Talking-up Scotland

Unusually, Kantar’s recent Westminster Voting Intentions poll doesn’t have a regional breakdown to allow consideration of the Scottish sub-poll. Readers will know that 6 out of the last 8 sub-polls have put SNP support in 45-50% range, up from the consistent 40-41% range in the previous 6.

When I asked to see the data, I was told:

‘Unfortunately, in each of our polls we have under 100 respondents from Scotland. This means that the voting intention estimates are very imprecise (we have estimated that the margins of error around each figure could be as much as +/- 10%pts). As such I am not able to share these results externally.’

Not able? Their Scottish subset was 102 from an online sample of 1 172 between 4^th and 8^th April 2019. The results were:

‘Kantar voting intention polls (4^th – 8^th April 2019) show Conservative 32% (-9 vs Mar 2019), Labour 35% (+4 vs Mar 2019), Lib Dems 11% (+3), UKIP 7% (+1), SNP 5% (nc), Green 4% (-2), Plaid Cymru 1% (+1), Other 6% (+4).’

https://uk.kantar.com/public-opinion/politics/2019/more-than-half-of-uk-public-want-final-brexit-deal-to-be-put-to-public-vote/

The 5% figure is high and caught my attention. 5% of 1172 is 58.6. 58.6 is 57% of 102.

I know, sub-polls are unreliable but with 7 out of the last 9 (nearly 1 000 people) now suggesting 45% to 50% and perhaps more, after a long period of stasis at 41%, is there really nothing here?