New party leaders normally enjoy a honeymoon in the polls. It’s noticeable for leaders taking over in opposition, on the relatively rare occassion that the party leadership changes hands in government the honeymoon is often remarkable. In the last fifty years there have been three previous occasions when the premiership changed hands between-elections:

  • Wilson-Callaghan, 1976. When Harold Wilson announced his resignation in the middle of March the polls were showing a Conservative lead of between two and five points. The polls immediately following Wilson’s resignation and during Callaghan’s first month in office showed Labour leads of between one and seven points, before returning to a steady Tory lead in May.
  • Thatcher-Major, 1990. Margaret Thatcher was famously removed by the Tory party in November 1990. In the month before the leadership election Labour had an average poll lead of thirteen points. In the month immediately following her resignation and replacement by John Major the Conservatives had an average lead of five points, peaking at 11 points. Over the next few months the polls settled down to an average Tory lead of four points or so.
  • Blair-Brown, 2007. The Blair-Brown handover was a more drawn out affair: Blair announced his resignation at the start of May 2007, when the Conservatives had a poll lead of around six points, and actually handed over to Gordon Brown at the end of June. Through July and August Brown enjoyed an average Labour lead of around five points, peaking in double-digit leads during the Labour conference at the end of September… and their rapid collapse afterwards. The Conservatives were ahead again by October, and remained so for the rest of the Parliament.

Every mid-term change of Prime Minister has been accompanied by a significant boost in polling figures – in the three historical cases, they’ve gone from trailing the opposition to a clear polling lead. The boosts have tended to be comparatively short though – Callaghan and Major only enjoyed a month or so before settling down into a new equilibrium, Brown enjoyed a honeymoon that lasted several months, but that was probably because he was seem to have responded well to the Glasgow Airport attack and Summer floods. There’s no clear pattern as to where the polls settle after the honeymoon: I suppose it depends very much on the leader. Once the honeymoons had passed the change in leader didn’t make that much difference in 1976 and 2007 (in both cases Labour’s position absolutely tanked a few months down the line… but for different reasons), in 1990 though there was a long lasting improvement in Tory support.

So to the current polling position. Today’s ICM poll has topline figures of CON 43%(+4), LAB 27%(-2), LDEM 8%(-1), UKIP 13(-1) (tabs are here). It follows on from an ICM poll last week showing the Conservatives ten points ahead, a YouGov poll giving the Conservtives an eleven point lead and an Opinium poll giving them a more modest six point lead. All four polls had Labour around or just below 30% and the Conservatives nearer 40%, UKIP down a little from the levels of support they’d been showing before the referendum.

Viewed together it certainly looks like the sort of boost a new Prime Minister normally receives, which is a good reason not to read too much into it. New Prime Ministers receive good poll ratings because they haven’t had to annoy too many people yet – the public can project their hopes onto them and convince themselves they really will be different, really will deliver this, that or the other. Before long, however, the shine will come off and they’ll have to start making compromises and disappointing people. This is one good reason for Theresa May not to plan for an early election (and the mistake Gordon Brown made in not shutting down such considerations) – the current polls look wonderful for her, but on past timescales they won’t necessarily be so rosy in a couple of months time. It’s also a crumb of comfort for Labour… though quite a small crumb.

UPDATE: YouGov have fresh voting intention figures that also show a strong lead for the Conservatives, albeit, not quite as big as ICM’s. Their topline figures are CON 40%, LAB 28%, LDEM 8%, UKIP 13%, GRN 4%. Tabs are here


Ipsos MORI’s monthly political monitor has a much closer race than ICM’s last poll. Topline figures are CON 36%, LAB 35%, LDEM 11%, UKIP 8%, GRN 4% (full tabs are here.)

The poll was conducted over the weekend before Theresa May became Prime Minister, though did include a question on whether people thought she had what it took to be a good Prime Minister (55% of people though she did, 27% did not).

Given it is being rampantly misrepresented on social media, I should also explain about MORI’s turnout filter and how they present their figures (and why, therefore, some people are tweeting entirely different MORI figures!). These days the overwhelming majority of opinion polls contain some sort of adjustment for how likely people are to vote. The general pattern is that older people and middle class people are more likely to vote than younger people and working class people; older people and middle class people are also more likely to vote Conservative, younger people and working class people more likely to vote Labour. This means if a poll just included everyone, with no reference to how likely or unlikely they actually are to vote, then it would overstate Labour when compared to actual election results.

Polling companies account for this by weighting by likelihood to vote (the more likely you are to vote, the more your answer is counted) or filtering by likelihood to vote (only taking people who say they are likely to vote), based either on how likely people say they are to vote, or on demographic modelling. In the case of MORI, their topline figures are based only on people who say they are at least 9/10 likely to vote AND that they always, usually or have sometimes voted in the past. This makes a substantial difference to their topline figures – without this adjustment they would have been showing a five point Labour lead.

MORI’s headline figure is the one that is adjusted for turnout – the one point Conservative lead – which they regard as a better indicator of actual voting intention. However, because MORI’s political monitor has been going since the 1970s they still publish the figures without the turnout adjustment to preserve the data trend, even if they don’t feel it paints an accurate picture in an era of lower turnouts.

In short, if you are looking at Ipsos MORI figures with a view to seeing how well the parties might do in a general election tomorrow, you need to look at the figures that account for how likely people actually are to vote, not take false solace from figures that don’t take turnout into account.


-->

The Conservative leadership election is abruptly over while a Labour leadership election begins. No doubt there will be polling on those over the next couple of weeks. In the meantime ICM put out a new voting intention poll today, with topline figures of CON 38%, LAB 30%, LDEM 8%, UKIP 15%.

The eight point lead for the Conservatives is a slightly larger Tory lead than at the general election, and was conducted over the weekend so does not yet account for any honeymoon Theresa May may or may not enjoy. During the leadership election May ruled out the opinion of an early general election, so if she keeps her word she’ll resist the temptation of an early election while Labour are at one another’s throats. If not it may be an interesting result.

ICM rolled out a couple of methodology changes for today’s poll. Firstly they’ve dropped weighting turnout based on self-reported likelihood to vote and replaced it with a turnout model based on demographics, secondly they’ve started weighting by level of political interest – the poll was also conducted online rather than by phone, which seems to be increasingly the case for ICM polls. According to ICM the impact of the changes is typically to increase Tory support by about a point and decrease Labour support by about a point. Full tabs are here.


ICM have again conducted two parallel polls for the Guardian, one online, one by telephone (tabs). The pattern is the same as last month, on Westminster voting intention the two ICM polls show the same two point lead, although the ICM online poll has a higher level of UKIP support:

ICM Online – CON 34%, LAB 32%, LDEM 7%, UKIP 17%, GRN 4%
ICM Phone – CON 36%, LAB 34%, LDEM 7%, UKIP 13%, GRN 4%

For the EU referendum ICM have the typical phone vs online contrast. They have a eight point lead for Remain by phone, a four point lead for leave online… a twelve point gap (the average gap between online and telephone polls since the start of April is about 10 points, so ICM is a little larger, but nothing to write home about).

ICM Online – Remain 43%, Leave 47%, Don’t know 10%
ICM Phone – Remain 47%, Leave 39%, Don’t know 14%

Martin Boon’s own take over on the ICM website is, as usual, both honest and somewhat bemused: “The narrative that phone polls are more likely to be right ignores some fundamental flaws in phone methods. Labour supporters are continually oversampled by phone, and that may matter more than those same phone polls missing out on supposedly pro-Remain types, who are disproportionately less likely to turn out to vote. Similarly, what’s lurking under online covers could be equally nasty, and we should not ignore that the fact the UKIP voters are again, as they have long since been, higher in online polls than phone (or indeed at recent elections).”

Incidentally, it’s probably worth flagging up that there are house effects beyond just the phone/online difference. There are differences between different online pollsters too. This is ICM’s sixth online poll in a row to show Leave ahead, and they are clearly showing a small Leave lead. In contrast the majority of online polls conducted by YouGov and TNS over the last six weeks have had Remain very narrowly ahead, it’s not a big gap, but it’s starting to look consistent. When it actually comes to learning lessons from the EU referendum, these smaller differences may end up being the more valuable: without much fuss, pollsters are taking quite different approaches to correcting their methods after last year and the referendum may teach us something useful about what corrections are (or are not) working for online; what corrections are (or are not) working for telephone.

Methodological concerns aside, what does ICM tell us about the state of public opinion? Well both their phone and online polls have the gap between Tory and Labour narrowing, down from five point leads a month ago. In the referendum race the four point leave lead in the online poll is ICM’s largest this year… but that trend isn’t echoed in the phone poll. We shall see if other EU polling this week shows any coherent trend.

There was also a new ComRes online poll at the weekend for the Indy and Sunday Mirror. This had topline figures of CON 36%(+1), LAB 30%(nc), LDEM 8%(nc), UKIP 17%(+1), GRN 4%(nc). On the face of it this is a stronger poll for the Tories, but this is largely methodological – ComRes’s online polls tend to produce the most positive results for the Tories of any company because of their demographic based turnout model. Full tabs are here.


Today there was a new London poll from Opinium and a new Scottish poll from Survation. However with only two days to go before Thursday’s elections I thought I would take a broader look at all the polling so far for this week’s contests.

London

Four companies have produced polls for the London election this year: ComRes, YouGov, Opinium and Survation – their latest figures are below (Opinium publish their first round data without removing don’t knows, so I’ve repercentaged it to make it comparable to other polling). Note that while Opinium described their poll as their final poll for the election, the polls from the other companies are not necessarily the final ones: I’m expecting to see some eve-of-election polls tomorrow.

First round . Second Round
Pollster Goldsmith Khan Pidgeon Whittle Berry Others Goldsmith Khan
Opinium (26th Apr-1st May) 35 48 4 5 5 3 43 57
Survation (21st-25th Apr) 34 49 3 5 3 6 40 60
YouGov (15th-19th Apr) 32 48 5 7 6 2 40 60
ComRes (30th Mar-3rd Apr) 37 44 7 5 4 3 45 55

The recent polls have Sadiq Khan convincingly ahead – the three most recent polls have him just short of winning on the first round, and on the second round he wins comfortably with a lead of 14 to 20 points. The ComRes poll is a little closer, but is a month old now and still had Khan winning by ten points on the second round. Note that only one poll, Opinium’s final call, has been conducted since Labour’s anti-Semitism row. Khan himself has thoroughly distanced himself from Ken Livingstone, but there is always a risk of guilt by association. Opinium’s poll doesn’t suggest it has damaged him (in fact it had Khan extending his lead) and Khan’s lead looks unsurmountable anyway, but we’ll see what the final polls tomorrow show.

Scotland

A broader range of companies have produced polls in Scotland, with figures from six different companies so far. Once again, these are by no means the final polls from each company and I am expecting final eve-of-election polls from some companies tomorrow.

Constituency . Regional
Pollster CON LAB LD SNP CON LAB LD SNP GRN
Survation (1st-2nd May) 19 21 7 49 20 19 6 44 7
Panelbase (23rd-28th Apr) 17 23 6 49 19 22 4 44 6
Ipsos MORI (18th-25th Apr) 18 19 6 51 19 17 7 45 10
TNS (1st-24th Apr) 17 22 7 52 18 22 5 45 8
BMG (11th-15th Apr) 16 21 6 53 16 20 6 46 7
YouGov (7th-11th Apr) 19 21 6 50 18 19 5 45 8

The SNP’s victory in Scotland is a foregone conclusion (hell, if they don’t win this would be the king of all polling errors). The more interesting question is who will come in second place – Scottish Labour’s stock has fallen so low they risk dropping behind the Scottish Conservatives. All recent polls now have Labour ahead of the Tories on the constituency vote, but several have the Conservatives ahead on the regional vote (and given that Labour will struggle to win constituency seats, the regional tally will likely have a greater impact on how many MSPs each party gets). Also keep an eye on the gap between the SNP’s constituency vote and regional vote – in 2007 and 2011 they were within a percentage point or two of each other, but the polls are suggesting the SNP will do between five and seven points worse on the regional vote, largely to the benefit of the Scottish Greens. This seems feasible enough (the Scottish Parliament’s electoral system means that if the SNP clean up on constituency seats they will struggle to win many list seats) but it will be interesting to see to what extent it is reflected in the actual results.

Wales

There is comparatively little polling in Wales and the only regular and recent figures are the YouGov polls for ITV Wales and Cardiff University (ably reported on by Roger Scully at his Elections in Wales blog). The most recent figures there are CON 19%, LAB 33%, LDEM 8%, Plaid 21%, UKIP 15% for the constituency and CON 19%, LAB 29%, LDEM 8%, Plaid 22%, UKIP 15% in the regional vote. There will be a final YouGov Wales poll on ITV news tomorrow night.

Northern Ireland

Despite the name of this blog, it actually tends to be GB Polling Report most of the time – Northern Irish polls are even rarer than Welsh ones. We do have one though! Lucidtalk had a poll of Assembly voting intention figures in the Northern Irish edition of today’s Sun – topline figures are DUP 27%, SF 26%, UUP 16%, SDLP 12%, Alliance 8%, TUV 4%, GRN 3%.

Police and local elections

The other two elections on Thursday are local authority elections – mainly in those districts councils that elect by thirds, including metropolitan councils outside London (just over a third of the country will have local elections) – and Police and Crime Commissioner elections, which take place throughout England and Wales with the exceptions of London and Greater Manchester. Neither contest has any published polling.