Measuring the Level and Direction of Split-Ticket Voting at the 1979 and 1997 British General and Local Elections: A Survey-Based Analysis
| Author | Michael Thrasher,Colin Rallings |
| Published date | 01 June 2001 |
| Date | 01 June 2001 |
| DOI | http://doi.org/10.1111/1467-9248.00316 |
| Subject Matter | Research Note |
Research Note:
Measuring the Level and Direction
of Split-ticket Voting at the 1979 and
1997 British General and Local Elections:
a Survey-based Analysis
Colin Rallings and Michael Thrasher
University of Plymouth
Only in 1979 and 1997 have British general elections coincided with the annual local government
electoral contests. This research note uses both survey and aggregate data to provide new estimates
of the extent of split-ticket voting in England at those two elections, and to compare similarities
and differences between them. It appears that no fewer than one in six of those electors who voted
for one of the three major parties in both contests made a different choice at national as opposed
to local level. The fact that the proportion of those splitting their vote in this way was so similar in
1979 and 1997 may surprise those expecting the phenomenon to have grown as part of a general
process of partisan dealignment. Rather it appears that changing patterns of contestation between
1979 and 1997, especially at local level, may be better able to account for those variations
observed.
In an earlier paper we examined aggregate level evidence about the incidence of
split-ticket voting at the synchronous 1997 general and English local authority
elections. Although the precise pattern varied between constituencies, we concluded
that at least 5 percent and up to 20 percent of electors had cast their available votes
for different parties. In those constituencies with ‘perfect, noise free competition’
at the local level between the three major parties nearly 10 percent of electors cast
split ballots (Rallings and Thrasher, 1998).
The only other occasion on which the general and local elections have coincided
was in 1979. Again, published analyses have been entirely based on aggregate data.
Waller examined local and national results in 100 constituencies and concluded
that ‘in general, electors did not split their … votes in the borough constituencies in
May 1979 [but] … in rural areas … the correlation between local and parliamentary
results is much less close’ (Waller, 1980). Liverpool, where Cox and Laver provide
compelling evidence of substantial split-ticket voting, was a clear exception, with
similar if less dramatic behaviour also being found in other towns (Cox and Laver,
1979).
A known weakness of aggregate voting data is that it necessarily hides any counter-
balancing switching by individuals. Indeed the evidence from other research using
survey data or the scrutiny of actual voting papers suggests that the true level of
ticket-splitting may be twice that revealed by the analysis solely of election results
POLITICAL STUDIES: 2001 VOL 49, 323–330
© Political Studies Association, 2001.
Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA
To continue reading
Request your trial