EU referendum roundup – is there a clock ticking?

Quite a few things to talk about tonight! We’ve had some new polls in the last couple of days:

The debate about methodology has rumbled on, with ICM’s Martin Boon describing methodological changes designed to counter response bias. Under the old method, quotas would fill up early, so created a bias towards early respondents. Under the new method, response time weighting will reduce the impact of this. In today’s poll, it produced a dead heat, which under the old methodology would have been unchanged with a four-point lead for Leave. YouGov responded with a statement to say that its current methodology wasn’t affected by the problem.

I made my current thinking public earlier, with a review of YouGov’s analysis of the Polls apart polls. To summarise, although the YouGov piece does highlight some important issues, it overestimates of the impact of the oversample of graduates in phone polls, and there are questions to be asked about its own phone poll. So I stand by the analysis in Polls Apart.

As for the state of the race, the NCP polling average now shows about a 6 point Remain lead. Based on the Polls apart analysis, and adjusting to include Northern Ireland, this translates into an 8 or 9 point Remain lead “nowcast”, with a central forecast unchanged at 56.6% for Remain, though with a great deal of uncertainty, due both to potential polling error and genuine changes in public opinion, in either direction.

Therefore it’s certainly not game over – Leave could still win this. But compared with the last forecast, there is less time for a surprise, so the probability of Brexit ticks down further to 17.4%.

Is that symbolic of time running out for the Leavers? Perhaps. On the campaign trail, Vote Leave has announced via Newsnight that its focus in the final month will be on immigration and not the economy. This is a risky strategy because of its potential to be divisive, and because:

… but then again:

So we’ll see how this plays out. Tomorrow I’ll be joining Keiran Pedley and Mike Smithson for the Polling Matters show, and watch this space for other interesting developments…

About The Author

Related Posts

  • mrodent33

    Hello Matt,

    (um… is it your policy never to comment on comments to your pages? That’d be a shame if so because I think below I raise some concerning angles…)

    I hope you’re right about these probabilities… but I invite you to take a look at this I think very intelligent comparison of the recent polls and graphs, including yours.

    … he suggests

    1) that you might be placing too much faith in your own model
    2) that undecideds, as they make up their minds, seem to be favouring Brexit
    3) that the passion of Brexiters will mean that they are far more likely to actually vote (this point has been made by many)

    And I also make one suggestion myself:

    4) is it possible that the large Remain volumes placed with Bookies are in fact being heavily influenced by none other than … YOURSELF! If one is into this sort of betting, one will of course know the name Matt Singh from GE 2015.

    Care to comment on the above points? Keep up the good work. God I hope you’re getting it right.

    PS I know which outcome you favour (being well-educated, realistic and sane) … dare one suggest there might be a tiny bit of wishful thinking influencing your modelling?

    PPS I’m currenlty thinking of placing a bet of £4k on Brexit at 4 to 1 or 5 to 1 (if it gets there). At least that would be some minor compensation if this disaster does in fact happen…

    • Matt Singh

      No, but as I’m sure you’ll understand I’ve been quite busy! I’ll respond to that article in due course, but to take your points:

      4) Quite possibly… They’ve certainty converged to my forecast. What’s driving it? I don’t and can’t know, so while it’s interesting to watch betting markets, I wouldn’t use them as a predictor.

      PS I write as a neutral – I think Remain is favourite, but that doesn’t mean it’s what I want. My only agenda is accuracy.

      PPS I don’t advise on bets, however I think (at this point) that Brexit is more likely than 5-1 against.

      • mrodent33

        Thanks… Apologies if I misrepresented your stance on the result… I seemed to have read something on your site where you said you favoured Remain… but it must have been that you **favoured** Remain (as the more likely outcome).

        Would be interested in your views on the linked article…

        • mrodent33

          Interesting: Brexit odds were as long as 9 to 2 two days ago. This morning every single company was at 4 to 1. Now (ominously) a couple have gone to 7 to 2. Maybe they’ve read your reply. Love to know how many daily hits your pages are getting!