Eternal Bandwagon: The Politics of

Presidential Selection 1st ed. Edition

Byron E. Shafer
Visit to download the full and correct content document:
https://ebookmass.com/product/eternal-bandwagon-the-politics-of-presidential-selecti
on-1st-ed-edition-byron-e-shafer/
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 Elizabeth M. Sawyer
Department of Political Science Department of Political Science
University of Wisconsin–Madison University of Wisconsin–Madison
Madison, WI, USA 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 informa-
tion 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—hypothet-
icals in this superficially anomalous year were generated by the Republi-
cans. 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 even-
tual 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 nomina-
tion, the same metaphor allows room for the personal idiosyncrasies,
state peculiarities, and/or strategic interactions which might still deter-
mine 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 nomi-
nating 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 band-
wagon,” 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 band-
wagon 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 time-
period favors a small set of occupations and careers that provide the crit-
ical 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 presiden-
tial 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 under-
stood 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 pres-
idential 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 succes-
sors 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 histor-
ical detail, and masterfully synthetic, drawing on almost 200 years of expe-
rience 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 elec-
tion 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 impor-
tant 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 them-
selves, 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 The Bandwagon Dynamic and the Causal Funnel:

Institutional Dynamics and Nominating Politics 1

2 Deep Channels to a Presidency: Occupations

and Careers 37

3 Factions, Constituencies, and Candidates: The

Democrats 85

4 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

Index 267

xv
 List of Tables

Table 1.1 Gain-Deficit calculations for 2016 9

Table 1.2 Delegate totals and Gain-Deficit Ratios: the
Republicans, 1860 11
Table 1.3 Gain-Deficit Ratios: the Republicans, 1920 13
Table 1.4 Gain-Deficit Ratios for major-party nominees across
time 18
Table 1.5 The Democratic nomination of 1912 27
Table 2.1 Preceding occupations of presidential nominees 41
Table 2.2 Preceding occupations of presidential nominees: the
pre-convention era, 1789–1828 44
Table 2.3 Preceding occupations of presidential nominees: the
antebellum era, 1832–1860 46
Table 2.4 Preceding occupations of presidential nominees, by
party 48
Table 2.5 Preceding occupations of presidential nominees: the
postbellum era, 1864–1908 52
Table 2.6 Preceding occupations of presidential nominees: the
mixed system, 1912–1968 57
Table 2.7 Preceding occupations of presidential nominees: the
plebiscitary system, 1972–2020 63
Table 2.8 Fortunes of the vice presidents who became president 70
Table 2.9 Fortunes of sitting presidents 75
Table 5.1 Delegation shares at the Republican convention of
1916: the impact of apportionment reform 166

xvii
xviii LIST OF TABLES

Table 5.2 Candidate votes after reapportionment: the Republican

nomination of 1916 167
Table 5.3 Delegation shares at the Republican convention of
1952: apportionment reform revisited 168
Table 5.4 Delegation shares at the Republican convention of
1976: the declining impact of apportionment reform 169
Table 5.5 State delegation shares at the Democratic convention of
1964: the impact of apportionment reform 170
Table 5.6 Delegation shares at the Democratic convention of
1968: the impact of apportionment reform 171
Table 5.7 Delegation shares at the Democratic convention of
1972: the impact of apportionment reform 172
Table 5.8 Candidate votes after abolition of the unit rule: the
Democrats, 1968 180
Table 5.9 The timing of nominations 184
Table 6.1 Nominating politics among incumbent presidents
seeking reelection 199
Table 6.2 Nominating politics among nonincumbents 201
Table 6.3 Iowa, New Hampshire, and the nomination:
1976–2016 211
Table 6.4 Number of presidential primaries across time 222
Table 6.5 A full plebiscitary picture: The changing matrix of
delegate selection 223
Table 6.6 State size by institutional form 224
Table 6.7 Participation levels by institutional form in 2016 225
Table 6.8 Clustering in the sequence of contests: The largest and
second-largest clusters by year 227
Table 6.9 Opening contests: Iowa and New Hampshire in the
plebiscitary era 229
Table 6.10 Super Tuesday and the resolution of nominating politics 231
Table 6.11 Strategic states and Super Tuesday: The role of the big
states 233
Table 7.1 Uncontested renominations of sitting presidents 244
Table 7.2 Contested renominations of sitting presidents 245
Table 7.3 Consensual nominations 249
Table 7.4 Front-runner victories 253
Table 7.5 Extended contests 254
Table 7.6 Gain-Deficit calculations for 2020 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 indi-
vidual 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 pres-
idential 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 explana-
tions or, of course, both. By itself, a longer time span would contribute to
explanations that were more fundamental and less idiosyncratic. By defini-
tion, 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 1

B. E. Shafer and E. M. Sawyer, Eternal Bandwagon,
The Evolving American Presidency,
https://doi.org/10.1007/978-3-030-51799-1_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 theo-
rizing 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, sepa-
rate 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 Nomina-
tions.”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 Presi-
dential 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 1848-
1948 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 nomi-

nations 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 Gain-
Deficit 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 scien-
tists 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 appar-
ently 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 crit-
ical 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 presi-
dential 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 relation-
ship between media coverage and delegate outcomes, between the stage
of the campaign and the level of uncertainty, between voter predispo-
sitions 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 candi-
date 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 compli-
cated 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 essen-
tially 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 band-
wagon 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. Subse-
quent 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 advan-
tages. 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 open-
ended 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 specula-
tions 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, poli-
tics had never denied nomination to such a candidate, and it was not
to deviate in 2016. Analysts concerned with different aspects of nomi-
nating 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 Donald Trump Ted Cruz Marco Rubio John Kasich

February 1 0.006 0.007 0.001 0.001

February 9 0.009 0.002 0.001 0.003
February 20 0.040 0 0 0
February 23 0.012 0.005 0.006 0.001
March 1 0.282 0.0218 0.085 0.017
March 5 0.071 0.074 0.012 0.008
March 6 0 0 0.021 0
March 8 0.095 0.068 0.001 0.014
March 10 0.001 0.001 0.002 0
March 12 0.001 0.027 0.010 0.008
March 15 0.420 0.064 0.006 0.074

B. The democrats
Date Hillary Clinton Bernie Sanders

February 1 0.010 0.009

February 9 0.004 0.006
February 20 0.009 0.006
February 27 0.017 0.006
March 1 0.293 0.178
March 5 0.033 0.027
March 6 0.005 0.009
March 8 0.058 0.040
March 12 0.002 0.001
March 15 0.327 0.193
March 22 0.046 0.054
March 26 0.034 0.078
April 5 0.035 0.037
April 9 0.006 0.005
April 19 0.148 0.092
April 26 0.301 0.165
May 3 0.057 0.046
May 7 0.006 0.003
May 10 0.016 0.019

(continued)
10 B. E. SHAFER AND E. M. SAWYER

Table 1.1 (continued)

B. The democrats
Date Hillary Clinton Bernie Sanders

May 17 0.086 0.071

June 4 0.012 0
June 5 0.065 0.027
June 7 2.18 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 spec-
ulations 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 assem-
bled 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 under-
line its utility for analyzing everything else that goes into presidential
nominating politics, including candidate backgrounds, factional struc-
tures, 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 majority-
rule 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 aspi-
rants 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 expec-
tations, 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

Seward Lincoln Cameron Chase Bates Others

1st Ballot 173.5 102 50.5 49 48 43

2nd Ballot 184.4 181 2 42.5 35 21.5
1st G-D Ratio +0.182 +0.598 −0.261 −0.070 −0.035 −0.146
3rd Ballot A 180 231 0 24.5 22 7.5
2nd G-D Ratio −0.090 +0.953 0 −0.094 −0.065 −0.066
3rd Ballot B 121.5 340 0 2 0 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 exam-
ples 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; Pennsyl-
vania 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 band-
wagon, 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 dele-
gates 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 Wood Lowden Johnson Harding Others

2 0.012 0.206 0.036 −0.015 −0.250

3 0.071 0.110 0.206 −0.001 −0.127
4 0.065 0.032 −0.021 0.007 −0.043
5 −0.008 0.088 −0.020 0.040 0
6 0.069 0.047 −0.062 0.072 −0.050
7 0.003 0 −0.027 0.243 −0.018
8 −0.067 −0.024 −0.031 0.078 −0.012
9 −0.206 −0.486 −0.012 0.965 0.015
10 −0.217 −0.202 −0.005 1.773 −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 nomi-
nation. 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 indif-
ference 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 forma-
tion 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 major-
party nominees would obviously—constitutionally—be demanded of a
president, so a majority inside the party seemed a reasonable corre-
sponding 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 geographi-
cally 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 total-
votes-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 Repub-
lican 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 renom-
ination 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 prospec-
tively 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 substan-
tially. 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 lead-
ership 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 reappor-
tionment 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 simple-
majority versus two-thirds rules, as well as by contested versus uncon-
tested 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 Gain-
Deficit 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 aspi-
rant 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

Republicans Democrats
Ratios for major-party
nominees across time A. The pure convention system

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

B. The mixed system

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

C. The plebiscitary system

1972 1972
1976 1976
1980 1980
1984 1984
1988 1988

(continued)
 1 THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL … 19

Table 1.4 (continued)

C. The plebiscitary system

1992 1992
1996 1996
2000 2000
2004 2004
2008 2008
2012 2012
2016 2016
2020 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 charac-
teristics 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 nomi-
nating 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 seven-
teen 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 desper-
ation. 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, hege-
mony 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 compro-
mise. 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 majority-
rule 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 require-
ments, 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 second-
largest 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 nomina-
tions 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 institu-
tional “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 conven-
tion 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 presiden-
tial 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 pres-

idential 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 pres-
idential 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 expe-
ditiously 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 intransi-
gent 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 fall-
back, 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 two-
thirds 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 candi-
dates, 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 consen-
sual 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 Repub-
licans never had any of these: a simple majority was the ultimate
decision.
• For the Democrats, there were even more cases where one candi-
date 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 accept-
able 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 renom-
ination 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 renom-
inated 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 candi-
date 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, Repre-
sentative 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 dele-
gates 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 stereo-
typically 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 guar-
antee 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 stan-
dard. 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 THE BANDWAGON DYNAMIC AND THE CAUSAL FUNNEL … 27

Table 1.5 The

A. Under a simple-majority rule
Democratic nomination
of 1912 BALLOT Wilson Clark

2nd 0.077 0.062

3rd 0.026 –
4th 0.013 0.02
5th 0.008 0
6th 0.016 0.02
7th – 0.048
8th – –
9th 0.005 0.038
10th – Nominated

B. Under a two-thirds rule

BALLOT Wilson Clark

2nd 0.041 0.021

3rd 0.014 –
4th 0.012 0.007
5th 0.004 0
6th 0.008 0.007
7th – 0.016
8th – –
9th 0.003 0.013
10th – 0.612
11th 0.011 –
12th – –
13th 0.005 0.041
14th 0.014 –
15th 0.004 –
16th 0 –
17th 0 –
18th – –
19th – –
20th 0.09 –
21st 0.021 –
22nd 0.003 –
23rd 0.008 –
24th 0.011 –
25th 0.008 –
26th 0.008 –

(continued)
28 B. E. SHAFER AND E. M. SAWYER

Table 1.5 (continued)

B. Under a two-thirds rule
BALLOT Wilson Clark

27th – 0.021
28th 0.107 –
29th – 0
30th 0.09 –
31st 0.062 –
32nd 0.001 0
33rd 0 0.004
34th 0.001 0
35th 0.064 –
36th 0.009 0.003
37th 0 –
38th 0.009 –
39th 0.013 –
40th 0 0.003
41st – 0.003
42nd – 0.02
43rd 0.855 –
44th 0.278 –
45th 0.043 0
46th Nominated –

was the numerical impact of a mechanical requirement, and not the insti-
tutionalized 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 immedi-
ately 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 two-
thirds 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 require-
ment 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 recur-
rent 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 two-
thirds 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 Gain-
Deficit 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 strik-
ingly 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 oper-
ating 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 asso-
ciated 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
Another random document with
no related content on Scribd:
kanssa! Hänen tunteensa ovat kuin lähteen kumpuamista, aavistaa,
että noiden ilmausten takana on todellinen suuri tunne. Itse
olemuksellaan, sillä tuskallisen kauniilla runoudella, joka säteilee
hänen persoonallisuudestaan, hän kirkasti koko kappaleen. Tämä
ilta oli huomattavin ja kaunein kaikista hänen näytäntöilloistaan.» —

*****

Alkujaan vuosien 1904—1905 kiertue oli suunniteltu laajemmaksi

kuin miksi se lopullisesti muodostui. Haastattelijoilleen Ida Aalberg
kertoi käyvänsä seurueineen m.m. Wienissä, Budapestissä,
Bukarestissa, ehkäpä Konstantinopolissakin. Matka Wieniin tuntuikin
varsin luonnolliselta; hänen seurueensa oli sieltä ja koko matkan
käytännöllisenä järjestäjänä toimi wieniläinen Ignaz Kugel. Tuosta
suunnitelmasta ei kuitenkaan tullut mitään, mutta sensijaan Ida
Aalberg antoi suuren joukon näytäntöjä Riiassa, Pietarissa ja
Moskovassa. Kiertueen ohjelmiston ytimenä olivat luonnollisesti
Ibsenin kappaleet ja Sudermannin »Koti», Suomessa oli esitetty
myöskin Strindbergin »Rikoksia», toisin paikoin taas Bernard Shaw'n
»Candidaa», »Kotia», »Eno Vanjaa», tai, kiertueen loppuaikoina
Venäjällä, Dumas'n »Kamelianaista» ja Zolan »Thérèse Raquin'ia».

Näytännöt Pietarissa ja Moskovassa menestyivät erinomaisesti.

Ida Aalberg näytteli Pietarissa useimmiten aivan täysille huoneille, ja
arvostelu oli erittäin kiittävää. Siellä hänellä olikin entuudestaan suuri
kannattajajoukko, mutta Moskovassa häntä ei aikaisemmin tunnettu.
Arvosteluista, joista lienee syytä ottaa tähän pari kohtaa, näkee kyllä,
että hänen kykynsä sielläkin saavutti täyttä tunnustusta.

»Ruskija Vjedomosti» kirjoitti hänestä:

 »Rouva Ida Aalberg, joka nykyjään antaa vierailunäytäntöjä
Moskovassa, on suuri ilmiö taiteellisessa maailmassa. Tähän saakka
taiteilija on esiintynyt Moskovan yleisölle kolmessa niin erilaisessa
osassa, että niiden täydellistä suorittamista varten vaaditaan hyvin
monipuolista lahjakkuutta. Sudermannin »Kodin» ja
»Kamelianaisen» esittävät monet näyttämötaiteilijat, mutta Ibsenin
»Rosmersholmiin» puututaan verraten harvoin. Tämä johtuu ehkä
siitä, että kolmesta mainitusta näytelmästä »Rosmersholm» vaatii
katsojalta mitä suurinta ajatuskyvyn, ei tunnekyvyn, jännitystä, ja
senvuoksi asettaa suorittajalleen erittäin vaikeita velvollisuuksia.
Seuratessanne rouva Aalbergin esitystä unohdatte nämä vaikeudet,
ja eteenne avautuu itsestään näytelmän aate. Taiteilijalla oli tässä
osassa suuri menestys. Eilen, »Kamelianaisen» esityksessä,
menestys oli yhtä ansaittu, kuten sen läsnäoleva yleisökin
yksimielisesti jälleen vakuutti. Jo tämän liiaksi näytellyn ja
vanhentuneen kappaleen ensi kohtauksesta katsoja tunsi taiteilijan,
joka osaa antaa elävyyttä ja todennäköisyyttä osan mitä
melodraamallisimmille momenteille. Onnettoman Margueriten
kärsimysten vaihteluun rouva Aalberg sulkee niin paljon vilpitöntä
tunnetta, naisellisuutta ja jotakin erityistä, hänelle ominaista
viehättävyyttä, että vanhaa juttua särjetystä sydämestä kuulee
pakostakin liikutettuna ja mielenkiinnoin. Joka näytöksen lopussa
huudettiin taiteilijatarta äänekkäästi esille.»

Arthur Luther arvosteli »Hedda Gableria» »Moskauer deutsche

Zeitungissa»;

»Kuinka on mahdollista, että niin suuren taiteilijan täytyy esiintyä

niin kurjassa huoneistossa! Tässä salissa [n.s. Romanovin
teatterisalissa], jossa toinen puoli ei kuule mitään, toinen puoli ei näe
mitään! Tällä näyttämöllä, missä dekoratsionit huojuvat, kun niihin
koskee! Vain meidän paras teatterimme olisi Ida Aalbergille kyllin
hyvä. Minä näin hänet eilen ensi kerran, mutta en tule tästä lähtien
laiminlyömään ainoatakaan hänen esiintymistään, sillä varmaankaan
hänen esiintymisensä ei tule supistumaan viiteen ilmoitettuun
vierailunäytäntöön.

Ida Aalberg on kauniin viivan taiteilija (ei kauniin pose'n). Jokaisen

yksityisen momentin tahtoisi saada ikuistetuksi. Tämä Hedda Gabler
todellakin kuolee kauneudessa. Naispaholainen hän kyllä on; hänen
pitää se olla, ja unohtumattomana tulee mielessäni säilymään kuva,
kuinka hän polvistuu uunin eteen ja palavat paperipalaset liekehtivät
hänen päänsä ympärillä. Mutta ei se, mikä on villin-saatanallista,
anna hänelle perussävyä (kuten Dumontille, jonka näimme viime
vuonna), vaan se, mikä on kissamaisen-mephistomaista. Kaunis
Mephistopheles? Niin, miksikäs ei? — kuten eräässä Heinen mitä
laajimmalle piireille tuntemattomassa tanssirunoelmassa. Kohtausta
oikeusneuvoksen kanssa toisessa näytöksessä ja loppua, missä hän
vetää sormiaan läpi alinomaa uurastavan Tean suortuvain, en ole
nähnyt vielä kenenkään niin esittävän. Tosin en tunne Dusea tässä
osassa,[29] mutta jo se, että täytyy mennä niin korkealle, kun yrittää
luonnehtia Ida Aalbergia, osoittaa riittävästi, mitä hän meille on.» —

Saksankielinenkään kritiikki ei Venäjällä ollut Ida Aalbergin

kielenääntämistä kohtaan niin ankara kuin vielä 1903 oli oltu
Berlinissä. Ibsenin runous ei ollut tuntematonta Pietarissa ja
Moskovassa, mutta siellä ei tehty sellaisia huomautuksia
käsitysvirheistä kuin Kööpenhaminassa, missä eräs arvostelija oli
sanonut:

»Vi bilder os dog ind at forstaa os paa Ibsen her hjemme —

kuvittelemme täällä kotonakin sentään ymmärtävämme Ibseniä.»
XVIII.

VIIMMEISIEN VUOSIEN TAITEELLISESTA TOIMINNASTA.

Ida Aalbergin kuolemasta on kulunut siksi lyhyt aika, että varsinkin

hänen elämänsä loppuvaiheista on liian vaikea laatia vähänkään
tyhjentävää ja samalla objektiivista kuvausta. Vuoden 1911 suuret
ristiriitaisuudet saivat aikaan, että hän kulutti kolmatta vuotta melkein
toimettomuudessa ja katkerassa ja tuskallisessa yksinäisyydessä,
mutta kuitenkin on hänen elämänsä viimmeinen vuosikymmen mitä
rikassisältöisin. Toistaiseksi on mahdollista käsitellä vuosien 1905—
1915 välistä aikaa vain perin suppeassa muodossa.

*****

Ida Aalberg oli kerran toisensa jälkeen yrittänyt näytellä vierailla

kielillä, mutta ei ole mikään ihme, että näin saavutetut voitot jäivät
puolinaisiksi ja ettei hän sitä tietä voinut saavuttaa pysyvää
eurooppalaista ja yleismaailmallista mainetta. Ainoa toinen todellinen
näyttelijä-suuruus, joka 1800-luvulla on yrittänyt samaa, näytellyt
vierailla kielillä, lienee Ristori. Ernest Legouvén puolittain
pakottamana italialainen näyttelijätär esiintyi Parisissa ranskan
kielellä, ja tästä kokeesta ei ainakaan hänellä itsellään ole huonoa
muistoa. Mutta kun Ristori myöhemmin mitä tarmokkaimpien
opiskelujen jälkeen näytteli Lady Macbethia englanniksi, havaittiin
hänen kielensä puutteelliseksi, ja kun hän Amerikassa uskalsi
eräässä saksankielisessä teatterissa tulkita englannin kielellä
pääosaa »Maria Stuartissa», onnittelivat yankeet häntä siitä, että
hän osasi niin hyvin saksaa! »Tuo viimeinen ponnistukseni
valtameren toisella puolella on antanut minulle sangen terveellisen
opetuksen taiteellisesta leikittelystä», kirjoitti Ristori, joka sen jälkeen
tyytyi näyttelemään vain italian kielellä.

Vuosien 1904—1905 suuri kiertue ei epäonnistunut taloudellisesti,

mutta ei myöskään siinä suhteessa antanut tulosta, joka olisi
houkutellut jatkamaan.

Kaikkialla, missä Ida Aalberg esiintyi saksan kielellä, hänet

kuitenkin saatiin tietää suomalaiseksi taiteilijaksi. »Suuri
suomalainen näyttelijätär» on hänen ainaisena nimityksenään,
esiintyipä hän missä tahansa ja minkälaisessa ympäristössä
tahansa. Ida Aalberg tunsi itsensä aina syvimmässä mielessä
suomalaiseksi, ja kun häntä monien Suomalaisen teatterin kanssa
sattuneiden ristiriitaisuuksien aikana syytettiin isänmaallisuuden
puutteesta, ei hän milloinkaan tahtonut myöntää tuota syytöstä
oikeutetuksi. Kaarlo Bergbomin teatteri ei hänestä vain ollut ainoa,
jossa suomalaista näyttämötaidetta voitiin palvella.

On joskus sanottu, että Kaarlo Bergbom oli joka kerta, kun Ida
Aalberg tuli Suomalaiseen teatteriin, äärettömän iloinen ja että hän
huokasi helpotuksesta joka kerta, kun näyttelijätär taas lähti pois.
Heidän kirjeenvaihtonsa antaa sangen paljon tukea tämmöiselle
väitteelle. Aina 1880-luvun alusta lähtien Suomalaisen teatterin
johtaja kirjoittaa niin nöyriä ja pyytäviä kirjeitä, että toisinaan tuntuu
melkein pahalta, kun näkee, miten kopeasti ja vaateliaasti Ida
Aalberg niihin vastaa. Kaarlo Bergbomia kohtaan Ida Aalberg esiintyi
loppuun asti kuin hemmoteltu lapsi, joka äksyilee ja niskottelee ja
kokeilee, kuinka pitkälle vanhempien rakkaus ulottuu. Suhde ei tosin
tullut milloinkaan myöhemmin niin huonoksi, kuin se oli ollut 1890-
luvun alkuvuosina, mutta liiankin selvää on, että Ida Aalbergin
käyttäytyminen vanhaa suosijaansa kohtaan myöhemminkin jätti
toivomisen varaa.

Vuonna 1902 Suomalainen teatteri muutti uuteen asumukseen ja

sai nimekseen Suomen Kansallisteatteri. Ida Aalberg oli antamassa
loistoa avajaisjuhlallisuuksille, mutta sai tuosta käynnistään
Helsingissä tyytymättömyydenaiheen Bergbomia vastaan.

Vuoden 1903:n alussa Kaarlo Bergbom kirjoittaa Ida Aalbergille:

»Arvoisa rouva Paroonitar!

Olen kolme viikkoa ollut vuoteen omana, sairastanut pahaa

keuhkokuumetta. Siinä syy, miksi en käynyt luonanne Helsingissä
ollessanne. Mutta hyvälle ystävälle Thérèse Hahlille olitte silloin
sanonut sanan, joka kovin kipeästi koski minuun. Syynä siihen,
ettette esiinny Helsingissä on muka »Bergboms bry sig icke om att
hafva mig» [Bergbomit eivät välitä minun tulostani sinne.]

Jos niin sanotte päästäksenne kysyjästä, jolle ette tahtonut antaa

vastausta, olkoon menneeksi. Mutta jos tosiaan sitä ajattelittekin,
erehdytte suuresti. Sekä minulle että sisarelleni on käyntinne
näyttämöllämme juhla. Tulkaa niin usein kuin tahdotte — aina olette
tervetullut. Meille iloksi, — teatterille kunniaksi. Muutamat
epäkohdat ovat Teitä suututtaneet ja Te laskette ne meidän
niskoillemme muistamatta, että me niistä saamme kärsiä paljon
enemmän kuin Te. Ne ovat Teille vastuksina muutamia viikkoja,
meille vuosien läpi.

Siis — jos tulette Maaliskuulla, tervetullut. Jos mieluummin tulette

Pääsiäisen perästä, tervetullut. Jos lykkäätte käyntinne syksyksi,
tervetullut silloinkin. Mainitkaa missä osissa tahdotte esiintyä, että
saamme valmistaa käännökset hyvän sään aikana. Jos tahdotte
suullisesti keskustella siitä, tulen Pietariin — mutta en nyt kohta,
olen vielä liian heikko. Mutta myöhemmin.

Kunnioittaen

Kaarlo Bergbom.»

Ida Aalberg kirjoitti vastaukseksi:

»Arvoisa Herra Tohtori!

Teidän kirjeenne 18 p:ltä olen vastaanottanut.

Te olette kummastunut kuinka minä olen uskaltanut sanoa

uskovani, ettei minua Kansallisteatterissa tarvita. — Minä olen sen
uskaltanut syystä, että olin saanut sen vakuutuksen ja iloitsen
suuresti nähdessäni, että siis se ei ole niin. Syytä tähän
vakaumukseen on minulle kylliksi teatterin avausjuhla-asiassa
annettu.

Tänä keväänä en enää tule.

Sitäpaitsi sanoitte syksyllä, ettei löydy minulle repertoiria. — Mitä

myöhempään tulee siitä voimme myöhemmin päättää. Iloitsen että
terveytenne on parempi ja toivon sydämellisesti pikaista
voimistumistanne voidaksenne edelleen kehittää ja nostaa
suomalaista taidetta.

Kunnioituksella

Ida Aalberg-Uexküll.»

Vaikka nämä kaksi kirjettä antavat verraten tyypillisen näytteen

heidän kirjeittensä yleisestä sävystä, on kuitenkin varmaa, että Ida
Aalbergin sydämen pohjalla oli mitä suurin hellyys Kaarlo Bergbomia
kohtaan. Kun tohtori kesällä 1903 on sairaana ulkomailla, tiedustelee
Ida Aalberg hätääntyneenä hänen osoitettaan saadakseen lähettää
hänelle sydämellisen osanottonsa ilmauksen. Ja saatuaan kuulla,
että Bergbom alkoi toipua, Ida Aalberg kirjoittaa Augusta af
Heurlinille: »Ole hyvä purista lämpimästi Tohtorin kättä ja sano
hänelle kuinka hartaasti toivon että hän meille takaisin annettaisiin,
suurta työtänsä ainakin suurella hengellänsä ohjaamaan.»

Kerrotaan, että Ida Aalberg itki kokonaisen vuorokauden saatuaan

tiedon Kaarlo Bergbomin kuolemasta. Varmaa on, että hän
jälkeenpäin tohtori Bergbomista puhuessaan ilmaisi mitä syvintä
kunnioitusta ja ihailua. Tämä myöhäinen kiitollisuus ja tunnustus on
vanha tarina hyvästä teatterinjohtajasta ja hyvästä näyttelijästä, Ida
Aalberg ei menetellyt paremmin eikä huonommin kuin hänen
ammattialallaan yleensä menetellään. Jos Ida Aalberg alusta lähtien
olisi mukautunut kaikkiin Kaarlo Bergbomin toivomuksiin, ei ole
lainkaan sanottu, että hän olisi näyttelijänä kasvanut niin suureksi
kuin kasvot. Esiintyessään vain vierailijana hän saattoi valita
mieleisensä ohjelmiston, sai välillä tarpeellisia lepoaikoja, säästyi
yleisön kyllästymiseltä enemmän kuin teatteriin vakinaisesti kuuluvan
näyttelijän on mahdollista. Ida Aalbergin henkilökohtainen suhde
Kaarlo Bergbomiin ei ole kuvaavaa vain yleiselle teatteripsyykelle
vaan sen takaa näkyy — jos nimittäin katsoo vain tämän suhteen
nurjaa puolta — itsesäilytysvaisto, jota ei voi muitta mutkitta tuomita.
Ristiriidat heidän välillään lienevät, niin näkyviä kuin ovatkin,
kuitenkin olleet verraten pinnallisia. On uskottavaa, että
kumpaisenkin suurimpana riemuna oli työskentely yhdessä, vaikka
Ida Aalberg ei ollut sitä aina kovin halukas tunnustamaan.

Kaarlo Bergbomin kuoltua Ida Aalbergista katosi halu

»eurooppalaistua». Tämän jälkeen hän esiintyi vain suomenkielisenä
näyttämötaiteilijana. Lukuunottamatta muutamia
hyväntekeväisyysnäytäntöjä Pietarissa, joissa hän suomenkielellä
tulkitsi Hugo von Hoffmansthalin »Elektraa», ja lyhyttä
suomenkielistä vierailumatkaa Unkariin ja Pietariin vuonna 1907
hänen koko taiteellinen toimintansa tapahtuu kotimaassa. Tuntuu
melkein siltä, että Bergbom vasta kuolemallaan kirkasti Ida
Aalbergille elämäntyönsä arvon.

*****

On joskus sanottu, että onnellisessa avioliitossa on vaikeata

määritellä, mikä osuus kummallakin puolisolla on yhteisessä
saavutuksessa.

Ida Aalbergin ja vapaaherra Uexküll-Gyllenbandin avioliitossa

tuntuu
erääseen aikaan olleen verraten varmasti määritelty työnjako.
Vapaaherra edusti yhteisessä taiteellisessa työssä teoreettista älyä,
Ida Aalberg oli toimeenpaneva voima.

Ida Aalberg oli erinomaisen henkevä ja älykäs nainen. Vapaaherra

Uexküll-Gyllenbandin suuresta vaatimattomuudesta johtui, että eräät
aikalaiset pitivät Ida Aalbergia perusluonteeltaan älyllisenä.
Tämmöinen erehdys on kyllä ymmärrettävää: on ollut suuri elämys
kuulla tuon henkevän naisen toistavan miehensä teoreettisia oppeja,
jotka koko luonteeltaan olivat erikoisia ja rohkeasti yksilöllisiä.

Ida Aalberg oli älykäs, mutta ei älyllinen. Kaikenlainen

teoretisoiminen oli hänen luonteelleen tuiki vierasta, hän oli nainen ja
hänellä oli — kaiken opiskelun ja lueskelun jälkeenkin — liian vähän
varsinaista älyllistä kouluutusta pystyäkseen ominpäin
muodostamaan teoriaa.

Ida Aalbergilla on kuitenkin ollut oma osuutensa miehensä

teoreettiseen ajatteluun. Hän on määrännyt sen suuntaa. Ida
Aalberg oli ollut aiheena siihen, että vapaaherra Uexküll-
Gyllenbandin filosofia kohdistui 1890-luvun lopulla
näyttämötaiteeseen. Hänellä on täytynyt olla määräävä osuus siinä,
että vapaaherra Uexküll-Gyllenbandin teoreettisessa kannassa
Kaarlo Bergbomin kuoleman jälkeen tapahtui eräs oleellinen muutos.

Vuonna 1908 julkaistiin »Ajassa» eräs kirjoitus Ida Aalbergin

nimissä. Artikkelin otsikkona oli: »Kansa ja teatteri». Se on pitkä
kirjoitus ja luonteeltaan sangen teoreettinen, vaikka lupaakin alussa
pysyä käytännön kannalla. Se saarnaa kosmopolitismia vastaan
taiteen alalla, se väittää, että hyvän taiteen tulee olla kansallista. Se
puhuu Kaarlo Bergbomin jättämästä perinnöstä, kansallisesta
teatterista, jota olisi kohotettava ja vaalittava.

Vapaaherra Uexküll-Gyllenband oli perusluonteeltaan

kosmopoliitti. Taide, ei kansallinen ja suomalainen taide, vaan
yleisinhimillinen taide, oli ollut hänen johtotähtenään. Asettuiko siis
Ida Aalberg miestään vastaan tuossa kirjoituksessa? Se ei ole
luultavaa, päinvastoin näyttää siltä, että koko kirjoitus on vapaaherra
Uexküll-Gyllenbandin laatima. Tyyli »Kansa ja teatteri» artikkelissa ei
ole Ida Aalbergin yksinkertaista ja havainnollista, vaan vapaaherra
Uexküll-Gyllenbandin filosofian ja juridiikan kyllästyttämää tyyliä.[30]
Mutta itse perusajatuksen muodostamisessa Ida Aalberg lienee ollut
»primus motor», aloitteen tekijä ja määräävä voima.

Ida Aalberg kirjoitti elämänsä lopulla eräitä erittäin kauniita

tunnelmakuvauksia, semmoisia kuin aikaisemmin mainittu
»Kotiseudun lumous» tai muistelma Oulunkoskien laskusta erään
kiertueen aikana. Kuitenkaan hän ei tunnu kernaasti tarttuneen
kynään, jos oli kysymys julkisuudesta. Esim. vuoden 1908 alussa
Kansallisteatterin »Elinan surman» esityksestä kirjoitettu ja hänen
nimissään julkaistu arvostelu saattaa kyllä sisältää hänen omat
mielipiteensä, mutta on alusta loppuun vapaaherra Uexküll-
Gyllenbandin laatima: on tallella vapaaherran saksalainen konsepti,
jota ei mitenkään voi selittää suomalaisen alkutekstin käännökseksi.

Jalmari Hahlin ollessa Suomen Kansallisteatterin johtajana Ida

Aalberg kevätkaudella 1907 vieraili kahdessa osassa:
tuntemattoman kotimaisen tekijän »Luvatussa maassa» ja Victor
Hugon »Lucrezia Borgiassa». »Valvojassa» Eino Leino merkitsi
edellisen kappaleen perin heikoksi, jälkimmäisen vanhentuneeksi ja
kirjoitti Ida Aalbergin osalle:

»Mitään erikoisen uutta eivät nämä Ida Aalbergin vierailut tällä

kertaa esityksen puolesta tarjonneet. Mutta luonnollisesti on hänen
oma taiteensa aina omiaan aateloimaan Kansallisteatterin
näyttämön siksi tunteiden ja väkevien mielikuvien kentäksi, joka sen
oikeastaan tulisi olla ja joksi ainoastaan Ida Aalberg sen vielä voi
tehdä. Mielestäni on näin ollen arvostelijan tehtävä vain merkitä
omansa ja epäilemättä myös koko teatteriyleisömme syvin
kiitollisuus.» —
 Vuoden 1907:n vierailut Kansallisteatterissa päättyivät kesken,
sillä Ida Aalberg suunnitteli siihen aikaan kiihkeästi vierailuretkeä
Unkariin. Saman vuoden syksyllä hän omine seurueineen esiintyi
sekä Helsingissä että useissa maaseutukaupungeissa näytellen
»Rosmersholmia», »Hedda Gableria», »Elinan surmaa» ja »Kotia».
Kotimaisen arvostelun voi ehkä lyhyimpänä lukea »Vasabladetista»,
joka kirjoitti:

»Ida Aalbergin näytteleminen ei kaipaa monia sanoja. Ei tarvitse

lisätä mitään siihen, minkä jokainen jo tietää, sillä hän on yhä sama
täydellinen näyttelijätär kuin ennenkin. Hänen näyttelemisensä on
viimeistä yksityiskohtaakin myöten viimeisteltyä, siinä on huippuunsa
kehitettyä tekniikkaa, eloa ja suloa kaikissa liikkeissä ja kaiken yllä
sitä sielukkuutta, jonka Ida Aalberg yksin voi luomilleen antaa.
Voimme siis, kuten aina ennenkin, vain tunnustaa, että meidät on
voitettu ja vallattu ja että olemme täynnä ihailua ja ihmettelyä.»

Helsinkiläisessä arvostelussa sentään oli havaittavissa jokunen

soraäänikin. Varsinkin Volter Kilpi suhtautui Ida Aalbergin ja hänen
seurueensa taiteeseen epäillen ja kielteisesti. Pitkän, kotimaassa
tehdyn kiertueen tarkoituksena oli valmistautuminen esiintymistä
varten Unkarissa, missä annettiin muutamia vierailunäytäntöjä
vuoden lopussa. Magyarien kritiikki Ida Aalbergin taiteesta oli
melkein yhtä innostunut ja ritarillinen kuin vuonna 1880. Puhuttiin
»Pohjantähdestä», otettiin vertauskohdiksi Duse ja Bernhardt ja
lisäksi etevä unkarilainen näyttelijätär Emilia Markus. Suomenkielen
sanottiin olevan yhtä kaunista kuin unkarin kieli, jopa joskus
muistuttavan kreikankieltä. »Neues Pester Journalissa» kirjoitettiin:

»Rouva Aalberg on suuri taiteilijatar, lahjakkaisuudessa aivan

kuuluisinten ammattisisartensa vertainen. Ihana, notkea vartalo,
suuret, puhuvat silmät, mieltäkiinnittävät kasvot, joissa kaikki
mielenliikunnot elävästi ja tarkasti kuvastuvat. Hänen koko olentonsa
muistuttaa Dusea, jonka kanssa hänellä on yhteistä tuo melodinen
persoonallisuus, ja Ida Aalberg kuuluu ehdottomasti niihin
näyttelijättäriin, joissa taiteen suuruus ja syvyys yhtyy hempeään
naisellisuuteen ja runolliseen suloon. Hänen puhelunsa, hänen
tapansa esittää inhimillisten tunteiden asteikkoa, on täynnä
intuitsionia ja hienoa taidetta; on yhdentekevää, näytteleekö hän
intohimoisesti liikkeillä vai ilmaiseeko hän mielensä kasvojensa ja
silmiensä ilmeillä.»

Parhaiten menestyi Unkarissa »Elinan surma». Yleisömenestys ei

muodostunut erikoisemmaksi, mutta unkarilaiset ylioppilaat juhlivat
Ida Aalbergia laulamalla hänelle teatterissa, riisumalla hevoset
hänen vaunujensa edestä ja vetämällä hänet loistohotellin eteen,
missä taiteilija Pestissä ollessaan majaili. Vähäinen reservatsioni
siitä, miten Ibseniä oikeastaan olisi näyteltävä, ilmestyi »Pester
Lloyd» lehdessä, mutta ei pahasti himmennä näiden Unkarin
näytäntöjen yleistä tunnelmaa. — Budapestista palattuaan Ida
Aalberg näytteli seurueineen vielä Pietarissa.

Vuonna 1909 Ida Aalberg saavutti huomattavan taiteellisen voiton

esiintymällä Helsingissä »Iphigeneia Tauriissa» draaman
nimiosassa. Hänen lausuntansa ilmehikkyys ja hänen käyttämiensä
asenteiden viivakauneus esiintyi edukseen tässä Goethen
kaihoisassa ja viileässä runoelmassa. Goethen draamassa ei ole
draamallista voimaa, ja Ida Aalberg esiintyi hänkin varsin vieraalla
alalla, mutta kuitenkin oli draaman tulkinta silloisissa oloissa jotakin
varsin poikkeuksellista. Ida Aalbergin oma innostus tarttui muihinkin
näyttelijöihin ja riitti aateloimaan esityksen, joka kuitenkaan ei
saavuttanut suuren yleisön suosiota.
 *****

Jalmari Finne on kertonut sanoneensa kerran Kaarlo Bergbomille:

»Teidän varjonne lyö kumoon viisi seuraavaa Kansallisteatterin
johtajaa.» Näitä sanoja ei oltu sanottu ainoastaan vakaumuksesta ja
todellisesta syvästä ihailusta Kaarlo Bergbomia ja hänen
elämäntyötään kohtaan, vaan myöskin lohdutukseksi Suomalaisen
teatterin perustajalle, joka elämänsä lopulla huokaili suurten
vastuksien ja surujen alla.

Yhtä vähän kuin Aleksanteri Suuri tai Bismarck kasvatti Kaarlo

Bergbom itselleen valtansa perijää ja työnsä jatkajaa. Koko kysymys
tuntuu olleen hänelle vastenmielinen. Aatteen miehenä hänen olisi
pitänyt valita itselleen seuraaja, mutta on hyvin anteeksiannettavaa,
ettei hän sitä tehnyt. Kaarlo Bergbom tiesi persoonallisesta
kokemuksestaan, että teatterinjohtajan vähäinen kunnia maksetaan
nöyryytyksillä ja kieltäymyksillä, joilla on vaikeata yrittää
onnellistuttaa toista ihmistä.

Kaarlo Bergbomin suurimmissa voitoissa oli ollut määräävänä

tekijänä Ida Aalberg. Ei ollut mitään suuruudenhulluutta, jos Ida
Aalberg katsoi itsensä Bergbomin lähimmäksi perijäksi ja Bergbomin
traditsionien suojelijaksi.

Vuonna 1909 Ida Aalberg valittiin Kansallisteatterin taiteelliseen

johtoon.

Jo 1904—1905:n kiertueen aikana Ida Aalberg lienee seisonut

tasavertaisena vapaaherra Uexküll-Gyllenbandin rinnalla —
vapaaherran itsensäkin mielestä. Ainakin jo sinä vuonna Ida Aalberg
ohjailee, kuten kirjeet osoittavat, saksalaisten näyttelijäin
roolianalyysiä. Tämän jälkeen hän yhä enemmän alkaa puuttua
ohjaukseen, ja esim. Kansallisteatterin ottaessa esitettäväksi
»Iphigeneia Tauriissa» Ida Aalbergin neuvot tuntuvat olleen sekä
määrääviä että hedelmällisiä. Vaikka vapaaherra Uexküll-Gyllenband
olikin Ida Aalbergin neuvonantajana ja mitä läheisimpänä
työtoverina, sisältyy Ida Aalbergin ohjaajatoimintaan vuosina 1909—
1911 paljon itsenäisyyttä ja omintakeista toimintaa.

Ida Aalberg saapui 1909 Helsinkiin täynnä innostusta ja uskoa ja

harrastusta, joka ei koskenut vain suomalaista näyttämötaidetta,
vaan suomalaista kulttuuria yleensä. Hän tahtoi koota ympärilleen
kansallisen sivistyksen suurimmat edustajat, hän tahtoi perustaa
oman »salongin», hän tahtoi suomalaista näyttämötaidetta
mitattavan vain kaikkein korkeimmilla mitoilla. Nämä optimistiset
toiveet pettyivät täydellisesti: emännän loistavasta seurustelutaidosta
huolimatta »salonki» ei ottanut oikein menestyäkseen, ja taiteellinen
toiminta teatterissa johti mitä surkeimpaan ristiriitaan, Ida Aalbergiin
nähden suorastaan katastroofiin.

*****

Ida Aalberg käsitti itsensä Bergbomin traditsionien suojelijaksi.

Hänen Kansallisteatterin kanssa tekemäänsä sopimukseen kuului,
että hän vuosittain ohjaisi neljä kappaletta, joissa myöskin itse
esiintyisi näyttelijänä.

Kaarlo Bergbomin johdon suurin saavutus lienee siinä, että hän loi
kotimaisen ohjelmiston, vaikutti itseensä draamatuotantoon. Siinä
suhteessa Ida Aalberg koetti olla ehdottomasti uskollinen
traditsioneille. Ottamalla näyttämölle Linnankosken »Ikuisen
taistelun», Maila Talvion »Anna Sarkoilan» ja Eino Leinon
»Alkibiades» näytelmän hän ei noudattanut vain omia mielitekojaan
ja primadonna-vaatimuksiaan, vaan myöskin, ja lähinnä, Kaarlo
Bergbomin traditsioneja. Onni ei kruunannut hänen rohkeita
pyrkimyksiään. Kritiikki ei tosin sivuuttanut hänen innostuneen
työnsä tuloksia kylmästi, vaan tunnusti, että ohjaustoiminnan takana
oli voimakas tahto ja rikas mielikuvitus, mutta sittenkin jäi menestys
puolinaiseksi. Ibsenin »John Gabriel Borkmanin» esityksestä
suomalainen kritiikki sanoi tuomionsanoja ja paljon jyrkemmässä
muodossa kuin aikaisempi ulkomainen kritiikki »Hedda Gablerista»
ja »Rosmersholmista». Vapaaherra Uexküll-Gyllenband avusti
parhaansa mukaan Ida Aalbergia tässä ohjaustoiminnassa, mutta
hänen »uusi tekniikkansa» tai »psykologinen» suhtautumisensa
tehtäviin ei tullut ainakaan yleiseksi tietoisuudeksi Helsingin
arvostelijain kesken, kun he arvostelivat Ida Aalbergin ohjausta ja
näyttelemistä. »John Gabriel Borkmanin» esitys oli Ibseniin
kohdistuvaa väkivaltaa, sen voi lukea suopeidenkin arvostelujen
rivien välistä, Anna Sarkoilaa Ida Aalberg ainakin toisten mielestä
tulkitsi liiaksi suureen tyyliin.

Suurisuuntaisin ja rohkein teko, minkä Ida Aalberg ohjaajana

suoritti, lienee ollut Linnankosken »Ikuisen taistelun» esittäminen.
Elämänsä lopulla taiteilija, joka niin usein oli näyttämöltä tulkinnut
rikollisia tyyppejä, otti esittääkseen itsensä pahuuden alkuvoiman,
Luciferin. Hän ei kuitenkaan valinnut kappaletta tuon osan vuoksi,
hän selitti nimenomaan, että hänen päätarkoituksensa oli kirjailijan
kannustaminen ja kotimaisen draaman vaaliminen. Ida Aalberg ja
vapaaherra Uexküll-Gyllenband eivät säästäneet vaivojaan,
ohjauskirjoista näkee, että työtä on tehty hellittämättömällä innolla.
Mutta niin epäitsekäs kuin Ida Aalbergin yritys olikin, »Ikuinen
taistelu» muodostui ehkä pahimmaksi kompastuskiveksi hänen
ohjaajatoiminnassaan. Suuret kustannukset, jotka kappale aiheutti
teatterille, herättivät pahaa verta. Taiteellinen tulos muodostui
mielenkiintoiseksi, mutta ei vakuuttavaksi. Draamaahan ei oltu
tehtykään näyttämön vaatimuksia silmälläpitäen.

V.A. Koskenniemi kirjoitti »Ikuisen taistelun» johdosta m.m.:

‒ ‒ ‒ »Ensimmäinen näytös, Luciferin öinen kokous, on

draamallisesti verraten heikko. Pimeyden ruhtinas on koonnut
ympärilleen oman seurakuntansa, Repokorvat, Vihreäsilmät, Täti
Lierot, Herjakielet, Satasormet j.n.e., ja suunnittelee heidän
kanssaan ihmissuvun viettelemistä Jehovan pelvosta. Taikka
oikeammin: koko näytös on Luciferin varsin epädraamallista
yksinpuhelua, jonka keskeyttävät siellä täällä henkien ulvahdukset
tai muut mielenosoitukset verraten merkityksettömine
vuorosanoineen. Näytelmän heikoimpia puolia on mielestäni se, että
Luciferin kuva kirjailijan ponnistuksista huolimatta jää varsin
vaaleaksi. Byronin Lucifer, Goethen Mefistofeles ja Andrejevinkin
Anatema — kirjallisuuden uusin viettelijätyyppi — ovat kuvatut
samalla kertaa sekä verevämmällä että hienommalla taiteella. Mikä
on Linnankosken Luciferin toiminnan ja tekojen sisin vaikutin? Jos
hän todella tietää alottaneensa »ikuisen taistelun» Jehovaa ja hänen
vaikutustaan vastaan, tuntuu hän tässä ensimmäisessä öisessä
kohtauksessa tuhlailevan yhdellä kertaa liiaksi sydämensä raivoa.
Samalla tekee hän liian kaunopuheliaan vaikutuksen. Hänestä
puuttuu jotain siitä tyynestä sisällisestä auktoriteetista, jonka
mielellään runoilisimme tähän Jehovan väsymättömään, ylpeään
vastustajaan. En tiedä onko runoilija ajatellut tätä tyyppiä luodessaan
paratiisin käärmettä, mutta ainakin tuli se usein Ida Aalbergin
esityksessä mieleen. Joka tapauksessa on »Ikuisen taistelun»
Lucifer mittasuhteiltaan vähäpätöisempi ja inhimillisesti
merkityksettömämpi kuin se ihmislapsi, Kain, johon hän pyrkii
vaikuttamaan. — — —
 *****

Esitys? Siinä pistää kenties ensimmäisenä silmään se ulkonainen

loisto, jolla tämä draama oli asetettu näyttämölle. Dekoratsionien
tyylikäs hienous ja komeus voittaa kaiken, mitä meillä
Kansallisteatterissa on nähty. Jo ensimmäinen kuvaelma idyllisine
maisemineen antoi rikkaan tunnelman. Hauskasti oli keksitty
uneksiva valmurivi rampin luona samoinkuin suuret, koristeelliset
sienet. Toisessa kuvaelmassa oli Luciferin luola niinikään pantu
näyttämölle komealla taiteellisella mielikuvituksella. Aivan yhtä
onnellista vaikutusta eivät tehneet ne henget, jotka tässä luolassa
pitivät kokoustaan. Heidän joukkohuudahduksissaan,
naurunrähäköissään ja koko yleisesityksessä tuntui kyllä huolellinen
ja harkittu johto, mutta koko kuvaelma on itsessään mahdoton
esitettäväksi. Ei voi kernaasti uskoa, että tuo päätön lauma pystyisi
ihmiskuntaa tuhoamaan. Ida Aalbergin komeasti, hillityllä energialla
lausumat vuorosanat antoivat tälle kohtaukselle kuitenkin
eräänlaisen loiston, joka osittain korvasi sen, mitä itse kuvaelmasta
sinänsä puuttuu. Kaiuin näyt toisessa näytöksessä olivat myöskin
huomattavia näyttämöteknillisiä saavutuksia, ehkä kuitenkin
lukuunottamatta muutamia taikalyhtykuvia, jotka olisivat voineet olla
parempia. Uhrikohtauksen stilisoitu maisema sekä viimeisen
kuvaelman syysmetsä olivat harvinaisen tunnelmallisia ja auttoivat
kaiketi osaltaan näytelmää sen taiteelliseen voittoon.

Osien esityksessä kiinnittää mieltä ensi sijassa Ida Aalbergin

Lucifer. On erinomaisen rohkeata naiselle käydä tulkitsemaan
tällaista osaa, jonka luonteeseen kaiken traditsionin mukaan kuuluu
juuri tärkeänä puolena maskuliininen uhma. Kukaan muu kuin Ida
Aalberg olisi tuskin tähän voimankoetteluun antautunut. Hänen
Luciferinsa oli enemmän käärmeellinen kuin uhmaileva, enemmän
raivokas kuin maltillisesti ja tyynesti harkitseva. Se oli eräänlainen
sukupuoleton Mefistofeles, kenties hiukan hermostunut ja
agitatoorinen, mutta monin paikoin, esim. kuiskatessaan Kainille
tämän maatessa maassa tai uhritoimituksen aikana, häikäisevän
demoninen.» — —

Se, että Ida Aalberg sekä ohjaajana että näyttelijänä saavutti

yhden suuren ja ehdottoman menestyksen, nimittäin Ostrovskin
»Ukkosilmassa», ei riittänyt vakiinnuttamaan hänen asemaansa
silloisen johdon silmissä. Ida Aalberg ei, kuten tavallisesti aina
ennen, ollut pystynyt täyttämään taiteensa voimalla katsomoa ja
kassaa, ja senvuoksi Kansallis-teatterin johtokunta kevätkaudella
1911 sanoi hänet irti vakinaisesta toimesta ja toivoi palattavan
entiselle kannalle, t.s. vierailunäytäntöihin, joilla Ida Aalberg
kahdenkymmenen vuoden aikana oli avustanut teatteria.

Sanottakoon Ida Aalbergin ohjausajasta 1909—1911 mitä muuta

tahansa, tämä ainakin on varmaa: hän oli työssään osoittanut niin
valtavaa innostusta ja antaumusta, että tuomio kohtasi häntä
ruhjovalla voimalla. Saamansa ero, jos mikään, muuttui hänelle
tunnekysymykseksi. Hän tunsi ansionsa teatteria kohtaan liian
elävästi kestääkseen näin saamaansa tuomiota, ja oli myöskin liian
vilpitön ja kuumaverinen kyetäkseen suhtautumaan siihen
ylimielisesti ja kylmästi.

Maila Talvio on kertonut, kuinka Ida Aalberg koetti, saatuaan

tiedon erostaan, tulla hänen luokseen. Ida Aalberg pääsi taloon ja
ovesta sisään, mutta lyyhistyi kokoon alimmille sisäportaille, tuijotti
siinä murtunein katsein eteensä ja toisteli raukealla yksitoikkoisella
äänellä: »Niin sitä vaan potkitaan — — pois!»
 Hän oli kuolemaan asti väsynyt, ja koko kohtaus oli kaamea ja
vilpitön kuin kuolema itse. Seuraavina päivinä ennen niin elegantti
Ida Aalberg pakkasi tavaroitaan puku ja tukka epäjärjestyksessä.

1911 Ida Aalberg pani toimeen viimeisen suuren kiertuensa

Suomessa. Lukuunottamatta Hämeenlinnassa sattuneita poliittisia
rettelöitä hän sai osakseen jakamatonta ja lämmintä innostusta.
Kuitenkin täytyy sanoa, että Ida Aalbergin taide alkoi loppuaikoina
olla epäilyksenalaista. Kun hän esitelmöi maaseutukaupungeissa
taiteellisen teatterin tarpeellisuudesta ja kulttuuriarvosta, sai hän
kritiikkejä, jotka olivat vastustavalla kannalla syyttäen Ida Aalbergin
puhuvan toista kuin hän taiteilijana esitti: maaseudullakin uusi aika
alkoi esittää »hillityn suunnan» vaatimuksia täysin tarmoin ja
kiivaasti. Kiivailu »hillityn taiteen» puolesta on naurettavaa, ja
varsinkin muutamassa maaseutulehdessä se esiintyy perin
koomillisessa valossa Ida Aalbergin esitelmämatkan aikana.

Vuosina 1912—1913 Ida Aalberg eli ylpeässä eristyneisyydessä

Viipurissa. Hän ei tahtonut puuttua mihinkään taidetta koskevaan,
hän kieltäytyi näyttelemisestä, lausumisesta, kirjoittamisesta.
Varmaa on, ettei hän tätä kaikkea tehnyt yksin omasta halustaan ja
voimastaan. Vapaaherra Uexküll-Gyllenband sai hänet muistamaan
vuoden 1911:n riitautumista kauemmin kuin hän itse olisi kyennyt
sitä muistamaan. Se oli epätoivon aikaa Ida Aalbergille. Hän
suunnitteli elämäkerran kirjoittamista, hän kävi keittokoulua, hän
lueskeli, hän koetti opiskella venäjän kielen alkeita — mistään ei
tullut vähintäkään tyydytystä.

Tarvittiin vain pieni kirje Juhani Aholta, kun Ida Aalberg unohti
ylpeytensä ja saapui vuoden 1914:n alussa Kansallisteatterin hänen
kunniakseen toimeenpanemiin suuriin juhliin.