Scottish Affairs
Scotland's Longest Running Journal
on Contemporary Political and Social Issues

Home >> Articles Online, by Author >> Articles Online, by date published online >> D.McCrone and L.Paterson, Scottish Affairs, No. 40, Summer 2002

Scottish Affairs, No. 40, Summer 2002

The Conundrum of Scottish Independence

by David McCrone and Lindsay Paterson

(Note: the analysis discussed in this article is up-dated and extended in Attitudes to Scottish Independence and to the SNP by Lindsay Paterson, online 27 August 2003)

Introduction

Scotland, as a 'stateless nation', is one of the few western European countries with a full-blown independence movement which is a major player in party politics. The Scottish National Party (SNP), as the main political challenger to the Labour party, is committed to full constitutional independence within the European Union, and can be contrasted with, for example, regionalist and nationalist parties in Wales and Catalonia which are committed to greater autonomy some way short of full independence. Thus, secession from the British state is more likely to come from Scotland, and yet Scottish nationalism appears, in comparative terms, to have some unusual features. First, it is often argued that it is a paradigmatic case of 'civic' rather than 'ethnic' nationalism insofar as conventional cultural markers such as language and religion do not mark it out as particularly distinctive from England. Benedict Anderson, for example, has argued that the Union of parliaments of the two countries in 1707 came about precisely because of the lack of strong cultural markers, at least between lowland Scotland and England (Anderson, 1996). Second, there is no simple relationship between preferring independence as a constitutional option, voting for the SNP, and defining oneself as Scottish. Thus, a minority of those who say they are in favour of Independence vote SNP; and barely half say they are Scottish not British (Paterson et al., 2001).

One must be careful, of course, in implying that Scotland is an outlier in these matters, and these features may well be shared with other so-called 'stateless nations' to a greater or lesser extent. What makes the Scottish case particularly interesting and relevant to this book is the recent establishing of a legislature in the form of a devolved parliament, set up in 1999. Whereas the constitutional debate prior to this event was structured mainly around whether Scotland should have a parliament or not, since then it has refocused around whether that parliament should be devolved or independent.

This chapter explores the possibility that what might seem like an 'extraordinary' choice, namely independence, may actually result from a series of incremental steps along a continuum of greater self-determination, framed by developments elsewhere in the UK and the European Union. 'Independence' is a term which carries its own historical and political baggage, and assumes that 'sovereignty' is a zero-sum game. Ordinary people may not share the common assumption that sovereignty is something which the polity either has completely or doesn't have at all. Sovereignty may be partial, shared, segmental. In terms of constitutional politics, in Scotland at least, such issues may not be separate from 'ordinary' politics but integral to them.

We will examine the degree to which support for independence has shifted over the last few years, its social base, and how it relates to political identity and to party identification. While much of our exploration depends on cross-sectional data, we are able to draw upon panel data collected as part of the British election panel study (BEPS) so as to explore the extent to which there is a core vote for independence or not. This has important theoretical implications because it is one of the assumed characteristics of nationalism that it grows outwards from a settled core. In his study of nationalism in Europe, for example, Miroslav Hroch argued that national movements had three phases: first, the search by intellectuals for national identity based on linguistic, cultural and historical attributes; second, the awakening of national consciousness by political activists; and third, the development of mass-scale nationalist movements (Hroch, 1985). If Hroch is correct, we ought to be able to place Scotland somewhere between stages 2 and 3, and to find that those who supported independence did so consistently, providing a core of political support which the nationalist party is able to exploit. In other words, we would expect to find support for independence to be solid rather than promiscuous over time.

Understanding Political Change

The data come from the Scottish Election Surveys of 1979, 1992 and 1997, the Scottish Referendum Survey of 1997, the Scottish Social Attitudes Surveys of 1999 and 2000, and the British Election Panel Study which followed up the respondents in the 1997 election study in spring 1998, spring 1999 and spring 2000. The cross-sectional surveys were multi-stage cluster samples, stratified at the cluster level, and drawn from the electoral register until 1992 and from the postcode address file from 1997. Data were collected by face-to-face interviews in respondents' homes, computer-aided from 1997. Full details of the sampling design etc are reported in the appendices of Paterson et al (2001) and Curtice et al (2001). The response rates were at least 60% and usually between 65% and 70%. The Scottish Social Attitudes Survey achieved samples of 1482 in 1999 and 1663 in 2000. The Scottish Election Survey achieved samples of 729 in 1979, 957 in 1992 and 882 in 1997. The Scottish Referendum Survey of 1997 achieved a sample of 676. Among the 882 respondents in Scotland to the 1997 election survey, the British Election Panel Study achieved response rates of 76% in 1998 (672 respondents), 71% in 1999 (626 respondents) and 66% in 2000 (586 respondents). Although the panel study thus had quite high levels of attrition, it does have the unique virtue of letting us study the ways in which individuals do or do not change their minds in their political attitudes. Moreover, the attrition did not vary by relevant attitudes (as measured in 1997): it was much the same among supporters of various constitutional options for Scotland, of various political parties, and in various demographic groupings defined by social class, gender and age: see Paterson et al (2001, p. 174) for further details. The surveys have been funded mainly by the UK Economic and Social Research Council and its predecessors, and have been run mostly by the National Centre for Social Research (formerly SCPR). The survey data can be obtained from the Data Archive at Essex University.

Support for Independence

Table 1 gives the context for patterns of attitudes to independence, showing the support for various constitutional options among people in Scotland between 1979 and 2000. The modal position has always been some kind of domestic parliament, the position that was eventually endorsed by the referendum of 1997. In that referendum, nearly all independence supporters voted for a devolved parliament: according to the Scottish Referendum Survey, 96% of people who supported either of the independence options voted in favour of a parliament. After the parliament was actually established in 1999, the new institution became the most popular single option, attracting around 50% support. After the referendum, the opposition to a parliament faded to around one in ten people, but throughout this period it had never been higher than one in four. Independence started in 1979 as an option with very low support: only 7%. By 1992, after a decade of Conservative government, that had grown to around one quarter. At the time of the 1997 referendum, it reached 37%, after which it fell back to around 30%. Moreover, a feeling that independence was likely to come about became the common view. At the time of the referendum in 1997, 59% thought that independence was 'very likely' or 'quite likely' in the ensuing two decades (made up of 76% of those who supported that option, and 48% of those who did not). In 1999, 51% took this view (75% of those who supported it and 43% of those who did not). For most of the last decade, independence within the EU has been about twice as popular as independence outwith the EU, although the temporary rise in support for independence at the time of the referendum was entirely accounted for by greater popularity for independence in the EU. Given the steady fall in support for no elected parliament, and the persisting low level of support since 1997 for a weaker form of parliament than is currently in place, the strengthening of support for independence almost certainly has come as a result of people shifting away from supporting a strong domestic parliament, being replaced in supporting that option by people who previously would have not supported any parliament. For the period since 1997, we look further at this question of the extent to which people change their views when we consider the panel study data later.

Table 1 Support for various constitutional options, 1979-2000   1979 1992 1997 (election) 1997 (referendum) 1999 2000 independence outwith EU* 7 6 8 9 10 11 independence in EU* 17 18 28 18 19 strong domestic parliament** 26 50 42 32 50 47 weak domestic parliament** 28 9 9 9 8 no elected body 26 24 17 17 10 12 sample size 729 957 882 676 1482 1665 * No distinction was made between the two types of independence in 1979. ** Strong domestic parliament was referred to in 1979 as 'Scottish Assembly which would handle most Scottish affairs', and from 1997 onwards as 'Scottish Parliament within the UK with some taxation powers'. Weak domestic parliament was referred to in 1979 as 'Scottish Assembly which would handle some Scottish affairs and would be responsible to Parliament at Westminster', and from 1997 onwards as 'Scottish Parliament within the UK with no taxation powers'. No distinction was made in 1992. Don't know and not answered included in the base. Sources: Scottish Election Surveys of 1979, 1992 and 1997; Scottish Referendum Survey of 1997; Scottish Social Attitudes Surveys of 1999 and 2000.

Attitudes towards independence also changed in a positive direction, as Table 2 shows (the question was asked in only 1979 and 1999). In 1979, 56% of people thought that independence would be 'bad' or 'very bad'; this had fallen to 45% in 1999. The proportion regarding independence as 'good' or 'very good' rose more sharply, from 29% to 45%, because the proportion taking no view on the matter fell from 17% to 11%. But the main reason for the growth in support for independence is not this: it is that far more of those who regard independence as attractive now also support it. In 1979, only 16% of those who saw it as 'good' also supported it; by 1999 this had risen to 45%. Similarly, the proportion supporting independence among those who viewed it as 'very good' rose from 40% to 77%. Nevertheless, in both years, the proportion regarding independence as 'good' or 'very good' was much higher than the proportion willing to support it: 29% against 7% in 1979, and 45% against 28% in 1999. Looking at this another way, despite the rise in overall support for independence, and despite the shift in a positive direction in people's views of independence, a consistent 28% of the majority who did not support independence regarded that option as 'good' or 'very good'.

 

Table 2 Attitude towards independence, and support for independence, 1979 and 1999   1979       1999 view of independence column % support for independence (% in cell) column % support for independence (% in cell) very bad 21 0 14 3 bad 33 0 31 6 good 21 16 33 45 very good 8 40 12 77 sample size 729 1482 The question in both years was 'If Scotland did become independent some time in the future, do you think that this would be a very bad thing [etc]'. Don't know and not answered included in the base. Sources: Scottish Election Survey of 1979 and Scottish Social Attitudes Survey of 1999.

There is a fair measure of consistency in the social basis of support for independence, as can be seen from Table 3. Support rose in all social classes, but, setting aside fluctuations that are probably the result of sampling variation, support has always been higher in working class groups than in middle class ones [1]. This was starkest at the peak of independence support at the time of the referendum in 1997, when around one half of working class people supported it. Nevertheless, in all years since 1997, around one quarter of middle class groups have supported independence. These social class differences were much more pronounced than in the support for any kind of elected parliament (Brown et al, 1998, p. 160): in 2000, for example, the proportions opposing any kind of parliament were 11% in the salariat, and 10% in the working class.

The gender difference in support is clear and stable: in all years, men have been more likely to support independence than women, usually by around five percentage points. This gender difference is not found in support for any kind of elected parliament, because women are more likely to support a domestic parliament than men (Brown et al, 1998, p. 161): in 2000, 12% of men and 11% of women opposed any kind of parliament.

In the last decade, younger people have been much more likely to support independence than older people. There is some evidence, however, that this is, at least in part, a cohort effect. This is clearest if we look at the 45-54 age group in 1992, who then showed 27% support. As they aged, to become roughly the 55-64 group in 2000, they maintained that level of support, and certainly did not fall to the 12% which the 55-64 group had in 1992. Something the same could be said of the 18-24 group in 1992, which had a 44% level of support by 2000. Nevertheless, if we go back to 1979, we can see that the main change has been related to period: between then and 1992, all age groups showed a rise of at least twice (the oldest group), and mostly by much more than that.

The conclusion from Table 3 is that, by 2000, there were substantial minorities supporting independence in all class, gender and age groups, but with the strongest support among working class people, men and young people, and with distinctly low levels of support only among old people.

 

Table 4 Support for independence by self-assigned class, 1979-1999   1979 1992 1997 (election) 1997 (referendum) 1999   % N % N % N % N % N middle class 3 78 17 214 17 221 27 156 23 389 working class 9 268 24 703 30 626 40 457 28 1026 Don't know and not answered included in the base (N). The question in the 1999 survey was: 'Most people say they belong either to the middle class or the working class. If you had to make a choice, would you call yourself middle class or working class?'. Questions in earlier years were similar. It was not asked in the 2000 survey. Sources: Scottish Election Surveys of 1979, 1992 and 1997; Scottish Referendum Survey of 1997; Scottish Social Attitudes Survey of 1999.

Table 4 shows that levels of independence support are less clearly related to people's sense of their class identity than it is to their objectively measured class. This is quite the opposite of what is typically found for levels of support for political parties, where opposition to the Conservatives, in particular, is more strongly related to subjective class than to objective class (Bennie et al, 1997, pp. 103-4).

There is some association of independence support with left-wing allegiance, as Table 5 shows. To define left, centre and right, we start with individual questions that relate to typical ideological issues separating left and right (Paterson, forthcoming). In the surveys of 1992, 1997 and 1999, these were six questions about inequalities of wealth and power, about the role of trade unions and of private enterprise, about government ownership of public services, and about government responsibility for creating full employment. The replies were then added up to give a scale of views: people who consistently took a left-wing position were at one end of the scale, and so on. In the survey of 2000, a different set of five questions was asked, but they were used in the same way: these asked for views about the extent of inequalities in wealth and power, about government's responsibility for redistributing income, about whether big business exploits workers, and about whether management takes advantage of workers. In each year, the left was then defined in each survey as the most left-wing third of people on this scale, the centre as the middle third, and the right as the most right-wing third. (Because of the way in which values were grouped on the scales, it was not possible to divide the samples into exactly equal thirds, as can be seen from the sample sizes in Table 5.) We can see from Table 5 that, on the left, independence has had around one third support or higher since 1992, in the centre it has had about a quarter of support, and on the right it is generally supported by no more than about one in five people.

Levels of support for independence are closely linked to nationalist ideology. Table 6 shows support according to whether people identify more strongly with a Scottish person of the same class, or an English person of the opposite class. Between 1979 and 1999, national identification grew at the expense of class identification (Paterson et al, 2001, p. 108): in 1979, 44% identified first with class, and 38% with nation, whereas by 1999 this had become 24% with class and 43% with nation. The national identifiers do show generally higher support for independence than the class identifiers, but the differences are in fact not as large as might be expected, except in 1979 when independence support was very low anyway. In one respect, there is a rather clearer relationship between independence support and a perception that there is serious conflict between Scotland and England (also in Table 6). That, too, grew (Paterson et al, 2001, p. 116): in 1979, 15% perceived 'very serious' or 'fairly serious' conflict. This became 30% in 1992, 43% in 1999 and 38% in 2000. The table shows that people who perceive very serious conflict are more likely to support independence than those who see less severe conflict. Nevertheless, support for independence is not strongly related to the other categories. Because the 'very serious' group never makes up more than about one in ten of the population, for most people attitudes to independence do not seem to be strongly shaped by their view as to the extent of any conflict between Scotland and England.

 

Table 6 Support for independence by national identification, and by perception of conflict between Scotland and England, 1979-1999   1979 1992 1997 (election) 1997 (referendum) 1999 2000   % N % N % N % N % N % N identify most with*:                       same class English 5 211 26 259 27 208 38 168 26 318 not asked opposite class Scot 11 185 24 429 32 395 39 262 30 610 not asked no preference 5 85 22 115 20 111 31 89 22 218 not asked perceived conflict between Scotland and England**:             very serious 18 28 35 69 not asked not asked 40 180 43 164 fairly serious 8 72 30 222 not asked not asked 29 468 34 462 not very serious 8 320 20 477 not asked not asked 24 727 27 868 none 6 234 18 178 not asked not asked 25 91 24 155 Don't know and not answered included in the base (N). * The question in the 1999 survey followed the question on self-assigned class noted in Table 4: ' Would you say that you had more in common with (same class) English people or with (opposite class) Scottish people?' Questions in earlier years were similar. ** The question in 2000 was 'Thinking about the Scots and the English, how serious would you say conflict between them is? Very serious [etc]'. The question in other years was very similar. Sources: Scottish Election Surveys of 1979, 1992 and 1997; Scottish Referendum Survey of 1997; Scottish Social Attitudes Surveys of 1999 and 2000.

Something the same can be said about the relationship with feelings of being Scottish or being British (Table 7). Identifying mainly as British does seem to have been a definite barrier to supporting independence throughout most of the last two decades. But Scottishness has been so pervasive since the 1990s - with 80% preferring this identity in 2000, for example - that independence support among such people is just a little above the overall percentage as shown in Table 1.

It might be thought that independence support could be readily explained by political mobilisation. Certainly supporters of the Scottish National Party are much more likely to support independence than supporters of other parties (Table 8). Since 1997, in fact, the SNP has consistently had around two thirds or more of its supporters also supporting the party's core policy. Yet that leaves one third not doing so. What is more, in the past decade, at least one half of independence supporters have not supported the SNP, as Table 9 shows.

When independence reached its recent peak of support, at the time of the referendum in 1997, a clear majority of these supporters would have voted for the Labour party, not the SNP. Back in 1979, things were markedly different. The SNP attracted two thirds of the small band of independence supporters (Table 9), but depended on opponents of independence for a much larger share of its vote than it has done recently (Table 8).

These results from the cross-sectional surveys suggests, then, that there might be a core of independence supporters who are left-wing, inclined to a nationalist ideology, young, working class and male. In any given year, these views and social characteristics certainly do characterise the people who are most likely to support independence. But how consistent are they in their support? Are they really a stable core around which an independence majority could be built? To study this empirically, we have to look at data from the panel study.

 

Table 10 Support for independence each year among those who support independence in other years, 1997-2000   Proportion supporting independence at wave:   among people who supported independence at wave: 1997 1998 1999 2000 sample size 1997 100 52 44 47 227 1998 61 100 59 59 148 1999 59 63 100 59 131 2000 53 54 54 100 136 Don't know and not answered included in the base. Source: Scottish respondents in British Election Panel Study, waves in spring 1997, spring 1998, spring 1999 and spring 2000.

If there were a core group of independence supporters, ideologically committed to that option, then we would expect them to be highly stable in their support. Table 10 shows the contrary to be the case. Each row corresponds to those people who supported independence in a particular wave of the survey. The columns show what they believed in each of the other three waves. The immediately striking thing is that, among those who support independence at any particular wave, only just over half support it at any other wave. It is not even a matter of initial support slowly draining away: continuity from one year to the next is not consistently greater than continuity between waves that are two or three years apart.

The same kind of conclusions can be reached about people moving towards independence from other options, as Table 11 shows.

 

Table 11 Support for independence each year among those who did not support independence in other years, 1997-2000   Proportion supporting independence at wave:   among people who did not support independence at wave: 1997 1998 1999 2000 sample size 1997 0 12 11 15 655 1998 16 0 10 13 524 1999 19 11 0 13 486 2000 18 11 11 0 446 Don't know and not answered included in the base. Source: Scottish respondents in British Election Panel Study, waves in spring 1997, spring 1998, spring 1999 and spring 2000.

Broadly speaking, around one in eight of such people in any given year would support independence in any other given year. Another way of looking at the same data as in Tables 10 and 11 is in Table 12, which shows the number of waves at which people supported independence, and the same for a domestic parliament and for no parliament. For example, in the first column, we see that 55% never supported independence, 20% supported it on one occasion, and so on. On the one hand, this means that, at some time during the four years, nearly half - 45% - of people did support independence. On the other, only one in fourteen - 7% - consistently supported it. The 45% can be contrasted with the 89% who supported a domestic parliament on at least one occasion, and with the 74% who on no occasion opposed some kind of parliament. If we take the core supporters of each option as those who supported it on at least three of the four occasions on which they were asked, independence commands 14% core support, no parliament at all 6%, but a domestic parliament 60%. We can deduce from the 3% in the final column who consistently opposed a parliament that 97% were not determinedly averse to one. By contrast, 55% were never inclined to support independence.

We can reach similar conclusions if we look further into the apparent social segmentation of support that we summarised cross-sectionally in Table 3. On the basis of that table, we might be inclined to say that the working class, men and young people made up the core of support. So far as class is concerned, that is not so. When we draw up a table analogous to Table 10, but separately for middle class and working class groups, we get a very similar pattern to Table 10. Middle class is defined to be the first three class categories in Table 3, working class to be the last three. Among the working class, usually only about one half of the supporters of independence on any given occasion also supported it another given occasion. As a result, the versions of Table 12 for different classes were also similar. The proportion of working class people who had supported independence on at least one occasion was 52%, not much larger than the 45% in the sample as a whole. For the middle class, the proportion was 40%. The proportion of people who consistently supported independence on all four occasions was very small in both groups - 7% in the working class and 9% in the middle class.

A similar pattern can be found with respect to age. To achieve adequate sub-sample sizes, age is grouped into 18-34 and 35 or older (following what appears to be a fairly clear break in levels of support for independence in the election survey in 1997: Table 3). But young people did not form any more of a stable core of support than the working class. People aged 18-34 were only slightly more likely to experiment with supporting independence than those aged 35 or over: 53% supported independence at least once, as against 42%.

The pattern with respect to gender is somewhat different, as Table 13 shows. We might expect from Table 3 that men would form a core group for independence. Men were indeed more stable in their allegiance to independence than women: for nearly all pairs of years, around 60% or more of male supporters of independence consistently supported it, in contrast to usually around 45-50% of women. Yet the consequence of that greater experimentation by women is that the cross-sectional differences shown in Table 3 exaggerate the gender effect. Nearly the same proportion of men and women had experimented with supporting independence at least once (as in the first column of Table 12): 46% of men and 44% of women. There was a larger core of men doing so, 10% maintaining support on all four occasions, against 5% of women, and 19% maintaining it on at least three occasions, against 11%. But this does not suggest that women are any more averse to independence than men: more women experiment with independence than men, but also more women experiment with other options too.

 

Table 13 Support for independence each year among those who support independence in other years, by gender, 1997-2000   Proportion supporting independence at wave:   among people who supported independence at wave: 1997 1998 1999 2000 sample size 1997 men 100 60 50 57 110   women 100 45 39 36 117 1998 men 60 100 60 65 81   women 64 100 58 50 67 1999 men 66 70 100 68 62   women 52 55 100 49 69 2000 men 63 62 58 100 68   wome