Selecting Sunak: Conservative MPs’ Nomination Preferences in the (Second) British Conservative Party Leadership Election

Abstract

This article utilises an original dataset covering all members of the Parliamentary Conservative Party (PCP) to analyse the basis of support for Rishi Sunak, Boris Johnson and Penny Mordaunt in the October 2022 British Conservative Party leadership election. The significance of our findings is that they form the basis for a new understanding of how conservatism in the UK is changing post-Brexit. Our political variables include constituency majority size, red wall status, constituency support for leave, an MP’s position in the June 2022 vote of confidence in Johnson’s leadership of the Conservative Party, ministerial status and rebellion rates on key legislation. Our ideological variables include EU referendum position and membership of the European Research Group and the Conservative Environment Network, along with whether MPs are classed as socially liberal, socially conservative or pro-levelling up. We find that candidates appealed to different, often rival, sections of the Conservative Party. In line with previous studies of this nature, social liberalism still represents a key dividing line within the party, and the European dimension of intra-party conflict has remained significant in light of the UK’s decision to leave the European Union.

This article offers the first empirical analysis of the preferences of Conservative MPs in the October 2022 Conservative Party leadership election. The election occurred following the resignation of Liz Truss as party leader and prime minister after just 49 days in office. Truss was succeeded by Rishi Sunak, the runner up in the previous leadership election and the only official candidate following the withdrawal of former prime minister Boris Johnson and Leader of the House of Commons Penny Mordaunt. Since all three candidates had public declarations of support from their fellow MPs, we are able to compare and contrast the bases of support for each candidate.

This article is positioned within an emerging body of academic literature on leadership selection, developing the tradition of research examining the motivations for candidate support among Conservative Party MPs at the parliamentary stage of leadership contests under the post-1998 rules (Heppell and Hill, 2008, 2009, 2010; Jeffery et al., 2018, 2022, 2023; Quinn, 2019; see also Roe-Crines et al., 2021 on confidence motions). From a comparative studies perspective, the focus of this literature has tended to be on leadership procedures and the composition of the electorate (see Kenig, 2009a; Cross and Blais, 2010), election competitiveness and leadership procedures (LeDuc, 2001; Kenig, 2009b), the factors for why and the means by which incumbents are evicted from their positions (‘t Hart and Uhr, 2011; Ennser-Jedanastik and Schumacher, 2021) and whether leadership transition can be beneficial to parties in terms of resolving ideological or policy disagreements (see Pedersen and Schumacher, 2015; Somer-Topcu, 2017).

From the perspective of leadership selection within the British Conservative Party, these themes have not been the primary drivers of academic debate, although some have been engaged with (see, for example, Alderman, 1999 and Quinn, 2005 on leadership rules; McAnulla, 2010 on evictions; and Heppell and McMeeking on leadership transitions, 2021). Rather the focus has been on evaluating campaigns and candidates and trying to explain outcomes (see, for example, Denham and O’Hara, 2008; Heppell, 2008). As stated above, our study specifically contributes to a tradition of research that examines the motivations for Conservative Party parliamentarians backing candidates at the parliamentary stage of leadership contests under the post-1998 rules, alongside studies of a similar kind from the pre-membership participation era of Conservative Party leadership elections (Cowley and Garry, 1998; Cowley and Bailey, 2000). To do this, we have created an original dataset of Conservative MPs that combines public declarations of support and synthesises these with political, ideological and demographic variables.

The article begins by contextualising the leadership contest via a brief overview of the leadership instability that characterised Conservative Party politics throughout 2022, before explaining how and why the vacancy came to exist and outlining the rules governing the selection process. We then identify the political and ideological variables our dataset will cover before explaining our methods and how we created our dataset. This is followed by our research findings and an analysis of what this means for our understanding of contemporary Conservative Party politics. We conclude by situating our findings within the broader history of leadership selection within the Conservative Party.

1. Leadership instability in the Conservative Party

The leadership election that took place in October 2022 was the third time in five months that Conservative MPs had been invited to comment on the party leadership.

The first occurred in early June, when the Conservative Party voted on whether they had confidence in Boris Johnson’s leadership. This was triggered as the party’s poll ratings, and Johnson’s satisfaction ratings, had declined significantly due to a series of governmental scandals, the chief of which was the ‘partygate’ scandal that came to dominate political debate in late 2021 and early 2022 (Elgot, 2022). Johnson had looked hypocritical by imposing lockdown restrictions during the COVID-19 pandemic, and then breaching them by holding social gatherings in Downing Street. His actions also resulted in a police investigation and a fine for breaking his own lockdown guidelines, and an accusation that he had lied to parliament when denying that he had attended these social gatherings in Downing Street (Judge, 2022). Despite an independent report condemning Johnson’s behaviour, he was unwilling to resign or to accept that he had broken the ministerial code of conduct. This, combined with a number of other scandals, including the government-backed (and then aborted) attempt to change the rules around the disciplinary process for MPs in order to protect backbencher Owen Paterson (who had breached paid advocacy rules), and questions over the source of funding of the refurbishment of Johnson’s prime ministerial apartment all contributed to a sense of governing degeneration after over a decade in power (Roe-Crines, 2022).

On 6 June Graham Brady, the chairman of the 1922 committee announced that the threshold of letters from Conservative MPs required to force a vote of no confidence in Johnson’s leadership had been reached (Boycott-Owen, 2022; Gutteridge, 2022). Johnson secured a victory, winning the confidence motion ballot by 211–148 votes. This meant he could not be subjected to another confidence challenge for a further 12 months, but with 41% of Conservative MPs voting against him, it was clear that his political authority had been badly eroded (Merrick, 2022). A further scandal, involving accusations of sexual misconduct against Chris Pincher, the deputy chief whip, and questions about whether Johnson knew about these claims or not before appointing him to the role, led to a series of ministerial resignations, starting with the health secretary, Sajid Javid, and closely followed by the chancellor of the exchequer, Rishi Sunak. In total, 62 members of the government resigned, which made it untenable for Johnson to remain in office and on 7 July 2022, Johnson resigned as both party leader and prime minister.

The second ballot process occurred because of Johnson stepping down and a leadership contest being held. Conservative parliamentarians had been asked to engage in a series of eliminative ballots to identify two candidates to present to the party membership. After five ballots, which eliminated Jeremy Hunt, Nadhim Zahawi, Suella Braverman, Tom Tugendhat and Mordaunt, the members were then presented with a choice between Sunak or Truss. Across the eliminative parliamentary ballots, Sunak was first placed in all five, but after being second to Sunak in the first four ballots, Mordaunt lost out to Truss in the final eliminative ballot. Sunak won 137 votes (38.3%), Truss 113 votes (31.6%) and Mordaunt 105 votes (29.3%) (Penna, 2022). By the end of the membership ballot round, however, more MPs had publicly declared for Truss than Sunak (158–138, respectively). After the six-week campaigning period, the members elected Truss over Sunak by 80,326 votes to 60,399 (or 57.4–42.6% on 82.6% turnout) (Crerar, 2022). As noted by Bale, ‘all politics inevitably involves not just ideas, interests, and institutions, but also individuals’, and in this instance, Sunak suffered in the membership round due to his ‘betrayal’ of Johnson, who was still broadly popular with the grassroots membership (Bale, 2023, p. 3).

Truss’ premiership was to be short-lived. The mini-budget delivered by her Chancellor, Kwasi Kwarteng, presented as The Growth Plan, was illustrative of what was to be known as ‘Trussonomics’ (Conway, 2022; HM Treasury, 2022). To stimulate economic growth, they set about implementing significant cuts in taxation, including the abolition of the 45p top rate of income tax, cuts to the basic rate of income tax and cancelling the proposed increases in corporation tax, but refused to provide a credible plan for how these measures would be financed. This disturbed the financial markets, leading to a drastic increase in the cost of UK government borrowing and mortgage rates alongside a collapse in public support for the Conservatives. Truss attempted to stabilise her administration by dismissing Kwarteng from his post (and replacing him with Hunt) and by abandoning the majority of the mini-budget (Culbertson, 2022). She then compounded concerns about her leadership credibility by a series of unconvincing performances in interviews, press conferences and parliament, alongside her mishandling of an opposition motion on fracking, whereby it was unclear whether the vote was a matter of confidence, reinforcing an impression of governmental chaos under her leadership. Under the existing rules, the option of holding a vote of no confidence against her was not open to Conservative MPs within the first twelve months of a leader’s tenue, although there were public discussions about the possibility of a rule change. The scale of discontent led her to stand down after discussions with Brady (BBC News, 2022).

Having created a leadership vacancy, consultations between Brady and Jake Berry, chairman of the Conservative Party, resulted in a shortened leadership election to replace Truss. Critically, it was decided that the leadership question should be concluded in one week, as opposed to the traditional long membership ballot stage. At stage one, the threshold to make it onto the first round of the ballot was changed from 20 to 100 (or from 6% to 28% of the Parliamentary Conservative Party [PCP]), to reduce the need for so many eliminative parliamentary ballots. If three candidates had made it past this threshold, one would be eliminated by MPs and an indicative vote between the final two would be held, to signal to members who had the support of the parliamentary party. Unlike the previous leadership election, the membership ballot would be conducted online, rather than via post, over the course of three days (Nevett and Gregory, 2022).

Stage two of the contest was averted as only Sunak submitted a nomination form, thereby ensuring that he became the leader of the Conservative Party and the next prime minister. Johnson opted not to formally submit his nomination, stating that although he had the support of over 100 MPs (a claim later backed by Brady), he could not unify his party (Andersson, 2022; Hill, 2022). Dramatically, Mordaunt released a statement just one minute before the deadline to submit nomination papers, saying she had fallen short of the 100 nominations required (Diver, 2022). Despite this, the number of public declarations of support for all three is sufficient for us to identify the bases of support for (and opposition to) each candidate.

2. Data collection and methods

In selecting our variables, we were influenced by the studies of Cowley and Bailey (2000) and Cowley and Garry (1998) on the 1975 and 1990 parliamentary leadership election ballots, which had also informed subsequent studies by Heppell and Hill (2008, 2009, 2010) on the parliamentary ballots in the 1997, 2001 and 2005 leadership elections. Their emphasis, however, on the ideological determinants of Conservatism around the wet-dry divide on the economy, the pro-European or Eurosceptic divide on European policy and the liberal-conservative divide on social, sexual and moral matters (Garry, 1995; Heppell, 2002) requires updating to make it more reflective of current debates within contemporary Conservative politics. So, for example, we wanted to embrace the debates around the culture wars, levelling up, the environment and Brexit, thus replicating the more recent approaches adopted by Jeffery et al. (2018, 2022, 2023) and Roe-Crines et al. (2021).

This study is innovative in that it draws on a unique dataset of MPs’ membership of extra-parliamentary party groupings to determine their ideological positioning, as opposed to the over-reliance on division lists that characterised previous studies on ideological distinctions within post-Thatcherite Conservatism (see Heppell, 2002, 2013; Heppell and Crines, 2016; Heppell et al., 2017). Group membership is an important and useful method because MPs are free to choose which groups they are affiliated with, and thus membership of a group should be understood as a clear public signal about the types of issues an MP finds important.

We divide our variables into political and ideological factors. To test for the significance of political factors, we first assumed that Conservative parliamentarians holding marginal constituencies would be more susceptible to opinion poll evidence on how well the Conservatives might perform under each of the three possible candidates. Opinion polling suggested that the Labour Party would maintain its lead regardless of the candidate selected, but the gap was largest if Mordaunt was selected (a 36-point lead to Labour), as compared to a Sunak premiership (20-point Labour lead), or a Johnson premiership (15-point Labour lead) (Redfield and Wilton Strategies, 2022).

We also concluded that alongside the vulnerability of each constituency we should consider the possible significance of the type of constituency, specially whether it was in the so-called red wall (Kanagasooriam and Simon, 2021) and the extent to which prior loyalty to Johnson—as the leader that led them to winning those seats at the 2019 general election—might be of significance. We further considered that MPs’ electoral considerations might be affected by the shadow of Brexit, which has polarised attitudes in the country (Sobolewska and Ford, 2020). Johnson and Mordaunt’s clearer association with the Leave campaign during the Brexit referendum (Sunak had backed Leave but was less of a public figure during the referendum) may have appealed to MPs representing more Eurosceptic areas.

We also consider behaviour in the vote of no confidence in Johnson’s leadership in June 2022. It seems obvious to expect that those who backed Johnson would do so out of support for his leadership, but there are two sets of motivations for MPs depending on their position in the executive. MPs who are members of the government may avoid making a public declaration of opposition to Johnson, for fear of reprisal in the form of sacking should he retain the confidence of the PCP. Backbench MPs, however, have no such worry and so their own decisions are more likely to be based on their actual views of Johnson. Thus we included a variable capturing the interaction between confidence vote and ministerial status.

Finally, the Conservative parliamentary party elected in 2019 has been surprisingly rebellious given the perceived unity of the party after the landslide general election. Major rebellions have taken place on issues as diverse as international aid, trade policy and the exclusion of Huawei from future 5G phone networks. Given that Mordaunt was less associated with government policy more generally in the 2019 parliament than Johnson or Sunak, we will explore whether she became the rebels’ choice.

We also wanted to consider the significance of ideological factors, beginning with an MP’s position on Brexit. Although all three potential candidates voted leave, Johnson and Mordaunt were more prominent in the leave campaign. Sunak had also been the preferred candidate of remain-voting Conservative MPs in the summer leadership contest (Jeffery et al., 2023).

Alongside considering the possibility that attitudes towards Brexit might structure nomination preferences, we also wanted to consider the attitudes towards cultural issues, levelling up and net zero. We identified socially liberal MPs by membership of Bright Blue and the Tory Reform Group, given both groups describe themselves as supporters of ‘liberal conservatism’. Conversely, we used membership of the Common Sense Group as a measure of social conservatism given their concern with political correctness, the ‘woke’ agenda and cancel culture, and that they present themselves as the defenders of traditional British values in the face of these developments. Given the disagreements over gender self-identification raised during the campaign—Mordaunt was criticised by social conservatives for previously adopting a more pro-trans stance and was seen as being closer to the socially liberal wing of the party (Boycott-Owen, 2022; Gimson, 2022; Payne and Parker, 2022)—we felt that this was a useful variable to consider.

To gain an insight into how the Johnsonian emphasis on levelling up might have structured candidate preference, we looked to membership of either the Onward task force for levelling up, Blue Collar Conservatism or the Northern Research Group. We expect MPs supportive of levelling up to tend to back Johnson, given his issue ownership of levelling up, Mordaunt’s lack thereof, and Sunak’s perceived hostility to the policy (Berry, 2022). Finally, we wanted to consider environmental issues. Johnson’s rhetoric was supportive of the net zero policy agenda, and although for many this rang hollow, it was a clear statement of intent and thus could appeal to specific ideological audiences within the PCP, thereby influencing leadership selection. As such, we looked at the membership of the Conservative Environment Network, which advocated market-based solutions to securing net zero.

MPs are under no obligation to publicly announce how they will vote in Conservative Party elections, nor are any public records of voting behaviour kept. Thankfully, many MPs do declare who they will be voting for and numerous websites keep track of these declarations. For this study, we collated the declarations made at the point just before Johnson publicly withdrew from MPs’ social media accounts, personal websites or declarations to the press. Out of a total of 358 MPs, we found 154 declarations for Sunak (43%), 65 for Johnson (18%) and 26 for Mordaunt (7%). One hundred and thirteen MPs did not make their position public (32%).

The starting point for our data collection process was to utilise the dataset constructed by Jeffery et al. (2022), which included Conservative MPs’ backgrounds and behaviour in previous leadership contests. We also drew on the Parlitools R package (Odell 2022) for MPs’ marginality and the mnis package (Odell 2021) for ministerial status. For group membership used to denote ideological positions, we drew on publicly available membership lists and constructed binary variables for whether MPs are socially liberal, socially conservative and pro-levelling up, as per the criteria established above. Data on MPs’ positions on the Brexit referendum were taken from Lynch (2020), red wall status was coded according to Kanagasooriam and Simon’s (2021) criteria, and estimated leave votes in MPs’ constituencies were taken from the estimates compiled by Hanretty (2017).

To construct a variable on MPs’ propensity to rebel in the 2019 parliament, data on seven major rebellions were gathered using the website Public Whip (2022).¹ Given that the majority of Conservative MPs did not rebel on any of the measures, and only a handful rebelled on at least five, we operationalised this as an ordinal variable according to whether MPs had not rebelled on any of these measures, rebelled rarely (once or twice) or rebelled frequently (three or more times). For our control variables, we drew on data on MPs’ gender, ethnicity (Uberoi and Burton, 2022, p. 16), sexuality (Peart, 2022) and educational backgrounds (Butler et al., 2021).

3. Research findings

Table 1 shows the descriptive results of the initial round of nominations. For our political variable, we see that support for Johnson, Mordaunt and Sunak differ on all the factors that we hypothesised: Sunak and Johnson were backed by MPs with a majority averaging 27%, compared with 33% for Mordaunt; red wall MPs made up a greater share of Johnson’s support than they did for other candidates and Sunak’s supporters, on average, had a lower level of support for leave in their constituencies. Those who had publicly stated they had confidence in Johnson made up a greater share of Johnson’s support than for other candidates, but it is interesting to note that while Johnson did win nearly half of the support of backbenchers who supported him publicly, he won just 28% of the ministers who supported him publicly. This suggests many ministers only made their decision in order to keep their ministerial role, rather than due to support for Johnson himself, or his policy platform. Finally, MPs who rebelled more frequently on key legislation made up a greater share of Mordaunt’s support base than MPs who did not.

Moving to ideological factors, we see that while Mordaunt’s support base contained a greater share of leave MPs than Sunak’s (50–42%), Johnson’s support base was overwhelmingly leave-based at 80%. Similarly, over half of Johnson’s public support came from ERG members, compared to 23% for Mordaunt and 19% for Sunak. Interestingly, and contrary to the expected relationship, MPs from the socially liberal wing of the party made up a larger share of Sunak’s base (52%) compared with Mordaunt’s (38%). However, members of the socially conservative wing of the party were overrepresented in Johnson’s support base and non-existent in Mordaunt’s base. Supporters of levelling up were overrepresented in Johnson’s base, but, and contrary to the hypothesised relationship, members of the Conservative Environment Network were underrepresented in Johnson’s support base.

To determine which of the political and ideological factors most significantly affected support for the three candidates, we ran a series of Firth logistic regression models where our dependent variable was a dummy for whether or not an MP supported a particular leadership candidate. The use of Firth logistic regression allows us to deal with small samples, missing data and multiple independent variables more effectively than classic logistic regression (Firth, 1993; Puhr et al., 2017). Table 2 shows the results for the initial round of nominations, alongside a stepwise regression model that allows us to identify the factors that best predicted support for the leadership candidates.

Starting with our political variables, we find no relationship between MPs with more vulnerable majorities and their support for candidates in the initial round of nominations. We find evidence that red wall MPs were significantly more likely to back Johnson than not, but this relationship was only present in the stepwise model, and evidence that these MPs were less likely to back Mordaunt, but this time only in the full model. We do find evidence that as an MP’s constituency leave vote share increased, MPs were more likely to support Mordaunt and less likely to support Sunak. There was no relationship for support for Johnson and constituency leave vote, but this is the only variable that makes it into the stepwise model predicting no public declaration of support—but even then it is not a statistically significant relationship.

In terms of ministerial status and position in the confidence vote on Johnson’s leadership, we find that relative to backbenchers who did have confidence in Johnson, backbenchers who did not have confidence were much less likely to support Johnson in the leadership contest, as were backbenchers who did not make their position known. This relationship was present in both the full model and the stepwise model. There was not, however, any statistically significant difference between backbenchers with confidence and either ministers who had confidence or had an unknown position and support for any candidate. In contrast, backbench MPs with no confidence were more likely to support Sunak or Mordaunt in the full model, but this relationship was not present in the stepwise model, suggesting it was not a strong predictor of support for either candidate.

Finally, when it comes to rebellion, relative to those who never rebelled, those who occasionally and frequently rebelled were more likely to support Mordaunt in both models in the stepwise model. As expected, there is no relationship for rebellion rate and support for Sunak or Johnson.

Moving to our ideological variables, we begin with an MP’s referendum position. Relative to voting leave, find a statistically significant relationship for supporting remain and supporting Sunak in the stepwise model, but apart from that referendum position did not structure support for Johnson or Mordaunt, perhaps reflecting the fact all three candidates did back leave. We do find, however, that membership of the ERG is a strong predictor of support for Johnson and opposition to Sunak. Obviously, there is a significant overlap between EU referendum position and ERG membership, and if ERG membership is removed from the regression model the relationship for the EU referendum variable plays out as expected. Thus, a European-related divide still permeates the PCP when it comes to leadership selection, but this is now embedded in an organisational structure vis-à-vis the ERG than simply just referendum voting behaviour.

Contrary to our hypothesised relationship regarding social liberalism, we see that MPs who could be classed as social liberals were less likely to back Johnson and more likely to back Sunak, while there was no statistically significant relationship for support for Mordaunt. We find no further relationships between other ideological groupings and support for candidates, in contrast to previous studies (Jeffery et al., 2023).

Finally, it is worth noting some interesting outcomes flowing from our control variables. Firstly, we see that female MPs were more likely to support Mordaunt and less likely to support Sunak—and despite claims from the commentariat of Johnson having a ‘woman problem’ (Courea and Dickson 2019; Balls, 2022), this does not seem to have filtered through to his support base in the PCP. We also see that Oxbridge-educated MPs were less likely to support Johnson and more likely to support Sunak. Interestingly, ethnicity did not play a statistically significant role in structuring support for any candidate—despite the well-reported fact that Sunak would be the first British Asian and non-white prime minister.

4. Analysis and conclusions

The aim of this article was to identify the divides within the PCP based on the public declarations of support for the three leading candidates in the October 2022 leadership election. With Truss’s ideologically driven government having crashed and burned, we wanted to know whether the selection of her replacement would be driven by ideology or whether political factors be more important to Conservative MPs. To do this, we explored MPs’ voting behaviour using a unique dataset of political and ideological variables, guided by a long academic tradition of studying leadership elections. Our work therefore contributes to a body of literature that systematically analyses leadership elections by investigating personal, ideological and electoral variables within the PCP.

In considering the variables relevant to determining support for a given candidate, we have identified clear divides: firstly, support for Johnson was primarily driven along the European dimension, with Johnson capturing the support of the hard-line Brexiteers represented by the ERG, rather than leave-backing MPs as a whole and, unsurprisingly, whether an MP was a supporter of his in the confidence vote or represented a red wall seat. We also find that Johnson, despite his claim to being a one nation Conservative, did not win the support of his one nation colleagues—in fact, they were less likely to support him. We find other factors to be broadly insignificant in explaining support for Johnson. Johnson’s support base can thus be described as loyal hard-line Eurosceptics.

Support for Sunak was largely driven by the same factors as Johnson, but operating in the opposite direction: ERG members were less likely to support him, and one nation Conservatives were more likely to. Sunak did gain support from backbenchers who did not have confidence in Johnson, but the magnitude of this effect was much smaller than the magnitude of the positive effect for Johnson. We also see that as a constituency’s leave vote increases, support for Sunak decreases, and that Sunak’s support was drawn from remain-backing MPs relative to leave-backing MPs. Thus Sunak’s support appears to be driven by socially liberal MPs who were not as stridently pro-Brexit as other parts of the PCP.

Mordaunt’s base is interesting insofar as it did not really look like either of the other two candidate’s bases. Whereas she did win more support from MPs who represent stronger leave-backing constituencies, the only other non-control variable that was statistically significant is an MP’s rebellion rate: Mordaunt seems to win support from among MPs who rebelled frequently on the votes included in the study, relative to those who rarely or never rebelled. This suggests Mordaunt appealed to those who were looking for an ideological fresh start or a change in direction from the Johnson agenda.

Furthermore, it is worth considering who did not make a public declaration of support. Our models identified no factors as being significant in predicting such MPs. This suggests that a reason for the lack of declarations was not due to ideological factors—e.g. that there was no candidate the CEN or social conservatives could back and so they remained silent publicly—but rather than the choice not to make a public declaration was due to other reasons.

Finally, there are clear parallels with previous Conservative Party leadership elections. There is a tradition of attitudes towards the European question influencing voting behaviour in Conservative leadership elections. Eurosceptics were aligned with the candidacies of Major in 1990, Hague in 1997 and Duncan Smith in 2001; pro-European MPs were aligned with Heseltine in 1990 and Clarke in 1997 and 2001. Voting behaviour in the referendum was relevant in the 2016 and 2019 leadership elections and in voting on May’s Withdrawal Agreement. Although the European divide has evolved over the 30 or so years mentioned here, it maintained its relevance in 2022, albeit this time partly mediated through membership of the ERG. Thus we see a further evolution of the Eurosceptic/pro-European divide, which then became the leave-remain divide, and now seems to be a split around whether an MP is a hard-line Eurosceptic or not. In a similar ideological vein, the socially liberal divide was significant in the leadership elections of 1997, 2001 and 2016 (Heppell and Hill, 2008, 2010; Jeffery et al. 2018), and we also found it to be significant in this contest, as social liberals were more likely to back Sunak and less likely to back Johnson.

Furthermore, we found an MP’s constituency leave vote to be significant, for Sunak and Mordaunt, which represents a departure from the 2016 contest when a similar variable, UKIP’s vote share in each constituency, was not found to be significant (Jeffery et al., 2018). Given that both Sunak and Mordaunt supported leave we deduce that this is not just about their positions on Europe, since the links between constituency factors and parliamentarians’ voting behaviour in leadership elections are quite tenuous. The significance of the leave vote in MPs’ constituencies coupled with the significance of ERG membership (for Johnson and Sunak) leads us to conclude that Europe remains an important political and ideological variable in the minds of many Conservative MPs. This is an important finding given that political discourse in the UK has mostly moved on from the era of politics being dominated by Brexit. Marginality was not found to have an impact on voting in the 1997, 2016 and 2019 leadership elections (Heppell and Hill, 2008; Jeffery et al., 2018, 2022), and marginality was again not found to be a factor at this leadership election. On the other hand, we find that gender is statistically significant in this leadership election—when it was always unlikely that the female candidate would win—unlike the 2016 leadership election where gender was not a determinant of candidate support, even though a female candidate had a very high chance of winning (Jeffery et al., 2018).

The October 2022 leadership election represented, in the hopes of many Conservatives at least, the end of a long period of intense leadership turmoil. After the poor ethical and moral standing of the Johnson government, and the shambolic policy execution of the Truss government, Sunak’s task was to build party unity, restore the party’s polling and govern competently. Twelve years in government had left the Conservatives open to charges of degeneration, especially when facing a Labour Party that, although not exciting the public, at least looks like a government in waiting. Although Sunak’s personal poll ratings compare favourably to Truss’ or Johnson’s in his later months (Redfield and Wilton Strategies, 2023), there has been no discernible shift in the Conservative’s poll ratings. Furthermore, Sunak has so far been unable to bridge the divides within his own party on European matters: with 22 rebellious MPs, including three former party leaders, rebelling over the Windsor Framework revisions to the Northern Ireland protocol (Parker et al., 2023). Other key legislation, for instance around self-ID for trans people, could allow for renewed conflict along the socially liberal axis of the party. While there is little appetite for a further leadership contest before the next general election, these ongoing divides coupled with the Conservatives’ continued abysmal poll ratings, means that Rishi Sunak’s time as prime minister will not be smooth sailing. While he may have ultimately gained the support of a majority of his own parliamentary party, Conservative MPs are far from united.

Acknowledgement

The authors thank Jon Tonge for his helpful feedback during the drafting process.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Conflict of interest

The authors have no conflicts of interest to report.

References

Footnotes