Reason to Doubt Stephen Flynn's Pro-Independence Polling

At PMQs yesterday the newly elected SNP Leader in Westminster, Stephen Flynn, was buoyed by news of a poll showing "Scottish independence has now hit 56% and support for the Scottish National party sits north of 50%". The poll in question came from Ipsos MORI, and you can understand why the SNP would've been pleased to promote its findings.

In addition to the strong margin in support of independence of 12%, excluding the undecideds, the SNP also sees large leads in Westminster. The poll puts them on 51%, up 6 on the 2019 election. This is a significant divergence from prior Scottish polling, on both independence and Scottish Westminster intention. As it is just one poll, and a sensational one at that, it should of course be treated with caution. Though there is further cause for scepticism.

In large part, this arises from methodology. Ipsos doesn't weight for past voting intention, be it for Holyrood, Westminster or independence. Whilst there are reasons for this, namely false recall, it can introduce some pretty dramatic sampling discrepancies.

In this case, the poll appears to significantly oversample both 2019 SNP and 2014 Independence supporters. Going by the recollections of those sampled in the poll, Yes would have won in 2014 by a margin of 54% to 46%. Considering the tables also show a higher proportion of former Yes voters switching to No than it does No voters switching to Yes, this should cast doubt on the headline finding in favour of independence.

While there may well still be a swing towards independence, this likely arises from changes to the electorate since 2014 - as young voters are shown to support independence by a margin of 54% to 31%. This change alone would likely be insufficient to lead to a victory for independence.

It's worth noting that some other polls have recently recorded majority support for independence. Redfield & Wilton puts the lead at 4%. Though given this is the only other national poll on Independence since Liz Truss resigned, it's far from a conclusive picture. Especially when "No" did have a somewhat consistent lead beforehand.

At the moment, the field of Scottish polling is sparse, especially considering the pace of political developments of the past few months. It will be interesting to see how the picture changes as more surveys are conducted and released. While it is easy to imagine independence support making gains, this new Ipsos poll should be treated with caution.