So far, two Scotland-only polls of vote intentions in the Westminster election on December 12th have been published. One, by Panelbase for Sunday Times Scotland was released in two tranches over the last two weekends. The other, by Ipsos MORI, was published by STV News last Thursday – the first poll from this stable for 18 months.
Between them they contain two seemingly contradictory messages. On the one hand they uncover signs of a revival in Conservative support. On the other hand, they also confirm that support for Scottish independence has risen.
Just a few weeks ago, shortly before the UK election was called, a poll conducted by YouGov painted a depressing picture for both the Conservatives and Labour. The Conservatives were credited with 22% of the vote, seven points down on what the party achieved in 2017, while Labour were put on just 12%. Indeed, Labour’s rating put the party one point behind the Liberal Democrats and thus in fourth place for the first time in Scottish polling history. The SNP were estimated to be on 42% of the vote, up five points on 2017.
Given that most seats in Scotland are marginal, these figures suggested the SNP might win a barrow-load of seats – perhaps as many as 50 MPs.
Moreover, these figures largely confirmed a picture that had been in evidence ever since the European election in May. Polls since then had on average put the Conservatives on 20% and Labour on 17%, while the SNP were on 41%.
However, once the election was called, the early weeks of the campaign were graced by Britain-wide polls that suggested that support for both the Conservatives and Labour was increasing. Meanwhile, small though the sample sizes were, the detailed figures for Scotland in those Britain-wide polls appeared to suggest that, in the case of the Conservatives at least, the trend was being replicated north of the border.
That expectation has, indeed, been confirmed by the two new Scotland-only polls. Panelbase put the Conservatives on 28%, down just a point on 2019. Ipsos MORI’s figure of 26% is a little less optimistic for the party, but still represented an improvement on any other poll reading in 2019.
Labour, in contrast, have advanced rather less than might have been anticipated given the party’s progress in the Britain-wide polls. Panelbase give the party 20%, while Ipsos MORI put the figure at just 16%. Meanwhile, both polls put the SNP on a healthy 40% and 44% respectively, while at 11% on both polls, the Liberal Democrats have eased back slightly.
These figures are largely replicated in an analysis of a mega 100,000 respondent poll conducted by YouGov and analysed using multi-level regression and post-stratification (MRP), a statistical technique that, inter alia, produces estimates of the outcome in individual constituencies. It suggests the Conservatives are on 28%, Labour on 18%, the SNP on 41% and the Liberal Democrats on 11%.
The improved figures for the Conservatives are reflected in the seat tally to which the polls now point. On the basis of the average of the two Scotland-only polls – and assuming that the change in party support that this implies is uniform across Scotland – the Conservatives could hope to retain eight seats, while the SNP would expect to win 46 seats, the Liberal Democrats four, and Labour just one. Meanwhile, YouGov’s modelling, which takes into account possible geographical variation in the ups and downs in party performance, suggests that the Conservatives could be heading for as many as 11 seats, with the Liberal Democrats down to three and Labour to two, leaving the SNP on a more modest 43 seats.
The difference between the two estimates of the possible outcome in seats, even though they are based on similar shares of the vote, is a reminder of how, when a country is full of marginal constituencies, even a modest amount of geographical variation in party performance can make a significant difference to the outcome in seats.
It might be thought that an improvement in the Conservatives’ electoral position would be accompanied by a rise in support for Scotland remaining part of the UK. Of this, however, there is no sign. As we have noted previously, polls conducted this year have consistently pointed to an increase in support for Yes in any second independence referendum. On average across seven polls conducted between April and the beginning of the UK election campaign, support for Yes averaged 49% (after excluding Don’t Knows), suggesting that the outcome of any such ballot would be on a knife-edge. That picture is confirmed by the two latest polls, with Panelbase putting support for Yes (after excluding Don’t Knows) on 49% and Ipsos MORI on 50%.
The data from both polls are also consistent with previous poll findings that the increase in support for independence has occurred primarily among those who voted Remain – and thus would seem to have been occasioned by the continuing debate about Brexit. Panelbase put support for Yes among those who voted Remain at 57%, while Ipsos MORI reckon it is 56%. In contrast, the two polls put support for Yes among those who voted Leave at just 33% and 30% respectively.
Rather than reflecting any change in attitudes towards the constitutional question, the increase in Conservative support has arisen because the party has improved its popularity among those who already back the party’s views – not only on the constitutional question but also on Brexit. Among those who voted No in 2014, Panelbase register a ten-point increase (from 36% to 46%) in support for the Conservatives since the company’s previous poll in early October. There has been just a two-point increase (from 4% to 6%) among those who voted No. Equally, support for the party among those who voted Leave is up from 40% in October to 55% now, whereas among Remain voters the equivalent figures are just 9% and 13% respectively.
The picture painted by Ipsos MORI is very similar. They report that the Conservatives enjoy 46% support among those who voted No in 2014 and just 5% among those who voted Yes, while the party has the support of 60% of those who voted Leave in 2016 but only 14% of those who back Remain. Rather than significantly expanding the size of the niche market to which it appeals north of the border, the party has simply managed to restore its popularity within that market.
However, the sharp Conservative advance among those who voted Leave is not without its significance. It is part of a broader picture whereby how people are minded to vote reflects their views on Brexit more strongly than was the case in 2017. On that occasion, the British Election Study internet panel estimated that 46% of people in Scotland who voted Leave in 2016 backed the Conservatives in 2017. Now, our two campaign polls point to an average figure of 58%, up 12 points. In contrast, at 14%, the current level of support for the party among Remain voters is four points down on 2017.
Conversely, on the same basis the Liberal Democrats have added some seven points (to 15%) to their tally among Remain voters, but slipped back two points (to just 3%) among Leave supporters. The SNP, with 51% support (up 10 points) among Remain voters but just 23% (down 2) among Leave supporters also seems to have found it easier to fish in the waters of Remain voters.
Indeed, the strength of the SNP among Remain voters helps explain why Labour has struggled to make much progress north of the border. In England & Wales Labour has gradually been scooping up some of the Remain vote as the election campaign has progressed. However, the SNP’s dominance of the Remain vote in Scotland leaves the party with much more limited scope for advancing in that way. Indeed, at 20% the party’s current level of support among Remain voters is not dramatically above the 14% the party enjoys among Leave voters.
However, in bridging the Brexit divide Scottish Labour remains the exception rather than the rule. For the most part, Brexit appears set to join the constitutional question as a major influence on how Scotland votes on December 12th.