Constitutional preferences (Yes or No to independence) in Scotland invoke polarisation far more than they do consensus. Identifying a middle ground on preferences might be difficult, but more than a decade on the from the independence referendum it is worth checking to see whether there is a middle ground on issues of referendum process, mandate and timing, particularly given the First Minister’s recent attempts to finesse a position on the matter. To assess this, the June 2025 Scottish Opinion Monitor survey (Scoop) fielded by the Scottish Election Study team included a range of questions about constitutional process.
This included questions about when Scots want a future independence referendum and who has the mandate to call one. Typically, in Scottish data, we see a rolling five-year window on timing. Supporters of the union generally wish it will never happen again, but among those backing change, the ideal time is in five years or so. Not now, but soon. An Augustinian “help me to be independent, but not yet”? More than a decade out from the 2014 referendum and ahead of the 2026 devolved elections, in which the issue of independence will likely feature, it is worth evaluating whether Scots’ preferred timing has changed.
Our timing question, presented to everyone in the sample of 1,204 residents of Scotland aged 16-and-over, asked whether a future referendum should be held as soon as possible after the next election, within the next two-and-a-half years, within the next five years, sometime after the next five years or never. Just under one third of the electorate said never; 18% said as soon as possible; 14% each said don’t know or within five years, or just after five years; and 11% percent said within two-and-a-half years. Put another way, just under 60% of Scots would like to see another independence referendum at some point in the future. The results vary considerably by constitutional preference. Over 60% of pro-union respondents never want another referendum, while over 40% of Yes supporters want one as soon as possible. The very polarised timing preferences by constitutional viewpoint can give an unhelpful sense of the “average” Scottish response.
We know, however, that issues of timing also relate to perceptions of mandates, specifically who has a right to call a referendum for Scotland and whether Scotland should have a say over its constitutional future in another referendum. In the June SCOOP we wished to evaluate two specific components of mandate attitudes:
- Is there a middle ground? Do Yes and No supporters entertain views other than the “never” and “ASAP” options?
- Does the framing of the referendum mandate matter?
Identifying the middle ground
To assess the mandate middle ground we split the sample in half, asking different questions of Yes and No supporters. To each side we offered a hard-line option, “as soon as possible” for Yes supporters and effectively “never again” for No supporters, but we also added a middle ground option for each:
For pro-independence respondents: “I don’t mind staying in the union for the moment so long as Scotland is able to choose whether it wants to stay or go“
For pro-union respondents: “While I support staying in the union I think Scotland should be able to choose whether it wants to stay or go at some point in the future“
In each instance, the middle ground option emphasises Scotland’s “right to choose”, which is the framing adopted by John Swinney.
The results make clear that the Scottish electorate is not split in half but into thirds, with one third impatient for another referendum, one third never wishing to see another, and one third open to another in the future when this is phrased as Scotland’s ability to choose.
The split in Figure 2 was based on current constitutional preferences, but to check whether switchers (No to Yes, Yes to No) behave differently we can check how current Yes and No supporters behave depending on their 2014 vote choice. Here we see that No-to-Yes switchers are less impatient than more consistent Yes supporters, with only 48% opting for “asap” (vs. 62% of consistent Yes supporters). These switchers were more likely to say they didn’t mind waiting in the union for a while so long as there is the option to go at some point (48% vs. 37%). Yes to No switchers are more open to holding another referendum in the future (40% vs. 21% of consistent No supporters). One important caveat here is the small sample size of switchers, fewer than 100 in each case, because attitudes to independence have remained relatively stable over time.
The importance of framing
Our second goal was to determine whether the framing of the referendum mandate mattered. To this end we created a split-half experiment with one half asked the basic question about mandates:
“Thinking about he right to decide whether a referendum on independence takes place, which comes closest to your view?“
Response options then included
- The Scottish Government should be able to decide whether Scotland holds another independence referendum
- The UK Government should have to give permission for Scotland to hold another independence referendum
- There should never be another independence referendum
- Don’t know“
The other half received the same response options but the question trunk included an additional prompt which emphasised Northern Ireland’s place in the union is rooted in the principle of consent (if ignoring the present lack of clarity on the triggers for such a referendum).
“As a result of the Good Friday Agreement, Northern Ireland can have a referendum in the future on whether it wishes to remain part of the UK. Scotland has no such agreement. Thinking about the right to decide whether a referendum on independence takes place, which comes closest to your view?“
By splitting our sample and asking these two versions of the question we can see whether referring to arrangements in other parts of the UK lead to different perceptions of mandates. The results show that there is an increase in support for a Scottish Government mandate when reference to Northern Ireland is made, but that it is not matched by a decrease in the perceived importance of a UK Government mandate. Instead, the Northern Ireland framing serves to depress the proportion of the sample that says a referendum should never be held.
Figure 3: Attitudes to mandate by experimental (red) and control (black) questions
Of course, we can imagine that the effect of this framing would be different for different groups. Yes supporters, already convinced of the Scottish government’s mandate, might be no more likely to hold this view if Northern Ireland is mentioned. To track this, we can look at the differences in support for each statement by different constitutional sides. Here we see that the shift in responses is less to do with Yes supporters reacting to the frame differently (there are modest differences between the two samples), and more because No supporters are much less likely to say “never” to another referendum when presented with the Northern Ireland frame.
Furthermore, we can examine this by strength of constitutional preference held, using the question about whether you think of yourself as a Yes or No supporter very strongly, fairly strongly or not very strongly. The samples for some groups are very small in part because so many Yes supporters are very strongly Yes, and neither side has a lot of weakly convinced adherents. Using the three largest groups (very strong Yes, very strong No, fairly strong No) we can see that the fairly strongly No group is likeliest to deviate from the control question framing. This group is almost 10 points more likely to say the Scottish Government has a mandate to hold a referendum and 15 points less likely to say a referendum should never be held, if reference is made to Northern Ireland and the Good Friday Agreement. In short, reminding voters about the existence of other constitutional arrangements tempers opposition to a referendum and does so because it encourages soft pro-union voters to depart from the tendency to say that a referendum should never be held.
So is there a middle ground on timing and mandates? On this evidence, it’s fair to say there is. When a middle ground option emphasising a future “right to choose” is shown, portions of Yes and No supporters opt for this rather than the “asap” or “never” option. Discussing it as such might take some of the polarisation out of the debate. We also have clear evidence that specific framing can help to shift supporters by invoking comparative examples within the UK. A reference to Northern Ireland’s right to choose results in lower proportions of No supporters saying another referendum should never be held, specifically because it encourages moderate No supporters to support the Scottish Government’s mandate to call one.
Of course there are other ways to identify a middle ground. On the specific issue of constitutional preferences this can include asking supporters of each side to identify the most compelling arguments made by the other. Our data do allow for this, but that’s for another blog.