Brexit's Scottish Dimension:

Dual Identities and Qualified Support for the EU

Marco Biagi

Abstract

Scotland’s ongoing support for the UK’s continued membership of the European Union since
2016 has been observed more than it has been explained. What caused the component nation
of the UK least supportive of Europe in the 1975 referendum to become the most supportive
41 years later? Using panel data from the British Election Study I argue for the primacy of two
factors: national identity and partisanship. There is little evidence of strong European identity
prior to the referendum but Scottish national identity also appears to have been more open to
dualism - i.e. feeling simultaneously both Scottish and European. The party system in Scotland
has however diverged from that of the United Kingdom and partisan cues over the course of
the referendum campaign transmitted Remain messages to many whose political
predispositions in England would have implied a high likelihood of voting Leave. Views on
Scottish independence are - perhaps surprisingly - secondary. Differences of opinion between
Scotland and England on immigration and the functioning of the EU are insufficient by
themselves to explain the greater Remain vote.
Introduction

That Scotland is the most pro-European Union of the United Kingdom’s four component nations
now verges on axiomatic. Although widely debated in the context of Scotland constitutional politics
and a potential second independence referendum, the question of why and how this came to be is
much less frequently asked. Scotland’s relative Europhilia has been observed more often than it has
been explained. Without such an explanation it is impossible to say whether the circumstances are
particular to the Scottish case or potentially replicable elsewhere – an important question given the
growing contemporary salience of right-populism across the consolidated democracies. Moreover, the
first time the United Kingdom’s membership of the then-European Community was put to a vote, in
1975, Scotland was in fact more Eurosceptic than England and Wales. While 58% of those voting
chose to retain membership, this compared to 67% across the UK and 69% in England. In four
decades therefore, relative levels of support for European integration within Great Britain have
reversed. Scotland fixed effects in wider regressions are useful shortcuts, but a meaningful explanation
of the Scottish Remain vote requires a closer examination of what took place north of the border in
June 2016. Data collected in the successive waves of the British Election Study show in particular a
different relationship between Scottish national identity and Europeanism than that of English, and
the difference in Scotland’s party system.

The European Union and domestic voting

Thus far, academic work on Brexit has settled on immigration and sovereignty as the proximate causes
of the Leave vote, while debating whether racial resentment/cultural backlash or economic anxiety
serves best as their prior cause (Clarke, Goodwin & Whiteley 2016; Inglehart & Norris 2016). Both
have in common the treatment of the question of European Union membership largely as the result
of a concern about something else, with opposition to the EU serving as a way to express opposition
to the status quo.

Scholarship is clear that domestic factors rather than the truly European regularly drive voting
behaviour. Elections to the European Parliament have long been observed to be ‘second-order’
elections in the eyes of the electorate, with lower turnouts and voting dominated by domestic rather
than European factors (Reif & Schmitt 1980; Schmitt 2005). Where the same electoral phenomena
occur simultaneously across Europe, they do so because of similar circumstances or institutional
incentives rather than a genuinely pan-European public discourse or the passing of retrospective
judgement on the performance of the Europe-wide administration of the institutions that one might
expect from a typical model of electoral accountability. One example is the consistently better
performance of small parties and those from certain ideological families in European Parliament
elections (Hix, Noury & Roland 2007; Hix & Marsh 2007, 2011). The lack of salience of European
integration with electorates has even been observed to extend to single-issue referendums held solely
on the topic, with decisions in a series of treaty ratification referendums seemingly strongly determined
by levels of approval of domestic governments (Franklin, Marsh & McLaren 1994). For Brexit this
would imply an important role for support for the UK Government as an explanatory factor. There
are major dissimilarities between these observed relationships and the 2016 referendum however.
Referendums on incremental, technocratic treaties, generally initiated by governments seeking
ratification of a change are potentially very different to one on the fundamental question of
membership of the Union and initiated by a government seeking a ratification of the status quo but
appearing to be divided on the issue.

An influential work by Hooghe & Marks (2005) provides a useful alternative framework, arguing that
attitudes to European integration can be explained by ‘calculations, community and cues.’ Broadly
these reflect and apply key schools of thought in the political science of voting behaviour – the classic
rational self-interest paradigm (Meltzer & Richard 1981); personal identification-based explanations
(Berelson, Lazarsfeld & McPhee 1954, Campbell et al 1960); and political and media elite leadership
heuristic effects (Lupia 1994; Lupia & McCubbins 1998; Zaller 1992).

All three were observable in the competing cases put forward in the British EU membership
referendum. Both sides openly argued for material benefits – economic benefits from trade on the
Remain side, the redirection of the UK’s contribution to the EU budget to domestic health spending
on the Leave side (embodied the infamous ‘£350m a week for the NHS’ bus). The Leave campaign
argument that an end to EU membership was needed to control immigration might be considered in
this category too as it was evidently an offer to satisfy an interest in many members of the public,
though whether it is a strictly material one leads into a wider question of what motivated those
interests.
Community appears to have played a less overt role, though Leave campaign appeals to British
national sovereignty against ‘European’ influence on domestic law would have been incoherent
without a foundation of British institutions commanding greater legitimacy with the public than the
EU’s. Ex-Prime Minister Gordon Brown made a late intervention (and published a book) putting
forward what he saw as the ‘patriotic’ case for ‘leading not leaving’ the European Union (Brown 2016).
This is remarkable, when any treatment of European identity shows that Britain is at or near the
bottom of any cross-national table (eg Risse 2015). At a more local level identitarian politics are
increasingly being attributed with significant contemporary effects on political behaviour through their
influence on the formation of understandings of the world. The observations of the American rural
right in Wisconsin and Louisiana (Walsh 2004; Hochschild 2016) have their British corollary in work
that has observed divergence since 1997 on key cosmopolitan-parochial issues between core and ‘left-
behind’ parts of England (Jennings & Stoker 2016).

If ‘community’ was promising ground for Leave, ‘cues’ were much less so. The Remain side was
supported by the leaders of all of the parties in the UK parliament except the United Kingdom
Independence Party and Northern Ireland’s Democratic Unionist Party. The campaign also
consciously deployed massive co-ordinated endorsements by cultural and business figures, the latter
including a letter signed by chief executives of 40 of the largest companies on the London Stock
Exchange. The Trades Union Congress also backed Remain, with their General Secretary a participant
in one of the eleven nationally televised debates. One caveat is needed to this however – the role of
the press and broadcasters as intermediaries. TV news, bound by impartiality regulations, had to give
equal time to the views put forward by either side; Britain’s print media, with their stridently partisan
presentation of news, was substantially more pro-Leave than pro-Remain if weighted by circulation.

The United Kingdom is not however the unitary state and monolithic polity it once was. Each of these
would have been filtered differently in Scotland. The dataset collected by the British Election Study,
a multi-wave longitudinal panel collected online from 2014-17 shows clear evidence originating
Scotland’s Remain support under both the ‘community’ and ‘cues’ headings.

The Case: Scotland

Within the UK Scotland is broadly considered a nation but is rather definitely not a state. The
contemporary British constitution is quasi-federal, but the devolution arrangements currently in place
do differ from most practices of federalism by granting through the Sewel Convention de facto
legislative authority to the Scottish Parliament in the fields devolved to its jurisdiction. Whereas for
example in the US, minimum wage law or education policy may be set by city, state or federal
government, in Britain the United Kingdom and Scottish Parliaments have separate legislative spheres.
Edinburgh legislates for and administers education, healthcare, criminal law, transportation and
environment, as well as setting income and property taxes, meaning the Scottish Government’s
responsibilities are a substantial part of the role of a typical European state. 57% of public spending
in Scotland is by the devolved institutions, and this comprises 27% of Scottish GDP (Scottish
Government 2017a). The UK Parliament retains de jure power to legislate in these areas contrary to
the will of the Scottish Parliament, but the exercise of such power would be a political and
constitutional nuclear option.

Over time, respondents to the regular Scottish Social Attitudes Survey have come to rate the Scottish
Parliament as having as much impact on their lives as the UK Parliament and wishing it had more
(Scottish Government 2017b), though they do continue to turn out for elections to the UK Parliament
at a higher rate. Scotland is therefore in a sui generis proto-statal constitutional arrangement that affords
considerable policy autonomy. This in turns offers the potential for a public sphere for political debate
that is at the same time distinctive and connected to that of the wider United Kingdom.

Between 1964 and 2010 Scotland gave a plurality of the vote to the centre-left Labour Party in every
election to the UK Parliament. The pro-independence Scottish National Party (SNP) became a
significant force from the 1970s, and elements within other parties also began organising for greater
autonomy within the United Kingdom. Following a referendum held by Tony Blair’s Labour
Government, the Scottish Parliament, elected by proportional representation, was established in 1999.
First governed by a Labour-led coalition, the SNP have formed the Scottish Government on a single-
party basis since 2007. An attempt by the SNP government to secure full independence through a
referendum in 2014 was unsuccessful, gaining 45% support and therefore falling short of a majority.
Over this protracted period Scotland’s party system has been noticeably more centre-left than that of
the wider UK, also featuring an additional political dimension of greater or lesser Scottish autonomy
vis-a-vis the UK as well as traditional economic and social left-right debate.

While social class has classically been seen as being the prime determinant of party support in British
politics (Butler & Stokes 1969), this has given way in recent years to arguments about the end of class
and the rise of valence models based on perceived ‘competence’ to deliver mutually agreed goods such
as economic growth (Green & Jennings 2012; Whiteley et al 2013; per Stokes 1963). That conventional
wisdom may be a product of the political convergence of the New Labour era and may not long
survive its recent demise at the hands of Jeremy Corbyn. So too in Scotland the additional party offers
a qualification to traditional class-based loyalty in favour of a nationality-based conception, though
again valence models have carried much of the weight of explanation of the SNP’s success (Johns et
al 2009; Johns, Mitchell & Carman 2013) – at least prior to the independence referendum of 2014.

Community: a European Scotland, a Remainer Scotland?

Scotland is a more European polity than England, but it is not a more European society. The regular
face-to-face Scottish Social Attitudes Survey has consistently shown less than 10% of respondents
choosing ‘European’ from a list of identities that they consider applies to them, a share that has been
consistently lower than the already unenthusiastic respondents in the rest of the United Kingdom
(Ormston 2015). The last British Election Study to ask the question was in 1997, where again only
9.5% of respondents in Scotland agreed, slightly lower than the already low 14.3% across Britain.
Qualitative research by the Institute for Public Policy Research conducted by Ipsos MORI on the eve
of the Brexit referendum concluded that the debate “lacked passion and conviction,” and that
participants did not see themselves as particularly ‘European’ (Diffley et al 2016). The current
generation of the British Election Study only began asking about European identity in Wave 7,
collected in April 2016, at a time when the referendum already loomed and would have added a
political dimension to responses. This is critical, when levels of European identity – ‘community’ in
the Hooghe & Marks framework or ‘affective’ links as distinguished from ‘instrumental’ (Easton 1975)
– are, unsurprisingly, strongly correlated with higher propensity to vote Remain.

If there is a lack of widespread or deep ‘Europeanness’ in Scottish individuals, could it instead be

something about ‘Scottishness’ that drives a different qualitative relationship with the European Union
and supranationality? A regular refrain of media commentary on the liberal left in Scotland is a self-
conception of the country and its identity as outward-looking. Scottish national identity does not itself
however appear to correlate with respondents being more political pro-European. In fact - with one
caveat - the opposite is the case.
As a multinational state, citizens of the United Kingdom have multiple national (and supranational
identities to choose from): that of their component country, British and, indeed, European. The graphs
above poss a challenge for the straightforward assumption that Scottish identity itself may have
directly contributed to a Remain vote. Signs for British, English and Scottish identity all show a
correlation between higher levels of identification with that nationality and with greater likelihood to
vote Leave, albeit to different levels of magnitude. High levels of Scottishness correlate less with Leave
voting than Englishness or Britishness, but the coefficient sign remains significantly positive. To be
sure, the relationship is not monotonic, but two thirds of respondents in the two highest levels of
identification, so it is that final slope between ‘6’ and ‘7’ that carries the greatest weight in the overall
model.

However, while higher levels of a respondent’s self-described Scottishness itself was, like higher levels
of Englishness, correlated with a Leave vote, that Scottishness was and is slightly more likely to exist
alongside a European identity.
Bivariate correlations between component country identity, Britishness and Europeanness are
strikingly consistent across all the waves of the BES. Respondents with higher English identity have,
on-balance, lower European identity. By contrast, intensity of Scottish identification is almost
irrelevant to how European those respondents feel. Again, an image of Scottishness as being actively
supportive of fostering Europeanness appears overblown from the data, but while Scottishness did
not in and of itself make people more likely to support EU membership, having a stronger sense of
Scottishness was not antithetical to having a stronger sense of European identity in the way that
intense feelings of Englishness were. Scottishness and Europeanness could exist alongside each other
in respondents more comfortably than Englishness and Europeanness. Where they did however, it
was the European identity that seems to have done the work of pushing Remain. Scottishness alone
was insufficient – indeed, on its own it is associated with Leave.

Which makes a national identity more ‘pro-European’ – creating support for the European Union
independently, or accommodating a parallel European identity? An influential development in
arguments around European integration is that of Risse, who argues that early integrationist dreams
of a single supranational European identity have given way to a widespread comfort with ‘dual’
identity. Instead of losing member state national identities, people would begin to feel ‘European and
German’, ‘European and Slovak’ and so on, and cites Eurobarometer datasets to support this
conceptualisation and its emergence (2015). By this definition, Scottishness is a more ‘European’
identity than Englishness by its compatibility as a dual identity, though with the caveat this case is easy
to overstate. In these BES results Scottishness does not appear to actively increase Europeanness, it
merely fails to inhibit it.

Of course, Scottishness and Englishness are not the only relevant national identities. Respondents in
both nations could also identify as British: yet Scottishness and Britishness are as negatively correlated
as Europeanness and Englishness. Suggestions that the UK may be home to ‘two dualisms’ – Scottish
and European in Scotland, English and British in England – should however be resisted. Scholarship
has shown the political significance of the relative strength of English and British identity in England,
and though the typical respondent will have modestly high levels of both, which is the greater has
significant consequences, not least with regard to attitudes to EU membership (Henderson et al 2017).

Faced with these results, advocates of Remain might be dismissive of whether any lessons can be
learned: Scottishness is simply odd, the product of a different set of historical circumstances than
Englishness, and impossible to replicate south of the border. Yet it is notable that, in the wider
perspective profiled by Risse, the relative compatibility between Scottishness and Europeanness is
actually the normality. The question then becomes why Englishness has come to deviate so much
from the other national identities around the continent.

Cues: partisanship

At UK general elections until the 1960s, the Scottish vote broadly corresponded to the Britain-wide
vote. Measured using a Gallagher index modified to show the difference between Scottish and British
voting 1, there are two stages at which the system diverged sharply: the entry of the Scottish National
Party to the political system at the start of the 1970s, and then the SNP’s surge in 2015. This is
unsurprising, since the SNP has had no direct corollary in the English party system.

1
A Gallagher index measures disproportionality between party vote shares and seat shares; in this case the
calculation is used to show disproportionality between party vote shares in Scotland and party vote shares across
Great Britain.
Partisanship can have an independent effect, a key theme in US research, ranging from inciting
motivated reasoning on key policy issues to the reifying effect on subsequent voting behaviour from
just publicly registering as a member of a political party (Bartels 2000; Green, Palmquist & Schickler
2004; Gerber, Huber & Washington 2010). In the post-referendum BES wave, 35% of English voters
identified with the two parties who furnished the Leave campaign’s high-profile leadership figures of
Johnson, Gove and Farage; only 18% of Scottish voters did. Similarly, the SNP cuing its supporters
to vote Remain and not being present in the English political system would have meant more exposure
to sympathetic pro-Remain cues in the Scottish electorate.

The presence of the SNP does provide a clear divergence from the rest of the United Kingdom in
terms of party system, but its relationship to the EU issue has been complicated. In 1975, the SNP
campaigned for a No vote and officially opposed EC membership into the 1980s, though with
prominent figures increasingly coming to support it over this period (Lynch 2002). The support for
‘independence in Europe’ that began in 1990 has been credited with strategic rather than ideological
origins (Keating 1996). One interpretation of European sub-state nationalist or regionalist parties’
general position is that they see the EU as an (unwitting) ally against the central state, which is alone
seen with acrimony as an illegitimate and centralised government (Jolly 2007), and this can be
expressed in terms of competing identities as well as competing administrative power centres.
Alternatively, the EU might offer a ‘safe space’ into which the sub-state unit can become independent
or at least more autonomous – an appeal that has declined since the ‘Europe of the Regions’ failed to
transpire (Elias 2008). Both of these show a chiefly pragmatic motivation. By recent times the SNP
had become one of the most volubly pro-Europe parties, but a survey of grassroots members in 2007
found that support for the EU was still somewhat qualified (Mitchell, Bennie & Johns 2012). Indeed,
during the 2016 referendum, while SNP leader Nicola Sturgeon participated in televised debates on
the pro-Remain side, the party itself later reported spending of just £90,830 in the referendum to the
Electoral Commission, raising allegations of ambivalence from opponents in the media (Gordon
2016). The SNP’s implications for the effect of the party system on Scotland’s greater propensity to
vote Remain is therefore not obvious.

Indeed, at first glance the evidence seems unhelpful that partisanship may have had an independent
effect. Those identifying as SNP supporters were indeed clearly Remain by a considerable majority,
but Labour, Conservative, Lib Dem and Green supporters in Scotland were all less likely to vote Leave
than their counterparts in England as well. Scottish Conservatives, Lib Dems and Greens also appear
to have been less likely to vote Leave than their English counterparts. To dismiss the effect of party
at this point is to overlook a key distinction - the sort of people in Scotland who identify with, for
example, the Labour Party, are not the same sort of people who identify with that party in England.

In the British Election Study, respondents were given two sets of statements, one corresponding to
an authoritarian-liberal dimension and the other an economic left-right, and asked to give their views,
from strongly agree to strongly disagree.

The table below uses an index constructed from these responses to classify individuals by
predisposition on these dimensions. Individuals were given scores of -2, -1, 0, 1 or 2, from strongly
agreeing with the position associated with the left to strongly agreeing with that associated with the
right. Then, due to the clustering of scores on the economic left and authoritarian right, the British
sample as a whole was divided based on relative position. Septiles allowed for broadly equal numbers
of respondents in each bin. Putting supporters of Labour and the SNP together might seem unusual,
given the differences between the parties, but they are both parties with social democratic
programmes, ethos and who have broadcasted a pro-European message over at least the past two
decades.
In England there is a clear pattern: voters who are relatively to the left economically are more likely to
identify with Labour, and those who are relatively to the right identify with the Conservatives.
Scotland’s pattern is similar but different in an important way: the SNP and Labour lead in six of the
seven septiles. Only the most extremely relatively economically conservative in Scotland are actually
Conservative.

In both countries the differences between parties on the authoritarian-liberal left-right scale are rather
less pronounced. Economic interest evidently remains Britain’s master cleavage. The contrast between
Scotland and England are again striking. Though voters most predisposed to social conservatism in
Scotland were more likely to identify with the Conservatives than either of the two centre-left parties
individually, collectively these party’s identifiers outnumbered Conservative even at the extreme end
of the spectrum.
Scotland’s party system has therefore seen centre-left parties able to win the affinity of those whose
predispositions in England would have made them far more likely to identify with a centre-right party.
This is more than people of certain demographics being more likely to support centre-left parties in
Scotland – this is evidence that people Scotland are nonetheless more likely to support centre-left
parties than people in England of the same worldview. A curious result of this ability by the centre-
left to compete more successfully with the centre-right for the affinity of voters in the centre is that,
by this measure, the typical supporter of the Scottish Conservatives is to the right of the typical
supporter of the English Conservatives – and the typical supporter of the SNP or Scottish Labour to
the right of the typical supporter of English Labour. If we assume that parties are able to provide cues,
it is clear that the pro-Remain cues were being received by those much further to the right on the
authoritarian-liberal scale in Scotland than England. Indeed, given the presence of Ruth Davidson, the
most popular leader of the Scottish Conservatives in the modern era, in UK-wide televised debates
for the Remain side, one could plausibly argue that the party cue transmitted to Conservatives in
Scotland was qualitatively different to the cue transmitted to their ostensible co-partisans in England.

The proportion of party identifiers or voters supporting Remain have remained broadly consistent
across the BES waves. Membership of the European Union has been a live issue in British politics
since long before this iteration of the BES however, and support or opposition may already have long
since been ‘baked-in’.

The driver of Scotland’s distinct party system is the emergence of the constitutional question. With an
extra dimension of political competition, the two economically centre-left parties can make appeals
on the basis of their staunch distinction on this political cleavage. Labour and the SNP are able to
connect with individuals of a relatively right-wing predisposition on grounds of opposition and
support for independence respectively. Again, it would be easy to dismiss this as unique to Scotland,
with no implications for competition in other polities, given that the question of Scottish
independence almost by definition cannot become a dimension of political competition in other
nations. That would be to take an excessively specific perspective. It is not intrinsically the question
of independence that affords economically centre-left parties this greater opportunity to compete, it
is the presence of an additional dimension of competition, something replicable in other polities with
multiple lines of political distinction or capable of developing them.
Scotland and England on key summary metrics, June 2016
Scotland England

Demographics

Median gross household income bracket £30,000 - £34,999 £30,000 - £34,999

White ethnicity 92% 88%
Owner-occupier 57% 59%
University degree 28% 26%
Median age of eligible voter 50 49

Issues % %
EU democracy Satisfied 32 24
Unsatisfied 68 76
Immigration preference Increase
Reduce
UK government Approve 20 27
Disapprove 59 46

National identification
European 5-7 (high) 42 36
1-3 (low) 39 45
‘British’ 5-7 (high) 56 80
1-3 (low) 31 9
‘Scottish’ / ‘English’ 5-7 (high) 79 78
1-3 (low) 13 12

Party identification
Labour 22 31
Conservative 16 30
Lib Dem 4 9
UKIP 2 5
Green 2 2
SNP 29 -
British Election Study 2014-17, Wave 9. Scotland n=3,663, England n=19,159.
Alternative explanations

The simplest alternative explanation for Scotland’s lower Leave vote would have been fewer people
in the categories that voted Leave. Immigration appears to be less of a feature in Scottish political
debate, and, indeed, both the SNP and Scottish Labour have been more willing to express active
support than the leadership of the English centre-left in recent years. Indeed, even the Scottish
Conservatives have joined with the other two main parties in support of the reinstatement of the post-
study work visa removed by their own co-partisan UK Government.

So is Scotland different? Could it be that there are simply different demographics in Leave-correlated
factors like age profile or education; or it could be fewer people had trigger attitudes like concern
about immigration? As the table above shows however, respondents from the two countries vary only
very slightly on demographic grounds – admittedly, some of this will be artificial due to the weighting
scheme, but that too is anchored in census research. The number of dimensions on which the Scottish
and English samples are relatively homogenous is also encouraging for the earlier findings, comparing
the two samples.

In addition to demographics, the prior scholarship already discussed has highlighted certain issue
positions as relevant to the decision to vote Leave or Remain, and while there are gaps evident on
these, they are only differences of degree. Respondents in Scotland and England were both unsatisfied
with the EU, both supportive of lower immigration and both disapproving of the UK government.
European identity is also marginally stronger in Scotland, and while here especially the effect of the
campaign is likely to be felt, if this is taken as a measure of susceptibility to priming of European
identity rather than long-run presence of European identity, it is still a useful variable. The really
substantial differences are on feelings of ‘Britishness’, and party identity.

To what extent could the other differences have driven the much lower Leave vote? The hypothesis
that the vote resulted from the difference in incidence of these underlying opinions can be tested by
building logit models from English respondents and creating predicted probabilities for all
respondents based on the characteristics in the model. The predicted probabilities for Scottish
respondents then allows the creation of an ‘implied Leave vote’ metric that will suggest their accuracy
in predicting the Scottish respondents, as an out-of-sample external validity test.
England logit models and predicted values for Scotland voters, 2016
Correctly Pseudo R2, Correctly Implied
classified, England classified, Leave vote,
England Scotland Scotland

% % % %

European attitudes 91.8 0.69 84.8 39.7

Europeanness, approval of EU, support for
integration, satisfaction with EU democracy
Immigration 84.3 0.46 75.6 47.6
Should immigration be lowered, will UK have
greater control of immigration if withdraws
National identification 80.2 0.36 78.9 46.3
Britishness, Europeanness, substate nationality
Economics 71.8 0.17 68.7 48.0
Can foresee being short of money or unemployed
personally, economy improving
Political attitudes 71.1 0.18 64.6 50.0
Left-right economic and social scales derived from
agreement/disagreement with series of statements.
Demographics 64.9 0.08 60.7 45.0
Age, income, gender, degree, owner-occupier,
ethnicity, born elsewhere in EU.
Newspaper readership 64.2 0.11 69.0 51.5
Title of newspaper mainly read, or non-reader
Government approval 59.1 0.03 59.1 51.8
Approval of UK Government, approval of UK
Prime Minister.

British Election Study 2014-17, Wave 9. Scotland n=3,663, England n=19,159.

The table above contains a series of parsimonious models2, which show that those that have better fit
for England, at the top of the table, also broadly predict Leave shares in Scotland somewhat closer to
the actual observed rate. The first model is predictive but not especially meaningful – one would hope
attitudes to the EU would be correlated well with voting behaviour on a referendum on EU
membership. Pure demographics on the other hand provide only a very basic fit with the data
provided, as also found by Clarke, Goodwin & Whiteley. Indeed, a demographic logit model using
Scottish respondents alone with those variables fails to generate a model with better fit in terms of
share correctly classified than the simple null of assuming all are Remain. Similarly, a model based on
UK government approval seems of little explanatory power.

The immigration result is striking, especially when it may seem more tangential to the European
project than questions of trade, or the operation of the institutions themselves. Though free
movement of people is a fundamental of the Single Market, one does wonder whether it is immigration
from elsewhere in the EU that chiefly motivates this sentiment. There is a possibility of a hidden
variable here. It may be that a propensity for closed-minded attitudes cause both opposition to
immigration and to membership of a supranational institution like the European Union. Yet such an
authoritarian-libertarian propensity is explicitly tested for in the political attitudes model, and does not
show the same strength of correlation.

The three-variable economics model is a modest fit, but it produces results prior to expectations.
Those who feel they are more at risk of unemployment were more likely to vote Remain, and those
who felt the economy was improving were more likely to vote Leave. Only the respondent envisaging
that there will be times when they would be short of money correlated with Leave in the way an
economic anxiety explanation would predict.

Crucially, together these models are suggestive that there is more at work than simply Scotland having
a different distribution of demographics or opinions. The latter divergence of opinion is present, to
be sure, and invites further research to explain why. But had Scottish voters chosen Leave or Remain
on the basis of their views to the same extent as English voters on, for example, immigration, then
these findings suggest that the Leave vote in Scotland would have been almost 10% higher. In short,

2 Following on from the statistical concerns surrounding large ‘garbage can’ additive regressions (Achen 2005) and the ever-increasing
problem of missing data in records as more and more variables are brought into the regression, I find these to be methodologically
preferable in this context.
these voters do not just hold slightly – and only slightly – different views, they also appear to be acting
on them differently. Immigration shows this most clearly, that in both Scotland and England higher
support for immigration was correlated with higher probability of voting Leave, but that at almost all
points those who have the same views on immigration were more likely to vote Leave in England than
Scotland.

Immigration to Scotland has been lower than to England, but the link between direct contact with
immigration and holding anti-immigration views remains tenuous on a cross-national basis (Norris
2005).

Scotland may also have greater economic exposure to Europe. Parts of Scotland have been major
recipients of European structural funds. The House of Commons Library, compiling various data
sources, however estimated 47% of Scotland’s exports went to the EU in 2016 and 39% of imports,
compared with UK-EU exports of 49% and imports 55% so by this one measure if anything Scotland
is less economically integrated than the rest of the UK (Ward 2017). This may not however have been
the perception. Since affective links with the European project seem tenuous, a greater perceived
benefit from continued membership would be a plausible alternative. But here the ‘calculation’ and
‘cues’ sections of the Hooghe & Marks framework blur into each other. A respondent who receives
Common Agricultural Policy payments and one who works in a sector unrelated to the European
Union may both well believe that their personal material wellbeing would be affected by the
referendum outcome, but by different mechanisms. The former would see the direct impact of leaving
the body that administers their funding, the latter might consider that the economy as whole would
be impacted and they would, along with the general population, suffer a decline in living standards.
The Remain campaign expended a great amount of time and effort on promoting the latter perception.
Both respondents might be calculating, but for the latter that calculation depends very heavily on prior
cues – accumulated trust in the established politicians and media figures who are spokespeople for the
public campaigns. This renders the effects of the two not quite observationally equivalent but at the
very least hard to disentangle.

Conclusion

Despite being demographically rather similar by what we classically associate with political behaviour,
the typical Scottish voter and the typical English voter reacted differently. They formed different views
on European Union membership and acted differently even where they held the same underlying
predispositions and even the same views on key triggers like reducing immigration. Partisanship and
the particular character of national identity emerge as key suspects to explain this. Scotland’s party
system clearly causes large numbers of people who can be measured as having right-leaning
predispositions nonetheless to support and identify with centre-left parties. The third, constitutional
dimension to contemporary politics may be part of the answer, but it is impossible to say whether
another cross-cutting cleavage in another system could have the same effect. Parties choosing not to
follow their electorates shows that they may be policy-seeking rather than vote-seeking (Strøm 1990)
but raises normative questions about representation in Scotland. Ironically, this sort of potentially
unrepresentative consensus may help trigger populist movements. Scottish national identity too
appears relatively compatible with a dual sense of European identity, but higher identification as
Scottishness without that dual identity correlates with greater likelihood to vote Leave. Scottishness
alone does not make a Europhile.
Bibliography

Achen, Christopher H. "Let's put garbage-can regressions and garbage-can probits where they
belong." Conflict Management and Peace Science 22, no. 4 (2005): 327-339.
Bartels, Larry M. “Partisanship and Voting Behavor, 1952-1996.” American Journal of Political
Science 44, no. 1 (2000): 35-50.
Berelson, Bernard R.; Paul F. Lazarsfeld; William N. McPhee. Voting: A Study of Opinion
Formation in a Presidential Election. Chicago: University of Chicago Press, 1954.
Brown, Gordon. Leading Not Leaving: A Patriotic Case for Britain Remaining in Europe. 2016.
Butler, David, and Donald Stokes. Political Change in Britain. New York: St. Martin's Press.
1969.
Campbell, Angus; Philip Converse; Warren Miller; Donald Stokes. The American Voter. Ann
Arbor: University of Michigan Press, 1960.
Clarke, Harold D.; Matthew Goodwin; Paul Whiteley. 2017. Brexit: Why Britain Voted to Leave
the European Union. Cambridge: Cambridge University Press.
Diffley, Mark; Sanah Zubairi; Colin Hockaday; Russell Gunson. The ‘Tin-man’ referendum: the
EU debate in Scotland is lacking heart. Edinburgh: Ipsos MORI / Institute for Public Policy
Research. Accessed online 4th August 2018:
https://www.ipsos.com/sites/default/files/migrations/en-uk/files/Assets/Docs/Polls/16-
026767-01%20EU%20Focus%20Groups%20-
%20Report%20Version%201%20Final%20to%20STV%20080616.pdf.
Downs, Anthony. ‘An Economic Theory of Political Action in a Democracy.’ Journal of
Political Economy 65(2), 135-150.
Easton, David. (1975). “A re-assessment of the concept of political support.” British Journal of
Political Science, 5(4), 435-457.
Electoral Commission. EU referendum results. Accessed online at
https://www.electoralcommission.org.uk/find-information-by-subject/elections-and-
referendums/past-elections-and-referendums/eu-referendum/electorate-and-count-
information.
Elias, Anwen. “From Euro-enthusiasm to Euro-scepticism? A Re-evaluation of Minority
Nationalist Party Attitudes Towards European Integration” Regional & Federal Studies 18,
no. 5 (2008): 557-581.
Franklin, Matt; Michael Marsh and L. McLaren. “Uncorking the Bottle: Popular Opposition to
European unification in the wake of Maastricht’ Journal of Common Market Studies 32, no. 4
(1994): 101-117
Gerber, Alan S., Gregory A. Huber, and Ebonya Washington. "Party affiliation, partisanship, and
political beliefs: A field experiment." American Political Science Review 104, no. 4 (2010):
720-744.
Goodwin, Matthew J. and Oliver Heath. 2016. ‘The 2016 Referendum, Brexit and the Left
Behind: An Aggregate-level Analysis of the Result’ Political Quarterly, 87(3).
Gordon, Tom. ‘SNP spent less on EU vote than it did fighting a by-election in Glenrothes.’
Herald, 29 November 2016.
Green, Donald P., Bradley Palmquist, and Eric Schickler. Partisan hearts and minds: Political
parties and the social identities of voters. New Haven: Yale University Press, 2004.
Green, Jane and Will Jennings “The dynamic of issue competence and vote for parties in and out
of power: An analysis of valence in Britain, 1979-1997”. European Journal of Political
Research 51 (2012): 469-503.
Hanretty, Chris. 2017 ‘Areal interpolation and the UK’s referendum on EU membership’ Journal
of Elections, Public Opinion and Parties, 27(1), 1-17.
Henderson, Ailsa; Charlie Jeffery; Dan Wincott; Richard Wyn Jones. ‘How Brexit was made in
England’ British Journal of Politics and International Relations 19(4), 631-646.
Hix, Simon and Michael Marsh. 2007. ‘Punishment or Protest? Understanding European
Parliament Elections.’ Journal of Politics 69(2), 495-510.
Hix, Simon and Michael Marsh. 2011. ‘Second-order effects plus pan-European political swings:
An analysis of European Parliament elections over time’ Electoral Studies 30, 4-15.
Hix, Simon, Abdul G. Noury, and Gérard Roland. Democratic Politics in the European
Parliament. New York: Cambridge University Press, 2007.
Hochschild, Arlie Russell. Strangers in Their Own Land: Anger and Mourning on the American
Right. New York: The New Press, 2016.
Hooghe, Liesbet and Gary Marks. ‘Calculation, Community and Cues: Public Opinion on
European Integration’ European Union Politics 6 (2004): 419-43.
Inglehart, Ronald F. and Pippa Norris. “Trump, Brexit, and the rise of Populism. Economic have-
nots and cultural backlash.” Presented at annual meeting of the American Political Science
Association, Philadelphia, 2 September 2016.
Jennings, Will and Gerry Stoker. “The Bifurcation of Politics: Two Englands” The Political
Quarterly 87 (2016): 372-382.
Johns, Robert; James Mitchell; David Denver; Charles Pattie. “Valence Politics in Scotland:
Towards an Explanation of the 2007 Election.” Political Studies 57 (2009): 207-233.
Johns, Robert; James Mitchell; Christopher J. Carman. “Constitution or Competence? The SNP’s
Re-election in 2011.” Political Studies 61 no 1. (2013): 158-178.
Jolly, Seth Kincaid. “The Europhile Fringe? Regionalist Party Support for European
Integration.” European Union Politics 8, no. 1 (2007): 109-130.
Keating, Michael. Nations Against the State: The New Politics of Nationalism in Quebec,
Catalonia and Scotland. New York: St Martin’s Press, 1996.
Lupia, Arthur. "Shortcuts versus encyclopedias: Information and voting behavior in California
insurance reform elections." American Political Science Review 88, no. 1 (1994): 63-76.
Lupia, Arthur, and Mathew D. McCubbins. The Democratic Dilemma: Can Citizens Learn What
They Need to Know?. New York: Cambridge University Press, 1998.
Lynch, Peter. SNP: The History of the Scottish National Party. Cardiff: Welsh Academic Press,
2002.
Meltzer, Allan H. and Scott F. Richard. ‘A Rational Theory of the Size of Government.’ Journal
of Political Economy 89, no 5 (1981): 914-927.
Mitchell, James, Lynn Bennie, and Rob Johns. The Scottish National Party: Transition to Power.
Oxford: Oxford University Press, 2012.
Norris, Pippa. Radical Right: Voters and Parties in the Electoral Market. New York: Cambridge
University Press. 2005.
Prosser, Chris; Jon Mellon, and Jane Green. What mattered to you most when deciding how to
vote in the EU referendum? British Election Study, 2016. Accessed online at
http://www.britishelectionstudy.com/bes-findings/what-mattered-most-to-you-when-deciding-
how-to-vote-in-the-eu-referendum/.
Reif, Karlheinz and Hermann Schmitt. 1980. ‘Nine second-order national elections: a conceptual
framework for the analysis of European election results’. European Journal of Political
Research 8(1), 3–45.
Risse, Thomas. A Community of Europeans?: Transnational Identities and Public Spheres.
Ithaca: Cornell University Press, 2015.
Schmitt, Herman. 2005. ‘The European Parliament Elections of June 2004: Still Second Order?’
West European Politics 28(3), 650-679.
Scottish Government. Government Expenditure and Revenue Scotland 2016. Edinburgh. 2017.
Scottish Government. Scottish Social Attitudes Survey 2016. Edinburgh. 2017.
Stokes, Donald. “Spatial models of party competition” American Political Science Review 57
no.2 (1963): 368-377.
Strøm, Kaare. "A behavioral theory of competitive political parties." American journal of
political science (1990): 565-598.
Taggart, Paul. ‘A touchstone of dissent: Euroscepticism in contemporary Western European
party systems’ European Journal of Political Research 33 (1998): 363-388.
Walsh, Katherine Cramer. Talking About Politics: Informal Groups and Social Identity in
American Life. Chicago: University of Chicago Press, 2004.
Ward, Matthew. Statistics on UK-EU trade. London: House of Commons Library, accessed
online at researchbriefings.files.parliament.uk/documents/CBP-7851/CBP-7851.pdf.
Whiteley, Paul; Harold D. Clarke; David Sanders; Marianne C. Stewart. Affluence, Austerity and
Electoral Change in Britain. New York: Cambridge University Press.
Zaller, John. The Nature and Origins of Mass Opinion. New York: Cambridge University Press,
1992.