Full November 2022 Scoop tables available here.
On the morning of Wednesday the 23rd of November, the Supreme Court of the United Kingdom released its judgement on the ability of the Scottish Parliament to legislate for another independence referendum without Westminster’s say so. To cut a very long story short, the court decided that this move wasn’t within Holyrood’s purview and that UK legislative consent, which had been obtained before the first referendum in 2014, would again be required for any fresh constitutional vote.
This put paid to the Scottish Government’s stated plan to hold another referendum in October 2023, since there is no realistic chance that the current UK government will allow another ballot on independence. Unless the expected 2024 Westminster election results in a hung parliament in which the Scottish National Party holds the balance of power, there is little prospect of any majority Conservative or Labour administration down south permitting such a contest this decade.
A cynic might suggest that this was part of the plan all along. Backed into a corner by a succession of her own campaign promises and impatient members of her constitutional tribe, Scotland’s First Minister Nicola Sturgeon had little choice but to take some concrete steps towards a referendum – despite public opinion remaining on a knife-edge and even many of her party’s supporters demonstrating reluctance to hold a fresh constitutional ballot on the proposed timescale.
Scotland’s leader is a pragmatist and, as such, understands that a premature second referendum loss, however narrow, could kill off the prospect of independence for good. The example of the Canadian province Quebec – where the pro-sovereignty side lost a second referendum on statehood by just 1.2 percentage points in 1995 and withered thereafter – looms large in the mind here. While senior SNP figures would doubtless deny this, it is unlikely many relished the prospect of another tumultuous referendum campaign at a moment of such domestic and international difficulties in the absence of a convincing and sustained polling majority for Yes.
As such, seeking the opinion of the Supreme Court was probably the nationalists’ least-worst option for now. Given the widespread assumption that the court would indeed rule that the referendum couldn’t be held (whether at this stage or after the legislation had passed through Holyrood), the question of legislative competence is at least settled and the Scottish Government has pivoted to arguing on the basis of “democracy” and the limitations on Scotland’s ability to make laws for itself. Out of options, the nationalists have also signaled that they intend to fight the next UK general election as a “de facto referendum” and, presumably, the democratic deficit will feature heavily in the campaign.
How effective might the argument argument be? This is one of the key questions for the indy movement. After all, it has some basic substance. The will of the democratically elected Scottish Parliament is currently being frustrated; pro-independence parties which stood on a manifesto pledge to hold an independence referendum control a majority in the chamber yet they can do nothing about it. Nationalists with an eye on the state of public opinion might well hope that this judgement is precisely what they need to persuade constitutional switherers.
We have previously discussed how entrenched each side of the Yes/No debate is. There are relatively few voters still up for grabs. That said, the Scottish electorate’s response to the COVID-19 pandemic – during the early stages of which both the SNP and Yes enjoyed substantial polling boosts – demonstrated that public opinion on independence does indeed move at the margins in response to real-world events. It also remains true that “don’t knows” hold the balance of power between the two constitutional options. It is not committed Yes or No supporters who will ultimately decide Scotland’s fate, but those who are currently torn on the issue.
Helpfully, we can shed some light on the immediate reaction, or lack thereof, to the Supreme Court’s judgement, since data for our latest Scottish Opinion Monitor (Scoop) survey was collected both immediately before and after the news broke.
Our new representative sample of 1,210 Scottish voting-age adults (16+), administered by YouGov between the 22nd and 25th of November, contains 382 respondents who started the survey before the judgement and 828 respondents who started the survey after the judgement. In social science terminology, we might call this a “quasi-experiment”, since a major intervention occurred during the course of data collection, effectively splitting the sample into two groups which were randomly exposed to different real-world conditions. This means the latest Scoop sample is an ideal testing ground for the idea that the argument about a democratic deficit had some impact, since there was a great deal of media coverage of the event and we would expect to see the biggest impact in the immediate aftermath of the decision.
It is important to note, firstly, that baseline support for independence (using question wording identical to that of the 2014 referendum) was higher overall in November than August, the time of our last Scoop survey. Yes actually pulled ahead this time around – a Scoop first in four surveys since December 2021 – albeit by an extremely narrow margin. Of 1,042 respondents who chose either constitutional option, 526 selected Yes and 516 selected No. With don’t knows removed and population weights are applied, this comes out at 50.2% Yes. It goes without saying that this result is within the margin of error, so this essentially represents a statistical tie. It is a modest improvement for Yes over the August survey, which showed 47.6% in favour of independence (once again with don’t knows removed and weights applied).
It is fair to say therefore, that between August and November, based on SES data, public opinion on Scotland’s future moved from a likely slight advantage for No to a coin-flip. Other opinion polls have shown something similar, or indeed relatively strong Yes leads, in recent weeks. Did the Supreme Court decision – the final nail in the coffin of a second referendum – really tilt the balance of public opinion in such a polarized environment? In recent weeks Yes has regained the advantage in opinion polling and headlines have attributed that rise to the judgement. We can test this directly, since two thirds of our sample was exposed to the news while one third was not, and we can compare the two groups’ levels of support.
Independence support from the Scottish Election Study’s last six surveys is shown in the table below. We can observe here that movements tend to be slight compared to, for example, the relatively volatile state of UK general election polling over a similar timespan. This is because, as we have discussed before, preferences on the issue of independence are very entrenched. However, this also means that small movements can be decisive. The latest Scoop is a case in point, with a modest pro-indy swing tilting the balance.
|SES 2021 Wave 1||April 2021||42.6||46.4||11|
|SES 2021 Wave 2||May 2021||41.7||48.4||9.9|
|Scoop 1||December 2021||43.1||47.2||9.6|
|Scoop 2||April 2022||43||47.9||9|
|Scoop 3||August 2022||42.9||47.3||9.8|
|Scoop 4||November 2022||45.7||45.4||8.9|
Respondents who indicated they would not vote removed in all cases
To what extent was this down to the news of the judgement? According to our data, if the story did have any impact at all, it was a very small one. We can look at the figures two different ways, with and without population weights. If we examine the unweighted figures – that is, those without statistical adjustments made to the sample to make it more closely resemble the population as a whole – indy support was a dead-heat prior to the decision and Yes takes a very slight edge after it. In this case, it would appear that a small number of undecideds broke in favour of independence. When weights are applied, the shift looks a bit bigger, with No leading among respondents before the decision and Yes afterwards. However, we have to be very cautious about interpreting this, since it may simply reflect slight differences in the sample composition of pre- and post-judgement groups.
Is this very modest shift a statistically meaningful one? In other words, is it distinguishable from a chance difference? However you slice it, the answer seems to be no. Chi-squared tests of the differences in vote intention between the two groups do not rise to the level of statistical significance. This also applies when we split the sample into three groups, slicing it up by responses from before the verdict, responses from the day of the verdict (after 10AM) and responses from the day after that (with a small number from the next morning as well).
The apparent lack of any statistically significant shift in opinion on independence also holds when we examine a much finer-grained measure of Yes/No support, a 0-10 scale running from “Definitely vote No” to “Definitely vote Yes”. T-tests of the difference in means between the before/after groups show no statistically significant difference between the two. Indeed, this measure shows that average support for independence was actually very slightly lower following the verdict, even though Yes pulled ahead in the binary vote intention question.
The graphs below illustrate how little changed in our sample’s average 0-10 indy support measure before and after the announcement, shown as daily groupings of the whole sample in the first figure and before/after groupings by indyref vote intention in the second. The only change which comes close to conventional statistical significance is a slight uptick in the 0-10 Yes/No measure among those who responded “Don’t know” to the vote choice question. But, again, the absolute gap in means is very small and it is not statistically distinguishable from a chance difference. If the most we can say about the Supreme Court decision is that it may have encouraged some undecided voters to view the Yes side more favourably, we’re not saying much at all. Once again, the story is overwhelmingly one of polarisation and entrenchment.
All of that said, headline support for independence does appear to have meaningfully increased since the last Scoop. What alternative explanation might there be? Pro-independence sentiment reached all-time polling highs in the early stages of the covid pandemic in mid-late 2020, when diverging perceptions of UK and Scottish Government handling of the crisis bolstered Yes support. It is not much of a stretch to think that something similar occurred in Autumn 2022 as the Conservative UK government descended into political chaos of its own making and the disastrous 44 day Truss Premiership permanently exacerbated the painful cost of living crisis engulfing the country.
To put this another way, undecided voters or those with “soft” constitutional preferences appear to move in favour of indy when the UK state exhibits instability and Conservative behaviour reinforces popular narratives about the party’s right wing ideological zeal and detachment from Scottish priorities. Although we cannot directly test this with our own data, it seems a more feasible explanation. Indeed, there was a fairly sizeable shift in both Scottish and UK government competence assessments between August and November as shown in the table below. Evaluations of UK government performance declined by around 15 points on net in the intervening time, while the Scottish government gained around half of that amount. This increased the Scottish Government’s relative advantage in net government approval ratings – the gap between perceptions of the two administrations – from 51 to 73.
|Good||Neither||Bad||Don’t know||Net (good-bad)|
|UK Govt. (Aug)||12.5||11.3||69.2||7||-56.7|
|UK Govt. (Nov)||5.3||10.6||76.3||7.8||-71|
|Scottish Govt. (Aug)||33.9||17.7||39.9||17.7||-6|
|Scottish Govt. (Nov)||37||18.7||35||9.4||+2|
Weighted figures shown
Of course, as stated, the nature of our data means we cannot directly estimate the full causal impact of these competence assessments on indy support. We also cannot rule out the possibility that the SNP’s appeals on the basis of the democratic deficit will not be persuasive over a longer timeframe or that the Supreme Court decision’s fallout served to raise the salience of the issue in such a way as to highlight perceived differences in UK and Scottish Government performance.
However, based on mounting circumstantial evidence and the apparent lack of immediate respondent reaction to the Supreme Court decision, it does seem to be the case that independence vote intention among the relatively small number of “persuadable” voters is more responsive to the perceived differences between the UK and Scottish Governments than relatively abstract procedural arguments. This obviously raises questions about the durability of Yes support if and when Labour enter government at Westminster. The permanence of the damage the Conservatives have done to the union in recent years will only become clear when they leave office.