Edinburgh Research Explorer Independence is not going away

Support for Scottish independence has remained stable since the change in the leadership of the Scottish National Party in March 2023, despite the fall in support for the party. The article analyses the sociological basis of independence support since 1979, using data from the Scottish Social Attitudes Survey and recent opinion polls. The recent stability re ﬂ ects three main changes: the pronounced growth in support among people born since the 1970s; the growth of education in these same cohorts; and the shift to the left of the independence campaign. These long-term trends suggest that the level of support for independence, and of opposition to it, are unlikely to be affected strongly or permanently by the transient fortunes of the SNP.


Introduction
THE CAMPAIGN FOR Scottish independence may be interpreted in two quite different ways. Which of these is the better description will determine what happens over the next few years, especially if the Scottish National Party (SNP) declines electorally and if the Labour Party forms the next UK government.
The first story is the more familiar externally and fits more readily into the much wider theme of what Roger Eatwell and Matthew Goodwin call 'national populism'. David Goodhart sums up this view succinctly: just as 'Brexit was a movement to reclaim control/sovereignty from a supranational EU', so Scottish nationalism 'is a movement to reclaim control/sovereignty from a multinational United Kingdom'. John Lloyd makes a similar point, adapting the slogan of the Brexiteers across the UK: nationalists in Scotland, as in other small nations across Europe, 'wish to "take back control" after, in many cases, several centuries of losing it'. If national populism is, in the words of Cas Mudde and Crist obal Rovira Kaltwasser, 'a confrontation between "the people" and "the establishment"', then in Scottish nationalism the people of Scotland confront the establishment at Westminster. Because that kind of rhetoric comes naturally to nationalist politicians in Scotland-and even to Scottish Labour politicians when faced with a Conservative UK government-it seems that nationalism in Scotland might be similar to that which led to Brexit, or which has supported such politicians as Marine le Pen or Giorgia Meloni. In a British context, the consequence would be that Scottish nationalism was a Scottish version of Brexit nostalgia, even though opposed to Brexit. It would be a movement of the old, the uneducated, the disengaged. 1 The obvious difference, however, is social ideology, or at least ideological rhetoric. Since the 1980s, Scottish nationalism has positioned itself on the left. However vaguely defined that term can be (and we return to definitions later), in this second story the Scottish people are enlightened and the English establishment is conservative. Scotland is young; England is old. Scotland is liberal; England is reactionary.
1 Scotland is on the side of history; England is steeped in post-imperial nostalgia.
This second story is much older than the recent rise to power of the SNP. The ideological formation of current Scottish nationalism is in the intellectual and cultural ferment of the 1970s and 1980s when a supposedly new Scotland was presented as an invigorating escape from stuffy old Britain, with its sclerotic institutions, its impermeable social hierarchies and its insufferable hypocrisy. The most influential of these theorists was Tom Nairn (who died this year), bringing the allure of the international new left to what had previously been Scottish nationalism's provincial dullness. Radical writers sought to link Scottish rebellion to anti-colonialism. Franz Fanon's notion of 'inferiorism' was invoked to explain Scotland's failure to challenge cultural norms emanating from the English metropolitan core. What seemed to be a new Europe of the regions after the Maastricht Treaty of 1992 became a way of linking Scottish independence to a European future, taking a modernised inspiration from the liberal nationalism of the mid-nineteenth century. It encouraged a story that Scotland had disastrously failed to be inspired by Europe's springtime of nations in 1848, which Nairn and others presented as the moment at which Scotland had chosen introverted backwardness over independent modernity. 2 The result was that, by the 1980s, almost all active support for any kind of elected Scottish parliament was on the left rhetorically, focussing on opposition to the governments of Margaret Thatcher. The shift happened first in support for devolution, attracting to it the whole panoply of post-1960s radical positions-the women's movement, the movement for sexual liberation, anti-racism, environmentalism. When the first devolved governments after 1999-led by the Scottish Labour Partyturned out to be as pragmatically reformist as UK Labour governments had always been, the most radical of these radical movements moved close to the new SNP government in 2007, followed naturally by their giving support to independence in the 2014 referendum. That is what then gave the independence side in that referendum an aura of national liberation, a sentiment that has persisted and that has separated itself from the SNP whenever that party in government has shown signs of being merely pragmatic.
Are these radical activists deluding themselves about the ideological character of the people whom they claim to lead? In assessing what might happen next to Scottish nationalism, this question is crucial. In a sociological sense, is nationalism based on young, radical idealism, or is it, despite the activists' selfbelief, a local instance of what has been called the movement of the left-behind?

Sources of evidence
The statistics analysed below come mainly from the series of Scottish Social Attitudes Surveys, which have been run by ScotCen (the Scottish arm of NatCen Social Research) almost every year since the Scottish parliament was established in 1999. 3 The series analysed here covers the years from then up to the referendum on the UK's membership of the European Union in 2016. Three earlier time points allow these results to be placed in the longer-term context of the modern campaigning for some kind of Scottish self-government-the Scottish election surveys that took place at the time of the UK general elections of 1997, 1992 and 1979. All these surveys were based on probability samples of people aged 18 or older (though 16 or older in 2016), and were conducted by in-person interviews in respondents' homes (though in some cases also with a self-completion supplement that was to be mailed to the survey organisation after the interview). The sample sizes were around 1,300 in each of the surveys 1999-2016, 780 in 1997, 811 in 1992 and 612 in 1979.
The surveys gathered views on a wide range of political issues and social questions, as well as the usual demographic information. Appropriate details are given below. The main question for our purposes is whether the respondent supported independence. This was derived from a question about the preferred constitutional future for Scotland. For 1997-2016, five response options were offered: independence outside or inside the EU; devolution with or without taxation powers; and no elected Scottish parliament. In 1992, there were four options (omitting devolution without taxation powers). In all these surveys, independence support is here defined to be choosing either of the first two options. In 1979, only one independence option was offered, along with a parliament responsible for most Scottish affairs, an assembly responsible for some affairs, and no elected assembly. Independence support in 1979 is defined to be that unique option.
The analysis here of these surveys is developed more fully elsewhere. 4 For some purposes, we can bring the story right up-to-date by evidence from opinion polls conducted in 2023, results from which are taken from the What Scotland Thinks website. 5 The percentage supporting independence is calculated after excluding people who did not give an opinion or would not vote, which was about 6 per cent of all respondents in each social attitudes survey, and around 19 per cent in each poll.

Independence support and birth cohorts
Overall support for independence was 8 per cent in 1979, rose to 23 per cent in 1992, and then fluctuated around 30 per cent between 1997 and as late as 2013. During the summer leading to the 2014 referendum in September, it sharply rose to 45 per cent, where it has mostly remained since, though it was also consistently above 50 per cent in 2020-21 during the disruption caused by Covid. For most of this time, men had a few percentage points higher support than women, but that difference vanished in the aftermath of the referendum.
That much is well-known. Also much discussed has been the difference in support by age. This pattern, too, has evolved over time, but it is useful initially to look at how clear it became at the referendum in 2014, and later. In the surveys 2013-16, average independence support was 38 per cent, which in that period happened to be the same as for people born in the decade 1957-66. The proportion among people born 1977-86 was 44 per cent. Those born 1927-36 had 21 per cent. In cohorts between these two, there was a steadily rising percentage from older to younger. Again, the male percentage in each birth cohort was a few points higher than the female. The association of independence support with youth was maintained after the referendum. Thus, in twelve opinion polls conducted between early April and early July 2023-after Humza Yousaf had become the new leader of the SNP (and First Minister)-the average support among people born 1959-68 was 40 per cent, and for those born 1979-88 was 57 per cent. The same gradient continued for people who had been too young to vote in the 2014 referendum (for which the voting age was 16): among people born 1999-2007, the percentage support was 66 per cent. In these polls, the female percentage had moved slightly ahead of the male.
These patterns of increasing independence support down the birth cohorts seems to have been a fairly permanent state of affairs, in that it may be seen throughout the series of surveys. Around the time that the Scottish parlia- What Scotland Thinks, Opinion Polls, Scottish Centre for Social Research, 2023; https://whatscotlandthinks. org/opinion-polls/. The polls which gave information in a consistent form on age were (along with fieldwork dates in 2023): Savanta (28/3-31/3 and 9/6-14/6); Redfield and Wilton (31/3-1/4, 30/4-2/5, 3/6-5/6 and 1/7-2/7); Survation (29/3-3/4, 27/4-3/5 and 23/6-28/6)); Ipsos Mori (15/5-21/5); Find Out Now (7/6-12/6 and 13/6-20/6). In each case, the reported percentage supporting independence excludes respondents who did not give an opinion; the percentages were also either weighted by likelihood to vote, or restricted to people who would be very likely to vote. The sample sizes of the polls on which these reported percentages are based were mostly around 800-1,200. The Political Quarterly young person's ideal has been consistently maintained. It has then not been implausible to attach to the idea of independence that it prefigures the future. Nevertheless, that perception was always about moments in time, not about trends. Most cohorts recently have caught up with the support in the youngest cohorts not long previously. For example, by the time of the 2014 referendum, the support in the 1957-66 cohort (38 per cent) was as great as in the youngest cohort (37 per cent) two decades earlier in 1992.

Education
Adding to this sense that independence belonged to the future because it was an idea of young people was the sense also that it was associated with growing amounts of education, as Scotland experienced the expansion of advanced education that took place throughout the world in the second half of the twentieth century. In the oldest survey here, only 9 per cent were graduates of any kind of higher education. This rose to around 30 per cent by the late 1990s, and over 40 per cent at the time of the referendum. The proportion with at most mid-secondary school certificates fell from 53 per cent to around 28 per cent in the same four decades. In the initial years of the new Scottish parliament, support for independence was notably lower among graduates than among those in that lower-attaining group. For example, in the surveys 1997-2003, the proportions were 23 per cent compared to 33 per cent. But by the time of the referendum the difference had vanished: both these education groups had around 35 per cent support in the spring of 2014, and by 2016 the proportions were 45 per cent among those with higher education and 43 per cent among those with at most midsecondary education. The pattern persisted: in the three Survation polls in early April, early May and late June 2023, the proportions were on average 49 per cent among people with higher education, and 48 per cent among people with at most mid-secondary attainment. In an Ipsos MORI poll in late May, independence support was higher (52 per cent) among graduates than among people with no qualifications (44 per cent).
This pattern interacted with age cohort. For the cohorts born 1927-36, independence support was always negatively associated with education level and fell around the time of the referendum. Between 2000 and 2016, support among male graduates in that cohort fell from 20 per cent to 17 per cent, whereas for those with at most secondary education it fell from 27 per cent to 20 per cent. The corresponding values for women were 12 per cent and 13 per cent (graduates) and 19 per cent to 16 per cent (low education).
On the other hand, for all cohorts born from the late 1930s onward, there is a simple conclusion: there was a rise in support for independence at all levels of education, but somewhat more rapidly among those with higher education. An example is shown in the table below, for the 1957-66 cohort. Independence support among female graduates went from 28 per cent in 2000 to 46 per cent in 2016. In that same cohort, the corresponding proportions among women with at most mid-secondary education rose less steeply, from 46 per cent to 59 per cent. The patterns were similar among men: 33 per cent to 45 per cent for graduates, and 48 per cent to 54 per cent for men with minimal education.
Therefore, whereas there used to be a large education divide-with independence support very much stronger among people with minimal education than among the highesteducated-in the aftermath of the 2014 referendum that divide was much weaker. The main reason for this was generational replacement. The rapid expansion of higher education from the 1960s, and especially from the late 1980s, meant that an increasing proportion of graduates in the population were from the younger cohorts, where the graduates shared in the general tendency for these younger cohorts to have greater independence support.

Political ideology
In short, independence support has been associated with youth and with expanding education. Further reinforcing the resulting sense that independence betokened the future is that it has come to be associated with being liberal ideologically, especially among graduates. This, in turn, is also associated with educational expansion (and thus with younger cohorts) because graduates tend to be more liberal than non-graduates. Ideology is measured in the social attitudes surveys by means of two scales that are constructed from suites of questionnaire items. To measure positions on a scale of liberal versus conservative values, the scale is constructed from questions that measure attitudes to sentencing of criminals, to respecting the rule of law, to children's respect for authority and to censorship. The other scale measures values ranging from laissez-faire positions on the economy to support for government intervention, which are commonly referred to as right-wing and left-wing positions. This scale is constructed from questions that record attitudes to the social distribution of wealth, to government action to redistribute it, and to the legal and managerial power of different social classes.
The values of these scales have changed over time and differ between education categories. The average position on the liberalconservative scale moved in a liberal direction between 2000 and 2016 and younger cohorts have been more liberal than older. These changes have partly been because people with higher education have always been more liberal on this scale than people with minimal education. Change has been more complex on the other scale. Graduates in later cohorts were more left-wing than graduates in older cohorts, whereas for people with minimal education later cohorts were more right-wing than older cohorts. As a result, the ideological as well as the educational composition of independence supporters changed.
We can illustrate the resulting patterns by dichotomising the liberal-conservative scale at its median (as calculated over the survey series 2000-16 as a whole) and looking at views of independence in 2015-16 in the cohort born 1977-86. Among people with at most mid-secondary schooling, there was no reliable evidence of any association of ideology with views about independence. But for graduates, liberals were much more likely to support independence than conservatives. Among male graduates, support was 54 per cent among liberals and 36 per cent among conservatives; among female graduates, the corresponding proportions were 48 per cent and 36 per cent.

National identity
In this context of youth, expanding education and liberal ideology, the further association of independence support with personal national identity seems very far from atavistic. It is more plausibly interpreted, rather, as the association of a Scottish identity with these other predispositions. National identity was measured in the survey series by a question asking what national identity best described the respondent from the list British, Scottish, Welsh, Northern Irish, Irish, European, and Other. In every year, at least 88 per cent of people chose either Scottish or British. From 1997 onwards, 20-25 per cent chose British and around 70 per cent chose Scottish, with little variation among birth cohorts.
Not surprisingly, support for independence was much lower among people who identified as British than those who chose Scottish (a point analysed in greater detail by Alex Scholes and John Curtice). 6 This is a notable contrast with attitudes towards devolution in the 1997 referendum on that topic, where the UK Labour government had successfully presented an elected Scottish parliament as being a modernisation of British governance. Holding to a Scottish identity pulled support for independence among young liberal graduates up to the level among non-graduates in the aftermath of the 2014 referendum. In 2015-16, the independence support among women who had liberal views and a Scottish identity, and who were born in the 1977-86 cohort, was 58 per cent with higher education, 59 per cent with secondary education, and 56 per cent with minimal education. For men of these same views and cohort, the proportions were 69 per cent, 69 per cent and 65 per cent.
All these changes together had consequences for the composition of the group who supported independence. By 2016, as much as a quarter of all independence supporters were liberal graduates with a Scottish identity. The next largest was 16 per cent who were liberal, Scottish and had completed secondary education. The same was broadly true of the left-right scale. The sense is then clear of a liberal and left leadership of the independence movement among people who are well-educated and have a Scottish identity. Because the graduates were disproportionately younger, with 40 per cent having been born since the mid-1960s, that leadership also tended to be young.

Religion and ethnic group
The association of independence support with youth and, among young people, increasingly with advanced levels of education was similar in different religious and ethnic groups. In the cohort born 1977-86, and in the survey years 2013-16, the average independence support among female graduates who were Catholics was 39 per cent, almost exactly the same as the 37 per cent among female graduates as a whole in that cohort. For male graduates, the corresponding proportions were 47 per cent and 41 per cent. All these graduate percentages had risen compared to previous years and older cohorts.
The sample sizes were too small to allow the detailed study of the changing views of individual groups defined by ethnicity. However, in a combined minority group, the patterns were similar to those for the whole population. In the 1977-86 cohort, the independence support among female graduates who were in the minority category was 32 per cent, while that in the non-minority category was 33 per cent. For men, the proportions were 37 per cent and 39 per cent.
These patterns for religion and ethnicity, then, further reinforced the sense that independence was the future, overcoming some of the invidious divisions that have afflicted Scotland in the past.

The future
Events in early 2023 suggest that the movement for Scottish independence is likely to stall for some time-until either the SNP recovers from the events around the departure of Nicola Sturgeon or a new party-political embodiment of the movement emerges. In a sense, it has been stalled since 2014. What happens next depends on which of the two stories told at the beginning is the more cogent.
If the first story is true, then it would be conceivable that the whole idea of independence could retreat severely. In that story, Scottish nationalism is essentially a matter of identity, differing from the national populism that has been evident elsewhere only in being on the left: its ideology would thus be an identitarian liberalism of the kind that allegedly obsesses university campuses, but does not really have much appeal beyond that supposedly immature and introspective world. According to this view, if Labour forms a successful UK government after 2024, with an updated version of the kinds of pragmatically reforming policies that were pursued by the governments of Tony Blair and Gordon Brown, then the rationale for independence might be dented in the eyes of the wider Scottish electorate. That indeed has been the hope of Labour opponents of independence throughout the whole history of the SNP's electoral success. John P. Mackintosh, Labour MP and professor of politics, put it with characteristic clarity as long ago as 1974: Only one thing will halt or reverse the onward march of the SNP and that is a period of government in London which is really successful so that it ends with a satisfied electorate eager to vote positively for a party that has once again restored the feeling that Britain is a successful, worthwhile country to belong to. 7 That remains essentially the dominant strand in Labour's response to Scottish nationalism, asserting the value of what Gordon Brown has called the pooling and sharing of resources in the wider British polity against the introspection and potential poverty of a smaller territory. According to this view, a Labour Party that was astute enough could assert pragmatic reform against identitarian nationalism.
The history of the period 1999-2007 does not suggest that this first version of the future is likely. The many social reforms of that period by the UK Labour government did not prevent the rise of the SNP. The further rise up to 2011 indicates, at least, a widespread perception in Scotland that a merely devolved government working with a merely reformist Westminster was neither strong enough nor permanent enough to be trusted. Indeed, even at that time, with a UK Labour government, people in Scotland tended to credit any successes in policy to the Scottish government and any failures to the UK. 8 These doubts about the first story might, nevertheless, be questioned as being too tied to events, the most prominent instance of which being the controversies over the Iraq war. Without that, the SNP may well not have achieved its one-seat majority in 2007. Then, by 2011, a still dominant Scottish Labour Party-winning forty-one out of fifty-nine Scottish seats in the 2010 UK general election-might have been able to present itself as the most reliable defender of Scotland against what had become the UK Conservative-led government. The SNP might then have been no more than a strengthened pressure group and Labour could have retained a weakened version of the hegemony that it had enjoyed since the 1960s. In that case, the first story about Scottish nationalism-that it is a version of national populism and thus not fundamentally distinct in its transient identitarianism even from the conservative movements that may be described by that term-would appear to be the best explanation of why its aim of independence was too fantastical for the pragmatic Scots.
But the main reason to doubt the cogency of that first story is not the happenstance of events, but the sociological evidence that has been summarised here. These patterns of support for independence bear almost no resemblance to those described for national populism in the first story. Independence is not, in the estimation of its supporters, a right-wing or conservative cause. Unlike Brexit support in England, it is not a cause associated with the old and the low-educated. According to those who favour it, independence is about changing Scotland in a liberal direction. When David Goodhart describes the ostensibly liberal ideology of the SNP as being largely a matter of the party's leadership, he is ignoring the role of liberal, educated activists in the wider campaign for independence. 9 The evidence suggests, in short, that a version of the second story is sufficiently close to the truth as to ensure its cogency, even when the main current carrier of the idea of independence is going through the travails that have afflicted the SNP this year. Support for independence is not declining. That has been noted by many commentators since the change of SNP leadership; what the analysis reported here offers is reasons to explain the persistence. The composition of support has changed, in the sense that people who voted against Brexit in 2016 have become more likely to support independence and people who voted for it have become less so. 10 The main source of scepticism about independence has been the economic prospects, where the SNP does not seem to have persuaded a majority that Scotland is economically viable without the large transfers from the south-east of England which sustain Scottish spending at a level that is higher per head than in England. 11 8 But this is a pragmatic argument, not one based primarily on identity or emotion. It is the current version of Mackintosh's view half a century ago. It reduces the question to a matter of rational calculation and there is no guarantee that the answer will always be on the side of the Union. The rise in support to more than 50 per cent during the disruption caused by Covid was attributed to a perception that the SNP government was more assured and reassuring in its handling of the crisis than the UK government. That tends to confirm what has long been the SNP's view that one way in which support could become more consistently dominant would be a settled view that Scottish government is more competent than anything which the UK can offer. The perception that the SNP has not, in fact, been very competent in government is currently one serious source of its malaise.
In any case, the survey evidence also suggests that, by the time of the referendum, the case was gradually being made across almost all age groups. What we have described here as a cohort effect is usually discussed in public as an age effect-that young people opt for independence and the old for nostalgia and the safety of the Union. It is true that, from around 2010, the oldest cohorts (born during or before the second world war) did show declining support for independence with age, although that was when they were already in their seventies and eighties. No such age-related decline was seen in any other cohort, even the postwar baby boomers. For example, the 1957-66 cohort had 27 per cent support in the early 1990s when they were aged in their thirties. That rose steadily in the following quarter century to reach 38 per cent around the referendum when they were in their fifties, an increase by a factor of 1.4. The 1977-86 cohort already had 40 per cent when they, in turn, were in their thirties (around the referendum), and recent polls suggest that this has risen to about 57 per cent, also an increase by a factor of 1.4.
The stability in overall support for independence since 2014 does suggest that this general rise has stopped, but cohort replacement continues at its ineluctably slow pace. For example, each year at present, about 51,000 people aged over 65 die and about 57,000 reach voting age. 12 If independence support is twothirds of the younger group and only one-third of the older (as the recent polls suggest), that would add perhaps 20,000 independence supporters every year, though this would be less in practice, because turnout tends to be lower in younger groups. So, the annual rise of independence support owing to cohort replacement is probably at most 0.4 per cent of the total electorate. The resulting change in overall support is thus slow and is swamped by the usual random variation in opinion polls. But, if it continues, it would eventually not be negligible.
Despite recent suggestions to the contrary, it is unlikely that in-migration will significantly work in the other direction. It is true that, around the time of the 2014 referendum, people in Scotland who were born in the rest of the UK had quite low proportions supporting independence: in a YouGov poll in early May 2015, 28 per cent of these migrants supported independence, compared to 49 per cent of people born in Scotland. Because net migration from the rest of the UK runs at about 10,000 annually, this differential could have become significant over time. But, even in that poll, there was no such difference among migrants from outside the UK (with 48 per cent support), and the net in-migration from that source is around 20,000 each year. In any case, the difference in support among migrants from the rest of the UK seems to have shrunk. In a Panelbase poll of voters in Scotland in late March 2023, independence support among people born in Scotland was 49 per cent, while among those born in England it was 41 per cent (and for those born elsewhere it was 53 per cent). In an Ipsos MORI poll of mid-May 2023, the proportion among the Scotland-born was 12 National Records of Scotland, Vital Events Reference Tables 2021 , Table 5.02; https://www.nrscotland. gov.uk/statistics-and-data/statistics/statistics-bytheme/vital-events/general-publications/vitalevents-reference-tables/2021/; Office for National Statistics, 2020-Based Interim National Population Projections; https://www.ons.gov.uk/peoplepopulation andcommunity/populationandmigration/population projections/bulletins/nationalpopulationprojections/ 2020basedinterim. 54 per cent and among those born anywhere else it was 50 per cent. 13 The ideological configuration that has developed in the past several decades has built on the older legacy from the second story outlined above. It is a utopian vision of Scotland as a liberal haven and is thus easily dismissed as naïve by its opponents. But it is a vision that has been so long-lasting-far back beyond the pragmatism of devolution-that it seems unlikely to wither away. 14 It has behind it, now, the liberal optimism of younger generations who have levels of education that their forebears lacked and who thus have a capacity for leadership that, until 2023, has enabled the SNP to sweep all before it electorally. If the SNP declines, that same educated judgement is perfectly capable of shaping the independence movement in new ways, building indeed on a rhetoric of betrayal that could as readily be directed at the old SNP as at unionists. Whether that would be a desirable politics, and whether independence is an attractive future for Scotland, will always remain debateable. But it is a future debate that is probably already being formed.