A Low Turnout Landslide: Abstention at the British General Election of 1997
| Author | Charles Pattie,Ron Johnston |
| Published date | 01 June 2001 |
| DOI | 10.1111/1467-9248.00314 |
| Date | 01 June 2001 |
| Subject Matter | Article |
A Low Turnout Landslide: Abstention at
the British General Election of 1997
Charles Pattie
University of Sheffield
Ron Johnston
University of Bristol
Turnout at the 1997 British General Election fell to a post-war low of 71 percent, raising worried
speculation about long term decline in political participation. On closer analysis, however, this
judgement seems premature. Most of the post-war decline in British turnout occurred between
1950 and 1970, and there is no evidence of long term decline in general election turnout between
1974 and 1997. The closeness of the electoral competition is a better predictor of national turnout
than a secular trend. Close elections produce high turnout, but widely anticipated landslides (as in
1997) produce low turnouts. The 1997 election was also notable for the small ideological gap between
the main parties. Analysis of individual voter abstention in 1992 and 1997 reveals that changes
from one year to the other in the perceived difference between Labour and the Conservatives
is crucial to accounting for the fall in turnout between the two contests. Turnout in 1997 was low
because the result was widely anticipated and because relatively few saw important policy
differences between the parties, not because British democracy is in crisis.
Although probably not a critical election in the generally accepted meaning of
the term, the 1997 British General Election produced several memorable results
(Evans and Norris, 1999). In addition to an unusually dramatic turn-round in party
fortunes, the contest produced a higher abstention rate than any previous post-war
British General Election (Denver and Hands, 1997; Heath and Taylor, 1999; Heath,
2000). Despite the unpopularity of the outgoing government (itself a post-war record),
the election did not seem to have generated widespread public involvement.
A number of possible explanations have been put forward for the high abstention
rate in 1997, from disillusioned Conservative voters to declining trust in the
political system. Some have relatively benign implications for the health of British
politics. The low turnout may have reflected electoral sophistication in the context
of a widely anticipated landslide election. Why vote if the result is a foregone
conclusion? The 1997 contest may just have been a ‘low interest’ election. If
this was the reason for the high abstention rate, turnout should increase in a less
predictable future election. There are past precedents for this. The 1970 election
for instance, was fought in the summer months, when many voters may have been
away on holiday. Turnout then fell to 72 percent, but recovered to almost 79
percent in February 1974. The October 1974 election, following hard on the heels
of the February contest, produced considerable ‘election fatigue’: turnout fell to
73 percent. But it had recovered once again by 1979, to 76 percent.
Other possible explanations are markedly less benign. In the run-up to the election,
the Conservative government had been subject to severe policy reversals, scandals,
POLITICAL STUDIES: 2001 VOL 49, 286–305
© Political Studies Association, 2001.
Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA
and splits. Labour, the main opposition party, had meanwhile accepted many of
the changes implemented by the Conservatives since 1979. It accepted that most
privatizations were now irreversible, it committed itself to operating within the
Conservatives’ projected spending plans for at least the first three years of a new
Labour government, and it was at pains to distance itself from its past reputation
as an advocate of high direct taxation. A substantial body of electors may have been
increasingly disillusioned and unhappy with the system of government as a whole,
perhaps feeling that there was little real choice on offer, that the parties did not
listen to voters, and that government was increasingly remote from their lives. Or
long-term Labour supporters might have been put off by their party’s swing into
the centre ground (although close analysis suggests this was not, in fact, a major
factor: Heath, 2000). To the extent that this was the case, 1997’s low turnout might
actually represent the beginning of an extended period of low participation rates.
No fully satisfactory explanation has yet been offered for the scale of the decline in
turnout in 1997, however. In the most comprehensive account so far, Heath and
Taylor (1999) demonstrate that changing turnout from election to election cor-
relates with the perceived closeness of the contest. Their claim rests on a comparison
of aggregate trends over time. Compelling though their evidence is, however, they
do not actually demonstrate that turnout change is caused by perceptions of
the intensity of electoral competition. In this paper, we move Heath and Taylor’s
argument on by examining the causal link. To do so, we start by examining the
1997 election in the context of turnout at post-war British elections. An aggregate
analysis investigates the claim that 1997 represents a low point in a long process of
secular decline in turnout. We then conduct individual-level analyses of turnout at
the last two British General Elections (1992 and 1997). Were the same factors
associated with individual abstention at each election, or was 1997 distinctive? And
did changing individual perceptions of the closeness of the contest account for the
fall in turnout between 1992 and 1997?
1997 in Context
A first step is to place 1997’s low turnout in the context of longer-term trends in
electoral participation. Was this a further step in a secular decline in participation
rates, or was it a one-off? An examination of turnout levels across time suggests
few clear trends (Figure 1). Turnout was low in the 1945 election. But the dis-
locations of war (the conflict in the Far East was still under way and many eligible
voters were overseas with the armed forces) and the effects of an inaccurate
electoral register in that year make this a special case. In some respects, the first
‘peacetime’ elections of 1950 and 1951 provide a better starting point. Comparing
1997 with these elections does suggest a dramatic decline in turnout, from 84 to
71 percent. But this belies a more complex pattern of change in the intervening
years. The decline in turnout has not been steady. Turnout averaged 80.5 percent
in the 1950s (though this is largely a product of the 1950 and 1951 turnouts). This
fell to 76.5 percent in the 1960s and to 74.9 percent in the 1970s. But the averages
for the 1970s, 1980s and 1990s are all remarkably similar, at between 74 and
75 percent. The four lowest post-1945 turnouts have all been after the 1960s. But
so have two of the highest, discounting the 1950 and 1951 elections. To adapt a
phrase used in another context, this seems to be trendless fluctuation.
A LOW TURNOUT LANDSLIDE 287
To continue reading
Request your trial