A variety of new and newish polls today.

Starting with the newest of the regular polls, Kantar‘s latest topline figures are CON 43%(+1), LAB 33%(-1), LDEM 11%(+2), UKIP 4%(nc). Fieldwork was between Thursday and Tuesday. The changes are not significant in themselves, but unlike most recent polls don’t show continuing movement towards Labour. Note also that there is a methodological change – Kantar now estimate how people who say don’t know will vote based on upon their demographics and whether they find May or Corbyn more trustworthy. The impact of this chance is to decrease the Labour vote by a point (so without it, the Conservative lead would barely have changed at all). Tables are here.

We also saw a Panelbase poll today. This is not actually new – it is the poll that was in the field during the Manchester bombing last week, which Panelbase made the decision to withhold in the light of the tragedy. Topline figures are CON 48%, LAB 33%, LDEM 7%, UKIP 4%. Fieldwork was back between the 19th and 23rd of May. I’ve not included changes as there are significant methodological changes here – Panelbase have tightened their turnout filter to only include people who say 10/10, and they reweight their voting intention question so the age matches the age profile of people who voted in 2015. As with other companies whose turnout model is based upon replicating the age profile of 2015 voters this has a substantial effect. Panelbase say without it their poll would have shown the lead narrowing by 6 points from their previous poll (implying they would otherwise be showing an eight point Tory lead on their old method!). Panelbase tabs are here. In their comments Panelbase also say they will be releasing a new poll in the next day or two which again has the Tory lead falling.

Thirdly there was a new Ipsos MORI Scottish poll. Topline voting intention there is SNP 43%, CON 25%, LAB 25%, LDEM 5%. As ever, the SNP are in a clear first place, but down from the last election. Where it had appeared that the Scottish Conservatives were now the clear second placed party, this suggests that Labour may have recovered into joint-second place (that would also be very good news for the SNP – under FPTP the SNP benefit from being the dominant pro-independence party when the unionist parties are split three ways). Full details are here.

There was also a new SurveyMonkey poll for the Sun. This has topline figures of CON 44%(nc), LAB 38%(+2), LDEM 6%(nc), UKIP 4%(-1). Now, SurveyMonkey are not members of the British Polling Council and we don’t have any tables or further methodological detail to examine. However, they did poll at the 2015 election so have a record to judge. Their method is unusual – sample is gathered by randomly selecting people at the end of other surveys hosted on the surveymonkey platform. Back in 2015 they were the only company whose pre-election poll got the Conservative lead about right…but because they got both Labour and the Conservatives too low their average error across all parties was the highest (and the BPC inquiry found that their sample was still heavily skewed towards the politically interested… though they may have corrected that since then). In short, make of that what you will – it may be that their approach does do something that traditional polling does not… or it may be they just got lucky in 2015.


Two new polls today:

Survation for Good Morning Britain this morning had topline figures of CON 43%(nc), LAB 37%(+3), LDEM 9%(nc), UKIP 4%(nc). Fieldwork was on Friday and Saturday and changes are from the previous week. Tabs are here.

ICM for the Guardian had topline figures of CON 45%(-1), LAB 33%(+1), LDEM 8%(nc), UKIP 5%(nc). Fieldwork was from Friday to Monday and changes are from the ICM/Sun on Sunday poll at the weekend. Tabs are here.

Both of today’s polls continue to show movement in Labour’s favour, thought the overall lead is different. A six point Tory lead would represent a small swing towards Labour, a twelve point Tory lead would still give them a stonking great majority.

The key difference between polls showing large and small leads is, as I wrote at the weekend down to how pollsters are treating turnout. There are lots of differences between different polling companies methods: they sample differently, weight by different things, do different things with don’t knows and so on. However, right now the one really huge difference is turnout. Weighted with all its normal demographic and political weights, ICM would have shown a Tory lead of only 3 points – that was transformed into a lead of 11 points by the turnout model, which predicts how likely respondents are to vote based on the estimates of turnout by age and class at the last election (the change from 11 to 12 points was the reallocation of don’t knows). That’s a big change, but given the errors in the polls in 2015 that may be necessary. On the other hand, if Jeremy Corbyn has managed to enthuse young people and there is a higher rate of turnout among younger voters than in 2015 then it risks understating Labour support. We shall find out next week…


-->

Saturday night polls

We should have a truckload of polls tonight. There is a new Opinium, a new ComRes for the Indy & Sunday Mirror, YouGov for the Sunday Times, probably an ORB and perhaps an ICM for the Sun on Sunday. I’ve seen rumours of Survation too (they normally poll for the Mail on Sunday) and we’re overdue a Panelbase poll. The thing to look for is whether polls continue to show a narrowing of the Conservative lead – keep an eye on the fieldwork dates, more recent polls could be showing an impact from reactions to the bombing (or, indeed, the effects of the dementia tax row fading). Also remember the house effects I wrote about earlier – ICM and ComRes tend to show larger Tory leads anyway, so even if they show a significant movement towards Labour it may still leave the Tories with a good lead.

The first poll we actually have figures for is Opinium, who have topline figures of CON 45%(-1), LAB 35%(+2), LDEM 7%(-1), UKIP 5%(nc). Changes are from the previous week and fieldwork was on Tuesday and Wednesday, so just after the Manchester bombing but before political campaigning had resumed. We have movement towards Labour, but the Conservatives still managing to cling onto a double-digit lead. Tabs are here.

ComRes for the Sunday Mirror and Independent have topline figures of CON 46%(-2), LAB 34%(+4), LDEM 8%(-2), UKIP 5%(nc). Changes are from a fortnight ago, and fieldwork was between Wednesday and Friday. The Tory lead has dropped by six points, but ComRes tends to give the Conservatives some of their better figures, so this still leaves them with a twelve point lead. Tabs are here.

ORB for the Telegraph have topline figures of CON 44%(-2), LAB 38%(+4), LDEM 7%(nc), UKIP 5%(-2). Changes are from a week ago and fieldwork was Wednesday to Thursday. Once again, we have a narrowing of the Tory lead, in this case down to six points.

YouGov for the Sunday Times have topline figures of CON 43%(nc), LAB 36%(-2), LDEM 9%(-1), UKIP 4%(nc). Changes are from the Times poll earlier in the week and fieldwork was Thursday to Friday. This is the most recent of the polls we’ve seen so far tonight, and it has Labour falling back a bit from the YouGov poll in the week. That said, it is only one poll, so don’t read too much into that unless we see other polls showing a similar pattern.

ICM for the Sun on Sunday has toplines of CON 46%(-1), LAB 32%(-1), LDEM 8%(-1), UKIP 5%(+1). Fieldwork was on Wednesday to Friday, and changes are since the ICM/Guardian poll conducted over last weekend. Changes here seem quite steady (ICM’s previous poll already showed a sharp narrowing of the lead). As I said earlier, ICM and ComRes tend to show the largest Tory leads because of their demographic based turnout model.

I’ll update this post through the evening as other polls appear.


Yesterday the British Polling Council had an event talking about how the polls had changed since 2015. This included collecting up data from all the companies on what they’ve done to correct the error and what they are now doing differently – all those summaries are collected here.

In looking at what’s changed it’s probably best to start with what actually went wrong and what problem the pollsters are trying to solve. As all readers will know, the polls in 2015 wrongly overstated Labour support and understated the Conservatives. The BPC/MRS inquiry under Pat Sturgis concluded this was down to unrepresentative samples.

Specially, it looked as if polls had too many younger people who were too engaged and too interested in politics. The effect of this was that while in reality there was a big difference between the high turnout among old people and the low turnout among young people, among the sort of people who took part in polls this gap was too small. In short, the sort of young people who took part in polls went out and voted Labour; the sort of young people who weren’t interested and stayed at home didn’t take part in polls either.

So, what have polling companies done to correct the problems? There is a summary for each individual company here.

There have been a wide variety of changes (including YouGov interlocking past vote & region, ICM changing how they reallocate don’t knows, ICM and ComRes now both doing only online polls during the campaign). However, the core changes seem to boil down to two approaches: some companies have focused on improving the sample itself, trying to include more people who aren’t interested in politics, who are less well educated and don’t usually vote. Other companies have focused on correcting the problems caused by less than representative samples, changing their turnout model so it is based more on demographics, and forcing it to more accurately reflect turnout patterns in the real world. Some companies have done a bit of both.

Changes to make samples less politically engaged…

  • ICM and YouGov have both added a weight by respondents level of interest or attention to politics, based upon the British Election Study probability survey. YouGov have also added weights by level of educational qualification.
  • Ipsos MORI haven’t added political interest weights directly, but have added education weights and newspaper readership weights, which correlate with political interest.
  • Kantar have added education weighting, and also weight down turnout to the level they project it to be as a way of reducing the overall level of political engagement in their sample.

Changes to base turnout on demographics…

  • ComRes have changed their turnout model, so it is based more on respondents’ demographics rather than how likely they claim they are to vote. The effect of this is essentially to downweight people who are younger and more working class on the assumption that the pattern of turnout that we’ve seen at past elections remains pretty steady. ICM have a method that seems very similar in its aim (I’m not sure of the technicalities) – weighting the data so that the pattern of turnout by age & social grade is the same as in 2015.
  • Kantar (TNS) have a turnout model that is partially based on respondents age (so again, assuming that younger people are less likely to vote) and partially on their self-reported likelihood.
  • ORB weight their data by education and age so that it matches not the electorate as a whole, but the profile of people who the 2015 British Election Study who actually voted (they also use the usual self-reported likelihood to vote weighting on top of this).
  • Opinium, MORI and YouGov still base their turnout models on people’s answers rather than their demographics, but they have all made changes. YouGov and MORI now weight down people who didn’t vote in the past, Opinium downweight people who say they will vote for a party but disapprove of its leader.
  • Panelbase and Survation haven’t currently made any radical changes since 2015, but Panelbase say they are considering using BES data to estimate likelihood to vote in their final poll (which sounds to me as if they are considering something along the lines of what ICM are doing with their turnout model)

In terms of actual outcomes, the pollsters who have adopted demographic turnout-models (ComRes, ICM and Kantar) tend to show larger Conservative leads than companies who have tried to address the problem only through sampling and weighting changes. We cannot really tell which is more likely to be right until June 8th. In short, for companies who have concentrated only on making samples more representative, the risk is that it hasn’t worked well enough, and that there are still too many of the sort of young engaged voters who are attracted to Jeremy Corbyn in their samples. For companies who have instead concentrated on demographic-based turnout models, the risk is that the pattern of turnout in 2017 differs from that in 2015, and that Jeremy Corbyn’s Labour really does manage to get more young people to come out to vote than Ed Miliband did. We will see what happens and, I expect, the industry will learn from whatever is seen to work this time round.


Two polls are out tonight. A Kantar poll conducted between last Thursday and Sunday (so before the bombing) has topline figures of CON 42%(-5), LAB 34%(+5), LDEM 9%(+1), UKIP 4%(-2). TNS has a turnout model based partially on age, so has tended to show larger Tory leads… but this poll has it dropping ten points and falling into single figures. Tabs are here.

YouGov’s weekly poll for the Times meanwhile has topline figures of CON 43%(-1), LAB 38%(+3), LDEM 10%(+1), UKIP 4%(+1) – a Tory lead of just five points. Fieldwork for this poll was conducted on Wednesday night and Thursday daytime, so is the first conducted entirely after the Manchester bombing. Tabs are here.

The Tory lead is clearly continuing to fall away at a rapid rate. On the face of it one might be tempted to conclude that the actual impact of the bombing was to help the Labour party, but I think it more likely that it’s to do with the disastrous Tory manifesto launch. I posted earlier about the negative impact of the Tory manifesto. In contrast the Tories still seem to have a good lead on security and terrorism – in today’s YouGov survey people say they trust Theresa May far more than Jeremy Corbyn to make the right decisions on terrorism (55% trust May, only 33% trust Corbyn) and the Tories have a strong lead on the issue of Defence and Security. That suggests to me the cause of the narrowing is far more likely to be the manifesto, row and u-turn.

As ever, all the usual caveats about one poll apply. Before one gets too excited wait and see if other polls show such a tight race, and whether or not other polls show any more impact from the bombing. As things stand though the election suddenly seems a little less of a foregone conclusion than it appeared at the beginning of the race.