Euroscepticism and religion before Brexit

British membership of a Community which . . . counts among its aims the reconciliation of European enmities, the responsible stewardship of European resources and the enrichment of Europe’s contribution to the rest of mankind, is to be welcomed as an opportunity for Christians.

Church of England General Synod on Europe in 1972 cited in Mudrov (2015: 515)

As with its departure 47 years after it joined, the UK's entry to the EC was legitimised through a national referendum in 1975 - the first nationwide referendum in the UK history While many of the arguments deployed in that campaign were similar to those in 2016, the results were very different: 67 per cent supported EC membership. While not a ringing endorsement of European integration, it reflected the belief of most voters that closerintegration with Europe was compatible with the UK’s economic, cultural and political interests (Gifford and Wellings 2018). As European integration continued apace, however, and EU citizens gained the right to move freely throughout Member States, as British voters became less loyal towards and more critical of their political parties, and as those parties increasingly took for granted or ignored socially conservative Eurosceptics, British hostility to European integration became more widespread.

This shift towards an increasingly Eurosceptic outlook is clear among Britain’s religious conununities, hr terms of religious communities’ support for EC/EU membership, the 1975 and 2016 referendums are almost mirror images. As shown in Chapter 4, in 2016, the staunchest supporters of EU membership were the religiously unaffiliated, while the most passionate advocates of Brexit were Anglicans, with free church Protestants not far behind. A clear majority of Catholics and Presbyterians, meanwhile, supported Remain. In 1975, however, the religiously unaffiliated were the most opposed to EC membership, while Britain’s Christian conununities (particularly Protestants) were more supportive. As the quote given earlier illustrates, the Church of England (which did not take an official stance in the 2016 referendum) expressed clear support for the principle and objectives of European integration just prior to the UK joining the EC. While the entire British electorate was to become more Emosceptic after 1975, the Anglican community stands out for a particularly pronounced change from being among the strongest supporters of European integr ation to among its fiercest critics - despite the stance of the Church of England’s leadership barely changing.

In this chapter, we use the BES series of surveys to examine the effects of religion on Euroscepticism, first through its relationship with support for EC membership in the 1975 referendum, and then through its effect on attitudinal support for that membership over the following 40 years. We show how Britain's Christian communities evolved from endorsing EC membership in 1975 to adopting increasingly varied views by the 2000s, in which some (such as Catholics and Presbyterians) were more supportive of European integration while others (such as Anglicans) were among its fiercest critics, as the consequences of EU membership increasingly interacted with religious voters’ distinct political and social ideologies and national identities. We also show that religious behaviour and belief have, in contrast with that of belonging, remained stable sources of sympathy for EC/EU membership, regardless of denominational membership. While their effects were consistently small, the perceptions that EU membership and mass migration were placing increasing strain on the homogeneity of British society and culture, as well as its political sovereignty and identity, were to some extent mitigated by the communal bonding effects and elite cues arising from greater engagement with religious institutions and values.

Religion and the 1975 referendum

Norton (2011: 53) called the issue of EC membership ‘one of the most politically contentious issues in British politics in the last half of the 20th century’: it split both the Conservative and Labour parties and cut though traditional left-/right-wing ideological allegiances. Vocal campaigners for and against EC membership could be found in both parties: Conservative Emosceptics (such as Enoch Powell) did not wish to undermine the identity of Britain as a distinct island nation and world power, while Labour Eurosceptics (such as Tony Benn) feared the potential limitations EC membership could impose on a socialist government (Norton 2011). Eventually, the decision to join the EC was enshrined in the 1972 European Communities Act, but this did not prevent the issue from splitting both major parties for decades to come.

Labour Party leader Harold Wilson illustrated the difficult balancing act the party was trying to maintain, leading opposition to the specific terms of EC membership defined in the 1972 Act while remaining supportive of membership in principle, drawing support and criticism from all sides of the House of Commons and even from within his Shadow Cabinet. In an (ultimately successful) effort to hold his party together, Wilson charted the course that David Cameron was to follow 40 years later and committed to renegotiate the terms of EC membership and put the outcome to a referendum if Labour won the 1974 general election - a policy that was to persuade the former Conservative minister Enoch Powell to vote for the party (Norton 2011; Gifford and Wellings 2018). While it took two general elections that year to deliver a Labour majority, the commitment was eventually delivered and the referendum held - although (just as David Cameron was in 2016) Wilson was forced to suspend Cabinet collective responsibility and allow ministers to campaign freely to keep his government intact (Geddes 1999: 3).

The themes of the 1975 referendum campaign closely resembled those of the 2016 Brexit referendum. The key argument of prominent Conservative ‘No’ campaigners was that EC membership would undermine the UK’s independence and status on the world stage, that it would be detrimental to parliamentary sovereignty and British democracy, and it would undermine the British national identity (Gifford and Wellings 2018). Left-wing Eurosceptics also focused on the impact of EC membership on parliamentary' sovereignty, arguing that EC membership would ‘erode the importance of the vote’ and weaken the capacity for people without wealth to ‘safeguard their futures’ (BBC 1975). ‘Yes’ campaigners emphasised the opportunity for EC membership to anest the decline of both the UK’s economy and its global influence, and dismissed concerns about sovereignty as ‘getting bogged down in a theoretical debate about paper sovereignty’ (BBC 1975). While the electorate endorsed continued EC membership, they did not completely reject the concerns of the ‘No’ campaign; while polling showed that voters agreed EC membership would reduce the sovereignty of the UK, however, they were so disillusioned with the major parties and the Westminster politics that they did not perceive such a loss to be particularly lamentable (Gifford and Wellings 2018). The polls also showed that voters placed great value on what they perceived as Britain’s special historic heritage, the culture and institutions (especially the Monarchy) that distinguished the UK from Europe, and its ‘unique national identity’ (Gifford and Wellings 2018: 271). They did not share the views of the ‘No’ campaign, however, that EC membership posed a threat to them.

If there is one reason, the results of the 1975 and 2016 referendums were so different despite the arguments for and against EC/EU membership being so similar, it is because the calculation of how great a risk European integration posed to the distinctive culture, institutions, heritage and identity of Britain changed substantially in the intervening 41 years. This explains why the relationship between religious belonging and Euroscepticisrn changed so markedly in that period - and in some cases, even more dramatically than in the wider electorate. While voters who identified with religious communities in 1975 were more likely to be socially conservative and (in the case of national church Protestants) to hold exclusive national identities (such as English/Englishness) just as in 2016, they were less likely to view EC membership as a threat to those values during the first referendum. In addition, the extent of European integration, and power afforded to European institutions, was far more limited before the Maastricht, Amsterdam and Lisbon Treaties of the 1990-2000s. There was less reason, therefore, for historic concerns about the overlap between European integration and Catholic dominance that would later underpin Protestant hostility to EU membership to trigger a Eurosceptic response.

Christian voters in 1975 were not only characterised, however, by less reason to be Eurosceptic than in 2016 - they were less Eurosceptic than non-religious voters at the time as well, especially Anglicans. Figure 3.1 shows the support for ‘Yes’ and ‘No’ of the largest religious communities in the 1975 referendum (as well as the religiously unaffiliated). Hostility to EC membership was greatest - at 34 per cent, just over a point higher than the national result - amongst the non-religious, while it was lowest

■ No, % DYes, %

Figure 3.1 Vote ill the 1975 referendum by religious denomination, per cent Source British Election Study 1975 Referendum Survey, post-election, face-to-face

Euroscepticisni and religion before Brexit 37 amongst the Christian communities, and particularly Protestants. The most pro-EC were Baptists, only 19 per cent of whom opposed membership, and Methodists, of whom 23 per cent opposed it. They were closely followed by Anglicans, Presbyterians and Catholics, of whom 24, 27 and 30 per cent, respectively, voted ‘No’. As we explain in Box 3.1, the difference between Anglicans and the non-religious is statistically significant, further supporting the view that at least some religious voters - and in this case, the largest Christian community in Britain - were more likely to support EC membership than the non-religious. While the data available to explain this pro-EU disposition among Anglicans are limited, it is likely to reflect the same values and identities that were to drive Anglicans towards a far more Eurosceptic position in later years. As described in Chapters 2 and 5, Anglicans are characterised by social conservatism and a strong sense of English national identity. While in 2016 these values led Anglicans to be more likely to perceive EU membership as a threat to the values, culture and institutions of then' conununities (local and national), in 1975 most voters felt that EC membership would protect and even promote the culture and prosperity of the UK. To the extent that Anglicans were more likely to prioritise British and English culture, institutions and identity, and to lament the strain placed on them by the decline of Britain’s empire and global standing, they may well have perceived EC membership as a way of preserving and protecting those elements of British and English life they most cared about - a perception that was to change dramatically over the following decades.

Text Box 3.1 Tests of statistical significance

Throughout this book, we present confidence intervals that surround our estimates of the attitudes and/or behaviour of religious voters. These are 95 per cent confidence intervals and represent a measure of the uncertainty that surrounds our estimate of a particular trait in the population based on our identification of it in a survey sample. In the BES sample, for example, 76 per cent of Anglicans supported EC membership in 1975, and so we estimate that 76 per cent of Anglicans in the electorate did so too - but there is a degree of uncertainty around that estimate, which is represented by our confidence interval (i.e. the error bars) in Figure 3.1. When comparing differences between groups within the same sample, we can usually assume that instances where those error bars do not overlap indicate a statistically significant difference, that is, a difference present in the population as well. The error bars surrounding the estimate of Anglicans’ support for ‘Yes’ in Figure 3.1 do not overlap with thosefor the non-religious, for example, and so we conclude that this is a meaningful, significant difference.

Frequently in our analysis, confidence intervals will overlap, and in some cases, even when the difference between two groups is as large as that between two other groups for which the confidence intervals do not overlap. In Figure 3.1, for example, 77 per cent of Methodists supported ‘Yes’ - even more than the 76 per cent of Anglicans - and yet the difference between Methodists and the non-religious is not statistically significant. This is because there are far fewer Methodists in the BES sample than there are Anglicans, meaning we have less data on which to base an estimate of support for EC membership in the wider electorate. This means we cannot be as confident in the specificity of our estimate, and the larger confidence intervals surrounding the Methodist estimate reflect this. This does not mean that there was no difference in the support for EC membership between Methodists and the non-religious, or that the estimate of 77 per cent of Methodists voting ‘Yes’ is worthless; it simply means that we cannot be confident the difference between Methodist and non-religious support for EC membership was replicated in the British electorate.

Where possible, we highlight differences that are statistically significant in our analyses, and we use a wide range of statistics and analyses to verify the importance of the data we present and the differences between voters exhibiting various religious characteristics. We also make the frill tabular outputs for our regression analyses, as well as our data and coding files, available on the project website, so that readers can investigate the data and scrutinise the importance of statistical findings for themselves.

There were also clear ‘religious effects’ stemming from other religious characteristics. The 1975 BES does not include distinct measures of religious behaviour or belief but includes a measure of religiosity instead. This represents respondents' degree of religious commitment and the importance of religion in their lives, and is bound up with both then behaviour and their beliefs: in short, higher religiosity implies more religious behaviour and stronger religious beliefs (McAndrew and Voas 2011; Bruce 2018). Figure 3.2 shows that respondents who are more religious tended to be more supportive of EC membership, and these differences were statistically significant: more than eight in ten ‘very religious’ respondents voted ‘Yes’ compared with seven in ten who were ‘not religious’.


20% -

  • - I
  • ---- Ï
  • — - T________
  • 1


Not really

To some extent

Very much so

-»-Yes No

Figure 3.2 Vote choice in the 1975 referendum by religiosity, per cent

Source: British Election Study 1975 Referendum Survey, post-election, face-to-face

As we detailed in Chapter 2, this is likely to reflect religious voters’ perception that values and objectives promoted by their faith are identifiable and achievable through European integration, such that the more strongly people held to those beliefs the more likely they were to support EC membership as a way of seeing them realised. It will also reflect the development of values emphasising tolerance, cooperation and the duty to help others even beyond one’s national community, which are promoted through religious interaction, as well as the influence of cues provided by religious elites during religious activities (such as attending church). The Church of England, for example, endorsed the spirit and the aims of Eur opean integration during the General Synod on Europe in 1972, reflecting a propensity to support EC membership that would largely persist for the following 40 years (Mudrov 2015; Grebe and Worthen 2019; Smith and Woodhead 2018). The Vatican also consistently expressed positive sentiment regarding European integration fr om Pope Pius XII, who led the Church between 1939 and 1958 and ‘gave all his support to the construction of the European community’ (Mudrov 2015: 523), and which has remained the official position adopted by the Holy See ever since (Nelsen et al. 2001). The more religiously active someone was, the more likely they were to interact with religious elites and other members of the congregation who were also exposed to such cues, instilling and reinforcing a tendency to view EC membership as a way of promoting and achieving Christian objectives.

In 1975, therefore, the effect of religion on Euroscepticisin was quite easy to characterise: religious belonging, behaviour and belief depressed Euroscepticisin and promoted support for EC membership. There were virtually no differences between the religious communities: what mattered most was whether people identified with a religious community, participated in religious activities and held religious beliefs. This is not to say that religion was the decisive characteristic that determined the outcome of the referendum. The differences between how members of religious communities, or those of differing religiosities or socialisation experiences, were small, and they are smaller still once we account for other characteristics related to Euroscepticisin and/or one’s propensity to express religious traits. Older people, for example, are more likely to identify with a religious community, hold religious beliefs and participate in religious activities, reflecting their socialisation into a society in which religion played a more prominent and widespread role (Clements 2015; Wuthnow 2002). Older people are also more likely to be Eurosceptic as a result of having been socialised into a UK that was not part of the EC or EU and when the ‘normal’ state of affairs was not for the UK to pool sovereignty with other countries (Fox and Pearce 2018; Down and Wilson 2013). We need to account for the influence of these characteristics when identifying the effect of religion, therefore, to avoid biased estimates of how important and influential religion actually was.

To do this, we use regression analysis: a statistical technique that allows us to examine how a trait such as religious identification and religiosity is related to support for EC membership while accounting for the effect of other characteristics also related to EC membership, such as age or gender.1 We modelled the likelihood of BES respondents voting ‘No’ in the referendum, depending on their religious denomination, religiosity and childhood religiosity, while controlling for age, education, gender, trade union membership, party identification and attitudes towards immigration.2 The results of the model are reported in Figure 3.4. We used the results of the regression analyses to predict the average probability of someone who identified with a religious community voting ‘No’ in the referendum (i.e. expressing Euroscepticisin) after the control variables were accounted for, and these probabilities are plotted in the graph.

The analysis shows that most of the differences in Euroscepticisin between people with different religious characteristics were actually the result of other characteristics exhibited by those who were more likely to identify with a religion or have been raised in a religious household - such as being older. That said, there remained clear differences in the probability of Christians voting ‘No’ and the non-religious, comparable in magnitude to those shown in Figure 3.1. On average, the religiously unaffiliated had a

  • 60%
  • 50%
Predicted probability of voting ‘No’ in the 1975 referendum by religious belonging and religiosity of childhood home, per cent

Figure 3.3 Predicted probability of voting ‘No’ in the 1975 referendum by religious belonging and religiosity of childhood home, per cent

Source British Election Study 1975 Referendum Survey, post-election, face-to-face. Predicted probabilities calculated using STATA ‘margins’ command

30 per cent likelihood of voting ‘No’, compared with 24 per cent for Anglicans, 23 per cent for Catholics, and between 19 and 22 per cent for Presbyterians, Baptists and Methodists, hi other words, while there was virtually no difference m Euroscepticism between Christian conununities, there was a difference between Christians collectively and the non-religious. While only the difference between Anglicans and the non-religious was significant -and only at the 90 per cent confidence level - the small number of respondents in other Christian communities inevitably means that the confidence intervals are very large, and there is no evidence of a substantial difference at all between Anglicans and the other Christian communities. The data support the impression given by Figure 3.1, therefore, that differences between Christians in 1975 were negligible, but on the whole Christians were more likely to support EC membership than the non-religious.

Overall, these analyses show religion did play a role in shaping Euroscepticism in 1975. In general, Christians were more likely to support EC membership than the non-religious (significant at the 90 per cent confidence level), as were those who were raised in a somewhat religious household compared to a non-religious one (significant at the 95 per cent confidence level). The characteristics were exhibited by a large proportion of the British electorate; just over half of BES respondents were raised in a religious household, while only a third were religiously unaffiliated, and four out of ten were Anglican. The influence of their religious characteristics was nonetheless relatively weak - certainly compared with age (younger people were more likely to vote ‘Yes’), education (those with more education were more likely to vote ‘Yes’), attitudes towards immigration (those who were tolerant of migration were more likely to vote ‘Yes’) and party identification (Conservatives and Liberals were more likely to vote ‘Yes’).3 To the extent that religion played a part in ensuring the UK remained a member of the EC in 1975, it had a positive but small effect on support for the institution - albeit one that was experienced by a majority of the electorate.

< Prev   CONTENTS   Source   Next >