When Whites Flock Together: The Social Psychology of White Habitus.” Critical Sociology 32, no. 2–­3: 229–­53. Bonilla-­Silva, Eduardo, and Amanda E. Lewis. 1999. “The New Racism: Racial Structure in the United States, 1960s–­1990s.” In Race, Ethnicity, and Nationality in the USA: Toward the Twenty-­First Century, edited by Paul Wong, 55–­101. Boulder, CO: Westview. Bonilla-­Silva, Eduardo, Amanda Lewis, and David G. Embrick. 2004. “‘I Did Not Get That Job Because of a Black Man . . .’: The Story Lines and Testimonies of Color-­Blind Racism.” Sociological Forum 19, no. 4: 555–­81. Bright, Cedric M., Corey A. Duefield, and Valerie E. Stone. 1998. “Perceived Barriers and Biases in the Medical Education Experience by Gender and Race.” Journal of the National Medical Association 90, no. 11: 681–­88. Brooks, Clem. 2000. “Civil Rights Liberalism and the Suppression of a Republican Political Realignment in the United States, 1972 to 1996.” American Sociological Review 65, no. 4: 483–­505. Brown, Timothy A. 2012. Confirmatory Factor Analysis for Applied Research. New York: Guilford Press. Bureau of Labor Statistics. 2014. Table E-­16. “Unemployment Rates by Age, Sex, Race, and Hispanic or Latino Ethnicity.” In Labor Force Statistics from the Current Population Survey. Washington, DC: United States Department of Labor. Butler, Daniel M., and David E. Broockman. 2011. “Do Politicians Racially Discriminate Against Constituents? A Field Experiment on State Legislators.” American Journal of Political Science 55, no. 3: 463–­77. Camera, Lauren. 2016. “More than 60 Years after Brown v. Board of  Education, School Segregation Still Exists.” U.S. News & World Report, May 17, 2016. https://www.usnews.com/news /articles/2016-05-17/after-brown-v-board-of-education-school-segregation-still-exists. Carmines, Edward G., Paul M. Sniderman, and Beth C. Easter. 2011. “On the Meaning, Measurement, and Implications of Racial Resentment.” ANNALS of the American Academy of Political and Social Science 634: 98–­116. Carmines, Edward G., and Richard A. Zeller. 1979. Reliability and Validity Assessment. Quantitative Applications in the Social Sciences, vol. 17. Newbury Park, CA: Sage Publications. Carr, Leslie G. 1997. “Colorblind” Racism. Thousand Oaks, CA: Sage Publications. Carson, Benjamin. 2016. Interview by Neil Cavuto, Cavuto: Coast to Coast, Fox Business Network, April 20, 2016. Quoted in https://www.thedailybeast.com/ben-carson-put -tubman-on-bill-no-one-uses. Carver, Marina. 2013. “NYPD Officers Say They Had Stop-­and-­Frisk Quotas.” CNN, March 22, 2013, updated March 26, 2013. http://www.cnn.com/2013/03/22/justice /new-york-stop-and-frisk-trial/.

300 / Works Cited Chen, Michelle. 2017. “Donald Trump’s Rise Has Coincided with an Explosion of Hate Groups.” The Nation, March 24, 2017. https://www.thenation.com/article/donald -trumps-rise-has-coincided-with-an-explosion-of-hate-groups/. Cheung, Gordon W., and Roger B. Rensvold. 2002. “Evaluating Goodness-­of-­Fit Indexes for Testing Measurement Invariance.” Structural Equation Modeling 9, no. 2: 233–­55. Chiasson, Dan. 2014. “Color Codes: A Poet Examines Race in America.” New Yorker, October 24, 2014. https://www.newyorker.com/magazine/2014/10/27/color-codes. Chudy, Jennifer. 2017. “Racial Sympathy in American Politics.” PhD diss., University of Michigan. Clement, Scott. 2015. “A Year after Ferguson, 6 in 10 Americans Say Changes Are Needed to Give Blacks and Whites Equal Rights.” Washington Post, August 5, 2015. http://www.wash ingtonpost.com/news/the-fix/wp/2015/08/05/what-changed-since-ferguson-americans -are-far-more-worried-about-black-rights/. CNN Politics. 2016. “Exit Polls: National President.” CNN, November 23, 2016. https:// www.cnn.com/election/2016/results/exit-polls/national/president. Cohen, Cathy J. 2010. Democracy Remixed: Black Youth and the Future of American Politics. Oxford: Oxford University Press. Cohen, Patricia. 2014. “For Recent Black College Graduates, a Tougher Road.” New York Times, December 25, 2014, Business. Communities United for Police Reform. 2012. “The Issue.” 2012. Accessed August 20, 2018. http://changethenypd.org/issue. Cottom, Tressie McMillan. 2017. Lower Ed: The Troubling Rise of For-­Profit Colleges in the New Economy. New York: New Press. Craig, Maureen A., and Jennifer A. Richeson. 2014a. “More Diverse Yet Less Tolerant? How the Increasingly Diverse Racial Landscape Affects White Americans’ Racial Attitudes.” Personality and Social Psychology Bulletin 40, no. 6: 750–­61. Craig, Maureen A., and Jennifer A. Richeson. 2014b. “On the Precipice of a ‘Majority-­ Minority’ America: Perceived Status Threat from the Racial Demographic Shift Affects White Americans’ Political Ideology.” Psychological Science 25, no. 6: 1189–­97. Craig, Maureen A., Julian M. Rucker, and Jennifer A. Richeson. 2018. “Racial and Political Dynamics of an Approaching ‘Majority-­Minority’ United States.” ANNALS of the American Academy of Political and Social Science 677: 204–­14. Cramer, Katherine J. 2016. The Politics of Resentment: Rural Consciousness in Wisconsin and the Rise of Scott Walker. Chicago: University of Chicago Press. Cutler, Stephen J., Sally Ann Lentz, Michael J. Muha, and Robert N. Riter. 1980. “Aging and Conservatism: Cohort Changes in Attitudes about Legalized Abortion.” Journal of Gerontology 35, no. 1: 115–­23. Dahl, Melissa. 2008. “Young Voter Turnout Likely Sets New Record.” NBC News, November 5, 2008. http://www.nbcnews.com/id/27562023/ns/politics-decision_08/t /young-voter-turnout-likely-sets-new-record/#.XNzJrC_MzGI. Darity, William. 2011. “Revisiting the Debate on Race and Culture.” Du Bois Review: Social Science Research on Race 8, no. 2: 467–­76. Davidov, Eldad. 2009. “Measurement Equivalence of Nationalism and Constructive Patri­ otism in the ISSP: 34 Countries in a Comparative Perspective.” Political Analysis 17, no. 1: 64–­82. Dawson, Michael C. 2012. “Racial Tragedies, Political Hope, and the Tasks of American Political Science.” Perspectives on Politics 10, no. 3: 669–­73. DeSante, Christopher D. 2013. “Working Twice as Hard to Get Half as Far: Race, Work Ethic, and America’s Deserving Poor.” American Journal of Political Science 57, no. 2: 342–­56.

Works Cited / 301 DeSante, Christopher D., and Candis Watts Smith. 2017. “Fear, Institutionalized Racism, and Empathy: The Underlying Dimensions of Whites’ Racial Attitudes.” Presented at the Midwest Political Science Association Conference, Chicago, IL, April 7, 2017. Desjardins, Lisa. 2017. “How Trump Talks about Race.” PBS, August 22, 2017. https://www.pbs .org/newshour/politics/every-moment-donald-trumps-long-complicated-history-race. Desmond-­Harris, Jenée. 2016. “The Upside to Overt Racism.” New York Times, May 1, 2016, Sunday Review, SR2. DeVellis, Robert F. 2016. Scale Development: Theory and Applications. 4th ed. Applied Social Research Methods, vol. 26. Newbury Park, CA: Sage Publications. DiAngelo, Robin. 2011. “White Fragility.” International Journal of Critical Pedagogy 3, no. 3: 54–­70. DiAngelo, Robin. 2018. White Fragility: Why It’s So Hard for White People to Talk about Racism. Boston, MA: Beacon Press. Dietrich, David R. 2015. “Racially Charged Cookies and White Scholarships: Anti-­ Affirmative Action Protests on American College Campuses.” Sociological Focus 48, no. 2: 105–­25. Dillon, Sam. 2011. “Students’ Knowledge of Civil Rights History Has Deteriorated, Study Finds.” New York Times, September 28, 2011. Dimock, Michael. 2018. “Defining Generations: Where Millennials End and Post-­ Millennials Begin.” Pew Research Center. March 1, 2018. Accessed July 1, 2018. http:// pewrsr.ch/2GRbL5N. Dovidio, John R. 2001. “On the Nature of Contemporary Prejudice: The Third Wave.” Journal of Social Issues 57, no. 4: 829–­49. Dowden, Sue, and John P. Robinson. 1993. “Age and Cohort Differences in American Racial Attitudes: The Generational Replacement Hypothesis Revisited.” In Prejudice, Politics and the American Dilemma, edited by Paul M. Sniderman, Philip E. Tetlock, and Edward G. Carmines, 86–­103. Stanford: Stanford University Press. Duane, Alwin, and Jacqueline Scott. 1996. “Attitude Change: Its Measurement and Interpretation Using Longitudinal Surveys.” In Understanding Change in Social Attitudes, edited by Bridget Taylor and Katarina Thomson, 75–­106. Aldershot: Dartmouth Publishing. Du Bois, W. E. B. 2007. Black Reconstruction in America: An Essay toward a History of the Part Which Black Folk Played in the Attempt to Reconstruct Democracy in America, 1860–­1880. New York: Oxford University Press. Edelman, Lauren B., Sally Riggs Fuller, and Iona Mara-Drita. 2001. “Diversity Rhetoric and the Managerialization of Law1.” American Journal of Sociology 106, no. 6: 1589–­641. Edwards, Breanna. 2013. “Map Shows Where the Most Racist and Homophobic Americans Live.” The Root, November 27, 2013. https://www.theroot.com/map-shows -where-the-most-racist-and-homophobic-american-1790899102. Eibach, Richard P., and Joyce Ehrlinger. 2006. “‘Keep Your Eyes on the Prize’: Reference Points and Racial Differences in Assessing Progress toward Equality.” Personality and Social Psychology Bulletin 32, no. 1: 66–­77. Embrick, David G. 2011. “The Diversity Ideology in the Business World: A New Oppression for a New Age.” Critical Sociology 37, no. 5: 541–­56. Feagin, Joe. 2013. Systemic Racism: A Theory of Oppression. New York: Routledge. Federico, Christopher M. 2006. “Race, Education, and Individualism Revisited.” Journal of Politics 68, no. 3: 600–­610. Feldman, Stanley, and John Zaller. 1992. “The Political Culture of Ambivalence: Ideological Responses to the Welfare State.” American Journal of Political Science 32, no. 1: 268–­307.

302 / Works Cited Firebaugh, Glenn, and Kenneth E. Davis. 1988. “Trends in Anti-­Black Prejudice, 1972–­ 1984: Region and Cohort Effects.” American Journal of Sociology 94, no. 2: 251–­72. Fisher v. University of  Texas at Austin. 2013. 570 U.S. ___, docket no. 11-­345. https://www .law.cornell.edu/supct/cert/11-345. Forman, Tyrone A. 2004. “Color-­Blind Racism and Racial Indifference: The Role of Racial Apathy in Facilitating Enduring Inequalities.” In The Changing Terrain of Race and Ethnicity, edited by Maria Krysan and Amanda E. Lewis, 43–­66. New York: Russell Sage Foundation. Forman, Tyrone A., and Amanda E. Lewis. 2015. “Beyond Prejudice? Young Whites’ Racial Attitudes in Post–­Civil Rights America, 1976 to 2000.” American Behavioral Scientist 59, no. 11: 1394–­428. Frankenberg, Ruth. 1993. White Women, Race Matters: The Social Construction of Whiteness. Minneapolis: University of Minnesota. GenForward Project. 2016. 2016 Election Survey, October. A Survey of the GenForward Project at the University of Chicago. Accessed on January 1, 2018. https://genforward survey.com. Gilens, Martin. 1999. Why Americans Hate Welfare: Race, Media and the Politics of Antipoverty Policy. Chicago: University of Chicago Press. Goff, Phillip Atiba, Jennifer L. Eberhardt, Melissa J. Williams, and Matthew Christian Jackson. 2008. “Not Yet Human: Implicit Knowledge, Historical Dehumanization, and Contemporary Consequences.” Journal of Personality and Social Psychology 94, no. 2: 292–­306. Goff, Phillip Atiba, Matthew Christian Jackson, Brooke Allison Lewis Di Leone, Carmen Marie Culotta, and Natalie Ann DiTomasso. 2014. “The Essence of Innocence: Consequences of Dehumanizing Black Children.” Journal of Personality and Social Psychology 106, no. 4: 526–­45. Goldfarb, Zachary A. 2014. “These Four Charts Show How the SAT Favors Rich, Educated Families.” Wonkblog. Washington Post, March 5, 2014. https://www.washingtonpost.com /news/wonk/wp/2014/03/05/these-four-charts-show-how-the-sat-favors-the-rich -educated-families/?utm_term=.0385fea989bc. Goldring, Ellen, and Claire Smrekar. 2000. “Magnet Schools and the Pursuit of Racial Balance.” Education and Urban Society 33, no. 1: 17–­35. Goldsmith, Pat Rubio. 2010. “Learning Apart, Living Apart: How the Racial and Ethnic Segregation of Schools and Colleges Perpetuates Residential Segregation.” Teachers College Record 112, no. 6: 1602–­30. Granberg, Donald, and Beth Wellman Granberg. 1980. “Abortion Attitudes, 1965–­1980: Trends and Determinants.” Family Planning Perspectives 12, no. 5: 250–­61. Grant, Tobin. 2014. “Poll: Most Whites Say Blacks Are Lazier or Less Intelligent than Whites.” Corner of Church and State, Religion News Service, December 8, 2014. https://religion news.com/2014/12/08/poll-whites-say-blacks-lazier-less-intelligent-whites-3-graphs/. Gratz v. Bollinger. 2003. 539 U.S. 244. https://www.law.cornell.edu/supremecourt/text /02-516. Grutter v. Bollinger. 2003. 539 U.S. 306. https://www.law.cornell.edu/supremecourt /text/539/306. Guinier, Lani. 2015. The Tyranny of the Meritocracy: Democratizing Higher Education in America. Boston: Beacon Press. Hagerman, Margaret A. 2018. White Kids: Growing Up with Privilege in a Racially Divided America. New York: NYU Press. Hall, Stuart. 1996. “Race, Articulation, and Societies Structured in Dominance.” In Black British Cultural Studies: A Reader, edited by Houston A. Baker Jr., Manthia Diawara, and Ruth H. Lindeborg, 16–­60. Chicago: University of Chicago Press.

Works Cited / 303 Hamilton, Darrick, William Darity, Anne Price, Vishnu Sridharan, and Rebecca Tippett. 2015. Umbrellas Don’t Make It Rain: Why Studying and Working Hard Isn’t Enough for Black Americans. Oakland, CA: Insight Center for Community Economic Development. Hancock, Ange-­Marie. 2011. Solidarity Politics for Millennials: A Guide to Ending the Oppression Olympics. New York: Palgrave Macmillan. Hannah-­Jones, Nikole. 2013. “What Abigail Fisher’s Affirmative Action Case Is Really About.” ProPublica, June 23, 2013. https://www.propublica.org/article/a-colorblind -constitution-what-abigail-fishers-affirmative-action-case-is-r. Hannah-­Jones, Nikole. 2014. “Segregation Now: Investigating America’s Racial Divide.” ProPublica, April 16, 2014. https://www.propublica.org/article/segregation-now-full-text. Harris-­Perry, Melissa V. 2011. Sister Citizen: Shame, Stereotypes and Black Women in America. New Haven, CT: Yale University Press. Hartley, Matthew, and Christopher C. Morphew. 2008. “What’s Being Sold and to What End? A Content Analysis of College Viewbooks.” Journal of Higher Education 79, no. 6: 671–­91. Hartmann, Douglas, Joseph Gerteis, and Paul R. Croll. 2009. “An Empirical Assessment of Whiteness Theory: Hidden from How Many?” Social Problems 56, no. 3: 403–­24. Henry, P. J., and David O. Sears. 2002. “The Symbolic Racism 2000 Scale.” Political Psychology 23, no. 2: 253–­83. Herek, Gregory M. 1988. “Heterosexuals’ Attitudes toward Lesbians and Gay Men: Correlates and Gender Differences.” Journal of Sex Research 25, no. 4: 451–­77. Herek, Gregory M., and Eric K. Glunt. 1993. “Interpersonal Contact and Heterosexuals’ Attitudes toward Gay Men: Results from a National Survey.” Journal of Sex Research 30, no. 3: 239–­44. Hetherington, Marc J. 1998. “The Political Relevance of Political Trust.” American Political Science Review 92, no. 4: 791–­808. Hetherington, Marc J. 2005. Why Trust Matters: Declining Political Trust and the Demise of American Liberalism. Princeton, NJ: Princeton University Press. Hicks, Gary R., and Tien-­Tsung Lee. 2006. “Public Attitudes toward Gays and Lesbians.” Journal of Homosexuality 51, no. 2: 57–­77. Hochschild, Jennifer L. 1981. What’s Fair? American Beliefs about Distributive Justice. Cambridge, MA: Harvard University Press. Hochschild, Jennifer L., Vesla Weaver, and Traci Burch. 2012. Creating a New Racial Order: How Immigration, Multiracialism, Genomics, and the Young Can Remake Race in America. Princeton, NJ: Princeton University Press. Hunt, Matthew O. 2007. “African American, Hispanic, and White Beliefs about Black/ White Inequality, 1977–­2004.” American Sociological Review 72, no. 3: 390–­415. Hutchings, Vincent L. 2009. “Change or More of the Same? Evaluating Racial Attitudes in the Obama Era.” Political Opinion Quarterly 73, no. 5: 917–­42. Hutchings, Vincent L., and Nicholas A. Valentino. 2004. “The Centrality of Race in American Politics.” Annual Review of Political Science 7: 383–­408. Ingraham, Christopher. 2014. “Three Quarters of Whites Don’t Have Non-­White Friends.” Washington Post, August 25, 2014. http://www.washingtonpost.com/blogs/wonkblog/wp /2014/08/25/three-quarters-of-whites-dont-have-any-non-white-friends/. Jackman, Mary R. 1994. The Velvet Glove: Paternalism and Conflict in Gender, Class, and Race Relations. Berkeley: University of California Press. Jackson, John L., Jr. 2008. Racial Paranoia: The Unintended Consequences of Political Correctness. New York: Basic Civitas. Jardina, Ashley. 2019. White Identity Politics. New York: Cambridge University Press.

304 / Works Cited Jardina, Ashley, and Spencer Piston. 2016. “Dehumanization and the Role of Biological Racism in American Politics.” Presented at the Midwest Political Science Association Conference, Chicago, IL, April 7, 2016. Jennings, M. Kent, and Richard G. Niemi. 1968. “The Transmission of Political Values from Parent to Child.” American Political Science Review 62, no. 1: 169–­84. Jensen, Robert. 2011. “Whiteness.” In The Routledge Companion to Race and Ethnicity, edited by Stephen M. Caliendo and Charlton D. McIlwain, 21–­28. New York: Routledge. Jimenez, Toms. 2017. The Other Side of Assimilation: How Immigrants Are Changing American Life. Oakland: University of California Press. Jones, Robert P., and Daniel Cox. 2012. “A Generation in Transition: Religion, Values and Politics among College-­Age Millennials.” Public Religion Research Institute, April 9, 2012. https://www.prri.org/research/millennial-values-survey-2012/. Jordan, Dave. 2014. “Exit Poll Angers Some SC Voters.” WSPA. November 5, 2014. https:// www.wspa.com/story/27294393/exit-poll-angers-some-sc-voters. Joseph, Suad, Benjamin D’Harlingue, and Alvin Ka Hin Wong. 2008. “Arab Americans and Muslim Americans in the New York Times, Before and After 9/11.” In Race and Arab Americans before and after 9/11: From Invisible Citizens to Visible Subjects, edited by Amaney Jamal and Nadine Naber, 229–­75. Syracuse, NY: Syracuse University Press. Joseph, Tiffany D. 2015. Race on the Move: Brazilian Migrants and the Global Reconstruction of Race. Palo Alto, CA: Stanford University Press. Junn, Jane. 2006. “Mobilizing Group Consciousness: When Does Ethnicity Have Political Consequences?” In Transforming Politics, Transforming America, edited by Taeku Lee, Karthick Ramakrishnan, and Ricardo Ramirez, 32–­47. Charlottesville: University of Virginia Press. Junn, Jane. 2007. “From Coolie to Model Minority: US Immigration Policy and the Construction of Racial Identity.” Du Bois Review: Social Science Research on Race 4, no. 2: 355–­73. Kreitzer, Rebecca, Candis Watts Smith, and Feiya Suo. 2018. “Dynamics of Racial Resentment across the 50 States.” Presented at the Midwest Political Science Association Conference, Chicago, IL, April 8, 2018. Katznelson, Ira. 2005. When Affirmative Action Was White: An Untold History of Racial Inequality in Twentieth-­Century America. New York: W. W. Norton. Keeter, Scott, Juliana Horowitz, and Alec Tyson. 2008. “Young Voters in the 2008 Election.” Pew Research Center, November 13, 2008. https://www.pewresearch.org/2008/11/13 /young-voters-in-the-2008-election/. Kendi, Ibram X. 2016. Stamped from the Beginning: The Definitive History of Racist Ideas in America. New York: Nation Books. Kinder, Donald R. 1986. “The Continuing American Dilemma: White Resistance to Racial Change 40 Years after Myrdal.” Journal of Social Issues 42, no. 2: 151–­71. Kinder, Donald R., and Tali Mendelberg. 2000. “Individualism Reconsidered: Principles and Prejudice in Contemporary American Opinion.” In Racialized Politics: The Debate about Racism in America, edited by David O. Sears, Jim Sidanius, and Lawrence Bobo, 44–­74. Chicago: University of Chicago Press. Kinder, Donald R., and Lynn M. Sanders. 1996. Divided by Color: Racial Politics and Democratic Ideals. Chicago: University of Chicago Press. Kinder, Donald R., and David O. Sears. 1981. “Prejudice and Politics: Symbolic Racism versus Racial Threats to the Good Life.” Journal of Personality and Social Psychology 40, no. 3: 414–­31. Kluegel, James R. 1990. “Trends in Whites’ Explanations of the Black-­White Gap in Socioeconomic Status, 1977–­1989.” American Sociological Review 55, no. 4: 512–­25.

Works Cited / 305 Kluegel, James R., and Lawrence Bobo. 1993. “Dimensions of Whites’ Beliefs about the Black-­White Socioeconomic Gap.” In Prejudice, Politics, and the American Dilemma, edited by Paul M. Sniderman, Philip E. Tetlock, and Edward G. Carmines, 127–­47. Stanford, CA: Stanford University Press. Kluegel, James R., and Eliot R. Smith. 1981. “Beliefs about Stratification.” Annual Review of Sociology 7: 29–­56. Kluegel, James R., and Eliot R. Smith. 1983. “Affirmative Action Attitudes: Effects of Self-­ Interest, Racial Affect, and Stratification Beliefs on Whites’ Views.” Social Forces 61, no. 3: 797–­824. Kluegel, James R., and Eliot R. Smith. 1986. Beliefs about Inequality: Americans’ Views of What Is and What Ought to Be. New York: Aldine De Gruyter. Kozol, Jonathan. 2005. The Shame of the Nation: The Restoration of Apartheid Schooling in America. New York: Broadway Paperback. Kreitzer, Rebecca J., Allison J. Hamilton, and Caroline J. Tolbert. 2014. “Does Policy Adoption Change Opinions on Minority Rights? The Effects of Legalizing Same-­Sex Marriage.” Political Research Quarterly 67, no. 4: 795–­808. Lax, Jeffrey R., and Justin H. Phillips. 2009. “Gay Rights in the States: Public Opinion and Policy Responsiveness.” American Political Science Review 103, no. 3: 367–­86. Lebron, Christopher J. 2017. The Making of Black Lives Matter: A Brief History of an Idea. New York: Oxford University Press. Levinson, Alana. 2013. “Polls Show Wide Racial Gap on Trayvon Martin Case.” Natio­nal Public Radio, July 22, 2013. https://www.npr.org/sections/itsallpolitics/2013/07/22 /204595068/polls-show-wide-racial-gap-on-trayvon-martin-case. Lewis, Amanda E. 2001. “There Is No ‘Race’ in the Schoolyard: Color-­Blind Ideology in an (Almost) All-­White School.” American Educational Research Journal 38, no. 4: 781–­811. Lewis, Amanda E. 2004. “‘What Group?’ Studying Whites and Whiteness in the Era of ‘Color-Blindness.’” Sociological Theory 22, no. 4: 623–­46. Loftus, Jeni. 2001. “America’s Liberalization in Attitudes toward Homosexuality, 1973 to 1998.” American Sociological Review 66, no. 5: 762–­82. López, Ian Haney. 2015. Dog Whistle Politics: How Coded Racial Appeals Have Reinvented Racism and Wrecked the Middle Class. New York: Oxford University Press. Lopez Bunyasi, Tehama, and Candis Watts Smith. 2019a. “Do All Black Lives Matter to Black People?: Respectability Politics and the Limitations of Linked Fate.” Journal of Race, Ethnicity, and Politics 4, no. 1: 180–­215. Lopez Bunyasi, Tehama, and Candis Watts Smith. 2019b. Stay Woke: A People’s Guide to Making All Black Lives Matter. New York: NYU Press. Luhby, Tami. 2018. “Millennials Born in 1980s May Never Recover from the Great Recession.” CNN Business, May 23, 2018. https://money.cnn.com/2018/05/22/news/econ ­omy/1980s-millennials-great-recession-study/index.html. Mannheim, Karl. 1952. “The Problem of Generations.” In Essays on the Sociology of Knowledge, edited by Karl Mannheim and Paul Kecskemeti, 276–­320. New York: Oxford University Press. Marx, Anthony W. 1998. Making Race and Nation: A Comparison of South Africa, the United States, and Brazil. New York: Cambridge University Press. Mason, Paul. 2016. “The Racist Hijacking of Microsoft’s Chatbot Shows How the Internet Teems with Hate.” The Guardian, March 29, 2016. http://www.theguardian.com /world/2016/mar/29/microsoft-tay-tweets-antisemitic-racism. Massey, Douglas S., and Nancy A. Denton. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press.

306 / Works Cited Massey, Douglas S., Jacob S. Rugh, Justin P. Steil, and Len Albright. 2016. “Riding the Stagecoach to Hell: A Qualitative Analysis of Racial Discrimination in Mortgage Lending.” City & Community 15, no. 2: 118–­36. Masuoka, Natalie, and Jane Junn. 2013. The Politics of Belonging: Race, Public Opinion, and Immigration. Chicago: University of Chicago Press. Mathias, Christopher. 2014. “Police Quotas Are Terrible, and the NYPD Still Seems to Be Using Them.” Huff Post, October 2, 2014. http://www.huffingtonpost.com/2014/10/02 /nypd-quotas_n_5916596.html. Mayorga-­Gallo, Sarah. 2014. Behind the White Picket Fence: Power and Privilege in a Multiethnic Neighborhood. Chapel Hill: University of North Carolina Press. Mayorga-­Gallo, Sarah. 2018. “‘People Were Not as Friendly as I Had Hoped’: Black Residential Experiences in Two Multiracial Neighborhoods.” In Black Politics in Transition: Immigration, Suburbanization, and Gentrification, edited by Candis Watts Smith and Christina M. Greer, 235–­62. New York: Routledge. Mayorga-­ Gallo, Sarah. 2019. “The White-­ Centering Logic of Diversity Ideology.” American Behavioral Scientist. Published ahead of print, April 18, 2019. https://doi .org/10.1177/0002764219842619. McConahay, John B., Betty B. Hardee, and Valerie Batts. 1981. “Has Racism Declined in America? It Depends on Who Is Asking and What Is Being Asked.” Journal of Conflict Resolution 25, no. 4: 563–­79. McDermott, Monica, and Frank L. Samson. 2005. “White Racial and Ethnic Identity in the United States.” Annual Review of Sociology 31: 245–­61. McIntosh, Peggy. 1992. “White Privilege: Unpacking the Invisible Knapsack.” In Multi­ culturalism 1992, edited by Anna May Filor, 30–­36. New York: New York State Coun­ cil of Educational Associations. https://files.eric.ed.gov/fulltext/ED355141.pdf?utm #page=43. Meredith, William. 1993. “Measurement Invariance, Factor Analysis and Factorial Invariance.” Psychometrika 58, no. 4: 525–­43. Mills, Charles. 2007. “White Ignorance.” In Race and Epistemologies of Ignorance, edited by Shannon Sullivan and Nancy Tuana, 13–­28. Albany: State University of New York Press. Millsap, Roger E. 2012. Statistical Approaches to Measurement Invariance. New York: Routledge. Moore, Wendy Leo, and Joyce M. Bell. 2011. “Maneuvers of Whiteness: ‘Diversity’ as a Mechanism of Retrenchment in the Affirmative Action Discourse.” Critical Sociology 37, no. 5: 597–­613. Morales, Mark, and Thomas Tracy. 2014. “NYPD Commissioner Bill Bratton Bags Bust Quotas in Favor of ‘Actual Impact.’” New York Daily News, January 28, 2014. http:// www.nydailynews.com/new-york/nyc-crime/nypd-bratton-bags-bust-quotas-favors -actual-impact-article-1.1594816. Morlin, Bill. 2018. “Study Shows Two-­Thirds of U.S. Terrorism Tied to Right-­Wing Extremists.” Southern Poverty Law Center, September 12, 2018. https://www.splcen­ ter.org/hatewatch/2018/09/12/study-shows-two-thirds-us-terrorism-tied-right-wing -extremists. Mosse, George L. 1995. “Racism and Nationalism.” Nations and Nationalism 1, no. 2: 163–­73. MTV Strategic Insights and David Binder Research. 2014. DBR MTV Bias Survey Summary. April 2014. http://d1fqdnmgwphrky.cloudfront.net/studies/000/000/001/DBR_MTV _Bias_Survey_Executive_Summary.pdf?1398858309.

Works Cited / 307 Mueller, Jennifer C. 2017. “Producing Colorblindness: Everyday Mechanisms of White Ignorance.” Social Problems 64, no. 2: 219–­38. Murray, Pauli. 1953. “The Historical Development of Race Laws in the United States.” Journal of Negro Education 22, no. 1: 4–­15. Nagourney, Adam. 2008. “Obama Elected President as Racial Barriers Fall.” New York Times, November 4, 2008, A1. Nagourney, Adam. 2014. “A Defiant Rancher Savors the Audience That Rallied to His Side.” New York Times, April 24, 2014, A1. National Public Radio. 2018. “Jason Kessler on His ‘Unite the Right’ Rally Move to D.C.” National Public Radio, August 10, 2018. https://www.npr.org/2018/08/10/637390626 /a-year-after-charlottesville-unite-the-right-rally-will-be-held-in-d-c. Neville, Helen A., Roderick L. Lilly, Georgia Durham, Richard M. Lee, and LaVonne Browne. 2000. “Construction and Initial Validation of the Color-­Blind Racial Attitudes Scale (CoBRAS).” Journal of Counseling Psychology 47, no. 1: 59–­70. New York Civil Liberties Union. 2011. “Stop and Frisk Facts.” 2011. Accessed August 20, 2015. https://www.nyclu.org/en/stop-and-frisk-facts. Newport, Frank. 2018. “Democrats More Positive about Socialism than Capitalism.” Gallup, August 13, 2018. https://news.gallup.com/poll/240725/democrats-positive -socialism-capitalism.aspx. Noble, Safiya Umoja. 2018. Algorithms of Oppression: How Search Engines Reinforce Racism. New York: NYU Press. Norris, Michele. 2018. “As America Changes, Some Anxious Whites Feel Left Behind.” National Geographic, April 2018. https://www.nationalgeographic.com/magazine/2018 /04/race-rising-anxiety-white-america/. Norton, Michael I., and Samuel R. Sommers. 2011. “Whites See Racism as a Zero-­Sum Game That They Are Now Losing.” Perspectives on Psychological Science 6, no. 3: 215–­18. Nteta, Tatishe M., and Jill S. Greenlee. 2013. “A Change Is Gonna Come: Generational Membership and White Racial Attitudes in the 21st Century.” Political Psychology 34, no. 6: 877–­97. Obama, Barack. 2013. “Inaugural Address by President Barack Obama.” United States Capitol, Washington, DC, January 21, 2013. https://obamawhitehouse.archives.gov /the-press-office/2013/01/21/inaugural-address-president-barack-obama. Obama, Barack. 2015. “Episode 613: President Barack Obama,” June 22, 2015, in WTF with Marc Maron, edited by Marc Maron, podcast. https://www.stitcher.com/pod cast/marc-maron/wtf-with-marc-maron-podcast-113810/e/episode-613-president -barack-obama-39454542. Obama, Barack. 2016. “Remarks by the President at Howard University Commencement Ceremony.” Howard University, Washington, DC, May 7, 2016. https://obamawhite house.archives.gov/the-press-office/2016/05/07/remarks-president-howard-university -commencement-ceremony. Obama, Barack. 2017. “Presidential Farewell Address.” McCormick Place, Chicago, IL, January 10, 2017. https://obamawhitehouse.archives.gov/farewell. O’Brien, Eileen. 2008. The Racial Middle: Latinos and Asian Americans Living Beyond the Racial Divide. New York: NYU Press. Oliver, Melvin, and Thomas Shapiro. 1995. Black Wealth/White Wealth: A New Perspective on Racial Inequality. New York: Routledge. Olshansky, S. Jay, Toni Antonucci, Lisa Berkman, Robert H. Binstock, Axel Boersch-­ Supan, John T. Cacioppo, Bruce A. Carnes, Laura L. Carstensen, Linda P. Fried, and Dana P. Goldman. 2012. “Differences in Life Expectancy Due to Race and Educational

308 / Works Cited Differences Are Widening, and Many May Not Catch Up.” Health Affairs 31, no. 8: 1803–­13. Omi, Michael, and Howard Winant. 1994. Racial Formation in the United States: From the 1960s to the 1990s. 2nd ed. New York: Routledge. Page, Susan, and Jenny Ung. 2016. “Poll Shows that Millennials Would Flock to Clinton against Trump.” USA Today, March 14, 2016. http://www.usatoday.com/story/news/pol itics/elections/2016/03/14/poll-millennials-clinton-sanders-trump-president/81612520/. Pager, Devah. 2007. Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration. Chicago: University of Chicago Press. Pager, Devah, and Hana Shepherd. 2008. “The Sociology of Discrimination: Racial Discrimination in Employment, Housing, Credit, and Consumer Markets.” Annual Review of Sociology 34: 181–­209. Parker, Christopher S., and Matt A. Barreto. 2013. Change They Can’t Believe In: The Tea Party and Reactionary Politics in America. Princeton, NJ: Princeton University Press. Pendergrass, Sabrina. 2013. “Routing Black Migration to the Urban US South: Social Class and Sources of Social Capital in the Destination Selection Process.” Journal of Ethnic and Migration Studies 39, no. 9: 1441–­59. Pérez, Efrén O. 2009. “Lost in Translation? Item Validity in Bilingual Political Surveys.” Journal of Politics 71, no. 4: 1530–­48. Pérez, Efrén O. 2016. Unspoken Politics: Implicit Attitudes and Political Thinking. New York: Cambridge University Press. Pérez, Efrén O., and Marc J. Hetherington. 2014. “Authoritarianism in Black and White: Testing the Cross-­Racial Validity of the Child Rearing Scale.” Political Analysis 22, no. 3: 398–­412. Perlstein, Rick. 2012. “Exclusive: Lee Atwater’s Infamous 1981 Interview on the Southern Strategy.” The Nation, November 13, 2012. https://www.thenation.com/article /exclusive-lee-atwaters-infamous-1981-interview-southern-strategy/. Pettigrew, T. F., and R. W. Meertens. 1995. “Subtle and Blatant Prejudice in Western Europe.” European Journal of Social Psychology 25, no. 1: 57–­75. Pettigrew, Thomas F. 2000. “Systematizing the Predictors of Prejudice.” In Racialized Politics: The Debate about Racism in America, edited by David O. Sears, Jim Sidanius, and Lawrence Bobo, 280–­301. Chicago: University of Chicago Press. Pettigrew, Thomas F. 2017. “Social Psychological Perspectives on Trump Supporters.” Journal of Social and Political Psychology 5, no. 1: 107–­16. Pew Research Center. 2005. “The Black and White of Public Opinion.” Pew Research Center, October 31, 2005. http://www.people-press.org/2005/10/31/the-black-and -white-of-public-opinion/. Pew Research Center. 2009. “Public Backs Affirmative Action, but Not Minority Preferences.” Pew Research Center, June 2, 2009. http://www.pewresearch.org/2009/06/02 /public-backs-affirmative-action-but-not-minority-preferences/. Pew Research Center. 2010. “Millennials: Confident. Connected. Open to Change.” February 24, 2010. https://www.pewsocialtrends.org/2010/02/24/millennials-confident -connected-open-to-change/. Pew Research Center. 2014. “Stark Racial Divisions in Reactions to Ferguson Police Shooting.” Pew Research Center, August 18, 2014. http://www.people-press.org/2014/08/18 /stark-racial-divisions-in-reactions-to-ferguson-police-shooting/. Pew Research Center. 2015. “Comparing Millennials to the Other Generations.” March 19, 2015. http://www.pewsocialtrends.org/2015/03/19/comparing-millennials-to-other -generations/.

Works Cited / 309 Picca, Leslie Houts, and Joe R. Feagin. 2007. Two-­Faced Racism: Whites in the Backstage and Frontstage. New York: Routledge. Piston, Spencer. 2010. “How Explicit Racial Prejudice Hurt Obama in the 2008 Election.” Political Behavior 32, no. 4: 431–­51. Pitofsky, Marina. 2018. “Are Women Changing ‘Unite the Right’ or Just ‘Rebranding’ the Movement?” USA Today, August 9, 2018. https://www.usatoday.com/story/news /2018/08/09/women-unite-right-rally-2018/874631002/. Plant, E. Ashby, and B. Michelle Peruche. 2005. “The Consequences of Race for Police Officers’ Responses to Criminal Suspects.” Psychological Science 16, no. 3: 180–­83. Poteat, V. Paul, and Lisa B. Spanierman. 2008. “Further Validation of the Psychosocial Costs of Racism to Whites Scale among Employed Adults.” Counseling Psychologist 36, no. 6: 871–­94. Pugliese, Joseph. 2002. “Race as Category Crisis: Whiteness and the Topical Assignation of Race.” Social Semiotics 12, no. 2: 149–­68. Rampell, Catherine. 2009. “SAT Scores and Family Income.” Economix (blog), New York Times, August 27, 2009. https://economix.blogs.nytimes.com/2009/08/27/sat-scores -and-family-income/. Ransby, Barbara. 2018. Making All Black Lives Matter: Reimagining Freedom in the Twenty-­ First Century. Oakland: University of California Press. Reardon, Sean F., and Ann Owens. 2014. “60 Years after Brown: Trends and Consequences of School Segregation.” Annual Review of Sociology 40: 199–­218. Regents of the University of California v. Bakke. 1978. 438 U.S. 265. https://www.law .cornell.edu/wex/regents_of_the_university_of_california_v_bakke_%281978%29. Ross, Stephen L., and Margery Austin Turner. 2005. “Housing Discrimination in Metropolitan America: Explaining Changes between 1989 and 2000.” Social Problems 52, no. 2: 152–­80. Rothstein, Richard. 2017. The Color of Law: A Forgotten History of How Our Government Segregated America. New York: Liveright Publishing Corporation. Rouse, Stella M., and Ashley D. Ross. 2018. The Politics of Millennials: Political Beliefs and Policy Preferences of America’s Most Diverse Generation. Ann Arbor: University of Michigan Press. Rugh, Jacob S., Len Albright, and Douglas S. Massey. 2015. “Race, Space, and Cumulative Disadvantage: A Case Study of the Subprime Lending Collapse.” Social Problems 62, no. 2: 186–­218. Ruiz-­Grossman, Sarah. 2017. “Most of America’s Terrorists Are White, and Not Muslim.” Huff Post, June 23, 2017. https://www.huffingtonpost.com/entry/domestic-terrorism -white-supremacists-islamist-extremists_us_594c46e4e4b0da2c731a84df. Rutten, Tim. 2008. “The Good Generation Gap.” Los Angeles Times, February 6, 2008. http://articles.latimes.com/2008/feb/06/opinion/oe-rutten6. Ryder, Norman B. 1965. “The Cohort as a Concept in the Study of Social Change.” American Sociological Review 30, no. 6: 843–­61. Saad, Lydia. 2009. “U.S. Abortion Attitudes Closely Divided.” Gallup, August 4, 2009. https://news.gallup.com/poll/122033/u.s.-abortion-attitudes-closely-divided.aspx. Sawyer, Mark Q. 2005. Racial Politics in Post-­Revolutionary Cuba. New York: Cambridge University Press. Schaefer, Richard T. 1990. Racial and Ethnic Groups. Glenview, IL: Little Brown Higher Education. 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, no. 1: 9–­34.

310 / Works Cited Schildkraut, Deborah J. 2017. “White Attitudes about Descriptive Representation in the US: The Roles of Identity, Discrimination, and Linked Fate.” Politics, Groups, and Identities 5, no. 1: 84–­106. Schmidt, Peter. 2007. Color of Money: How Rich White Kids Are Winning the War over College Affirmative Action. New York: Palgrave MacMillan. Schneider, Anne, and Helen Ingram. 1993. “Social Construction of Target Populations: Implications for Politics and Policy.” American Political Science Review 87, no. 2: 334–­47. Schofield, Janet Ward. 1986. “Causes and Consequences of the Colorblind Perspective.” In Prejudice, Discrimination, and Racism, edited by John R. Dovidio and Samuel L. Gaertner, 91–­126. New York: Academic Press. Schuman, Howard, Charlotte Steeh, Lawrence Bobo, and Maria Krysan. 1997. Racial Attitudes in America: Trends and Interpretations. Revised ed. Cambridge, MA: Harvard University Press. Schwadel, Philip. 2011. “Age, Period, and Cohort Effects on Religious Activities and Beliefs.” Social Science Research 40, no. 1: 181–­92. Schwadel, Philip, and Michael Stout. 2012. “Age, Period and Cohort Effects on Social Capital.” Social Forces 9, no. 1: 233–­52. Scott, Eugene. 2018. “Six Times President Trump Said He Is the Least Racist Person.” Washington Post. January 17, 2018. https://www.washingtonpost.com/news/the-fix/wp /2018/01/17/six-times-president-trump-said-he-is-the-least-racist-person/?noredirect =on&utm_term=.05ee25ca65fa. Sears, David O. 1988. “Symbolic Racism.” In Eliminating Racism: Profiles in Controversy, edited by Phyllis A. Katz and Dalmas A. Taylor, 53–­84. New York: Plenum Press. Sears, David O. 1993. “Symbolic Politics: A Socio-­Psychological Theory.” In Explorations in Political Psychology, edited by Shanto Iyengar and William J. McGuire, 113–­49. Durham, NC: Duke University Press. Sears, David O., and P. J. Henry. 2003. “The Origins of Symbolic Racism.” Journal of Personality and Social Psychology 85, no. 2: 259–­75. Sears, David O., John J. Hetts, Jim Sidanius, and Lawrence Bobo. 2000. “Race in American Politics: Framing the Debates.” In Racialized Politics: The Debate about Racism in America, edited by David O. Sears, Jim Sidanius, and Lawrence Bobo, 1–­43. Chicago: University of Chicago Press. Sears, David O., and Donald R. Kinder. 1970. “The Good Life, ‘White Racism,’ and the Los Angeles Voter.” Paper presented at the conference of the Western Psychological Association, Los Angeles, CA, 1970. Sears, David O., Richard R. Lau, T. Tyler, and H. Allen. 1980. “Self-­Interest vs. Symbolic Politics in Policy Attitudes and Presidential Voting.” American Political Science Review 74, no. 3: 670–­84. Sears, David O., Jim Sidanius, and Lawrence Bobo. 2000. Racialized Politics: The Debate about Racism in America. Chicago: University of Chicago Press. Sides, John, Michael Tesler, and Lynn Vavreck. 2017. “The 2016 US Election: How Trump Lost and Won.” Journal of Democracy 28, no. 2: 34–­44. Smith, Candis Watts. 2014. “Shifting from Structural to Individual Attributions of Black Disadvantage: Age, Period, and Cohort Effects on Black Explanations of Racial Disparities.” Journal of Black Studies 45, no. 5: 432–­52. Smith, Candis Watts, and Sarah Mayorga-­Gallo. 2017. “The New Principle-­Policy Gap: How Diversity Ideology Subverts Diversity Initiatives.” Sociological Perspectives 60, no. 5: 889–­911.

Works Cited / 311 Smith, Elsie J. 1991. “Ethnic Identity Development: Toward a Development of a Theory within the Context of Majority/Minority Status.” Journal of Counseling and Development 70, no. 1: 181–­88. Smith, Tom W. 1990. “Liberal and Conservative Trends in the United States since World War II.” Public Opinion Quarterly 54, no. 4: 479–­507. Smith, Tom W., Peter V. Marsden, and Michael Hout. 2011. General Social Surveys, 1972–­ 2010 Cumulative Codebook. Chicago: National Opinion Research Center. Sniderman, Paul M., and Richard A. Brody. 1977. “Coping: The Ethic of Self-­Reliance.” American Journal of Political Science 21, no. 3: 501–­21. Sniderman, Paul M., Richard A. Brody, and James H. Kuklinski. 1984. “Policy Reasoning and Political Values: The Problem of Racial Equality.” American Journal of Political Science 21, no. 3: 75–­94. Sniderman, Paul M., Gretchen C. Crosby, and William G. Howell. 2000. “The Politics of Race.” In Racialized Politics: The Debate about Racism in America, edited by David O. Sears, Jim Sidanius, and Lawrence Bobo, 236–­79. Chicago: University of Chicago Press. Sniderman, Paul M., and Thomas Leonard Piazza. 1993. The Scar of Race. Cambridge, MA: Belknap Press of Harvard University Press. Sniderman, Paul M., and Philip E. Tetlock. 1986. “Symbolic Racism: Problems of Motive Attribution in Political Analysis.” Journal of Social Issues 42, no. 2: 129–­50. Soss, Joe, Richard C. Fording, and Sanford F. Schram. 2008. “The Color of Devolution: Race, Federalism, and the Politics of Social Control.” American Journal of Political Science 52, no. 3: 536–­53. Soss, Joe, Richard C. Fording, and Sanford F. Schram. 2011. Disciplining the Poor: Neoliberal Paternalism and the Persistent Power of Race. Chicago: University of Chicago Press. Southern Poverty Law Center. 2011. Teaching the Movement: The State of Civil Rights Education in the United States 2011. Montgomery: Southern Poverty Law Center. Spanierman, Lisa B., and Mary J. Heppner. 2004. “Psychosocial Costs of Racism to Whites Scale (PCRW): Construction and Initial Validation.” Journal of Counseling Psychology 51, no. 2: 249–­62. Spanierman, Lisa B., V. Paul Poteat, Amanda M. Beer, and Patrick Ian Armstrong. 2006. “Psychosocial Costs of Racism to Whites: Exploring Patterns through Cluster Analysis.” Journal of Counseling Psychology 53, no. 4: 434–­41. Steeh, Charlotte, and Howard Schuman. 1992. “Young White Adults: Did Racial Attitudes Change in the 1980s?” American Journal of Sociology 98, no. 2: 340–­67. Stegmueller, Daniel. 2011. “Apples and Oranges? The Problem of Equivalence in Comparative Research.” Political Analysis 19, no. 4: 471–­87. Stoker, Laura. 1998. “Understanding Whites’ Resistance to Affirmative Action: The Role of Principled Commitments and Racial Prejudice.” In Perception and Prejudice: Race and Politics in the United States, edited by Jon Hurwitz and Mark Peffley. New Haven, CT: Yale University Press. Tajfel, Henri, and John C. Turner. 1986. “The Social Identity Theory of Intergroup Behavior.” In Psychology of Intergroup Relations, edited by Stephen Worchel and William G. Austin, 7–­24. Chicago: Nelson-­Hall Publishers. Tate, Katherine. 2010. What’s Going On? Political Incorporation and the Transformation of Black Public Opinion. Washington, DC: Georgetown University Press. Tatum, Beverly Daniel. 2003. “Why Are All the Black Kids Sitting Together in the Cafeteria?”: and Other Conversations about Race. New York: Basic Books. Taylor, Keeanga-­Yamahtta. 2016. From #BlackLivesMatter to Black Liberation. Chicago: Haymarket Books.

312 / Works Cited Taylor, Paul, and Pew Research Center. 2014. The Next America: Boomers, Millennials, and the Looming Generational Showdown. New York: Public Affairs. Tesler, Michael. 2012. “The Spillover of Racialization into Health Care: How President Obama Polarized Public Opinion by Racial Attitudes and Race.” American Journal of Political Science 56, no. 3: 690–­704. Tesler, Michael. 2013. “The Return of Old-­Fashioned Racism to White Americans’ Partisan Preferences in the Early Obama Era.” Journal of Politics 75, no. 1: 110–­23. Tesler, Michael. 2016. Post-­Racial or Most-­Racial? Race and Politics in the Obama Era. Chicago: University of Chicago Press. Tesler, Michael, and David O. Sears. 2010. Obama’s Race: The 2008 Election and the Dream of a Post-­Racial America. Chicago: University of Chicago Press. Thernstrom, Abigail, and Stephen Thernstrom. 2008. “Racial Gerrymandering Is Unnecessary.” Wall Street Journal, November 11. The Sentencing Project. 2017. “The Facts: State-­by-­State Data.” The Sentencing Project. Accessed November 1, 2018. https://www.sentencingproject.org/the-facts/#map?dataset -option=BWR. Thornton, Arland, and Linda Young-DeMarco. 2004. “Four Decades of  Trends in Attitudes toward Family Issues in the United States: The 1960s through the 1990s.” Journal of Marriage and Family 63, no. 4: 1009–­37. Tolson, Jay. 2008. “The Obama Effect: His Success Is Challenging the Rules of Race and Politics.” U.S. News and World Report, February 25, 2008, 38. Tuch, Steven A. 1984. “A Multivariate Analysis of Response Structure: Race Attitudes 1972–­ 1977.” Social Science Research 13, no. 1: 55–­71. Twenge, Jean M., W. Keith Campbell, and Elise C. Freeman. 2012. “Generational Differences in Young Adults’ Life Goals, Concern for Others, and Civic Orientation, 1966–­ 2009.” Journal of Personality and Social Psychology 102, no. 5: 1045–­62. Underhill, Megan R. 2018. “Parenting during Ferguson: Making Sense of White Parents’ Silence.” Ethnic and Racial Studies 41, no. 11: 1934–­51. Unnever, James D., and Francis T. Cullen. 2007. “The Racial Divide in Support for the Death Penalty: Does White Racism Matter?” Social Forces 85, no. 3: 1281–­301. US Census Bureau. 2016. “Table 2. Poverty Status of People by Family Relationship, Race, and Hispanic Origin: 1959 to 2016.” In Historical Poverty Tables: People and Familes -­1959 to 2017. Washington, DC: United States Census Bureau. https://www .census.gov/data/tables/time-series/demo/income-poverty/historical-poverty-people .html. Valentino, Nicholas A., and Ted Brader. 2011. “The Sword’s Other Edge: Perceptions of Discrimination and Racial Policy Opinion after Obama.” Public Opinion Quarterly 75, no. 2: 201–­26. Vega, Irene I. 2014. “Conservative Rationales, Racial Boundaries: A Case Study of Restrictionist Mexican Americans.” American Behavioral Scientist 58, no. 13: 1764–­83. Verba, Sidney, Kay Lehman Schlozman, Henry Brady, and Norman H. Nie. 1993. “Race, Ethnicity, and Political Resources: Participation in the United States.” British Journal of Political Science 23, no. 4: 453–­97. Wallace, George. 1963. “The Inaugural Address of Governor George C. Wallace.” Alabama Capitol, Montgomery, AL, January 14, 1963. http://digital.archives.alabama.gov/cdm /ref/collection/voices/id/2952. Warikoo, Natasha K., and Janine de Novais. 2014. “Colour-­Blindness and Diversity: Race Frames and Their Consequences for White Undergraduates at Elite US Universities.” Ethnic and Racial Studies 38, no. 6: 860–­76.

Works Cited / 313 Wilder, Craig Steven. 2013. Ebony and Ivy: Race, Slavery, and the Troubled History of America’s Universities. New York: Bloomsbury Publishing USA. Wilkerson, Isabel. 2015. “When Will the North Face Its Racism?” New York Times, Jan­ uary 11, 2015, Sunday Review. Wilson, David C., and Darren W. Davis. 2011. “Reexamining Racial Resentment: Conceptualization and Content.” ANNALS of the American Academy of Political and Social Sciences 634: 117–­33. Wilson, Erika. 2018. “Charters and Choice: The New White Flight.” Lecture, UNC Policy Research Group, Chapel Hill, NC, September 7, 2018. Wilson, Thomas C. 1996. “Cohort and Prejudice: Whites’ Attitudes toward Blacks, Hispanics, Jews, and Asians.” Public Opinion Quarterly 60, no. 2: 253–­74. Wilson, Valerie, and William M. Rodgers III. 2016. Black-­White Wage Gaps Expand with Rising Wage Inequality. Washington, DC: Economic Policy Institute. Winograd, Morley, and Michael D. Hais. 2011. Millennial Momentum: How a New Generation Is Remaking America. New Brunswick, NJ: Rutgers University Press. Wise, Tim. 2011. “A Bad Year for White Whine: College Scholarships and the Cult of Caucasian Victimhood.” TimWise.org. March 6, 2011. http://www.timwise.org/2011 /03/a-bad-year-for-white-whine-college-scholarships-and-the-cult-of-caucasian -victimhood/. Wong, Cara, and Grace E. Cho. 2005. “Two-Headed Coins or Kandinskys: White Racial Identification.” Political Psychology 26, no. 5: 699–­720. Van den Berghe, Pierre L. 1967. Race and Racism: A Comparative Perspective. New York: Wiley. Yadon, Nicole, and Spencer Piston. 2018. “Examining Whites’ Anti-­Black Attitudes after Obama’s Presidency.” Politics, Groups, and Identities. March 5, 2018. https://doi.org/10 .1080/21565503.2018.1438953. Yang, Alan S. 1997. “Trends: Attitudes toward Homosexuality.” Public Opinion Quarterly 61, no. 3: 477–­507. Yang, Yang, Wenjiang J. Fu, and Kenneth C. Land. 2004. “A Methodological Comparison of Age-­Period-­Cohort Models: The Intrinsic Estimator and Conventional Generalized Linear Models.” Sociological Methodology 34, no. 1: 75–­110. Yang, Yang, and Kenneth C. Land. 2013. Age-­Period-­Cohort Analysis: New Models, Methods, and Empirical Applications. Boca Raton: Chapman & Hall/CRC. Yang, Yang, Sam Schulhofer-­Wohl, Wenjiang J. Fu, and Kenneth C. Land. 2008. “The Intrinsic Estimator for Age-­Period-­Cohort Analysis: What It Is and How to Use It.” American Journal of Sociology 113, no. 6: 1697–­736. Yinger, John. 1995. Closed Doors, Opportunities Lost: The Continuing Costs of Housing Discrimination. New York: Russell Sage Foundation. Zaller, John, and Stanley Feldman. 1992. “A Simple Theory of the Survey Response: Answering Questions versus Revealing Preferences.” American Journal of Political Science 36, no. 3: 579–­616. Zepeda-­Millán, Chris. 2017. Latino Mass Mobilization: Immigration, Racialization, and Activism. New York: Cambridge University Press. Zepeda-­Millán, Chris, and Sophia J. Wallace. 2013. “Racialization in Times of Contention: How Social Movements Influence Latino Racial Identity.” Politics, Groups, and Identities 1, no. 4: 510–­27. Zogby, John. 2009. “Barack Obama: America’s First Global President.” Campaigns & Elections, February 28, 2009. https://www.campaignsandelections.com/campaign-insider /barack-obama-america-s-first-global-president.

Index

Page numbers in italics refer to figures and tables. abortion rights, 39, 41, 44–­45, 68; age effects and, 45 advantage/disadvantage: Asian Americans and, 108–­9, 128–­29, 138, 167; diversity and, 134–­36, 136–­40; FIRE battery and, 235–­43; knapsack of privilege and, 132–­ 40; Latinx people and, 109, 117, 118, 144, 145, 147, 151, 154, 155; use of term, 8; White millennials and, 132–­40. See also racial privilege affective-­orientation measure: Black Americans and, 54–­57; data and methodological strategies, 48–­49, 190–­91; feeling thermometer and, 40, 49, 54–­55, 91, 93; gay and lesbian Americans and, 46–­47, 54–­57, 286; generational cohorts and, 76–­77, 216–­22, 218, 219, 220; in-­group/ out-­group, 55, 58, 70, 146–­47; period effects and, 56, 58; pro-­White affect, 55, 56, 57, 59, 286; racial resentment and, 91; symbolic politics and, 76–­77; use of term, 27, 49; women’s liberation and, 54–­57, 285. See also PCRW (Psychosocial Costs of Racism to Whites) affirmative action: age effects on, 63–­69; Black and Latinx support for, 24; cohort effects on, 63–­69; college admissions and, 156–­58, 162; color-­blind racial ideology and, 161, 162, 165–­66; di­ ver­sity and, 158–­60, 162–­64; egal­i­tar­ ianism and, 1–­2, 41, 44–­45, 49, 65; in­terracial marriage and, 63–­67; opposition to, 26, 164–­67; period effects on, 63–­69; policy preferences and, 24,

162–­71, 202–­9; predicted probability of support for, 208; as quota system, 156–­ 60, 162, 164–­66, 171; racial segregation and, 133–­34, 137–­39; reverse discrimination vs., 29, 133–­34, 165; support for, 24, 26, 65, 167–­70, 252, 289, 290; use of term, 143–­44, 145; White millennials and, 65, 134, 137–­39, 143–­44; women and, 164, 290 age effects: abortion rights and, 45; affective­orientation measure and, 55–­58, 56; affirmative action and, 63–­69; on egal­ itarianism, 49, 58–­69, 60, 62, 64; gay and lesbian Americans and, 55; gender equality and, 44–­45; in-­group/out-­group, 55, 146–­47; racial resentment and, 80–­ 83; women’s rights and, 55 age-­period-­cohort intrinsic estimator (APC-­IE). See APC-­IE (age-­period-­cohort intrinsic estimator) Alexander, Michelle, 151, 294n3 American National Election Studies (ANES). See ANES (American National Election Studies) Andolina, Molly W., 34–­35 ANES (American National Election Studies), 11, 48–­49, 55–­58, 61–­69, 78–­86, 89–­95, 96, 265–­68, 293n2, 294n8. See also APC-­IE (age-­period-­cohort intrinsic estimator); feeling thermometer anti-­Black animus: attitude shifts and, xiv–­xv, 1–­2, 47–­48, 87–­95; color-­blind racial attitudes and, 179–­82, 187–­90, 202–­7; evolution of, 30; generational

316 / Index anti-­Black animus (cont.) cohorts and, 146–­47; measurement strat­ egies, 177–­79, 182–­85; racial empathy and, 182–­85, 190–­92, 202–­7; racial re­ sentment and, 185–­86, 192–­93, 202–­7, 233; symbolic racism and, 225; White guilt and, 182–­85, 190–­92, 202–­7. See also racism anti-­racist principles, xvi, 36, 96, 147, 183–­84 APC-­IE (age-­period-­cohort intrinsic estimator): generational difference analysis, 80–­83; methodological strategy and data, 44, 48–­49, 261–­68; supplemental tables, 286–­92 Apfelbaum, Evan P., 95, 97 Asian Americans: color-­blind racial ideology and, 255; disadvantage and, 108–­9, 128–­29, 138, 167; immigration and, 15, 75, 108, 228; “model minority” stereotype, xv, 75, 108, 147; policy preferences and, 16, 24, 172–­73; racial-­group growth, 15–­16 Atwater, Lee, 33–­34 authoritarian personality theory, 25. See also principle-­policy gap baby boomers, 4, 5, 7, 33, 44, 54, 75, 147, 171, 293n1 Benedict, Ruth, 105 biological racism: cohort effects and, 33, 49–­54; generational cohorts and, 27–­31, 33, 49–­54, 73, 105, 264–­65; White millennials and, 13, 54, 95–­97, 105–­8, 140–­ 41. See also anti-­Black animus; racism Black Americans: affective-­orientation mea­ sure and, 54–­57; affirmative action and, 40–­64, 69–­70, 72–­76; criminal justice system and, 34, 39, 134–­35, 148–­56; demographic shifts among, 74–­75; diversity and, 134–­36, 136–­40; egalitarianism and, 40, 41, 47–­48, 51, 52; immigration and, 15–­16; race as special case and, 39–­42; racial stasis and, 31–­37; racism as Black American concern, 108–­9; shift in attitudes toward, 40–­64, 69–­70; stop-­and-­frisk and, 144–­46, 148–­56; White-­Black feeling thermometer, 40, 41. See also anti-­Black animus Black Lives Matter movement, 91, 93–­95, 93, 94, 256, 293n2

blatant racism, 181–­82, 190, 196–­97, 202–­9, 209–­14, 213, 222–­23. See also anti-­Black animus; racism Blinder, Scott B., 35, 36, 96, 147, 219 Bonilla-­Silva, Eduardo, 6, 29, 104, 118, 129, 150, 162, 165 boomers, 4, 5, 7, 33, 44, 54, 75, 147, 171, 293n1 Brooks, Clem, 47 Brown, Michael, 10, 24, 101, 293n2 Bundy, Cliven, 32 Carson, Ben, 247 CCES (Cooperative Congressional Election Study) 2014: CCES sample data (2014), 12, 187, 189, 195, 196, 226; CoBRAS and, 195–­98, 201, 205–­6, 208; cognitive/emotional dimensions and, 216–­19; EXR and, 185–­86, 196; generational status and, 200–­201, 216–­22; old-­fashioned racism and, 201–­7, 217–­ 19; overview of, 195–­201; PCRW and, 182–­85; policy-­preference predictions, 219–­22; racialized policy preferences and, 207–­9; racial resentment and, 198–­ 201; RRS and, 196; second-­order model analysis, 209–­14, 214–­16; validity of, 216–­19 CCES (Cooperative Congressional Election Study) 2016: affirmative action and, 252; conceptualization of racism and, 104; fear and, 236, 254; generational status and, 68, 72, 104, 229, 231; interracial marriage and, 234; overview of, 4, 12, 226, 295n1; partisanship and, 254; po­litical ideology and, 72, 236, 254; presidential campaign (2016) and, 72, 236, 238, 239; racial resentment and, 254; structural inequalities and, 104. See also FIRE (Fear, acknowledgment of Institutional Racism, and Empathy) battery Charlottesville rally, 15, 32, 101–­2, 249, 256–­57, 293n2 Clinton, Hillary, 236–­43, 247–­49 CoBRAS (Color-­Blind Racial Attitudes Scale): blatant racism, 182, 190, 196–­ 97, 202–­9, 213; color-­blind racial ideology and, 179–­82, 187–­90; confirmatory factor analysis, 196–­201, 197; EXR and, 185–­86, 196; fear and, 202–­9; genera-

Index / 317 tional cohorts and, 187–­90; institutional discrimination, 181–­82, 188, 196–­97, 202–­9; interview data, 189; overview of, 180–­82, 187–­90, 295n1; political ideology, 202–­9; predicted subscales, 196–­201, 209–­14; racial privilege, 180–­81, 188, 196–­97, 202–­9; second-­order model and, 209–­14; White millennials and, 187–­90, 191. See also FIRE (Fear, acknowledgment of Institutional Racism, and Empathy) battery; racial-­attitudes survey (2014) Cohen, Cathy J., 151 cohort effects: biological racism and, 33, 49–­54; diversity and, 117–­32; on egal­ itarianism, 39, 59, 60; environmental factors, 31–­37, 77–­78, 115, 117–­32, 146– ­48; gender equality and, 44–­45, 59; pro-­ White affect and, 58; racial resentment and, 95–­97; use of term, 31–­32 cohort replacement: APC-­IE and, 44; biological racism and, 33, 49–­54; and gay and lesbian Americans, 39, 47; liberalization and, 42–­44, 54, 146; paradox of generations and, 146–­47; policy preferences and, 47–­48; racial attitudes and, 36–­37, 47–­48, 95–­98; racial tolerance and, 22–­ 23; and women’s rights, 39; use of term, 9. See also APC-­IE (age-­period-­cohort intrinsic estimator); generational cohorts Color-­Blind Racial Attitudes Scale (CoBRAS). See CoBRAS (Color-­Blind Racial Attitudes Scale) color-­blind racial ideology: affirmative action and, 161, 162, 165–­66; anti-­Black animus and, 179–­82, 187–­90, 202–­7; Asian Americans and, 255; CoBRAS and, 179–­82, 187–­90; discrimination and, 96–­97; evolution of, 28–­29; intergenerational transfer of, 36; Latinx people and, 255–­56; liberalism and, 29–­30; racial resentment and, 14, 88, 95–­98; reverse racism and, 30, 91; RRS and, 91, 93, 95–­97; strategic color-­blindness, 97; two-­ track socialization and, 96; use of term, 14, 29–­30, 180–­81. See also diversity countervailing forces, theory of: bumper stickers and, 247, 248; bumper sticker visualization of, 247, 248; color-­blind racial ideology and, 7; diversity discourse and, 7, 117–­20; paradox of generations

and, 102, 144, 146–­48; racial stasis and, 7, 14, 170–­73, 189–­90, 244–­45, 249–­50 criminal justice system, 34, 39, 134–­35, 148–­56. See also policy preferences Cutler, Stephen, 43, 45 Davis, Darren W., 186, 198–­99 Davis, Kenneth E., 47 Desmond-­Harris, Jenée, 259 DiAngelo, Robin, 115 discrimination. See institutional discrimination diversity: advantage and, 134–­36, 136–­40; affirmative action and, 158–­60, 162–­ 64; cohort effects and, 117–­32; college campuses and, 122–­24; as commodity, 124–­26; countervailing forces and, 7, 117–­20; critical diversity literature, 250–­ 51; diversity dilemma, 158–­60; friend groups and, xiv, 46, 126–­32, 192, 199; majority-­minority demographics and, 15–­16, 36, 133–­34; policy preferences and, 171; racial isolation vs., 126–­32, 157–­58, 217–­19, 251; residential segregation and, 117–­20, 119, 126; White millennials and, 7, 117–­32; White privi­ lege and, 132–­40 Duane, Alwin, 43 egalitarianism: age effects and, 49, 58–­69, 60, 62, 64; attitudes toward Black Americans and, 40, 41, 47–­48, 51, 52; cohort replacement and, 42–­44; gender roles and, 40, 41, 44–­45; liberalization and, 40–­42, 41, 69–­70; measurement strategy, 44, 48–­58, 59–­69; millennial generation and, 4–­6, 14, 41, 102–­3, 132–­34, 250; principle-­policy gap, 16, 25–­27, 35–­37, 58–­69; race as special case, 14, 25–­27, 35–­37, 40, 47–­48, 58–­ 69; racial attitudes and, 39–­42, 41, 47–­48; racial progress and, 1–­4, 35–­ 37, 39–­42, 47–­48; support for, 287; women’s rights and, 41, 44–­45, 49; use of term, 1–­4, 58 Ehrlichman, John, 33 empathy: anti-­Black animus and, 182–­85, 190–­92, 202–­7; cognitive/affective dimensions and, 209–­14, 216–­22, 225–­26; fear-­empathy dimension, 214,

318 / Index empathy (cont.) 217–­19, 221–­22, 226–­28, 253; PCRW scale and, 190–­92, 198, 209; second-­ order model of racial attitudes and, 209– ­14; stop-­and-­frisk policy and, 148–­56. See also FIRE (Fear, acknowledgment of Institutional Racism, and Empathy) battery epistemology of ignorance, 104–­5, 140–­41, 293n2 equal treatment. See egalitarianism EXR (Explicit Racial Resentment) scale, 185–­86, 192–­94, 199–­201, 209–­14. See also racial-­attitude measurement factor analysis, 83–­87, 196–­201, 209–­16, 269–­73 fear: CCES and, 236, 254; CoBRAS and, 202–­ 9; cognitive/affective dimensions and, 209–­14, 216–­22, 222–­23; fear-­empathy dimension, 214, 217–­19, 221–­22, 226– ­28, 253; interracial marriage and, 233–­ 35; old-­fashioned racism and, 202–­7; PCRW and, 182–­85, 190–­92, 197–­99, 202–­9; policy preferences and, 207–­9, 235–­43; political behavior and, 228, 235–­43; presidential election (2016) and, 235–­45; racial-­attitude-­measurement question and, 228; second-­order racial attitudes and, 209–­14, 222–­23. See also FIRE (Fear, acknowledgment of Institutional Racism, and Empathy) battery feeling thermometer, 41, 47, 49, 54–­58, 91, 93. See also affective-­orientation measure Feldman, Stanley, 104, 167 Ferguson uprising, 10, 24, 255, 256, 293n2 FIRE (Fear, acknowledgment of  Institutional Racism, and Empathy) battery: advantage/disadvantage and, 235–­43; CCES sample (2016) data, 226, 295n1; educational attainment and, 230–­32; fear and, 226–­28, 228–­32, 253–­54, 295n6; generational status and, 228–­32; interracial marriage and, 233–­35; as multidimensional measure, 226–­28, 232–­33, 252–­54; overview of, 8–­9, 223, 225–­28, 232–­33, 295n6; partisanship and, 230–­32; political ideology and, 230–­32; predictive power of, 235–­43, 253–­54; presidential election (2016) and, 40, 235–­43, 244; racial-­attitude

distribution and, 228–­32; racial empathy and, 226–­28; racial/ethnic-­group specificity and, 228; racial stasis and, 242–­44; respondent demographics, 275; RRS and, 228, 233–­35, 252–­54; presidential election (2016) and, 226, 235–­43, 244–­45; validity of, 226, 232–­ 43. See also racial-­attitude measurement Firebaugh, Glenn, 47 Fisher v. University of Texas at Austin (2013), 134, 165 Forman, Tyrone A., 35 Frankenberg, Ruth, 28, 95–­96 friend groups, xiv, 46, 126–­32, 192, 199. See also racial isolation; social distance gay and lesbian Americans: affective-­ orientation measure and, 46–­47, 54– 5 ­ 7, 286; feeling thermometer and, 41, 47; generational cohorts and, 39, 47, 58–­63; same-­sex marriage, 4, 46, 293n2; shift in attitudes toward, 12, 31, 39–­42, 44, 45–­47, 48–­58, 59, 250–­53; social-­ force influences on, 31; workplace protection, 58–­63, 288 General Social Survey (GSS). See GSS (General Social Survey) generational cohorts: affective-­orientation measure and, 76–­77, 216–­22, 218, 219, 220; anti-­Black animus and, 146–­47; biological racism and, 27–­31, 33, 49–­54, 73, 105, 264–­65; CoBRAS and, 187–­90; gay and lesbian Americans and, 58–­63; interracial marriage and, 291; latent racism and, 87–­95; Mannheim’s theory of generations, 33; old-­fashioned racism and, 217–­19; paradox of generations and, 146–­47; political ideology and, 200–­201; racialized policy preferences, 219–­23; racial resentment and, 87–­95; racial stasis and, 34–­35; social forces and, 31–­37; use of term, 9, 293n1 Generation X, 34–­35, 293n1 Generation Z, 257–­58 Goldsmith, Pat Rubio, 118 Greenlee, Jill S., 251 GSS (General Social Survey), 11, 48–­54, 63–­ 68, 118, 264–­65, 268, 293n2, 294n1 Hais, Michael D., 5–­6 Hall, Stuart, 6

Index / 319 Hannah-­Jones, Nikole, 165 Hartmann, Douglas, 133 Heppner, Mary J., 183, 184, 196 Heyer, Heather, 32, 101, 249 Hochschild, Jennifer L., 167 Horton, Avialae, 102 Hutchings, Vincent L., 23–­24 immigration: Asian Americans and, 15, 75, 108, 228; Black Americans and, 15–­16; implicit bias and, 254; support for amnesty for undocumented immigrants, 219–­21, 220, 221 institutional discrimination: CoBRAS and, 181–­82, 188, 196–­97, 202–­9; cognitive/affective dimensions of, 209–­14, 222–­23; PCRW and, 196–­97, 205–­8; predicted probability of support for, 207–­9; second-­order model of, 209–­16 interracial marriage: affirmative action and, 63–­67; fear and, 233–­35; FIRE battery and, 233–­35; generational cohorts and, 291; Latinx people and, 255, 296n1; shifting attitudes on, 291; White millennials and, 6, 63–­67, 71 Jackman, Mary R., 140 Jardina, Ashley, 251 Junn, Jane, 75 Katznelson, Ira, 161 Kinder, Donald R., 26, 74, 95, 147, 177–­ 78, 225, 253, 294n2; “Prejudice and Politics,” 225 Kluegel, James R., 162, 163 knapsack of privilege, 132–­40 Latinx people: affirmative action and, 24; color-­blind racial ideology and, 255–­56; friendship groups and, 127; interracial marriage and, 255, 296n1; policy preferences and, 23–­24; presidential elections and, xiv; racial disadvantage and, 109, 117, 118, 144, 145, 147, 151, 154, 155; racial-­group growth, 15–­16, 95, 108, 228; White racial attitudes toward, 118, 122, 147, 172–­73 Lewis, Amanda E., 29, 36 liberalism: abstract liberalism, 29, 82, 164–­ 67, 180; color-­blind racial ideology and, 29–­30, 179–­82; racialized policies and,

47–­48; social-­force influences and, 31; use of term, 40–­42; White millennials and, 76–­78 liberalization: cohort replacement and, 42–­ 44, 54, 146; egalitarianism and, 40–­42, 41, 69–­70; gay and lesbian Americans and, 40, 41, 45–­47; racial-­attitude trends and, 39–­42; racial progress and, 40–­42 Loftus, Jeni, 46 majority-­minority demographics, 15–­16, 36, 133–­34 Mannheim, Karl, 9, 33, 42–­43 marriage: interracial marriage, 6, 63–­67, 71, 233–­35, 255, 291, 296n1; same-­sex marriages, 4, 46, 293n2; social-­force influences on, 31 Martin, Trayvon, 24, 107–­8, 278 Mason, Paul, 258 Mayer, Jeremy D., 34 Mayorga-­Gallo, Sarah, 118, 122, 126, 132 McIntosh, Peggy, 132–­34, 135 methodology: APC-­IE analyses, 261–­68; data sets, 10–­12, 48, 261–­68; factor analysis, 83–­87, 196–­201, 209–­16, 269–­73; feeling thermometer, 40, 49, 54–­55, 91, 93; focus on White Americans, 15–­17; Generation Z, 257–­58; interview schedule/respondents, 10–­12, 13–­14, 102, 275–­83, 294n1, 296n2; methodological concerns, 7–­9; millennials of color, 254–­57; non-­Hispanic White millennial focus, 4–­9, 10–­12, 15–­17, 254–­55. See also ANES (American National Election Studies); GSS (General Social Survey) millennials of color, 5, 254–­57 Mills, Charles, 293n2 moving walkway of racism, 102–­17 Mueller, Jennifer C., 293n2 Murray, Pauli, 1 Neville, Helen A., 180–­81, 195, 196 Nixon, Richard, 33–­34 Nteta, Tatishe M., 251 N-­word, xiv, 22, 106, 293n1 Obama, Barack: color-­blind racial ideology and, 34; millennial voters and, 71; “post-­racial” America and, xiv, 1, 21, 34, 35–­36, 69, 81–­83, 259; predicted probability of support for, 207–­9;

320 / Index Obama, Barack (cont.) presidential campaign of, xiv, 35–­36, 40, 71, 88, 294n3; presidential election in 2016 and, 240–­43; racial animus and, 40, 81–­83; use of N-­word, 21, 22 Obergefell v. Hodges (2015), 46, 293n2 old-­fashioned racism, 87–­95, 202–­7, 217–­ 19. See also anti-­Black animus; biological racism; blatant racism paradox of generations, 102, 144, 146–­48 PCRW (Psychosocial Costs of Racism to Whites): CCES and, 195–­201; con­fir­ matory factor analysis, 196–­201; em­ pathetic reactions and, 190–­92, 197–­99, 202–­9; empathy and, 190–­92, 198, 209; fear and, 182–­85, 190–­92, 197–­ 99, 202–­9; generational cohorts and, 190–­92; institutional discrimination and, 196–­97, 205–­8; old-­fashioned racism and, 202–­7; overview of, 182–­85, 190–­92; racial resentment and, 202–­9; second-­order model of, 209–­14; White guilt and, 191–­92, 197–­99, 202–­9. See also racial-­attitude measurement people of color: majority-­minority demographics of, 15–­16, 36, 133–­34; White millennials’ attitudes toward, 172–­73 Pérez, Efrén O., 254 period effects: abortion rights and, 45; affective-­orientation measure and, 56, 58; on egalitarianism, 60, 62, 66; gay and lesbian Americans and, 47, 55–­56; gender equality and, 44–­45; in-­group/ out-­group, 55, 58; pro-­White affect and, 56, 57, 58; racial resentment and, 80–­83; use of term, 31; on women’s rights, 55 Piston, Spencer, 251 policy preferences: affirmative action and, 24, 143–­44, 156–­71, 202–­9; Asian Americans and, 16, 24, 172–­73; cohort replacement and, 47–­48; fear and, 207–­ 9, 235–­43; generational cohorts and, 219–­23; Obama approval, 202–­9; predicted probability of, 207–­9; principle-­ policy gap, 16, 25–­27, 35–­37, 58–­69; race-­neutral aspects and, 26–­27; racial antipathy and, 26; racial empathy and, 148–­56; racialized policy preferences, 202–­9; social welfare and, 202–­9; stop-­ and-­frisk and, 143–­46, 170–­71

political behavior: fear and, 228, 235–­43; presidential election (2016) and, 40, 226, 235–­43, 244–­45, 247–­49, 259 political ideology: CoBRAS and, 202–­9; cognitive/affective dimensions of, 209–­ 14, 222–­23; confirmatory factor analysis of, 200–­201; generational cohorts and, 200–­201; predicted probability of support for, 207–­9; racial prejudice and, 200–­201; racial resentment as, 87–­95; second-­order model and, 209–­14 “Prejudice and Politics” (Kinder and Sears), 225 presidential election (2016), 40, 226, 235–­ 43, 244–­45, 247–­49, 259 principle-­policy gap, 16, 25–­27, 35–­37, 58–­69 Psychosocial Costs of  Racism to Whites (PCRW) scale. See PCRW (Psychosocial Costs of Racism to Whites) push-­pull dyads: knapsack of privilege, 132–­40; moving walkway of racism, 102–­17; paradox of generations, 102, 144, 146–­48 race, as special case, 39–­42, 69–­70 racial antipathy, 2, 25–­26, 40, 86–­87 racial apathy, 35, 40, 140, 172, 222 racial-­attitude measurement: cognitive/affective dimensions and, 216–­22, 225–­26; color-­blind language and, 178; countervailing forces and, 6–­7, 13–­14; emerging racial grammars and, 13–­15; emotional dimensions of, 214–­16, 222–­23; generational cohort status and, 77; implicit bias and, 254; measurement strategies, 7–­9, 228, 252–­54; multi­dimensional measures, 8–­9; need for new measures, 177–­79, 193–­94; second-­order model of, 209–­14; symbolic politics and, 76–­77; two-­track socialization and, 36, 96, 147, 219; use of fear question, 228. See also CoBRAS (Color-­Blind Racial Attitudes Scale); EXR (Explicit Racial Resentment) scale; FIRE (Fear, acknowledgment of Institutional Racism, and Empathy) battery; PCRW (Psychosocial Costs of Racism to Whites); racial-­attitudes survey (2014); RRS (racial resentment scale) racial attitudes: changing nature of, 11, 27–­ 31, 30; cohort replacement and, 36–­37,

Index / 321 95–­98; demographic shifts and, 172–­ 73; liberalization trends and, 39–­42, 41; principle-­policy gap, 25–­27; racial stasis and, 31–­37, 223; sociopsychological models of prejudice and, 25–­26; symbolic racial attitudes, xv; two-­track socialization and, 36, 96, 147, 219. See also anti-­Black animus; color-­blind racial ideology; racial resentment; racism racial-­attitudes survey (2014): affirmative action and, 208–­9; cognitive dimension of, 214–­16, 222; countervailing forces and, 223; emotional dimension of racial attitudes, 214–­16, 222; empathy and, 198, 223; generational cohorts and, 223; interview schedule/respondents, 10–­12, 13–­14, 102, 275–­83, 294n1, 296n2; old-­fashioned racism and, 217–­19; PCRW and, 196–­98; political conservatism and, 200–­201; racialized policy preferences and, 202–­9; racial privilege and, 223; racial resentment scales and, 198–­200; racial stasis and, 223; respondent demographics, 280–­83; second-­order model of racial attitudes, 209–­14; social-­welfare spending and, 208–­9; White guilt and, 198 racial isolation, 126–­32, 157–­58, 217–­ 19, 251. See also friend groups; social distance racial privilege: CoBRAS and, 180–­81, 188, 196–­97, 202–­9; cognitive/affective dimensions of, 209–­14, 222–­23; knapsack of privilege, 132–­40; predicted probability of support for, 207–­9; White guilt and, 184–­85; White millennials and, 132–­40 racial resentment: affective-­orientation measure and, 91; age effects and, 80–­83; anti-­Black animus and, 185–­86, 192–­ 93, 202–­7, 233; CCES and, 198–­201, 254; cognitive/affective dimensions of, 209–­14, 222–­23; cohort effects and, 80–­83, 81, 82, 95–­97; color-­blind racial ideology and, 14, 88, 95–­98; evolution of, 28–­29, 30; EXR scale, 185–­86, 192–­94; generational status and, 87–­95; as old-­fashioned racism, 87–­95; PCRW and, 202–­9; period effects and, 80–­83; policy preferences and, 25–­26; as political ideology, 87–­95; socialization of

prejudice and, 25–­26; structural racism and, 185–­86; use of term, 35–­36; White millennials and, 76–­87. See also EXR (Explicit Racial Resentment) scale; RRS (racial resentment scale) racial stasis: Black Americans and, 31–­37; countervailing forces and, 7, 14, 170–­ 73, 189–­90, 244–­45, 249–­50; FIRE battery and, 242–­44; generational cohorts and, 34–­35; measurement strategies, 7–­9, 249–­50, 254–­59; use of term, xvi, 12, 31–­37, 250–­51; White millennials and, 6–­8, 12–­15, 31–­37 racism: blatant racism, 181–­82, 190, 196–­ 97, 202–­9, 213; class-­stratification approach to, 115–­17; emerging racial grammars, 13–­15; moving walkway of racism, 102–­17; old-­fashioned racism, 87–­95, 201–­7, 217–­19; reverse racism, 30, 91; structural racism, 7, 10, 27, 28–­29, 48, 155–­56, 172, 183; symbolic racism, 25–­26, 35–­36; systemic racism, 36, 96, 147, 219; use of term, 10, 102–­8, 250; White guilt and, 202–­9; White millennials and, xvi, 14, 35–­36, 102–­17, 155–­56, 167, 172–­73. See also anti-­Black animus; biological racism; color-­blind racial ideology RRS (racial resentment scale): as accurate assessment measure, 13–­14; ANES methodology and data set, 78–­79, 80–­81, 89–­91; Black Lives Matter and, 91; Blacks, stereotypes of, 89–­91; color-­ blind racial ideology and, 91, 93, 95–­97; confirmatory factor analysis, 200–­201; continuous socialization model and, 76; critiques of, 74, 87–­97, 252–­54; cross-­ generational validity of, 74–­75, 77–­78; FIRE battery and, 228, 233–­35, 252–­54; generational cohorts and, 13, 77, 78–­79; measurement invariance, 83–­87, 85; overview of, 72–­76; predicted probability of support for racialized policies, 207–­9; racial frames and, 13–­14; racial survey and, 198–­200; reliability of, 95–­98; second-­order model of, 209–­14; social-­distance questions, 72–­73; symbolic politics theory and, 76–­77, 78. See also racial-­attitude measurement Rutten, Tim, 71 Ryder, Norman B., 9

322 / Index Schnatter, John, 22 Schuman, Howard, xv, 35, 47–­48, 73, 146 Scott, Jacqueline, 43 Sears, David O., 26, 74, 76–­77, 95, 147, 177–­78, 225, 253, 294n2; “Prejudice and Politics,” 225 segregation, 47, 70, 72–­73, 95 Sidanius, Jim, 25, 171 silent generation, 5, 7, 75, 293n1 Smith, Eliot R., 162, 163 Smith, Tom W., 42, 69 Sniderman, Paul M., 26 social distance, 47, 70, 72–­73, 95. See also racial isolation social identity theory, 25, 55 socioeconomic status, 28, 74, 104–­5, 115–­20, 131–­32, 255, 257 Spanierman, Lisa B., 183, 184, 196 Steeh, Charlotte, 47–­48 Sterling, Donald, 22 stop-­and-­frisk, 148–­56, 170–­71 structural racism, 7, 10, 27, 28–­29, 48, 155–­56, 172, 183 symbolic politics, 76–­77 systemic racism, 36, 96, 147, 219 system justification theory, 25. See also principle-­policy gap system of equations, 261–­63, 262 Tatum, Beverly Daniel, 103 Tay (chatbot), 258 Tesler, Michael, 75, 82, 88, 251 Thornton, Arland, 40, 44 Trump, Donald: presidential election (2016) and, 40, 226, 235–­43, 244–­45, 247–­49, 259; racial animus of, 32, 40, 228, 248; Trump administration actions, 32, 40, 228, 247–­49 two-­track socialization, 36, 96, 147, 219 Valentino, Nicholas L., 23–­24, 35 Van den Berghe, Pierre L., 105–­6

PCRW and, 191–­92, 197–­99, 202–­9; policy preferences and, 207–­9; and predicted probability of support for racialized targets, 207–­9; racial disparity and, 183–­85, 191–­92; racial privilege and, 184–­85; second-­order model of racial attitudes and, 209–­14 White millennials: anti-­racist principles and, xvi, 36, 96, 147, 183–­84, 187, 191, 250; class-­stratification approach to racism, 115–­17; demographics of, 4–­9, 36–­37; diversity discourse and, 7, 117–­32; and emerging racial grammars, xvi, 6, 13–­15, 172–­73; interracial marriage and, 6, 63–­67, 71; liberalism and, 76–­78; methodological focus on, 4–­9, 10–­12, 15–­17, 254–­55; new racial logics of, 101–­2; paradox of generations, 144, 146–­47; people of color, attitudes toward, 172–­73; post-­racial narratives, 71–­72; as qualitative data sources, 10–­ 12, 13–­14; racial ideology development, 140–­41; as racially progressive, 7, 11, 31–­37, 48, 58, 65–­70; racial stasis and, 6–­8, 12–­15, 31–­37; two-­track socialization and, 36, 96, 147, 219; understanding of racism and, xvi, 14, 35–­36, 102–­ 17, 155–­56, 167, 172–­73; use of term, 9, 11. See also millennials of color Wilson, David C., 186, 198–­99 Wilson, Thomas C., 34 Winograd, Morley, 5–­6 women’s rights: affective-­orientation mea­ sure and, 54–­57, 285; affirmative action and, 164, 290; age affects and, 56; at­titu­ dinal trends and, 44–­45; cohort effects and, 59; cohort replacement and, 39; egalitarianism and, 41, 44–­45, 49; pe­ riod effects and, 57; shift in attitudes toward, 39; social-­force influences on, 31 workplace protections, 58–­63, 288, 289 Young-­DeMarco, Linda, 40, 44

White fragility, 115, 250 White guilt, 222; anti-­Black animus and, 182–­85, 190–­92, 202–­7; cognitive/ affective dimensions of, 209–­14, 222–­ 23; old-­fashioned racism and, 202–­9;

Zaller, John, 104, 167 zero-­sum game, 25. See also principle-­ policy gap Zogby, John, 71