Introduction

It is now over a decade since the curtain came down on the Prime Ministerial tenure of John Major and he decided to leave the political stage. By his immediate resignation, Major plunged the Conservative Party into a highly personalised, ideologically divisive and traumatic leadership election. Conservative parliamentarians then rejected the following political and electoral indicators: the most experienced candidate available was the erstwhile Chancellor, Kenneth Clarke; the candidate most popular with the electorate was Clarke; the candidate most popular with the membership of the Conservative Party was Clarke; and the candidate who the Labour Party most feared was Clarke.¹ However, the war of the Major succession was to be won by William Hague, an inexperienced political lightweight and moderate economic dry Eurosceptic, as compared to the experienced heavyweight, economic wet Europhile, Clarke.² His ascent was a truly remarkable event. First, he would not have stood had Michael Portillo (his own first preference) been available. Second, he was a deeply reluctant candidate who had to be persuaded not to endorse Michael Howard (his second preference) and stand himself. Third, he was assumed to have been elected at the behest of the Thatcherite Eurosceptic right, but he only secured the endorsement of Thatcher when all other non-Clarke options had been eliminated. By selecting Hague over Clarke, the Conservatives prolonged their adherence to an outmoded model of statecraft (i.e. Thatcherism), and in doing so, they delayed the necessary adaptations to the Conservative narrative. Therefore, understanding the final ballot of the Conservative Party leadership election in the war of the Major succession is essential to understanding the evolution of Conservative party politics in the era of New Labour hegemony.

So how did Hague do it? Or why did the parliamentary Conservative Party (PCP) do it? Our motivation to answer these dilemmas stems from the fact that the most significant article on Hague's elevation to the party leadership to date, by Alderman, is an anecdotal narrative rather than a systematic evaluation of voting behaviour. Having offered detailed appraisals of the background to the contest, the candidates and the multiple ballots, Alderman (1998, 1–16) concludes that Hague was a surprise victor in a debilitating and divisive contest. As such, Alderman does not offer (nor does he attempt to) an explanation of voting motivations of Conservative parliamentarians, covering the centrality of political characteristics as well as ideological disposition. Although two short non-peer review papers that do make brief attempts to offer a systematic appraisal of this issue have also been published, by Cowley (1997) and Norton (1998), this paper differs from these in two substantive ways. First, Cowley and Norton do not systematically assess the relevance of voting behaviour for candidates in relation to the political characteristics of Hague and Clarke vis-à-vis their electorate (i.e. Conservative parliamentarians); and second, while both systematically (albeit briefly) assess the significance of ideology as a voting determinant, this paper challenges their findings in this area. Our challenge embraces both how they define Conservative parliamentarians in ideological terms, and how their papers concentrate exclusively on the Eurosceptic ideological divide, to the exclusion of the economic and morality ideological divides (Cowley, 1997, 91–95; Norton, 1998, 10–14).

Therefore, this paper seeks to systematically re-examine the Hague–Clarke final ballot face off and to explain the how and why of this event. In doing so, the paper will be broken down into three substantive sections. First, to provide a context to the subsequent evaluation, the paper will begin by providing a historical narrative of the succession campaign and the multiple ballots. Second, the paper will determine the voting behaviour of the PCP in the final parliamentary ballot between Hague and Clarke. Finally, having established a data set of voting behaviour, the paper will seek to analyse the voting motivations of Conservative Party MPs. The paper will seek to test six hypotheses, based around political characteristics and ideological disposition, of what motivated this behaviour. The hypotheses are that the key factors involved were:

1. The age of the voters (i.e. Conservative parliamentarians) vis-à-vis the age of Hague and Clarke; namely, that younger voters would coalesce around Hague and older voters around Clarke. Aligned to this assumption is the expectation that recently elected Conservative parliamentarians would gravitate to Hague and long serving Conservative parliamentarians would endorse Clarke. If so, this would obviate the common assumption that 'young cardinals vote for old popes'.
2. The career status of Conservative parliamentarians. Political insiders who had recently held government office (the possessed) would be predisposed to Clarke as the establishment insider; while political outsiders (the never possessed) would tend towards Hague, who had only recently acquired high office and who was more identifiable with the mentality of the political outsider.
3. The electoral vulnerability of Conservative parliamentarians. Here, it is assumed that Conservative parliamentarians holding marginal constituencies would be inclined towards Clarke given the polling evidence that suggested he would be a more popular candidate with the electorate. Given that Hague was deemed to be inferior to Clarke in terms of electoral appeal, it is assumed that Hague would acquire his support base from Conservative parliamentarians holding safer constituencies.
4. Attitudes towards social, sexual and morality-based issues. Here we hypothecate that Clarke's long-standing association with the left of the Conservative Party would mean that social liberals would be inclined towards him. Conversely, we assumed that because Hague was a social conservative,³ this would aid his capacity to secure endorsements from the socially conservative wing of the PCP.
5. Attitudes towards the economic policy ideological divide. We assumed that Clarke, as one of the leading economic wets during the Thatcher/Major era, would secure support from the economically wet wing of the PCP. It is assumed that given his associations with the economically dry wing of the PCP, economic dries would endorse Hague.
6. Attitudes towards the European policy ideological divide. Here we assumed that pro-European Conservative parliamentarians would endorse Clarke, as the leading exponent of Europhilia, and that Eurosceptics would gravitate towards Hague.

The fact that Hague was not expected to emerge as the new party leader in succession to Major provides the rationale for these six hypotheses.⁴ Relative to Clarke the candidature of Hague had three obvious political limitations: first, his age and parliamentary inexperience; second, his limited ministerial and Cabinet experience, as well as his peripheral status; and third, the fact that he was less likely than Clarke to improve the Conservatives' position in the opinion polls. On the ideological factors, the standard interpretation is that an essentially Thatcherite PCP (i.e. socially conservative, economic dry and Eurosceptic) would not permit a non-Thatcherite (i.e. socially liberal, economic damp Europhile) in the shape of Clarke to assume the party leadership. This provides the basis for the three ideologically motivated hypotheses on the social, sexual and morality ideological divide; on the economic policy ideological divide; and finally, the European policy ideological divide. The justification for these hypotheses becomes more evident when we offer a historical narrative of the party leadership campaign and explain how the ballots evolved.

Determining the Succession: The Leadership Election Campaign

The dynamics of the war of the Major succession were fundamentally altered by two key events. First was the political demise of Michael Portillo. The anointed candidate of the Eurosceptic Thatcherite right proved to be the iconographic electoral casualty of the electoral meltdown in May 1997. Second, Michael Heseltine was admitted to hospital having suffered an angina attack in the immediate aftermath of the general election defeat. He immediately abandoned any plans that he had to stand for the leadership of the Conservative Party.

The enforced absence of the prime candidate of the Thatcherite Eurosceptic right (Portillo) and the standard bearer of the Tory Europhile left (Heseltine), combined with the fact that the incumbent leader had resigned, created a large field of candidates. By the deadline for nominations there were five candidates for the party leadership.⁵ First, there was Kenneth Clarke, who had been a minister since 1979 and a Cabinet minister since 1985, during which time he had served as Home Secretary and Chancellor. Clarke was the sole candidate of the Tory left: a social liberal, economic wet Europhile. Second, there was the pure Thatcherite (socially conservative, economic dry, Eurosceptic), Michael Howard, who had gained notoriety as an uncompromising Home Secretary between 1993 and 1997. The third and fourth candidates for the succession, Peter Lilley and John Redwood, were also strongly aligned to the Thatcherite cause. Lilley had served as Trade and Industry and then Social Security Secretary. Redwood had served as Welsh Secretary before resigning to challenge John Major for the party leadership in 1995. The fifth and final candidate was William Hague. Hague was considerably younger and less experienced than the four other candidates, with only 4 years worth of ministerial experience, of which only two were in the Cabinet. He had to be persuaded to enter the succession contest after initially intending to endorse the candidature of Howard. Although Hague was aligned to the Thatcherite Eurosceptics he was less identifiable with the cause than Howard, Lilley and Redwood (Crowson, 2007, 64).

Of the five candidates Clarke was the most intriguing. He was the most experienced and had developed a reputation for being a competent and capable Cabinet minister and parliamentarian. Despite the fact that Clarke could offer experience, competence and the greatest electoral appeal, Conservatives agonised over whether he would be a divisive party leader. His committed pro-Europeanism would make it difficult for him to manage a PCP that was (or was at least viewed as being) overwhelmingly Eurosceptic. Given this, it was assumed that the Eurosceptic Thatcherite right might be inclined towards any candidate other than Clarke. The question was which one?

Prior to his entry into the succession battle, Hague had assumed that Howard would defeat Lilley and Redwood to win the primary battle to be the Thatcherite Eurosceptic right candidate for the final ballot, thus leading to a Howard/Clarke final ballot face off for the ideological future of the Conservative Party. Hague made this initial assumption based on the following factors. First, he felt that Redwood was a divisive figure,⁶ and believed that Lilley lacked leadership credibility.⁷ Second, he recognised that Howard had established a strong reputation among Thatcherite social conservatives during his tenure as Home Secretary. Finally, he assumed that Howard was being backed by Thatcher, albeit discreetly (Nadler, 2001, 9–15; Crick, 2005, 368–370).

The viability of the Howard candidature was based on establishing a bandwagon effect and making himself the preferred candidate of the Eurosceptic Thatcherite right and the candidate best placed to prevent Clarke from acquiring the party leadership. A key plank of the bandwagon strategy involved securing the endorsement of Hague. Howard offered Hague the deputy leadership and the position of Party Chairman. Much to Howard's disappointment, however, Hague dramatically changed his mind. He abandoned the arrangement after initially agreeing to support Howard. Hague had eventually calculated that Howard could not defeat Clarke but that he, himself, could (Nadler, 2001, 9–15). The appeal of a Hague candidature reflected the fact that many Conservative parliamentarians were attracted to the notion of skipping a political generation and making a clean and symbolic break with the Thatcher and Major eras, of which other candidates such as Howard, Lilley, Redwood and Clarke were clearly associated (Alderman, 1998, 5).

The possibility that a gravitational pull towards Hague was occurring, and that the Howard campaign had imploded, was confirmed with the results of the first ballot, which caused both Howard and Lilley to withdraw. Although Clarke had come first in the ballot, it was also felt that he needed to have secured a bigger lead over Hague than the eight vote margin that he had attained. The elimination of Howard and Lilley freed up 47 votes as the three remaining candidates entered the second ballot. Given that those 47 votes had been built around the Eurosceptic Thatcherite credentials of Howard and Lilley, many questioned whether Clarke would be able to penetrate into this with sufficient depth, over the next two ballots, to retain his lead over Hague. With the majority of Conservative parliamentarians wedded to Eurosceptism the psychological winner of the first ballot was Hague, who was the most moderate of the four Eurosceptics on offer (ibid., 10). This was confirmed by the outcome of the second ballot. Forty-seven votes were freed by the elimination of Lilley and Howard: Hague's vote increased by 21, with Clarke increasing his vote by 15. The failure of the Redwood candidature was also apparent: although 47 presumed Eurosceptic votes were available, Redwood only secured 11 further votes. Clarke and Hague then proceeded to the final ballot. Both candidates needed to obtain a majority of the 38 votes that Redwood had secured in the second ballot. Given the Eurosceptic basis of the Redwood candidature, it was assumed that this favoured William Hague.

In anticipation of the crucial third ballot, Clarke attempted to formulate a tactical alliance with Redwood which would provide him with the majority of these 38 votes. If Redwood's vote base was relatively cohesive then he had it in his power to be king-maker and hand the party leadership to the leading exponent of Europhilia. In ideological and policy terms, however, the alliance seemed implausible, and indeed the single European currency, the issue that divided them most, was fudged; the pair agreeing that the direction of European policy was no longer within their sphere of influence given that they were in opposition. In any case, they argued, British participation within the single European currency was unlikely in the short term, and should it arise in the future then they would offer a free vote on the issue. The political necessity of this hastily arranged tactical alliance was obvious from the perspective of Clarke, although its attractiveness to Redwood was open to speculation.⁸ One assumption was that if Clarke should emerge victorious, then Redwood would be rewarded with the position of Shadow Chancellor (Williams, 1998, 218–221) (Table 1).⁹

Table 1 - The parliamentary ballots.

Full table

The majority of the Eurosceptic Thatcherite right were contemptuous of the Clarke–Redwood alliance. Redwood was derided as a careerist politician who had abandoned his principles for personal advancement. The intervention of Thatcher extinguished any possibility that the Clarke–Redwood pact could be the solution to the ideological conflict that had bedevilled post-Thatcherite Conservatism. The prospect of a Europhile figure of the Tory left in the shape of Clarke acquiring the party leadership was sufficient for Thatcher to abandon her stance of public neutrality. She instructed her ideological bedfellows to endorse Hague (Campbell, 2003, 788).

The ensuing third and final ballot confirmed the failure of the Clarke–Redwood alliance: Hague secured a comfortable victory as the Redwoodites mostly switched their allegiance to him rather than Clarke. Hague was elected as the new leader of the Conservative Party with 93 votes, a majority of 22.

Few Conservative parliamentarians from the 1992 to 1997 PCP would have anticipated that Major's successor would be Hague (Brandreth, 1999, 488). He was a reluctant candidate whose candidature emerged due to the elimination of his preferred choice (Portillo) and his sense that his second choice (Howard) would not be able to win. So how did he do it? The remainder of this paper seeks to explain how Hague triumphed by addressing two issues: first, who was responsible (i.e. who voted for Hague and who did not and instead endorsed Clarke); and, second, why did they vote in the way that they did: what variables explained their voting patterns — that is, political characteristics such as age and experience, career status, electoral vulnerability and ideological factors such as the social, sexuality and morality ideological divide, the economic ideological divide, and the European ideological divide.

Determining Voting Behaviour: Who Voted for each Candidate?

The viability of this exercise is entirely dependent upon producing an accurate data set of voting behaviour by Conservative parliamentarians in the final ballot between Hague and Clarke. Given the anonymity of a secret ballot this was a significant methodological challenge. We circumnavigated this methodological hurdle by the following means. First, we gathered together a range of newspapers (The Times, Daily Telegraph, The Guardian, The Independent) on the day of the final ballot and the day after the ballot. These provided a raft of information on the voting patterns of Conservative parliamentarians, which could then be cross-referenced. Second, we were able to supplement this with an analysis of published interviews with Conservative parliamentarians, and with a range of secondary material referring to the Hague–Clarke final ballot. Finally, we wrote to Conservative parliamentarians of whom we had little firm evidence, which helped to plug some of the gaps in our data set (Tables 2, 3 and 4).

Table 2 - Voted for William Hague.

Full table

Table 3 - Voted for Kenneth Clarke.

Full table

Table 4 - Abstainers and unclassified.

Full table

Through this we were able to place each Conservative parliamentarian into one of three classifications: definite, probable and unclassified. Conservative parliamentarians were placed in the definite classification if they had been identified as voting the same way in three or more separate sources. The probable classification includes Conservative parliamentarians identified by one or two different sources. We had one caveat to our one source rule: any Conservative parliamentarian who replied to us and informed us of their voting behaviour, and of whom we had no other data, was automatically defined as a definite. The unclassified classification contains Conservative parliamentarians about whose intentions nothing has been discovered or about whom there is contradictory data.¹⁰ These unclassified Conservative parliamentarians have been omitted from this analysis and there has been no attempt to deduce for whom they were likely to have voted.¹¹

The critical issue was that the unclassified grouping needed to be very small, and we needed to ensure that the probable classification was not too large. Our data set had definite information on the voting behaviour of 60% of Conservative parliamentarians, which, when the probable voting behaviour of Conservative parliamentarians was added, thus gave us a good indication of the voting behaviour of nearly 90% of the PCP. Eventually, we excluded only nine Conservative parliamentarians, whose voting behaviour we were unable to ascertain.

Having constructed our data set, we began testing our six motivational hypotheses for voting behaviour: that is, the political characteristics of age and experience, career status, electoral vulnerability, and the ideological factors, such as the social, sexuality and morality ideological divide; the economic ideological divide and the European ideological divide. When we tested our data set findings against these variables we utilised bivariate analysis by using the chi-square test of association.

Determining Voting Motivations: Political Characteristics

The political characteristics hypothesis demanded that we assess the validity of the following. First, was there evidence to suggest that younger and recently elected Conservatives coalesced around Hague, and that older and more experienced Conservative parliamentarians gravitated to Clarke? Second, did insider status contribute to voting for Clarke and did outsider status cause Conservative parliamentarians to gravitate towards Hague? Finally, did Clarke garner support from Conservative parliamentarians holding marginal constituencies because he was the more popular candidate with the wider electorate?

The age and experience hypothesis

The age and experience hypothesis assumes that Conservative parliamentarians will vote for candidates with similar political characteristics to their own. At the time of the leadership election, Clarke was 57 years old and had been in parliament for 27 years. If the hypothesis were proved Clarke would have benefited from being a member of the largest age group in the PCP, as indicated in Table 5. At 36, Hague was much younger than Clarke, and he was also a member of the smallest age group within the PCP. However, with only 8 years in parliament, his relative inexperience could have been beneficial as 69 of his colleagues had less than 9 years service. Clarke was also closest to the median Conservative parliamentarian who was aged 50.

Table 5 - The age and parliamentary experience of conservative parliamentarians 1997.

Full table

The median age of Clarke supporters was 53, while the median age of Hague supporters was 45, as indicated in Table 6. This demonstrates that Clarke was more popular among older Conservative parliamentarians. He scored more votes than Hague in the 50–59, and over 60 age brackets. Conversely, younger Conservative parliamentarians (those aged between 30 and 49) were more likely to vote for Hague.

Table 6 - Age and voting behaviour.

Full table

Analysis of these figures shows that the relationship between the ages of two candidates and the ages of their electorate may have played a role in the voting behaviour of Conservative parliamentarians during the party leadership election. Therefore, the hypothesis is supported: ²=16.124, d.f.=3, P=0.001.

How does the length of parliamentary service fit into this analysis? In this case we hypothecated that the more inexperienced Conservative parliamentarians, with the shortest length of service, would back Hague (8 years in parliament), while the more experienced Conservative parliamentarians would vote for Clarke (27 years in parliament). The effects of the 1997 election defeat and the unusually high number of retiring Conservative parliamentarians is demonstrated in Tables 5 and 6. The median Conservative parliamentarian had been in parliament for 10 years: the median Hague supporter had 9 years service, while the median Clarke voter had 14 years experience. Table 7 shows that Hague attracted a majority among Conservative parliamentarians with 0–9, and 10–19 years service, while Clarke was more popular with the most experienced Conservative parliamentarians, who had more than 20 years service.

Table 7 - Parliamentary experience and voting behaviour.

Full table

Analysis of the variables also supports the hypothesis that there is a relationship between years of service and voting behaviour. Less experienced Conservative parliamentarians tended to vote for Hague, while their longer serving colleagues tended towards Clarke. Therefore, the hypothesis is supported: ²=15.062, d.f.=3, P=0.002.

The insider/outsider hypothesis

Another aspect of the political characteristics hypothesis is government experience, which alludes to the insiders/outsiders dichotomy. Political insiders were the Conservatives' frontbench representatives, who have held government office; while political outsiders represented those who had yet to begin climbing this particular career ladder, many of whom recognised that they were now never going to progress beyond the position of backbencher. We had hypothecated that Clarke, as a senior figure with 12 years of continuous Cabinet service, would be able to garner a strong level of support from Conservative parliamentarians who also had government experience. If Clarke was the candidate of the political insider, we assumed that Hague, who had only 2 years worth of Cabinet experience, would draw more support from political outsiders within the PCP.

Applying this insider/outsider distinction to the PCP, we can see that it was relatively evenly divided between those who had ministerial experience (70) and those who had no ministerial experience (82). Of the 70 with ministerial experience, the split between Clarke at 37 and Hague at 33 was not as pronounced as the hypothesis had assumed, although Hague was able to secure a lead over Clarke among those without ministerial experience, securing 50 of the available 82 political outsiders votes, as demonstrated in Table 8.

Table 8 - All ministerial experience and voting behaviour.

Full table

Just over half of all Conservative parliamentarians who had previously been ministers in their parliamentary careers voted for Clarke, whereas support for Hague came from Conservative parliamentarians who had no ministerial experience. However, the analysis demonstrates that this relationship is not statistically significant — ²=2.915, d.f.=1, P=0.088 — and therefore the hypothesis is not supported.

The insiders/outsiders dichotomy can be reframed however. Defining political insiders as those who have ministerial experience may be misplaced. Some of those with ministerial experience may have been removed from their ministerial responsibilities against their wishes and may be resentful at their demotion (those who Major famously defined as the dispossessed). These Conservative parliamentarians may adopt the mindset of political outsiders despite their aforementioned insider status. This may have skewed the findings. Therefore to overcome this potential distortion, we can reframe the insider/outsider status by limiting insider status to those Conservative parliamentarians who held ministerial office when the Major government was removed from office. This reframes political outsiders as those who have never held office, plus those who have previously held office but did not have ministerial positions at the end of the Major era, due to their own voluntary resignations or because they had been dismissed. Those in the dispossessed categorisation may have been dispossessed for a considerable length of time — that is, they may have been dismissed early in the Major era, or even in the Thatcher era.

As the Conservatives entered opposition, they had 43 parliamentarians who had lost their ministerial positions when New Labour were swept to office (i.e. political insiders), alongside 122 who were either backbenchers prior to the general election, or retread parliamentarians, or new parliamentarians (i.e. political outsiders). Here we were hypothecating that outsiders may have blamed Clarke for their marginal status, or believed that their career progression would be further stymied under a future Clarke leadership and would therefore have supported Hague. Conversely, we can hypothesise that the insiders who had held government posts under Major would be more likely to vote for Clarke in the belief that he would choose more of his fellow insiders for senior opposition posts (Table 9).

Table 9 - Recent ministerial experience (to May 1997) and voting behaviour.

Full table

However, analysis shows that there is no significant relationship between voting behaviour and Conservatives who had held ministerial office up until a few weeks earlier. This second, and reframed insider/outsider definition, also proved not to be statistically relevant — ²=0.805, d.f.=1, P=0.370 — and, therefore, the insider/outsider hypothesis is not supported.

The electoral vulnerability hypothesis

The final variable within the political characteristics hypotheses relates to electoral marginality. Vulnerability to electoral rejection shapes parliamentary careers and frontbench opportunities: those with a healthy parliamentary majority have been traditionally more able to attain high office than those holding more marginal constituencies (Hill, 2007, 94–97). Our electoral marginality or vulnerability hypothesis assumed that Conservative parliamentarians in marginal constituencies would tend towards the most popular leadership candidate with the wider electorate, as this may be crucial to their capacity to retaining their constituency. As Clarke was more popular with the wider electorate than Hague, it was assumed that Conservative parliamentarians in marginal constituencies may have been inclined towards Clarke rather than Hague.

However, the median Hague voter among Conservative parliamentarians had a majority of 11.6%, while the median Clarke endorser had a majority of 12.3%. Indeed, the candidates' support is fairly evenly spread across the range of majorities, as is demonstrated in Table 10. Furthermore, analysis shows that there is no relationship between the size of Conservative parliamentarians majorities and their voting behaviour — ²=3.549, d.f.=5, P=0.616 — and therefore, the hypothesis is not supported.

Table 10 - Electoral vulnerability and voting behaviour.

Full table

Having dealt with hypotheses that consider the political characteristics of the candidates and voters, we can turn to possible evidence that ideological factors influence voting behaviour.

Determining Voting Motivations: Ideological Factors

The hypotheses relating to ideological factors advanced three key dilemmas. First, were we correct in assuming that on the social, sexual and morality policy ideological divide, social liberals would endorse Clarke, and social conservatives would gravitate towards Hague? Second, were we correct in assuming that on the economic policy ideological divide, economic wets would coalesce around Clarke, while economic dries would tend towards Hague? Finally, of greatest significance, did pro-Europeans endorse Clarke, and Eurosceptics Hague, as we assumed?

Before we can address these dilemmas and determine whether these hypotheses were substantiated or not we need to determine the ideological composition of the PCP. This was more complex than the political characteristics evidence (which was easily obtained from The Times Guide to the House of Commons). We needed to determine the attitudes of each Conservative parliamentarian on each of the three ideological policy divides in order to determine whether they were socially liberal or socially conservative. While social conservatives tend to oppose homosexual rights, freedom of contraception and abortion, while supporting the family, capital and corporal punishment, censorship and restrictions on immigration, social liberals tend to adopt a less authoritarian and more tolerant attitude to issues of morality. We also needed to determine where Conservative parliamentarians stood on the economic policy ideological divide. Interventionist wets favoured measures to reduce unemployment and maintain growth. Although Thatcherite dries would also doubtless claim that they favoured measures to reduce employment and ensure growth, the crucial difference here was that their measures emphasised control of the money supply and supply-side reforms, coupled with promoting labour market flexibility, to boost employment in the longer term (albeit of a low-paid or insecure variety), while advocating a positivist approach towards the public sector and the maintenance of good relations with the trade unions. Finally, we needed to clarify whether they were Europhiles or Eurosceptics. Europhiles advocated interdependence and argued that further integration, with an implicit pooling of sovereignty, was essential to renewed British influence on the world stage. Eurosceptics opposed further European integration because it entailed surrendering sovereignty to a supranational body that Britain could not control (Cowley and Garry, 1998, 479).

The following four sources were examined to ascertain the attitudes of Conservative parliamentarians on each of the three ideological policy divides: first, division lists; second, early day motions,¹² third, membership of party groups,¹³ and fourth, public comment.¹⁴

Having explained the range of sources utilised, it is necessary to explain the process through which we ascertained the ideological disposition of each Conservative parliamentarian on each of the ideological divides. Each Conservative parliamentarian starts from an assumption that they are ideologically neutral and have an ideological score of zero. If a Conservative parliamentarian is a member of a dry economic group they are given a score of +1, conversely if they are a member of a wet group they are given a score of -1. This method of scoring is then repeated for public and private comment expressed during the course of the 1997 general election and the subsequent Conservative Party leadership campaign. Through this method, a statement supporting a dry position scores +1, while a statement supporting a wet position scores -1. The same process is applied to Early Day Motions and division lists. Conservative parliamentarians with a positive rating are regarded as economically dry; those with a negative rating are wet, while those with a score of zero can be considered to be neutral. The same method can be applied to determine whether a Conservative parliamentarian is of a Eurosceptic or Europhile persuasion, or whether they are socially liberal or socially conservative.

Table 11 traces the ideological disposition of Conservative parliamentarians across the three ideological policy divides. From this it is clear that on the social, sexual and morality ideological policy divide, the social conservatives were in the ascendant, with 122 Conservative parliamentarians adhering to that ideological position, with the social liberals restricted to only 29 and the remaining 13 being defined as agnostic. On the economic policy ideological divide it was clear that the advocates of a limited state, or Thatcherite dry economic liberals, were in the ascendant. The number of Conservative parliamentarians adhering to the interventionist position of justifying an extended state, or a wet position, was a mere 40. The agnostics were limited to 13, leaving the remaining 111 as advocates of a limited state. Finally, on the European ideological policy divide the dominance of the Eurosceptics over the Europhiles was evident: 139 Conservative parliamentarians were identified with the Eurosceptic cause, leaving a rump of just 14 Europhiles.

Table 11 - The ideological composition of the parliamentary conservative party 1997.

Full table

The social, sexual and morality policy ideological hypothesis

Our research on the ideological composition of the PCP concluded that 29 Conservative parliamentarians were socially liberal, 122 were socially conservative, and the remaining 13 were agnostic. As Table 12 demonstrates Clarke was endorsed by two-thirds of socially liberal Conservative parliamentarians, while a similar proportion of Conservative parliamentarians of a socially conservative persuasion, voted for Hague.

Table 12 - Social, sexual and morality ideological divide and voting behaviour.

Full table

Of the three ideological policy divides, the social, sexual and morality divide was not a central part of the campaigning approaches of any of the candidates (a sharp contrast to the succession battle 4 years later). Despite the fact that it did not dominate the party leadership campaign, our research has demonstrated that Clarke was more attractive to social liberal Conservative parliamentarians, and Hague was more appealing to Conservative parliamentarians of a socially conservative persuasion. There was a significant relationship between social, sexual and morality-based attitudes and voting behaviour — ²=8.219, d.f.=2, P=0.016 — and, therefore, the hypothesis is supported.

The economic policy ideological hypothesis

Our research on the ideological composition of the PCP concluded that 37 Conservative parliamentarians were economic wets; 103 were economic dries, while 12 were beyond ideological categorisation. As Table 13 demonstrates, Clarke secured three-quarters of the economic wets, while Hague secured two-thirds of economic dries. Although agnostics also favoured Hague, the margin was close to that predicted by the chi-square test. This indicates that there is strong association between the wet/dry dichotomy and voting behaviour, which supports the hypothesis — ²=19.058, d.f.=2, P<0.0001

Table 13 - The economic policy ideological divide and voting behaviour.

Full table

The European policy ideological hypothesis

Of all of the ideological policy divides the European ideological policy divide was the most intriguing to us. The fratricidal conflict between the Eurosceptics and Europhiles was the defining (and disfiguring) dynamic of the Major era: indeed, it had been a contributory factor in his decision to dramatically hold a party leadership election 2 years prior to the Conservatives' electoral meltdown. Therefore, it would be surprising if the European policy ideological divide was not a key determinant in the war of the Major succession.

Our European policy ideological hypothesis had assumed that pro-European Conservative parliamentarians would gravitate to Clarke, while Eurosceptic Conservative parliamentarians would coalesce around Hague. Given that the ideological composition of the PCP was overwhelmingly Eurosceptic this should have aided the candidature of Hague. The unlikely pact that Clarke attempted to formulate with Redwood, in preparation for the final ballot, demonstrated the need for the Clarke candidature to find a way to appeal to the Eurosceptic wing of the PCP, given that his career and reputation had been built on the pillars of Europhilia.

Table 14 demonstrates that all pro-Europeans within the PCP did indeed endorse Clarke, as did the majority of agnostics on the European policy ideological divide. Hague secured the party leadership at the behest of the Eurosceptics; but, critically, not all Eurosceptics endorsed him. Despite this the analysis shows that there is an association between the European policy variable and voting behaviour — ²=23.721, d.f.=2, P=0.000 — and, therefore, the hypothesis is supported.

Table 14 - The European policy ideological divide and voting behaviour.

Full table

Analysis

We believe that through our data set of voting behaviour, and its application to the political characteristics and ideological disposition of the candidates vis-à-vis their electorate, we can advance four arguments relating to the war of the Major succession that have not been advanced in the existing academic accounts to date.

First, no previous paper on the 1997 Conservative Party leadership election has considered the relationship between political characteristics, candidates and voting behaviour. Our research revealed that there was a statistical correlation between age and parliamentary experience and voting behaviour. Younger Conservatives and recently elected parliamentarians coalesced around Hague and older Conservatives and long serving parliamentarians gravitated towards Clarke. This arguably reflects the respective campaigns. Clarke based his appeal around his experience and political gravitas, while Hague sought to portray himself as the candidate of the next generation or the fresh start candidate. However, our research did not substantiate the hypotheses relating to the insider/outsider dichotomy or electoral marginality. The absence of a correlation between the candidates' voting motivation and electoral marginality was illuminating. The electoral marginality hypotheses assumed that voters with marginal constituencies would gravitate towards the candidate who was the most electorally appealing in the country at large, as this provided them with the best opportunity of retaining their constituencies at the next general election. That the superior electoral appeal of Clarke failed to act as a motivating factor in the voting behaviour of Conservative parliamentarians was intriguing given that the party had just been overwhelmingly rejected by the national electorate.

Second, the paper also confirmed that ideological disposition was a considerable influence and did act as a motivational factor on the voting behaviour of Conservative parliamentarians. Yet this had been implied by both Cowley and Norton in their earlier short papers, so in what sense does this paper break new ground? In their articles, both the Philips had focused on the centrality of the European ideological divide as an influence on voting behaviour. Unlike them, this paper now confirms that attitudes across all three of the ideological determinants of post-Thatcherite Conservatism were a motivational influence on voting behaviour. On the social, sexual and morality divide there was a clear relationship: social liberals tended towards Clarke and social conservatives towards Hague. On the economic policy divide there was also a clear relationship: economic wets gravitated towards Clarke and economic dries towards Hague. Finally, on the European policy divide there was also a clear relationship. Clarke gained support from Europhiles and Hague from Eurosceptics. Hague, therefore, obtained the party leadership at the behest of the pure Thatcherites: the socially conservative, economic liberal and Eurosceptics. However, Hague did not secure their absolute support. The PCP was overwhelmingly Thatcherite in its ideological orientation on the key ideological divides of contemporary Conservatism, as indicated in Table 11.

Third, the research underpinning these tables, demonstrates that a remarkable socio-ideological transformation had occurred within post-Thatcherite Conservatism. Research on the ideological disposition of the PCP in the 1992 to 1997 Parliaments suggested that there were 101 social liberals, 230 social Conservatives on the social, sexual and morality ideological divide, 112 interventionist wets, 188 Thatcherite dries, plus 31 agnostics/loyalists on the economic policy ideological divide, and 98 pro-Europeans, 192 Eurosceptics, plus 41 agnostics/loyalists on the European ideological policy divide (Heppell, 2002, 299–324). The traditional Tory left had been eroded by the commencement of the 1997 Parliament: as a percentage social liberalist thinking had fallen from 30.5% in the previous Parliament to 17.6% in the new Parliament; interventionist wet thinking had fallen from 33.8% to 23.8%; and critically pro-Europeanism had fallen dramatically from 29.6 to 7.9%.

Fourth, we can contextualise our data set against those used by Norton and Cowley in their papers (Cowley, 1997, 91–95; Norton, 1998, 10–14). While both implied that roughly 50–60% of the PCP had Eurosceptic leanings, we believe that the scale of Eurosceptism was more pronounced. Through a combination of examining voting behaviour, early day motions, membership of party groupings, and examining public comment (most significantly campaign literature), our data set suggests that this underestimates the strength of Eurosceptism, which we calculate at over 80% of the PCP. Through the same methods we believe that that the scale of Europhilia was far smaller than they assumed. Our finding of roughly 7% Europhilia may seem a significant deduction from the 25% alluded to by Norton, or the 20% implied by Cowley, but can be validated by the work of Turner, who implied that loyalist Europhilia had slipped below 5% in the 1997 PCP (Turner, 2001, 214).

Our paper suggests that because Cowley has significant numbers (30%) of Conservative parliamentarians in the faithful/centrist/agnostic bloc, this skews his conclusions. It is because this faithful/centrist/agnostic bloc is bloated and the scale of Europhilia is over stated that Cowley concludes that the final ballot was not characterised by as much ideological polarisation as the previous ballots of the contest (Cowley, 1997, 94). As we have demonstrated, when the scale of Eurosceptism vis-à-vis agnosticism and Europhilia is reassessed, it is clear that the final parliamentary ballot was characterised by ideological polarisation, with the support for Hague being virtually exclusively Eurosceptic.

Conclusion

This paper established a data set of the voting behaviour of Conservative parliamentarians in the final ballot of the 1997 party leadership. The data sets of the respective candidates were then applied to political characteristics (age and parliamentary experience, career status and electoral vulnerability) and ideological disposition (the economic ideological divide, the European ideological divide, and the social, sexual and morality ideological divide), to determine whether our hypotheses could be validated. By testing our data sets against these variables and utilising bivariate analysis this enabled us to demonstrate that interpretations of how and why Hague emerged as the new leader of the Conservative Party that attribute his emergence to Europe alone are intellectually limiting: age, experience as well as ideology (in its widest sense) explain how the youthful, inexperienced, Thatcherite right Eurosceptic, Hague defeated the senior, experienced, Tory left Europhile, Clarke.

Traditionally, the main preoccupation of the Conservatives had been the acquisition and retention of power. In that pursuit they had been prepared to adapt their statecraft strategies to achieve those ends. This paper helps us to locate the war of the Major succession within the context of both post-Thatcherite Conservative party politics and their decade-long spell in opposition. It provides us with evidence to suggest that the final ballot of the Conservative party leadership of 1997, and the defeat of Clarke, was a victory for ideological dogmatism over traditional Conservative statecraft, that is, pragmatism in the pursuit of power (Hayton, 2006, 2). By rejecting the more senior, experienced, electoral attractive, but economically damp, socially liberal and pro-European Clarke, the Conservatives prevented themselves from steering themselves back towards the political centre ground. By electing the junior, inexperienced, less electorally attractive, economically dry, socially conservative, Eurosceptic, Hague, the Conservatives ensured that they would drive down the cul-de-sac of the common sense revolution and the core vote strategy (Bale, 2006, 2, 4). The dominance of Thatcherite informed ideological forces within the PCP, primarily located among the younger generation of Conservative parliamentarians, ensured that the necessary processes of adaptation, and the reconfiguring of the Conservative narrative, would be delayed by a decade. Ideological acceptability mattered more than electoral appeal or political experience.

Notes

¹ Numerous Gallup and Mori opinion polls conducted during the party leadership election stated that Clarke was the most favoured candidate among both Conservative and non-Conservative voters. A constant refrain by political commentators throughout the campaign was the assertion that Clarke was the candidate that Labour feared most.

² Clarke was defined as an economic wet, pro-European, social liberal, on the basis of an appraisal of the basis of his positioning in the Heppell typology of Conservative thought (see Heppell, 2002, 299–324).

³ The economic dry and moderate Eurosceptism of Hague was widely acknowledged by political commentators. His attitudes to morality were less well documented. We deemed him to be a social conservative due to the fact that on conscience issues he had voted for or expressed views that indicated that he was for the death penalty, against abortion, for corporal punishment, sceptical towards reforms that made divorce easier, concerned about the liberalisation of Sunday trading, and against embryo research (see Heppell, 2002, 299–324).

⁴ When considering the election of Hague as the new party leader, it may be worth reflecting on the view of an anonymous Conservative parliamentarian. Just 3 months before the Conservative Party leadership election of 1997, it was suggested to him that Hague might emerge as the next party leader. They responded disdainfully with: 'please, you cannot be serious' (Brandreth, 1999, 488).

⁵ The former Health Secretary, Stephen Dorrell, had publicly declared his intention to stand but decided to withdraw prior to the deadline.

⁶ Redwood attempted to argue that his status outside of the cabinet between 1995 and 1997 absolved him of any responsibility or association with the discredited Major era and the resultant electoral meltdown. However, the association that resonated was of disloyalty resulting from the challenge that Redwood had made to Major. By constantly reminding Conservative parliamentarians of the failings of the Major era, an indissoluble connection between the candidature of Redwood and the cause of the implosion of the Major era was established. Redwood was a symbol of ideological division and disloyalty; to many his advancement would be an impediment to unity and renewal (Williams, 1998, 188).

⁷ Conservative peer and political academic, Norton, argued that Lilley lacked electoral appeal, that is, there was no evidence that he would not be a liability to the standing of the party (Norton, 1998, 13). Polling by Gallup and Mori throughout the period between the general election defeat and the first ballot of the Conservative party leadership election, showed that Lilley had a low public recognition factor and was the least appealing of the candidates to the electorate (Williams, 1998, 200).

⁸ Williams has eloquently outlined how intensely Redwood disliked Hague. Throughout the campaign Redwood had been openly hostile to Hague and his cohorts, as he found the idea of Hague assuming the party leadership as preposterous (Williams, 1998, 216, 218).

⁹ Redwood argued that he was attracted to the alliance with Clarke as it was imperative to end the 'Balkanisation' of the Conservative Party and that only Clarke offered an approach that was 'inclusive' enough to achieve this. By adopting a free vote on the hypothetical question of a parliamentary division of the single European currency advocates of the Clarke–Redwood alliance suggested that this was more inclusive and unifying than the stance that Hague had shifted towards; that is, his insistence that members of his shadow cabinet would have to endorse his negative stance on the hypothetical question of the single European currency (Alderman, 1998, 12).

¹⁰ A possible concern might be the reliability of the information provided from Conservative parliamentarians, the so-called 'fib factor'. Like Cowley and Garry, when analysing the second ballot of the 1990 Conservative Party leadership election, we agree that there is little need to publicly lie about ones' support (intention or behaviour) when there was no threat to an incumbent leader (i.e. in a vacant contest). Therefore, they argued that lying did not constitute a methodological problem, a position that we agree with (Cowley and Garry, 1998, 498–499).

¹¹ If all of our calculations are correct, we can deduce that the excluded nine must have voted for Hague, as we have identified all of the Clarke supporters and the two abstainers, plus all but nine of the Hague supporters. After much discussion, we decided to retain their status as unclassified in our data set.

¹² The methods through which divisions lists and early day motions can be exploited to ascertain the ideological disposition of Conservative parliamentarians are explained in detail in Heppell (2002, 299–324).

¹³ For example, membership of the No Turning Back group (which included John Redwood and Peter Lilley) was indicative of economic neo-liberalism, while membership of the Tory Reform Group (which included Kenneth Clarke and Michael Heseltine) was indicative of an economic interventionist damp mentality. Both these groups can be defined as dual purpose: for example, the Tory Reform Group can be indicative of attitudes to both the economic and the European ideological policy divides as it represents the economic damp and Europhile strands of thought within contemporary Conservatism. Alternatively, the Conservative Way Forward articulates both an economic dry and a Eurosceptic viewpoint. Supplementing these sources on the European policy ideological divide was membership of pro-European groupings (e.g. Conservative Group for Europe and the Tory Euro Network) or membership of Eurosceptic groupings (e.g. the Bruges Group and the 92 Group). On the social, sexual and morality ideological policy divide membership of the all-party parliamentary pro-choice and sexual health group was indicative of a socially liberal mentality. Alternatively, membership of the Conservative Christian Fellowship and the all-party parliamentary pro-life group was seen to be indicative of socially conservative thinking.

¹⁴ Given that the war of the Major succession occurred in the first 6 weeks of a new Parliament, we relied heavily on the 1992–1997 Parliament with regard to the first and second sources (i.e. division lists and early day motions). New parliamentary entrants were ideologically determined primarily through the third and fourth sources (i.e. membership of party groups and public comment, most notably campaign literature).

References

1. Alderman, K. (1998) 'The conservative party leadership election of 1997', Parliamentary Affairs 51(1): 1–16.
2. Bale, T. (2006) 'What took you so long? Why the conservative party needed nearly a decade to come to its senses', Paper presented to the Annual Conference of the Elections, Public Opinion and Parties specialist group, University of Nottingham, September 2006.
3. Brandreth, G. (1999) Breaking the Code: Westminster Diaries, London: Phoenix.
4. Campbell, J. (2003) Margaret Thatcher. Volume Two: The Iron Lady, London: Jonathan Cape.
5. Cowley, P. (1997) 'Just William? A supplementary analysis of the 1997 conservative leadership contest', Talking Politics 10: 91–95.
6. Cowley, P. and Garry, J. (1998) 'The British conservative party and Europe: the choosing of John major', British Journal of Political Science 28(3): 473–499. | Article |
7. Crick, M. (2005) In Search of Michael Howard, London: Simon and Schuster.
8. Crowson, N. (2007) The Conservative Party and European Integration since 1945, Abington: Routledge.
9. Hayton, R. (2006) 'Conservative party strategy in opposition: a strategic-relational analysis', Paper presented to the Political Studies Association Conference, University of Reading, 5 April 2006.
10. Heppell, T. (2002) 'The ideological composition of the PCP 1992 to 1997', The British Journal of Politics and International Relations 4(2): 299–324. | Article |
11. Hill, M. (2007) 'The PCP and conservative party leadership: The elections of William Hague and Iain Duncan-Smith', Unpublished Doctoral Thesis, Centre for Democracy and Governance, University of Huddersfield.
12. Nadler, J. (2001) William Hague: In His Own Right, London: Politicos.
13. Norton, P. (1998) 'Electing the leader: the conservative leadership contest 1997', Politics Review 7(4): 10–14.
14. Quinn, T. (2005) 'Leasehold or freehold? Leader-eviction rules in the British conservative and labour parties', Political Studies 53(3): 793–815. | Article |
15. Turner, J. (2001) The Tories and Europe, Manchester: Manchester University Press.
16. Williams, H. (1998) Guilty Men: Conservative Decline and Fall 1992–1997, London: Aurum Press.