What do we know about the Conservative Party leadership election? [Second Round]
This information is correct as of just after the start of the second ballot — 15:45, 18/06/2018. The data (and hence the graphs) may be updated as more information becomes available, but the text will not.
If you’re interested in reading an academic journal article I’ve written about the 2016 Conservative leadership election, click here!
My blog on the first round of the contest can be found here.
The first round of the Conservative Party leadership election saw seven candidates make it through to the next round: Boris Johnson (114 votes), Jeremy Hunt (43 votes), Michael Gove (37 votes), Dominic Raab (27 votes) Sajid Javid (23 votes), Matt Hancock (20 votes) and Rory Stewart (19 votes).
Andrea Leadsom won 11 votes, Mark Harper won 10, and Esther McVey came last with 9. As none of these three won over 5% of the vote, they were eliminated. Matt Hancock then decided he would withdraw, and backed Johnson — almost certainly because he shared his vision (despite running against him) and not because he wants a job when this is all over.
This blog will outline the geographic, social, and ideological splits in support as it currently stands, drawing on Guido Fawkes’ collection of public backing (data).
In terms of public declarations, Johnson is clearly in the lead — with 105 MPs publically backing him, he is guaranteed to make it to the members’ ballot round.
Changes in Support from Round 1 to Round 2
In terms of former candidates, we can see a general seeping of support to Boris Johnson — he certainly had the momentum. Matt Hancock’s support has has split: 8 of the 17 public backers are now undeclared, 4 have gone to Johnson (including Hancock himself), but 5 have switched to Rory Stewart. None have gone to Sajid or Hunt, as the media speculated they would.
Leadsom’s public backers broke 3 for Johnson and 1 for Javid, whilst of McVey’s 6, 4 went to Johnson, 1 to Javid, and 1 is undeclared. Harper’s 8 backers went 5 undeclared, 3 to Johnson.
Interestingly, while Hunt, Raab, Javid and Stewart have all held on to their first-round support, Gove lost an MP — Bob Seely — to Johnson. That must hurt.
Map of Support
So far, it seems that support does not have a geographic bent to it — although if you can notice one, do let me know! (You can click through to the full map here. Hold down Ctrl to zoom.)
For all of the tables below, hover over the bars to see the percentage share — and you can click on a header to sort by column. Nifty.
We can see that Johnson is languishing in terms of support from women — his support base is 14% female, compared to a party average of 20%. In contrast, Gove, Raab, Javid and Stewart all draw a greater share of their support from women than the party as a whole.
Just 13% of Conservative MPs represent constituencies in the North of England, despite accounting for 24% of constituencies.
Middle and South England throw up few surprises — Raab underperforms in the former and overperform in the latter, whilst the reverse is true for Javid and Stewart.
Gove’s support is Scotland-heavy — winning nearly 50% of Scottish MPs at the moment — he fails to win any support in Wales. Of the 8 Welsh Conservative MPs, 3 have backed Johnson, 1 Raab, 1 Javid, 1 Hunt, and 2 are keeping us guessing. For now.
Who are new MPs supporting?
The most interesting thing about this chart — in my opinion, anyway — is that Gove is drawing 27% of his support from MPs who entered the Commons under May. Javid’s support is the most “Cameronite”, with 81% of his support entering the House when Cameron was party leader, and interestingly Stewart’s support is mostly pre-Cameronite — 57%.
There are clear brexit divides here: Stewart’s support is the most Remain-y, followed by Hunt, Javid, and Gove. This is interesting, since Gove led the campaign to leave the EU.
The PCP was split 45/54 in terms of Leave/Remain. Only Boris Johnson and Dominic Raab are more Leave-y than the parliamentary party at large, with Raab’s support being the most Leave-heavy (although it’s worth remembering Johnson has more leave MPs supporting him than Raab has public backers at all).
The next measure is a Euroscepticism Score, based on the series of indicative votes held in the House of Commons at the end of March. This is taken from the Twitter feed of Alexandre Afonso:
I’ve taken this data and put it in the below chart. You can click on the labels at the top to remove Remain or Leave MPs from the chart.
Interestingly, we can see that of the main candidates, Raab has the most concentrated support — he draws overwhelmingly from those who have a higher Euroscepticism score. Johnson has the widest spread, with Gove failing to win over the most Eurosceptic MPs, and he shares the same space as Hunt.
This chart shows the Eurosceptic score of each candidate’s supporters, with the colour representing their own Eurosceptic score. We can see three clear patterns here: Raab and Johnson are the most Eurosceptic, Hunt, Gove, and Javid’s support is all broadly as mildly-Eurosceptic as each other’s, whilst Stewart’s support is clearly and unashamedly remainy.
This seems to represent the key divide in this contest.
Raab and Johnson seem to have won over the largest chunk of ‘no meaningful votes’, whilst Gove, Hunt, Hancock and Stewart are supported by those who were most likely to consistently support it.
Interestingly, most of the undeclared candidates voted in favour of the deal — this might present a challenge for Raab in appealing to more MPs.
The Payroll Vote
Javid and Stewart seem to be the candidates of the payroll vote — despite making up just 30% of Conservative MPs, they make up 48 and 43% of their support respectively.
Raab, on the other hand, is the candidate least supported by the payroll vote — something he might see as a badge of honour — with just 4% (1 MP) of his support coming from those who take the whip.
This measure of loyalty is taken from The Public Whip, which measures it as:
a vote against the majority vote by members of the MP’s party. Unfortunately this will indicate that many members have rebelled in a free vote. Until precise data on when and how strongly each party has whipped is made available, there is no true way of identifying a “rebellion”. We know of no heuristics which can reliably detect free votes.
So, in a situation where the Government whips one way, but backbenchers are having none of it and rebel in greater numbers, then it is the Government MPs who will be classed as rebels. Until better data becomes available, the below chart should be read with that in mind. Link here.
Joyously, I’ve had to delete Ken Clarke from this analysis because he rebelled so much that he skewed the whole chart. His rebellion rate was 24%. YOLO indeed.
Stewart’s tendency to pick up rebellious MPs is down to MPs who back remaining in the EU having to vote against the majority of their colleagues when they rebel on key Brexit votes.
Confidence Vote on Theresa May
This is another polarising metric: Stewart’s support is made up solely of MPs who voted to support Theresa May, whereas just 28% of Raab’s support is. Again, the MP closest to Raab is Johnson.
Javid, Gove and Hunt again draw from a similar profile of MPs — it is hard to see how they are distinguishing themselves here.
European Reform Group
We see the Brexit divide playing out here too — Raab support is the most ERG-heavy as a share of total support, whilst Johnson comes a close second.
Javid, Gove and Hunt have broadly similar levels of ERG support, whilst Stewart has no ERG MPs supporting him. Funny that.
2016 Leadership Vote
There are a few interesting facts here — firstly, Raab’s support is more 2016-Gove-heavy than Gove’s own support, whilst in terms of raw numbers Boris has more 2016-Gove backers than Gove has.
Both Raab and Johnson are drawing on Leadsom’s 2016 support to make up 1/3 of their support (despite neither of them being mothers), whilst Stewart, Javid, and Hunt have primarily 2016-May-backing support.
Second Placed Party
In terms of the second-placed party in their supporters’ constituencies, we see that Gove — through having the support of the most Scottish MPs — has a larger than average proportion of ScotNats in second place.
Javid has larger proportion of Liberal Democrats coming second place too.
There are three key clusters in this election so far, none of which are surprising.
Firstly, Rory Stewart is unashamedly drawing from the remain section of the party.
Secondly, Boris Johnson might have a greater level of absolute support, but his support is more Brexit-y than the party’s average, and he is threatened on his Brexit flank by Dominic Raab, who has an even more Brexit-y level of support.
Finally, Gove, Hunt and Sajid seem to be competing on the same soft-Eurosceptic ground. They are not out of step with the party on average on most issues, but it is unclear what exactly it is they are offering.