Archive for the 'Pollsters/polling' Category


The Corbyn polling could be the 2015 version of what happened to Hilary Benn in 2007

Thursday, July 30th, 2015

Don’t attach too much credence to numbers at this stage

There have been only two Labour elections in recent times where there has been polling and we are able to look back and compare the survey numbers with the actual votes received.

In 2010, as I’ve reported before, the final YouGov members’ survey taken after the voting had started showed EdM with a 4% lead in this part of the electoral college. David actually won this segment by 8.8%.

Three years earlier when the party elected Gordon Brown leader without him having the inconvenience of facing another candidate all the focus was on the deputy battle.

The deputy leader polls throughout, as shown above, all had the son of that great Labour icon, Tony Benn, the then international development secretary Hilary. The Benn name resonated very strongly with the party members and in each of the three polls he came top.

    Notice how in the first two polls Benn chalked up three times the members’ first preferences that he eventually received and twice as many in the final survey.

    What members were telling pollsters beforehand was very different from hat they actually did

There’s another interesting comparison with Corbyn. Benn had a great struggle securing enough MP nominations to get on the ballot.

We are now 45 days from the election. There is a lot of time for the mood to change. Things can be very different when those eligible to vote are faced with the actual ballot paper.

Mike Smithson


The issues facing Britain: Immigration and NHS down/ Defence-terrorism sharply up

Wednesday, July 29th, 2015


ORB/Indy poll finds that 76% think that LAB less electable now than it was on May 7th

Tuesday, July 28th, 2015

But does this poll really tell us anything?

An ORB poll for the Independent carried out over the weekend finds that 76% of those who had a view believe that LAB is less electable now than it was at the general election.

We’ve not yet seen the dataset or the precise question wording but the overall picture looks gloomy for the red team and sets out very clearly the challenge facing the new leader when he/she takes over the party on September 12th.

    Aren’t we just seeing what happens to most political parties less than three months after a devastating election defeat?

I can’t recall a similar post general election poll on a party that has lost power and is going through the process of finding a new leader.

How, for instance would the Tories have performed in a similar survey eleven weeks after their 1997 election defeat by Tony Blair’s new Labour or in the aftermath of GE2001 when IDS, Ken Clarke and Michael Portillo were slugging it out. In the latter the blue team ended up with the leader who was the most unelectable – something that was blindingly obvious to many inside and outside the party

Inevitably leadership contests highlight divisions because that’s their very nature and we know that voters are more reluctant to give their support to split parties.

The big question is how LAB will be seen when the new leader is in place.

Mike Smithson


Why getting a credible leader is so important to LAB: YouGov polling on why the party lost

Monday, July 27th, 2015


The data that underlines the importance of the current election

Whichever of the four ins he/she will have to be perceived a lot better than Ed was if the red team is to have any chance whatsoever.

This polling should be at the heart of the leadership campaign. A non-credible leader means a likely third consecutive general election defeat.

Mike Smithson


Just one in 20 Corbyn supporters tell YouGov LAB poll that the chances of him winning GE20 was a key factor

Thursday, July 23rd, 2015


The qualities LAB leadership voters are looking for – knowing how to win elections ranks 6th

Wednesday, July 22nd, 2015

So not a lot for Osbo to worry about for GE2020

Mike Smithson


YouGov Poll: Why Corbyn is winning but still unlikely to win

Wednesday, July 22nd, 2015


Keiran Pedley on the sensational survey for the Times

Last night, YouGov released a poll on the Labour leadership that has thrown something of a hand grenade into the contest. After rumours that private polling was showing Jeremy Corbyn ahead we now have a poll showing exactly that. In fact, the first preference numbers in this poll are:


Not only does Corbyn win on first preferences (and by a country mile) but on these numbers he actually wins the entire contest, beating Andy Burnham by 53% to 47% in the final round.

When sensational polls such as these are released they usually cause a sensational reaction. We can expect nothing less from this poll but it is worth pausing and considering how much weight we should place on these results. At the time of writing we have limited information on methodology and no data tables so for now here are some thoughts which I will likely update at a later time once we have more information.

The problem with this poll is that we are not likely to get many others to compare it to. It will stand alone as the voice of the Labour (s)electorate until there is another one. Given the results, it will undoubtedly shape the immediate agenda in the Labour leadership race as candidates seek to respond to these numbers and establish or re-establish momentum.

On this point I have some sympathy for Lord Foulkes. This one poll will shape the immediate political agenda but as any pollster will tell you it is still ‘just one poll’. In an ideal world there would be several others to compare to (though I suspect this is unlikely). Corbyn’s lead could be exaggerated. We have learned from the Scottish referendum and General Election that focusing on one poll showing ‘Yes’ ahead or a couple showing ‘The day the polls turned’ is folly. Perhaps the media hasn’t learned.

It is worth remembering that this is a very difficult audience to poll. Getting a representative sample of the wider electorate is hard enough (as we pollsters well know by now) but doing so of those choosing the leader of a political party is even harder.

There are many variables to control for. For a start, we have the usual issues of age, gender and region for which YouGov would need to know the makeup of the Labour membership. This ought to be fine if you have that information but there are other more complex issues to address such the balance between affiliates, Trade Union members and party members and also when Labour members joined the party. All of which, if skewed, could impact the final result. Also, as an aside, building a large enough body of Labour members on an online panel is quite a challenge. Personally, I would love to know how many they have.

It is very brave of YouGov to release a poll like this at this time. I know the team at YouGov, if anyone can do it they can but it is a very difficult job. After all, the pollsters face a tough time at the moment given the 2015 General Election polling disaster and subsequent BPC inquiry. It will be interesting to see if they release any more of these and how close to the day of the result. Of course, there is not a ‘polling day’ as such so it is different to General Election polling.

Corbyn ahead – for now

So what do the results actually mean? Well clearly we should take the prospect of a Corbyn win seriously now. It could be the ‘vent before the vote’ of course but it is clear that the Labour membership has little time for a move to the centre. I have made my thoughts on this known here. It is, however, worth considering that even being so far ahead on first preferences Corbyn only narrowly wins overall. This means he must suffer once second preferences are reallocated and that he needs a big lead on first preferences to actually win. It could be that this poll overstates his position, or that his lead decreases over time. Either way, although this poll is sensational, I still do not think he will win. The difference now is we really cannot be sure of that.

Keiran Pedley is an elections and polling expert at GfK and tweets about politics at @keiranpedley


The British Election Study suggests that differential turnout the most likely cause of the GE2015 polling failure

Thursday, July 16th, 2015

IMG_20150513_104124 (1)

LAB supporters more likely to have overstated certainty to vote

There’s an important paper just out from the British Election Study on what went wrong with the polls on May 7th. Why did it happen?

The report by Jon Mellon and Chris Prosser does not give much credence to the “late swing”, “don’t knows” and “shy Tory” theories and argues that differential turnout was most to blame. Those over-stating their likelihood to vote were more likely to be those saying LAB.

A key part of the examination has been to check whether polling respondents who said they would voted actually did so. They do this by looking at the publicly available marked register. They also looked out for other inconsistencies. Some interesting numbers emerged.

  • 20% of respondents in areas without local elections claim to have voted in them in 2015
  • 3-6% of respondents in the campaign wave of BES polling claim to have voted by post before the postal ballots were actually issued
  • 46% of respondents who could not be verified as registered to vote in June 2014 claim to have voted in the 2014 European Election

  • In all of these cases, the fibbers leant significantly more Labour than other respondents.

    The report goes on:

    “The evidence in the BES suggests that the reason for the increased impact of differential turnout is not due to a change in the relative enthusiasm between Labour and Conservative supporters since 2010. 84% of Labour supporters in 2015 said that it was “very likely” that they would vote, compared to 86% of Conservative supporters, while in 2010 the figures were 87% and 90% respectively. Rather the data suggest that the increase in the turnout gap between Labour and the Conservatives can be explained by shifts in party support amongst those who are actually less likely to turnout to vote, even if they say they will. This evidence strongly suggests that differential turnout was a major factor in the polling miss.”

    Overall they say “this is relatively good news for pollsters. It should be possible for pollsters to fix many of their by using turnout weighting that accounts for the wider set of factors we have identified.”

    Mike Smithson