What can we take from the MRP seat projections?

13 seats.

That’s what the  You Gov MRP seat projection says we’re going to get.

Not what we wanted to hear.  YouGov only spoke to a small number of people in each seat. Our campaign teams on the ground will have spoken to thousands more people over the last few weeks. They are likely to have a much more accurate idea of what is going on and I think that they will find these projections surprising.

But note the caveat from YouGov

The idea behind MRP is that we use the poll data from the preceding seven days to estimate a model that relates interview date, constituency, voter demographics, past voting behaviour, and other respondent profile variables to their current voting intentions. This model is then used to estimate the probability that a voter with specified characteristics will vote Conservative, Labour, or some other party.

This is a Brexit election though. Brexit is an issue which has split the country into Leave and Remain voters. Traditional patterns of voting for each party may well not apply.

I’m grateful to Morgan Griffith-David for doing this work for me so I don’t have to, but the poll only gives us 13 seats, it puts us in touching distance in another 23. This should concentrate the minds of pro-Remain voters. For example, Brecon and Radnorshire is projected as Conservative 49%, Lib Dem 35%, Labour 14%. It doesn’t take a genius to work out what Labour voters need to do to stop the Tories, whose vote is not split by the Brexit party this time.

These, and others, are the seats where Remainers really need to get behind the Lib Dems to stop Boris Johnson getting a majority which will enable him to inflict Brexit hell on the country.

Its top of the range guesstimate for Labour is well short of any sort of a majority for Jeremy Corbyn. So many people, Labour and Tory voters alike, do not want to see him in Downing Street. If there is no prospect of that happening, they may well feel that they can safely vote for the Lib Dems. Sky’s Lewis Goodall made the point that Conservative voters in the home counties are not happy:

Stephen Bush from the New Statesman agrees with that analysis:

If you are a Remainer who lives in Winchester, Cheltenham, South East Cambridgeshire, Cheadle or Guilford, all of which are within striking distance according to this, and would prefer to switch your Conservative MP for a Liberal Democrat one, but don’t want a Labour government, then this model is a pretty powerful argument to the Liberal Democrats that they can do so safely.

Former Lib Dem communications chief Phil Reilly noted:

The only thing that’s clear is that it shows that many seats are still in play. You can plough your way through the csv file here. 

Bear in mind that this is a projection and there are still two weeks to go.





* Caron Lindsay is Editor of Liberal Democrat Voice and blogs at Caron's Musings

Read more by or more about , or .
This entry was posted in News.


  • Only two weeks till we find out…

  • The MRP confirms that neither our campaign nor our leader have had the impact we hoped. Plenty of stuff for the post-mortem.

    The potential failure of the model to pick up local campaigns was a point made over and over by LibDems when the first of its kind survey came out in 2017. And of course the MRP was right and we were wrong.

    The good news is that we have recovered good second places in a batch of seats, mostly where we have been traditionally strong, plus a few obvious remain surges. The key now is to focus our campaign and use this information to try and beat the Tories in those seats.

    There are two weeks to go. In any previous information, being told which of our targets were way behind (some still in third place) with so much time to cut our losses and focus our efforts elsewhere would have been golddust for which we would have paid handsomely. Let’s hope we don’t waste the information!

  • “This is a Brexit election though. Brexit is an issue which has split the country into Leave and Remain voters. Traditional patterns of voting for each party may well not apply.”

    While for some people it will affect who they now vote for, so long as Brexit opinion is correlated with some measurable demographics it should be possible to account for that.

    The most obvious correlations are with Age and 2016 Referendum Vote, which are both used as inputs to the model – see https://yg-infographics-data.s3-eu-west-1.amazonaws.com/ZAfbtHgj42wx4reHnaMtbBamoKdMxkFMpz4gnWMjiZCUAxDX66MsCB38K/2019_data/MRP_Tables_2019_Election_Public_Release.pdf for the detailed tables – so it should be okay there.

    If you look at table 10, it has the Lib Dems picking up 21% of those voters who voted Remain in 2016 but Conservative in 2017, for example.

  • Yeovil Yokel 28th Nov '19 - 8:43am

    I don’t have time to look at this in more detail, and I accept that very sophisticated modelling was used for this poll, but I have two problems with it:
    (1) as Caron alludes to above, this poll is making a projection two weeks ahead based upon data which is one week old, in a situation which is dynamic;
    (2) Yeovil doesn’t even make it into the ‘Near LD misses’ above, and yet we know here on the ground that it’s a seat very much in play.

  • Bill le Breton 28th Nov '19 - 8:59am

    Let’s not lose heart.

    Interestingly, if you believe in the wisdom of crowds, the index betting sites have been selling (ie the buy price) Labour at 210 seats for a week now, righton the MRP projection. It is the index for LD seats @ circa 40 that is at most variance with this MRP projection.

    A friend tells me we are 2nd in 100 seats.

    I can see the line that Stephen Bush writes above and I remember using exactly this message to Conservatives in the Isle of Wight in 1983 when we faced a Thatcher landslide …

    But this also needs to be a wake-up call to us to persuade Labour Remainers in those seats in which YouGov are saying we are second to Conservatives.

    Can Morgan tell us which these are?

  • “For example, Brecon and Radnorshire is projected as Conservative 49%, Lib Dem 35%, Labour 14%. It doesn’t take a genius to work out what Labour voters need to do to stop the Tories, whose vote is not split by the Brexit party this time.”

    And what is our message to those labour voters? Nationally at least it seems to be that we wont work with Corbyn under any circumstances, but we will flirt with the Tories if they offer a referendum?

    That might secure Winchester, South East Cambridgeshire, Cheadle or Guilford but is it costing us elsewhere?

    This isnt (enough of) a Stop Brexit election for most people – it’s a Stop Corbyn one and I dont think we help our own cause by joining in with that (Tory) narrative.

  • Jonathan Linin 28th Nov '19 - 9:05am

    The EU referendum was 48-52, the 48% then disregarded with no attempts at bridge building or a softer Brexit as the majority trumps everything. So Boris might win a large majority on, guess 42-44%. How long before he calls it an overwhelming endorsement from the British people ?

  • Richard Underhill 28th Nov '19 - 9:26am

    There is no official Brexit candidate in Tory held seats but The Times of Tunbridge Wells (free on Wednesdays) shows two Independent candidates, both Leavers, one of whom says he left the Tory party because of the re-adoption of Greg Clark as Conservative candidate. The Labour candidate is from London. Greg Clark is quoted as saying he voted Remain in 2016, voted against his party to prevent No Deal, but will back Boris Johnson’s deal.
    In 2016 the local council area voted Remain as I blogged on LDV at the time. We were the only area in Kent to do so although Tonbridge and Sevenoaks are closer to London.
    We heard Nigel Farage on the radio saying that a Remain win looked likely (nationally) but he would continue his campaign for Leave. He now argues that we should do the opposite of what he said he would do.

  • “Our campaign teams on the ground will have spoken to thousands more people over the last few weeks. They are likely to have a much more accurate idea of what is going on and I think that they will find these projections surprising.”

    This is prety much what was said all the time in ’15 – that polls didn’t reflect the vibe on the doorstep. I wonder if the art of rigorous canvass analysis has been lost over the years.

  • The media are pushing the Presidential two horse race narrative. Ironically for many it is effectively signing their own death warrant, won’t be much left of the BBC after five years of Depeffle. The rest of the electronic media will struggle in a country getting poorer and isolationist. The dead tree media have speeded up their own demise and reduced their chance of joining the electronic world. Turkeys voting for Christmas just about sums them up.

  • nigel hunter 28th Nov '19 - 11:13am

    The trouble with polls is that they can INFLUENCE people who to vote for. The info is then transmitted to the right wing papers who can then manipulate the voter into voting the way the papers want which is shown in the final election results. Being a week behind in publishing it can be that weeks polling that influence the result.WE HAVE EVERYTHING TO FIGHT FOR RIGHT UP TO THE CLOSE OF POLLS

  • Do the pollsters only contact registered voters? If so, the late surge in registrations (2m or so) might not be reflected in the current findings – but will start to impact later.

  • Yousuf Farah 28th Nov '19 - 12:37pm

    If this poll is even remotely true; then it will reflect a complete failure of tactical voting. Originally this was an idea pushed by the Corbynistas, and since they’ve been controlling the narrative in this election, and making gains in the opinion polls, they’re now saying vote Labour. This MRP poll could make them think twice, but I heavily doubt that, like I’ve always doubted that they were a remain party, or that they ever wanted to vote tactically in the first place. Tribalism and bitterness goes hand in hand with those lot.

    Though on the bright side, this poll might make remainer Tories feel more free to cast a vote for the Lib Dems. But still, if people don’t tactically vote, then Boris, Gove and Priti will have five years with a sizeable majority. I see Lib Dems voting tactically, I just don’t see Corbynistas doing it.

  • I think Wokingham – which voted 58% Remain and is represented by John Redwood- might be considerably tighter than MRP is predicting. Certainly the Liberal Democrats are comfortably winning the stakeboard count.

  • Agreed Yousuf.

    We will be involved with or read about many conversations about tactical voting, but it is only ever going to be a small proportion of voters who think to do it and then half will be confused about how best to do so.

    I also suspect you are right that many see it as a one way street, or using the lack of clarity in tricky to determine seats as an excuse not to support it anywhere even where we are the clear front runner.

    I only hope that this poll is a wake up call for those convinced that we are the enemy and that while attacks on us might boost the Labour vote share, it mainly helps the Tories get a majority.

  • Sacha Griffiths 28th Nov '19 - 2:04pm

    The revoke Article 50 policy is fine if you really promote it. The Swinson campaign has been really lack lustre. The ‘B*****ks to Brexit’ in the Euro elections was bold, strident and attention grabbing. We should also point out the £50bn cost of Brexit so far and the £33bn of the Bojo deal. The Lib Dems do not want to waste your money on this Brexit nonsense. That is money not spent on the NHS and all the rest of it. Equally, a People’s Vote where you put on the table the 4 options of Remain, No-deal, Customs Union Deal and the Bojo deal in an elimination vote referendum similar to Strictly or I’m a Celebrity would have broad popular public appeal

  • Paul Murray – In the 1980s when I lived in Richmond on Thames we always seemed to win the skateboard war by a country mile but the Tories always seemed to win the vote.

  • @Paul Murray

    Again in 2015 there were some pictures of some really impressive stakeboard displays in several seats (including in Wells which I knew from working there in the past) posted on social media with comments about how the polls weren’t reflecting this. We lost all of them.

    Whether that will happen this time I don’t know – things can change. But in a run-off between polls and feel on the doorstep, polls won hands down in 2015.

    My other point would be people who are saying two weeks to go. There isn’t. One pretty consistent pattern is that the Lib Dems poll in real votes lower than their opinion polls on the final weekend. The last week squeeze is definitely a thing. SO if you aren’t ahead on the final weekend it is IMO very hard to reverse that.

    If (big if) last minute tactical voting shifts happen that could be different this time. But how many places have the data to do that.

  • Peter Watson 28th Nov '19 - 4:51pm

    @George Flaxman “In the 1980s when I lived in Richmond on Thames we always seemed to win the skateboard war by a country mile”
    Now that’s a race I’d like to have seen!! 😉

  • Tony Greaves 28th Nov '19 - 9:10pm

    This poll at this election is very substantial bollocks. It may work for the typical north of England old industrial towns for instance but it will not work for seats where there is a very recent and very different contest going on (as in some London seats). I am not saying we will win 50 seats but it will be rather more than 13. Also it’s well known that where there is a strong squeeze on third and fourth parties it tends to happen in the last few days.

    This is not to say that we are not suffering badly from the failure of the national “air” campaign and the vicious attacks on us, notably from Labour, on social media. I am sticking wi9th my original prediction.

  • What we are looking at in these comments is an object lessons in how easy it is to manipulate people with disinformation. These polls may be accurate, but equally they may not. A lot of opinion polls historically have proved misleading. To say that the same methodology got it right once is hardly replication. It needs to be tried dozens of times before we can consider it reliable.

    Reasons have already been given as to why we should treat estimations of Lib Dem support in individual constituencies with caution. Here his another. A number of long-term trends can be identified that affect voting behaviour and are not necessarily picked up by pundits. For instance, there is a long-term trend for Tory support to decline in major urban areas. That is why I am more hopeful about our London targets than targets elsewhere. Similarly, there is a long-term trend for older people not to support us. That is causing problems in the West Country (outside Bath and Cheltenham), and could conceivably cost us Eastbourne and North Norfolk.

    All the critics can do is carp and say “I told you so”. Do they have any useful suggestions? Can they do things better? The national campaign has already been readjusted in ways that will make it more effective. Firstly, going for Johnson on character. He has no answer to that. He is a proven liar, and he has published numerous comments that show him to be an unpleasant right-wing snob. How can you trust a serial liar? Do you want a Prime Minister who clearly regards ordinary people with contempt? Secondly, more emphasis is now being given to a People’s Vote rather than revocation. It is easy for people to become confused by our position. Thirdly, there is now a greater focus on the environment, which will boost support among younger voters and some soft Tories. The Labour Party is also giving us a helping hand by shifting more decisively towards Brexit and effectively withdrawing its campaigning efforts from its weaker areas.

    Even if this poll is 100% spot on, which I doubt, it is insufficient reason to give up and go home, which is what the critics appear to think we should do.

  • Richard Underhill 29th Nov '19 - 10:21am

    YouGov are right to exclude Northern Ireland on the grounds that a different election is happening there, but wrong not to do a separate survey.
    The new Speaker is ex-Labour.
    They say that they spoke to 100,000 people in 632 seats, so the mean (average) size of each constituency sample is 100,000/632. Not very many really.

  • Richard Underhill 29th Nov '19 - 11:09am

    One of the pollsters was on BBC tv Politics Live. Britain Thinks. She said that Boris’ Leave slogan comes back to them in focus groups.
    His refusal to accept the rules about time when debating Corbyn one to one was followed up in headlines from his friends in the press. He plays by the rules of the Eton Wall Game (if any).

  • Neil Sandison 3rd Dec '19 - 11:02pm

    What we need in these final days is a hearts and souls message for those with a liberal disposition about what they will lose under a Conservative /Brexit led government when Farage pulls Johnsons strings on issues like migration ,international trade and the removal of workers rights . A Singapore on Thames econony dictated to by Uncle Trump . A reminder about why the governments of europe came together to build trade and commerce and resist walls and barriers to liberty and freedom .Hearts and Souls and not facts and figures can help us cut through this sterile contest .

  • This hasn’t aged well either.

    Again making easy conclusions based on ‘the feel on the ground’. This was apparently the most data driven campaign the party has run. But if that data is rubbish then so will be the conculsions. Was canvassing accurate or not analysed right,

    There is little value in thousands of people out on the streets if what they are doing is of little use.

Post a Comment

Lib Dem Voice welcomes comments from everyone but we ask you to be polite, to be on topic and to be who you say you are. You can read our comments policy in full here. Please respect it and all readers of the site.

This post has pre moderation enabled, please be patient whilst waiting for it to be manually reviewed. Liberal Democrat Voice is made up of volunteers who keep the site running in their free time.

To have your photo next to your comment please signup your email address with Gravatar.

Your email is never published. Required fields are marked *

Please complete the name of this site, Liberal Democrat ...?


Recent Comments

  • Roland
    Re 2008 financial crash… We should not overlook the role the Conservatives played, by making the UK economy overly dependent upon financial services, resul...
  • Martin Gray
    'Sunak fits well into the neocon mould of Meloni, Farage and Trump – Politicians who believe their own evil myths of Western triumphalism and denialism, who a...
  • Peter Martin
    @ David Allen, " Like Scholz in Germany, {Starmer} risks becoming paralysed by his lack of ambition. But what’s the alternative?" As this is ...
  • Peter Martin
    @ David Allen, " Like Scholz in Germany, {Starmer} risks becoming paralysed by his lack of ambition. But what’s the alternative?" As this is ...
  • Peter Martin
    "I know perfectly well the financial crash in 2008 was not NL’s responsibility." It was mainly the result of what happened in America but it was...