LEADERSHIP POLL: Swinson ahead of Davey

Over the past week, I have been running an informal poll for the 2019 Liberal Democrat leadership election. I did the same in 2015, and came within 2% of the correct result, so thought I’d have another go this time round.

The results are now in, and before I go into them, I’ll just put a caveat here that although I have tried to do demographic weighting as accurately and diligently as possible, because I don’t have precisely accurate membership data (and have had to rely on the QMUL Party Members Project for information), the weightings may be slightly out. As such, this poll should not be relied upon as a formal, professional-standard poll, but its results should be relatively indicative.

I surveyed 846 participants between 24th June and 1st July 2019. This gives a theoretical margin of error of 3%. The results are weighted based on gender, age and joining period, as well as implied 2015 leadership vote based on stated ideological self-identification, corroborating the data with my 2015 leadership poll. This last weighting mechanism replaces a question about past vote (simply because more than half of members today were not members in 2015). Because it is a more experimental weight than the other three, I have also run the data without it: the results of that are under the headline section.

Full tables are available here.


The headline figures are Jo Swinson on 54%, and Ed Davey on 46%. This is a narrower race than some were expecting, but still a relatively comfortable margin for Swinson. Along with Mark Pack’s survey putting her on 60% (albeit with a large number of undecideds), my poll seems to justify her status as favourite for this race.

NB: Even without the weighting by implied 2015 leadership vote, weighting only for gender, joining period and age, Jo Swinson still leads with 53% to Ed Davey’s 47%.


Perhaps most interesting are the demographic trends.

Men appear to be voting overwhelmingly for Jo Swinson, with 62% of them joining Jo, and only 38% going wavy for Davey. Women, on the other hand, are splitting in the opposite direction, contrary to what many expected at the beginning of this campaign. 57% of them are voting for Ed, while 43% of them are voting for Jo.

The other trend to note is one of age. A whopping 68% of 18-24 year olds say they will be voting for Jo Swinson – perhaps reflective of her much-publicised time involved in LDYS and her frequent appearances at Young Liberals Conference. On the opposite end of the spectrum, a massive 78% of those over 65 say they will vote for Ed Davey – which could be crucial, as over 65s make up the largest proportion of the party’s membership. If Davey manages to turn them out en masse, while youth voters stay at home, his big margin here could be the difference. The trend is the startling thing however – in each category, the older you are, the more likely you are to be voting for Davey. The only exception to this is under 18s, who Ed is winning handily – but this should be caveatted, because the raw sample is so small that it may not be accurate to draw statistical conclusions from it.

While there isn’t any reliable data available for sexuality to be able to weight based on it, some interesting trends do emerge from the raw data. Amongst heterosexuals, Ed Davey has the edge with 52% to Jo’s 48%. With those self-identifying as LGBTQ+, Jo Swinson is ahead with 66% to Ed’s 34%.

Finally, amongst the BAME community, Jo is ahead with 53% to Ed’s 47%, though again, and unfortunately, this should be caveatted due to a small sample size. The low number of BAME respondents to this poll was of deep disappointment to me, and shows that we need to continue to work to make our party reflect the wide diversity of the country we’re seeking to represent.


Unlike in 2015, where my poll showed that Norman Lamb won a clear majority of those identifying themselves with the ‘classical liberal’ tradition, and Tim Farron won a huge majority of those identifying with the ‘social liberal’ tradition, the picture is much more mixed.

Those placing themselves in the left wing of the party make up 38% of the membership, according to my poll. And 64% of them say they will be voting for Jo Swinson, which is a big lead. There has been much discussion of the Coalition legacy at this election, and with two Coalition ministers fighting it out, the belief was at the beginning of the race that there was no clearly ‘left’ candidate. But for whatever reason, Jo is seemingly on course to win this essential bloc of voters.

Those placing themselves in the centre of the party are the largest share of the membership, making up 47% of the party. Ed Davey is on course to win 54% of these voters, which is another piece of evidence to suggest that the race will be closer than anticipated. Although this margin is much narrower than Jo’s popularity with the left of the party – and thus doesn’t make up for the size of her win there – if Ed can consolidate this lead with the party’s self-identifying centrists, this could be his path to victory.

Those placing themselves on the right of the party – the aforesaid ‘classical liberals’ – make up a smaller portion of the membership compared with 2015, with only 15% belonging to this group (cf. 22% in 2015). Among these voters, Ed and Jo are effectively tied – Jo ekes ahead with 51% of the vote, while Ed is on 49%.

In terms of joining date, Jo is winning with every group except those who joined pre-Coalition, where she is tied with Ed on 50% apiece. Jo’s lead by 65% to Ed’s 35% amongst those members who joined during the Coalition is particularly notable in the context of Ed’s decision to campaign hard on his record as a Coalition minister. Of post-Coalition joiners, the gap is at its most narrow amongst the 20,000 who joined in the 2019 surge, perhaps reflecting the fact that they are by a long way the most centrist group of new joiners (with a full 50% of 2019 joiners identifying as centrist rather than left or right of the party).

2 thoughts on “LEADERSHIP POLL: Swinson ahead of Davey

  1. The s issue about turnout of the over 65s voting while youth voters stay at home is surely irrelevant where a considerable proportion of the electorate will have voted electronically soon after receiving their e-mails. Indeed evidence elsewhere would be that younger voters are happier to indulge in on-line voting than older voters. However, I do not know what percentage of members now have an e-mail address registered with the party.

Leave a Reply

Your email address will not be published. Required fields are marked *