Explaining Split-Ticket Voting at the 1979 and 1997 General and Local Elections in England
| Published date | 01 October 2003 |
| Date | 01 October 2003 |
| Author | Michael Thrasher,Colin Railings |
| DOI | 10.1111/1467-9248.00441 |
| Subject Matter | Article |
Explaining Split-Ticket Voting at the
1979 and 1997 General and Local
Elections in England
Colin Rallings and Michael Thrasher
University of Plymouth
Only in 1979, 1997 and 2001 have British general elections coincided with the annual local gov-
ernment elections. Uniquely, this study uses both survey and aggregate data to examine aspects
of split-ticket voting at the simultaneous elections in 1979 and 1997. Through the use of bi- and
multi-variate analysis, it suggests that ticket-splitting is a product of both voter attitudes and party
strategies: although it is almost wholly confined to the less partisan, the electoral context in which
those votes are being cast can play an additional and significant role in stimulating such
behaviour.
Split-ticket voting has long attracted the attention of political scientists working in
countries where electors have the opportunity to cast simultaneous votes for dif-
ferent tiers of government or to express preferences between political parties at the
same election. Theories to account for this behaviour have focused either on indi-
vidual voters or on the activities of candidates and parties. Individuals, for example,
are seen to be prone to ticket-splitting as a result of factors such as weak parti-
sanship, the desire to moderate policy outputs and the use of different decision
rules to determine their preferred party at different types of election (Petrocik and
Doherty, 1996; Fiorina, 1988; Jacobson, 1990). In multi-party systems, an addi-
tional motivation might be provided by a voter’s desire to ensure the defeat of their
least preferred party at one of the sets of elections (Evans et al., 1998). Conversely,
other research into split-ticket voting has claimed it to be a reaction to the activi-
ties of, and messages received from, candidates and parties, which have the effect
of persuading targeted voters away from a straight party choice (Beck et al., 1992;
Petrocik, 1991; Wattenberg, 1991). In Britain, at least until the advent of two-vote
ballots for the Scottish Parliament and Welsh Assembly in 1999 and for the Greater
London Mayor and Assembly in 2000, ticket-splitting has rarely been possible. Nor-
mally, under plurality voting, a single vote is allocated to a single party at just one
election. However, in 1979, 1997 and 2001, general and local elections coincided,
so electors had to decide whether and how to distribute multiple votes.
This is the third in a series of papers examining split-ticket voting at simultaneous
British general and local elections in 1979 and 1997. The first paper (Rallings and
Thrasher, 1998) used aggregate-level data to compare and contrast the results in
76 English constituencies where a local election was held in May 1997 and where
the boundaries of the local electoral division and parliamentary seat were coter-
minous. The second paper (Rallings and Thrasher, 2001) sought to extend the
analysis of split-ticket voting in two ways: (i) by using survey data from the British
© Political Studies Association, 2003.
Published by Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
S P L I T- T I C K E T V O T I N G
559
Election Study (BES) to examine the reported behaviour of individual voters;
and (ii) by explicitly comparing the level and direction of split-ticket voting at the
1979 and 1997 elections. We found that, at each election, some one in six of those
who cast both their available votes for a major party opted to split their vote. The
present paper moves the story forward by using both survey and aggregate-level
data to develop a clearer understanding of which electors are more likely to split
their vote between parties and the reasons behind such behaviour. We will begin
by addressing five different types of potential explanation for ticket-splitting. We
will then assess the degree to which the various factors interact with each other
and whether they form a basis for modelling the differences between those elec-
tors who stay loyal to a single party and those who split their vote between com-
peting parties.
The Data
Individual-level data were taken from the BES cross-section surveys conducted fol-
lowing the general elections of 1979 and 1997. In addition to the usual range of
attributional and attitudinal questions, respondents were asked whether there was
a local election in their ward, whether they had voted in it and how they had
voted. Aggregate-level data on constituency and ward results were drawn from the
files of the Local Government Chronicle Elections Centre at the University of
Plymouth. In 1979, the vast majority of electors in England outside London had
the opportunity to vote at a local election. In 1997, the local elections were less
widespread, with about half the English electorate having simultaneous local and
general contests. There were no local elections on either occasion in Scotland,
Wales or London, nor, in 1997, in metropolitan England.
The two types of data were matched according to the codes used in the surveys to
identify each respondent’s ward and/or polling district. For 1979, we were obliged
to take on trust the respondents’ claims to have voted or not, but all those who
lived in a ward where there was not a local election, those who failed to provide
full information about their voting behaviour and those who reported a party
choice at either local or national level that was in fact unavailable are excluded
from the analysis. In 1997, respondents were subjected to the same tests, with
those who could not be proved to have recorded a vote following validity checks
conducted by the BES team added to the excluded category.1 We concentrated on
a core of respondents at each election who both passed these tests and reported
voting for one of the three major parties at both general and local elections.
Although a simple distinction between party loyalists and ‘splitters’ can be made,
more interesting findings are likely to come from dividing respondents according
to the parties for which they cast their votes. The number of cases in some of these
categories is admittedly small (Table 1), but they do represent groups between
which we would expect to find rather different characteristics and motivations.
Potential Explanations of Split-Ticket Voting
We will look at five different types of potential explanation for ticket-splitting in
turn.
560
C O L I N R A L L I N G S A N D M I C H A E L T H R A S H E R
Table 1: Categories for Analysis of Ticket-Splitting
Cases Cases
in 1979
in 1997
N
(%)
N
(%)
1. Conservative general/Conservative local
135
42.5
241
32.3
2. Conservative general/Labour local
8
2.5
6
0.8
3. Conservative general/Liberal (Democrat) local
16
5.0
33
4.4
4. Labour general/Conservative local
5
1.6
20
2.7
5. Labour general/Labour local
98
30.8
254
34.0
6. Labour general/Liberal (Democrat) local
14
4.4
39
5.2
7. Liberal (Democrat) general/Conservative local
5
1.6
19
2.5
8. Liberal (Democrat) general/Labour local
5
1.6
15
2.0
9. Liberal (Democrat) general/Liberal (Democrat) local
32
10.1
120
16.1
Total ‘splitters’
53
16.7
132
17.6
Total ‘non-splitters’
265
83.4
615
82.3
Partisan Attachment
It would be surprising if the ‘onward march of dealignment’ (Crewe and Thomson,
1999) was not reflected in how electors reacted to the opportunity to cast multi-
ple votes at a single point in time. In both 1979 and 1997, nine in 10 of those who
claimed to be ‘very strong’ partisans report giving both their available votes to their
chosen party, and all voted for that party on at least one occasion.2 The number of
‘not very strong’ partisans who split their vote was considerably higher, ranging
from 35 percent of Conservative identifiers to 60 percent of Liberals in 1979 and
from 25 percent of Labour identifiers to more than four in 10 of Liberal Democ-
rats in 1997.3 Of course, the direction of causation possibly runs the other way, in
that the act of ticket-splitting might itself have prompted the reporting of a weaker
party identification.
In terms of more general partisan attitudes, those who voted for a single party at
both elections were likely to be most positive about their own party and its leader
and, at least as far as Labour and Conservative loyalists were concerned, among
the most negative about their opponents. In 1979, respondents were asked to mark
each party out of 10. Conservative loyalists scored their own party at 8.4 and
Labour at 4.2; Labour loyalists returned the compliment, marking their party at
9.2 and the Conservatives at 4.8. Liberal loyalists gave both a lower mark to their
own party (7.2) and higher ones to their two opponents. In 1997, respondents
were asked how they felt about each party on a scale running from 1 (‘strongly in
favour’) to 5 (‘strongly against’). Conservative loyalists gave a mean score of 2.0
for their own party and one of 4.2 for Labour, and their Labour counterparts scored
their own party at 1.5 and the Conservatives at 3.5.
S P L I T- T I C K E T V O T I N G
561
Splitters tended to fall between these two extremes. However, in the majority of
cases, the...
To continue reading
Request your trial