Elections and Voters in Britain 1St Ed 2022 Edition Denver Full Chapter

Elections and Voters in Britain 1st ed.
2022 Edition Denver

Visit to download the full and correct content document:
https://ebookmass.com/product/elections-and-voters-in-britain-1st-ed-2022-edition-de
nver/
Elections and Voters
in Britain
David Denver · Robert Johns

Elections and Voters in Britain
David Denver · Robert Johns
Elections and Voters

in Britain
David Denver Robert Johns
Department of Politics, Philosophy University of Essex
and Religion Colchester, Essex, UK
Lancaster University
Lancaster, UK
ISBN 978-3-030-86491-0 ISBN 978-3-030-86492-7 (eBook)

https://doi.org/10.1007/978-3-030-86492-7
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer
Nature Switzerland AG 2022
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 credit: steven gillis hd9 imaging/Alamy Stock Photo
This Palgrave Macmillan imprint is published by the registered company Springer Nature
Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
This fifth edition of Elections and Voters in Britain has been a long time
forthcoming. The normal rhythm was to update the book every four
or five years, after each general election. This time, however, no sooner
had we updated the text to reflect one election than another was fast
approaching. The sense of trying to hit a moving target was reinforced
by the fact that more recent elections were unusual in their outcomes
as well as their frequency. Between them, the 2017 and 2019 elections
challenged a variety of received wisdoms about, for example, the role of
the economy, the limited impact of national campaigns, the advantages of
ideological moderation and the rarity of single-issue voting. Since those
received wisdoms had been carefully laid out in previous editions, the
events of the past few years called for a good deal of updating. Mean-
while, the powerful impact of the votes on Scottish independence and
EU membership has meant that referendums have been promoted from
footnote status to a chapter of their own.
Rehearsing all this explains the delayed arrival of this edition but also,
we hope, why it was worth the wait. In addition to a new chapter on
referendums, the discussions of electoral geography and electoral systems
have now been separated and expanded such that there is a second new
chapter. There is also much new material in each of the other chapters,
with results and survey data from the 2019 election fully incorporated
into the discussion as well as relevant tables and figures. There are many
v
vi PREFACE
analyses also covering the 2017 election as well as contrasts with the rather
more orthodox contest in 2015.
We used the term ‘updating’ rather than ‘rewriting’ because it is not
clear how far the most recent elections mark a temporary interruption of
normal service rather than a new electoral order. For example, part of the
very powerful Brexit effect in 2019 is likely to die away as the issue slowly
recedes, with other issues such as the economy liable to regain impor-
tance. On the other hand, part of what looked like a Brexit effect was
more a reflection of longer-term trends in the sociology and geography
of British public opinion, especially relating to the issue of immigration,
and this looks to be potentially a more enduring change. In updating the
chapters, then, we do not abandon received wisdom but acknowledge the
severity and estimate the longevity of the challenges posed to it by recent
elections.
As noted in the Preface to the first edition, this book is not primarily
for colleagues actively researching and writing about British elections.
Although there is some fresh data analysis, most specialists in the field
will already know most of what is written in the following pages. Our
hope is that they find it a useful overview which can be recommended
to their students. The broader intention is to summarise and simplify the
work of these colleagues for non-specialists and students in both schools
and universities (as well as people who just happen to be interested in
elections). One consequence of this focus is that we have tried to keep
tables as clear and uncluttered as possible. In some cases this has meant,
for instance, not showing in tables the raw numbers on which calcula-
tions are based. This would be unforgiveable in a research article, of
course, but seems, on balance, to be a sensible strategy in this case. As
before, we have also included a glossary of statistical terms, and items
appearing in the glossary are printed in bold when they first appear in
the text. The explanations and descriptions of various statistical measures
and techniques offered are intended to be non-technical. They attempt to
enable readers to understand why various techniques are used and what
the results can tell us.
Everyone writing about electoral behaviour in Britain has to acknowl-
edge a debt to the successive teams which, from 1963 onwards, have
undertaken the burden of directing the British Election Study (BES)
surveys. Their task is not getting any easier, especially in respect of the
post-election face-to-face surveys that have been the primary bases for
comparisons over time in previous editions of the book. Arranging and
PREFACE vii
fielding these surveys at short notice is difficult and expensive enough

at the best of times; doing so amid a pandemic is well-nigh impossible.
For this reason, where our analysis of the 2019 election required survey
evidence, we have drawn largely on the Internet surveys also conducted by
the BES. We are, therefore, grateful for the speed with which the current
BES team, like its predecessors at Essex, made those online survey data
available to colleagues. Also much appreciated are the swift responses by
BES researchers (notably Jon Mellon, Chris Prosser and Jack Bailey) to
our many queries. Since we have not always shown the same prompt-
ness, we are grateful for the patience showed by our publishers, especially
Lloyd Langman. Finally, we should thank Chris Carman who co-authored
the previous edition of the book and also worked on the stillborn 2015
election version, and whose suggestions for updates and additions were
valuable.
The first edition of the book began with the words ‘Elections are fun’.
Almost thirty years later they have become even more fun—certainly they
are more unpredictable, in timing as well as outcome. We hope that this
book will start to develop in at least some readers the same interest and
enthusiasm that we retain for the subject.
Lancaster, UK David Denver

Colchester, UK Robert Johns
Contents
1 Studying British Elections 1

Why Study Elections? 4
Studying Elections 7
Election Results and Survey Data Compared 15
Theories of Voting 20
References 27
2 Turnout: Why People Vote (or Don’t) 29
Turnout Variations Over Time 34
Turnout Variations Across Constituencies 36
Survey Studies of Non-voting 41
The Consequences of Low Turnout 52
Conclusion: What Is to Be Done About Low Turnout? 54
References 57
3 The Changing British Party System 61
Class and the Two-Party System 63
Partisan Alignment 68
Dealignment 71
The Causes of Dealignment 77
Dealignment and Electoral Volatility 86
From Dealignment to Realignment? 91
References 94
ix
x CONTENTS
4 Issues, Policies and Performance 97

Issue Voting 98
Position Issues, Policies and Ideology 103
Valence Issues and Performance 115
The Economy and Voting 118
Conclusion 126
References 127
5 Party Images and Party Leaders 131
Party Images 132
The Content of Party Images 133
Party Images and Party Choice 139
Party Leaders 140
Why Leadership Matters (More) 142
Leader Evaluations and Party Choice 145
The Sources of Party and Leader Images 151
Conclusion 156
References 158
6 Campaigning and the Media 161
National Campaigning 162
Opinion Polls 169
The Media 175
Television 180
The National Press 183
Constituency Campaigning 188
The Digital Campaign 190
Conclusion 194
References 195
7 The Geography of British Elections 201
Regional Variations 201
Urban–Rural Variations 207
Explaining Regional Variations 209
Constituency Variations 212
Tactical Voting 218
Localities and Neighbourhoods 221
References 225
CONTENTS xi
8 Elections and Electoral Systems 229

The Operation of First-Past-the-Post 231
In Defence of First-Past-the-Post 239
Elections and Voting Under Different Electoral Systems 242
Conclusion 249
References 250
9 Referendums 253
Definition and Role 254
The British Experience 256
Referendums and Voters in Britain 258
The Role of the Campaign 273
In Defence of (Some) Direct Democracy 277
References 279
10 Elections and Party Choice in Britain Today 283
Explaining Party Choice 284
Explaining Election Outcomes 304
Conclusion 308
References 310
Glossary of Statistical and Technical Terms 313

Author Index 319
Index 325
List of Figures
Fig. 1.1 Party shares of votes in general elections, 1959–2019

(Great Britain) 9
Fig. 1.2 Butler–Stokes model of party choice (individual voters) 25
Fig. 2.1 Turnout in UK general elections, 1959–2019 35
Fig. 2.2 Turnout (per cent) by strength of party identification,
1992–2019 52
Fig. 3.1 Alford index class voting scores for Labour
and Conservatives, 1964–2019 74
Fig. 3.2 Extent and strength of party identification, 1964–2019 75
Fig. 4.1 Conservative and Labour left–right positions, 1950–2019 107
Fig. 4.2 Five parties located in two-dimensional policy space, 2019 113
Fig. 5.1 A model of party and leader image formation 151
Fig. 6.1 Total campaign expenditure by parties, 2001–2017 165
Fig. 6.2 Voting intentions in campaign polls, 2017 167
Fig. 6.3 Voting intentions in campaign polls, 2019 168
Fig. 6.4 Average error in final polls’ prediction
of Conservative-Labour lead, 1966–2019 172
Fig. 6.5 Direct effects and filter models of media influence 178
Fig. 7.1 Deviations from national (GB) Conservative lead
over Labour, 1959–2019 203
Fig. 7.2 Regional deviations in England from national (GB)
Conservative lead over Labour, 1959–2019 205
Fig. 7.3 Urban–rural deviations from national Conservative lead
over Labour, 1959–2019 207
xiii
xiv LIST OF FIGURES
Fig. 9.1 Percentage of ‘don’t knows’ in vote intention polls

at three referendums 264
Fig. 10.1 The added explanatory power of variables representing
different theories of voting, 2015–2019 303
List of Tables
Table 1.1 Results of the 1959 and 2019 general elections in Great
Britain 2
Table 1.2 Party shares of votes in general elections, 1959–2019
(Great Britain) 8
Table 1.3 The ‘flow of the vote’ between 2017 and 2019 19
Table 2.1 Turnout in UK elections, 2015–2019 32
Table 2.2 Turnout in UK general elections, 1959–2019 34
Table 2.3 Correlates of constituency turnout (Great Britain) 1987
and 2019 37
Table 2.4 Turnout of social groups in 2019 general election (%)
and change from 2017 45
Table 2.5 Binary logistic regression analysis of voting
versus non-voting in 2019 49
Table 2.6 Comparing hypothetical choice among non-voters
with election results, 1970, 2017 and 2019 (%) 53
Table 3.1 Vote in 1964 by objective and subjective class (row
percentages) 66
Table 3.2 Party identification in 1964 by parents’ party preference
(per cent) 69
Table 3.3 Vote in 2017 and 2019 by objective and subjective class
(row percentages) 72
Table 3.4 Average net satisfaction ratings for governments,
1951–2019 81
Table 3.5 Trends in inter-election volatility (per cent) 88
xv
xvi LIST OF TABLES
Table 3.6 Percentage of a party’s 2017 voters reporting a different

or uncertain vote intention in mid-term panel waves 89
Table 3.7 Age, education and Conservative versus Labour choice
in 1987 and 2019 92
Table 4.1 Issue salience rankings in 2015, 2017 and 2019 (%
citing issue as ‘most important’) 101
Table 4.2 Party leads on handling valence issues, 2010–2019 116
Table 4.3 Retrospective sociotropic evaluations in 2015 and 2019 123
Table 4.4 Percentage voting Conservative by retrospective
economic evaluations and EU referendum vote, 2015
and 2019 124
Table 5.1 Class images of labour and the conservatives, 1960s
and 2010s (%) 134
Table 5.2 Percentages rating the major parties as looking
after the interests of three groups, 2015–2019 135
Table 5.3 Five valence aspects of party image, 2015–2019 (%) 137
Table 5.4 Leaders’ average ratings (0–10) on likability, competence
and integrity, months before and immediately
after polling 146
Table 5.5 Best person for Prime Minister, 1983–2019 (%) 148
Table 5.6 Party choice in 2019 by preferred Prime Minister,
controlling for party identification 149
Table 6.1 Average error in final polls’ prediction of party vote
shares, 1970–2019 171
Table 6.2 Endorsements, slant of coverage and party choice
among readers of national daily newspapers at the 2019
election 184
Table 7.1 Conservative share of the votes in regions, 1959
and 2019 202
Table 7.2 Conservative and Labour voting in English regions
in 2019 (row percentages) 206
Table 7.3 Correlations between socio-economic characteristics
of constituencies and party shares of vote, 1966
and 2019 213
Table 7.4 Multiple regression equations of party shares of vote
in 2019 216
Table 8.1 UK electoral systems 231
Table 8.2 Index of disproportionality scores (UK general elections
since 1966) 233
Table 9.1 Turnout in selected referendums in the UK since 1975 259
Table 9.2 Correlations between socio-demographic profile
and referendum turnout across local authorities 262
LIST OF TABLES xvii
Table 9.3 Campaign trends in certainty (per cent ‘very’ or ‘fairly’

sure) about the consequences of referendum outcomes,
2014 and 2016 275
Table 10.1 Social characteristics and party choice in 2019 (row
percentages) 285
Table 10.2 Effects of socio-demographic variables in logistic
regression analyses of Conservative vs. Labour voting
in 2019 288
Table 10.3 Effects of issue/ideological variables in logistic
regression analyses of Conservative vs. Labour voting
in 2019 296
Table 10.4 Effects of valence variables in logistic regression analyses
of Conservative vs. Labour voting in 2019 300
CHAPTER 1
Studying British Elections
Since the previous edition of this book, there have been three general
elections and a major nationwide referendum in the UK. The latter, on
the country’s membership of the European Union (EU), produced an
outcome which almost paralysed ‘normal’ politics for three years and led
directly to two further elections (2017 and 2019) following that of 2015.
Not since the 1920s had three general elections been held within less than
five years. It is clear, then, that voters—and the decisions they took—
played a central role in the dramatic political developments that occurred
in this period. Moreover, the results of at least two of those elections
and the referendum surprised most commentators and pundits, as well as
politicians, to varying degrees and in different ways.
A recurring theme of the book is that elections and voters in Britain
have become much more unpredictable and volatile than they used to be.
To illustrate this at the outset, it is useful to contrast the December 2019
general election with one held sixty years before, in October 1959. The
latter, no doubt, will seem like ancient history to many but it provides
an apt point of comparison. For one thing, it was held during an era of
relative calm. There was a broad consensus over policy among the parties
and little switching of votes between elections. In addition, it exemplified
the pattern of two-party dominance that had characterised British politics
from the end of the Second World War and had come to be accepted as
the norm which was unlikely to be threatened in the foreseeable future.
© The Author(s), under exclusive license to Springer Nature 1

Switzerland AG 2022
D. Denver and R. Johns, Elections and Voters in Britain,
https://doi.org/10.1007/978-3-030-86492-7_1
2 D. DENVER AND R. JOHNS
Table 1.1 Results of the 1959 and 2019 general elections in Great Britain
1959 2019
% of votes Candidates Seats % of votes Candidates Seats
Conservative 48.8 613 352 44.7 631 365

Labour 44.6 618 258 32.9 631 202
Liberal (Democrats) 6.0 215 6 11.8 611 11
SNP 0.1 5 0 4.0 59 48
Plaid Cymru 0.3 20 0 0.5 36 4
Communist 0.1 18 0 0.0 0 0
Greens (incl. Scottish – – – 2.8 494 1
Greens)
Brexit – – – 2.1 275 0
Others/Independents 0.2 18 2 1.2 488 1
Turnout 79.0% 67.7%
Note Results exclude Northern Ireland. By convention, the Speaker is not opposed by the main
opposition parties and is counted as an ‘other’ (although in 2019 both a Green and an independent
pro-Brexit candidate stood in the Speaker’s constituency)
Source Leeke (2003); Uberoi et al. (2019)
A simple comparison of the results in the two elections illustrates some

of the major changes that have occurred (see Table 1.1) and shows that
in many respects there were very marked differences between them—as,
of course, might be expected. There is one key similarity, however, which
relates to the position of the Conservative and Labour parties. In both
elections they were the largest parties in terms of shares of the votes and
seats won. Even so, they did not completely dominate the electoral land-
scape in 2019 as they had in 1959. In the earlier election, together they
obtained 94% of the votes and 99% of the seats; in 2019, their combined
share had declined to 78% of votes and 90% of seats. Unlike in 1959
when the two main parties effectively were the House of Commons, in
2019 smaller parties won 64 seats (excluding those in Northern Ireland),
giving them more relevance in parliamentary processes even if the govern-
ment had a clear majority. This significant shift, from a world in which
Labour and the Conservatives needed to worry only about fighting each
other to one in which they are merely leaders among multiple players, is
discussed in Chapter 3.
Elections from 2015 have actually seen something of a recovery for
the major parties. At its lowest point, in the 2010 general election, their
combined share was just 67%. The improvement in their popularity is
1 STUDYING BRITISH ELECTIONS 3
related to one other similarity between 1959 and 2019 which conceals
a lot of change in the meantime. The old Liberal Party emerged from
the 1959 election with an unimpressive vote share and only six Members
of Parliament (MPs). By 2019, the name had changed to the Liberal
Democrats but the electoral showing was not a great deal better: 12%
of the votes and eleven seats. In the 2010 election, however, the party
had polled over 20% of the votes, won 57 seats, and become part of the
government as junior coalition partner to the Conservatives.
Two other changes relating to party competition are clear in Table 1.1.
First, in Scotland, and to a lesser degree in Wales, the part played in elec-
tions by the nationalist parties was much more significant in 2019 than it
had been in 1959. In the 1950s the Scottish National Party (SNP) and
Plaid Cymru were generally regarded as rather peculiar parties, supported
by somewhat eccentric individuals and on the fringes of politics. Few
candidates were put forward, especially in Scotland, and those that were
won few votes. From the 1970s, however, both began to contest all seats
in their respective countries (although in 2019 Plaid stood aside in four).
Not only that, but by 2019 the SNP was the dominant party in Scotland
by some distance, winning 45% of the vote and 48 of the 59 seats. Plaid
has been much less successful but was able to take almost 10% of the
Welsh vote and hold on to its four seats.
Second, there has been a massive increase in the number and variety of
‘others’ participating as candidates in British elections. In 1959 the only
other player worth even footnoting was the Communist Party of Great
Britain (which won more than 30,000 votes). It has now disappeared
but has been more than replaced by an abundance of new parties. Some-
times, these can be highly significant participants. The United Kingdom
Independence Party (UKIP) won 13% of the vote in the 2015 general
election and, while the similarly anti-EU membership Brexit Party was
much less successful overall in 2019, it was a relevant player in many
of the constituencies in which it stood. The Greens typically now stand
in most constituencies and have held one seat, Brighton Pavilion, since
2010. In addition, in 2019 there was a flurry of candidates representing
much smaller parties. These included splinters from the major parties
(such as The Independent Group for Change), old parties that refuse
to die off (like the Social Democratic Party), new parties trying to gain
a foothold (the Women’s Equality Party, for example), local campaigners
(like the Heavy Woollen District Independents) or colourful perennials
(such as the Monster Raving Loony Party) whose primary role seems to
be to entertain (or annoy, according to taste) viewers of election night

broadcasts. Finally, there were many more people simply representing
themselves as independent candidates.
Why has there been this explosion in the number of candidates from
small parties? One reason, discussed in Chapter 3, is that many people
have lost faith in the major parties. Another more mundane reason is
that the deposit required for candidacy (retained by the authorities if the
candidate does not obtain 5% of the votes in the relevant constituency) is
no longer a serious barrier. In 1959 the deposit was £150. This was equiv-
alent to about £3,500 in today’s money but the deposit has increased to
only £500—which it has been since 1985. Moreover, small parties and
some individuals have realised that the deposit buys a lot of publicity.
Every nominated candidate is entitled to a free postal delivery to every
household in the relevant constituency; a party contesting one-sixth or
more of constituencies nationwide is guaranteed a free national radio and
television broadcast.
Finally, the table shows that the turnout of electors in 2019 was
lower—by more than eleven percentage points—than in 1959. The two
figures are not exactly comparable in that the voting age was lowered
from 21 to 18 just before the 1970 general election. Hence, the 2019
percentage turnout is based on a larger electorate, containing a much
larger proportion of young people (a group which has a poor turnout
record), than the 1959 figure. Nonetheless, there has been a clear decline
in turnout. As with the two-party vote share, the comparison here under-
states the extent of change over time. The lowest post-war turnout was
recorded in 2010 (59.1%) which provoked much discussion and concern
at the time. These worries eased, however, as turnout inched upwards
in subsequent elections, without ever approaching the figures regularly
recorded from the 1950s through to the 1990s.
Why Study Elections?

One good reason for studying elections is that they are fun events. In the
first half of the nineteenth century, part of the fun involved the voters
getting wildly drunk at the candidates’ expense (being ‘treated’, as this
was known), pelting them with rotten fruit and brawling in the street
with opponents. The practice of candidates treating voters was effectively
ended by the Second Reform Act of 1867 which greatly enlarged the elec-
torate. There were now simply too many voters to treat. In addition, in
1872 the Ballot Act made voting secret (previously voters had to declare
their choice in public), so that candidates could no longer check that
the voters they had treated actually voted for them. Finally, in 1883 the
Corrupt Practices Act outlawed treating (much to the disappointment of
many voters, one suspects). Despite this, elections continued to be a form
of public entertainment until well into the twentieth century. In 1922 a
crowd of 10,000 assembled to hear the declaration of the general elec-
tion result in Dundee and was rewarded with the news that the sitting
member, Winston Churchill, had been defeated by a Prohibitionist (in a
city notorious for drunkenness).
Popular enthusiasm for elections does not reach those standards today.
Nonetheless, general elections are still major national events which precip-
itate greatly increased political activity, discussion, interest and media
coverage. On election night, the attention of much of the country is
at least partly engaged by the election. Next day, the front pages of
all newspapers worthy of the name are entirely devoted to election
news. Television channels provide round-the-clock campaign coverage
and social media platforms are inundated with election-related material.
All of this is well over the top for some people. One ‘vox pop’, Brenda
from Bristol, became a spokesperson for many with her plaintive cry of
‘not another one!’ when told about the calling of a snap election in 2017.
For others, however, general elections remain a major spectator sport:
they devour the news coverage, discuss the election online and offline,
maybe even bet on the outcome, and then stay up all night to watch the
results and drink in celebration or sorrow.
Part of the fascination is the sheer mass of information available. In
general elections there are results for 650 (UK) constituencies to be
looked at, but in addition there are elections to the Scottish Parlia-
ment, and the Welsh, Northern Ireland and London Assemblies as well
as local council and mayoral elections involving many thousands of wards
and electoral divisions. The resulting numbers can be analysed endlessly:
aggregated, averaged, correlated, graphed and used to construct maps.
Some academic studies of electoral behaviour reach such levels of statis-
tical sophistication that they are well-nigh incomprehensible to all but a
few specialists. All of these data, together with the increasing accessibility
of statistical software, partly explains the huge growth in the academic
literature focusing on elections over the past 60 years or so.
There is more to it than simply providing data for political geekery,
however. Put simply, elections are studied because they are important.
Most people would agree that it is the existence of free, competi-

tive elections which distinguishes political systems that we normally call
‘democratic’ from others. Different versions of democratic theory attach
different weights to elections and assign them different functions, but all
see elections as central to democracy.
In traditional democratic theory, elections give sovereignty or ultimate
power to the citizens. They make governments accountable to the elec-
torate, at least once every five years in Britain. Voters can pass judgement
on the government and either keep it in office or replace it. Other demo-
cratic theorists view elections as only one among a number of channels
of citizen influence, stressing the indirect nature of influence through
the electoral process, and an influential version of democratic theory
suggests that elections simply allow citizens a choice between competing
elites. Free elections remain, nonetheless, the essential difference between
democratic and non-democratic states and, at a very minimum, they
provide a peaceful way of changing governments. Notwithstanding the
rash of referendums in recent years (see Chapter 9), elections also remain
the main means by which the great mass of citizens can participate in the
political process and in Britain millions of citizens do participate in this
way. For most, voting is the only overtly political action which is regularly
undertaken.
These would be reasons enough for studying elections but it is also
clear that elections do make a difference to the policies pursued by
governments and hence to the lives of most people. Exactly how and
how much elections affect what governments do is a matter of some
debate. Voters themselves disagree about how much difference it makes
when one party rather than another wins an election and it is certainly
true that governments of any party are constrained by external events
over which they have little control. Nonetheless, the political impact of
recent election results is clear. The loss of Theresa May’s majority in the
2017 election abruptly changed the kind of Brexit deal that she could
aspire to get through Parliament. It is too early to ascertain the impact of
the 2019 outcome, especially since COVID-19 would have interrupted
the plans of Labour under Corbyn as it did those of the Conservatives
under Johnson. Nevertheless, given the ideological distance between the
two major parties, it strains credulity to think that the two parties would
otherwise have charted a similar course.
As David Butler (1998: 454) observed, ‘History used to be marked off

by the dates of Kings …Now it is marked by the dates of [general] elec-
tions’. Elections are key to democracy, occasion mass political behaviour,
determine who governs and thus affect the lives of all of us. By studying
them we seek to deepen our understanding of how a key process of
democracy operates, to discover how citizens make their voting decisions
and to explain election outcomes. We may also learn how to protect these
central mechanisms of democracies, by trying to ensure that elections
remain a free and fair method of selecting representatives in the face of
external threats to electoral security.
Studying Elections
Unsurprisingly, given their importance and fascination, there is a vast liter-
ature dealing with elections and voting in Britain. This covers almost every
conceivable aspect of the subject, including whether the weather has an
influence on turnout and whether candidates whose surnames come first
alphabetically (and hence appear at the top of the ballot paper) fare better
than others. Here we focus on the two main sources of data for studying
voting: election results and sample surveys of voters.
Analysing Election Results

For as long as there have been elections, there have been results to
analyse, and so this approach has a very long pedigree. There is, for
example, a famous analysis of the relationship between votes and seats
won in British elections dating from 1905 (Edgeworth, 1905). Until the
late 1960s, such analysis was normally confined to the election results
themselves. The reason for this was that it was not until the census
of 1966 that census data were made available on a constituency basis.
From that point, however, it was possible to see how the patterns of
party support or turnout varied with the socioeconomic or demographic
characteristics of a constituency. We will consider some examples in later
chapters.
The analysis of election results can vary from the very simple to the very
complex. At the simple end, we begin by filling in the gap between 1959
and 2019 in order to assess trends in party support. The relevant data
are shown in Table 1.2 and the parties’ shares of the votes are graphed
in Fig. 1.1. Obviously, the picture is rather more complicated than that
8
Table 1.2 Party shares of votes in general elections, 1959–2019 (Great Britain)
1959 1964 1966 1970 Feb Oct 1979 1983 1987 1992 1997 2001 2005 2010 2015 2017 2019
1974 1974
% % % % % % % % % % % % % % % % %
Conservative 48.8 42.9 41.4 46.2 38.8 36.7 44.9 43.5 43.3 42.8 31.5 32.7 33.2 36.9 37.7 43.4 44.7
D. DENVER AND R. JOHNS
Labour 44.6 44.8 48.8 43.9 38.0 40.2 37.8 28.3 31.5 35.2 44.3 42.0 36.2 29.7 31.2 41.0 32.9
Liberal (Lib 6.0 11.4 8.6 7.6 19.8 18.8 14.1 26 23.1 18.3 17.2 18.8 22.7 23.6 8.1 7.6 11.8
Dems)
Others 0.6 0.9 1.1 2.3 3.4 4.3 3.2 2.2 2.1 3.7 7.0 6.5 8.0 9.9 24.6 8.0 10.6
Pedersen 3.3 6.7 4.3 6.0 13.3 2.7 8.4 11.8 3.2 5.3 12.4 2.3 5.9 6.5 15.2 15.5 8.1
index
Lab-Con 1.2 −3.1 −2.8 4.9 −0.8 −2.2 5.3 4.1 −1.7 −2.1 −10.2 1.8 3.2 5.1 −0.4 −2.1 4.7
swing
Standard 2.3 2.4 1.7 2.1 2.9* 1.5 3.1 3.0* 3.2 2.8 3.4* 2.6 2.4 3.7* 4.4 4.4 3.5
deviation of
swing
Notes The figures for the Liberals in 1983 and 1987 are for the ‘Alliance’ between the Liberals and the Social Democratic Party (SDP). Standard
deviation of swing figures for elections in which there were comprehensive boundary revisions are asterisked. In these cases, constituency swings are
based on estimates of voting in the preceding election
Source Data from Rallings and Thrasher (2012) and Uberoi et al. (2019). Swing and Pedersen index statistics are calculated from the original data
55
50
45
40
35 Conservative
% of vote
30 Labour
25 Liberals/LD
Others
20
15
10
0
1959 1964 1966 1970 1974 1974 1979 1983 1987 1992 1997 2001 2005 2010 2015 2017 2019
Feb Oct
Fig. 1.1 Party shares of votes in general elections, 1959–2019 (Great Britain)
suggested by looking at only the first and last elections in the series.
The period might be divided into four main phases. The first, which ran
until the two elections in 1974, featured a series of close races between
the Conservatives and Labour with just small fluctuations in their vote
shares. Understandably, these fluctuations captured most attention at the
time. In hindsight, however, the bigger story was the gains made by other
parties at the expense of the main two. The 1959 election would mark
the last time that they would win 90% of the British vote between them.
By October 1974, not only were the Liberals close to 20% but the nation-
alist parties had surged in Scotland and Wales, the SNP winning 30.4% of
Scottish votes.
The second and third phases saw dominance by the Conservatives and
then Labour. While the Liberals (in various incarnations) maintained a
healthy vote share throughout, they and the other parties were side-lined
by the big victories won first by Margaret Thatcher’s Conservatives and
then Tony Blair’s Labour. The pattern was one of big majorities, at first
eroded only slowly but then suddenly overturned as voters finally lost
patience with the governing party. The 1997 and 2010 elections thus
marked painful ends to long periods in office for the Conservatives and
Labour, respectively.
The final phase, from 2010 onwards, is harder to characterise given

some major fluctuations in vote shares, especially for the smaller parties.
It has something in common with both the 1959–1974 and the 1979–
2005 periods. On the one hand, there are some signs of the two-party
pendulum swinging slowly back to the Conservatives. This did not deliver
sustained dominance as in the 1980s, however, first because of the UKIP
spike in 2015 and then because Labour enjoyed an unexpected revival in
2017. Meanwhile, the close races and the small or non-existent majori-
ties—at least until 2019—are more reminiscent of the 1960s and 1970s,
as is the fact that the biggest shifts often involved parties other than the
big two.
Overall, the trend depicted by Fig. 1.1 is one of fragmentation and
growing volatility in British electoral politics. Ever since 1974, it has
clearly been a mistake to think of elections in terms of a simple two-party
system. The share of votes going to ‘others’ has fallen back somewhat
in the last two elections but this could be a temporary phenomenon,
driven by Brexit and the collapses of the Liberal Democrats and UKIP
in turn. There is no sign of any mass re-attachment to the Conservatives
and Labour and certainly no return to the relative calm of the 1950s
and 1960s. Each of the last five general elections has seen a big surge or
slump in the vote share of at least one of the key players and, overall,
British elections look less predictable than ever.
We can capture this growing volatility more precisely using a measure
of the extent of electoral change called the Pedersen index (see Pedersen,
1979). This is calculated by summing the changes in each party’s share of
the vote in successive elections and dividing by two. The relevant index
scores for elections since 1959 are shown in Table 1.2. They are compar-
atively low until February 1974, an election which abruptly ended the
apparent tranquillity of post-war electoral politics. There was also excep-
tional volatility in 1997, when the huge swing to Labour ended eighteen
years of Conservative government, and in 2015 and 2017 when, respec-
tively, the Liberal Democrats and UKIP saw their vote shares collapse. As
these cases highlight, the big swings implied by volatility need not involve
only the major parties. The 2019 election looks quite sedate by compar-
ison, with Labour’s losses marking the only major change. Nonetheless,
that election’s Pedersen index score of 8.1 is higher than that for every
election between 1945 and 1974. The greater volatility of today’s elec-
tions is underlined by the fact that the final two Pedersen scores in the
series come from pairs of elections held just two years apart.
A more familiar measure of change between elections is ‘swing’. This

is similar to the Pedersen index but, rather than considering all parties
involved, it is focused on just two. Swing is a measure of the net change
in support for two parties in a pair of elections. It was developed by David
Butler, is simple to calculate and is defined as follows:
(C2 − C1 ) + (L 1 − L 2 )
2
In this formula, C 1 is the percentage share of the total vote obtained
by the Conservatives at the first election and C 2 the percentage at
the second; L 1 is Labour’s share at the first election and L 2 Labour’s
percentage at the second. By convention the parties are put in the order
shown and the effect of this is that a positive figure denotes a swing to the
Conservatives and a negative figure a swing to Labour. The parties could
appear in any order, however, and any two parties could be substituted
for Conservative and Labour.
Table 1.2 shows the Labour-Conservative swing between successive
elections and is calculated on the basis of the national distribution of
votes. The last point means that the swing shown is the ‘overall’ or
‘national’ swing. Like the Pedersen index, these swings tended to be
rather small in the 1950s and 1960s but more variable thereafter. A new
post-war record was set in 1979 with a swing of 5.3% to the Conser-
vatives, but this was dwarfed by the massive swing to Labour (10.2%)
in 1997. Compared to that, more recent figures have been more modest,
but the five-point swings to the Conservatives recorded in 2010 and 2019
would have been startling by pre-1970s standards.
In the past, swing was a widely used and very useful measure of elec-
toral change. National swing provided a neat summary of the overall
trend; constituency-level swings could then be compared to identify devi-
ations from that trend. The main reason why swing became such a
ubiquitous feature of electoral analysis, however, is that there were few
such deviations. As Crewe (1985: 101–103) put it, ‘To know the swing
in Cornwall was to know, within a percentage point or two, the swing
in the Highlands; to know the results of the first three constituencies to
declare on election night was to know not only which party had won –
but by how many seats’. The last point was something of an exaggeration
but it was a pardonable one. Swing did indeed tend to be in the same
direction and of the same magnitude over the country as a whole. Thus,
once the results in a relatively small number of seats had been declared,
a reasonable estimate of the national swing could be made, and the final
result of the election could be predicted with some confidence, even to
the extent of identifying which seats would change hands.
Again, things have changed. This is confirmed by the final statistic in
Table 1.2: the standard deviation of constituency swings. This gives an
indication of the extent to which swing varied across constituencies in
each election: whether it was broadly similar (lower score) or varied a
good deal (higher score). Amid some short-term fluctuations, the long-
term pattern is one of greater variability in more recent elections than
in the early elections reported in Table 1.2. Overall shifts in support
between the major parties have become less uniform across the country.
In 2019, for example, despite the big overall swing to the Conserva-
tives, 31 constituencies recorded a swing to Labour. Every single one
of these is estimated as having had a ‘Remain’ majority in the 2016 refer-
endum, confirming the cross-currents of Brexit as a major reason for
recent variations in swing.
Another reason why swings vary more, and why national swing is a
less useful notion in the context of more recent elections, is the growing
vote for the Liberal Democrats and other parties. Swing was developed
in a situation in which elections were essentially contests between two
parties only. It cannot tell us about relative change between three (or
more) national parties, plus strong regional contenders. There have been
attempts to devise three-way swing figures (Miller, 1981) but these are
complicated to calculate and lack the elegance and simplicity of tradi-
tional swing. The commonest way of measuring aggregate or net electoral
change nowadays is simply to use the changes in each party’s percentage
share of the vote from one election to another.
Surveys of the Electorate

The second major source of data about elections and voters is sample
surveys. Using the information collected from a sample of individuals,
generalisations can be made, within limits, about the population from
which the sample was taken. People unfamiliar with survey research
frequently express disbelief when it is pointed out that, on the basis of a
sample of around 1,000 people, broad generalisations can be made about
the whole British electorate. Yet the principle is the same as that of a
doctor needing just a few millilitres of blood in order to draw conclusions
about the five litres circulating in someone’s body. Without going into
sampling theory, we simply assert that if the sampling methods are appro-
priate this can be done with a fair amount of accuracy and that surveys
are, therefore, among the most important tools of empirical social science.
At the same time, the experience of polling errors at recent elections—
covered in Chapter 6—teaches us that we must be careful consumers of
survey results.
Academic survey studies of voting behaviour were pioneered in the
1940s in the United States. They arrived in Britain in the 1950s but were
initially local studies, surveying individual constituencies like Greenwich
(Benney et al., 1956) or Bristol North East (Milne & Mackenzie, 1958).
The first national survey study of the British electorate was undertaken in
1963 under the direction of David Butler and Donald Stokes. Although
1963 was not an election year (the survey was one of two focusing on the
1964 election), this proved to be the first of a series of national surveys
carried out at every general election since 1964 under the auspices of the
British Election Study (BES). Butler and Stokes themselves covered the
1964, 1966 and 1970 elections. A team from the University of Essex took
over for the elections of 1974 (two) and 1979. The study returned to
Oxford where, from 1983 until 1997, Anthony Heath, Roger Jowell and
John Curtice were the principal investigators. A new Essex-based team ran
the study from 2001 to 2010 and then the duopoly was finally broken:
the current BES team, with five principal investigators, is based at the
University of Manchester (see www.britishelectionstudy.com). Over the
years, the BES surveys have formed the bases of a number of major works
on voting in Britain and hundreds, possibly thousands, of book chapters,
scholarly articles and papers.
One core component of the BES has remained constant over the
decades: a post-election survey based on face-to-face interviews with a
representative sample of British voters, designed not only to record but
also to explain whether, how and why they voted. Nonetheless, the scale
and complexity of the studies have increased considerably. From 2001 the
BES has commissioned the online survey agency YouGov to collect daily
survey data during the ‘short’ campaign. The 2015–2017–2019 edition of
the study involved almost 100,000 respondents, interviewed face-to-face
or online, and over 20 ‘waves’ of data collection beginning a year before
the 2015 election and still running at the time of writing. Beyond the
BES, there are individual election studies for the Scottish Parliament and
the Wales and Northern Ireland Assembly elections, and in 2010 there
was an ethnic minority add-on to the BES, providing much larger samples
of Britain’s minority groups than are available from standard surveys. In

addition, there is a steady stream of more ad hoc surveys designed to
address questions about elections and voting behaviour.
Opinion polls are another important source of survey data about indi-
vidual voters. While the public usually pays attention to polls only at
election time, the leading political pollsters (such as Ipsos MORI, YouGov
and Opinium) monitor the opinions of the electorate on a regular basis,
producing monthly ‘trackers’ which, in addition to current voting inten-
tion figures, record details of the voters’ perceptions of party leaders,
government performance, current issues and much else. Polling firms are
commercial agencies, of course, and carry out political polls on behalf of
clients (mainly newspapers and television). Their clients are not particu-
larly interested in obtaining the kind of detailed information about voting
choice and the factors affecting it that academics require. Poll interviews
tend, therefore, to be much shorter than interviews for major academic
surveys and are mostly conducted online or by telephone rather than face
to face. Nonetheless, polls constitute a valuable source of individual data,
not least because the methods used mean that those data are available
very rapidly. Almost as soon as an election is over, commentators use poll
results to analyse voting patterns. In contrast, it may take many months,
if not years, for reports on major academic surveys to become available.
The reliability and accuracy of survey results vary with the type and size
of sample used. A first important distinction is between probability and
non-probability samples. In a probability sample, respondents are chosen
at random from as comprehensive a list of the electorate as is available,
such that everyone has a more or less equal probability of being selected.
This has been the method used for the BES’s face-to-face post-election
surveys. Non-probability samples lack that element of randomness and so
some people are likelier to be selected than others. An obvious case is the
Internet surveys run by YouGov and others, in which only those who sign
up to be part of that agency’s panel can be selected. Since these volunteers
are not a remotely random subset of the electorate, notably being more
politically engaged and likelier to vote than the average, polling compa-
nies have to apply various adjustments to their samples to try to make
them more representative of the electorate. As discussed in Chapter 6
when assessing the polls’ performance in recent elections, this process can
present difficulties.
Even if a poll is conducted with the purest random sample, there is
still a degree of unreliability involved. The rule of thumb is that, given
a sample of around 1,000 respondents, it is highly probable (95% likely)

that the support found for each party will be within about three points
either way of that party’s true level of support in the electorate as a whole.
So, if an opinion poll of that size puts the Conservatives on, say, 48%,
that means that it is highly probable that the Conservatives’ true level
of support is between 45 and 51%. This estimate of the accuracy of the
poll estimates is the margin of error. Incidentally, there is nothing in
the term ‘error’ that implies mistakes on the part of pollsters here. All
sampling involves this kind of random ‘noise’ (in the same way that, if
someone tosses a fair coin 1,000 times and gets heads 48% of the time,
they would not immediately conclude that the coin is biased—because
the expected 50% is within the margin of error). The important point to
note is that figures derived from surveys should not be regarded as being
precise. They are estimates of the true situation among the population
being studied. This is a fact which journalists and commentators appear
to forget when they come up with detailed political explanations for shifts
of just one or two percentage points—the kind of shifts that might easily
simply reflect sampling error.
Surveys are also liable to other sources of error (see Heath & Johns,
2010). Questions may be ambiguous or unclear, lead respondents to give
particular answers, be incomprehensible to the respondent or be inter-
preted in a way that was not intended by the questionnaire designer.
Often respondents obligingly answer a question even if they do not under-
stand it or have no opinion on the matter. They may also give the answer
that they suspect the interviewer or researcher wants to hear, and avoid
admitting to opinions that they perceive as socially unacceptable. More
mundanely, mistakes may be made by interviewers in recording answers.
Even so, sample surveys are generally reliable and powerful research
tools. They have become an indispensable part of electoral analysis and
have played a crucial role in advancing our understanding of electoral
behaviour.
Election Results and Survey Data Compared

There is a key distinction between election results and surveys. Surveys
provide individual-level data. They tell us, for a sample of individuals,
both who they voted for and usually also some other information about
their background and opinions. Election results are a form of aggregate-
level data: that is, data that relate to an aggregate or collectivity such
as a constituency, region or country. We know, for example, that in the

Lancaster and Fleetwood constituency at the 2019 general election there
was a 64.7% turnout and the distribution of votes was 46.8% Labour,
41.6% Conservative and 11.6% for the others combined. This result was
obtained by staff at the election count totting up, or aggregating, the
number of people who voted and the party that they voted for. From the
final result we do not know whether or how any individual voted, let alone
what they thought about any other issue, but we do know something
about voters overall in the constituency. Other examples of aggregate data
are the percentage of council tenants in a ward, the number of people
aged 65 and over in a constituency, the percentage of manual workers
in the North of England and the change in Labour’s share of the vote
between 2017 and 2019 over the country as a whole.
There are some important advantages in analysing aggregate data. It
costs many thousands of pounds to employ a firm to undertake a major
national survey of the British electorate, whereas there is an abundance of
aggregate data freely available for analysis by anyone armed with a laptop
and a knowledge of some elementary statistical techniques. They also refer
to the total population being studied and are therefore not susceptible to
sampling error and certain other biases in the way that survey data are. We
know from election results the precise percentage of the electorate which
turned out to vote, while a survey would enable us only to estimate the
percentage which claimed to have done so.
There are some important limitations to aggregate data, however. One
practical point is that it restricts the analyst to material that has been
collected and published by official bodies. Until an appropriate ques-
tion was included in the 2001 census, for example, there were no reliable
figures showing the distribution of different religions or Christian denom-
inations across constituencies (except in Northern Ireland). In surveys, on
the other hand, the investigator can ask respondents for any information
that seems appropriate. Then there are two more fundamental points.
They might be summed up by saying that, while aggregate data are very
good at answering ‘how many?’ questions about voting behaviour, they
are much less useful for the ‘who?’ and ‘why?’ questions that are central
to our understanding of elections.
On the ‘who?’ point, it is an important feature of these data that they
tell us only about the aggregates in question—be they wards, constituen-
cies, regions or countries. Any inferences we make about individual voters
from these are just that, inferences, and may be incorrect. There is a
technical term, the ecological fallacy, for the error of drawing conclu-
sions about individual voters from the patterns in aggregate data. A vivid
example comes from the rise of the Nazi Party in 1930s Germany. This
went hand in hand with soaring national unemployment rates and the
party also polled especially heavily in districts where there was high unem-
ployment. Most observers concluded that the Nazis must have appealed
in particular to the angry or desperate jobless. Yet that conclusion goes
beyond the data. Later and more finely-grained analysis indicated instead
that unemployed voters turned to the Communists while the employed
middle classes, seeing in the lengthening dole queues the threat of unrest
and revolution, turned to a Nazi Party promising order, security and a
steadfast opposition to communism.
The other crucial advantage of survey data is that they can begin to
tell us why people vote the way they do. Aggregate statistics refer only to
the objective characteristics and behaviour of a population; voting deci-
sions are driven by people’s beliefs, attitudes and judgements. Consider
the hypothesis discussed just above—that unemployment drives support
for parties on the far right of politics—which remains a live debate in
studies of voting in Britain and beyond (Rooduijn, 2018). Underlying
that hypothesis are various ideas about why the jobless might turn to a
radical right-wing party like UKIP or the Brexit Party. One is that unem-
ployed people are prone to blame immigrants for their plight and so
are attracted by an anti-immigration party. Another is that unemploy-
ment leads to frustration and anger with the system and, in turn, to
sympathy with the anti-establishment stances of radical right parties. Atti-
tudes like hostility to immigration and anger with the system can only be
measured (and, even then, only imprecisely) using surveys. Thus, if we
want to know why certain characteristics are associated with party choice,
individual-level data from surveys are crucial. Without them, theories
about why people vote the way they do would remain highly speculative.
The limitations of aggregate data are well illustrated by returning to
the topic of swing. As explained above, swing is a summary measure of
net electoral change based on aggregate data (namely, election results).
It cannot tell us how individuals behaved. The five per cent Labour-to-
Conservative swing between 2017 and 2019, for example, does not mean
that five per cent of the electorate switched their vote from Conserva-
tive to Labour. It is possible that more than five per cent did so; it is
also possible that no one did. This can be easily understood if we distin-
guish the different components of electoral change. Whether over the
country as a whole or in individual constituencies or wards (assuming

no boundary changes) the aggregate swing between two consecutive
elections is produced by four sorts of change by individuals:
1. Switching between major parties. Some people who voted Labour in

the first election will vote Conservative in the second, and vice versa.
2. Other party traffic. Here again switching parties is involved but this
time from the smaller parties (such as the Liberal Democrats or
Greens) to one of the major parties, and vice versa. (There may also
be some movement between different smaller parties but this will
not affect the Labour-to-Conservative swing calculation.)
3. Non-voting traffic. Some people who did not vote in the first elec-
tion will vote in the second; others who voted first time round fail
to do so the second time.
4. The physical replacement of the electorate. The numbers fluctuate a
little but, every year, around 750,000 people in Britain turn 18
and therefore become eligible to vote, while about 600,000 people
die. Immigration to and emigration from the country or a particular
constituency can have a similar effect.
Measures like swing simply summarise the effects of all of these ebbs
and flows. Only if there is imbalance in one or more type of switching—
one major party gaining more from other parties, or losing less through
abstention or physical replacement—will that individual-level switching
translate into aggregate-level change. For more detailed information
about the various components of change we have to turn to individual-
level data produced by surveys.
Ideally, a survey study designed to analyse electoral change would inter-
view a sample of voters after one election and then re-interview the same
people after a second election. This is a panel survey. It is well estab-
lished that many voters do not accurately recall which party they voted for
some years previously and the use of a panel reduces this problem. Panel
surveys have their own problems, however. They suffer from ‘attrition’ as
many respondents—especially those less interested in politics—drop out
and the sample becomes unrepresentative; and even this approach strug-
gles to measure the impact of the physical replacement of the electorate
(since it is impossible to re-interview people who have died or emigrated
between the two elections). Nonetheless, the panel design is an extremely
powerful tool for those analysing electoral change. The usual method is
to construct a two-way table showing exactly what people did at the two
elections concerned. A table of this kind is sometimes called a ‘flow of
the vote’ table and Table 1.3 is an example of this based on BES Internet
panel respondents who were asked about their votes in the 2017 and
2019 elections. The table shows how survey data can provide informa-
tion about the various elements of electoral change which could not be
derived from election results alone.
In most British elections, switching between the two major parties
is relatively rare. That was less true in 2019 (and in 2017) than usual
because the cross-pressures of Brexit provided an unusual powerful
inducement for Labour Leavers and Conservative Remainers to cross to
the other side. However, since a larger proportion of former Labour
voters switched to the Conservatives (10%) than vice versa (4%), this
already accounts for part of the overall swing between the two. Labour
also lost more voters to the other parties than did the Conservatives,
although it picked up more of those who had voted Liberal Democrat
in 2017. On the other hand, the Conservatives made much greater gains
from ‘others’ (mainly former UKIP voters). Finally, Labour also lost out
from the switching between voting and abstention. Since Conservative
supporters tend to turn out more dependably, the general pattern is that
Labour loses more of its former voters to abstention but also gains more
of those who return to the polls. Only the first of these happened in 2019,
however, the Conservatives proving slightly more popular among erst-
while abstainers. The likely reason is explored in the next two chapters:
Table 1.3 The ‘flow of the vote’ between 2017 and 2019
2017 vote
2019 vote Did not vote Conservative Labour Lib Dem Other
% % % % %
Did not vote 75 11 17 12 18

Conservative 10 76 10 11 21
Labour 9 4 61 16 9
Liberal Democrat 3 6 7 53 5
Other 3 4 6 8 47
N 8,835 11,314 11,175 2,207 2,832
Source BES Internet panel, Waves 13 and 19

in a Brexit-flavoured election, the Conservatives made headway among

those groups with lower socioeconomic status that are typically less likely
to vote.
The analysis above concerns the flow of votes between two elections.
Provided that panel survey data are available, the same method can be
applied to electoral change between any two points in time. For example,
many election surveys include a ‘pre-post’ panel, in which voters are asked
about their vote intention—and many other things—a few weeks before
the election and then, following polling day, asked about how they actu-
ally voted. This allows for a flow-of-the-vote table tracking the effects of
the campaign. This reveals not only how many people changed their mind
but also the patterns of switching, indicating which parties made gains
and where these came from. This was particularly useful in 2017 when,
although there was a big overall swing from the Conservatives to Labour
during the campaign, flow-of-the-vote data revealed that very little of this
was directly between the two parties (Mellon et al., 2018). Rather, Labour
squeezed support from elsewhere, notably the Liberal Democrats and the
Greens and a group of people who had previously voted Labour but were
contemplating abstaining. Survey data also helped to identify the reasons
for the shift, notably the much-improved evaluations of Jeremy Corbyn
by previously wavering voters. Entering the 2019 campaign with similarly
dismal ratings, this time there was no miraculous recovery in Corbyn’s
popularity among potential Labour supporters and this goes some way
to explaining why the party itself did not rally as dramatically as it had
in 2017. This underlines the point made earlier: surveys are essential for
testing our theories and understanding why people end up voting the way
they do.
Theories of Voting
The study of elections and voting behaviour is not just about collecting
facts and ‘number crunching’ on the basis of complicated statistical anal-
ysis. As in all social sciences, facts have to be selected, collected, ordered,
analysed and interpreted. For that, theory is required. We have never yet
seen a survey study of voting which asked respondents what colour their
eyes are, for the simple reason that there is no theory suggesting that
eye colour may affect a person’s choice of party. Theories give guidance
as to which facts should be collected and how they should be organised,
interpreted and explained (see Sanders, 2017).
In electoral analysis there is an assortment of theories at different levels

of generality. At a simple level, someone might have a theory that married
people are more likely to vote because they are more likely to conform
to socially approved behaviour. This is better described as a hypothesis
and it could be tested by asking a sample of electors whether they were
married, whether they voted, and about their reactions to various exam-
ples of conventional and unconventional behaviour. At a slightly more
general level, there are various theories which seek to explain why older
people tend to be more likely to vote Conservative than younger people,
the changing pattern of party choice among Britain’s BAME voters, the
effect of newspapers on opinions and so on. These will be considered in
later chapters. Here, however, we introduce two even more general theo-
ries or approaches which have strongly influenced how voting in Britain
has been understood and explained, and which inform much of the rest
of the book.
The Sociological Approach and the Michigan Model

The first of these theories suggests that voting is largely a product of the
social situation of the voter. The authors of the first-ever survey study
of voting behaviour in the United States, The People’s Choice (Lazarsfeld
et al., 1944), had intended to focus upon short-term factors affecting
voting choice in the presidential election of 1940—the book was subtitled
‘How The Voter Makes Up His Mind in a Presidential Campaign’—
but in the course of their research they became more impressed by the
importance of social characteristics such as class, religion and race in struc-
turing party choice. Lazarsfeld and his colleagues discovered that they
could predict a person’s vote with considerable accuracy from knowl-
edge of just a few social characteristics. Their blunt conclusion was that
‘a person thinks, politically, as he [sic] is socially. Social characteristics
determine political preference’ (p. 27). Describing relationships between
various social and demographic characteristics and party choice is not in
itself very useful, however. As with the example of unemployment and far-
right voting described above, what is required is some theory that explains
why there should be a link between specific social or personal characteris-
tics and voting. We need an answer to the question: ‘Why are some social
differences associated with political differences whereas others are not?’.
In 1948, the authors of The People’s Choice carried out a second study
(Berelson 1954), extending their original argument. In particular, they
provided an answer to the above question by suggesting that for a social

difference to be translated into a political cleavage three conditions need
to be fulfilled:
1. ‘Initial social differentiation such that the consequences of political

policy are materially or symbolically different for different groups;
2. Conditions of transmittibility from generation to generation; and
3. Conditions of physical and social proximity providing for continued
in-group contact in succeeding generations’ (p. 75).
The first condition requires that the social groups concerned must have
differing material or symbolic interests which are affected by government
policy. Thus council tenants and owner-occupiers might have different
interests with regard to housing policy. Policy in areas such as abortion
or embryo research might not directly affect many people but are of
symbolic importance to some groups (such as Roman Catholics). The
second and third conditions relate to the processes by which social and
political divisions are maintained and reinforced. In Chapter 3, we show
how these conditions used to be fulfilled in the context of social class in
Britain—but are much less clearly in evidence today.
We have here, then, what might be termed an ‘interests plus sociali-
sation’ theory or model. Different social groups have different interests
and hence different policy preferences. They tend, therefore, to vote
for different parties which they perceive as representing these inter-
ests. Awareness of a group’s distinctiveness and of group-party links is
sustained by regular contact with fellow group members in the family,
among peers and in the community.
This model is appealingly simple and has played a large part in
voting research in Britain. However, it is not without difficulties. One of
these is the issue of overlapping group memberships. People belonging
to the same social class may differ in terms of religion, ethnicity or
income bracket, calling into question the idea that they all share the
same interests. This might explain why the sociological approach is often
confounded by ‘deviants’: that is, the (often large) minorities who do not
conform to group voting norms. A case in point is British Indians. In
the 2017 and 2019 elections, around 40 per cent of that group voted
Conservative (Martin & Khan, 2019). This looks surprisingly high if
we assume that ethnicity was their primary group identity, given that the
Conservative vote share among BAME voters as a whole was only 20 per
cent. However, if social class had driven the decisions of these predomi-
nantly affluent, suburban voters, we might have expected an even larger
Conservative vote share. The theory offers few clues about which group
membership will be decisive in determining an individual’s party support.
That is partly because it risks giving the impression that voters do all the
work, automatically calculating their interests and finding the party that
best serves them. This ignores the active role that political parties play in
mobilising and structuring the electorate—that is, in encouraging voters
to feel certain identities while downplaying others. Third, the theory also
neglects the key part played by national identities. Voting in Scotland has
long been partly a competition between Scottish and British identities
which largely transcend the kinds of social divisions discussed by Berelson
and his colleagues (Carman et al., 2014, ch. 2). Meanwhile, an assertion
of English—as opposed to British—identity was part of the UKIP surge
in 2015 that attracted votes across social classes (Hayton, 2016).
Some of these difficulties were addressed—at least in part—by an
important extension to the sociological model. This, too, derives from
the study of voting in the United States, and specifically from a famous
book—The American Voter (Campbell et al., 1960). Campbell and his
colleagues developed what has come to be known as the ‘Michigan
model’ (since the original research was directed from the University of
Michigan). It shared the same foundations as the sociological model in
its suggestion that long-term factors are most important in determining
party choice. However, in the Michigan model there is no simple step
from social location to voting behaviour. Rather, the social position that
an individual occupies affects the kinds of influences that he or she will
encounter in interacting with family, friends, neighbours, work colleagues
and so on. As a consequence of these interactions, especially those within
the family, the individual acquires a party identification. This means a
sense of attachment to a party, a feeling of commitment to it, being a
supporter of the party and not just someone who happens to vote for
it from time to time. It has something in common with national iden-
tity: just as most of us think of ourselves as English, Scottish, Welsh,
British or whatever, many also think of themselves as Conservative or
Labour. Other analogies might be support for a football team or the kind
of ‘brand loyalty’ that some consumers have for a particular make of car
or chocolate bar.
It is important to grasp that identifying with a party is not the same as

voting for it; indeed, it is possible to identify with one party and vote for
a different one. When an election comes along, voters may feel conflict
between their long-term party identification and various short-term influ-
ences, such as current political issues, campaign events, the personalities of
party leaders or candidates, or the tactical situation in their constituency.
The Michigan team were at pains to emphasise, however, that voting
‘against’ party identification is rare. This is partly because partisanship
powerfully influence how voters interpret and evaluate those short-term
factors—issues, party leaders and so on. For example, if a Labour and
Conservative identifier were to sit together and watch the leader’s speech
at the Labour Party conference, they would probably have very different
views not just on the substance of the speech but also on the quality of
the performance, the sincerity of the body language and even the calibre
of the jokes. The term ‘perceptual screen’ was coined to describe this
tendency for identifiers to interpret politics in a manner consistent with
their partisan sympathies. A blunter description would be ‘bias’.
When party identification is widespread, it has important effects on
the political system as a whole. Most obviously it provides an element
of stability and continuity. If people identify with a party they are not
likely to shoot off in all directions at successive elections (just as people—
at least, adults—who identify themselves as Manchester United supporters
don’t suddenly switch their affections to Liverpool). Rather, they will have
a ‘normal’ vote which, in most cases, will remain stable from election to
election and across different types of election.
Party identification was central to the model of party choice in Britain
that was developed by Butler and Stokes (1969, 1974) in their pioneering
survey studies. Simplifying drastically, the picture that Butler and Stokes
drew of the British voter at this time was as shown in Fig. 1.2. The promi-
nence of both social class and party identification makes it a combination
of the sociological approach and the Michigan model. The starting point
of the model is the class and party of voters’ parents and it is under-
pinned by the theory of political socialisation, which suggests that families
are particularly important in transmitting political attitudes and beliefs
to succeeding generations. Butler and Stokes saw party identification as
being inherited to a large extent through the family and leading almost
automatically to support for one party or another in elections.
Fig. 1.2 Butler–Stokes model of party choice (individual voters)
Voting as Rational Choice

One of the most striking aspects of the sociological approach is that it
makes little reference to voters’ opinions about the policies or perfor-
mance of parties. Although the Michigan model does include ‘issue
orientation’ and ‘candidate orientation’, and Butler and Stokes also allow
for short-term influences on party choice, their role is very much down-
played. The two models are largely concerned with voting as a function of
social and psychological processes. A quite different broad approach stems
from rational choice theory (see Eriksson, 2011). In its modern form
rational choice theory derives from economics but it has been applied in
a variety of ways in a large number of fields. In economics, the starting
point of the theory is an assumption that economic actors (whether firms
or individuals as consumers) act rationally: that is, before deciding on a
course of action (such as buying a particular product) they weigh up the
costs and benefits of the various alternatives and take the decision which
maximises the benefits and minimises the costs to themselves. On the basis
of this assumption, economic behaviour can be successfully predicted.
When applied to political activity, this elegant and simple idea—and
propositions deduced from it by logical reasoning—has been fruitful
in yielding insights in a number of areas, including the behaviour of
parties, pressure groups, politicians and voters. As with consumers, it can
be hypothesised that voters weigh up the pros and cons of voting for
different parties (or of voting at all) and opt for the course of action
which they think will bring them the greatest benefit. This apparently
simple suggestion has importantly influenced recent electoral analysis,
spawning a variety of more specific approaches to understanding party
choice such as issue voting, spatial or proximity voting, economic voting,
and valence or ‘performance’ politics—the notion that voters choose the
party that looks most capable of governing effectively. Admittedly, it is a
broad definition of rational choice that includes all of these approaches.
For example, the valence model of voting has acquired an importance in
recent studies of British voting such that it could reasonably have been
singled out as a third model in itself. The different approaches that we
have collected under the heading of ‘rational choice’ represent different
ways of specifying how voters evaluate the ‘benefits’ available from the
competing parties; otherwise, saying that voters choose the party that will
bring them the greatest benefit is trivial or even tautological. The reason
for collecting them together is that, in this book, we use the term ‘rational
choice’ largely as a contrast with the sociological or Michigan approach
and the word ‘choice’ is just as important as the word ‘rational’. The point
is that, rather than voting more or less automatically in line with long-
standing attachments, voters instead make a choice anew at each election
on the basis of what the parties are offering at that time. It is also worth
noting that ‘rational choice’ does not imply arduous and complex calcu-
lations of the costs and benefits associated with each party. It may well
be more rational, especially for those less engaged with politics, to find a
quick and simple way of judging the parties.
To summarise, then, theories or models are required to provide frame-
works within which appropriate data can be collected and understood
and also to provide explanations for the empirical relationships that are
discovered. Whether the data are at the aggregate or individual level,
theory guides us about which variables are likely to matter—which socioe-
conomic characteristics of a constituency to investigate, which survey
questions to ask. Having outlined the most influential general theories
of voting behaviour here, we will discuss them again, offering some eval-
uation of their relative usefulness, in the final chapter. Before examining
patterns of party choice, however, it is important to discuss another deci-
sion that electors have to make—whether to vote at all—which logically
must precede a decision about which party to vote for. The next chapter,
therefore, is concerned with turnout in elections.
References
Benney, M., Gray, A. P., & Pear, R. H. (1956). How people vote. Routledge &
Kegan Paul.
Berelson, B., Lazarsfeld, P., & McPhee, W. (1954). Voting. University of Chicago
Press.
Butler, D. (1998). Reflections on British elections and their study. Annual
Review of Political Science, 1, 451–464.
Butler, D., & Stokes, D. (1969). Political change in Britain (1st ed.). Macmillan.
Butler, D., & Stokes, D. (1974). Political change in Britain (2nd ed.).
Macmillan.
Campbell, A., Converse, P., Miller, W., & Stokes, D. (1960). The American
Voter. Wiley.
Carman, C., Johns, R., & Mitchell, J. (2014). More Scottish than British: The
2011 Scottish Parliament Election. Palgrave Macmillan.
Crewe, I. (1985). Great Britain. In I. Crewe & D. Denver (Eds.), Electoral
change in Western democracies (pp. 100–150). Croom Helm.
Edgeworth, F. (1905). The law of error. Transactions of the Cambridge Philo-
sophical Society, 20, 35–65 and 113–44.
Eriksson, L. (2011). Rational choice theory: Potential and limits. Palgrave
Macmillan.
Hayton, R. (2016). The UK independence party and the politics of Englishness.
Political Studies Review, 14(3), 400–410.
Heath, O., & Johns, R. (2010). Measuring political behaviour. In M. Bulmer, J.
Gibbs, & L. Hyman (Eds.), Social measurement through social surveys (pp. 47–
68). Ashgate.
Lazarsfeld, P., Berelson, B., & Gaudet, H. (1944). The people’s choice. Duell,
Sloan and Pearce.
Leeke, M. (2003). UK election statistics: 1945–2003. House of Commons
Library.
Martin, N., & Khan, O. (2019). Ethnic minorities at the 2017 British general
election. Runnymede Trust.
Mellon, J., Evans, G., Fieldhouse, E., Green, J., & Prosser, C. (2018). Brexit or
Corbyn? Campaign and inter-election vote switching in the 2017 UK General
Election. Parliamentary Affairs, 71(4), 719–737.
Miller, W. (1981). The end of British politics. Clarendon Press.
Milne, R. S., & Mackenzie, H. C. (1958). Marginal seat. Hansard Society.
Pedersen, M. N. (1979). The dynamics of European party systems: Changing
patterns of electoral volatility. European Journal of Political Research, 7 , 1–26.
Rallings, C., & Thrasher, M. (2012). British electoral facts 1832–2012. Biteback.
Rooduijn, M. (2018). What unites the voter bases of populist parties? Comparing
the electorates of 15 populist parties. European Political Science Review, 10(3),
351–368.
Sanders, D. (2017). Behaviouralism. In V. Lowndes, D. Marsh, & G. Stoker

(Eds.), Theory and methods in political science (4th ed., pp. 20–38). Palgrave
Macmillan.
Uberoi, E., Baker, C., Cracknell, R. (2019). General election 2019: Results and
analysis (Library Briefing Paper, no. 8749). House of Commons.
CHAPTER 2
Turnout: Why People Vote (or Don’t)
Electoral turnout is a variable. At the aggregate level it varies from

country to country and within the UK it varies from one type of election
to another and from one election to the next. In any one election it varies
across constituencies or wards. Turnout also varies at the individual level:
some people vote and some don’t. Before discussing aggregate varia-
tions, however, it is worth noting that accurately measuring turnout—the
number of people who cast a vote as a percentage of the eligible elec-
torate—is not as straightforward as it might appear as there are problems
relating to both of the figures involved.
To start with the numerator (the number of people who cast a vote),
the media and even the House of Commons Library (which publishes
reports of election results) usually take this to be the sum of votes cast
for all parties and candidates, i.e. the number of valid votes. However,
this excludes those who cast a vote but whose ballots were rejected for
some reason—if, for example, they had deliberately spoiled their ballot
paper or accidentally voted for more candidates than they were supposed
to. Usually, of course, the number of rejected ballots is relatively small
(totalling almost 113,000 across Britain in 2019) but our view is that
they should be included in the turnout calculation. A bigger ‘numer-
ator problem’ relates to postal voting. Following the introduction of ‘on
demand’ postal voting just before the 2001 general election, the propor-
tion of votes cast in this way rose steadily (and constituted almost 21%
© The Author(s), under exclusive license to Springer Nature 29

Switzerland AG 2022
D. Denver and R. Johns, Elections and Voters in Britain,
https://doi.org/10.1007/978-3-030-86492-7_2
of ballots included in the count in 2019). However, a number of well-

publicised cases of corruption came to light and in 2006 the rules relating
to postal voting were tightened. Applicants now had to provide their
date of birth and signature on the relevant form and also on the state-
ment accompanying their ballot, which is then checked for validity. Voting
by post, therefore, is more complicated than voting in person and, as a
consequence, many postal ballots—in the 2019 general election almost
150,000 (2.2% of those returned)—are disqualified on technical grounds
and excluded from the count. Whether these (attempted) votes should be
included in the turnout calculation is a moot point. Some may be genuine
attempts which involved a minor slip; others may be fraudulent.
Problems relating to the denominator (the ‘eligible electorate’) are
more serious. In many countries the figure used is an estimate of the total
population of voting age but in the UK it is the number of names on
the electoral register. The latter (commonly known as ‘the voters roll’)
used to be revised annually on the basis of forms which the ‘head of
household’ in every residence was legally required to complete. When the
register came into force (in February) it was already some three months
out of date and, since no names could be added or deleted for a year,
it then became increasingly inaccurate. In 2014, however, three impor-
tant changes to the system were introduced. First¸ every individual had
to apply separately to be put on the register. Second, ‘rolling registration’
meant that people could apply at any time up to a few weeks before an
election and third, applications could now be made online.
Despite these changes, the accuracy of the electoral register remains
questionable. On the one hand, it is legal to be registered (but not to
vote) in more than one place. Students, for example, may be registered at
home and at their university. Since electoral registration is the respon-
sibility of local government, however, there is no central register and
hence no way of knowing the number of names entered more than once.
On the other, there are citizens legally entitled to vote but not on the
register—perhaps because they have moved house, don’t realise that they
have to register or simply never get round to (the relatively simple task of)
doing it. Estimates of the numbers involved run into millions, with non-
registration being especially prevalent among young people, those who
rent their homes privately and (to a lesser extent) ethnic minorities (Elec-
toral Commission, 2019). Since these groups usually disproportionately
favour Labour, the normally obscure and technical matter of electoral
registration has acquired a political dimension and online ‘registration
Another random document with
no related content on Scribd:
“Will Taft, what do you think we waked you up for? You can’t go
back to sleep. We want you to stay awake and comfort us!”
“All right, Maria,” said he, with the utmost good nature;
whereupon he sat up, changed his position to one more comfortable,
and proceeded to lapse again into peaceful slumber.
The next morning Maria and I drove down through the town to see
the effects of the typhoon. Three trees were uprooted in our own
garden, and across the street a house was flattened out. Groups of
Filipinos stood here and there talking and gesticulating in their usual
manner, but nobody seemed unduly excited. We saw many houses
unroofed, and once in a while we met a native with a piece of nipa or
tin roofing balanced on his head, quietly carrying it back where it
belonged.
We drove down through the Escolta and into the crowded Tondo
district beyond, and there we suddenly found ourselves hub-deep in
a flood. The below-the-sea-level quarters were under several feet of
water, and we got a sudden revelation as to why all the nipa houses
are built on such high and unsightly stilts. Crowds of Filipinos were
paddling through the flood, most of them carrying some part of a
house, or other belonging, and nearly all of them playing and
splashing like pleased children. Bancas—long canoes from the river
—were plying from house to house as if it were an everyday affair and
conditions were quite normal.
I had heard a great deal about the severity of typhoons, but as I
had passed a whole season in the East and had crossed the China Sea
during the typhoon season without encountering one, I began rather
to scoff at the general fear of them. But I never did after that; when
anybody said typhoon I knew exactly what it meant. The water
subsided rapidly and in a day or two Manila showed few signs of the
fury which had passed, but for several days the Commission
continued to receive reports pf the damage done and the lives lost
throughout the surrounding country. It was the worst and the last
storm of that year.
When we arrived in Manila we found the social atmosphere
somewhat peculiar. Members of our own party, who had crossed the
Pacific on the Hancock, welcomed us at once with dinners and teas
and other kinds of parties; also a number of Army ladies called
without delay, and our circle broadened rapidly. But General
MacArthur, who was the Military Governor and lived at Malacañan
Palace, did not entertain anybody except a select military circle. He
sent an aide with cards, of course, and he accepted our invitations to
dinner, but that was all. Not that we minded, except that it made it
rather awkward and added something to the “feeling” that all was not
well between the Army and the new civil government.
The Commission had been for three months busily engaged in
investigating conditions, as directed by the President, before they
assumed any authority, and then they acted with no haste. We were
impatiently awaiting news from America with regard to the
Presidential election. It was thought to be futile to take any definite
steps toward the establishment of local governments and the
inauguration of far-reaching reforms until the status of American
control should be settled. Mr. Bryan had promised political
independence, and if Mr. Bryan were elected all the Commission’s
plans would go for naught.
The provincial and municipal codes were completed; certain
important questions between the Church and the people were being
considered, and many open sessions were held for discussion, with
the purpose of advising the people that they would be listened to by a
civil government. In the meantime the insurrectos were keeping
things lively in a guerilla warfare with small squads of greatly
harassed and very much disgusted American soldiers. There were
occasional rumours about uprisings in Manila—when the guard at
our gate would be doubled—but Mr. Taft assured us that Manila was
as safe as New York or Chicago and we really had few fears.
General MacArthur continued to resent the coming of the
Commission and to consider himself personally humiliated by their
being appointed to divide his power. He was still in command of
about seventy thousand men and had the general executive control of
a large civil force, but this, apparently, was not enough. The tone he
adopted in his correspondence with the Commission kept them in a
constant state of controlled anger. They were very careful in return to
observe every courtesy and to manifest an earnest desire for
harmony and co-operation. They were tremendously interested in
their problems and wanted much to succeed, but their efforts at
conciliation did little good. The General objected to almost every
suggestion put forward by them and did not hesitate to tell them in
plain words that he did not welcome advice from them concerning
military or any other matters. It was really a very difficult situation.
The Commission thought General MacArthur took an entirely
erroneous view of the attitude of the Philippine people in general,
and that in everything he did he moved with an exasperating
slowness. They wanted a large native constabulary which they knew
could successfully be organised and relied upon to render great
assistance in the pacification of the Islands. He did not agree with
them and held the matter up for many months. He was not in
sympathy with any move they made, and his greatest cross was that
he had no power to veto their legislation. He saw military dangers in
all manner of things without being able to state just what they were,
and he was always calling for more troops, while the Commission
was entertaining hopes that it would not be a great length of time
before a large part of the troops already there could be recalled. I find
my husband writing at this time:
“General MacArthur, knowing that we differ from him as to the
condition of things in the Islands, makes it a point to send me an
account of each disaster as if it vindicated his view. This is not the
spirit of a man who is likely to succeed in giving energy to a
campaign which will bring about successful results, but the matters
will solve themselves in spite of his slowness of movement and lack
of enthusiasm....
“The minute the policy with respect to these Islands is settled by
Bryan’s defeat and the election of McKinley, the leniency which has
been almost too great towards ladrones and these murdering
generals will have to be changed. They must be given an opportunity
to come in and if they do not come in in a short time, they ought to
be deported from the country and sent to Guam. This will have an
effect so healthy that a short time will see accomplished what we
desire. There will be a great awakening for some of these men who
have come to rely on the supineness of the Americans, and who do
not understand that we can be severe when we choose....
“It was General Otis who inaugurated the plan of laughing at the
insurrection, of capturing men and letting them go, and the result is
that they have laughed at us, but with a little tightening of the reins
their laugh will cease....
“They dread deportation more than anything else and I have
written to Secretary Root and asked him to have a prison constructed
at Guam to which we may send those whom we think worthy of a less
punishment than hanging. The insurrection must be suppressed for
the benefit of the United States and, still more, for the benefit of the
Filipino people. The lenient methods, having been tried for two
years, must be changed to those more severe....
“The insurrection, such as it is now, is nothing more than a
conspiracy against the sovereignty of the United States sustained by
murder and assassination of Filipinos by Filipinos....
“MacArthur is drawing the reins a little tighter, though not as tight
as we think he ought to draw them, and he has now imprisoned
about fifteen hundred insurgents. There have been a great many
arrests made in Manila, which has been the head centre of the
insurrection in the way of raising money. I should think there have
been fifty or sixty insurgent officers arrested in the city....
“I sent a telegram to the Secretary of War on Sunday night which
was signed by Buencamino and other prominent Filipinos, about a
dozen of them, in which they spoke out with emphasis about the
continuation of the insurrection. They propose to organise what they
call a counter-revolution; that is, they mean they will organise a
military movement among the Filipinos against Filipinos. They are
getting very tired and weary of this murder and assassination policy
without which the insurrection could not last a week....
“You could hardly believe the closeness with which the
Presidential matters are being watched by the Filipinos, and how
they follow the speeches made against the Republican cause. General
Smith, away down on the island of Negros, told me he had found
speeches by Hoar and Bryan, and other anti-expansionists and anti-
imperialists, in the most remote mountains of his district....
“Every one is waiting and it is not impossible that should Bryan be
elected there might be some riotous demonstration among the
natives. The Chief Justice of the Supreme Court, Señor Arellano, has
made arrangements, should Bryan be elected, to leave the islands
three days after the announcement. He is the ablest Filipino in the
islands, by far the best lawyer and a man of the highest probity. He
says that much as he is interested in the success and prosperity of his
fellow citizens, he knows that they are utterly incapable of self-
government and should the guiding hand of the United States be
withdrawn, chaos, conscription and corruption would follow
inevitably....”
I have taken these excerpts at random from my husband’s letters
to his brother during the months of September and October, 1900,
and they serve to show the situation which existed and will illustrate
the fact that we were living in interesting times. But they deal only
with the insurrection, while the main body of his correspondence
shows that the Commissioners were engaged upon legislative
matters of the gravest import which would be rendered entirely
superfluous should Mr. Bryan be elected and his announced policies
be carried into effect. In that event they proposed immediately to
turn matters back to the military government and withdraw, leaving
Mr. Bryan to face the problems which they knew he would soon
discover had to be dealt with from the standpoint of constructive
statesmanship.
In the meantime the peace movement was rapidly gaining
adherents among the people in spite of the still active insurrectos, or
rather, because of them and their methods; while everybody seemed
to welcome the change from a strictly military to a partially civil
government.
The popularity of the Commission, as offering a change from the
strictness of military rule, was becoming every day more marked.
Juan de Juan, a Spaniard, and editor of the lively organ El Progreso,
which was always in opposition to anything American, said that on
the first of September when the Commission began to exercise its
authority, he intended to devote the whole front page of his paper to
just three words: “Gracias a Dios,”—Thanks to God! Juan de Juan
was a good deal of a Bohemian and really cared little what happened
so long as he got a sensation out of it. September first came and
went, and I don’t remember whether he made good this extravagant
threat or not. I presume he didn’t for, though I had been in Manila
less than a week, I surely would have remembered.
After the Commission had been in power for just a month, and
while the excited interest in events in the United States was at its
height, Juan de Juan broke out in a characteristic Spanish editorial, a
translation of which has been preserved. We had entertained Juan de
Juan at dinner, and he evidently was impressed. We made it a rule
from the beginning that neither politics nor race should influence
our hospitality in any way, and we came thus to have a very wide and
diverse acquaintance. The editorial in El Progreso gives such a
curious picture of attitude and conditions in general, as well as of my
husband, my family and my home, that I think I must quote it,—at
least in part. It is headed simply:
SEÑOR TAFT
The most uncompromising jingoes; the rabid partisans of militarism, as well as
the men of democratic sentiments who consider the occupation of the Philippines
as an odious Cæsarism, respect and venerate the President of the Civil
Commission, whose surname serves as the caption of these lines. Uprightness and
bonhommie always demand recognition.
Before the Hancock, bearing this statesman, had anchored in Manila Bay, the
echo of his reputation and the radiations of the brilliant aureole which his success
in the judiciary of his country had imposed upon him—and we underline the word
imposed because the characteristic trait of Mr. Taft is his modesty—had reached
the Philippines. The Filipinos awaited him with the same pleasing curiosity with
which a child opens a toy with a concealed surprise, and the foreigners as the
incarnation of those American patriarchal, democratic ideas with which Castelar
portrayed to his followers the country of Lincoln.
Behind that spacious brow of the thinker, between his liberal tendencies and the
incomparable exactions of the enormous burdens which his country undertook in
Paris, fierce struggles are waging. The President of the American Civil Commission
has broad shoulders, but the weight of a people whom patriotism endows with the
strength of a colossus is very great.
We must concede to all the leading authorities whom America has sent to the
Philippines the trait of being industrious. We know that General Otis worked more
than twelve hours a day; MacArthur, that Daban of the American Army through
the rapidity of his advancement, follows the same course as his predecessor, and
Mr. Taft leaves his house every morning at eight and, as unostentatiously as a
clerk, proceeds to become a part of his chair in the Ayuntamiento. There his first
occupation is glancing over the American press, and what is of interest in the
Spanish papers.
Then the show begins. Paterno, Macabulos, Montenegro, some envoy from Cebu,
for example, who come to sound him, as the slang saying goes, arrive. Mr. Taft has
the same respectful smile for all, the same courtesy, and addresses them all in the
same terms, which his athletic Secretary, Mr. Fergusson, repeats in Spanish with
the gravity of a Sphinx and the fidelity of a phonograph. When the matter warrants
it, Mr. Pepperman, the chief stenographer of the Commission, enters the office and
proceeds to take notes of the interview.
In this way the Americans are forming a luminous record which, united to what
were our archives, which they preserve through the terms of the Treaty of Paris,
will guide them well in the administration of the Philippines.
Later Mr. Taft becomes engulfed in the examination of the bills which the other
members of the Commission present for him to study; he discusses their text with
his colleagues, listens to all their observations, and judging them by a standard
most favorable to the interests of the Philippines, the most liberal within the
instructions from Washington—it is proper to say that Mr. Taft is the most
democratic element of the Commission—he expresses his opinion, generous, calm
and noble, which assuredly, in view of his personal prestige, must carry great
weight in the framing of the bills, whose execution is entrusted to the Military
governor.
To dissipate the gloomy smoke of the conflagration, to still the groans of those
who fall in this immense expoliarium into which fatality has converted the
Philippine fields, is the mission which the men composing the American
Commission desire to bring to a successful issue. To make peace. For this they
came, and if fortune does not reserve for them the happy chance of accomplishing
so beautiful an ideal, they will retire, and the factor they represent in the problem
to be solved, with its distinguishing traits of civil moderation, will be substituted as
a system that has failed, by another, wherein the martial power will prevail over
political wisdom.
As General MacArthur undoubtedly spends many hours over maps of the
Philippines, Mr. Taft also often rests his gaze on a map covering one of the walls of
his office, tracing, in mente, a railroad which, crossing the island, shall drown with
the cheery whistle of the locomotive the moans of the victims of war. Thus would
Mr. Taft like to pacify the Philippines.
It is now one o’clock P.M. and Mr. Taft is at home, where this personage stands
out more boldly before us, since the trials through which the country is passing do
not permit us yet to judge him politically.
The President of the Commission, in his private life, has many points of
similarity with Count de Caspe, that stainless gentleman the Filipinos still recall
with veneration. Excepting the brilliancy of those splendid entertainments with
which he endeavoured to blot out all racial differences by mingling in fraternal
embrace Filipinos and Spaniards at the Malacañan villa, there ordinarily reigned in
the governor’s mansion the placid silence of the home of a well-to-do retired
merchant. The Countess, who on Thursdays did the honours of her salon with
exquisite tact, was during the other days of the week a housekeeper who did not
disdain to go to a grocery store to make purchases, or to look over the laundry list.
The same thing happens in the elegant chalet at Malate where Mr. Taft lives.
This is a quiet and peaceful home, a temple erected to the affections, under whose
roof Mr. Taft rests some hours after the efforts which his political work demands.
His table reflects his modest character. Four courses, two kinds of fruit, a dessert
and sauterne compose the menu of the luncheon where Mr. Taft is always
accompanied by some guest, either Filipino, American or Spanish. During the meal
politics are banished; if the guest is a Filipino who speaks French Mrs. Taft
interrogates him on the customs of the archipelago; if he is Spanish, as to the
toilettes worn in Manila by the ladies at the most brilliant receptions held here; as
to the favourite musical composer of the Hispano-Filipino society; and this
conversation increases in attraction when Miss Herron, sister-in-law of Mr. Taft
and the incarnation of the modern woman’s education, takes part therein. Miss
Herron speaks French correctly, has travelled much, and journeyed through Spain
like an intelligent tourist. The architectural lace-work of the Alhambra charmed
her, and she went into ecstasies over the orange blossoms growing along the banks
of the Guadalquivir. With what Miss Herron was not in harmony, and she berates
them like an unsubsidised journalist, were the Spanish railroads. Miss Herron is
right.
The children, Robert, about eleven years old; Helen, a girl of nine, and Charles, a
baby of three, who is the king of the household:—the McKinley, as it were, of this
patriarchal republic—do not come to the table; they eat with the governess.
After the meal, in the fine gallery overlooking the sea, sipping the coffee, Mr.
Taft talks of the education of his children, of the difficulties met in the Philippines
in the solution of so interesting a problem; and his wife converses of the charitable
work she expects to undertake when she shall have assumed a more permanent
place in the Archipelago, which Magellan discovered for Spain, and which, through
a horrible fatality, is no longer ours. Politics are also eschewed on the gallery.
Needless to say this extraordinary editorial afforded us all
boundless amusement; we began to caution Mr. Taft frequently
about the careful preservation of his “aureole” and Maria and I
decided that we would have to walk warily indeed, if we were
destined to be so minutely reported.
CHAPTER VI
A STRANGE ENVIRONMENT
In the Far East one meets certain expressions the significance of

which may be described as adamantine. Each represents a racial
attitude against which it is useless to contend. In Japan it is the
equivalent of it cannot be helped; a verbal shrug of the shoulders
with which the Japanese tosses off all minor and many grave
annoyances. “Masqui,” down the China coast, has the same import,
but with the added meaning of “what difference does it make.” In the
Philippines the phrase which must be met and which cannot be
overcome by any system of reform is “el costumbre del pais”—the
custom of the country.
If it is el costumbre del pais it has to be done and there is nothing
more to be said about it. The manaña habit—putting everything off
until to-morrow—is, perhaps, to Americans, the most annoying of all
the costumbres del pais in the Philippines, but it yields to pressure
much more readily than do many others, among which is the custom
of accumulating parientes; that is, giving shelter on a master’s
premises to every kind and degree of relative who has no other place
to live. This is, I suppose, a survival of an old patriarchal
arrangement whereby everybody with the remotest or vaguest claim
upon a master of a household gathered upon that master’s doorstep,
so to speak, and camped there for life.
In my first encounter with this peculiarity of my environment I
thought there was a large party going on in my cochero’s quarters;
and an indiscriminate sort of party it seemed to be. There were old
men and old women, young men and young women, many small
children and a few babes in arms. We had only Chinese servants in
the house, but the stables were in charge of Filipinos and, as I soon
discovered, the “party” was made up entirely of our stablemen’s
parientes.
(LEFT TO RIGHT) GENERAL WRIGHT,

MR. TAFT AND JUDGE IDE AS
PHILIPPINE COMMISSIONERS
I had a pair of ponies and a Victoria; Mr. Taft had his two little
brown horses and a Victoria; besides which there was an extra horse
to be used in case of accident to one of the others, as well as a pony
and calesa for the children. This rather formidable array was
necessary because we found it impossible to take a horse out more
than twice a day, and usually not more than once, on account of the
sun. My ponies were taken out only in the early morning or the late
evening, and those of Mr. Taft had all they could do to take him to
the office and bring him home twice a day. Distances were long and
there were no street cars which ran where anybody wanted to go.
This number of conveyances made a good many stablemen
necessary and all of them, with their families, lived in quarters
attached to the stables. These families consisted of fathers, mothers,
sisters, brothers, uncles, aunts, cousins near and far removed, wives,
children, grandchildren, and a few intimate and needy friends with
their family ramifications. Besides our three cocheros and the stable
boys, there was a gardener with his parientes, so it is no wonder that
on my first inspection of the lower premises I should have thought
that some sort of festivity was in progress. I might have lived in
Manila twenty years without being able to straighten out the
relationships in this servant colony; it was not possible to learn who
had and who had not a right to live on the place; and my protest was
met with the simple statement that it was el costumbre del pais, so I,
perforce, accepted the situation.
Filipino servants never live in the master’s residence; they never
want to; they want the freedom of a house of their own, and these
houses are, as a rule, built on the outer edges of the garden, or
compound. I believe Americans now are learning to meet the
pariente habit by having room for just as many people as they need,
and no more. But those who live in the old places, with their ample
quarters, still gather the clans and are permitted to enjoy a most
expansive and patriarchal sensation.
My horses, when I first saw them, were a source of the greatest
pride. A beautifully matched pair of coal-black, stylishly-paced and
glossy little stallions, hardly larger than Shetland ponies, they looked
as if they had been washed in some sort of shrinking soap and had
come out in perfect condition except that they were several sizes
smaller than they ought to have been. These Philippine ponies are
doubtless descendants of the Arabian horses brought over by the
Spaniards and have been reduced to their present size by the change
of climate and the difference in food and environment, but they still
have the fine lines and the general characteristics of their
progenitors.
Mr. Taft secured mine from Batangas, where all the best ponies
come from, through the kindness of Mr. Benito Legarda, the
staunchest of Americanistas. Batangas was a most unquiet province,
the last, in fact, to become pacified, and Mr. Legarda had to pay an
insurrecto for bringing the horses through the insurgent lines and
delivering them at Calamba, near Manila. Although he did not know
their exact origin when he bought them, Mr. Taft said that if the facts
became known he would be accused, in certain quarters, of giving
indirect aid to the revolutionists; but he wanted the ponies so he did
not return them.
When they were hitched to the shining little Victoria which had
been built for them, they were as pretty as a picture and, as I did not
propose to have such a turn-out ruined by a couple of Filipinos on
the box in untidy camisas hanging outside of as untidy white
trousers, I had made for my cochero and boy, or coachman and
footman, a livery of white and green in which they took such
inordinate pride that they seemed to grow in stature and dignity.
Maria and I felt a sense of the utmost satisfaction the first time we
stepped into this carriage for a drive down to the Luneta where we
were sure to see everybody we knew and hundreds of people besides;
but our vanity was destined to be brought to a sudden termination.
As we were driving along with much satisfaction, a bit of paper
floated down alongside the blinkers of the little ebony steed on the
right and he made one wild leap into the air. His companion gave
him an angry nip, and then the fight was on. Maria and I jumped out,
which was not difficult in a low-built Victoria, and no sooner had we
done so than we saw the complete wreck of all our grandeur. With all
the leaping and plunging and biting and kicking, in the vicinity of a
handy lamp-post, the smash-up was fairly complete. Neither of the
ponies was hurt, except by the lash of the whip, and I must say the
little wretches looked rather funny; like very pretty and very bad
children, sorry for what they had done. But their characters were
established and they proceeded after that to live up to them. We
never could have any confidence in them and my coachman was the
only person who could do anything with them. He was a most
unsatisfactory man in many ways and used often to call for us at
dinner parties in a state of gay inebriety, but we didn’t dare discharge
him because everybody else in the stables stood in awe of the blacks
while he seemed greatly to enjoy his constant and spectacular
struggles with them.
The Filipinos are a most temperate people; there is no such thing
as drunkenness among them; but coachmen seem to be an exception
in that they allow themselves a sufficient stimulation of the fiery vino
to make them drive with courage and dash, sometimes minus all care
and discretion. The drivers of public vehicles seem to love their little
horses in a way; they are inordinately proud of a fast paced or
stylish-looking pony; yet they are, as a rule, quite harsh to them.
They overload them and overdrive them, and under all conditions
they lash them continuously.
No Filipino cochero likes to have another cochero pass him, and
the result is constant, indiscriminate racing, on any kind of street,
under any circumstances,—and never mind the horse.
My children were driving with their governess to the Luneta one
evening, when two caromatas came tearing down behind them, each
driver hurling imprecations at the other and paying no attention to
what was ahead of him. The result was a violent collision. The two
caromatas went plunging on, the cocheros not stopping to see what
damage they might have done—which was very characteristic—and
the children narrowly escaped a serious accident. Charlie was hurled
out and fell under the children’s calesa and Robert and Helen both
declare they felt a sickening jolt as a wheel passed over him. The
baby, too, vowed that the calesa “went wight over me, wight dere,”
indicating a vital spot; but upon the closest examination we could
discover nothing more serious than a few bruises. However, it made
us very much afraid to trust the children out alone.
The gardener had two little boys, José and Capito, who were a few
years older than Charlie, but about his size, and he took a
tremendous fancy to them. They were clad, simply, in thin gauze—or
jusi—shirts which came down a little below their waists, and I think
Charlie envied them this informal attire. He used to order them
around in a strange mixture of Spanish, Tagalog and English which
made me wonder at my wholly American child; but it was an
effective combination since he seemed to have them completely
under his thumb and, as he revelled in his sense of power, he never
tired of playing with them.
Maria and I soon adopted the universal habit of driving down to
the Escolta in the early morning to do such shopping as was
necessary. We found a variety of interesting shops, but with very
little in them to meet the ordinary demands of an American woman.
There were delightful Indian bazaars and Chinese tiendas where all
manner of gaudy fabrics and strange oriental articles were on sale,
while the Spanish shops upon which everybody had to depend in
those days, and which had such grandly European names as Paris-
Manila and La Puerta del Sol, catered largely to the Filipino taste for
bright colours.
The Escolta at that time was full of saloons, established by the
inevitable followers of a large army, and the street being very narrow
and the old, rickety, wooden buildings being very wide open, the
“beery” odour which pervaded the atmosphere at all hours was really
dreadful. Mr. Taft decided that as long as this was the only street in
town where women could go shopping, the saloons would have to be
removed. There was opposition on the Commission to the bill which
provided for their banishment, and it was fought from the outside
with great vigour and bitterness, but a majority were in favour of it,
so it passed, and the saloons had to move. There has not been a
saloon on the Escolta from that day to this and, indeed, they have
ever since been under such satisfactory regulation that there is little
evidence left of their existence in the city.
I am afraid it is going to be very difficult to convey an adequate
picture of Manila society during the first years of American
occupation. There had been, in the old days, a really fine Spanish and
rich mestizo society, but all, or nearly all, of the Spaniards had left
the Islands, and the mestizos had not yet decided just which way to
“lean,” or just how to meet the American control of the situation. I
may say here that most of the educated, high-class Filipinos are
mestizo; that is, of mixed blood. They may be Spanish mestizo or
Chinese mestizo, but they have in them a strong strain of foreign
blood. Besides the Spanish- and Chinese-Filipinos, there are a
number of British mestizos who are very interesting people. Mr.
Legarda, Chief Justice Arellano, Dr. Pardo de Tavera and Mr.
Quezon, the Filipino delegate to the United States Congress, are
Spanish mestizos, while Mr. Arañeta, the Secretary of Finance and
Justice, as well as the Speaker of the Philippine Assembly and many
able lawyers and successful business men are of Chinese descent.
The mestizos control practically all the wealth of the Philippines, and
their education, intelligence and social standing are unquestioned. It
is the only country in the world that I know about—certainly the only
country in the Orient—where the man or woman of mixed blood
seems to be regarded as superior to the pure blooded native.
Dating back also to the Spanish days was quite a numerous foreign
society consisting of a few consuls, some professional men, the
managers of banks and large British and European mercantile firms,
and their families. The leaders of the British colony were Mr. and
Mrs. Jones—Mr. Jones being the manager of the Manila branch of
the Hongkong and Shanghai Banking Corporation. Mrs. Jones, a
very beautiful and charming woman, gave some very elaborate
parties during that first winter. Bank House, the residence
maintained by the bank for its manager in Manila, is a beautiful
place in Uli-Uli, a district on the picturesque banks of the upper
Pásig, and it is finely adapted for balls and large receptions. Then
there were several German families who also entertained quite
lavishly, and I remember, especially, one Austrian exile; indeed, I
shall never be able to forget him because my husband took such joy
in pronouncing his name. He was Baron von Bosch.
This was the “set” which entertained the Commission most
cordially during our first season in Manila, while the Army officers,
following the lead of their Commanding General, held themselves
somewhat aloof. I kept up a constant round of parties of different
kinds in my house, and gave a dinner at least once a week at which
were gathered companies of a most interestingly cosmopolitan
character. And we did not fail to observe all the desirable forms. Both
Filipinos and Europeans expect a certain amount of ceremony from
the representatives of government and are not at all impressed by
“democratic simplicity”; so believing in the adage about Rome and
the Romans, we did what we could. Beside the spic and span guard at
the outer gate of the illuminated garden, we always, on dinner party
nights, stationed coachmen, or other stable boys disguised as liveried
footmen, on either side of the entrance, to receive guests and conduct
them to the dressing-rooms, and up the stairs to the reception room.
Our house was nicely adapted for a dinner of twelve and I usually
tried to confine myself to that number. We always had an orchestra,
orchestras being very plentiful in Manila where nearly every native
plays some sort of instrument, and the music added greatly to the
festive air of things, which was enhanced, too, by a certain oriental
atmosphere, with many Japanese lanterns and a profusion of potted
plants and great, hanging, natural ferneries and orchids which were
brought in from the forests by the Filipinos and sold on the streets.
My husband is supposed to be the author of the phrase: “our little
brown brothers”—and perhaps he is. It did not meet the approval of
the army, and the soldiers used to have a song which they sang with
great gusto and frequency and which ended with the conciliating
sentiment: “He may be a brother of William H. Taft, but he ain’t no
friend of mine!”
We insisted upon complete racial equality for the Filipinos, and
from the beginning there were a great many of them among our
callers and guests. Their manners are models of real courtesy, and,
while their customs are not always like ours, wherever they are able
they manifest a great willingness to be conforme,—to adapt
themselves,—and their hospitality is unbounded.
I shall never forget my first call from a Filipino family. They
arrived shortly after six in the evening: el señor, la señora and four
señoritas. We went through a solemn and ceremonious handshaking
all around. I received them first, then passed them on to my husband
who, in turn, passed them on with a genial introduction to my sister
Maria. We had been sitting on the verandah, and when a semi-circle
of chairs had been arranged, the six of them sat down; el señor
noisily cleared his throat a couple of times while the ladies calmly
folded their little hands in their laps and assumed an air of great
repose. It was as if they had no intention of taking any part whatever
in the conversation.
El señor explained in Spanish that they were our near neighbours
and that they had called merely to pay their respects. Mr. Taft had
been studying Spanish diligently ever since he left the United States,
but he is not conspicuously gifted as a linguist, and he had not yet
waked up—as he so often expressed a wish that he might—to find
himself a true Castilian. However, his ready laugh and the cordiality
of his manners have always had a peculiar charm for the Filipinos,
and he was able on this occasion, as he was on many future ones, to
carry off the situation very well. We all nodded and smiled and said,
“Si Señor” and “Si Señora,” to long and no telling what kind of
speeches from our guests; then Maria and I complimented the ladies
on their beautifully embroidered camisas, which started things off
properly. They praised everything in sight, and what we didn’t get
through the little Spanish we knew, we got from gesture and facial
expression. They got up and wandered all around, feeling of my
Japanese tapestries and embroideries, breathing long “ahs!” of
admiration over my gold screens and pictures and curios, and acting
generally like callers who were being very well entertained. Then the
children came in and they broke out afresh in voluble praise of them.
I assumed the proper deprecatory mien in response to their
laudation of my children, and altogether I felt that we were
acquitting ourselves rather well in this first inter-racial social
experience.
But at the end of half an hour the strain was getting a little severe
and I was wondering what to do next, when our six callers arose and
said they must be going. I breathed an inward sigh of relief and was
making ready to escort them to the top of the stairs, when my
husband cordially exclaimed:
“Why, no! Porque? Tenemos bastante tiempo. Why hurry?” And—
they—all—sat—down!
I regretted then even the little Spanish Mr. Taft had learned,
though, of course, he didn’t expect them to heed his polite protest.
He knew nothing at all about Filipino manners; he didn’t know they
expected to receive some sign from him when it was time to go and
that they would consider it discourteous to go while he was urging
them to stay. He kept up, without much assistance, a brave if
laboured conversation, and the minutes slowly passed. Our dinner
hour approached and I darted warning glances at him, for I had a
horrible fear that he just might ask them to remain and dine. But at
the end of another hour a strained expression began to spread itself
over even his face, and there was not a word of protest from him
when, at a quarter past eight, our little brown neighbours once more
indicated an intention of going home. We entertained Filipino callers
nearly every day after that, but never again did we urge them to
reconsider their sometimes tardy decision to depart.
With regard to Filipino manners and customs; I am reminded that
we were nonplussed, though greatly amused by the costumbre del
pais which decreed that some return be made by a Filipino for any
and all favours bestowed upon him. We grew accustomed to this
before we left the Islands, and came to expect a few offerings of sorts
almost any day in the week, but in the beginning it was usually most
embarrassing.
One time, soon after our arrival, a very loyal Americanista was
shot down in the street, during the peaceful discharge of his duty, by
an insurrecto. His widow, with her children, came into Manila in a
state of utter destitution, to secure some recompense from the
government for her husband’s services, and while her case was
pending Mr. Taft, in great pity for her, sent her money enough to live
on. The next day the whole family, from the wide-eyed boy to the
babe carried astride the mother’s hip, came to call on their
benefactor, bringing with them as a gift a basket containing a few
eggs, some strange Philippine fruits and a lot of sea-shells. Mr. Taft
was deeply touched, and with the brusqueness of a man who is
touched, he told her he had given her the money to buy food for
herself and her children and not for him, and he refused her offering.
I know, by the light of a fuller knowledge of the character of the lowly
Filipino, that she went away feeling very much cast down.
But in connection with such gifts there were always more laughs
than sighs. We invited to luncheon one day a dashing Filipino named
Tomaso del Rosario. Señor Rosario, a man of wealth and prominence
who had a fine Spanish education and was well dressed in the high-
collared, patent-leathered and immaculate-linened Spanish style,
was quite self-confident and enjoyed himself very much. He seemed
attracted to Maria and she, being linguistic, was able to talk to him in
a mixture of many languages. The next day she received from Señor
Rosario, not a floral offering, but a basket filled with nuts, a canned
plum-pudding, some canned chocolates and preserved fruits. This
attention did not seem so remarkable, however, when we learned, to
our amusement, that he had sent exactly the same present to Alice
Worcester, then five years old.
Our life, on the whole, was intensely interesting in its unusual
atmosphere and curious complications, but throughout everything
we were made to feel the deep significance of our presence in the
Islands; and the work of the Commission was first, last and always to
us the subject of the greatest moment. Even in our daily round of
social affairs we dealt with tremendous problems whose correct
solution meant the restoration of peace and prosperity to what then
should have been, and what we knew could be made, a great country.
That for which the American flag had always stood began to assume,
for many of us, a broader and a finer meaning; and being so much a
part of our flag’s mission in a strange field a certain zest was added to
our patriotism which we had never felt before. I believe, and I think
all those who know the truth believe, that Americanism, in its highest
conception, has never been more finely demonstrated than in the
work done by the United States in the Philippine Islands; work, the
broad foundation for which the Commission was engaged in
constructing during the period of which I write.
So many were the problems to be met and dealt with that in the
beginning the Commissioners were each given a set of subjects for
investigation and study, their findings being submitted for debate
and consideration in the general meetings.
Taxation, civil service, provincial and municipal organisations,
currency and finance, police, harbour improvements, roads and
railways, customs, postal service, education, health, public lands, an
honest judiciary and the revision of the code of laws; these were
some of the vital problems, but underlying them all was the
immediate necessity for the establishment of tranquillity and
confidence throughout the archipelago.
In order to make clear, in any degree, the Philippine situation as
we found it, it is essential that, briefly, the position of the Catholic
Church and its representatives, the Friars, be explained. For the first
time in its history the American government found itself compelled
to adjust a seemingly insurmountable difficulty between a church
and its people.
With us the Church is so completely separate from the State that it
is difficult to imagine cases in which the policy of a church in the
selection of its ministers, and the assignment of them to duty could
be regarded as of political moment, or as a proper subject of
comment in the report of a public officer, but in the first reports of
the Philippine Commission to Washington this subject had to be
introduced with emphasis.
The Spanish government of the Philippine Islands was a
government by the Church through its monastic orders, nothing less.
In the words of the Provincial of the Augustinians, the Friars were
the “pedestal or foundation of the sovereignty of Spain” which being
removed “the whole structure would topple over.” The Philippine
people, with the exception of the Mohammedan Moros and the non-
Christian tribes, belonged, during the Spanish dominion, to the
Roman Catholic Church, and the Church registry of 1898 showed a
total membership of 6,559,998. The parishes and missions, with few
exceptions, were administered by Spanish Friars of the Dominican,
Augustinian and Franciscan orders, and it was to the nature of this
administration that Spain owed the insurrections of 1896 and 1898,
the latter of which terminated only upon our assuming control of the
islands.
In 1896 there were in the Philippines 1,124 monks of the
Augustinian, Dominican and Franciscan orders, which body included
a company of Recolletos, who are merely an offshoot of the order of
St. Augustine and differ from the Augustinians only in that they are
unshod. In addition to these there were a few Jesuits, Capuchins,
Benedictines and Paulists, but they engaged in mission and
educational work only and did not share with the other orders the
resentment and hatred of the people. Filipinos were not admitted to
any of the orders, but they were made friar curates and served as
parish priests in some of the smaller places.
When a Spanish Friar curate was once settled in a parish he
remained there for life, or until he was too old for service, and
because of this fact he was able to establish and maintain an
absolutism which is difficult to explain in a few words. He was simply
everything in his parish. As a rule he was the only man of education
who knew both Spanish and the native dialect of his district, and in
many parishes he was the only Spanish representative of the
government. In the beginning, through his position as spiritual
guide, he acted as intermediary in secular matters between his
people and the rest of the world, and eventually, by law, he came to
discharge many civil functions and to supervise, correct or veto
everything which was done, or was sought to be done in his pueblo.
He was Inspector of Primary Schools, President of the Board of
Health and the Board of Charities, President of the Board of Urban
Taxation, Inspector of Taxation, President of the Board of Public
Works, Member of the Provincial Council, Member of the Board for
Partitioning Crown Lands, Censor of Municipal Budgets, and Censor
of plays, comedies or dramas in the dialect of his parish, deciding
whether or not these were against the public peace or morals. In a
word, he was the government of his parish; and in addition to all
things else, it was he who, once a year, went to the parish register,

Elections and Voters in Britain 1St Ed 2022 Edition Denver Full Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Elections and Voters in Britain 1St Ed 2022 Edition Denver Full Chapter

Uploaded by

Copyright:

Available Formats

Elections and Voters in Britain 1st ed.

2022 Edition Denver

David Denver · Robert Johns

Elections and Voters

ISBN 978-3-030-86491-0 ISBN 978-3-030-86492-7 (eBook)

Cover credit: steven gillis hd9 imaging/Alamy Stock Photo

fielding these surveys at short notice is difficult and expensive enough

Lancaster, UK David Denver

1 Studying British Elections 1

4 Issues, Policies and Performance 97

8 Elections and Electoral Systems 229

Glossary of Statistical and Technical Terms 313

Fig. 1.1 Party shares of votes in general elections, 1959–2019

Fig. 9.1 Percentage of ‘don’t knows’ in vote intention polls

Table 3.6 Percentage of a party’s 2017 voters reporting a different

Table 9.3 Campaign trends in certainty (per cent ‘very’ or ‘fairly’

Studying British Elections

© The Author(s), under exclusive license to Springer Nature 1

Conservative 48.8 613 352 44.7 631 365

A simple comparison of the results in the two elections illustrates some

be to entertain (or annoy, according to taste) viewers of election night

Why Study Elections?

Most people would agree that it is the existence of free, competi-

As David Butler (1998: 454) observed, ‘History used to be marked off

Analysing Election Results

The final phase, from 2010 onwards, is harder to characterise given

A more familiar measure of change between elections is ‘swing’. This

Surveys of the Electorate

of Britain’s minority groups than are available from standard surveys. In

a sample of around 1,000 respondents, it is highly probable (95% likely)

Election Results and Survey Data Compared

as a constituency, region or country. We know, for example, that in the

country as a whole or in individual constituencies or wards (assuming

1. Switching between major parties. Some people who voted Labour in

Did not vote 75 11 17 12 18

Source BES Internet panel, Waves 13 and 19

in a Brexit-flavoured election, the Conservatives made headway among

In electoral analysis there is an assortment of theories at different levels

The Sociological Approach and the Michigan Model

provided an answer to the above question by suggesting that for a social

1. ‘Initial social differentiation such that the consequences of political

It is important to grasp that identifying with a party is not the same as

Fig. 1.2 Butler–Stokes model of party choice (individual voters)

Voting as Rational Choice

Sanders, D. (2017). Behaviouralism. In V. Lowndes, D. Marsh, & G. Stoker

Turnout: Why People Vote (or Don’t)

Electoral turnout is a variable. At the aggregate level it varies from

© The Author(s), under exclusive license to Springer Nature 29

of ballots included in the count in 2019). However, a number of well-

In the Far East one meets certain expressions the significance of

(LEFT TO RIGHT) GENERAL WRIGHT,

You might also like