Sunday polls round-up

Four voting intention polls in the Sunday papers.

YouGov/Sunday Times – CON 39%(+3), LAB 26%(+1), LDEM 17%(nc), BREX 10%(-1) (tabs)
Deltapoll/Mail on Sunday – CON 41%(+1), LAB 29%(+1), LDEM 16%(+2), BREX 6%(-5) (tabs)
Opinium/Observer – CON 41%(-1), LAB 29%(+3), LDEM 15%(-1), BREX 6%(-3) (tabs)
Panelbase – CON 40%(nc), LAB 30%(+1), LDEM 15%(+1), BREX 8%(-1) (tabs)

In most cases changes are from last week, YouGov’s changes are from their midweek poll for the Times & Sky. All four show the upwards trend in Labour support continuing (though given some of them also show the Conservatives gaining support, we cannot say that the Tory lead is narrowing). All four also show support for the Brexit party falling, particularly in the Deltapoll and Opinium figures. Apart from one outlier in August, 6% is the lowest the Brexit party have recorded since the European elections.

Following on from my post about MRP in the week, the Observer also reports the topline results from a second MRP model, again carried out by an organisation campaigning for tactical voting – this time Gina Miller’s Remain United. The model currently predicts 347 Conservative seats, 204 Labour seats and 24 Liberal Democrat seats. In comparison the BestforBritain MRP results that Chris Hanretty scraped from their website seemed to imply a better showing for the Conservatives, with around 358 Conservatives seats and just 188 Labour and 19 Lib Dems (note that the Best for Britain model only covered England & Wales). The average vote shares in the Remain United model imply a Conservative lead of only around 6 points, so the difference in seat numbers may very well just be down to projecting a higher level of Labour support, rather than a different pattern of swings. For all the fuss about “rival tactical voting sites”, by my count there are only 13 seats where RemainUnited suggest voting Labour and BestforBritain suggest voting Lib Dem.

There is a methodological explanation on the remain united website here that says it is not actually an MRP model, but an RRP model, which is apparently “similar” to an MPR model (I don’t know what the technical differences are to the approach, but the explanation they give does indeed sound very similar). The model is based on a ComRes sample of only 6097 responses, significantly smaller than the 46,000 sample that Best for Britain used and the 50,000 or so samples that YouGov were using at the last election. There was a similar ComRes/ElectoralData RRP model for the European elections earlier this year which did not perform particularly well – while it got the share of Brexit party support correct, it overstated Labour support by 10%, understated the Greens by 6% and Liberal Democrats by 5%, which would be rather a problem is your aim is to work out which remain party is best placed to win in seats. That said, the data for their European election model was collected more than a week before polling day, and they may well have finessed the model since then.


Given the success of the approach at the 2017 election I expect we’ll see several MRP seat models this time round. The first one to emerge however is one constructed by Focaldata, using data from mixed sources, including YouGov, that Best for Britain have used to drive a tactical voting website. It has caused some controversy – particularly on the comment pages of the Guardian – with people arguing over the validity of its recommendations. I won’t get too far into that (vote for whoever the hell you want), but thought it was probably worth making a few comments about MRP itself, considering it will crop up again through the campaign.

What is MRP?

First, we need to understand what MRP is. It stands for multilevel regression and post-stratification, which almost certainly doesn’t help. There is a academic paper by Ben Lauderdale and his colleagues who run the YouGov MRP that explains it in great detail here, however the short version is that it’s a modelling technique aimed at producing robust estimates for small geographic areas from large national samples. In the context of elections, that means coming up with estimates of vote share in single constituencies based on a big national sample.

Using traditional techniques even very, very large samples don’t contain enough respondents to be a good guide to individual seats. If you had a huge sample of 50000, divided by 632 seats it would still give you less than 100 people a seat – which wouldn’t be enough to produce decent data. I’ve seen this as a naive criticism of the Best for Britain MRP model (there are only 70 people per seat!) but in fact that is exactly the problem that MRP is intended to solve.

MRP works by modelling the relationship between demographic and political variables and voting intention (the multilevel regression part), and then applying that to the demographics and political circumstances in each individual constituency (the post-stratification). So in this case, an MRP model would look at how demographics like age, gender and education relate to vote intention, and how that differs based on political variables (Is there an incumbent MP? Is it a remain or leave area?). That model is then applied to the known characteristics each seat. What that means is the projection in an individual seat is not just based upon how respondents in that seat say they would vote, it’s effectively also based on how respondents with the same demographics in seats with similar political circumstances say they would vote.

How well does it work?

At the last election YouGov had an MRP model that performed very well – correctly predicting the hung Parliament and some of the more unusual election results like Canterbury and Kensington. Clearly, given the accuracy of the YouGov model, it is possible to use MRP successfully to produce decent seat level estimates from a big national sample.

Best for Britain’s defence of their tactical recommendations relies heavily on how well the YouGov MRP model did in 2017. However, not all MRP models are necessarily equal. It isn’t one single model, it’s a technique, and it’s possible to do it well or badly. It is not certainly not a magical guarantee of accuracy. If we look back to 2017 the YouGov MRP model got all the attention, but it wasn’t the only MRP model out there. Lord Ashcroft also commissioned an MRP model, but that wrongly predicted a Tory majority. Just as some polls have been more accurate than others in recent years, some MRPs may be more accurate than others.

The things that drive the quality of a MRP model should be the quality of the data that’s going into it, and the quality of the model itself – have those designing it picked demographics and political factors that allow them to accurately model voting intentions? As an external observer however, it is quite hard to judge that. For the YouGov model there is its track record from 2017. From other MRP models, we’re driving a bit blind. We know it is a technique that can be very successful if done well, but we won’t really know if it is being done well until it’s compared to actual election results.

Are tactical voting recommendations based on an MRP model sensible?

In principle, yes. MRP is obviously not perfect or infallible – nothing is – but it is an established technique for producing estimates of support in small geographical areas from a larger national poll. Certainly it should be better than using a crude uniform swing, or just basing recommendations on what the levels of support were at the previous election and assuming nothing has changed since.

In practice, of course, it depends on the quality of the model and the tactical decisions that people make based upon them – I certainly don’t intend to get into that debate, especially since I expect there will be various rival tactical voting sites with different recommendation, and perhaps different aims and motivations.


-->

At the weekend we had a positive glut of national polls. In the last couple of days they’ve been joined by London and Wales polls from YouGov.

The Welsh poll for ITV Wales has Westminster voting intentions of CON 28%(-1), LAB 29%(+4), LDEM 12%(-4), BREX 15%(+1), Plaid 12%(nc). Changes are since mid-October, and show Labour retaking the lead over the Conservatives in Wales. While recent movement is in Labour’s favour, compared to the result at the 2017 general election these would be terrible figures for Labour. Compared to the shares of the vote in the 2017 general election in Wales the Conservatives are down six points, Labour are down twenty(!) points, the Liberal Democrats up seven, Plaid up two. So while the Tories are losing support, the slump in Labour support would likely result in many Labour seats falling to the Tories. As ever, Roger Awan-Scully has more in depth analysis here.

The YouGov London poll for Queen Mary University London shows similar dynamics in the capital. Current vote shares are CON 29%, LAB 39%, LDEM 19%, BREX 6%. Compared to the 2017 general election results in London that represents a drop of four points for the Conservatives, a drop of sixteen for Labour, an increase of ten for the Liberal Democrats. While the Conservatives are losing support, the large scale movement of voters from Labour to the Liberal Democrats may well win them a significant number of seats. It is a reminder that while people have been looking towards the more “leave-inclined” Labour seats in the North and Midlands for potential Tory gains, it is perfectly possible for them to win in more remain-inclined seats where they are losing support, so long as Labour are losing more support.

How it actually translates in terms of seats is difficult to know (especially in a city as politically diverse as London, where the dynamics of the race may be radically different in the inner-city seats, the leafy Lib Dem-Con marginals of South-West London and the more typical Con-Lab marginals in North London). Over the last few days we’ve also seen a drip-drip of constituency polls by Survation, primarily conducted for the Liberal Democrat party. So far they have published polls for South East Cambridgeshire (showing an 11 point Tory lead, a Con>LD swing of 11.5%), North East Somerset(a 16 point Tory lead, a Con>LD swing of 15%), Portsmouth South(a 3 point Lib Dem lead, a Lab>LD swing of 15%) and Cambridge (a 9 point Lib Dem lead, a Lab>LD swing of 16%). Obviously they all show the Lib Dems doing well, but I would urge some caution in their interpretation, as I would with any political party commissioned polls. It is impossible to know how many constituency opinion polls the Liberal Democrats have commissioned, so it’s perfectly possible that they have commissioned another ten, twenty, thirty constituency polls in seats where they weren’t doing quite so well, and choose never to publish them. We’re probably only seeing the constituency polls that the Lib Dems want us to see.

Finally, while I am not going to update with every individual national poll – the best way of looking at voting intention polls will always be to look at the broad trend – I’ll just update with those we’ve seen since my last post.

There is a new ICM poll for Reuters, the first of a regular series for the election campaign. Topline figures are CON 38%, LAB 31%, LDEM 15%, BREX 9%, GRN 3%. Fieldwork was over the weekend. It’s been almost a month since the last ICM poll (their regular voting intention polls seemed to peter out somewhat after Martin Boon left to set up Deltapoll), so changes since their last poll aren’t really relevant. This poll got some attention from the single digit Labour lead, though given the paucity of ICM polling in the last year we can’t really tell if that’s movement, or just ICM’s methodology.

Secondly there was a new YouGov poll for the Times. Topline figures were CON 38%(-1), LAB 25%(-2), LDEM 16%(nc), BREX 11%(+4), GRN 5%(+1). Fieldwork was over the weekend and changes are from last Thurs-Fri. The YouGov/Sunday Times poll at the weekend had some sharp movements: a six point increase for Labour, a six point drop for the Brexit party. Today’s poll partially reverses those changes, suggesting it was probably something of an outlier… though that means Labour are still four points up on the YouGov/Times poll last week.


The first Sunday of the election campaign, and as you expect, several polls in the Sunday papers:

The Telegraph have a poll from ORB. Topline figures are CON 36%, LAB 28%, LDEM 14%, BREX 12%. Fieldwork was Wednesday and Thursday. While they’ve published some trackers on support for Brexit, the last time I recall seeing a voting intention poll from ORB was back in April, so changes here aren’t really relevant.

ComRes in the Sunday Express have topline figures of CON 36%(+3), LAB 28%(-1), LDEM 17%(-1), BREX 10%(-2). Fieldwork was Wednesday and Thursday, with changes from mid-October, at the time of Johnson’s deal.

The regular Opinium poll for the Observer has topline figures of CON 42%(+2), LAB 26%(+2), LDEM 16%(+1), BREX 9%(-1), GRN 2%(-1). Fieldwork was Thursday and Friday, and changes are from last week.

YouGov in the Sunday Times has topline figures of CON 39%(+3), LAB 27%(+6), LDEM 16%(-2), BREX 7%(-6). Fieldwork was Thursday and Friday, with changes from midweek.

Finally, Deltapoll in the Mail on Sunday (they’ll be running weekly polls for them for the campaign) have topline figures of CON 40%(+3), LAB 28%(+4), LDEM 14%(-5), BREX 11%(nc). Fieldwork was Thursday to Saturday, with changes from mid-October, just after Johnson’s deal.

All the polls continue to show a sizeable Conservative lead, though as ever this varies somewhat from pollster to pollster. The eight point leads that the ORB and ComRes polls show would be quite tight between a Conservative majority and a hung Parliament on a uniform swing (not that I’d expect a uniform swing); the 12 and 16 point leads that YouGov, Deltapoll and Opinium show should give them a solid majority.

Note that all the polls are showing the Conservatives gaining support (so did the Survation, Panelbase and MORI polls published on Thursday and Friday), and most of them are showing the Labour party gaining too – suggesting perhaps that now an election is a reality, some wavering voters are moving their support behind the two main parties after all (something we also saw in the 2017 election, when UKIP support fell off a cliff almost as soon as the election was called).

Note also that the fieldwork for the YouGov, Deltapoll and Opinium polls was conducted on and after October 31st, when Britain had obviously not left the European Union on time. It does not appear to have either damaged the Tories or boosted the Brexit party. That looked like the case anyway, after all, it’s been very obvious for about a week and a half that Britain was not going to leave on the 31st, but there was always that small chance that there would be an impact when the date actually passed. There wasn’t. (And I told you those hypothetical “How would you vote if Britain hadn’t left the EU by Oct 31st” questions showing Tory support slumping didn’t have any predictive value. Next time people do them – and they will – please do remember that they don’t work!)


The first voting intention polls published since the election was called were in this morning’s papers: Survation for the Mail, Ipsos MORI for the Standard and YouGov for the Times. Topline figures were

Survation – CON 34%, LAB 26%, LDEM 19%, BREX 12%, GRN 1% (tabs)
Ipsos MORI – CON 41%, LAB 24%, LDEM 20%, BREX 7%, GRN 3% (tabs)
YouGov – CON 36%, LAB 21%, LDEM 18%, BREX 13%, GRN 6% (tabs).

There’s quite a spread between the results – Ipsos MORI have the Conservatives up above 40, their highest in any poll since August. YouGov and Survation have them in the mid-thirties. Labour’s support varies between 26% in Survation down to 21% in YouGov. All three have the Lib Dems between 18%-20%. This means while the Conservative lead varies, there is a consistent Conservative lead across the board as we start the campaign.

It’s worth noting that that Tory lead is largely down to a split opposition. Even in the MORI poll the Conservatives have lost support since the election (in the YouGov and Survation polls they’ve lost a lot of support). This is not a popular government – in the MORI poll, their satisfaction rating is minus 55 – it’s just that the main opposition have lost even more support. The healthy Conservative lead is down to the fact that the Conservatives are retaining the bulk of the Leave vote, while the remain vote is split between Labour, the Liberal Democrats, the Greens, the SNP, Plaid and so on.

For as long as this is the case, the Conservatives should do well. If it should change they’ll struggle. If the Brexit party manage to get back into the race and take support from the Tories it would eat into their lead. The other risk for the Tories is if the Remain vote swings more decisively behind either Labour or the Liberal Democrats (or that there are signs of more effective tactical voting, winning seats off the Conservatives despite a split vote). Essentially Boris Johnson needs to keep the Leave vote united and the Remain vote divided.

It is also worth considering how the Conservative lead might translate into seats. In 2017 the Conservative lead over Labour was only two and a half percentage points. You would therefore expect an eight point Conservative lead to translate into a majority, and a fifteen or seventeen point lead to be a landslide. In reality that Survation poll could easily be touch-and-go for a Tory majority and, while the bigger leads would likely get a Tory majority, it may not be landslide territory.

The reason that the Conservatives translated votes more effectively into seats in 2015 and 2017 was to do with the distribution of the vote. The Conservative re-emergence in Scotland meant that Tory votes up there were no longer wasted (but Labour votes increasingly were), the collapse of the Liberal Democrats in the South-West meant that the Tories vote there returned more MPs. If at the coming election we see those trends reverse, and the Conservatives lose seats to the SNP in Scotland and the Lib Dems in the South, then suddenly their votes won’t be translated so effectively into seats, and they’ll need to win more seats off Labour to make up for it.

Right now we have little evidence of how uniform or not the changes in support are, of whether there is evidence of tactical voting (Survation have released a couple of constituency polls they have conducted for the Liberal Democrats showing them doing very well in individual seats, but I don’t think it’s too cynical to imagine that the Lib Dems may have selectively published seats they are doing particularly well in). In the fulness of time I expect we will see the publication of MRP models along the lines of those YouGov conducted in 2017 that may give us a better steer, but I’ll come to that another day.

In the meantime, as we cross the starting line the Conservatives have a clear lead in the polls, but how it translates into seats is unclear. In the polls with the smaller Tory leads, it may not produce a majority at all. Equally, their lead is dependent upon the Leave vote remaining relatively united, and the Remain vote remaining divided, if that changes, the race could end up being far closer.