Uncategorized

Latest Westminster and EU Referendum Evidence

Amidst all the excitement of last week’s National Assembly election, and the on-the-day poll that we conducted with YouGov, some things got missed out from last week’s final pre-election Welsh Political Barometer poll. We once again asked people our standard questions about how they would vote in a UK general election, and about their voting intentions for the forthcoming EU referendum. But unless I missed it – which is very possible given everything else that was going on – then I don’t think these results have been reported or discussed. So a brief blogpost for you on those.

First Westminster. These were the figures for voting intention, with changes on the previous Barometer poll in brackets:

 

Labour 36% (-1)

Conservative 25% (+2)

UKIP 17% (no change)

Plaid Cymru 12% (-1)

Liberal Democrats 7% (no change)

Others 3% (no change)

 

In short, we see nothing but minor changes since the previous poll in late April, all well within the ‘margin of error’. The small uptick in Conservative ratings form the previous poll was very much in line with the changes we also saw at the time for the Assembly. Given what we now now about the Assembly election result, it would seem most likely that these small differences simply reflect sampling variation between the late-April Barometer poll and this one conducted in early May. The latter poll simply happened to produce a slightly more pro-Conservative sample, and one that was marginally less favourable to Plaid Cymru, than did the previous Barometer poll. That’s polling; it happens.

Anyway, for what it is worth (and I’ll be publishing soon a detailed review of the performance of uniform national swing, and ratio swing, in the Assembly election) if we apply the changes since the May 2015 general election implied by this poll uniformly across Wales, we get the following projected result (with all seats won by a party at last year’s general election remaining in their hands unless stated otherwise):

 

Labour: 26 seats (losing Ynys Môn but gaining Gower and Vale of Clwyd)

Conservative: 9 seats (losing Gower and Vale of Clwyd to Labour)

Plaid Cymru: 4 seats (gaining Ynys Môn)

Liberal Democrats: 1 seat

 

Under the alternative assumption of Ratio Swing, the outcome is identical except for  Ynys Môn being projected to be narrowly held by Labour, rather than narrowly lost as under UNS.

 

For the EU Referendum, these were the figures (with changes from the previous Barometer poll in brackets):

 

Remain: 41% (no change)

Leave: 39% (+2)

Would not vote: 6% (-1)

Don’t Know: 15% (no change)

 

So once again these are only minor changes, well within the ‘margin of error’. Moreover, the slight increase in the Leave share here would be very consistent with the changes in the party support levels discussed above: getting a more pro-Conservative sample in this poll would probably also help lift the Leave share very slightly. The broad picture looks like one of essentially no change between these two polls.

 

To help you see the trends in EU Referendum preferences over time, I’ve put together this chart (click on the chart for a larger version):

(The chart shows responses to all polls with an EU Referendum voting intention question run by YouGov since February 2013. Where more than one poll was run in the same month, I have averaged the figures for that month).

Comments

  • John R Walker

    But what you’re not telling us that YouGov are not terribly confident they are calling it right in the EU ref polls.

    Top Pollster: ‘I’m Not Confident Brexit Result Will Be As We Predict’

    http://www.breitbart.com/london/2016/05/10/top-uk-pollster-im-not-confident-brexit-result-will-predict/

    I suspect they’re under-reading the anger against the EU and our political class. And the Commonwealth immigrants who are now 2nd class citizens when it comes to further immigration are seriously unhappy… Nobody is obviously factoring them in or trying to reach them from either side as far as I can see. And they do tend to vote! Matt Singh is more likely to understand this…

Leave a Reply

Your email address will not be published. Required fields are marked *