Still A Brexit Election?

Posted on 11 December 2019 by John Curtice

We argued at the beginning of the election campaign that Brexit seemed set to play a key role in shaping voters’ choices. At that point, the polls suggested that those who voted Leave were for the most part saying they intended to vote for the Conservatives or the Brexit Party, while those who voted Remain were inclined to support one of the parties willing to support a second EU referendum.

Since the beginning of the campaign, however, support for both the Conservatives and Labour – both of whom have been more likely to draw support from both sides of the Brexit divide than some of their rivals – has increased by five points on average, while that for both the Liberal Democrats and the Brexit Party has fallen away. Given that the movement has been towards parties with a less distinctive electorate on Brexit, we might wonder whether how people intend to vote is less reflective of their views about Brexit than it was at the beginning of the election. But is that really the case?



The table shows – separately for Remain and Leave voters in 2016 – how people said they intended to vote in polls conducted in the first week of the election campaign (see the columns headed ‘Beginning’) and how – on average – they have said they would vote in polls conducted in the second half of last week (see the columns labelled ‘End’).

A number of important patterns emerge. First, almost all of the progress the Conservatives have made during the course of the campaign has occurred among those who voted Leave. Among that group support for the party has increased by 58% to 70%. In contrast the party’s support has barely changed at all among those who voted Remain.

Indeed support for the Conservatives among Leave voters is ten points up on the position in 2017, while that among Remain voters is four points lower.

Meanwhile, the picture for Labour is almost the exact opposite. True, the party has made a little progress among Leave voters – up from 13% at the beginning the campaign to 16% now. However, its advance among those who voted Remain is much stronger – an increase from 42% to 49%.

That still leaves the party at a lower level of popularity than it enjoyed among Remain voters in 2017, but at four points the drop is smaller than the ten-point loss of support that the party is still suffering among Leave voters.

So, despite the fact that they began the campaign with some support from both Remain and Leave voters, both the Conservatives and Labour have largely only managed to increase their support among Leave and Remain voters respectively.

That of course means that the losers in the election have been the two parties whose support at the beginning of the campaign consisted almost entirely of those on one side or the other of the Brexit divide.  Support for the Liberal Democrats has fallen by seven points among those who voted Remain, while the Brexit Party has seen its support fall from 20% to 6% among those who voted Leave.

However, taken together these patterns mean that the alignment between how people propose to vote and how they voted in the 2016 EU referendum is still almost as strong as it was at the beginning of the campaign. The proportion of Leave voters who are backing either the Conservatives or the Brexit Party has only slipped slightly from 78% to 76%, while the proportion of Remain supporters backing one of the parties in favour of a second referendum is, at 80%, barely any different from the position at the beginning of the campaign (81%). Meanwhile, both figures are still up on the equivalent statistics for the last election.

At the same time, we should note the extent to which Brexit has shaped the dynamics of vote choice during the election. Most of the movement has consisted of voters switching between pro-Brexit parties and switching between pro-second referendum ones. That perhaps is the clearest testament that Brexit is indeed playing a key role in shaping voters’ choices at this election.


John Curtice

By John Curtice

John Curtice is Senior Research Fellow at NatCen and at 'UK in a Changing Europe', Professor of Politics at Strathclyde University, and Chief Commentator on the What UK Thinks: EU website.

18 thoughts on “Still A Brexit Election?

  1. Judging the result, in hindsight it would seem it wasn’t really a Brexit election as a battleground to form a government at all. The recent yougov polls showing whom voted for who indicates that Brexit was already decided by the public and that opinion polls since the 2017 election had misinformed the opposition into thinking that the public were largely okay with Brexit being stymied to setup another question or revocation – which is now a bit of a spotlight on polling credibility. To what extent do polls get used as a propaganda tool for those with power?

    As a former Labour voter, I genuinely feel sorry for Mr Corbyn being muzzled by factions within his party on the issue of Brexit. Despite him now carrying the can for the massive defeat, he’s achieved a lot in his time as opposition leader: He’s largely vanquished the duplicitous New Labour spin doctoring machine from the Labour party, he’s seen off the embers of Thatcher-ism in Cameron and May’s governments with Clarke et al being removed. He has forced the Conservatives to reverse their further leaning to the right, having won the argument to end austerity, the argument on the importance of public ownership, the argument on investment in publicly owned infrastructure projects and he’s even won the argument on healthy protectionist measures in the UK to balance great capitalism with protecting British industry and British jobs – in a post-brexit Britain.

    Much like Farage, Corbyn has probably achieved more of what he wanted to achieve by indirect means. His only true failing is that he failed to ensure that the Unite union voted and favoured leaving the EU for the last few years – which probably would have cost him his leadership much sooner – but in doing so would be making the rebuilding process for Labour significantly easier. The shortlist wouldn’t contain people like Stammer that are openly at odds with about 60% of the electorate and seemingly don’t believe in democracy. For Labour to recover from this defeat, it is probably more important that Unite re-vote and change their position on Brexit, so that no matter which Labour MP becomes the new leader, they can then do so taking an official line on Brexit that is harmonious with the electorate.

    Whether Brexit is the right choice or not, the electorate, like scientists, are entitled to set a path that might be wrong and then learn from it to inform their next choice – and this IMHO is a huge advantage of strong First Past the Post goverment compared to the middling compromised everyone-is-a-loser Proportional representation.

    If Brexit turnouts out to be the wrong choice – (I don’t think it is) as the scaremongering remainers have prophecised – then when we rejoin the EU, we will do so with a new appreciation and consent that could never have occurred by remaining in against our collective will.Report

  2. @ Simon Thorpe
    Perhaps it may help if you take note of Richard Burgon (Shadow minister) when he says ” I think the biggest mistake the Labour Party made was perhaps underestimating the desire for people who had voted Leave to leave the European Union.”
    Remainers have been very vociferous over the past 3years, but Leavers are , generally, a different type. They expect the democratic rules to be respected without further clamour on their part.

  3. Today’s featured poll (Ashcroft 11-12 Dec) asks if people voted in the EU referendum.
    5 % did not.
    54% think we should leave (combined leave and remain voters)
    40% want to remain – (combined remainers and ex-leave now remain supporters)

    This show that a ‘How would you vote…’ question is over-simplistic, in my view.Report

  4. On the very day that Boris Johnson got his “stomping” victory, the latest opinion poll that asked the question “Should the United Kingdom remain a member of the European Union, or leave the European Union?” came up with the result 54% Remain, 46% Leave (ignoring the 10% who said they didn’t know). Check the other page on this website if you don’t believe it

    I suppose that it’s quite possible that many of those people who don’t want to leave are among the 1/3 of people who didn’t vote. I guess that when pollsters ask people to respond to such a question it is quite possible that they fail to ask two other basic and obvious questions, namely, (1) are you entitled to vote? (not true if you an EU national living as a permanent resident in the UK), and (2) can you be bothered to vote?

    Maybe we need to change the way that polling works? Or even better, change the way the UK’s electoral system works.Report

  5. So by the looks of it corbyn has lost so whats he going to do for all of us that voted for him .
    Typical the rich will get richer and the poor poorer .Scotland have to get another referendum Report

  6. I got yelled at the last time I said this – three weeks ago – but I honestly think that Labour are going to follow up a massive election defeat with an epic error. They are going to blame the whole thing on Corbyn, with McDonnell in a supporting role, and then go on as though nothing major has changed.

    I think we are now going to hear a lot about how well things were going, until Corbyn made some mistakes, and so all they need to do is to elect a new leader and move on.

    That will be a huge mistake because until 10pm this evening we were hearing that Corbyn was just the boy to push through Labour’s wonderful policies, policies that the whole country was yearning for.

    But Corbyn wasn’t a cause at all, but a symptom. Labour policies since Ed was elected leader have consisted in portraying British economic health as sickness, record low unemployment as a jobs crisis, attracting 300k immigrants a year as a sign of a bankrupt country, growing as a sign of collapse, profits as a cancer.

    Labour have no positive policies, so they have adopted a nihilistic idea that whatever is, is wrong. If they now tell us that, Oh, by the way, the Corbyn Saviour figure we have been promoting for the past three years was wrong as well, what’s left for them?Report

  7. Well, the exit polls are predicting an 86 seat majority for the Conservatives. Boris Johnson will therefore be able to plough ahead with Brexit, and even a very hard one if the EU is not very cooperative.

    So, could John Curtice please tell me (and the millions of people watching the television this evening), how this squares with the 113 consecutive polls listed on his site saying that the majority of people in the UK don’t want to leave the EU? Are the people really saying that this is what they really want?

  8. John Livesay – One other comment. Yes, the polls were consistently giving a lead for Remain just before the referendum in 2016, and the result clearly surprised just about everyone. I remember Boris Johnson and Michael Gove looking totally shell shocked by the result on the morning after. This does indeed make it clear that polls are not necessarilly good predictors. But it is also fair to point out that the referendum result was probably also influenced by some pretty devious tactical targeting of particular groups – remember Cambridge Analytica etc. In addition, it could well be that some voting was motivated in part by a desire to give the “establishment” a kick in the teeth. Such behaviour is encouraged if you believe that the actual result will be Remain because the polls have been giving a highly consistent prediction for months. It’s also a factor in Trump’s surprise victory when everyone was saying that Clinton was almost sure to win.

    These are yet more reasons for believing that the solution will not be found in polling, but in a referendum where everyone is fully informed about the options and implications. That was definitely not the case in 2016.Report

  9. Specific request to John Curtice. Could you please comment, either on the BBC, or here on your website, on the remarkable fact that the last 113 EURef2 polls have all given Remain in front. Do none of the polls simply ask the people who they are polling whether they are allowed to vote? If the polling does include the 3 million EU nationals who are resident in the UK and who were not allowed to express their view in 2016, this could indeed explain the discrepency. But that would demonstrate that polls in the UK are done very badly. It would be like asking people under 16 what they think and including those results in the data.

    That said, I think that it is a scandal that the views of those people was ignored. I also think that it was unfair that UK nationals like myself, who live elsewhere in the EU were ignored too. But I’m prepared to accept that the rules are the rules. But that string of 113 polls certainly convinces me of the absolute neccesity of putting the decision to a second referendum, and not relying on the results of today’s general election to make such a momentous decision.Report

  10. John Livesay – you suggest that if you prefer Leave, your only option is to vote for the Tories (or maybe the Brexit party). But that assumes that you think that Boris Johnson’s deal is the best possible one. Labour is proposing to renegotiate the deal, and then offer the second referendum. Why is that such a daft idea? If they come up with a much improved proposal, then maybe the string of 113 polls that consistently demonstrate that the majority want to Remain would come to an end. That would be finally demonstrated once and for all in the second referendum. That’s not such a bad prospect, even for those who want Brexit.Report

  11. Paul – to the best of my knowledge, there are no restrictions on who is allowed to organise a poll on the Remain/Leave question. If UKIP, the Brexit Party, or the Tories had paid for an organisation to do a poll, their results would have been included here, because this website is described as non-partisan. There are several different polling organisations who have been involved in generating the string of 113 consecutive polls since 16th March 2018 that have all given Remain a consistent leave. So, I really think that it is difficult to imagine that this is not a reflection of the real state of affairs. One poll done according to a selection process that you would find acceptable would give a least one counter example. There are none.

    Yes, the UK population is split roughly half and half on Brexit. But that split is apparently very clearly and reliably slightly in favour of Remain.

    What I find incredible is that this run of 113 polls does not get mentioned anywhere. John Curtice, who was on the Andrew Neil show last night and will be very much in evidence on the BBC during today’s election, has not (to the best of my knowledge) ever mentioned this fact – even though it is on his website! Even Jo Swinson and the Lib Dems, who are explictly pushing for remain, don’t seem to have mentioned this remarkable run of consistent results.

    Now, I certainly don’t want to claim that you can use a string of 113 polls that have given Remain an average advantage of 5.6% over Leave to make an important decision. The only acceptable way to test that would be to hold a second referendum. It seems quite possible that the UK’s dismal First Past the Post voting system will give Boris Johnson the majority that allows him to push through his deal. But if that happens, it will be in the face of a string of 113 polls that have said that this is clearly not what the majority want. That result would scar the nation for decades to come. Faith in our political system really would be crippled.


    1. Simon Thorpe – maybe you can advise me here. I know that on-line polls generally use a database of people who have signed up to (and get rewarded for answering surveys). I am on one such databse myself.
      Do you not think it likely that there may therefore be an introduced bias amongst this group – a certain type of person?Report

  12. @Simon

    I think you’re over stating the validity of opinion polling if you are suggesting they would be a good substitute for actual polls. We were never asked for consent to use our sovereignty to join the EU, and were only given the 2016 referendum because it was believed (statistically 🙂 ) to be a rhetorical question, this many years on from joining. So even the word ‘Remain’ in the context of Brexit is a misnomer. We have only ever been asked the question about EU membership once, and that was in 2016 when we voted ‘No’ by a significantly larger number of voters (3-5Million) more than have elected any UK government in looking as far back as 1945, and certainly more than any since we joined the common market and had EU treaties later signed in our name without consent by said goverments.

    Assuming Johnson is returned with a good majority he will have a double mandate to untie us from the shackles of EU subservience – one from the referendum and one from the GE. Whether we have further vote/s to rejoin the EU is a completely separate matter entirely.

    Our democracy will be lost and damaged for a long period if the losers of the 2016 referendum don’t experience the political loss and accept their defeat. Even if we had another referendum you can’t win the argument; that’s been clear since you lost. The EU’s actions and statement since 2016 strengthen UK nationalist thinking, not least the recent sound-bite from Barnier being distasteful electioneering that breaches the EU’s own rules on interference. Leave voters and politicians continue to make the argument for leaving stronger, even if only half-heartedly now, with stuff about women’s sanitary products being EU taxed or shark fin soup barbaric practices being impossible to outlaw, but that’s understandable because why would the side that won the argument need to up their game before they needed to?

    On the issue of opinion polling reliability, it occurred to me that the reason the brexit polls are probably so far out(and got the EuRef1 wrong) is probably because pollsters fit the ‘remain’ voter demographic (like BBC news staff did/do according to retired BBC veteran John Humprys) and so the hidden bias in the polling is at the staffing level. And even though the country are more for UK independence, unless pollsters were randomly selected from the population to avoid skewing pollster staff brexit views, bias was going to creep in to the poll results. Comparing brexit polling to a non-brexit-general-election of the past, the polling staff were probably better aligned in voting intention to the population and so despite a greater percentage of pollsters being Conservative voters(in all likelihood), there wasn’t a big enough pollster view bias to bias the polling methodology towards inflating Conservative poll numbers.Report

  13. I just realised that there is something even weirder than I thought about Simon Thorpe’s proposal. If I want Brexit, am I really going to vote against the Tories in order to sabotage a deal that has already been made with the EU, so as to string out Brexit for more years and probably for good? I’m going to put Brexit at risk to salve his “festering sore”? Because a Remainer modestly and innocently proposes that I ought to?

    I may be just a simple little voter, but I’m not that daft.Report

  14. Simon Thorpe should remember that the polls gave a 52/48 majority to Remain the day before they lost the referendum by the same margin, It ought not to be necessary to point it out, but polling is a rough indication of the wishes of the voters; only the votes themselves tell us what the voters really think.

    Oh, and on the topic of the polls telling us the same story for three years, the Remainers have been making the same bad argument about polls for the same length of time.Report

  15. I’ve just posted this comment on another section of this website, but I think the points are very relevant here.

    I just downloaded your complete set of EURef2 poll data. The conclusions are absolutely clear. The last time Remain and Leave tied was on the 16th March 2018. Since that time, there have been no less than 113 consecutive polls that have all given a majority for Remain. Not a single one has put Leave ahead. The average vote has been 52.8% for Remain, to 48.2% for Leave. That’s a difference of 5.6%. And this site is giving its running poll of polls at 53% to 52% – very close to the average of the 113 polls.

    Surely, noone should be allowed to claim that Boris Johnson would be respecting the will of the people by forcing Brexit through – even if he did get a solid majority. If we want to heal the divisions that Brexit has produced, there is only one solution. It is to have an second referendum. If the majority is in favour of Johnson’s deal (or some Labour negotiated variant), then so be it. But that won’t make that run of 113 consecutive polls all saying people want to remain go away. The only explanation that I can see would be that opinion polls also take into account the 3 million UK residents who are prevented from expressing their opinion. That would explain the discrepency. Maybe Sir John Curtice would like to comment?? But if that is the explanation, it is extremely unfair. People who have been living in the UK for years and paying taxes should have a say as well.

    In the meantime, I would plead with everyone in a position to reduce the Tory majority by tactical voting to do so. Even if you want brexit, you ought to agree that a second referendum is the only way to make the festering sore demonstrated by that string of 113 polls go away. Report

Leave a Reply

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