Eternal Bandwagon: The Politics of Presidential Selection [1st ed.] 9783030517984, 9783030517991

Table of contents :
Front Matter ....Pages i-xviii
The Bandwagon Dynamic and the Causal Funnel: Institutional Dynamics and Nominating Politics (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 1-35
Deep Channels to a Presidency: Occupations and Careers (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 37-84
Factions, Constituencies, and Candidates: The Democrats (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 85-120
Factions, Constituencies, and Candidates: The Republicans (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 121-157
The Usual Suspects: Rules for the Nominating Contest (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 159-192
The Usual Suspects: Strategies and Sequences (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 193-239
Eternal Bandwagon: Nominating Politics and the General Election (Byron E. Shafer, Elizabeth M. Sawyer)....Pages 241-266
Back Matter ....Pages 267-273

Citation preview

THE EVOLVING AMERICAN PRESIDENCY

Eternal Bandwagon The Politics of Presidential Selection Byron E. Shafer Elizabeth M. Sawyer

The Evolving American Presidency

Series Editors Michael A. Genovese Loyola Marymount University Los Angeles, CA, USA Todd L. Belt Graduate School of Political Management George Washington University Washington, DC, USA

This series is stimulated by the clash between the presidency as invented and the presidency as it has developed. Over time, the presidency has evolved and grown in power, expectations, responsibilities, and authority. Adding to the power of the presidency have been wars, crises, depressions, industrialization. The importance and power of the modern presidency makes understanding it so vital. How presidents resolve challenges and paradoxes of high expectations with limited constitutional resources is the central issue in modern governance and the central theme of this book series.

More information about this series at http://www.palgrave.com/gp/series/14437

Byron E. Shafer · Elizabeth M. Sawyer

Eternal Bandwagon The Politics of Presidential Selection

Byron E. Shafer Department of Political Science University of Wisconsin–Madison Madison, WI, USA

Elizabeth M. Sawyer Department of Political Science University of Wisconsin–Madison Madison, WI, USA

The Evolving American Presidency ISBN 978-3-030-51798-4 ISBN 978-3-030-51799-1 (eBook) https://doi.org/10.1007/978-3-030-51799-1 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Cover image: © jamtoons/DigitalVision Vectors/Getty Image This Palgrave Macmillan imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To Carol McBride, H. Lee Sawyer, Jr., and Wanda Shafer, for their many and varied contributions.

Preface

Brave New World or Latest Twists on a Familiar Narrative? Presidential nominating contests are extremely fertile ground for theorizing about alternative outcomes in presidential selection, and the struggles of 2016 provided a particularly rich example. The Democrats generated their share of fanciful options. Perhaps Bernie Sanders could indeed surpass Hillary Clinton for the Democratic nomination if only this or that event happened. Or perhaps he could do so if only this or that state contributed a surprise. Yet the real—and really fanciful—hypotheticals in this superficially anomalous year were generated by the Republicans. Surely the polling front-runner as Christmas turned to New Years, Donald Trump, would fold in the face of the “real candidates”—Jeb Bush, Marco Rubio, Ted Cruz, or a raft of further hopeful replacements. Perhaps Trump would instead continue his poll-based rise into the early contests, forcing the rest of the party to consolidate around the eventual nominee. Or perhaps no one would consolidate a delegate majority, producing the first real nominating convention in forty years. The practical point, of course, is that none of that happened. Yet the practical collapse of these and so many other hypotheticals—pick your counterfactual favorite—does raise some insistent theoretical questions in its aftermath, and only more so with another quadrennial contest under way as this is written. In the narrowest of terms, were the nominating

vii

viii

PREFACE

dynamics of 2016 ever very different from those of previous years? More broadly and leading on to everything that follows, how far can we go in setting out a generalized picture of nominating politics? Such a picture needs to begin with an underlying and continuing—and in that sense, inherent—nominating dynamic. This dynamic then has to contribute a more informed understanding of the seedbeds, that is, the occupational backgrounds and career trajectories, which produce ultimate nominees. That generalized picture must next tease out the key structural influences that shape an actual choice among the specific aspirants who emerge from these political seedbeds. Lastly, the changing formal rules of the game, along with changing strategies among the major players—the focus of so many analysts—have to be folded into an emerging composite picture. Elucidating this picture is the purpose of everything that follows. Such an enterprise is often summarized as a quest for the fundamentals versus the ephemerals.1 The former are presumably deep-seated and recurrent. The latter are instead transient and idiosyncratic. Without dissenting in any way from that particular dichotomy, we shall recast its way of thinking by means of another familiar metaphor, the funnel of causality.2 At the broad end of this funnel but farthest from the ultimate outcome, there are deep-seated and recurrent influences, presumably the fundamentals. They, not rules reforms or candidate strategies, are what winnows an abstractly limitless field on the way to a singular nomination. Though at the narrow end of the funnel and the one most proximate to a specific nomination, the same metaphor allows room for the personal idiosyncrasies, state peculiarities, and/or strategic interactions which might still determine an individual nomination in those (presumably rare) cases where the fundamentals leave room for an ephemeral impact. In their time, the nominating contests of 2016 were a bonanza for those on both sides of the partisan aisle who believe essentially that there are no fundamentals, certainly none sufficient to discipline nominating politics in a recurrent fashion. To be fair, many of those who are charged with interpreting any given nominating contest as it unfolds are almost compelled to believe in a portentous idiosyncrasy. If there is instead an inherent and continuing dynamic to nominating politics, if it strongly favors particular occupations and careers at a given point in time, and if a small set of structural influences really makes crucial choices within the resulting field of contenders, then most individual options for nominating politics are in effect eliminated before the candidates assemble, and certainly before the delegates are chosen. All else qualifies only as

PREFACE

ix

ephemeral inputs to long-established dynamics and patterns, with limited interpretive value. So Chapter 1 must isolate the long-running contours to nominating politics, contours that add up to what is known colloquially as “the bandwagon,” a dynamic that can in fact be perceived for most of American history. This bandwagonbandwagon is easiest to see in the Republican Party, surfacing in Republican nominating contests for the entire lifetime of the party. But it extends easily to those parties that were Republican predecessors. And it encompasses Democratic nominating contests from 1936 onward, where the explanation for why it cannot reach farther back is a crucial part of the evidence confirming the very nature of the bandwagon itself. In that sense, Chapter 1 must also be more abstract than the chapters that succeed it, since it must reduce all those contests to one common, dominant, and continuing dynamic. Yet if the operations of such a bandwagon can indeed be isolated, then Chapter 2 must move immediately to consider the occupational and career seedbeds generating the candidacies that unfold within this bandwagon dynamic, a task that is more specific to particular points in time, with a product that is always more changeable. The universe of individuals who might conceivably run for president at any given moment remains immense, a fact underlined once again by the huge field of candidates who announced that they were running for the Democratic nomination of 2020, though the universe of those who had credentials equal to many of these announced candidates was of course larger still. Yet each timeperiod favors a small set of occupations and careers that provide the critical seedbed of those candidacies with a reasonable chance of producing a nomination, thanks to a set of resources that are crucial to coalition building inside the two major parties at that point. These aspiring candidates, along with many of those who report on their progress, almost inescapably want to believe that it is either candidate virtues or candidate strategies which produce an eventual nomination. But in fact, a small set of structural influences, deriving most centrally from factional alignments and factional struggles inside the two major parties, actually shapes the choice among contenders—not the behavior of the candidates themselves, nor of the strategists who work for them, nor of the media that cover them, nor even of the party leaders or party voters who opt for one candidate rather than another. Chapter 3 looks at these key influences for the Democratic Party across time. Chapter 4 turns to these influences for the Republican Party (and its predecessors).

x

PREFACE

All that said, the more usual way to organize an inquiry into the politics of presidential nomination begins at the opposite end of the causal funnel, with the formal rules of the nominating contest and the strategic behavior of its candidate contenders. So Chapter 5 doubles back to address major rules, most especially those involving the apportionment of convention delegates to the states and the allocation of those delegates to presidential aspirants. Just as Chapter 6 moves on to examine the strategies of the candidates as they try to secure a nomination, in interaction with the strategies of the states as they attempt to maximize their own clout in nominating politics. The politics of ultimate presidential election, the politics unfolding after the major-party candidates have been nominated, is widely understood to differ from the politics of presidential nomination in major ways: in the contours of their politicking, in the strictures on their strategies, and often in the policy issues at their center.3 But there is one last aspect of nominating politics that can reach into this general election campaign in important ways as well. This involves the impact on the general election of a prior nominating politics, the one producing what has become the major-party candidates for president. Some of this follows from the nature of the successful nominating campaign itself. More of it follows from the aftermath of those nominating campaigns that did not produce the desired result. So Chapter 7 closes an examination of the politics of presidential nomination by looking at these after-burns from the nominating struggle. Madison, USA

Byron E. Shafer Elizabeth M. Sawyer

Notes 1. A marvelously provocative invocation of these notions is John Zaller, “Monica Lewinsky’s Contribution to Political Science,” PS: Political Science and Politics 31(1998), 182–189. An extended deployment of these notions is Richard Johnston, Michael G. Hagen, and Kathleen Hall Jamieson, The 2000 Presidential Election and the Foundations of Party Politics (Cambridge: Cambridge University Press, 2004). 2. The best-known application for students of American politics is probably Angus Campbell, Philip E. Converse, Warren E. Miller, and Donald E. Stokes, The American Voter (New York: John Wiley, 1960). An inquiry into the concept on its own terms is Matt Wilder, “Whither the Funnel

PREFACE

xi

of Causality,” Canadian Journal of Political Science 49 (December 2016), 721–741. 3. The godfather of all studies that consider nomination and election jointly is probably Nelson W. Polsby, Aaron Wildavsky, Steven E. Schier, and David A. Hopkins, Presidential Elections: Strategies and Structures of American Politics, 15th ed. (Landover: Rowman & Littlefield, 2019). Related successors include Stephen J. Wayne, The Road to the White House 2020, 10th ed. (Boston: Cengage Learning, 2015), and John Sides, Daron Shaw, Matt Grossman, and Keena Lipsitz, Campaigns and Elections, 2nd ed. (New York: W. W. Norton, 2015).

Praise for Eternal Bandwagon

“Shafer and Sawyer have given us the most comprehensive account of presidential nomination in decades. It is theoretically astute, rich in historical detail, and masterfully synthetic, drawing on almost 200 years of experience with a keen eye for what is particular to a specific era, party, or set of rules and what is, as the authors put it, eternal.” —Larry M. Bartels, Professor, Vanderbilt University, USA “What sets this study apart from virtually everything else that has recently been written about presidential nominations is the remarkable scope of the inquiry. Where most recent work on this topic has been about one election or one component of the nomination process, Shafer and Sawyer’s book seeks to examine every major component of the process, for the entire period from 1832 to the present, and organizes those components according to the point at which they enter the ‘causal funnel.’ This book raises the bar for all future studies of the presidential nomination process.” —William G. Mayer, Professor of Political Science, Northeastern University, USA “Many books on the politics of the presidential nomination process focus on the modern era in which candidates vie for the nomination by competing in a series of state-by-state contests. This book offers an important new take by linking the modern era to the past. Through an original analysis of nearly 200 years of nomination contests, Shafer and Sawyer xiii

xiv

PRAISE FOR ETERNAL BANDWAGON

show how a dominant nominating dynamic runs through them all. Most critically, they demonstrate how factional alignments and struggles within the two major parties shape the choice among contenders. The authors argue that it is these deep structural influences—not the candidates themselves, the media who cover them, or the decisions of party leaders—that provide the most insight into the logic of the presidential nomination process.” —Amber Wichowsky, Associate Professor, Marquette University, USA

Contents

1

2

The Bandwagon Dynamic and the Causal Funnel: Institutional Dynamics and Nominating Politics

1

Deep Channels to a Presidency: Occupations and Careers

37

Factions, Constituencies, and Candidates: The Democrats

85

Factions, Constituencies, and Candidates: The Republicans

121

5

The Usual Suspects: Rules for the Nominating Contest

159

6

The Usual Suspects: Strategies and Sequences

193

7

Eternal Bandwagon: Nominating Politics and the General Election

241

3

4

Index

267

xv

List of Tables

Table 1.1 Table 1.2 Table 1.3 Table 1.4 Table 1.5 Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 2.5 Table 2.6 Table 2.7 Table 2.8 Table 2.9 Table 5.1

Gain-Deficit calculations for 2016 Delegate totals and Gain-Deficit Ratios: the Republicans, 1860 Gain-Deficit Ratios: the Republicans, 1920 Gain-Deficit Ratios for major-party nominees across time The Democratic nomination of 1912 Preceding occupations of presidential nominees Preceding occupations of presidential nominees: the pre-convention era, 1789–1828 Preceding occupations of presidential nominees: the antebellum era, 1832–1860 Preceding occupations of presidential nominees, by party Preceding occupations of presidential nominees: the postbellum era, 1864–1908 Preceding occupations of presidential nominees: the mixed system, 1912–1968 Preceding occupations of presidential nominees: the plebiscitary system, 1972–2020 Fortunes of the vice presidents who became president Fortunes of sitting presidents Delegation shares at the Republican convention of 1916: the impact of apportionment reform

9 11 13 18 27 41 44 46 48 52 57 63 70 75 166

xvii

xviii

LIST OF TABLES

Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9 Table 6.1 Table 6.2 Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7 Table 6.8 Table 6.9 Table 6.10 Table 6.11 Table Table Table Table Table Table

7.1 7.2 7.3 7.4 7.5 7.6

Candidate votes after reapportionment: the Republican nomination of 1916 Delegation shares at the Republican convention of 1952: apportionment reform revisited Delegation shares at the Republican convention of 1976: the declining impact of apportionment reform State delegation shares at the Democratic convention of 1964: the impact of apportionment reform Delegation shares at the Democratic convention of 1968: the impact of apportionment reform Delegation shares at the Democratic convention of 1972: the impact of apportionment reform Candidate votes after abolition of the unit rule: the Democrats, 1968 The timing of nominations Nominating politics among incumbent presidents seeking reelection Nominating politics among nonincumbents Iowa, New Hampshire, and the nomination: 1976–2016 Number of presidential primaries across time A full plebiscitary picture: The changing matrix of delegate selection State size by institutional form Participation levels by institutional form in 2016 Clustering in the sequence of contests: The largest and second-largest clusters by year Opening contests: Iowa and New Hampshire in the plebiscitary era Super Tuesday and the resolution of nominating politics Strategic states and Super Tuesday: The role of the big states Uncontested renominations of sitting presidents Contested renominations of sitting presidents Consensual nominations Front-runner victories Extended contests Gain-Deficit calculations for 2020

167 168 169 170 171 172 180 184 199 201 211 222 223 224 225 227 229 231 233 244 245 249 253 254 261

CHAPTER 1

The Bandwagon Dynamic and the Causal Funnel: Institutional Dynamics and Nominating Politics

The traditional approach to the analysis of nominating politics in the United States, and a very long tradition at that, was to focus on the specifics of individual contests. These were judged to be consequential in their own right. After all, they were what gave us our presidents. Beyond that, individual contests could also be studied comparatively, usually with the most recent version interpreted by comparison to various predecessors. A second, more contemporary approach sought to move toward generalization while still retaining the specifics of contemporary contests. Though implicitly, this remained an approach valorizing individual contenders, distinctive strategies, and the interactions of both with institutional strictures and campaign practices.1 In the 1980s, a third approach to understanding the politics of presidential nomination began to appear, likewise comparative as among individual contests but seeking to work at what we are calling the opposite end of the causal funnel. To that end, it sought to generalize either across a greater temporal span or with a greater focus on theoretical explanations or, of course, both. By itself, a longer time span would contribute to explanations that were more fundamental and less idiosyncratic. By definition, an attempt to isolate a small set of shaping influences and apply them to multiple contests would contribute to those same ends. This approach did not yet dub itself a separate school of thought, and it certainly did not go on to use its results to organize and “discipline” the products of © The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_1

1

2

B. E. SHAFER AND E. M. SAWYER

previous research, but it was incipiently foundational for any effort to do precisely that. In a sense, then, everything in this chapter is an attempt to take that next step, searching for evidence of a long-running bandwagon effect in presidential nominating politics, coupling this with some strong theorizing about how this bandwagon might have developed and why it rolled so consistently across time, and using the result to gather, discipline, and reorganize an established way of thinking about presidential nominations. To that end, this chapter needs to reach back and resurrect that first set of initial (and critical) contributions toward a different way of thinking about nominating politics. It needs to drive their insights as far across American history as the course of practical politics will allow. And having teased out a long-running commonality to the politics of presidential nomination, it needs to turn around and search for limits, institutional or temporal, on this long-running essence of nominating politics.2

The Bandwagon Dynamic Two landmarks in the literature of presidential nominations played an outsized role in the new effort to understand those nominations more systematically. So they provide an essential introduction to the general contours of nominating politics as these have come to be understood, especially when the effort is to organize a comprehensive approach to nominating politics that can move on from this understanding, separate fundamentals from ephemera, and align both from the broad and distant to the narrow and proximate ends of a funnel of causality. The first landmark was a journal article from Donald S. Collat, Stanley Kelley, Jr., and Ronald Rogowski on “The End Game in Presidential Nominations.”3 The second was a book-length treatment from Larry M. Bartels on Presidential Primaries and the Dynamics of Public Choice.4 The most remarkably condensed summary of a grand dynamic underpinning the politics of presidential nomination—and hence the inescapable evidence not just of its existence but of its power—was gathered by Collat, Kelley, and Rogowski in “The End Game in Presidential Nominations.” Looking at the majority-rule nominating contests from 1848, effectively the point by which national party conventions with detailed records were institutionalized, up to 1980, the most recent contest before their own publication date, these authors searched

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

3

for—and found—evidence that a common set of shared calculations, contributing a common dynamic, governed every one: This study shows that a simple, empirically derived rule accounts well for the outcome of multi-ballot, majority-rule conventions in the period 18481948 and that the same rule would have enabled one to predict correctly the outcome of all contested presidential nominations since 1952 before any candidate had achieved majority support in the Associated Press’ poll of delegates.5

What they elicited from over a hundred years of presidential nominations by way of national party conventions was nothing less than a law-like description of its behavioral operation across time. In this, when the most recent increments to delegate support for particular candidates were added to their existing totals in any given year, there would always come a point when the resulting new totals could no longer be overcome by any candidate who was not in the lead at that particular point in the process. Yet now, this point could be specified precisely, and it was always the same, at least for parties which operated under the requirement of a simple majority for nomination. There might be many ballots to come in earlier conventions or many delegates to be chosen in a later era, but nothing that occurred with regard to either would ever derail a candidate who had already passed this one particular statistical point. Moreover, what was being elicited here, and what proved its power in a parallel fashion in all such presidential nominations, could be given a precise statistical description. In some vague and abstract sense, the twists and turns of the real world of nominating politics could in theory have caused the resulting summary statistic to be misleading. But in actual practice, no matter how different the relevant nominating campaigns were, the same formula could not just always be applied; it would always be accurate, revealing, and decisive when the evidence could be presented in the appropriate format. The authors called their product “the GainDeficit Ratio,” and they could offer it in one simple formula, whose key terms could be specified explicitly: St−1 − St−2 Q − St−1

4

B. E. SHAFER AND E. M. SAWYER

S t –1 represented a candidate’s delegate support in the most recent test. In the early years, this was the most recent convention ballot; in later years, it would be the most recent delegate survey. S t –1 − S t –2 represented the change in a candidate’s support from the immediately preceding test, be it a ballot or a survey. And Q represented the minimum proportion of the vote sufficient to nominate. In practical political terms and in their words, “The ratio takes account of where the candidate is, how close victory is, and how fast the candidate is progressing towards it.”6 Even more to the practical point, this ratio could be given a precise figure, namely 0.360. Given both the power of this finding and the role that it will play in everything that follows, it is important here to note what this result is, but also what it is not. It is not a “theory” about how nominating politics works. There were no hypotheses generating the search for this ratio: it was purely an empirical hunt for the point at which current delegate leadership could no longer be reversed. And there was no set of propositions about why this result should arrive, not to mention why it should be resolved at 0.360. This was instead a behavioral law, pure and simple: it specified what always happened, no matter the specifics of whatever was driving it in any given contest. What this implied, in the maddening world where social scientists confronted media analysts, was that it should always be possible to discard—or at least to specify the point at which they should be discarded—all those impressionistic judgments about the possible future of any given nominating contest. In other words, there was (and so far is ) a point at which it was simply no longer true that: • “The Republicans would never ultimately nominate a man who had to be ruled out of order when he protested that he was not a candidate” (James A. Garfield); • “The Democrats would never ultimately nominate an ethnic metropolitan Catholic wet” (Alfred E. Smith); • “The Democrats could never rally around a one-term small-town peanut farmer” (James E. Carter); or, of course, • “In the end, the Republicans will never ultimately choose a populist plutocrat star of reality TV” (Donald J. Trump).

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

5

It was now possible to know not only that all would indeed do the apparently unthinkable in each of these cases, but precisely when this result could have been known—and should have been recognized. So the challenge was to unpack retrospectively the ongoing behavioral aspects of a recurrent nominating dynamic—remembering that the formal process itself would change monumentally across time—aspects which could in principle drive an informal politics that would be summarized in this precise and law-like fashion. Fortunately, an abstract description of the nominating process, sufficient to provide precisely this theoretical framework, would also appear in very short order. For what was being elicited through the Gain-Deficit Ratio was a formal embodiment of a long-time metaphor among practicing politicians, the “bandwagon,” in which most players could be expected to trade off their preferred outcome against their judgment of the likely result, trying in effect to move toward that outcome (or stave it off) at a point where their move would maximize the reward to their constituents, their ideologies, or themselves. In other words, what was needed was a theoretical description of the bandwagon dynamic, along with a theoretical explanation for why it should reliably recur. Given the time span in this search for some common dynamic, something short of 200 years of American political history, the fact that the ratio could produce a simple summary figure that always yielded this dynamic—with its crucial turning point—was analytically impressive. It is this statistical figure that will provide the critical tool not just for extending the Collat, Kelley, and Rogowski analysis to subsequent contests, including the most recent one, but much more consequentially, for further deconstructing major developments within those contests, old and new. Yet within a further three years, a theoretical formulation that could explain why it was possible to generate this figure and why it should reliably recur would in fact appear. In this second crucial piece of the new analytic ferment around presidential nominations, the one attempting to systematize an explanation of nominating politics, the elucidation of an apparently inevitable dynamic was to be provided most centrally by Larry M. Bartels in Presidential Primaries and the Dynamics of Public Choice. Working from the modern rather than the historic record of presidential nominating politics, Bartels was to answer the question of why this bandwagon should always roll. To do so, he homed in on a concept central to understanding its dynamic (and hence to explaining the Gain-Deficit Ratio), namely momentum, “the demonstration effect of important primary victories.”7 For Bartels,

6

B. E. SHAFER AND E. M. SAWYER

this crucial mechanism was the inescapable product of a selection system nearly unique in world politics but fundamental to the nominating process for the American presidency: … the modern presidential nominating process has a dynamic aspect unmatched by any other electoral process. Its key institutional feature is that individual state primaries are spread over a period of three and a half months. As the focus of attention moves around the country from week to week, politicians, journalists, and the public use the results in each state to adjust their own expectations and behavior at subsequent stages of the process. One week’s outcome becomes an important part of the political context shaping the following week’s choices. Thus, each primary must be interpreted not as a final result but as a single episode in the series of interrelated political events that together determine the nominee.8

Yet in offering this theoretical explanation for why the bandwagon should roll reliably in an era when primary elections were dominant, Bartels was implicitly offering a framework that could be generalized and extended much farther back in time, indeed all the way back to when presidential nominations were purely the product of internal party machinations. In other words, by reducing the alleged complexity of the modern world into a handful of straightforward explanatory notions, Bartels was able to accomplish two superficially inconsistent tasks: (a) to deconstruct major aspects of the contemporary nominating process, which rolls overwhelmingly by way of primary elections, in a fashion that (b) nevertheless provides direct analogs to an earlier process, which rolled principally among state party organizations and their chosen convention delegates. Along the way, Bartels himself would go on to unpack the relationship between media coverage and delegate outcomes, between the stage of the campaign and the level of uncertainty, between voter predispositions and candidate choice, and indeed among the mix of candidate strategies that characterized any given contest. However, what he had done even more centrally was to provide a theoretical explanation for the effect (and power) of the Gain-Deficit Ratio. This is the explanation that allows our analysis to move on from a perennial bandwagon to the prior occupations and career pathways that generate a field of contenders for a presidential nomination; to the key structural factors that effectively choose among these contenders, whatever analysts absorbed in the details

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

7

of contemporary contests may believe; and to the rules reforms and candidate strategies that are so often treated as decisive, rather than as final addenda to a process that has been more or less familiar since 1848. So that in the end, in Bartels’ own summary words: … actual experience with the presidential nominating process has been exactly the opposite of what we are led to expect on theoretical grounds. Every recent campaign—no matter how many candidates it involved, what specific rules were used to translate votes into delegates, or how complicated the electorate’s preferences may have been—has produced a clear winner. The problem of preference aggregation central to the classical theory has been overcome, time and time again, by the dynamic features of the nominating process…9

The Bandwagon in Operation An effort to deploy the bandwagon dynamic and its Gain-Deficit Ratio as central tools for analyzing presidential nominating politics must begin by confirming that the dynamic and its ratio have continued in essentially the same form since their initial discovery. If this continuity can indeed be confirmed, and thereby extended, the next task is to double back and align all the relevant contests, that is, those with recorded votes plus nominating rules that required only a simple majority. So aligned, this composite record of nominating politics can use the bandwagon dynamic and its Gain-Deficit Ratio to go on and examine the very different institutional arrangements that have nevertheless featured this common long-run dynamic, while simultaneously isolating the one great historical exception that was capable of annulling an otherwise general dynamic. Evidence in the original Collat, Kelley, and Rogowski piece ended with the 1980 contest, now getting on toward forty years ago. Subsequent analyses, most especially a pair of articles from Barbara Norrander, extended their approach and its findings through 2004, while in effect looking at the bandwagon dynamic the other way around, that is, by way of candidate attrition rather than majority formation.10 So an effort to finish this sequence by bringing it up through 2016—we shall attend to 2020 specifically in the closing chapter—proves fully consistent with previous findings from all those earlier years, and has two added advantages. In the process, this detailed exposition can provide a full example

8

B. E. SHAFER AND E. M. SAWYER

of how the Gain-Deficit Ratio is calculated, in this and all other years. More pointedly, these examples and their measurement can be teased out of the most recent year, 2016, that featured serious nominating contests in both parties, contests that appeared at the start to be sufficiently openended as to allow multiple differing trajectories—and multiple differing interpretations of their dynamics. Or, said the other way around: given that the nominating contests of 2016 were the fountainhead for so much speculation about nominating dynamics that paid no attention to—and thus was unconstrained by—an ongoing bandwagon effect, and given that the ubiquity of these speculations was the immediate stimulus for a return to the Gain-Deficit Ratio, it seems additionally beneficial to use 2016 as the concrete example of how the ratio and its interpretation will work in practice in everything that follows.11 To that end, Table 1.1 offers the rolling record of the Gain-Deficit Ratio in 2016, first for the Republicans and then for the Democrats.12 Among Republican aspirants, Donald Trump registered a major surge in the ratio, one unmatched by any other aspirant, on March 1, “Super Tuesday,” a day featuring primaries in Alabama, Alaska, Arkansas, Georgia, Massachusetts, Minnesota, Oklahoma, Tennessee, Texas, Vermont, and Virginia (Table 1.1A). Two weeks later, on March 15, with primaries in Florida, Illinois, Missouri, North Carolina, and Ohio, he moved past the required figure of 0.360. The requisite number had been reached; the race was over. In all of American history, politics had never denied nomination to such a candidate, and it was not to deviate in 2016. Analysts concerned with different aspects of nominating politics might attend to the responses of the other candidates to the evolution of Trump’s Gain-Deficit Ratio. But in the headline event— a major-party nomination for President of the United States—it was all over but the clean-up on March 15. Among Democratic aspirants, Hillary Clinton registered a counterpart surge in her Gain-Deficit Ratio on Super Tuesday as well (Table 1.1B). On the one hand, the number of delegates available on that date was so large that any candidate might have been expected to show some improvement there, and Bernie Sanders also did. On the other hand, it is worth noting that such a result was not automatic. Only Donald Trump secured a similarly positive result among Republicans. Despite the huge delegate availability, gains by Ted Cruz, Marco Rubio, and John Kasich were comparatively trivial. In any case, despite a narrow loss with great

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

9

Table 1.1 Gain-Deficit calculations for 2016 A. The republicans Date February 1 February 9 February 20 February 23 March 1 March 5 March 6 March 8 March 10 March 12 March 15

Donald Trump

Ted Cruz

Marco Rubio

John Kasich

0.006 0.009 0.040 0.012 0.282 0.071 0 0.095 0.001 0.001 0.420

0.007 0.002 0 0.005 0.0218 0.074 0 0.068 0.001 0.027 0.064

0.001 0.001 0 0.006 0.085 0.012 0.021 0.001 0.002 0.010 0.006

0.001 0.003 0 0.001 0.017 0.008 0 0.014 0 0.008 0.074

B. The democrats Date February 1 February 9 February 20 February 27 March 1 March 5 March 6 March 8 March 12 March 15 March 22 March 26 April 5 April 9 April 19 April 26 May 3 May 7 May 10

Hillary Clinton

Bernie Sanders

0.010 0.004 0.009 0.017 0.293 0.033 0.005 0.058 0.002 0.327 0.046 0.034 0.035 0.006 0.148 0.301 0.057 0.006 0.016

0.009 0.006 0.006 0.006 0.178 0.027 0.009 0.040 0.001 0.193 0.054 0.078 0.037 0.005 0.092 0.165 0.046 0.003 0.019

(continued)

10

B. E. SHAFER AND E. M. SAWYER

Table 1.1 (continued) B. The democrats Date May 17 June 4 June 5 June 7

Hillary Clinton

Bernie Sanders

0.086 0.012 0.065 2.18

0.071 0 0.027 0.544

press attention in the Michigan primary a week later, Clinton came close to duplicating Trump’s decisive win on March 15. Her Gain-Deficit Ratio was 0.327, ever so slightly short of what had otherwise always predicted a nomination. She was to be outpaced by Sanders until the New York primary on April 19, a long month of marking time. Clinton surged again the following week, but the Gain-Deficit Ratio, while registering the surge, likewise confirmed that she had not yet closed the deal. That occurred on June 7, when she took the two major contests, including the huge California primary, and smashed to 2.18 on this gain-deficit calculus. The lateness of her victory had provided some continuing rationale for the continuation of the Sanders campaign. Afterward, its continuation had to be based on concerns other than nominating politics. Along the way and comparatively, the Clinton nomination underlined the earliness, the power, and the definitiveness of the Trump victory, insistent press speculations notwithstanding. For the Republicans, their candidate had been decisively nominated on March 15. What continued to the convention was only speculative commentary to the contrary. So that is the form of an analysis of the bandwagon dynamic at work in a year that was contested in both parties, one falling in the modern plebiscitary period of delegate selection, that is, in the years after 1972 when both processes were dominated by primary elections. Yet for a much longer period, one hundred twenty years (1848–1968) rather than just forty-four (1972–2020), the two processes had been conducted by way of a very different institutional matrix, to which we will attend more comprehensively below. Here, note just that there was a long early period, really 1832 through 1908, when delegate selection was purely an internal party matter for the individual state parties, with the overall result assembled subsequently at the national party convention. And this was to be

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

11

followed by another extended era, 1912 through 1968, when the two processes, delegate selection and presidential nomination, resulted in a mixed system of state party conventions and state presidential primaries, albeit still dominated by the former. So perhaps a specific example of each would add value here, as a means to confirm the historical generality of the bandwagon dynamic and underline its utility for analyzing everything else that goes into presidential nominating politics, including candidate backgrounds, factional structures, and strategic interactions among presidential aspirants and the state parties. In this regard, perhaps the iconic contest of the pure convention era, at least for national conventions that did indeed use a majorityrule standard, was the one nominating Abraham Lincoln for president in 1860.13 This is the one simultaneously instantiating the Republican Party in an evolutionary form that could thereafter be traced from Lincoln’s nomination to the current moment. So Table 1.2 offers the gain-deficit calculation for these Republicans in 1860.14 The apparent front-runner for the 1860 Republican nomination was Senator William H. Seward of New York, and he was to be the leader on the opening ballot of the convention, followed at some distance by former congressman Abraham Lincoln of Illinois, with favorite sons in Senator Simon Cameron of Pennsylvania, Governor Salmon P. Chase of Ohio, and former congressman Edward Bates of Missouri, plus another five aspirants who would achieve more than single-digit support at some point. Line 1 of Table 1.2 presents the vote on this opening ballot for these five named figures plus a catch-all category for the others. Confirming expectations, Seward did show a solid lead over these aspirants on this opening ballot, albeit well short of the 234 needed for a convention majority. Yet Table 1.2 Delegate totals and Gain-Deficit Ratios: the Republicans, 1860

1st Ballot 2nd Ballot 1st G-D Ratio 3rd Ballot A 2nd G-D Ratio 3rd Ballot B

Seward

Lincoln

Cameron

Chase

Bates

Others

173.5 184.4 +0.182 180 −0.090 121.5

102 181 +0.598 231 +0.953 340

50.5 2 −0.261 0 0 0

49 42.5 −0.070 24.5 −0.094 2

48 35 −0.035 22 −0.065 0

43 21.5 −0.146 7.5 −0.066 2

12

B. E. SHAFER AND E. M. SAWYER

a disciple of the Gain-Deficit Ratio should have recognized by the second ballot that the contest was over—with Lincoln as the nominee. Indeed, Table 1.2 is so brief in comparison to most of the other examples offered subsequently because the obvious front-runner was to be so quickly demolished. Seward continued to lead on this second ballot; he hardly wilted. But the big gainer was already Lincoln, whose initial raw increment came mainly from the withdrawal of Cameron; Pennsylvania Republicans had decided that they preferred Lincoln to Seward. But methodologically and practically, the Gain-Deficit Ratio signified that Lincoln had not just picked off Pennsylvania but had in fact stormed past the necessary threshold and would be the eventual nominee in short order. Which indeed he was on the third ballot. Lincoln was still two and a half votes short of an absolute majority when this ballot was tallied, but major shifts across the board left him with nearly everything except the big New York delegation and a smattering of delegates elsewhere, who stayed with Seward to the end. Accordingly, despite the fact that Lincoln was not the front-runner going into the convention, nor even a particularly impressive second leader in a diverse and divided field, the bandwagon as registered in the Gain-Deficit Ratio needed only two further ballots to make Lincoln the nominee and lead on to the first Republican presidency, the American Civil War, and a thoroughly recast party system afterward. Different variants of the same story could be elicited from many contests in the mixed system of delegate selection as well, the one that emerged after the addition of presidential primaries in 1912. For more than fifty years, delegate selection through primary elections would remain a secondary aspect of nominating politics. Some of the states which adopted primaries offered only presidential preference polls, selecting the delegates themselves elsewhere. Conversely, some of the states which adopted primaries elected only their delegates, without a presidential preference poll. Others among the states which adopted primaries had both a preference poll and a delegate selection but provided no way to connect the two. With the result that very few states offered what would be the consensual understanding of a true presidential primary in the modern era. So an example from this long period of the mixed system, 1912 through 1968, might be expected largely just to reconfirm the power of an ongoing bandwagon, even though the institutional arrangements through which its power was confirmed were formally and noticeably

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

13

different. One of the most dramatic contests from this new institutional era, one that should have challenged this continuity if anything did and hence one of the best examples of the undisturbed power of the bandwagon, came with the Republican nominating contest of 1920, the one resulting ultimately in the nomination of Governor Warren G. Harding of Ohio.15 Table 1.3 follows the (mis)fortunes of three main contenders: Major General Leonard Wood of New Hampshire, Governor Frank Lowden of Illinois, and Senator Hiram Johnson of California, along with the usual smattering of favorite sons. The table also isolates the vote for one of three favorite sons who had some modicum of support—the weakest of the three, Governor Warren G. Harding of Ohio, who was of course to be the ultimate nominee.16 Harding had entered (and won) the primary in his home state, but he had entered no others. He was not an active campaigner for the nomination. And the two other serious favorites sons, Governor William C. Sproul of Pennsylvania and Nicholas M. Butler of New York, long-time president of Columbia University, both began with more delegates than Harding. The three leaders dueled on relatively stable terms through the first seven ballots, with their feeble Gain-Deficit Ratios suggesting that all three were going nowhere. Among the others, Harding shed a few delegates and then received a few—until the seventh ballot when he became the leader among the remaining favorite sons, albeit achieving only a third of the delegates for Wood or Lowden. Moreover, his Gain-Deficit Ratio remained unimpressive, actually declining after the eight ballots. Table 1.3 Gain-Deficit Ratios: the Republicans, 1920 Ballot 2 3 4 5 6 7 8 9 10

Wood

Lowden

Johnson

Harding

Others

0.012 0.071 0.065 −0.008 0.069 0.003 −0.067 −0.206 −0.217

0.206 0.110 0.032 0.088 0.047 0 −0.024 −0.486 −0.202

0.036 0.206 −0.021 −0.020 −0.062 −0.027 −0.031 −0.012 −0.005

−0.015 −0.001 0.007 0.040 0.072 0.243 0.078 0.965 1.773

−0.250 −0.127 −0.043 0 −0.050 −0.018 −0.012 0.015 −0.046

14

B. E. SHAFER AND E. M. SAWYER

Yet the night between the eighth and ninth ballots was to feature the famous “smoke-filled room,” in which a number of party leaders, having noted Harding’s progress by comparison to their other apparent options, agreed to attempt to break an apparently bootless deadlock by means of the Ohio Governor.17 The ninth ballot reflected their deal: Harding’s delegate total surged, and while he was still well short of a majority, his Gain-Deficit Ratio exploded, crossing the threshold that predicts a nomination. Which he did on the next ballot, the tenth, when only a rump of Wood supporters failed to join his bandwagon.

The One Reform That Mattered Is this long-running bandwagon effect, then, invariant to all changes in nominating regimes, large or small, institutional or behavioral? Its indifference to the institutional matrixes for delegate selection, to the nature (or even the identity) of the political parties, and to the historical issues of the day, will indeed be confirmed in subsequent chapters. Yet there has been one formal shift in the procedural mechanics of nominating politics which did affect the behavioral character of the bandwagon in a major way. This was the creation of a diagnostic and long-running division between the Republicans and the Democrats in their numerical requirements for a presidential nomination, and then the erasure of that difference in the modern period. The Republican Party (along with its short-lived predecessors) began and maintained a simple majority as the standard for nomination. But the Democratic Party began instead with a two-thirds requirement, lasting through 1936 when it finally joined the Republicans under a simple majority, which it has retained ever since.18 The logic of majority formation under the two standards was to prove very different over time. Moreover, the differential impact of that logic can be measured: the Gain-Deficit Ratio would work unfailingly for the Republicans through all their history and for the Democrats after 1936. But the Ratio just as clearly would not apply systematically to the Democrats during their long two-thirds era. The Republican preference for a simple majority appears to be a more or less unconsidered response of the moment. A majority among majorparty nominees would obviously—constitutionally—be demanded of a president, so a majority inside the party seemed a reasonable corresponding demand of its nominee. And no one dissented in a serious

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

15

way. By contrast, the Democrats featured initial and then intermittent, explicit arguments about a proper nominating standard, albeit settling early on two-thirds. The original rationale was to create nominees with widespread support in an aspiringly national party that was geographically dispersed and forced to operate in a world where communication among state parties, the key operative elements of this national party, was difficult. Later, key party factions would instead rally to the two-thirds standard in the belief that it protected what were essential policy interests to these factions. So the difference went all the way back to the two foundings. Whig rules for their first national party convention in 1839 required only a simple majority. The Republican Party, formed officially in 1854, carried the rule into their first national convention in 1856, albeit only by way of a judgment from the chair, to the effect that a simple majority was all that was required.19 Republicans needed a further elaboration of that ruling in 1860, where their Rules Committee endorsed a simple majority of totalvotes-allocated but a successful minority report argued that because a number of (slave) states were not even present, nomination should require only a simple majority of the vote among states actually represented at the convention. Thereafter, there would be occasional queries at Republican conventions as to the proper base on which a majority should be calculated, but the simple-majority provision was not in question.20 A bandwagon dynamic for presidential nominations among Whigs and then Republicans, stretching from 1840 through 2020 and counting, was the result. Yet this was to be a one-party dynamic from 1836 until 1940, courtesy of the choice by the Democrats of a different nominating standard. That standard was the two-thirds rule, and it would lead to a hundred years of a different nominating dynamic before the two parties ultimately converged. Under the two-thirds rule, as we shall see, the Democrats were no less likely than the Republicans to possess first ballot, consensual nominees. But when they did not possess these easy choices, nominating politics would differ radically from the Republican pattern. Moreover, this difference in the logic of majority formation would live through the pure convention era and into the mixed system of delegate selection, before being finally, formally, and officially eradicated in 1936. The two-thirds rule was not only challenged but momentarily repealed by a narrow margin at the first nominating convention of Martin Van Buren in 1836, though it was quickly reinstated by a voice vote from the floor as Van Buren supporters closed ranks to prevent any surprise

16

B. E. SHAFER AND E. M. SAWYER

events thereafter. Van Buren had no problem with the rule in his renomination in 1840, though it was the barely implicit reason why the vice presidential nomination was left to the individual state parties, since Richard M. Johnson, sitting vice president, would have faced major convention controversy. The Van Buren forces were then behind an effort to repeal the two-thirds requirement in 1844, and when that effort failed, Van Buren became the first Democratic aspirant for a presidential (re)nomination to secure a simple majority but fail to reach the necessary two-thirds.21 The rule was explicitly and solidly affirmed in 1848. It was affirmed once again in 1860, when a judgment from the chair that this rule required two-thirds of all convention votes, rather than just of the votes allocated for states that were actually present, became part of the disaster that sent the convention into recess for six weeks without a nominee, before reconvening and nominating Stephen Douglas by voice vote.22 The rule was reaffirmed once more in 1868, in the same form as 1860. It was revisited in 1884, through a proposal that it be abolished prospectively from 1888 onward, but that proposal had so little support that the vote on it was not even completed. By 1932, the implicit political impact of the rule had changed substantially. Before the Civil War, the rule had been viewed by many inside the party as a bulwark against regional balkanization. A major emerging north–south division might yet be staved off, the argument went, because the two-thirds rule forced the party to find a nominee who could at best unite, or at worst straddle, that particular divide. After the Civil War and once the Democratic Party was again nationally competitive, the nature of this regional divide shifted, away from North versus south and toward south-plus-West versus Northeast. The populist surge uniting the first two regions was the explanation for this change, and it made the two together so reliably dominant internally that rules reform became an irrelevance: nothing could derail the new populist majority.23 And there matters stood until 1932, when Franklin Roosevelt, entering the convention with an absolute majority but not two-thirds, was pressed by the more liberal state delegations to come out against the two-thirds standard. By then, the key division had reverted to south versus North, implicitly on economics and explicitly on race, where the Roosevelt leadership was not inclined to risk the unity of their bandwagon over a convention rule. Yet a majority of the Roosevelt coalition was in favor of repeal, and when Roosevelt returned for what was only a trivially

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

17

contested renomination in 1936, his Rules Committee agreed to end what had long been a distinguishing procedural feature.24 The report recommending its abolition was sweetened with the promise of reapportionment for the convention as a whole, so as to reflect the strength of state party results, a potential major boon to the south. Otherwise, a distinguishing standard, one that had given the party a distinctive internal politics since 1832, finally expired. Table 1.4 begins the effort to isolate the impact of different majority standards by arraying all presidential nominations from 1848 onward.25 To that end, the table distinguishes nominations conducted under simplemajority versus two-thirds rules, as well as by contested versus uncontested nominating struggles, where the point of this second distinction is to prevent differences or similarities in the basic character of the parties themselves from artificially suggesting different impacts from the rules defining a majority requirement. In this latter regard, neither the GainDeficit Ratio as a statistic nor the bandwagon as an underlying concept and behavioral force are necessary to elaborate on the (non)dynamics of consensual nominations, those for which there was no actual change to explain. Yet grand institutional differences over time do require a further initial distinction within the designation of “contested”: • For earlier years, contested means going to at least a third ballot at the national party convention. Conventions with less than three ballots can be ‘predicted’ by almost any indicator: their first ballot is either the final outcome or a kind of straw poll for confirmation on the second ballot. So for these earlier years, only conventions with three or more ballots can properly test the Gain-Deficit Ratio and comment on its associated dynamic. • For later years, contested instead means having more than one aspirant who won more than one primary over the entire sequence. Local candidates who capture only their home state or aspiringly national candidates who win only one isolated contest are never treated (by anyone other than themselves) as serious contenders. Yet a serious alternative challenger is again necessary to check the Gain-Deficit Ratio and examine its associated bandwagon.

18

B. E. SHAFER AND E. M. SAWYER

Table 1.4 Gain-Deficit Ratios for major-party nominees across time

Republicans

Democrats

A. The pure convention system 1848 1852 1856 1860 1864 1868 1872 1876* 1880* 1884 1888 1892 1896 1900 1904 1908

[1848] [1852] [1856] [1860] [1864] [1868] [1872] [1876] [1880] [1884] [1888] [1892] [1896] [1900] [1904] [1908]

B. The mixed system 1912 1916 1920 1924 1928 1932 1936 1940 1948 1952 1956 1960 1964 1968

[1912] [1916] [1920] [1924] [1928] [1932] [1936] 1940 1948 1952 1956 1960 1964 1968

C. The plebiscitary system 1972 1976 1980 1984 1988

1972 1976 1980 1984 1988

(continued)

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

Table 1.4 (continued)

19

C. The plebiscitary system 1992 1996 2000 2004 2008 2012 2016 2020

1992 1996 2000 2004 2008 2012 2016 2020

KEY Contested races are left-justified; uncontested races are offset Races requiring a two-thirds majority are in brackets [] *Races where the Gain-Deficit Ratio could not predict have asterisks

The first findings to emerge from this reconstitution are two characteristics that are shared by, not distinguished between, the two parties. In the first, a finding so obvious as to be almost true by definition, both parties have always reached an ultimate nomination. Which is to say: no national party convention has ever dispersed without a nominee, though the Democrats endured a number of contests which must have made many delegates wonder. And in the second, a commonality that might seem almost counterintuitive for parties that used such sharply different requirements for nomination, neither party was disproportionately likely to have contested as opposed to consensual pathways to a presidential nomination. With regard to the first of these implicit findings, on the ultimate ability to produce a nominee by both parties, the Democrats had, in truth, been the party at apparently greater risk of failing to find an eventual nominee. The had actually been forced to adjourn and reconvene in 1860, before turning to a voice vote to make a nomination otherwise unattainable under the established interpretation of the two-thirds rule. A dozen years later, the Democrats of 1872 would feel compelled to solve their nominating challenge by accepting as their standard-bearer the candidate of another party, Horace Greeley of the short-lived Liberal Republican Party. And a half-century after that, the Democrats of 1924 would need seventeen days to reach their ultimate nominating compromise, via John W. Davis of West Virginia, former Solicitor General of the United States, who had to wait through 103 ballots for his coronation.26

20

B. E. SHAFER AND E. M. SAWYER

A second common characteristic, much more operationally important to the analysis that follows, is that both parties did generate extended stretches of consensual nominations, different nominating requirements notwithstanding. In truth, these were generally the product of either feast or famine, that is, periods of reliable hegemony or of continual desperation. Hegemony was the story for the Republicans after the Civil War and desperation the story for the Democrats, as the latter searched for any candidate who might be competitive nationwide. Conversely, hegemony was the story for the Democrats after the Great Depression and desperation the story for the Republicans, who crashed out of a short stretch of more or less automatic wins to face their own extended search for someone who could even potentially challenge Democrat Franklin Roosevelt. But the point here is that the difference between requiring a simple majority or a two-thirds majority was not sufficient to cause one party to reach an early consensus often and the other to reach it only rarely. In the abstract, a two-thirds requirement might have driven the Democrats to seek a pre-convention consensus more (not less) often, or it might have led Democratic Party factions or Democratic presidential aspirants to resist any early agreement, perhaps hoping to spearhead a later compromise. Yet neither effect is evident. Indeed, the fact that the two parties do not differ powerfully in the generation of consensual nominations argues that it was not their internal structure but the actual rules for creating a majority when there was no consensus that was ultimately decisive. A series of further, major, analytic points are then embedded within this historical array. The first has already been noted: there were no majorityrule contests from 1848 through 2020 for which the Gain-Deficit Ratio produced an incorrect prediction. The closest thing to an exception were the Republican contests of 1876 and 1880, where the ratio could not generate a prediction at all. So the Gain-Deficit Ratio could generate predictions in forty-eight out of fifty cases, and these predictions were never incorrect. Though the more impressive (and honest) comparison would involve only contested nominations, where the ratio was able to generate predictions for thirty-four out of thirty-six, and where these predictions too were never incorrect. In any case, among simple majority but not among two-thirds requirements, the Gain-Deficit Ratio has failed to make a prediction only twice, when no candidate reached the requisite figure of 0.360. Said the other way around: so far in American history for those nominating races where

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

21

the ratio can be applied, it has always told the truth in advance of an actual majority. This result could of course be extended to the secondlargest cluster of nominating contests, those featuring an uncontested nomination, but its further success there would be trivial. Conversely, it cannot be applied before the emergence of the modern party system and creation of the national party convention, since parallel processes did not exist. Beyond that, the ratio requires a different sort of attention in all those earlier Democratic contests that were governed by a two-thirds rule, which it will get in the next section of this chapter. Otherwise, the result is unfailing. A second major analytic point inherent in Table 1.4 might be less immediately obvious but is nearly as important. Presidential nominations were the product of very different institutional matrixes across time. That is the point of further subdividing them by institutional eras. Yet the bandwagon dynamic was so integral to nominating politics and so dominant within it that this bandwagon rolled inside and across all three eras without evident difficulty. Since 1832, with the creation of mass political parties and their national party conventions, there have in fact been three different institutional configurations—three different institutional “regimes,” if you will—through which to pursue a presidential nomination: • A pure convention system characterized nominating mechanics from 1848 through 1908. In this system, every state chose its delegates to the national party convention through some form of state convention system, although states varied in the degree of openness in the selection of their party officialdom. Regardless, the resulting process was entirely internal to the official party structure, and party officials were uniformly the dominant players. • A mixed system then characterized nominating mechanics from 1912 through 1968. Primary elections arrived in some individual states for the selection of state delegations to the national party convention, though most states retained the older convention system and some primary systems were more “advisory” than effective. Nevertheless, this mix of primaries and conventions gave some voters a role in presidential selection and some candidates an alternative route to the nomination. • Ultimately, a plebiscitary system succeeded this mixed system, and gave us the modern world. In this world, the vast majority of states

22

B. E. SHAFER AND E. M. SAWYER

used presidential primaries to select (the vast majority of) delegates to national party conventions, so that the outcome of these primaries was the sole effective and not just an advisory means to the presidential nomination. The story of presidential nominations thus became the story of sequential primary elections. The lettered divisions in Table 1.4 (A, B, and C) distinguish presidential nominating politics by way of these three institutional eras. What they simultaneously confirm is that institutional arrangements for delegate selection, as structurally fundamental as these clearly were in their own right, do not distinguish nominating politics by comparison to an underlying, continuing, and common nominating dynamic under majority rule, nor do they distinguish the two parties once the Democrats adopted that standard. But what otherwise looks like—and indeed is— two very different strategic calculations in earlier years can be seen to be derivative of two powerfully different procedural constraints and not of the differing character of two parties, as different as this often was across the same period of time. Said the other way around, parties could experience extended periods when their nominations were effectively uncontested under both the pure convention and the mixed systems of delegate selection. This again implied that the majority standard could determine the manner by which nominating competitions actually ran, but it even more evidently did not demand that nominations were more likely to be contested in one party rather than the other. Neither party was actually to feature any further uncontested stretches after 1968 and the arrival of the plebiscitary system, so this absence does appear to be the product of an institutional change. At the very least, party officialdom in the modern world could no longer forestall even unpropitious nominating campaigns. Yet the basic point remains: the plebiscitary era, like both preceding eras, did not gainsay the power of a continuing bandwagon.

Nominating Politics Under the Two-Thirds Rule Candidate beliefs and pundit comments to the contrary, no candidate who surpassed the Gain-Deficit Ratio under a majority-rule standard was ever denied a nomination after 1848. Just as no candidate who did not surpass that standard by breaching its precise measure—regardless of their

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

23

credentials, regardless of their resources, regardless of their strategies— ever defeated a candidate who did surpass it. Though at the same time, the critical limitation on this otherwise striking regularity lay in the definition of a nominating majority. Indeed, we began by isolating and commenting on the impact of the requirement of a simple majority in part to be able to go on and isolate its differences from the requirement of a two-thirds majority instead. And what results is the perception that during the long history of divergent nominating requirements, both the operation of the bandwagon dynamic and the strategic logic of nominating politics would be different under the two majority principles, and hence between the two major parties. For the Republicans under their rules, the goal of serious contenders was to move to front-runner status at their earliest opportunity. So presidential aspirants—and even more crucially, their supporters—needed an early test of their prospects for victory. If these were negative, they were encouraged by the comparative ease of majority formation to shift expeditiously to the next-best alternative, while calculating what both the candidates and their supporters could extract in return for this altered support. Since everyone else would be doing the same, the ability to win with a simple majority meant that candidates who remained intransigent but not victorious risked being left out of the crucial decisions. Said differently, the requirement of a simple majority biased the Republicans toward a next-best-alternative among early aspirants and then, as a fallback, toward a subsequent shift to some compromise candidate who gave offense to none. For the Democrats under their rules, the task of serious contenders likewise began with an early test of their ability to move more or less directly to a nomination, a consideration that was even more pressing given the augmented challenge to majority formation inherent in the twothirds requirement. Yet if they could not hope to win expeditiously, this more-demanding standard then counseled against any immediate shift. Even if there was no ill will toward those who had in effect vetoed their own prospects—and this too was a more likely implication of a two-thirds requirement—it remained desirable to wait (and to hold supporters) until the succession of possible compromises had clarified, in order to avoid dissipating initial influence by going with another candidate who might yet have the same difficulty in reaching two-thirds. So all the main players

24

B. E. SHAFER AND E. M. SAWYER

were well advised to wait for the specifics on these compromise candidates, who would inevitably appear and who might bring a more concrete prospect of garnering the necessary two-thirds. In that sense, the internal nominating dynamics for the two parties ran recurrently in opposite directions. Both could still begin with consensual nominees, as Table 1.4 confirms. Yet the nature of the resolution for nominations that were actively contested was recurrently different under long-standing Democratic rules—that is, behaviorally different— when compared to the resolutions characterizing Republican contests, just as it would go on to be different as between Democratic contests of the two-thirds era and Democratic contests from a later shift to a simple majority. Yet in the long interim before partisan convergence on a simple-majority standard, all sorts of implicitly rules-based outcomes sprang to life for the Democrats under the two-thirds standard, outcomes that appeared unlikely-to-impossible with a simple majority: • The most dramatic of these involved Democratic contests where a presidential aspirant actually attained an absolute majority, but could not move on and accumulate two-thirds. By definition, the Republicans never had any of these: a simple majority was the ultimate decision. • For the Democrats, there were even more cases where one candidate after another moved into the lead as a nominal front-runner but likewise could not move on to a two-thirds majority. While the Republicans too were capable of generating sequential front-runners, their challenge in moving to a nomination still faced a much lower bar. • Moreover, while the identity of the key internal factions would differ between the two parties—see Chapters 3 and 4—it was not that the Republicans never generated serious factional disputes. It was just that the Democrats made the resolution of such disputes more numerically challenging, while simultaneously feeling more fraught to the participants. • Not surprisingly, then, it was the Democrats who were to gift the nation as a whole with the notion of a “dark horse,” an eventual nominee who surfaced only late in the process, as a broadly acceptable rescue from previously deadlocked efforts to achieve the basic nominating requirement, which in their case was two-thirds of all convention delegates.

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

25

An example rolling all of these dramatic fall-outs from a two-thirds standard into one cluster arrived early, with the fourth Democratic national convention in 1844, set off by the unsuccessful effort at renomination by former president Martin Van Buren.27 Van Buren had been nominated and elected in 1836, as the heir-inherent to Andrew Jackson, nominal founder of the Democratic Party. Van Buren was then renominated but not reelected in 1840. He aspired to be nominated again and return to the presidency in 1844, a hope reinforced by the death of William Henry Harrison, the Whig president who had defeated him. And Van Buren actually went on to achieve an absolute majority on the first ballot of the 1844 convention, with only Lewis Cass, Senator from Michigan, as even an apparently serious alternative.28 Yet that would prove to be the high-water point of the Van Buren campaign. Cass would gain and Van Buren would lose, both marginally, on the second ballot. Cass would actually move to front-runner status by the fifth, displacing Van Buren, though the accompanying gain of two delegates was nowhere near sufficient to cause the Gain-Deficit Ratio (not shown) to move consequentially, since the main cause of his new lead was a modest decline in support for Van Buren. Van Buren managed to staunch his losses thereafter and remained roughly stable into the eighth ballot, when it was Cass who began to shed delegates, albeit modestly, and when James K. Polk, former Speaker of the House and former Governor of Tennessee, was introduced as a possible way forward. Which he did indeed become on the very next ballot, the ninth, when all of the former Van Buren support and most of the Cass support shifted to him. So Van Buren managed to become the first example of a candidate gaining an absolute majority but not two-thirds at a Democratic convention, while Cass became the first example of a rising challenger who seized front-runner status from the original leader but could never reach two-thirds, and Polk completed the story by becoming the first successful “dark horse”—uninvolved at the start, a distant third on the penultimate ballot, but achieving a healthy two-thirds-plus on the very next vote. Yet as dramatic as that composite outcome was, the most famous example of an absolute majority that ultimately could not be harvested belongs to the unfortunate Champ Clark of Missouri, Speaker of the US House, in the Democratic nominating contest of 1912, under the mixed system and not the pure convention system of delegate selection.29 Clark was to lead solidly on the first ballot, followed by Governor Woodrow

26

B. E. SHAFER AND E. M. SAWYER

Wilson of New Jersey with serious if initially trailing support, followed by a set of favorite sons: former Governor Judson Harmon of Ohio, Representative Oscar W. Underwood of Alabama, and Governor Thomas R. Marshall of Indiana.30 Clark continued to move forward over the next nine ballots, reaching an absolute majority on the tenth, and he continued to reproduce that majority through the twenty-fourth.31 Nevertheless, Wilson was also gaining across all twenty-four ballots, while Underwood and Marshall remained static, and Harmon faded when his New York base shifted, albeit to Clark rather than Wilson. Wilson nevertheless moved ahead of Clark on the thirtieth ballot, with his largest increment coming from the withdrawal of Marshall, who would be his vice president. Wilson achieved an absolute majority of his own on the thirty-sixth ballot, only to lose it again on the forty-second. Yet this time, there would be no dark horse. Instead, Illinois shifted all 58 of its delegates to Wilson on the following ballot, setting off a stampede one ballot later, culminating in a two-thirds majority for Wilson on the forty-sixth ballot. Moreover, this 1912 contest, historically important in its own right, gains an even larger place in political history because it captures so stereotypically not just the differing fortunes of Democratic versus Republican candidates under different majority rules, and not just the way that the Gain-Deficit Ratio could measure a bandwagon dynamic under one rule but not the other, but because it brings the two fundamentally different behavioral dynamics behind these divergent results into focus. Or, said the other way around, the 1912 contest is especially good at showing that what distinguished nominating politics in the two major parties for a very long time was not a different nominating requirement by itself, but rather the different behavioral dynamic to nominating politics that followed in the wake of those differing rules. On its own terms, the 1912 contest was just another impressive example of how the acquisition of an absolute majority could not guarantee nomination for a Democratic contender. To that end, Table 1.5 concentrates on the fortunes of Clark and Wilson, as a backdrop to the difference between the two majority standards. Table 1.5A reruns the actual ballot as if it had been produced under a simple-majority standard. Needless to say, Clark is nominated on the tenth ballot, where he acquired his numerical majority at the actual convention but failed to be nominated under its two-thirds requirement. Yet if all that this reflected

1

27

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

Table 1.5 The Democratic nomination of 1912

A. Under a simple-majority rule BALLOT

Wilson

Clark

2nd 3rd 4th 5th 6th 7th 8th 9th 10th

0.077 0.026 0.013 0.008 0.016 – – 0.005 –

0.062 – 0.02 0 0.02 0.048 – 0.038 Nominated

B. Under a two-thirds rule BALLOT

Wilson

Clark

2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th 13th 14th 15th 16th 17th 18th 19th 20th 21st 22nd 23rd 24th 25th 26th

0.041 0.014 0.012 0.004 0.008 – – 0.003 – 0.011 – 0.005 0.014 0.004 0 0 – – 0.09 0.021 0.003 0.008 0.011 0.008 0.008

0.021 – 0.007 0 0.007 0.016 – 0.013 0.612 – – 0.041 – – – – – – – – – – – – –

(continued)

28

B. E. SHAFER AND E. M. SAWYER

Table 1.5 (continued)

B. Under a two-thirds rule BALLOT

Wilson

Clark

27th 28th 29th 30th 31st 32nd 33rd 34th 35th 36th 37th 38th 39th 40th 41st 42nd 43rd 44th 45th 46th

– 0.107 – 0.09 0.062 0.001 0 0.001 0.064 0.009 0 0.009 0.013 0 – – 0.855 0.278 0.043 Nominated

0.021 – 0 – – 0 0.004 0 – 0.003 – – – 0.003 0.003 0.02 – – 0 –

was the numerical impact of a mechanical requirement, and not the institutionalized behavior that came with it, there was a simple way to restore the analytic utility, of the Gain-Deficit Ratio. For the formula for this Ratio actually included a specific provision to handle the situation, and thus theoretically to restore its ability to predict presidential nominations under different majority requirements. To see this in action, recall the precise definition of the formula itself: St−1 − St−2 Q − St−1 Recall likewise that S t –1 represented a candidate’s delegate support in the most recent test, being the most recent convention ballot in the early years, the most recent delegate survey in more modern times. S t –1 − S t –2 then represented the change in a candidate’s support from the immediately preceding test, be it a ballot or a survey. And Q represented the minimum proportion of the vote sufficient to nominate, which, when an

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

29

absolute majority was the nominating standard, required one-half (0.50) of the total. It is then a mechanically simple task to adjust Q to require a twothirds majority, that is, 0.67 + 1, rather than a simple majority of 0.50 + 1 (Table 1.5B). And making this change moves the analysis directly to an understanding that the Gain-Deficit Ratio fails under a two-thirds requirement not because it was calibrated on a different nominating requirement: it can be easily recalculated on any new standard. Rather, the logic of the ratio as elucidated fails because the recurrent behavior that accompanies the need for a simple majority is fundamentally different from the recurrent behavior associated with the need for two-thirds. That is the message inherent in Table 1.5B, derived by rerunning Table 1.5A under a twothirds and not a simple-majority standard, that is, with a Q of 0.67 rather than 0.50. The point—now punctuated—is that adjusting the formula to some other majority requirement still does not allow it to work. The GainDeficit Ratio is specific to the behavior that emerges in response to the need for a simple majority. In other words, what adjusting the formula to a two-thirds requirement confirms is that the dynamic that emerges in response to a much stiffer nominating requirement is not something that can be recaptured by adjusting a statistical formula. Rather, it is the recurrent behavior associated with a particular numerical requirement that does the political work here, and not the requirement itself. Most strikingly and despite a radically adjusted formula, it is again the tenth ballot that stands out, showing a huge jump in the adjusted Gain-Deficit Ratio for Champ Clark, to 0.612. Yet Woodrow Wilson is still nominated on the forty-third ballot, having shown no serious action in even this adjusted Gain-Deficit Ratio until the forty-third ballot. So just as breaching the critical standard when Q was 0.5 yielded an incorrect prediction of ultimate victory for a party operating under the two-thirds rule, that same breach still yields an incorrect prediction when Q is moved up to 0.67, that is, when the behavior associated with a two-thirds requirement is run through a formula reflecting that formula as well. Once again, then, it is the behavior associated with divergent cut-offs, not the cut-offs themselves, that continues to make one work and the other fail. Under the two-thirds rule, there would be other classic dark horses. Franklin Pierce was to follow James K. Polk only eight years later, as the dark-horse nominee of 1852. Those two unforeseen nominations in the

30

B. E. SHAFER AND E. M. SAWYER

course of only three contests became the reason why it was the Democrats and not the Republicans who inaugurated the notion of a “dark-horse candidate.” The milder version of the same phenomenon was to be realized even more frequently, whereby a succession of incipient bandwagons appeared, each achieving an actual lead but none threatening to reach two-thirds. The Democratic convention of 1852, the one ultimately producing Pierce, proved to be archetypal in this regard, featuring surges into the lead first for Lewis Cass of Michigan, then for James Buchanan of Pennsylvania, then for Stephen A. Douglas of Illinois, back again to Cass, and then to William L. Marcy of New York—before Franklin Pierce of New Hampshire was introduced on the thirty-fifth ballot, from whence he rolled to the nomination.32 The general pattern, of clear front-runners who could not ultimately progress and thereby contributed to the nomination of horses of various darknesses, actually became familiar. An unsuccessful renominating effort for President Andrew Johnson at the 1868 Democratic convention was followed by three other boomlets, first for former congressman George H. Pendleton of Ohio, then for General Winfield Scott Hancock of Pennsylvania, then for Senator Thomas H. Hendricks of Indiana, before finding former Governor Horatio Seymour of New York, the permanent chairman of the convention who announced that he was not a candidate but was nominated anyway. Yet another embodiment of the pattern would still be around in 1920, for the fratricidal conflict between William Gibbs McAdoo of California, former Secretary of the Treasury, and A. Mitchell Palmer of Pennsylvania, former Attorney General, to the eventual benefit on the forty-fourth ballot of James M. Cox, Governor of Ohio. Though the extended combat in 1924 between McAdoo and Alfred E. Smith, Governor of New York, would be even more impressive, resolved on the 103rd ballot for John W. Davis, Governor of West Virginia.33 What unites these diverse patterns—simple majorities that never reached two-thirds, extended successions of temporary front-runners, and ultimate dark-horse nominees—is that the final outcome seemed unpredictable in every case. Yet what appeared to be going on, while opaque in its time, was recognizable and recurrent by hindsight. Contested nominations, whenever there was enough organized opposition that the front-runner could not dismiss it quickly, proved to be extremely hard on all those with major early support under the two-thirds standard. Unlike the counterpart situation among the Republicans, where an initial leader might reasonably hope to consolidate early support by acquiring

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

31

the backers of less prominent early champions, and initial opponents might hope either to coalesce against that early leader or to strike a deal with him, initial leaders among the Democrats had much less hope of claiming the nomination once their momentum had slowed. What they could expect was to be more or less completely sure that they did not have to endure the nomination of an early rival, especially one who was involved in frustrating their own nomination. Yet the other main point implicit in all these composite examples, as individually idiosyncratic as they obviously were, was that they rendered collective testimony to the ability of one large institutional arrangement to derail an established and otherwise long-running bandwagon dynamic. The choice of a simple majority or a two-thirds requirement was diagnostic. And the recurrent behavior that emerged in response would not be disrupted by different institutional regimes for delegate selection— three of them across time—nor by the radically changing character of the political parties, explored here in Chapters 3 and 4, nor even by reforms and strategies that were often aimed explicitly at making the bandwagon run differently. Indeed, as if to underline the power of that continuity and the importance of the one institutional rule that could disrupt it, the Democrats themselves would finally abandon the two-thirds rule after 1936, bringing the Democratic Party into complete conformity with an established bandwagon dynamic and its confirmatory Gain-Deficit Ratio, while eliminating a long-established difference between Democratic and Republican nominating politics.

Simultaneous Mutual Adjustments to Sequential Struggles The demise of the two-thirds rule after 1936 moved both parties, by definition, to the same numerical standard for presidential nominations. Yet what it did simultaneously—the largest impact of a structural reform on nominating politics and surely the least recognized—was to put to sleep the only major impediment to the bandwagon dynamic in all of American history. Nominees would still have personal characteristics idiosyncratic to themselves. Party factions would still evolve divergently. Procedural tinkerers and strategic wizards would never cease proposing—and often implementing—intendedly new approaches to shaping a presidential nomination. Yet none of these would alter a bandwagon dynamic that is rolling on toward its two-hundredth birthday.

32

B. E. SHAFER AND E. M. SAWYER

Once, its key summary formula could be applied to Republicans, because they required only a simple majority for nomination, but not to Democrats, because they required two-thirds. After 1936, this one and only fundamental constraint on the bandwagon dynamic was to disappear. In that sense, nominating politics would become more, not less, varied as between the two major parties. There would be other changes, before and after 1936, that were capable in principle of reshaping this politics: different institutional arrangements, different party organizations and factional coalitions, different economies, technologies, and social structures, and of course different approaches to all of the above by the major players. Many of these evolutionary aspects did make contributions to the specifics—the ephemera—of nominating politics. None were sufficient to derail an underlying dynamic. Instead, the politics of presidential nomination in the United States was and is one of simultaneous mutual adjustment to the ongoing results of a sequential struggle, as registered through the creation of delegates to a national party convention. The convention as an institution once played much more of a role in assembling a nominating majority. Over the last fifty—but really seventy-five—years, it has mainly just confirmed a majority assembled in advance of the official gathering. Yet the underlying dynamic has been so stable and so consistent in its internal logic that it could burn through institutional reform and social change without noticeable impact, with that one great exception of the choice of something than other than a simple majority to confirm these nominations. At this point, accordingly, it is important to return to the opening metaphor of a funnel of causality. The bandwagon dynamic lives at— but really defines—the wide end of the funnel, the one farthest from the choice of a particular individual as a presidential nominee and the one within which all other factors exercise their influence. So an opening focus on any of these other alleged influences would be an analytic mistake: they are recurrently trumped by a grand and long-running bandwagon dynamic. On the other hand, being nearer to an ultimate nomination and more focused on specific individuals, these other factors recurrently look more consequential to a candidate or a strategist who cannot, after all, reshape the dynamic within which each must play their part. So it becomes essential to follow the funnel down from the bandwagon dynamic and toward those ultimate nominations. And the next step in doing so is to attend to the changing seedbeds, that is, the

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

33

occupational backgrounds and career trajectories, which have generated the fields of presidential aspirants over time, fields from which one or another individual nominee ultimately emerges. This is the central task of Chapter 2.

Notes 1. The resulting chain of interpretative comparisons could be additionally squeezed for suggestive break points, when an established pattern could be argued to have shifted, an approach deriving from the realignment synthesis in presidential elections, also on the rise at that point in time. See Walter Dean Burnham, Critical Elections and the Mainsprings of American Politics (New York: W.W. Norton, 1970), and James L. Sundquist, Dynamics of the Party System: Alignment and Realignment of Political Parties in the United States (Washington, DC: Brookings, 1973). 2. An earlier version of our own approach to the topic is Byron E. Shafer and Elizabeth M. Sawyer, “Institutional Dynamics and Nominating Processes”, Presidential Studies Quarterly 48(2018), 586–610. We are grateful to the editors of that journal for the ability to use elements of that article in this chapter here. 3. Donald S. Collat, Stanley Kelley, Jr., and Ronald Rogowski, “The End Game in Presidential Nominations”, American Political Science Review 75(1981), 426–435. 4. Larry M. Bartels, Presidential Primaries and the Dynamics of Public Choice (Princeton: Princeton University Press, 1988). 5. Ibid., 426. 6. Ibid., 428. 7. Bartels, Presidential Primaries and the Dynamics of Public Choice, 26. 8. Ibid., 5–6. 9. Ibid., 305. 10. Barbara Norrander, “The End Game in Post-Reform Presidential Nominations”, Journal of Politics 62(2000), 999–1013, and Norrander, “The Attrition Game: Initial Resources, Initial Contests, and the Exit of Candidates during the US Presidential Primary Season”, British Journal of Political Science 36(2006), 487–507. 11. For the nominating contests of 2016 as composite events, see James W. Ceaser, Andrew E. Busch, and John J. Pitney, Jr., Defying the Odds: The 2016 Elections and American Politics (Lanham: Rowman & Littlefield,

34

B. E. SHAFER AND E. M. SAWYER

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

22. 23.

2017), and John Sides, Michael Tesler, and Lynn Vavreck, Identity Politics: The 2016 Presidential Campaign and the Battle for the Meaning of America (Princeton: Princeton University Press, 2018). For details on delegate selection and their specific results, an extremely helpful source is The Green Papers, Election 2016: Presidential Primaries, Caucuses, and Conventions, https://www.thegreenpapers.com/P16/. For early analyses making use of these data, Michael Nelson, ed., The Elections of 2016 (Washington, DC: CQ Press, 2017). For the larger political context of that cataclysmic year, Michael F. Holt, The Election of 1860: “A Campaign Fraught with Consequences” (Lawrence: University Press of Kansas, 2017), most especially Chapter 5, “Republicans Choose an ‘Available’ Man”. For the ballots themselves, Richard C. Bain and Judith H. Parris, Conventions Decisions and Voting Records (Washington, DC: Brookings Institution, 1973), under “1860 Republican” in the section on “The Voting Records”, beginning at page 350 but thereafter unpaginated. Sources on a much-neglected president include Francis Russell, The Shadow of Blooming Grove: Warren G. Harding in His Times (New York: McGraw-Hill, 1968), and John W. Dean, Warren G. Harding (New York: Times Books, 2004). For the ballots themselves, Bain and Parris, op. cit., “1920 Republican”. For an effort to unpack the metaphor in its historical setting, Wesley M. Bagby, “The ‘Smoke-Filled Room’ and the Nomination of Warren G. Harding”, Mississippi Valley Historical Review 41(1955), 657–674. Rules adopted by the Anti-Masons in 1831 actually required a threequarters majority for nomination. Yet both their presidential and vice presidential candidates achieved three-quarters without requiring so much as a second ballot in 1831, while by 1836, they had declined to the point where they merely endorsed the Whig candidacy of William Henry Harrison. For the formation of the Republican Party more generally, William E. Gienapp, The Origins of the Republican Party, 1852–1856 (New York: Oxford University Press, 1988). Republican conventions in the Civil War era actually barred some southern states, and then readmitted former Confederate states gradually, so the majority continued to be calculated on the total vote apportioned to states actually present. Donald B. Cole, Martin Van Buren and the American Political System (Princeton: Princeton University Press, 1984), and Joel H. Silbey, Martin Van Buren and the Emergence of American Popular Politics (Lanham: Rowman & Littlefield, 2005). Nine southern states had bolted the convention, making two-thirds almost a mathematical impossibility. See Chapter 6 for the larger story.

1

THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL …

35

24. Bain and Parris, op. cit., “Twenty-Sixth Democratic Convention”, 238– 244, and “Twenty-Seventh Democratic Convention”, 247–250. 25. The most accessible condensed survey is “Convention Chronology”, Chapter 4 in National Party Conventions, 1831–2008 (Washington, DC: CQ Press, 2010). More detail is available in Bain & Parris, op. cit. 26. Though one might argue that the Whigs of 1852 had been so seriously stressed by their effort to find a nominee—or by the social divisions making that effort so difficult—that the result was a contribution all its own to the fact that the party would be effectively defunct by 1856. 27. For the ballots themselves, Bain and Parris, op. cit., “1844 Democrats”. 28. For the larger context, Sam W. Haynes, James K. Polk and the Expansionist Impulse (New York: Harper Collins, 1997), and John Seigenthaler, James K. Polk (New York: Times Books, 2004). 29. For the larger context, Lewis L. Gould, Four Hats in the Ring: The 1912 Election and the Birth of Modern American Politics (Lawrence: University Press of Kansas, 2008). See also Kendrick A. Clements, The Presidency of Woodrow Wilson (Lawrence: University Press of Kansas, 1992, especially Chapter 2, “1912”. 30. For the ballots, Bain and Parris, op. cit., “1912 Democrats.” 31. For the record, Clark would never actually secure the jump in support necessary to pass the 0.360 standard under a simple-majority rule. Though that fact hardly mattered, since he did secure what would have been an actual nominating majority on the 10th ballot under such a standard. 32. For the larger context, Michael F. Holt, Franklin Pierce (New York: Times Books, 2010). For the votes, Bain and Parris, op. cit., “1852 Democrats”. 33. For the ballots, Bain and Parris, op. cit., “1868 Democrats”, “1920 Democrats”, and “1924 Democrats”.

CHAPTER 2

Deep Channels to a Presidency: Occupations and Careers

Who seeks the presidential nominations that will be most crucially shaped by a long-running bandwagon dynamic? Or more precisely, which occupational categories and career trajectories regularly produce the fields of presidential aspirants who collectively constitute the next level of the funnel of causality, a funnel that ultimately disgorges one particular individual at one specific point in time? Alleged breeding grounds for a presidential nomination are in fact a long-time focus in the study of the politics of presidential selection, though the product is often more anecdotal than analytic, concentrating largely on the prior occupations of the most recent nominees.1 Yet candidate occupations and career pathways do provide the obvious next further route into a search for patterns in the pursuit of the presidency.2 The bandwagon dynamic undergirding presidential nominations has itself been impressively stable. That was the central message of Chapter 1. Yet this diagnostic stability has been masked from contest to contest by the great variety of candidates, issues, and strategies that turn up inside actual nominating campaigns. And first among these varying details are the identities of the aspirants who run for president, most especially the identities of the winners, where the most common analytic way to cut into these candidate identities is by way of their occupational choices and career pathways.

© The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_2

37

38

B. E. SHAFER AND E. M. SAWYER

So we shall begin with this focus too. But the point here is not ultimately to see whether there are occupations and careers that facilitate a presidential nomination. There are, and they do, but not by much. From one side, occupational backgrounds can indeed bring advantages that are critical at particular points in time. Yet from the other side, these impacts remain largely outside the control of presidential contenders. The point here is instead to turn this focus around, using the success of various occupational categories and career trajectories to get at the larger structure of the nominating politics that produce them, the one that does the real work of constraining the field of plausible nominees, though one that—unlike the bandwagon—changes gradually but ineluctably. With this as a focus, the questions become straightforward. Do aspiring contenders need to possess a career in politics? If they do, when must it begin? If not, what other occupational preparations will suffice? Within explicitly political careers, does the holding of one public office as opposed to another advantage (or disadvantage) some contenders over others on the way to a major-party nomination? Does this differ by party? Does it change over time? In more aggregate terms, do certain career pathways facilitate (or, again, retard) ultimate success in seeking a majorparty nomination? Does this too differ by party? Does it also change over time? To jump ahead of the story: there are in fact three major factors shaping the specific occupations and career pathways that become the dominant seedbeds for presidential nominations. These are the mechanics of the nominating process, the internal character of the political parties, and the balance between them in the larger society. Combining and recombining these three structural elements goes a long way toward explaining the occupational and career advantages that characterize five distinct institutional periods to presidential nominating politics. Along the way, federal executives, congresspersons, governors, and senators— along with the intermittent military hero—all have their day. A distinctly modern era likewise emerges, with a tantalizing twist at the very end. Candidates do occasionally amend their career progressions to conform to what they perceive to be the dominant occupational advantage of their time. Yet such choices remain overwhelmingly a product, not a cause, of the structure of nominating politics. Put more theoretically, early occupational choices by eventual presidential aspirants do not shape the overall structure of the nominating politics within which they will subsequently compete. But key elements of that structure do provide resources

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

39

and impose constraints on occupational and career successes and failures. So there is good reason to believe that prior occupations and facilitative careers remain a good route into the larger structure of nominating politics.

Seedbeds for Presidential Candidacies The simplest way to begin such an analytic effort is to set up and examine the record of occupations held previously by ultimate presidential nominees, and the resulting taxonomy has two immediate if unsurprising findings. First, it confirms that these nominees were overwhelmingly likely to have arrived by way of prior political offices, though the handful of nonpartisan occupations that have vaulted their occupants into a nomination will require a bit of attention on its own. Second, and apart from those rare exceptions, the prior careers of successful nominees appear to be much better used as evidence of the structure of politics at the time these individuals were nominated than as shapers of that structure— always remembering that these diagnostic careers do change over time in tandem with the politics that produced them. Within this total picture, three aspects of structural change will prove continuingly and recurrently important. Changing institutions for delegate selection regularly influence the evolutionary change in the occupations and careers of nominees. So do the internal structures of the political parties, principally the organizational character of the active parties, the ones doing the actual work of politics, but also the social coalitions that constitute their rank and file. So, thirdly, does the evolving balance between the major parties, and hence their prospective electoral fortunes. Ultimately, of course, these three are always joined by the particular challenges facing the American government at any given point, along with the instruments available to address those challenges. That said, even a simple tabular presentation of the occupations of major-party nominees becomes, immediately and perhaps inevitably, not simple at all. Just how should both occupations and careers be tabulated? Does holding many different offices count the same as holding one office, if they are held for the same period of time? If it does, should both be counted the same as the position held by a contender who served only one term in one office before successfully securing a nomination for president? One term in each of the two offices?, etc. Do six years as a Senator twenty years ago count the same as six years as a Senator directly

40

B. E. SHAFER AND E. M. SAWYER

before being nominated for president? Do the most recent six years as Senator trump twenty previous years spent in the House? Should both be counted? Should both be counted the same? Are they additive? Potential elaborations on these questions are endless, even in principle but certainly in practice, as a careful tour of major-party nominees from 1832 (or 1789 in some cases) will confirm. On the other hand, the purpose of the analysis does provide some immediate guidance. To wit: if the central purpose is to ask which occupations at which times helped to provide a connection to national party politics sufficient to underpin a presidential nomination—again remembering that the meaning of “sufficient political connection” can itself vary with the structure of nominating politics—then there would seem to be two features of any incipiently political occupation that seem especially worth noting: • First and most obvious is the post or office held immediately before an effort at the presidential nomination. This most proximate position is in some sense the temporal springboard for that effort. • The main alternative is the occupation that first catapulted a presidential aspirant to national attention. Usually, first occupations mainly open subsequent opportunities for other political offices. But sometimes, they overshadow everything that follows. Within those strictures, it ought to be possible to do a crude tally of the offices nearest to a successful nomination that was held by the ultimate nominee, an effort that will provide a kind of perverse reassurance. In principle, such a tally is most useful as a route into unpacking the structure of nominating politics around and behind it at various points in time. In practice, the resultant tally will behave in just that fashion, doing nothing that would counteract an underlying bandwagon dynamic— with the partial exceptions of the presidency and the vice presidency themselves—yet contributing a noteworthy influence on the individual identities of the candidates who ultimately benefit from this dynamic. Table 2.1 attempts this initial form of presentation. Column A covers the period before presidential nominations were formally conveyed at national party conventions, indeed before they were produced by any systematic and regularized process. This opening period, 1789–1828, does not lend itself to some of the analyses in the chapters to follow. Yet it is possible to put occupational backgrounds for these years

– – – 4 – 2 – 1 – 6

– – – 9 – 4 – – – –

3 3 1 1 – 4 1 1 – 2

Na. 3 3 1 3 – 5 1 – – –

Br. 1 3 5 1 1 3 1 – 2 7

Na. 2 4 9 1 1 4 1 – 2 –

Br.

i. Antebellum Br.

1864–1908 ii. Postbellum

1832–1860

1789–1828

Na.

B. Pure convention system

A. Pre-convention

Preceding occupations of presidential nominees

Senator Congressman Governor Executive Judiciary General None Vice Pres Nominee President

Table 2.1

4 – 9 2 1 1 1 2 – 10

Na.

1912–1968

C. Mixed system

7 – 14 4 1 2 1 1 – –

Br.

6 – 6 1 – – 1 3 – 9

Na.

1972–2020

8 1 10 2 – – 2 3 – –

Br.

D. Plebiscitary system

14 6 21 9 2 10 4 7 2 34

Na.

1789–2020

20 8 34 16 2 15 5 4 2 –

Br.

E. All-time results

2 DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

41

42

B. E. SHAFER AND E. M. SAWYER

into Table 2.1, and doing so will at a minimum confirm the distinctiveness of the period.3 Column B is then the long period best gathered as the “pure convention system” of presidential nominations. These years, 1832–1908, are the time when national party conventions were created and remained comprised of delegates generated solely by state party machinery, that is, internal party mechanics, usually in the form of state party conventions. On the other hand, this long period was divided by the Civil War, that great rupture of the American political process, so it seems important to compare the two behavioral sides of this era. Accordingly, Table 2.1 further divides Column B into: (i) the antebellum years leading up to the Civil War and (ii) the postbellum years leading away from it. Party structures are not dissimilar across this great break, but party balance certainly is. Column C then covers the extended period, 1912–1968, when presidential primaries were injected into this institutional picture, resulting in what is often referred to as the mixed system of presidential nominations. Most delegates to national party conventions were still created by state convention systems, as they had been since nomination by national party convention was institutionalized. Yet a minority of states adapted the primary election to presidential nominations, allowing the expression of a presidential candidate preference or choosing national convention delegates through the state primary ballot, and it was this mix of state conventions and state primaries that formally created a national nominee. Column D, for the period 1972–2016, is the modern world. A small set of reformed convention systems, usually referred to as participatory caucuses, did survive. Yet presidential nominations were overwhelmingly dominated by presidential primaries, additionally reformed to guarantee that actual candidate choices were what shaped the state delegations sent forward to the national convention. The result was rightfully called a plebiscitary system, whereby a sequence of state primaries influenced each other and produced a nomination that was formalized but no longer truly created at the national convention. Column E, lastly, puts the whole picture back together, summing Columns A–D. In a sense, this creates the whole American story. In a different sense, it underlines the opposite, that is, the degree to which different periods featured evidently different routes to a presidential nomination. Either way, it leads on to a focus on the linkage between the occupations of presidential nominees and the larger structure of American

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

43

politics, within which those nominations were realized and through which the individual identities of presidential contenders were determined. All five columns (and six temporal divisions) are further divided to show the application of a narrow versus a broad definition of prior occupational backgrounds. The narrow definition refers solely to the occupation nearest in time to the most recent appearance on a national ticket. One quirk of this (narrow) definition is that the most recent office held by a sitting president seeking reelection is, of course, “president,” and this is rendered accordingly. Just as the most recent office held by a vice president who subsequently sought the presidency must be classified as “vice president.” By contrast, the broad definition stays with the occupation closest to the initial appearance on a national party ticket. Thus incumbent presidents standing for reelection who were, for example, governors at the time they were first nominated are counted as governors each time they run, and the same can be said of vice presidents. This prevents occupational backgrounds and career pathways—the springboards for a presidential nomination—from being artificially reduced and oddly sampled by presidential incumbencies.

The Oldest World of Presidential Nominations What strikes the eye most immediately in Table 2.1 is the degree to which the years before nomination via national party convention, the years 1789–1828, are different in kind from all those that followed. This stands out starkly in the broad definition of preceding occupations, that is, either the position held immediately before first being nominated for president or the preceding position that effectively defined a future candidate by occupation (Table 2.1.A). In this earliest period, all presidents of the United States had served in major executive positions in the new government, counting military leadership as a governmental occupation. Generals would provide presidential contenders on a disproportionate basis through the nineteenth century, but federal executive office as a category would never again serve as such a serious springboard to the office—a modest irony, since the federal executive branch would grow only larger and more consequential. This early period also serves as a good introduction to the difference between a narrow and a broad definition of the relationship between previous occupations and presidential nominations, making the impact of this distinction especially easy to see (Table 2.2). In the narrow defini-

44

B. E. SHAFER AND E. M. SAWYER

Table 2.2 Preceding occupations of presidential nominees: the pre-convention era, 1789–1828

Senator Congress Governor Executive Judiciary General None Vice Pres Nominee President

Na.

Br.

– – – 4 – 2 – 1 – 6

– – – 9 – 4 – – – –

tion of previous occupation—the office held immediately before standing as a nominee for president—literally all of these candidates would have had to answer “the presidency” at least once. George Washington, John Adams, Thomas Jefferson, James Madison, James Monroe, and John Quincy Adams were all presidents when they ran for reelection, though both Adamses were unsuccessful in this regard.4 So the narrow definition reminds us how common in this period (or uncommon in others) a prior presidency was, while the broad definition is otherwise more helpful in talking about occupational and career patterns across time. With a focus on this narrow definition and on the larger political structure that generated its results, this “pre-convention era”, could be made to look even more distinct. For in fact—a situation without analogy in any other period—the first five presidents in this opening era had not just served in major executive office. Two of them, Washington and Madison, had been present at the Constitutional Convention, the one producing the overall framework of government within which they became the highest elected official and chief executive. Two others, Adams and Jefferson, were already serving in federal executive office at that point, while Monroe occupied a series of national offices in company with the members of this “founding generation”. And just to put icing on the metaphorical cake, the sixth president, John Quincy Adams, was the son of the second president, making him a sort of ‘founding offspring’.5 Moreover, in the interim between the Constitutional Convention and its gift of a presidency, all but George Washington, the initial president, could provide prior national service within the new federal government by way of conventional executive office. Though Washington in some

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

45

sense provided the greatest additional service of all by commanding the federal army in the war for independence. Thus Adams had been Minister Plenipotentiary and then Minister to Britain before becoming Vice President to Washington. Jefferson had been Minister to France and then Secretary of State under Washington before becoming Vice President to Adams. Madison had been a major author of “The Federalist Papers” in support of ratification for the Constitution, then Secretary of State under Jefferson. Monroe had been both Secretary of State and Secretary of War under Madison. And J.Q. Adams had been Secretary of State under Monroe, negotiating major international agreements and actually drafting what became otherwise known as the “Monroe Doctrine.” So that was a picture of the formal resources that facilitated a presidential nomination in this early period. From the informal side, the key buttress to this biographical pattern was a proto-party system, albeit one that quickly collapsed. Many of the founders had hoped to avoid a comprehensive party system entirely, but the very success of their new constitution was to lead more or less directly to partisan coalitions whose members desired to manage the nascent government. These, however, remained loose elite coalitions, Federalists versus DemocraticRepublicans, while Jefferson was to begin a sequence of DemocraticRepublican presidents.6 Even more to the practical point, even these proto-parties were to collapse early as devices for generating and managing a major-party contest between presidential nominees, and the absence of continuing party competition made it additionally easy to transfer the presidency among Democratic-Republicans who possessed the appropriate background credentials. Jefferson, Madison, and Monroe still saw some electoral votes cast for presidential alternatives—nominally for C.C. Pinckney, DeWitt Clinton, and Rufus King, respectively, as Federalists—but oppositional activity was largely confined to New England and disappeared entirely in 1820.

The Pure Convention System The Antebellum World The occupational backgrounds and career trajectories on the way to a presidential nomination would be very different in the successor period, the antebellum years of the pure convention system (Table 2.3.A). What would come most notably to life as the leading alternative to executive

46

B. E. SHAFER AND E. M. SAWYER

Table 2.3 Preceding occupations of presidential nominees: the antebellum era, 1832–1860 A. Narrow vs. Broad

Senator Congress Governor Executive Judiciary General None Vice Pres Nominee President

B. Dems vs. Whig/Reps

Na.

Br.

Dem

W/Rep

3 3 1 1 – 4 1 1 – 2

3 3 1 3 – 5 1 – –

3 – 1 3 – 1 –

– 3 – – – 4 1

–

–

office in this successor period was service in Congress. Both the House and the Senate were to jump up as seedbeds for presidential nominations, though the House would be the more important, across both antebellum and postbellum years. Service in the executive branch fell away, but a military background did not. So the central point is that the coming of the pure convention system elevated service in Congress as a key element of national introduction for presidential aspirants in the antebellum years, and sustained it into the postbellum period7 (Table 2.1.B.i and B.ii). The two contests between John Q. Adams and Andrew Jackson, in 1824 and 1828, were responsible for bringing the politics of presidential selection out of the shadow of the founding fathers and toward a more regularized process of both nomination and election.8 And this time, though not every time as we shall see, a major institutional change was central to the shift in requisite candidate backgrounds. General Andrew Jackson lost the first of these contests, winning the states which had already put their electoral votes on the ballot but losing those which continued to allow state legislatures to choose electors. Yet when Jackson won the rematch, his victory heralded a set of major shifts in the larger nature of nominating politics. The first part of the institutional transition that replaced Adams, scion of the old order, with Jackson, harbinger of the new, was the overwhelming spread of direct public choice of presidential electors. In 1828 as in 1824, Jackson won heavily wherever there was a mass electorate.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

47

But this time, the balance had swung solidly toward direct election, so Jackson’s popular vote was sufficient for an indisputable victory. Yet there was to be a second great institutional transition associated with the 1832 election, the one for which Jackson was renominated. This further shift was the arrival of national party conventions as the means for conveying major-party nominations for president, a transition likewise destined to last into the modern world. In passing, the 1832 conventions kicked off the long “pure convention era,” to be marked by a major change in the occupational and career patterns of presidential nominees.9 Yet direct institutional shifts in the mechanisms for presidential nomination as well as election were to be accompanied by—driven but also facilitated by—a social shift of the most profound sort. This break point in the larger society was stereotypically symbolized by the replacement of Adams—wealthy, eastern, highly educated, and connected by birth to the pre-convention era—with Jackson: frontiersman, westerner, perfunctorily educated, and a founding father to the pure convention era instead. All of this was in turn to be embedded in new party structures, a major change in the organizational character of the political parties—finally and collectively registered through a major change in the previous occupations and prior careers of the main presidential contenders. For the antebellum part of this pure convention era, these changes would be accompanied by a new party balance. The party conflict underpinning the change in occupational and career seedbeds would feature an ongoing Democratic Party, symbolized by Jackson though most centrally constructed by his vice president, Martin Van Buren. This enduring party would be confronted by a sequence of opposition parties, shifting but self-conscious coalitions that formed and reformed in response to the Democrats. First to arrive were the short-lived National Republicans. Then came the fissiparous but longer-lived Whigs. Who finally gave way to the Republicans, part and parcel of the process by which not just the party system but the nation itself came apart in a Civil War.10 Despite the organizational instability of opposition forces in these antebellum years, those forces were sufficiently overlapping to play a full part in generating the change in the career base of presidential nominees. As a result, they were likewise continuous enough to make it necessary to break out and check the occupational and career patterns of these nominees by party. Table 2.4 offers this in the same comprehensive format as Table 2.1. Democratic and opposition parties now get separate columns within their periods, with the single entries for each of these taken from

3 – 1 3 – 1 – – –

– 3 – – – 4 1 – –

– 1 5 – 1 2 1 2 –

D 2 3 4 1 – 2 – – –

R 4 – 10 1 – – – – –

D 3 – 4 3 1 2 1 – 1

R

ii. Postbellum

i. Antebellum

W/R

1912–1968

1864–1908

1832–1860

D

C. Mixed system

B. Pure convention system

6 – 5 1 – – – – 1

D

1972–2020

2 1 5 1 – – 2 – 2

R

D. Plebiscitary system

Preceding occupations of presidential nominees, by party

Senator Congressman Governor Executive Judiciary General None Nominee Vice Pres

Table 2.4

13 1 21 5 1 3 1 2 1

D

1832–2020

E. All-time results

7 7 13 4 1 8 4 – 3

R

8 – 5 – – – – – –

D

1972–2020

4 1 5 1 – – 2 – –

R

F. Plebiscitary recalculations

48 B. E. SHAFER AND E. M. SAWYER

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

49

our original “broad” definition. In other words, Table 2.4 gathers the positions held by presidential nominees directly before their first nominations, except where their public identities were self-evidently created by a specific occupation or position previous to this nomination. Service in the two houses of Congress, considered together, clearly displaced cabinet secretaryships as the leading producers of presidential nominees in the antebellum period. The need was still for political figures with a standing sufficient to attract national support. But now, it was the support of far-flung party leaders in the states that was necessary, in order to allow them to sell the nominee to their voting publics. And now, in an established governmental system, the obvious place to find these figures was Congress, both the House and the Senate. From the other side, the founding fathers were simply gone, accidentally but effectively taking the automatic role of major administrative office away with them. Breaking these results into Democratic and opposition clusters does show further distinctions between the two houses of Congress, and these are broken out for the antebellum years in Table 2.3.B. Yet these initial partisan distinctions also owe a great deal to one individual politician. That politician was Henry Clay, whose political prominence (and putative presidential caliber) were in turn owed chiefly to his role as Speaker of the House during three nonconsecutive terms. The Speakership more or less automatically made him a national figure, with the result that Clay was the national Republican candidate against Andrew Jackson in 1832; he was an unsuccessful challenger for the Whig nomination in 1840; he returned as the actual Whig candidate for President in 1844; and he returned unsuccessfully as a major challenger for the Whig nomination in 1848. For taxonomic purposes, the other presidential nominee in this period whose prior office was in the House was Abraham Lincoln in 1860. Yet Lincoln served only a single distant term, 1847–1849, and his prepresidential prominence owed more to a losing campaign for the Senate in 1858, the one that spawned the Lincoln–Douglas debates and brought him national attention inside the newly emergent Republican Party. In truth, Andrew Jackson, Democrat, had also been elected to the House, but he had found it unattractive and stayed for only nine months. While just to complete the inventory, John Quincy Adams—with Jackson, bridging the transition between pre-convention and pure convention systems—served eighteen years in Congress after his presidency, forging a long and distinguished congressional career.

50

B. E. SHAFER AND E. M. SAWYER

Formalistically, on the other side of the aisle, the Democrats put forward three different Senators, not Representatives, as their presidential nominees during the antebellum period: Lewis Cass of Michigan in 1848; Franklin Pierce of New Hampshire in 1852—when he bested Cass as well as Stephen Douglass of Illinois; and Douglas in his own right in 1860 when the Democrats splintered off a second proto-party for the general election. In losing the presidential election of 1848, Cass remained in the Senate. Yet Pierce had been a congressman as well as a Senator, and was out of both offices when he became a presidential nominee. Douglas likewise had both a House and a Senate career, so that the rise of Congress as a seedbed for presidential nominations looks a lot clearer than any further distinction by chamber or party. In the same way, summary statistics (in Table 2.3.B.i) might suggest that a tendency to elevate presidential candidates by way of executive office survived within the Democratic but not the Republican Party during these years. But in fact, two of the three apparent Democratic nominations with a prior executive background were of Martin Van Buren, Secretary of State in the first Jackson administration but a man whose real job had been to create the machinery for a national Democratic Party and whose formal job by the time he himself ran for president was Vice President under Jackson.11 In practical terms, then, holding major executive office at the federal level would never again recover an influence anything like the opening period, though it would also never die away completely either: Hillary Clinton in 2016, with a last previous office as Secretary of State, would still offer a distant echo. The generals were a very different story, as well as an important part of the reason why Congress contributed many nominees but few presidents in this period. Though any attempt to come to grips with the overall category must begin by dividing it in two. Throughout all of American history, and scattered among all the presidential nominees in Tables 2.1 and 2.4, have been three nation-risking military conflicts: the Revolutionary War, the Civil War, and World War II. These are in turn responsible for three nominees (and three resultant presidencies), each reelected comfortably. From the Revolutionary War, there is George Washington; from the Civil War, Ulysses S. Grant; and from World War II, Dwight D. Eisenhower. A desire by one or the other political parties to front them as a presidential nominee was still necessary for their nomination, but that desire was always forthcoming. Indeed, there was talk of Eisenhower as a substitute nominee for Democrat Harry Truman in

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

51

1948, while it had been unclear in the year before his nomination that Grant was a Republican and not a Democrat. A more common but very different subcategory of generals as presidential aspirants featured military heroism less as an indisputable virtue and more as the practical solution to a specific political problem. We have already addressed General Andrew Jackson, the lone Democrat in this tally for the antebellum years. But in fact, the total of generals as presidential nominees reached its apogee in this period, and all four of the other military candidacies—William Henry Harrison twice, then Zachary Taylor, then Winfield Scott—were a response to strategic dilemmas within the opposition Whig Party. Always more loosely coalitional than the Democrats, the Whigs had reasonable success in gaining congressional majorities from what was a geographically dispersed coalition but greater difficulty in promoting individuals to the presidency from their congressional leadership—Henry Clay was the archetype, both positive and negative—because these individuals tended to have strong geographic associations that were nationally problematic.12 So former military heroes became the more or less natural solution. Widely known and generally admired, they also tended not to come with programmatic baggage. The Whig Party was still in its formative period in 1836 when it turned to General William Henry Harrison of Ohio as one of its three national nominees.13 It returned to him in 1840, when Harrison succeeded in defeating President Martin Van Buren, only to die in office. The party managed to coalesce around their leading congressional politician, Henry Clay, in 1844, but that venture was unsuccessful in the general election. In response, they turned back to another military leader, General Zachary Taylor, in 1848. Taylor, like Harrison, was successful at the ballot box but again died in office. The party was to turn one more time to a military figure, General Winfield Scott towards the end of his fifty-three-year career in uniform, and he became the nominee in 1852, albeit unsuccessfully against Democrat Franklin Pierce.14 Otherwise, the period featured one stray Governor for the Democrats, former Governor James K. Polk of Tennessee, the original dark horse, and none for the opposition parties. Conversely, it featured one nominee who was effectively outside of politics, from the opposition this time. John C. Fremont had actually served part of a term as Senator from California, though he found it distasteful and abandoned it quickly. His national recognition was in any case based on a high-profile role as an explorer of the American West, while he too had the advantage of being almost free

52

B. E. SHAFER AND E. M. SAWYER

of programmatic attachments. So a man without a true political career became the first nominee of the new Republican Party at its first national nominating convention. The Postbellum World Some major aspects of the occupational and career underpinnings of presidential selection merely resumed after the Civil War15 (Table 2.5.A). That was the story of the continuing place of Congress as a launching pad for presidential nominations. The demise of federal executive office as the main springboard to the presidency, in conjunction with the demise of the early proto-party system, had created a need for some new source of presidential nominees during the antebellum part of the pure convention era. Congress had filled this void, buttressed by its place at the center of prewar conflicts over territorial expansion, financial and infrastructure management, and, more and more, the place of slavery. The issues would change in the postwar period, but Congress would continue to play roughly the same role. Likewise continuing was the intermittent retreat to military heroes, though both types of political generalship would surface in this postbellum era. There was the heroic champion of a nation-saving campaign in Ulysses S. Grant, commander of the victorious Union army, successful Table 2.5 Preceding occupations of presidential nominees: the postbellum era, 1864–1908 A. Narrow vs. Broad

Senator Cong’man Governor Executive Judiciary General None Vice Pres Nominee President

B. Dems vs. Reps

Na.

Br.

Dem

Rep

1 3 5 1 1 3 1 – 2 7

2 4 9 1 1 4 1 – 2 –

– 1 5 – 1 2 1

2 3 4 1 – 2 –

2

–

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

53

nominee and president in 1868 and 1872, but failed contender for a third nomination when he tried to return in 1880. Yet there were numerically more of the second type of general, the ones whose military career gifted them with a respectability separate from their party, yet available for partisan contests where the lack of both a factional and a policy history was a virtue. This time, however, it would be the Democrats and not the Whigs who saw virtue in this type of candidate. Yet what distinguished the period when occupational background and career trajectories were the focus was the rise of governorships. Governors had remained comparatively inconsequential in the antebellum part of the pure convention era, as they had been in the pre-convention period. Yet governorships would come into their own in these postbellum years, an occupational rise that would last all the way into the modern period. Along the way, governors would surpass members of both houses of Congress, even when the latter were added together. So the major change that needed to be explained—by institutional arrangements, party structures, or interparty competition—was the rise of these governors. Institutional arrangements had little or nothing to contribute to this story: the pure convention system of delegate selection and presidential nomination extended across the Civil War in an effectively unchanged fashion. Yet the internal nature of the parties that did their business through these institutional arrangements did change in powerful and diagnostic ways. Indeed, it was this structural change that would tie most directly to the rise of the governors. Though beyond that, the competitive balance between those two parties actually reversed. Where the opposition to the Democrats had faced the greater challenges in nomination politics during most of the years before the war, it was the Democrats themselves who faced the greater challenges afterward. In the particular structural shift that crucially underpinned the rise of the governorships, the postbellum years were to be the great era of machine politics in the United States, now for two stable and institutionalized parties. Yet these remained profoundly state-based organizations. National links were not irrelevant, but the concrete rewards from growth, urbanization, industrialization, and immigration, generating the huge array of capital projects that became the “spoils” of machine politics, were largely based in state government. Governors were not always the operative leaders of the organized parties that resulted, though even when they were not, their identities remained a central concern of state politics. Gubernatorial incumbents might still lack a nationwide reputation with

54

B. E. SHAFER AND E. M. SAWYER

the general public—something they would come to acquire in the mixed system and then keep for the plebiscitary system—but they could readily build such a reputation among party leaders in other states.16 Table 2.5.A confirms the rise of these governorships to occupational primacy. Table 2.5.B goes on to confirm that this rise was truly bipartisan. So where executive office had distinguished the pre-convention period for presidential nominations and where Congress had jumped up as an institutional source in the antebellum part of the pure convention era, it was governorships—a state-based and not a national office—that were the obvious gainers in the postbellum years. Congress held onto its nonnegligible role, though this now required separate attention within the two parties. And a military background continued to have value, though this continuity too was registered in a different way as between the parties. Such partisan differences were actually a good introduction to the second major structural impact on occupational advantages. Put simply, these years would be a period of Republican dominance, lasting well into the mixed system of presidential nominations.17 This would make the occupational and career story of Republican nominations comparatively simple and straightforward. Across the partisan aisle, however, the period would feature only one Democratic President, Grover Cleveland—though he would secure three nominations—with the result that the Democratic electoral challenges stretching across the entire postbellum world divide further into mini-eras, each requiring some attention. First came the immediate postwar years, when the central task of the Democratic Party was to recover from association with the losing side in the Civil War and regain national competitiveness. Its first gambit in 1864, while the Civil War still raged, involved one of its military candidates, General George McClellan. The party would then try its first governor in 1868, with Horatio Seymour of New York. His party was still in such disarray in 1872 that it simply endorsed Horace Greeley, editor of the New York Tribune, one of the handful of major-party nominees across all of American history with no previous partisan office, who was already the nominee of the short-lived Liberal Republican Party. This was followed by another governor from New York, Samuel Tilden, who actually won the popular vote in 1876, but lost the disputed electoral vote within the halls of Congress. The party would then make its last attempt at solving its problems in this first mini-era with a second military hero, General Winfield Scott Hancock, who had previously contested its nomination

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

55

in 1868 and 1876 but was finally successful—as a nominee but not a president—in 1880. In the end, however, two generals, two governors, and one nonpolitician did not add up to a reconstituted winning coalition at the presidential level. Yet change—or at least, a second mini-era—did arrive in 1884 in the person of Grover Cleveland, Governor of New York, whose nomination and election were the product of two interacting structural influences. A reform Governor from New York had potentially one of the few pedigrees capable of finessing the factional split among northeastern immigrants, western progressives, and southern regionalists in the party as a whole, an asset greatly magnified by the fact that by far the largest state which could in principle be captured by either side was New York.18 Grover Cleveland was just such a candidate, and he did in fact capture both a nomination and a presidency in 1884. He was defeated for reelection by Benjamin Harrison in 1888, himself a Senator from Indiana, the next-largest state that could be captured by either party in this period. Yet Cleveland was able to turn the tables on Harrison four years later, returning as both nominee and president. On the other hand, similar successors, that is, Democratic reform governors from New York, proved to be in short supply. Indeed, another popular reform governor from New York would not appear on a presidential ticket until 1900, in the person of Republican Theodore Roosevelt. As a result, the remainder of the postbellum period became a third mini-era for the Democrats, though this time the party lacked even a cohort of governors with any serious hope of turning the situation around. Instead, it was to rally behind William Jennings Bryan of Nebraska for three of the four remaining presidential contests.19 Bryan had previously been a one-term congressman, so he must be registered as the lone congressional product for the Democrats in this period. Yet his second and third nominations were really due to his first presidential run and not to that one distant congressional term, so it would not be wrong to list all three of his runs under the category “nominee”—meaning that his performance at the 1896 Democratic convention, capped by the famous “Cross of Gold” speech, was really what underpinned all three of his presidential nominations. By contrast, the Republican story for this period was happier and simpler, escaping any need for mini-eras to describe its facilitating occupations.20 First came the two nominations and inescapable presidencies

56

B. E. SHAFER AND E. M. SAWYER

of the victorious Ulysses S. Grant in 1868 and 1872, though Grant partisans were to try unsuccessfully to bring him back for a third term in 1880. After Grant, with the exception of William Howard Taft in the final contest of these postbellum years, all nominees had either governorships or congressional service as the critical elements in their resumes. Indeed, among Republicans much more than among Democrats in this period, nominees frequently featured both of these congressional tracks: • William McKinley, two-term Republican president in 1896 and 1900, is tallied here as a governor because he was Governor of Ohio when he ran for the presidency, though he actually had an extended and influential career in the House before that. • Conversely, Rutherford B. Hayes, also twice elected Governor of Ohio, likewise began his political career with a term in the House before the first of two successful gubernatorial runs. • James A. Garfield, a nine-term Representative, had just been elected to the Senate when he stood as the Republican candidate for president, though he declined the former post when he accepted the latter. • James G. Blaine, contender for the nomination in 1876 and 1880 and nominee in 1884, had made his name in national politics by being Speaker of the House, though he was actually in the Senate when he was successfully nominated. • And in a final mix of careers, Theodore Roosevelt, Governor of New York when he became Vice President under William McKinley in 1900, owed much of his public prominence to a military career, when he headed the “Rough Riders” in the Spanish–American War.

The Mixed System of Presidential Nominations Governors would only increase in statistical prominence as nominees and presidents when the mixed system of presidential nomination arrived (Table 2.6.A). Their totals were artificially augmented by the four nominations (and elections) of Governor Franklin Roosevelt of New York. But in fact, governors were to lead the tally of offices in Table 2.6 under both the broad and the narrow definition of occupational background, that is, even when three of the four Roosevelt nominations are recorded as “president” rather than “governor.” Yet what was most clearly different in this

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

57

Table 2.6 Preceding occupations of presidential nominees: the mixed system, 1912–1968 A. Narrow vs. Broad

Senator Cong’man Governor Executive Judiciary General None Vice Pres Nominee President

B. Dems vs. Reps

Na.

Br.

Dem

Rep

4 – 9 2 1 1 1 2 – 10

7 – 14 4 1 2 1 1 – –

4 – 10 1 – – –

3 – 4 3 1 2 1

–

1

successor period was the rise of senatorships as an occupational forerunner to both nomination and election. Remarkably, Republican Warren G. Harding in 1920 was the first sitting senator ever elected president, approximately one hundred thirty years after the first occupant took his seat in the Senate. So the arrival of the mixed system of presidential nominations simultaneously marked the political coming-of-age of US Senators. Afterward, these two offices together, governor and senator, were to overwhelm all other occupational and career routes to a presidential nomination. Though this joint fresh development was shaped by a very different mix of the usual influences, namely institutional arrangements, internal party structures, and external party balance. Unlike the story in the antebellum era, a change in the occupational and career seedbeds of presidential aspirants in the era of the mixed system was to be accompanied by a major shift in the institutional arrangements for delegate selection and presidential nomination. Those arrangements had been effectively stable between the antebellum and postbellum worlds. Now, they made a major institutional change on the way into the mixed system via the introduction of presidential primary elections, the change that was actually responsible for making the new era a mix of convention systems and, now, primary elections.21

58

B. E. SHAFER AND E. M. SAWYER

Conversely, what was not different was the internal structure of the political parties. They were reliably under assault in their traditional machine-type embodiment by progressive reformers, beginning well before the mixed era began. Moreover, the new period was marked by obvious reform triumphs, as with the imposition of primary elections, a progressive favorite.22 Yet the operative nature of both parties was to bend but not break, allowing them to compensate for new reforms when they could not prevent them, even as the need for compensation attested to a changing climate for party politics. All of which added up to noteworthy change coupled with major aspects of stability, a combination that would fit comfortably and logically with the persistence of governors and the rise of senators as nominating contenders. The precise nature of both this persistence and that rise was further colored by an actual reversal of the party balance characterizing the previous era. The antebellum world had featured the national dominance of the Republican Party, with important but only occasional Democratic breakthroughs. The world of the mixed system would see this Republican hegemony continue in its opening years, then shift dramatically, thanks to the Great Depression and policy responses to it. This second stage was a Democratic hegemony that would last through the remainder of the period—and cause the occupational backgrounds of Republican versus Democratic nominees to reverse. The dominant party still had an easier time than the opposition party with the principal occupational and career demands of the period. It was just that the identities of those two parties would be reversed. But in the beginning, what made the mixed system “mixed,” and most clearly distinguished it from the antebellum era, was the arrival of those presidential primary elections.23 Formally, this arrival promised a whole new option in pursuit of a presidential nomination on what became known as “the primary trail.” In theory, candidates could circumvent the regular party at the start and force themselves upon it in the end. Practically, however, the reality was much more limited. The primaries were indeed a reform device, often forced upon the regular party. Yet party regulars hardly lost the ability to fight back,24 often by creating primary elections that offered presidential preference polls but selected the actual delegates through an old-fashioned party convention or by allowing the election of their national convention delegates but only in the absence of any presidential attachment.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

59

Regardless, during all the years of the mixed system, the bulk of the delegates still came from traditional party conventions. Which was part of the reason that many presidential aspirants looked askance at the newly available primaries. A national aspirant, unknown locally as most were, could easily demonstrate early weakness on the primary trail but could rarely hope to win enough actual delegates to make it worth much risk—though here, out-state senators were at least more likely to have a recognized profile than out-state governors. By extension, presidential primaries were commonly converted into what might best be called “demonstration events,” whereby individual candidates received a potential opportunity to confirm their putative electoral prowess if they performed well in (a few) carefully selected venues. Still, the primary route had its applications and their successes. Warren Harding entered and carried his home state of Ohio in the Republican primary in 1920, which proved important to his remaining a potentially electable alternative when a three-way contest arose among initially stronger contenders. Calvin Coolidge ran actively (and successfully) in the Republican primaries of 1924, thereby quieting any residual doubts about the democratic legitimacy of a vice president who had been elevated to the presidency only by the death of Harding. And Herbert Hoover followed by winning the big states that lacked favorite sons in his 1928 nominating campaign, though this was as much confirmation of a general move to Hoover within the party as it was a stimulus to that move.25 At the other extreme, there were cases where a presidential aspirant took to the primary trail because it appeared to be the only route to a nomination, and these were never happy stories. The most dramatic came in the first nominating contests under the mixed system of delegate selection. Former president Theodore Roosevelt, unhappy with the policy agenda of his handpicked successor, William Howard Taft, attempted to replace Taft as the Republican nominee in 1912. But since President Taft had become the champion of the regular party, Roosevelt turned to the nascent primary trail in order to demonstrate superior electability.26 He did in fact win nine of the twelve available Republican primaries in 1912, with Senator Robert M. LaFollette of Wisconsin winning two and President Taft securing only one. Yet the regular party did not flinch, and Taft was overwhelmingly renominated. Fifty years later, in what would be the last round of contests under the mixed system of delegate selection, Senator Eugene McCarthy of

60

B. E. SHAFER AND E. M. SAWYER

Minnesota attempted to use the primary trail to drive out sitting president Lyndon Johnson. Johnson won the opening primary, but the result drew Senator Robert Kennedy of New York into the contest, and when Johnson then withdrew, it was McCarthy plus Kennedy (later replaced by Senator George McGovern of south Dakota) attempting to overcome Vice President Hubert Humphrey. Humphrey became the champion of the regular party after the withdrawal of Johnson, more or less by default and too late to enter any presidential primaries. Yet while there were sixteen of these on the Democratic side of the aisle, Humphrey was nevertheless solidly nominated on the first ballot.27 On the other hand, the record of those cases where primaries did make a serious and partially autonomous contribution to a presidential nomination hints at a further partisan bias to this effect: consequential primary contests were heavily concentrated on the Republican side of the aisle (Table 2.6.B). Until the Democrats changed the two-thirds rule for nominations, even a chain of primary wins, if such could be extracted, would have left the winner far from the number of delegates needed for nomination, while simultaneously alerting other aspirants of their need to coalesce around alternatives and build the opposition by the time the convention actually convened. Not surprisingly, then, Democrats were better than Republicans at skipping the primaries entirely but being nominated anyway. All that said, the main occupational and career point is that the previous strategic and resource advantages of a governorship for presidential nominations in the postbellum part of the pure convention era proved adaptable to a system mixing party conventions and presidential primaries, though the sources of this adaptability did shift. Then, governors had featured the potential to lead state party machines, with their capacity to deliver convention delegates and mount electoral campaigns. Now, governors had the potential to bring proven electability plus a record of policy-making. So the main practical point for gubernatorial aspirants, both Republican and Democratic, was that an office which had possessed clearly relevant assets in the antebellum period could and did blend old and new assets successfully under the mixed system as well. Yet the new institutional forum—the presidential primary—proved even more consequential to at least some among the senators who rose to national prominence during this period. In the old world, it had ordinarily been governors and not senators who managed both inter- and intraparty business. Indeed, senatorships were often used as a reward from the

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

61

machine rather than a directorship of it. But the arrival of the presidential primary, coinciding as it did with the uniform arrival of direct election of Senators—and even the direct nomination of those Senators in many states—began to make potentially national figures out of an occupation that had been available in principle since ratification of the Constitution itself.28 Senators would continue to trail governors as leaders of the state party. But senatorial acquaintance with a wide range of policy issues and their ability to play to a growing national network of mass media increasingly compensated. Indeed, it was changes in the mass media of information that interacted with the public agenda available to these senators and helped fuel their rise, first by way of radio, then by way of television. In principle, both senators and governors could take advantage of a lessmediated access to the general public, hoping to “get around” reporters and talk to what was at least aspiringly “their public.” In practice, senators appeared better able than governors to use the events of their daily political lives to forge this connection, most especially with out-state publics.29 All that said, the other main structural contribution to the politics of presidential selection under the mixed system of presidential nominations—and with it, to the career patterns of successful nominees—once again came from party balance in the nation as a whole. The postbellum years had featured a straightforward and ongoing one-party domination by the Republicans, broken only by Grover Cleveland and not reliably even then. The opening years of the mixed system were kicked off by a second crack in that pattern of partisan domination, when the third-party candidacy of former President Roosevelt allowed Woodrow Wilson, the Democratic nominee, to secure the presidency. Wilson, however, would prove to be an exception and not a harbinger, with Republicans reclaiming the next three presidencies. So change would await the arrival of the Great Depression, blamed in some sense on the Republican Party, along with the arrival of the New Deal and the institutionalization of the welfare state. Together, these produced a new pattern of partisan dominance that would last for almost forty years. As a result, the mixed system of delegate selection and presidential nomination was to be characterized by two single-party stretches, a short stretch for the Republicans from 1920 to 1932 and a longer one for the Democrats from 1932 to 1968.

62

B. E. SHAFER AND E. M. SAWYER

From 1932 until 1968, the default result of any national election could be expected to produce a Democratic and not a Republican president.30 So the situation characterizing the postbellum years was effectively reversed. In the abstract, it was the dominant party that could be expected to tap the characteristic occupational base for presidential nominations, and the opposition party that would need to cast its nominating net more broadly. In specific occupational and career terms, this meant that the Democrats could rely on either their governors or their senators to provide not just nominees but ultimate presidents, while the Republicans would need to find candidates who could best counter the changed partisan expectation, by tapping into a more diverse occupational or career base (Table 2.6.B). The Democratic list was of course heavily colored by four nominations (and elections) for Governor Franklin Roosevelt of New York. He would be succeeded by his vice president, Senator Harry S. Truman of Missouri, who won his own presidency in 1948; followed by another governor, Adlai E. Stevenson of Illinois in both 1952 and 1956, as the lone essentially sacrificial candidate against the one Republican success story of the period, General Dwight Eisenhower; followed in turn by Senator John F. Kennedy of Massachusetts in 1960; then by Kennedy’s Vice President, Senator Lyndon B. Johnson of Texas in 1964; followed finally in 1968 by Johnson’s Vice President, Senator Hubert H. Humphrey of Minnesota. Like the Democrats in the opening years of the mixed system, the Republicans were to succeed only once. General Dwight Eisenhower, the American victor in World War II, was successfully nominated in 1952 and then reelected in 1956. Alf Landon, Governor of Kansas, had made the first (and most seriously sacrificial) effort to stem the new partisan tide in 1936. After that, the Republican list was indeed more varied than its Democratic counterparty. In 1940 the Republicans followed with the only presidential nominee under the mixed system without prior public office, in Wendell Willkie, corporate president and attorney in private practice. Then came Governor Thomas E. Dewey of New York in 1944 and 1948, General Eisenhower in 1952 and 1956, Vice President Richard Nixon in 1960, Senator Barry Goldwater of Arizona in 1964, and Nixon again in 1968.

2

63

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

The Plebiscitary System and the Modern World The plebiscitary period of presidential nominations, the one in which we all still live as this is written, would see only an exaggeration of the occupational and career favoritism that characterized the mixed system of delegate selection (Table 2.7.A). From 1972 to 2020 and counting, governors and senators would become an overwhelming share of the total story of presidential nominations, especially if the analyst used the broad definition of occupational background, the one assigning incumbent presidents to their previous position. Yet this continuity, favoring governors and senators even more than the preceding period, appeared initially to be accompanied by a large but ultimately superficial puzzle. The puzzle was that all three of the recurrently dominant influences on occupational preference, namely institutional arrangements, internal party structures, and external party balance, shifted substantially as the new era arrived, while the seedbeds for presidential nomination changed very little. Yet this initial puzzle could be quickly dismissed by noting that the reason that major shifts in all three recurrent influences drove intensified rather than altered results for occupational backgrounds and career trajectories was that all three intensified their movement in the same direction as they had exercised it previously. The matrix of institutions was indeed overhauled dramatically, to a far greater degree than the counterpart change at the beginning of the mixed system, but in the same general Table 2.7 Preceding occupations of presidential nominees: the plebiscitary system, 1972–2020 A. Narrow vs. Broad

Senator Congress Governor Executive Judiciary General None Vice Pres Nominee President

B. Dems vs. Reps

Na.

Br.

Dem

Rep

6 – 6 1 – – 1 3 – 9

8 1 10 2 – – 2 3 – –

6 – 5 1 – – –

2 1 5 1 – – 2

1

2

64

B. E. SHAFER AND E. M. SAWYER

direction. The internal nature of the political parties was likewise altered sharply, favoring the forces that had been only beginning to challenge the regular party in the previous era but were to secure a sweeping victory over them in this one. While the balance between the two parties shifted back toward the Republicans, creating a much more competitive landscape, one favoring partisan convergence in occupational preferences and career trajectories. Within this welter of powerful but parallel changes, the institutional arrangements for delegate selection and presidential nomination were overhauled in a fashion second only to the creation of the party system itself in the 1830s.31 Every individual piece of the process was altered, as was the balance among these pieces. Primary elections exploded, to become overwhelmingly the dominant institution in a reformed process. But now, these had to be reformed primaries, with delegate selection tied tightly to—and sometimes completely submerged in—public preferences among presidential aspirants. In the same way, the state convention systems that survived, in a small minority what tended to be smaller states even then, were forcefully reconstituted to feature open public meetings rather than meetings of party officialdom.32 The result was an almost pure plebiscitary system, in which state primary elections, plus the occasional caucus proceeded in sequential fashion, such that results in one week’s contests could inform the judgment of all the major players by the following week. Candidate strategists, party activists, financial donors, the mass media of information, and, last but not least, the participating general public were now caught up in what was effectively the modern version of the historic bandwagon. In other words, polished to an only slightly higher level of generalization, a thoroughly transformed nominating politics could still fit the description of a bandwagon dynamic that had rolled across American political history since the 1830s. This bandwagon now rolled through primary elections rather than convention ballots, yet even an institutional reform on this scale did not otherwise alter its fundamental nature. At the same time, the political parties themselves were transformed. As with institutional arrangements, so with party structures: the original creations of the 1830s had been extended comfortably into the postbellum world, thereby relegating the nominating politics of the mixed system to a kind of interim between these two worlds, the postbellum and the plebiscitary. But now, that future had arrived. Once and for a very long time, American political parties had been built around the

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

65

distribution of concrete benefits, often described by their opponents as the “spoils” of government but viewed from the opposite perspective as reflecting major divisible products of public policy. Regardless, these were parties now built around the participation of the more highly (but autonomously) active portion of the citizenry, the part more reliably mobilized by abstract policy issues.33 The coming of this modern world of the political parties had been at the center of party conflict in American politics for a very long time, at least since the appearance of the progressives in the late nineteenth century.34 The triumph of party reformers over party regulars, and hence of party activists over party officials, had been forecast again and again, as an allegedly inevitable development, since the late nineteenth century. Yet during the later years of the mixed system, American social scientists had inquired into its progress—and been surprised to find that the old order survived in a far more vigorous state than expected. But these professional students of party structure agreed that around 1970, just in time both to drive and to benefit from institutional reform, the activist side had finally begun to win this battle on a regular basis.35 Yet in the plebiscitary era, both forces, that is, changing institutional arrangements and changing party structures, came to their ascendancy in a world unlike either the Republican dominance of the antebellum period or the Democratic dominance of the mixed system. Instead, at least by hindsight, this Democratic dominance had already broken down by the election and reelection of Republican Richard Nixon in 1968 and 1972 (Table 2.7.B). The Democrats would recapture the presidency with Jimmy Carter in 1976, the Republicans would take it back again in 1980 with Ronald Reagan, and the overall party dominance that had characterized the American political world since roughly 1860 would disappear in favor of a much more competitive balance, where each party would be favored in some contests and where party control of the presidency would swing back and forth between them.36 This plebiscitary era was already a world where productive presidential seeds were planted largely among senators and governors. That was evident even by the narrow definition of occupational background, where sitting presidents were counted as presidents if they ran for renomination, even if they had previously been senators or governors. Inevitably, then, the broad definition, assigning these presidents to the positions they held before they were president, makes this joint senatorial and gubernatorial predominance much more evident (Table 2.7.A). Yet in a world where

66

B. E. SHAFER AND E. M. SAWYER

party support and party victories were more balanced than they have been since the antebellum period, that is, before the Civil War, it should not be surprising that separate Democratic and Republican pictures still captured mostly the same overall story, with a little more variety offered by the Republicans than by the Democrats (Table 2.7.B). Even then, our method of counting prior occupations suggests more partisan variety than a practical reality might justify since, while both parties achieved the presidency from senatorial roots by way of intermediary offices like the vice presidency, the Republicans were more likely to do so. In response, Table 2.4 adds a column (Column 4.F) that counts these intermediary routes during the plebiscitary era—as with senator to vice president to president—as belonging to the Senate and not the immediately intermediary office.37 Tallied this way, the Democrats have never nominated anyone for president in the plebiscitary era who has not been either a senator or a governor, with twelve such nominations out of the twelve possibilities. The Republicans show a bit less uniformity, with nine senatorial or gubernatorial nominations out of twelve, leaving room for the likes of Congressman Gerald Ford, the lone appointed president, and Donald Trump, the lone nominee in the modern world who lacked a formal political career. Many factors beyond institutional arrangements, internal party structures, and external party balance contributed to the social change that was nascent at the coming of the mixed system for delegate selection but instantiated subsequently with the arrival of the plebiscitary era. The dominance of senatorships and governorships remains one inescapable product of this collective social change—the crucial one for an examination of the occupational and career seedbeds of presidential nominations. Though other highly consequential elements of a comprehensive change might include the rise of mass media of information, the rising affluence and educational levels of a general public, the nationalization of many, perhaps most, aspects of American politics, and even the simple growth of government itself.38 Yet there is one other diagnostic product of nominating politics that surfaces at least implicitly in any account of the politics of presidential nominations. This is not just a product that is best approached through this politics. It is one that puts a very specific and concrete face on collective social change. This alteration in the nature of nominating politics involves the effective departure of presidential nominations from national

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

67

party convention in the 1950s and its effective migration into the sequential process of delegate selection. The Democratic convention of 1952 still featured the actual assembly and not just the official confirmation of nominating majorities at the convention itself, while the Republican nomination was realized only through switches after the opening ballot. After 1952, none of this would ever be true again, though commentators and especially the candidates and their campaigns would continue to believe otherwise for a very long time,39 up into the nominating contests of 2020.40 Nevertheless, from 1956 onward, all presidential nominations, again in both parties, would be resolved within the process of delegate selection, to be no more than confirmed by a majority on the opening ballot of their conventions. Early on, confirmation retained some suspense for some commentators; as time passed, even this would prove less and less true. Either way, that, internal party reform, or external party rebalancing, this by-product actually preceded—and in that sense set the stage for—the collective changes that would define the modern plebiscitary era. Measured by convention ballots, the Democrats had lived in a world all their own in the early years of the mixed system, requiring 46, 44, and 103 ballots to produce a nominee in 1912, 1920, and 1924, respectively. Yet Franklin Roosevelt would still need four ballots in his initial nominating success in 1932, and Adlai Stevenson would still need three in 1952, though his nomination would put the long period of multiballot contests to sleep forever among Democrats. The Republicans had never reached these heights of ballot conflict, but the nomination of the great nonpolitician of the time, Wendell Willkie in 1940, had required six ballots; Thomas Dewey required two in his renomination in 1948; and Dwight Eisenhower might well have needed a second ballot in 1952, before late shifts after a closely divided first ballot allowed him to avoid a second. Ironically, it was Stevenson, multi-ballot nominee for the 1952 Democrats, who actually (if accidentally) demarcated the departure of the nomination by the time of the Democratic convention of 1956. In 1952, Stevenson had been a late archetype of the old order, having eschewed the primaries—except for the one in his home state that he entered and lost—but surging to nomination by the third ballot anyway. Yet by 1956, Stevenson was instead to be a harbinger of the new world. Having overcome Senator Estes Kefauver of Tennessee as his leading challenger four years previously, where Kefauver had not only entered but won most primary contests, Stevenson felt the need to enter them when Kefauver tried again four years later. Winning the lion’s share this time, he in effect

68

B. E. SHAFER AND E. M. SAWYER

freed the regular party to line up once more behind him, which they did expeditiously at the 1956 convention, without needing so much as a second ballot.41 This story was then more powerfully underlined in 1960, when Senator John F. Kennedy of Massachusetts felt compelled to take to the primary trail because his youth, but especially his Catholicism, raised doubts in the minds of many in party officialdom that he could actually go on to defeat the emerging Republican nominee, Vice President Richard Nixon. Still, many of these party regulars preferred Kennedy in private, and they were watching closely as he collected some major primary wins, most especially in Wisconsin and then in West Virginia. That appeared to give them the reassurance necessary to come out actively for Kennedy, and he too would harvest this emerging support easily on the opening ballot.42 The story would be confirmed for the Republicans too in the nominating contest of 1964. Senator Barry Goldwater of Arizona, like Kennedy in 1960 among Democrats, raised even stronger fears within Republican party officialdom that he could not pose a serious challenge to incumbent President Lyndon Johnson. So Goldwater really had no choice but to go out on the primary trail, though his activist supporters in local Republican parties across the country also mounted deliberate challenges to party officeholders in the delegate selection contests of nonprimary states. The contest was long-running and apparently open-ended for much of the time. But when Goldwater was victorious in the huge, late, winner-take-all primary in California, he too needed only to confirm an emergent result on the first ballot of the national convention.43

Two Occupations That Mattered On the one hand, then, Stevenson, Kennedy, and Goldwater were all concrete incarnations of a set of social changes with powerful impacts on nominating politics, impacts that were institutionalized in the plebiscitary era after 1972. On the other hand, and much more important in closing the chapter, this plebiscitary era otherwise merely intensified the occupational backgrounds and career trajectories that had already come to characterize nominating politics, while the institutional arrangements, party structures, and electoral balance that did the work of institutionalizing this new era did nothing to suggest that the bandwagon dynamic as it rolled through the plebiscitary period had been altered or even redirected in any serious way.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

69

So the penultimate question becomes only: is there no prior office or career pathway that actually could overcome the influence of an ongoing bandwagon and thereby circumvent its impact on the preferred positions of the modern era? It should be said from the start that it is not clear that there is any positive answer to those questions. Yet it is important to note that one prime contender presents the clearest challenge, surpassing all others in its potential to produce a presidency merely by virtue of holding an office and thus the one that might go on to alter the actual policy outputs associated with the relevant presidency. That office is, of course, the vice presidency, a position that has acquired a great deal of disrespect across American history yet one that intermittently contravenes the outcomes expected from a particular presidential nomination. The individual identity of any given vice president will still be a product of the associated presidential nomination, but sometimes it will reflect the needs of a successful presidential aspirant to complete a nominating majority or pursue a general election. Thereafter, a vice president can acquire, not usually but more than occasionally, an autonomous role in the fallout from nominating politics, and in the policy-making of the elected administration. Fourteen vice presidents have actually gone on to become President of the United States. (Table 2.8) Five did so by standing subsequently for election in their own right: John Adams in 1796, Thomas Jefferson in 1800, Martin Van Buren in 1836, Richard Nixon in 1968, and George H.W. Bush in 1988. So this list confirms that the vice presidency has not been particularly kind to the direct electoral ambitions of vice presidents: two presidencies after 1836 is almost a form of failure, not success. Yet the far more common pattern for vice presidential promotion was not direct election. Rather, the other nine acceded to the office during the term of the president with whom they had first been elected. For the record, five of them ascended to the office because the head of their ticket died of natural causes: John Tyler for William Henry Harrison in 1841, Millard Fillmore for Zachary Taylor in 1850, Theodore Roosevelt for William McKinley in 1901, Calvin Coolidge for Warren Harding in 1923, and Harry Truman for Franklin Roosevelt in 1945. Three others were elevated because the head of their ticket was assassinated: Andrew Johnson for Abraham Lincoln in 1865, Chester A. Arthur for James A. Garfield in 1881, and Lyndon B. Johnson for John F. Kennedy in 1963. Lastly, one acceded to the office after a presidential resignation: Gerald Ford for Richard Nixon in 1974.44

70

B. E. SHAFER AND E. M. SAWYER

Table 2.8 Fortunes of the vice presidents who became president

1796 John Adams

1800 Thomas Jefferson

1836 Martin Van Buren

1844 John Tyler

1848 Millard Fillmore

1864 Andrew Johnson

1880 Chester A. Arthur

1900 Theodore Roosevelt

1920 Calvin Coolidge

Elected in his own right; Re-nominated but not re-elected Elected in his own right; Re-nominated & re-elected Elected in his own right; Re-Nominated but not re-elected Ran again in 1848 as Free Soil candidate Succeeded W.H. Harrison upon his death; hence 1st to be elevated without election in his own right Not re-nominated by Whigs or Dems Supported Dem nominee J.K. Polk Succeeded Zachary Taylor upon his death; Not re-nominated Ran again in 1856 as Know-Nothing candidate Succeeded Abraham Lincoln upon his death; Not nominated by Democrats in 1868 Succeeded James A. Garfield upon his assassination; Not re-nominated in 1884 Succeeded William McKinley upon his death; Re-nominated & re-elected in his own right Did not seek second full term Ran again in 1912 as Progressive candidate Succeeded W.G. Harding upon his death Re-nominated and re-elected in his own right;

(continued)

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

Table 2.8 (continued) 1944 Harry Truman

1960 Lyndon Johnson

1968 Richard M. Nixon

1974 Gerald Ford

1988 George H.W. Bush

71

Did not seek second full term Succeeded F.D. Roosevelt upon his death Re-nominated & re-elected in his own right; Did not seek second full term Succeeded J.F. Kennedy upon his assassination; Re-nominated & re-elected in his own right; Did not seek second full term Elected in his own right in 1968, after being defeated for election in 1960; Re-elected in 1972; Resigned in 1974 Never elected in his own right; 1st to be elevated without election as VP Re-nominated but not re-elected Elected in his own right; Re-nominated but not re-elected

So questions about prior occupations and previous careers for these nine presidents were not about institutional arrangements, party structures, or electoral balance, at least not directly. Instead, the relevant analytic questions are two. Did the vice presidency create presidents who could not otherwise have been nominated in their own right? And did their elevation to the presidency change the policy character of administrations which they had not originally been elected to lead? Both of these questions do have an insistent counterfactual element to them. If they had not been elevated, how can we know what might have ensued for them? And if their president had not disappeared, what would administration policy have looked like? But in truth, the electoral fortunes of

72

B. E. SHAFER AND E. M. SAWYER

these successor (vice) presidents and the policy conflicts of the administrations which they inherited make it much simpler in practice than it is in theory to specify some answers. To begin with, it is relatively easy to say of the first four on this list— John Tyler, Millard Fillmore, Andrew Johnson, and Chester Arthur—that they would not have been president save for their vice presidencies. In that sense, this office was indeed different from all others. Their only route to a presidency was by way of a vice presidency. Moreover, two of them, Tyler and Johnson, were always in active tension with the policy programs pursued by their presidents. Though this is less impressive than it might seem abstractly, since both had been selected to be somewhat divergent, serving as members of a unity ticket intended to draw disenchanted Democrats to Harrison the Whig and to Lincoln the Republican, respectively, a paper unity that would be torn up in both cases when they assumed the higher office. In any case, all four were refused renomination when their formal (and accidental) terms came to an end. Fillmore in 1852 for the Whigs and Arthur in 1884 for the Republicans did seek the endorsement of parties that were at least their own, but they were denied. Even as incumbent presidents, they could not be nominated in their own right. Tyler in 1844 for the Whigs and Johnson in 1868 for the Republicans could muster no support at their next party conventions, though as minority members of previous unity tickets this was less surprising. Yet both had continued to champion policy positions associated more with their prior parties, to the point where both ultimately aspired to be nominated for president by the opposition party when their formal term of office ended. In that regard, both failed. Four other vice presidents who were elevated to the presidency upon the death of their original president are less simply dismissed. Theodore Roosevelt in 1904, Calvin Coolidge in 1924, Harry Truman in 1948, and Lyndon Johnson in 1960 were all renominated and reelected. In that limited sense, they were by definition electable in their own rights. Moreover, all but Roosevelt clearly pursued the policies characteristic of the administration in which they were originally slated as vice presidents, and even Roosevelt, spokesman for the progressive wing of the Republican Party as he clearly was, became much more obstreperously progressive and factionally distinctive after his term in office had ended.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

73

Nevertheless, most analysts would rate Truman and Johnson clearly, and Roosevelt probably, as unlikely to have achieved a presidential nomination absent the vice presidency. Roosevelt, after all, had not been McKinley’s choice for vice president in his first term, and when that choice (Garret Hobart of New Jersey) died, McKinley declined to indicate a preference. Truman never considered seeking the presidency in his own right, though his replacement of Henry Wallace by Franklin Roosevelt in 1944 would have major consequences for pursuit of the Cold War. Lastly, Johnson, having failed decisively in 1960 despite a long-running interest in the presidency, was rescued only when John Kennedy chose him as vice president for geographic and ideological balance. So on this entire list, only Coolidge should be considered likely to have been nominated in his own right, quite apart from his vice presidency. He was a popular figure in the national Republican Party, where his handling of the police strike in Boston had given him a national profile. Moreover, his election was clearly an extension of national party policy, just as Truman consciously meant to extend the program of Franklin Roosevelt and Johnson the program of John Kennedy. As above, the “program” of Theodore Roosevelt was more accurately described as a factional rather than a dissident view until after his term of office ended. Last comes the one vice president, Gerald Ford, who had never been elected to the office, who was seriously challenged for renomination after his appointment, and who survived that challenged only to lose the election. Like Truman, he almost surely would not have sought the presidency on his own, despite being the (minority) leader of his party in the House. Though like Coolidge—and very unlike John Tyler or Andrew Johnson—he too represented a direct extension of the orthodox program of his national party. So his vice presidency and then presidency were surely not contributions to something other than an ongoing politics.45 Overall, that is a picture which allows vice presidents to be individually idiosyncratic but not otherwise to alter the occupational backgrounds and career trajectories that shape the fields for presidential nominations. But what of the presidency itself? Which is to say: has winning and holding the office been at least a further influence on renomination and reelection? For a total of thirty-one contenders, the office itself has, after all, been the most immediate predecessor to another run at nomination and election.46 So there is a recurrent justification for asking whether a first term as president begets a second.

74

B. E. SHAFER AND E. M. SAWYER

Being president is in no reasonable sense a “career route” to being president. One cannot stand for reelection without having previously held the office, though sitting presidents who cannot be dismissed as mere ticket-balancers have failed even at renomination. Martin Van Buren sought and failed at re-nomination in 1844 after having failed at reelection in 1840,47 though Grover Cleveland succeeded at the same maneuver: nominated and elected in 1884; renominated but defeated in 1888; renominated and (re)elected in 1892. At the same time, being president does reliably raise the question in the minds of all the other major players of whether a sitting president will seek a further term and what his prospects are. So a search for the role of prior occupations and political careers should probably close with this question for us too. Even here, simple answers require immediate qualification (Table 2.9). At the most rudimentary level, among presidents who were elected in their own right and sought a second term, sixteen were successful and ten were not.48 Moreover, both cohorts are scattered across time, so that this relationship is not just a picture of a particular era with a specific model of party structure, nor of one particular stretch of party hegemony. And this relationship does not appear to be a product of one or another matrix of institutions for selecting delegates and nominating presidents, that is, of the pre-convention, pure convention, mixed, or plebiscitary framework. Though this is a case where the current moment tantalizes, by suggesting a change. Or at least, the successful reelection of the three most recent presidents who stood for a second term, coupled with the fact that this has simultaneously alternated partisan control of the office— Democrat Bill Clinton for two terms, Republican George W. Bush for two terms, and Democrat Barack Obama for two terms, followed as this is written by the reelection effort of Republican Donald Trump—might just suggest that a hold on the presidency itself is gaining value in the modern world. On the other hand, that hold was not sufficient to save Democrat Jimmy Carter or Republican George H.W. Bush during even the modern plebiscitary era, so the jury must remain out.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

Table 2.9 Fortunes of sitting presidents

Elected to Two Terms George Washington Thomas Jefferson James Madison James Monroe Andrew Jackson Abraham Lincoln Ulysses Grant Grover Clevelanda William McKinley Franklin Rooseveltb Dwight Eisenhower Richard Nixon Ronald Reagan Bill Clinton George W. Bush Barack Obama Reelected After Acceding to the Presidency Theodore Roosevelt Calvin Coolidge Harry Truman Lyndon Johnson Chose Single Term James Polk James Buchanan Rutherford Hayes

75

Renominated but Lost Reelection

John Adams John Q. Adams Martin Van Buren Grover Clevelanda Benjamin Harrison William H. Taft Herbert Hoover Gerald Ford Jimmy Carter George H.W. Bush Desired a Second Term But Not Renominated John Tyler Millard Fillmore Andrew Johnson Chester Arthur Died in Office W.H. Harrison Zachary Taylor Abraham Lincoln James Garfield William McKinley Warren Harding Franklin Rooseveltb John Kennedy

a Only President to be elected to two nonconsecutive terms b Only President to be elected to more than two terms

Occupations, Careers, and Presidencies There is a distinctive patterning across time to the seedbeds for presidential nominations. Or, said the other way around, different occupational choices and different career pathways provided optimal springboards to a presidency at different points in time. Some of this comes close to

76

B. E. SHAFER AND E. M. SAWYER

being definitional: public offices betoken a career in politics, and individuals with a career in politics are more likely to run for—and win—its highest office, especially as compared to individuals with an absence of prior connection to politics generally. Though there have been exceptions even at this quasi-definitional level, as with the unsuccessful John C. Fremont for the Republicans in 1856, the unsuccessful Horatio Seymour for the Democrats in 1868, and the unsuccessful Wendell Willkie for the Republicans in 1940, not to mention the lone winner without a public office or an explicitly political career, namely Donald Trump in 2016. Among the multitude of others, there was a further grand patterning that could be specified far more precisely: executive office to congressional office to gubernatorial office to senatorial office. On its own terms, this patterning was an important next move down the causal funnel. In effect, it focused an enduring bandwagon dynamic on a much-reduced body of presidential aspirants. Yet what it did was to focus that dynamic: these occupations by themselves were not driving anything. Rather, their patterning was a goad to search for the structural aspects of nominating politics that favored federal executives in one era, governors in another era, and senators in another. Just as the resulting pattern was an inherent reminder that the occupations or careers that were dominant at one or another point in time could hardly produce the changing succession of their own dominance. Rather, it was the presence of critical resources attached to the offices in favor during a specific era that were the ones that had become disproportionately more useful, causing an increasing number of individuals with realistic presidential aspirations to come from, say, a governorship rather than the House. A taxonomy of prior political offices must thus be interpreted not principally as a measure of what strategic candidates were thinking at the time, so much as a reflection, even a key indicator, of the political character and requirements of an era. In that regard, three elements of this incentive structure stood out with regard to presidential nominations. The institutional arrangements for delegate selection were a major recurrent element of this structure, and these mechanics have lurked close to the surface in the preceding analysis. But the organizational nature of the political parties—where they got their own internal resources and how they governed themselves—regularly mattered as well. As did the electoral balance between these parties, since the likelihood of ending up on a

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

77

winning ticket, a losing ticket, or in a close race affected the strategic calculations of most of the main players. Seen the other way around, occupations and careers, nominations and elections, and their collective patterning are evidence toward the larger structure of politics in their eras: • Within this overall framework, the pre-convention period had relied on federal executives and military generals for its presidential nominees. Proximity to the creation of the constitution itself, a partisan politics of elite negotiations in the making of public policy, and the early collapse of an incipient competitive party system all contributed to these occupational priorities. • The antebellum start to the pure convention system added members of Congress, especially of the House of Representatives, and kept the military generals, but exiled the federal executives. This rise of presidential nominees with their roots in Congress did attest implicitly to the rising consequence of government itself and, of course, to the centrality of the new Congress within it. Just as they inaugurated the rise of political careers based in party politics. • The postbellum part of the pure convention system then injected governors into this picture. This rise of governorships was most directly a reflection of the appearance of machine type parties built around jobs, contracts, and favors, and based profoundly in the states. Governors were often their central figures, though in a period of partisan imbalance, it was the Republicans who hewed most closely to the dominant seedbeds for presidential nomination, the Democrats who spread the net more broadly. • Governorships continued to grow as a seedbed for presidential nominations during the mixed system of delegate selection, where they were increasingly joined by members of the Senate, though the House more or less completely disappeared. In a period of reform struggles and social change, both ascendant offices now had the advantage of demonstrating proven electoral prowess, while providing a podium for public statements that could be covered (and magnified) by a rising media of information. • Finally, the modern world, that is, the plebiscitary era, intensified and cemented the dominance of both governors and senators. This reflected the power of a system of sequential primary elections that provided a stream of ever-changing evidence of candidate prospects

78

B. E. SHAFER AND E. M. SAWYER

for all the major players to consider. Parties driven by issue activists were especially responsive to this dynamic. On the other hand, a closer balance between the two parties meant that they no longer differed noticeably in their occupational and career seedbeds. Along the way, the roster of successful occupations and careers began to put names and faces on the faceless bandwagon that ran across nominating politics in such a constant pattern from the 1840s to the current moment. If the name and the face were yours, this might be all you cared about. But for the analyst, the process of channeling by occupation and career begins to explain how a bandwagon dynamic running across most of American history could be regularly infused not just by personalities and strategies but also by the constituencies and issues that characterized American politics at any given point in time. So that in the end, the changing seedbeds for presidential nominations tell two great stories, and not just one. The first and most immediate involves generating the fields of contenders for major subsequent aspects of nominating politics: factional struggles within each of the parties, formal rules to channel those struggles, and strategies by both presidential aspirants and state parties either to evade or to capitalize on all of that. The second and more fundamental story then turns back to the bandwagon dynamic that runs through them, selecting a winner. In statistical terms, the occupational and career seedbeds for nominating contests are hugely important: they shape the ability of presidential aspirants to build an extended career, while they make the “wrong” occupations nearly irrelevant at various—albeit changing—points in time. Yet after all of that, these occupational choices and career trajectories have not altered the underlying bandwagon dynamic that still runs through them, at least not since the 1830s. Instead, they push the funnel of causality on to the next level, involving the factional struggles that occur within each party. These are the struggles that most powerfully and directly select a specific individual presidential nominee from within these seedbeds and their fields.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

79

Notes 1. The outstanding journalistic student of occupation backgrounds and career trajectories in modern American politics is probably Theodore H. White. The template for an ongoing series of specific observations was The Making of the President 1960 (New York: Atheneum, 1961), but White himself did attempt one career-long aggregation of the same phenomena in America in Search of Itself: The Making of the President, 1956–1980 (New York: Harper & Row, 1982). 2. Many related aspects of these questions are helpfully gathered in John H. Aldrich, “Who Runs for President and Why”, Chapter 2 in Aldrich, Before the Convention: Strategies and Choices in Presidential Nominating Campaigns (Chicago: University of Chicago Press, 1980). A more narrowly focused but much longer reach is Robert L. Peabody, Norman J. Ornstein, and David W. Rohde, “The United States Senate as a Presidential Incubator: Many Are Called but Few Are Chosen”, Political Science Quarterly 91(1976), 237–258. 3. This was once taken as an evident route into a much more general approach to American politics and not just American parties, as with William Nesbit Chambers and Walter Dean Burnham, eds., The American Party Systems: Stages of Political Development (Oxford: Oxford University Press, 1967). 4. For the shifting and diverse means by which these individuals were nominated, Richard P. McCormick, The Presidential Game: The Origins of American Presidential Politics (New York: Oxford University Press, 1982), and M.J. Heale, The Making of American Politics, 1750–1850 (London: Longman, 1977). For their fraught possibilities, James Roger Sharp, American Politics in the Early Republic: The New Nation in Crisis (New Haven: Yale University Press, 1993). 5. For the politics around them, Ronald P. Formisano, “Federalists and Republicans: Parties, Yes—System, No”, in Paul Kleppner et al., The Evolution of American Electoral Systems (Westport: Greenwood Press, 1989; John C. Miller, The Federalist Era, 1789–1801 (New York: Harper & Row, 1960); Lance Banning, The Jeffersonian Persuasion: Evolution of a Party Ideology (Ithaca: Cornell University Press, 1978); David Hackett Fisher, The Revolution in American Conservatism: The Federalist Party in the Era of Jeffersonian Democracy (New York: Harper & Row, 1965). 6. Paul Goodman, “The First American Party System”, Chapter 3 in Chambers and Burnham, eds., Political Parties in a New Nation, and Ronald P. Formisano, “Deferential Participant Politics: The Early Republic’s Political Culture, 1789–1840”, American Political Science Review 68(1974), 473–487.

80

B. E. SHAFER AND E. M. SAWYER

7. For the emergence of national party conventions as the definitive institution for presidential nominations, see James S. Chase, Emergence of the Presidential Nominating Conventions, 1789–1832 (Urbana: University of Illinois Press, 1973). For politics across the entire pure convention period, see Richard L. McCormick, The Party Period and Public Policy: American Politics from the Age of Jackson to the Progressive Era (New York: Oxford University Press, 1986), and Joel H. Silbey, The American Political Nation, 1838–1893 (New York: Oxford University Press, 1985). 8. For these critical elections and the critical break, see Donald Ratcliffe, The One-Party Presidential Contest: Adams, Jackson, and 1824’s Five-Horse Race (Lawrence: University Press of Kansas, 2015), and Lynn Hudson Parsons, The Birth of Modern Politics: Andrew Jackson, John Quincy Adams, and the Election of 1828 (Lawrence: University Press of Kansas, 2009). 9. For politics in the antebellum years as a whole, see Daniel Walker Howe, What Hath God Wrought? The Transformation of America, 1815–1848 (New York: Oxford University Press, 2009); Sean Wilentz, The Rise of American Democracy: Jefferson to Lincoln (New York: W. W. Norton, 2005); Joel H. Silbey, The Partisan Imperative: The Dynamics of American Politics Before the Civil War (New York: Oxford University Press, 1985). 10. Richard P. McCormick, The Second American Party System: Party Formation in the Jacksonian Era (Chapel Hill: University of North Carolina Press, 1966); Robert V. Remini, The Era of Good Feelings and the Age of Jackson (Arlington Heights: AHM Publishing, 1979); Michael F. Holt, Political Parties and American Political Development (Baton Rouge: Louisiana State University Press, 1992). 11. John Niven, Martin Van Buren: The Romantic Age of American Politics (New York: Oxford University Press, 1983); Donald B. Cole, Martin Van Buren and the American Political System (Princeton: Princeton University Press, 1984); Robert V. Remini, Marin Van Buren and the Making of the Democratic Party (New York: Columbia University Press, 1951). 12. Daniel Walker How, The Political Culture of the American Whigs (Chicago: University of Chicago Press, 1979); Michael Holt, The Rise and Fall of the American Whig Party: Jacksonian Politics and the Onset of Civil War (New York: Oxford University Press, 1999); George R. Poage, Henry Clay and the Whig Party (Chapel Hill: University of North Carolina Press, 1936). 13. Unlike Zachary Taylor or Winfield Scott, Harrison did have an extended and more orthodox political career.

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

81

14. Scott had already been a serious favorite of some party leaders in the contest for a Whig nomination in 1840 against Harrison and again in 1848 against Taylor.sas. 15. On the politics of that period, Peter H. Argersinger, “The Transformation of American Politics: Political Institutions and Public Policy, 1865–1910”, and Richard Jensen, “Democracy, Republicanism, and Efficiency: The Values of American Politics, 1885–1930”, Chapters 5 and 6 in Byron E. Shafer and Anthony J. Badger, eds., Contesting Democracy: Substance and Structure in American Political History, 1775–2000 (Lawrence: University Press of Kansas, 2001). More generally, Robert W. Cherny, American Politics in the Gilded Age, 1868–1900 (Arlington Heights: Harlan Davidson, 1997). 16. The two great—and opposing views of this development in its time were James Bryce, The American Commonwealth, 2 vols. (London: Macmillan and Company, 1888), and Henry Jones Ford, The Rise and Growth of American Politics (New York: The Macmillan Company, 1898). On the resources that were at issue, see Mark W. Summers, The Era of Good Stealings (New York: Oxford University Press, 1993). 17. Leonard D. White, The Republican Era (New York: Macmillan, 1958); Robert Marcus, Grand Old Party: Political Structure in the Gilded Age, 1880–1896 (New York: Oxford University Press, 1971); A. James Reichley, “The Republican Era”, Chapters 6–11, in Reichley, The Life of the Parties: A History of American Political Parties (New York: The Free Press, 1992). 18. On the problem generally, Alan Ware, The Democratic Party Moves North, 1877–1962 (Cambridge: University of Cambridge Press, 2006). On the only successful solution(s) of the period, see Richard E. Welch, Jr., The Presidencies of Grover Cleveland (Lawrence: University Press of Kansas, 1988), but also Charles W. Calhoun, Minority Victory: Gilded Age Politics and the Front Porch Campaign of 1888 (Lawrence: University Press of Kansas, 2008). 19. Michael Kazin, A Godly Hero: The Life of William Jennings Bryan (New York: Knopf, 2006); on populism as a movement, Peter H. Argersinger, Populism and Politics: William Alfred Peffer and the People’s Party (Lexington: University Press of Kentucky, 1982); and more generally, Robert C. McMath, Jr., American Populism: A Social History, 1877–1898 (New York: Hill & Wang, 1992). 20. Which is not to say that there were not other developments (and conflicts) in the American politics of the time, as with Gretchen Ritter, Goldbugs and Greenbacks: The Antimonopoly Tradition and the Politics of Finance in America, 1865–1896 (New York, 1997), and Elisabeth Clemens, The People’s Lobby: Organizational Innovation and the Rise of Interest Group Politics in the United States (Chicago: University of Chicago Press, 1997).

82

B. E. SHAFER AND E. M. SAWYER

21. For the larger context of politics during the mixed system of presidential nominations, Otis L. Graham, Jr., The Great Campaigns: Reform and War in America, 1900–1928 (Englewood Cliffs: Prentice-Hall, 1971); Michael McGerr, A Fierce Discontent: The Rise and Fall of the Progressive Movement in America, 1870–1920 (New York: Oxford University Press, 2005); Ellis W. Hawley, The Great War and the Search for a Modern Order (Prospect Heights: Waveland Press, 1997). 22. For the Progressives as a political movement, Arthur S. Link and Richard L. McCormick, The Progressives (Arlington Heights: Harlan Davidson, 1983); for their reform program, especially with regard to presidential selection, James W. Ceaser, “Woodrow Wilson and the Origin of the Modern View of Presidential Selection”, Chapter 4 in Ceaser, Presidential Selection: Theory and Practice (Princeton: Princeton University Press, 1979). 23. An early look at the operation of this new institution is Louise Overacker, The Presidential Primary (New York: Macmillan, 1926). 24. Byron E. Shafer and Regina L. Wagner, The Long War Over Party Structure: Democratic Representation and Policy Responsiveness in American Politics (Cambridge: Cambridge University Press, 2019). 25. These and many other results relevant to this analysis can be easily followed in Chapter 10, “Presidential Primaries”, Congressional Quarterly’s Guide to U.S. Elections, 2 Vols., 7th ed. (Washington, DC: CQ Press, 2016). 26. Geoffrey Cowan, Let the People Rule: Theodore Roosevelt and the Birth of the Presidential Primary (New York: W. W. Norton, 2016); also Lewis L. Gould, “Roosevelt Versus Taft in 1912”, Chapter 3 in Gould, Four Hats in the Ring. 27. Thedore H. White, The Making of the President 1968 (New York: Atheneum, 1969), and Lewis Chester, Godfrey Hodgson, and Bruce Page, An American Melodrama: The Presidential Campaign of 1968 (London: Andrew Deutsch, 1969). 28. An early—and slightly premature—look at this particular institutional reform is George H. Haynes, The Election of Senators (New York: Henry Holt, 1906). Its successful adoption as Article XVII of the US Constitution was confirmed by Secretary of State William Jennings Bryan on May 31, 1913, making it nearly coterminous with the arrival of the presidential primary and the mixed system of presidential nominations, as part of the same thrust toward direct participation. 29. On radio, its opportunities and demands, Douglas B. Craig, Fireside Politics: Radio and Political Culture in the United States, 1920–1940 (Baltimore: Johns Hopkins University Press, 2000); on television, Austin Ranney, Channels of Power: The Impact of Television on American Politics (New York: Basic Books, 1983).

2

DEEP CHANNELS TO A PRESIDENCY: OCCUPATIONS AND CAREERS

83

30. Two interpretations of this great and widely recognized change are Robert S. Erikson and Kent L. Tedin, “The 1928–1936 Partisan Realignment: The Case for the Conversion Hypothesis”, American Political Science Review 75(1981), 951–962, and Kristi L. Andersen, The Creation of a Democratic Majority, 1928–1936 (Chicago: University of Chicago Press, 1979). An effort to locate the precise change point is Helmut Norpoth, Andrew H. Sidman, and Clara H. Suong, “Polls and Elections: The New Deal Realignment in Real Time”, Presidential Studies Quarterly 43(2013), 146–166. 31. For that story, Byron E. Shafer, Quiet Revolution: The Struggle for the Democratic Party and the Shaping of Post-Reform Politics (New York: Russell Sage Foundation, 1983). 32. See especially Table 3.1 The Spread of Participatory Politics: A Changing Institutional Matrix for Delegate Selection, Byron E. Shafer, The American Political Pattern: Stability and Change, 1932–2016 (Lawrence: University Press of Kansas, 2016), 86. 33. Peter B. Clark and James Q. Wilson, “Incentive Systems: A Theory of Organizations”, Administrative Science Quarterly 6(1961), 129–166, elaborated and focused on parties in “Political Parties”, Chapter 6 in Wilson, Political Organizations (New York: Basic Books, 1973). 34. For the early opposite positions, see James Bryce, The American Commonwealth, 2 vols. (London: Macmillan & Co., 1888) versus Henry Jones Ford, The Rise and Growth of American Politics: A Sketch of Constitutional Development (New York: Macmillan, 1898). 35. David R. Mayhew, Placing Parties in American Politics: Organization, Electoral Settings, and Governmental Activity in the Twentieth Century (Princeton: Princeton University Press, 1986), and Alan Ware, The Breakdown of Democratic Party Organization, 1940–1980 (New York: Cambridge University Press, 1985). 36. For politics in the transition, Gareth Davies, From Opportunity to Entitlement: The Transformation and Decline of Great Society Liberalism (Lawrence: University Press of Kansa, 19960), and Iwan W. Morgan, Beyond the Liberal Consensus: A Political History of the United States Since 1965 (New York: St. Martin’s, 1994). For the larger context afterward, James T. Patterson, Restless Giant: The United States from Watergate to Bush v Gore (New York: Oxford University Press, 2005). 37. Note that this also moves Hillary Clinton in 2016 from “executive” to “senator,” because being Senator from New York immediately preceded her being Secretary of State in the Obama Administration.

84

B. E. SHAFER AND E. M. SAWYER

38. For a more extended attack on the question of why the nomination disappeared from the convention without the stimulus of formal change, see Byron E. Shafer, “The Nomination and the Convention: An Informal Disappearance”, Chapter 1 in Shafer Bifurcated Politics: Evolution and Reform in the National Party Convention (Cambridge, MA: Harvard University Press, 1988), gathered most concisely at Figure 1.1, 21. 39. A prescient notice of the development was William Carleton, “The Revolution in the Presidential Nominating Contest”, Political Science Quarterly 72(1957), 224–240. 40. Addressed in the concluding chapter. 41. For 1952, Paul T, David, Malcolm Moos, and Ralph M. Goldman, “The National Story”, Vol. 1 in David, Moos, and Goldman, Presidential Politics in 1952 (Baltimore: Johns Hopkins Press, 1954). For 1956, Charles A.H. Thomson and Frances M. Shattuck, The 1956 Presidential Campaign (Washington, DC: Brookings Institution, 1960). 42. Theodore H. White, The Making of the President 1960 (New York: Atheneum, 1961), and W.J. Rorabaugh, The Real Making of the President: Kennedy, Nixon, and the 1960 Election (Lawrence: University Press of Kansas, 2009). 43. Theodore H. White, The Making of the President 1968 (New York: Atheneum, 1965), and Robert Alan Goldberg, Barry Goldwater (New Haven: Yale University Press, 1997). 44. A quick tour of each individual story is Philip Abbott, Accidental Presidents: Death, Assassination, and Democratic Succession (New York: Palgrave Macmillan, 2008). 45. The extreme diversity present in the situations of these nine “accidental presidents” is a major reason that institutional analysts have had great difficulty in theorizing an appropriate replacement for a president during his term of office. One such exercise, explicitly acknowledging those difficulties, is Allan P. Sindler, Unchosen Presidents: The Vice President and Other Frustrations of Presidential Succession (Berkeley: University of California Press, 1976). 46. For a summery of modern efforts, see Kathryn Dunn Tenpas, “How Incumbent Presidents Run for Re-election”, Chapter 4 in William G. Mayer, ed., The Making of the Presidential Candidates 2004 (Lanham: Rowman & Littlefield, 2003). 47. The indomitable Van Buren would reappear one more time, as the presidential candidate of the Free Soil Party in 1848. 48. The tally puts Grover Cleveland in both categories, unsuccessful in 1888 but successful in 1892.

CHAPTER 3

Factions, Constituencies, and Candidates: The Democrats

Press coverage of aspirants for a presidential nomination has long featured their personal backgrounds and campaign characteristics, actual and asserted. Part of this focus is a response to the expected curiosity of the political audience; part is an effort to find prospective indicators of nominating fortunes. In the modern world, given expanded technical possibilities, this press attention is often idiosyncratic, voluminous, picayune, and compulsive. Yet the temptation to reason backward in this sense has always been present, that is, the temptation to treat personal characteristics and individual campaigns as explanations for nominations and then elections. The goal of this analysis is different: to isolate ongoing influences that recurrently shape the final choice within a field of contenders. Isolated successfully, these influences become the recurrent factors that actually select a specific nominee from within those fields. In the process, again if successful, these influences relegate candidate characteristics and individual strategies to the far end of the funnel of causality, the end closest to an ultimate nominating decision but also the one that explains the smallest amount about the ultimate outcome.1 Chapters 3 and 4 will suggest that party factions, that is, coordinated subgroups within the two parties that have both common interests and a shared sense of those interests, are the decisive factors here.2 From one side, it is the struggles among these regularized factions that produce a nominee. From the © The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_3

85

86

B. E. SHAFER AND E. M. SAWYER

other side, nominees are best understood, not through their personalities, strategies, or programs, but as individual incarnations of the factional structure at a given point in time. In that way, a factional structure becomes the essential link between the occupational and career seedbeds that produce a field of aspirants for a presidential nomination and the two specific individuals who end up as nominees. Such a focus fits simultaneously and comfortably with an effort to move to the next level in the causal funnel: a bandwagon dynamic runs through changing seedbeds that generate fields of presidential contenders within which it is the struggles of the factions that actually make the crucial nominating choice. Accordingly, this chapter will focus on the main factional struggles shaping presidential nominations within the Democratic Party, that is, the established influences that made Democratic nominations roughly interchangeable over multiple contests but that always broke up this continuity when the factions shifted. In the process, the chapter becomes a kind of template for unpacking the structural influences on presidential nominations that have characterized American politics across time. For an analysis concerned with fundamentals versus ephemerals, there is a certain frustration in this. Presenting the evidence for the power of these changing factions requires a chronicle of contests, one sufficiently detailed to make an ongoing factional structure stand out. On the other hand, the analysis aspires to place as much of this historical detail as possible in a larger framework, where the nature of the factions plays the crucial deciding role. So the bulk of the chapter must set out the evidence delineating a universe of party factions and factional politics across time. But the chapter must close by reassembling this chronicle to highlight its more general contributions.

A Democratic Party Founding the Modern Party The replacement of the Democratic-Republican Party after 1824, that loose elite coalition which had snuffed out the Federalists nationally while consolidating the presidencies that became known as “the Virginia Dynasty,” was in effect the founding of the modern Democratic Party. The associated break between old and new was so profound—a difference in geography, policy, culture, and, last but not least, political organization—that the five elections from 1824 through 1840 were to be

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

87

as narrowly circumscribed in the specific identities of the Democratic nominee as any five-election period in American history, rivaled only by the five contests from 1932 through 1948.3 For five elections, an original Jacksonian coalition, uniting the south and the west against the northeast, and thus proponents of an expansionist agricultural economy against a commercial market economy, was sufficient to tell the story of internal Democratic factions and their presidential nominees. The opening presidential contest in this five-election sequence was testament to the problems inherent in what had become a one-party system, with five major contenders inside the Democratic-Republican Party, four of whom were stereotypical products of the old regime: John Quincy Adams of Massachusetts, Secretary of State; John C. Calhoun of South Carolina, Secretary of War; Henry Clay of Kentucky, Speaker of the House; and William J. Crawford of Georgia, Secretary of the Treasury. The wildcard was General Andrew Jackson of Tennessee, western not eastern, self-made not educated, and a popular hero rather than a government official.4 There being no agreed process for making one of these men the official nominee of the Democratic-Republican Party, the choice was effectively left to the Electoral College. While the Jackson candidacy had begun as the most speculative of the five, the fact that he led in the popular vote but lost the election, thrown into the House of Representatives where the result was additionally disconnected from the voting tally, made him the figurehead for opponents of the winner, President John Q. Adams. The institutional effects of the effort that followed, to wrest away the presidency, were to be profound and lasting, though these owed as much to Senator Martin Van Buren of New York, their real organizing focus, as they did to Jackson himself.5 First, an incipient Jackson candidacy for 1828 became an important part of the effort to remove selection of the remaining Electors from the hands of state legislators while putting them formally on the ballot. Four more states did so in time for the 1828 election, including the vote-rich state of New York. Then, the Jackson campaign focused on generating organized committees around the country, on a scale unprecedented in its time. By the end of this five-election sequence, the Democratic Party would feature a systematic presence in much of the country, built not upon the socioeconomic positions of its notables but on the concrete rewards which the control of public office could directly convey.6

88

B. E. SHAFER AND E. M. SAWYER

In order to fuse the west and the south against the northeast, Jackson had accepted Calhoun as his vice presidential running mate in 1828. Yet in running again four years later, he would turn to a national party convention as the means for asserting his renomination, and it would officially endorse Martin Van Buren as his potential vice president. In winning the presidency in 1828, Jackson had already in effect created the central party program as well, combining his general populist persona with an endorsement of governmental support for internal improvements and an aggressive hostility to the Bank of the United States, the latter a symbol of catering to a northeastern commercial elite. The public appeal of this total package was sufficient to allow what became the Jackson-Van Buren ticket to be reelected in 1832, and sufficient after 1832 to allow Jackson to transfer the coalition to Van Buren at the 1836 convention, where the latter was confirmed unanimously as Jackson’s successor. Van Buren was likewise to be unanimously renominated in 1840, though the controversial status of his ticket-balancing vice president, congressman Richard M. Johnson of Kentucky, did cause Van Buren to leave the vice presidential slot to the discretion of the state parties, most of whom ended up with Johnson anyway. A Rising Factional Puzzle The defeat of Martin Van Buren in 1840, in some sense the last definitional Jacksonian, was to usher in a very different period for the politics of presidential selection. The inaugural consensus on policy issues—localized improvements, decentralized banking, and low tariffs—was increasingly cross-cut rather than aligned by party factions, while the burgeoning divide over slavery would insistently color them all. The result was a fiveelection kaleidoscope for presidential nominations inside the Democratic Party. Two of these, James K. Polk in 1844 and Franklin Pierce in 1852, would be the first “dark horse” nominees in American politics, individuals who were not well known outside their home states and not serious candidates when nominating politics began. Two others, Lewis Cass in 1848 and James Buchanan in 1856, were straddlers, Northern Democrats with southern principles in the rhetoric of the time, who specialized in obscuring their positions on the slavery issue. And the last of the five, Stephen Douglas in 1860, would become a nominee-by-subtraction, as major pieces of the old Democratic coalition splintered into separate campaigns.

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

89

The first contest after the defeat of Van Buren in the election of 1840 was the perfect introduction to this new era of fragmented and cross-cutting factions and constituencies. Desirous of regaining the White House, Van Buren took the lead on the first ballot of the 1844 convention with an actual majority. Yet he had lost a prior procedural battle to repeal the two-thirds rule, so Van Buren became the first Democrat to reach a majority but fail to reach two-thirds. His main opponent was Lewis Cass of Michigan, Secretary of War under Jackson and a classic straddler,7 who acquired a plurality but not a majority on the fifth ballot. After the 8th ballot, Tennessee introduced its “favorite son,” former governor James K. Polk, a state-based figure previously perceived only as a potential ticket-balancer for either Van Buren or Cass. Democratic delegates who had cast eight ballots without much change in the balance between Van Buren and Cass, with no hope of a twothirds majority for either, then turned Polk into the nominee, taking him from 15% of the vote on the 8th ballot to 89% on the 9th, before the convention made him unanimous. A promise of only one term as President and a focus on national expansion8 helped to keep the slavery issue at bay, though new territories would quickly come to exacerbate the slavery issue. But in the short run, Polk was to be true to his one-term promise, and Cass returned to secure the nomination in 1848. Beginning with a plurality on the first ballot, Cass achieved a majority on the second and was nominated on the fourth, though the convention chair did have to rule that a two-thirds majority of those present and voting (and not of the full convention) was sufficient for nomination. A second dark horse nomination was to follow hard upon the defeat of Cass in the general election. Like Van Buren, Cass too wished for a second shot at the presidency and was the leading candidate on the first ballot in 1848, followed by James Buchanan of Pennsylvania, previously Minister to the United Kingdom, and favorite sons William Marcy, Governor of New York, and Stephen Douglas, Senator from Illinois. After nineteen inconclusive ballots, Buchanan took the lead on the twentieth, Douglas took the lead on the thirtieth, and Marcy took the lead on the fortyfifth, but none approached a majority, much less two-thirds. Franklin Pierce of New Hampshire, a little-known former congressman and senator who was not central to any faction, was introduced by Virginia on the thirty-fifth ballot and sustained thereafter by New Hampshire and Maine. Having seen all the original aspirants achieve a plurality but little else, the convention moved more or less as a body to Pierce on the forty-ninth ballot.

90

B. E. SHAFER AND E. M. SAWYER

Unlike Polk, Pierce as President proved unable to duck the slavery issue, and his signing of the Kansas–Nebraska Act plus endorsement of the Fugitive Slave Act seriously eroded his northern support. Yet also unlike Polk, Pierce went back on his one-term promise, standing second to Buchanan on the opening ballot of the 1856 convention, with Stephen Douglas a distant third. As it became clear that Pierce would never regain a two-thirds majority, Douglas emerged as the alternative to Buchanan. After sixteen ballots, as it became clear that Douglas could not even displace Buchanan as plurality leader, the convention moved to the Pennsylvanian and he became the Democratic nominee (and then President) for 1856. The convention finished by adding congressman John C. Breckinridge of Kentucky for vice president as a ticket-balancing gift to the south. Buchanan’s Presidency was to be an unhappy one, and the Democratic National Convention of 1860 was perhaps the most unhappy in the long life of the party. Six southern delegations bolted the convention after a platform fight over slavery, and while the opening ballot showed Douglas in possession of a two-thirds majority of those delegates present and voting, the convention chair ruled this time that his majority had to be of total votes allocated. There was no hope of any such majority, and after three days of balloting and fifty-seven votes, the convention recessed. When it reassembled the following month, credentials challenges produced a new walkout, adding three western delegations to six southern counterparts and making the mathematics even more hopeless. In response, the convention merely moved by voice vote that Douglas, having achieved two-thirds of the vote cast, was officially the nominee. The aftermath left the Democratic Party splintered into its factional pieces. The northern faction belonged to Douglas. Some of the southern delegates who had bolted the original gathering reassembled in Richmond and nominated Vice President John C. Breckinridge of Kentucky as the southern Democratic candidate for president, making Breckinridge a symbol of the shift from a desperate attempt to hold the factions together to a situation where each fronted its own contender for the general election. The picture was completed by a congeries of dissidents— Whigs, Know-Nothings, and Democrats—who aspired to be the absent unifying force and nominated former Senator John Bell of Tennessee as the candidate of the Constitutional Union Party.

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

91

The Refounding A Truncated Party The next three nominations, while individually diverse, were united by the need to respond to the Democratic disaster of 1860.9 So in one sense, they were the ultimate extension of the problems bedeviling the Democratic coalition on the way into the Civil War. Yet the context for addressing these problems changed so greatly, courtesy of the war and its immediate aftermath, that efforts to address them, desperate as these were, bore little resemblance to initiatives characterizing the years leading up to the war. The challenge was now nothing less than the reconstruction—in essence, the refounding—of the party itself. Perhaps inevitably, the first response was a series of experiments, some creative, all quixotic, and none successful. For 1864, there was a failed military general; for 1868, there was a noncandidate who attempted to disavow his support as it emerged; and for 1872, there was a rubber-stamp endorsement of the candidate of a nascent third party. The first challenge for a party that lacked all eleven of the Confederate (and reflexively Democratic) states for 1864 was to find a candidate with any potential for mobilizing an electoral majority within the remainder. General George B. McClellan of New Jersey, former commander of the Union Army, received more than three-quarters of the vote on the first ballot as a candidate who could be dissident but not disloyal. The real hope of the convention, however, rested on a generalized war-weariness in the American public, and the scheduled gathering was actually postponed in hopes of improving that prospect. In the event, Rep. George Pendleton of Ohio, recognized as a leading ‘Copperhead’ from the peace faction which favored a quick negotiated settlement with the south, joined McClellan to form the Democratic ticket.10 All but three southern states had been restored to presidential voting by 1868, but the party now had to face off against the hero of the Union victory, General Ulysses S. Grant, as its Republican opponent. Pendleton returned to lead on the first ballot, ahead of sitting President Andrew Johnson, a Democrat from the unity ticket with Abraham Lincoln and respected more among Democrats than Republicans for his attitude toward Reconstruction. When Johnson’s support proved to be largely complimentary and Pendleton proved unable to advance, the contest became an extended duel between General Winfield Scott Hancock of Pennsylvania and Senator Thomas A. Hendricks of Indiana.

92

B. E. SHAFER AND E. M. SAWYER

On the twenty-second ballot, Ohio shifted its entire vote to Horatio Seymour, former Governor of New York and permanent chairman of the convention. Despite his disavowal of candidacy, his bandwagon rolled aggressively and by the end of the ballot, Seymour was the nominee. He too would prove unable to break Grant’s hold on the Presidency, improving on 1864 mainly by virtue of an increase in southern states that were now allowed to vote, plus the capture of his home state of New York. The much-improved popular vote did contain some promise of a workable Democratic strategy for the future, but before that, the party would follow one very different strategic possibility for its 1872 nomination. The new Liberal Republican Party, appearing first in Missouri as a fusion ticket between Democrats and reform Republicans—anticorruption, anti-carpetbag—managed to assemble a national ticket in 1872 led by Horace Greeley, editor of the New York Tribune. The Democratic national convention, in session for only six hours, simply accepted both the platform and the nominees of these Liberal Republicans, even though Greeley, despite a life-long interest in politics, was a frequent critic of the Democratic Party and possessed only a three-month career in elected office. A First Solution to the Factional Challenge The next five elections, from 1876 through 1892, would be different in kind. By 1876, the party faced a very different strategic problem, one that was simply described but operationally complex. It could now capture genuine majorities in Congress, suggesting the presence of a national majority, but it could not easily transfer support for this congressional majority to the presidential level. With three great factions—northeastern immigrants, western populists, and southern regionalists—the party was fully capable of capturing a congressional majority by aggregating its support in each factional homeland. Yet a presidential candidate who was attractive to one of the three factions was reliably unattractive to the other two. And of course, a presidential candidate who was not particularly attractive to any of the three was more or less a nonstarter. In 1876, the national party in effect blundered upon a solution that would work. In that year, Samuel J. Tilden, reform Governor of New York, had a solid (but not two-thirds) majority on the opening ballot. His main opponent was Thomas A. Hendricks, Governor of Indiana, plus a smattering of state and regional favorites. When Missouri switched from Hendricks along with a favorite son to Tilden within the first ballot,

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

93

the bandwagon kept rolling into the second and nominated Tilden. The convention quickly made Hendricks his vice president, and the party had a ticket. What it also had, by hindsight and somewhat by accident, was a winning strategy. The Tilden-Hendricks ticket was to win an undisputed majority of the popular vote. But the presence of conflicting electoral slates for Florida, Louisiana, and South Carolina would allow an ad hoc commission to award those states (and the presidency) to Rutherford B. Hayes, the Republican nominee.11 Yet if the Democratic ticket of 1876 was thereby denied, the internal Democratic solution would endure. In effect, a reform governor from New York,12 not tightly attached to any of the three great factions, could be acceptable to them all. Moreover, just to lock that solution in place, the two largest states that could actually be captured by either party, in a period where most states leaned heavily to one party or the other, were none other than New York and Indiana. After his (disputed) defeat, Tilden was unwilling to announce an intention for 1880, though Henry G. Payne, former Ohio Representative, did stand in as a stalking horse. Yet the old Pennsylvania favorite, General W.S. Hancock, led on the first ballot, followed closely by Thomas F. Bayard, Senator from Delaware, serving, as he did so many times, as a resting place for southern delegations but never a plausible nominee in his own right. Tilden formally announced his noncandidacy before the second ballot, but by then, Hancock’s total had doubled, and switches before the end of the ballot made him a nearly unanimous nominee with only Indiana dissenting, behind its favorite son, Thomas Hendricks. Despite an attempt to mollify Indiana by adding former representative William H. English (also of Indiana) to the ticket, Hancock’s subsequent electoral support proved largely limited to the southern and border states, and he went down to defeat. By 1884, however, the Democrats had another reform governor from New York, Grover Cleveland. He too led on the first ballot, followed once again by Thomas Bayard as a southern stalking horses and then by former senator Allen G. Thurman of Ohio as a favorite son. Pennsylvania switched from its favorite son, Speaker Thomas J. Randall, to Cleveland on the second ballot, swelling his total, and when North Carolina voted first for Bayard and then switched to Cleveland, the bandwagon brought him the nomination. On that same second ballot, Indiana switched from former senator Joseph E. McDonald to former senator Thomas Hendricks, and despite his assertion that he was not a candidate,

94

B. E. SHAFER AND E. M. SAWYER

the convention moved quickly to make Hendricks vice president under Cleveland.13 The 1888 convention unanimously endorsed Cleveland plus former senator Thurman of Ohio, Vice President Hendricks having died. Yet this time, the effects of an economic downturn during the first Cleveland administration could not be overcome, and the Democratic ticket lost New York, Indiana, and the election as a whole.14 Cleveland sought to return in 1892, and unlike most defeated presidents, he was nominated easily on the first ballot, with the leading cluster of dissenters going for David B. Hill, the regular (machine) Governor of New York. There was yet again a serious Indiana contender for the vice presidency, former governor Isaac P. Gray, but he was defeated by Adlai E. Stevenson of Illinois, Postmaster General in the first Cleveland Administration. The Cleveland-Stevenson ticket would once more carry New York, Indiana, and the country as a whole. Accordingly, a structure of factions and constituencies that was first clarified in 1876 was still alive and functional in 1892. Moreover, this resolution had been foreshadowed even earlier, in 1868 by the nomination of Horatio Seymour, reform governor of New York, and then again in 1872 through the Electoral Votes cast for Thomas Hendricks, former governor of Indiana.15 Said differently, the Democratic Party fronted some variant of this ticket for 1868 through 1892, realizing this combination but then losing in 1876; failing to realize it in 1880 and then losing; realizing it again in 1884 and winning; failing to harvest it for a second term in 1888 thanks to an economic downturn; but it bringing it back and winning again in 1892.

Morphing Factions The Great Populist Shift Yet the subsequent era and its struggles would suggest that only the possibility of a Cleveland (re)nomination had staved off what would soon be a very different factional struggle, a different nominating coalition in response, producing a different electoral strategy and, last but not least, a very different result. A broad-based populist surge in 1892 had already seen a genuine Populist Party carry four states and drive the Democratic ticket into third place in four others. Nevertheless, the recapture by Cleveland of New York and Indiana, along with Illinois, had been enough to

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

95

allow the party to shed Democratic support to the Populist ticket and still win the election. Four years later, the populist surge was still growing. And this time, for the first time since 1880, there was no Grover Cleveland to sustain the old compromise and no fresh reform governor of New York to substitute for him.16 By 1896, in fact, the Democratic national convention would be so thoroughly infused by the supporters and policies of this populist surge that the only serious question was which populist would it to carry its banner in November. Once that question was answered, the only question for the next three Democratic nominations would be whether the winner in 1896, William Jennings Bryan, was going to stand again. The Democratic national convention of 1896, and associated struggles, could hardly have been more different from those characterizing the Democratic convention of 1892 only four years previous, the last to put forward former Governor and former President Cleveland. The issues were different, the factions were different, even the political roots of the candidates were different. In the search for a nominee among these aspirants, the specific rise of currency, initially the central policy issue for populism, allowed an easy commonality between southern and western delegations. Residual resistance on behalf of hard money and high tariffs was present in the northeast, but there was no reform governor in New York to broker a compromise.17 When the first item of business for the convention was the successful replacement of Senator David Hill, regular party leader from New York, as temporary chairman by someone favoring the unlimited coinage of silver, it became clear that a standard-bearer for the silver forces was destined to be the nominee. Fourteen pro-silver candidates were then nominated—gold supporters declined to name even one—but the leading contenders were sitting congressman Richard P. “Silver Dick” Bland of Missouri versus a former one-term congressman, William Jennings Bryan of Nebraska. Bryan was a delegate thanks only to a successful credentials challenge, yet he not only managed the platform debate for the free-silver forces, designating himself as the final speaker. He leveraged that opportunity further with one of the great convention orations of all time, his “Cross of Gold” speech.18 Bland led solidly on the first ballot, with Bryan following. The balance began shifting on the second and third ballots. Bryan pulled into the lead on the fourth, and the resulting bandwagon was sufficient to pick

96

B. E. SHAFER AND E. M. SAWYER

up nearly everything on the fifth and final ballot except the Pennsylvania vote for a favorite son.19 The electoral result was to be much less happy, with William McKinley holding the northeast and the industrial midwest for the Republicans against a Bryan vote in the south and mountain states. Yet the balance of forces inside the Democratic Party was unchanged four years later, when Bryan was renominated without opposition and the convention brought back Adlai Stevenson of Illinois, Vice President in the last Cleveland administration, as Bryan’s running mate. The silver issue itself survived only because Bryan was attached to it, while the populist policies unifying the convention were refocused on anti-monopoly, anti-trust, and anti-imperialism.20 Another solid loss in November preceded the withdrawal of Bryan as an active candidate for renomination in 1904. In looking for some other solution, the Democratic Party did go back to New York for a nominee, in the person of Alton B. Parker, chief justice of the New York Court of Appeals, a very conservative choice. Bryan spoke against Parker from the convention podium, yet his personal unwillingness to support William Randolph Hearst, the leading progressive alternative, more or less sealed the outcome, and switches brought Parker the nomination at the end of the first ballot. Parker brought even worse results in the 1904 election, and the easy return of Bryan as the nominee in 1908 suggests that little had changed in an underlying party structure: returning for an effectively uncontested third nomination, he gathered 89% of the convention vote on the first ballot. John W. Kern, former gubernatorial candidate in Indiana, was chosen by acclamation for vice president, and the Democratic Party had its ticket, essentially rerunning the presidential contests of the previous three elections. The headline story of the 1912 convention would be the extended struggle between House Speaker Champ Clark of Missouri and Governor Woodrow Wilson of New Jersey, a story covered thoroughly in Chapter 2 where it provides excruciating testimony to the power of the two-thirds rule. Yet Bryan would remain a force inside the Democratic convention. He was narrowly defeated as its temporary chairman; he turned down the chairmanship of the platform committee; and later in the balloting, he announced a change of support from Clark, to whom he was bound by Nebraska primary rules, to Wilson. The latter did ultimately prevail on the forty-sixth ballot, and he offered one late and distant echo of the factional compromise uniting the party under Grover Cleveland when he took Governor Thomas R. Marshall of Indiana as his running mate, the last of a long line of Indiana ticket-balancers.

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

97

New Division, Old Problem A dramatic split in Republican ranks in 1912, more than any obvious factional rearrangement among the Democrats, would rescue the Democratic presidential ticket and break this string of Republican victories. Woodrow Wilson, Governor of New Jersey, would have sufficient attraction to the populists—even Bryan himself—to gain the nomination and hold the presidency in both 1912 and 1916. Yet had the Great Depression not arrived in 1929, the New Deal not arrived in 1933, and a huge shift in party identification not arrived by the end of the (four) presidencies of Franklin Roosevelt, the years after Wilson might have looked like nothing so much as an exaggerated return of the factional complexity that had bedeviled the Democrats in the years leading up to the Civil War. As it was, those the post-Wilson contests were to provide a foreshortened incarnation of exactly that. In the aftermath of Wilson—really in the aftermath of Cleveland, Bryan, and Wilson—the party had no obvious leading figure. What it had instead was a crazy quilt of cross-cutting internal divisions. There were three older regional blocs, northeastern, southern, and western. There was a continuing split between ideological populists and party regulars. On top of which there was a major emergent division between rural and urban areas. The latter would map onto both of the other factional divides, while it moved to the center of internal conflict for the Democratic Party. But it would be the Great Depression, and not any factional bargain, that would provide the solution this time. The renominating convention of 1916 was unconflictual, with effectively unanimous approval of both Wilson and Marshall. Wilson actually desired to be nominated a third time in 1920, but declining health coupled with declining popularity kept him from going public with a campaign, thus keeping him outside the efforts to find a successor. This allowed twenty-four candidates to receive votes on the opening ballot of the convention, four of them with serious but cross-cutting support. William Gibbs McAdoo of California, former Secretary of the Treasury, was a strong progressive and a committed supporter of Prohibition. Attorney General A. Mitchell Palmer of Pennsylvania was the creator of the “Palmer Raids” against domestic radicals and a spokesman for social conservatism. James M. Cox, Governor of Ohio was a progressive reformer and supporter of Wilsonian internationalism. And Alfred E. Smith, Governor of New York, was a progressive Catholic, a favorite of Tammany Hall, and a strong ‘wet’ on Prohibition.

98

B. E. SHAFER AND E. M. SAWYER

The story of their changing support became as good an introduction as any to the internal chaos of the national Democratic Party in the years after Wilson. McAdoo and Palmer dueled through the first six ballots. McAdoo and Cox dueled through the next five. Cox led McAdoo from the twelfth through the twenty-ninth. McAdoo retook the lead on the thirtieth through the thirty-eighth. A fresh boomlet for Palmer appeared on the thirty-fifth but died shortly thereafter. Cox regained the lead on the thirty-ninth, as Indiana shifted from McAdoo to Cox,21 and while the two candidates continued to struggle through the forty-third ballot, major shifts to Cox in other states made him the presidential nominee at that point. His choice for vice president was Franklin D. Roosevelt, Assistant Secretary of the Navy, who was backed by acclamation. All of which was simplified in the worst way in the nominating contest of 1924. The convention quickly settled into a two-man contest between McAdoo and Smith, pitting McAdoo as Protestant, rural, dry, nativist, and Klan-tolerant, against Smith as Catholic, urban, wet, immigrant, and anti-Klan. In a convention spanning 17 days and 103 ballots, McAdoo led Smith through the eighty-fifth ballot, lost the lead to him from the eighty-sixth through the ninety-third ballot, then regained it from the ninety-fourth through the ninety-eighth ballot. Always in the second tier was John W. Davis of West Virginia, Solicitor General and then Ambassador to Britain, who gained through the twenty-fourth ballot, receded until the ninety-fourth ballot, then surged from the ninety-fifth to the one-hundred-third, when he emerged as the nominee. The convention turned to Charles W. Bryan, Governor of Nebraska and younger brother of William Jennings Bryan, as its Vice President. That experience led party leaders of all stripes to look for a different bargain for 1928. In it, the rural side received the convention itself, in Houston, the first to be conducted in the south since the disastrous gathering of 1860. While the urban side got the candidate, Al Smith, Governor of New York. Though the deal was cemented only when William Gibbs McAdoo, clearly capable of disrupting it but sharing the general revulsion with the 1924 convention, declined to be a contender. Nominated by Franklin Roosevelt, Smith came within ten votes of the necessary two-thirds, Ohio quickly shifted from its favorite son to Smith, and the latter ended this first nominating ballot with 77% of the total.22 Joseph T. Robinson of Arkansas, minority leader of the Senate, was easily confirmed as his running mate, the first southerner to be nominated for national office since the Civil War.

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

99

Another New Democratic Party The Roosevelt Presidencies The disastrous factional decade of the 1920s, capped by a huge loss by Smith to Republican Herbert Hoover in 1928, when Smith fell so far as to cede half the south, appeared to have the potential to roll on and on. Instead, it was to be replaced by the longest period of Democratic presidential dominance in American history: from 1932 to 1968, the Republicans were to secure exactly one president. The immediate stimulus for this Democratic hegemony was the Great Depression, often treated, along with the Civil War, as one of the two great crises in the history of an independent United States. In response, both parties would change the configuration of their internal factions in major ways. Political change itself came in two pieces. The first was tied tightly to the candidacy of Franklin D. Roosevelt—a reform governor from New York yet again. Roosevelt was to secure four terms as president, 1932– 1944, and his vice president, Senator Harry S. Truman of Missouri who succeeded him upon his death in 1945, was to secure renomination (and reelection). So this half of the long New Deal order really was more or less one united piece. Though if there was little variation to the ultimate nominees during this period, the factional structure of the party changed in a major way. Franklin Roosevelt entered the convention in 1932 with an absolute majority but not the necessary two-thirds. His leading opponent was the man he had nominated in 1928, Al Smith, and the only other candidate who was more than a favorite son was John Nance Garner of Texas, Speaker of the House. Neither were plausible nominees, so the main strategic question was whether the Roosevelt forces could maintain sufficient control of the convention to strike the deals that would garner enough from a lesser array of favorite sons to get him from 58 to 67%. The successful defeat of credentials challenges to two Roosevelt delegations and the successful imposition of their preferred permanent chairman suggested that they would. The nominating vote itself was nevertheless static for three ballots, before McAdoo announced that California was switching from Garner to Roosevelt. Supporters of all those other favorite sons, along with most of the support for Garner, followed quickly. Garner was then the unanimous choice for vice president, and the platform promised not what became the New Deal, but rather a cut of 25% in federal spending and pursuit of a balanced budget.23

100

B. E. SHAFER AND E. M. SAWYER

Despite assembling in the shadow of major pieces of the New Deal, the ones that would become the core of an American welfare state, the 1936 Democratic convention would then feature no floor debates and no roll-call votes. A proposal to eliminate the two-thirds nominating majority did get some contentious attention, but this was defanged with a promise of abolition prospectively, for 1940, with a side payment to the south in the form of a change in the allocation formula for state delegations. Both Roosevelt and Garner were thus renominated by acclamation, and the convention could spend its entire time lauding its ticket. The 1940 convention likewise resulted in the overwhelming renomination of Roosevelt, now for a third term, but only after a surface display of reluctance by the candidate and some marginal opposition from former allies, Vice President Garner and Postmaster General James A. Farley.24 More conflictual were struggles over the vice presidency and the platform. For the former, Roosevelt shifted from Garner to Henry A. Wallace, his Secretary of Agriculture and a leading liberal ideologue. Wallace did secure a majority over his closest competitor, Speaker of the House William B. Bankhead of Alabama, though most of the southern and border states plus a serious smattering of others deserted him, to the point where he was not invited to address the convention. For the platform, Roosevelt struggled to conciliate a strong isolationist minority that went beyond opposing participation in the Second World War to opposing American rearmament. For 1944, Roosevelt moved early to confirm his desire for a fourth term. With no show of reluctance this time, he secured 92% of the vote on the first ballot, with the rest going to Senator Harry F. Byrd of Virginia in a civil rights protest. In effect, the nominee then threw the vice presidential nomination open, affirming Henry Wallace but stating his willingness to serve with Senator Harry S. Truman of Missouri or Supreme Court Justice William O. Douglas. Wallace led Truman on the first ballot, though a vote equal to that for Wallace was scattered among a wide array of favorite sons. Truman caught Wallace narrowly on the second ballot, courtesy of fresh support from the south. When a further round of shifts was finished, Wallace was left with only Iowa and Minnesota. Harry Truman, who had to be recruited explicitly by Roosevelt as vice president, actively sought (re)nomination in his own right in 1948.25 Some party leaders flirted with Dwight Eisenhower, who proved uninterested, and with William O. Douglas, who had confirmed his lack of interest at the previous convention. So Truman was an easy first ballot

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

101

victor, losing only ten states from the former Confederacy which voted for Senator Richard B. Russell of Georgia, along with Mississippi which withdrew from the convention. Truman added Alben W. Barkley of Kentucky, the convention keynoter, as vice president, and the remaining convention fight was over the civil rights plank, stiffened by Mayor Hubert H. Humphrey of Minneapolis in an amendment that lost mainly the south, the border, and some smaller western states. The Late New Deal Era A second cluster of New Deal presidencies, from 1952 through 1968, would lack the personal thread connecting its predecessors but otherwise be characterized by the evolution of this factional structure, now as it applied to internal party conflicts. The dominant faction would remain the major industrial states of the northeast, especially those that possessed old-fashioned organized parties. The dissent would surface on the right from Democratic Parties with an even longer lineage, in the south and on the border, and from Democratic Parties on the left with a volunteer activist structure, scattered throughout the country but most commonly found in the midwest and west. Yet the dominant faction, operating through a shared understanding among their official party structures rather than through any coordinating leader and unified by their focus on economic policy and social welfare, ordinarily remained the essence of the story. The Roosevelt thread connecting all these nominations made it harder than it would otherwise have been to delineate the factional structure of the Democratic Party as bequeathed by Roosevelt, one organized around attitudes toward the welfare state and one that favored labor liberalism rather than populism as its ideology. So the Democratic nominating contest of 1952, the first with no evident Roosevelt connection, became a good introduction to the factional structure of what had become a lasting environment. From the start, party leaders from major industrial states with old-style organized parties were thought to favor Governor Adlai E. Stevenson of Illinois, though Stevenson averred that he wanted only to run for reelection. Senator Estes Kefauver of Tennessee, just as clearly the candidate of independent activists and volunteer parties and not of regular party leaders, took actively to the primary trail in an effort to demonstrate electability while acquiring actual delegates in states that offered the possibility.

102

B. E. SHAFER AND E. M. SAWYER

Kefauver did lead on the opening ballot, with Stevenson second and Senator Richard B. Russell of Georgia third, having collected much of the south. Averill Harriman, a favorite son from New York, followed in fourth place, though after the second ballot, Harriman and the next-leading favorite son, Governor Paul A. Dever of Massachusetts, shifted their delegations to Stevenson. The third ballot magnified this with serious gains in Michigan and Pennsylvania, and Stevenson achieved his majority. The candidate then chose Senator John J. Sparkman of Alabama as his running mate, in a familiar north–south balance.26 While all of this was a factional story destined to be familiar through 1968, this contest also featured several secondary characteristics that would be far more critical to the succeeding period than to this late New Deal era. Foremost among these was the rise of a factional politics inside the Democratic Party that increasingly pitted issue activists against the regular party. Eventually, this rise would split the major states, with California and New York and activist leaders and Pennsylvania, Illinois, and Ohio as regular bulwarks. In the beginning, however, Stevenson was a sufficiently issue-oriented candidate with a cerebral style, making him satisfactory to issue activists, while the primary trail contained too many contests that were largely disconnected from actual delegate selection to allow an activist outsider like Kefauver to overcome the regular party base. The rise of the civil rights issue, and the increasingly strident policy positions that came with that rise, did render the south an autonomously dissident faction on its own, but one that was decreasingly likely to triumph over the other two. Stevenson effectively suppressed these distinctions when he sought the nomination again in 1956, by contesting Kefauver in the primaries.27 There remained few contests where the two men could go head to head and produce delegate results, but the Oregon primary, where candidates were entered in spite of themselves, and then the giant—and effective— California primary were both won by Stevenson. He was solidly selected on the first convention ballot, and the only real voting contest arrived with the vice presidential nomination, which Stevenson left open to the convention. Estes Kefauver led on the first ballot, followed by Senator John F. Kennedy of Massachusetts, followed by Senator Albert A. Gore of Tennessee. Kennedy gained the lead on the second ballot with major increments across the south, and when Gore was recognized during this ballot to endorse Kefauver, the bandwagon continued to roll in the other direction courtesy of the midwest and west, taking Kefauver (not Kennedy) to a majority.

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

103

Having narrowly failed to be nominated for vice president, Kennedy sought to move up in 1960.28 His vice presidential effort in 1956 had made him the tepid front-runner, though he was forced to follow the Kefauver strategy this time, pursuing the nomination through the primaries because many regular party leaders, who otherwise preferred him, were unsure that his Catholicism was not an insuperable negative. Senator Stuart Symington of Missouri made himself publicly available as a candidate, as did Senator Lyndon Johnson of Texas, Majority Leader in the Senate, while Adlai Stevenson was always waiting in the wings—he took no action but his supporters were vocal—should there be a deadlock. Yet only Senator Hubert H. Humphrey of Minnesota, as the champion of the activist forces, felt it necessary to join the primary contests, where crucial encounters occurred in Wisconsin, a majority Protestant state with a serious Catholic minority, and West Virginia, overwhelmingly Protestant and implicitly anti-Catholic. When Kennedy won both, he was able to enter the convention as the emergent choice. An immediate preconvention debate with Johnson did nothing to diminish the possibility, and Kennedy delivered a clear majority on the first ballot, capturing all the major non-southern states. The Massachusetts Senator then turned around and took Johnson, with his generalized southern support, as his vice president. The assassination of Kennedy in 1963 meant that it was President Johnson who stood for (re)election in 1964, and his convention became another in which there were literally no roll-call votes. Nominated by acclamation, Johnson appeared shortly afterward to reveal that Humphrey, a lion of the liberal activists, was his choice for running mate. A credentials challenge in Mississippi and a rules provision for a loyalty oath provided some conflict in committee, to the point where the Mississippi delegation and part of Alabama withdrew in protest, but even these required no ballots on the convention floor.29 So the usual factions were present, mixed in a slightly different way—Johnson a southerner who had acquired regular party credentials though his time as vice president and picked up the activists by adding Humphrey to the ticket. The Democratic convention of 1968 proved to be the last instantiation of the New Deal era, though the old pattern was already fraying badly. Senator Eugene McCarthy of Minnesota stood first for the activist part of the party to challenge President Johnson for renomination. When Johnson announced that he would not stand,30 Senator Robert

104

B. E. SHAFER AND E. M. SAWYER

F. Kennedy of New York entered the contest after the New Hampshire primary; President Johnson exited; and Kennedy won most of the next round of primaries, including the giant head-to-head contest in California where he was assassinated, leaving Senator George S. McGovern of south Dakota to stand in for him at the convention. Nevertheless, most of the primaries were still of the old-fashioned variety, conveying preferences but not delegates, and the familiar New Deal coalition of strong parties from major industrial states was now supportive of Humphrey rather than McCarthy or McGovern. Accordingly, and despite struggles over rules, credentials, and platform coupled with serious disturbances in the street, Hubert Humphrey was nominated on the first ballot with 67% of the vote.

The Changing Nature of Party Factions Established Factions in a New Nominating Process At some point during the long mixed system of delegate selection, after 1912 but before 1968, the nominating politics captured by the notion of a bandwagon dynamic began to change informally. Formally, this process was still weighted toward old-fashioned party conventions with a leavening of presidential primaries, many of which were not very good at connecting public preferences to delegate outcomes. But if party officeholders were not ceding influence directly to rank-and-file voters, they were becoming more aware of—and following more closely—the progress of the various nominating campaigns at collecting delegates in states other than their own. The Democratic nominating contest of 1960, producing Senator John Kennedy of Massachusetts, or the Republican nominating contest of 1964, producing Senator Barry Goldwater of Arizona, were classic examples.31 The nationalization of American politics was both cause and effect of this informal development.32 So was the growth of national media of information. And so was a better educated public, increasingly interested in following a politics that, while still being managed by the official party, was now widely covered by mass media—and hence more attached to this interested public—than it had previously been. One major by-product, occurring a full generation before the arrival of the plebiscitary era but cemented in place by that arrival, was the informal departure of the nomination from the national party convention and its lodging instead in the process of delegate selection. After 1952, a nominating majority would

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

105

always be assembled before the national convention, and no convention would ever again go so far as a second ballot to confirm it. After 1968, in an effort to deal with the disastrous events of the previous convention, the Democratic Party moved toward comprehensive institutional reform. The most obvious product was a sharply different process of delegate selection,33 and the plebiscitary era had arrived by the time of the 1972 convention. As it turned out, however, the mutually adaptive process that had long been capable of being summarized by the notion of a bandwagon dynamic was unaffected, operating as easily in the new world as the old—despite the second-largest institutional change in the history of nominating politics.34 Yet many peripheral things were different. Favorite-son candidates disappeared, as delegates not won by one national aspirant would invariably be won by another. And party leaders had to exercise their influence not by holding party office but by being able to deliver primary electorates in their respective states. So the means of identifying party factions had to change in conjunction with all these other changes, formal and informal. Factions that would previously have been recognized through lasting coalitions of party leaders now had to be recognized as recurrent coalitions of state electorates instead. Located this way, the three key factions inside the Democratic Party remained recognizable from preceding contests. Yet each would evolve differently in this succeeding era, and all would have to express their preferences indirectly, through mass constituencies tapped by the primary process, rather than directly, though party officials who were already in place. The regular party remained rooted in economic welfare concerns and concentrated in the major industrial states in the northeast, though even they were now cross-cut by tensions between the organized versus volunteer parties. Just as the west and midwest would remain leading producers of volunteer parties driven by issue activists, though it would no longer be geographic region per se that explained their behavior, since they would continue to produce a minority of organized parties as well. And the south remained a separable and distinctive bloc, though the long resistance of the southern Democratic Party to civil rights was increasingly a rear-guard action, suggesting that a southern distinction, if it survived, would have to have a very different focus. The first contest of the new plebiscitary era would showcase this factional structure in its altered institutional world. By standing in for

106

B. E. SHAFER AND E. M. SAWYER

Robert Kennedy in 1968, George McGovern had achieved the status of a serious contender, one aspiring to be the champion of issue activists within volunteer parties in 1972.35 Both members of the 1968 ticket, Hubert Humphrey and Edmund Muskie, returned to seek the nomination, though Muskie faltered early, leaving Humphrey as the champion of the regular party. Governor George C. Wallace of Alabama completed the picture as the southern champion, in one more effort at opposition to civil rights. For 1972, this all came down to the late, giant, winner-take-all primary in California. When McGovern took this crucial contest and came away a slight lead in committed delegates, he was the apparent nominee— unless his emergent majority could be derailed at the convention itself. When it turned out that it could not, he was assured of nomination. The resignation of Republican President Richard Nixon opened the way in 1976 for a candidate who might transcend the factions on behalf of a new national unity, and James E. “Jimmy” Carter proved to be this wild card. A one-term former Governor of Georgia, he was to leverage a second-place finish in the Iowa caucuses plus a win in the New Hampshire primary into a successful nominating campaign.36 Carter added the southern faction by defeating George Wallace, most critically in the Florida primary. And when he won the Illinois and then the Pennsylvania primaries over Senator Henry M. Jackson of Washington, the aspiring champion of the regular party, the contest was effectively over. Carter’s conservatism on some signature Democratic policies did unsettle the activist wing of his party, causing some late losses to candidates who aspired to represent the issue activists and their volunteer parties. Yet that proved too little, too late: Carter secured 74% of the total vote on the first ballot, and his choice of Senator Walter Mondale of Minnesota, protégé of Hubert Humphrey, garnered 94%. Carter in 1980 became the second post-reform President to be challenged for renomination within his party, following Gerald R. Ford, Republican President, in 1976. For Carter, the challenge came from Senator Edward M. “Teddy” Kennedy of Massachusetts and was to be effectively dismissed early in the primary process, though not without leaving a residue of conflict and disappointment for the convention itself.37 Kennedy would win primaries in the major states of New York, Pennsylvania, and California, but Carter would prove untouchable in the south, pick up much of the midwest and west, and cut into the industrial states with some wins of his own. The drift of delegate totals appeared decisive, and when the Kennedy forces failed in a crucial rules challenge

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

107

at the convention itself, their candidate withdrew. So the Carter-Mondale ticket was effectively (re)confirmed, though the Carter forces lost so many platform fights to previous Kennedy proposals as to undermine any picture of ultimate unity. Walter Mondale came back to seek the presidency on his own in 1984, and while he began as the consensual front-runner, the primaries were to reopen the classic three-cornered factional struggle.38 Senator Gary Hart of Colorado, touting a “new politics,” became the spokesman for issue activists when he upset Mondale in New Hampshire and produced a string of early victories in primary states. Mondale righted the situation by winning in the delegate-rich states of New York, Pennsylvania, and Illinois in the northeast and industrial midwest. And the Rev. Jesse L. Jackson—about as different from George Wallace as he could be— gave the southern bloc an incipiently new twist by winning the Louisiana primary and picking up delegates in black areas across the south. Hart resurged slightly in the midwest and west at the end, including a narrow win in California, but proportional allocation of those delegates plus a solid win for Mondale in New Jersey on the same day gave the latter a clear convention majority. Michael S. Dukakis, Governor of Massachusetts, established himself early as the candidate of the regular party in 1988 by winning the New Hampshire primary and then enough delegates on “Super Tuesday”, the giant aggregate of primaries three weeks later, to sustain his campaign. Otherwise, there were two surprises. The more widely noticed was that a concentration of southern primaries within Super Tuesday strongly boosted the Jackson campaign, which harvested them solidly. So what had been the transition to a new southern faction in 1984 was a reality by 1988. And the other was that issue activists within the party never found their champion. Many thought that Gary Hart would return to play this role or that Senator Paul Simon of Illinois would inherit it, but Hart was forced to withdraw and Simon proved able to win only his home state. So Dukakis had a large and easy majority by convention time, and chose Senator Lloyd Bentsen of Texas as his running mate, a nod to an older form of ticket-balancing.39 The final election confirming the new presidential nominating process, a revised bloc structure within the party, and a new practical politics to go with them was 1992.40 The absence of an indisputable front-runner was exemplified by the fact that there were five separate winners in the first six contests. But when Governor William J. “Bill” Clinton of Arkansas picked

108

B. E. SHAFER AND E. M. SAWYER

up South Carolina on the Saturday before “Super Tuesday” and went on to win six out of eight contests on the day itself, he surged out from the pack. When he won both Illinois and Michigan the following week, he added major industrial states to his initial southern base and went on to win twenty-two of the twenty-three remaining contests. In that sense, the puzzle was not that a candidate who could put the south and the industrial states together was nominated—any candidate who could do that would be nominated—but only that the activist faction again acquired no serious champion. California Governor Edmund G. Brown, Jr., aspired to that mantle, but one of Clinton’s late wins was a head-to-head battle with Brown in his home state of California. When Clinton won that, he would go on to tally over 80% of the vote on the opening convention ballot. Clinton was to choose another moderate southern Democrat as his running mate, Senator Albert “Al” Gore, Jr., of Tennessee, thus asserting a new-found moderation for the national Democratic Party,41 and the renomination of the Clinton-Gore ticket in 1996 was to be uncontested and unanimous. Evolving Factions for a Plebiscitary Era In that peculiar sense, the 1996 non-contest for the Democrats marked the division between factional structures in the early years versus the more modern period of the plebiscitary era. This division could be observed even procedurally at the national party conventions. In earlier years, candidate adaptations to changed formal arrangements plus changed practical politicking could still introduce substantial conflicts over rules, credentials, and platform into what were otherwise predetermined nominating fights.42 In successor years from 2000 onward, the convention would largely surrender even these secondary conflicts and become mainly a device for publicizing the nominees, their programs, and their party.43 Yet the division between the two periods was better and more practically observed through the shaping of the politics of presidential nomination, where the factional structure underpinning internal conflicts would change in important ways. In the earlier years of the plebiscitary era, there had been three great if amorphous factions inside the national party, mapping roughly onto its internal geography: northeast versus south versus midwest/west. In the years after 1996, this picture would evolve into a much simpler and more generalized division, one more obviously structural and less directly regional, with the exception of

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

109

one giant constituency group that could move between the two sides as events dictated. On one side of this divide were the liberal activists, always more concerned with reformed political processes and proper policy positions, especially emergent positions. On the other side were the inheritors of the regular party, lacking many of the fungible assets of an older politics but still concerned principally with winning elections and making policy.44 And swinging between them with the potential to be decisive was that large constituency group of black Democrats, who were actual majorities of the party in many places but who could usually be decisive regardless, that is, if they moved as a bloc from one factional preference to the other in any given contest. The nominating contest of 2000 was to show all this in clear and simple form.45 Vice President Al Gore very much desired the presidency, and brought with him not just the regular elements of party organization but the main interest groups allied with it, namely union labor, women’s groups, and, most especially, the civil rights organizations. His main challenger—really the only serious one—was former Senator William W. “Bill” Bradley of New Jersey. Bradley aspired to be the champion of the activist wing of the party, talking about a new politics, reformed approaches, and emergent issues. In the event, Gore defeated Bradley solidly in the initial caucuses in Iowa, then more narrowly in the opening primary in New Hampshire. Bradley persevered until “Super Tuesday”, five weeks later, when Gore won all eleven primaries plus four caucuses, and Bradley suspended his campaign. The nominating contest of 2004 began as if it would follow a scenario every bit as clear-cut as that of 2000. Heading into the Iowa caucuses, the liberal activists had their candidate in Governor Howard Dean of Vermont, opposed by a major party regular in Rep. Richard Gephardt, Minority Leader of the House and a favorite of organized labor. Yet those opening caucuses were to change the identities of the two individuals who would thereafter play these roles. Senator John Kerry of Massachusetts became the regular party champion and Senator John Edwards of North Carolina the activist standard-bearer. Though if the specific identity of these role-players switched dramatically, the roles themselves—that is, the factional structure in the background and the nominating politics that went with it—did not. When Kerry won the New Hampshire primary, he became the front-runner. He and Edwards would then be one-two in a smattering of contests over the next three weeks. And when Kerry took

110

B. E. SHAFER AND E. M. SAWYER

all but one of the nine contests on “Super Tuesday”, including New York and the newly front-loaded California, Edwards withdrew. By contrast, the Democratic contest of 2008 had an evident champion for the party regulars but some ambiguity about an activist counterpart.46 Senator Hillary R. Clinton of New York was the regular champion, and the few other possibilities for that role disappeared quickly. John Edwards returned with his theme of “The Two Americas”, but it was Senator Barack Obama of Illinois who generated the early activist excitement, who had been a consistent opponent of the Iraq War, and who proved an excellent fund-raiser. When Obama upset the Iowa caucuses, forcing Clinton into third, his stock as the activist champion went up sharply. When she recovered in New Hampshire, with Edwards driven into third, the two contenders for what would be a long delegate contest had been confirmed. And this time, a “Super Tuesday” with no less than twentythree contests, rather than resolving the matter, split thirteen for Obama and ten for Clinton. No results on either side would thereafter qualify as knock-outs, but Obama did maintain a small delegate lead, and after the final contest in Montana, Clinton conceded. She could claim to have gotten more votes, though this included two major primaries, Florida and Michigan, which the Democratic National Committee had proscribed after rules violations, while also reflecting the disproportionate success of Obama in the caucus states, where small votes could deliver larger delegate margins.47 The factional story, in any case, resided elsewhere. The third great faction in the modern Democratic Party, black Americans, had initially supported Clinton as its established favorite. But as Obama became a live alternative, they shifted en bloc. In the process, they melded the activist wing with the black constituency—and awarded the nomination to Obama. He was to take Delaware Senator Joseph R. “Joe” Biden, Jr., an older and more experienced representative of the regular party, to balance his ticket, and the Obama–Biden ticket was to win in 2008 and be renominated without opposition in 2012. So it would be the Democratic nominating contest of 2016 where a remix of the three main factions constituting the Democratic Party would acquire its next incarnation—and where the usual role of all three pieces would somehow be a surprise to most observers. Hillary Rodham Clinton, Secretary of State under Obama, was judged to be such a strong representative of the regular party that none of what were taken to be potentially serious alternatives stood against her. So the surprise

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

111

came from an essentially unknown ‘minor’ challenger, Senator Bernard “Bernie” Sanders of Vermont, who exploded to prominence in a field otherwise characterized by the lack of established contenders but who then sustained his challenge onto the convention floor.48 The possibility that the activist faction was going to generate a lasting challenger was underlined in the Iowa caucuses, where Clinton edged Sanders narrowly from 49.9 to 49.6%. A week later, he turned the tables decisively in the New Hampshire primary, beating her from 60 to 38%. Clinton would right herself on “Super Tuesday”, winning seven of nine primaries, most by two-to-one. Thereafter, Clinton would win all of the major primaries with the exception of Michigan, but often narrowly enough that proportional rules provided only small delegate margins. The other thing that kept this contest grinding on was the activist character of Sanders support which, once mobilized, was less affected by individual outcomes, an asset magnified in the caucus states where low voter turnout made activist support a particular advantage. Yet the ultimate firewall against a Sanders possibility was the third major faction in the Democratic coalition, black Americans, who supported Clinton overwhelmingly, both in southern primaries where this support was determinative and in northern counterparts where it was a hugely valuable increment. So Clinton as a candidate and black Democrats as a faction provided striking confirmation of the modern factional structure of their party. In 2008, when Clinton began as the champion of the regular party and Barack Obama managed to become the champion of party activists, the outcome was effectively determined when black Democrats moved from Clinton to Obama. But in 2016, when Clinton again began as the champion of the regular party and Bernie Sanders emerged as the champion of party activists, her long-time connection to black Democrats and his failure to create a counterpart gave her two of the three party pillars, and the nomination. The Democratic nominating struggles of 2020 would be anticlimactic by comparison to either 2008 or 2016, with a bandwagon that was formulaic, decisive, and short. Though within this narrow window for factional conflict, many observers would still have time to believe that they saw two different, successive, inevitable nominees. At the start, the contest had multiple contenders for all three factional leads: (1) Former vice president Joe Biden had the nominal support of both the regular and the black factions, but this appeared tepid enough that he was joined among the regulars by Senator Amy Klobuchar of Minnesota and Mayor

112

B. E. SHAFER AND E. M. SAWYER

Peter Buttigieg of south Bend; (2) Senator Bernie Sanders of Vermont returned as an activist spokesman, but he was vigorously challenged for that mantle this time by Senator Elizabeth Warren of Massachusetts; and (3) there were two black aspirants with serious elective credentials in Senator Kamala Harris of California and Senator Corey Booker of New Jersey. Sanders began with a near-tie in the Iowa caucuses, a clear win in the New Hampshire primary, and another win in the Nevada caucuses, making him the obvious champion of the activists, with a lead in delegate totals among all contenders. Yet Biden righted his campaign with a huge win in the South Carolina primary, and his overwhelming support among black party regulars had two major portents. In one, party regulars now rallied to him, as both Klobuchar and Buttigieg endorsed Biden and withdrew. In the other, the black electorate gave him further major support three days later on Super Tuesday. That left only two contenders standing, Biden and Sanders, and the former had a sufficiently commanding lead that the latter withdrew before the next major contest in Wisconsin, a state that Sanders had won decisively four years before. We shall return to the 2020 contests and their Gain-Deficit Ratios in a different way in the concluding chapter. Here, the background story is just another win for the contender who captured two of the three factional pillars of the modern national party.

Factional Struggles and Candidate Fortunes Two major points about party factions and presidential nominations burn through this chronicle of nominations and factions for the Democratic Party, all the way from 1828, which is really before its founding, until the current moment. One is the power of factional arrays and the struggles within them to determine the specific identity of presidential nominees. Candidates always bring idiosyncratic characteristics to the nominating contest, but it is the factional structure of their political party that effectively chooses among them. The other is that while these factional structures are simultaneously central to linking major developments in society with its political and governmental institutions, they are themselves undergoing constant change. From one side, they are tasked implicitly with selecting one individual nominee from a field of presidential aspirants across time. Though from the other side, they cannot be

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

113

tasked, explicitly or implicitly, with directing the process by which this history unfolds. So what results is in some sense the last key piece in a description of how presidential nominations work over time, or at least the penultimate piece before tossing personal quirks or strategic anomalies into the final picture of any given year. In the funnel of causality that underpins this description, there is a bandwagon dynamic governing the actual politics of presidential nomination. This bandwagon rolls through a field of contenders that is largely a product of the occupational preferences and career trajectories dominant in any given period, a dominance that is in turn shaped by institutional arrangements, internal party structures, and external party balance. And finally, it is the factional structure of the party in question—the two parties do not need to present the same internal structure, and they usually do not—that actually selects within these fields and thus nominates a single potential president. Factions Over Candidacies The dominance of factional structures and factional struggles over the aggregate mass of ephemera that characterize news coverage in any given year is not a fact that would burn through the careful, detailed, but very idiosyncratic nature of this coverage as any individual contest unfolds. Yet few candidates have ever had the personal characteristics sufficient to overcome these demands from a factional structure. Among Democrats, Andrew Jackson probably qualifies. Or at least, he rather evidently circumvented—defeated—the entire previous structure for presidential nominations, and he certainly was the founding father of the Democratic Party in much more than a metaphorical sense, giving it not just programmatic direction but an operational structure—including the national party convention—that can be traced all the way up to the current moment. The four nominations of Franklin Roosevelt—five if you add the renomination of Harry Truman as an obvious Roosevelt derivative— would put him on many lists. He certainly holds the record for personal control of nominating politics across time. Beyond that, and more even than Jackson, he proved able to violate the main procedural norm governing all the years before him, to the effect that no president should seek and secure more than two terms. Yet it is hard to believe that Roosevelt was anywhere near as essential to Democratic victory in 1932

114

B. E. SHAFER AND E. M. SAWYER

as Jackson had been in 1828. Would Herbert Hoover really have been reelected if Roosevelt had not stood? Conversely, would Roosevelt reliably have been elected if he had sought the nomination one election earlier, in 1928, or even reliably nominated if he had waited one election, until 1936, to see how the Great Depression was playing out? At the other extreme is a larger body of nominees who were so thoroughly a product of factional negotiations that their nominations did not directly involve the candidates themselves. The list begins in the early days with those famous dark horses, James K. Polk and Franklin Pierce. Later, it includes the nominees whose elevation represented little more than a desperate default to some available standard-bearer, as with Horatio Seymour, who announced that he was not a candidate, and Horace Greeley, who was actually the candidate of another party. The modern world has destroyed the possibility of becoming known only after you are nominated, but if a candidate who was unknown to the general public when the presidential primaries began is treated as the modern version of the dark horse, then Jimmy Carter surely qualifies—and Barack Obama just might. Even more striking are the successful candidacies which embodied little beyond the factional structures out of which they emerged. The three Grover Cleveland nominations are archetypal here, being the solution to a factional deadlock whereby each of three main party factions would not tolerate a candidate from the other two but could stomach a reform governor from New York with his apparent loyalties to none. Yet there was really a fourth nomination that grew out of the same factional straightjacket, of Samuel Tilden in 1876, who was—what else?— a reform governor from New York. And one could reasonably add Horatio Seymour in 1872 to the same list. There were reform governors from other states; they could not succeed. There were non-reform governors from New York; they could not succeed. What mattered was being a New York Democratic reform governor, of which there could be a maximum of one at any given time.49 Factions as Evolutionary Structures On the other hand, if those pictures suggest that a factional structure was dominant and candidate characteristics were effectively incidental—which in fact they do and they were—then it is necessary to note immediately

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

115

that this factional structure itself was hardly unchanging. Factional structures and factional struggles would go on to connect (and simultaneously filter) grand societal developments on their way into nominating politics. Yet in doing so, they invariably changed with those societal developments. The factional configuration behind Democratic nominations from Tilden through Cleveland has already made this point implicitly: these were obviously unlike the dominant factional considerations that applied either before or after the years from 1872 through 1892. Yet there was far more dramatic testimony to the impermanence of factional structures in the face of social change, in the form of true national disruptions like the Civil War or the Great Depression. In both cases, the factional structure of the party beforehand did not long survive in the world after. Thus the gradual rise of abolitionism as an issue, increasingly entangled as it was with national expansion, forced a change in party politics on both sides of the partisan aisle, though in very different ways. For the Democrats, it is worth remembering that civil war was most definitely not viewed as inevitable, but rather as a pressing strategic problem for party politicians to evade, straddle, or manage. On the other hand, a factional structure that still allowed Polk in 1844 to escape the toughest choices through conscious diversion began more and more to force indigestible options that neither Pierce in 1852 nor Buchanan in 1856 could avoid. In turn, the war would leave the Democratic Party with a desperate rebuilding imperative. Responses were to differ, but not until 1896 would these aftereffects be implicitly exiled from factional calculations inside the party, at least as they applied to presidential nominations. Though at that point, the populist surge that changed the factional structure of the party, the policy program of the party, and even the social composition of the party was noteworthy for the way that it led Democratic presidential nominees not into office but into the political wilderness. Only the Great Depression brought to power a Democratic Party for which the two Cleveland presidencies had provided only transient relief and whose only other rescue was courtesy of the Republican implosion of 1912, creating a Wilson interregnum before the Democratic Party returned to the wilderness. What followed, despite a presidential campaign promise in 1932 to cut spending and reduce the deficit, was the New Deal, the creation of an American welfare state, and a factional balance inside the party that would still be in evidence almost forty years later, up through 1968. In that

116

B. E. SHAFER AND E. M. SAWYER

sense, not just the four Roosevelt presidencies and the Truman renomination, but also the Stevenson nominations, the Kennedy and Johnson presidencies, and the Humphrey nomination were all connected by the same factional dynamic—set in motion by the personal characteristics of none of these post-Roosevelt nominees but harvested by all. Nevertheless, there were conscious efforts over time to reconstruct the factional structures of the parties by remaking procedures, policies, and even constituencies within them. These centered on the deliberate reform of rules and procedures, and what has to be said is that most were effectively irrelevant to the larger nominating bandwagons that rolled through (and over) them across American history. Though there were exceptions, a few formal strictures and a few institutional shifts that were basic enough to determine how factional negotiations were conducted. The Democrats actually produced the great historical example of the autonomous power of a party rule, in the two-thirds requirement for a presidential nomination. A small set of nomination contenders were thereby condemned to achieve an absolute majority but be denied a nomination by failing to reach two-thirds. More amorphously but with equal force, institutional forms could determine the manner in which key factions expressed themselves. Here, the outstanding example is the long evolution from institutions of delegate selection that expressed factional wishes by way of party officialdom, that is, as one very concrete product of a state party structure, toward institutions of delegate selection that instead expressed factional wishes through voting constituencies, instantiating the factions by way of primary electorates. There were surely Democratic nominees who succeeded under the pure convention system but would not have prospered on the primary election trail, just as there have been recent nominees who succeeded in a dominant primary system but were evidently not favorites of the official party even when they were nominated. Yet in the end, what distinguished even these impressively different institutional arrangements was mainly the fashion in which they reflected the factional struggles of their respective eras. For in every case, it was still those struggles that determined the identities of the individual nominees chosen by way of their respective frameworks. At the same time, it is important to remember that all of this has been elicited from the factional story of one party, the Democrats. So Chapter 4 must turn to the counterpart story for the Republicans. This will add cases to the total account. It will allow testing generalizations based so far only on one party. And it

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

117

will permit the analysis to move on and ask how much of either story was due to competition between the two parties, rather than developments within one or the other.

Notes 1. Tremendously helpful compilations toward that goal include: Presidential Elections, 1789–2008 (Washington, DC: CQ Press, 2010); National Party Conventions, 1831–2008 (Washington, DC: CQ Press, 2010); Richard C. Bain and Judith H. Parris, Conventions Decisions and Voting Records, 2nd ed. (Washington, DC: Brookings Institution, 1973); “Presidential Primary Returns, 1912–2000”, in Guide to U.S. Elections (Washington, DC: CQ Press, 2001). 2. Daniel DiSalvo, Engines of Change: Party Factions in American Politics, 1868–2010 (New York: Oxford University Press, 2012). 3. M.J. Heale, The Presidential Quest: Candidates and Images in American Political Culture, 1787–1852 (New York: Longman, 1982). 4. Donald Ratcliffe, The One-Party Presidential Contest: Adams, Jackson, and 1824’s Five-Horse Race (Lawrence: University Press of Kansas, 2015). 5. Joel H. Silbey, Martin Van Buren and the Emergence of American Popular Politics (Lanham: Rowman & Littlefield, 2002). 6. Lynn Hudson Parson, The Birth of Modern Politics: Andrew Jackson, John Quincy Adams, and the Election of 1828 (New York: Oxford University Press, 2009). 7. Note that throughout this period, in the years before the Civil War, the south would express its regional preferences through such individuals rather than through direct regional champions. 8. Joel H. Silbey, Party Over Section: The Rough and Ready Presidential Election of 1848 (Lawrence: University Press of Kansas, 2009). 9. Joel H. Silbey, A Respectable Minority: The Democratic Party in the Civil War Era, 1860–1868 (New York: W. W. Norton, 1977). 10. Jennifer L. Weber, Copperheads: The Rise and Fall of Lincoln’s Opponents in the North (New York: Oxford University Press, 2006). 11. Michael F. Holt, By One Vote: The Disputed Presidential Election of 1876 (Lawrence: University Press of Kansas, 2008). 12. But not a regular (or machine) governor, who symbolized northeastern immigrants. And since New York was one of the rare states that could be carried by both parties, it might at any given time have neither form of Democratic governor but a Republican instead. 13. Richard E. Welch, Jr., The Presidencies of Grover Cleveland (Lawrence: University Press of Kansas, 1988). 14. Charles W. Calhoun, Minority Victory: Gilded Age Politics and the Front Porch Campaign of 1888 (Lawrence: University Press of Kansas, 2008).

118

B. E. SHAFER AND E. M. SAWYER

15. Though Hendricks acquired those votes only because the head of the ticket, publisher Horace Greeley, had died between election day and the official meeting of the Electoral College. 16. Robert C. McMath, Jr., American Populism: A Social History, 1877–1898 (New York: Hill and Wang, 1992). 17. Indeed, the Governor, David B. Hill, was a hardline machine product, raising all the problems that Grover Cleveland had always finessed. 18. Richard F. Bensel, Passion and Preferences: William Jennings Bryan and the 1896 Convention (Cambridge: Cambridge University Press, 2008). 19. Michael Kazin, A Godly Hero: The Life of William Jennings Bryan (New York: Anchor Books, 2007). 20. In a curious sense, the Populist Party of 1892, with 8% of the popular vote and 5% of the Electoral Vote had swallowed up the Democratic Party. 21. Indiana had begun with its own favorite son, Vice President Thomas R. Marshall, went to Cox on the 7th ballot, went to McAdoo on the 30th ballot, then moved back toward Cox on the 39th ballot, beginning his ultimately successful bandwagon. 22. A solid majority of those who never moved to Smith were from the south, and North Carolina, Texas, and Virginia did ultimately defect to Herbert Hoover, the first time any of the three had voted Republican since Reconstruction. 23. Michael Barone, Our Country: The Shaping of America from Roosevelt to Reagan (New York: Free Press, 1990), Chapters 6–14, and David M. Kennedy, Freedom from Fear: The America People in Depression and War (New York: Oxford University Press, 1999), Chapters 1–12. 24. John W. Jeffries, A Third Term for FDR: The Election of 1940 (Lawrence: University Press of Kansas, 2017). 25. Irwin Ross, The Loneliest Campaign: The Truman Victory of 1948 (New York: New American Library, 1968). 26. Paul T. David, Malcolm Moos, and Ralph M. Goldman, Presidential Nominating Politics in 1952, 5 vols. (Baltimore: Johns Hopkins University Press, 1954). 27. Charles A. Thomson and Frances M. Shattuck, The 1956 Presidential Campaign (Washington, DC: Brookings Institution, 1960). 28. Theodore H. White, The Making of the President 1960 (New York: Atheneum, 1961), begins his series of books on presidential selection that would run through his capstone volume in 1980, America in Search of Itself: The Making of the President, 1950–1980 (New York: Harper & Row, 1981). 29. Robert David Johnson, All the Way with LBJ: The 1964 Presidential Election (Cambridge: Cambridge University Press, 2009). 30. Albert Eisele, Almost to the Presidency: A Biography of Two American Politicians (Blue Earth, MN: Piper, 1972).

3

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE DEMOCRATS

119

31. For Kennedy, Theodore H. White, The Making of the President 1960 (New York: Atheneum, 1961); W.J. Rorabaugh, The Real Making of the President: Kennedy, Nixon, and the 1960 Election (Lawrence: University Press of Kansas, 2009). For Goldwater, Theodore H. White, The Making of the President 1968 (New York: Atheneum, 1965); Robert Alan Goldberg, Barry Goldwater (New Haven: Yale University Press, 1997). 32. William M. Lunch, The Nationalization of American Politics (Berkeley: University of California Press, 1987), and Byron E. Shafer, “The Nomination and the Convention: An Informal Disappearance”, Chapter 1 in Shafer Bifurcated Politics: Evolution and Reform in the National Party Convention (Cambridge, MA: Harvard University Press, 1988). 33. Byron E. Shafer, Quiet Revolution: The Struggle for the Democratic Party and the Shaping of Post-Reform Politics (New York: Russell Sage Foundation, 1983). 34. On the change and its politics, Byron E. Shafer, Quiet Revolution: The Struggle for the Democratic Party and the Shaping of Post-Reform Politics (New York: Russell Sage Foundation, 1983). 35. Bruce Miroff, The Liberals’ Moment: The McGovern Insurgency and the Identity Crisis of the Democratic Party (Lawrence: University Press of Kansas, 2007). 36. Gerald M. Pomper, ed., The Election of 1976: Reports and Interpretations (New York: David McKay, 1977), begins a series that would run quadrennially through 2000. 37. Timothy Stanley, Kennedy vs. Carter: The 1980 Battle for the Democratic Party’s Soul (Lawrence: University Press of Kansas, 2010). 38. Michael Nelson, ed., The Elections of 1984 (Washington, DC: CQ Press, 1985), begins a series that was still continuing in 2016. See also the analyses of the 1984 nominating contest in Larry M. Bartels, Presidential Primaries and the Dynamics of Public Choice (Princeton: Princeton University Press, 1988). 39. Austin Ranney, ed., The American Elections of 1980 (Washington, DC: AEI Press, 1981), with a 1984 volume as well. 40. This was the election that generated the series of analyses from James Ceaser and Andrew Busch, beginning with Upside Down and Inside Out: The 1992 Elections and American Politics (Lanham: Rowman & Littlefield, 1993), and going on to include Losing to Win (1996), The Perfect Tie (2000), Red Over Blue (2004), Epic Journal (2008), After Hope and Change (2012), and Defying the Odds (2016). 41. Further aggravating the liberal activist wing of the party that had so noticeably lacked a champion in 1992. Kenneth S. Baer, Reinventing Democrats: The Politics of Liberalism from Reagan to Clinton (Lawrence: University Press of Kansas, 2000).

120

B. E. SHAFER AND E. M. SAWYER

42. Paul T. David, Ralph M. Goldman, and Richard C. Bain, The Politics of National Party Conventions (Washington, DC: Brookings Institution, 1960). 43. Byron E. Shafer, “The Pure Partisan Institution: National Party Conventions as Research Sites”, in L. Sandy Maisel and Jeffrey M. Berry, eds., Oxford Handbook of American Political Parties and Interest Groups (New York: Oxford University Press, 2010). 44. Byron E. Shafer and Regina L. Wagner, The Long War Over Party Structure: Policy Responsiveness and Democratic Representation in American Politics (Cambridge: Cambridge University Press, 2019). 45. Richard Johnston, Michael G. Hagen, and Kathleen Hall Jamieson, The 2000 Presidential Election and the Foundations of Party Politics (Cambridge: Cambridge University Press, 2004). 46. Dan Balz and Haynes Johnson, The Battle for America 2008 (New York: Viking, 2009). 47. Texas provided the perfect picture of this difference, since it featured a primary during the day and caucuses in the evening. The regular faction tended to win the primary, while the activist faction tended to win the caucuses, as they did with Clinton and Obama respectively and as they could also be seen to have done in individual primaries versus caucuses around the country. 48. John Sides, Michael Tesler, and Lynn Vavrick, Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America (Princeton: Princeton University Press, 2018). 49. The argument might even be extended to the reluctant Horatio Seymour in 1868, yet another reform governor from New York.

CHAPTER 4

Factions, Constituencies, and Candidates: The Republicans

Across the full sweep of American history, it would be hard to argue that careful consideration of one major party would obviate the need for attention to the other. The simple flow of gains and losses suggests a differential importance to one party rather than the other at differing points in time. In turn, even the idiosyncratic details of each unfolding pair of major-party nominees, details that are otherwise useful mainly as evidence of the factional struggles that produced them, speak to an expectation of distinction rather than commonality by party in the course of these paired struggles. And in the end, even a deliberate attention to one party (and inattention to the other) would turn up recurrent cases where the individual story of one is powerfully shaped by the individual story of the other. Chapter 3 began with the longer-lived of the two contemporary American political parties, the Democratic Party. So Chapter 4 must bring in its main opposition, initially the Whigs and then the Republicans, giving them the same treatment. Using a common template for the two chapters does more or less double the number of cases that can help unpack the character of nominating politics across time, and especially of the factional struggles that were their main proximate drivers. Yet consideration of the other major party allows three not just two further relationships to emerge:

© The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_4

121

122

B. E. SHAFER AND E. M. SAWYER

• Democratic and Republican nominating politics can simply be responding to two very different internal situations, via two distinctive sets of factional struggles; • Democratic and Republican nominating politics can be responding instead to the same external context, yet react in very different ways; • Or Democratic and Republican nominating politics can be responsive to a counterpart politics inside the opposing party, rather than to their own partisan innards or even to major outside stimuli that might appear to be common.

The Search for an Organized Opposition The Precursors The success of Andrew Jackson in securing the presidency and launching an organized party that would long outlive him was by definition a problem—and a continuing one—for those who did not support him or his party. Yet the success of Jackson in coming back from defeat in 1824 and securing the presidency in 1828 left the anti-Jackson forces in serious disarray. The defeated campaign of John Quincy Adams, last of the Democratic-Republicans, was not a framework for ongoing opposition. Worse yet, the incipient anti-Jackson forces were not just diverse in their politics but less immediately inclined—and less able in the absence of a unifying leader—to copy the systematic organization that Martin Van Buren was putting together in support of Jackson.1 The result would be an early hodge-podge of aspiring alternatives, each of which would find it too difficult to hammer the various constituencies and interests unhappy with the ascendant Jacksonians into a coherent piece. One initial framework for opposing the new Democratic Party was offered by the Anti-Masonic Party, united mainly by its opposition to rising immigration and the group that in fact pioneered the national party convention. Yet their first choice for president, Henry Clay, Senator from Kentucky and Secretary of State under Adams, turned them down, and the party eventually appeared on only five (of twenty-three) state ballots, carrying only Vermont. A broader alternative was the short-lived National Republicans whose 1831 convention unanimously endorsed Clay, who accepted this particular nomination. The National Republicans would appear on eighteen state ballots, and Clay was to carry five states in the otherwise very solid reelection of Jackson.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

123

Four years later, organized remnants of this proto-party plus antiJackson figures from the Democrats and the Anti-Masons would be ideologically coherent enough to describe themselves as another protoparty, the Whigs, but not yet organizationally coherent enough to hold a national convention or even to rally behind one nominal leader.2 Instead, the Whigs ran three regional candidates nominated by state legislatures, in hopes of securing enough Electoral Votes to throw the election into the House of Representatives, where sympathetic congressmen might be able to create the missing single spokesperson. Senator Daniel Webster of New Hampshire was the New England incarnation of this strategy, Senator Hugh L. White of Tennessee was the southern incarnation, and General William Henry Harrison of Ohio stood for all the rest. Harrison managed to carry seven states, White two, and Webster one, but that was not nearly enough to overcome the fifteen generally larger states carried by Democrat Martin Van Buren.3 The Whigs These Whigs would nevertheless be the first to go on and make a consistent effort at uniting policy programs and regional interests behind an aspiring leadership core, even if the new party had arrived in that halting tripartite fashion. On the one hand, they would survive long enough to be recognized as the leading organized opposition. On the other hand, their triumph was always conditional, exemplified by the fact that the main factional divide inside the party would pit its emergent leadership and their aspiringly national program against everybody else in the coalition, that is, all of those who were not in the original leadership cadre and/or not lined up behind the original program. This might suggest that the Whigs were fated to be consigned quickly to the status of proto-party. Yet by 1839, elements of a putative Whig coalition instead felt coherent enough to hold a first national convention. The person who would provide a kind of long-run organizational glue was again Henry Clay of Kentucky, destined to be both Speaker of the House and Senator during the short but noteworthy lifetime of the Whigs. And its program was to be Clay’s “American System,” including a national bank to underwrite economic development, federal infrastructure projects to connect the differing regional economies, and tariffs to protect the entire project from outsiders. This became the essence of the Whig program, though its differential impact on various interests was often the

124

B. E. SHAFER AND E. M. SAWYER

explanation when Clay himself failed in his recurrent efforts to be the party’s presidential candidate. At its first national convention, the leader on the first ballot was indeed Henry Clay, followed by William Henry Harrison, who had outperformed expectations in the 1836 election, followed by General Winfield Scott of Virginia, veteran of the War of 1812 and subsequent Indian conflicts. Clay was believed to have even more support among minority clusters in a number of states, but because the party had adopted a rule compelling each state to cast its full vote as a unit, this support could never be realized. Faced with a choice, supporters of the third leading candidate, Scott, moved to Harrison rather than Clay, and Harrison emerged with the nomination. As partial compensation, former Democrat John Tyler of Virginia, not just a geographic ticket-balancer but also a close friend of Clay, became Harrison’s vice presidential candidate. Harrison was to defeat President Martin Van Buren in the general election, only to die in office a mere thirty-one days after assuming it. The constitutional result was the elevation of Vice President Tyler to the presidency, but the political result was to disrupt what had looked to be the most serious effort at building a national opposition party. Tyler did share some Whig programs, but as President he remained more of a states’ rights Democrat, actually vetoing some major Whig initiatives. In return, the Whigs expelled him from the party, making Clay their de facto national leader. So when the 1844 Whig convention returned to the party-building task, it began with the unanimous nomination of Henry Clay, despite complaints that he had already lost both a nominating contest and the presidency itself.4 Clay returned after losing the 1844 election to be in the hunt for a nomination again in 1848, once more opposed by two generals, both from Virginia this time and both with assets and liabilities opposite to his. On one side, Clay still possessed that comprehensive Whig program and the partisan allies to go with it. On the other side, Zachary Taylor, hero of the Mexican War, and Winfield Scott, combatant in the War of 1812, subsequent Indian conflicts, plus the Mexican War, were distinguished by few known policy positions and could hope to gather up all those for whom the Clay program did not fit. Taylor led on the first ballot, with Clay second and Scott third. On the second ballot, Clay supporters began to bleed away to both Taylor and Scott. And on the third, Taylor achieved his majority. To balance the ticket geographically and placate disappointed Clay partisans, the convention turned to Millard Fillmore, a Clay ally and State Comptroller of New York, as vice president.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

125

The splits that had bedeviled Clay in particular reached what would prove to be their apogee by 1852.5 The Compromise of 1850, designed by Henry Clay and driven through Congress by Clay and Senator Stephen Douglas of Illinois, had temporarily defused regional tensions over slavery and its application to new states and territories for the party in Congress. But the Whig delegates of 1852 were disproportionately Northerners who opposed the compromise strongly and southerners who supported it with equal intensity. Moreover, each side had a candidate: General Winfield Scott for ideological northerners, versus Millard Fillmore, who had become President upon the death of Zachary Taylor, for ideological southerners. These attachments were ironic. A Virginian, Scott was unpopular in the south, given his ambivalence on the Compromise plus his support by New York politicos. Conversely, Fillmore, a New Yorker, had major allies in the south, thanks to unhesitant support of the Compromise as President. Beginning at 133 for Fillmore and 132 for Scott, with a rump of 29 for Daniel Webster, the two leaders remained close through fortysix ballots. At that point, the Fillmore leadership signaled a willingness to support Webster if he could find another twenty delegates somewhere. He could not, and enough of Fillmore’s delegates then bled off to Scott to give the Mexican War hero a majority by the fifty-third ballot. After a resounding defeat in the presidential election of 1852 and with internal divisions exacerbated catastrophically by a legislative struggle over the Kansas-Nebraska Act, repealing the Compromise of 1850, the Whigs were essentially defunct by 1856. The thread uniting both main factions all along the way had been Henry Clay, usually through programmatic developments and coalition building, alternatively through the efforts of others to avoid him. Clay had been the nominee of the (pre-Whig) National Republicans in 1832, the first ballot leader but not the nominee in 1840, the nominee himself in 1844, one of three serious contenders in 1848, and the critical shaping figure for the internal contest of 1852.6 So it was almost a fitting coda to the demise of the Whig Party after 1852 that Clay himself was to die in that inauspicious year.7

126

B. E. SHAFER AND E. M. SAWYER

The Republican Founding The demise of the Whigs once again left opposition to the Democratic Party in disarray.8 There were again divergent policy preferences within this incipient opposition, split further among divergent geographic regions, though there was at least a cadre of national politicians who saw no future for themselves as Democrats and were looking for alternatives. One of two aspiring frameworks for coordinating this opposition was the American Party, more commonly recognized as the Know-Nothings, whose central concern was to limit burgeoning immigration from Europe, most especially Germany and Ireland. Yet the slavery issue had already reached the point where, having hastened the death of the Whigs, it would fatally damage the prospects of a new proto-party that simply ducked the issue. In their one and only national convention in 1856, the Americans easily nominated Millard Fillmore, former president, coupling him with Andrew Jackson Donelson of Tennessee, adopted son of Andrew Jackson. The party did manage to appear on the ballot in twenty-three of the thirty states, but the Fillmore-Donelson ticket managed to carry only Maryland. The other incipient possibility was the new Republican Party, unified by its opposition to slavery, though this was coupled with a more general penumbra of opposition to the policies of the Pierce Administration.9 Yet if slavery was not a principled problem for these incipient Republicans, who differed mainly in the intensity of their opposition and thus the policy implications that they were willing to consider, the programmatic benefits of a single-issue unity remained unclear. Offering a sharp-edged antislavery position immediately wrote off large parts of the nation (and its electorate). Though in return, the party might hope to receive whatever political residues remained from the former Liberty Party, with national conventions and a presidential candidate in 1840 and 1844, plus the Free Soil Party, with national conventions and a presidential candidate in 1848 and 1852, all marginal third-place finishers.10 Fittingly, the first Republican convention featured delegations from northern and border but not southern states. A preliminary informal ballot pitted John C. Fremont of California, famous not for his truncated senate term—he had resigned the only political office he ever held—but as a famous explorer of the American west, against John McLean of Illinois, Supreme Court Justice. A formal ballot after this preliminary test was then be resolved decisively, Fremont 520 and McLean 47. A similar informal

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

127

ballot for vice president showed William L. Dayton, former Senator from New Jersey, leading Abraham Lincoln, former congressman from Illinois, where the formal follow-up was equally decisive: 532 Dayton to 44 Lincoln. Being absent in the southern and most border states, the Republican Party appeared on fewer ballots than the American Party, yet Fremont did carry eleven of the sixteen where it was present, including the major states of Ohio and New York. By 1860, the American Party was defunct and this Republican Party had consolidated. At its second national convention, the main conflict was between Senator William H. Seward of New York and former congressman Abraham Lincoln, along with favorite son support for Senator Simon Cameron of Pennsylvania, Governor Salmon P. Chase of Ohio, plus a handful of others. Seward was backed by the powerful New York organization, while Lincoln was widely known for his series of debates with Stephen Douglas in a losing contest for the Senate, as well as for his work in helping to build Republican parties around the country. Seward led Lincoln on the first ballot, 174-102, with Cameron 51 and Chase 49. Pennsylvania switched from Cameron to Lincoln on the second ballot, bringing Seward and Lincoln essentially even. On the third ballot, Ohio (along with Kentucky) abandoned Chase for Lincoln, and much of the remainder followed, giving him a victory over Seward, 340 to 123. Lincoln was to take all four main challengers into his presidential cabinet, with Seward as Secretary of State, Chase as Secretary of the Treasury, Cameron as Secretary of War, and Edward Bates (favorite son from Missouri) as Attorney General, in the process keeping them on the inside rather than outside as potential challengers in 1864. In the meantime, he worked actively to continue building the Republican Party, and the combination produced an overwhelming renomination: Lincoln lost only the 22 delegates from Missouri, who cast a sentimental vote for their favorite, General of the Union Army Ulysses S. Grant. In part because he was rebranding the party as “Union” rather than “Republican,” Lincoln allowed his vice presidential choice to be an open ballot, where Democrat Andrew Johnson of Tennessee, the only member from a seceding state to remain in the Senate, replaced Hannibal Hamlin of Maine, Lincoln’s previous vice president. The assassination of Lincoln in 1865 brought Johnson to the presidency, but his early, consistent, and determined opposition to the Republican leadership over issues of reconstruction for the south made him an irrelevancy for the 1868 nomination.11 It might not have mattered, since

128

B. E. SHAFER AND E. M. SAWYER

General Grant was to secure the Republican nomination unanimously, even though his partisanship was sufficiently murky that some Democratic leaders in the previous year had hoped to make him the Democratic nominee instead. Like Lincoln, Grant left the vice presidency to the convention, where eleven candidates received votes on the first ballot. The eventual two-candidate race between Schuyler Colfax of Indiana, Speaker of the House, and Benjamin F. Wade, Senator from Ohio—won by Colfax—was noteworthy mainly because both were Radical Republicans with strict views on southern reconstruction. The 1872 renomination of Grant was again unanimous, facilitated this time by the fact that what would later be recognized as the reform faction of the Republican Party, the faction most offended by the rampant corruption of the Grant presidency, had already deserted to joint the short-lived Liberal Republican Party. There was a touch of conflict in a genuine contest over the vice presidency, which Schuyler Colfax held narrowly. Otherwise, the most notable aspect of change was a further toughening by what would come to be recognized as the regular faction, headquartered among the big northern machine parties, of their demands for actual black suffrage as a central aspect of southern reconstruction.

The Factionalization of the Republicans Years of Hegemony The first Republican contest without a Civil War icon was to confirm that below a previously unified surface, the Republican Party, apparently the new majority nationwide, was nearly as factionalized as the Democrats had been before the war. Though this factionalism did express itself through three clear and continuing major pieces. The regulars, the half-breeds, and the liberals would structure the conflict over presidential nominations well into the 1890s, before a new division finally recast this picture. With hindsight, this suggested that a majority party that had been brought to power by the Civil War would almost inevitably, if it remained the evident majority, acquire a factional structure to go with representing the largest share of a diverse nation. With the two great heroes of the Civil War no longer (apparently) available as nominees and with the party in its majority status, it would quickly become clear that the party was constituted from three clear-cut and continuing factions:

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

129

• The regulars were based in major states with strong party organizations, where the concrete rewards of politics were their central concern, though they were also disproportionately the ones still devoted to civil rights and strict reconstruction, a combination that gave them leverage over the southern delegations. • The half-breeds, in the terminology of the time, were focused on the policy program of a postwar party, including hard money and a protective tariff, and found their base in a congressional party that was tasked with generating actual legislative majorities for those positions. • And the liberals were the reform faction that had first revolted against the corruption of the Grant Administration, when this faction became central to the short-lived Liberal Republican Party, but they would otherwise rally reliably around a “clean politics” aspirants for the Republican nomination.12 Remarkably, the struggle to acquire the support of two of these three factions—and thus guarantee a nomination—would produce three distinctive combinations in the first three contests of this period. For 1876, Senator Roscoe Conkling of New York and Senator Levi P. Morton of Indiana, leaders of their respective party organizations, carried the banner of the regulars. The standard-bearer for the half-breeds was James G. Blaine of Maine, the Speaker of the House. And the liberals were led by Benjamin H. Bristow of Kentucky, reforming Secretary of the Treasury. Blaine led with 285 votes on the first ballot, followed by Morton with 124, Bristow with 113, Conkling with 99, and Gov. Rutherford B. Hayes of Ohio, a favorite son with no clear factional connection, at 61. Three more ballots did little to dislodge that ordering, but the fifth saw Blaine and Hayes progress at the cost of Morton and Bristow. After one more iteration, the seventh ballot saw supporters of everyone but Blaine move to Hayes, making him the candidate of the regulars plus the liberals and awarding him the nomination. Hayes did not seek renomination in 1880, Blaine returned as the champion of the half-breeds, but the contest was shaped most critically by an effort from the regulars to bring back Ulysses S. Grant for a third term. Grant did lead Blaine on the first ballot, followed at a distance by John Sherman of Ohio, Secretary of the Treasury, nominated by his largely unknown home-state congressman, James A. Garfield. A set of previous conflicts over rules and credentials, however, where the Grant

130

B. E. SHAFER AND E. M. SAWYER

forces attempted to reinforce the influence of major party organizations, had been turned back (albeit narrowly) by an anti-Grant coalition, and this ultimately proved crucial. Despite that setback, Grant sustained his lead through the thirty-fifth ballot, when Garfield was brought forward as a candidate, not by his home state but by Indiana and Wisconsin. He became the nominee one ballot later, as all of the anti-Grant people except a small rump of Blaine supporters—that is, the liberals plus the half-breeds this time—shifted to Garfield. Garfield took Chester A. Arthur, former Collector of the Port of New York, as his vice president, to heal divisions with the Grant forces and the regular faction. Yet when Garfield was assassinated and Arthur became president, the latter supported civil service reform, which jumbled the usual alignments for 1884. Arthur desired renomination, but his support was now centered in the south, thanks largely to the patronage power of the federal government with a skeletal southern Republican Party. Blaine returned for the half-breeds in a third try at the nomination, assuming the lead over Arthur on the first ballot, with two Senators at a distance: George F. Edmunds of Vermont for the reformers and John A. Logan of Illinois for the regulars. That result was roughly stable for three ballots, when the Logan forces shifted toward Blaine and enough others followed to create a nomination. Blaine then took Logan as his vice president—the combination of a half-breed plus a regular, the third possible combination in three elections—and the Republicans had their ticket. The rise of two major challenges to the party, first from the resurgent Democrats and then from new social movements in Republican circles, would prove more than sufficient to guarantee that this neat and enduring factional structure could not continue undisturbed. In the first, the return of the Democrats to competitive status would be limited to the three elections fronted by Grover Cleveland. When the Democrats were thereafter forced to absorb a major factional challenge from the surging populist movement, the Republicans would find a fresh extended restoration before an internal challenge of their own, from progressives rather than populists but with many common elements, would finally put the factional struggles of the immediate postbellum world to sleep for good. The defeat of Blaine in the general election of 1884 by Democrat Grover Cleveland did suggest that the Republicans needed to attend to shoring up their electoral majority. An implicit response began with the announcement by Blaine that he, the defeated standard-bearer, would not stand in 1888. The first upshot for the 1888 Republican convention

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

131

was an opening ballot scattered among fourteen candidates and foreshadowing a new factional ambiguity. Senator John Sherman of Ohio led, but with only a little over a quarter of the vote, followed by favorite sons from Michigan, New York, Iowa, and two from Indiana, one of whom was principally supported by Illinois. Among these, Chauncey Depew, attorney for the New York Central Railroad, withdrew after the third ballot as New York switched to another state’s favorite son, Senator Benjamin Harrison of Indiana. Harrison gained the lead on the seventh ballot, when delegates holding out for the return of Blaine gave up hope. And when the supporters of yet another favorite son, Senator William B. Allison of Iowa, broke toward Harrison on the eighth ballot, enough others followed to make him the nominee.13 Factional tensions inside the Republican Party hardly disappeared during the Harrison presidency to follow, but the president was in firm control of its 1892 convention, where he won renomination solidly on the first ballot. Two other candidates acquired a substantial vote, though curiously, both denied that they had ever sought the nomination. These were once again James G. Blaine, then Secretary of State and playing his last convention role, along with William McKinley, Governor of Ohio and actually permanent chairman of the 1892 convention. Harrison had fallen out with his vice president, Levi Morton, most pointedly over Morton’s lack of support for a major civil rights initiative, and he replaced Morton with Whitelaw Reid, also of New York, owner of The New York Tribune and ambassador to France. The ticket, however, was to fall to Grover Cleveland in the general election, coming back successfully in his third nomination and suggesting that Republican victories were still hardly automatic. Years of Restoration Consistent efforts to round up a delegate majority before the convention would not become a central feature of nominating politics until the 1950s, and not be institutionalized until the 1970s. Moreover, such efforts in a year without a sitting president were understood to be optimally challenging. Yet the most famous exception, and in that sense the most famous predecessor to the modern world, belonged to Mark Hanna, Ohio businessman, in his efforts to build an advance majority for William McKinley, Governor of Ohio, at the 1896 Republican convention. Prior conversations among party leaders who happened to know each other,

132

B. E. SHAFER AND E. M. SAWYER

along with individual introductions to potential candidates, were hardly novel, but Hanna is usually regarded as taking these to a whole new level. The result proved overwhelming, giving McKinley 72% of the vote on the opening ballot of the 1896 convention and leading on to a Republican come back and a McKinley presidency.14 The McKinley triumph was to usher in another period of apparent hegemony, while bringing an end to the antebellum Republican world, with its three-cornered factional conflicts. The period of uncertainty and instability in the years leading up to the McKinley nomination disappeared with that old factional world, and in its place, there would be an extended era of factional calm, before a greater internal storm finally broke this up too. Yet during this extended calm, a sequence of nominating knockons made it look as if McKinley (and Hanna) had found the secret to banishing Republican factionalism: first through the McKinley nomination; then through the McKinley renomination; into the nomination of McKinley’s vice president, Theodore Roosevelt; and on to the nomination of Roosevelt’s vice president, William Howard Taft. That was an impressive series of consensual hand-offs from one nominee to another. Already by 1896, there remained only tiny hints of incipient divisions of any sort, as with small contrary efforts on behalf of Thomas Reed of Maine, another Speaker of the House, plus Levi P. Morton, Governor of New York, who both became part of a failed strategy by old-line regulars to force McKinley to bargain further with them. He ignored the possibility. Just as there were only tiny harbingers of programmatic conflicts to come, conflicts between conservatives and progressives that would ultimately destroy any notion of a simple and ongoing intraparty consensus. The best example was two minority planks supporting bimetallism rather than the gold standard, defeated in such a lopsided fashion as to suggest the absence of any serious policy disputes. In the meantime, McKinley was to win the election and be renominated unanimously in 1900. His previous vice president, Garret A. Hobart of Indiana, had died in late 1899, and McKinley left the choice of a running mate to the convention this time, asking even Hanna not to intervene. From the perspective of its time, the result was another overwhelmingly consensual nomination, for Governor Theodore H. Roosevelt of New York, hero of the Spanish–American War and reform governor from a machine state—whose Republican Party was happy to see a strong rival promoted up and out. Even the platform was not just

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

133

undisputed but essentially unchanged, supporting the gold standard and a protective tariff while praising geographic expansionism and foreign trade. McKinley was to be assassinated in September after reelection, and Roosevelt became president. Yet despite some incipient programmatic— and huge stylistic—differences between the two, consensus and unanimity remained the dominant motif through the convention of 1904. Roosevelt collected all 994 convention votes in his renomination, while the party favorite for vice president, Senator Charles Fairbanks of Indiana, was endorsed by acclamation. Moreover, for a presidential nominee who would later return as the lion of progressivism, the platform was again essentially a repeat—the gold standard and a protective tariff, an expansionist foreign policy, and expanded foreign trade. The party appeared to be united and self-satisfied. Which it would be one more time in 1908.15 Roosevelt declined to stand for reelection but worked actively on behalf of his Secretary of War, William Howard Taft, who was nominated overwhelmingly. A 72% vote for Taft on the opening ballot did leave room for a host of states to honor favorite sons—Senator Philander Knox of Pennsylvania, reform Governor Charles Evans Hughes of New York, Speaker Joseph G. Cannon of Illinois, Vice President Charles W. Fairbanks of Indiana, and Senator Robert M. LaFollette of Wisconsin—an array hinting at far more variety in Republican ranks than the preceding three conventions would have suggested. Yet nothing interfered with the easy endorsement of Rep. James S. Sherman of New York as vice president or with adoption of the familiar platform one more time, where a comprehensive succession of minority planks from LaFollette were essentially flicked away: 952-28, 880-94, 917-63, 866-114.

The Factions Reconstituted The stunning rupture of Republican unity in 1912, second only to the Democratic implosion of 1860 in its intensity, would attest to a very different factional divide emerging beneath the surface calm of these extended nominating hand-offs, a divide quite unlike anything that had preceded them. Needless to say, once this new division appeared, on a scale suggesting that it could not simply go away, it was necessary to treat internal Republican politics in terms of two emergent grand factions, progressives and conservatives, as the key to understanding subsequent nominations. Some means for reintegrating the progressives

134

B. E. SHAFER AND E. M. SAWYER

in a larger Republican consensus became essential, and it would take a pair of conventions to carpenter that together. Though when this had been accomplished, the 1920s would go a long way toward obscuring the depth of the factional division that had so recently preceded them, and that still bubbled along under a new surface calm. Nothing in the apparently lasting McKinley reconstitution of the party would have prefigured the Republican convention of 1912. Already by 1908, Roosevelt had come to recognize a growing split beneath the surface between progressive and conservative factions, but he viewed Taft as an appropriate compromise to keep this in abeyance. By 1912, on the other hand, Roosevelt had decided that Taft was increasingly an agent of the conservatives, and thus part of the problem, just as he himself leaned increasingly to the progressives. Yet because Taft and not Roosevelt now controlled the party machinery, Roosevelt had to challenge Taft for renomination by way of the presidential primaries. Formally, Roosevelt defeated the President in nine of the ten where both names were entered, losing only Massachusetts. Practically, however, most of these were merely preference polls, conveying no actual delegates. Early convention battles over the temporary chairmanship and over the right of challenged delegates to vote on other credentials matters were close but decided in favor of the Taft forces, as almost all of the south lined up with Taft, most of the midwest and west lined up with Roosevelt, while the big industrial states were split and contentious— though Roosevelt’s home state of New York was unanimously against him. When the credentials challenges that did make it to the floor were uniformly dismissed, Roosevelt asked his delegates to remain in the hall but cease participating. Taft thus received 52% of the total convention but 76% of those who voted, and the tally renominating Vice President James S. Sherman was similar. Roosevelt responded by underlining his availability for an insurgent campaign, and a new Progressive Party was announced shortly thereafter, with Roosevelt for president and Governor Hiram W. Johnson of California as his running mate. Roosevelt would ultimately push Taft into third place in both the popular and the Electoral vote, but be crushed in both regards by Democrat Woodrow Wilson.16 In that context, what distinguished the Republican convention of 1916 would be its dual interactive negotiations. The Republican and the Progressive parties held their conventions in Chicago in the same week in June. Yet the committee to work out a joint ticket was initially stymied by the insistence of the Progressives

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

135

that the nominee had to be Roosevelt, and the counterpart refusal by the Republican leadership to tolerate that result. So the Republican convention began, with Supreme Court Justice Charles Evans Hughes of New York in the lead as a moderate progressive who at least had no role in the preceding debacle. The regular wing of the party was nevertheless uncomfortable with Hughes, and seven candidates had a consequential vote on the opening ballot, including Elihu Root of New York, Secretary of War under both McKinley and Roosevelt and the temporary convention chair for the Taft forces in 1908. Roosevelt then reentered the process, leveraging negotiations in a surprising fashion. The joint committee suggested Hughes to him as a compromise. Roosevelt rejected Hughes but suggested Henry Cabot Lodge, the far more conservative Senator from Massachusetts. The Progressive convention responded by nominating Roosevelt by acclamation. Roosevelt responded by saying that he could in fact support Hughes if the latter brought policy positions that were acceptable. The next day and in that light, opposition to Hughes within the convention hall essentially disappeared, and he went on to collect 96% of the vote on what was the third ballot. Charles W. Fairbanks of Indiana, vice president under Roosevelt, was then chosen to come back as vice president under Hughes. The Republicans had their ticket, and the Progressive Party effectively disappeared, ultimately acquiring about 0.002% of the total vote, half of which came from Georgia. These new factional differences were still clearly visible at the Republican convention of 1920, though they were less intense than they had been in the two preceding contests and their impact was muddled by strategic complications specific to each faction. On one side, there were two early progressive champions, Major General Leonard Wood of New Hampshire and Senator Hiram Johnson of California, and the fact that neither would cede to the other became a lasting problem.17 On the other side, Governor Frank Lowden of Illinois had been expected to be the consensual champion of the regulars, but he had tested his appeal in the primaries and did badly. He would still attain a narrow lead on the fifth ballot but never be able to expand it. Partly as a result of those earlier inauspicious showings by Lowden, most of the major states, except for Illinois but including New York, Pennsylvania, Ohio, and Massachusetts, simply sat behind favorite sons and awaited developments. These came during an overnight recess after the fifth ballot,18 when Governor Warren G. Harding, the Ohio favorite

136

B. E. SHAFER AND E. M. SAWYER

son, began to attract fresh supporters. Harding moved into third on the seventh ballot, into first on the ninth as Lowden supporters broke heavily in his direction, and went over the top on the tenth. Governor Calvin Coolidge, the Massachusetts favorite son, was then an overwhelming choice to be Harding’s vice president. Coolidge was president by the time of the 1924 Republican convention, Harding having died the preceding August. Moreover, Coolidge had entered a number of demonstration primaries where, unlike Lowden in 1920, he performed well, suggesting that his popularity among party professionals could translate to the voting public. His nomination was thus overwhelming: Coolidge 1065; LaFollette 34; Johnson 10. Former Governor Lowden of Illinois was brought back as vice president, securing a clear majority on the second ballot. Yet Lowden had previously claimed that he would not accept, and when the convention checked, he did in fact refuse. After Idaho Senator William Borah also spurned informal inquiries—first Lowden for the conservatives, then Borah for the progressives—the third ballot produced Charles G. Dawes, inaugural director of the Bureau of the Budget, and the Republicans had their ticket. When President Coolidge announced that he would not stand for reelection, Herbert Hoover of California, Coolidge’s Secretary of Commerce, became the apparent favorite of party leaders for 1928. Hoover blended a previous career in private business with a sterling record of public service, having been head of a sequence of international relief efforts over the previous fifteen years. Like Coolidge, he tested his appeal on the primary trail; like Coolidge, he was successful. An early convention challenge to Hoover delegates from Texas suggested that his forces were in full control, losing only the farm states, which were mired in a severe and widespread agricultural downturn. Those impressions were confirmed on the opening ballot, where Hoover secured 837 votes (77% of the total), with Frank Lowden a distant second at 74, half of which came from Illinois. Vice President Charles Dawes was available for a second term, but Coolidge spoke privately against Dawes, having split strongly with him over the farm crisis—Dawes, more of a progressive, favored greater governmental intervention—and Senator Charles Curtis of Kansas became the effectively unopposed vice presidential choice.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

137

The Long New Deal of the Out Party The Wilderness Years No one at the 1928 Republican convention could have foreseen the scale of the disaster that became the Great Depression by the time of the 1932 gathering. So no one could have predicted the partisan disaster that came with it for the Republican Party, as Democrat Franklin Roosevelt rode his programmatic response to four successive presidential nominations. During the first three, 1932 through 1940, Republican nominating politics more or less ceased to be about melding conservative and progressive factions–a residual effort would be visible only around the edges–and shifted instead to finding a nominee sufficient, if not to unhorse Roosevelt, then at least to mount a credible campaign. To that end, the party would try three different approaches to a counterpunch. Herbert Hoover would stand for reelection in 1932, offering a forceful defense of what had been a far more active government than opponents would acknowledge. Governor Alf Landon of Kansas would succeed him in 1936, mixing a defense of private enterprise with some major progressive proposals plus the actual endorsement of a number of New Deal programs. And Wendell Willkie, attorney in private practice, would change the entire focus in 1940, pulling the Republican Party away from a persistent isolationism as war loomed while mixing his new internationalism with the most clear-cut attack yet on the economic thrust of the New Deal. Hoover was to collect 98% of the vote on the opening ballot at the 1932 Republican convention, despite a generalized pessimism about his reelection. His vice president, Charles Curtis, was renominated with him but much more narrowly, getting only 54% of the vote as the farm states of the midwest and west came out even more strongly against Curtis than they had in 1928. Still mired in a long-running agricultural depression, theses states went with Hanford McNider of Iowa, national commander of the American Legion. So that was a residual progressive residuum of the old factional tension between conservatives and progressives. The conservative version came from New York, with the largest individual delegation, which stayed with Major General James G. Harbord, whose long military service had been followed by the Chairmanship of RCA and who was prepared to offer a stronger return to conservative economics. Governor Alfred M. “Alf” Landon of Kansas, one of the few Republican governors to be reelected after the onset of the Great Depression,

138

B. E. SHAFER AND E. M. SAWYER

was the next choice to carry the presidential banner, in 1936. The draft platform which awaited him focused on an “America in Peril.” But once Landon had collected support from all the states except Wisconsin, he sent a telegram to the assembled delegates calling for a constitutional amendment to provide safe working conditions for women and children, a broad extension of civil service, and support for major parts of the New Deal. Senator William Borah of Idaho had been the nominal champion for progressive forces in the primaries, but he had garnered nearly no delegates, so Landon went with Col. Frank Knox of Illinois, publisher of the Chicago Daily News, who had actually beaten Borah in the Illinois primary. For 1940, one last wilderness strategy emerged from a very different issue environment. The looming policy concern was not domestic economics but international war, where Republican Parties in many states had an established hesitation about rearmament and an ongoing preference for isolationism. So a widespread perception that the world had changed would go a long way toward explaining the surprise presidential nomination of Wendell L. Willkie of Indiana, an outspoken internationalist who was better known for his opposition to public power and his general criticism of Franklin Roosevelt. Despite being an evident insurgent with no prior public office, Willkie did not even begin his campaign until well after the primaries were closed to new entrants. Nevertheless, what he did acquire, in the brave new era of opinion polling, was an explosion in his polling numbers during May and June. His two main opponents were in effect a preview of major party divisions to come, pitting moderate regulars against the conservative activists, this time in the persons of Manhattan District Attorney Thomas E. Dewey of New York for the moderates and Senator Robert A. Taft of Ohio for conservatives. Dewey led on the first ballot (360), followed by Taft (189) and trailed by Willkie (105), along with two favorite sons, Senator Arthur H. Vandenberg of Michigan (76) and Governor Arthur M. James of Pennsylvania (74). Dewey retained a declining lead on the second ballot. Willkie moved ahead of Taft on the third. New York went to Willkie on the fourth ballot as Dewey dropped out. And when Michigan deserted Vandenberg on the sixth ballot, followed quickly by Pennsylvania abandoning James, the nomination of Willkie was assured. Charles W. McNary of Oregon, Minority Leader in the Senate and a supporter of some aspects of the New Deal, became his vice president and completed the ticket.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

139

The Moderate Response Republican nominating politics in 1944 was to be not just different from these desperate predecessors. It was to be the opening realization of a new division that had been only implicit in 1940 but would last into the 1970s. For more than a generation, this division could be simply described. In the belief that the Democrats could be defeated by a moderate Republican, one who did not call for a roll-back of popular programs, the organized parties of the big northeastern industrial states, essential to any defeat of the Democratic candidate, were to line up in support of a succession of just such moderate champions. Though their acceptance of the parameters of the new political world would remain unacceptable, and increasingly so, to Republican Parties that were driven more by issue activists, parties found more reliably in the midwest and west. These struggles were to be reliably settled in favor of the big-state moderates, though never so decisively as to put this new factional cleavage to rest.19 In the first such contest, Thomas E. Dewey, now Governor of New York, returned as the front-runner for the Republican nomination of 1944. Before the convention, he was opposed by Governor John W. Bricker of Ohio for the conservatives and by Harold E. Stassen, former governor of Minnesota, for the remaining progressives. Both withdrew before the first ballot, however, where Dewey received all but one vote. The delegates followed with a unanimous vote for Bricker, a clear-cut isolationist and an ideological conservative, as vice president. So the ticket contained leaders for each major side of what would be a long-running internal division, with the moderate spokesman at its head. The same initial array was to characterize the 1948 convention. Dewey was again the front-runner, but this time his leading opponents did not drop out. Stassen was back to lead the rump of remaining progressives, but it was Senator Robert A. Taft of Ohio, the main Dewey challenger among conservatives in 1940, who came back to carry the more consequential conservative banner. On the first ballot, it was Dewey 434, Taft 224, and Stassen 157, with Senator Arthur Vandenberg of Michigan (62), Governor Earl Warren of California (59), and Governor Dwight H. Green of Illinois (56) as favorite sons. In a recess after the second ballot, the Taft and Stassen forces attempted but failed to form an alternative coalition. Having failed, both candidates then withdrew, leaving Dewey to be unanimously nominated on the third ballot, while confirming Governor Warren as his running mate by acclamation.

140

B. E. SHAFER AND E. M. SAWYER

The closest and most tense of the string of contests from 1944 through 1960 came at the 1952 Republican convention.20 Taft was back to lead party conservatives, while his moderate opponent this time was General Dwight D. Eisenhower, Supreme Commander of the Allied Expeditionary Forces (AEF) during World War II and Supreme Commander of the North Atlantic Treaty Organization (NATO) when he resigned to come home and contest the nomination. There had been some speculation about Eisenhower as a Democratic nominee in 1948, but he came home as a Republican and split a series of head-to-head primaries with Taft. Early roll-calls on credentials and rules went narrowly to Eisenhower, and the first ballot featured Eisenhower 595 and Taft 500, with Warren as a favorite son from California at 81 and Stassen as a favorite son from Minnesota at 20. Eisenhower was thus nine votes short of a majority when 19 of those Minnesota delegates switched their support, making him the nominee. His choice for vice president, Senator Richard M. Nixon of California, was then endorsed by acclamation, though many believed that the real “ticket” was Eisenhower and California Governor Warren, whom the new president put on the U.S. Supreme Court at his first opportunity. In any case, Eisenhower was back seeking renomination in 1956, when the real question was medical rather than strategic, involving his health after a heart attack.21 When the prognosis was judged reassuring, he received 100% of the vote on the first ballot. His vice president, Richard Nixon, had to weather some tentative early maneuvering toward a replacement, but when this came to nothing, he too acquired a unanimous vote. So the question for 1960 had to revolve mainly around the ease or difficulty with which Nixon could succeed to a presidential nomination of his own. On the formal record, he received 99% of the vote, against two incipient opponents who were unwilling even to be officially nominated. In a different year, Governor Nelson A. Rockefeller of New York would have championed the moderates and Senator Barry M. Goldwater of Arizona the conservatives, exaggerating the usual moderate-conservative split in the process–exactly what they would do four years later. But for 1960, a late rejigging of the platform, especially on civil rights and national security, would reassure Rockefeller and infuriate Goldwater, feeding into the explosive conflict that was to come in the next Republican nominating contest.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

141

The Changing Nature of Party Factions, Republican Styles The Activist Insurgency The Republican nominating contest of 1964 was to provide the first triumph for party conservatives over party moderates since this division came onto the agenda in 1940. Yet this had from the start been as much a structural as an ideological conflict, in that it pitted organized parties against volunteer parties, and hence party regulars against issue activists. The main asset of the regulars remained control of party office itself, where their motivating goal was to win elections. The main asset of the activists remained their volunteer energy, where clear-cut programmatic positions were paramount for energizing additional activists.22 Though the sequence of defeats that the Republicans had endured for a generation also played into the gradual alteration of the balance between the two sides, as activist conservatives became more and more frustrated by the repeated electoral failures of organization moderates. Seen from above, those defeats had deprived the party of federal appointive powers and of the ability to designate targeted capitol projects. Seen from below, they meant that the party needed to rely more and more on individuals who were motivated by ideological goals rather than concrete rewards. The initial triumph of the activists in 1964 would nevertheless be tempered by the subsequent two-term Nixon presidency, where Richard Nixon brought strong ties to both sides. So it would be the contest following the Nixon resignation that brought this factional division between regulars and activists back to the center of Republican–and not just Democratic–politics. Though even then, there was a major irony: a sitting president based in the regular party would win that particular struggle, but an activist icon would return to secure both the nomination and the election four years later. Yet at the beginning of this chain of factional events, Nixon as the defeated nominee of 1960 was no longer present for the Republican contest of 1964. Instead, Barry Goldwater and Nelson Rockefeller, shadow challengers to Nixon in the previous contest, returned to seek the nomination explicitly.23 Their struggle was initially unsettled when Henry Cabot Lodge of Massachusetts, the former American ambassador to the United Nations, upset both men in the New Hampshire primary. Thereafter, Lodge, Goldwater, Rockefeller, and occasionally Governor

142

B. E. SHAFER AND E. M. SAWYER

William W. Scranton of Pennsylvania would win intermittent contests, many of which were still not strongly linked to real delegate outcomes, a fact that itself was going to matter. Near the end, Goldwater would defeat Rockefeller head-to-head in the giant, winner-take-all, California primary, though his ultimate victory owed as much to the successful but decentralized pursuit of delegate status by conservative activists, not least in some of those states where someone other than Goldwater nominally “won” the primary. Yet with 67% of the delegate vote, Goldwater triumphed easily on the opening ballot of the Republican convention, and subsequent attempts to moderate his platform fell short by roughly the same margin. Goldwater in 1964 was to pull the worst Republican vote since Alf Landon in 1936, and with that as context, many party leaders sought a successor who was evidently more moderate, while not being an affront to conservative activists. The available man proved once again to be Richard Nixon, the defeated candidate of 1960 who ran more or less unchallenged in the 1968 primaries, though Rockefeller did beat him as a write-in in Massachusetts and the Nixon forces did feel the need to leave the huge winner-take-all primary in California to Governor Ronald W. Reagan. Nixon was nevertheless to win on the first ballot: Nixon 692, Rockefeller 277, Reagan 182, nine others 182. The nominee then took Governor Spiro T. Agnew of Maryland as his vice president, another organizedparty leader with rhetorical skill at satisfying conservative activists, and the Republicans had their ticket. Nixon would be re-nominated without effective opposition in 1972, when he got all but one vote on the opening ballot of the convention (for an anti-war challenger, congressman Paul N. McCloskey of California). In turn, Spiro Agnew got all but two abstentions in the ballot for vice president. Yet the sequential resignations of both Agnew and Nixon himself created a very different contest for 1976. The champion of the regulars was congressman Gerald R. Ford of Michigan, now holding office as the first unelected president in American history. But this time, the champion of the activists, Ronald Reagan, would not wait until late in the contest as he had in 1968, but instead challenged Ford from the start. And this time, the Republican primaries were institutionally different, being widely reformed to link presidential preferences directly to delegate selection.24

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

143

A narrow win by Ford in New Hampshire leveraged a sequence of wins in northeastern and midwestern contests. Reagan managed a win in North Carolina, then a bigger win in the Texas caucuses plus a series of lesser southern primaries. Ford would resume winning as the contest shifted back to the northeast and midwest, while Reagan would ordinarily be the winner in the west and remaining south. Reagan would take the huge California contest on the final primary day, but Ford would balance that with wins in both New Jersey and Ohio. At the convention, the Reagan forces countered a close but apparently adverse balance of delegates with a proposed reform that would have forced the aspiring nominees to reveal their vice-presidents in advance. When this amendment failed narrowly, 1180 to 1069, Ford went on to defeat Reagan for nomination the next day by essentially the same margin, 1187 to 1070. When Reagan turned away some informal inquiries about the vice presidential slot—potentially the ultimate regular/activist balance—Ford went with Senator Robert J. “Bob” Dole of Kansas. In turn, the magic moment for Reagan, who had run for the nomination with escalating degrees of seriousness in 1968 and 1976, arrived with the Republican nominating contest of 1980.25 This time, Reagan opened with a solid win in the New Hampshire primary and would not be seriously threatened thereafter. George H.W. Bush, former congressman from Texas and the occupant of a long line of major executive posts, came closest to being a real challenger, winning six scattered primaries along the way and giving the regulars a voice. But Reagan won the other twentynine and had an overwhelming delegate lead entering the convention. A residual bit of ticket-making drama raised the possibility that Gerald Ford, turned down by Reagan in 1976, might return as his vice president in 1980. Yet that passed in an evening, and Reagan turned to the more moderate Bush as his vice presidential choice, symbolically linking activists and regulars but with the former clearly on top this time. A New Incarnation of an Old Factional Conflict These developments acquired a special intensity in one geographic region during these years. Republican presidential contenders had picked off some southern electoral votes with Herbert Hoover in the late 1920s and again with Dwight Eisenhower in the 1950s, while George Wallace took some away from both parties in the 1960s. But none of these

144

B. E. SHAFER AND E. M. SAWYER

losses signified a gain for Republican parties at the state level. Instead, from the 1870s into the 1970s, southern Republican parties were essentially skeletal structures. Rather than sustaining themselves with statebased resources, they depended on federal resources transferred back to the states. So their delegates to national party conventions were often bargaining chips for out-state parties, most often in support of sitting presidents, who had the greatest influence over federal resources. That became strikingly different from the 1970s, as the south acquired active Republican parties with a continuing organizational existence, that is, parties which existed to contest state and local public offices and not just to funnel outside patronage. Yet along the way, the transition in these southern regional branches reflected national change in a particularly stereotypical fashion: they went from being formalistically “regular,” in the sense of being office holding skeletons, to being disproportionately activist-based, the regional incarnation of networks of issue activists that were growing nationwide. So southern Republican Parties became, more or less by definition, weighted toward what looked more like activist than like regular party branches outside the south. Uncontested renominations are often as much coronations as conventions, and the 1984 Republican version fit the pattern well, with Reagan and Bush going so far as to collect all the votes cast by way of a joint roll-call. But now, it could not be surprising that a conservative nominee brought a conservative platform with him, nor that the vice presidential (re)nominee went out of his way to emphasize his own, integral, conservative place within the administration being celebrated. On the other hand, only hindsight can convey a sense of how far the insurgency of the conservative activists had come. For in fact, the next four nominating contests would demonstrate that the Republican Party had a new dominant core, such that the real contest on the way to nomination was essentially an effort to determine which candidate most clearly embodied it. At the same time, the ideological location of this new core had clearly shifted, away from the space occupied by nominees during the moderate hegemony and far enough to its right to satisfy the activists on whom the party increasingly depended.26 George H.W. Bush was to find himself, happily, as the heir-apparent to Reagan in 1988, boasting a lead in major endorsements, financial contributions, and campaign organization to go with this potentially inheritable mantle. The most likely alternative appeared to be Senator Bob Dole of Kansas, vice presidential nominee of 1976 and another man with broad

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

145

party connections.27 Dole did upset the opening Iowa caucuses, but Bush righted himself in New Hampshire. They traded a few tiny wins, until Bush took all sixteen primaries on “Super Tuesday,” most by a two-thirds margin. The active contest being over, Bush strengthened his claim on a party where he had once been the moderate alternative by taking the clearly conservative Senator J. Danforth Quayle of Indiana as his vice president and by fronting a platform that left no doubt of his devotion to the new agenda that characterized new party norms. Four years later, the usual advantages of a sitting president—name recognition, party connection, and media access—were augmented by Bush’s leadership role in construction of the international coalition that had been victorious in the short Persian Gulf War. That combination was enough to deter any major party figures from challenging his renomination. Patrick J. Buchanan, conservative commentator, did stand against him in New Hampshire and did enter a wide array of other contests. Yet Bush was to defeat Buchanan in that first primary and go on to win literally all of the thirty-seven to follow, rendering Buchanan numerically irrelevant well before the end. Ronald Reagan came back to underline his support of the sitting president at the convention, Dan Quayle came back to run for reelection as Bush’s vice president, and the Republicans again had their team without anything resembling a serious challenge. The field for the Republican nominating contest in 1996 was in turn to look remarkably like the field in 1988. There was an experienced candidate with a long career and wide connections, Senator Bob Dole of Kansas, along with an aspirant who wanted that role for himself, Senator Lamar Alexander, Jr., of Tennessee. They were flanked by a strong conservative prioritizing cultural values, via Pat Buchanan in a return engagement, along with a strong conservative prioritizing economics, in Malcolm “Steve” Forbes, Jr., of Forbes Magazine.28 The contest was temporarily unsettled when Dole and Buchanan traded wins in Iowa and New Hampshire. But Dole succeeded in a number of smaller primaries— South Carolina became effectively the show-down state—then won all eight of the lesser primaries on what was known in 1992 as “Junior Tuesday,” followed by all seven on the more familiar (and delegate-rich) “Super Tuesday.” When he followed the next week with wins in Illinois, Michigan, and Ohio, there would be no challengers left standing.

146

B. E. SHAFER AND E. M. SAWYER

The 2000 Republican nominating contest appeared initially to be of a piece with its predecessors, being led by Governor George W. Bush of Texas, a party regular with good conservative credentials. In the apparently reasonable expectation that this contest would follow the trajectory of its immediate predecessors, Bush moved early to solidify his standing and pivot toward the general election by offering a fresh comprehensive program, labeled “compassionate conservatism” and tailored to expand a Republican base that had contracted in the defeat of his father in 1992 and of Senator Dole in 1996. What would prove different for 2000, and jumble these plans to be the early consensus contender, was that his main opponent on the way to a nomination would be neither a cultural nor an economic ideologue but rather a party maverick, Senator John S. McCain, III, of Arizona. In a major additional twist, McCain would augment his status as leading challenger through his attraction for independents and Democrats in those Republican primaries where they could vote, making him the resting place for moderates and forcing Bush to tack back to the right, arguing that he could offer a more genuine version of Republican values. Bush won the Iowa caucuses, skipped by McCain, but McCain won the New Hampshire primary, solidly. They split the six further primaries before “Super Tuesday,” which was to prove decisive. Of the eleven contests on that date, McCain would win only the four in New England, with Bush carrying the other seven, including California, New York, and Ohio. A few days later, McCain suspended his campaign, and Bush became the de facto nominee. Both Bush and his vice president, former Rep. Richard B. Cheney of Wyoming, went on to win all 2066 votes at their convention.

A Factional Structure for the Modern World For the prescient, that would have been grounds for wondering whether the factional structure inside the Republican Party was shifting once again. If so, a test of that possibility would have to wait for 2008. The last serious challenge to the renomination of a sitting Republican president had been from Ronald Reagan to Gerald Ford in 1976. Like most modern others, George Bush and Dick Cheney would be unchallenged in 2004, a status powerfully reinforced by their role in rallying the nation after the terrorist attack on the World Trade Center in 2001. So Bush actually received all but one convention vote, and Cheney was endorsed unanimously.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

147

Nothing about that untroubled renomination would have suggested that Republican factional politics—the structure of nominating politics inside the Republican Party—was to shift after 2004, though in fact it was. The ability to settle relatively early on a nominee sitting at the ideological center of the party would not change much, remaining roughly stable from the 1980s through the 2010s. The dominance of conservative activists, of various stripes, would likewise remain. But the nature of the factional opposition on the way to a resolution would be tweaked in one noteworthy way. What was essentially an ideological split on cultural grounds would prove to be a recurrent trend in nominating politics, buttressed by a social split putting the evangelical Protestants ascendant in Republican politics against those holding conventional economically conservative preferences. This new split was first registered in the larger context for nominating politics that no longer involved 9–11 and a response to it, but rather what became recognized as “the great recession.” Among many Republicans, a major economic downturn under a sitting Republican president would add to the attractions of John McCain, party maverick, who would return as a presidential aspirant, begin as the front-runner, falter early on, but end up as the nominee. Yet key trials for McCain along the way would begin to register the rise of an altered set of intraparty divisions, divisions that could not as easily be suppressed or dismissed as they had been in the 1980s and 1990s. This effect would be clear but modest in 2008. It would become less disputable in 2012. And it would survive but be additionally tweaked in 2016. For 2008, having consciously mended his fences with President Bush, McCain was seen as the effective front-runner for the nomination, an impression that dissipated when his campaign imploded during late 2007, losing staff, finances, and public standing. In his place, former Mayor Rudolph W. Giuliani of New York City became the new leader in preference polls, though his rise would bedeviled by the fact that his national poll standing was only weakly reflected in the states with opening delegate contests, namely Iowa and New Hampshire. The opening Iowa caucuses, however, would be upset by a third figure, Michael D. Huckabee, former Governor of Arkansas. After Iowa, Huckabee became the clear champion of social conservatives, and his staying power would be the first indicator that raising a recentered party banner might have become more difficult than it was previously.

148

B. E. SHAFER AND E. M. SAWYER

McCain righted his campaign with a win in the New Hampshire primary, but lost to Mitt Romney, a classic champion for the remaining regular party, a week later in Michigan, before picking off South Carolina. Florida, the last contest before “Super Tuesday” and the one intended by the Giuliani forces as their launching pad, was instead won by McCain, followed by Romney. And “Super Tuesday” proved to be the practical end in numerical terms, with McCain winning ten of fifteen contests and an even larger share of their delegates. Only Huckabee would continue on, in a lesser struggle that looked like a sideshow in 2008, but was actually indicative of the factional pattern that would extend into at least the next two nominating contests. Huckabee went on to score victories in a number of caucus states plus southern and western primaries. Though by convention time, even he would have endorsed McCain, who gathered 98% of the delegate vote. His vice president, Governor Sarah L. Palin of Alaska, was then a nod to the former Huckabee constituency. Another open contest for the Republican presidential nomination four years later was distinguished first by the sheer number of aspiring presidential nominees and then by the temporal extension of intraparty debates among them, though both of these aspects would go quickly from being a deviation to a harbinger. The contest within these multiple introductory debates would be distinguished by wild gyrations in support for a sequence of nearly unknown contenders, surging, and then imploding29 : congresswoman Michele Bachmann of Minnesota, Governor Rick Perry of Texas, businessman Herman Cain of Georgia, and former Speaker Newt Gingrich of Georgia. Yet as the real contests for delegate selection approached, this chaotic pre-story resumed a pattern familiar from— indeed almost identical to—the one from 2008. Mitt Romney would play the role of John McCain this time, as the only candidate who escaped the surge-and-decline dynamic. This might still have implied only an ongoing second place, but Romney also had a consistent lead in endorsements, finances, and media attention. Senator Rick Santorum of Pennsylvania was to inherit the part of Mike Huckabee, winning the Iowa caucuses, carrying the second-largest number of states, and surviving the longest against the Romney bandwagon. Romney would ultimately win nearly three-quarters of the states, including all those with more than ten Electoral votes. Conversely, Santorum would be advantaged by low turnout and intense supporters in caucus states generally, just as he would be advantaged in primaries in the south, the

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

149

agricultural midwest, and the mountain states, where there was a disproportionate presence of cultural conservatives or most commonly rooted among evangelical Protestants. In the end, Santorum in 2008 outperformed Huckabee in 2004, but there were simply not enough of these favorable states to make him a threat to the Romney nomination. In formalistic terms, the Republican nominating contest of 2016 would begin with another, even larger crop of announced candidates, participating in an even longer sequence of introductory debates.30 But rather than a surge-and-decline dynamic, this pre-contest was characterized by the rise of Donald Trump of New York, property developer and reality TV star, to a clear lead in preference polls by the end of 2015. And when the real contest arrived, Trump was to prove so dominant as to marginalize many of the considerations that had characterized the Republican nominating struggles of 2008 and 2012. It took a bit longer to determine that the roles of Huckabee in 2004 and Santorum in 2008 would be played this time by Senator Ted Cruz of Texas. Cruz did narrowly beat Trump in the Iowa caucuses, though Trump righted himself in the New Hampshire primary, crushing all of his main plausible opponents. Thereafter, Trump would go on to win sixteen of the next nineteen primaries, and thirty-three of a grand total of thirty-eight, nearly rubbing out further distinctions. Like Huckabee and Santorum, Cruz did better—quite well, actually—in caucus states, and he did have his best performances in primary states with culturally conservative and/or evangelical electorates. Yet this impact was smaller than usual in the context of losing thirty-four of thirty-eight primaries, while Trump partially but consciously retailored himself as a cultural conservative, an identification he solidified by taking Governor Michael R. Pence of Indiana as his running mate.

Factional Struggles and Candidate Fortunes Revisited Factions Over Candidates In Chapter 3, a chronicle of presidential nominations in the Democratic Party suggested that the central outcomes of nominating politics, namely the presidential nominees themselves, were not the products of their individual backgrounds, their individual resources, their individual strategies, or their individual skills. Rather, these human outcomes were products

150

B. E. SHAFER AND E. M. SAWYER

of the patterning within and the struggles among major party factions at any given point in time. In earlier days, these factional struggles were expressed through bargaining among formal leaders of the official party. In the modern world, they are instead expressed through voting constituencies in presidential primaries. In both cases, however, the central point is that less ambitious, less experienced, and less abstractly attractive candidates who met these factional and constituency requirements were entitled to succeed, while more ambitious, more experienced, and more attractive candidates who did not meet those requirements should be expected to fail. A chronicle of presidential nominations in the Republican Party and associated predecessors now tells the same tale and makes the same points, only more so. To begin with, there really is no Republican counterpart to Andrew Jackson, a Democratic presidential contender who created a political party with a lineage that runs down to the current moment. That certainly qualifies Jackson as a figure who, in his person, overcame the previous dynamics of nominating politics. For the Whigs, the counterpart figure was Henry Clay, the leading congressional figure of the antebellum years and a man who hoped to go on and play this Jacksonian role. Clay did become the main architect of the Whig program, yet his personal successes as a presidential aspirant were spotty at best, and the party he nurtured was effectively defunct by 1856. Many presidential nominees thereafter would grapple with important national policy matters. Some would even be important to bringing these policy conflicts into government, though most major conflicts would be driven by a much wider range of influences and actors. But if the focus is the politics of presidential nomination, there are two key questions that stand out, and that constitute the ultimate test of individuals versus factions as shaping forces for nominating politics. First, did any given nominee alter, or at least rechannel, the bandwagon dynamic that has characterized (and explained) this politics? And second, did this nominee go on to reshape, rather than just reflect, the factional struggles that would produce individual nominees within this altered bandwagon dynamic? So if those are the measure of a genuinely personal and idiosyncratic redirection of nominating dynamics, then Clay, rather than taking his party to a new era, lived only to see it destroyed. Because the Republican Party then became the lasting opposition that Clay had hoped to create, this successor party demands the same attention. Yet Abraham Lincoln, revered to this day as its founding figure,

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

151

entered a nominating process that had in effect already been institutionalized by the Democrats, while his particular nomination conformed to an already-familiar bandwagon dynamic, one that has since characterized both parties. (See Table 1.2 of Chapter 1) Lincoln was an important figure in the creation of the Republican Party, and the internal character of the resulting vehicle had a great deal to do with the coming of the Civil War. Yet even a conflict on that scale did not obviously change the nominating dynamics of an underlying bandwagon, while it was the factional structure created by the war itself that would characterize Republican nominating politics for a long time thereafter, and whose evolution would be central to the products of this nominating politics. This leaves only a short list of other nominees who accomplished anything beyond the normal electoral maximum–two terms as president– or beyond the ultimate nominating ambition, namely a second chance to run for president as the nominee of a major party. Indeed, within the resulting list, the only nominee who could pass the first test, that of being elected more than twice, was Franklin Roosevelt, Democratic nominee in 1932, 1936, 1940, and 1944. Roosevelt did go on to produce the only institutional reform in American party history that actually rechanneled the bandwagon dynamic, when he supported abolition of the two-thirds requirement for Democratic nominations in 1936 (Chapter 1). And Roosevelt could also receive some credit for reshaping the factional structure of the Democratic Party, moving it from an old world of regional divisions overlaid by an urban–rural split to a new one of class interests overlaid by conflicts between organized versus volunteer political parties. Yet even Roosevelt came into both roles through an extremely orthodox nominating politics. His nomination had all the hallmarks of a classic bandwagon, including initial support of the two-thirds rule in order to attract the southern Democratic faction along with the selection of one of his two main rivals as his vice president, John Nance Garner of Texas, the Speaker of the House, in order to acquire the necessary delegates to reach two-thirds. And while a major factional alteration would indeed occur on his watch, it was more of a reaction to, than a creation of, a much broader range of stimuli and supports. In other words, if the Great Depression and the New Deal did reshape the Democratic Party internally, it this new factional structure, and not subsequent Democratic nominees, that would institutionalize the struggles set off by the coming

152

B. E. SHAFER AND E. M. SAWYER

of the welfare state and carry them forward. While the revised Republican factions that sprang up on the other side of the partisan aisle were basically reacting to Democratic developments. The list of those with some potential to alter the bandwagon dynamic under the second standard, that of being nominated more than twice, is even less promising. Indeed, this list is more of testament to the power of factional demands over candidate characteristics and candidate strategies: • Grover Cleveland was indeed nominated three times, in 1884, 1888, and 1892, but he serves as the perfect embodiment of one type of nominee, whose selection was explicitly dictated by the structural demands of the party factions of his time. A reform governor from New York proved to be the only political type who could unite a party of northeastern immigrants, western populists, and southern regionalists, where a favorite from any of these individual factions was unlikely to be nominated, impossible to elect. • William Jennings Bryan was nominated three times, in 1896, 1900, and 1908, but he serves as the embodiment of a very different archetype, by satisfying a dominant party faction at the cost of ignoring the demands of a general election. Beaten solidly for election in 1896 after being confirmed at the convention as the favorite of Democratic populists, Bryan managed to be renominated more easily in 1900 only to go on and lose the election even more solidly. He then returned after a one-election hiatus to be renominated with equal ease before going on to extend his margin of electoral defeat one last time. • Richard Nixon–often the forgotten man in these regards–was the last to be nominated three times, in 1960, 1968, and 1972, though rather than disrupting an institutionalized bandwagon or reconstructing his party factions, he stands out as the ideal “available man.” Which is to say: in all three of his nominations, his strength lay in being an acceptable compromise for the factional balance inside the Republican Party where, if he had championed one major faction or challenged another, he was unlikely to have been nominated at all. At the other extreme, the Republican Party was fully the equal of the Democrats in creating nominees who were almost pure products of factional interactions, bringing little or no personal input to their own

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

153

nominating politics. It was the Democratic Party that inaugurated the notion of such “dark horses,” first with James K. Polk, then with Franklin Pierce. Yet the Republicans generated a counterpart cluster after the Civil War, with Rutherford B. Hayes, James A. Garfield, and Benjamin Harrison. All five were essential compromises emerging in the course of a national party convention, once the established factions could not secure the nomination for one of their own. Moreover, all five were to be oneterm presidents—Polk and Hayes by choice, Garfield via assassination, Pierce from rejection by his own party, and Harrison through electoral defeat. Factions in Retreat Where the Republican Party did produce a set of nominees who were noteworthy and idiosyncratic, a cohort that was essentially absent among Democrats,31 was in the small cluster whose members, while discovered and ultimately delivered by existing party factions, had no prior connection to public office. The first of these was the great western explorer, John C. Fremont, at the inaugural Republican convention of 1856. His lone brush with public office had been a California senatorship which he abandoned shortly after election when it proved uninteresting. In a later era, Wendell Willkie would come truly off the unknown list, with no prior public office, no prior party position, and not even a primary election campaign in 1940, the year of his nomination. Though by those standards, the most impressive remained Donald Trump in 2016. A candidate who was not entered in early preference polls and who, if he was known to some Republican Party officials, was treated by them as an annoying curiosity, Trump rose to the top of those polls by the end of 2015 and moved on to a rapid nomination once actual contests for delegate selection began. In the process, Trump kept open the theoretical possibility that he might prove to have altered the bandwagon dynamic in a fundamental way. Or at least, he was the lone total outsider to be elected and hence, by definition, the only one who could be reelected. Yet win or lose in the general election of 2020, the truth about any lasting contribution has to await the arrival of the 2024 contests for any actual evidence of change that was not idiosyncratic to the candidate. In the meantime, there was another set of nominating contests where the influence of party factions and factional struggles were at their least impactful. These are the election years when consensual crises imposed

154

B. E. SHAFER AND E. M. SAWYER

universal demands for a policy response, effectively the same demands on both parties though they could still respond differently. And the two that rank as historically indisputable are the Civil War and the Great Depression. Moreover, while their immediate foci were the nominating contests of 1860 and of 1932 respectively, each would spread the impact of this shared stimulus over four or more presidential elections: the two of Abraham Lincoln in 1860 and 1864 plus the two of Ulysses S. Grant in 1868 and 1872, along with the four of Franklin Roosevelt in 1932, 1936, 1940, and 1944 plus the reelection of Harry Truman in 1948. Both parties had to have a response to both crises. What differed crucially was the identity of the beneficiary and the victim. In each case, the party in power was rejected; in each case, the party that came to power responded; and in each case, the party out of power found itself in a kind of electoral wilderness afterward. Thus the three elections of 1864, 1868, and 1872 were a low point of presidential nominations for the Democratic Party, in the same fashion that the three elections of 1936, 1940, and 1944 were low points for the Republicans. In both cases, the strategic situation was sufficiently dire that the response by the party out of power was not so much a product of factional struggles as just the pursuit of three wildly varied experiments with nominees and programs, experiments that would not have been characteristic before the crisis and would not be characteristic afterward. A weaker version of the same pattern, driven less by immediate crisis and more by a shared desire to fend one off, prevailed in the years leading up to the Civil War. Both major parties faced the challenge of retaining the ability to act in a coordinated fashion in the face of rising national divisions over slavery and abolition. For a fully established party whose leaders were consciously concerned with being a national presence, Democratic fortunes in particular turned on the ability to manage their factional struggles. The Whigs faced the same imperative, but it became clear early that they could not manage the necessary straddle, and they disappeared. While the Republicans, not even seeking to have branch parties in the south, were free to follow a more uncompromising policy, with little concern for anything the Democrats might be doing. At first, this looked like nothing so much as an opportunity to replicate the failures of the Liberty and then the Free Soil parties. Though when the Democrats ultimately faltered, dramatically at their 1860 convention, these Republicans would come to power.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

155

Yet the final point from this return to the changing factions and constituencies of the two parties remains how hugely varied was the resulting mosaic of idiosyncratic personal characteristics that could coexist comfortably with a changing structure of party factions, where the latter more or less dominated and selected among the former, despite personal virtues, vices, and idiosyncrasies. Candidates who did not fit the structural demands of one or the other party at a given point in time were simply not nominated. Candidates who were nominated became crucial evidence for the influence of the factions that make up that internal structure. And candidates who succeed but do not appear to fit this structure–if any– become evidence either that the analyst is mistaken about an underlying factional structure, or that this structure has shifted yet again.

Notes 1. Richard P. McCormick, The Presidential Game: The Origins of American Presidential Politics (New York: Oxford University Press, 1982), and M.J. Heale, The Presidential Quest: Candidates and Images in American Political Culture, 1787–1852 (New York: Longman, 1982). 2. Michael F. Holt, The Rise and Fall of the American Whig Party: Jacksonian Politics and the Onset of the Civil War (New York: Oxford University Press, 1999). 3. The second stage of the Whig strategy was to run Francis Granger of New York, Postmaster General under Thomas Jefferson on a vice presidential ticket with Harrison and Webster, along with Senator John Tyler of Virginia, who resigned from the Democratic Party just in time to stand as a Whig vice president. It was Francis Granger, however, who gained the distinction of being the only candidate ever to lose the vice presidency in the Senate, in a conditional election when the Electoral Vote for Vice President was tied. 4. Clay was to lose yet again, to the dark horse Democrat James K. Polk, who completed one of Tyler’s major policy initiatives, the annexation and integration of Texas. 5. Daniel Walker Howe, The Political Culture of the American Whigs (Chicago: University of Chicago Press, 1979). 6. James C. Klotter, Henry Clay: The Man Who Would Be President (New York: Oxford University Press, 2018). 7. Indeed, pulling it all together, he had come to see his “American system” as a device not just for national unity on its own terms, but as a program to unite north and south—and forestall the burgeoning conflict between them. That too was clearly failing within the party that was a long-run project for him, as he too expired.

156

B. E. SHAFER AND E. M. SAWYER

8. Michael F. Holt, The Political Crisis of the 1850s (New York: W. W. Norton, 1978). 9. William E. Gienapp, The Origins of the Republican Party, 1852–1856 (New York: Oxford University Press, 1987). 10. The Liberty Party had secured 0.3% of the national vote in 1840 and 2.3% in 1844. The Free Soilers had secured a full 10.1% in 1848, driving the Democrats into third in Massachusetts, New York, and Vermont, though much of this was owed to its presidential nominee, former President Martin Van Buren. Without him, it fell back to 4.9% in 1852. 11. In fact, Johnson had hoped to be a compromise Democratic nominee for president in 1860, and did run 2nd on the opening ballot of the Democratic convention in 1868. 12. Daniel DiSalvo, Engines of Change: Party Factions in American Politics, 1868–2010 (New York: Oxford University Press, 2012) is the great source on political factions generally but is especially helpful with the post-Civil War story. 13. Charles W. Calhoun, Minority Victory: Gilded Age Politics and the Front Porch Campaign of 1888 (Lawrence: University Press of Kansas, 2008). 14. R. Hal Williams, Realigning America: McKinley, Bryan, and the Remarkable Election of 1896 (Lawrence: University Press of Kansas, 2010). 15. For continuing factional tensions of the time, see Richard L. McCormick, “Political Change in the Progressive Era,” Party 3 in McCormick, The Party Period and Public Policy: American Politics from the Age of Jackson to the Progressive Era (New York: Oxford University Press, 1986). 16. Lewis L. Gould, Four Hats in the Ring: The 1912 Election and the Birth of Modern American Politics (Lawrence: University Press of Kansas, 2008). 17. Exacerbated by the intransigence of Wisconsin and Senator Robert H. LaFollette, who stayed off on their own from beginning to end. 18. The famous, if possibly apocryphal, “smoke-filled room.” 19. Gary W. Reichard, Politics as Usual: The Age of Truman and Eisenhower (Arlington Heights: Harlan Davidson, 1988), and David W. Reinhardt, The Republican Right Since 1945 (Lexington: University Press of Kentucky, 1983). 20. Paul T. David, Ralph M. Goldman, and Richard C. Bain, The Politics of National Party Conventions (Washington, DC: Brookings Institution, 1960). 21. Clarence G. Lasby, Eisenhower’s Heart Attack: How Ike Beat Heart Disease and Held on to the Presidency (Lawrence: University Press of Kansas, 1997). 22. The Democrats would have their own version of this grand conflict, but it actually arrived earlier in Republican ranks. 23. White, The Making of the President 1964.

4

FACTIONS, CONSTITUENCIES, AND CANDIDATES: THE REPUBLICANS

157

24. See the analyses of the 1976 nominating contest in Larry M. Bartels, “Momentum Triumphant: The Case of Jimmy Carter,” Chapter 8 in Bartels, Presidential Primaries and the Dynamics of Public Choice (Princeton: Princeton University Press, 1988). 25. Andrew W. Busch, Reagan’s Victory: The Presidential Election of 1980 and the Rise of the Right (Lawrence: University Press of Kansas, 2005). 26. W. Elliott Brownlee and Hugh Davis Graham, eds., The Reagan Presidency: Pragmatic Conservatism and Its Legacies (Lawrence: University Press of Kansas, 2003). 27. The leading economic conservative was Rep. Jack F. Kemp of New York, the leading cultural conservative was the Rev. Marion G. “Pat” Robertson of Virginia, but while those were divisions that would matter in subsequent years, they were largely blotted out by Bush and Dole in 1988. 28. This ignores Senator William Philip “Phil” Gramm of Texas, a phenomenal fund-raiser who crashed out of the race before it really began, after using his resources to publicize goals and expectations in popular standing and delegate support which he could not meet. 29. Well described as “discovery, scrutiny, and decline” in John Sides and Lynn Vavreck, The Gamble: Choice and Chance in the 2012 Presidential Election (Chicago: University of Chicago Press, 2013). 30. John Sides, Michael Tesler, and Lynn Vavreck, Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America (Princeton: Princeton University Press, 2018). 31. A partial exception was the nomination of Horatio Seymour, who was already the nominee of the short-level Liberal Republican Party—and a Democratic critic to boot—but who was simply rubber-stamped by the 1872 Democrats, in their shortest national convention ever.

CHAPTER 5

The Usual Suspects: Rules for the Nominating Contest

There is a governing dynamic to the politics of presidential nomination in the United States, one that has run regularly and relentlessly since nominating politics moved outside of legislatures and inside political parties. The essence of this dynamic involves mutual adjustments among all the major players to a sequence of political tests—once convention ballots, now primary elections—until the point where the succession of these responses produces a presidential nominee. There is a further structuring to the fields of contenders who enter these contests and whose fortunes are shaped by that governing dynamic, a structuring characterized by the occupational backgrounds and career trajectories that are privileged at various points in time. These are in turn shaped most centrally by the institutional arrangements for delegate selection and presidential nomination, by the internal organization of the parties themselves, and by the electoral balance between the two parties. Yet around these favored backgrounds was always a further set of influences that were critical to making a particular executive, congressman, governor, senator, or occasional general into the presidential nominee of a major political party. Party factions were crucial to this further narrowing of the options for a presidential nomination within both parties, though they usually differed substantially as between the two. So in the end, it was the interaction of those three grand influences—a dominant nominating dynamic, a diagnostic seedbed for successful nominees, and a © The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_5

159

160

B. E. SHAFER AND E. M. SAWYER

factional context for nominating politics itself—that did the actual work of choosing individual winners, whatever aspirants, strategists, pundits, or reporters may have believed. But what about all those other factors, so beloved by professional observers, so fascinating (at least fleetingly) to a public audience, and so central to the livelihood of campaign strategists? Especially the changing rules of the game, the various strategies applied to them, and the sequence by which the interaction of rules and strategies played out. These are the questions that frame this chapter on the rules of the nominating contest, and Chapter 6 on the strategies and sequences that appeared within them. Both chapters proceed on two tracks. On track one, the analysis must search for specific aspects of rules, strategies, or sequences that could in principle have played a consequential role inside those grand factors that did recurrently shape presidential nominations. On track two, the analysis must attempt to place all such impacts within a funnel of causality, leading in this case from the wider society to a specific nominee while contributing a useful tool for judging how consequential or trivial these other factors actually were. Accordingly, this chapter begins with the rules governing the nominating contest. These rules, or at least any with significant scope, inevitably levy differential costs and advantages on specific candidacies. Individual candidates, after all, come with differing ties to the political parties; they come with differing attractions for various social groups; and they come with differing policy attachments, to be variously harvested or escaped. Yet many of the major rules governing the nominating struggle were developed on the fly, in the process of creating and then institutionalizing the political parties themselves. And most of the rest were developed in response to a particular problem at a peculiar point in time, or sometimes afterward as a side payment to those who had lost the struggle over that particular problem. The point here is that both types tend to live on after their creation, with the ability in principle to shape subsequent nominating politics. Many of these rules were to differ between the two major parties and within each of them across time, though in most cases such distinctions proved to be relatively minor, however heated the struggles over their implementation may have been. For example, each party needed a procedure for certifying the credentials of convention delegates, most especially when they were under challenge from aspiring replacements. Likewise,

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

161

each party needed a procedure for promulgating and enforcing the regulations governing the convention itself as it went about its business. And lastly, each party needed a procedure for publicizing the program of the party, even if individual nominees would vary greatly in their attachment to it. Beyond those operational prerequisites, however, there were two institutional rules that were (and are) formally fundamental to the creation of each convention, much more than the procedures for certifying credentials, promulgating rules, or proclaiming a platform. So these become the two that offer the greatest scope for an autonomous rule-based impact on presidential nominations. Moreover, they have indeed changed over time, offering hope that their impact can be measured concretely: • First is the formula for apportioning delegates to the national party convention among its component state parties. The very first conventions came close to accepting anyone who turned up as a convention delegate. But in short order, the number of delegates— or more precisely, the number of votes that could be cast toward a presidential nomination—had to be stipulated in detail, to prevent geographic proximity from generating political chicanery. • And the other formally fundamental rule is the formula by which state parties in turn allocate these delegates to individual candidates. This allocation process too was recognized early as having intrinsic implications not just for the specific outcomes of nominating politics but also for the continuing fortunes of major-party actors. So the possibility that various arrangements would generate unintended consequences led to the creation of standards for what was and was not acceptable in this allocation process.

The Politics of Delegate Apportionment Almost from the start, apportionment of national convention delegates to state political parties required some explicit metric. Conventions that were little more than an inaugural rally for a proto-party could afford to leave this formula to fate, but even second conventions, suggesting that the new party might achieve continuing political impacts, were rarely that cavalier. On the other hand, for both the Democrats and the Whigs, the proper standard seemed intuitive, namely state size in the Electoral

162

B. E. SHAFER AND E. M. SAWYER

College. Early conventions might still draw hugely varying numbers of observers from the individual states.1 Yet the delegation votes that could be cast for a presidential nominee were intended to be proportionate to state influence in the Electoral College. Later, there would be disputes over what state size itself should mean: size in the Electoral College or actual population? Later still, there would be arguments over whether state size, however measured, captured the support and interests that the party was trying to foster: raw size, however measured, or some mix with actual party support? But in the beginning, Electoral College size was what seemed intuitive. The question of whether the Electors themselves were to be chosen by the state legislature or on a public ballot had been central to the Jacksonian effort to create the Democratic Party. Direct election had then been central to a Jacksonian presidency. Yet the Electoral College as an institution remained at the center of presidential politicking, to the point where the Whig strategy against Democratic nominee Martin Van Buren in 1840 was constructed not around a single candidate but around three separate Whig politicians, in an effort to collect the Electoral Votes of states where one or another was disproportionately popular. In the years before the Civil War, this Electoral College standard was powerfully reinforced by the determination of the Democrats and the aspiration of the Whigs to be truly national parties, that is, competitive everywhere.2 The Electoral College standard was to survive unadulterated into the twentieth century. Yet incipient strains that would force its adjustment in the very longest run were already present by the end of the Civil War. The possibility of building a nationally competitive party effectively died with that war, or rather, with the postwar appearance of a huge, one-party, Democratic south, bringing internal strategic difficulties for both parties. For the Republicans, the question was how much representation to give states that compulsively and overwhelmingly supported the opposition, states that they could never carry. For the Democrats, the problem was instead that a southern Democrat could not hope for victory in the general election,3 making a southern nomination pointless, while at the same time it remained difficult to assemble a nominating majority without southern support. The postwar addition of new western states with very small populations made Electoral College size implicitly discordant in a different way. These new western states would be strikingly over-represented in national party conventions under a simple Electoral College standard. Yet in this case,

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

163

practical politics inside the Republican Party actually came to the rescue of an established rule. In this, new western states entered the Union as reliably Republican, while the two new Senators that each brought with them helped to keep the Senate in Republican hands in an era when the Democrats were returning to competitive status nationwide. So as long as the big organized parties of the Northeast could dominate national party conventions anyway, they had little incentive to injure lightly populated but strategically valuable western counterparts by reducing their apportionment.4 For the Republicans, a major change of the apportionment formula, the only one in party history, came in 1916, in the aftermath of the disastrous convention and election of 1912, the one producing an independent presidential candidacy for Theodore Roosevelt while driving the Republican nominee, President William Howard Taft, into third place in the popular and Electoral votes. The question of the proper representation of southern Republican Parties, implicit in the very creation of a national Republican Party, had lain largely dormant since the Civil War. In effect, delegations from these parties served either as supports for the major organized parties of the northeast, with whom they shared a focus on distributing the spoils of politics, or as supports for the renomination of sitting presidents, who were their lifeline to federal patronage—there being nearly no state-level patronage for them to distribute on their home grounds. The disastrous Republican convention and then election of 1912 changed all this. One of the main arguments of the Roosevelt progressives had been that President Taft could manipulate—and was in effect saved by—these skeletal southern party outposts. Party regulars outside the south who supported Taft in 1912 were not prepared to accept this argument and forfeit the nomination. Yet four years later, they were prepared to placate losing state parties and dissident progressive delegates by moving to a formula based fifty percent on size in the Electoral College and fifty percent on presidential vote for the Republican nominee. The Republicans would continue to tinker with the delegation size of non-states—the territories, variously defined, plus the District of Columbia—but these would be tiny adjustments, and the apportionment reform of 1916 would live comfortably into the twenty-first century.5 For the Democrats, issues of proper delegate apportionment were to lie dormant even longer. In the years following the Civil War, the Democratic Party was focused overwhelmingly on gambits and strategies that

164

B. E. SHAFER AND E. M. SAWYER

might allow it to return to competitive status nationally. By the time of the populist revolt, a division that would eventually lead to reapportionment for the Republicans, the Democratic Party was increasingly able to carry the lightly populated western states whose delegations were disproportionately augmented by an Electoral College standard. While at the same time, a one-party south was untroubled by the soft-money/hard-money divide that deviled both parties elsewhere, such that the region could be relied upon to provide Democratic support at the general election, however “hard” or “soft” its national nominee might be. As a result, the apportionment issue lay dormant until 1936, when conflict over the twothirds requirement for nomination at Democratic national conventions brought it to life more or less accidentally.6 This internal battle over the two-thirds rule pitted programmatic liberals, found largely in the north, against programmatic conservatives, based largely in the south. The large northeastern states resented the implicit veto power that a two-thirds standard granted to the south, and supporters of Franklin Roosevelt urged him to support its abrogation as he came to power in 1932. Roosevelt, however, needed southern support to complete his nomination, and he demurred, albeit for what proved to be only one convention. So in 1936, a century-long tradition finally ended, and the rule was indeed abrogated. Though as a side payment to the south, there was a promise to address the aspect of apportionment most harmful to the region as a whole, namely the absence of any delegate bonus for actually carrying their states in the general election. The 1940 convention, however, tasked with providing this reapportionment, took no action. In response, the 1944 convention promised to add Democratic votes for president to the formula prospectively. But the 1948 convention honored this promise only in the breach, adding two delegates to the totals of each state that had voted Democratic in 1944. Given the preceding Roosevelt landslide, the result was trivial.7 So only in the 1950s, as the civil rights issue became more insistent for internal party politics while uniform southern voting for president could no longer be guaranteed—the solid south actually cracked in 1948 and again in 1952—did the national party truly begin to experiment with greater reapportionment. For 1956, the party added rewards for electing Democratic Governors and Senators, not just presidents, to a formula still taking off from the Electoral College. In 1960, it swept away this jerrybuilt formula, replacing it for one convention with pure population. For 1964, it then

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

165

turned to a mix of size in the Electoral College plus Democratic votes for the presidential ticket, where it would essentially remain. The party would tickle the balance between these factors intermittently thereafter, and the formula was never as neatly mechanical as its Republican counterpart, yet this became the major change in delegate apportionment in the long history of the Democratic Party, one that lives on as this is written.8

The Impact of Apportionment Reform The Republicans That is a chronicle of reform politics in two major parties, ending up in the same place but otherwise originating at different times, driven by different forces, and moving along different trajectories. Yet in the end, did it matter? Which is to say: were any of these apportionment formulae sufficient to alter the dynamic characterizing the politics of presidential nomination in a fashion anywhere near the one rules difference that truly did matter, the one involving requirements for a nominating majority? Did any of these apportionment amendments come close to altering the occupational and career seedbeds of presidential nomination? And did any of them diminish or redirect the main structural influences that produced individual nominees from among that field of presidential aspirants? There are two main elements to any test of the impacts from reapportionment. The first involves simple before-and-after calculations. Were comparative state delegation shares notably different under old versus new rules? If this difference was statistically inconsequential, the possibility of a resulting shift in nominating politics could be said to have fallen at the first hurdle. If it did not, then the second main element of any test involves the recalculation of nominating outcomes themselves, in whole or even in part, under alternative apportionments. This is a far more fraught exercise, since the point of a long-running bandwagon dynamic is precisely the constant and ongoing mutual adjustments that would presumably occur under any formula. Yet because we do know how every state divided its delegates in the actual selection process, it is simple enough to retain this division but apply it to state delegations of the size created by different apportionment formulae. Even then, there were many nominating outcomes, before and after reform in both parties, that could not plausibly be traced to any change in the apportionment of state delegations. These were the ones that settled

166

B. E. SHAFER AND E. M. SAWYER

overwhelmingly on the ultimate nominee at an early point in the nominating process, especially those where an evolving consensus preceded the convention and was merely ratified there. Even among the other outcomes, those that were not consensual, the collective impact of dominant structural influences on nominating politics was always changing, such that a formula conducing in one direction at one point in time could easily conduce in a different direction at another point. Still, a focus on the immediate before-and-after contests associated with delegate reform, plus a side look at a few other contests that turned out to be close, ought to isolate possible impacts from reapportionment, if any, while simultaneously establishing limits on these impacts. The Republican Party moved first with its major change of nominating formulae, and this first test of a reform impact from reapportionment is facilitated by the fact that the “before” and “after” contests of 1912 and 1916 left most state delegations at an effectively identical size, thereby narrowing the focus of any real change. To that end, Table 5.1A offers delegation sizes for 1916 under both the reformed formula and its older (discarded) counterpart. Nine states clearly did suffer major losses as a result of reform, all of them in the old Confederacy. By contrast, Table 5.1 Delegation shares at the Republican convention of 1916: the impact of apportionment reform

A. States losing more than a third of their delegates

AL FL GA LA MS NC SC TX VA

1912

1916

24 12 28 20 20 24 18 40 24

16 8 17 12 12 21 11 26 15

Change (%) −33 −33 −39 −40 −40 −29 −39 −35 −38

B. Regional fortunes after reapportionment

N’East West South

1912 (%)

1916 (%)

Change (%)

45 27 28

49 28 22

+9 +4 −21

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

167

only four others saw any change at all. So the impact of reform was evident, targeted, and substantial. Table 5.1B summarizes the result: the Republican south had been sharply reduced as a presence at Republican conventions, and the first of two tests for real impact was effectively passed. Yet the basic question is still whether this statistical effect from a changing apportionment formula had any practical impact on nominating politics. And there, the answer appears to be no. In fact, the marginal impact that it did have was ironic, running opposite to the rationale for initiating reform. In the Republican nominating contest of 1916, Charles Evans Hughes, Supreme Court Justice and former New York Governor, was to become the candidate of the progressives, opposed by Elihu Root, former Secretary of War, Secretary of State, and Governor of New York, who as a key supporter of ex-President Taft became the main rallying point for party conservatives.9 These two were joined by a bevy of aspiring compromise options, of whom John W. Weeks, Senator from Massachusetts, secured the most votes. Yet the main message of Table 5.2 is that a changed apportionment formula had little effect on the actual candidate outcomes of nominating politics in 1916, while the candidate representing the faction that was expected to be advantaged by the new formula, Justice Hughes, would actually have done better under the old (unreformed) arrangement— an implicit warning that structural reform, even on this statistical scale, might work quite differently within any given bandwagon. On the first ballot, Hughes was the clear front-runner, yet he would have been more so under the old formula. On the second ballot, with Hughes as the rising candidate, the same remained true: leading under both formulae but leading by more under the old arrangement. The third ballot resulted Table 5.2 Candidate votes after reapportionment: the Republican nomination of 1916 A. First ballot

B. Second ballot

Formula

Hughes

Root

Weeks

Hughes

Root

New Old Change

253.5 277.5 +24

103 110 +7

105 120 +15

328.5 363.5 +35

98.5 101.5 +3

168

B. E. SHAFER AND E. M. SAWYER

in actual nomination, by acclamation, after Theodore Roosevelt removed his objections—an action having nothing to do with reapportionment and resulting from something very different, highly personalistic, and hence idiosyncratic. Because all the years to follow would occur under this apportionment formula, the reform itself would soon become an integral but invisible influence on the terrain through which subsequent bandwagons rolled. Moreover, another truly close Republican tally would not even reappear until 1952. Dwight D. Eisenhower, president of Columbia University and former Allied commander, would then lead Robert A. Taft, Senator from Ohio, by a margin of 595-500 on the opening ballot, before a cascade of delegate shifts confirmed the Eisenhower nomination.10 Many surface impacts from apportionment would remain familiar. Two generations later, the impact—the cost—of a change in the apportionment formula was still registered overwhelmingly in the south (Table 5.3A). On the other hand and once again, this actual impact made nearly no difference to the nominating tally, while marginally repeating the original irony: Eisenhower, the type of candidate favored by the reformers, would have been a tiny beneficiary of the old (unreformed) formula (Table 5.3B). By the time there was another close nominating vote at a Republican convention, in 1976, the nature of the Republican vote for president had itself changed so substantially that any previous impact from apportionment reform had been effectively eliminated11 (Table 5.4). Now, the south was the rising Republican region, and growing southern Republicanism obliterated the initial reform impact on what was now a partisan Table 5.3 Delegation shares at the Republican convention of 1952: apportionment reform revisited

A. Regional fortunes after reapportionment

N’East West South

New (%)

Old (%)

Change (%)

47 33 20

45 28 28

+4 +18 −29

B. Candidate votes after reapportionment

New formula Old formula

Eisenhower

Taft

595 601

500 494

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

Table 5.4 Delegation shares at the Republican convention of 1976: the declining impact of apportionment reform

N’East West South

169

New (%)

Old (%)

Change (%)

41 32 27

42 30 27

−2 +7 0

growth sector and not a partisan wasteland propped up by rules. Beyond that, the west was growing modestly in demographic terms, while the northeast was declining a bit more, though it retained the overall regional plurality. So three main points about the impact of apportionment reform stand out within the Republican Party when the available diagnostic contests— 1916, 1952, and 1976—are considered together: • In the first, apportionment reform did indeed affect delegation size, and in the intended fashion, originally but not eternally. Those fighting over alternative formulae were thus not deluded with regard to the direction of its impact. • But second, a simple change in the identities (and attractions) of subsequent candidates could neutralize any such impact, while adding a sting in the tail by making reapportionment work modestly opposite to the intentions of reformers. • And third, by the time there was another plausible test, patterns of party support had changed so substantially that a reformed apportionment formula no longer produced even the same statistical result that had accompanied its inception. The Democrats The Democrats would stay with their original formula for delegate apportionment for a much longer time—essentially 1832 to 1960—before adopting an effectively parallel arrangement to the one that had captured the Republican Party forty years earlier. Late in the day, there would be three transitional formulae for these Democrats during the 1950s, each of which proved to be a one-time experiment before the party settled on a mix of size in the Electoral College plus support for the presidential ticket, the mix that would characterize the party from 1964 through

170

B. E. SHAFER AND E. M. SAWYER

Table 5.5 State delegation shares at the Democratic convention of 1964: the impact of apportionment reform

N’East West South

New (%)

Old (%)

Change (%)

47 28 25

43 30 27

+9 −7 −7

the current moment. The real world would then deny any immediate test of its practical impact, since there were quite literally no recorded votes at the 1964 Democratic National Convention, up to and including the presidential ballot.12 Nevertheless, the immediate statistical impact of apportionment reform remained easy to measure, even in the absence of practical consequences for nominating politics (Table 5.5). In this, the northeast did prosper numerically under reformed rules, as their framers had intended. The south did likewise suffer, though far less than southern Republican counterparts in an earlier period. And the west sustained marginal damage, though continuing population growth would then go on to generate a test of practical rather than just statistical impact at the next available contest, the one where Hubert H. Humphrey, vice president under Lyndon B. Johnson, attempted to move on to the presidency in 1968.13 Statistical effects from reapportionment appeared once again, though this particular real world was already making these look different even just from 1964, another reminder of how much of practical politics is not an epiphenomenon of formal rules. The pattern of presidential voting from 1964 meant that the south would have possessed more delegates as a region under the old formula, while the west would (and now clearly did) possess more delegates under the new (Table 5.6A). Yet a very familiar nonimpact of reform resurfaced in 1968, wiping out any practical consequence from a modest regional redistribution of delegate votes. For, in fact, recalculating the nominating ballot on the basis of old versus revised apportionment formulae was to confirm, now for Democrats as well as Republicans, that the victor in the presidential tally would have prospered even more under the previous (unreformed) arrangement (Table 5.6B). This added a fresh irony to reform politics, in that Hubert Humphrey had long been one of the leading reform forces in the Democratic Party, though he was running in 1968 as the champion of the regulars. Either way, the critical point remains that Humphrey would have been a substantial winner under either metric: no formula capable

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

Table 5.6 Delegation shares at the Democratic convention of 1968: the impact of apportionment reform

171

A. Regional shares after reapportionment

N’East West South

New (%)

Old (%)

Change (%)

44 33 23

44 29 27

0 +14 −15

B. Candidate votes after reapportionment

New formula Old formula

Hubert Humphrey

All others

1759 1839

841 761

of adoption would have altered the actual result. Humphrey would in fact have done marginally better under the old rules, so this time, reform had at least secured a small impact on nominating politics in the expected direction, just far too small to have a noticeable effect on the ultimate result. As ever, the potential for autonomous impacts from apportionment reform would inevitably decline as the new formula became integrated into—part and parcel of—an overall nominating dynamic. Yet the Democrats were destined to have another highly contested nomination in the very next presidential year, 1972, a contest still close enough to the reform moment to capture any unadulterated impact from reapportionment.14 And once again, there were statistical effects from a change in the apportionment formula (Table 5.7A). In these, the Democratic south continued to be substantially disadvantaged by the new formula, while it was the northeast this time that appeared to have been advantaged. Yet putting the contests of 1964, 1968, and 1972 together underlines how much of any actual apportionment effect was dependent on the varying impacts of presidential voting, in this case, because the northeast had stayed Democratic in 1968, while neither the south nor the west had done so. Regardless, the actual nominating tally at the national convention remained the crucial second step in demonstrating real consequences from an altered formula, and there the story remained much the same (Table 5.7B). Dividing this tally into George McGovern versus all others did show a McGovern benefit from the reformed formula, a result generally in line with original (reformist) intentions. Yet McGovern, like

172

B. E. SHAFER AND E. M. SAWYER

Table 5.7 Delegation shares at the Democratic convention of 1972: the impact of apportionment reform

A. Regional shares after reapportionment

N’East West South

New (%)

Old (%)

Change (%)

48 30 23

42 29 28

+14 −3 −18

B. Candidate votes after reapportionment

New formula Old formula

George McGovern

All others

1728 1621

1288 1395

C. Winner-take-all primary after reapportionment

New formula Old formula

Yeah

Nay

1618 1537

1238 1321

Humphrey before him, would have won solidly under the old formula as well, so once again, either formula would have accommodated the same ultimate result. Yet participants at the time might have argued that the 1972 nominating ballot was the wrong test of any ultimate reform impact. For this nominating ballot followed not just a long sequence of presidential primaries and state caucuses, but a key rules decision at the convention itself. In this line of argument, the critical indicator of apportionment impact arrived in the guise of a minority dissent from the majority report of the Rules Committee. The majority report argued that new party rules meant that all state results, including those from the giant winner-take-all primary in California, were forbidden to follow a “unit rule,” whereby all delegates were delivered to the plurality winner. By that definition, the California primary had been arguably the greatest application of the unit rule in all of American history. So Table 5.7C recalculates this prior skirmish, giving each side the same share of individual state votes as it actually obtained, but applying this percentage to the size of state delegations under both the new and the old apportionment formulae. Drama and precedent aside, the result remains familiar. On the one hand, the McGovern forces did better under the

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

173

new formula than they would have done under the old. That is a reform impact running in the expected direction. But on the other hand, and critically, this difference was not close to decisive: the McGovern forces would have won under either formula. In reality, having lost this vote, Hubert Humphrey would almost surely have withdrawn under the old formula as he did under the new. And a McGovern triumph would have looked much the same either way.

Rules for Delegate Allocation Yet there was a second major category of rules governing the formal process of presidential nomination, with its own prospects for altering the grand nominating dynamic, its own struggles over stasis or change, and its own potentially autonomous contribution to nominating outcomes in any given year. Once again, such procedures would inevitably levy costs or benefits on candidates with differing political backgrounds, differing group support, and differing policy preferences. Though beyond that, there was a key difference. Explicit and universal rules for delegate apportionment were necessary almost from the beginning, else national party conventions could not define their own boundaries. Yet while some minimal ground rules for the allocation of those delegates would likewise prove necessary, else competing cohorts of state delegates would repeatedly tie up the national convention, the need for uniformity among these procedures was much less pressing. Indeed, the basic assumption within both parties for much of American history was that this was an internal matter, that is, a matter best left to state party decisions. It was the state parties, after all, that would select individual delegates to the national convention, where these delegates would in effect become the convention votes that were allocated to aspiring nominees. Sometimes decisions about candidate support would be reached collectively within a state; sometimes they would be reached individually. Sometimes delegates would be chosen with candidate support in mind; other times they would look to acquire this attachment on the way to the convention or at the convention itself. Yet decisions on all such matters were internal party business, and this would remain true for the Democrats until the late twentieth century and for the Republicans more or less forever, at least as this is written.

174

B. E. SHAFER AND E. M. SAWYER

In that sense, the two major categories of party rules were instantly different in the nature of their associated politics. Yet as with delegate apportionment, so with delegate allocation: there was an implicit assumption at the start about the way votes apportioned to the states should be cast (and in that sense allocated) at the convention itself. Moreover, just as the opening assumption about delegate apportionment was that size in the Electoral College was the proper standard for a process in which presidential votes would ultimately be cast en bloc by the Electors, so there was a counterpart analogy with regard to the process of delegate allocation. To wit: states should poll their delegates and then be able to cast their convention vote as a bloc for any presidential aspirant with a (state) majority, again just as in the Electoral College. This would maximize state influence with aspiring national nominees, allowing this influence to be diminished only when there was no candidate majority. Efforts to institutionalize this analogy by means of a ‘unit rule’ would thus become the long-running, if highly intermittent, flash point for conflict over rules governing delegate allocation. Initially, the rule seemed sufficiently intuitive as to remain in the background, more or less as an unwritten standard. Yet from the start, there were inherent tensions associated with its application. Delegates were, after all, chosen as individuals, and while these particular individuals would normally salute the importance of maximizing state influence, they were simultaneously people with careers, interests, and preferences of their own. Moreover, while some states selected all of their delegates at a state convention, the counterpart to the national gathering, others selected them in a more decentralized fashion, at county or district conventions where regional differences in policy stakes and candidate preference were indirectly built into their national delegations.

The Politics of Delegate Allocation The Republicans The Democrats would be unbothered by fallout from these incipient tensions for a very long time. The party would have to visit and revisit the two-thirds requirement for a presidential nomination in its early conventions. Yet the unit rule for allocating convention delegates to presidential aspirants would not attract even tangential disputes until after the Civil War. The opposition party, however, was not to be so conflict-free.

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

175

Indeed, the Whigs ran into questions about the imposition of the unit rule almost from the start, when their first truly national convention, in 1840, produced a nominating politics forcing the issue to the fore. The front-runner for their nomination was Senator Henry Clay of Kentucky, followed by General William Henry Harrison of Ohio and trailed by General Winfield Scott of Virginia. As the best known and best connected of the three, thanks to his years as Speaker of the House, Clay had support dispersed among the largest number of individual states. Yet each of his two main opponents possessed concentrated blocs of delegates in particular states, and they had more support than Clay in total, when (if) their partisans were added together. So after extended debate, the Harrison and Scott forces imposed state balloting by unit rule on the full convention. This deprived Clay of minority support in many places, and the nomination ultimately went to Harrison.15 Those politics and that result suggested that the unit rule had the ability in practice and not just in principle to alter a nominating dynamic. Clay himself was to recover soon enough, becoming the unanimous choice of the 1844 convention, though he would go on to lose the general election. Yet in the process of nominating him and despite its unanimity behind the nominee, this 1844 convention retreated from convention-wide imposition of a unit rule, returning the issue of its application to the individual states. And there it would remain for the Republicans until after the Civil War, when the convention of 1876 brought it back to party-wide attention, ultimately generating a resolution that has lasted into our time. The issue reemerged at the Republican convention of 1876 in a fashion again having as much to do with candidate prospects as allocation rules per se. James G. Blaine of Maine, Speaker of the House, was the front-runner in a multicandidate contest, when three delegates from Pennsylvania appealed to the convention chair, asking to be freed from a state rule requiring a vote for the majority preference of the total state delegation. The chair ruled in their favor, a voice vote confirmed that ruling, and a roll-call reconsidered it, before a second roll-call upheld the power of the chair to suspend the rule. The short-run result was a few more votes for Blaine, albeit nowhere near enough to prevent the eventual concentration of his opponents around Rutherford B. Hayes, favorite son and Governor of Ohio.16

176

B. E. SHAFER AND E. M. SAWYER

The issue came back as a conscious procedural matter but with additionally clear-cut candidate implications at the Republican convention of 1880. Old-line regulars within the party were attempting to bring back Ulysses S. Grant in a third run for president, and while Grant, like Blaine, was the early leader of a multicandidate field, the political logic was reversed. This time, it was the Grant forces who wanted to impose a unit rule convention-wide, because they had majority support in a number of major states. Their effort was to spill over into everything on the convention agenda. It began by coloring the battle over a convention chair, where the Grant preference was beaten by a neutral substitute. The struggle moved on to numerous credentials challenges, often pitting delegates chosen in district conventions, favoring various Grant opponents, against delegates chosen (for the same seats) at state conventions, favoring the former president.17 When most of those challenges were won by the districting side, the conflict moved directly to the unit rule, through a majority report from the rules committee recommending that the rule not be used. The Grant forces made a late effort to duck this by moving that the nominating ballot precede the adoption of the rules. That too was unsuccessful, though Grant still led on multiple opening ballots, ahead of none other than James G. Blaine of Maine plus John Sherman, Secretary of the Treasury from Ohio. Grant remained in the lead through the thirtyfifth ballot, but on the thirty-sixth, Congressman James A. Garfield of Ohio, who had originally nominated Sherman and had been gathering strength only late in the contest, overcame the former president for the nomination. In an effort to avoid the multiple procedural knots that had characterized 1880, the Republican convention of 1884 would then produce the first codified party rules for delegate selection nationwide. Along the way, the report of its rules committee repeated the decision of the 1880 convention on the unit rule—that it was a matter for decision by the individual states—carrying this provision over without further discussion and in effect integrating it into official party procedure. And there the matter rested within the national Republican Party, with the exception of some modern situations (addressed below) where Democratic reforms to delegate allocation, requiring passage by various state legislatures for their realization, would be applied by some of these legislatures to the Republican process as well.

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

177

The Democrats The convention of 1876 would also be the one to see the unit rule emerge as a matter of contention among Democrats, though for some years this would be largely just as a by-product of disputes within the giant New York delegation, local party disputes played out on the national convention stage. Unlike the Republicans, the Democrats would not really address the rule as an aspect of institutional structure until 1968, a hundred years later. Though when they finally did so, they would be far more procrustean, with an attempt to abolish the practice at every level and in every locale. In the meantime, where the Republican response was destined to be short and decisive, the Democratic response was instead to become contingent and protracted. Application of a unit rule for delegate allocation did pop up as an issue of dispute at the Democratic and not just the Republican convention of 1876, in the first of what would be a series of state-based battles within the giant New York delegation. On one side was Tammany Hall, citadel of the regular Democratic Party, based in New York City, and led throughout by John Kelly. On the other side were supporters of those reform governors who would serve as potential saviors for the Democratic Party as it tried to return to nationally competitive status. Samuel J. Tilden would appear in this latter role twice, albeit once in absentia, while Grover Cleveland would appear in it three times. And the unit rule was integral to most of the resulting disputes.18 In 1876, the full New York delegation invoked the rule in order to cast all its votes (including Kelly’s) for Tilden, while Kelly decried imposition of the rule on otherwise legitimately selected delegates, as one further element in his campaign to demonize Tilden. For 1880, Tammany would come back to focus more directly on delegate credentials, asking to be awarded 20 of New York’s 70 delegates. Yet the full convention turned him down, and New York cast all its votes, courtesy of the unit rule, for a succession of stand-ins for Tilden, who never himself ultimately entered. Tammany brought the rule back explicitly in 1884, by moving a minority report abolishing it and thereby setting Tammany members free to vote against Grover Cleveland. The full convention rejected this initiative, with New York voting 0-72 against the rules minority report and 72-0 for the nomination of Cleveland. There was to be no serious conflict over the renomination of Cleveland in 1888, but Tammany would receive one last hurrah in 1892, when its forces finally captured the New

178

B. E. SHAFER AND E. M. SAWYER

York delegation behind sitting Governor David B. Hill. Cleveland was nevertheless overwhelmingly chosen, 671.5 for the president versus 114 for Hill on the first ballot, where 72 of those 114 votes resulted from imposing the unit role on the New York delegation. The rule resurfaced, idiosyncratically and more or less in passing, at the Democratic convention of 1912, the one featuring a forty-five-ballot struggle between Champ Clark, Speaker of the House, and Woodrow Wilson, Governor of New Jersey.19 The Wilson forces sought to allow district delegates from Ohio who favored Wilson to cast their votes for him, even though the state convention had bound the delegation to a favorite son, Governor Judson Harmon. The full convention did allow those delegates to vote their individual choice, being careful to focus the decision narrowly on Ohio. So that decision had no effect on uses of the unit rule elsewhere, allowing the large New York delegation to cast all 90 of its votes against Wilson up through the forty-fifth ballot. In this way, the rule continued to be a nonissue nationally, in effect an option for individual states but not for the convention as a whole. The biggest states remained partial to it, though mainly as a tool for holding their delegations behind favorite sons until the structure of candidate conflict had clarified, rather than as a way to maximize their initial preferences. The southern states remained partial to it as well, as a further means of leveraging their clout within the two-thirds requirement for nomination. And there the situation rested until the issue broke through again in the Democratic convention of 1968, where delegate allocation by means of the unit rule produced the opening salvo in a much larger battle over party rules, while being an institutional olive branch whereby the Humphrey leadership could join the McCarthy and McGovern forces in a shared pursuit of reform—and an ultimately vain pursuit of unity. The demise of a rule that had characterized nominating politics inside the Democratic Party for the previous hundred and thirty-six years was largely lost to the outside world amidst the plethora of other internal conflicts—and external dramas—characterizing that 1968 convention. Yet despite the endorsement of all the major candidates, the vote on abolition was comparatively close—1350 in favor and 1206 opposed—while containing a set of nice ironies. New York and California, now the two biggest delegations, cast unanimous votes for reform, the New York vote being a product of the unit rule and the California vote being the result of a winner-take-all primary, the electoral counterpart to that rule.20

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

179

The Impact of Allocation Reform The Unit Rule Specifically How much did it matter? If answering that question about apportionment rules was a fraught exercise, answering it about allocation rules, as embodied most forcefully and for the longest time in the unit rule, is nearly impossible. The ideal test would have come right at the beginning, with the Whig convention of 1840. There, the fact that the rule was applied to every state was widely believed, by major players at the time and political historians afterward, to have been fatal to Henry Clay, the candidate with minority support in multiple places, which means that Clay may indeed have been defeated not by a bandwagon but by a rule. Unfortunately, its presence convention-wide also means that there can be no evidence of the extent of divisions beneath a forced conformity, and hence no plausible estimate of the scale of rules-based damage to Clay.21 So the lasting point is only that any impact from use of the unit rule convention-wide could not be a recurrent effect, since there would never again be uniform application of the rule. That left plenty of room for individualized effects in specific states or even in larger geographic blocs, and Chapter 6 will go in search of these as part of an inquiry into state and regional strategies. But after 1840, the rule reverted to being purely a state option, which it would remain until it was abolished. In the long interim, some states would use it. Other states would not. Some states would generate a unanimous vote without the need for a unit rule. Others would adopt the rule but allow designated individuals to deviate from its strictures. So in the face of this variety, there would never again be an obvious metric for comparison, while the analyst who imposed one on paper could not know what share of unified delegations would have remained that way without imposition of a unit role or, more tellingly, what share of minimally divided delegations would have been more divided if their apparent candidate splits had been not just personal exemptions but implicit indicators of greater minority dissidence. Ironically, the closest thing to a plausible concrete test would not arrive until the very end of the institutional life of the rule itself. This came at the Democratic convention of 1968, where the party summarily banned the unit rule not just inside the convention but inside individual states as well. As a preliminary to this ban, there was an actual vote assessing state party orientations toward the rule, asking states to choose between banning the rule universally or preserving a right to apply it case by case,

180

B. E. SHAFER AND E. M. SAWYER

Table 5.8 Candidate votes after abolition of the unit rule: the Democrats, 1968 Unit rule Absent Present

Humphrey

McCarthy

McGovern

Others

1759 1828

601 543

147 141

68 63

as a state option. If a state had intermittently used the unit rule in earlier years and voted heavily to retain it—for purposes of Table 5.8, by more than three to one—then that state might reasonably have been expected to apply the rule to its own presidential tally, had it been allowed to do so. Fortunately for the analysis, because state delegations could not apply the rule thereafter—abolition did succeed—the analyst also possesses an actual division of candidate preferences within the individual states. This allows the analysis to create two contrasting versions of the 1968 nominating ballot: (a) a real one, in which states were forbidden to use the unit rule, designated here as rendering the rule “absent” and (b) a hypothetical one, in which states that were supportive of the unit rule in principle were automatically assigned to apply it, marking the rule as “present” in these states but not in the others (Table 5.8). From the nominating vote recast in this way, two findings emerge.22 First, the difference between disallowing the rule or allowing it at state discretion still appears, and in the expected direction. Thus Hubert Humphrey would have gained delegates if the rule was imposed, while self-described reformers would have shed those delegates. But second, as with most impacts from recalculating results under different apportionment formulae, the impact of allowing or forbidding a unit rule is small. It shifts about 70 out of 2600 delegates, which carries no threat, even in the heated atmosphere of an explosive year, of producing a different nominating outcome. So nothing about the particular nominating politics of 1968 was anywhere close to being altered by what was otherwise a historic reform in narrowly institutional terms. Procedurally, this resolution marked the end of an allocation rule that had been available to Republican state parties until 1884 and to Democratic counterparts until 1968. Yet operationally, while the unit rule itself was gone in both parties in the aftermath of this vote, elimination would go on to trigger a string of subsequent struggles over knock-on standards for delegate allocation, generating one more set

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

181

of possible impacts on nominating politics, one that showed a continuing difference between the parties this time. Fortunately, that would also permit further concrete tests of the impact of alternative allocation rules. In other words, the demise of the unit rule, rather than terminating disputes over permissible ways to divide and allocate delegates, actually extended these disputes, driving established reform arguments into fresh procedural territory. In the new plebiscitary era, the vast bulk of national convention delegates were now chosen in presidential primaries, all of which would be candidate primaries, where every delegate was connected to a presidential preference. So fresh reform arguments were inevitably focused on these primaries. Yet their implementation now pitted champions of the right to elect delegates, one great democratic principle, against champions of the precise reflection of candidate preferences, another great principle and the one that drove Democratic reformers—and would ultimately divide the parties. Believers in the right to select individual delegates would come to prefer what are known as districted elections, run on a first-past-the-post basis by electoral district, though the size of these districts could still vary among the states. Conversely, believers in the primacy of presidential preference would prefer strict proportional formulae, likewise applied to electoral districts whose size varied by state but privileging accurate preference registry over direct delegate control. The former could thus be attacked for producing delegate outcomes that did not precisely track expressed presidential preferences. Just as the latter could be attacked for sometimes going so far as to remove the identities of their delegates from the ballot entirely, but always removing the ability to pick them individually. Proportionality in General It was the state of California that would convert this philosophical difference into a practical struggle, by returning in 1972 to an established strategy for expanding its own influence at national party conventions. In California, a presidential primary with a large number of delegates— then the second-largest number—was moved to the end of the primary calendar with the goal of being decisive by offering all these delegates in a bloc to the victor, winner-take-all, thereby terminating the contest. On its own terms, this embodied a nice irony, with party reformers defending an allocation rule more extreme than the old unit rule, while party regulars

182

B. E. SHAFER AND E. M. SAWYER

attempted to enforce the very standards that reformers had articulated at the previous convention. A change of candidates, with George McGovern as the champion of the reformers in 1968 but the winner of the California primary in 1972, was the proximate explanation for this particular switch. For 1972, the McGovern forces succeeded in having it both ways, keeping California winner-take-all but affirming that this particular allocation rule was indeed illegal going forward. For 1976, then, winner-take-all primaries, now formally banned, actually disappeared inside the Democratic Party. They declined within the Republican Party as well, in part because state Republican Parties were responding to elements in their own constituencies that were attracted by the proportionality argument, in part because legislatures in states with Democratic majorities sometimes imposed the same rules on both parties. Yet for Democratic reformers, the 1976 results, rather than capping their triumph within a long-running struggle, only brought into focus a second alleged miscarriage of reform politics. For with winner-take-all rules banned, many state Democratic Parties had moved to adopt districted primaries, where national convention delegates were elected in subdistricts of the state, most often congressional districts. This added inherent elements of proportionality, but reform proponents argued that it continued to violate a true proportional standard by allowing a candidate to capture all the delegates in any one district, a system which they derided as a “loophole primary.” In response, the Democratic convention of 1976 would ban direct election of delegates by district as well, pressing its state units toward some version of explicit, that is, formulaic, proportionality.23 Though the size of individual states coupled with the size of election districts—the party was not prepared to require that all delegates be chosen statewide—would still leave some variation in delegate results. Across the partisan aisle, however, the proportionality argument made less headway, with state units of the Republican Party more inclined to feature direct election as the essence of internal democracy. For them, this still guaranteed major elements of proportionality, unless one candidate really was winning everything, in which case the entire argument was moot. At the same time, it avoided interposition of a mechanical formula that was beyond the control of—and rarely even understood in its details by—primary voters. There was still some movement in the direction of the Democrats for the usual reasons, with the reform argument gaining traction among some Republicans. or Democratic legislatures delivering more

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

183

proportionality than the other side desired. Yet in the end, the Republicans retained considerably more districted elections for the plebiscitary era.24 So both the philosophical and the institutional arguments were resolved differently as between the two major parties, with the Democrats extending uniform national regulations deep into the internal mechanics of their state parties, while the Republicans left a greater variety of alternative mechanisms to state party choice. Yet this variation in allocation rules as between the two parties raised the usual questions. For practitioners, the question became whether proportionality extended or districting compressed the time-period for a presidential nominating process, and whether this altered timeline, if indeed it existed, contributed to the success or failure of the eventual presidential nominee. For analysts and interpreters, the question was instead whether this difference had a serious impact on the character of the nominating bandwagon, the occupational seedbeds of presidential aspirants, or the factional behavior that effectively chose among them. The easy part of generating an answer would involve comparing the two extremes, that is, recalculating results on a winner-take-all versus proportional basis, though even this becomes highly problematic within a very short time inside any given nominating contest.25 Yet the real world had already moved past that. Winner-take-all primaries had effectively disappeared, so a more realistic test has to involve the differential impact of explicit proportionality versus direct election. This is a much less satisfying comparison even in formal terms, because it turns on the presumption that a few delegates here might have generated a few more delegates there, which might have gone on to alter the result by a few other delegates there, perhaps until there was a different nominee but certainly in a manner that notably extended the nominating contest. So recalculation on these terms risks becoming hypothetical very quickly, and thus extremely tendentious. On the other hand, the presence of presidential primaries weighted toward proportionality among Democrats but toward districting among Republicans does allow one mechanical recalculation that is less purely hypothetical and generates a more automatic comparison. For this, it should be possible to take the actual vote within the party favoring proportionality, in this case the Democrats, and rerun this vote on paper using rules prioritizing districted election, in this case those of the Republicans.26 The resulting comparison still overstates the difference, since it

184

B. E. SHAFER AND E. M. SAWYER

does not kill off marginal candidates who might have disappeared even earlier if they had been forced to produce district majorities in order to survive. Yet the basic calculation is not difficult. Alternative rules already exist in the two existing processes of delegate selection, and there is a neutral and automatic measure of their impact, the Gain-Deficit Ratio, designed explicitly to make temporal comparisons. So it should be possible to answer the question of whether one or the other allocation formula produces earlier or later attainment of the Gain-Deficit Ratio, the point at which the nomination has been resolved for all practical purposes.27 To that end, Table 5.9A sets out the dates when the Gain-Deficit Ratio Table 5.9 The timing of nominations A. Gain-Deficit success for Democrats versus republicans Year

Dems

Reps

1976 1980 1984

June 8 May 6 May 8

June 8 May 3

1988 1992

June 7 June 2

1996

—— March 8 —— March 12

—— March 7 March 2

2000 2004 2008 2012

March 7 —— February 5 March 6

February 5 —— March 22 March 3

2016 2020

March 15 ——

B. Gain-Deficit success under proportionality versus districting

1988 2016

Actual Democratic date

Democratic date under Republican rules

Actual Republican date

June 7 March 22

May 23 March 15

March 8 March 15

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

185

was achieved for Democratic and Republican nominees across all the years after the Democratic Party explicitly banned winner-take-all contests. The results come close to terminating the analysis at the start. Or at least, in six of these twelve presidential years, only one party had a contested nomination, eliminating any meaningful comparison. Worse yet, in four other years, when the two parties did both have genuine contests, their Gain-Deficit Ratios were achieved within the same week, and ordinarily on the same day. That is a nil result in ten of twelve cases, denying even a potential chance to reorient bandwagons, change careers, or overcome factions. That leaves two years, 1988 and 2016, when there was a gap of a full week or more between the parties in their resolution dates, and both of these did feature Republican resolutions earlier than their Democratic counterparts. Accordingly, Table 5.9B offers Gain-Deficit Ratios for Democratic nominees under Republican rules in those two years. To repeat: recalculating these results the other way around—taking Republican results and running them under Democratic rules—is effectively impossible, requiring draconian assumptions because it would demand the restoration of the vote for lesser candidates who, in the very argument being tested, should have been rubbed out by districted rules. Some such restoration is abstractly possible for the very earliest contests in a few further years, but none provide a basis for systematic comparison across even a serious stretch of the nominating struggle. So real Democratic results under actual Republican rules have to be the basis for any defensible comparison. For that comparison in 1988, the hypothesized effect was in fact present: these recalculation suggest that the Democratic race would have effectively ended more than two weeks before its actual end if the Democrats had proceeded according to Republican rules. Though note that there would still have been a difference between the two contests, in that the Republican resolution would have been two weeks earlier than the Democratic resolution even under the same rules. So there was clearly a contribution both from a formal difference in the rules and from practical politics inside the parties. For a parallel comparison in 2016, a shadow of the same effect reappears, but even less consequentially. Under Republican rules, the Democratic contest would have been resolved one week earlier than it actually was. On the one hand, this is now purely a rules impact: take away the rules difference, and you take away the timing difference as well. But on the other hand and in a far more practical sense, one of the only two

186

B. E. SHAFER AND E. M. SAWYER

contests that show any real temporal difference between proportionality and districting still reduces that difference to exactly one week, a distinction unlikely to impress either the main players, the institutional analysts, or the nominees themselves. So in the end, this becomes yet another result in the familiar form. On the one hand, an influence from alternative allocation rules on the length of nominating contests—unlike the frequent result with apportionment rules—does run in the expected direction. On the other hand, this effect is small, it appears only rarely, and when it does appear, it has so far always run in the direction of affirming the actual result, not diminishing it, much less reversing it. Said concretely, Michael Dukakis in 1988 and Hillary Clinton in 2016 would have been the nominees under either plan, just earlier under the rules governing the Republican contest rather than the rules governing their own. Said more pungently, an effect worthy of mention once every twelve contests—that is, once every forty-four years—is hard not to dismiss as effectively trivial, even random.

Delegate Apportionment, Delegate Allocation, and Nominating Politics Was there never anything of consequence for nominating politics inherent in the rules of the game themselves? A summary answer requires two immediate cautions. The first is just a reminder: this analysis has established (in Chapter 1) that there was one great rules-based effect that mattered hugely and repeatedly. That was the difference between requiring a simple majority versus a two-thirds majority for nomination, and it affected the contours of nominating politics—the very nature of the bandwagon—for a very long time. Which returns to the purposes of this entire analysis, and moves us to the second caveat. If the question is whether any other rules beyond the formal specification of a majority affected the recurrent nature of nominating politics, then the answer is just as effectively no. Not even the huge institutional shifts from a pure convention system to a mixed system to a plebiscitary system can accomplish this. Which is to say: • The continuing nature of a bandwagon dynamic was not altered by changed rules for delegate apportionment among the states nor by changed rules for delegate allocation among the candidates.

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

187

• The occupational backgrounds and career trajectories that characterized the fields of presidential aspirants were neither created nor altered by the rules for delegate apportionment or delegate allocation. • And the structural factors that consistently explained the choice of a particular nominee from among these fields, most especially internal partisan factions and external electoral demands, were not altered by the rules for delegate apportionment or the rules for delegate allocation either. Indeed, neither apportionment nor allocation could even reliably produce lesser impacts within nominating politics on a recurring basis. On the other hand, if the question is about anything of idiosyncratic consequence for nominating politics, for example, “Why did my favorite candidate win (or more often lose) in 1996?”, then any or all of these rules might have made a contribution. One way to think about this is to say that the bandwagon dynamic, the occupational seedbeds within it, and the structural influences that choose among the aspirants who arose from those seedbeds were the major shaping influences in the funnel of causality, an old but easy way to talk about progression from grand social forces at the broad end of the funnel, down to single particularistic influences right before the ultimate decision. Grand social forces qualify as such because they recur constantly, all the while narrowing the possible outcomes to a few—and sometimes no— alternatives. Yet after the bandwagon dynamic, its preferred occupations, and the structural demands of a point in time have all played this grand winnowing role, the explanation of why one aspirant rather than another was ultimately nominated—in 1888, in 1952, last time, or next time— might yet be determined by one or another embodiment of what are usually taken as “rules of the game.” Among these, rules for the apportionment of delegates to the national party convention and rules for the allocation of those delegates to presidential aspirants loom large. And in the occasional case where the major enduring forces have somehow netted out to an essentially even contest, a given difference in these rules—or alternatively, as in Chapter 6, in the strategies of presidential aspirants or the contrary strategies of state parties—might still be the final nudge toward one specific nominee rather than another. Yet the weakness of these random further influences is simultaneously confirmed by the fact that, unlike dynamics, seedbeds, and structures,

188

B. E. SHAFER AND E. M. SAWYER

they do not reliably recur. Beyond that, changes in apportionment formulae on any serious scale are rare: no candidates would be well advised to tie their ultimate fortunes to a prospective reapportionment. A few major changes in apportionment formulae did reliably shift the comparative size of regional blocs at national conventions, as the Republicans did in 1916 and the Democrats did in 1968. Those are real— measurable—impacts. It is just that the best available measures suggest that they did not affect ultimate nominating outcomes, even in years of major reform. Sometimes their impact at least ran in the expected direction. Other times, this impact actually ran in the direction opposite to reform intentions. Both times, however, their practical impact, if real, was small. Much the same can be said of major rules governing the allocation of convention delegates to presidential aspirants. Serious changes in these allocation strictures are likewise rare, though here, candidates do occasionally tie their fortunes to various incipient reform efforts. Thus the opponents of Henry Clay in 1840 did significantly diminish his nominating prospects by imposing a convention-wide unit rule, while the supporters of Ulysses S. Grant in 1880 sought crucial support for a third nomination by reimposing that rule, though they were in fact denied. The possibility of some such effect reappeared after 1972, when the Democrats decided that banning the unit rule not only barred winnertake-all primaries but actively required proportionality. Yet the result ultimately paralleled that for reapportionment: districted primaries, largely the Republican practice, as opposed to proportional primaries, the Democratic preference, made Democratic campaigns run longer (or Republican campaigns run shorter) only once in a blue moon and even when they did, the difference was nowhere close to producing a different nominee. So what has to be said about the rules for both delegate apportionment and delegate allocation is that these strictures, or changes in them, never displace factors farther back in the causal funnel. The latter recur reliably; the former do not. Though in rare and idiosyncratic cases when those grand prior factors—the bandwagon dynamic, occupational seedbeds, and factional structures—net out to rough equality among two or more aspirants for a presidential nomination, some lesser characteristic could in principle push the ultimate result in one direction or another. Moreover, if the rules of the game offered two obvious candidates for this late and idiosyncratic role, so did the strategies of the major players and the sequencing of nominating contests, which are subject matters for Chapter 6.

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

189

Notes 1. Especially for the Democrats, especially to Baltimore, where the Democrats staged their first convention in 1832 and which they did not leave until 1856. 2. The Democrats actually achieved this goal for a while, before issues of slavery and statehood stressed it to the breaking point. The Whigs shared at least the same aspiration before their demise, while the Republicans, lacking most of the slave states from the start, were unconcerned in principle. 3. Lyndon Johnson in 1964 would finally breech this informal embargo. 4. Charles Stewart, III, and Barry Weingast, “Stacking the Senate, Changing the Nation: Republican Rotten Boroughs and American Political Development in the Late 19th Century”, Studies in American Political Development 6(1992), 223–271. See also Byron E. Shafer, Regina L. Wagner, and Par Jason Engle, “The 2014 Midterm in the Longest Run: The Puzzle of a Modern Era”, The Forum: A Journal of Applied Research on American Politics 13(2015), Article 1. 5. At various points in time, the Republican national convention contained delegates not just for Alaska, Hawaii, and the District of Columbia, but also Indian Territory, Puerto Rico, the Philippine Islands, the Canal Zone, and the Virgin Islands. 6. Every state from the old Confederacy voted Democratic for president from 1888 though 1944, except for 1928, when Alfred E. Smith, Governor of New York and a Catholic, a “wet,” and a northeastern ethnic proved culturally indigestible. The sole other exception was a single deviation by Tennessee in 1920. 7. For a total of 72 delegates out of 1234, only 22 of which were southern. 8. A simple way to follow the evolution of these formulae in both parties is the series of reports from the Secretary of the Senate, published by the U.S. Government Printing Office, on Nomination and Election of the President and Vice President of the United States, Including the Manner of Selecting Delegates to National Party Conventions . 9. In addition to “1916 Conventions”, in National Party Conventions, 1831– 2008 (Washington, DC: CQ Press, 2010), 101–103, and “1916” in Richard C. Bain and Judith H. Parris, eds., Convention Decisions and Voting Records (Washington, DC: Brookings Institution, 1973), 193– 199, see “The Selection of Hughes and the Launching of the Republican Campaign”, Chapter 5 in Lewis l. L. Gould, The First Modern Clash over Federal Power: Wilson Versus Hughes in the Presidential Election of 1916 (Lawrence: University Press of Kansas, 2016).

190

B. E. SHAFER AND E. M. SAWYER

10. Most comprehensively, Paul T. David, Malcolm Moos, and Ralph M. Goldman, comps., Presidential Nominating Politics in 1952, V Vols. (Baltimore: Johns Hopkins Press, 1954). For the larger context, see John Robert Greene, I Like Ike: The Presidential Election of 1952 (Lawrence: University Press of Kansas, 2017). 11. Gerald Pomper, ed., The Election of 1976: Reports and Interpretations (New York: Longman, 1977), and Jules Witcover, Marathon: The Pursuit of the Presidency, 1972–1976 (New York: Viking Penguin, 1977). See also the very trenchant analyses of 1976 scattered throughout Larry M. Bartels, Presidential Primaries and the Dynamics of Public Choice (Princeton: Princeton University Press, 1988). 12. Robert David Johnson, All the Way with LB J: The 1964 Presidential Election (Cambridge: Cambridge University Press, 2009). For the actual delegation sizes, see Table 1. “Apportionment of Votes, Delegates, and Alternates for the 1968 Democratic Convention”, which includes the apportionments for 1964 as well. Nomination and Election of the President and Vice President of the United States, 37–38. 13. Richly detailed pictures of the context are Theodore H. White, The Making of the President 1968 (New York: Atheneum, 1969), and Lewis Chester, Godfrey Hodgson, and Bruce Page, American Melodrama: The Presidential Campaign of 1968 (London: Andre Deutsch, 1969). 14. For extended context, Bruce Miroff, The Liberals’ Moment: The McGovern Insurgency and the Identity Crisis of the Democratic Party (Lawrence: University Press of Kansas, 2007). For the specifics of the struggle, Theodore H. White, The Making of the President 1972 (New York: Atheneum, 1973). 15. An analysis taking Clay through the 1840s, and providing a rich portrait of the Whig Party around him, is Michael F. Holt, “Harrison and Prosperity or Van Buren and Ruin”, Chapter 5 in Holt, The Rise and Fall of the American Whig Party: Jacksonian Politics and the Onset of the Civil War (New York: Oxford University Press, 1999). 16. Michael F. Holt, “Republicans’ Nomination”, Chapter 5 in Holt, By One Vote: The Disputed Presidential Election of 1876 (Lawrence: University Press of Kansas, 2008). 17. Charles W. Calhoun, “Third Term Dreams”, Chapter 22 in Calhoun, The Presidency of Ulysses S. Grant (Lawrence: University Press of Kansas, 2017). 18. The Tammany thread can be followed through all five of these elections, 1876–1892, in Bain and Parris, eds., Convention Decisions and Voting Records, and in CQ Press, National Party Conventions, 1831–2008.

5

THE USUAL SUSPECTS: RULES FOR THE NOMINATING CONTEST

191

19. Lewis L. Gould, “Woodrow Wilson Against the Democratic Field”, Chapter 4 in Gould, Four Hats in the Ring: The 1912 Election and the Birth of Modern American Politics (Lawrence: University Press of Kansas, 2008). 20. For the specifics of rules reforms at the 1968 convention itself, Byron E. Shafer, “The Idea of a Reform Commission: The McCarthy Campaign and the ‘Mandate’,” Chapter 1 in Shafer, Quiet Revolution: The Struggle for the Democratic Party and the Shaping of Post-Reform Politics (New York: Russell Sage Foundation, 1983). For the convention vote on abolition, see CQ Press, National Party Conventions, 1831–2008, 253. 21. See Note 16 above. The specifics of the actual nominating vote in 1840 are missing from the two standard compilations, Bain and Parris, eds., Convention Decisions and Voting Records, and CQ Press, National Party Conventions, 1831–2008. That vote is available for 1848, the last serious run by Clay, but an attempt to generate two nominating ballots there, one with and one without the unit, requires much stronger assumptions than such an attempt would for 1840, while the impact of the unit rule runs in the opposite direction this time. That is, a comparison suggests that it was the front-runner Zachary Taylor, who would have gained votes under a convention-wide imposition of the rule, the largest number of which would have come at the expense of none other than Henry Clay. 22. As in Note 21 above, not just the vote on abolition but also the ultimate nominating ballot can be found in CQ Press, National Party Conventions, 253. An even larger of votes from the 1968 Democratic Convention of 1968 is offered by Bain and Parris, eds., Convention Decisions and Voting Records, in “Appendix C. The Voting Records”, unpaginated. 23. For a sense of the limitless character of potential reforms, Barbara Norrander, The Imperfect Primary: Oddities, Biases, and Strengths of U.S. Presidential Nomination Politics (New York: Routledge, 2010), gathers much of that universe in one place for consideration. 24. For the tangled story on both sides of the aisle, but with particular attention to Democratic struggles that were driving the overall change, see Elaine C. Kamarck, “Proportional Representation: Why Democrats Use It and Republicans Don’t”, Chapter 4 in Kamarck, Primary Politics: How Presidential Candidates Have Shaped the Modern Nominating System (Washington, DC: Brookings Institution, 2009). 25. One effort along these more mechanically conservative lines is James I. Lengle and Byron E. Shafer, “Primary Rules, Political Power, and Social Change,” American Political Science Review 70(March, 1976), 25–40.

192

B. E. SHAFER AND E. M. SAWYER

26. Recalculating these results the other way around—taking Republican results and running them under Democratic rules—is much harder, demanding more draconian assumptions since districted results discourage what might well be obvious minority votes under proportional rules. In fact, within a very few weeks in most years, generating additional minority votes is quite liberally impossible. 27. For 2016, and indeed for most recent years, a hugely useful compilation of state party rules for delegate selection can found the website of The Green Papers: thegreenpapers.com. For 1988, the weekly issues of National Journal and Congressional Quarterly provide good procedural information, supplemented by convention briefing materials from the major television networks and the two conventions themselves.

CHAPTER 6

The Usual Suspects: Strategies and Sequences

Changes in the major rules governing the apportionment of national convention delegates among the states or the subsequent allocation of those delegates to presidential aspirants have not evidently altered the bandwagon dynamic that has characterized presidential nominating politics since the 1830s. These changes have likewise failed to alter the nature of the prior occupations and career trajectories that have dominated extended stretches of time within that long political run, though presidential seedbeds do change with the larger context of politics around them. Nor, finally, have amended rules changed the key structural influences that actually determine an ultimate choice within the winnowed fields of presidential aspirants, though again, key structural elements like internal party factions do themselves shift in separate but important ways over time. Most rules battles are otherwise a form of “inside baseball.” They can, at least intermittently, get attention from the main players, especially campaign strategists and professional analysts, but this rarely extends into even the interested public. On the other hand, there is a further set of recurrent, putatively influential factors that do get a great deal of attention from reporters, pundits, and professors. And here, their reports, comments, and extrapolations reach more often and more regularly into the general public, or at least that part of it which follows delegate selection politics and which ultimately turns out in candidate primaries or © The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_6

193

194

B. E. SHAFER AND E. M. SAWYER

participatory caucuses. These purportedly influential factors, the ones that do attract some public attention, are the strategies that reliably characterize the nominating campaigns of presidential aspirants and that, at one remove, characterize the efforts by state political parties at augmenting their own nominating influence. Like rules of the game, these strategic factors have the potential to alter one or another given outcome on the way to a presidential nomination. At a minimum and very concretely, the combination of this second set of maneuvers, that is, the interaction of candidate strategies and state calculations, is what produces the specific sequence of contests— and eventual outcomes—on the way to a presidential nomination. In this inter-action, state parties seek to optimize the influence of their delegate apportionment, just as presidential contenders seek to maximize their own allocation of these delegates. So state calculations do recurrently shape the sequence of contests within which the nominating bandwagon must role, while the strategies of contenders who respond to this sequence inevitably provide the detailed politicking that characterizes any given presidential year. From one side, the sequence by which the key events of a nominating contest unfold will powerfully shape the strategies—mutual and interactive, after all—which the various participants produce. From the other side, the strategies of a field of presidential contenders will always alter the practical implications of any particular sequence of events. So the most concrete and specific way to organize a search for the contributions of both is to focus first on the actions of presidential contenders, then on the actions of state political parties. Candidate strategies are more numerous, more varied, and more nimble, in the sense of changing more quickly. State strategies are larger and more slow-moving, but exhibit a greater (and continuing) impact when they do occur. These two moving pieces, being the ones highlighted by commentators of all sorts, are the other main purported influences on the politics of presidential nomination, even if they usually turn out to be a minor element in any comprehensive analysis. In the eyes of all too many observers, they appear to determine the final choice between one specific individual and another. And they may very occasionally be the final advantage or disadvantage for a specific aspirant in a particular year—in effect the remaining influence, the error term if you will, that necessarily completes a comprehensive analysis. Regardless, for an observer who does not care about how nominating politics works but cares a great deal about who turns up

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

195

at the end of the causal funnel—not to mention for the individual aspirants themselves—these lesser factors may be the main point of the entire exercise. So an analysis that focuses on the bandwagon dynamic, candidate seedbeds, and structural factors need to pay attention to these other purported influences too.

Strategic Maneuvers Among Presidential Aspirants Whether to Run Every potential candidate must decide whether to enter the nominating contest, how to enter the nominating contest, and when to enter the nominating contest. Because candidates vary widely in the strategic answers that are practically available to them, but also because the three answers are usually linked, observers tend to register and remember only whether any given individual was “in” or “out,” along with their eventual fortunes. Yet while strategies do inevitably change as the contest unfolds, initial strategies ordinarily contain answers to all three opening questions. These almost inevitably emerge out of occupational backgrounds and prior careers. They are colored by the key structural influences shaping candidate selection. And they follow from judgments about the potential response of state parties, as well as of the other major players, who are simultaneously making judgments on the same matters. The universe of those who seriously considered, or might well have considered, or should have considered the run for a presidential nomination remains infinite and unknowable. Moreover, the most basic strategic decisions made by aspiring candidates for a presidential nomination are perhaps the ones most quickly lost to history. Beyond this historical amnesia, most election years produce some actual entrants whose calculations could neither have been anticipated nor perhaps even comprehended. Just as many years feature the absence of one or two candidates who might have been expected to enter and were widely judged to be capable of doing so. Yet as diverse and amorphous as this process may seem, specific answers can be given a consistent patterning across time by the three major institutional regimes for presidential nominating politics: the pure convention system, 1832–1908; the mixed system, adding presidential primaries to state conventions, 1912–1968; and the modern plebiscitary system, dominated by explicitly candidate-based primaries, 1972–2020.

196

B. E. SHAFER AND E. M. SAWYER

Perhaps the outstanding modern example of the decision not to run, as well as a good introduction to the analytic puzzle and its practical implications, involved Mario Cuomo, Democratic Governor of New York from 1983 to 1994. Widely viewed as the front-runner for the Democratic presidential nomination of 1992, with substantial support from both party activists and the regular party, Cuomo did not enter.1 So that contest ultimately produced the far less anticipated nomination of Bill Clinton, Governor of Arkansas, accompanied by frequent assertions in its aftermath that the absence of Cuomo was a major factor in the success of Clinton. Though the “ghost” of Mario Cuomo was even longer-lived than this headline example, running from his keynote address at the 1984 convention, to widespread elite speculation in 1988, to his nominating speech for Clinton in 1992, through his frequently referenced attack on congressional Republicans at the 1996 gathering. An earlier example, with actual and more evident measures of the real impact of a nonentry, involved Samuel J. Tilden, reform Governor of New York and Democratic nominee for president in the disputed general election of 1876.2 Having been deprived of what many regarded as a victory in that election, Tilden was widely expected to make a second run in 1880, and informed opinion dubbed him the incipient front-runner. Yet Tilden kept his intentions to himself, to the extent that when the 1880 Democratic convention convened, supporters had to settle initially for a “stalking horse,” Henry J. Payne, former congressman from Ohio. Payne was expected to play that role only until Tilden entered. Instead, the role itself disappeared with Tilden’s public statement that he would not seek the nomination, which in turn allowed Gen. Winfield Scott Hancock, the delegate front-runner at that point, to proceed to a nomination. The modern plebiscitary era has converted choices by non-entrants into less of a topic for analysis. Because candidates must enter in order to acquire delegates, there are no longer any early non-entrants who are merely waiting for their moment. Instead, having demurred, non-entrants quickly become irrelevant. Delegates who have been acquired as personal loyalists by one candidate can rarely be acquired by another, while the bandwagon will begin rolling early among those who did not hesitate to enter. Moreover, this compression of the timeline for entry appears to have achieved a second level of impact: the modern era is much harder on what were once the “perennial possibilities,” that is, major figures who might enter in one or more of an extended sequence of quadrennial contests.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

197

Both the pure convention period and the mixed system still generated such figures, who seemed a perennial possibility for a presidential nomination, to the point where other aspirants had to take their expected actions into account. In that sense, they were omnipresent in collective strategic decisions and not just on their own. In the years before the Civil War, Henry Clay of Kentucky, Speaker of the House, became the prime example. Clay stood first as the presidential candidate of the National Republicans in 1832; endorsed the three-candidate strategy of the Whigs in 1836 and otherwise stayed clear of the presidential race; actively sought the Whig nomination in 1840 but was defeated, in part by imposition of the unit rule; was unanimously chosen as the Whig nominee in 1844; and came into the 1848 convention as the front-runner, before the supporters of Gen. Zachary Taylor and Gen. Winfield Scott united to deny him the nomination.3 In the years after the Civil War, this role was instead played by James G. Blaine of Maine, likewise Speaker of the House. Blaine brought a wide lead into the Republican convention of 1876, before opponents united behind Rutherford B. Hayes as a dark horse. He was the lead opponent of a third nomination for Ulysses S. Grant in 1880, before the ultimate strategy for avoiding Grant came to involve uniting behind James A. Garfield. Blaine became the nominee in 1884, after the death of Garfield and over the opposition of sitting President Chester Arthur. He took himself out of the running in 1888, thereby encouraging multiple candidacies and producing the unexpected nomination of Benjamin Harrison. And Blaine seriously considered challenging the sitting but unpopular President Harrison in 1892, before deciding that this would guarantee electoral defeat for either of them.4 For the late nineteenth century and comfortably into the twentieth, the potential contender who always had to be considered was William Jennings Bryan, former one-term congressman from Nebraska. Bryan stormed to an unexpected and highly contested Democratic nomination in 1896; was renominated without opposition in 1900; announced his noncandidacy in 1904, then tried unsuccessfully to block the eventual choice, Judge Alton B. Parker; retook control of the convention and its nomination in 1908 over only token opposition; did not stand in 1912 but threw his support to Woodrow Wilson, the eventual nominee; lost a contest for a delegate seat at the 1916 convention, which invited him to speak anyway; and returned as a delegate but otherwise as an operational shadow of his former self in 1920 and 1924, proposing various major alternative platform planks that were variously ignored or crushed.5

198

B. E. SHAFER AND E. M. SAWYER

No one after 1920 would have the strategic omnipresence of a Bryan, a Blaine, or a Clay, but the general phenomenon continued through the mixed system of presidential nominations. Thus in the 1940s and into the 1950s, each major faction of the Republican Party generated a continuing—a repeat—champion. For the moderate regulars, this was Governor Thomas E. Dewey of New York; for the ideological activists, it was Senator Robert A. Taft of Ohio; and the struggle to elevate or to displace one or the other would color Republican nominating politics for a long time.6 Dewey, then Manhattan District Attorney, and Taft, already a Senator, inaugurated this struggle as front-runners at the 1940 convention, before insurgent Wendell Willkie elbowed first Dewey and then Taft aside. Dewey, now Governor of New York, came back as the successful consensus choice in 1944, successful, that is, in the nomination but not the election. Dewey sought the nomination for a third time in 1948, again challenged by Taft, whose inability to forge a coalition with the other candidates, especially former Governor Harold E. Stassen of Minnesota, ultimately presented Dewey with his second successful nomination. And in a close contest in 1952, Taft returned as the champion of the activist wing of his party to stand against Dwight Eisenhower, who had replaced Dewey as champion of the regulars, while Stassen returned to lead the shift to Eisenhower, which completed the nomination of the latter. That did prove to be an end of a mini-era: thereafter, no candidate would seriously enter the strategic calculations of other possible contenders for four or more consecutive conventions, at least as this is written. Conversely, if those men produced the most unpredictable contributions to candidate decisions in nominating politics—other potential contenders could know that these repeat entrants were potentially disruptive; they just could not know what they would do in any given year—then the least conditional contributions came from sitting presidents. Across time, most sought renomination, meaning that all other potential contenders had to begin with an estimate of the likely fortunes of the incumbent President (Table 6.1). Two sitting presidents did actually commit to a single term as part of their initial nomination, James K. Polk in 1844 and Rutherford B. Hayes in 1876; both honored that commitment. One other, Andrew Johnson in 1868, ultimately sought the nomination of the other party, albeit unsuccessfully. Everyone else sought renomination and most achieved it, though it was the pure convention era, not the plebiscitary era, which produced the exceptions this time:

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

Table 6.1 Nominating politics among incumbent presidents seeking reelection

Institutional era Pure conventions Mixed system Plebiscitary system

199

Success

Failure

8 10 7

5 0 0

• The Whigs in 1840, in the process of coalescing as a national opposition, created a unity ticket, combining William Henry Harrison for president, a committed Whig, with John Tyler for vice president, a dissident Democrat. When Harrison died and Tyler ascended to the presidency, he dissented so thoroughly from his new nominal party that they expelled him from membership.7 • In 1852, the party had a genuine Whig as incumbent president in Millard Fillmore, though he too had ascended to the presidency on the death of the original candidate, Zachary Taylor. Yet rising tensions from issues involving slavery and statehood had become too hard for a sitting president to straddle, and Fillmore was displaced as nominee by General Winfield Scott at the 1852 Whig convention. • For the Democrats in that same year, Franklin Pierce became the nominee—and then president—as only the second true dark horse.8 But four years later, he had acquired a record on the same poisonous disputes over slavery and statehood that had doomed Fillmore, and he too could not reconstruct a nominating majority from a factionalized party that needed desperately to avoid those poisonous issues. • Another unity ticket, for the nascent Republicans this time, appeared in 1864, when Abraham Lincoln as the Republican nominee for president took Andrew Johnson, a dissident Democrat, as his vice president. Lincoln was to die in office; Johnson was to ascend to the presidency; and his dissent on Reconstruction would put him so clearly at odds with his nominal party that Johnson actually stood for renomination at the Democratic convention in 1868 before his support ebbed away. • The only other sitting president to try and fail at renomination was Chester A. Arthur, vice presidential nominee in 1880 on a Republican ticket with President James A. Garfield, but once again undone by the curse of being a ticket-balancer. Arthur too ascended to the presidency on the death of Garfield, but in 1884 much of

200

B. E. SHAFER AND E. M. SAWYER

the party preferred James G. Blaine, leaving Arthur—too regular for the reformers, too reformist for the regulars—with a substantial base in the skeletal southern Republican delegations but without the possibility of renomination. • The other incumbent presidents who ascended to the presidency on the death of the president—Theodore Roosevelt in 1904, Calvin Coolidge in 1924, Harry Truman in 1948, and Lyndon Johnson in 1964, along with Gerald Ford, who ascended on the resignation of Richard Nixon—would all content themselves with completing the terms of their predecessors and then seeking a further full term of their own. In that sense, they were never denied renomination. All but Ford secured that further term, though the prospects of Truman and Johnson would have been cloudy had they sought a second full term.9 Said differently, after the pure convention era, incumbent presidents were always renominated if they sought to be so. The demands of party formation and the stresses associated with what became the Civil War did produce multiple exceptions in that first institutional period. Though even these always coincided with the death of the president who had actually been elected. No elected incumbent standing for renomination was denied that honor even during this first formative era. In any case, even these exceptions would disappear by the successor period, the mixed system of presidential nominations, and they would stay gone in the modern period, the plebiscitary era. Those are the messages of Table 6.1. How to Run Successfully renominated presidents did not always go on to win the general election. Martin Van Buren in 1840, Grover Cleveland in 1888, Herbert Hoover in 1932, and Jimmy Carter in 1980 all failed in November after succeeding in their respective conventions. Yet more to the point here, bracketing the renominations of sitting presidents highlights a much more varied story of presidential nominating struggles and results among the much larger cohort of nominees who remain. Though again, these individual stories too present some common overall patterns when further arrayed according to the three great institutional periods of presidential nomination. So Table 6.2 goes on to underline the very real differences by era when there was no sitting president.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

201

Table 6.2 Nominating politics among nonincumbents

Pure conventions Mixed system Plebiscitary system

First ballot

Front runner

Early entrant

Late entrant

14 10 15

8 4 0

4 4 0

6 2 0

Seen this way, it becomes clear that in the two earlier periods, containing the bulk of American political history, the majority of presidential nominations were not confirmed on the opening ballot. They required successive ballots, sometimes realizing the purported lead of the apparent front-runner on the opening ballot, sometimes elevating an entirely different candidate who at least had serious support when the convention opened, and sometimes finding an ultimate nominee who was not even in evidence when the opening ballot took place. Only the modern era would see all nominations without an incumbent president effectively realized in advance of the convention, that is, within the process of delegate selection. In fact, no convention in either party during this period was to require so much as a second ballot to confirm the result of that prior process. Yet there were further strategic calculations specific to each period. In the pure convention era, the further question for a strategic presidential aspirant was whether to attempt to be a serious presence on the opening ballot or to adopt a strategy of being “available” but not demonstrably active. Under the mixed system, not otherwise very distinctive from the first era in terms of actual outcomes, an additional strategic question arose—whether to enter the presidential primaries—and this became a further calculation piled on top of the continuing choice about a general level of activity in seeking the nomination. By contrast in the plebiscitary period, where waiting to be recognized for your “availability” disappeared as an option and where mandatory primaries were effectively ubiquitous, the sole remaining strategic choice boiled down to entering one, the other, or both of two opening contests In the pure convention period, the first three candidates who were not even put in nomination on the opening ballot but ultimately ended up as the nominee were all Democrats, though none of them featured much in the way of a calculated strategy:

202

B. E. SHAFER AND E. M. SAWYER

• James K. Polk, the inaugural dark horse in 1844, did not appear until the eighth ballot but was nominated on the ninth. A former state official in Tennessee, his nomination was largely a response to the continuing deadlock between Martin Van Buren, former president, and Lewis Cass, Senator from Michigan. What made him attractive— anonymity plus a reputation for partisan reliability—could hardly qualify as a “strategy.” • Franklin Pierce, the second dark horse, made Polk look positively ambitious, by not appearing among the candidates at the 1852 convention until the thirty-fifth ballot, when Virginia (and not his home state of New Hampshire) put his name in play. A former Congressman and former Senator, Pierce’s attraction likewise lay in the lack of a policy profile in a conflictual convention with four major contenders. There was again no evidence of strategy on his part, and little room for it. • Horatio Seymour, the Democratic nominee in 1868, resembled Pierce in that he did not appear until the twenty-second ballot and was put in play not by his home state (New York) but by Ohio, while New York was actively backing Thomas Hendricks of Indiana. The attraction of Seymour came most critically from his behavior over the previous days as permanent chairman of the convention, where an even-handedness in earlier struggles among four major (and intentional) candidates made him attractive as a compromise. • While he was a first ballot nomination, the 1872 Democratic choice, Horace Greeley, former editor of the New York Tribune, was perhaps the least deliberately strategic of these four. Greeley was already the nominee of the Liberal Republican Party, one of the short-lived responses to opposition party dilemmas in the aftermath of the Civil War; he was a frequent critic of the Democratic Party in the pages of his newspaper; yet he went on to be rubber-stamped by a six-hour Democratic convention. After Greeley, the next three presidential nominees who were not initial entrants in a nominating process were all Republicans—Rutherford B. Hayes in 1876, James A. Garfield in 1880, and Benjamin Harrison in 1888—and all were products of the internal divisions within the dominant party into three continuing factions:

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

203

• The 1876 Republican convention featured two champions for the regulars, along with the indomitable James G. Blaine for the moderates, plus a further champion for the reformers. Governor Rutherford B. Hayes of Ohio was present from the start, but only as a favorite son. He began to move on the fifth ballot when Michigan shifted to him, and he went over the top on the seventh, when both New York and Pennsylvania made further major contributions. So Hayes clearly was a strategic marker for Ohio, and must have been at least agreeable to being one. • Hayes having died, the leading contender in 1880 was former President Ulysses S. Grant, as opposed by the redoubtable Blaine and by Treasury Secretary John Sherman of Ohio, nominated by fellow Ohioan James A. Garfield. The latter did not come to life as a presence in his own right until the thirty-fifth ballot, and even then only through support from Indiana and Wisconsin rather than his home state. Yet Ohio did lead a plunge to him on the next ballot, leading to his nomination, all of which looked less like a strategy and more like a pure opportunity of the moment. • Eight years later, after the successful nomination but electoral defeat of Blaine in 1884, fourteen candidates were proposed at the Republican convention of 1888. The eventual nominee, former Senator Benjamin Harrison of Indiana, did have the support of his home state, though he was fifth among fourteen on the opening ballot. A “hidden” Harrison vote then proved to be scattered among delegates holding out for Blaine: they settled on Harrison on the seventh ballot, and he went over the top on the eighth. So while the nominee was a favorite son, he could not conceivably have gone into the convention with the strategy of hoping that Blaine would have substantial support but would not enter, and that disappointed Blaine delegates would therefore turn to him en masse. The strategic repertoire for active nomination contenders was broadened with the arrival of the mixed system of delegate selection after 1908, since candidates now needed a conscious decision, not just about running for president and how actively to do so, but about adding one or more of what was a growing number of presidential primaries.10 Because most of these were preference polls only, allowing no connection between expressed preferences for a nominee and the election of actual delegates, candidates generally chose not to enter. Still, there was an enhanced

204

B. E. SHAFER AND E. M. SAWYER

range of options. At the extreme, a candidate could go all-in, entering as many contests as possible. Alternatively, a candidate could enter a carefully selected subset of contests, ideally demonstrating voter attraction without risking repeat losses. Or the candidate could rest content with a home-state primary, if available, aiming to confirm an electoral attraction for the general election. The strategy of going all-in attracted a peculiarly bifurcated clientele, made up of those who had no other hope of gaining the nomination and those who wanted just some final confirmation of what appeared to be an incipient coronation. Among those who really had little other choice, former president Theodore Roosevelt turned to this strategy in 1912, since sitting President Taft could be expected to dominate the regular party and its internal selection processes. In a different manner but with the same bleak alternative prospects, Senator Estes Kefauver chose the same strategy among the Democrats in 1952, since as an investigator of labor corruption through his Senate committee, Kefauver had no hope for autonomous support from the regular party and its main organized interest groups. In the end, both men gained the same ultimately unproductive outcome, winning significant primaries, receiving fewer delegates from them than their vote would have suggested, and losing the nomination at the convention to candidates who had largely avoided the primary trail. On the other hand, Roosevelt in 1912 did trigger the only relegation of a major-party nominee to third place at the general election, and Kefauver in 1952 did force Adlai Stevenson to answer a second challenge from him in 1956, where Stevenson, who had consciously avoided the primary trail in 1952, went out on it in 1956 as the price of dispatching Kefauver. Occasionally, two leading contenders, under less immediate compulsion, chose the full primary trail in a conscious effort to diminish the apparent attraction of the other. Thus Woodrow Wilson and Champ Clark entered most Democratic presidential primaries in 1912, though the result was largely a wash, and the long convention battle afterward was little influenced by these preceding contests. Dwight Eisenhower and Robert A. Taft produced much the same result from the Republican primaries in 1952: entered in most, splitting them into roughly equal parts, and seeing a short but intense convention turn on things other than preference expressions at the ballot box.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

205

This weakness of the practical (which is to say, the delegate) side of presidential primaries was perhaps best demonstrated by William Gibbs McAdoo, former Secretary of the Treasury, in the Democratic primaries of 1924. McAdoo won almost all of the preference polls inside these contests, but he and Governor Alfred E. Smith of New York, who won no primaries at all, were to produce the most protracted convention in American party history—17 days and 104 ballots—before both lost to John W. Davis, former Solicitor General. Davis began essentially as a favorite son for West Virginia, saw his support grow through the thirty-fifth ballot, saw it decline through the ninety-fourth ballot, moved into the lead on the hundred-and-second ballot, and was nominated one ballot later. Despite those hugely diverse roles for the presidential primary— compulsive pursuit, close division, or complete irrelevance—the most common function of the primaries was to reassure the rest of the party about the electoral potential of an apparent front-runner. Thus Calvin Coolidge, having ascended to the presidency on the death of Warren Harding, put party anxieties to rest by entering and winning the Republican preference polls in 1924. Herbert Hoover did the same with his emergent candidacy to succeed Coolidge in 1928. By contrast, their predecessor, Warren Harding, had contented himself in 1920 with winning his home-state primary in Ohio, though it was seriously contested, a result that came back to reassure supporters when he became an unforeseen compromise at an old-fashioned brokered convention. So those were some successful cases of an essentially confirmatory strategy using the presidential primary trail. Yet even in this context and even for candidates who were not forced to undertake it and could choose their venues carefully, the strategy was hardly foolproof. Thus Governor Frank Lowden of Illinois, who did manage to stand second on the opening ballot at the 1920 Republican convention, had aspired to be the champion not just of Midwestern states but of party regulars generally. But while he did win the preference poll in his home state of Illinois, nearly uniform losses in a number of other states damaged him badly, with the result that expected support in other large organization states was divided and either arrived late or never arrived at all as those states stayed with favorite sons.11 Despite these multiple and varied examples of life on the primary trail, the majority of all nominating contests during the entire period of the mixed system of delegate selection were not seriously contested in the

206

B. E. SHAFER AND E. M. SAWYER

primaries. Though this reticence did have a further partisan cast: Republican presidential aspirants traveled the primary route much more often than Democratic counterparts. Republicans appeared to see the value in demonstrating electoral clout. Democrats, implicitly reinforced by the additional demands from the need for a two-thirds majority, were more likely to see the down-side from primary losses. Yet at the end of the day and finally, only two contests truly turned on presidential primary outcomes, and these came right at the end of the era: the Democratic nomination of John Kennedy in 1960 and the Republican nomination of Barry Goldwater in 1964. Kennedy, the private favorite of much of the regular party in the runup to the 1960 Democratic convention, carried major potential liabilities that were widely recognized by these same regulars. His comparative youth and inexperience were on this list; so, especially, was his Catholicism. So he needed to find some way to demonstrate that these were not serious electoral liabilities. Kennedy found the solution by defeating Hubert Humphrey, a serious, more senior, Protestant contender, first in Wisconsin and then in West Virginia, the latter believed to be hostile in principle to a Catholic candidate. In the aftermath of those wins and when none of the other contenders—Senator Lyndon Johnson of Texas, Senator Stuart Symington of Missouri, former Governor Adlai Stevenson of Illinois—were willing to take the same risks on the primary trail, Kennedy proceeded to an early nomination at the convention itself.12 The Republican counterpart arrived four years later, when Senator Barry Goldwater of Arizona, champion of the activist wing of his party, needed a means to pry the nomination away from the great regular parties of the northeast, more moderate in ideology and more reliant on internal party structures to generate their delegates. The primaries would confirm Governor Nelson Rockefeller of New York as the electoral champion of that big, regular, northeastern wing, but in a head-to-head contest in the giant winner-take-all primary in California, Goldwater was to defeat Rockefeller, narrowly in percentage terms but massively in the associated delegate result. The convention itself would still be conflictual, with efforts to undo the emergent Goldwater majority. But California would prove to have been definitive, and a Goldwater majority going into the convention was not undone.13

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

207

So both Kennedy and Goldwater owed their nominations in part to the availability of a primary trail. That said, the mixed system of delegate selection still allowed for many contests in the same mode as the pure convention system, where the front-runner did not triumph at the opening of the convention and the eventual nominee arose from inside the ensuing struggle. None of these were as dramatic as the instances in an older world where the eventual winner was not even nominated when the balloting began, as with James K. Polk for the Democrats in 1844 or James A. Garfield for the Republicans in 1880. Yet the point remains that if the newer world of mixed selection systems brought some true products of the primary trail, it also featured many contests on the old model, where an initial contender who was not a pre-convention leader eventually collected the nomination. Wendell Willkie for the Republicans in 1940 was the most striking example, having held no prior public office and having entered no presidential primaries. But this account has already touched on numerous others: Warren Harding for the Republicans in 1920, entering only his home-state primary in Ohio and standing sixth among the initially nominated contenders; Governor James M. Cox for the Democrats in that same year, likewise entering only Ohio and needing forty-four ballots to achieve his (two-thirds) majority; John W. Davis for the Democrats at the following convention, entering no presidential primaries and having to endure a hundred-ballot contest between two leading opponents; and Adlai E. Stevenson as late as 1952, entering no preference primaries outside his home state of Illinois and losing that one to Estes Kefauver, yet managing to wrest the nomination away from Kefauver on the third ballot. Where to Run The coming of the plebiscitary era after 1968 would put an end to all this by altering the strategic calculations of any and all presidential aspirants. Some fundamental things did not change. The Gain-Deficit Ratio would continue to measure the impact of the bandwagon dynamic within this reformed institutional framework. The preferred occupational backgrounds and career trajectories would continue down a path inaugurated in the mixed and not the plebiscitary system, with its continuing potential for governors and a growing prospect for senators. The main structural

208

B. E. SHAFER AND E. M. SAWYER

influences would change, especially the factional structure of the individual parties, but not in coordination with the arrival of a plebiscitary system. Rather, this key shift in the factional structure of the individual parties preceded, and in that sense was part of the impetus for, a reformed era of delegate selection, while a later change in these divisions would not be accompanied by any new institutional arrangements for presidential nomination. What would indisputably change was the near-universal implementation of state presidential primaries. Moreover, these were now to be candidate primaries, primaries that reliably linked the outcome of presidential preference polls to the creation of state delegations to the national party convention. The fall-out was to be immediate, generalized, and distinctive. Never again would the identity of the nominee be unknown when the convention assembled. Indeed, never again would eventual nominees enter the convention without already possessing their majority, though the Democratic contest of 1972, the Republican contest of 1976, and the Democratic contest of 1980 would sustain vain hopes that an obvious and ineluctable result might still be undone. After these opening years of the plebiscitary era, losers would be deprived even of this distant prospect.14 By extension, the strategic options of presidential aspirants were sharply constrained. No candidate could any longer decide to wait for entry until the convention itself: delegates not already acquired by this candidate were guaranteed to have been acquired by someone else. No candidate could even hope to lock up his home state and then sit tight. In the new institutional and behavioral milieu, such favorite sons became anomalous, and quickly disappeared. State voters who could choose among “real” presidential contenders would prove to be thoroughly resistant to a local candidate who could not demonstrate the prospect of an impending nomination, even if they knew said candidate from home-state politics. There would no longer be anything even as conditional as a delayed nomination of the front-runner. That said, there could still be conventions that were not unanimous on the first ballot. So losers might yet be put in nomination and enjoy their moment before the cameras.15 But the real nomination would always have been decided beforehand, to be merely confirmed on this ballot. The original question of whether to enter the nominating contest remained, of course, but only in the sense that anyone could decide not to run. Otherwise, the new world demanded early, active, and constant pursuit of the

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

209

nomination: no candidate who skipped more than the first two contests would ever again be nominated, at least as this is written. Nothing could banish death, disability, or disqualification as theoretical exceptions, but those were the only foreseeable conditions under which the winner from the primary trail would not prove to be the ultimate nominee—and they have never yet been realized. Along the way, the question of how to enter also disappeared. The one and only option was to contest the full range of delegate selection contests, most of them candidate primaries with a few participatory caucuses thrown in.16 The odd contest could be skipped for idiosyncratic reasons—Hillary Clinton did not campaign against Bernie Sanders in 2016 in his home state of Vermont—but those moves had to be both idiosyncratic and isolated. By subtraction, strategic considerations for the candidates were reduced to the question of when to enter, and in the modern world, even this had only three generic answers: via the Iowa caucuses, via the New Hampshire primary, or after both. Though in practice, even the last of these, while early enough in the abstract to collect a majority of delegates, would prove to be chimerical. The roots of this modern narrowing of strategic possibilities can be traced all the way back to New Hampshire in 1916, when that state initiated its presidential primary.17 The resulting contest was always early, becoming a local diversion in the long dark winter. But the most common result for both parties during the extended years of the mixed system was still a slate of delegates that was not publicly pledged to any national contender. This changed in the run-up to the 1952 nominations, when the state legislature amended its ballot rules to allow easier entry and, especially, to provide for an explicit candidate connection. Thereafter, New Hampshire became not just the temporal opening of the nominating contest but the first concrete indication of which candidates had votegetting prowess, a powerful lure for a mass media and a general public which reliably wanted to know what was “really happening.” That impact was magnified by the fact that this opening contest arrived, early and by itself, in a state still small enough to feature “retail” politics. What would become recognized as “wholesale” politics, relying principally on commercial advertising and supplemented by flying visits from the candidates, would become the norm for nominating politics once the delegate selection process reached the point where there were multiple primaries on the same day in larger states. But in a small state that was

210

B. E. SHAFER AND E. M. SAWYER

not sharing candidate attention on the day, retail politics, featuring inperson campaigning and direct candidate contact, remained possible. As other states woke up to the strategic requirements of a sequential plebiscitary system, the New Hampshire state legislature would take an additional step, setting the date of their contest by law at one week before the next-earliest primary, so as to maintain its initial character—as well as the attention from both the candidates and the media that went with it. Though the surprise of the plebiscitary era, one largely accidental, was the emergence of the Iowa caucuses as the other true opening contest in this new system.18 Iowa too had traditionally begun the process of selecting its delegates early. But with a multistage process spread over a number of subsequent months and without any candidate implications that could be readily calculated, the state had remained a backwater in candidate, media, and public attention. Yet the same reforms that were to impel an overwhelming sequence of candidate primaries also required the remaining caucus states to register presidential preferences and tie them directly to delegate selection. Iowa not only complied but actively facilitated the ability of commentators to see what had just happened at the local level and to project it through the subsequent process. And a second, regular, opening test of candidate prospects had been born. Most candidates thereafter began in Iowa and moved on to New Hampshire. A few, expecting to do much better in the latter than in the former, skipped Iowa and waited to open their campaigns in the New Hampshire primary. A few others, with reverse expectations, opened in the Iowa caucuses but did not invest seriously in New Hampshire. Over the coming years, a variety of states would jockey to be third, and an occasional challenger would aspire to break into the top two. Yet these latter challengers reliably failed in the face of an accepted tradition plus explicit resistance by two states that were militant in defense of their privileged status, while no state that went third, fourth, or later ever came close to achieving the continuing impact of these first two. In effect, then, the strategic defaults became: Iowa-to-New-Hampshire, Iowa-but-not-New-Hampshire, or New Hampshire-but-not-Iowa. Table 6.3 compares the pay-off from these strategies, results that can in turn be made to tell a set of lesser stories, mixing the possible strategies and outcomes:

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

211

Table 6.3 Iowa, New Hampshire, and the nomination: 1976–2016 Both parties Outcome 1st 1st 1st 1st 1st

in IO + NH in IO in NH in IO or NH or 2nd in IO or NH

Dems only

Reps only

Nom’d

Not

Nom’d

Notv

Nom’d

Not

5 12 11 17 18

0 8 8 2 1

3 6 5 8 9

0 4 5 2 1

2 6 6 9 9

0 4 3 0 0

• At one extreme, winning in Iowa and New Hampshire (1st in IO + NH) was tantamount to nomination. However long they managed to stagger onward, all other candidates were dead after New Hampshire under these conditions, and the contest was effectively over. Two opening wins was, however, an uncommon result, perhaps because Iowa and New Hampshire were so very different demographically, perhaps just because participatory caucuses tend to sample their electorates so very differently from primary elections. • At the other extreme, winning in Iowa (1st in IO) or New Hampshire (1st in NH) were not by themselves especially portentous, given that these were reliably the two contests with the largest number of contestants, that is, the largest number that would prove to be irrelevant. A candidate who won one or the other was more likely to be nominated than a candidate who did not win either, yet there were so many more of the latter than the former, given the usual raft of aspirants who would prove to be going nowhere, that a single win was not all that strong a prediction of ultimate nomination. • Though combing the two together, that is, winning one or the other by comparison to losing both, was again portentous. The collection of Democrats and Republicans who won at least one of the two (1st in IO or NH) was overwhelmingly likely to contain the ultimate nominee. An early win more or less automatically improved name recognition while simultaneously suggesting electability. At the same time, it created a veritable field of losers, many of whom had invested all their resources in these opening contests, to the point where most would never go on to win anything. Anyone who then demonstrated an inability to win another contest quickly—even though

212

B. E. SHAFER AND E. M. SAWYER

there were more than forty yet to come—would have a very difficult time attracting fresh activists, fresh donors, and fresh voters by comparison to the candidates who were in fact winning. • Before Joe Biden captured the 2020 Democratic nomination, it could also have been said that no candidate who had not come in first or second in Iowa or New Hampshire (1st or 2nd in IO or NH) had ever been nominated in the plebiscitary era. In that limited sense, the Democratic contest of 2020 was precedent-breaking. The critical contest, the South Carolina primary, actually did come after Iowa and New Hampshire, while the winner, Joe Biden, had not placed either first or second in these prior contests. Yet even Biden could not break the apparent behavioral rule that every ultimate winner had to have contested one or the other of those first two delegate selection events. Rudolph Giuliani, former Mayor of New York City, did produce the essential ultimate contrary test. Leading in the polls in 2011 but lagging in both Iowa and New Hampshire, Giuliani gambled on waiting until the Florida primary, fifth in the sequence in 2012, where he was leading when his campaign decided to pass on the first two contests. Yet the interim was to see John McCain, Senator from Arizona, win the New Hampshire primary and Mitt Romney, former Governor of Massachusetts, place second in both New Hampshire and Iowa. So twenty-six days after Iowa and twenty-one days after New Hampshire, Giuliani found himself opening his campaign in the face of the apparent front-runner in McCain and the anointed challenger in Romney, while he himself was to finish an undistinguished third in Florida, garnering a vote less than half that of either of those others.

Strategic Maneuvers by State Parties Those are the key strategic decisions facing—and inevitably made by— the individual aspirants for a presidential nomination across time. Some of these aspirants had little choice in the matter: individual decisions were more or less made for them by their own inherent constraints. Others had more leeway, being better-connected or better-resourced, though by the coming of the plebiscitary era, even the best-off had little ability to reshape the overall course of nominating politics on their own. Yet there was another set of strategic decisions that interacted with these candidate

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

213

calculations from the other side, a set that belonged not to the presidential aspirants, individually or as a cohort, but rather to the individual states and their state parties. As decisions, theirs were more episodic than those available to the candidates, but simultaneously more fundamental in that they tended to be larger in their reach and more lasting in their impact. These state choices put the details into the specific course of a nominating conflict, and even when the national parties were willing to impose some nationwide standards and restrictions, much remained largely the province of the states. Their decisions involved state-based choices for the institutions of delegate selection, state-based systems for delegate allocation, and state-based efforts to maximize their own influence within the unfolding collective sequence of delegate selection contests. In turn, state decisions inevitably interacted with the strategic choices made by individual candidates. It is that sense in which the state side of nominating politics fills in the specifics of any given contest, to the point where the two influences together—individual campaign strategies coupled with state party adjustments—sometimes went so far as to provide the final shove favoring one aspirant over another. These alternative decisions for the states and their parties once again differed by institutional era. In fact, one major instance—one major strategic choice and hence one key process—characterizes each temporal period. Under the pure convention system, in the years from 1832 to 1908, the main option available to the individual states involved applying the unit rule to their respective delegations or not. During the mixed system for delegate selection, in the years from 1912 through 1968, the main option was instead whether to adopt a presidential primary or stay with a traditional convention, and how to structure a primary if they made that choice. And in the modern plebiscitary era, 1972–2020 and counting, the major remaining choice was where to situate their state in the long chain of presidential primaries and occasional party caucuses through which nominating politics ran—and where the combination of these state choices was what produced the actual sequence of delegate contests.

214

B. E. SHAFER AND E. M. SAWYER

Imposition of a Unit Rule The pure convention system featured a general uniformity in the intraparty mechanics of delegate selection across the states. Still, there were some further but lesser options in the detailed workings of selection politics among the individual parties in this long opening period. Thus states differed in the locus of ultimate selection. Some used regional party gatherings plus a further increment from the state convention; others assigned creation of the full delegation to the state party in convention; and some went on to reserve the option of selecting the full delegation to the state central committee, without so much as a formal convention. Yet the main option for state parties in this opening era was whether to apply a unit rule to any given result. States with small delegations tended not to do so. From one side, interpersonal relations could handle the proceedings of the delegation. From the other side, those relations could be actively damaged by compelling the behavior of colleagues who were personal acquaintances and whose coordinated contributions might well be needed for other party business. So the states that did use this further power of constriction tended to fall into two categories: • First were the big states with classic machine-type parties, found mainly in the northeast. More than the others, these states arrived at the national party convention with the conscious—and apparently plausible—intention of maximizing their influence over the ultimate result. The unit rule was one more potential tool in pursuit of that goal. • The other reliable users were the southern states. Their results in November were known in advance, for which southern Democrats wished to be rewarded and southern Republicans excused. Yet they also had policy concerns specific to the region, and both Democratic and Republican parties could hope to use the unit rule to pursue these. But how much did it matter? That is, how good were either of these state clusters at securing the outcomes they desired—or desired to avoid? And more consequentially for nominating politics, how much did this record of success or failure alter an ongoing bandwagon dynamic, the shifting backgrounds of nominees, or the influence of party factions?

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

215

In the case of the major northeastern states, the crucial test is perforce focused on the three outsized examples, New York, Pennsylvania, and Ohio. Though among Republicans, a tour of national party conventions in the pure convention era suggests that practical opportunities to apply the unit rule were more limited than its theoretical availability would suggest. Before 1876, the Republican Party would in principle tolerate imposition of the unit rule convention-wide; after 1876, implementation was purely a state matter. Yet there was no uniform imposition for the three major states—and no uniform success when the rule was imposed. New York did apply the rule in 1860 at the first serious Republican convention, yet in doing so it stayed with Senator William H. Seward, its favorite son, even through the final ballot when the vast bulk of the others shifted to Abraham Lincoln. New York had a favorite son and stayed with him again in 1876, until Rutherford B. Hayes became the only remaining option to James G. Blaine. After that, the state was normally so divided as to make imposition impossible. New York did move early as a bloc to Benjamin Harrison, the ultimate nominee, in 1888. Yet it was also the lone major hold-out against Harrison when he was renominated in 1892. At the opposite extreme, Ohio was equally devoted to its favorite sons, but more successful in using procedural weapons to support them. The Buckeye state possessed a favorite son in 1860 but nevertheless broke to Lincoln on the third ballot, a break that was important to his nomination. Ohio stayed with its favorite son in 1876, Governor Rutherford B. Hayes, which gave him support sufficient to survive until the convention was ready to seek a compromise nominee. Ohio likewise stayed with Treasury Secretary John Sherman, yet another favorite son, against the Grant renomination effort in 1880, stalling the Grant drive until a second Ohioan, Congressman James A. Garfield, came into play. It lost this uniform winning record in 1888, when it stayed with another favorite son until the nomination of Benjamin Harrison of Indiana was an inescapable result. And it came back to stay with home-state Governor William McKinley in 1892, even as Harrison was overwhelmingly renominated. Lastly, Pennsylvania had an even longer but less imposing roster of favorite sons across this period, yet it ended up with the winner, James G. Blaine—not a favorite son—only in 1876. Otherwise, imposition of the unit rule on behalf of also-rans from Pennsylvania resulted in a reliable trip to the political wilderness. For the three states as a group, then, there was only intermittent application of the rule. Yet when it could be

216

B. E. SHAFER AND E. M. SAWYER

imposed, the record of strategic pay-offs was spotty at best. New York had the hardest time of the three in imposing the rule at all. Pennsylvania had the easiest time but the worst pay-off. So only Ohio wielded this strategic power with real, albeit still intermittent, impact on the eventual nomination. In theory, the skeletal Republican parties of the south during this period might have resurrected the rule on a regional basis. But in practice, the southern states were often split in their judgments of which candidate would best provide the essential federal patronage that kept these skeletal parties alive, while individual delegates were further split over their prospective personal advantages. So imposition of the unit rule was comparatively uncommon. The exceptions were 1880, where the possible return of Grant held out the hope of bringing back a patronage heyday for regional Republicans, and 1884, when these states provided the essential base, albeit another unsuccessful one, for the effort by President Chester Arthur to secure renomination in his own right. On the other hand, the unit rule was always more of a favorite among Democratic parties, where its clout was potentially magnified by the need for a two-thirds nominating majority. In the antebellum years, the bigthree states were unanimous for Van Buren in his attempt at a third nomination in 1844, before all three broke to Polk, the inaugural dark horse. The three were again united in imposing the rule for 1848, but two of them, New York and Pennsylvania, went with favorite sons and not with Lewis Cass, the nominee. Ohio and New York were both split in 1852 and 1856, while Pennsylvania applied the unit rule on behalf of its favorite son, James Buchanan, unsuccessfully in 1852 but successfully in 1856. Conversely, Ohio and New York applied the rule in 1860 on behalf of Stephen Douglas, the ultimate nominee, while Pennsylvania was the one that was split throughout. So the big northern states were more often able to apply the unit rule in this period among Democrats as opposed to Republicans, though their record of success in return was equally spotty. After the Civil War, the main interparty decisions for the Democrats involved the attempt to find a nominee who might restore the national party to competitive status, which quickly focused convention politicking on the fortunes of any available reform governor from New York. Yet as consensual—and successful—as that particular strategic judgment turned out to be, the bigthree states, with presumably a great deal to gain, repeatedly splintered in response.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

217

A unanimous Ohio delegation in 1872 contributed the major shift on the twenty-second ballot that started the bandwagon for Horatio Seymour, former reform governor of New York. In 1876, it was instead New York that delivered a unanimous vote for Samuel J. Tilden, sitting reform governor, while both Ohio and Pennsylvania sat with their own favorite sons on the critical opening ballot. In 1880, the three went their separate ways, with New York floating a pair of possible successors to Tilden, Ohio remaining behind its favorite sons, and Pennsylvania split and marginal to the nomination, despite the fact that Winfield Scott Hancock, the eventual nominee, was a Pennsylvanian. New York again turned to the unit rule in 1884 to support Grover Cleveland, while Ohio was split and Pennsylvania opened with a favorite son, before both became part of a general move to Cleveland on the second ballot. The Democratic convention of 1896 registered the first coming of the populist revolution to the national Democratic Party. This time, a strong dissenting view from regular parties in the northeast actually unified the big-three states. Yet imposition of the unit rule on top of a common cause would still leave all three on the losing side for a generation. All three big-state parties were again in vigorous opposition as the populist impulse steamrolled the Democratic convention of 1896, with both New York and Pennsylvania bootlessly invoking the unit rule first in support of the gold standard and then against William Jennings Bryan as the nominee. Bryan was so evidently unanimous as the (re)nominating choice in 1900 that none of the big-three bothered with an even more bootless opposition. Though when the opportunity returned to vote against Bryan at the 1904 Democratic convention on procedural issues and against populism via alternative champions, New York, Pennsylvania, and Ohio all stayed with the unit rule and with New York Justice Alton Parker, the lone traditional conservative during the 1896–1908 stretch of Bryan nominations. By contrast, most southern states had an easy time imposing the unit rule in these antebellum years, where an underlying homogeneity as a regional bloc could be focused on the one particular candidate who was ordinarily judged most sympathetic to the region as a whole. Lewis Cass in 1844 and 1848, a classic northerner with southern sympathies, was easy. The southern states remained en bloc in support of Franklin Pierce, rewarded for his regional neutrality as president in his unsuccessful renomination bid in 1856. And by 1860, the real question was neither the unit rule nor regional unity, but simple southern continuation in the national Democratic Party, as Alabama, Arkansas, Florida, Louisiana,

218

B. E. SHAFER AND E. M. SAWYER

Mississippi, South Carolina, Texas, and Virginia each walked out at one or more points during that extended and disastrous convention.19 After the war, the southern states, more isolated than ever as a region, were often content with a reform governor from New York as the way to appease three main party factions: southern loyalists, northeastern immigrants, and western populists. Yet they usually added a wrinkle to their intraparty strategy by rallying to an explicitly southern champion first, before moving to an ultimate compromise. Thomas F. Bayard, Senator from Delaware, was the most common vehicle for this, and he attracted a small, largely southern vote at the 1876 convention. He reappeared as the more conscious instrument of a “southern strategy” in 1880, when Bayard placed second to Grover Cleveland on the opening ballot. And he appeared once more, standing second on the opening ballot in 1888, before the southern states came back into line behind Cleveland.20 That particular strategy expired with the convention of 1896, when the rise of populism and the accompanying candidacies of William Jennings Bryan proved comfortable to these southern Democratic parties. The unit rule continued to be generally imposed across the region, but a generalized regional preference for candidate Bryan meant that this had little separable impact on the bandwagon dynamic, candidate backgrounds, or major structural influences. In sum, the south as a region was even more likely to impose the unit rule than were its big northeastern brethren, but it was harder to separate the unifying effect of this imposition from the unifying impact of the search for a candidate who would protect or advance common southern interests. Adoption of a Primary The arrival of the presidential primary in 1912 was to be the opening move in what would become two, not just one, extended institutional regimes for delegate selection and presidential nomination: • The first of these became the mixed system of delegate selection, stretching through the half-century from 1912 to 1968. It was “mixed” because state parties received their first major institutional option since creation of the national party convention itself in the 1830s. In turn, some states did indeed avail themselves of the choice, though most were to stay with the traditional state party convention.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

219

• A second institutional regime displaced the mixed system after 1968, morphing quickly into the modern plebiscitary system and itself extending through the next half-century. In principle, states could choose between a reformed presidential primary or a reformed state convention. But in practice, the primary would become so dominant as to justify calling the result a plebiscitary era. In the beginning, however, the states gained a major strategic option, most aspects of which would actually be taken away in the successor period. Yet during what became the long mixed period for delegate selection, the appearance of this new institutional option was to have two general but contrary effects on state strategies. From one side, state parties could now choose a whole different mechanism for delegate selection, while simultaneously tickling its details to suit local preferences.21 Yet from the other side, the practical impact of new institutional arrangements would remain limited and intermittent. Most states in fact experienced no change at all: the bulk of national convention delegates would continue to be chosen by way of established—internal—party machinery through 1968. Still, the mixed system of presidential nominations had arrived, courtesy of those states that did move to a presidential primary, and the minority that did adopt the new primary was large enough to make the composite institutional matrix look different. Within its boundaries, states could choose not just between the new primary or the traditional convention to create their national convention delegates. Those who switched to a primary could choose additionally among three variants: one linking a presidential preference poll directly to delegate election, one offering both a preference poll and delegate election but without an automatic connection, and one focusing on direct election of the delegates, absent a presidential preference poll. A few states injected not just the letter but the spirit of nominating reform into their institutional processes. This implied not just a presidential preference poll on the public ballot, but a link between this poll and the selection of individual delegates. Wisconsin was an early example, adopting this form of (true) presidential primary at its first opportunity in 1912 and retaining it through 1968. Though Wisconsin was simultaneously good at reminding analysts—and presumably the candidates themselves—that primary voters did not have to opt for an obvious national contender, and that when they did so, there could still be

220

B. E. SHAFER AND E. M. SAWYER

substantial flexibility in the link between preference poll and delegate result. In fact, in a first test in 1912, Wisconsin Republicans went on to cast 73% of their newly empowered vote for a favorite son, Senator Robert M. LaFollette, Jr., apparently giving him all the delegates, though these delegates would ultimately abstain from voting on a presidential nominee at the convention itself. That same year, Wisconsin Democrats were to favor Woodrow Wilson over Champ Clark, giving him 56% of their vote but 77% of their delegates.22 Fifty years later, Wisconsin Republicans would give 80% of their preference votes and 100% of their delegates to Richard Nixon, while Wisconsin Democrats gave 56% of their preference votes along with 83% of their delegates to Eugene McCarthy. More states offered both a presidential preference poll and direct election of their delegates, but on a ballot that provided no automatic connection between the two. In principle, an aggressive nominating campaign on the part of a presidential contender could forge that connection. But for an out-state contender (which most were) who was not widely known in-state (as most were not), this was a serious operational challenge, made only more difficult by the fact that such campaigns were implicitly challenging the existing party organization. Pennsylvania was an early proponent of this system, adopting it in 1912 and continuing with it through 1968. In its initial round for the Republicans in 1912, Theodore Roosevelt gathered 60% of the preference poll in the Keystone state and 84% of state delegates, while Woodrow Wilson was the sole entrant in the Democratic preference poll and came away with 93% of the delegates. In the final round of this arrangement in 1968, Richard Nixon won the Republican primary on a write-in, scoring 60% of the vote, yet Nelson Rockefeller acquired 41 of the 64 state delegates, for 64% of the total. At the same time, Eugene McCarthy won 56% of the official Democratic preference poll, while Hubert Humphrey acquired 104 of 130 delegates, 80% of that total, despite having no one on the ballot who was publicly committed to him. Lastly, some states, including New York with the largest delegation during this entire period, dispensed with the preference poll and simply offered the names of all contenders for delegate status. Again in principle, individual candidates for national convention delegate were free to publicize their preferences, though that was easier said than done by any one individual. More commonly, voters chose delegates based on general

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

221

public recognition or ongoing party connection. New York stayed with this arrangement from the first moment of the mixed system in 1912 through its final moment, or in some sense past the final moment, since New York did not initially comply with Democratic reform rules passed for the 1972 contest. Absent a preference poll, it was only possible to say that William Howard Taft for the Republicans got 76 of the 90 New York delegates in 1912, while Champ Clark, thanks to imposition of the unit rule, received all 90 of its Democratic delegates. A half-century later, Nelson Rockefeller would acquire 88 of 90 delegates for the Republicans, while Hubert Humphrey would outgain Eugene McCarthy for the Democrats, 97 to 87, despite not announcing his campaign for president until the calendar deadline for putting delegate names on the ballot had passed. That institutional world would be worth more than a half-century, but would ultimately be swept away by the wide-ranging reforms of the 1970s, almost immediately in the case of the Democrats, more gradually in the case of the Republicans, yet swept away in both cases. The result was the plebiscitary era of presidential selection, beginning in 1972 and heading for another half-century as this is written. The new era would be characterized most simply by the overwhelming dominance of a sequence of presidential primary elections. Yet this overall dominance risked obscuring the real engines for the constriction of institutional choice inside the overall shift, engines that involved the type of presidential primary that a given state could utilize or the type of party caucus it could retain if it chose not to go with the drift toward primaries. The overall move away from internal convention-based systems and toward external primary-based systems was dominant, self-evident, and overwhelming23 (Table 6.4). For a very long time beforehand—really the entirety of the mixed system of delegate selection—there had been only individualistic further fluctuations in the adoption of the presidential primary. Primaries had burst on the scene in a major way in 1912, giving the succeeding era of nominating politics its “mixed” character. Yet from there through the end of the era in 1968, there was only random further variation: a primary added here, a primary subtracted there, but a long-run picture of impressive stability. Measured by the simple presence or absence of presidential primaries, 1968 looked more or less indistinguishable from 1912.24

222

B. E. SHAFER AND E. M. SAWYER

Table 6.4 Number of presidential primaries across time

1912 1936 1952 1968 1972 1980 2000 2016 2020

Dems

Reps

12 14 16 15 21 34 40 39 41

13 12 13 15 20 34 43 39 41

That was instantly changed by the first big round of Democratic reforms in the run-up to 1972, with an institutional fall-out reaching into both parties. So by comparison to the long run of institutional stasis over the previous fifty-six years, the scale of change was already noteworthy. Yet within two further iterations, that is, by 1980, about two-thirds of the state parties on both sides of the partisan aisle had gone to the primary. Moreover, this aggregate of states understated the dominance of the shift, since larger states were more likely to make the move. By 2020, about forty of the fifty states could be expected to use primaries rather than caucuses, selecting close to ninety percent of the delegates. Nevertheless, that still hid the larger half of the institutional impact that came with—but really created—a new plebiscitary era. For in fact, the nature of the presidential primaries inside this burgeoning primary gain, coupled with the nature of the party caucuses that remained among states that did not move to the primary, brought the larger share of the practical impact from overall change, rendering the composite result truly and strikingly different25 (Table 6.5). In the previous world of the mixed system, delegate primaries, those that elected their national convention delegates without an explicit presidential attachment, far outnumbered candidate primaries, the ones where presidential attachment produced the delegate result. Just as traditional conventions, with their essentially internal selection mechanics, solidly outnumbered participatory caucuses, where open public meetings began the process instead. The balance among these four institutional arrangements—along with an occasional selection limited to a state central committee—looked essentially as it had since 1936, the year the Democrats finally eliminated the two thirds rule, up through 1968, the effective end of the

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

223

Table 6.5 A full plebiscitary picture: The changing matrix of delegate selection Year

Committee selections (%)

A. The Democrats 2008 0 1976 0 1972 2 1968 13 1952 8 1936 8 B. The Republicans 2008 0 1976 1 1972 3 1968 5 1952 4 1936 4

Traditional conventions (%)

Delegate primaries (%)

Participatory caucuses (%)

Candidate primaries (%)

0 0 2 24 27 31

0 9 14 19 26 31

17 24 36 21 19 15

83 66 46 23 19 14

0 4 16 24 26 31

0 11 20 23 25 32

16 24 24 28 25 20

84 60 37 20 19 14

mixed system of delegate selection. Reconstitution was clearly underway by 1972 when, if the generic primary had not yet fully dominated the composite process, the new additions were all candidate primaries, and in both parties. By 1976, after only one more iteration, the delegate primary was nearly extinct, salvaged among Democrats only by the state of New York. Both the delegate primary and traditional party convention— along with those stray committee selections—would die off entirely within the decade, by which time the dominance of the candidate primary had nearly quadrupled among Democrats and more than quadrupled among Republicans. State parties that wished to maximize their individual clout in nominating politics were effectively left with two remaining options. The first involved the choice between a candidate primary and a participatory caucus, where the choice of the latter could be important to the individual state but was truly minor—with the great exception of Iowa—inside the larger process. The second involved the temporal placement of any individual contest, which requires a separate section below. Here, note merely that, as the lone recurrent option remaining to state parties in the modern era, it would be more frequently attempted.

224

B. E. SHAFER AND E. M. SAWYER

The first option, the choice between a candidate primary and a participatory caucus, was institutionally fundamental. The two institutions not only created their delegates in entirely different fashions; they were even linked to the general public through differing routes. That said, there was never much actual competition between the two. The tide of reform in the 1970s and after, propelled by a desire to increase public participation in party affairs, ran heavily in favor of candidate primaries rather than participatory caucuses.26 Primaries reliably involved far more individual citizens than caucuses, while potential voting publics were far less attracted by the argument that the selection process as a whole would be improved if they themselves were removed, or at least sharply reduced, by means of a participatory caucus. The institutional outcome was clear enough. Not only did the number of states turning to the primary continue on an upward path. The comparative size of the states utilizing primaries continued to rise as well (Table 6.6). Already by 1976, the states that had turned to candidate primaries for delegate selection contained more than two-thirds of all Electoral Votes; only Texas among the larger states had not joined the parade. By 2016, forty years later and deep in the plebiscitary era, this Electoral Vote imbalance, two to one at the start, had reached eight to one.27 No major states and very few of the middle-sized remained with the participatory caucus. So the caucus effectively relegated its remaining users to spectator status within nominating politics, again with the dramatic exception of Iowa. The lone compensation to the lesser rump of small states that still clung to the caucus option, beyond managing to keep delegate selection focused on party activists and out of the hands of their rank and file, was that the delegate reward per caucus participant was far greater than the delegate reward per primary voter28 (Table 6.7). Which is to say: it required far Table 6.6 State size by institutional form A. Primaries

B. Caucuses

Year

States

EVs

States

EVs

1976 2016

28 39

363 478

23 12

175 60

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

225

Table 6.7 Participation levels by institutional form in 2016 1. Primaries States

2. Caucuses Turnout

A. Democrats 39 28,872,000 B. Republicans 39 29,877,000

Dels.

Per

States

Turnout

Dels.

Per

3383

8534

12

858,000

455

1886

2995

9975

12

578,000

300

1927

more voters in a primary to earn a single delegate and far fewer participants in a caucus to accomplish the same thing: nearly five times as many for the Democrats and more than five times as many for the Republicans by 2016. Said the other way around, if caucus states had needed to produce the same number of participants per delegate as the primary states, then the Democratic caucus states would have been entitled not to 455 delegates in 2016 but to 103, while Republican caucus states would have seen their delegate total fall from 300 to 95.

Selection of a Date–State Strategies The major option remaining for state parties that were trying to maximize their nominating clout involved the choice of an individual date for what was now almost always a candidate primary. In principle, this choice provided substantial degrees of freedom, in that individual states could choose to go early, middle, or late within the sequence of state primaries as a whole, while calibrating this placement additionally by considering the clustering of the other states, though their temporal strategies were simultaneously in flux. Yet the cumulative impact of such moves, widely attempted and constantly adjusted though they were, would prove much smaller than the often-excited predictions of reformers, strategists, commentators, and participants. To begin with, the more a reformed process encouraged states to move their delegate selection contests, the more the simultaneous collective maneuvering that resulted tended to cancel out these individual efforts. More of this underwhelming return on placement strategies was due to the fact that one of the three theoretical options—going early— was largely foreclosed: the successful assertion of temporal primacy by Iowa among the caucuses and New Hampshire among the primaries

226

B. E. SHAFER AND E. M. SAWYER

had severely reduced the possible impact of going early on the part of everyone else. Nevada did increasingly succeed in asserting a right to be the second caucus, and South Carolina did increasingly succeed in asserting a parallel right to be the second primary.29 Yet the shaping influence of Iowa and Hampshire was so overwhelming in terms of campaign investment and media attention, and so reliably determinative in narrowing the field of presidential aspirants, that these next two states, even together, had little hope of a consistent comparable impact.30 Twenty-two aspirants qualified to participate in the televised Democratic debates before the first actual contest in 2020; exactly five were even vaguely plausible as nominees after Iowa and New Hampshire. Indeed, the impact of the two opening contests was ordinarily sufficient to shift the focus of nominating campaigns away from wins or losses in individual states, especially small ones, and toward the raw accumulation of delegate totals, which made large states or large clusters the central target for presidential aspirants. In the process, room for strategic maneuvers by way of the choice of a temporal location had effectively been compressed to “middle” versus “late,” though conversely, what was the middle of an evolving post-reform sequence would creep forward over time. Which is to say: a growing consensus on choosing middle rather than late would produce a generalized push toward “front-loading,” a pressure that only increased over time while simultaneously reducing any incentive to wait. Though at the start of the plebiscitary era and for some elections to come, the logic of going late retained some real attraction. The contests of 1972, the first ones after sweeping party reform, still featured a hybrid quality in this regard. From one side, most states that had previously possessed primaries kept them in the same temporal location but merely switched, as necessary, from delegate to candidate format. From the other side, states that were introducing a new primary tended to look at what appeared to be the available—that is, the open—slots within the existing calendar.31 The initial result for 1972 and the next two contests kept the largest single aggregate of delegates at the end of the selection process (Table 6.8 A and B). Though in truth, this was not so much an aggregation of multiple states as a register of strategic decisions by a few states with large delegations. For them, the logic still seemed simple: late contests could be the ones that chose among the serious remaining contenders and crowned a specific nominee. The 1984 nominating contests would break this mold, setting off what became a chain reaction of strategic choices, leading to a clearly different

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

227

Table 6.8 Clustering in the sequence of contests: The largest and secondlargest clusters by year

A. Democrats 1972 Largest Second 1976 Largest Second 1980 Largest Second 1984 Largest Second 1988 Largest Second 1992 Largest Second 2000 Largest Second 2004 Largest Second 2008 Largest Second 2012 – 2016 Largest Second 2020 Largest Second B. Republicans 1972 – 1976 Largest Second 1980 Largest Second 1984 – 1988 Largest Second 1992 –

Date

States

Delegates

Share of total (%)

June 6 April 25 June 8 April 6 June 3 March 25 March 13 June 5 March 8 June 7 March 10 June 2 March 7 March 14 March 2 February 3 February 5 March 4 – March 1 May 15 March 3 June 2

4 2 3 2 3 2 9 6 20 4 11 6 14 6 10 6 22 4 – 11 5 14 8

415 284 535 342 696 336 602 583 1483 547 1002 835 1535 671 1407 314 2064 287 – 1002 358 1336 500

27 19 18 11 21 10 15 15 38 14 24 20 36 16 33 7 48 7 – 22 15 34 13

– June 8 April 6 June 3 March 25 – March 8 June 7 –

– 3 2 9 2 – 15 3 –

– 331 199 428 158 – 730 265 –

– 15 9 21 8 – 32 12 –

(continued)

228

B. E. SHAFER AND E. M. SAWYER

Table 6.8 (continued)

1996 2000 2004 2008 2012 2016 2020

Largest Second Largest Second – Largest Second Largest Second Largest Second –

Date

States

Delegates

Share of total (%)

March 12 March 19 March 7 March 14 – February 5 March 4 March 6 June 7 March 1 May 15 –

7 4 11 6 – 21 4 11 5 11 5 –

383 229 554 341 – 1124 265 466 299 595 358 –

29 14 27 17 – 48 11 21 13 25 15 –

overall sequence of state contests and eventually to the much-debated front-loading (Table 6.8 A and B). For 1984, the immediate impetus was an attempt by southern states within the Democratic Party to coordinate their primaries in order to maximize regional influence over the identity of the Democratic nominee.32 On its own terms, this effort was not successful: only three states conformed to what was mooted as an ostensible region-wide strategy. Though in passing, their moves drew the attention of several non-southern states, so that the upshot, for the first time ever, was the single largest concentration of states and delegates in an earlier temporal location, March 13, 1984. The pattern would be sustained through 1988 and 1992, with the largest concentration of delegates available in early March and the secondlargest concentration available in early June. Though by 1988, the early concentration had become massively the largest concentration of primary delegates ever seen on one day in the selection process: 38% among the Democrats and 32% among the Republicans. The two nominating contests of 2000 then altered the second piece of these patterns, by moving the main secondary concentration of delegates away from the end of the process and far forward. In that year, Super Tuesday was on March 7 for both parties, but the second main clustering was now March 14, exactly one week later, taking with it any serious support for the strategy of going late.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

229

This secondary cluster then remained in March through 2008, but the main cluster continued to shift forward, getting to February 5 for both parties in 2008. Both national parties would eventually respond to this “calendar creep” by specifying limits within which states could and could not locate their primaries, on pain of having the credentials of their delegates rejected at the national party convention.33 In response, the Republicans in 2012 and the Democrats in 2016—the Democratic nomination in 2012 was uncontested—were to see Super Tuesday migrate back to early March, a bit later in contemporary terms but still very early by historical standards. That date continued to capture the largest single cluster of state parties, along with the largest single cluster of convention delegates, though these were below totals for previous years. In the process, the old secondary cluster of available delegates—toward the end but not at the very end this time—would be restored as well. Along the way, the evolution of this front-loading proved sufficient to produce a feedback effect in Iowa and New Hampshire, driving them to move the two benchmark “early” contests earlier and earlier (Table 6.9). New Hampshire was the lead actor here, shifting its primary in 1976 from the traditional mid-March location to late February, in sync with the arrival of the plebiscitary era; then to early February or late January from 2000, as front-loading took hold across the party; then to early January in 2008—before the national parties took control of the entire calendar. A new, national, and explicit counterattack on calendar creep Table 6.9 Opening contests: Iowa and New Hampshire in the plebiscitary era

1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 2016 2020

Iowa caucuses

New H. Primary

[January 24] January 19 January 21 February 20 February 8 February 10 February 12 January 24 January 19 January 3 January 3 February 1 February 3

March 11 February 24 February 26 February 28 February 16 February 18 February 20 February 1 January 27 January 8 January 10 February 9 February 11

230

B. E. SHAFER AND E. M. SAWYER

guaranteed that New Hampshire would not have to continue its migration forward in time in order to maintain its first-in-the-nation status, and New Hampshire settled back into early March in 2012. Iowa had made its original—and even more consequential—contribution to this process in 1976, the year it switched to revealing candidate preferences in what as a result became the opening contest in the entire nominating sequence. At the start, this put Iowa a full month ahead of New Hampshire, but from 1984 onward, under informal pressure from the national party to move its contest closer to the total sequence, it settled on being essentially a week—eight days—before the date of the New Hampshire primary. Thereafter, aping the positioning of New Hampshire, Iowa would move to early February, then to late January, then to early January—all the way forward to January 3. At that point, the same national reconstitution of 2012 that moved New Hampshire back a month would move Iowa back a month too, to early February. Selection of a Date: Strategic Effects But once again, how much did all such maneuvering matter? The coming of the plebiscitary era did reconstruct the institutional context for delegate selection, eliminating the traditional party convention along with the delegate primary, while replacing them with a sequence dominated overwhelmingly by candidate primaries. And inside that sequence, the strategic gambits of individual state parties went on to rearrange the ordering of these selection contests and, even more crucially, its clustering. Yet Chapter One confirmed that a bandwagon dynamic with its diagnostic measure, the Gain-Deficit Ratio, had characterized nominating politics since the 1830s, burning across three sharply different institutional frameworks. So the coming of the plebiscitary era did not alter, much less derail, an ongoing dynamic, while reorganizing the sequence and clustering of contests inside newly plebiscitary institutions was even less likely to achieve any such effect. Table 6.10 demonstrates this basic continuity despite all such state party maneuvering, by offering the specific points when the Gain-Deficit Ratio was achieved in all twenty-eight of the nominating contests in the modern era. These are the points when the nomination was effectively settled, aggregate vote totals notwithtanding and regardless of contrary interpretations by candidates and commentators. Seen this way, the selection of an aspiringly strategic date for one or another state primary looked

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

231

Table 6.10 Super Tuesday and the resolution of nominating politics The Democrats

The Republicans

Year

Largest cluster

Gain-deficit ratio

Largest cluster

Gain-deficit ratio

1972 1976 1980 1984 1988 1992 1996 2000 2004 2008 2012 2016 2020

June 6 June 8 June 8 March 13 March 8 March 10 – March 7 March 2 February 5 – March 1 March 3

June 6 June 8 May 6 May 8 June 7 June 2 – March 7 March 2 February 5 – March 22 March 3

– June 8 June 8 – March 8 – March 7 March 7 – February 5 March 1 March 1 –

– June 8 May 3 – March 8 – March 12 March 7 – February 5 March 6 March 15 –

like just another case of deliberate reform or strategic calculation, with their usual disappointments. All such gambits had concrete and measurable impacts on the formal mechanics of nominating politics. Yet all had only very secondary impacts on its practical pursuit, frequently proving irrelevant if not actually counterproductive. The collective impact of individual state efforts at the strategic placement of their contests did alter the formalistic sequence of delegate selections, most obviously by moving the date of the largest cluster of available delegates, ranging from June 8 at the latest to February 5 at the earliest. Yet that did not prevent the bandwagon from running in a parallel and familiar fashion across all twenty-eight shifting contests in this modern era. By comparison, individual state calculations, or even of the collective front-loading that followed from it, were inconsequential. Said the other way around, the impact of front-loading on the bandwagon dynamic was exiguous. As ever, this did not mean that the details by which each contest unfolded were the same, nor that the institutional reordering of these contests had no effects. The bandwagon did role later when the bulk of the delegates were selected later, and earlier when they were selected earlier. Though even within this mechanical impact, there were years when

232

B. E. SHAFER AND E. M. SAWYER

later clustering meant later effective nomination, as with 1972 and 1976. Just as there were years when earlier clustering meant earlier nomination, as with 2000, 2004, 2008, and 2020. Yet there were also years, the modal category, when the ratio was achieved after the date of the largest delegate cluster, as with 1984, 1988, 1992, 1996, 2012, and 2016. And there was even a year when the ratio was achieved before that date, in 1980. That composite picture underlines the inefficiency of individual state strategies as a collective whole. And the associated irony is evident. If states remained where they had been historically, they quickly discovered that their contests had become irrelevant. But if states sought a fresh advantage from temporal placement, this meant that the process would become increasingly front-loaded, with the result that individualized gains for particular states within this augmented clustering could only get smaller and smaller, not larger and larger. But were there no individual successes within this largely self-canceling collective failure? The prospects of any such effect among the smaller states brought another immediate irony. The statistically inconsequential states of Iowa and New Hampshire were the two states of any size that captured in practical politics what they lacked in raw numbers. But once they had secured the only temporal positions that could make up for their small size, by creating, defending, and harvesting their opening locations, Iowa and New Hampshire doomed other small states to disappear among their larger counterparts. So if there was to be some strategic effect elsewhere— some individualized triumphs buried within a collective failure—it could come in only two places. In the first, it had to come from coordination among a bloc of mid-sized states with some common (unifying) characteristic. Or it had to come from the largest states once again, the ones with some numerical hope of manipulating the total process on their own. The best example of the bloc strategy—albeit ultimately just a great new instance of self-defeating efforts—surfaced in 1984 and reached its apogee in 1988, when a growing group of southern states attempted to create a regional cluster that would wield more practical influence than any of them could generate individually. After a halting effort in 1984, the southern states largely succeeded in moving as a group to what became Super Tuesday in 1988, injecting twelve southern and border states into that one day. Yet when these newly clustered states actually voted, their strategic gambit fell afoul of two inherent defects. In the first, even the south proved to be much more heterogeneous in its candidate preferences than strategists had imagined, allowing a range of candidates to

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

233

Table 6.11 Strategic states and Super Tuesday: The role of the big states Year

Large states

Share of super T. (%)

Small states

Share of super T. (%)

2000 2004 2008

CA, NY, OH CA, NY, OH CA, NY, IL, NJ

59 59 51

11 7 18

41 41 49

survive comfortably: Senator Al Gore of Tennessee, the moderate leader in the field; the Rev. Jesse Jackson of Illinois, the most liberal; and Governor Michael Dukakis of Massachusetts, the ultimate nominee. And once the bulk of southern states had voted, the south as a region was of course devalued in the remainder of the contest, having few regional representatives left to register their preferences (Table 6.11). The big-state story was different, but ultimately came with the same self-defeating moral. Commentators at the time did not fail to notice the onward march of front-loading; indeed, they anguished over it. What they did fail to notice was that this new phenomenon, this front-loading, was actually caused by the behavior of the cluster of large states that had kept the largest delegate cluster anchored at the end of the nominating sequence for a long time. When these states finally abandoned that position, recognizing the growing importance of a couple of early delegate clusters and moving to join them, it was in fact this move that converted a statistically prominent date, Super Tuesday, into a frequently decisive one, while simultaneously crushing the individual efforts not just of the smallest states but also of aspiring blocs among the middle-sized. California, New York, and Ohio could take the most immediate credit for this effect in 2000 and 2004, while California, New York, Illinois, and New Jersey took the result to its zenith in 2008, the year when 48% of the delegates in both parties were available on Super Tuesday. But the key fact inside all this was that these three or four states were more than half of the delegate concentration on that critical date in all three years. There were still 11, 7, and 8 other states located on this date in 2000, 2004, and 2008 respectively, so it would have garnered the largest number of states without the contribution of a handful of giants. Yet those remaining states had become small change once these giants moved in. So the lead story—and the essence of the new dynamic—was the behavior of the big states that had previously anchored the final stage of the selection process.

234

B. E. SHAFER AND E. M. SAWYER

For a student of the impact of state strategies, the upshots were cruel. Once again, the collective impact of a set of state maneuvers proved to be largely self-canceling. The mid-sized states that had begun a coordinated move toward the front of the plebiscitary sequence, in order to escape the domination of Iowa and New Hampshire on the front end and of a cluster of giants at the back, did set off the creation of a couple of greater and greater (and earlier and earlier) delegate clusters, the largest of which was usually dubbed Super Tuesday. Initially, this was an improvement, albeit a limited one because these states were not nearly as homogenous as strategists had imagined. Yet any such advantage dissipated quickly when the largest states, recognizing the new delegate distribution inherent in this new sequence of state contests, move to join the couple of major dates with rising influence over the nomination—where they simultaneously became the explanation for this rising influence and the crucial elements within these new delegate clusters.

Strategic Aspirants, Strategic Parties, and Nominating Politics What was not too early to perceive was the continuing inability of even the largest states to use their strategic possibilities to overcome, alter, or so much as divert the behavioral logic of a long-running bandwagon dynamic. In that sense, analyses of the impact of the rules of the game and analyses of the strategies adopted in response to them, by state parties and presidential candidates, were very similar. Rules reform, albeit very intermittent until the modern era, did have real and measurable impacts on the apportionment of delegates among the states and on the allocation of those delegates among presidential aspirants. It was just that these impacts were reliably minor, when they were not actually contrary to the intentions of rules reformers. Adding strategic impacts to this picture changes its specifics but not its overall conclusion. From the earliest days, presidential aspirants did have to decide whether to run for a presidential nomination, when to enter the nominating contest if they wanted to run, and how to proceed once they had answered the first two questions. Yet their realistic options were powerfully constrained by their own assets and liabilities and then by their indifferent success at gaming an actual nominating contest, as diverse as these contests reliably were and where most of the other contestants were weighing the same strategic considerations. As a result, most strategies

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

235

were unsuccessful in the ultimate sense of securing a nomination, while many that looked like successes had to be considered largely accidental. One of the biggest things that serious strategists had to consider was the specific course of a nominating contest, that is, the temporal placement of state contests within the overall process, the way one contest might (or might not) leverage another, and the potential collective impact of the resulting sequence. This aspect of their strategic environment did become more varied, or at least more complex in its detail, as time passed. Moreover, state parties were not just aware of being integral to the strategies of potential nominees. They worked constantly to adjust their own place within the nominating sequence so as to maximize state influence over it. With the result that the individual pieces were often in motion, while the sequence itself changed regularly, courtesy of those changing attempts at state influence. Yet candidate strategies proved to be more constrained in practice than an abstract description might suggest, in part because the sequence of nominating contests brought its own constraints, in larger part because the bandwagon dynamic easily absorbed a changing sequence. Beyond that, state strategies proved not only to have inherent constraints of their own—leeway for strategic placement was effectively quite narrow—but often to become self-canceling when any significant minority of states, or just a small set of states with a significant number of delegates, reached similar conclusions about how to game the overall contest. So in the end, strategies and sequences told essentially the same story as apportionment and allocation. There were still real effects from both rules reform and strategic adjustment, albeit often with unanticipated consequences. Yet they remained reliably insufficient to alter a nominating dynamic forged over time, not to mention the occupational and career seedbeds of those who ran for a presidential nomination within that dynamic and the structural influences— especially factional preferences and factional struggles—that owned the real selection of a nominee from within that field of contenders. The interaction of strategies and sequences, or the impact of major rules of the game, might have some influence on the result of any given contest, in the sense of making a final idiosyncratic contributions to it. In a sufficiently balanced struggle, this might even help to produce one aspirant rather than another as the nominee.

236

B. E. SHAFER AND E. M. SAWYER

Yet in the long causal funnel that runs from a historically repetitive bandwagon dynamic to the anointment of one particular presidential nominee, these candidate strategies and state sequences, like formulae for delegate apportionment and delegate allocation, proved to be far down the causal chain. As portentous as one or another of these rules battles may have seemed while it was being waged, or as insightful as a particular strategic gambit may have seemed when it was announced, they could qualify at best—and usually not at all—as final adjustments to the longrunning process that recurrently produced presidential nominees. In that sense, it remained the structural fundamentals that deserved this decisive designation, and they remained largely outside the reach of even the major individual actors.

Notes 1. A biography of the years before 1992 is Robert S. McElvaine, Mario Cuomo: A Biography (New York: Scribner, 1988). An argument that the long will-he-or-won’t-he ballet was important is Saladin Ambar, American Cicero: Mario Cuomo and the Defense of American Liberalism (New York: Oxford University Press, 2017). 2. Michael F. Holt, By One Vote: The Disputed Presidential Election of 1876 (Lawrence: University Press of Kansas, 2008). 3. James C. Klotter, Henry Clay: The Man Who Would Be President (New York: Oxford University Press, 2018). Clay is inevitably also a major connecting thread in Michael F. Holt, Rise and Fall of the American Whig Party: Jacksonian Politics and the Onset of the Civil War (New York: Oxford University Press, 1999). 4. Daniel DiSalvo, “Gilded Age Factions”, 66–69, in DiSalvo, Engines of Change: Party Factions in American Politics, 1868–2010 (New York: Oxford University Press, 2012). For the man himself, Neil Rolde, Continental Liar from the State of Maine: James G. Blaine (Thomaston, ME: Tilbury House Publishers, 2007). 5. Michael Kazin, A Godly Hero: The Life of William Jennings Bryan (New York: Anchor Books, 2007). For the larger movement around him, see Robert C. McMath, Jr., American Populism: A Social History, 1877–1898 (New York: Hill & Wang, 1992), and Peter H. Argersinger, The Limits of Agrarian Radicalism: Western Populism and American Politics (Lawrence: University Press of Kansas, 1995). 6. For Dewey, Richard Norton Smith, Thomas E. Dewey and His Times (New York: Simon & Schuster, 1982); for Taft, James T. Patterson, Mr. Republican: A Biography of Robert A. Taft (New York: Houghton Mifflin, 1972).

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

237

7. Richard C. Bain and Judith H. Parris, “First Whig Convention”, 24– 27, in Bain and Parris, Convention Decisions and Voting Records, 2nd ed. (Washington, DC: Brookings, 1971), and “1839–40 Conventions: Whigs”, 50–51, in National Party Conventions, 1831–2008 (Washington, DC: CQ Press, 2010). For thumbnail accounts of nominating politics and their conventions in what follows, especially when these are used principally in lists of supportive examples, the two simplest references are Bain and Parris, Convention Decisions and Voting Records, along with the CQ Press volume on National Party Conventions. The latter is more condensed, the former more comprehensive. 8. James K. Polk in 1844 being the first. 9. In a very narrow sense, Theodore Roosevelt was “denied renomination” in 1912, though he was actually blocked by the renomination of the sitting president at the time, William Howard Taft. 10. For a view of the nominating world shortly after their arrival, see Louise Overacker, The Presidential Primary (New York: Macmillan, 1926), with a reprint edition by Arno Press in 1974. 11. “1920 Primaries”, 324–326, in Congressional Quarterly’s Guide to U.S. Elections (Washington, DC: CQ Press, 2001). 12. Theodore H. White, The Making of the President 1960 (New York: Atheneum, 1961). Also W.J. Rorabaugh, The Real Making of the President: Kennedy, Nixon, and the 1960 Election (Lawrence: University Press of Kansas, 2009). 13. Theodore H. White, The Making of the President 1968 (New York: Atheneum, 1965). Also Robert Alan Goldberg, Barry Goldwater (New Haven: Yale University Press, 1997). 14. Byron E. Shafer, “Institutionalizing the Disappearance: The Nomination After Reform”, Chapter 2 in Shafer, Bifurcated Politics: Evolution and Reform in the National Party Convention (Cambridge, MA: Harvard University Press, 1988). 15. For this struggle over podium appearances and their timing, Byron E. Shafer, “The Pure Partisan Institution: National Party Conventions as Research Sites”, Chapter 14 in L. Sandy Maisel and Jeffrey M. Berry, eds., Oxford Handbook of American Political Parties and Interest Groups (New York: Oxford University Press, 2010). For an even larger context to these same developments, Costas Panagopoulos, ed., Rewiring Politics: Presidential Nominating Politics in the Media Age (Baton Rouge: LSU Press, 2007). 16. See Table 6.5. 17. Gary R. Orren and Nelson W. Polsby, eds., Media and Momentum: The New Hampshire Primary and Nomination Politics (Chatham, NJ: Chatham House, 1987).

238

B. E. SHAFER AND E. M. SAWYER

18. Peverill Squire, ed., The Iowa Caucuses and the Presidential Nominating Process (Boulder: Westview Press, 1989). 19. Beyond Bain and Parris plus the CQ Press National Party Conventions, see Michael F. Holt, The Election of 1860: “A Campaign Fraught with Consequences” (Lawrence: University Press of Kansas, 2017). 20. Charles Callan Tansill, The Foreign Policy of Thomas F. Bayard, 1885–1897 (New York: Fordham University Press, 1940), and in his own words, Thomas F. Bayard and Edward Spencer, An Outline of the Public Life and Services of Thomas F. Bayard, Senator of the United States from the State of Delaware, 1869–1880 (Victoria, Australia: Leopold Classic Library, 2016). 21. The early story for the mixed system is the subject of Louise Overacker, The Presidential Primary (Note 10). The situation at the end of this era is captured in James W. Davis, Presidential Primaries: Road to the White House (New York: Thomas Y. Crowell, 1967). 22. “1912 Primaries”, in from “Presidential Primaries”, Chapter 10 of Congressional Quarterly’s Guide to U.S. Elections, 4th ed. (Washington, DC: CQ Press, 2001). Specific results for all direct subsequent references to primary entrants and primary outcomes are drawn from the same compilation. 23. Drawn from Congressional Quarterly’s Guide to U.S. Elections. 24. For a marvelously rich description of the operation of this system before the plebiscitary era, state by state, see Paul T. David, Malcolm Moos, and Ralph M. Goldman, comps., Presidential Nominating Politics in 1952, 5 vols. (Baltimore: Johns Hopkins University Press, 1954). 25. Table 6.5 is drawn from Table 14.2 in Shafer, “The Pure Partisan Institution: National Party Conventions as Research Sites” (Note 15), 269. 26. For a view of these developments in the larger context, see Byron E. Shafer and Regina L. Wagner, “Party Structures and Representational Impact”, Chapter 2 in Shafer and Wagner, The Long War Over Party Structure: Democratic Representation and Policy Responsiveness in American Politics (Cambridge: Cambridge University Press, 2019). 27. For most of the tables in this chapter and this book, the nominating contests of 2020 provide the most recent entry to the relevant tabulations. Yet 2020 was disturbed powerfully by the coronavirus and its associated pandemic, such that the 2020 process could not always be entered in parallel with preceding years—in which case 2016 became the end of the data series. In some cases, the problem was that states changed the date of their own delegate selection while the total process was unfolding, making it unclear which date should be used. In other cases, state presidential primaries were actually cancelled if there was no longer a challenger to the obvious nominee. In no cases was 2020 excluded from tabulations in which choices about its details would alter or amend any result.

6

THE USUAL SUSPECTS: STRATEGIES AND SEQUENCES

239

28. In several modern contests, Texas has provided a sense for the scale of the difference between using the caucus and using the primary, by selecting its delegates in both ways on the same day. See, for example, Byron E. Shafer and Amber Wichowsky, “Institutional Structure and Democratic Values: A Research Note on a Natural Experiment”, The Forum 7(Issue 2, 2009), Article 2. For a far more comprehensive examination of the impact of doing state party business by way of participatory caucuses, see Seth E. Masket, No Middle Ground: How Informal Party Organizations Control Nominations and Polarize Legislatures (Ann Arbor: University of Michigan Press, 2011). 29. Caitlin E. Jewett refers to them as “the other carve-out states” in Jewitt, The Primary Rules: Parties, Voters, and Presidential Nominations (Ann Arbor: University of Michigan Press, 2019), 225. 30. The ability of South Carolina to achieve just such an impact for the Democrats in 2020 was to be both anomalous and dependent on an extremely unlikely concatenation of prior circumstances. See the final section of Chapter 7. 31. Five of the six states created new presidential primaries in 1972 placed them in May, and all five were on days that had not previously hosted a primary in any other state. 32. William G. Mayer and Andrew E. Busch unpack this southern effort in “The Rise of Front-Loading”, Chapter 2 in Mayer and Busch, The Front-Loading Problem in Presidential Nominations (Washington, DC: Brookings, 2004). 33. The larger struggle over placement within the nominating sequence, and the recurrent role of fights over the “window” within which various contests would occur, is richly considered in Elaine C. Kamarck, “Sequence as Strategy” and “The Fight to be First,” Chapters 2 and 3 in Kamarck, Primary Politics: How Presidential Candidates Have Shaped the Modern Nominating System (Washington, DC: Brookings, 2009).

CHAPTER 7

Eternal Bandwagon: Nominating Politics and the General Election

Nominating politics are hardly the end of the politics of presidential selection: there is still no president until one or another of these nominees has survived a general election campaign and garnered a majority in the Electoral College. Yet much of the totality of this politics is completed by the time the second major-party candidate accepts a presidential nomination. The vast bulk of the winnowing of possible presidents is over by the end of the nominating process. The bandwagon characterizing the politics producing two—or very occasionally three—serious options for president has run its course. The occupational and career seedbeds that generated an available field of contestants have long since done their work. And the factional structures of the two parties have made the remaining final choices within those fields. Nevertheless, a number of major differences between the politics of a presidential nomination and the politics of a general election remain.1 Said differently, the politics of the general election are not some logical extension of nominating politics. The field of contenders always shrinks, modestly in some years, substantially in others. The policy differences characterizing what are now two major-party nominees almost inevitably change from the policy differences that characterized the previous arrays of alternative aspirants for a presidential nomination. Different constituency groups must now be targeted—in fact, two different sets of different groups—and their preferences are generally stable enough to © The Author(s) 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1_7

241

242

B. E. SHAFER AND E. M. SAWYER

shape general election politics all the way from its advertising thematics to the way the two candidates allocate their travel time. Even the elementary mechanics differ fundamentally. Which is to say: convention delegates and Electoral Votes, even when they are apportioned by the exact same formula, require a different strategy for their collection. The convention delegates from each and every state will always be won by some Democratic contender among the Democrats and some Republican contender among the Republicans. Yet many Electoral Votes are beyond the reach of either any Democratic or any Republican nominee. So a strategic focus for the general election must differ from the options available in the nominating process, even for the same individual who stood for a nomination and now stands for president. Perhaps the one place where the two kinds of politics, nominating versus electing, remain linked is through the question of how the former shapes, and is sometimes shaped by, the latter. From one side, some of the ultimate outcomes can surely be traced back to the nominating politics that preceded it. While from the other side, some of this nominating politics is surely a response to challenges from the general election campaign that could be foreseen in advance. So the final task in a comprehensive look at nominating politics is to move on and see what can be said about the links between the two, that is, the limitations on an eventual election campaign that were imposed by nominating politics, and the limitations on nominating politics that were imposed by anticipating aspects of an eventual election campaign. That task will absorb the first and larger part of this concluding chapter. A final section can then return and attempt to gather the entire chronicle—bandwagon dynamic, candidate seedbeds, factional structures, and electoral links—into one concise closing summary. Counterfactual problems in the analysis of alternative formulae for apportioning delegates among the states or for allocating those delegates among presidential aspirants proved to be a serious challenge for Chapters 5 and 6. Yet they pale before the counterpart problems inherent in the analysis of links between nominating politics and the politics of the general election. Would election results have been different if a different nominee had triumphed inside one, the other, or both parties? Would nominating politics have been different if the coming general election seemed destined to play out one way rather than another? Those are questions that, in principle, cannot be answered.

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

243

On the other hand, the ability to explain real results remains more valuable than any putative effort to explain hypotheticals. Moreover, there are two major aspects of the link between nominating and electing politics that can potentially be analyzed. At a bare minimum, an analysis can classify presidential contests through the patterning of their nominating politics, while comparing the rise, fall, and interaction of such patterns across time. In the process, it should be possible to examine the factional maneuvering that accompanied each of those patterns, looking for the ones that need little explanation beyond the desires and tactics of individual factions, versus those where perceived electoral demands appeared to feed back into—and actually shift—factional behavior. So, to cut directly to the chase: in the hunt for nominating patterns and their drivers, there have been five such patterned outcomes, each needing some attention in its own right but all ultimately distinguishable only through attention to the others. These begin with the two patterns that involved sitting presidents, associated with the simple division between uncontested and contested presidential renominations. After that, for the considerably larger number of contests without a sitting president, there were three further alternatives: nominations that were consensual from the start, front-runners who went on to secure a nomination, and contested nominations where a nominee was produced only after extended conflict, often as a totally unforeseen product. In each case, the central question is how much of a particular pattern could be explained by factional maneuvering and how much owed something substantial to perceived demands from the coming general election. Answers will vary between the parties at various points in time; they will vary within each of these parties across time; and they will vary with the main institutional periods for nominating politics: pure convention system, mixed system, and plebiscitary system. In the process of unpacking this story, there will be clear periods within each party where an impending electoral politics actually did feed back into prior nominating struggles. Yet on the average and over time, it is the power of party factions, and of factional struggles, that continues to stand out.

244

B. E. SHAFER AND E. M. SAWYER

Nominating Politics with a Sitting President Uncontested Renomination Efforts Easiest to analyze in these terms are the uncontested renominations of sitting presidents. Most sitting presidents have wanted to be renominated; most presidents seeking renomination have secured it without serious opposition; and most renominated presidents have gone on to be reelected (Table 7.1). The exceptions get their attention in Table 7.2, but the general pattern is of incumbent presidents—and overwhelmingly if they have previously been elected in their own right—who tend to be at peace with major-party factions while being beneficiaries of a generalized legitimacy in the voting public. In the days when internal party machinery Table 7.1 Uncontested renominations of sitting presidents

Democrats

Whig/Republicans

Jackson 1832 Van Buren 1840* Lincoln 1864 Grant 1872 Cleveland 1888* Cleveland 1892

Harrison 1892 McKinley 1900 T. Roosevelt 1904

————————— Wilson 1916 Hoover 1932* F. Roosevelt 1936 F. Roosevelt 1940 F. Roosevelt 1944 Eisenhower 1956 L. Johnson 1964 ————————– Nixon 1972 Reagan 1984 H.W. Bush 1992 Clinton 1996 W. Bush 2004 Obama 2012 *Went on to lose presidential election

7

Table 7.2 Contested renominations of sitting presidents

ETERNAL BANDWAGON: NOMINATING POLITICS …

Democrats

245

Whig/Republicans Tyler 1844*** Fillmore 1852**

Pierce 1856** A. Johnson 1868*** Arthur 1884** ————————— Taft 1912* Truman 1948 ————————— Ford 1976* Carter 1980* *Renominated but went on to lose presidential election **Failed to secure renomination ***Stood for Renomination (unsuccessfully) in the other party

created most convention delegates, these presidents were usually in a position to control the machinery. In the days when external primary contests instead created most convention delegates, these presidents had the best name recognition coupled with a record, at the very least, of having done the job. Moreover, impending electoral demands clearly did not alter their nominating fortunes. There has been an impressive minority of presidents who died in office: William Henry Harrison in 1840, Zachary Taylor in 1848, Abraham Lincoln in 1864, James A. Garfield in 1880, William McKinley in 1900, Warren Harding in 1920, Franklin Roosevelt in 1944, and John Kennedy in 1960. Just as there have been presidents who, while elected in their own right, did not seek a second term: James K. Polk in 1844, Rutherford B. Hayes in 1876, Theodore Roosevelt in 1904,2 Calvin Coolidge in 1924, Harry Truman in 1948, and Lyndon Johnson in 1968. Some might have faced fresh electoral challenges if they had sought a second full term, but none appeared to face serious risks of not being renominated.3 So elected presidents who sought a second term succeeded overwhelmingly at renomination. And indeed, they tended to succeed afterwards at the ballot box afterward as well. Moreover, the three who failed cannot be said to have done so because of factional dissent inside their party.

246

B. E. SHAFER AND E. M. SAWYER

Herbert Hoover in 1932 was the consensus choice for what looked likely to be a desperate reelection contest. Grover Cleveland in 1888 was the consensus choice for reelection in a year when a seriously contracting economy would have weighed on the fortunes of any sitting president, yet Cleveland remained so consensual that, even after electoral defeat, he would be renominated without opposition (and successfully elected) in 1892. Lastly, Martin Van Buren, “founding father” of the Democratic Party along with Andrew Jackson, could be renominated without opposition at the 1840 convention but, like Cleveland forty-eight years later, could not survive the impact of a steep economic downturn, in his case via the Panic of 1837.4 Still, there appear to be no exceptions to the generalization that sitting presidents preferred to be uncontested if they were actively seeking renomination. In the abstract, there might have been years when a rudimentary initial opposition could have provided a kind of ‘tune-up’ for the general election campaign. But in practice, no president in either Table 7.1 or Table 7.2 appears to have looked at the world this way. Certainly none acted on that view, though presidential wishes hardly guaranteed that sitting presidents would secure an uncontested renomination, as Table 7.2 will confirm. Contested Renomination Efforts Intraparty challenges to a sitting president were to occur with any regularity only during the pure convention system, and these always reflected most directly the factional struggles inside the party in office. Moreover, in this period but in this period only, serious intraparty challenges actually resulted in the failure of the sitting president to secure even renomination. Thereafter, such challenges would be rare, involving the occasional factional split that could not be suppressed by the incumbent. All four of these later challenges, to the renomination of William Howard Taft in 1912, Harry Truman in 1948, Gerald Ford in 1976, and Jimmy Carter in 1980, were to be overcome by the president in question, yet three of the four—Truman being the lone exception—led on to their defeat at the general election.5 The larger half of these intraparty challenges to a sitting president sprang from resurgent factional problems triggered when a sitting president died in office, though factional disappointments rather than electoral pressures appeared to drive not only these early exceptions but all the

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

247

rest as well. More pointedly, four of the five challenges during the antebellum period were products of a peculiar nexus of party-building and candidate mortality. To begin with, all four contested presidents were in office only because they had been vice presidents under an elected counterpart who had died in office: John Tyler for William Henry Harrison in 1844, Millard Fillmore for Zachary Taylor in 1852, Andrew Johnson for Abraham Lincoln in 1868, and Chester A. Arthur for James A. Garfield in 1884. So it could safely be said that they had never been the choice of a factional majority within their party, nor had they ever achieved an electoral majority at the ballot box. Beyond that, two of the four had actually been nominated for vice president as the lesser half of an actual cross-party ticket: Tyler as an erstwhile Democrat with the Whig Harrison and Johnson as a continuing Democrat with the Republican Lincoln. Yet the critical point for Fillmore and Arthur too was that they had been “ticket-balancers,” that is, active members of their party but chosen as a partial side payment to losing factions. Franklin Pierce thus became the lone exception, a dark-horse compromise in 1852 after an extended factional struggle between four initially serious contenders. Yet Pierce and Fillmore were united not just by their failures at renomination but by the way that rising internal party divisions over territorial expansion and the place of slavery within it had characterized their presidencies, miring them in deep factional splits which they could not ultimately finesse.6 The same could be said less dramatically of Chester Arthur: nominated as a ticket-balancer in compensation to a losing faction, but forced to conduct a presidency that could not please any of the three entrenched Republican factions. While after Arthur, contested renominations for sitting presidents would be rare. The most dramatic belonged to William Howard Taft in 1912, where the aggrieved dissent of the progressive faction would be enough to drive Republican Taft into third place in the general election. Clearly, even the threat of electing the Democratic nominee was not sufficient to cause progressive Republicans under Theodore Roosevelt or regular Republicans under Taft to unite with the opposing faction.7 By the plebiscitary era, factional opponents of a sitting president were more or less compelled to assert that they were acting to salvage a general election, else they could be portrayed to a primary electorate as simple spoilers. So both the activist challenge to the renomination of Republican President Ford in 1976 and the activist challenge to the renomination

248

B. E. SHAFER AND E. M. SAWYER

of Democratic President Carter in 1980 duly asserted the centrality of electoral calculations. Yet in both cases, factional desires on the part of dissident activists were evidently dominant: there was no evidence that a large chunk of the Ford vote would come to Reagan but not to Ford, or that a large chunk of the Carter vote would come to Kennedy but not to Carter.8 Both Ford and Carter were to survive this challenge to their renomination, Ford by a whisker and Carter more easily, but both would go on and lose the general election. In that sense, a preelection “tune-up” was clearly not a benefit. So the lone survivor of an internal contest for renomination by a sitting president was to be Democrat Harry Truman in 1948. From one side, he was beset by party figures who feared electoral defeat enough to try to attract Dwight Eisenhower, the Republican nominee for president four years later, as a replacement for Truman. From the other side, he was beset by factional leaders from the southern Democrats, who actually bolted the convention and then abandoned its nominee, in an ultimately failed effort to deprive him of an Electoral Vote majority.9 What distinguished Truman from everyone else on this list, from John Tyler through Jimmy Carter, was that he was the only one to be not just renominated but reelected—in the period following on from the New Deal when it was difficult for a Democrat not to be reelected, even when beset by the array of difficulties that challenged Truman.

Nominating Politics Among Non-Presidents Yet the bulk of presidential nominations inevitably came in years without a sitting president. Even in years with such a president, one of the two nominations was still generated by the out-party, while years without a sitting president from either party featured two nonincumbent nominees, and there was that handful of contests in the early period when an incumbent president was not renominated. Within this much greater share of nominating contests, then, there were three available patterns to the politics of presidential nomination: • About a third of the time, these became consensual nominations despite the absence of a sitting president, coalescing behind an aspirant chosen on the first ballot during the pure convention and mixed systems or in the early primaries in the plebiscitary era.

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

249

• About a third of the time, there was no imminent consensual nominee but there was an obvious front-runner, leading on the first ballot under those older systems or in preprimary polls in the modern world and converting this front-runner status into a nominating majority. • Lastly, another third of the nominating struggles without a sitting president did possess someone who had to be considered the frontrunner but who failed to harvest that status, producing a protracted conflict in the convention hall or on the primary trail for a winner who would often have been unexpected when the process began. Consensual Nominations The first of these patterns, consensual nomination for an aspirant who was not a sitting president, confirms the recurrent possibility of consensual nominating majorities even in “open” contests (Table 7.3). As such, these majorities were inescapable evidence of factional unity, though the Table 7.3 Consensual nominations

Democrats

Whig/Republicans Clay 1844*

McClellan 1864* Grant 1868 Greeley 1872* Cleveland 1888* Cleveland 1892 Bryan 1900* Parker 1904* Bryan 1908*

B. Harrison 1892* McKinley 1896

Taft 1908

————————— Smith 1928*

Coolidge 1924 Hoover 1928 Landon 1936* Dewey 1944*

Stevenson 1956* Nixon 1960* —————————— *Went on to lose presidential election

250

B. E. SHAFER AND E. M. SAWYER

reasons for this unity could vary not just between the two parties but within each over time. On rare occasions, the explanation was the appearance of an irresistible hero-candidate, as with Ulysses S. Grant after the Civil War. Yet even Franklin Roosevelt after the Great Depression and Dwight Eisenhower after World War II were not consensual nominations. More often, the pattern was actively misleading when viewed in terms of subsequent electoral prospects: two-thirds of these apparently untroubled nominating paths ended in defeat come November. In their numerous consensual nominations during the pure convention era, the Democrats actually featured two clusters with effectively opposite balances between the power of factional maneuvering and the power of electoral strategy. At one extreme was an entire sequence of contests where external constraints dominated the nominating politics beforehand. These are the contests involving a dominant and draconian electoral math in the years after the Civil War, years of impressive Democratic recovery in Congress but only contingent recovery with the presidency. In response, the factional logic was simple: a reform governor from New York could unite a party with three sharply divergent factions by being a member of none but acceptable to all.10 Moreover, factional logic was reinforced by the fact that New York, the biggest state in the union, was simultaneously one of the few that could be shifted between parties from election to election.11 The result was five nominees who fit the requisite model: Horatio Seymour in 1872, Samuel Tilden in 1876, and Grover Cleveland in 1884, 1888, and 1892. As if to underline the point, Cleveland would go on to win the two elections where he carried New York, and lose the one where he did not. At the opposite extreme was an even longer sequence of Democratic contests, likewise characterized by consensual nominations. But this time, that resulted from a more or less total capitulation to one faction, coupled with a more or less total disregard for electoral considerations. This was the long chain of nominations controlled by what began as the populist faction. The nominating contest of 1896 inaugurated the chain with a struggle over who would stand as the figurehead for a populist explosion. William Jennings Bryan became the answer, repeated in 1900 despite its electorate consequences in 1896 and repeated again in 1908 despite decreasing returns from the previous two efforts.12 The ultimate escape from this populist dominance took a great deal longer, though that fact was masked by one major electoral success along the way. This success belonged to Woodrow Wilson in 1912, who proved

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

251

acceptable to the populists, now rebranded as progressives. Yet while Wilson did end up as president, that was mainly courtesy of a dramatic factional implosion within Republican ranks. Wilson returned easily as a re-nominated president in 1916, though the election itself was close. But after that, factional struggles to reintegrate the party returned to a new version of their pre-Wilson grimness. Wilson’s son-in-law, William Gibbs McAdoo of California, then carried the progressive banner in 1920, until a deadlocked convention turned to Governor James Cox of Ohio. McAdoo came back as one of two leading contenders in 1924, helping to drive that convention to 103 ballots before it turned to John W. Davis of New York. In 1928, McAdoo deferred to Governor Al Smith of New York, his main opponent the previous time out, to avoid another donnybrook—secure in the conviction that Smith would lose the general election. And the progressive faction, along with its dire electoral prospects, was finally transitioned out of existence in 1932, when it was McAdoo who launched the bandwagon that put Franklin Roosevelt—former progressive but incipient labor-liberal—over the top.13 The Republicans featured their own, quite different versions of the same stories: driven by factions with little attention to the election at one point, driven by the logic of electoral competition at another. In the first, the long postbellum period saw a Republican electoral dominance so sweeping that the factions were free to play out nominating conflicts among themselves whenever they lacked a consensual nominee, without much perceived electoral constraint.14 The Hayes, Garfield, and B. Harrison presidencies were all a product of this partisan dominance, and the factional coalitions that made each into the nominee underlined this power of three contending factions to focus only on their own dominance: a coalition of the liberals and the regulars for Hayes, a coalition of the liberals plus the half-breeds for Blaine, and a coalition of the half-breeds plus the regulars for Harrison. That situation would be reversed after the coming of Franklin Roosevelt, the New Deal, and a Democratic majority. After a dozen years when the main challenge of Republican conventions was just to find someone credible enough to front an invariably losing campaign, the Republicans, having already begun to recapture some major-state governorships, began to look for a different sort of nominee. Accordingly, this incipient resurgence was fueled by a determined effort led by the regular party to find moderate nominees who were potentially capable of carrying the major industrial states. That was the strategic straitjacket that followed

252

B. E. SHAFER AND E. M. SAWYER

from a still-daunting but less dire electoral environment, and the associated tactical response would underpin the nominations of Dewey in 1944 and 1948, Eisenhower in 1952 (and 1956), and Nixon in 1960. The other main thing to note about consensual nominations is that in the modern plebiscitary era, where victory could only be conferred through an extended sequence of presidential primaries, the prospect of consensual nominations in open contests effectively disappeared. There could still be sitting presidents who were consensual choices when they sought renomination, as with Richard Nixon in 1972, Ronald Reagan in 1984, George H.W. Bush in 1992, Bill Clinton in 1996, George W. Bush in 2004, Barack Obama in 2012, and Donald Trump in 2020. But even some modern renomination efforts by sitting presidents were not consensual, as with Gerald Ford in 1976 and Jimmy Carter in 1980, and all open contests produced a field of contenders for their sequence of primary elections, not some consensus choice who could hope to walk unopposed through that sequence. Front-Runner Victories Yet for all the many years when there was no evident consensual choice as the nominating contest opened, there was still by definition a frontrunner, and it was the fortunes of these front-runners that contributed the two remaining patterns to nominating politics. Even in the plebiscitary era, someone had to carry a polling lead, however fragile, into the first real contest for delegate selection. So the modern world could once again be treated in the same terms as earlier eras where the front-runner had instead been the leader on the opening ballot of conventions that likewise lacked an imminent consensual choice (Table 7.4). The critical distinction between the two remaining patterns—each was about as numerous as uncontested renominations for sitting presidents or consensual nominations in open contests—was whether the initially leading contender could convert front-runner status into a nominating majority. The analytic questions could remain the same. How much of this particular pattern resulted mainly from factional struggles? And how much was instead constrained by advance perceptions of the general election to follow? When there had been a consensus choice and not just some de facto front-runner, the Democrats and the Republicans had answered those questions with periods when the likely nature of the general election fed back actively into nominating politics and other periods when

7

Table 7.4 Front-runner victories

ETERNAL BANDWAGON: NOMINATING POLITICS …

253

Democrats

Whig/Republicans

Cass 1848*

Taylor 1848 Fremont 1856*

Tilden 1876* Hancock 1880* Cleveland 1884

Blaine 1884*

————————— Hughes 1916* F. Roosevelt 1932 Stevenson 1952* Kennedy 1960 Humphrey 1968*

Dewey 1948* Eisenhower 1952 Goldwater 1964* Nixon 1968

—————————— Reagan 1980 Mondale 1984* Dukakis 1988* Gore 2000* Clinton 2016*

H.W. Bush 1988 Dole 1996* W. Bush 2000 Trump 2016

*Went on to lose presidential election

prospective electoral politics just as clearly did not. Table 7.4 will gather the cases where a front-runner who was not fortunate enough to be an imminent consensual choice was still able to convert this initial edge into a nomination. Table 7.5 will close this section by gathering the remaining years, distinguished by the fact that the formal front-runner could not convert early support into a nominating majority. By extension and unsurprisingly, front-runners who were ultimately nominated (and hence successful) were generally able to consolidate their support quickly, after only a few convention ballots or with a major delegate lead early in the primary sequence. Though this difference is better stated the other way around: front-runners who did not accomplish this feat quickly tended to fade, yielding an eventual leader who emerged from the more-extended conflict that followed (and whose fortunes are covered in Table 7.5).15 Note that, in an important aside, the fact that successful front-runners crossed the Gain-Deficit Ratio at an early point

254

B. E. SHAFER AND E. M. SAWYER

Table 7.5 Extended contests

Democrats

Whig/Republicans W.H. Harrison 1840

Polk 1844 Pierce 1852 Buchanan 1856 Douglas 1860* Seymour 1868*

Scott 1852* Lincoln 1860 Hayes 1876 Garfield 1880 B. Harrison 1888

Bryan 1896* ——————————— Wilson 1912 Cox 1920* Davis 1924*

Harding 1920 Willkie 1940*

——————————— McGovern 1972* Carter 1976 Clinton 1992 Kerry 2004* Obama 2008

McCain 2008* Romney 2012*

*Went on to lose presidential election

did not necessarily mean that challengers reliably surrendered, exited, and disappeared. What it did mean was that those challengers would never succeed. For a variety of reasons, one or another defeated aspirant might continue on, albeit without further hope of affecting the identity of the nominee.16 In earlier periods, declining to withdraw when there was an obvious winner meant simply tallying one more time the shrinking cohort of delegates who were willing to stay with the loser(s). In the modern world, declining to withdraw meant instead insisting on a pro forma nomination, presumably to use press coverage at the convention for other purposes—for a policy, for a constituency, or for personal advertising.17 But the point here is that while an impending election had not been sufficient to drive the party to a consensus nomination at the start,

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

255

the efforts of the front-runner had been sufficient to bring a majority into line behind this initially advantaged aspirant. Nominations where a majority had to be built by the front-runner, like consensual confirmations, told different stories as between the two parties and at different points in time, with associated implications for the power of factional versus electoral pressures. For the Democrats, it was the long late New Deal era from 1932 through 1968 when frontrunners—that is, front-runners who were not sitting presidents seeking renomination—were regularly successful. Franklin Roosevelt in 1932, Harry Truman in 1948, John Kennedy in 1960, and Lyndon Johnson in 1964 would go on to win the general election, while Adlai Stevenson in 1952 and Hubert Humphrey in 1968 would secure the nomination but lose the general election. Yet nothing about their collective fortunes in November, winning or losing, did much to disrupt an ongoing triumph within nominating politics by regular parties from major industrial states, whose successful factional preferences were reliably dominant. Conversely, the Republicans had had nearly no nominations resulting from front-runner conversions in the long postbellum world, where the party was a heavy favorite to win in the general election.18 In essence, this favorability was what left the factions free to indulge themselves in nominating politics, without so much as a side glance at the coming election, though the composite result was ironic. There were consensual nominations for Abraham Lincoln in 1964, Ulysses S. Grant in 1868, Benjamin Harrison in 1892, and William McKinley in 1896, when the factions were in agreement; no electoral considerations disturbed the consensus; and the ticket proved in due course to be elected anyway. Yet when the factions were not in agreement, front-runners generally failed to secure the nomination: even Grant failed, despite being the front-runner in his attempt at a third nomination in 1880. In other words, these were years when factional disputes, absent a consensus when the convention assembled, had to be resolved through an extended internal struggle, albeit still conducted largely free of any shared electoral concerns. Extended Contests What distinguished the fifth and final pattern to nominating politics was the absence of opening consensual choices and the inability of formalistic front-runners to forge that consensus as the nominating struggle unfolded. In practical terms, this fifth and final pattern contained the

256

B. E. SHAFER AND E. M. SAWYER

nominees who as a cohort were the least secure in their own prospects when they entered and the least projected by the rest of the party to end up with a presidential nomination. In theoretical terms, the process by which these individuals ended up as nominees simultaneously highlights the Gain-Deficit Ratio at its most analytically powerful, since the ultimate outcome was least obvious to observers in their own time and since the major players were least likely to agree that any given contest was actually over, even when the Gain-Deficit Ratio said that it was. The resulting list of winners in an extended contest inevitably begins with the stereotypical dark-horse presidents from the antebellum period: James K. Polk in 1844 and Franklin Pierce in 1852. The list continues with the compromise presidents from the postbellum period: Rutherford B. Hayes in 1872, James A. Garfield in 1880, and Benjamin Harrison in 1888. It contains two successful products of extended nominating conflict under the mixed system of delegate selection: Woodrow Wilson in 1912 and Warren Harding in 1920, neither of whom led on the opening ballot. And it concludes with the three counterparts from the plebiscitary era to date: Jimmy Carter in 1976, Bill Clinton in 1992, and Barack Obama in 2008. Though this still does not include the long list of candidates from extended contests who ended up as nominees but not presidents, again in all three eras: there were nearly as many of these as there were ultimate electoral winners. Once more, the two parties produced differing embodiments of this nominating pattern at different points in time. The Democrats were particularly likely to generate this type of nominee in the years before the Civil War, when the party was determined to remain truly national but was increasingly stressed by issues of slavery and abolition. Major factions were disinclined to accept the leading preference of other factions, and extended conflict plus unforeseen nominees were the logical result. The Republicans instead had their first cluster of this type of nominee in the years after the Civil War, a period of reliable party dominance where if the factions did not begin as aligned behind a consensual contender, they were likely to have to fight their way to some compromise which by-passed the front-runners. Again, factional divisions were to drive the specific individual results. Under the mixed system of delegate selection, it was then the Democrats who went back to producing stereotypically protracted contests, with 46 ballots for Woodrow Wilson in 1912, 44 for James Cox in 1920, and a stupendous 103 for John W. Davis in 1924. Though this

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

257

remained an overall period favoring front-runners who did not need to endure such a trial-by-ballot. The modern plebiscitary world would then be very different in this regard, while once more differing between the two parties as well. Delegate selection by way of a sequence of presidential primaries was now prone to extended conflicts, resulting frequently in someone other than the initial front-runner. With these, the Democrats would prove even more prone than the Republicans, with the latter more likely to generate front-runners who could clean up the primary sequence quickly. This partisan difference in the appearance of extended nominating conflict appears to be rooted in the earlier merger of the regulars and the activists among Republicans, which was in many ways the triumph of the activists over these regulars. In sustaining three major factions—regulars, activists, and racial minorities— the Democrats were instead particularly good at fueling and supporting extended factional conflicts. Though if the Republicans appeared to be blessed on the front end by having less of this insistent factional conflict, they tended to pay on the back end by losing the general election when intraparty conflicts did characterize their nominating politics.19

Winnowing Toward a Nominee Because the five distinguishable patterns to nominating politics are isolated here in a fashion aimed at paying particular attention to the balance between factional maneuverings and electoral constraints, these patterns group nominees in a manner that could look very different if the same contests were being categorized for other purposes. Moreover, because these patterns encompass the entire universe of major-party nominees since 1832, they would allow a huge further range of individual comparisons. Yet when the focus is the larger funnel of causality running from an underlying bandwagon dynamic to the coronation of two specific individuals, the central point remains the comparative influence of factional maneuvering rather than electoral constraint as the analysis approaches the point where two major-party nominees are confirmed. Through this lens, a focus on the five main patterns begins by confirming, over and over, the omnipresence of party factions and factional politics as the last major shaping influence on presidential nominations. Apparently obvious electoral needs—though some of politics is always a struggle to define the obvious—could sometimes overcome these

258

B. E. SHAFER AND E. M. SAWYER

factional desires and resulting factional struggles. The Democrats after the Civil War are one obvious example, with their quest for a reform governor from New York. The Republicans after the opening years of the New Deal and its new Democratic majority are another, with their search for a moderate candidate from the large industrial states. On the other hand, electoral needs were sometimes explicitly dismissed by dominant factions, who simply wanted one of their own to be the nomine—and never mind the presidential election. The Democrats in the long populist/progressive era, beginning before 1896 and expiring only after 1932, provide the leading example here. Yet the Republicans offered the opposite version of deliberate electoral obliviousness in the half-century after the Civil War, though this was a matter whereby expected electoral victories were what left the factions free to fight out nominating politics among themselves. One further large lesson from this tour of the five major patterns characterizing the politics of presidential nomination is their inherent variety, across time and between parties. A familiar and long-standing dynamic by which a bandwagon runs through changing fields of candidates whose fortunes are centrally in the hands of party factions does not require one dominant set of personalities, strategies, or issues. It does not even cause a single pattern to emerge in individual eras. Nor does it contribute to patterns that are diagnostic of nominating politics in one party as opposed to the other. In the same way, intense factional conflicts can sometimes be overcome by consensual candidates, who then go on both to win and to lose subsequent elections. Just as intense factional conflicts can produce compromise candidates, who are likewise capable both of winning and losing the presidency. So in the end, what remains is a governing dynamic for the politics of presidential nomination, one that has run regularly and relentlessly since nominating politics moved outside legislatures and inside political parties. From the 1830s onward, presidential nominations have been a product of sequential mutual adjustments among all the major players: presidential aspirants, campaign managers, state parties, state electorates, financial contributors, and the mass media of information. Each successive round in this contest, whether it involves convention ballots, primary elections, or some mix thereof, provides new strategic considerations for everyone involved. In response, they adjust, until the succession of these adjustments produces a presidential nominee.

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

259

Over the years, the occupational backgrounds and career patterns of these nominees have changed. Different points in time were inevitably characterized by prior offices and career trajectories that best captured the changing mix of resources that were a response to three key political considerations: the institutional forms for delegate selection, the internal structure of the political parties, and the external balance between the two major parties. In the days before institutionalized political parties with a formalized process of delegate selection, federal executive office was the crucial seedbed for presidential nominations. Once formal and explicit nominating procedures appeared, unfolding within political parties and culminating in national party conventions, these seedbed offices changed. A career in politics would remain central, with only the rarest of exceptions. But the leading institutional generator of presidential nominees would shift: first to Congress, especially the House of Representatives; then to Governorships, for a long stretch of time; then to Senatorships in the modern era. Around these favored backgrounds was a set of further influences that were critical to making a particular Congressman, Governor, or Senator into the presidential nominee of a major party. Shifting policy conflicts were one of these evolving influences, a kind of external shock to the nominating politics of the time that included national expansion, economic development, and cultural standards for both. Shifting struggles among social groups were another such constantly evolving input, as policy conflicts pitted different groups against each other more or less in spite of themselves: free versus slave, industrial versus agricultural, blue-collar versus white-collar, black versus white, progressive versus traditionalist, and on into the future. Though the central point here remains that all of these were simultaneously channeled by the political logic of the time, first by way of the party factions that had to be assembled in order to result in a nomination—these repeatedly came to the fore—then by way of the state coalitions and voter constituencies that had to be assembled in order to result in the election of any given nominee. Indeed, it was the interaction of three grand and continuing intermediary influences—a dominant nominating dynamic, a diagnostic seedbed for successful aspirants, and the factional struggles to nominating politics itself—that did the actual work of choosing individual winners, whatever candidates, strategists, pundits, reporters, or (even!) professors may have believed.

260

B. E. SHAFER AND E. M. SAWYER

In fairness, these basic structural influences were the hardest things for observers of contemporary politics to perceive in their own time. So it remains understandable that candidates, their staffs, those covering them, plus a wide swath of the interested public persisted in focusing on things like the rules of the national contest, the strategies pursued by the major players within those rules, and the sequence by which the delegate selection contests unfolded. Said less politely, most observers were focused on idiosyncratic characteristics that reflected and expressed but also obscured the dynamics actually producing any given individual nominee. If the only point was explaining what had just happened to candidate X in primary Y, this alternative approach still offered some value, though a limited value even then. But if the point is instead to explain how presidential nominating politics works, then such an approach, torn erratically and idiosyncratically from the funnel of causality, has little value. In the recurrent process leading from an aggregate nation to two presidential nominees, the vast bulk of decision-making—that is, of winnowing toward those potential presidents—is done by bandwagon dynamics, candidate seedbeds, and factional structures. Idiosyncrasy and random variation is what is left for anything else to explain.

A Gain-Deficit Afterword on 2020 For most purposes, the nominating contests of 2020 have been folded into the appropriate tables on the bandwagon dynamic, candidate careers, and party factions, becoming one more data-point in an effort to unpack the funnel of causality as it applies to presidential nominations across American history. But there was one aspect of the 2020 contests that might merit a separate closing comment. This involved their coverage, which was impressively discordant by comparison to what actually happened. When examined through the Gain-Deficit Ratio, the 2020 Democratic race looked extremely orthodox. Yet it seemed dramatic and surprising to a general public informed by a discordant consensus among professional interpreters. So it seems worth returning to 2020 because its nominating politics—orthodox dynamic, familiar factions, quick resolution—seemed an ultimate surprise to so many. In its specifics, and as if to add an exclamation point to the story of nominating politics across most of American history, the nominating struggles of 2020 actually featured not one but two stereotypical, albeit stereotypically opposite, contests. One made the

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

261

dominant bandwagon dynamic as irrelevant as it ever gets: this was the uncontested renomination of Donald Trump, sitting president. The other showcased this dynamic at its most operatively powerful, when Joe Biden was quickly and decisively nominated after a couple of early and idiosyncratic losses (Table 5.6A). The renomination of a sitting president has become nearly automatic in modern American politics, and Donald Trump in 2020 was only the latest beneficiary. Sitting presidents begin with serious name recognition, a sine qua non of the plebiscitary system. Most are actually admired as well, at least within their own party, while the others have at least evidently done the job. Moreover, they have serious concrete resources by way of government itself. And there is no alternative mechanism for easing them out: a party-splitting effort is essential; it is hard to mount; and it is likely to doom the nominee in November, whether this is the incumbent or the challenger. Recall that the Reagan challenge to Gerald Ford in 1976 was followed by the presidency of Democrat Jimmy Carter, just as the Table 7.6 Gain-Deficit calculations for 2020 A. The Republicans Date

Donald Trump

All others

Feb 3 Feb 11 Feb 22 March 3 March 10 March 17

NA 0.018 0.021 1.938 1.485 Reached nominating threshold

NA 0 0 0 0 0

B. The Democrats Date

Joe Biden

Bernie Sanders

Elizabeth Warren

Pete Buttigieg

Amy Klobuchar

Feb 3 Feb 11 Feb 22 Feb 29 March 3 March 10 March 14 March 17

NA 0.000 0.004 0.020 0.513 0.201 0.002 0.384

NA 0.005 0.012 0.008 0.413 0.112 0.003 0.134

NA 0.000 0.000 0.000 0.037 NA NA NA

NA 0.005 0.002 0.000 NA NA NA NA

NA 0.003 0.000 0.000 NA NA NA NA

262

B. E. SHAFER AND E. M. SAWYER

Kennedy challenge to Carter in 1980 was followed by the elevation of Reagan. Yet the story of nominations without a sitting president is different, being much more reliably a product of factional struggles. The plebiscitary system of presidential nomination has all but eliminated uncontested nominations for the out-party, even when there is an obvious frontrunner with broad party support and strong polling numbers. Even in that situation, the plebiscitary system provides what is effectively an institutionalized invitation to enter the contest, and the Democratic nominating field of 2020 began by confirming the power of this standing invite. For the Democrats this time—the Republicans had seen a lesser version of the same phenomenon in 2012 and 2016—the presence of a long series of preliminary candidate debates encouraged a plethora of aspirants to announce that they were running for president, so many that the media televising these debates had to divide them into two (large) panels. Unfortunately, commentators went on to conclude that this plethora of candidates betokened an extended contest, probably resulting in the nomination of an insurgent, possibly reaching all the way to the convention without a predetermined majority.20 Yet when the real delegate selection finally began, the contest was over quickly, in fact in record time. Rare few Democratic contests in the modern world have been this orthodox and this abrupt. Chapter 5 has already recognized this resolution as the earliest among the contested nominations in the entire plebiscitary era, now more than forty years old. Table 7.6B goes on to present the details of this 2020 result. The Iowa caucuses were on February 3; Super Tuesday was on March 3; and on that date, Joe Biden was effectively nominated. The next morning, as the results were confirmed and translated into delegates, Biden clearly led Bernie Sanders while having breached the Gain-Deficit threshold. No one has ever failed to be nominated once that has occurred. Yet Sanders did actually cross this threshold too, which allowed supporters to hang onto the prospect of a different result. So Table 7.6B adds the next two weeks, when Biden again met the nominating standard, while crushing out Sanders completely. How could what was roughly consensual reportage have ended up so far off the actual mark in so short a time? Rude answers abound. But in fact, if the Gain-Deficit Ratio made the underlying reality explicit, it also provided a simple explanation for a shared misjudgment. In 2020, this shared fallacy came from an implicit but generalized belief, to the

7

ETERNAL BANDWAGON: NOMINATING POLITICS …

263

effect that many contenders meant a longer nominating struggle. When a couple of early contests seemed anomalous, this “wisdom” seemed confirmed. But in fact, the bandwagon dynamic underlying all such contests, long or short and not just this one, tells the opposite story. Which is to say: the bandwagon rolls more rapidly in a fragmented field, as it did for the Democrats in 2020 and as it had for the Republicans in 2016. What allows contests to run longer is the opposite situation, that is, the early appearance of a field of aspirants with two serious contenders, as was the case for the Democrats in both 2008 and 2016, but not in 2020. That is what makes it possible for a leader and a challenger—sometimes swapping roles along the way—to extend the contest by trading wins over a longer period. For 2020, by contrast, the bandwagon dynamic and its associated causal funnel had done what they always do. It was just that they had done it more rapidly than in more seriously contested nominations.

Notes 1. Among many, Nelson W. Polsby, Aaron Wildavsky, Steven E. Schier, and David A. Hopkins, Presidential Elections: Strategies and Structures of American Politics, 15th ed. (Landover: Rowman & Littlefield, 2019), and John Sides, Daron Shaw, Matt Grossman, and Keena Lipsitz, Campaigns and Elections, 3rd ed. (New York: W. W. Norton, 2015); Candice J. Nelson and James A. Thurber, eds., Campaigns and Elections America Style: The Changing Landscape of Political Campaigns, 5th ed. (New York: Routledge, 2018). 2. Roosevelt did not seek a second consecutive full term, passing the baton to William Howard Taft in 1908, but he did come back to seek just such a second term, unsuccessfully, in 1912. 3. Recall that both Truman and Johnson received a first partial term upon the deaths of their presidents, Franklin Roosevelt and Lyndon Johnson, respectively; were then elected to a full term of their own; but declined to seek a second full term. By hindsight, both appear fully capable of securing a second renomination of their own. Still, Truman’s fortunes against Dwight Eisenhower, the Republican nominee in 1952, must otherwise be judged to have been tenuous, and Johnson did withdraw after the New Hampshire primary in 1968, even though his performance in that opening primary was impressive, winning it as a write-in against Senator Eugene McCarthy of Minnesota, at that point the leading challenger for the Democratic nomination and formally on the ballot.

264

B. E. SHAFER AND E. M. SAWYER

4. Jessica M. Lepler, The Many Panics of 1837: People, Politics, and the Creation of a Transatlantic Financial Crisis (Cambridge: Cambridge University Press, 2013), and Alisdair Roberts, America’s First Great Depression: Economic Crisis and Political Disorder After the Panic of 1837 (Ithaca: Cornell University Press, 2012). 5. Making Truman the only one to survive as president: Irwin Ross, The Loneliest Campaign: The Truman Victory of 1948 (New York: New American Library, 1968), and Zachary Karabell, The Last Campaign: How Harry Truman Won the 1948 Election (New York: Random House, 2000); most especially Andrew W. Busch, Truman’s Triumphs: The 1948 Election and the Making of Postwar America (Lawrence: University Press of Kansas, 2012). 6. In 1852, Fillmore’s association with the Compromise of 1850 (under his presidency) made him indigestible to much of his party, while in 1856, Pierce’s association with the Kansas–Nebraska act had the same effect. Paul Finkelman, Millard Fillmore (New York: Times Books, 2011), and Michael F. Holt, Franklin Pierce (New York: Times Books, 2010). 7. Lewis L. Gould, Four Hats in the Ring: The 1912 Election and the Birth of Modern American Politics (Lawrence: University Press of Kansas, 2008), and Geoffrey Cohen, Let the People Rule: Theodore Roosevelt and the Birth of the Presidential Primary (New York: W. W. Norton, 2016) is good on the mercurial character of Roosevelt. 8. Introduced in Chapters 5 and 6. See also, Larry M. Bartels, “Trench Warfare: The Reagan and Kennedy Challenges”, Chapter 9 in Bartels, Presidential Primaries and the Dynamics of Public Choice (Princeton: Princeton University Press, 1988), and Jules Witcover, Marathon: The Pursuit of the Presidency 1972–1976 (New American Library, 1977); Timothy Stanley, Kennedy vs. Carter: The 1980 Battle for the Democratic Party’s Soul (Lawrence: University Press of Kansas, 2010). 9. W.H. Lawrence, “Truman, Barkley Named by Democrats; South Loses on Civil Rights, 35 Walk Out”, The New York Times, July 26, 1948, and Kari Frederickson, The Dixiecrat Revolt and the End of the Solid South, 1932–1968 (Chapel Hill: University of North Carolina Press, 2003). 10. And the problem was acute: the Democrats had returned to a majority in the House of Representatives by 1876, yet between 1876 and 1912, they could duplicate this majority in the presidency only under a reform governor from New York. 11. This electoral logic reached even farther in the sense that the other state of any size that could be captured by either party was Indiana, and Thomas A. Hendricks, variously congressman, senator, and governor from Indiana was the vice presidential candidate with Samuel Tilden in 1876 and with Grover Cleveland in 1884, while in the interim election in 1880, it was

7

12.

13.

14.

15.

16.

17.

ETERNAL BANDWAGON: NOMINATING POLITICS …

265

former congressmen William H. English of Indiana who was vice presidential candidate with Winfield Scott Hancock. Vice President Hendricks died in 1885 or he would have been the running mate for Grover Cleveland in 1888 as well. Robert C. McMath, American Populism: A Social History, 1877–1898 (New York: Hill & Wang, 1992), and Michael Kazin, A Godly Hero: The Life of William Jennings Bryan (New York: Alfred A. Knopf, 2006). The party did take one election off, from this Bryan sequence, under conservative jurist Alton Parker in 1904, but when electoral results were no better, it returned consensually to Bryan in 1908. Michael McGerr, A Fierce Discontent: The Rise and Fall of the Progressive Movement in America, 1870–1920 (New York: Oxford University Press, 2005), and Douglas B. Craig, Progressives at War: William G. McAdoo and Newton D. Baker, 1863–1941 (Baltimore: Johns Hopkins University Press, 1913). Robert W. Cherney, American Politics in the Gilded Age, 1868–1900 (Arlington Heights: Harlan Davidson, 1997), and Robert D. Marcus, Grand Old Party: Political Structure in the Gilded Age, 1880–1896 (New York: Oxford University Press, 1971). The leading exception—and as a result, the leading challenges to being placed in Table 4 instead of Table 5—were the two nominees in 2016, Hillary Clinton for the Democrats and Donald Trump for the Republicans. Both had clear leads in the polls leading into the opening delegate selection contests in Iowa and then New Hampshire. Both were ultimately to convert their front-runner status into nominations. Yet both had to endure an extended nominating contest, albeit in large part because their remaining opponents had reasons for continuing the contest well beyond the point when they themselves could possibly be nominated. For serious attention to those who did indeed persevere beyond the signal from the Gain-Deficit Ratio that the contest was over, see Byron E. Shafer and Elizabeth M. Sawyer, “Institutional Dynamics and NomiNating Processes”, Presidential Studies Quarterly 48(2018), 586–610. For another systematic looks at the losers, see Barbara Norrander, “The Attrition Game: Initial Resources, Initial Contests, and the Exit of Candidates During the US Presidential Primary Season”, British Journal of Political Science 36(2006), 487–507. For a look at alternative uses of national party conventions for publicrelations purposes, see Byron E. Shafer, “The Pure Partisan Institution: National Party Conventions as Research Sites”, Chapter 14 in L. Sandy Maisel and Jeffrey M. Berry, The Oxford Handbook of American Political Parties and Interest Groups (New York: Oxford University Press, 2010).

266

B. E. SHAFER AND E. M. SAWYER

18. The lone exception was James G. Blaine in 1884, who converted frontrunner status into a Republican nomination, but went on to lose the general election to the Democratic nominee, Grover Cleveland. 19. These differential resolutions for factional disputes revolving around the activist wing of the party are covered in Chapter 5 for the Democrats and Chapter 6 for the Republicans. 20. “The Democratic Race Is Now Sanders Versus the Field, and a Contested Convention Possible Awaits”, The Washington Post, February 20, 2020, and “Moderate Democratic Candidates Wrestle Against Leftward Lurch”, Fox News, February 20, 2020; “Debate Shows Bernie Sanders Could Win Most Votes but be Denied Nomination”, The Guardian, February 20, 2020; “How to Win the Democratic Nomination and Why It Could Get Complicated”, The New York Times, February 22, 2020.

Index

A Adams, John, 43–44 Adams, John Quincy, 46–47, 87–88, 122 Agnew, Spiro T., 142 Aldrich, John H., 79 Alexander, Lamar Jr, 145 Allison, William B., 131 American (Know Nothing) Party, 127 Andrew E. Busch, 33 Anti-Masonic Party, 122 Arthur, Chester A., 69–72, 130, 197, 199, 200, 216, 246–247 B Bachmann, Michelle, 148 Baer, Kenneth S., 119 Bain, Richard C., 34, 35, 117 Bandwagon Introduced, ix as a dynamic, 2–7 summarized, 32, 37, 193, 230, 241, 263

Bankhead, William B., 100 Barkley, Alben W., 101 Bartels, Larry M., 7, 119, 157 Bates, Edward, 11 Bayard, Thomas F., 93–94, 218 Bell, John, 90 Bentsen, Lloyd, 107 Biden, Joseph R. “Joe”, 110 Blaine, James G., 56, 129–131, 175–176, 198, 200, 203–204, 215 Bland, Richard, 95 Borah, William, 136, 138 Bradley, William W., 109–110 Breckinridge, John C., 90 Bricker, John W., 139 Bristow, Benjamin H., 129 Brown, Edmund G. Jr., 108 Bryan, Charles W., 98 Bryan, William Jennings, 55, 95–97, 152, 197, 198, 217–218, 250 Bryce, James, 81 Buchanan, James, 30, 88, 89–90, 216

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 B. E. Shafer and E. M. Sawyer, Eternal Bandwagon, The Evolving American Presidency, https://doi.org/10.1007/978-3-030-51799-1

267

268

INDEX

Buchanan, Pat, 144–145 Burnham, Walter Dean, 33, 79 Bush, George H.W., 69, 143–144 Bush, George W., 74, 145–146 Butler, Nicholas M., 13 Byrd, Harry F., 100 C Cain, Herman, 148 Calhoun, John C., 88 Cameron, Simon, 11–12, 127 Campbell, Angus, x Candidate Strategies, 194–195 How to run, 200–204 Iowa & New Hampshire, 212–213 Iowa caucuses, 210–212 New Hampshire primary, 210 Where to run, 207–212 Whether to run, 195–200 Cannon, Joseph G., 133 Carter, James E. “Jimmy”, 4, 106–107, 200, 246, 248 Cass, Lewis, 30, 50, 88, 89, 202, 217 Ceaser, James W., 33, 82 Chambers, William Nesbitt, 79 Chase, Salmon P., 127 Cheney, Richard B. “Dick”, 146 Civil War, effect on candidate careers, 42 Democratic nominations during postbellum years, 91–92 effect on Democratic Strategy, 162 Clark, James B. “Champ”, 25–29, 96–97, 178, 204, 221 Clay, Henry, 49, 87, 123–125, 150, 175, 179, 197 Cleveland, Grover, 55, 93, 94, 114, 130, 177–178, 200, 217–218, 246, 250 Clinton, DeWitt, 45 Clinton, Hillary R., 8–10, 50, 110–112

Clinton, William “Bill”, 74, 107–108, 196, 209 Colfax, Schuyler, 128 Collat, Donald S., 2 Compromise of 1850, 125 Congresspersons as Presidential Nominees, 47–49, 51 Conkling, Roscoe, 129 Constitutional Union Party, 90 Contested Nominations, definition of, 16–19 Coolidge, Calvin, 59, 69, 72–73, 135–136, 200, 205 Cox, James M., 30, 97–98, 207, 251–252 Crawford, William J., 87 Cuomo, Mario, 196 Curtis, Charles, 136, 137

D ‘Dark Horse’ Candidacies, 30, 152–153, 202 David, Paul T., 84 Davis, John W., 30, 98, 205, 207, 251 Dawes, Charles G., 136 Dayton, William L., 127 Dean, Howard, 109 Delegate Allocation democratic reform, 177–178 evolution of the rules, 173–174 Introduction of, 161 Reapportionment of 1936, 17, 163–164 Republican reforms, 174–176 Delegate Apportionment democratic reforms, 163–165, 169–173 evolution of the rules, 165 Republican reforms, 163, 165–169 Democratic-Republicans, 45

INDEX

Depew, Chauncey, 131 Dever, Paul A., 102 Dewey, Thomas, 62, 67, 138–140, 198 DiSalvo, Daniel, 117, 156, 236 Dole, Robert J. “Bob”, 144–146 Donald E. Stokes, x Donelson, Andrew Jackson, 126 Douglas, Stephen A., 16, 30, 50, 88, 89–90, 216 Dukakis, Michael S., 107 E Edmunds, George F., 130 Edwards, John, 109–110 Eisenhower, Dwight D., 50, 62, 67, 140, 168, 198, 204 English, William H., 93 F Factions. See Party Factions, Democratic faction; Party Factions, Republicans faction defined, 85–86 Fairbanks, Charles W., 133, 135 Farley, James A., 100 Federal Executives as Presidential Nominees, 42–44 Federalists, 45 Fillmore, Millard, 69–72, 124–125, 126, 199, 247 Forbes, Malcolm “Steve” Jr, 145 Ford, Gerald, 66, 71, 73, 106, 142–143, 200, 246–248 Ford, Henry Jones, 81 Fremont, John C., 51, 76 Funnel of Causality and candidate backgrounds, 37, 78 and nominating strategies, 195 and party structure, 85–86 and rules reform, 160

269

Introduced, viii G Gain-Deficit Ratio 2016, 7–10 adjusted, 28–29 discriminatory power, 260–261 general success of, 11–12 Introduced, 3–5 under altered delegate rules, 184–186 Garfield, James A., 4, 56, 69, 130, 176, 197, 199, 203, 215, 247 Garner, Nance, 23–24 General Elections and Nominating Politics, 242–243 consensual nominations, 250–252 contested renominating efforts, 247–248 extended contests, 255–257 front-runner victories, 252–255 uncontested renominating efforts, 244–246 Generals as Presidential Nominees, 51–52 Gephardt, Richard A., 109 Gienapp, William E., 34, 156 Gingrich, Newt, 148 Giuliani, Rudolph W., 147, 212 Goldwater, Barry, 68, 104, 140–142, 206, 207 Gore, Albert A. “Al”, Jr., 108–109, 233 Gore, Albert A., Sr., 102 Gould, Lewis L., 35, 82, 189 Governors as Presidential Nominees, 52–54, 61–62 Grant, Ulysses S., 50–52, 56, 127–128, 154, 176, 197, 203, 216 Gray, Isaac P., 94 Greeley, Horace, 54, 92, 202

270

INDEX

Green, Dwight H., 139 The Green Papers, 34, 192 Grossman, Matt, 263 H Half-breed Republicans, 129 Hancock, Winfield Scott, 30, 91, 93, 196, 217 Hanna, Mark, 131, 132 Harbord, James G., 137 Harding, Warren G., 12–14, 57, 59, 69, 135, 136, 205, 207 Harmon, Judson, 178 Harrison, Benjamin, 55, 130–131, 251–252 Harrison, William Henry, 51, 69–72, 123, 124, 175, 199, 247 Hart, Gary, 107–108 Hayes, Rutherford B., 56, 129–130, 175, 197, 198, 203, 215 Heale, M.J., 79 Hearst, William Randolph, 96 Hendricks, Thomas H., 30, 91–94, 202 Hill, David B., 94, 95 Hobart, Garret A., 132 Holt, Michael F., 34, 35, 80, 117, 155, 156 Hoover, Herbert, 59, 99, 136–137, 200, 205, 246 Howe, Daniel Walker, 80, 155 Huckabee, Michael D., 147 Hughes, Charles Evans, 133, 135, 167–168 Humphrey, Hubert, 60, 62, 101, 103–104, 106, 170–171, 180, 206, 220–221 I Iowa caucuses created, 210

influence, 210–212

J Jackson, Andrew, 46–51, 87–89, 113, 114, 122 Jackson, Henry M., 106 Jackson, Jesse L., 107–108, 233 James, Arthur M., 138 Jamieson, Kathleen Hall, x Jefferson, Thomas, 43–45, 69 John J. Pitney, Jr., 33 Johnson, Andrew, 30, 69–73, 91, 127–128, 198–199, 247 Johnson, Hiram, 13, 134, 136 Johnson, Lyndon, 60, 62, 68–73, 103–104, 200, 206 Johnson, Richard M., 16, 88 Judith H. Parris, 34, 117

K Kefauver, Estes, 101–102, 204, 207 Kelly, John, 177 Kennedy, Edward M., 106 Kennedy, John F., 62, 68–73, 102–103, 104, 205–207 Kennedy, Robert F., 60, 104, 106 Kern, John W., 96, 109–110 King, Rufus, 45 Kleppner, Paul, 79 Knox, Frank, 138 Knox, Philander, 133

L LaFollette, Robert M., 59, 133, 220 Landon, Alf, 62, 137, 138 Liberal Republican Party, 54, 92, 128, 129 Lincoln, Abraham, 11–12, 49, 69–72, 126–128, 150–151, 199, 215, 247

INDEX

Lipsitz, Keena, 263 Lodge, Henry Cabot, 135, 141 Logan, John A., 130 Lowden, Frank, 13–14, 135–136, 205–206

M Madison, James, 43–45 Majority Rule in operation, 29 Introduced, 14–15 Marcy, William L., 30, 89 Marshall, Thomas R., 97 Mayer, William G., 84, 239 McAdoo, William Gibbs, 30, 97–98, 99, 205, 251–252 McCain, John S. III, 146–148, 212 McCarthy, Eugene, 59, 103, 104, 220, 221 McClellan, George, 54, 91 McCloskey, Paul N., 142 McCormick, Richard L., 82, 156 McCormick, Richard P., 79, 80 McDonald, Joseph E., 93 McGovern, George, 60, 104, 105–106, 171–173, 182 McKinley, William, 56, 69–73, 96, 131–133, 215 McLean, John, 126 McNary, Charles W., 138 McNider, Hanford, 137 Michael G. Hagen, x Mixed System, 11, 12–13 and candidate backgrounds, 42, 56–62 defined, 11 Momentum Introduced and defined, 5–7 Mondale, Walter, 107 Monroe, James, 43–45 Moos, Malcolm, 84

271

Morton, Levi P., 129, 131, 132 Multi-Ballot Conventions, 67–68, 201

N National Republican Party, 122 New Hampshire primary created, 209–210 influence, 210–212 Nixon, Richard, 62, 65, 68–69, 140–142, 152 Nominating Contests, Democrats 1824-1860, 86–90 1864-1892, 91–94 1896-1928, 94–98 1932-1968, 99–104 1972-1992, 104–108 1996-2020, 108–112 Nominating Contests, Republicans 1856-1872, 126–128 1876-1892, 128–131 1896-1912, 131–133 1916-1928, 133–136 1932-1940, 137–138 1944-1960, 139–140 1964-1980, 141–143 1984-2004, 143–146 2008-2016, 146–149 Nominating Contests, Whigs, 123–125 Nomination and Election of the Presidents and Vice President of the United States, including the Manner of Selecting Delegates to the National Party Conventions , 189 Norrander, Barbara, 7, 33

O Obama, Barack, 74, 110–111 Overacker, Louise, 82, 237, 238

272

INDEX

P Palin, Sarah L., 148 Palmer, A. Mitchell, 30, 97–98 Parker, Alton B., 96, 197 Party Factions. See also Nominating Contests, Democrats; Nominating Contests, Republicans and two-thirds rule, 15–16 defined, 85–86, 159 in operation, 257–258 Republicans faction, 152 Payne, Henry G., 93, 196 Pence, Michael R., 149 Pendleton, George H., 30, 91 Perry, Rick, 148 Phillip E. Converse, x Pierce, Franklin, 30, 50–51, 88, 89–90, 199, 202, 247 Pinckney, C.C., 45 Plebiscitary System and Candidate Backgrounds, 42–43, 108–114 defined, 21 Polk, James K., 51, 88, 89–90, 198, 202, 216 Polsby, Nelson W, 263 Populist Party, 94 Pre-Convention System and Candidate Backgrounds, 40–45 defined, 42 Presidential Primary Elections., 58–60, 64–65. See also New Hampshire Primary timing, 226–234 Presidential Renominations, 73–74 Progressive Party, 134–136 Proportionality Rules, 181–186 Pure Convention System, 11 and Candidate Backgrounds, 42–43 defined, 11 in Antebellum Era, 47–51 in Postbellum Era, 52–56

Q Quayle, J. Danforth, 145 R Ralph M. Goldman, 84 Randall, Thomas J., 93 Reagan, Ronald, 65, 142–143 Reed, Thomas, 132 Regina L. Wagner, 82 Reid, Whitelaw, 131 Robinson, Joseph T., 98 Rockefeller, Nelson A., 140–148, 206, 220–221 Rogowski, Ronald, 2 Romney, Willard “Mitt”, 148, 212 Roosevelt, Franklin D., 16, 56, 62, 67–73, 97, 98–101, 113, 114, 138, 164, 251 Roosevelt, Theodore, 55–57, 59, 69–72, 132–135, 163, 168, 200, 204, 247 Root, Elihu, 135 Rules Reform, 186–188 Russell, Richard B., 101, 102 S Sanders, Bernard “Bernie”, 8–10, 110–111 Santorum, Rick, 149 Scott, Winfield, 51, 54, 124, 125, 175, 196, 199, 217 Scranton, William W., 142 Senators as Presidential Nominees, 57, 60–63 Seward, William H., 11–12, 127, 215 Seymour, Horatio, 30, 54, 92, 94, 202, 217, 250 Shafer, Byron E., 33, 81, 83, 119 Shaw, Daron, 263 Sherman, James S., 133–134 Sherman, John, 129, 176, 215

INDEX

Sides, John, 34, 120, 157 Silbey, Joel H., 34, 80, 117 Simon, Paul, 107 Smith, Alfred E., 4, 30, 97–98, 205, 251 ‘Smoke-Filled Room’, 14 Sparkman, John J., 102 Stanley Kelley, Jr., 2 Stassen, Harold E., 139–140, 198 State Party Strategies, 213 adoption of a presidential primary, 218–222 ‘front-loading’, 229–234 imposition of a unit rule, 214–218 Institutional Options, 213–226 Selection of a Date, 225 temporal placement, 230–232 Stevenson, Adlai E., I, 62, 94, 96 Stevenson, Adlai E., II, 67, 101–102, 204, 207 Sundquist, James L., 33 Symington, Stuart, 103, 206 T Taft, Robert A., 138, 140, 168, 198, 204 Taft, William Howard, 56, 59, 132, 133–135, 163, 204, 221, 246, 247 Taylor, Zachary, 51, 69, 124, 199, 247 Tesler, Michael, 34, 120, 157 Thurman, Allen G., 93, 94 Tilden, Samuel, 54, 92–93, 177, 196, 217, 250 Truman, Harry, 50, 62, 69–73, 100, 200, 248 Trump, Donald J., 4, 66, 74, 76, 149, 153 Gain-Deficit Ratio and, 7–10 Two-Thirds Rule end of, 163–164 Introduced, 14–15

273

struggles over, 15–17 Tyler, John, 69–72, 124, 199, 247 U Underwood, Oscar W., 26 Unit Rule, 174–182 in operation, 214, 215 V Van Buren, Martin, 15, 16, 47, 50, 69, 74, 87–89, 123, 200–202, 216, 246 Vandenberg, Arthur H., 138 Vavreck, Lynn, 34, 120, 157 Vice Presidents as Presidential Nominees, 69–73 W Wade, Benjamin F., 128 Wallace, George, 106 Wallace, Henry, 73, 100 Warren, Earl, 139–140 Warren E. Miller, x Washington, George, 43–44, 50 Webster, Daniel, 123 Weeks, John M., 167 Whig Party, 51, 125–128 White, Hugh L., 123 White, Theodore H., 79, 82, 84, 118, 119, 190, 237 Wildavsky, Aaron, 263 Willkie, Wendell, 62, 67, 76, 137, 138, 153, 198, 207 Wilson, Woodrow, 26, 29, 61, 96, 97, 134, 178, 197, 204, 220, 250, 251 Wood, Leonard, 13–14, 135 Z Zaller, John, x