Sociology data analyze paper " why trump got elected in 2016)

sharon1997

AldrichPrefacechs1-3.pdf

Home >Social Science homework help >Sociology homework help >Sociology data analyze paper " why trump got elected in 2016)

Change and Continuity in the 2016 Elections

John H. Aldrich Duke University

Jamie Carson University of Georgia

Brad T. Gomez Florida State University

David W. Rohde Duke University

To Paul R. Abramson

Sara Miller McCune founded SAGE Publishing in 1965 to support the dissemination of usable knowledge and educate a global community. SAGE publishes more than 1000 journals and over 800 new books each year, spanning a wide range of subject areas. Our growing selection of library products includes archives, data, case studies and video. SAGE remains majority owned by our founder and after her lifetime will become owned by a charitable trust that secures the company’s continued independence.

Change and Continuity in the 2016 Elections

John H. Aldrich Duke University

Jamie Carson University of Georgia

Brad T. Gomez Florida State University

David W. Rohde Duke University

To Paul R. Abramson

All rights reserved. No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher.

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

Names: Aldrich, John H., 1947– author.

Title: Change and continuity in the 2016 elections / John H. Aldrich, Jamie Carson, Brad T. Gomez, David W. Rohde.

Description: Los Angeles : SAGE/CQ PRESS, 2018. | Includes bibliographical references and index.

Identifiers: LCCN 2017042445 | ISBN 9781544320250 (pbk. : alk. paper)

Subjects: LCSH: Presidents—United States—Election—2016. | United States. Congress—Elections, 2016. | Voting—United States. | Elections—United States.

Classification: LCC JK526 2018 .A54 2018 | DDC 324.973/0932—dc23 LC record available at https://lccn.loc.gov/2017042445

This book is printed on acid-free paper.

15 16 17 18 19 10 9 8 7 6 5 4 3 2 1

FOR INFORMATION:

CQ Press

An Imprint of SAGE Publications, Inc.

2455 Teller Road

Thousand Oaks, California 91320

E-mail: [email protected]

SAGE Publications Ltd.

1 Oliver’s Yard

55 City Road

London EC1Y 1SP

United Kingdom

SAGE Publications India Pvt. Ltd.

B 1/I 1 Mohan Cooperative Industrial Area

Mathura Road, New Delhi 110 044

India

SAGE Publications Asia-Pacific Pte. Ltd.

3 Church Street

#10-04 Samsung Hub

Singapore 049483

Acquisitions Editor: Monica Eckman

Editorial Assistant: Zachary Hoskins

Production Editor: Christine Dahlin

Copy Editor: Pam Schroeder

Typesetter: C&M Digitals (P) Ltd.

Proofreader: Alison Syring

Indexer:

Cover Designer: Janet Kiesel

Marketing Manager: Erica DeLuca

CONTENTS

Tables and Figures 000

Preface 000

About the Authors 000

INTRODUCTION 1

Change and Continuity 5

Voters and the Act of Voting 9

Survey Research Sampling 11

Plan of the Book 12

PART I • THE 2016 PRESIDENTIAL ELECTION 17

CHAPTER 1 • The Nomination Struggle 19

Who Ran 20

The Rules of the Nomination System 24

Strategy and the Candidates’ Campaigns in 2016: The Electoral Setting and How the Candidates Won Their Nominations 35

CHAPTER 2 • The General Election Campaign 47

The Strategic Context and Candidates’ Choices 47

Political Context, Overall Strategy, and Opening Moves 49

From the Conventions to the Debates 53

The End Game and the Struggle over Turnout 60

Did the Campaign Make a Difference? 65

CHAPTER 3 • The Election Results 69

The Election Rules 76

The Pattern of Results 80

State-by-State Results 85

Electoral Change in the Postwar South 90

The Electoral Vote Balance 93

PART II • VOTING BEHAVIOR IN THE 2016 PRESIDENTIAL ELECTION 97

CHAPTER 4 • Who Voted? 99

Voter Turnout, 1789–1916 102

Voter Turnout, 1920–2016 105

Voter Turnout among Social Groups 112

Changes in Turnout after 1960 121

Election-Specific Factors 125

Does Low Voter Turnout Matter? 128

CHAPTER 5 • Social Forces and the Vote 135

How Social Groups Voted in 2016 137

How Social Groups Voted during the Postwar Years 146

Why the New Deal Coalition Broke Down 164

CHAPTER 6 • Candidates, Issues, and the Vote 167

Attitudes toward the Candidates 168

Prospective Evaluations 172

Issue Positions and Perceptions 178

Issue Voting Criteria 182

Apparent Issue Voting in 2016 183

Conclusion 192

CHAPTER 7 • Presidential Performance and Candidate Choice 193

What Is Retrospective Voting? 194

Evaluations of Government Performance on Important Problems 197

Economic Evaluations and the Vote for the Incumbent 199

Foreign Policy Evaluations and the Vote for the Incumbent 204

Evaluations of the Incumbent 205

The Impact of Retrospective Evaluations 207

Conclusion 211

CHAPTER 8 • Party Loyalties, Policy Preferences, and the Vote 213

Party Identification as Loyalty: The Original View 214

Party Identification as Retrospective Evaluation: A Second View 215

Party Identification, Polarization, and Identity: A Synthesis? 216

Party Identification in the Electorate 217

Hispanic Partisanship in 2008, 2012, and 2016 223

Party Identification and the Vote 224

Policy Preferences and Performance Evaluations 228

Conclusion 238

PART III • THE 2016 CONGRESSIONAL ELECTIONS 243

CHAPTER 9 • Candidates and Outcomes in 2016 245

Election Outcomes in 2016 246

Candidates’ Resources and Election Outcomes 258

The 2016 Elections: The Impact on Congress 269

The 2018 Elections and Beyond 278

CHAPTER 10 • The Congressional Electorate in 2016 291

Social Forces and the Congressional Vote 291

Issues and the Congressional Vote 294

Party Identification and the Congressional Vote 295

Incumbency and the Congressional Vote 296

The Congressional Vote as Referendum 299

Presidential Coattails and the Congressional Vote 302

Conclusion 304

PART IV • THE 2016 ELECTIONS IN PERSPECTIVE 307

CHAPTER 11 • The 2016 Elections and the Future of American Politics 309

The Great Continuities: The Electoral System and the Party System 309

The Great Change: Depolarization and the Return of Partisan Polarization 312

Change and Continuity in Turnout 315

Continuities in Electoral Partisanship 316

Changes in the Partisan Electorate 317

Change and Continuity in the U.S. Congress 318

Appendix 321

Notes 341

Index 407

TABLES AND FIGURES

Tables

1-1 Candidates for Nomination to the Presidency by the Democratic and Republican Parties, 2016, with Various Aspects Pertinent to Their Candidacy 22

1-2 Current or Most Recent Office Held by Declared Candidates for President: Two Major Parties, 1972–2016 23

1-3 Democratic Nomination Results, 2016: Bound Delegates Won in State Primaries and Caucuses—Clinton and Sanders 40

1-4 Republican Nomination Results, 2016: Bound Delegates Won in State Primaries and Caucuses—Trump, Cruz, Rubio, and Kasich 43

2-1 Change in Turnout in Battleground States and Nationally, 2012–2016 67

3-1 Presidential Election Results by State, 2016 71

3-2 Presidential Election Results, 1832–2016 82

4-1 Voter Turnout in National Elections, 1945–2017 (Percent) 100

4-2 Percentage of Adults Who Voted for Each of the Major-Party Candidates, 1920–2016 106

4-3 Percentage of Electorate Who Reported Voting for President, by Social Group, 2016 115

4-4 Percentage of Whites Who Reported Voting for President, by Strength of Party Identification and Sense of External Political Efficacy, 2016 125

4-5 Percentage of Electorate Who Reported Voting for President, by Party Identification, Issue Preferences, and Retrospective Evaluations, 2016 131

5-1 How Social Groups Voted for President, 2016 (Percent) 138

5-2 Relationship of Social Characteristics to Presidential Voting, 1944–2016 149

6-1 Relative Ranking of Presidential Candidates on the Feeling Thermometer: Response Distribution and Vote Choice, 2016 170

6-2A Distribution of Responses on Presidential Candidate Trait Evaluations, 2016 (Percent) 171

6-2B Major-Party Vote for Clinton and Trump by Presidential Candidate Trait Evaluations, 2016 (Percent) 173

x Change and Continuity in the 2016 eleCtions

6-3 Most Important Problem as Seen by the Electorate, 2012 and 2016, and Reported Vote for Clinton and Trump in 2016 (Percent) 178

6-4 Four Criteria for Issue Voting, 2016, and Comparisons with 1972–2012 Presidential Elections 184

6-5 Major-Party Voters Who Voted for Clinton, by Seven-Point Issue Scales, 2016 (Percent) 186

6-6 Apparent Issue Voting, 2016, and Comparisons with 1972–2012 Presidential Elections (Percent) 187

6-7 Distribution of Electorate on Net Balance-of-Issues Measure and Major-Party Vote, 2016 190

6-8 Percentage of Major-Party Voters Who Voted for Clinton, by Opinion about Abortion and What They Believe Trump’s and Clinton’s Positions Are, 2016 191

7-1 Evaluation of Government Performance and Major-Party Vote, 2012 and 2016 (Percent) 197

7-2 Evaluation of Party Seen as Better on Most Important Political Problem and Major-Party Vote, 2012 and 2016 (Percent) 199

7-3 The Public’s Assessment of Their Personal Financial Situation and Major-Party Vote, 2000–2016 201

7-4 The Public’s View of the State of the National Economy and Major-Party Vote, 2000–2016 203

7-5 Evaluations of the Incumbent’s Handling of the Economy and Major-Party Vote, 2000–2016 204

7-6 President’s Handling of Foreign Relations and Major-Party Vote, 2016 (Percent) 205

7-7 President’s Handling of the Job and Major-Party Vote, 2000–2016 206

7-8 Summary Measure of Retrospective Evaluations of the Obama Administration and Major-Party Vote, 2016 209

7-9 Major-Party Voters Who Voted for Clinton, by Balance-of-Issues and Summary Retrospective Measures, 2016 211

8-1 Party Identification in Presidential Years, Pre-election Surveys, 1980–2016 (Percent) 218

8-2 Party Identification among Whites in Presidential Years, Pre-election Surveys, 1980–2016 (Percent) 220

8-3 Party Identification among Blacks in Presidential Years, Pre-election Surveys, 1980–2016 (Percent) 221

8-4 Party Identification among Latinos in Presidential Years, Pre-election Surveys, 2008–2016 (Percent) 225

8-5 White Major-Party Voters Who Voted Democratic for President, by Party Identification, 1980–2016 (Percent) 227

8-6 Approval of Incumbent’s Handling of the Economy among Partisan Groups, 2016 (Percent) 232

tables and Figures xi

8-7 Balance-of-Issues Positions among Partisan Groups, 2016 (Percent) 234

8-8 Retrospective Evaluations among Partisan Groups, 2016 (Percent) 236

8-9 Percentage of Major-Party Voters Who Voted for Clinton, by Party Identification and Summary Retrospective Evaluations, 2016 237

9-1 House and Senate Incumbents and Election Outcomes, 1954–2016 247

9-2 House and Senate General Election Outcomes, by Party and Incumbency, 2016 (Percent) 251

9-3 Party Shares of Regional Delegations in the House and Senate, 1953, 1981, and 2017 (Percent) 252

9-4 Success in House and Senate Elections, Controlling for Office Background, Party, and Incumbency, 2016 (Percent) 259

9-5 Average Vote Percentages of House Incumbents, 1974–2016 263

9-6 Incumbents’ Share of the Vote in the 2016 House Elections, by Challenger Campaign Spending (Percent) 266

9-7 Percentage of Incumbents by Vote Share, Challenger Experience, and Spending in the 2016 House Elections 268

9-8 House Seat Losses by the President’s Party in Midterm Elections, 1946–2014 279

9-9 Percentage of Vote Received by Winning House Candidates, by Party and Type of Race, 2016 283

10-1 How Social Groups Voted for Congress, 2016 (Percent) 292

10-2 Percentage of White Major-Party Voters Who Voted Democratic for the House, by Party Identification, 1952–2016 297

10-3 Percentage Who Voted Democratic for the House and Senate, by Party Identification and Incumbency, 2016 300

10-4 Percentage of Voters Who Supported Incumbents in House Voting, by Party Identification and Evaluation of Incumbent’s Performance, 2016 301

10-5 Percentage Who Voted Democratic for the House, by Evaluation of Obama’s Job Performance, Party Identification, and Incumbency, 2016 301

10-6 Percentage Who Voted Democratic for the House and Senate, by Party Identification and Presidential Vote, 2016 303

10-7 Percentage Who Voted Democratic for the House, by Presidential Vote, Party Identification, and Incumbency, 2016 304

A7-1 Evaluation of Government Performance on Most Important Problem and Major-Party Vote, 1972–2008 322

A7-2 Evaluation of Party Seen as Better on Most Important Problem and Major-Party Vote, 1972–2000 and 2008 323

xii Change and Continuity in the 2016 eleCtions

A7-3 The Public’s Assessment of Their Personal Financial Situation and Major-Party Vote, 1972–2016 324

A7-4 The Public’s View of the State of the National Economy and Major-Party Vote, 1980–2016 325

A7-5 Evaluations of the Incumbent’s Handling of the Economy and Major-Party Vote, 1984–2016 326

A7-6 President’s Handling of the Job and Major-Party Vote, 1972–2016 327

A8-1 Party Identification in Presidential Years, Pre-election Surveys, 1952–2016 (Percent) 328

A8-2 Party Identification among Whites, 1952–2016 (Percent) 329

A8-3 Party Identification among Blacks, 1952–2016 (Percent) 330

A8-4 White Major-Party Voters Who Voted Democratic for President, by Party Identification, 1952–2016 (Percent) 331

A8-5 Approval of Incumbent’s Handling of the Economy among Partisan Groups, 1984–2016 (Percent) 332

A8-6 Approval of Incumbent’s Handling of the Economy among Partisan Groups, 1984–1996 (Percent) 334

A8-7 Balance-of-Issues Positions among Partisan Groups, 1976–2016 (Percent) 335

A8-8 Approval of Incumbent’s Handling of Job, by Party Identification, 1972–2016 (Percent) 339

Figures

1-1 Length of Multicandidate Campaigns: Two Major Parties, 1976–2016 28

1-2 Front-Loading: Comparing Democratic Party Delegate Totals Weekly, 1976, 2000, and 2016 31

1-3 Difference in Average Thermometer Rating: Selected Candidates, January 1980 and 2016 36

1-4 Warm Feelings from Opposite Partisans: Selected Candidates, January 1980 and 2016 37

2-1 States That Voted Democratic in at Least Four Out of Five Elections, 1996–2012, with Number of Electoral Votes 48

2-2 Location of Presidential Campaign Field Offices, 2012 and 2016 64

3-1 Electoral Votes, by State, 2016 77

3-2 Trump’s Margin of Victory over Clinton, 2016 87

3-3 Results of the 2000, 2004, 2008, 2012, and 2016 Presidential Elections 94

3-4 Competitive Balance in the Electoral College, 1988–2016 96

4-1 Voter Turnout in Presidential Elections, 1828–1916 103

tables and Figures xiii

4-2 Percentage of Voting-Age Population and of the Politically Eligible and the Voting-Eligible Population Voting for President, 1920–2016 109

4-3 Percentage of Voters Who Voted Early, by State, 2016 111

5-1 Major-Party Voters Who Voted Democratic for President, by Race, 1944–2016 (Percent) 148

5-2 White Major-Party Voters Who Voted Democratic for President, by Region, 1952–2016 (Percent) 153

5-3 White Major-Party Voters Who Voted Democratic for President, by Union Membership, 1944–2016 (Percent) 155

5-4 White Major-Party Voters Who Voted Democratic for President, by Social Class, 1944–2008 (Percent) 157

5-5 White Major-Party Voters Who Voted Democratic for President, by Religion, 1944–2016 (Percent) 160

5-6 White Major-Party Voters Who Voted Democratic for President, by Social Class and Religion, 1944–2008 163

6-1 Median Self-Placement of the Electorate and the Electorate’s Placement of Candidates on Issue Scales, 2016 180

8-1A Approval of Democratic Incumbents’ Handling of Job, by Party Identification, 1980, 1996, 2000, 2012, and 2016 229

8-1B Approval of Republican Incumbents’ Handling of Job, by Party Identification, 1976, 1976, 1984, 1988, 1992, 2004, and 2008 230

9-1 Democratic Share of Seats in House and Senate, 1953–2017 (in Percentages) 250

9-2 Competitive House Districts, 2015–2016 255

PREFACE

On November 8, 2016, Republican Donald Trump was elected president of the United States despite losing the popular vote. Trump’s Electoral College victory over the Democratic nominee, former Secretary of State Hillary Clinton, marked the first time since 2000 that the popular vote winner did not also win a majority of votes in the Electoral College. The Democrats gained two Senate seats in 2016, but this was insufficient for them to regain majority control of the upper chamber. The Democrats also gained six seats in the House of Representatives, but the Republicans maintained a sizable majority and thus control of that body.

In the short term, the reemergence of unified partisan control of govern- ment created new opportunities for a Republican majority in adopting key aspects of their political agenda. Yet with recent legislative struggles in Congress and a new president besieged by his inability to deliver on major campaign promises— somewhat unusual circumstances for a first-term president with unified govern- ment—Democrats remain cautiously optimistic about winning back control of government in the future. Democrats have now won the popular vote in six of the last seven presidential elections. The party currently draws support from a coali- tion of the highly educated, women, African Americans, and Latinos. The latter two groups are expected to increase as a share of the U.S. population over the next twenty years, and electoral participation among both groups has increased steadily in recent years. Yet any talk of a long-term electoral advantage for the Democrats presupposes that current voters will maintain their loyalties to the two political par- ties and that group allegiances will be stable over time. The past shows that this can be a tenuous assumption.

Is America in the midst of an electoral transformation? What were the sources of Trump’s victory in 2016, and how do they differ from Republican coalitions of the past? Does his victory signal a long-term positive trajectory for Republicans’ chances in presidential elections? And are the electoral forces at play in presidential elections similar to those that structure congressional elections? These are the sorts of ques- tions that we seek to answer here.

OUR ANALYSIS

In our study of the 2016 elections, we rely on a wide variety of evidence. Because the bulk of our analysis focuses on individuals’ voting decisions, we rely extensively on survey evidence—four surveys in particular. In studying voter turnout, we employ the Current Population Survey (CPS) conducted by the U.S. Census Bureau. The CPS

xvi Change and Continuity in the 2016 eleCtions

provides information on the registration and voting behavior of more than 131,000 individuals from more than 80,000 households. In examining voting patterns, we rely heavily on a survey of more than 24,500 voters interviewed as they exited the voting booths; this survey, conducted by Edison Research for a consortium of news organi- zations, is commonly referred to as the “pool poll.” We employ pool poll data in our analysis of the 2014 congressional midterm elections as well. These data were also collected by Edison Research and reflect a combination of exit and telephone inter- views with more than 19,000 voters. In studying the party loyalties of the American electorate, we also analyze data from the General Social Survey (GSS) conducted by the National Opinion Research Center at the University of Chicago, which measured party identification twenty-seven times from 1972 through 2008, usually relying on about 1,500 respondents.

Our main source of survey data is the 2016 American National Election Studies (ANES) survey based on 1,181 face-to-face and 3,090 web-based interviews con- ducted before the 2016 election and 1,059 face-to-face and 2,590 interviews con- ducted after the election, using the version of the data released for analysis on May 2, 2017. This 2016 ANES is part of an ongoing series funded mainly by the National Science Foundation. These surveys, carried out originally by a team of scholars at the University of Michigan, began with a small study of the 1948 election; the first major study was in 1952. The ANES investigative team has studied every subsequent presi- dential election, as well as all thirteen midterm elections from 1954 to 2002. The 2016 ANES was conducted jointly by Stanford University and the University of Michigan. In the course of our book, we use data from all thirty-one surveys conducted between 1948 and 2016.

The ANES data are available to scholars throughout the world. Although we are not responsible for the data collection, we are responsible for our analy- ses. The scholars and staff at the ANES are responsible for neither our analyses nor our interpretation of these data. Similarly the organizers and researchers of the CPS, GSS, and national pool poll bear no responsibility for our analyses or interpretation.

ACKNOWLEDGMENTS

Many people assisted us with this study. We deeply appreciate the hard work of our research assistants, Ryan Williamson at the University of Georgia, Mark Dudley at Duke University, Jessica Sullivan at the North Carolina School of Science and Math- ematics (serving on a mentorship program at Duke) and now at Duke, and David Macdonald and Joshua Scriven at Florida State University. Ryan assisted with the data analysis for Chapters 2, 6, 7, 8, 9, and 10; Mark provided special assistance in integrating the 2016 ANES with earlier election studies; and David assisted with the data analysis for Chapters 3, 4, and 5.

In our study of turnout, we were greatly assisted by Michael P. McDonald of the University of Florida, who for several years has provided scholars with a valu- able resource on voter turnout. McDonald’s website, the United States Elections

PreFaCe xvii

Project (http://www.electproject.org), presents detailed national- and state-level estimates of voter turnout based on both voting-age and voting-eligible population estimates.

Several years ago Russell J. Dalton of the University of California at Irvine and the late Robert W. Jackman of the University of California at Davis helped us locate information about cross-national estimates of voter turnout. Abraham Diskin of the Hebrew University of Jerusalem helped us locate turnout data for Israel and provided us with updated data for this volume. Corwin D. Smidt at Michigan State University exposed us to several recent studies on religion and politics. And we thank Gary C. Jacobson of the University of California, San Diego for sharing his data on 2016 House and Senate race types.

We are grateful for support from the Department of Political Science at Duke University, the Department of Political Science at the University of Georgia, the Political Institutions and Public Choice Program at Duke University, and the Department of Political Science at Florida State University.

At CQ Press we are grateful to Charisse Kiino and Monica Eckman for encour- agement and Michael Kerns for help in the early editorial stages. We are espe- cially grateful to them for finding reviewers who had assigned our book in the past, thereby allowing us to receive input from instructors and, indirectly, from students. The reviewers were Suzanne Chod (North Central College), Abbie Erler (Kenyon College), Alison Howard (Dominican University of California), Brad Lockerbie (East Carolina University), and Thomas R. Marshall (University of Texas Arlington). We are grateful to Christine Dahlin, our production editor, Zachary Hoskins, and Erica DeLuca for her efforts in marketing our book. Pam Schroeder did an excellent job of copyediting the manuscript.

This book continues a series of books that we began with a study of the 1980 elections. In many places we refer to our earlier books, all of which were published by CQ Press. Some of this material is available online through the CQ Voting and Elections Collection, which can be accessed through many academic and public libraries.

Like our earlier books, this one was a collective enterprise in which we divided the labor. With the volume preceding this one, however, membership in the collective changed. With David Rohde’s retirement from the writing and data analysis, Jamie Carson was invited to join the authorship team. Brad Gomez had primary responsi- bility for the Introduction and Chapters 3, 4, and 5; John Aldrich for Chapters 1, 6, 7, and 8; and Jamie Carson for Chapters 2, 9, and 10. Aldrich, Carson, and Gomez collaborated on Chapter 11.

Finally we dedicate this book to our friend and colleague, Paul R. Abramson of Michigan State University. Paul coauthored seventeen volumes in the Change and Continuity series, beginning with the first volume in 1980. Although he retired from writing on this series, Paul’s influence remains. We are indebted to him for his contri- butions to this project. Paul and his wife, Janet, lost their beloved son, Lee, in 2016 to his long, brave fight with ALS (Lou Gehrig’s Disease). Lee also contributed to these volumes from time to time with assistance in gathering data. Paul, Janet, and their daughter, Heather, remain in our thoughts. Shalom, friends.

xviii Change and Continuity in the 2016 eleCtions

We appreciate feedback from our readers. Please contact us if you disagree with our interpretations, find factual errors, or want further clarification about our meth- ods or our conclusions.

John H. Aldrich Duke University

[email protected]

Jamie L. Carson University of Georgia

[email protected]

Brad T. Gomez Florida State University

[email protected]

David W. Rohde Duke University [email protected]

xix

ABOUT THE AUTHORS

John H. Aldrich is Pfizer-Pratt University Professor of Political Science at Duke University. He is author of Why Parties? A Second Look (2011) and Before the Conven- tion (1980) and coeditor of Positive Changes in Political Science (2007), and he has also published numerous articles, chapters, and edited collections. He is past president of the Southern Political Science Association, the Midwest Political Science Association, and the American Political Science Association.

Jamie L. Carson is professor of political science at the University of Georgia. His research interests include congressional politics and elections, American political development, and separation of powers. He is coauthor of Ambition, Competition, and Electoral Reform (2013) and Electoral Incentives in Congress (2018) and has published articles in the American Political Science Review, American Journal of Political Science, Journal of Politics, and other journals.

Brad T. Gomez is associate professor of political science at Florida State University. His research interests focus on voting behavior and public opinion, with a particular interest in how citizens attribute responsibility for sociopolitical events. His published work has appeared in the American Political Science Review, American Journal of Political Science, Journal of Politics, and other journals and edited volumes.

David W. Rohde is Ernestine Friedl Professor of Political Science and director of the Political Institutions and Public Choice Program at Duke University. He is coeditor of Why Not Parties? (2008) and Home Style and Washington Work (1989), author of Parties and Leaders in the Postreform House (1991), and coauthor of Supreme Court Decision Making (1976).

INTRODUCTION

P residential elections in the United States are partly ritual, a reaffirmation of our democratic values. But they are far more than rituals. The presidency confers a great deal of power, and those powers have expanded during most of the twentieth century and into the twenty-first century. It is precisely because of these immense powers that presidential elections have at times played a major role in determining public policy and in some cases altered the course of American history.

The 1860 election, which brought Abraham Lincoln and the Republicans to power and ousted a divided Democratic Party, focused on whether slavery should be extended to the western territories. After Lincoln’s election eleven southern states attempted to secede from the Union, the Civil War broke out, and, ultimately, the U.S. government abolished slavery completely. Thus an antislavery plurality—Lincoln received only 40 percent of the popular vote—set in motion a chain of events that freed some four million black Americans.

In the 1896 election Republican William McKinley defeated the Democrat and Populist William Jennings Bryan, thereby beating back the challenge of western and agricultural interests to the prevailing financial and industrial power of the East. Although Bryan mounted a strong campaign, winning 47 percent of the popular vote to McKinley’s 51 percent, the election set a clear course for a policy of high tariffs and the continuation of the gold standard for American money.

Lyndon B. Johnson’s 1964 landslide over Republican Barry M. Goldwater pro- vided the clearest set of policy alternatives of any election in the twentieth century.1 Goldwater offered “a choice, not an echo,” advocating far more conservative social and economic policies than Johnson. When Johnson received 61 percent of the popu- lar vote to Goldwater’s 38 percent, he saw his victory as a mandate for his Great Society programs, the most far-reaching social legislation since World War II. The election also seemed to offer a clear choice between escalating American involve- ment in Vietnam and restraint. But America’s involvement in Vietnam expanded after Johnson’s election, leading to growing opposition to Johnson within the Democratic Party, and four years later he did not seek reelection.

Only the future can determine the ultimate importance of the 2016 election. Some scholars argue that American elections have become less important with time, and there is some truth to their arguments.2 Yet elections do offer important choices on public policy, choices that may affect the course of governance—even if only in the short term.

Despite the continued, fifteen-year-long presence of American combat forces in Afghanistan, the 2016 presidential election focused mainly on economic issues. Nearly a decade removed from the Great Recession of 2007–2009, the U.S. economy had experienced a prolonged period of economic growth, but growth had been slow

2 Change and Continuity in the 2016 eleCtions

and uneven. The average annual growth rate in real GDP between 2010 and 2016 was 2.1 percent, well below the average 3.1 percent growth rate experienced during the decade before the economic collapse or the 3.2 percent growth rate experienced in the 1990s.3 Between 2009 and 2016, corporate profits in the United States had grown by nearly 50 percent, and the Dow Jones Industrial Average signaled a bullish stock market, increasing a staggering 12,000 points over the seven-year period.4 But these corporate gains were not always felt in the pocketbooks of average citizens. By elec- tion day 2016, the U.S. unemployment rate had declined to 4.6 percent, well below its 2009 high mark of 10 percent and similar to 2007 levels. But the number of long-term unemployed individuals was higher than pre-recession levels, and so too was labor force participation, meaning that many Americans who could not find jobs simply withdrew from the labor market.5 Real median household income in the United States declined in four of the seven years between 2009 and 2016, and economic inequality in the United States, which has increased markedly since 1980, reached levels not seen since the 1920s.6

These economic issues provided the backdrop for one of the most remarkable electoral events in U.S. history—one not soon to be forgotten. For the first time ever, one of the major American parties nominated a candidate with no prior political or military experience, Donald J. Trump. Trump, the Republican Party nominee, was no stranger to the American public; the billionaire New York real estate developer had been a fixture on Americans’ televisions and tabloid magazines since the early 1980s, and for fourteen seasons, he hosted a reality TV show, The Apprentice, on NBC. And now he was an American Silvio Berlusconi, a billionaire populist, speaking for the “common man” against the “rigged system.”7 Trump argued that America’s economic woes were the product of unfair international trade deals, taxation, and illegal immi- gration. According to Trump, “[America’s] politicians have aggressively pursued a policy of globalization—moving our jobs, our wealth, and our factories to Mexico and overseas.”8 Trump’s anti-free trade message was counter to conventional Republican Party doctrine, and he promised to renegotiate America’s trade deals with China and “tear up” the 1994 North American Free Trade Agreement (NAFTA) with Mexico and Canada.9 Trump’s views on taxation were more in step with his party, favoring significant reductions in corporate tax rates, a lowering of the top individual tax rate from 39.6 percent to 33 percent, and a repeal of the estate tax. Perhaps his most noted (and divisive) policy position was on the issue of immigration. Even before announc- ing his candidacy, Trump had warned conservatives in his party against comprehensive immigration reforms that would create a path to citizenship for illegal immigrants already in the United States: “They’re taking your jobs. You better be careful.”10 Instead Trump promised, “I will build a great, great wall on our southern border. And I will have Mexico pay for that wall.”11

For the most part Democratic Party nominee Hillary Clinton’s economic posi- tions were more in line with her party’s typical positions. Clinton called for increased government spending for job training programs and community college education; the former First Lady, senator, and secretary of state argued that spending in these areas could help retrain workers who had lost jobs in manufacturing industries. Clinton’s position on taxation was antithetical to Trump’s, arguing that “we need to

introduCtion 3

get the wealthy and the corporations to pay more for their fair share.”12 Clinton also argued that Trump’s trade policy would start a trade war with China. But Clinton was forced to take a more nuanced approach to trade policy than perhaps she would have liked. For instance, as secretary of state, Clinton had originally supported America’s participation in the Trans-Pacific Partnership (TPP), calling it the “gold standard” of trade agreements. But after fighting back a primary challenge from Vermont sena- tor Bernie Sanders, a socialist and only recent affiliate of the Democratic Party, she withdrew her support from the TPP rather than risk alienating Sanders’s supporters: “I oppose it now. I’ll oppose it after the election, and I’ll oppose it as president.”13 Clinton’s position on immigration also contrasted sharply with Trump’s. Clinton claimed that the U.S. border with Mexico was “the most secure border we have ever had” and that Trump’s plan for a wall across the entire border was pure “fantasy.”14 And Clinton supported immigration reform efforts to provide lawful status to the children of illegal immigrants and a path to citizenship for undocumented immigrants who did not provide a security threat.

Another issue on which the candidates differed was on the future of the cen- tral legislative achievement of the Obama administration, the Affordable Care Act of 2010, commonly known as “Obamacare.” The legislation marked the most significant change to the nation’s health care system since the creation of Medicare and Medicaid during the 1960s. It mandated that all individuals who are not currently insured or already covered under government insurance programs buy a private health insurance policy or pay a penalty (a “tax” according to the United States Supreme Court). And to lower the costs of policies, Obamacare promoted the creation of state-level health insurance exchanges to foster competition among insurance providers and grant sub- sidies to low-income individuals and families to offset costs. Clinton hoped to amend the program, among other things, expanding the tax credits individuals could receive to offset out-of-pocket health expenses, increasing funding for community health cen- ters, and increasing government control over drug price increases.15 Trump offered few details about how he would reform the health care system other than his oft-stated pledge to “immediately repeal and replace Obamacare.”

The contrast between the two candidates who squared off in the 2016 presidential election could not have been sharper, both in style and on substance. On the Republican side was a highly confrontational political novice who often engaged on the political trail in both bombast and profanity, pledging to “make America great again” and claiming during his nomination acceptance speech that “I alone can fix it.” On the Democratic side was the ultimate political insider, a policy wonk who had not only served a president but also had been married to one. However, for all of their differences, Trump and Clinton had one thing in common—Americans didn’t seem to like either one. On the eve of the election, the Gallup Poll reported that the candidates suffered from the lowest favorabil- ity ratings recorded by the company since it starting asking the question back in 1956. According to Gallup, 52 percent of Americans viewed Clinton unfavorably, whereas a record 61 percent found Trump unfavorable.16 Yet, despite the public’s consternation, the election went on . . . just as the Constitution mandates. On November 8, 2016, Donald J. Trump was elected president of the United States, winning a comfortable Electoral College victory despite losing the popular vote to Clinton, 48.2 to 46.1 percent.

4 Change and Continuity in the 2016 eleCtions

To be sure, the 2016 presidential campaign was unusual; from the rise of an unorthodox candidate to the presence of a woman at the head of a party’s ticket (a historic first), and there were missing emails, stolen emails, FBI investigations (of both candidates we would learn later!), charges of Russian interference, and abhorrent conversations recorded on buses. Yes, 2016 was unusual (or, at least, one might hope)!

But none of this means that the election itself was unusual. Indeed, as we hope to show, the 2016 election is in many ways the product of electoral continuity. Although American parties have become more ideologically disparate over the past few decades, no party holds a clear advantage. The 2016 election marked only the second time since Franklin D. Roosevelt that one of the parties, the Democrats, won the popular vote in three straight elections—the Republicans did so in 1980, 1984, and 1988. But the popular vote balance was so close that it became feasible for the second-place finisher in the popular vote, Trump, to capture the Electoral College and the White House. Once considered a historical aberration, this divergence between the popular vote and electoral vote has now happened in two of the last five U.S. elections.

With the election of Republican majorities in both the House and Senate, along with Trump’s victory, 2016 brought unified partisan control of government. Whereas unified government has not been the norm in modern American politics, it is impor- tant to note that the Republican majorities in both the House and Senate are relatively small. The Republicans lost six House seats in 2016, bringing their majority down to 23 seats (roughly 5 percent of the chamber). Senate Republicans also lost seats (two) in 2016 but maintain a slim, two-seat majority. Although the election outcome did advantage the Republican Party, it would be difficult to claim that the GOP now holds an electoral advantage. Indeed, with only small majorities in both houses of Congress, there is no guarantee that President Trump will be able to push through his legisla- tive agenda with ease. The first six months of Trump’s administration have borne this out. Consider Trump’s campaign promise to quickly repeal and replace Obamacare. During the Obama presidency, and knowing full well that the president would veto their efforts if necessary, Republicans voted on 54 separate occasions to repeal all or part of the Affordable Care Act.17 Yet with unified Republican control, the Congress has failed to repeal the law. In March 2017, during what should have been the new president’s “honeymoon period,” House Speaker Paul Ryan (R–WI) cancelled a vote to replace Obamacare (a full repeal was not in the offing to the consternation of some conservatives) because he did not have the votes.18 Although the House would even- tually pass a replacement bill in May, the Senate failed to pass a similar bill in July by one vote.19 An apparently final attempt for the year in the Senate was pulled from consideration in September after the Republicans failed to receive commitments for at least a bare fifty-vote majority. There are prospects for major policy changes during the remainder of Trump’s term, of course, particularly on issues such as tax reform, but majority status alone is not sufficient for legislative success.20 If the Democrats are able to gain control of the House and/or Senate during the 2018 midterm election, the Trump agenda clearly would be in peril.

Is America in the midst of an electoral transformation? What were the sources of Trump’s victory in 2016, and how do they differ from electoral coalitions of the past? Does his victory signal a long-term negative trajectory for Democrats’ chances

introduCtion 5

in presidential elections? And are these electoral forces similar to those that structure congressional elections? These are the sorts of questions that we seek to answer here.

This book continues a series of seventeen books that we began with a study of the 1980 elections. Our focus has always been both contemporary and historical. Thus we offer an extensive examination of the 2016 presidential and congressional campaigns and present a detailed analysis of individual-level voting behavior, examining those factors that lead citizens to vote as well as those that affect how they vote. We also aim to place the 2016 elections in proper historical and analytical contexts.

CHANGE AND CONTINUITY

Elections are at once both judgments on the issues of the day and the product of long-term changes in the relationship between the political parties and voters. For example, Democrats’ aspirations for an emerging electoral majority following their 2012 presidential victory were not unfounded. If one is to believe the projections of the U.S. Census Bureau, many of the social groups that have supported Democrats in recent elections, particularly Latinos, are growing as a percentage of the overall U.S. population. And turnout among these groups has increased in recent decades. So, for some Democrats, their party’s future success in presidential elections over the next few decades seemed all but assured. Then Donald Trump won, and the questions became: Was 2016 the dawn of an emergent Republican majority? Had Trump realigned the American party system?21

It is not uncommon for winning parties to make hyperbolic claims about the “historic” nature of their victories or to assert that their win was a sign of impending electoral dominance. Indeed, in 2008, Democrats were exuberant over Obama’s siz- able victory over John McCain and were even more pronounced in their claims of a bright Democratic future. Some observers saw the election as restoring Democrats to their status as the majority party, which they had enjoyed between 1932 and 1968. Lanny J. Davis, a former special counsel to President Clinton, wrote follow- ing the 2008 election: “Tuesday’s substantial victory by Barack Obama, together with Democratic gains in the Senate and House, appear to have accomplished a funda- mental political realignment. The election is likely to create a new governing major- ity coalition that could dominate American politics for a generation or more.”22 Two years later the Democrats lost sixty-three seats and their majority status in the House of Representatives—the largest seat change since 1946—and six seats in the Senate, where they maintained a slim majority.23

In 2004, following incumbent President George W. Bush’s victory over Democrat nominee, John Kerry, scholars speculated about a pro-Republican realignment. Indeed speculation about Republican dominance can be traced back to the late 1960s, when Kevin P. Phillips, in his widely read book, The Emerging Republican Majority, argued that the Republicans could become the majority party, mainly by winning support in the South.24 Between 1969, when his book was published, and 1984, the Republicans won three of the four presidential elections, winning by massive landslides in 1972, when Richard M. Nixon triumphed over George S. McGovern, and in 1984, when

6 Change and Continuity in the 2016 eleCtions

Ronald Reagan defeated Walter F. Mondale. In 1985 Reagan himself proclaimed that a Republican realignment was at hand. “The other side would like to believe that our victory last November was due to something other than our philosophy,” he asserted. “I just hope that they keep believing that. Realignment is real.”25 Democratic victories in the 1992 and 1996 presidential elections called into question the claims of a pro- Republican realignment.

Obviously not all elections are transformative. So how is electoral change—not simply the ebbs and flows from election to election but changes in the fundamental factors that link parties and voters—to be understood?

For generations of political scientists, theories of electoral change have centered on the concept of political realignment.26 Political scientists define realignment in dif- ferent ways, but they are all influenced by V. O. Key Jr., who developed a theory of “critical elections” in which “new and durable electoral groupings are formed.”27 Elections like that in 1860 in which Lincoln’s victory brought the Republicans to power, in 1896 in which McKinley’s victory solidified Republican dominance, and in 1932 in which the Democrats came to power under FDR are obvious candidates for such a label.

But later Key argued that partisan shifts could also take place over a series of elections—a pattern he called “secular realignment.” During these periods, “shifts in the partisan balance of power” occur.28 In this view the realignment that first brought the Republicans to power might have begun in 1856, when the Republicans displaced the Whigs as the major competitor to the Democrats and might have been consolidated by Lincoln’s reelection in 1864 and Ulysses S. Grant’s election in 1868. The realignment that consolidated Republican dominance in the late nineteenth century may well have begun in 1892, when Democrat Grover Cleveland won the election, but the Populist Party, headed by James D. Weaver, attracted 8.5 percent of the popular vote, winning four states and electoral votes in two others. In 1896 the Populists supported William Jennings Bryan and were co-opted by the Democrats, but the electorate shifted to the Republican Party. The pro-Republican realign- ment might have been consolidated by McKinley’s win over Bryan in 1900 and by Theodore Roosevelt’s victory in 1904.

Though the term New Deal was not coined until Franklin Roosevelt’s campaign of 1932, the New Deal realignment may have begun with Herbert C. Hoover’s tri- umph over Democrat Al Smith, the first Roman Catholic to be nominated by a major political party. Although badly defeated, Smith carried two New England states, Massachusetts and Rhode Island, which later became the most Democratic states in the nation.29 As Key points out the beginnings of a shift toward the Democrats was detectable in Smith’s defeat.30 However, the “New Deal coalition” was not created by the 1932 election but after it, and it was consolidated by Roosevelt’s 1936 land- slide over Alfred M. Landon and his 1940 defeat of Wendell Willkie. The New Deal coalition structured the distribution of party support within the electorate during the earliest decades of the post-World War II period, and its decline and eventual replacement are important to understanding the changes and continuities of modern electoral politics.

Past partisan realignments in the United States have had five basic character- istics. First, realignments have traditionally involved changes in the regional bases

introduCtion 7

of party support. Consider, for instance, the decline of the Whig Party and rise of the Republicans. Between 1836 and 1852, the Whigs drew at least some of their electoral support from the South.31 The last Whig candidate to be elected, Zachary Taylor in 1848, won sixty-six of his electoral votes from the fifteen slave states. In his 1860 victory Lincoln did not win a single electoral vote from the fifteen slave states. Regionalism may be less important to future electoral changes, however. Today televi- sion and other media have weakened regionalism in the United States, and politics is much more nationalized. Two-party competition has diffused throughout the country, and the issues on which the parties compete tend to be more national in scope.32

Second, past party realignments have involved changes in the social bases of party support. Even during a period when one party is becoming dominant, some social groups may be moving to the losing party. During the 1930s, for example, Roosevelt gained the support of industrial workers, but at the same time, he lost support among business owners and professionals.

Third, past realignments have been characterized by the mobilization of new groups into the electorate. Indeed the mobilization of new voters into the electorate can result in significant electoral volatility.33 Between Calvin Coolidge’s Republican landslide in 1924 and Roosevelt’s third-term victory in 1940, turnout among the voting-age population rose from 44 percent to 59 percent. Although some long-term forces were pushing turnout upward, the sharp increase between 1924 and 1928 and again between 1932 and 1936 resulted at least in part from the mobilization of new social groups into the electorate. Ethnic groups that were predominantly Catholic were mobilized to support Al Smith in 1928, and industrial workers were mobilized to support Franklin Roosevelt in 1936.

Fourth, past realignments have occurred when new issues have divided the elec- torate. In the 1850s the Republican Party reformulated the controversy over slavery to form a winning coalition. By opposing the expansion of slavery into the territories, the Republicans contributed to divisions within the Democratic Party. Of course no issue since slavery has divided America as deeply, and subsequent realignments have never brought a new political party to power. But those realignments have always been based on the division of the electorate over new issues.

Last, most political scientists argue that partisan realignments occur when voters change not just their voting patterns but also the way they think about the political parties, thus creating an erosion of partisan loyalties. During the Great Depression in 1932, for example, many voters who thought of themselves as Republicans voted against Hoover. Later many of these voters returned to the Republican side, but oth- ers began to think of themselves as Democrats. Likewise, in 1936 some voters who thought of themselves as Democrats disliked FDR’s policies and voted against him. Some of these defectors may have returned to the Democratic fold in subsequent elec- tions, but others began to think of themselves as Republicans.

Not all scholars believe that the concept of realignment is useful. In 1991 Byron E. Shafer edited a volume in which several chapters questioned its utility.34 More recently David R. Mayhew published a monograph critiquing scholarship on realignment, and his book received widespread critical acclaim.35 Mayhew cites fifteen claims made by scholars of realignment and then tests these claims. He argues that many of these claims

8 Change and Continuity in the 2016 eleCtions

do not stand up to empirical scrutiny, questions the classification of several elections as “realigning,” and suggests that the concept of realignment should be abandoned.

Although we agree with some of the claims made by Mayhew, we see no reason to abandon the concept completely. Some electoral changes may correspond to the critical election-realignment dynamic—a long period of stability in the party system is altered by a rapid and dramatic change, which leads to a new, long-term partisan equi- librium. Using biological evolution as a theoretical analogue, Edward G. Carmines and James A. Stimson argue that partisan realignments of this type are similar in form to the evolutionary dynamic known as cataclysmic adaptation.36 But the authors note that biological examples of the cataclysmic adaptation dynamic are extraordinarily rare and suggest that critical election realignments are likely to be rare also.

Carmines and Stimson articulate two additional evolutionary models of partisan change. The authors argue that Key’s secular realignment dynamic is consistent with the model of Darwinian gradualism. In this view electoral change does not result from a critical moment but instead is “slow, gradual, [and] incremental.”37 As noted in Key’s original work, the secular realignment dynamic “operate[s] inexorably, and almost imperceptibly, election after election, to form new party alignments and to build new party groups.”38

The third model of partisan change espoused by Carmines and Stimson is consis- tent with the “punctuated equilibrium” model of evolution.39 In this dynamic process,

the system moves from a fairly stationary steady state to a fairly dramatic rapid change; the change is manifested by a ‘critical moment’ in the time series—a point where change is large enough to be visible and, perhaps the origin of a dynamic process. Significantly, however, the change—the dynamic growth—does not end with the critical moment; instead it continues over an extended period, albeit at [a] much slower pace.40

In our view the punctuated equilibrium model best captures the dynamic nature of electoral change in the United States since the 1960s.

The 1960s were a critical moment in American politics. The events of the decade were the catalysts for fundamental changes in the rules that govern politi- cal parties and the partisan sentiments that would govern voters for years to come.41 Of particular interest is the transformative power of the issue of race. By 1960 the national Democratic Party’s sponsorship of civil rights for African Americans had cre- ated a schism between the more-liberal elements of the party and white southern Democrats. But it had also allowed the party to chip away at black voters’ allegiance to the Republican Party, “the party of Lincoln.” The partisan loyalties of African Americans had been shaped by the Civil War, and black loyalties to the Republican Party—where and when allowed to vote—lasted through the 1932 election. By 1960 a majority of African Americans identified with the Democratic Party, but there was still a substantial minority of Republican identifiers. Between 1960 and 1964, how- ever, African American loyalties moved sharply toward the Democrats. The civil rights demonstrations of the early 1960s and the eventual passage of the 1964 Civil Rights Act solidified the position of the Democratic Party as the party of civil rights. By late

introduCtion 9

1964 more than 70 percent of African Americans identified as Democrats, a level of loyalty that persists today. The change in partisanship among blacks and the subse- quent mobilization of black voters following the passage of the 1965 Voting Rights Act provided the rapid, critical moment that disrupted the stable equilibrium created by the New Deal Coalition. And, as the punctuated equilibrium dynamic suggests, the electorate continued to change in a direction set forth by the critical era of the 1960s, but it did so at slower rate, and it continues to have ramifications for politics today.

The political events of the 1960s also had an effect on white partisanship, but the change was neither immediate nor decisive. From the mid-1960s to the mid-1970s, there was a substantial erosion in party loyalties among whites. The proportion of the white electorate who considered themselves “independent” increased noticeably. By 1978 nearly 40 percent of whites said they were either pure independents or inde- pendents who “leaned” toward one of the two parties, nearly double that found in the late 1950s and early 1960s.42 These changes led some scholars to use the term dealign- ment to characterize American politics during the period.43 The term was first used by Ronald Inglehart and Avram Hochstein in 1972.44 A dealignment is a condition in which old voting patterns break down without being replaced by newer ones. Yet, beginning in the 1980s, the proportion of whites claiming to be pure independents declined as whites nationally began to lean toward the Republican Party. In the once “solid Democratic South,” whites have become decidedly Republican. Voters appear to be aligned.

Despite these changes the Republicans have never emerged as the majority party among the electorate. Democrats, however, saw a growth in political loyalties between 2004 and 2012, and in 2016, the party once again emerged as the majority party among two party identifiers, albeit a small majority.45 This is not to say that Republicans cannot win, of course; it simply means that the GOP has entered recent elections at a numeri- cal disadvantage. Democrats’ electoral gains have largely been the product of the criti- cal events of the 1960s, which established them as the party of civil rights. As America’s non-white population has increased—more than half of the growth in the U.S. popula- tion between 2000 and 2010 was due to an increase in the nonwhite population—and Democrats have been the beneficiaries. For instance roughly two out of every three Latino voters in the United States identify with Democratic Party, and 54 percent of Latino voters say that Democrats have more concern for them, compared to only 11 percent who say that Republicans do.46 America’s racial and ethnic minorities continue to view the Democrats’ adherence to the civil rights agenda of the 1960s as providing them with a natural political home, and America’s whites are increasingly more likely to side with the Republicans. In our view the 2016 elections do not represent a fundamen- tal change in America’s electoral politics. Instead the 2016 elections continue to reflect electoral alignments set in motion by a critical era that occurred nearly a half century ago.

VOTERS AND THE ACT OF VOTING

Voting is an individual act. Indeed the national decision made on (or before) November 8, 2016, was the product of more than 230 million individual decisions.47

10 Change and Continuity in the 2016 eleCtions

Two questions faced Americans eighteen years and older: whether to vote and, if they did, how to cast their ballots. These decisions, of course, are not made in iso- lation. Voters’ decisions are influenced by the social, economic, and information contexts in which they live; they are influenced by the political attitudes that they have acquired throughout their lifetime; and they are influenced by the voting decisions they have made in the past.48 Voters’ decisions are also constrained by America’s electoral rules and two-party system—these are the primary sources of continuity in our political system.

How voters make up their minds is one of the most thoroughly studied subjects in political science—and one of the most controversial.49 Voting decisions can be stud- ied from at least three theoretical perspectives.50 The first approach is sociological in character and views voters primarily as members of social groups. Voters belong to primary groups of family members and peers, secondary groups such as private clubs, trade unions, and voluntary associations, and broader reference groups such as social classes and religious and ethnic groups. Understanding the political behavior of these groups is central to understanding voters, according to Paul F. Lazarsfeld, Bernard R. Berelson, and their colleagues. Social characteristics determine political preferences.51 This perspective is still popular, although more so among sociologists than political scientists.52

A second approach places greater emphasis on the psychological (or, more aptly, attitudinal) variables that affect voting. The “socio-psychological model” of voting behavior was developed by Angus Campbell, Philip E. Converse, Warren E. Miller, and Donald E. Stokes, scholars at the University of Michigan Survey Research Center, in their classic book The American Voter.53 The Michigan scholars focused on attitudes most likely to have the greatest effect on the vote just before the moment of decision, particularly attitudes toward the candidates, the parties, and the issues. An individu- al’s party identification emerged as the most important social-psychological variable that influences voting behavior. The Michigan approach is the most prevalent among political scientists, and party identification continues to be emphasized as one of the most influential factors affecting individual vote choice, although many deemphasize its psychological underpinnings.54

A third approach draws heavily from the work of economists. According to this perspective, citizens weigh the costs of voting against the expected benefits when deciding whether to vote. And when deciding for whom to vote, they calculate which candidate favors policies closest to their own policy preferences. Citizens are thus viewed as rational actors who attempt to maximize their expected utility. Anthony Downs and William H. Riker helped to found this rational choice approach.55 The writ- ings of Riker, Peter C. Ordeshook, John A. Ferejohn, and Morris P. Fiorina are excel- lent examples of this point of view.56

Taken separately none of these approaches adequately explains voting behavior; taken together the approaches are largely complementary.57 Therefore, we have cho- sen an eclectic approach that draws on insights from each viewpoint. Where appro- priate we employ sociological variables, but we also employ social-psychological variables such as party identification and feelings of political efficacy. The rational choice approach guides our study of the way issues influence voting behavior.

introduCtion 11

SURVEY RESEARCH SAMPLING

Because of our interest in individual-level voting behavior, our book relies heavily on surveys of the American electorate. It draws on a massive exit poll conducted by Edison Research for the National Election Pool, a consortium of six news organizations, as well as surveys conducted in people’s homes by the U.S. Census Bureau, and telephone polls conducted by the Pew Research Center. But our main data source for 2016 is the 1,181 face-to-face and 3,090 web-based interviews conducted before the election and 1,059 face-to-face and 2,590 interviews conducted after the election as part of the American National Election Studies (ANES) Time Series Survey.58 Originally conducted by The Survey Research Center (SRC) and Center for Political Studies (CPS) at the Univer- sity of Michigan, the ANES surveys have been conducted using national samples in every presidential election since 1948 and in every midterm election between 1954 and 2002.59 The 2016 ANES was conducted jointly by Stanford University and the University of Michigan, with funding by the National Science Foundation. Since 1952 the ANES surveys have measured party identification and feelings of political effective- ness. The CPS, founded in 1970, has developed valuable questions for measuring issue preferences. The ANES surveys are the best and most comprehensive for studying the issue preferences and party loyalties of the American electorate.60

Readers may question our reliance on the ANES surveys of just over 4,200 people when some 230 million Americans are eligible to vote. Would we have similar results if all adults eligible to vote had been surveyed?61 The ANES uses a procedure called multistage probability sampling to select the particular individuals to be interviewed. This procedure ensures that the final sample is likely to represent the entire popula- tion of U.S. citizens of voting age, except for Americans living on military bases, in institutions, or abroad.62

Because of the probability procedures used to conduct the ANES surveys, we are able to estimate the likelihood that the results represent the entire population of noninstitutionalized citizens living in the United States. Although the 2016 ANES survey sampled only about one in every 55,000 voting-eligible Americans, the repre- sentativeness of a sample depends far more on the size of the sample than the size of the population being studied, provided the sample is drawn properly. With samples of this size, we can be fairly confident (to a level of .95) that the results we get will fall within three percentage points of that obtained if the entire population had been sur- veyed. For example, when we find that 52 percent of respondents approved of the job Barack Obama was doing as president, we can be reasonably confident that between 49.7 percent (52 – 2.3) and 54.3 percent (52 + 2.3) approved of his performance. The actual results could be less than 49.7 percent or more than 54.3 percent, but a confi- dence level of .95 means that the odds are nineteen to one that the entire electorate falls within this range. The range of confidence becomes somewhat larger when we look at subgroups of the electorate. For example, with subsets of about five hundred (and the results in the 50 percent range) the confidence error rises to plus or minus six percentage points. Because the likelihood of sampling error grows as our subsamples become smaller, we sometimes supplement our analysis with reports of other surveys.

12 Change and Continuity in the 2016 eleCtions

Somewhat more complicated procedures are needed to determine whether the difference between two groups is likely to reflect the relationship found if the entire population were surveyed. The probability that such differences reflect real differ- ences in the population is largely a function of the size of the groups being com- pared.63 Generally speaking, when we compare the results of the 2016 sample with an earlier ANES survey, a difference of three percentage points is sufficient to be reason- ably confident that the difference is real. For example, in 2008 during the final year of the George W. Bush presidency and during the onset of the “great recession,” only 2 percent of respondents said that the economy had improved in the last year; in 2016, 28 percent did. Because this difference is greater than three percentage points, we can be reasonably confident that the electorate was more likely to think the national econ- omy was improving in 2016 than they were to think it was improving back in 2008.

When we compare subgroups of the electorate sampled in 2016 (or compare those subgroups with subgroups sampled in earlier years), a larger percentage is usually nec- essary to conclude that differences are meaningful. For example, 35 percent of whites who did not complete high school favored Hillary Clinton; among those who graduated high school but did not continue their education, 33 percent favored Clinton. We can- not be confident this is real, however, because the subsample sizes are quite small—only seventy-seven people are in the first category, whereas there are 463 people in the latter. With subsample sizes of this size, we would need to see a difference of thirteen points to be confident in the results. Thankfully, the relatively large sample size that we obtained in 2016 by using the full (face-to-face and Internet) sample means that statistical confi- dence in our subgroup comparisons is much easier to achieve. For instance, among vot- ers, we have 2,131 men and 2,392 women.64 With subsamples of this size, a three-point difference is large enough to conclude that the gender difference was real. However, it is important to recognize that in previous years, our sample sizes (face-to-face only) were much smaller. Generally speaking, comparisons of men and women using data from previous ANES studies require a difference of five percentage points. Similarly it is important to be mindful of racial differences in the sample. In 2016 our full sample contains 4,563 whites (71.6%) and 598 blacks (9.4%). Thus, to be confident in racial difference, we require a spread of more than four percentage points. When using data from previous years with smaller sample sizes, a difference of at least eight percentage points is needed to conclude that differences between whites and blacks are meaningful.

This discussion represents only a ballpark guide to judging whether reported results are likely to represent the total population. Better estimates can be obtained using the formulas presented in many statistics textbooks. To make such calculations or even a rough estimate of the chances of error, the reader must know the size of the groups being compared. For that reason, we always report in our tables and figures either the number of cases on which our percentages are based or the information needed to approximate the number of cases.

PLAN OF THE BOOK

We begin by following the chronology of the campaign itself. Chapter 1 examines the battle for the Democratic and Republican Party presidential nominations. Three

introduCtion 13

major Democratic candidates and twelve major Republican candidates campaigned for the chance to square off in the general election. As is typical when no incumbent president stands for reelection, both parties featured heated contests for the nomina- tion. In Chapter 1 we discuss the regularities in the nomination process that explain why some candidates run and others do not. We then examine the rules governing the nomination contests, and we also assess the importance of campaign finance. The dynamics of multicandidate contests and the concept of momentum to discuss nomi- nation contests in the 1970s are covered in Chapter 1 as well.

Chapter 2 moves to the general election campaign. Because of the rules set forth by the U.S. Constitution for winning presidential elections, candidates must think about how to win enough states to gain a majority (270) of the electoral vote (538 since 1964). We examine the Electoral College strategies adopted by the campaigns. There were three presidential debates and one vice presidential debate, and we discuss their impact. Last, we turn to the end game of the campaign, the battle over turnout. We examine the “ground game” undertaken by each campaign in an effort to get out the vote, and we will examine how these strategies differ from previous presidential campaigns.

Chapter 3 turns to the actual election results, relying largely on the official elec- tion statistics. Our look at the electoral vote is followed by a discussion of the elec- tion rules, noting that the U.S. plurality vote system supports “Duverger’s law.” We examine the pattern of results during the eighteen postwar elections as well as those in all forty-seven elections between 1832 and 2016. We then analyze the state-by-state results, paying particular attention to regional shifts in the elections between 1980 and 2016. We focus special attention on electoral change in the postwar South because this region has been the scene of the most dramatic changes in postwar U.S. politics. Finally we study the results of the last five presidential elections to assess the electoral vote balance.

Chapter 4 analyzes what is perhaps the most important decision of all: whether to vote. We examine the dynamics of electoral participation in U.S. politics, particu- larly changes in turnout during the postwar period. Although turnout grew fairly con- sistently between 1920 (the year women were enfranchised throughout the United States) and 1960, it fell in 1964 and in each of the next four elections. We show that the decline in turnout during this period coincides with steep declines in partisan attach- ment and political efficacy in the electorate. As partisan attachments have increased in recent decades, turnout has risen, but it remains lower than its 1960 high. Turnout is low in the United States compared with other advanced democracies, but it is not equally low among all social groups. In Chapter 4 we examine social differences in turnout in detail, using both the 2016 ANES survey and the Current Population Survey conducted by the U.S. Census Bureau.

In Chapter 5 we examine how social forces influence the vote. The ANES sur- veys enable us to analyze the vote for Clinton and Trump by race, gender, region, age, education, income, union membership, and religion. The impact of these social fac- tors has changed considerably in the postwar period as the New Deal coalition broke down and new partisan alignments emerged after the critical era of the 1960. We show that minorities—specifically blacks and Latinos—are now central to the modern Democratic coalition.

14 Change and Continuity in the 2016 eleCtions

Chapter 6 examines attitudes toward both the candidates and the issues. We begin by examining voters’ feelings toward the candidates before turning our atten- tion to their appraisals of the candidates’ personal traits. We then attempt to assess the extent to which voters based their votes on issue preferences. We conclude that voters’ issue concerns were particularly important in determining their vote choices in 2016.

We then examine how “retrospective evaluations” influence voting decisions. Existing research suggests that many voters decide how to vote on the basis of past performance. In other words voters decide mainly on the basis of what the candidates or their parties have done in office, not what they promise to do if elected. In Chapter 7 we show that retrospective evaluations, particularly those related to the performance of the economy, were a powerful reason underlying citizens’ vote decisions. Perhaps most interesting we find that just one in four American voters in 2016 thought that the country was on the right track.

In Chapter 8 we explore the impact of party loyalties on voting using the ANES data. Since the 1980s there was a substantial shift in whites’ partisan loyalties— particularly in the South—toward the Republican Party. The clear advantage Democrats once held among whites dissipated. Although the 2008 election that ini- tially brought Obama to office saw a resurgence in whites’ Democratic identification, that advantage proved temporary as whites’ party loyalties reverted to near parity in 2012. Remarkably the Democrats were able to reestablish an advantage in party loyal- ties in 2016. This is a striking (and unusual) finding, not just because the Republican won the Electoral College vote but also because Clinton won the popular vote by only a few percentage points, hardly a sweep toward the Democratic Party. We examine partisanship among whites and blacks separately, tracking change from 1952 to 2016. This analysis reveals that the patterns of change among whites and blacks have been markedly different. We also compare Latino partisanship in recent elections. Finally we take a close look at the role of party loyalties in shaping issue preferences, retro- spective evaluations, and voting preferences. We find that the relationship between party identification and the vote was very strong in every U.S. election since 2000, including 2016.

In Chapters 9 and 10 we are reminded that election day 2016 featured many elec- tions. In addition to the presidential election, there were twelve gubernatorial elections, elections for thousands of state and local offices, as well as thirty-four elections for the U.S. Senate and elections for all 435 seats in the U.S. House of Representatives.65 We focus our analysis on the 2016 House and Senate elections, which are by far the most consequential for national public policy.

Chapter 9 examines the pattern of congressional outcomes for 2016 and brings to light those factors that affect competition in congressional elections. We review the pattern of incumbent success in House and Senate races between 1954, the first Democratic victory in their forty-year winning streak, and 2016. Despite citizens’ low levels of trust in government and the large portion of voters who believed the country was heading in the wrong direction, congressional incumbents were returned to office in droves. In the House, 96.4 percent of incumbents were reelected in 2016, whereas the success rate for Senate incumbents was 93 percent—not the “anti-Washington” fervor that one might infer from the outcome at the presidential level. We examine the

introduCtion 15

interplay of national and regional factors in structuring congressional election out- comes. And, of course, we give particular attention to the critical factors of candidate recruitment, incumbency, and campaign finance. Finally we speculate on the future of congressional elections and party polarization in Congress in 2018 and beyond.

Chapter 10 explores how voters make congressional voting decisions. Using the same ANES surveys we employed to study presidential voting, we examine how social factors, issues, partisan loyalties, incumbency, and retrospective evaluations of con- gressional and presidential performance influence voters’ choices for the House and Senate. We also try to determine the existence and extent of presidential “coattails,” that is, whether Democrats were more likely to be elected to Congress because of Obama’s presidential victory.

Finally, in Chapter 11, we attempt to place the 2016 elections in the proper his- torical context. Although we examine changes and continuities in American elections over the course of the nation’s history, the great advantage of our analysis is its use of high-quality surveys of the electorate over the last sixty years. This wealth of data pro- vides extraordinary insights regarding the political preferences of the America people, how those preferences have varied with time, and how they relate to voting behavior. Thus we explore the long-term changes and continuities in the politics of American national elections.

PART ONE

THE 2016 PRESIDENTIAL ELECTION

CHAPTER ONE

THE NOMINATION STRUGGLE

P residential nomination campaigns are the contests through which the two major political parties in the United States select their presidential nominees. As they have done since 1832 (Democrats) and 1856 (Republicans), the delegates who are chosen to be seated at the national party conventions do the actual selecting. However, since about 1972, both parties have used public campaigns for popular support as a way of selecting and/or instructing most delegates to the convention on how they should vote. Many people think of these primary contests as formal elections, just like those in general elections in the fall. Whereas presidential primary elections are, indeed, run by the government, they are actually designed solely to help each politi- cal party select delegates to choose its presidential nominee, and that applies only to the roughly half of the states that use primary elections to select or instruct their delegates.1 States that use the alternative means, caucus or convention procedures, instead of primaries (see what follows) do so without involving the government at all. Presidential nominations are thus a mixture of public and private selections, and they are conducted at the state level only, even though their ultimate outcome is to select the two major parties’ nominees for the only national offices that Americans elect.

In this, America is nearly unique. In almost no other country have the leaders of the major political parties’ leaders ceded so much control over candidate selection to the general public. While now and then there are primary elections run by political parties in other nations, they are rare, typically isolated to one or a few parties, and are often used only once or twice before being discarded. American nominations, on the other hand, have run this way for Democrats and Republicans since the 1970s and have become entrenched in the public’s and the political leaderships’ minds. It would be very difficult for a party to nominate someone the public did not support at near or actual majority levels in the primary season. The leadership has, in that sense, ceded its control over its own party to the general public.2 In turn that has empowered the media who seek to inform the public and the many activists, supporters, and financial donors of the presidential nomination campaigns who provide the wherewithal for most candidates to have any chance of reaching the public to win their support.

The 2016 campaigns in many respects were like all of those since the 1970s, that is, in the era of the “new nomination system,” as we call it. As we shall see there were per- haps a surprising number of similarities between the two campaigns of 2016 and their

20 Part one • the 2016 Presidential election

predecessors. Most people, however, when they speak of 2016, talk with wonder about specific and individual aspects of the campaigns regardless of the similarities to other contests. They ask “How could someone like Donald J. Trump win the Republican nomination?” and (if they disliked the outcome) “Why couldn’t Republican leaders prevent his nomination?” On the Democratic side the question more often seemed to be “Why didn’t Hillary R. Clinton win nomination more easily and quickly instead of appearing unable to reach out to larger numbers of Democrats?” or (if the out- come was viewed as negative) “How could the party fail to nominate someone more at the heart of the Democratic Party and end up with someone who so epitomizes the ‘establishment’ in this anti-establishment year?” As we will see the answers to these questions are that the two parties’ campaigns largely unfolded in replication of the many and well-established continuities established since the empowering of the public and consequent loss of party leadership control over nominations. But it is the unique properties of the two winners, especially in comparison to their major party opponents, that made the two campaigns unlike previous ones and in sometimes very important ways.

In short reforms in the late 1960s and early 1970s brought about a new form of nomination campaign, one that required public campaigning for resources and votes. The new nomination system has shaped many aspects of all contests from 1972 onward, and we examine the similarities that have endured over its more than forty- year existence. Each contest, of course, differs from all others because of the electoral context at the time (e.g., the state of the economy or of war and peace) and because the contenders themselves are different. And in the new nomination system, the rules change to some degree every four years as well. The changes in rules and the strategies that candidates adopt in light of those rules combine with the context and contenders to make each campaign unique.

WHO RAN

A first important regularity of the nomination campaign is that when incumbents seek renomination, only a very few candidates will contest them, and perhaps no one will at all. In 1972, although President Richard M. Nixon did face two potentially cred- ible challengers to his renomination, they were so ineffective that he was essentially uncontested. Ronald Reagan in 1984, Bill Clinton in 1996, George W. Bush in 2004, and Barack Obama in 2012 were actually unopposed. They were so, in large part, because even a moderately successful president is virtually undefeatable for renomina- tion. Conversely Gerald R. Ford in 1976 and Jimmy Carter in 1980 each faced a most credible challenger.3 Ford had great difficulty defeating Reagan, and Carter likewise was strongly contested by Democratic senator Edward M. Kennedy of Massachu- setts.4 Of course Obama was ineligible to run for a third term in 2016, and so there was no incumbent running in either party. President Trump may well run for reelection in 2020 or perhaps join the few incumbents who chose not to run for reelection even though eligible, such as Harry S Truman in 1952 and Lyndon B. Johnson in 1968.

The second major regularity in the nomination system concerns the contests— such as those in 2016—in which the party has no incumbent seeking renomination.

chaPter one • the noMination strUGGle 21

In such cases a relatively large number of candidates run for the nomination. For our purposes we count candidates as “running” if they were actively campaigning on January 1, 2016 (or entered even later, although none did this time). That definition means that there were twelve major candidates who sought the Republican Party’s nomination in 2016. There were actually quite a few more in 2015—by most counts seventeen—although that means that five were sufficiently “defeated” (or at least believed their chances of winning were too remote) so that they dropped out before January 1, 2016.5 By our counting procedure there were three Democratic candidates in 2016.6 Thus, in this section, we will be considering fifteen major party contenders. The numbers are higher on the Republican side and lower on the Democratic side than usual but not substantially out of the ordinary in either case.

Since 1980 there have been thirteen campaigns in which there was no incumbent seeking a major party’s nomination, and the number of major candidates that were in the race as the year began varied remarkably little: seven in 1980 (R); eight in 1984 (D); eight (D) and six (R) in 1988; eight in 1992 (D); eight in 1996 (R); six (R) and two (D) in 2000; nine in 2004 (D); eight in both parties’ contests in 2008; eight in 2012 (R); in addition to the twelve Republicans and three Democrats in 2016. Thus most such races featured at least six candidates. Only 2000 (D) and 2016 (D) had notice- ably fewer, whereas 2016 (R) had a third more candidates running than the next most crowded field (2004, D).7 We will discuss why there were fewer candidates in those two races, but note that both had larger numbers of declared candidates before our January 1 date for counting (as did most other races).

The three candidates on the Democratic side were: Hillary Clinton, who most recently served as secretary of state in the Obama administration;8 Bernie Sanders, senator from Vermont; and Martin O’Malley, former governor of Maryland. The large number of Republicans was somewhat unusual in that the list included three candidates who had held no previous political office experience and very unusual in that such candidates (such as Ben Carson and Carly Fiorina in 2016) generally fare poorly, whereas Trump went on to win the nomination and election. There were also three incumbent senators (Ted Cruz, TX; Marco Rubio, FL; and Rand Paul, KY), two incumbent governors (John Kasich, OH; and Chris Christie, NJ); three former governors (Jim Gilmour, VA; Jeb Bush, FL; and Mike Huckabee, AR), and a former senator (Rick Santorum, PA). See Table 1-1 for these and other details we will discuss shortly. We have so far illustrated two regularities: few or no candidates will challenge incumbents, but in most cases many candidates will seek the nomination when no incumbent is running. In this 2016 is not particularly exceptional.

A third regularity is that among the candidates who are politicians, most hold or have recently held one of the highest political offices. This regularity follows from “ambition theory,” developed originally by Joseph A. Schlesinger to explain how per- sonal ambition and the pattern and prestige of various elected offices lead candidates to emerge from those political offices that have the strongest electoral bases.9 This base for the presidential candidates includes the offices of vice president, senator, gov- ernor, and of course, the presidency itself. Note that even with a large number of contenders, there were no sitting members of the U.S. House who chose to run for the presidential nomination in 2016. House members do not have as strong an electoral

22 Part one • the 2016 Presidential election

Table 1-1 Candidates for Nomination to the Presidency by the Democratic and Republican Parties, 2016, With Various Aspects Pertinent to Their Candidacy

Name Last Political Office

Withdrawal Datea

Campaign Expenditures (in Millions of Dollars)b

Independent Expenditures (in Millions of

Dollars)c

Democrats Clinton Sec o f State None $187 $12

O’Malley Gov (former) 1-Feb $6 $0.40

Sanders Sen (current) 12-Jul $213 $6

Republicans Bush Gov (former) 20-Feb $87

Carson None 4-Mar $6 $5

Christie Gov (current) 10-Feb $8* $22

Cruz Sen (current) 3-May $85 $27

Fiorina None 10-Feb $11 $4

Gilmour Gov (former) 12-Feb $0.40 NA

Huckabee Gov (former) 1-Feb $4 $3

Kasich Gov (current) 4-May $19 $21

Paul Sen (current) 3-Feb $12 $5

Rubio Sen (current) 15-Mar $52 $49

Santorum Sen (former) 3-Feb $0.30 $0.20

Trump None None $62 $44

Source: Compiled by authors.

aInformation obtained from the New York Times, http://www.nytimes.com/2016/65/election/ 2016-presidential-candidate.

bInformation obtained from the Federal Election Commission, http://www.fec.gov/disclosurep/pna tional.do—and various subpages from there; accessed March 20, 2016.

cInformation obtained from OpenSecrets,org, https://www.opensecrets.org/outside-spending/summ .php?cycle-2016&disp-C&type-P.

chaPter one • the noMination strUGGle 23

base from which to run for the presidency, and they may well have to abandon a safe House seat to do so. As a result few House members run, and fewer still are strong contenders. The most prominent exception to the strong electoral base of ambition theory—Trump having had no experience in politics—will be at the center of our account of the unique features of his victory.

Most candidates in 2016, as in all earlier campaigns under the new nomination system, emerged from one of the strong electoral bases. Table 1-2 presents the data for 2016 and for all campaigns from 1972 to 2016 combined. More than two-thirds of the presidential candidates had already served as president, vice president, senator, or governor; another one in eight was a member of the U.S. House. In 2016 those ratios were largely true again, although no member of the House from either party was still a candidate as 2016 opened.10 Many of the presidents in the early years of the nation were chosen from the outgoing president’s cabinet (especially the sitting secretary of state) and other high level presidential appointees, but the cabinet is no longer a common source of presidential candidates, and the same is true for the nation’s many mayors.11 About one in seven candidates run for president without ever holding any elective office. That percentage was a little higher in 2016 as one in four of the

Table 1-2 Current or Most Recent Office Held by Declared Candidates for President: Two Major Parties, 1972–2016

Office Held

Percentage of All Candidates Who Held That

Office Number, 1972–2016 Number, 2016

President 6 8 0

Vice President 3 4 0

U.S. Senator 36 53 5

U.S. Representative 12 18 0

Governor 24 35 6

U.S. Cabinet 3 5 1

Other 6 9 0

None 10 14 3

Total 99 146 15

Sources: 1972–1992: Congressional Quarterly’s Guide to U.S. Elections, 4th ed. (Washington, D.C.: CQ Press, 2001), 522–525, 562. 1996: Paul R. Abramson, John H. Aldrich, and David W. Rohde, Change and Continuity in the 1996 and 1998 Elections (Washington, D.C.: CQ Press, 1999), 13. 2000: CQ Weekly, January 1, 2000, 22. 2004: CQ Weekly, Fall 2003 Supplement, vol. 61, issue 48. The 2008–2016 results were compiled by the authors.

24 Part one • the 2016 Presidential election

Republican candidates in 2016 (and no Democrats) had not held office previously. The big change, then, was not in the numbers but that one of those relatively politi- cally untested contenders actually won the nomination in 2016, whereas few had left any visible mark at all on the contests in earlier years.

A fourth regularity, also consistent with ambition theory, is that of the many who run in nomination contests without incumbents, only a few put their current office at risk to do so. In 2016 only two senators, Paul and Rubio, were up for reelection. Paul withdrew on February 3, after the first contest of the campaign (the Iowa cau- cuses). Rubio said he would not run for reelection as a senator, but perhaps because the Florida senatorial primary was so late (August 30), he reentered the senatorial contest after withdrawing from the presidential race and won renomination and then reelection to the Senate.12

THE RULES OF THE NOMINATION SYSTEM

The method that the two major parties use for nominating presidential candidates is unique and includes an amazingly complicated set of rules. To add to the complication, the various formal rules, laws, and procedures in use are changed, sometimes in large ways and invariably in numerous small ways, every four years. As variable as the rules are, however, the nomination system of 1972 has one pair of overriding characteris- tics that define it as a system The first is that whereas delegates actually choose their party’s nominee, it is the general public, at least those who vote in the primaries and attend the caucuses, that chooses the delegates and often instructs them as to how to vote. The second characteristic is that the candidates, as a consequence, campaign in public and to the public for their support, mostly by heavy use of traditional media, such as television and newspapers, and, increasingly, social media, such as Facebook and Twitter. The dynamics of the technology of the media make campaigning in the media dynamic as well. Obama pioneered fund-raising and campaign contacting on social media in 2008 and 2012. Trump adroitly used the “free media” of television and newspaper coverage in lieu of buying campaign ads on them, and he pioneered the use of Twitter, especially, in 2016.

The complexity of the nomination contests is a consequence of four major fac- tors. The first of these, federalism, defines the state as the unit of selection for national nominees and has been central to party nominations for nearly two centuries now. The second factor is the specific sets of rules governing primaries and caucus/con- vention procedures—established at the level of the national party in terms of general guidelines and then more specifically by state parties and/or state laws—these rules are at the heart of the nomination system of 1972. These rules govern delegate selec- tion (and sometimes dictate instructions for delegates’ presidential voting at the con- vention). The third factor is the set of rules about financing the campaign, which are also the oft-revised products of the reform period itself, starting in 1972. The fourth factor is the way in which candidates react to these rules and to their opponents, strat- egies that grow out of the keen competition for a highly valued goal. These factors are described in more detail in the sections that follow.

chaPter one • the noMination strUGGle 25

Federalism or State-Based Delegate Selection

National conventions to select presidential nominees were first held for the 1832 election, and for every nomination since then, the votes of delegates attending the conventions have determined the nominees. Delegates have always been allocated at the state level; whatever other particulars may apply, each state selects its parties’ del- egates through procedures adopted by state party organizations whether they choose to use caucuses and conventions, by state law, or the party organization wants to use a primary election, or both. Votes at the convention are cast by a state’s delegation, and in general the state is the basic unit of the nomination process. Thus there are really fifty separate delegate selection contests in each party.13 There is no national primary, nor is there serious contemplation of one.

The fact that there are more than fifty separate contests in each party creates numerous layers of complexity, two of which are especially consequential. First, each state is free to choose delegates using any method consistent with the general rules of the national party. Many states choose to select delegates for the parties’ conventions via a primary election. States not holding primaries use a combination of caucuses and conventions, which are designed and run by each political party and not by the state government. Caucuses are simply local meetings of party members. Those attend- ing the caucuses report their preferences for the presidential nomination and choose delegates from their midst to attend higher-level conventions such as at the county, congressional district, state, and eventually national levels.

The second major consequence of federalism is that the states are free (within the bounds described as follows) to choose when to hold their primaries or caucuses. These events are thus spread out over time, although both parties now set a time period—the delegate selection “window”—during which primaries and caucuses can be held. Both parties began delegate selection on February 1, 2016, with the Iowa caucuses (a month later than in 2012), Republicans closed their delegate selection process with five states (CA, MT, NJ, NM, and SD) holding primaries on June 7, whereas Democrats in DC held a primary on June 14. The Republicans, concerned about how long the Romney nomination in 2012 took to unfold to victory, not only favored this shortening of the length of the primary season but also tried to regulate front-loading even further. In particular they required that states holding their pri- maries before March 15 had to use some kind of proportional allocation method so that the delegates awarded to candidates were to some degree proportionate to the votes those candidates received in the primary or caucus. It was not until March 15 that states could use the winner-take-all (WTA) rule, such that the candidate with the most votes wins all that state’s delegates.14 WTA rules are often favored by GOP states, due to the larger impact that state’s delegation might have on the race, concentrating their vote on a single candidate.15

The Nomination System of 1972: Delegate Selection

Through 1968 presidential nominations were won by appeals to the party leader- ship. To be sure public support and even primary election victories could be impor- tant in a candidate’s campaign, but their importance stemmed from the credibility

26 Part one • the 2016 Presidential election

they would give to the candidacy in the eyes of party leaders. The 1968 Democratic nomination, like so many events that year, was especially tumultuous.16 The result was that the Democratic Party created a committee, known as the McGovern- Fraser Commission, which proposed a series of reforms that were proposed to the Democratic National Committee between 1969 and early 1972 and then finally adopted by the party convention in 1972. The reforms were sufficiently radical in changing delegate selection procedures that they, in effect, created a new nomina- tion system. Although it was much less aggressive in reforming its delegate selection procedures, the Republican Party did so to a certain degree. However, the most con- sequential results of the Democratic reforms for our purposes—the proliferation of presidential primaries and the media’s treatment of some (notably the Iowa) caucuses as essentially primary-like—spilled over to the Republican side as well.

In 1968 Democratic senators Eugene J. McCarthy of Minnesota and Robert F. Kennedy of New York ran very public, highly visible, primary-oriented campaigns in opposition to the policies of President Lyndon B. Johnson, especially with respect to the conduct of the Vietnam War. Before the second primary, held in Wisconsin, Johnson surprisingly announced, “I shall not seek and I will not accept the nomi- nation of my party for another term as your President.”17 Vice President Hubert H. Humphrey took Johnson’s place in representing the presidential administration and the policies of the Democratic Party generally. Humphrey, however, waged no public campaign; he won the nomination without entering a single primary, thereby splitting an already deeply divided party.18 Would Humphrey have won the nomina- tion had Robert Kennedy not been assassinated the night he defeated McCarthy in California, effectively eliminating McCarthy as a serious contender? No one will ever know. Democrats including Humphrey himself did know, however, that the chaos and violence that accompanied Humphrey’s nomination clearly indicated that the nomi- nation process should be opened to more diverse candidacies and that public partici- pation should be more open and more effective in determining the outcome. He thus offered a proposal to create the McGovern-Fraser Commission, as it was popularly called, which was accepted by the Democratic National Committee.

The two most significant consequences of the reforms were the public’s great influence on each state’s delegate selection proceedings and the proliferation of presi- dential primaries. Caucus/convention procedures, however, also became timelier, were better publicized, and in short, were more primary-like. Today the media treat Iowa’s caucuses as critical events, and the coverage of them is similar to the coverage of primaries—how many “votes” were “cast” for each candidate, for example. Indeed the party organizations formally recognized this fact. The Iowa Republican Party, for example, held a secret balloting among caucus attenders that determined how the delegates to subsequent levels of conventions were to be allocated among supporters of the candidates.19 Iowa Democrats, in their turn, conducted a standing “vote” of attenders to the same effect.

Whereas the McGovern-Fraser Commission actually recommended greater use of caucuses, many of the state party officials concluded that the easiest way to conform to the new Democratic rules in 1972 was to hold a primary election. Thus the number of states (including the District of Columbia) holding Democratic primaries increased from fifteen in 1968 to twenty-one in 1972 to twenty-seven in 1976, and the number

chaPter one • the noMination strUGGle 27

of Republican primaries increased comparably. The numbers peaked in 2000, when forty-three states conducted Republican primaries, and Democratic primaries were held in forty states. In 2016 there were thirty-nine primaries on each side. Thus it is fair to say that the parties’ new nomination systems have become largely based on primaries or in more primary-like conventions.

The only major exception to this conclusion is that about 15 percent of delegates to the Democratic National Convention were chosen because he or she is an elected officeholder or a Democratic Party official. Supporters of this reform of party rules (first used in 1984) wanted to ensure that the Democratic leadership would have a formal role to play at the conventions of the party. These “superdelegates” may have played a decisive role in the 1984 nomination of Walter F. Mondale, in the nomination of Obama over Clinton in 2008, and again for Clinton’s nomination in 2016, when she, like Mondale and Obama, at one point had a majority of the non-superdelegates but not a majority of all delegates.20 Each candidate needed only a relatively small number of additional superdelegates to commit to vote for them to win the nomina- tion. All three received those commitments soon after the regular delegate selection process ended, and with that, they were assured the nomination.21

The delegate selection process has, as noted, become considerably more front- loaded.22 The rationale for front-loading was clear enough: the last time California’s (actual or near) end-of-season primary had an effect on the nomination process was in the 1964 Republican and the 1972 Democratic nomination contests. Once can- didates, the media, and other actors realized, and reacted to, the implications of the reformed nomination system, the action shifted to the earliest events of the season, and nomination contests, especially those involving multiple candidates, were effec- tively completed well before the end of the primary season. More and more state parties and legislatures (including, for a while, California’s) realized the advantages of front-loading, bringing more attention from the media, more expenditures of time and money by the candidates, and more influence to their states if they held primaries sooner rather than later.

Soon, however, other factors started to affect state decisions. First, the rewards for early primaries were concentrated in a relatively small number of the very earli- est primaries. And as we have noted, the national parties regulated which ones could go when and threatened to penalize states that violated the national party decisions. Indeed Michigan and Florida were actually penalized in 2008 and 2012 for holding their contests too early in the season. In addition the very early presidential primaries forced states to make an increasingly difficult choice. If they held their presidential primaries early in the year, they had to decide whether to hold the primary elections for all other offices at the same time, which was proving quite a bit earlier than made sense for candidates for local, state, and even national congressional posts, or to pay the costs of running two primaries, one for the president and one much later for all other offices.23 Some states like California, for example, which were not able to reap the major benefits of being among the very earliest of events, chose to return to late in the season.

If the rationale for front-loading was clear by 1996, when it first became con- troversial, the consequences were not. Some argued that long-shot candidates could be propelled to the front of the pack by gathering momentum in Iowa and

28 Part one • the 2016 Presidential election

New Hampshire and could, before the well-known candidates had a chance to react, lock up the nomination early. The alternative argument was that increasing front- loading helps those who begin the campaign with the advantages associated with being a front-runner, such as name recognition, support from state and local party or related organizations, and most of all, money. The dynamic of this adjustment, described in the following paragraphs, can be seen clearly in Figure 1-1, which reports the week in which the winning candidate was assured nomination in contested nomi- nation campaigns since 1976.

Indeed as the primary season has become more front-loaded, the well-known, well-established, and well-financed candidates have increasingly dominated the pri- maries. Sen. George S. McGovern of South Dakota and Carter won the Democratic nominations in 1972 and 1976, even though they began as little-known and ill- financed contenders. George H. W. Bush, successful in the 1980 Iowa Republican cau- cuses, climbed from being, in his words, “an asterisk in the polls” (where the asterisk is commonly used to indicate less than 1 percent support) to become Reagan’s major contender and eventual vice presidential choice and his successor to the presidency. And Colorado senator Gary Hart nearly defeated former Vice President Mondale in 1984. In 1988 the two strongest candidates at the start of the Republican race, George H. W. Bush and Bob Dole, contested vigorously, with Bush winning, while

Figure 1-1 Length of Multicandidate Campaigns: Two Major Parties, 1976–2016

D 19

R 19

D 19

R 19

D 19

R 19

R 20

D 20

R 20

D 20

Party and Year

N um

be r

of W

ee ks

a fte

r N

ew H

am ps

hi re

P ri

m ar

Source: Compiled by authors.

chaPter one • the noMination strUGGle 29

their presence basically locked other lesser-known contenders out. Gov. Michael S. Dukakis of Massachusetts, the best-financed and best-organized (albeit little known) Democrat, won the nomination surprisingly easily. Bill Clinton’s victory in 1992 appeared, then, to be the culmination of the trend toward an insuperable advantage for the strongest and best-financed candidates. Clinton was able to withstand scandal and defeat in the early going and eventually cruise to victory.

The campaign of former Democratic Senator Paul Tsongas of Massachusetts in 1992 illustrates one important reason for Clinton’s victory. Tsongas defeated the field in New Hampshire, and as usual, the victory and the media attention it drew opened doors to fund-raising possibilities unavailable to him even days earlier. Yet Tsongas faced the dilemma of whether to take time out of daily campaigning for the public’s votes so that he could spend time on fund-raising or to continue campaigning in the upcoming primaries. If he campaigned in those primaries, he would not have the opportunity to raise and direct the funds he needed to be an effective competitor. Front-loading had simply squeezed too much into too short a post-New Hampshire time frame for a candidate to be able to capitalize on early victories as, say, Carter had done in winning the nomination and election in 1976. The events of 1996 supported the alternative argument—that increased front-loading benefits the front-runner—even though it took nearly all of Dole’s resources to achieve his early victory that year.24

This lesson was not lost on the candidates for 2000, especially George W. Bush. In particular he began his quest in 1999 (or earlier!) as a reasonably well-regarded governor but one not particularly well-known to the public outside of Texas (although, of course, sharing his father’s name made him instantly recognizable). He was at that point only one of several plausible contenders, but he worked hard to receive early endorsements from party leaders and raised a great deal of money well ahead of his competition. When others sought to match Bush’s early successes in this “invisible pri- mary,” they found that he had sewn up a great deal of support. Many, in fact, withdrew before the first vote was cast, suddenly realizing just how Bush’s actions had length- ened the odds against them. Bush was therefore able to win the nomination at the very opening of the primary season. Incumbent Vice President Al Gore, on the other side, also benefited from the same dynamics of the invisible primary made manifest by front-loading, although in the more classical role of one who began the nomination season as the odds-on favorite and therefore the one most able to shut the door on his opposition well before it was time for most voters to cast their ballots.25

In 2004 there was no strong leader of the contest before the Democratic cam- paign began. Howard Dean burst on the scene and rather surprisingly into a lead before dropping nearly as suddenly.26 As a result there was a period of uncertainty in the shape of the contest, followed by solidifying support around long-time sena- tor John Kerry, who thereby benefitted more from lack of anyone able to compete strongly against him than any rule.

The pre-primary period on the Republican side in 2008 was quite variable, with first McCain, then Giuliani, then Romney surging to the front. McCain’s campaign was considered all but dead in the water by that point, but it regathered strength before 2007 ended. There was, then, no strong front-runner in the GOP; the cam- paign was wide open. In fact some pundits imagined former Arkansas Governor Mike Huckabee had become a favorite to win, and so McCain’s victory in the Iowa

30 Part one • the 2016 Presidential election

caucuses was a genuine surprise (at least from the perspective of, say, October 2007). On the Democratic side, Hillary Clinton was a clear front-runner. In retrospect it was also clear that Obama had developed an impressive organization both by mobiliz- ing support across the nation and by fund-raising, especially through adroit use of the Internet. Thus once his organizational strength became publicly visible, it was no surprise that he and Clinton easily defeated their rivals. Having boiled down to a two-candidate contest, each had carved out their own bases of support, and neither could decisively defeat the other. Obama did have a slight lead throughout much of the primary season, but because of the superdelegates, it was too slight a lead to be able to secure an outright majority of delegates until after the primary season ended. As heretofore unbound superdelegates determined their choices, they soon favored Obama sufficiently to put him over the top.

The 2012 Republican contest had some similarities to 2008, with Romney mov- ing from his also-ran slot to replace McCain as the candidate who early on seemed strong, lost steam, and then resurged back to victory. One effect of the modest reversal in front-loading was that Romney, even though ahead, was not able to completely shut the door on his opposition until much later in the season. Simply too few delegates were selected as early in 2012 as in, say, 1980. This extended length of time had several effects. The most important appears to have been that the slowing of the delegate selection process, although still relatively highly front-loaded, permitted Romney’s opponents to run negative campaigns against him, quite possibly hurting his ability to shape his own image and providing fodder for attacks in the general election campaign before the campaign had selected enough delegates for him to claim what proved to be a rather straightforward nomination victory.

Much the same appeared to happen again in 2016 on the Republican side. The unusual nature of someone like Trump emerging as the leading contender (even after losing the Iowa caucuses but righting his campaign and its dynamic growth in New Hampshire) led to calls for the remaining candidates (fairly soon into the sea- son, the race reduced effectively to Trump versus Cruz, Kasich, and Rubio) and the “Republican establishment” to figure out a way to stop Trump. When that failed to happen, the divided opposition allowed Trump to build his delegate lead to victory.27 On May 3, the night of the Indiana primary, his last major opponent, Ted Cruz, with- drew his candidacy, although Trump was still short of having a majority of delegates on his side. But from that night onward, he was unchallenged and thus the “presumptive nominee.” Thus continued active opposition until May did yield a longer period in which Republicans were criticizing the eventual nomination, sometimes quite strenu- ously, in spite of a relatively straightforward and convincing win by Trump.

The slowed rate of delegate selection also affected the Clinton-Sanders contest on the Democratic side. Clinton, as she had in 2008, began her quest for nomination as a very strong front-runner, especially after those who appeared likely to be among her strongest opponents, Sen. Elizabeth Warren (MA) and Vice President Biden, decided not to run. Of the remaining actual candidates, Sanders effectively had the liberal wing of the party on his own, and the race narrowed almost immediately to a two-person contest. In such races it is typically the case, as here, that both candidates have their own constituency in their party’s base and are thus difficult to defeat. That is to say that these races—in 2016 like 2008 and others before them—take a long time

chaPter one • the noMination strUGGle 31

to resolve. Even when Clinton had secured an outright majority of delegates (which was at the end of the season anyway), Sanders failed to concede and thus continued to be able to criticize Clinton and to remain a holding place for liberal Democrats who were disenchanted with her.

These effects can be seen in Figure 1-2, which reports the cumulative selection of delegates. As can be seen there, 1976 (the first primary season defined by the rules adopted at the 1972 Democratic convention) shows a slow, gradual increase in the number of delegates selected. It is not until week thirteen, just over a month before the season ends, that 50% of the delegates were selected, and even later that a sufficiently large proportion of the delegates had been selected to make a majority likely to be held by the leading candidate, if he or she faced any opposition at all. The 2000 sea- son was dramatically different, with the 50% mark being reached in week six (indeed reaching nearly two-thirds of the delegates selected by that week). Finally the slight retreat from such heavy front-loading in 2016 is visually apparent, but it is also appar- ent that it is rather slight, looking far more like the 2000 apogee than the 1976 perigee.

The final consequence—and possibly the most important for differentiating the nomination system of 1972 from its predecessors—is “momentum,” the building of success over time during the extended campaign period, such that every nomination has, so far, always been decided before the convention balloting and always going to the candidate who won the greatest support from the party’s electorate.

The most significant feature of the nomination process, from the candidates’ per- spectives, is its dynamic character. This system was designed to empower the general

Figure 1-2 Front-Loading: Comparing Democratic Party Delegate Totals Weekly, 1976, 2000, and 2016

Number of Weeks Starting with the lowa Caucuses

Pe rc

en ta

ge o

f D el

eg at

es S

el ec

te d,

Cu m

ul at

iv e

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

2016 2000 1976

Source: Compiled by authors.

32 Part one • the 2016 Presidential election

public, giving it opportunities to participate more fully in the selection of delegates to the national party conventions. The early state delegate selection contests in Iowa and New Hampshire allowed largely unknown candidates to work a small state or two using the “retail” politics of door-to-door campaigning to achieve a surprising success that would attract media attention and then money, volunteers, and greater popular support. In practice this was exactly the route Jimmy Carter followed in 1976.

John H. Aldrich developed this account of “momentum” in campaigns, using the 1976 campaigns to illustrate its effect. He first showed that there is no stable balance to this process.28 In practical terms he predicted that one candidate will increasingly absorb all the money, media attention, and public support and thereby defeat all oppo- nents before the convention. He further showed that the tendency for this process to focus rapidly on a single winner increases the more candidates there are. This finding was just the opposite of the original speculation and, indeed, what at the time seemed obvious: the greater the number of candidates, the longer it would take to reach vic- tory. But commonsense was not a helpful guide in this case. Like other contests with large numbers of contenders, the Republican race of 2016 illustrates the power of momentum. Trump did not start off the campaign with a large lead in popular support, but he built that over the course of the campaign, eventually all but crushing even his strongest opponents and forcing their mostly early exits.

There is one exception to this pure “momentum” result: the possibility of an unstable but sustainable balance with two candidates locked in a nearly precise tie. Early campaigns offered two illustrations compatible with two candidates in (unsta- ble) equipoise, the 1976 Republican and 1980 Democratic contests. In both the 1984 Democratic and 2008 Democratic contests, the campaigns began with a large num- ber of candidates. Each featured a strong, well-financed, well-known, well-organized candidate (former Vice President Mondale and Hillary Clinton, respectively) who, it turned out, was challenged strongly by a heretofore little-known (to the public) can- didate who offered a new direction for the party (Sen. Gary Hart and Sen. Barack Obama, respectively). The multicandidate contest quickly shrank to just two viable can- didates. The 2016 Democratic contest fits the pattern of balanced two-party contests very nicely, with neither bloc of voters willing to move from Sanders to Clinton nor from Clinton to Sanders in any great numbers, as inevitably happens in a momentum- driven contest.

The Nomination System of 1972: Campaign Finance

Campaign finance is the third aspect of the reform of the presidential nomina- tion process. In this case changes in law (and regulation in light of the law) and in the technology for raising money in nomination contests have made the financial context widely different from one campaign to the next. The 2016 campaign was no excep- tion. These candidates were able to learn some of the lessons from strategies tried in 2012, which was the first run under a new (de-)regulatory environment in light of the Supreme Court case popularly known as Citizens United (2010), and so in 2012 candidates tried a large variety of new or modified strategies for campaign financing in response. Two major changes were the increased reliance on what are known as independent expenditures by a number of candidates, and Trump’s strategy, which

chaPter one • the noMination strUGGle 33

focused less on raising money but instead in getting the media to cover his campaign much more highly than those of other candidates. This was a strategy he believed to be a more effective use of “free” media than what impact higher expenditures for purchasing time on the paid media would offer.

Our story begins, however, with the Federal Election Campaign Act of 1971 and especially amendments to that act in 1974 and 1976. The Watergate scandal dur- ing the Nixon administration included revelations of substantial abuse in raising and spending money in the 1972 presidential election (facts discovered in part in imple- menting the 1971 act). The resulting regulations limited contributions by individuals and groups, virtually ending the power of individual “fat cats” and requiring presi- dential candidates to raise money in a broad-based campaign. The federal govern- ment would match small donations for the nomination, and candidates who accepted matching funds would be bound by limits on what they could spend.

These provisions, created by the Federal Election Commission to monitor cam- paign financing and regulate campaign practices, altered the way nomination cam- paigns were funded. Still, just as candidates learned over time how to contest most effectively under the new delegate selection process, they also learned how to campaign under the new financial regulations. Perhaps most important, presidential candidates learned—although it is not as true for them as for congressional candidates—that “early money is like yeast, because it helps to raise the dough.”29 They also correctly believed that a great deal of money was necessary to compete effectively.

The costs of running presidential nomination campaigns, indeed campaigns for all major offices, have escalated dramatically since 1972. But a special chain of stra- tegic reactions has spurred the cost of campaigning for the presidential nomination. The Citizens United case accelerated the chain reaction by creating a much more fully deregulated environment.

When many states complied with the McGovern-Fraser Commission reforms by adopting primaries, media coverage grew, enhancing the effects of momentum, increasing the value of early victories, and raising the costs of early defeat. By 2008 very few candidates were accepting federal matching funds because doing so would bind them to spending limits in individual states and over the campaign as a whole, and these limits were no longer realistic in light of campaign realities. By 2012, only one candidate, former Louisiana Governor Buddy Roemer, applied for federal fund- ing, and his candidacy was considered sufficiently hopeless that many debates did not even bother to include him among the contestants. No major candidates accepted matching funds in 2016.

Much money was being raised, however. Through May 2008, for example, the fund-raising totals for the three major contenders were $296 million for Obama, $238 million for Clinton, and $122 million for McCain.30 By the same point in 2012, Romney reported raising $121 million, with Paul having raised $40 million, Gingrich $24 million, and Santorum $22 million. See Table 1-1 for reports on campaign expen- ditures in 2016. Note that, for example, Clinton and Sanders spent much more than Romney raised in 2012.

The 2008 campaign also marked a dramatic expansion in the use of the Internet to raise money, following on the efforts of Democrat Howard Dean, the former gov- ernor of Vermont, in 2004 (and, to an extent, McCain in 2000). Ron Paul, for example,

34 Part one • the 2016 Presidential election

raised more than $6 million on a single day, December 6, 2007, through the Internet. But Obama’s success in 2008 served as the model for future campaigns, such as the $55 million he raised in February at a critical moment for the campaign.31

The Citizens United decision in 2010 changed the landscape dramatically. In the narrow it overturned the 2002 Bipartisan Campaign Reform Act and held that cor- porations and unions could spend unlimited money in support of political objectives and could enjoy first amendment free speech rights, just as individuals could. These organizations, however, continued to be banned from direct contribution to candi- dates and parties. The case, and especially a subsequent one decided by the U.S. Court of Appeals in light of this case, spurred the development of what are known as “super PACs,” which are political action committees that can now accept unlimited contri- butions from individuals, corporations, and unions and spend as much as they like so long as it is not in explicit support of a candidate or party’s election campaign or coordinated with their campaign organization.32

According to data from the Center for Responsive Politics, expenditures on behalf of the three major nomination contenders were quite large. In 2012 about $14 million was spent on behalf of Romney, $19 million for Gingrich, and $21 mil- lion for Santorum. Data from Open Secrets are reported in Table 1-1 for the 2016 campaign. Note that the expenditures on behalf of many candidates, especially Republicans, had as much, or even more, spent on behalf of their campaigns than they spent themselves.33 These organizations altered the terms of the campaign in that their expenditures had to be independent of the candidates and their (and their party’s) organizations. It is therefore not necessarily the case that the candidate and, in the fall, the party will retain total control over the campaign and its messages.

Another consequence of these changes is that what were previously dubbed “fat cats” are once again permitted. The 2012 exemplar was Sheldon Adelson, a casino magnate and a strong supporter of Israel. He contributed $10 million to the Winning Our Future super PAC in support of Newt Gingrich, contributing about half that total before the South Carolina primary and the other half before the next primary in Florida. His public support is rare, however. Most of the super PACs are funded and led by small numbers of individuals, and we often do not know their names.

Note that in 2016, although Trump did raise and spend a good deal of money, much of his expenditures came later in the game, and he made a very public case for not spending a dime of his own money until late into the campaign. Certainly he spent much less than either of the two major Democrats, both of whom raised sums comparable to the Obama-Clinton race in 2008. But he did spend much more than his opponents, with only Rubio being at all close behind. And, of course, he eventually received a lot of support from super PACs, even though Bush also had a great deal spent on his behalf (even if ineffectually). The lessons are that money is very helpful, that early money still must be better than that raised late, that candidates are still try- ing to figure out the best configuration in this largely deregulated campaign finance regime, and that, as Trump’s approach shows, it is not money that is important, but what it will buy. We will discuss his campaign strategy in a little while, but this also raises the final lesson for the future, that if candidates come to rely on super PACs, they risk control over their campaign, or they simply agree to adopt the stances of their party or its backers as their own. This concession to the party and its “image” is

chaPter one • the noMination strUGGle 35

greatly strengthened due to the dramatic increase in partisan polarization that began around 1980 and continues to increase today.

STRATEGY AND THE CANDIDATES’ CAMPAIGNS IN 2016: THE ELECTORAL SETTING AND HOW THE CANDIDATES WON THEIR NOMINATIONS

The Strategic Context: One of the most dramatic changes of the last half century has been the increase in partisan polarization, which generally means an increasing similarity of attitudes and preferences within each party and a substantial increase in divergence of opinion between the two parties. The leading indicators of this increase in partisan polarization have been among the party elites and especially their elected officeholders.34 What is less clear is whether the electorate has followed polarization among elites (or, even less obviously, led elite polarization), and if so, how much the electorate has followed (or led). Some argue that there has been little change for decades, especially on such key measures as issue and ideological prefer- ences. In this view polarization in the electorate is relatively small, with the result that the electorate continues to be basically moderate in its views.35 Others point to at least some increased polarization in preferences between partisan identifiers, particularly among the more attentive and engaged in politics, such as those among the most likely to participate in primaries and caucuses.36

The clearest evidence of partisan polarization in the electorate lies in divergences between the two parties in other ways than their attitudes toward issues and even their ideological views. For example, Marc Hetherington and Jason Husser showed that there has been a dramatic decline in trust across party lines, whereas Shanto Iyengar and colleagues have shown that emotional responses have become much more polar- ized along party lines in the electorate.37 Finally Gary Jacobson demonstrated that the so-called approval ratings of presidents (something we analyze in Chapter 7 in detail) went from having only a modest amount of partisan differences to becoming deeply divided by party.38

Here we illustrate that the context for the 2016 presidential nomination campaign has become much more deeply polarized along party lines than it was in 1980 in terms of overall affective evaluations of the candidates running for the presidential nomina- tion. The ANES ran nation-wide surveys in January 1980 and in January 2016.39 These years turn out to be especially appropriate ones for this look at partisan polarization of candidate evaluations for two reasons. The 1980 presidential election, as it happens, was the year in which elite partisan polarization turned and began its sharp increase, and thus we have data from the beginning and (current) end points of elite polariza- tion. In addition both parties’ nominations were strongly competitive. In both years the Democrats witnessed a strong two-person contest that lasted throughout the primary season. In both years the Republicans chose over a larger number of candidates that more quickly ended with Reagan and Trump’s victories, respectively, but were nonethe- less hotly contested in January. Especially on the Republican side, contenders argued for their candidacy in part by claiming to be supported by Democrats.

36 Part one • the 2016 Presidential election

The survey data reported in Figures 1-3 and 1-4 show an increase in partisan polarization in the electorate in two ways. The figures report data using the so-called candidate thermometers, which ask how “warmly” or “coolly” the respondent feels toward the candidate, where 100 is the warmest possible feeling, 50 is neutral (nei- ther warm nor cool), and 0 is the coldest possible feelings toward that candidate.40 Figure 1-3 reports the difference between how the average Democrat and average Republican evaluated that candidate.41 This is probably the most direct measure of partisan polarization of candidate evaluations. Those who are concerned about parti- san polarization often point to the decline in the ability to work across party lines—the decline in bipartisanship—which is compounded by an apparent growth in emotional hostility to those on the other side of party lines. In Figure 1-4, therefore, we report the percentage of partisans who rate the relevant candidate of the opposite party posi- tively (i.e., warmly or above 50 degrees). Even if there are large gaps in evaluations, as in Figure 1-3, the ability to see the opposition candidate positively bodes more favor- ably for bipartisanship.

The two figures have strong and reinforcing findings. In 1980, although each party felt more positively toward its own candidates than did those identifying with the other party, the difference between the two parties was fairly muted, with

Figure 1-3 Difference in Average Thermometer Rating: Selected Candidates, January 1980 and 2016

Bu sh

1980

Ab so

lu te

V al

ue o

f P ar

tis an

D iff

er en

ce s

in T

he rm

om et

er S

co re

2016

Ca rte

r Do

Ke nn

ed y

Re ag

an Cr

Ru bio

Sa nd

ers

Tru mp

Cl int

Source: Authors’ analysis of “feeling thermometers” is from the respective ANES surveys.

Note: Data are weighted.

chaPter one • the noMination strUGGle 37

differences typically under 20 degrees on the 100-point scale. Only evaluations of Kennedy were higher, at 27 points. None of these were as large as the smallest partisan gap in 2016, and the partisan polarization between the two eventual nominees was great—more than 40 points for Trump and more than 50 for Clinton. Perhaps even more dramatic, in 1980 majorities of the opposition felt warmly toward Carter and toward eventual vice presidential nominee Bush, with Reagan evaluated positively by many Democrats. Conversely, in 2016, none of the candidates were evaluated posi- tively by a quarter of the opposition, and fewer than one in five partisans felt warmly toward the candidate who eventually won the other party’s nomination. There was very little chance that either party could nominate a candidate with any appreciable support among the opposition in 2016, quite unlike Reagan’s success in winning over “Reagan Democrats” in his campaigns and presidency.42 Even at the start of the cam- paign, that is, the contenders were in a strategic context that rewarded focus on one’s own party with no incentive to build toward a cross-party coalition, either in open primary states or for the general election—or thereafter when in office. The public has become deeply divided emotionally over our electoral contests even before they have barely begun in a way that simply was not true a generation ago.

Figure 1-4 Warm Feelings from Opposite Partisans: Selected Candidates, January 1980 and 2016

Pe rc

en t o

f O pp

os ite

P ar

tis an

s Sc

or in

g Ca

nd id

at e

ab ov

e 50

D eg

re es

o n

Th er

m om

et er

1980 2016

Bu sh

Ca rte

r Do

Ke nn

ed y

Re ag

Cl int

on Cr

Ru bio

Sa nd

ers

Tr um

Source: Authors’ analysis of “feeling thermometers” is from the respective ANES surveys.

Note: Data are weighted.

38 Part one • the 2016 Presidential election

Whereas, therefore, it is always true that nomination politics leads candidates to focus on their party to win, this was truer in 2016 than ever before. How, then, did the candidates win? We begin with Democrats. Hillary Clinton started both the 2008 and 2016 campaigns in the enviable position of an unusually well-known candidate in the public, with many areas of support in the Democratic electorate already won and with a great deal of support from leading Democrats and those with access to fund- ing sources. Her position was thus well-defined with appeal to moderate Democrats, women (especially older women), African Americans, and those who have long been “Clintonians.” As such if she had a vulnerability in the Democratic primary electorate, it was on the party’s left.

Sanders was able to clearly fashion his appeal to that very constituency, even though not nearly as well-known to Democratic voters in January. He had long been on the left. Indeed his original election as mayor of Burlington, VT, was as a Socialist, and he had long served in Congress as an Independent who caucused with the Democrats but retained his independent status. Only recently had he formally and publicly affiliated as a Democrat, making his potential nomination viable. Unlike earlier nomination contests, few contested for the liberal portion of the Democratic electorate, with the apparently strongest contenders, especially Sen. Elizabeth Warren (MA) declining to run. Given his late and, in some measures, begrudging entry into the Democratic Party, he lacked the close interactions and shared service to the party leadership that Clinton had with so many of them. And both his lack of seeking sup- port from Democratic donors (Vermont did not require the same level of campaign expenditures as larger states) and his vocal stance against super PACs and other organizations that serve as sources of campaign resources, he also faced obstacles to expanding upon his electoral base. And, indeed, that is how the campaign worked out.

Perhaps the biggest surprise of the Democratic campaign was how Sanders was able to make a strong appeal, especially to younger voters on the left, and to turn college students, among others, into active supporters. As a further result, Clinton, even as she emphasized the more liberal parts of her agenda and adopted more left- wing positions on key issues, was unable to expand her base on the left, as Sanders was demonstrably a liberal (even socialist) candidate, and she was forced to publicly change her stances to try to reach Sanders’s supporters. Conversely Sanders had too little standing among Democratic leaders (such as superdelegates) or more moderate Democrats in the public, nor even among the large constituency of African Americans. Thus he too was unable to cut into Clinton’s strengths and expand his base of support.

Clinton won the Iowa caucuses (a real victory, given her loss there in 2008) and held Sanders to a relatively small victory in New Hampshire, sitting next door to Sanders’s home (see Table 1-3 for delegates won by these two candidates over the nomination campaign). Of course that meant that Sanders did reasonably well in Iowa and won New Hampshire, cementing him as a credible candidate, able to be con- sidered by voters over the long haul. Still, his inability to shake much of her support meant that in the March 1 “Super Tuesday” primaries, most of which were in the South and thus featured two sources of Clinton strength, moderate white Democrats and African Americans, followed by her largely similar victories in the large, industrial states of the “Rust Belt,” meant that Sanders fell behind in the count of delegates won, even as he very slowly approached her standing in the public opinion polls. Because

chaPter one • the noMination strUGGle 39

the Democratic rules require some form of proportional selection of delegates (i.e., roughly in proportion to the percentage of votes received), Clinton’s delegate lead became simply insurmountable.

As noted earlier, however, it took Clinton until June for her to win an absolute majority of the delegates and thus achieve victory, a victory strengthened by heavy support of superdelegates, those party leaders with whom she had so long worked. In short two candidates with clear and distinct appeals were able to hold their own sup- port, but both were unable to expand into that of the opponent. As a result the early lead Clinton had in public opinion polls (8 points in the February CBS/New York Times [NYT] poll) held steady throughout the season (the same poll in May had her with a 7-point lead), and that relatively small lead in the national polls became a small but winning majority in the delegate count as state after state selected its delegates.

The Republican side was, of course, rather different in many ways. Still, Trump held a 17-point lead over Cruz, his closest competitor in the January (and February) CBS/NYT poll that in April, just before Cruz withdrew, stood at 13 points. Once delegate selection started, that is, all the sound and fury of Republican candidates attacking each other on increasingly personal grounds had at best minor effects on Trump’s public standing and lead in the delegate count. To be sure, in 2016 or 2017, many different Republicans got their day in the sun, but none were able to close the Trump lead.

Perhaps surprisingly Trump lost to Cruz in Iowa (and nearly fell to third place there) but righted his ship in New Hampshire, South Carolina, and virtually every- where thereafter, consistently winning most of the states with pluralities (only occa- sionally with actual majorities) but picking up the bulk of the delegates in state after state. Trump did lose two large states, Ohio and Texas, and with those losses in votes, he also lost even larger percentages of their delegates, but these were divided between Kasich and Cruz, respectively (each winning their home states). Even so, Trump was able to carry many other larger states (perhaps most significantly, Rubio’s home state of Florida). While opponents considered ways to unite their forces to maximize lever- age against Trump, no plan was able to be worked out. Further, after the earliest states had chosen, the Republican Party rules permit states to use WTA rules so that the candidate who wins more of the larger states wins a far higher percentage of the del- egates needed to win nomination. Thus Trump was able to move consistently and smoothly toward victory, as can be seen in Table 1-4, which reports the results of each Republican contest.

This relatively placid and straightforward account of how Clinton and Trump won nomination belies the media frenzy that accompanied both campaigns—and especially these two candidates in particular. These circumstances are those that most remember, even though their consistent and largely unchecked (and apparently uncheckable) drives to victory are the real story of how to win nominations in the post-1972 nomination system.

Still, both were tagged with problems (quite reasonably understood as of their own doing) that would dog their campaigns throughout the spring, summer, fall, and in Trump’s case, into the White House itself as we discuss in subsequent chapters. Clinton was tarnished with three charges that yielded appearances of corruption— the financing of the Clinton Foundation, the events that led to the deaths of four

40 Part one • the 2016 Presidential election

Table 1-3 Democratic Nomination Results, 2016: Bound Delegates Won in State Primaries and Caucuses—Clinton and Sanders

Date State Clinton Sanders

1-Feb Iowa 23 21

9–Feb N.H. 9 15

20-Feb Nev. 20 15

27-Feb S.C. 39 14

1-Mar

Ala. 44 9

Ark. 22 10

Colo. 25 41

Ga. 73 29

Mass. 46 45

Minn. 31 46

Okla. 17 21

Tenn. 44 23

Texas 147 75

Vt. 0 16

Va. 62 33

5-Mar

Kan. 10 23

La. 37 14

Neb. 10 15

6-Mar Maine 8 17

8-Mar

Mich. 63 67

Miss. 31 5

15-Mar

Fla. 141 73

Ill. 79 77

Mo. 36 35

N.C. 60 47

Ohio 81 62

22-Mar

Ariz. 42 33

Idaho 5 18

Utah 6 27

chaPter one • the noMination strUGGle 41

Date State Clinton Sanders

26-Mar

Alaska 3 13

Hawaii 8 17

Wash. 27 74

5-Apr Wis. 38 48

9-Apr Wyo. 7 7

19-Apr N.Y. 139 108

26-Apr

Conn. 28 27

Del. 12 9

Md. 60 35

Pa. 106 83

R.I. 11 13

3-May Ind. 39 44

10-May W.Va. 11 18

17-May

Ky. 28 27

Ore. 25 36

5-Jun P.R. 37 23

7-Jun

Calif. 254 221

Mont. 10 11

N.J. 79 47

N.M. 18 16

N.D. 5 13

S.D. 10 10

14-Jun D.C. 16 4

Listed numbers are for bound delegates.

Source: Kevin Schaul and Samuel Granados, “The Race to the Democratic Nomination,” Washington Post, October 10, 2017, https://www.washingtonpost.com/graphics/politics/2016-election/prima ries/delegate-tracker/democratic/, accessed June 1, 2017.

Americans in Benghazi, Libya, and her use of a private e-mail account while she was secretary of state (and possible misuse of classified material). The latter, of course, con- tinued right up to election day itself. In the spring, as well as in the fall, Trump regu- larly referred to her as “Crooked Hillary,” and his audiences chanted “Lock her up! Lock her up!” Trump made a series of what we would ordinarily have been imagined

42 Part one • the 2016 Presidential election

to be candidacy-ending gaffes, but (just as Clinton’s poll numbers stayed fixed at a high level in the nomination campaign) no matter how vindictive (calling Cruz’s wife “ugly”), mean-spirited (“Little Marco”), lascivious,43 factually inaccurate, or seemingly outrageously racist his words, Trump simply marched toward victory in the spring. Or, as he put it himself, “I could stand in the middle of 5th Avenue and shoot somebody, and I wouldn’t lose voters,” Trump said in Sioux City, Iowa, January 24, 2016.44 Yet these unique features of these two candidates seemed to have had little effect on their nomination campaigns.

National Party Conventions: As we noted earlier the purpose of the state primary or caucus convention procedures is to select who will be the delegates from that state to attend their national party convention and/or to instruct those delegates on how to vote for presidential nomination. The delegates are those entrusted with voting on all the convention’s major pieces of business. These include resolving any remaining problems that arose in selecting one state’s or another’s delegations, adopting rules that will govern the party for the next four years, voting on the pro- posed party platform, and choosing the presidential and vice presidential nominees. Thus the delegates are entrusted with essentially all of the party’s major decisions. But, as we have already seen with respect to the presidential nomination, they may cast the formal ballots—and it could well be some day that they will in fact play active roles—their decision making is so tightly constrained that they almost invari- ably have no real choices to make. Their choice for presidential nominee is con- strained by the vote of the public in their state.45 The presidential nominee selects a candidate she or he would like to see serve as a running mate, and it has been a very long time since there was any real opposition to that choice.46

Party platforms once were regularly contended, as this was the one time when the party leadership could interact and work out just what the party stood for. Although this has not been true in recent years, both parties have had protests over the plat- form committee’s proposals on one issue or another (e.g., the change in the 1980 Republican platform from its long-held stance of endorsing an equal rights amend- ment to the Constitution for women to opposing it), whereas the last truly contended (nearly violently contended) battle over a platform plank was the debate over the Vietnam War in the 1968 Democratic Convention.

Instead of the traditional role of party conventions serving as the one time the party gathers from around the nation to debate and decide party business, the con- ventions have changed in recent decades to serve as major public presentations of the party to the nation. This leads the party and its leadership to seek to downplay internal divisions (although when they are really there, they are typically not able to be completely hidden) and present a united front to the public. Their other central role is to serve as the end of the intra-party competition of nominations and the transi- tion to the general election campaign. The acceptance speeches of the nominees (and certainly of the presidential nominee) are times generally used to showcase the major themes of the candidates for the general election campaign.

In 2012 the conventions were held late in August, which put the Republicans, especially, at a disadvantage as their nominee was restricted in spending in opposition

chaPter one • the noMination strUGGle 43

Table 1-4 Republican Nomination Results, 2016: Bound Delegates Won in State Primaries and Caucuses—Trump, Cruz, Rubio, and Kasich

Trump Cruz Rubio Kasich

1-Feb Iowa 7 8 7 1

9-Feb N.H. 11 3 1 4

20-Feb S.C. 50 0 0 0

23-Feb Nev. 14 6 7 1

1-Mar Ala. 36 13 1 0

Alaska 11 12 5 0

Ark. 16 15 9 0

Ga. 42 18 16 0

Mass. 22 4 8 8

Minn. 8 13 17 0

Okla. 13 15 12 0

Tenn. 33 16 9 0

Texas 48 104 3 0

Vt. 8 0 0 8

Va. 17 8 16 5

5-Mar Kan. 9 24 6 1

Ky. 17 15 7 7

La. 25 18 0 0

Maine 9 12 0 2

6-Mar P.R. 0 0 23 0

8-Mar Hawaii 11 7 1 0

Idaho 12 20 0 0

Mich. 25 17 0 17

Miss. 25 15 0 0

10-Mar V.I. 1 0 0 0

12-Mar D.C. 0 0 10 9

Wyo. 1 23 1 0

(Continued)

44 Part one • the 2016 Presidential election

Trump Cruz Rubio Kasich

15-Mar Fla. 99 0 0 0

Ill. 54 9 0 6

Mo. 37 15 0 0

N.C. 29 27 6 9

M.P. 9 0 0 0

Ohio 0 0 0 66

22-Mar Ariz. 58 0 0 0

Utah 0 40 0 0

5-Apr Wis. 6 36 0 0

9-Apr Colo. 0 30 0 0

19-Apr N.Y. 89 0 0 6

26-Apr Conn. 28 0 0 0

Del. 16 0 0 0

Md. 38 0 0 0

Pa. 17 0 0 0

R.I. 12 2 0 5

3-May Ind. 57 0 0 0

10-May Neb. 36 0 0 0

W.Va. 30 0 0 1

17-May Ore. 18 5 0 5

24-May Wash. 41 0 0 0

7-Jun Calif. 172 0 0 0

Mont. 27 0 0 0

N.J. 51 0 0 0

N.M. 24 0 0 0

S.D. 29 0 0 0

Listed numbers are for bound delegates.

Table 1-4 (Continued)

chaPter one • the noMination strUGGle 45

to Pres. Obama by the rules of the nomination season. Thus they were especially keen to hold their convention earlier in 2016. They choose to hold it in Cleveland July 18–21, whereas the Democrats held theirs in Philadelphia July 25–28.47

Trump selected the governor of Indiana, Michael Pence, to be his running mate on July 15. Pence is as understated as Trump is flamboyant and has had considerable experience in politics to balance Trump’s outsider status. He has particularly deep reli- gious beliefs, which guide many of his policy positions and, of course, appeals strongly to the large and important religious right in the party. That he hails from a combined Rust Belt, agricultural Midwestern state balanced the ticket, as is a common tradi- tion, counterbalancing a New York City, high-rolling businessman with little formal connections to religion. Whereas the Trump and Pence nominations (and adoption of the party platform) went smoothly enough, there were moments of contention. Perhaps the most obvious was Cruz’s unwillingness (often described as “defiance”) to endorse Trump’s nomination on prime-time television, which resulted in loud booing and heckling. Trump, for his part, stuck pretty closely to the script of his acceptance speech, which outlined a dark vision of contemporary America, leading those who agreed to the conclusion (he hoped) that one needed to vote for him to reverse course.

Clinton, for her part, selected Senator Tim Kaine, Virginia, as her running mate. This choice had less ticket balancing as compared to selecting a candidate from the liberal wing of the party, such as Sanders or Sen. Elizabeth Warren (both of whom had featured speeches—Warren gave the keynote address). Although perhaps not quite as similar as Sen. Al Gore was to Bill Clinton, Kaine was less about uniting the party (although he certainly did not divide it) than about trying to win the general election. Any worries about major disruption from the left wing were unfounded, and thus the convention presented a united image to the public and allowed Clinton to use her acceptance speech to complete the uniting and begin the general election campaign.

CHAPTER TWO

THE GENERAL ELECTION CAMPAIGN

Once they have been nominated, candidates choose their general election cam-paign strategies based on their perceptions of what the electorate wants, the rela- tive strengths and weaknesses of their opponents and themselves, and their chances of winning. A candidate who is convinced that he or she has a dependable lead may choose very different strategies from those used by a candidate who believes he or she is seriously behind. A candidate who believes that an opponent has significant weak- nesses is more likely to run an aggressive, attacking campaign than one who does not perceive such weaknesses.

After the 2016 conventions Hillary Clinton maintained a modest lead over Donald Trump in national polls (although many polls were within the margin of error). Most political observers thought that Clinton would win the election and that the stark differences in campaign styles for the two candidates would make a differ- ence. Chapters 4 through 8 of this book will consider in detail the impact of particular factors (including issues and evaluations of President Obama’s job performance) on the voters’ decisions. This chapter will provide an overview of the fall campaign—an account of its course and a description of the context within which strategic decisions were made.

THE STRATEGIC CONTEXT AND CANDIDATES’ CHOICES

One aspect of the strategic context that candidates must consider is the track record of the parties in recent presidential elections. In presidential races the past is certainly not entirely prologue, but it is relevant. From this perspective the picture was slightly more encouraging for the Republicans than for the Democrats. From 1952 through 2012 there had been seventeen presidential elections, and the Republicans had won ten of them. On the other hand the Democrats had won three of the last five races since 1996, and in 2000 they secured a narrow popular-vote margin despite falling short in the Electoral College.

48 Part one • the 2016 Presidential election

States won in all five elections States won in four elections

States won in fewer than four elections

N.D. 3

S.D. 3

Texas 38

Minn. 10

La. 8

Ind. 11 W.Va.

Del. 3

N.J. 14

Conn. 7

R.I. 4

Maine 4

Mass. 11

Mich. 16

Wash. 12

Mont. 3

Wyo. 3

Idaho 4

Utah 6 Colo.

9 Kan. 6

Calif. 55

Ariz. 11

N.M. 5

Nev. 6

Ore. 7

Neb. 5

Okla. 7

Iowa 6

N.M. 5

Iowa 6

Mo. 10

Ark. 6

Wis. 10

Ill. 20

Miss. 6

Ohio 18

Ala. 9

Ga. 16

Ky. 8

Tenn. 11

Fla. 29

N.C. 15

Va. 13

S.C. 9

Pa. 20

N.Y. 29

N.H. 4

Vt. 3

Md. 10

Alaska 3 Hawaii

D.C. 3

The nature of the American system for electing presidents requires that we exam- ine the state-by-state pattern of results. U.S. voters do not directly vote for president or vice president. Rather they vote for a slate of electors pledged to support a presi- dential and a vice-presidential candidate. Moreover, in every state except Maine and Nebraska, the entire slate of electors that receives the most popular votes is selected. In no state is a majority of the vote required. Since the 1972 election, Maine has used a system in which the plurality-vote winner for the entire state wins two electoral votes. In addition the plurality-vote winner in each of Maine’s two House districts receives that district’s single electoral vote. Beginning in 1992 Nebraska allocated its five electoral votes in a similar manner: the statewide plurality-vote winner gained two votes, and each of the state’s three congressional districts awarded one vote on a plurality basis.1

If larger states used the district plan employed by Maine and Nebraska, the dynamics of the campaign would be quite different. For example, candidates might target specific congressional districts and would probably campaign in all large states, regardless of how well they were doing in the statewide polls. But given the WTA rules employed in forty-eight states and the District of Columbia, candidates cannot safely ignore the pattern of past state results. A state-by-state analysis of the five presi- dential elections from 1996 through 2012 suggests that the Democrats had reason to be hopeful about the effort to win the 270 electoral votes required for victory.

Figure 2-1 States That Voted Democratic in at Least Four Out of Five Elections, 1996–2012, with Number of Electoral Votes

Source: Compiled by authors.

Chris

Text Box

<PEs; please change the legend box for "States won in all five elections" to white; also NM needs the stripes>

chaPter two • the General election caMPaiGn 49

As Figure 2-1 reveals eighteen states plus the District of Columbia voted Democratic in all five of these elections. Only fifteen states were equally loyal to the GOP. (See Chapter 3 on long-term voting patterns.) These perfectly loyal states pro- vided a prospective balance of 242 electoral votes for the Democrats to only 121 for the Republicans. Less problematic for the GOP candidates were the next groups of states. Nine states had voted Republican in every election but one, with a total of eighty-five electoral votes. Balancing these were only three states with fifteen electoral votes that had supported the Democrats in four of the five contests. Thus if each of these state’s political leanings were categorized solely on the basis of the last five elec- tions, one might expect that 257 electoral votes were likely to go to the Democrats in 2016, whereas only 206 were as likely to go to the Republicans, placing Clinton fifty-one votes ahead of Trump and only thirteen votes short of the number required to win.

If this past pattern had completely controlled the 2016 election, the GOP ticket would have been at a serious disadvantage. But, of course, things were not that simple, and many factors made Republican chances considerably better than they looked based on these numbers. Most obviously they had won two of the four previous elec- tions, and the loss in 2008 occurred in the context of the worst economic downturn since the Great Depression, for which many blamed President Bush and his party. Moreover, the economic recovery over the past few years had been modest, and many potential voters were unhappy with the president’s performance generally and his stewardship of the economy in particular.

Although most observers thought Clinton had a distinct advantage once Trump secured the Republican nomination, both campaign organizations viewed the same set of states determining the outcome of the election. These would be the “battleground” states, where both campaign organizations would concentrate the lion’s share of their time, money, and effort. Indeed, even before the beginning of 2016, the two parties had already focused their attention on a set of twelve or thirteen states, and most of the other states would be largely ignored until election day.2 The larger states in this group—particularly Florida, Michigan, Ohio, and Pennsylvania—would be the main focus of their efforts. Many of the non-battleground states, on the other hand—even large ones like California, New York, and Texas—would see little evidence that a presi- dential campaign was in progress. A state perspective through the lens of the Electoral College would dominate the strategy of the 2016 campaign.3

POLITICAL CONTEXT, OVERALL STRATEGY, AND OPENING MOVES

The strategic choices of candidates and parties are shaped by the particular context of the election. One feature of that context is whether an incumbent is running. Races without an incumbent, like the one in 2016, are very different from those where an incumbent is seeking reelection. In many respects the 2016 presidential race was one of the most bizarre and unprecedented elections in American history. Donald Trump, a political novice, became the presumptive nominee on May 3, when Ted

50 Part one • the 2016 Presidential election

Cruz withdrew from the race following his defeat in the Indiana primary. Trump later secured enough delegates to officially clinch the Republican nomination for president on May 26, 2016, after besting fourteen other candidates in the primaries. When Hillary Clinton clinched the Democratic nomination two weeks later, she made his- tory as the first woman ever to win a major party’s presidential nomination but only after a long and grueling struggle against Vermont senator Bernie Sanders. The fact that both Trump and Sanders had done so well early in the campaign reflected a mix of populist anger and rhetoric that would continue to influence the general election campaign.

Clinton had a distinct fund-raising advantage over Trump in the fall campaign. As of July 21, 2016, the Clinton campaign had raised $264.4 million compared to the $89.0 million raised by the Trump campaign. Super PACs also contributed a large sum of money to Clinton’s presidential campaign by this time, totaling nearly $122 mil- lion. By comparison the Trump campaign had only received about $5.2 million from super PACs in late July. This trend would continue throughout the fall campaign as Clinton continued to outraise and spend more money than her Republican opponent. When the electoral dust settled, Clinton would ultimately have nearly a 2:1 spending advantage over Trump in the general election campaign.4

Much of the $1.2 billion that the Clinton campaign would ultimately spend dur- ing the fall was directed toward television advertising and the get-out-the-vote opera- tion. Her campaign believed that mobilizing voters to go to the polls was key and that being able to outspend her opponent was critical in this strategy. By contrast the Trump campaign raised only about $646.8 million from individual contributors and super PACs (which was less than any major party presidential nominee since John McCain in 2008). Trump’s ability to dominate media headlines and his prolific use of Twitter in an attempt to control the narrative of the campaign were clearly factors in his ability to keep the polls close despite spending considerably less than Clinton in the fall campaign.5 One estimate suggests that he may have received as much as $2 billion worth of free media coverage from the beginning of his campaign through February 2016.6

Both candidates faced highs and lows after the primary elections. One of the Trump campaign’s biggest challenges came about a month before the Republican National Convention when Donald Trump fired Corey Lewandoski, his initial cam- paign manager. Lewandoski had been instrumental in Trump’s early wins during the primaries, but influential party officials, such as RNC Chairman Reince Priebus, routinely criticized his controversial and divisive campaign style. Lewandowski lacked experience running a national campaign, and many of his critics within the party organization felt that his strategies for fund-raising and staffing the campaign were insufficient to wage a competitive operation against Hillary Clinton, who was clearly excelling in both of these areas. “Republicans across the spectrum welcomed [Lewandoski’s] firing as a positive step, but they suggested that it needed to be fol- lowed by consistent changes in performance from the candidate himself.”7

On July 1, the Trump campaign hired Kellyanne Conway as a political advi- sor, who had previously worked as a super PAC strategist for Ted Cruz’s campaign. Conway was a veteran GOP pollster and long-time political strategist who had expe- rience working with Republican lawmakers. Trump had previously offered Conway

chaPter two • the General election caMPaiGn 51

a job in 2015, but she turned down the offer to work on behalf of Ted Cruz instead. Conway decided to join the Trump campaign in early May once Cruz dropped out of the presidential race. She was initially brought on board in her long-time capac- ity of advising Republicans on how to better appeal to female voters.8 On August 17, 2016, Conway was named as Donald Trump’s campaign manager, making her the first woman to ever be selected to run a Republican presidential campaign. On the same day Steve Bannon, former Breitbart news executive, was appointed as the CEO of Trump’s campaign. Bannon has a long history of populist rhetoric, including criticiz- ing the establishment wing of the Republican Party, but is perhaps best known for his tireless crusade against the Clintons.9

Amid the controversy and shake-ups within the Trump campaign, Clinton was dealing with her own challenges stemming from a year-long FBI investigation. On March 2, 2015, the New York Times reported that Clinton had used a private email server when she was serving as secretary of state during the Obama administration.10 Clinton defended her use of the private server several days later, claiming it was for convenience, so she could use one device for both business and personal use. On August 11, 2015, the Clinton campaign revealed that her server had been turned over to the Justice Department as per their request. More than nine months passed before the State Department’s inspector general issued a report highly critical of Clinton’s use of a private email server saying that “she had not sought permission to use it and would not have received it if she had.” The report, which had been issued to Congress, undermined some of Clinton’s previous statements and gave new political fodder to Republicans and the Trump campaign.11

Based largely on the inspector general’s report, Hillary Clinton was interviewed by the FBI in Washington, DC, for more than three hours on July 2, 2016, about her previous use of the private email server. Three days later FBI Director James Comey, who had been at the center of the investigation since 2015, announced that the FBI would not recommend that Clinton be indicted on any charges but called her behavior “extremely careless” with respect to how she handled classified information during her time as secretary of state.12 Clinton and her Democratic allies praised the decision by Comey, whereas Trump and his followers were highly critical of the outcome, refer- ring to it as a “total miscarriage of justice.”13 Many Democrats hoped this would put an end to the discussion of emails and private servers, but the topic continued to remain relevant throughout the campaign and would resurface in a big way later on in the fall.

Selecting the Vice Presidents

Most political scientists agree that the presidential candidate’s choice of running mates has little or no effect on the outcome of the election because most voters are inclined to vote primarily on the basis of which candidates are at the top of the ticket (and typically along party lines). Nevertheless, the selection of a candidate’s vice presi- dential nominee often receives considerable attention by the media because it invites potential speculation about geographical or ideological balancing on the ticket.14 The process also tends to occur a few days before the national conventions during the summer, when news coming out of the campaigns is far less regular and the media are looking for anything newsworthy to report upon.

52 Part one • the 2016 Presidential election

In the days leading up to Donald Trump’s choice of a vice presidential candidate, there was considerable speculation about who he would select for the position. Among some of the possible contenders discussed early in the summer were Ohio Governor John Kasich, Oklahoma Governor Mary Fallin, New Jersey Governor Chris Christie, and former U.S. house speaker from Georgia, Newt Gingrich.15 Although there were potential advantages and disadvantages with respect to each of these picks, many within the establishment wing of the party felt that most of these candidates would be in a good position to mitigate some of the more controversial aspects of the Trump candidacy. At the same time the “right” candidate for vice president might help ensure that Republican voters would not stay home on election day, a worry that had con- cerned party loyalists ever since Trump had sailed through the Republican primaries.

Just prior to the Republican National Convention, Trump seemed to narrow his choice down to one of three potential vice presidential nominees—Christie, Gingrich, and Indiana Governor Mike Pence. Pence had not been part of the initial discussion of potential nominees earlier in the summer, but he had gained some traction after a series of meetings with Trump, where the two seemed to hit it off. Trump had planned to wait until the convention to announce that Pence was joining the ticket but ended up making the official announcement the Friday before the convention started. Many establishment Republicans were thrilled with the selection of Pence because he was viewed as a “safe” choice who had been a consistent voice of conservative orthodoxy since his early days in talk radio. Pence’s selection was also viewed by many as an attempt to shore up support, especially among social conservatives, who remained somewhat skeptical of Trump as the Republican candidate for president and perhaps an opportunity to win over voters in many of the Rust Belt states. According to one initial report, “In tapping Pence, Trump adds to the GOP ticket a politician with ties to the Koch brothers and other influential donors who have so far stayed away from Trump.”16 Pence’s prior experience in office may have also helped to appease those Republicans who worried about Trump’s lack of executive experience.

On the Democratic side Hillary Clinton waited until the Republican conven- tion ended before naming her own choice for vice president. In the weeks leading up to the conventions, five different individuals had regularly been discussed as poten- tial Clinton nominees for vice president—Virginia senator Tim Kaine, Ohio senator Sherrod Brown, New Jersey senator Cory Booker, Agriculture Secretary Tom Vilsack, and Massachusetts senator Elizabeth Warren. Some of these individuals, including Vilsack, were viewed as a safe choice by the Clinton campaign, whereas others—like Kaine and Brown—were considered at or near the top of the list given the strategic advantage that might come from having someone from states like Ohio and Virginia, which were very much in play in the presidential election.17

In the end Clinton went with the individual that potentially could offer her the greatest strategic advantage on the Democratic ticket, especially in carrying the state of Virginia—Senator Tim Kaine. Other than Vilsack, Kaine was easily among the safest choices for Clinton, and most within the Democratic Party readily acknowl- edged this. Given his strong ties to the party and his past electoral success in the state of Virginia, Kaine’s selection as Clinton’s nominee for vice president made a lot of strategic sense. Additionally his fluency in Spanish was definitely a plus in Clinton’s attempt to reach out to Hispanic voters.18 Interestingly enough Bernie Sanders spoke

chaPter two • the General election caMPaiGn 53

out the day after the announcement indicating that he would have preferred seeing Clinton select Elizabeth Warren as her running mate rather than Kaine. His com- ments seemed to reflect others’ views from within the left wing of the party who were clearly disappointed by Clinton’s selection.19

FROM THE CONVENTIONS TO THE DEBATES

The Conventions

The Republicans held their convention in Cleveland, Ohio, from July 18 to 21, amid potential protests and turmoil both inside and outside the convention hall. Many worried that the protests outside the convention would be large and unruly but ulti- mately proved to be much smaller than anticipated.20 On the first day of the con- vention, tensions within the Republican Party became apparent when officials were forced to adopt the rules of the proceedings by a loud voice vote. Anti-Trump forces within the chamber sought to further derail the proceedings, but things quickly set- tled down, and no further disruptions were immediately apparent. Perhaps the most conspicuous event at the convention was the sheer number of notable Republicans who willingly chose not to attend, including George W. Bush and his brother Jeb along with John Kasich and John McCain. “From the party’s former presidents to the host state governor, many leaders were staying away from the convention stage, or Cleveland altogether, wary of being linked to a man whose proposals and tempera- ment have sparked an identity crisis within the GOP.”21

This potential identity crisis was further illustrated when Ted Cruz spoke on the third night (July 20) of the convention. Cruz had been an outspoken critic of Donald Trump during the presidential primaries, and many within the party wondered if he would be willing to set aside his differences with the presumptive nominee and pub- licly endorse him for president. Much of his rhetoric echoed his comments from the campaign, especially when it came to issues such as health care, immigration, religious freedom, and taxes. Near the end of his convention speech, however, he stopped short of endorsing Trump when he offered the following remarks: “If you love our country, and love our children as much as you do, stand, and speak, and vote your conscience, vote for candidates up and down the ticket who you trust to defend our freedom, and to be faithful to the Constitution.”22 Many within the party viewed the “vote your conscience” comment as an implicit rejection of Trump as the nominee, which served to further highlight the divisions within the Republican Party rather than signify unity as is typically the case at national conventions.23

Trump’s acceptance speech the following night was a mix of fiery rhetoric and partisan imagery that marked a significant departure from the optimistic tone char- acterizing Republican convention speeches dating back to the presidency of Ronald Reagan. Early in his speech he remarked, “Our convention occurs at a moment of crisis for our nation. The attacks on our police, and the terrorism in our cities, threaten our very way of life. Any politician who does not grasp this danger is not fit to lead our country.” Trump reiterated that he would be the voice of American citizens who felt that they had been left behind over the past eight years. He also said he would take the

54 Part one • the 2016 Presidential election

necessary steps to keep America safe and bring back prosperity to the nation.24 Finally Trump emphasized that he would present the facts “plainly and honestly. . . . So if you want to hear the corporate spin, the carefully crafted lies and the media myths, the Democrats are holding their convention next week. Go there.”25

The Democratic National Convention met in Philadelphia, Pennsylvania, the following week. Just prior to the convention, however, WikiLeaks released a large number of hacked emails from the Democratic National Committee suggesting that DNC chairwoman Debbie Wasserman Schultz may have taken steps to help Hillary Clinton win the Democratic primary. These allegations led to significant fallout among Democrats, especially those who had supported Bernie Sanders in the pri- maries, as some of the emails made several disparaging remarks about Sanders and his campaign manager, Jeff Weaver. The outrage over the emails was enough to force Wasserman Schultz to announce she would be stepping down after the Democratic National Convention in light of her involvement in the scandal.26 The timing of the email leaks did little to support the Democrats’ message of unity following the chaos and discord at the Republican convention the week before.

Once the Democratic National Convention got under way on July 25, it was clear that the rhetoric would be significantly different from the message at the Republican convention the previous week. The speakers, which included a number of promi- nent Democrats such as former President Bill Clinton, vice presidential nominee Tim Kaine, and President Barack Obama made a strong case for why the country needed to select Hillary Clinton to be the next president. Many viewers agreed that one of the most memorable and impassioned speeches was given by First Lady Michelle Obama on the third night of the convention. In her remarks she mentioned that “this election and every election is about who will have the power to shape our children for the next four or eight years of their lives.” She continued by adding, “I am here tonight because in this election there is only one person who I trust with that responsibility, only one person who I believe is truly qualified to be president of the United States, and that is our friend Hillary Clinton.”27

On the following evening Hillary Clinton made history as she became the first woman to accept the nomination for president of the United States. Her speech was a sharp contrast to the one given by Donald Trump the week before. Clinton empha- sized the need for unity as well as putting aside our differences: “We have to decide whether we all will work together so we all can rise together.” She also contrasted her message from that of her Republican opponent, who she mentioned was try- ing to divide the nation for political gain. In her own words, she stated, “That’s why ‘Stronger Together’ is not just a lesson from our history. It’s not just a slogan for our campaign. It’s a guiding principle for the country we’ve always been and the future we’re going to build.”28 According to the Nielsen ratings, approximately 33.7 mil- lion people tuned in to watch Clinton’s convention speech, which was just shy of the 34.9 million who had watched Trump’s speech the week before.29

It is often the case that nominating conventions provide a short-term boost in the polls to the candidate of the party holding them. After all, the party and its candi- date receive a lot of attention, and they largely control what is seen and heard during the convention. In 2016, with the conventions so close together, it is difficult to be sure of the effects, but data from a CNN/ORC poll conducted immediately after the

chaPter two • the General election caMPaiGn 55

convention gave Trump a bounce of six points, three points ahead of Clinton in a head-to-head matchup.30 Shortly after the Democratic National Convention wrapped up on July 28, an NBC News poll had Clinton leading Trump by eight points (50 to 42 percent) after a narrow lead several days prior following the conclusion of the Republican convention.31 Clinton continued to maintain a small but modest lead over Trump during the month of August, when he received a growing amount of negative news coverage.

A series of news stories over the next four weeks would lead to a disastrous August for the Trump campaign. The downward spiral began on the third night of the Democratic convention when Trump tweeted, “If Russia or any other country or person has Hillary Clinton’s 33,000 illegally deleted emails, perhaps they should share them with the FBI!” The campaign was forced to deny that Trump had “encouraged” a foreign nation to interfere with the U.S. election, a concern that would continue to hound them for the foreseeable future. The following evening the parents of Army Captain Humayun Khan, who had been killed in 2004 by an Iraqi suicide bomber, spoke at the Democratic convention and questioned whether Trump had even read the U.S. Constitution before pulling out a pocket-size version and stating, “I will gladly lend you my copy.”32 Trump could not resist the temptation to respond, and the situation continued to escalate over the next few days with most of the media criticism clearly directed at Trump’s mishandling of the situation.

Following campaign rallies in both Ohio and Pennsylvania, Trump then sug- gested that the election might be “rigged” against him, a charge that he would bring up again later in the campaign. He also referred to Clinton at one of the rallies as “the devil” before suggesting that a crying baby be removed from a campaign event (to which he responded later that he was only kidding).33 Then, in an August 2 interview with the Washington Post, Trump implied that he was not yet ready to endorse House Speaker Paul Ryan or Senator John McCain in their GOP primaries, which only served to further illustrate the deep divisions within the Republican Party.34 During a press conference following Trump’s remarks, President Obama declared Trump “unfit to serve as president” and “woefully unprepared to do this job.” He also challenged Republican leaders to withdraw their support for their nominee in light of the latest series of critical news stories.35

Two days later a new series of polls showed that Clinton had a growing lead over Trump in states such as Florida, Michigan, New Hampshire, and Pennsylvania. Then, on August 5, a new Atlanta-Journal Constitution poll showed that Clinton was leading Trump in Georgia, a state that had supported every Republican presiden- tial candidate since Bill Clinton in 1992.36 Republicans were growing increasingly nervous at this point about Trump’s chances of winning, but he showed little or no signs of greater discipline on the campaign trail. The following week a new CNN poll released on August 8 suggested that Clinton had a ten-point national lead over Trump, which further cemented mounting concerns within the Republican Party over Trump’s candidacy.

Trump did little to assuage his critics when he spoke at a campaign rally in Wilmington, North Carolina, the following day. During the speech, he remarked, “Hillary wants to abolish—essentially abolish the Second Amendment. By the way, if she gets to pick, if she gets to pick her judges, nothing you can do, folks.” After

56 Part one • the 2016 Presidential election

the crowd started booing, Trump added, “Although the Second Amendment people, maybe there is. I don’t know.”37 Not surprisingly, Trump’s remark set off a firestorm of protests among Democrats who quickly denounced suggestions of violence against Hillary Clinton or liberal judges. It also attracted the attention of the Secret Service, who takes any type of threat against presidential candidates very seriously. Clinton’s running mate, Senator Tim Kaine, said that Trump’s statement offers “a window into the soul of a person who is just temperamentally not suited to the task.”38 Even Republicans remarked that Trump’s comment was a very poor joke at best.

Still reeling from his comment about the “Second Amendment people,” Trump continued to push ahead by claiming that Obama and Clinton were the “cofounders” of ISIS and that any fallout was clearly their fault. It took several days before Trump backed off of this provocative claim and began referring to it as “sarcasm.” At the same time he continued to mention that the election was “rigged” and that the media was completely biased against him.39 When asked if he might adjust his approach during a CNBC interview in light of poll numbers that continued to trend downward, Trump rejected this suggestion and said that his plan was to “[j]ust keep doing the same thing I’m doing right now.” He went on to add that “[a]nd at the end, it’s either going to work or I’m going to, you know—I’m going to have a very, very nice, long vacation.”40

Just days after insinuating that he could actually lose the election, Trump announced the most serious shake-up in his campaign staff since Lewandowski was fired as campaign manager in June. As noted earlier in the chapter, Trump brought Steve Bannon onboard as the new CEO of his campaign, which many viewed as a strong signal that Trump wanted to continue to pursue a more aggressive style. Bannon had previously served as the executive chairman of Breitbart News, which had a reputation for hard-hitting conservative news. He also promoted senior adviser Kellyanne Conway to campaign manager. Whether or not these two actions would help to right the ship were not yet clear, but many within the Republican Party believed that it could hardly make things worse at this point in the campaign.41

Although August proved to be problematic for the Trump campaign, Clinton had to deal with two events in early September that generated much more media attention than her campaign probably would have expected. At a New York City fund-raising event held on September 9, Clinton remarked, “You know, just to be grossly generalis- tic, you could put half of Trump’s supporters into what I call the basket of deplorables. Right? They’re racist, sexist, homophobic, xenophobic, Islamophobic—you name it. And unfortunately there are people like that. And he has lifted them up.” The follow- ing day Trump responded to Clinton’s statement by tweeting, “Wow, Hillary Clinton was SO INSULTING to my supporters, millions of amazing, hard working people. I think it will cost her at the Polls!42 Clinton’s “basket of deplorables” comment may have had a deleterious effect on the polls as the Trump campaign repeatedly used a clip of this statement in campaign ads throughout the fall.

At an event two days later commemorating the victims of the September 11 ter- rorist attacks, Clinton was recorded nearly collapsing afterward and was helped into her van by Secret Service agents. Two hours later she emerged from her daughter Chelsea’s Manhattan apartment looking refreshed and ready to return to the cam- paign. Later that day her doctor reported that she was being treated for pneumonia and dehydration. This event in particular renewed concerns about whether Clinton

chaPter two • the General election caMPaiGn 57

was healthy enough to serve in office. “The incident, which occurred after months of questions about her health from her Republican opponent, Donald J. Trump, and his campaign, is likely to increase pressure on Mrs. Clinton to address the issue and release detailed medical records, which she has so far declined to do.”43 This issue would continue to come up during the next few weeks as questions of Clinton’s health and “stamina” would be raised by the Trump campaign.

The Debates

In the days leading up to the first presidential debate, the presidential race had tightened up considerably, with Clinton clinging to a narrow two-point lead. Both candidates saw an opportunity for turning their polls around during the debate held at Hofstra University. In preparation for the first debate, Clinton maintained a lighter campaign schedule so that she could review prior debate performances by Trump and brush up on the important issues that would likely come up in the debate. “For her prep sessions, Clinton has reportedly surrounded herself with a team of strategists, including Ron Klain, former chief of staff to Vice President Joe Biden, and Karen Dunn, a Clinton adviser and former White House aide.”44 She also sought a variety of media and communication specialists such as Jim Margolis, Mandy Grunwald, and campaign chairman John Podesta, who could help her focus her message during the debate.45

In contrast to Clinton, Trump seemed to favor a different style of preparing for the first debate. When asked about his strategies for debating Clinton, Trump often seemed to rely on his comfort and past exposure to the media spotlight. In one of the more revealing statements about his debate preparation, Trump mentioned, “Obviously I will be practicing, but I don’t want to put so much practice in that all of a sudden, you’re not who you are.”46 In a story reported by National Public Radio on the debate, it was suggested, “One way Trump may have been preparing already is by giv- ing a number of scripted speeches on the campaign trail in recent weeks, as opposed to the freewheeling style he adopted at rallies through most of the campaign. Those speeches have allowed Trump to practice delivering more measured and detailed ver- sions of his pitch to voters.”47

During the actual debate a number of topics were addressed ranging from jobs, gender, national security, and taxes. Early on in the debate, when Trump was asked to clarify a remark he had made that Clinton lacked a certain presidential “look,” he responded by saying, “She doesn’t have the stamina. To be president of this country, you need tremendous stamina. You have to be able to negotiate.” In response to this specific critique, Clinton remarked, “As soon as he travels to 112 countries and negoti- ates a peace deal, a cease fire, a release of dissidents, an opening of new opportunities in nations around the world, or even spends 11 hours testifying in front of a congres- sional committee, he can talk to me about stamina.” To this she added, “He tried to change from looks to stamina, but this is a man who has called women pigs, slobs and dogs.”48

Throughout the course of the debate, both candidates managed to get in a few memorable one-liners when answering questions and discussing their specific policy positions. When asked if he would release his tax returns, Trump repeated the refrain

58 Part one • the 2016 Presidential election

that he was facing a routine audit that precluded him from distributing this informa- tion. The moderator challenged this assertion, to which Trump responded by saying that he would go against his lawyers’ wishes and release his returns if Clinton agreed to release approximately 33,000 emails that had been deleted from her private server.49 Later on in the debate, when discussing his plans to lower taxes on wealthy Americans while also mixing in additional critiques of Clinton, she remarked, “I have a feeling that by the end of this evening, I’m going to be blamed for everything that’s ever hap- pened,” to which Trump replied, “Why not?”50

The audience for the first presidential debate was substantial. In total approxi- mately 84 million viewers tuned into the thirteen stations that carried the debate live, making it the most watched presidential debate in American history. In reality the viewership was even higher because many individuals ended up watching the debate online.51 By the next morning the media consensus was that Clinton was the winner of the first debate. According to a Gallup poll released later in the week, “Americans who saw the debate believed Clinton was the runaway victor, 61 percent to Trump’s 27 percent.”52 Clinton also seemed to dominate Trump on several dimensions. “On all the characteristics Gallup asked voters about, Clinton prevailed over Trump: she appeared to be more inspiring (46 percent to Trump’s 34 percent), was more likable (55 to 36 percent), appeared presidential (59 percent to 27 percent), and exhibited a better understanding of the issues (62 to 26 percent).”53

Next in the debate sequence was the vice-presidential debate held on October 4. Not surprisingly, it received far less attention than the presidential events. Trump’s running mate, Mike Pence of Indiana, sought to soften his image and “put a calmer, gentler face on the 2016 Republican ticket.”54 Senator Tim Kaine, in contrast, used the forum to remind voters about all of the negative comments Trump had made on the campaign trail and asked Pence to explain them. As a former talk-radio host, Pence was able to deflect most of Kaine’s criticisms, and a CNN/ORC poll that came out days later suggested that 48 percent of those who tuned into the debate thought Pence had won compared with only 42 percent saying Kaine won.55 In the end, however, it did little to move poll numbers for the two presidential candidates, as is typically the case.56

Just days before the second presidential debate, a shocking news story broke that turned out to be the first “October surprise” in the campaign. “Trump’s campaign was sent reeling on Friday after a private tape was published in which the reality TV star bragged about groping, kissing and attempting to have sex with married women—and said he was entitled to do so because he’s a ‘star.’ ”57 The Washington Post first reported the story, indicating that the video was from a 2005 Access Hollywood interview that host Billy Bush conducted with Trump.58 Following the release of the shocking video, described by some as the worst October surprise that any campaign has ever suf- fered, there were immediate calls for Trump to step down as the Republican nominee. Numerous insiders expressed concerns that they did not know how their candidate could survive this.59

In response to the firestorm of criticism that erupted, Trump issued a rare apol- ogy late Friday evening. “I never said that I’m a perfect person nor pretended to be someone I’m not. I’ve said and done things I regret and the words released today on this decade old video are one of them. Anyone who knows me knows these words don’t reflect who I am. I said it, I was wrong, and I apologize.” Had Trump stopped

chaPter two • the General election caMPaiGn 59

there, the fallout might have been very different. But Trump continued with the fol- lowing controversial remarks, “I’ve said some foolish things but there’s a big differ- ence between the words and actions of other people. Bill Clinton has actually abused women and Hillary has bullied, attacked, shamed, and intimidated his victims.”60

Not surprisingly these extra remarks set off additional protests that did little to silence critics of Donald Trump in the days ahead. In fact as many as nine differ- ent women came forward following the release of the tape claiming that Trump had assaulted them at some point in the past.61 Although some within the Trump campaign questioned the timing of these charges and accusations, they forced Trump and his team to immediately go on the defensive. Trump responded to the growing criti- cisms by tweeting, “The media and establishment want me out of the race so badly—I WILL NEVER DROP OUT OF THE RACE, WILL NEVER LET MY SUP- PORTERS DOWN.”62

Two days later the second presidential debate took place in St. Louis, Missouri. Still stinging from news stories surrounding the tape, Trump elected to go on the offensive by inviting several women who had made sexual harassment accusations in the past against Bill Clinton to attend the debate and sit in the front row. The subject of the 2005 tape came up early on in the debate, and Trump continued to downplay his comments—“It’s just words, folks”—whereas Clinton used his remarks to remind the audience how little respect he had for women. She also added, “He never apologizes for anything to anyone.”63 Later on in the debate, when Clinton stated it was a good thing Trump is not in charge of the law, Trump responded, “Because you’d be in jail.”64 In the aftermath of the debate, the public’s view of the results were much closer than after the first debate, with respondents to an NBC poll choosing Clinton as the winner by 44 to 34 percent.

The final presidential debate, on October 19, was supposed to focus on enti- tlements and the debt, immigration, the economy, and the Supreme Court, but the discussion quickly veered off into Russian influence in the campaign, whether the election was rigged, and the candidates’ fitness to be president. At one point, while dis- cussing Russia’s role in the presidential election, Clinton remarked, “Russia is trying to influence the election. Putin would prefer to have a puppet,” to which Trump imme- diately responded, “You’re the puppet.” Later on in the debate, when asked repeatedly by the moderator whether Donald Trump would accept the results of the election, he simply responded by saying, “I will look at it at the time. I will tell you at the time. I will keep you in suspense.”65 Finally, while Clinton was answering a question about social security taxes at the end of the debate and remarking that her “social security payroll contributions will go up as will Donald’s assuming he can’t figure out how to get out of it,” Trump leaned into the microphone and stated, “Such a nasty woman.” This comment served to undermine his earlier remarks about respecting women and became a rallying cry for women in the waning days of the campaign.66 According to a CNN/ORC poll, Clinton was the winner of the debate by a thirteen-point margin, giving her a clean sweep over Trump in all three presidential debates.67

The consensus among political scientists is that presidential debates usually do not have a significant impact on a race.68 The most prominent explanation is that by the time the debates occur, the vast majority of voters have made up their minds and are thus unlikely to have their position reversed by the event. There are, however, a

60 Part one • the 2016 Presidential election

few exceptions where some analysts perceive a greater impact. These include 1960 (Kennedy vs. Nixon), 1976 (Ford vs. Carter), and 1984 (Reagan vs. Mondale), and 2012 (Obama vs. Romney).69 Robert Erikson and Christopher Wlezien took a sys- tematic look at the ten presidential elections with debates (1960 and 1976–2008), comparing the poll standings of candidates before and after the debates. They found that with one exception, the pre-debate polls were closely matched by the post-debate polls (the exception was 1976, when Carter was already in decline before the debates and the decline persisted). They conclude that debates do not have as great an impact as the conventions (the effect of which they find to be substantial) but that they may have as much or more effect than other campaign events.70

It appears, however, that 2016 may be another exception. Data from Real Clear Politics (which averages results for all major polls over a time interval) shows that on September 26, the day of the first debate, Clinton had a 2.3 percent lead in the poll averages, while on October 19 (the last debate’s date), Clinton led by 6.5 percent.71 That amounted to a four-point swing in favor of Clinton during the course of the debates, which initially suggested that Trump might have a difficult time turning the tide around before the election.

THE END GAME AND THE STRUGGLE OVER TURNOUT

The Final Two Weeks

In the days following the third debate, both candidates continued to campaign in swing states around the country, repeating their respective messages about how each of them offered a better alternative for the country than their opponent. Clinton seemed to capitalize immensely on her final debate performance as a ABC News tracking poll that came out a few days following the third debate showed her leading Trump 50 percent to 38 percent in a four-way race (with Libertarian Party nominee Gary Johnson earning 5 percent and Green Party nominee Jill Stein earning about 2 percent in the poll). The poll reflected the highest level of public support for Clinton of any survey taken by ABC News during the fall and the lowest for Trump.72 Some analysts interpreted this new poll as evidence that Clinton was making inroads among voters that had traditionally supported Trump throughout the fall campaign.73

The good news for the Clinton camp ended up being short-lived, however, as potentially unsettling news broke just eleven days before the election. “The FBI on Friday dropped a bombshell on Hillary Clinton’s campaign less than two weeks before Election Day, with Director James Comey telling lawmakers that the agency is reviewing new evidence in its investigation into her use of a private email server as secretary of state.”74 The new evidence was prompted by an investigation into an unrelated case where the FBI discovered emails that could be relevant to Clinton’s case. Although Comey did not elaborate on where the emails originated from, it was later learned that they were discovered “out of the probe into (former Congressman) Anthony Weiner, the estranged husband of Clinton aide Huma Abedin who is under investigation for allegedly sexting a teenage girl.”75 According to a New York Times

chaPter two • the General election caMPaiGn 61

story, the FBI seized several devices from Weiner as part of their ongoing probe and discovered the additional emails.76

News of the renewed FBI investigation into Clinton’s emails broke as she was on the way to a campaign event in Iowa, where she waved off reporter’s questions as she was getting off the plane. “Clinton’s campaign initially appeared blindsided by the development.”77 Nevertheless, it did not take long before her surrogates began hitting back about the news, especially given the close proximity to the November election. Clinton’s campaign chairman, John Podesta, referred to the timing of the announce- ment as “extraordinary.” In an interview on MSNBC, Brian Fallon, Clinton’s national press secretary said, “It boggles the mind why this step was taken today. As it is, we now have the worst of all worlds here. He aired this in an extraordinary step and the public doesn’t have any way to judge the significance of this and what it has to do with Clinton.”78 Tim Kaine, Clinton’s running mate, also mentioned that he found the FBI’s handling of the renewed investigation as “very, very troubling” and that the FBI “should give a clear accounting of what’s going on right now.”79

While the Clinton team was scrambling to deal with the late-breaking October surprise, Trump took full advantage of the situation when the news broke. During a campaign rally in New Hampshire, Trump led off by saying, “Hillary Clinton’s cor- ruption is on a scale that we have never seen before. We must not let her take her criminal scheme into the Oval Office. I have great respect for the fact that the FBI and the Department of Justice are now willing to have the courage to right the horrible mistake that they made.”80 Trump’s comments were met with cheers of “lock her up,” a now common refrain at his campaign rallies. He continued his remarks by adding, “This was a grave miscarriage of justice that the American people fully understood, and it is everybody’s hope that it is about to be corrected.”81

In the days following Comey’s announcement, the polls for the presidential race began to tighten up significantly. What started out as a five-point advantage for Clinton on the day the story broke quickly turned into a narrow one-point lead over Trump according to polling trends on Real Clear Politics.82 Clinton continued to try to downplay the story in the days before the election, whereas the Trump campaign repeatedly emphasized it at campaign rallies around the country. When FBI Director Comey announced on November 6, two days before the election, that his agency’s review of the new emails did not change his previous conclusion that Clinton should not be prosecuted for her behavior, many on both sides wondered if the damage had already been done.83

Mobilizing the Vote

In 2004 the Republicans had a distinct advantage with respect to mobilizing and turning out voters on election day. The Democrats significantly reversed this trend in 2008, leading to the highest turnout in a presidential election since 1968.84 Four years later the Democrats’ voter identification and mobilization efforts took a big leap forward in terms of both technology and effort.85 When Jim Messina took on the job of Obama’s campaign manager, he said: “We are going to measure every single thing in this campaign.”86 Messina “hired an analytics department five times the size that of the 2008 campaign.”87 These analysts believed that the product of their efforts—their

62 Part one • the 2016 Presidential election

data—was the principal advantage President Obama had over his opponent, and they guarded it diligently.

In 2016 both parties continued using new data analytic techniques to mobilize voters, but this strategy has become increasingly difficult as the country has become more polarized. In the past campaigns sought to persuade undecideds to go to the polls, while they simultaneously reached out to loyal partisans to vote. More recently, however, candidates and parties have had to adapt to a new reality in light of the decreasing number of “swing voters.” “The goal is now to mobilize the most loyal voters rather than lure in the undecided or persuade the other party’s voters to change sides.”88 One direct consequence of this change is that an increasing number of voters tend to be stronger partisans who are less likely to defect than independents or those who only weakly lean toward one party in particular.

The shift away from persuasion to one of increasing turnout among loyalists had vastly different effects across the two campaigns. “For the Trump campaign, this mobilization strategy meant trying hard to inspire disenchanted working-class whites in heavily Republican areas, giving them something to get excited about in Donald Trump’s anti-establishment white identity politics.”89 Trump did not attempt to reach out to voters who were not already disposed to vote for him as his campaign recog- nized the futility of such effort. Clinton, on the other hand, sought to repeat the suc- cess of the Obama campaign in 2008 and 2012 in getting voters to turn out on election day. “For Democrats, this meant investing in large-scale get-out-the-vote operation on the premise that the ‘Obama coalition’ of black and Hispanic voters and young educated whites could be sustained by enough field offices and data analytics.”90

Whereas the Trump campaign was relying largely on the support of white voters to win the election, Clinton and her surrogates recognized the value of reaching out to a broad coalition of ethnic voters, including Hispanics. Even before Trump emerged as the Republican nominee, the Democrats had been laying the groundwork to attract new Hispanic voters by spending millions of dollars in outreach efforts. Clinton also saw an advantage in carefully positioning herself on immigration, especially when it came to the issue of how to deal with undocumented immigrants. “From the begin- ning of her campaign, Mrs. Clinton and her team saw untapped potential in the 27 million Hispanics who would be eligible to vote in 2016, a 26 percent increase since 2012.”91 Although turnout among Hispanics is traditionally lower than among other groups, the task was perhaps easier in 2016 in light of Trump’s repeated asser- tions about illegal immigration as well as his description of immigrants as “murderers and rapists.” Indeed her efforts appeared to pay off during the initial weeks of early voting. “Energized by anger at Mr. Trump and an aggressive Democratic campaign to get them to the polls, Latinos are turning out in record numbers and could make the difference in the outcome in several highly contested states.”92

One can see the differences in the candidates’ campaign strategies as reflected by their placement of presidential campaign field offices during 2016 as shown in Figure 2-2 (data from 2012 is included for comparison purposes). For instance Clinton had 511 field offices across the country in forty-nine different states with Wisconsin being the only exception. In twenty-seven of these forty-nine states, she had only one field office, presumably because these states were not considered competitive elector- ally. In contrast Trump had only 145 field offices across the country in eighteen total

chaPter two • the General election caMPaiGn 63

states. In five states he had only one office, and in thirty-two he had none. Although the pattern for the placement of Trump’s field offices is less clear, it appears that many were located in the Midwest or mid-Atlantic region, where the Trump campaign believed those states were in play.93 Overall the 2016 placements reflect a significant decline from the 2012 campaign, in which both candidates had more field offices— Obama had a total of 755 offices across the country, whereas Romney had 283, nearly twice as many as Trump in 2016.

As noted earlier Clinton had a distinct fund-raising advantage over Trump in the general election, which continued into the final month of the presidential campaign. This, in turn, led many to believe that Clinton would have an edge in getting voters to the polls. During the third quarter of 2016, Clinton’s joint fund-raising with the Democratic National Committee yielded slightly more than $261 million based on Federal Election Commission reports. “By comparison, the equivalent Trump Victory Fund, which coordinates with the Republican National Committee and state parties, raised less than one-quarter of Clinton’s haul—$61 million.”94 Nevertheless, Trump did receive a tremendous amount of free media coverage over the course of the pri- mary and general election campaign as a result of his “celebrity” status. “According to data-driven analytics firm mediaQuant, Trump received around $5 billion in free media coverage, more than twice that garnered by Clinton.”95 How that much free coverage from the media would translate into increased voter turnout was unclear in the days leading up to the election.

The Final Days

As noted the Real Clear Politics average of polls showed Clinton with slightly more than a five-point lead over Trump two weeks before the election. Once news broke of the FBI’s renewed interest in Clinton’s emails, however, the presidential race immediately began to tighten up. By November 3, Clinton’s lead over Trump had been reduced to 1.3 percent. Then, over the next few days, Clinton began to open up a small lead, and on election day, she led by a margin of 3.2 points in the poll average. It was hardly a safe cushion, and the outcome remained in doubt, but many Democrats felt a growing confidence that Clinton would eventually win. The fact that the race had tightened up considerably gave both candidates reason to be optimistic, and the campaigns launched the final effort to appeal to voters in the last days before the election.

As is often the case, presidential candidates maintain a rigorous schedule dur- ing the last weekend before an election in an attempt to deliver their personal mes- sage to voters one last time. “Hillary Clinton and Mr. Trump used the final Saturday before Election Day to make their closing pitches to voters, with Mrs. Clinton in South Florida and Philadelphia and Mr. Trump dashing to four states across three time zones—the sort of barnstorming tours presidential candidates have traditionally made in the last 72 hours before Election Day.”96 Both candidates and their surrogates also spent a considerable amount of time in New Hampshire just prior to the election given its status as an important swing state.97

During the closing days of the campaign, Clinton avoided discussing her emails as well as Comey’s decision to reopen the investigation so close to election day. Instead

Fi gu

re 2

-2

Lo ca

ti on

o f

Pr es

id en

ti al

C am

p ai

gn F

ie ld

O ffi

ce s,

2 01

2 an

d 2

01 6

So ur

ce : F

ie ld

o ffi

ce in

fo rm

at io

n a

n d

a d

d re

ss es

w er

e ga

th er

ed b

y th

e au

th or

s fr

om t

h e

of fi

ci al

c am

p ai

gn w

eb si

te s

on t

h e

w ee

ke n

d s

be fo

re t

h e

re sp

ec ti

ve e

le c-

ti on

d ay

s. T

h e

au th

or s

u se

d G

oo gl

e M

ap s

to id

en ti

fy t

h e

lo n

gi tu

d e

an d

la ti

tu d

e co

or d

in at

es f

or e

ac h

o ffi

ce .

T ru

m p

F ie

ld O

ffi ce

s C

lin to

n F

ie ld

O ffi

ce s

R o

m ne

y F

ie ld

O ffi

ce s

O b

am a

F ie

ld O

ffi ce

20 12

P re

si d

en ti

al C

am p

ai g

20 16

P re

si d

en ti

al C

am p

ai g

chaPter two • the General election caMPaiGn 65

she spent most of her time on the campaign trail talking about salient issues such as raising the minimum wage, providing affordable child care, and tuition-free college education for the middle class. She also sought to reassure voters that in light of the harsh rhetoric used throughout the fall campaign, it was important for the next presi- dent to find a way to bridge the growing divide in this country. “We will have some work to do to bring about healing and reconciliation after this election. We have to begin listening to one another and respecting one another.”98

Trump’s message in the final days sought to reiterate many of the same themes that he had emphasized throughout the campaign. In addition to reiterating the claim that Hillary was “too corrupt” to be president, he discussed the importance of cutting taxes on businesses to help stimulate the economy, increasing funding for the military, terminating NAFTA, and building the wall along the Mexican border. Trump also reemphasized the need to deal with what he perceived as the growing immigration problem in the country. At one campaign stop over the weekend in New Hampshire, he declared, “And we will keep radical Islamic terrorists the hell out of our country.”99

DID THE CAMPAIGN MAKE A DIFFERENCE?

It is appropriate to ask whether the general election campaign made any difference, and the answer depends on the yardstick used to measure the campaign’s effects. Did it determine the winner? Did it encourage voters to go to the polls? Did it affect the choices of a substantial number of voters? Did it put issues and candidates’ positions clearly before the voters? Would a better campaign by one of the major party candi- dates have yielded a different result? Did the campaign produce significant events that will have a lasting impact on American politics? We cannot provide firm answers to all of these questions, but we can shed light on some of them.

Regarding the outcome and voters’ decisions, it seems clear that the campaign did indeed have an effect.100 As noted the relative standing of the candidates ebbed and flowed from the conventions to November, and these changes seemed to be linked in part to events in the campaign. Both candidates clearly got a boost from their respective conventions, but Clinton’s seemed to last longer, especially given the series of gaffes that Donald Trump and his campaign endured during the month of August. Clinton also seemed to do well in each of the three debates after a rough start in September. However, the renewed investigation by the FBI into Clinton’s emails with less than two weeks until the election clearly affected the closeness of the race, especially because the FBI did not seem to be investing equal time into allegations that Russia might be trying to influence the presidential election. In the final few days of the campaign, most of the undecideds seemed to move toward Trump, which may have played a key role in him carrying states like Michigan, Pennsylvania, and Wisconsin in the Rust Belt.101

Perhaps the best evidence of the campaign’s impact relates to turnout. As noted earlier many Republicans thought they had an advantage in 2016 because the Democrats had controlled the White House for the previous eight years and the electorate would be less favorable to them than it had been in either 2008 or 2012. Although Trump’s missteps throughout the fall left a number of Republicans worried

66 Part one • the 2016 Presidential election

about the election outcome, the returns coming in at and around 9:00 p.m. on elec- tion night seemed to suggest the race would be a lot closer than expected. In key states such as Michigan, North Carolina, Ohio, Pennsylvania, Virginia, and Wisconsin, the race was simply too close to call based on exit poll data, suggesting that much of the previous polling in these states had been mistaken. If turnout among white voters was higher than expected across these states and African Americans and Hispanics did not turn out in as great of numbers as some anticipated, there was a chance that Trump could turn the tide in the election. In the end this is what happened as exit polls showed Trump won the white vote by a record margin (58 to 37 percent) and turnout among ethnic groups and women was lower than expected.102

The success of the presidential candidates’ mobilization effort is indicated by the turnout data compiled for 2012 and 2016 by Michael McDonald of the University of Florida.103 The data show that the national turnout rate increased by 1.6 points, from 58.6 percent to 60.2 percent. The increase, however, was not equal across the battleground states. Table 2-1 lists the thirteen battleground states and their turn- out in the two elections. In seven of the thirteen states, turnout increased relative to 2012, whereas it declined in the remaining six. Trump ended up winning a majority of the battleground states—including key Rust Belt states that traditionally vote for the Democratic candidate—which is surprising given the limited number of field offices in many of these states relative to Clinton and despite lingering concerns about his ability to win. At the same time it appears that Trump did better in states with slightly lower-than-average turnout.104

There is also the question of whether a better campaign by a candidate, specifi- cally by Clinton, would have led to a different result. Many observers expressed the view that a better campaign by Clinton could have carried the race. “In anointing Clinton, the Democrats went all in with a candidate despised by a good portion of the country. . . . Warning signs of trouble came early and often in her campaign, but she failed to heed them. She was overly cautious, effectively staying on a course of political destruction rather than learning from mistakes along the way.”105 Critics also claimed that her campaign did not go far enough in attempting to reach out to voters who had supported Sanders over her in the primary by making a bolder vice presidential pick such as Sanders or Elizabeth Warren. “Failing to fire up her own side only exacerbated how much she fired up her opposition. . . . Her incendiary ‘basket of deplorables’ and other remarks flourished on the internet and fit into the narrative of people who spread an anti-Hillary gospel.”106

Some were even of the dubious opinion that had Sanders won the Democratic primary, he might have had a better shot of defeating Trump in the election. When asked by the Washington Post several days after the election if he could have beaten Trump, Sanders responded by simply saying, “I hesitate to be a Monday morning quarterback. In my heart of hearts, I think there’s a good chance I could have defeated Trump, but who knows.”107 When his wife Jane was asked the same question the week before on CNN, she replied, “Absolutely, but it doesn’t matter now.”108 Not every- one shared these opinions about Sanders’s likely victory over Trump in the election, however. At least one postelection commentary criticized this perspective by saying that most of the candidates that Sanders endorsed ended up underperforming in the

chaPter two • the General election caMPaiGn 67

Turnout Turnout Change 2012–20162012 2016

National 58.6 60.2 +1.6

Colorado 70.6 72.1 +1.5

Florida 63.3 65.8 +2.5

Iowa 70.6 69.0 −1.6

Michigan 65.4 65.7 +0.3

Minnesota 76.4 74.8 −1.6

Nevada 56.5 57.3 +0.8

New Hampshire 70.9 72.5 +1.6

New Mexico 54.8 55.2 +0.4

North Carolina 65.4 65.2 −0.2

Ohio 65.1 64.2 −0.9

Pennsylvania 59.5 63.0 +3.5

Virginia 66.6 66.1 −0.5

Wisconsin 72.9 69.4 −3.5

Table 2-1 Change in Turnout in Battleground States and Nationally, 2012–2016

Source: Data are from the United States Elections Project, http://www.electproject.org/, accessed February 7, 2017.

election, which says a lot about how he might have done himself had he been on the ballot instead of Clinton.109 Moreover, this perspective fails to account for the fact that Clinton won the popular tally by more than 2.86 million votes but ended up losing the Electoral College as a result of near misses in states like Michigan, Pennsylvania, and Wisconsin, where she failed to spend time or money in the waning days of the campaign.110

Additionally campaigns without an incumbent are usually close, and close elec- tions can go either way. One should not expect the positives from Obama’s per- formance to easily transfer to Clinton, his former secretary of state. Nevertheless, Clinton did carry the popular vote despite failing to achieve a majority in the Electoral

68 Part one • the 2016 Presidential election

College (like Al Gore did in 2000). This was the seventh presidential race with no incumbent since the Second World War (the others were 1952, 1960, 1968, 1988, 2000, and 2008). The elections of 1960, 1968, and 2000 were closer races in terms of the popular vote, and even though Bush’s father won fairly easily in 1988, he had trailed in the polls that year as well.

Finally, and perhaps most consequential for the outcome of the election, Clinton seemed to be at a strategic disadvantage as a result of having to run against Sanders in the primary and having both Gary Johnson and Jill Stein in the race during the fall campaign. Many Democrats who had strongly supported Sanders, especially younger voters, seemed to reluctantly embrace Clinton once she won the Democratic nomi- nation. In the fall Johnson and Stein were polling as high as 8 and 4 percent, respec- tively, in some pre-election polls, which may indicate that voters were looking for candidates other than Clinton and Trump to support in the election. The fact that Johnson earned at least 2.4 percent of the vote in very close states like Pennsylvania and Michigan, may have been just enough to shift the Electoral College vote in favor of Trump.

CHAPTER THREE

THE ELECTION RESULTS

In the closing days of the bitter campaign, it appeared that Hillary Clinton held an electoral advantage. Analysis of early voting patterns in several swing states suggested that Clinton’s ground game advantage was paying dividends—registered Democrats were turning out at higher rates than registered Republicans.1 Donald Trump, who had been calling the election “rigged” in the weeks leading to election day, responded to reports of early voting trends with additional skepticism: “Do you think these bal- lots are properly counted,” Trump asked supporters in Colorado; “do you think?”2 The polls also suggested a Clinton victory was on the horizon. The average (mean) of ten national public opinion polls conducted during the final week of the campaign showed Clinton with 46.8 percent of the vote and Trump with 43.6 percent, a seem- ingly comfortable lead for the Democratic nominee.3 Indeed these ten polls were decidedly pointed in Clinton’s favor: six predicted a Clinton victory of greater than four percentage points, three predicted a Clinton victory in the one to three point range, and only one national poll predicted a Trump victory (by a margin of three points). A series of academic models were similarly predictive of a Clinton electoral victory.4 In October 2016, James E. Campbell published the results of eleven statisti- cal forecasts for the presidential election.5 The median prediction from these models projected that Clinton would win 51.1 percent of the two-party vote. Nine of the eleven models forecasted a Clinton win, whereas only two predicted Trump’s popular vote victory. Importantly, none of these polls or models of the election predicted the Electoral College outcome, only the popular vote.

All eyes were on New York on election day, November 8, as for the first time since 1944 both major party nominees hailed from the same state.6 Both candidates held last- minute campaign events that bled into election day before returning to New York to cast their ballots. Clinton had taken part in an election-eve rally in Philadelphia that featured her husband, as well as President Obama and First Lady Michelle Obama, and was punc- tuated by musical performances by Jon Bon Jovi and Bruce Springsteen.7 She then jetted to Raleigh, North Carolina, where she was joined by Lady Gaga for a 1:00 a.m. rally.8 Trump made his final campaign stop shortly after midnight in Grand Rapids, Michigan, a state that last voted for a Republican presidential candidate in 1988. It was his elev- enth campaign stop in two days, a trip that covered seven states and three time zones.9 Despite being the oldest two candidates to ever square off in a presidential contest, Clinton and Trump both sprinted to the proverbial finish line before returning to their

70 Part one • the 2016 Presidential election

respective homes in the wee hours of election day morning.10 And both were back at it within hours. Clinton and her husband arrived at the Douglas Grafflin Elementary School in Chappaqua, New York, at 8:00 a.m. to vote. She described casting a ballot for herself as “the most humbling experience, you know, I know how much responsibil- ity goes with this and so many people are counting on the outcome of this election.”11 Trump and his wife, Melania, arrived at P.S. 59 in midtown Manhattan a few hours later to cast their votes. Trump joked with reporters that it was a “tough decision” and then noted presciently, “We’ll see what happens; it’s going to be an interesting day.”12

The campaigns’ election night parties were held a mere twenty-five blocks from each other.13 Trump supporters—many of whom were wearing the campaign’s now ubiquitous, red “Make American Great Again” baseball caps—had gathered at the New York Hilton Midtown. But the lasting images of the night are likely those that were witnessed at the Clinton campaign party at the Javits Convention Center on Manhattan’s West Side, where excitement turned to shock and then despair. The Clinton supporters’ optimism as the night began was not unfounded; the electoral map seemed to present a very difficult path to victory for Trump. The toss-up states on the electoral map were Florida, Iowa, North Carolina, and Ohio—three of these had voted Democrat in the last two elections, whereas North Carolina supported Obama in 2008 but not in 2012. The Trump campaign believed that Michigan, Pennsylvania, and Wisconsin were also in play, but these states had voted Democrat in each election since 1992. There are, of course, various routes to an Electoral College victory, but the numerical advantage that Clinton garnered from “safe” Democratic states meant that she needed far fewer of the toss-up states to win than Trump required.14 Expectations started to give way to reality around 10:21 p.m., however, when Ohio was called for Trump. At 11:07 p.m., North Carolina went Trump’s way, followed soon thereafter by Florida and Iowa.15 Early returns from the other toss-up states—all of them!—were also leaning toward Trump. Around 2:00 a.m., Clinton’s campaign chairman, John Podesta, took the Javits Center stage to inform the shocked and tearful crowd that their candidate did not plan to speak and that “everybody should head home.”16 Forty minutes later, Clinton called Trump to concede the race.

Donald Trump had pulled off a shocking and decided Electoral College vic- tory, capturing 304 electoral votes to Hillary Clinton’s 227. But the Electoral College outcome does not adequately capture the level of political division within the 2012 electorate. Indeed Trump became the fourth presidential candidate to win the Electoral College without winning the popular vote.17 In the final tally, Clinton won a 48.2 percent plurality of the popular vote, while Trump won 46.1 percent. Discontent with the two major-party candidates was made evident by the fact that independent and third-party candidates garnered 5.7 percent of the total popular vote in 2016, four percentage points higher than in 2012. Leading the way were Libertarian Gary Johnson, who garnered 3.3 percent of the national popular vote, and Green Party nominee, Jill Stein, who won 1.1 percent.18 Clinton’s popular vote margin of 2.1 per- centage points makes 2016 the ninth closest presidential election in U.S. history and the closest since 2000, when Democrat Al Gore claimed a 0.51 percentage point popu- lar vote victory over Republican George W. Bush, only to lose the Electoral College.

Table 3-1 presents the official 2016 election results by state and includes those for Maine’s two congressional districts and Nebraska’s three districts.19 Clinton won

Ta bl

e 3-

1 Pr

es id

en ti

al E

le ct

io n

R es

u lt

s by

S ta

te , 2

01 6

T o

ta l V

o te

(% )

S ta

te T

o ta

l V

o te

C lin

to n

em .)

T ru

m p

ep .)

O th

er T

w o

-p a

rt y

if fe

re n

ti al

D em

. R

ep .

A la

ba m

a 2,

12 3,

37 2

72 9,

54 7

1, 31

8, 25

5 75

,5 70

58 8,

70 8

R 34

.4 62

A la

sk a

31 8,

60 8

11 6,

45 4

16 3,

38 7

38 ,7

67 46

,9 33

R 36

.6 51

A ri

zo na

2, 57

3, 16

5 1,

16 1,

16 7

1, 25

2, 40

1 15

9, 59

7 91

,2 34

R 45

.1 48

A rk

an sa

s 1,

13 0,

63 5

38 0,

49 4

68 4,

87 2

65 ,2

69 30

4, 37

8 R

33 .7

60 .6

C al

if or

ni a

14 ,1

81 ,5

95 8,

75 3,

78 8

4, 48

3, 81

0 94

3, 99

7 4,

26 9,

97 8

D 61

.7 31

C ol

or ad

o 2,

78 0,

24 7

1, 33

8, 87

0 1,

20 2,

48 4

23 8,

89 3

13 6,

38 6

D 48

.2 43

C on

ne ct

ic ut

1, 64

4, 92

0 89

7, 57

2 67

3, 21

5 74

,1 33

22 4,

35 7

D 54

.6 40

D el

aw ar

e 44

3, 81

4 23

5, 60

3 18

5, 12

7 23

,0 84

50 ,4

76 D

53 .1

41 .7

Fl or

id a

9, 42

0, 03

9 4,

50 4,

97 5

4, 61

7, 88

6 29

7, 17

8 11

2, 91

1 R

47 .8

49 .0

G eo

rg ia

4, 11

4, 73

2 1,

87 7,

96 3

2, 08

9, 10

4 14

7, 66

5 21

1, 14

1 R

45 .6

50 .8

H aw

ai i

42 8,

93 7

26 6,

89 1

12 8,

84 7

33 ,1

99 13

8, 04

4 D

62 .2

30 .0

Id ah

o 69

0, 25

5 18

9, 76

5 40

9, 05

5 91

,4 35

21 9,

29 0

R 27

.5 59

(C on

ti nu

ed )

T o

ta l V

o te

(% )

S ta

te T

o ta

l V

o te

C lin

to n

em .)

T ru

m p

ep .)

O th

er T

w o

-p a

rt y

if fe

re n

ti al

D em

. R

ep .

Ill in

oi s

5, 53

6, 42

4 3,

09 0,

72 9

2, 14

6, 01

5 29

9, 68

0 94

4, 71

4 D

55 .8

38 .8

In di

an a

2, 73

4, 95

8 1,

03 3,

12 6

1, 55

7, 28

6 14

4, 54

6 52

4, 16

0 R

37 .8

56 .9

Io w

a 1,

56 6,

03 1

65 3,

66 9

80 0,

98 3

11 1,

37 9

14 7,

31 4

R 41

.7 51

K an

sa s

1, 18

4, 40

2 42

7, 00

5 67

1, 01

8 86

,3 79

24 4,

01 3

R 36

.1 56

K en

tu ck

y 1,

92 4,

14 9

62 8,

85 4

1, 20

2, 97

1 92

,3 24

57 4,

11 7

R 32

.7 62

Lo ui

si an

a 1,

99 4,

06 5

78 0,

15 4

1, 17

8, 63

8 35

,2 73

39 8,

48 4

R 39

.1 59

M ai

ne a

74 7,

92 7

35 7,

73 5

33 5,

59 3

54 ,5

99 22

,1 42

D 47

.8 44

M ar

yl an

d 2,

78 1,

44 6

1, 67

7, 92

8 94

3, 16

9 16

0, 34

9 73

4, 75

9 D

60 .3

33 .9

M as

sa ch

us et

ts 3,

32 5,

04 6

1, 99

5, 19

6 1,

09 0,

89 3

23 8,

95 7

90 4,

30 3

D 60

.0 32

M ic

hi ga

n 4,

79 9,

28 4

2, 26

8, 83

9 2,

27 9,

54 3

25 0,

90 2

10 ,7

04 R

47 .3

47 .5

M in

ne so

ta 2,

94 4,

81 3

1, 36

7, 71

6 1,

32 2,

95 1

25 4,

14 6

44 ,7

65 D

46 .4

44 .9

M is

si ss

ip pi

1, 20

9, 35

7 48

5, 13

1 71

0, 74

6 13

,4 80

22 5,

61 5

R 40

.1 58

Ta bl

e 3-

1 (C

on ti

n u

ed )

T o

ta l V

o te

(% )

S ta

te T

o ta

l V

o te

C lin

to n

em .)

T ru

m p

ep .)

O th

er T

w o

-p a

rt y

if fe

re n

ti al

D em

. R

ep .

M is

so ur

i 2,

80 8,

60 5

1, 07

1, 06

8 1,

59 4,

51 1

14 3,

02 6

52 3,

44 3

R 38

.1 56

M on

ta na

49 7,

14 7

17 7,

70 9

27 9,

24 0

40 ,1

98 10

1, 53

1 R

35 .7

56 .2

N eb

ra sk

ab 84

4, 22

7 28

4, 49

4 49

5, 96

1 63

,7 72

21 1,

46 7

R 33

.7 58

N ev

ad a

1, 12

5, 38

5 53

9, 26

0 51

2, 05

8 74

,0 67

27 ,2

02 D

47 .9

45 .5

N ew

H am

ps hi

re 74

4, 29

6 34

8, 52

6 34

5, 79

0 49

,9 80

2, 73

6 D

46 .8

46 .5

N ew

J er

se y

3, 87

4, 04

6 2,

14 8,

27 8

1, 60

1, 93

3 12

3, 83

5 54

6, 34

5 D

55 .5

41 .4

N ew

M ex

ic o

79 8,

31 9

38 5,

23 4

31 9,

66 7

93 ,4

18 65

,5 67

D 48

.3 40

N ew

Yo rk

7, 72

1, 45

3 4,

55 6,

12 4

2, 81

9, 53

4 34

5, 79

5 1,

73 6,

59 0

D 59

.0 36

N or

th C

ar ol

in a

4, 74

1, 56

4 2,

18 9,

31 6

2, 36

2, 63

1 18

9, 61

7 17

3, 31

5 R

46 .2

49 .8

N or

th D

ak ot

a 34

4, 36

0 93

,7 58

21 6,

79 4

33 ,4

18 12

3, 03

6 R

27 .2

63 .0

O hi

o 5,

49 6,

48 7

2, 39

4, 16

4 2,

84 1,

00 5

26 1,

31 8

44 6,

84 1

R 43

.6 51

O kl

ah om

a 1,

45 2,

99 2

42 0,

37 5

94 9,

13 6

83 ,4

81 52

8, 76

1 R

28 .9

65 .3

O re

go n

2, 00

1, 33

6 1,

00 2,

10 6

78 2,

40 3

21 6,

31 8

21 9,

70 3

D 50

.1 39

P en

ns yl

va ni

a 6,

16 5,

47 8

2, 92

6, 44

1 2,

97 0,

73 3

26 8,

30 4

44 ,2

92 R

47 .5

48 .2

(C on

ti nu

ed )

T o

ta l V

o te

(% )

S ta

te T

o ta

l V

o te

C lin

to n

em .)

T ru

m p

ep .)

O th

er T

w o

-p a

rt y

if fe

re n

ti al

D em

. R

ep .

R ho

de Is

la nd

44 6,

14 4

25 2,

52 5

18 0,

54 3

13 ,0

76 12

2, 47

3 D

56 .6

40 .5

S ou

th C

ar ol

in a

2, 10

3, 02

7 85

5, 37

3 1,

15 5,

38 9

92 ,2

65 30

0, 01

6 R

40 .7

54 .9

S ou

th D

ak ot

a 37

0, 09

3 11

7, 45

8 22

7, 72

1 24

,9 14

11 0,

26 3

R 31

.7 61

Te nn

es se

e 2,

50 8,

02 7

87 0,

69 5

1, 52

2, 92

5 11

4, 40

7 65

2, 26

3 R

34 .7

60 .7

Te xa

s 8,

96 9,

22 6

3, 87

7, 86

8 4,

68 5,

04 7

40 6,

31 1

80 7,

17 9

R 43

.2 52

U ta

h 1,

13 1,

43 0

31 0,

67 6

51 5,

23 1

30 5,

52 3

20 4,

55 5

R 27

.5 45

Ve rm

on t

31 5,

06 7

17 8,

57 3

95 ,3

69 41

,1 25

83 ,2

04 D

56 .7

30 .3

Vi rg

in ia

3, 98

4, 63

1 1,

98 1,

47 3

1, 76

9, 44

3 23

3, 71

5 21

2, 03

0 D

49 .7

44 .4

W as

hi ng

to n

3, 31

7, 01

9 1,

74 2,

71 8

1, 22

1, 74

7 35

2, 55

4 52

0, 97

1 D

52 .5

36 .8

W es

t V irg

in ia

71 4,

42 3

18 8,

79 4

48 9,

37 1

36 ,2

58 30

0, 57

7 R

26 .4

68 .5

W is

co ns

in 2,

97 6,

15 0

1, 38

2, 53

6 1,

40 5,

28 4

18 8,

33 0

22 ,7

48 R

46 .5

47 .2

W yo

m in

g 25

5, 84

9 55

,9 73

17 4,

41 9

25 ,4

57 11

8, 44

6 R

21 .9

68 .2

D is

tr ic

t o f C

ol um

bi a

31 1,

26 8

28 2,

83 0

12 ,7

23 15

,7 15

27 0,

10 7

D 90

.9 4

U ni

te d

S ta

te s

13 6,

66 9,

23 7

65 ,8

53 ,5

16 62

,9 84

,8 25

7, 83

0, 89

6 2,

86 8,

69 1

D 48

.2 46

So u

rc e:

F ed

er a

l E

le ct

io n

C o

m m

is si

o n

, “O

ff ic

ia l

20 16

P re

si d

en ti

a l

G en

er a

l E

le ct

io n

R es

u lt

s, ”

Ja n

u a

ry 3

0, 2

01 7,

h tt

p s:

// tr

a n

si ti

o n

.f ec

.g o

v /p

u b

re c/

fe 20

16 /2

01 6p

re sg

er es

u lt

s. p

d f.

B as

ed o

n r

ep or

ts o

f th

e se

cr et

ar ie

s of

s ta

te o

f th

e fi

ft y

st at

es a

n d

t h

e D

is tr

ic t

of C

ol u

m bi

a I n

M ai

n e

th e

st at

ew id

e p

lu ra

li ty

v ot

e w

in n

er g

ai n

ed t

w o

vo te

s, a

n d

e ac

h o

f t h

e st

at e’

s tw

o co

n gr

es si

on al

d is

tr ic

ts w

as a

w ar

d ed

o n

e vo

te o

n a

p lu

ra li

ty b

as is

h e

of fi

ci al

r es

u lt

s re

p or

te d

b y

co n

gr es

si on

al d

is tr

ic t

d o

n ot

m at

ch t

h e

of fi

ci al

s ta

te -w

id e

to ta

ls r

ep or

te d

Ta bl

e 3-

1 (C

on ti

n u

ed )

M ai

74 7,

92 7

35 7,

73 5

33 5,

59 3

54 ,5

99 22

,1 42

D 47

.8 44

1s t D

is tr

ic t

39 2,

10 7

21 0,

97 9

15 4,

12 7

27 ,0

01 56

,8 52

D 53

.8 39

2n d

D is

tr ic

t 35

1, 83

4 14

3, 73

9 18

0, 81

8 27

,2 27

37 ,0

79 R

40 .9

51 .4

b I n

N eb

ra sk

a th

e st

at ew

id e

p lu

ra li

ty v

ot e

w in

n er

g ai

n ed

t w

o vo

te s,

a n

d e

ac h

o f

th e

st at

e’ s

th re

e co

n gr

es si

on al

d is

tr ic

ts w

as a

w ar

d ed

o n

e vo

te o

n a

p lu

ra li

ba si

s (t

h e

of fi

ci al

r es

u lt

s re

p or

te d

b y

co n

gr es

si on

al d

is tr

ic t

d o

n ot

m at

ch t

h e

of fi

ci al

s ta

te -w

id e

to ta

ls r

ep or

te d

N eb

ra sk

84 4,

22 7

28 4,

49 4

49 5,

96 1

63 ,7

72 21

1, 46

7 R

33 .7

58 .7

1s t D

is tr

ic t

28 2,

33 8

10 0,

13 2

15 8,

64 2

23 ,5

64 58

,5 10

R 35

.5 56

2n d

D is

tr ic

t 29

1, 68

0 13

1, 03

0 13

7, 56

4 23

,0 86

6, 53

4 R

44 .9

47 .2

3r d

D is

tr ic

t 27

0, 03

4 53

,2 90

19 9,

65 7

17 ,0

87 14

6, 36

7 R

19 .7

73 .9

76 Part one • the 2016 Presidential election

roughly 65.8 million votes, a total that is similar to the 65.9 million votes won by Barack Obama four years earlier. Yet, compared to Obama in 2012, Clinton and the Democrats saw declines in two-party vote share in 45 of the 50 states, by an average decrease of 5.1 percentage points. Trump won roughly 62.9 million votes nationally, nearly 3 million fewer votes than Clinton. However, Trump’s popular vote total dem- onstrated gains for the Republican Party. Trump garnered just over two million more popular votes than the 2012 Republican nominee, former Massachusetts Governor Mitt Romney. And in half of the states carried by Obama in 2012, Trump improved the Republican’s two-party vote share by an average increase of 2.2 percentage points.

Despite her favorable popular vote margin, Clinton lost six of the twenty-six states won by Obama in 2012, costing her the Electoral College and the presidency.20 As Figure 3-1 shows, Trump won 304 electoral votes to Clinton’s 227.21 Among the 538 electors chosen to represent their states in the Electoral College, there were a record seven “faithless electors” who cast ballots contrary to their states’ popular votes— five who were faithless to the Democratic nominee and two to the Republican.22 Nevertheless, Trump’s 304 electoral votes easily surpassed the 270 votes needed for election. Trump carried thirty of the fifty states and received one electoral vote for winning Maine’s second congressional district. Trump won each of the twenty-four states that Romney won for the Republicans in 2012. Trump’s average two-party vote share margin in these twenty-four states was 22.6 percentage points—only Arizona and North Carolina were within a five-point margin. To this electoral coalition Trump was able to add Florida, Iowa, Michigan, Ohio, Pennsylvania, and Wisconsin—states that comprise a total of 99 electoral votes. Although he won Iowa and Ohio rather comfortably, Trump’s margin in the remaining four states was extraordinarily thin. Trump’s electoral margin in Florida was 1.2 percentage points, and he won Michigan, Pennsylvania, and Wisconsin by an average of only 0.57 points. Indeed in Michigan, Trump bested Clinton by only 10,704 votes. No matter how close his victories in these battleground states may have been, it is remarkable that Trump was able to put all of these states in the Republican column, especially when one considers that the GOP has not won Michigan or Pennsylvania since 1988 and Wisconsin since 1984.

THE ELECTION RULES

Rules matter. And this is certainly the case in U.S. presidential elections.23 Electoral rules, specifically those pertaining to the Electoral College, structure the nature of party competition and voter behavior, they influence the strategic actions of candi- dates (as we saw in both Chapters 1 and 2), and as was the case in 2016, they some- times dictate who wins and who loses. Whereas Hillary Clinton won the popular vote, Donald Trump won the Electoral College and was elected president. Simply put the rules governing U.S. presidential elections do not guarantee that the candidate who receives the most votes wins. In addition to the 2016 election, on four other occasions in American history—the elections of John Quincy Adams in 1824, Rutherford B. Hayes in 1876, Benjamin Harrison in 1888, and George W. Bush in 2000—the plural- ity winner of the popular vote failed to achieve a majority in the Electoral College and lost the presidency.24 Confusion or a lack of knowledge about the electoral rules that

chapter three • the election results 77

govern the United States can cause consternation among voters, and calls for Electoral College reform (or its elimination) resound when the outcome of the electoral process does not seem to coincide with the general wishes of the electorate, but the rules are well defined.25

As we saw in Chapter 2, voters do not vote directly for president. Rather they vote for a slate of electors pledged to support certain presidential and vice presidential candidates. In every state except Maine and Nebraska, the slate that receives the most popular votes (a plurality) is awarded all of the state’s electoral votes. In no state is a majority required to win. In 2016 Trump won majorities in twenty-three of the thirty states he carried; Clinton won majorities in thirteen of the twenty states in which she led voting. Plurality winners are most likely to occur when a prominent independent or third-party candidate (or candidates) is on the ballot. In 2016 a robust 5.7 percent

Figure 3-1 Electoral Votes, by State, 2016

Source: Federal Election Commission, “Official 2016 Presidential General Election Results,” January 30, 2017, https://transition.fec.gov/pubrec/fe2016/2016presgeresults.pdf. Based on reports of the secretaries of state of the fifty states and the District of Columbia.

Note: Donald J. Trump won 304 electoral votes; Hillary Clinton, 227.

*Trump was awarded one electoral vote from Maine for winning the state’s second congressional district. In 2016 there were seven “faithless electors”: a Democratic Party elector from Hawaii cast a ballot for Bernie Sanders; a Republican Party elector from Texas cast a ballot for John Kasich, and another voted for Ron Paul; and in Washington three Democratic electors cast ballots for former Secretary of State Colin Powell, a Republican, and another voted for Faith Spotted Eagle, a Native American activist.

ND 3

SD 3

TX* 38

MN 10

LA 8

IN 11 WV

DE 3 NJ 14

CT 7 RI 4

ME* 4

MA 11

MI 16

WA* 12 MT

WY 3

ID 4

UT 6 CO

9 KS 6

CA 55

AZ 11

NM 5

NV 6

OR 7

NE 5

OK 7

IA 6

MO 10

AR 6

WI 10

IL 20

MS 6

OH 18

AL 9

GA 16

KY 8

TN 11

FL 29

NC 15

VA 13

SC 9

PA 20

MI 16IA

WI 10

OH 18

FL 29

PA 20

NY 29

NH 4

VT 3

MD 10

AK 3 HI*

DC 3

States carried by Clinton (Obama in 2012) States carried by Trump (Obama in 2012)

States carried by Trump (Romney in 2012)

78 Part one • the 2016 Presidential election

of the national popular vote went to candidates outside of the two main parties, result- ing in plurality winners in fourteen of the fifty states. Compare this to 2012, when only 1.7 percent of the national popular vote went to candidates outside of the two main parties, or 2008, when the independent/third-party vote share was 1.4 percent. In 2012 both Obama and Romney won majorities in each of the states in which they were victorious. In 2008 Obama won a majority in all but two of the twenty-eight states he carried.

The plurality rule, winner-take-all system usually transforms a plurality of the popular vote into a majority of the electoral vote. And it takes a majority of the elec- toral vote (currently 270 votes) to produce a winner. If no candidate wins a majority of the electoral vote, the U.S. House of Representatives, voting by state delegations, chooses among the three candidates with the highest number of electoral votes. Yet the House has not chosen a winner since 1825. In forty-four of the forty-eight elec- tions from 1828 to 2016, the candidate with the most popular votes has won a majority of the electoral vote. During this period there were fourteen elections in which a can- didate won a plurality of the national vote—but not a majority—and won a majority of the electoral vote.26 So why does the plurality rule, winner-take-all system typically produce a majority winner? The answer lies in the tendency for plurality rule voting systems to yield a two-party system.

The U.S. plurality vote system is a confirmation of Duverger’s Law, a proposition advanced by French jurist and political scientist Maurice Duverger in the 1950s.27 According to Duverger, “the simple-majority single-ballot system favours the two- party system.”28 Indeed Duverger argued that “the American procedure corresponds to the usual machinery of the simple-majority single-ballot system. The absence of a second ballot and of further polls, particularly in the presidential election, constitutes in fact one of the historical reasons for the emergence and the maintenance of the two-party system.”29

According to Duverger, this principle applies for two reasons. First, a plurality vote system produces a “mechanical factor”: third parties may earn a large number of votes nationally but fail to gain a plurality of the votes in many electoral units. Scholars agree that this effect is important, except in countries where smaller parties have a geographic base. Second, some voters who prefer a candidate or party they think cannot win will cast their votes for their first choice between the major-party candidates, which Duverger labels the “psychological factor.” This behavior is called “sophisticated” or “strategic” voting and in Britain is referred to as “tactical” voting. William H. Riker defines strategic voting as “voting contrary to one’s immediate tastes in order to obtain an advantage in the long run.”30 Whether strategic voting occurs to any significant extent is controversial, yet evidence suggests that a substantial number of voters who preferred a third-party or independent candidate in the 1968, 1980, 1992, 1996, and 2000 elections wound up voting for one of the major-party candidates instead of voting their “sincere” preferences.31 As we will see in Chapter 6, it seems highly likely that a substantial number of voters in 2016 who preferred a third-party candidate voted for one of the major-party candidates instead.

The plurality rule system thus places a heavy burden on third-party or indepen- dent candidates. Even a relatively successful third-party candidate typically receives a far smaller share of the electoral vote than of the popular vote.32 Here it is useful to

chaPter three • the election results 79

review the fates of the four most successful third-party and independent candidacies (in popular vote results) since World War II: George C. Wallace won 13.5 percent of the popular vote in 1968, John B. Anderson won 6.6 percent in 1980, and H. Ross Perot won 18.9 percent in 1992 and 8.4 percent in 1996. Despite relatively high levels of popular support among these candidates, only Wallace was able to win enough votes in a state to obtain electoral votes. Wallace came in first in five states (winning majorities in Alabama and Mississippi) and gaining forty-six electoral votes (including one from a faithless elector from North Carolina). Yet Wallace won only 8.5 percent of the electoral vote, significantly less than his popular vote share.33

The rules had a direct and clearly discernable effect on the outcome of the 2016 election, and the rules also create the conditions that have led to major-party domi- nance; they help shape the strategies that campaigns employs, and they also constrain the choices that voters make. Choosing the president by presidential electors is a central part of these rules, and a strong case can be made for eliminating the Electoral College.34 Some critics, such as presidential scholar George C. Edwards III, argue in favor of direct election of the president. Direct election would force candidates to campaign nationally (although candidates would likely concentrate their efforts in densely populated, urban areas) and would promote equality by making every vote in every state count. Moreover, direct election would eliminate questions of fairness that arise when popular-vote winners do not win the presidency. The main obstacle to adopting a direct election system is that it requires a constitutional amendment, which is unlikely because gaining approval of three-fourths of the states would be difficult in a system that overrepresents the smaller states.35

An alternative reform would retain the Electoral College but would diminish its importance by establishing a compact among states that would guarantee that elec- tors would vote for the national popular-vote winner regardless of the outcome within their own state. This compact would come into effect only after it is enacted by states collectively possessing a majority of the electoral votes. As of July 2017 ten states and the District of Columbia with a total of 165 electoral votes (roughly 61 percent of the 270 electoral votes needed) had agreed to the National Popular Vote Interstate Compact.36 Because an interstate compact requires congressional approval, there would still be an additional hurdle, but only a majority of both chambers is required to approve an interstate compact, not the two-thirds supermajority to initiate a con- stitutional amendment.

As described earlier Maine and Nebraska both have district systems for choos- ing electors, and widespread state-level adoption of this method is often put forward as a way of reforming the Electoral College. The district system does away with the winner-take-all rule for assigning electors and makes it possible for candidates to split a state’s electoral vote. However, this method does not guarantee that electors will be divided in proportion to candidates’ popular-vote shares. Indeed, because partisan majorities in state legislatures typically gerrymander congressional district lines to make the districts uncompetitive and unbalanced in their favor, the adoption of the district method would likely bias Electoral College outcomes relative to the popu- lar vote.37 A study by The Cook Political Report shows that had there been a uniform application of the district method in 2016, Donald Trump would have been elected.38 Trump won more congressional districts than Clinton (230 of 435) while carrying

80 Part one • the 2016 Presidential election

30 states for a total of 290 electoral votes. Clinton would have received 248 electoral votes based on 205 congressional district wins, twenty state popular-vote wins, and the District of Columbia. Obviously this result is in line with the real Electoral College outcome, but note that neither method is consistent with the national popular vote. Moreover, the district method is not always in line with the Electoral College. In 2012 Barack Obama won the popular vote (51.1 percent) and the Electoral College (332 electoral votes to Mitt Romney’s 206), yet the district method would have resulted in a Romney victory because he carried 226 of 425 districts and twenty-four states, a total of 274 electoral votes.39 Widespread adoption of the district system is unlikely, how- ever, because most states do not want to diminish their potential influence by making it likely that their electoral votes will be split.

THE PATTERN OF RESULTS

Political observers often focus their attention on those factors that make a particular election appear unique. The 2016 election certainly featured many events that were unprecedented in American electoral history. Indeed one headline in a major newspa- per referred to the presidential contest as “unusual, unexpected, strange, weird, and . . . bizarre.”40 Yet systematic analysis also requires that commonalities be observed. Thus the 2016 election can be placed in perspective by comparing it with previous presiden- tial elections. Three conclusions emerge. First, the election further demonstrates the competitive nature of postwar elections in the United States, which exhibit a relatively even balance between the two major parties. Second, although 2016 provided the Democrats with their third consecutive popular vote win, postwar elections continue to display a pattern of volatility. Third, although incumbent presidents seem to have an electoral advantage, incumbent parties do not.

Perhaps the most remarkable feature of American presidential elections in the postwar period is the competitive balance between the two major parties. In the eigh- teen elections held since World War II, the Republicans have been victorious in ten, whereas the Democrats have won eight. If one considers popular vote winners, the score is tied eight to eight. The Republicans have been slightly more successful in establishing electoral majorities, winning a majority of the popular vote in seven of these elections (1952, 1956, 1972, 1980, 1984, 1988, and 2004). The Democrats have won a popular-vote majority only four times, including both of Obama’s victories (1964, 1976, 2008, and 2012). The average (mean) level of popular support shows the competitive balance: the Republicans have won 48.7 percent of the popular vote, and the Democrats have won 47.0 percent. This division of popular support also demon- strates the dominance of the two major parties in presidential elections. During the postwar period, third-party and independent candidates have only garnered an aver- age of 4.4 percent of the popular vote.

Examining electoral history is like looking at clouds: if you look hard enough, you’ll find something that looks like a pattern. Yet, with a few important historical exceptions, electoral history is best described as volatile. This is especially true in the postwar period, where competitive balance would appear to place a party’s chances at winning the presidency at 50-50.41 Evidence of electoral volatility should give pause to

chaPter three • the election results 81

pundits and partisans alike who are too quick to label any one election, such as 2016, as the dawn of a new era of electoral dominance. Declarations such as this seem ill- advised and ill-informed.

Table 3-2 presents presidential election results since 1832, the first election in which political parties used national nominating conventions to select their candidates. From 1832 to 1948 there are four instances in which the same party won three elec- tions or more in a row. Scholars often associate these episodes with partisan realign- ments. Walter Dean Burnham, for instance, identifies the elections of 1860, 1896, and 1932 as realigning elections during this period.42 From 1948 to 2004, however, despite the election of five two-term presidents, the same party won three elections in a row only once (the Republicans in 1980, 1984, and 1988). The 2008, 2012, and 2016 elections marked three consecutive popular vote victories for the Democrats, but the party’s inability to secure an Electoral College win in 2016 is greater evidence of partisan balance and electoral volatility than one party advantage.

The postwar period is an era of sustained electoral volatility. From 1952 to 1984, neither party was able to win more than two elections in a row. The Republicans won 1952 and again in 1956, the Democrats won in 1960 and 1964, and the Republicans won in 1968 and 1972. In all three sets of wins, the second win was by a larger margin than the first. Volatility increased in 1980, when the Democrats, who had won the White House in 1976, failed to hold it. The 1980 and 1984 elections reverted to the pattern of back-to-back party wins when Ronald Reagan was reelected in a landslide.43 Then in 1988 George H. W. Bush’s election gave the Republicans three elections in a row, breaking the pattern of postwar volatility. But the Republicans could not sustain their control of the presidency. The Democrats recaptured the White House with Bill Clinton’s victory in 1992. Clinton was reelected in 1996, and he did so by capturing a larger popular-vote margin than in his first election. George W. Bush’s defeat of Al Gore in 2000 continued the postwar volatility, and Bush, who finished second in the popular vote in 2000, improved his vote share in 2004. Obama’s victories in 2008 and 2012 gave some Democrats hope that a period of Democratic Party dominance had emerged. Yet Obama’s reelection was achieved with a smaller popular-vote margin than his first (unlike all other reelected incumbents during the postwar period) and did not carry with it the legislative majorities generated in 2008.44 Hillary Clinton’s popular vote plurality in 2016 fell short of Obama’s electoral majorities and demon- strated a softening of electoral support for the party. By capturing an Electoral College majority, Donald Trump was able to recapture the White House for the Republican Party and add to the pattern of volatility.

The electoral volatility of the postwar period is not without precedent. In fact two periods in the nineteenth century were more volatile. From 1840 to 1852 the incumbent party lost four consecutive elections—a period of volatility between the Democrats and the Whigs. This occurred again from 1884 to 1896, when the Republicans and the Democrats alternated elections. Both of these periods, how- ever, were followed by party realignments. In 1854 just two years after the decisive defeat of the Whigs, the Republican Party was founded, and by the 1856 election the party’s nominee, John C. Fremont, came in second behind James Buchanan, the Democratic winner.45 By 1860 the Republicans had captured the presidency, and the Whigs were extinct.46 Although many Whigs, including Abraham Lincoln himself,

82 Part one • the 2016 Presidential election

Table 3-2 Presidential Election Results, 1832–2016

Election Winning Candidate

Party of Winning Candidate

Success of Incumbent Political Party

1832 Andrew Jackson Democrat Won

1836 Martin Van Buren Democrat Won

1840 William H. Harrison Whig Lost

1844 James K. Polk Democrat Losta

1848 Zachary Taylor Whig Lost

1852 Franklin Pierce Democrat Lost

1856 James Buchanan Democrat Won

1860 Abraham Lincoln Republican Lost

1864 Abraham Lincoln Republican Won

1868 Ulysses S. Grant Republican Wonb

1872 Ulysses S. Grant Republican Won

1876 Rutherford B. Hayes Republican Won

1880 James A. Garfield Republican Won

1884 Grover Cleveland Democrat Lost

1888 Benjamin Harrison Republican Lost

1892 Grover Cleveland Democrat Lost

1896 William McKinley Republican Lost

1900 William McKinley Republican Won

1904 Theodore Roosevelt Republican Won

1908 William H. Taft Republican Won

1912 Woodrow Wilson Democrat Lost

1916 Woodrow Wilson Democrat Won

1920 Warren G. Harding Republican Lost

1924 Calvin Coolidge Republican Won

1928 Herbert C. Hoover Republican Won

chaPter three • the election results 83

Election Winning Candidate

Party of Winning Candidate

Success of Incumbent Political Party

1932 Franklin D. Roosevelt Democrat Lost

1936 Franklin D. Roosevelt Democrat Won

1940 Franklin D. Roosevelt Democrat Won

1944 Franklin D. Roosevelt Democrat Won

1948 Harry S. Truman Democrat Won

1952 Dwight D. Eisenhower Republican Lost

1956 Dwight D. Eisenhower Republican Won

1960 John F. Kennedy Democrat Lost

1964 Lyndon B. Johnson Democrat Won

1968 Richard M. Nixon Republican Lost

1972 Richard M. Nixon Republican Won

1976 Jimmy Carter Democrat Lost

1980 Ronald Reagan Republican Lost

1984 Ronald Reagan Republican Won

1988 George H. W. Bush Republican Won

1992 Bill Clinton Democrat Lost

1996 Bill Clinton Democrat Won

2000 George W. Bush Republican Lost

2004 George W. Bush Republican Won

2008 Barack Obama Democrat Lost

2012 Barack Obama Democrat Won

2016 Donald J. Trump Republican Lost

Source: Presidential Elections, 1789–2008 (Washington, D.C.: CQ Press, 2009); 2012–2016, compiled by authors.

aWhigs are classified as the incumbent party because they won the 1840 election. In fact their presidential candidate, William Henry Harrison, died a month after taking office and his vice president, John Tyler, was expelled from the party in 1841.

bRepublicans are classified as the incumbent party because they won the 1864 election. (Technically Lincoln had been elected on a Union ticket.) In fact after Lincoln’s assassination in 1865, Andrew Johnson, a war Democrat, became president.

84 Part one • the 2016 Presidential election

became Republicans, the Republican Party was not just the Whig Party renamed. The Republicans had transformed the political agenda by capitalizing on opposition to slavery in the territories.47

The 1896 contest, the last of four incumbent party losses, is usually consid- ered a critical election because it solidified Republican dominance.48 Although the Republicans had won five of the seven elections since the end of the Civil War, after Ulysses S. Grant’s reelection in 1872, all their victories had been by narrow margins. In 1896 the Republicans emerged as the clearly dominant party, gaining a solid hold in Connecticut, Indiana, New Jersey, and New York, states that they had frequently lost between 1876 and 1892. After William McKinley’s defeat of William Jennings Bryan in 1896, the Republicans established a firmer base in the Midwest, New England, and the Mid-Atlantic states. They lost the presidency only in 1912, when the GOP was split, and in 1916, when the incumbent, Woodrow Wilson, ran for reelection.49 The Republicans would win again in 1920, 1924, and 1928.

The Great Depression ended Republican dominance. The emergence of the Democrats as the majority party was not preceded by a series of incumbent losses. Instead the Democratic coalition forged in the 1930s relied heavily on the emerging working class and the mobilization of new groups into the electorate.

As the emergence of the New Deal coalition demonstrates, a period of electoral volatility is not a necessary condition for a partisan realignment. Nor perhaps is it a sufficient condition. In 1985 Ronald Reagan himself proclaimed that his reelection was indicative of a realignment. Political scientists were skeptical about that claim mainly because the Democrats continued to dominate the U.S. House of Representatives. George H. W. Bush’s election in 1988 suggested that Republican dominance indeed might have arrived. But Clinton’s 1992 victory called this thesis into question, and his 1996 victory cast further doubts on the idea that a realignment had occurred. After the 2000 election, the Republicans held control of the House, the Senate, and the presidency for the first time since 1953, although they temporarily lost control of the Senate between June 2001 and January 2003.50 But the closeness of the election called into question any claim of Republican dominance. The Democrats regained the presidency—by a comfortable margin—with Obama’s victory in 2008, and they also won relatively large majorities in both houses of Congress, only to lose control of the House and a sizable portion of their advantage in the Senate in the 2010 midterm elections. The election of 2012 gave Obama and the Democrats another four years in the White House, but Congress remained under divided party control. Following the 2016 election the Republicans control the White House and both houses of Congress. Yet, just as unified political control of government following the 2008 election did not launch a period of Democratic dominance, it is unlikely that unified government under the Republicans signals the dawn of GOP ascendancy. No party currently dom- inates American politics. Volatility persists.

One clear pattern that does emerge when one examines presidential elections across history is that incumbent candidates appear to have an advantage. Between 1792 and 2016, in-office parties retained the White House about two-thirds of the time when they ran the incumbent president but only won half the time—a coin flip— when they did not run an incumbent.51 Obama’s victory in 2012 made him the third straight incumbent president to win reelection. In fact eight of ten postwar incumbent

chaPter three • the election results 85

presidents seeking reelection have won, with Ford losing in 1976 and Carter in 1980. The 1976 and 1980 elections were the only successive elections in the twentieth cen- tury in which two incumbent presidents in a row lost. The only other elections in which incumbent presidents were defeated in two straight elections were in 1888, when Benjamin Harrison defeated Grover Cleveland, and in 1892, when Cleveland defeated Harrison. With no incumbent on the ballot in 2016, the presidential election was seemingly either party’s to win. In Chapter 7, we examine voters’ evaluations of presidential performance and how it relates to voting.

STATE-BY-STATE RESULTS

The modern electoral map is a conglomeration of Republican “red states” and Demo- cratic “blue states.” Yet this color pairing has no real historical meaning and, in fact, has only become convention in recent years.52 In 1976, for instance, election-night news coverage on NBC classified Republican (Ford) wins in blue and Democratic (Carter) victories in red. ABC News featured an electoral map that colored Demo- cratic states in blue and Republican states in yellow.

Whereas the colors on the electoral map may be meaningless, the political geography of presidential elections most certainly is not. Because states deliver the electoral votes necessary to win the presidency, the presidential election is effectively fifty-one separate contests, one for each state and one for the District of Columbia. With the exception of Maine and Nebraska, the candidate who wins the most votes in a state wins all of the state’s electors. Regardless of how a state decides to allocate its electors, the number of electors is the sum of its senators (two), plus the number of its representatives in the U.S. House.53 Since 1964 there have been 538 electors and a majority, 270, is required to win. In 2016 the distribution of electoral votes ranged from a low of three in Alaska, Delaware, Montana, North Dakota, South Dakota, Vermont, Wyoming, and the District of Columbia to a high of fifty-five in California.

Because each state, regardless of population, has two electoral votes for its sena- tors, the smaller states are overrepresented in the Electoral College and the larger states are underrepresented. The twenty least-populated states and the District of Columbia were home to roughly 10.5 percent of the U.S. population according to the 2010 Census, but these states had 16.5 percent of the electoral votes. The nine most- populated states, which had 52.2 percent of the population, had only 44.8 percent of the electoral vote.

Even though smaller states are overrepresented in the Electoral College, presi- dential campaigns tend to focus their resources on larger states unless pre-election polls suggest that a state is unwinnable. Consider the two most populous states, California and Texas. California’s fifty-five electoral votes represent one-fifth of the votes needed to win the Electoral College. Texas has thirty-eight electoral votes, one-seventh of the votes necessary to win. Clearly both are vital for building an Electoral College victory. Yet pre-election polls suggested landslide wins for Clinton in California and Trump in Texas, and neither campaign spent significant time or money in either state. During the general election campaign (from the end of their respective conventions to elec- tion day), neither presidential candidate made a public appearance in California, and

86 Part one • the 2016 Presidential election

Trump visited Texas only twice.54 Florida, North Carolina, Ohio, and Pennsylvania, on the other hand, were competitive, large states with a total of eighty-two electoral votes at stake, and both candidates visited the four states more than any other. Clinton made twenty-one public appearances in Florida, for instance. Trump, who spent more time on the campaign stump overall, made twenty public appearances each in Florida, North Carolina, and Pennsylvania.55 Trump won all four states.56

States are the building blocks of winning presidential coalitions, but state-by- state results can be overemphasized and may sometimes be misleading for three rea- sons. First, although much attention is given to battleground states, the nature of broadcast and social media coverage means that candidates must run national cam- paigns. Candidates can make appeals to specific states and regions, but those messages are likely to be reported across geographic boundaries. Thus, whereas battleground contests and regional bases of support may color a campaign’s message and strategy, most campaigns seek to form a broad-based coalition throughout the nation. Indeed, given that forty-four of the forty-eight elections between 1828 and 2016 have resulted in the candidate with the largest number of popular votes also winning a majority of the electoral votes, it would appear that successful campaigns have always been national rather than regional in scope.

Second, comparing state-level election results over time can be misleading and may even conceal change. To illustrate this point we compare the results of two close Democratic victories—John Kennedy’s defeat of Richard Nixon in 1960 and Jimmy Carter’s defeat of Gerald Ford in 1976—that have many similarities. In both 1960 and 1976, the Republicans did very well in the West, and both Kennedy and Carter needed southern support to win. Kennedy carried six of the eleven southern states—Arkansas, Georgia, Louisiana, North Carolina, South Carolina, and Texas—and gained five of Alabama’s eleven electoral votes, for a total of eighty-one electoral votes. Carter car- ried ten of the eleven southern states (all but Virginia) for a total of 118 electoral votes. Yet the demographic basis of Carter’s support was quite different from Kennedy’s. In 1960 only twenty-nine percent of African Americans in the South were registered to vote compared with sixty-one percent of whites. According to our analysis of the American National Election Studies, only about one in fifteen of the Kennedy voters in the South was black. In 1976, 63 percent of African Americans in the South were registered to vote compared with 68 percent of whites.57 We estimate that about one in three southerners who voted for Carter was black. A simple state-by-state compari- son would conceal this massive change in the social composition of the Democratic presidential coalition.

Third state-by-state comparisons do not tell us why a presidential candidate received support. Of course such comparisons can lead to interesting speculation, especially when the dominant issues are related to regional differences. Following the 2016 election, for example, some observers speculated that Trump’s victory, particu- larly in states such as Michigan, North Carolina, Ohio, and Pennsylvania, could be attributed to manufacturing job losses in those states.58 Yet exit polls from these states show that voters who reported that the economy was the “most important issue facing the economy” were actually more likely to vote for Clinton than Trump.59 Inference based solely on who won the state may be fallacious and lead to mischaracterizations of the electorate. Indeed similarly constructed inferences often lead to hyperbolic

chapter three • the election results 87

comparisons of “red states” versus “blue states,” creating an illusion of a deeply divided electorate.60 State-level election results should not be used to infer voters’ preferences; for this we must examine individual-level survey responses—as we do in later chapters.

With these qualifications in mind, we now turn to the state-by-state results. Figure 3-2 shows Trump’s margin of victory over Clinton in all states. As noted ear- lier both of the parties maintain regional bases of strong support, whereas relatively few states are competitive and truly in play. A continuing base of strength for the Democrats was the Northeast, sweeping eight of the nine states in the region (by an average popular vote margin of 15.3 points).61 The Democrats have dominated the Northeast in presidential elections since Bill Clinton’s election in 1992.62 But Hillary Clinton was unable to hold the region’s second-largest electoral prize, Pennsylvania, which Trump won by a slim 0.72 points. Despite the party’s electoral strength in the Northeast, the region is proving to be a precarious base of electoral support for the Democrats. Whereas comparison of vote shares across the last seven elections sug- gests that the region has not waned in its support of the Democratic Party, the region

Figure 3-2 Trump’s Margin of Victory over Clinton, 2016

Note: We classify Maine as being carried by Clinton in 2016. Clinton actually won the statewide race and one of Maine’s two congressional districts, thus winning three of the state’s four Elec- toral College votes.

NJ CT

IL OH

MI IA

TX LA

IN WV

WY ID

MS AL GA

PA OH

GA SC

PA NV

More than 20%

Between 15% and 20%

Between 10% and 15%

Between 5% and 10%

Less than 5%

Clinton win

88 Part one • the 2016 Presidential election

is declining in population and thus carries less weight in the Electoral College. In 1992 the region offered 106 electoral votes. In 2016 that number was only ninety-six, and by one estimate—based on U.S. Census population projections—by the 2024 election, the Northeast’s share of electoral votes will decline further to ninety-one.63 Whether the region continues to vote consistently for the Democratic Party remains to be seen.

Clinton’s coalition was not restricted to one region. Outside of the Northeast she carried seven other states, as well as the District of Columbia, by 10 percentages points or more, including the electorally rich states of California and Illinois. Clinton’s sizable victories in Oregon and Washington continue the Democrat’s dominance in those states, which have voted Democratic in each of the last eight presidential elec- tions. Clinton also won pluralities in Nevada and Colorado, although her victories in each did not approach the comfortable majorities secured there by Obama in both 2008 and 2012. Yet Clinton’s ability to keep these states in the Democratic column is notable because Nevada had voted Democrat only twice (for Bill Clinton in 1992 and 1996) between 1964 and 2008, and Colorado voted Democrat only once (1992) during that time frame. Clinton was also able to maintain Virginia for the Democrats. Until Obama’s victory there in 2008 and 2012, Virginia had not voted Democratic since Lyndon Johnson’s 1964 landslide, although the state had been inching toward the Democratic column because of the suburban growth from D.C. But the rest of the South remained unfertile ground for the Democrats in 2016. Clinton was only able to capture 13 of the South’s 160 electoral votes.

Clinton’s largest electoral prize came from California, where she won 61.7 per- cent of the vote and the state’s 55 electoral votes. In fact California was the only state in which Clinton’s margin of victory was bigger than Obama’s in 2012. The Golden State has now voted for the Democratic candidates in each of the last seven presidential elections. Between 1952 and 1988, the state voted for the GOP presidential candidate in nine of the ten elections. But in those elections, California did not differ much from the country as a whole: its average level of Republican support was the same as that of the nation as a whole. One reason for this political change is the state’s growing Latino population, which increased from 19 percent in 1980 to 39 percent in 2016, overtaking whites to become California’s largest ethnic group.64 According to Mark Baldassare, based on exit polls in 1990, only 4 percent of California voters were Latino; by 2000 14 percent were.65 In 2016 exit polls indicate that Latinos accounted for 31 percent of the California electorate. Of those voters 71 percent voted for Clinton and only 24 percent for Trump.66 Perhaps California’s greatest effect on the 2016 presidential election is its contribution to the disparity between the national popular vote and the Electoral College outcome. Clinton received a whopping 4.3 million more votes than Trump in California. That surplus of votes alone is greater than that received by Trump in his seven most decisive victories combined.67 If California were taken out of the vote tally, Trump would have won the national popular vote by 1.4 million votes.68 Thus California’s place in the 2016 election helps demonstrate the logic inherent in the Electoral College, an institution designed by the founders to prevent regional candidates from dominating a national election.69

As noted previously Trump’s narrow victories in Michigan, Pennsylvania, and Wisconsin—all by less than one percentage point—brought three reliably Democratic states back into the Republican’s win column. All three of these states

chaPter three • the election results 89

supported Reagan in both 1980 and 1984, and Michigan and Pennsylvania sided with George H. W. Bush in 1988. In the six elections that followed, all three states voted for the Democratic nominee, although not always by comfortable margins. These states represented a substantial haul for Trump, totaling 46 electoral votes (17 per- cent of the 270 needed for victory). Other states that Trump was able to win, such as Florida and Ohio, are far more variable electorally. Since, 1992, for instance, neither state has voted for a presidential candidate of the same party more than twice in a row. Unsurprisingly, these states are typically defined as among the most competitive from election to election. And Ohio continues to be a bellwether for the nation; the Buckeye State has voted for the presidential (Electoral College) winner in every elec- tion since 1960, when it voted for Richard Nixon over John F. Kennedy.

The remaining states captured by Trump in 2016 are the core of the Republican electoral coalition. These twenty-two states, which Trump won by an amazing aver- age margin of 23.6 percentage points, all voted Republican in each of the last five presidential elections.70 Thirteen of these states (Alabama, Arkansas, Idaho, Kansas, Mississippi, Nebraska, North Dakota, Oklahoma, South Carolina, South Dakota, Utah, and Wyoming) have voted Republican in every election dating back to 1980. The problem for Republicans, generally, is that the twenty-two solidly Republican states tend to be smaller in population and result in 180 electoral votes. Texas, with its thirty-eight electoral votes, is the largest state in this coalition, followed by Georgia with sixteen electoral votes, but five of these states have only three electoral votes each. To make the point clearer, consider the subset of thirteen states that have voted Republican since 1980. These thirteen states cumulatively represent sixty-seven elec- toral votes. This is not much of an advantage when we recall that Democratic-leaning California alone has fifty-five votes.

The region that offered Trump his greatest electoral reward was the South. One hundred forty-seven of Trump’s electoral votes were from the South, nearly half of his total (48.4 percent). In the last half century, the South has been transformed into the base of the Republican Party, and this transformation is the most dramatic change in postwar American politics. Clinton’s only victory in the South was in Virginia. In 2008 and 2012 Obama cut into the Republican’s southern base by winning both Florida and Virginia, and in 2008, he also carried North Carolina. In each of those elections, however, Obama would have won the Electoral College without winning a single southern state. The same is true of Bill Clinton’s victories in 1992 and 1996. Yet it is infeasible for the Republicans to win the presidency without southern electoral votes.

Republican strength in the South and Democratic advantage in the Northeast does not mean that sectionalism has beset the country. Indeed regional differences in presidential voting have declined in the postwar period and are currently low by histor- ical standards. This can be demonstrated by statistical analysis. Joseph A. Schlesinger has analyzed state-by-state variation in presidential elections from 1832 through 1988, and we have updated his analyses through 2016.71 Schlesinger measures the extent to which party competition in presidential elections is divided along geographic lines by calculating the standard deviation in the percentage of the Democratic vote among the states.72 The state-by-state variation was 10.35 in 2016, slightly higher than the 10.29 deviation in 2012 and the 9.54 deviation in 2008. This suggests that states were slightly more divided in their support for Trump in 2016 than in recent elections.73

90 Part one • the 2016 Presidential election

Schlesinger’s analysis clearly reveals the relatively low level of state-by-state varia- tion in the postwar elections.74 According to his analysis (as updated), all fifteen of the presidential elections from 1888 to 1944 displayed more state-by-state variation than any of the seventeen postwar elections. To a large extent, the decline in state-by- state variation has been a result of the transformation of the South and the demise of local party machines, which has allowed partisan cleavages to become more consistent across states and allowed party competition to increase across the country.75

ELECTORAL CHANGE IN THE POSTWAR SOUTH

The South is a growing region that has undergone dramatic political change. Even though five of the eleven southern states have lost congressional representation since World War II, Florida and Texas have made spectacular gains. In the 1944 and 1948 elections, Florida had only eight electoral votes, but in the 2016 election, it had twenty-nine. In 1944 and 1948, Texas had twenty-three electoral votes; in 2016 it had thirty-eight. Since the end of World War II, the South’s electoral vote total has grown from 127 to 160. The South gained seven electoral votes following the 2010 Census and projections suggest that it may gain an extra four by 2024.76

The political transformation of the South was a complex process, but the major reason for the change was simple. As V. O. Key Jr. brilliantly demonstrated in Southern Politics in State and Nation in 1949, the main factor in southern politics is race. “In its grand outlines the politics of the South revolves around the position of the Negro. . . . Whatever phase of the southern political process one seeks to understand, sooner or later the trail of inquiry leads to the Negro.”77 And it was the national Democratic Party’s sponsorship of African American civil rights that shattered the party’s domi- nance in the South.78

Between the end of Reconstruction in 1877 and the end of World War II, the South was functionally a one-party system. Unified in its support of racial segrega- tion and in its opposition to Republican social and economic policies, the South was a Democratic stronghold—the “Solid South.” Indeed in fifteen of the seventeen elec- tions from 1880 to 1944, all eleven southern states voted Democratic. Between 1896 (the first election after many southern states adopted the “white primary”) and 1944, the average Democratic Party vote share in presidential elections was 71.6 percent.79 The only major defections were in 1928, when the Democrats ran Alfred E. Smith, a Roman Catholic. As a result the Republican candidate, Herbert Hoover, won five southern states. Even then six of the most solid southern states voted for Smith, even though all but Louisiana were overwhelmingly Protestant.

After Reconstruction ended in 1877, many southern blacks were prevented from voting, and in the late nineteenth and early twentieth centuries, several southern states changed their voting laws to further disfranchise blacks. The Republicans effec- tively ceded those states to the Democrats. Although the Republicans garnered black support in the North, they did not attempt to enforce the Fifteenth Amendment, which bans restrictions on voting on the basis of “race, color, or previous condition of servitude.”

chaPter three • the election results 91

In 1932 a majority of African Americans in the North remained loyal to the Republicans; although by 1936 Franklin D. Roosevelt had won the support of north- ern blacks. But Roosevelt made no effort to win the support of southern blacks, most of whom remained disfranchised. Even as late as 1940, about 70 percent of the nation’s blacks lived in the states of the old Confederacy. Roosevelt carried all eleven of these states in each of his four victories. His 1944 reelection, however, was the last contest in which Democrats carried all eleven southern states.

World War I led to massive migration of African Americans from the agrar- ian South and into the industrial North, where—given the absence of laws restrict- ing their suffrage—many would enjoy the franchise for the first time. The influx of African Americans alarmed some Democratic politicians in the North, who would likely see their electoral prospects decline unless they were able to siphon a share of African American voters who were loyal to the party of Lincoln. In 1932 African American voters in most major cities in the North voted for Herbert Hoover by a roughly two-to-one margin.80 To appeal to African American voters, many northern Democrats encouraged their party to adopt a supportive position toward civil rights. By 1948 President Harry Truman was making explicit appeals to blacks through his Fair Employment Practices Commission, and in July 1948 he issued an executive order ending segregation in the armed services.81 These policies led to defections from the “Dixiecrats” and cost Truman four southern states in the 1948 election; he still won the seven remaining southern states by an average margin of 26.2 points. In 1952 and 1956 the Democratic candidate, Adlai E. Stevenson, de-emphasized appeals to blacks, although his opponent, Dwight Eisenhower, still made inroads in the South. In 1960 Kennedy also played down appeals to African Americans, and southern electoral votes were crucial to his win over Nixon.82 Kennedy also strengthened his campaign in the South by choosing a Texan, Lyndon Johnson, as his running mate. Clearly Johnson helped Kennedy win Texas, which he carried by only two percentage points.

If Johnson as running mate aided the Democrats in the South, Johnson as presi- dent played a different role. His explicit appeals to African Americans, including lead- ing the Civil Rights Act into law in 1964, helped end Democratic dominance in the South. Barry Goldwater, the Republican candidate, had voted against the Civil Rights Act as a member of the Senate, creating a sharp difference between the two candidates. Goldwater carried all five states in the Deep South.83 The only other state he won was his native Arizona. In 1968 Hubert Humphrey, who had long championed black equality, carried only one southern state, Texas, which he won with only 41 percent of the vote. He was probably aided by George Wallace’s third-party candidacy because Wallace, a segregationist, won 19 percent of the Texas vote. Wallace carried Alabama, Arkansas, Georgia, Louisiana, and Mississippi, while Nixon carried the remaining five southern states. Nixon won every southern state in 1972, and his margin of victory was greater in the South than in the rest of the nation. Although Carter won ten of the eleven southern states in 1976 (all but Virginia), he carried a minority of the vote among white southerners.

In 1980 Reagan won every southern state except Georgia, Carter’s home state. In his 1984 reelection victory, Reagan carried all the southern states, and his margin of vic- tory in the South was greater than his margin outside it. In 1988 George H. W. Bush

92 Part one • the 2016 Presidential election

was victorious in all eleven southern states, and the South was his strongest region. Four years later, in 1992, Clinton, a native of Arkansas, made some inroads in the South and somewhat greater inroads in 1996. All the same the South was the only predominantly Republican region in 1992, and in 1996 Bob Dole won a majority of the electoral vote only in the South and mountain states. In 2000 the South was the only region in which Bush carried every state, and more than half of his electoral votes came from that region. Bush again carried every southern state in 2004, along with all of the states in the Mountain West. As was the case four years earlier, more than half of his electoral votes came from the states of the old Confederacy. Despite slippage in 2008 and 2012, Republicans have won every southern state in five of the twelve elections (1972, 1984, 1988, 2000, and 2004) between 1972 and 2016.

Although the transformation of the South is clearly the most dramatic change in postwar American politics, the 2016 election underscores that the Republicans do not hold the same level of dominance in the region that the Democrats once enjoyed. The average Republican vote share in the South between 1972 and 2016 was 54.1 points—much smaller than the 71.6 vote share that we reported earlier for the Democrats from 1896 to 1944. Florida is highly competitive. Bill Clinton won the Sunshine State in 1996, and in 2000 George W. Bush carried the disputed con- test by a negligible margin. Obama narrowly won Florida in both 2008 and 2012, and Trump won the state by 1.2 percentage points. Trump won North Carolina by 3.7 points, but it too remains competitive. The state narrowly voted for Obama—by a 0.3 percentage-point margin—in 2008 and for Romney—by a slim two-point mar- gin—in 2012. Although Virginia did not vote Democratic between 1968 and 2004, the growing number of suburbanites in northern Virginia has made the state more com- petitive, and the Democrats have now captured the state in three straight elections. Even in Georgia Democrats see Atlanta and its close-in suburbs as fertile ground to make the state competitive.84 Clinton carried Georgia in 1992, and McCain won by only 5.2 percentage points in 2008. Republicans have an advantage in the South, to be sure, but Democrats are competitive in few a southern states, thus allowing the South to keep its place of prominence in modern presidential politics.

Some scholars predict that the South will play a part in the next major transfor- mation in American politics, one they argue could make the Electoral College less competitive. John Judis and Ruy Teixeira contend that shifting demographics, specifi- cally a growing professional class and an increase in America’s nonwhite population, are setting the stage for an “emerging Democratic majority.”85 Central to this argu- ment is that in the next two decades, the proportion of Latinos in the electorate is likely to double.86 Latino growth in the South, where African Americans already com- pose a large share of the electorate, could greatly benefit the Democrats. Three south- ern states—Arkansas, North Carolina, and South Carolina—were among the top five states in Hispanic population growth between 2000 and 2010 according to the U.S. Census.87 And Texas (19 percent) and Florida (8 percent) already have the second- and third-largest Hispanic populations, respectively. It is assumed by Judis and Teixeira that further growth in the Latino population could make Texas a Democratic-leaning state and Florida a safe Democratic state by the 2024 election cycle, giving Democrats, who already hold advantages in the electorally rich states of California and New York, an easier path to victory in the Electoral College.88

chaPter three • the election results 93

We have heard predictions of impending electoral realignment before, and as in the past, we encourage caution when evaluating these claims.89 As noted earlier, after his reelection in 1984, Ronald Reagan proclaimed that his victory represented a Republican realignment. Indeed some scholars went so far as to argue that the Republicans held an electoral vote “lock.”90 But the Democrats won two consecu- tive elections in the 1990s. The scenario outlined by Judis and Teixeira offers rea- son for optimism for the Democrats and pessimism for the Republicans. Yet there are two major assumptions undergirding this scenario that complicate things for the Democrats. First, it should not be assumed that Latino voting participation will increase proportionately with Latino population growth; it has not thus far. As noted in a Pew Research Center report, the “[n]umber of Latino eligible voters is increas- ing faster than the number of Latin voters in presidential election years.”91 Hispanics were 24 percent of Florida’s population but only 18 percent of voters on election day. In Texas Hispanics were 39 percent of the population but only 24 percent of the electorate.92 For Democrats to make real substantial gains in the near future, voter participation among Latinos—particularly those in Texas—must grow at a faster rate to become commensurate with the group’s share of the population. Second, it should not be assumed that Latinos will continue to support Democrats at the same levels. The Latino vote is not monolithic and has changed somewhat over time (we will have more to say about Latino political preferences in Chapter 8). For instance, Cuban Americans in Florida, many of whom fled their homeland to escape Fidel Castro’s dic- tatorship, have long been a reliable voting bloc for Republicans. “In Florida, Cubans were about twice as likely as non-Cuban Latinos to vote for Donald Trump” in 2016.93 But recent evidence shows that second- and third-generation Cuban Americans are more liberal and more likely to vote Democratic than the elder generation.94 This is good news for Democrats, of course, but it also serves to remind both parties that old loyalties are not easily maintained and that voters respond to changing issues and interests, not simply on the basis of ethnicity. However, this battle for Latino votes plays out, it appears the South will be the focus of both parties’ attention for many elections to come.

THE ELECTORAL VOTE BALANCE

Elections often conclude with the winning party making hyperbolic claims of electoral mandates and partisan realignments. However, as we have seen, comparing presiden- tial elections results over time suggest that party competition for the presidency is high in the postwar period. Today’s presidential elections are national in scope, and the Electoral College provides no significant barrier to either political party. Moreover, Andrew Gelman, Jonathan N. Katz, and Gary King present compelling evidence that since the 1950s partisan biases created by the Electoral College are negligible.95

Despite Trump’s victory in 2016, short-term factors suggest that Democrats are cur- rently advantaged. Since 1988 the Democrats have won the presidency in four of seven elections and the popular vote in six out of the seven. But the recent past is not always a guide to the future. Consider the fact that in the six elections between 1968 and 1988, the Republicans held the advantage, winning five of the six and several by significant

94 Part one • the 2016 Presidential election

margins.96 Republican strength in the 1980s was soon replaced by Democratic victories in the 1990s. Yet because competition for the presidency has always rested upon some assessment of a candidate’s relative strength in each of the states, recent election results often guide how parties develop future electoral strategies.97 In Figure 3-3, we illustrate how the states have voted in each of the last five elections.

Figure 3-3 creates an illusion of Republican dominance. Twenty-two states have voted Republican in each of the past five elections—an impressive number, to be sure. But the geographic expanse of these Republican states belies their electoral power. These solidly Republican states only tally 180 electoral votes (33 percent of all avail- able electoral votes). Texas with thirty-eight electoral votes is the only large state among the twenty-two, and five of these states have the minimum of three votes. Two other states, Indiana and North Carolina, have voted Republican in four of the five elections, accounting for twenty-six electoral votes. Both states voted for Obama in 2008. The Republicans have won Florida and Ohio (forty-seven electoral votes com- bined) in three of the last five elections, but both voted for Obama in 2008 and 2012. In total these twenty-six “Republican states” represent 253 electoral votes, seventeen votes shy of the 270 needed to win the Electoral College.

Figure 3-3 Results of the 2000, 2004, 2008, 2012, and 2016 Presidential Elections

Source: Compiled by authors.

IN DE

NJ CT

MI ID

KSCA

NV NE

TN NC

NHVT

WY .

NV IA

OH PA

Five Republican wins

Four Republican wins; one Democratic win

Five Democratic wins

One Republican win; four Democratic wins

Two Republican wins; three Democratic wins

Three Republican wins; two Democratic wins

chaPter three • the election results 95

This leaves Democrats with twenty-four states and the District of Columbia, which total 285 electoral votes, exceeding the majority needed. Of these, fifteen states and the District of Columbia have voted Democratic in five straight elections. Although seemingly less impressive than the twenty-two states held comfortably by the Republicans, the solidly Democratic group represents 285 electoral votes (53 percent of all the available electoral votes). Five states, Michigan, New Hampshire, New Mexico, Pennsylvania, and Wisconsin (55 electoral votes collec- tively) have voted Democratic in four of the five elections. And four states, Colorado, Iowa, Nevada, and Virginia (34 votes in sum) have voted Democratic on three of the last five occasions.

If we were solely to use these five elections as the basis for determining battle- ground states for 2020, it would be a narrow field of play. Only six states fall into the “three-out-of-five” categories: Colorado, Florida, Iowa, Nevada, Ohio, and Virginia (three voted for Clinton; three voted for Trump in 2016). In this hypotheti- cal, the Democratic candidate in 2020 would only have to win Florida to win the Electoral College. Indeed with 251 electoral votes on the safer side of their ledger, the Democrats could reach 270 electoral votes with a wider variety of coalitions than the Republicans. The Republicans, on the other hand, start with only 206 electoral votes in the safe column and cannot create a winning electoral coalition without Florida. Even with Florida in the win column, the Republicans would need a mini- mum of three additional states in their coalition to reach 270. Of course the odds of a Republican victory were just as long in 2016.

Undoubtedly the current electoral map presents a challenge for the GOP. But this is not the first time in recent history that the Electoral College map has appeared so uninviting for one of the parties and generally uncompetitive. Figure 3-4 illustrates over- time changes in the electoral balance of the Electoral College. Similar to Figure 3-3, we calculated how each of the states voted in the prior five elections but did so for each election from 1988 to 2016.98 We then categorized states that voted for the same party in three of the five previous elections as “highly competitive.” States that voted for the same party in four of the five elections were labeled “sometimes com- petitive,” and those that voted for the same party in each of the five were labeled “uncompetitive.”99

The figure shows that the number of uncompetitive states has more than doubled since the 2000 election. Thirty-eight of the fifty states currently appear to be uncom- petitive, having voted for one of the two parties consistently over each of the last five elections. This rise coincides with a decline in the number of highly competitive states, which as noted earlier is now down to six (in 1996 the number of highly competitive states was twenty-six!). This alone would suggest a lack of electoral competition. But perhaps the clearest message conveyed by the figure is that the current electoral vote balance is likely subject to change . . . indeed, perhaps rapid change. The competi- tive balance in 1988 was similar to what we see today. Following that election, only eight states were considered highly competitive, and twenty-six were uncompetitive. Yet the challenge confronting Democrats following the 1988 elections was arguably more daunting than that confronting Republicans today. Whereas there are thirty- eight uncompetitive states following the 2016 elections, sixteen are Democratic and twenty-two are Republican states. In 1988, however, twenty-five of the twenty-six

96 Part one • the 2016 Presidential election

uncompetitive states were Republican states. Based on the electoral vote balance alone, no one could have reasonably predicted Democratic victories in 1992 and 1996.

To assess the future prospects of the major parties, we must go beyond analyzing official election statistics. Although these statistics are useful for understanding com- petitive outcomes and future party strategies, they tell us little about why Americans vote the way they do or whether they vote at all. To understand how social coalitions have changed over time, as well as the issue preferences of the electorate, we must turn to surveys. They reveal how Trump was able to win election despite the public’s relatively positive assessment of the Obama presidency. Furthermore, to determine the extent to which Americans really are polarized along party lines, we must study surveys to examine the way in which the basic party loyalties of the American elector- ate have changed during the postwar period. Thus, in the next five chapters, our study turns to survey data to examine the prospects for continuity and change in American electoral politics.

Figure 3-4 Competitive Balance in the Electoral College, 1988–2016

1988 1992 1996 2000 Election Year

N um

be r

of S

ta te

2004 2008 2012 2016

Highly Competitive Sometimes Competitive Uncompetitive

Source: Calculated by the authors.

ALDRICH_CHANGE REV_I Proof (1)_Part1
ALDRICH_CHANGE REV_I Proof (1)_Part2
ALDRICH_CHANGE REV_I Proof (1)_Part3
ALDRICH_CHANGE REV_I Proof (1)_Part4
3