Usual caveats apply

Survation have put out a new poll, the topline voting intention figures are CON 24%(-5), LAB 35%(-1), LD 11%(-1), UKIP 22%(+6). The 22% for UKIP is the first poll to show them breaking the twenty percent mark.

In many ways the high UKIP score here shouldn’t come as a surprise, for methodological reasons Survation tend to show the highest levels of UKIP support so if ICM have them at 18% and ComRes at 19% I would have expected Survation to have them in the low twenties. Striking it may be, but the increase in UKIP support is actually in line with what weve seen elsewhere, just using a method that is kinder to UKIP.

More interesting is the drop in Tory support, down five points on Survation’s poll in April. The poll was conducted on Friday and Saturday so at least partially after the “swivel eyed loon” story broke (it came out in Saturday’s papers, so broke about 10pm on Friday night). All the usual caveats I apply to any poll showing sharp or unusual results apply. Sure, it might indicate a shift in support, but just as likely its a blip – wait to see if it is reflected in any other polling. As Twyman’s Law of market research says “anything surprising or interesting is probably wrong”.

Two further comments, I’ve written before about people making the error of looking at the changes in a poll over a month and assuming that events in the last few days are the cause. Survation’s last poll was at the end of April before the local elections, so changes are just as likely to be down to the local elections and the Conservative infighting over Europe as anything more recent.

Secondly there is a tendency for the media and the denizens of Twitter to get all excited about unusual polls that give newsworthy stories when this is, of course, the exact opposite of what you should do if you actually want to understand public opinion. The correct approach is to look at the broad underlying trend and ignore the odd looking polls, the media normally do the opposite. The trend is that UKIP support has jumped substantially following their local election success, and that the Labour lead has been narrowing. The Conservative figure here may yet suggest a new direction, but let’s wait and see.


Following the Populus/Times Eastleigh poll that showed the Liberal Democrats five points ahead, there is a new Survation poll of Eastleigh in the Mail on Sunday tomorrow that shows the opposite picture, with the Conservative party four points ahead. Their topline voting intentions for Eastleigh are CON 33%(nc), LAB 13%(nc), LDEM 29%(-7), UKIP 21%(+5) – changes are from the previous Survation poll of Eastleigh a fortnight ago.

Both polls were conducted on the telephone and while I haven’t seen the Survation tables both companies tend to use a similar methodology in terms of weighting and reallocating don’t knows to the parties they supported at the last election. I understand they were carried out at about the same time, so it shouldn’t be a “Rennard effect”. The two polls show UKIP with the same level of support, and no significant difference in Labour support – the only difference is the figures for the Conservatives and Liberal Democrats.

I’ll have a proper look when the Survation tables appear, but at first glance the most likely explanation for the difference between the two polls is just the normal variation within the margin of error… suggesting that the race really could be neck-and-neck.

I don’t know if there are any polls of Eastleigh to come – I’ve heard rumours of a poll in the field over this weekend, but we shall see.

UPDATE: The tables for the Survation poll are up here. There are some minor differences in approach (Survation reallocated don’t knows at a lower rate, and didn’t weight up people who didn’t vote in 2010 to as a high a proportion of the sample as Populus did), but none that would explain the difference. Not that the difference really needs a fancy explanation – once you take into account the high level of don’t knows the difference between the two polls can easily be explained by normal margin of error.


The post-speech polls

The first poll conducted after Cameron’s Europe speech, YouGov’s on Friday, didn’t appear to show any impact on voting intention. However, this weekend we have a further four polls conducted after the speech (the Opinium one was mostly done prior to the speech). Here they are:

Angus Reid/Sunday Express have the Conservatives on 30% (up three) and Labour on 39% (down three) – the online version of the article doesn’t mention the UKIP or Lib Dem scores. The poll was conducted on Thursday and Friday.

ComRes/IoS/Sunday Mirror has the Conservatives up 5 on 33%, Labour unchanged on 39%, UKIP down 4 points on 10%. As I wrote yesterday, some of this appears to be due to ComRes treating likelihood to vote differently in their December poll, but even with consistent treatment of likelihood to vote the poll would have shown the Conservatives up 4 or 5 points, though it would probably also have shown Labour down slightly.

Survation/Mail on Sunday has the Conservatives on 31%, up two, Labour unchanged on 38% and UKIP down two on 14%.

YouGov in the Sunday Times has topline figures of CON 35%, LAB 41%, LDEM 12%, UKIP 7%. Compared to the average of YouGov’s polls last week that equates to the Conservatives being up by about two, Labour down by about two and UKIP down by about one.

In each individual poll the changes are relatively small, but they are consistent across the pollsters, so we can be fairly confident that the Conservatives have enjoyed a small boost from the referendum promise and the positive publicity around it. UKIP appear to have dropped slightly, but not massively (the biggest drop they had, that in the ComRes poll, was mostly due to methodological variation). The shift is hardly a game changing degree though.

Looking at the other questions, there is some stark variation in how people say they’d vote in an EU referendum. YouGov’s figures are very similar to what they were showing early in the week – in a straight referendum question 37% say they would vote to stay, 39% say they would vote to leave. If David Cameron were to renegotiate and recommend a yes vote then 50% of people say they would vote to stay, compared to 25% who would vote to leave – the contrast is almost entirely down to Conservative voters, who would currently vote to leave, but would vote to stay on renegotiated terms.

Angus Reid asked how people would vote if Cameron did manage to repatriate some powers, 34% said they would vote to stay in, 34% said they would vote to leave.

Survation found 50% saying Britain should leave the EU, 36% that they should stay. If Cameron wins back some powers, 43% of those saying Britain should leave say they would consider voting to stay.


Polls tonight

Tonight we are due at least three polls – the monthly online ComRes poll for the Independent on Sunday & Sunday Mirror, the fortnightly Opinium poll for the Observer and the weekly YouGov poll for the Sunday Times. I believe there may also be a Survation poll for the Mail on Sunday plus, of course, any ad hoc polls that newspapers have commissioned on the back of the referendum announcement in the week. We had our first poll conducted entirely after Cameron’s referendum announcement yesterday, and that didn’t show any obvious impact… but as ever, that was only one poll. We’ll have a much better idea tonight. I’ll update as the polls come in.

UPDATE: ComRes’s monthly online poll has topline figures of CON 33%(+5), LAB 39%(nc), LDEM 11%(+2), UKIP 10%(-4). Changes are from their last online poll in December. There is also an improvement in David Cameron’s own figures – 32% agreed with the statement that he was turning out to be a good Prime Minister (up 5 from a month ago), 46% disagreed (down 6).

UPDATE2: The ComRes tables are now up here. Regular readers may remember that last month ComRes changed their methodology so that when they weighted by likelihood to vote, they treated the minor parties the same as all the other parties (previously they only included BNP, Green and UKIP voters if they said they were 10/10 certain to vote, but included Tory, Lab and LD voters who were 5+/10 likely to vote, weighted proportionally). The effect was to give UKIP a hefty boost. Well, looking at this month’s tables it looks to me as if they’ve reversed the change, and gone back to only including UKIP voters who are 10/10 certain to vote. If that is the case, it would suggest a lot of the increase in UKIP’s support in ComRes’s poll last month, and their drop in support this month, is just switching methodology back and forth.

UPDATE3: By my rough, back-of-a-fag-packet reckoning, if ComRes had used the methodology they used last month they’d have shown topline figures of CON 32%, LAB 37%, LD 11%, UKIP 13% – so it would still have shown a Conservative increase, but UKIP would have suffered a smaller drop, with Labour also falling slightly.

UPDATE4: The fortnightly Opinium poll for the Observer has topline figures of CON 28%(-3), LAB 41%(nc), LDEM 8%(+1), UKIP 14%(+2). Unlike the ComRes poll, which was done between Wednesday and Friday, so wholly after Cameron’s referendum speech, the Opinium poll was done between Tuesday and Thursday and, according to the Observer, most of the fieldwork was completed before news of the speech.

UPDATE5: According to Sky the Survation poll in the Mail on Sunday has topline figures of CON 31%(+2), LAB 38%(nc), LDEM 10%(-1), UKIP 14%(-2). The movement is in the direction of the Conservatives, but the changes are all well within the margin of error. The other poll due tonight is YouGov in the Sunday Times, but that often doesn’t surface till the morning. I’ve seen mention of an Angus Reid poll in the Sunday Express too, but no idea what that shows or if it even exists.


We are a year and a half away from the European elections, but the recent rise in UKIP support in Westminster polls has brought some attention to them, meaning we are already seeing voting intention polls for the European elections. ComRes have an online European elections poll in the People tomorrow, and Survation had one in the Mail on Sunday last weekend. Whether they mean anything at this stage is a different question, but here they are for the record.

ComRes/People – CON 22%, LAB 35%, LDEM 8%, UKIP 23%, GRN 5%, Others 8%
Survation/MoS – CON 24%, LAB 31%, LDEM 11%, UKIP 22%, GRN 6%, Others 6%

Both show Labour topping the poll – what we’d expect given their strong lead in the Westminster polls and the mid-term nature of the elections. What will get more attention is that both polls also show the Conservatives and UKIP closely placed for second place – Survation had the Conservatives narrowly ahead, ComRes UKIP narrowly ahead.

The other questions in the ComRes poll seem to show a slight movement towards support for the European Union since ComRes last asked them in October 2011. The percentage of people agreeing that EU membership was a costly mistake fell from 54% to 48%, people disagreed with the statement that they’d vote in favour of withdrawal by 42%-33%, when in October 2011 it had been evenly balanced at 37%-37%. I’ll write more about this tomorrow…

This weekend we are also due the weekly YouGov/Sunday Times poll and, assuming they are sticking to their current timetable, the fortnightly Opinium/Observer poll.

UPDATE: Opinium’s fortnightly poll for the Observer has topline figures of CON 31%(+2), LAB 41%(+2), LDEM 7%(-1), UKIP 12%(-3) – Observer write up here.