Membership Ballots in Party Leadership Elections in Britain
Thomas Quinn
Department of Government
University of Essex
Wivenhoe Park
Colchester
CO4 3SQ
This is an Author's Accepted Manuscript of an article published in Representation, Vol. 46,
No 1 (Jan 2010): 101-117 [copyright Taylor & Francis], available online at:
http://www.tandfonline.com/doi/full/10.1080/00344890903570967.
ABSTRACT
Given the propensity of party members to be ideologically less centrist than voters, British
parties’ shift to membership ballots in leadership elections might have been expected to result
in radical leaders. This paper explains why this development has largely not occurred, using
data from four recent leadership elections. Following Stark (1996), it argues that members’
selection criteria are unity, electability and competence. The paper shows that members are
not blind ideologues but are sensitive to the need for party unity.
1
Membership Ballots in Party Leadership Elections in Britain
All three major political parties in Britain now permit formal voting rights to their individual
members in leadership elections. The Liberal Democrats have operated a pure one-memberone-vote (OMOV) system since the merger between the old Liberal Party and the Social
Democratic Party in 1988. Since 1993, Labour’s electoral college for leadership selection has
set aside one-third of the votes for party members in an OMOV ballot (see Quinn, 2004).
Finally, in 1998 the Conservatives adopted a hybrid system of preliminary parliamentary
ballots to select two candidates, who would go forward to an OMOV ballot to select the
leader (see Alderman, 1999).
There is a growing literature on party leadership selection in Britain, partly as a consequence
of this widespread institutional change and partly because the three major parties have
conducted a total of 15 contests since 1988. This literature tends to fall into two main
categories: examination of institutional rules (e.g. Alderman, 1999; Punnett, 1992; Quinn,
2004, 2005) and detailed accounts of individual contests (e.g. Alderman and Carter, 2002;
Denham and Dorey, 2006, 2007; Denham and O’Hara, 2008; Heppell, 2008). Both types of
literature have discussed the shift from selection by MPs to selection by party members.
What is largely missing, however, is a detailed examination of the criteria by which party
members choose leaders. This issue is important because it is frequently assumed that
changing the institutions of leadership selection influences the types of leaders chosen (e.g.
Punnett, 1992). For example, given the long-standing argument that party members tend to be
ideologically more radical than MPs (e.g. May, 1973), moving selection away from MPs to
members might result in more radical leaders.
One existing work in which selection criteria are at the forefront of the analysis of leadership
contests is Leonard Stark’s Choosing a Leader (1996). Stark’s conclusion was that
institutions do not matter much because when MPs and members choose leaders they have the
same goals in mind: in order of importance, these are party unity, electoral victory and policy
development. However, Stark’s book is now quite old and does not cover more recent
membership ballots – the Conservatives still allowed MPs to choose their leaders when it was
published. Moreover, much of Stark’s evidence is qualitative, relying on elite interviews.
There was very little data on party members’ preferences at the time. That has now changed.
2
Pollsters such as YouGov and ICM conduct in-depth surveys of party members during
leadership contests, data that are freely available to researchers.
This paper uses Stark’s framework of analysis to analyse four recent leadership elections in
the Conservative (2001 and 2005) and Liberal Democratic (2006 and 2007) parties. These
contests all entailed membership ballots. (The Conservative contest of 2003 and the Labour
contest of 2007 are excluded because in each case only one candidate secured sufficient
nominations.) For each contest, polling evidence on members’ preferences is consulted. The
paper shows that Conservative and Liberal Democrat members chose leaders primarily on the
basis of their ability to unite their parties, in those contests where there was an overriding
need for unity. When party unity was not a key consideration, the members chose leaders on
the basis of their perceived electability. The paper supports Stark’s argument that party
members, like MPs, choose leaders on the basis of predictable and hierarchical criteria.
The paper is organised as follows. The first section discusses the various criteria that actors
use in choosing party leaders. The second section examines the Liberal Democrats’ leadership
contests of 2006 and 2007. The third section analyses the Conservatives’ leadership elections
of 2001 and 2005. The paper concludes with some lessons drawn about party members’
selection criteria from the four cases studied.
Membership Ballots and Selection Criteria
In his analysis of leadership contests in Britain up to 1994, Stark (1996: 125-6) follows
Gunnar Sjoblom (1968) in identifying a hierarchy of three strategic goals for major parties in
parliamentary systems: unity, electoral victory and policy implementation (see Table 1).
Paramount is internal unity, the first-order goal, because it is a prerequisite for everything else
parties do. Without unity, parties find it difficult to win elections, because voters dislike and
distrust divided parties. They also struggle to agree on a coherent set of policies. If a new
leader is to achieve intra-party unity, he must be broadly acceptable to the main groups and
factions in his party. Thus, the primary selection criterion in leadership contests is
acceptability. A party’s second-order goal is winning elections because only then can it form
a government and implement policies. Moving out of the intra-party arena into the electoral
arena, the corresponding criterion for selectors choosing between rival candidates is thus their
electability. Finally, there is the parliamentary arena, where the third-order objective of
parties, once they are unified and have won elections, is to implement policies in government.
3
Formulating effective policies requires selectors to assess the candidates’ competence, which
is thus the third selection criterion.
Table 1: Strategic Goals and Selection Criteria in Leadership Contests
Order
Arena
Goal
Criterion
First
Internal
Unity
Acceptability
Second
Electoral
Victory
Electability
Third
Parliamentary
Policy
Competence
Source: Stark, 1996: 126.
Stark argues that the first-order goal of unity is a party’s priority but it is usually an explicit
consideration only in parties that are badly divided (Stark, 1996: 126). In such parties, unity
normally trumps all other considerations. Michael Foot was deemed better able than Denis
Healey to unite Labour in 1980, but Healey was more appealing to the electorate and would
have made a better prime minister. Similarly, Neil Kinnock was regarded as being more likely
than Roy Hattersley to unite Labour in 1983, but Hattersley was stronger in the electoral and
parliamentary arenas (Stark, 1996: 128).
More often, parties are not dramatically divided and so leadership elections turn on questions
of electability and competence. Ideological extremists are not usually appealing to voters,
whereas ideological centrists are generally more so. Competence is crucial, not merely in
Stark’s narrow sense of designing policies, but also in terms of leadership skills. Competent
leaders need, among other things: the ability to perform well in parliament; the appearance of
being an effective prime minister (less so for Liberal Democrat leaders); managerial skills;
and sound judgment, decisiveness and the right balance between strength and flexibility.
Ideology is usually a crucial factor underlying strategic goals and selection criteria. This point
was alluded to in relation to electability because radical leaders may repel floating voters.
However, ideology is normally a central consideration when selectors determine the extent to
which candidates are acceptable and able to unite their party. Foot and Kinnock were seen as
4
more acceptable than Healey and Hattersley because they were from Labour’s ascendant
centre-left, whereas the latter two candidates were from the discredited centre-right.
Stark’s framework was, in principle, applicable to all selectors, whether they were MPs or
party members. Stark supported his case with considerable interview evidence from MPs, but
the available evidence on party members was less extensive. There was little in the way of
polling data in those contests in which party members and activists took part. Yet, it is
important to know whether party members put unity and electability at the forefront of their
minds when choosing leaders or whether they want ideologues. This latter view was held by
many observers who cited the case of Iain Duncan Smith’s victory in the Conservative
leadership contest of 2001 (see below). The availability of better polling data enables this
argument to be examined more closely.
The Liberal Democrat Leadership Contest of 2006
The 2006 Liberal Democrat leadership contest followed the fall of Charles Kennedy as leader.
Despite increasing his party’s number of parliamentary seats in the 2005 general election, the
result was disappointing, given the unpopularity of Labour and the Conservatives (Russell,
2005: 87). The party’s leading economic liberals, associated with the controversial Orange
Book of essays that argued for less statism (Marshall and Laws, 2004), believed that it needed
to reposition itself to protect its electoral gains, particularly after David Cameron’s accession
to the Tory leadership (Denham and Dorey, 2007: 32-3, 35). Compounding this political
context was concern about Kennedy’s drinking habits. Although not widely known about in
the party, Kennedy’s drinking appeared to have an impact on his performance to the point
where it became a major issue for his critics (Hurst, 2006: 1-23, 218-42). These factors
combined in late 2005 to create a full-scale crisis for Kennedy, as he lost the confidence of his
senior colleagues. In a desperate attempt to stay on, Kennedy called a leadership contest in
which he hoped to compete. Eventually, he bowed to the inevitable and resigned.
Sir Menzies Campbell was the obvious choice to take over from Kennedy. He was the Liberal
Democrats’ elder statesman and had enjoyed ‘a good war’ during the Iraq conflict, being the
face and voice of the party’s opposition to the invasion. Ideologically, he was a centrist who
could hold together the diverging social and economic liberals (Denham and Dorey, 2007:
35). He garnered significant support from young, right-leaning ‘modernising’ MPs associated
with The Orange Book. They saw Campbell as a stop-gap, who could appoint modernisers to
5
the front-bench and take the party through to the next election, by which time they would
have gained experience and be well-positioned to take over (The Times, 2 February 2006).
Among the other figures mentioned as candidates, Simon Hughes, the Liberal Democrats’
president, was on the left of the party and appealed to its activists, while Mark Oaten, the
home affairs spokesman, was a contributor to The Orange Book, although he dropped out
after securing only one firm endorsement among his parliamentary colleagues and shortly
before he became embroiled in an embarrassing personal scandal. One other serious candidate
to emerge was Chris Huhne, a former MEP first elected to parliament in 2005. Huhne was a
contributor to The Orange Book and initially seen as right-leaning: he wanted to abandon the
Liberal Democrats’ policy of raising the top rate of income tax, although he wanted a shift to
green taxes, emphasising his environmental credentials (Daily Telegraph, 21 January 2006).
Campbell enjoyed the greatest support among Liberal Democrat MPs – he ended up with 35
declared supporters, including most leading front-benchers (Denham and Dorey, 2007: 43).
Hughes and Huhne were a long way behind. However, when the contest turned to the OMOV
stage, there was no guarantee that party members would back the preferred candidate of the
MPs, especially in a party with a history of a vocal and influential membership. The Liberal
Democrats’ members had a reputation for being left-wing in their ideological leanings.
However, a large-scale survey of the membership had found that it was fairly centrist, if
slightly skewed to the left: on a nine-point left-right scale of British politics, where 1 was the
most left-wing position, 9 the most right-wing and 5 in the centre, the mean self-placement
score was 4.1. Party members were more left-leaning on equality and redistribution, as well as
being green and strongly pro-European (see Whiteley et al., 2006: 49-54, 61).
It is not the purpose of this paper to provide a full account of the leadership campaign in
which the candidates appealed for party members’ votes (see Denham and Dorey, 2007). It is
sufficient to state that Campbell began the contest as favourite and although he suffered a
‘wobble’ when it looked as though Huhne might overtake him, he pulled through and won
convincingly. Hughes was eliminated in the first count of the alternative-vote system and after
his votes were reallocated, Campbell finished 15 points ahead of Huhne (Table 2).
6
Table 2: Liberal Democrat Leadership Election Result (2006)
First Count
%
Second Count
%
Sir Menzies Campbell
23,264
44.7
29,697
57.9
Chris Huhne
16,691
32.1
21,628
42.1
Simon Hughes
12,081
23.2
-
-
Total
52,036
100.0
51,325
100.0
Notes: Alternative-vote electoral system used. After no candidate achieved over 50% of the
vote on the first count, Hughes was eliminated and his votes reallocated on the basis of
voters’ second preferences. There were 711 non-transferable votes in the second count.
Turnout: 72.2%.
Selection Criteria
An in-depth poll by YouGov, conducted mid-way through the leadership contest, provided
information on how Liberal Democrat members rated the candidates on key selection criteria.
The poll found a narrow plurality backing Huhne (32 percent) over Campbell (29 percent) and
Hughes (23 percent), and so the data on selection criteria probably understate Campbell’s true
standing. Nevertheless, they are highly instructive. Table 3 summarises some of the most
important findings of the poll, in which respondents are provided with various statements and
asked to say which candidate they apply to most. The top half of the table contains items
related to key selection criteria and the bottom half of the table contains items related to
specific qualities that are desirable in Liberal Democrat leaders.
On the first-order criterion of ‘acceptability’, Campbell was regarded by almost half of party
members as the candidate most likely to unite the Liberal Democrats, with Huhne on 26
percent and Hughes on 15 percent (Table 3). Although the party was not extremely split on
ideology, the division between the economic and social liberal factions was real enough and
was an undercurrent to Kennedy’s resignation. The fact that the membership was more
socially liberal ensured that this factional division had the potential to turn into something
nastier, a split between MPs and party members. It is interesting to note, therefore, that the
left-leaning Hughes was seen as unable to unite the party. During the campaign, economic
7
liberals had briefed the press that there could be funding cuts and defections to the
Conservatives among some Liberal Democrat MPs if Hughes became leader (Sunday Times,
29 January 2006). The membership was aware of the dangers and eschewed a candidate long
regarded as the voice of the activists. Campbell was seen as the party’s elder statesman and a
steadying influence at a time of upheaval in the party. In this respect, he was acceptable to
both the MPs and the members, and to economic and social liberals.
Table 3: Liberal Democrat Members’ Views of Candidates’ Leadership Qualities (2006)
Campbell
Hughes
Huhne
48
15
26
30
22
30
33
25
31
36
26
27
He has right kind of experience to lead Lib Dems
58
21
16
He knows about real life outside politics
19
23
52
He is honest
46
14
28
He is committed to the right kind of policies
27
31
34
He would make effective government minister
50
19
25
Selection Criteria: Candidate best placed to…
Lead a united Liberal Democratic Party
[acceptability]
Boost the Liberal Democrats’ chances of gaining
seats at the next election [electability]
Come over well on television & radio, and in the
press [electability]
Challenge Blair & Cameron most effectively in the
House of Commons [competence]
Leadership Qualities
Source: YouGov, ‘Survey of Liberal Democratic Party Members’, poll for John Stevens, 7-9
February 2006, available at <http://www.yougov.com>.
Notes: Q on leadership qualities: ‘Here are some statements about the candidates. For each
statement, and putting aside your own voting preference, please indicate which candidate you
think it applies to most.’ All figures are percentages. N = 401
8
On the second-order criterion of electability, Campbell and Huhne were closely matched, with
Hughes not far behind. Campbell and Huhne were seen as equally likely to improve the
Liberal Democrats’ chances of gaining seats at the next election, while Campbell had a slight
lead on the candidate most likely to perform well in the media – a key consideration for
parties in seeking to broaden their electoral appeal. However, on both these items, only 8
points separated the best and worst candidates, and no-one polled higher than 33 percent. That
indicated uncertainty among party members as to which candidate would promote the party’s
electoral chances at a time when the Conservatives were advancing under Cameron. On the
basis of these figures, it is unlikely that Campbell won because of his perceived electoral
appeal in comparison with the other candidates (cf. Denham and Dorey, 2007: 43).
On the third-order criterion of competence, Liberal Democrat leaders are not usually seen as
likely to enter government, except in hung parliaments and coalition governments (Campbell
was seen as most likely to make an effective government minister – see lower half of Table
5). However, competence is still crucial because the leader is expected to challenge the prime
minister and the leader of the main opposition party in the House of Commons. Campbell
enjoyed a 9-10 point advantage over the other candidates on this criterion, although again, his
backing was only 36 percent. On leadership qualities, Campbell was regarded as the most
experienced of the candidates (58 percent) and the most honest (46 percent). He did not
electrify party members, but he did not repel them either.
Campbell became leader because the Liberal Democrats were in turmoil after Kennedy’s
departure. The factional disagreements between economic and social liberals had the potential
to become something worse and so a safe pair of hands was needed at this critical juncture.
The requirement for unity was thus real and that helped Campbell. Electoral considerations,
although important, were more of a medium-term problem; if the contest had been decided on
that basis, Campbell could have struggled to defeat Huhne.
The Liberal Democrat Leadership Contest of 2007
Campbell survived less than two years as Liberal Democrat leader. During his time at the
helm, the party’s support in the opinion polls plummeted to below 15 percent. The Liberal
Democrats found themselves squeezed by a resurgent Conservative Party and by Labour
during Gordon Brown’s ‘honeymoon’ period at the start of his leadership in 2007. Campbell’s
9
own performance had been severely criticised. He appeared uncertain and nervous in
parliament. He looked older than his 67 years and suffered in comparisons with the youthful
David Cameron. He had survived most of 2007 only because Brown was expected to call a
general election in the Autumn, but when that plan was abandoned, Campbell found himself
facing the beginnings of a putsch (Daily Telegraph, 14 October 2007). In October 2007, he
decided to resign before he was forced out.
The leadership contest that followed Campbell’s resignation was a mainly sedate affair. Only
two candidates secured the required seven nominations from MPs – Chris Huhne, who had
participated in the 2006 contest, and Nick Clegg. Both candidates were associated with the
‘modernising’ wing of the Liberal Democrats, and shared a broad agreement on policy.
Nevertheless, Huhne was quickly established as the more left-leaning of the two candidates,
with Clegg regarded as slightly more right-leaning (The Independent, 20 October 2007).
Huhne’s perceived position was ironic given that he had been seen as the right-leaning
candidate in the 2006 leadership election. His interest in green politics facilitated this shift in
perceptions, as did his focus on fairness and social justice. Where Clegg emphasised social
mobility, Huhne spoke of the need to achieve equality and end child poverty.
Table 4: Liberal Democrat Leadership Election Result (2007)
Votes
%
Nick Clegg
20,988
50.6
Chris Huhne
20,477
49.4
Total
41,465
100.0
Notes: Single transferable vote electoral system used. Turnout: 64.1%.
Clegg was quickly identified as the front-runner in the contest by journalists (and
bookmakers). He attracted the support of at least 39 of the Liberal Democrats’ 63 MPs,
compared with at least 11 endorsing Huhne (The Guardian, 23 November 2007). One of the
reasons for Huhne’s lower support among MPs was the perception that he had plotted against
Campbell in order to force a contest before Clegg had established himself as a rival (Sunday
10
Times, 21 October 2007). However, Huhne enjoyed greater support among party members. A
YouGov poll for Sky News found Clegg leading Huhne by 56 percent to 44 percent among
party members (for poll details, see ‘source’ for Table 5). The result of the contest turned out
to be closer than anyone anticipated, Huhne finishing just 1.2 percent behind Clegg (Table 4).
Selection Criteria
Clegg’s victory was largely down to the perception that he would boost the Liberal
Democrats’ flagging electoral fortunes. The party had slumped to 11 percent in the polls
under Campbell and there were fears that the Conservatives would make major gains from the
Liberal Democrats at the following general election. Dealing with the Tory threat was thus a
key issue in the leadership contest. Table 5 shows party members’ views of the candidates’
main qualities. There is no data on which candidate was most acceptable to the party and able
to unify it, but it was not especially divided at any rate. Unlike the eviction of Kennedy, the
resignation of Campbell created no great trauma in the party, either among MPs or activists.
Although the two candidates did have some differences over policy, both were modernisers.
The major factor behind Campbell’s resignation was not party unity but electability. Huhne
edged out Clegg on competence, although a big majority of members believed there was not
much difference between them. Clegg was marginally seen as better able to oppose the
Labour prime minister, but he had a much bigger lead on being able to oppose the
Conservative leader. Huhne’s policies were seen as better, although again, almost half of
party members could not discern much difference. However, the most striking finding was on
the candidate judged most voter-friendly, and here Clegg led 53-9. In combination with his
lead on opposing Cameron, it is clear that Clegg won the leadership on the criterion of
electability, especially in relation to the Tory threat. Nevertheless, it is noteworthy that party
members came extremely close to electing Huhne, despite Clegg’s advantage on electability.
Doing so would have put them at loggerheads with Liberal Democrat MPs, who had clearly
identified Clegg as their preferred leader on the basis of his electoral appeal.
11
Table 5: Liberal Democrat Members’ Views of Candidates’ Leadership Qualities (2007)
Candidate who…
Clegg
Huhne
Not much
difference
Has more voter-appeal [electability]
53
9
33
Has a better policy programme
19
28
48
Is more competent [competence]
16
21
60
Would more effectively oppose Gordon
33
31
30
40
26
28
Brown [competence]
Would more effectively oppose David
Cameron [competence/electability]
Source: YouGov, ‘Liberal Democrat Leadership Contest’, poll for Sky News, 3 December
2007, available at <http://www.yougov.com>.
Notes: All figures are percentages. N = 678
The Conservative Leadership Contest of 2001
The Conservative leadership contest of 2001 followed the party’s defeat in that year’s general
election and the subsequent resignation of the incumbent leader, William Hague. It would be
the first to be conducted under a new hybrid selection system in which MPs would whittle
down the field to two candidates, who would then face an OMOV ballot of party members to
decide the winner (Alderman and Carter, 2002; Quinn, 2005: 804-6). Numerous possible
candidates were mentioned in the press, but five formally entered the contest. Michael Portillo
was the early favourite, having returned to the Commons in 1999 after losing his seat two
years earlier. Since then, he had softened his hard-line Thatcherite image, adopting a more
socially liberal approach to such issues as homosexuality and multiculturalism. The other
major candidate to declare was Kenneth Clarke, the Tories’ principal pro-European in a party
that had become increasingly Eurosceptic. Clarke’s campaign would emphasise his electoral
popularity and vast governmental experience. Three other candidates came forward, of whom
the most significant was Iain Duncan Smith, a traditional Thatcherite and social conservative.
He had been a rebel against the Major government over the Maastricht Treaty but had served
in Hague’s shadow cabinet. He hoped to collect the votes of the Tory right. Competing
12
against him for the same constituency was David Davis, another right-wing Eurosceptic.
Finally, Michael Ancram presented himself as a moderately Eurosceptic pragmatist.
Portillo and Clarke were the two ‘big beasts’ in the contest, but each faced major problems.
Clarke had already run for the leadership in 1997, but lost to Hague largely because an
increasingly Eurosceptic parliamentary party would not accept a pro-European as their leader
(Norton, 2002: 72). Clarke’s hope was that his greater electoral appeal would be sufficient to
overcome this opposition. He emphasised his long years in government and his reputation for
economic competence. In contrast, Portillo was aligned with his party on Europe, but his
political ‘reinvention’ after 1997 – and his subsequent public acknowledgement of
homosexual experiences in his youth – created suspicion among social conservatives
(Heppell, 2008: 139-40). He also had a reputation for disloyalty to the two previous Tory
leaders, John Major and Hague (Cowley and Green, 2005: 49). The main beneficiary of the
opposition to Portillo was Duncan Smith, who attracted support largely because he was
‘sound’ on Europe and socially conservative.
Table 6: Conservative Party Leadership Contest of 2001
First MPs’
Re-run first
Second MPs’
ballot
MPs’ ballot
ballot
Votes
%
Votes
%
Votes
Iain Duncan Smith
39
23.5
42
25.3
Kenneth Clarke
36
21.7
39
Michael Portillo
49
29.5
David Davis
21
Michael Ancram
Abstentions/spoilt
Total
OMOV ballot
%
Votes
%
54
32.5 155,933
60.7
23.5
59
35.5 100,864
39.3
50
30.1
53
31.9
-
-
12.7
18
10.8
-
-
-
-
21
12.7
17
10.2
-
-
-
-
0
0
0
0
0
0
166
100.0
166
100.0
166
100.0 256,797
100.0
Notes: Eliminative ballots in parliamentary votes. Turnout in OMOV ballot: 79%.
13
Despite these worries about Portillo, the results of the parliamentary ballots were still a shock.
Portillo finished first in the initial (and then the re-run) ballot, but not by as much as expected.
The elimination of Davis and Ancram saw their votes go to Clarke and Duncan Smith, with
the former topping the final parliamentary ballot and the latter overtaking Portillo to take
second place by one vote (Table 6).
The OMOV ballot was thus to be between Clarke and Duncan Smith. Ideologically, the latter
was much more closely attuned to Tory members, particularly on the all-important issue of
Europe. Clarke tried to make the argument that he was the more experienced candidate and
the more likely to win back Tory defectors who had voted for New Labour. However, he
could not dispel the belief that he was a divisive figure. In the event, the membership’s
suspicion of the pro-European Clarke proved too great and Duncan Smith eased to a
comfortable victory, winning 60.7 percent of the votes to Clarke’s 39.3 percent.
Selection Criteria
The most common interpretation of the 2001 Conservative leadership contest is that an
ideological mass membership put itself at odds with the MPs in choosing an inadequate leader
who reflected their own views over an electorally appealing candidate who did not (e.g.
Heppell, 2008: 151-2; Denham and O’Hara, 2008: 65; cf. Cowley and Green, 2005: 48-50).
However, although ideology was important, its influence should be seen in the context of the
desire for party unity rather than any lack of concern among members with electability. In a
party that had become intensely ideological, only a candidate acceptable to the dominant
coalition could preserve party unity. Clarke’s pro-Europeanism made him unacceptable to
most members and thus unable to maintain intra-party unity. As Norton (2006: 38) argues:
However popular Clarke may have been in the country, he carried too much
political baggage for most Conservative Party members. It was not simply
that they disagreed with him on the issue of European integration, though
they did, but also that they realized that his stance on the issue was bound to
split the party, certainly badly and possibly catastrophically. They were
simply not prepared to take the risk. They voted for Duncan Smith.
Tory members felt that, irrespective of Clarke’s appeal to voters and his undoubted
competence, a Clarke-led Conservative Party would be beset by endless rows over Europe.
14
That in turn would make the party unelectable, as voters dislike divided parties. The pollsters,
ICM, conducted two opinion polls during the OMOV campaign. One was of ordinary voters,
with separate results for voters as a whole and ‘core Tories’. The latter were likely to be the
closest in their views to Tory members. A second poll was of Conservative members,
although the sample was very small (229), it was conducted in four local associations, and the
results were unweighted and not claimed to be fully representative. Some ‘buyer beware’ is
therefore in order when consulting this poll, but it is nevertheless highly suggestive.
Figure 1: Conservative Party Members’ Views of Who Would Split the Party (2001)
%
40
35
30
25
20
15
10
5
0
Very likely
Quite likely
Quite unlikely
Clarke
Very unlikely
Duncan Smith
Source: ICM Research, ‘Conservative Party Membership Poll’, poll for the Sunday
Telegraph, 26 August 2001, available at <http://www.icmresearch.co.uk>.
Notes: Q. ‘Do you think it is very likely, quite likely, quite unlikely or very unlikely that
[name] would split the party if he became leader?’ N = 229
The significance of Europe and the single currency was evident when members ranked it as
the key issue of the campaign and 86 percent regarded it as very or quite important, while 13
percent thought it not very or not at all important. Although the members were not asked in
this poll for their own views on Europe, they were known to be overwhelmingly Eurosceptic
(as polls in 2005 demonstrated – see next section). On that basis, a pro-European would be
expected to split the party. When asked directly whether or not a given candidate would split
the party, 58 percent thought Clarke very/quite likely to do so while 28 percent thought him
very/quite unlikely to do so (net +30). In contrast, only 18 percent believed Duncan Smith
15
very/quite likely to split the party, whereas fully 73 percent regarded him as very/quite
unlikely to do so (net -55). Figure 1 provides the full break down of figures.
It might appear strange that Duncan Smith, a standard-bearer of the Thatcherites with an
uncompromising position on Europe, could be seen as a unifier. However, uniting a party in
which one faction is much stronger than the others requires a different approach from uniting
a party in which the rival factions are evenly balanced. In 1990, John Major was seen as the
candidate best able to hold together the warring pro- and anti-European factions (Alderman
and Carter, 1991). After 1997, however, the Eurosceptics were dominant and by 2001 even
more so. Uniting the Conservatives required recognition that the Eurosceptics had won the
argument inside the party and that required a Eurosceptic leader.
Portillo, meanwhile, failed in his leadership bid largely because he too was deemed unable to
unite the party. Unlike Clarke, Portillo was acceptable to his party on Europe, but it was on
social policy (e.g. drugs, multiculturalism, gay rights), the emerging fissure within the
Conservative Party, that he found himself at odds with his colleagues (Kelly, 2004). In
addition to ideological differences, Portillo was also widely distrusted because of his
perceived disloyalty to previous Tory leaders. One of his most vocal critics in the
parliamentary party was Widdecombe, who accused him and his allies of backbiting and
destabilising the party during Hague’s leadership (Heppell, 2008: 143). This distrust was
shared by party members (Alderman and Carter, 2002: 584).
Preserving party unity became the over-riding selection criterion in the 2001 leadership
contest because the two candidates initially seen as the front-runners, Clarke and Portillo,
were both deemed too divisive (which would have raised serious questions about the
legitimacy of the selection system in party members’ eyes had both reached the OMOV
ballot, as nearly happened). Duncan Smith was far from being the ideal candidate, but he
trumped both his major rivals on the issue of acceptability, which was a more basic priority
than electoral appeal or governmental experience. Consequently, he was the default choice for
party members: had he faced another Eurosceptic in the OMOV ballot, unity would no longer
have been the major criterion, and selectors would have decided on the basis of electability.
It is unlikely that the identity of the selectorate affected the outcome of the contest. Although
Duncan Smith won only a third of MPs’ votes in the parliamentary ballots, Clarke hardly won
16
much more. Duncan Smith also won the crucial backing of the still-revered Lady Thatcher
during the OMOV stage. Even Clarke’s supposedly stronger electoral appeal, which might
have been expected to appeal to MPs, was not as great as he implied. An ICM poll of voters
for the Daily Telegraph in August 2001 found that only 13 percent said a Clarke leadership
would make them more likely to vote Tory, as opposed to 10 percent for Duncan Smith, while
9 percent said Clarke would make them less likely to support the Tories (12 percent for
Duncan Smith). In both cases, two-thirds of voters said it would make no difference.
Similarly, only 9 percent thought the Conservatives would have a good chance of winning the
next election under Clarke, with 8 percent saying the same of Duncan Smith. These findings
emphasise how little the 2001 contest had to do with electoral considerations. Given that fact,
and given the Eurosceptic profile of Conservative MPs (Heppell, 2008: 152), it is doubtful
that Clarke would have defeated Duncan Smith in a parliamentary ballot, particularly as this
method had delivered Hague a comfortable victory over Clarke four years earlier.
The Conservative Leadership Contest of 2005
The Tory leadership contest of 2005 took place in the months following the Conservatives’
general-election defeat in May. Michael Howard, who had replaced Duncan Smith following
the latter’s eviction in 2003 (see Heppell, 2008: 155-70), announced he would be stepping
down in the autumn but first wanted to change the leadership selection rules. The experience
of Duncan Smith, who turned out to be a poor leader, had convinced most Conservative MPs
that the experiment with intra-party democracy had been a mistake. Consequently, for four
months following the general election, no one knew precisely what the rules would be in the
leadership contest, although there was an expectation that MPs would once again make the
decision. However, the move failed to win sufficient support at the party’s constitutional
convention and so OMOV was retained (Denham and O’Hara, 2008: 127-9).
Numerous candidates were touted early on, but eventually four would participate in the
contest. Kenneth Clarke took part in his third leadership contest and again offered experience.
However, his pro-Europeanism would again prove to be a major obstacle to his success.
Meanwhile, David Davis ran on a Thatcherite ticket of tax cuts, Euroscepticism and social
conservatism. Davis was the early favourite and was far ahead of the other candidates in
endorsements from MPs. Liam Fox, a right-winger, appealed to the same constituency as
Davis. Finally, David Cameron, the youngest and least experienced of the candidates,
represented the Conservatives’ ‘modernising’ wing and was widely seen as Howard’s
17
preferred successor (Denham and O’Hara, 2008: 108). Unlike Davis and Fox, Cameron was
socially liberal; unlike Clarke, Cameron accepted the majority Eurosceptic line in the party,
although he gave less impression of being obsessed with the issue.
The highlight of the campaign was at the party conference, where the four leadership
candidates (together with a fifth, Sir Malcolm Rifkind, who had not yet dropped out) each
addressed delegates for twenty minutes. Clarke and Cameron delivered good speeches, which
were well received. Clarke stressed his experience and electability while Cameron delivered
his speech without notes, showcasing his modernising credentials and communication skills.
Fox made a good populist speech, in which he urged the party to stop apologising for the past.
However, Davis performed poorly. Television pictures appeared to show some members of
the audience falling asleep. The conclusion of the speech was so low-key that many people
did not realise it had ended and Davis had to motion to them to applaud (Montgomerie, 2005).
The immediate consequence of the speeches was a dramatic change in the candidates’ poll
ratings among Tory members. In a YouGov poll a week before the conference, Davis and
Clarke were tied for first place on 30 percent, with Cameron in third place on 16 percent
(Davis led Clarke 48-44 head-to-head). But another YouGov poll taken straight after the
conference found Cameron on 39 percent, Clarke on 26 percent and Davis on 14 percent.
Cameron’s successful conference gave him momentum to take into the parliamentary ballots.
The first ballot ended in the predicted victory for Davis, but his 62 votes were less than
expected. In second place was Cameron, just six votes behind. Fox edged out Clarke by four
votes to go through to the second ballot, to be held two days later (Table 7). By that time,
Cameron was firmly established as the frontrunner, rallying to his cause most of those MPs
who voted for now eliminated Clarke in the first ballot. Cameron duly won the second ballot
with 90 votes, far ahead of Davis who dropped five votes to finish on 57, just six ahead of
Fox. Cameron and Davis thus went through to the OMOV ballot, which lasted from October
to December. Cameron was by now firmly established as the favourite, with a two-to-one lead
over Davis in polls of party members. The OMOV campaign, consequently, was a sedate
affair with little change in the candidates’ poll ratings. Cameron secured a convincing victory,
taking the votes of 67.6 percent of Conservative members, to Davis’s 32.4 percent.
18
Table 7: Conservative Party Leadership Contest of 2005
First MPs’ ballot
Second MPs’ ballot
OMOV ballot
Votes
%
Votes
%
Votes
%
David Cameron
56
28.3
90
45.5
134,446
67.6
David Davis
62
31.3
57
28.8
64,398
32.4
Liam Fox
42
21.2
51
25.7
-
-
Kenneth Clarke
38
19.2
-
-
-
-
Abstained/spoilt
0
0
0
0
198
100.0
198
100.0
198,844
100.0
Total
Notes: Eliminative ballots in parliamentary votes. Turnout in OMOV ballot: 77%.
Selection Criteria
The lesson of 2001 was that only a Eurosceptic leader would be acceptable to the Tories,
given the ideological profile of MPs and members alike. Clarke remained the only proEuropean with a realistic prospect of running for the leadership and that owed everything to
his perceived electability. However, as a staunch Eurosceptic, Davis was more ideologically
acceptable to the party and his participation in the 2001 contest had established him as an
important figure in the party. Going into the conference, Davis was the media’s frontrunner,
with Clarke second. Aside from Clarke, other potential candidates were regarded mainly as
also-rans, with even Cameron’s candidacy seen in terms of setting down a marker for the
future (Denham and Dorey, 2006: 36).
There were, nevertheless, doubts about Davis. The nature of those doubts was evident in a
YouGov poll on the eve of the party conference (28-30 September). At this point, Davis and
Clarke were seen as the two main candidates and YouGov asked Conservative Party members
to assess them on five criteria that translate well into the acceptability-electability-competence
criteria set out earlier (see Table 8). Clarke led Davis on electability (boosting the Tories’
election chances (9 points) and performing well on television (21 points)) and by 31 points on
competence (challenging Blair and Brown in parliament).
19
Clarke’s problem was that he failed to convince his party on the first-order criterion of
acceptability: his pro-European views were seen as a major impediment to uniting the party.
On this vital question, Davis led Clarke by 19 points. Davis found himself in a similar
position to Duncan Smith four years earlier: he was seen as less voter-friendly or competent
than Clarke, but he became the early default choice of the party because Clarke’s proEuropeanism was unacceptable. Only the memories of Duncan Smith enabled Clarke to
remain competitive going into the 2005 conference.
The existing misgivings about Davis were confirmed by his performance at the conference,
particularly in relation to electability and his ability to communicate effectively with voters.
However, it was only with the emergence of Cameron as a credible candidate that Davis’s fate
was sealed. A poll after the elimination of Clarke gave Cameron absolute majorities on all the
key leadership qualities, even after the inclusion of Fox (see 18-19 October poll in Table 8).
Cameron’s biggest leads were on electability criteria, viz. performance on television and
boosting the Tories’ election chances. Most important, he led Davis in the one category that
Davis had led Clarke, on being best placed to unite the party. Davis had acquired a reputation
for plotting against Hague and Duncan Smith, and these suspicions about his loyalty, rather
than ideology, created doubts about his ability to unite the party (Heppell, 2008: 189-90).
Cameron, like Clarke, enjoyed a strong lead on important leadership attributes, but without
being burdened by opinions on Europe that were unacceptable to his party. His speech, and its
favourable coverage, demonstrated that he could communicate effectively with voters.
20
Table 8: Conservative Members’ Perceptions of Candidates’ Leadership Qualities (2005)
28-30 Sept Poll
18-19 Oct Poll
9-11 Nov Poll
Candidate best placed to:
Davis
Clarke
Davis
Cameron
Fox
Davis
Cameron
Unite Conservative Party
53
34
19
55
15
27
59
39
48
13
63
15
21
64
34
55
8
70
17
17
71
31
62
20
52
16
35
50
n/a
n/a
17
56
19
32
53
[acceptability]
Boost Tories’ election chances
[electability]
Perform well on TV & radio
[electability]
Challenge Blair & Brown in
Commons [competence]
Make a good Prime Minister
[competence]
Sources: YouGov, polls for the Daily Telegraph: ‘Conservative Members Survey 2’, 3 October 2005; ‘Conservative Leader Election’, 20
October 2005; ‘Conservative Leadership Part 5’, 14 November 2005, available at <http://www.yougov.com>.
Notes: All figures are percentages. N = 615 (Sept); N = 665 (Oct); N = 748 (Nov)
21
Table 9: Ideological Stance of Members and their Perceptions of Final Three Candidates
Tory members’
Tory members’ descriptions of…
self-description
Cameron
Davis
Fox
7
9
3
2
Eurosceptic
80
40
65
72
Neither
13
36
19
14
1
15
13
12
‘One Nation’ Centre
45
57
20
9
‘Clear Blue Water’ Right
48
15
45
53
Neither
n/a
21
28
31
7
7
7
6
Europe
Pro-European
Don’t know
Electoral Strategy
Don’t know
Source: YouGov, ‘Conservative Leader Election’, poll for the Daily Telegraph, 20 October
2005, available at <http://www.yougov.com>.
Notes: All figures are percentages. N = 665
These impressions find support from polling evidence. Table 9 shows the results of a YouGov
poll of party members’ ideological stances and their perceptions of the candidates’ stances,
carried out after Clarke’s elimination. On Europe, 80 percent of Tory members described
themselves as Eurosceptic, while just 7 percent described themselves as pro-European, 13
percent saying they were neither. The members saw both Davis and Fox as being in tune with
them. Only 40 percent believed Cameron was a Eurosceptic, while 36 percent said he was
neither Eurosceptic nor pro-European. Just 9 percent saw him as pro-European, so his image
among party members was not that of a Europhile, but a pragmatic, moderate Eurosceptic.
On electoral strategy, the members were almost evenly divided between ‘one nation’ centrists
and ‘clear blue water’ right-wingers. Both Fox and Davis were regarded as ‘clear blue water’
Tories, but Cameron was seen as a ‘one nation’ Conservative. Ideology was thus important in
22
that MPs and activists had serious problems accepting a pro-European, but hard-line
Euroscepticism was not essential. Similarly, a ‘one nation’ strategy was seen as preferable to
a ‘clear blue water’ strategy, regardless of activists’ personal beliefs.
The Conservatives were not as divided in 2005 as they had been for much of the post-1997
period. However, party unity remained an important consideration as long as Clarke was still
in the contest because there was the risk that the party would be ungovernable with a proEuropean as leader. Once Clarke had been eliminated, unity was no longer a pressing concern.
The contest could then be decided on the second-order criterion of electability, which it was.
Conclusion
Although only four leadership contests were considered in this paper, some tentative
conclusions can be drawn. First, and most important, the perceived unifying candidate
normally wins in divided parties. That was most strongly the case in the Conservative Party in
2001, when Duncan Smith easily defeated Clarke despite being weaker on the criteria of
electability and competence. The conventional wisdom that Tory members were irrational to
choose Duncan Smith overlooks the basic requirement for party unity as a prerequisite for
electoral success. A pro-European would have split the party and left it vulnerable to
debilitating attacks from the Labour government and the media. Deep down, Conservative
members probably knew that they had little chance of winning the next general election, but
their more immediate aim was to preserve party unity. Therefore, they took a rational decision
in choosing Duncan Smith, irrespective of his obvious weaknesses.
In this respect, Conservative members in 2001 replicated the behaviour of Conservative MPs
in 1997, when the latter group chose William Hague over Clarke, again for reasons of
ideological acceptability and party unity. It is therefore ironic that Tory members were much
maligned for their choice: they were accused of choosing a candidate with limited support in
the parliamentary party because only 54 of 166 Tory MPs made Duncan Smith their first
choice in the preliminary ballots (Heppell, 2008: 148-9). However, Clarke had only 59 votes,
and so he too would have faced questions over his legitimacy had he won the OMOV ballot.
Many Tory MPs were sufficiently unimpressed with the 2001 vote to seek to change the rules
and revert to MP ballots, a move narrowly defeated. Yet those same Conservative members
went on to choose Cameron over Davis in 2005, confounding the notion that they were blind
ideologues. They were ideological, but not blindly so: they had a sophisticated understanding
23
of the necessity for internal unity before electoral considerations. Once they were offered
candidates who did not threaten the broad Eurosceptic consensus in the party in 2005, they
chose on the basis of electability. If a candidate such as Cameron had been offered to them in
2001, it is highly liked they would have chosen him, certainly over Clarke and probably over
Duncan Smith. Despite being a Eurosceptic, Portillo was not such a candidate, because he
divided his party on social policy, homosexuality and (lack of) loyalty.
Unity was also a consideration in the Liberal Democrat leadership contest of 2006. Campbell
was the choice of both MPs and individual members to take over from Kennedy because he
was acceptable to both major factions in the party. Hughes might have been expected to be
the choice of the left-leaning members, but he was unacceptable to most MPs, especially the
economic liberals. The members were more amenable to Huhne, but they opted for Campbell
partly in order to avoid further discord within the party – Campbell was seen as a safe pair of
hands and his selection would postpone any conflict between the economic and social liberals.
Again, party members had an acute sense of the need for party unity. The notion that they act
significantly differently than MPs do in choosing leaders cannot be sustained.
A second conclusion to draw from these contests is that when party unity is not pressing,
members tend to opt for leaders whom they rate highly on electability. Cameron in 2005 and
Clegg in 2007 were both examples of this tendency. In Cameron’s case, his opponent, Davis,
was a Eurosceptic, and so both were ideologically acceptable – hence, the decision was made
on electability. Clegg’s victory over Huhne owed a lot to his greater electable appeal,
something of huge significance after the dramatic drop in the Liberal Democrats’ poll ratings
during Campbell’s tenure. The members liked Huhne’s policies more, but they narrowly
decided that an improvement in the party’s electoral performance was more important.
OMOV ballots in Britain have not led to the election of radical leaders, or at least, no more
radical than MPs would likely have chosen. This point largely follows from the members’
concern with party unity. However, the question arises as to what would happen if a party’s
MPs and members had different ideological profiles. Would they be likely to interpret
ideological ‘acceptability’ in different ways? If so, then a split between MPs and members
could become inevitable. It is noteworthy that although Liberal Democrat members chose
Clegg over Huhne in 2007, they did so by just 1.2 percent. If 256 members had switched their
votes, Huhne would have become leader despite Clegg being the overwhelming choice of
24
MPs. As it was, there was not a huge ideological difference between the candidates, but there
could be in a future contest. That suggests that some caution is required before claiming that
OMOV ballots have changed little in the way of candidate selection in British parties.
References
ALDERMAN, KEITH. 1999. Revision of leadership election procedures in the Conservative
Party. Parliamentary Affairs 52 (2): 260-74.
ALDERMAN, KEITH and NEIL CARTER. 1991. A very Tory coup: The ousting of Mrs
Thatcher. Parliamentary Affairs 44 (2): 125-39.
ALDERMAN, KEITH and NEIL CARTER. 2002. The Conservative Party leadership
election of 2001. Parliamentary Affairs 55 (4): 569-85.
COWLEY, PHILIP and JANE GREEN. 2005. New leaders, same problems: The
Conservatives, in Andrew Geddes and Jonathan Tonge (eds), Britain Decides: The UK
General Election 2005. Basingstoke: Palgrave Macmillan, pp. 46-68.
DENHAM, ANDREW and PETER DOREY. 2006. The ‘caretaker’ cleans up: The Liberal
Democrat leadership election of 2006. Parliamentary Affairs 60 (1): 26-45.
DENHAM, ANDREW and PETER DOREY. 2007. A tale of two speeches: The Conservative
leadership election. Political Quarterly 78 (1): 35-41.
DENHAM, ANDREW and KIERON O’HARA. 2008. Democratising Conservative
Leadership Selection: From Grey Suits to Grass Roots. Manchester: Manchester
University Press.
HEPPELL, TIMOTHY. 2008. Choosing the Tory Leader: Conservative Party Leadership
Elections from Heath to Cameron. London: Tauris Academic, 2008.
HURST, GREG. 2006. Charles Kennedy: A Tragic Flaw. London: Politico’s.
KELLY, RICHARD. 2004. The extra-parliamentary Tory Party: McKenzie revisited.
Political Quarterly 75 (4): 398-404.
MARSHALL, PAUL and DAVID LAWS (eds). 2004. The Orange Book: Reclaiming
Liberalism. London: Profile Books.
MAY, JOHN. 1973. Opinion structure of political parties: The special law of curvilinear
disparity. Political Studies 21 (2): 135-51.
25
MONTGOMERIE, TIM. (2005). How Cameron won… and Davis lost, 6 December, available
at <http://conservativehome.blogs.com/toryleadership/2005/12/how_cameron_won.html>.
NORTON, PHILIP. 2002. The Conservative Party: Is there anyone out there?, Anthony King
(ed.), Britain at the Polls 2001. New York/London: Chatham House Publishers, pp. 68-94.
NORTON, PHILIP. 2006. The Conservative Party: The politics of panic, in John Bartle and
Anthony King (eds), Britain at the Polls 2005. Washington DC: CQ Press, pp. 31-53.
PUNNETT, R. M. 1992. Selecting the Party Leader: Britain in Comparative Perspective.
Hemel Hempstead: Harvester Wheatsheaf.
QUINN, THOMAS. 2004. Electing the leader: The Labour Party’s electoral college. British
Journal of Politics and International Relations 6 (3): 333-52.
QUINN, THOMAS. 2005. Leasehold or freehold? Leader-eviction rules in the British
Conservative and Labour parties. Political Studies 53 (4): 793-815.
RUSSELL, ANDREW. 2005. The Liberal Democrat campaign, in Pippa Norris and
Christopher Wlezien (eds), Britain Votes 2005. Oxford: Oxford University Press, pp. 87100.
SJOBLOM, GUNNAR. 1968. Party Strategies in a Multiparty System. Lund:
Studentlitteratur.
STARK, LEONARD. 1996. Choosing a Leader: Party Leadership Contests in Britain from
Macmillan to Blair. Basingstoke: Macmillan.
WHITELEY, PAUL, PATRICK SEYD and ANTONY BILLINGHURST. 2006. Third Force
Politics: Liberal Democrats at the Grassroots. Oxford: Oxford University Press.
26