#EURef note: BIG new poll from Survation!

As we sit in the middle of a key, cold month in Cameron’s botched re-negotiation plans, we’re given a BIG new poll by pollsters Survation. We offer this brief note on the key parts of it

Why is it BIG?

Literally, this poll is a whopper! Simply as it’s based on a sample of 10’000 people – ten times larger than your average poll of political preferences. From such a large poll we can extrapolate some key points.
First the figures. With undecideds not included the ‘outies’ lead with 51% to the ‘innies’ 49%. With the undecideds included, the numbers break down as 42% out, 40% in. 
(Tables here: EU ref poll Survation Dec 15)

Key points

It is important to note that these leads are not statistically significant; within margin of error sort of stuff. The main pull out from this poll is the degree to which these mirror Survation’s (and others’) previous much smaller polls, both in the top line numbers and the breakdowns. It shows little movement across the key demographics, so at least the two sets of campaigns will know what to target. Not only do the innies and outies appear to split in the same way, but the percentage of undecideds also mirrors that of previous polls.
The one key difference between this poll and previous ones is that Survation have now decided to poll in Northern Ireland – by far the most pro-EU province of the UK  (see Chartist’s note on this from September’s blog on Survation’s EU polling).
This has not altered the national numbers. One note however, all the non-English parts of the UK are far more pro-EU than England. The size of the electorates across the four nations might be marginally different, but English respondents, a snip more outie than innie (40 to 39% in most of the tables) take up 83% of the sample, but its population only 75% of the total UK citizens. If we take out total population sizes out of it and look just at electorate sizes we must ask: does England have a full 8% more of the UK’s current electing population? We haven’t fished for these numbers yet, but it would seem statistically very unlikely.
The tentative conclusion is that England has been over sampled relative to the other four nations. A quick run down of the numbers suggests it is Northern Ireland that has been under-sampled (Wales and Scotland count up ok relative to England). This is probably a symptom of northern Ireland being harder to poll, hence why we’ve got so few polls in the past with the 6 counties included.
Comments welcome. but thanks to Survation again for their work. following on from the Scottish referendum, they’re clearly making themselves a key part of our political polling and social research landscape.

Leave a comment...

This site uses Akismet to reduce spam. Learn how your comment data is processed.