Trends in national identity
We begin by looking at what has happened to the incidence of British national identity in Scotland since the advent of devolution. Scottish Social Attitudes has regularly presented its respondents with a list of all the national identities associated with one or more parts of Great Britain and Ireland and asked which best describes themselves. Respondents can select more than one if they wish, but are then asked to choose one single identity. A few surveys conducted before 1999 presented their respondents with much the same list but asked them to choose just one identity in the first place. By combining the two sources of information we can see how the most salient national identity of people living in Scotland has evolved since 1979. As by far the two most popular answers throughout have been "Scottish" and "British" and given too that these are the two responses that are of most interest to us, we only show in Figure 6.1 the proportion choosing one or other of those.
British identity clearly plays second fiddle to Scottish identity north of the border. Forced to choose a single identity, at most only around one in five say that they are British, while typically around three-quarters or so indicate that they are Scottish. There is, though, nothing new about this. Much the same pattern has been in evidence since the late 1990s; there is no evidence of a secular increase in adherence to a Scottish identity since the advent of devolution. However, throughout the last decade and a half, far fewer people have regarded themselves as British than did so in the 1970s. Evidently Britishness has been in decline north of the border over the longer term, but it would seem that devolution is better seen as a consequence of Scotland's distinct sense of identity rather than something that has led to its development.
However, forcing people in Scotland to choose just one identity runs the risk of underestimating the degree to which they are willing to acknowledge at least some sense of Britishness. When Scottish Social Attitudes has given them the chance to choose more than one identity, typically around 40 per cent have chosen both Scottish and British, with the most recent reading (for 2012) standing at 45 per cent. The possibility that people might feel both identities is explicitly acknowledged in a different approach to asking about national identity that has also been implemented on a regular basis on Scottish Social Attitudes. This is the so-called Moreno question, which was originally inspired by the existence of a pattern of dual identities in much of Spain (Moreno, 1988; 2006). Respondents are presented with a set of options that range from exclusively Scottish at one end to exclusively British at the other, while at the same time also offering various possible combinations of feeling both Scottish and British:
Which, if any, of the following best describes how you see yourself?
Scottish not British
More Scottish than British
Equally Scottish and British
More British than Scottish
British not Scottish
The responses to this question confirm the impression that Scottishness is the more strongly felt of the two identities. Typically no more than around one in ten place themselves in one of the last two rows of Table 6.1, thereby indicating that they are exclusively or predominantly British. In contrast, the proportion claiming to be exclusively or predominantly Scottish has often exceeded 60 per cent. However, there is no sign that this proportion has increased since devolution has been in place. Indeed, if anything, in recent years it has tended to be a little lower, averaging 58 per cent since 2007 as opposed to 66 per cent between 1999 and 2006. True, the proportion saying they were Scottish and not British was lower in 1997 than in all but one year thereafter, but this is the only evidence that can possibly be cited in support of any claim that there is a link between the introduction of devolution and a decline in Britishness. Overall, our evidence suggests that devolution has not served to undermine Britishness in Scotland, though it is clearly relatively weak compared with Scottishness.
While many people in Scotland might be willing to acknowledge more than one identity, in Northern Ireland identity is often regarded as singular and therefore a potential source of conflict. Those who say they are British are thought to deny any sense of being Irish, and vice versa. Indeed when in 2003 the Northern Ireland Life and Times survey gave respondents the opportunity to select more than one identity, only 11 per cent chose both, far fewer than we have seen claim in Scotland to be both Scottish and British. However, on the couple of occasions, in 2007 and 2012, when the Northern Ireland Life and Times survey administered the Moreno question, as many as between 47 per cent (2012) and 58 per cent (2007) claimed to be some mixture of British and Irish. Though still rather less than the three-fifths or so in Scotland who claim to be some mixture of Scottish and British (see Table 6.1). This suggests that in practice many people in Northern Ireland may in fact have at least some sense of both British and Irish identity.
Further evidence pointing in this direction emerges from Figure 6.2, which shows the responses to the question on national identity that has been asked on a regular basis on the Northern Ireland Life and Times survey. Respondents are asked to choose which one of four possible identities best describes themselves: British, Irish, Northern Irish or Ulster. It is not just British and Irish that prove to be relatively popular, Northern Irish proves to be too, and especially so since 2005. Since then, on average, 27 per cent have said they are Northern Irish, compared with 20 per cent in the period between 1989 and 2004. Meanwhile, further analysis reveals that over three-quarters of this group (78 per cent in 2012) say that they are some mixture of British and Irish when presented with the Moreno question. It appears then that Northern Irish is an identity that is particularly likely to be adopted by those who do not feel that British and Irish are necessarily mutually exclusive identities. The increase in its popularity might therefore be regarded as some evidence of a decline in the potential of national identity to act as a source of division and conflict - in line with the hopes of many supporters of the Good Friday/Belfast Agreement.
But while claiming a Northern Irish identity may signal some sense of being both British and Irish, the increase in the identity's popularity seems to have been wholly at the expense of the proportion who simply say they are British. Before 2005, on average, 45 per cent said that they were British. Since then only 38 per cent have done so. In contrast, the proportion saying they are Irish has not declined at all. Indeed it has in fact been slightly higher, on average, since 2005 (29 per cent) than it was beforehand (26 per cent). The rise in the proportion of people saying they are Northern Irish seems to represent an increased willingness to acknowledge having an Irish as well as a British heritage rather than an increased acceptance of having a British as well as Irish one.
Indeed, if we examine separately the trends in identity among those who claim adherence to a Protestant denomination and those who say they are Catholic (see Table 6.2), we find that the increase in adherence to a Northern Irish identity has occurred more or less exclusively among Protestants. As a result, whereas once a Northern Irish identity was more common among Catholics than Protestants, now, if anything, it is more likely to be claimed by the latter. Not only does claiming a Northern Irish identity seemingly enable people to avoid having to make the historically polarised choice between British and Irish, but also it appears to help bridge the divide between the region's two religious communities.
All of this might well be thought beneficial to reducing the potential for conflict in Northern Ireland, without necessarily undermining support for its continued membership of the UK - although intriguingly the shift in national identities started a little before the successful conclusion of the political agreement that finally lead to the resumption of devolution in 2007. Still, as Table 6.2 shows, the distributions of identities among Protestants and Catholics continue to look very different from each other - Britishness remains almost exclusively a preserve of the former while Irishness remains largely confined to the latter. In fact, the most recent reading for 2012 suggests that there has been something of a reversal of the increase in a Northern Irish identity (and especially so among Catholics) while both British and Irish identities have become more common within the religious communities with which they are traditionally associated. This may represent a reaction to a sequence of events that took place shortly before or during the period when the 2012 survey was conducted, including a row about the number of days the Union flag was to be flown over Belfast City Hall. This reversal is a reminder that despite the growth in acknowledgement of what appears to be a more cross-community form of identity, issues of identity can still cause conflict in the region.
If in Northern Ireland national identity often appears to be a source of conflict, in England it is perhaps most commonly seen as a source of confusion. National identity in the UK's largest nation has been described as 'fuzzy', and the terms 'English' and 'British' called a distinction without a difference (Cohen, 1995; Kumar, 2003). But perhaps the emergence of devolved institutions in the rest of the UK has helped make people in England more aware of the fact that the two are not synonymous and, more importantly, more inclined to say that they are English rather than British?
In Figure 6.3 we show how people in England have responded when presented with a list of identities and asked to initially choose more than one if they wish, but then asked to pick just one. In this case we can add one further reading, for 1992, from a survey that only invited people to choose one identity in the first place. Once again we focus on the two most common identities, British and English.
It would appear that English has proved relatively more popular, and British less so, since and including 1999. Before that, the proportion choosing English ranged between 31 per cent and 37 per cent; it has been above that range on all but two subsequent occasions. Conversely, the proportion saying they are British has usually been lower than it was at any time before 1999. To that extent the advent of devolution elsewhere in the UK appears to have coincided with some increase in the relative popularity of Englishness as opposed to Britishness. However, we should also note that the reduction in the proportion selecting a British identity is only evident when respondents are required to choose a single identity; the proportion choosing British as one of their identities has not discernibly changed at all, standing usually at a little under 70 per cent. Moreover there is no evidence of any further change in the relative popularity of the two identities since 1999; the figures for 2012 (43 per cent English, 43 per cent British) are almost exactly the same as those for 1999 (44 per cent English, 44 per cent British). In short, any effect that devolution has had on national identity in England has only been in the form of a one-off step change, rather than a continuous secular change.
Unfortunately, evidence on the pattern of responses to the Moreno question is only available for one year prior to 1999 (see Table 6.3). However, this proves to be consistent with the suggestion that a one-off step change occurred in 1999. In 1997 just seven per cent said that they were English not British; since then the figure has never been less than 17 per cent. However, once again there is no sign since 1999 of any secular trend towards more people saying they are predominantly or wholly English. At 29 per cent,
the most recent reading for the proportion saying either that they are English not British or that they are more English than British, is much the same as the equivalent figure for 1999, 31 per cent.
Still, we should also note that, contrary to what we might have expected from the evidence of Figure 6.3, the proportion saying they are exclusively or primarily English is typically greater than the proportion indicating they are exclusively or primarily British. One possible explanation is that some of those who say they are British in response to the question on which Figure 6.3 is based are stating a fact about their legal citizenship rather than indicating their identity. Perhaps it is only when asked a question that explicitly asks them to weigh the two that their Englishness emerges. That said, we should bear in mind that by far the most common answer to the Moreno question is "Equally English and British". In truth, identity still seems to be much 'fuzzier' in England than it is in the rest of the UK.
One point is clear: devolution has certainly not proved to be the harbinger of any strengthening of Britishness. Rather that identity seems to have weakened somewhat in both England and Northern Ireland, while in Scotland it has remained as weak as it has ever been. However in Northern Ireland, the decline has been accompanied by an increase in people's willingness to say they are Northern Irish, a change that seems motivated more by a degree of willingness to acknowledge having a British and an Irish heritage rather than any rejection of Britishness. (It is also a change that potentially still seems to be capable of being reversed when inter-communal disagreements and disputes break out.) And although in England, the relative importance of being English and being British tipped a little in favour of the former in 1999, since then the two identities have lived side by side with each other in much the same way as before. Still, arguably the acid test of whether devolution has undermined the popular basis of the Union is what has happened to constitutional preferences since 1999. It is to that topic that we now turn.
- Download chapter
- Entitled the Northern Ireland Executive.
- Wales is not included because surveys using a methodology similar to that deployed by those analysed here have not been conducted there since 2007. For information on the surveys that have been conducted since then and the trends in respect of national identity and constitutional preference they suggest have occurred see Curtice (2013), and Wyn Jones and Scully (2012).
- There is one small difference between the Northern Ireland Life and Times survey and the two social attitudes surveys so far as their reporting conventions are concerned. In the case of British Social Attitudes and Scottish Social Attitudes the standard practice is to include in the denominators on which percentages are based those who refused to answer a question or are otherwise recorded as not having answered a question. In the case of the Northern Ireland Life and Times survey, however, they are usually excluded. To avoid the risk of us quoting in this chapter figures for any of these surveys that might be slightly different from those given elsewhere we have followed as appropriate each survey's usual practice.
- Though we should note that the fieldwork for the 2012 survey, when there was a particularly marked drop to 53 per cent, was undertaken during and in the weeks immediately following the 2012 London Olympics and that we cannot reject the possibility that that event may have helped to foster a short-term increase in feelings of British identity.
- It should also be noted that as long ago as 1996, a System Three poll reported that 39 per cent said that they were Scottish not British, many more than did so in our 1992 and 1997 surveys. However this System Three poll was conducted using a rather different methodological approach and thus we cannot be sure that the difference represents a methodological artefact rather than evidence that the incidence of an exclusive Scottish identity had at some point been just as high before the advent of the Scottish Parliament as it has proven to be subsequently (Moreno, 2006).
- On the events in 2012 that might help account for the lower level of Northern Irish identity in that year see Note 8.
- Bases for Table 6.2 are as follows:
- Belfast City Council voted in December 2012 to fly the Union flag only on a limited number of special days rather than, as hitherto, every day. The decision occasioned some rioting. There was also some serious rioting in Belfast and elsewhere the previous summer in the wake of that season's Orange Order parades (Nolan, 2013: 161). The year 2012 also saw the centennial commemoration of the events leading up to the signing of the Ulster Covenant in opposition to Home Rule.
- Wyn Jones et al. (2012) claimed on the basis of an internet 'Future of England' survey conducted by YouGov in July/August 2011 that there had been a marked increase in English as opposed to British identity. They found that 40 per cent said they were exclusively or predominantly English (compared with 33 per cent on the 2009 British Social Attitudes survey). However, a subsequent YouGov poll conducted for Channel 4 News in January 2012 found only 33 per cent saying they were exclusively or predominantly English while a second Future of England survey in November 2011 put the figure at 35 per cent (Wyn Jones et al., 2013). Even leaving aside the many methodological differences between the two exercises, it would appear that Wyn Jones et al.'s relatively high 2011 figure could well have been the result of no more than sampling variation.
- It should be noted that in 2012 the English version of the Moreno question was not administered to a small number of respondents who had been born in Scotland or Wales (N=76). However, given their place of birth they might reasonably be expected to be more likely to claim to be exclusively or predominantly British rather than exclusively or predominantly English. Their exclusion thus should not have diminished the proportion saying they are predominantly or wholly English.
On the whole, do you think that England's economy benefits more from having Scotland in the UK, or that Scotland's economy benefits more from being part of the UK, or is it about equal?
England benefits more
Scotland benefits more
Would you say that compared with other parts of the United Kingdom, Scotland gets pretty much its fair share of government spending, more than its fair share, or less than its fair share of government spending?
Please choose your answer from this card.
Much more than its fair share of government spending
A little more than its fair share of government spending
Pretty much its fair share of government spending
A little less than its fair share of government spending
Much less than its fair share of government spending
- Bases for Table 6.4 are as follows:
- In 1991 talks with the parties (other than Sinn Fein) were instigated by the then Northern Ireland Secretary of State, Peter Brooke. Sinn Féin were excluded from these talks, but at the same time Brooke authorised secret contact be made with the Irish Republican Army. Meanwhile the UK government appeared to recognise nationalist sentiment by declaring that the Britain had "no selfish strategic or economic interest in Northern Ireland", a declaration that helped pave the way for talks between John Hume of the Social Democratic and Labour Party and Gerry Adams of Sinn Fein that in turn were eventually able to help pave the way towards the Good Friday/Belfast Agreement (Nolan, 2012: 20).
- The bases for Table 6.6 are as follows:
- We should also note that there has since 2007 been a marked decline among Catholics in the level of support for reunification. In that year no less than 47 per cent backed that view, but now only 32 per cent do so. A similar decline, from 21 per cent to eight per cent, has occurred among those of no religion. One possible explanation is that the prospect has come to look less attractive, for the time being at least, as a result of the particularly adverse consequences that the banking crisis of 2008 visited upon the Irish Republic. Recent demands from Sinn Fein that another poll be held on Northern Ireland's constitutional status may also have encouraged people to consider the possible practical consequences of reunification.
- We do not, however, find intensification on the scale claimed by Wyn Jones et al. (2012; 2013) who reported the results of two Future of England internet surveys that asked, in 2011 and 2012, the same question about the voting rights of Scottish MPs as that asked by British Social Attitudes and then compared these surveys' findings with those obtained by British Social Attitudes up to and including 2007. The Future of England surveys reported that in 2011 no less than 53 per cent strongly agreed that Scottish MPs should not vote on English laws, and in 2012 that as many as 55 per cent did so, far higher figures than obtained by British Social Attitudes either before or since 2007. The figure obtained by the 2011 survey might have been thought to have been occasioned by the fact it failed to offer respondents a "neither agree nor disagree" option, but this option was included on the 2012 survey and an even higher figure obtained. However, in both years the remaining response options on the Future of England survey read, "Strongly agree" and "Tend to agree" together with "Strongly disagree" and "Tend to disagree", whereas the options on British Social Attitudes are, "Agree strongly" and "Agree" together with "Disagree strongly" and "Disagree". We would suggest that respondents are more likely to say that they "Agree strongly" when the alternative is to say "Tend to agree" rather than "Agree", and that consequently there must be severe doubt about the merits of drawing substantive conclusions from any comparison of the findings of the two series. We would also note that those who evince a relatively high level of interest in politics are more likely to agree strongly with the proposition that Scottish MPs should not vote on English laws, and that while just 36 per cent of the 2012 British Social Attitudes sample in England said that they had "a great deal" or "quite a lot" of interest in politics, as many as 67 per cent of the 2012 Future of England sample said they were "very" or "fairly interested". In short part of the reasons for the difference between the two sets of findings may well be that the internet survey contained more people with a high level of interest in politics.
- The Future of England surveys (Wyn Jones et al., 2012; 2013) also record relatively high levels of don't know responses on this issue, 31 per cent in 2011 and 26 per cent in 2012. However, they also report rather higher proportions saying that Scotland gets more than its fair share, 45 per cent in 2011 and 52 per cent in 2012, suggesting that attitudes have become yet more critical since 2007. However, we should note that the question is administered somewhat differently on the Future of England survey than on British Social Attitudes; respondents are simply invited to say whether Scotland gets pretty much its fair share, more than its fair share or less than its fair share whereas on British Social Attitudes respondents are presented with a showcard that lists all five options shown in Table 6.9. In addition, on the Future of England surveys the question is asked immediately after a question that asks respondents whether England gets its fair share of public spending and this may have helped cue some respondents into saying in the following question that Scotland gets less than its fair share.
- In 2004-2006 the second option read "that makes decisions about the region's economy, planning and housing". The 2003 survey carried both versions of this option and demonstrated that the difference of wording did not make a material difference to the pattern of response. The figures quoted for 2003 are those for the two versions combined.
- Bases for Table 6.11 are as follows:
- Bases for Table 6.12 are as follows:
- Bases for Table 6.13 are as follows:
- Bases for Table 6.14 are as follows:
- If we undertake a loglinear analysis of the data in Table 6.15, we find that the data are fitted adequately at the 5% level of probability without fitting a term for the interaction between national identity, constitutional preference and year. The residual chi-square for such a model is 13.0, which, with eight degrees of freedom, has a p value of 0.11.
- Note that in their attempt to argue the contrary position neither Wyn Jones et al. (2012) nor Wyn Jones et al. (2013) demonstrates that the link between Moreno national identity and attitudes towards devolution is stronger now than previously. They merely demonstrate that there is a link between national identity and such attitudes now, a point that is not in dispute. What has to be demonstrated for their argument to be sustained is that the link has grown stronger.
- Bases for Table 6.15 are as follows:
- Related links