News |

Religion and the British Social Attitudes 2016 Survey

Posted on October 11, 2017 by Ben Clements

The British Social Attitudes (BSA) 2016 survey dataset has been released via the UKDS. This post updates the long-term religious data available from the BSA surveys.

Figure 1 charts the data on affiliation for the period 1983-2016. Key features include the long-term decline in the proportion identifying as Anglican (which stood at 40% in 1983 and had declined to 15% in 2016), increased identification with non-Christian faiths over recent decades (3% in 1983, 6% in 2016), broad stability in levels of Catholic affiliation (10% in 1983, 9% in 2016), and the increase in the proportion with no affiliation (32% in 1983 and 53% in 2016). The proportion of other Christians has also increased over time, from 15% in 1983 to 17% in 2016. However, the composition of this group has shifted. The proportion identifying as non-denominational Christians has risen over time, with a decreasing share professing a denominational affiliation – in particular, with the Nonconformist churches.

Figure 1: Religious affiliation in Britain, 1983-2016

Source: Author’s analysis of BSA surveys.

Figure 2 shows levels of religious attendance between 1983 and 2016. Attendance has been divided into three categories: attending once a month or more often (or frequent attendance); attending less often (infrequent attendance); not attending. The proportion reporting that they never attend religious services (beyond going for the traditional rites of passage – baptisms, marriages and funerals) increased from 56% in 1983 to 66% in 2016. There has been some decline in the levels of frequent and infrequent attending: attending once a month or more fell from 21% in 1983 to 18% in 2016. The proportion attending on an infrequent basis declined from 23% in 1983 to 16% in 2016.

Figure 2: Religious Attendance in Britain, 1983-2016

Source: Author’s analysis of BSA surveys.

Looking at patterns of attendance in more detail, Figure 3 charts, over time, the proportions of Anglicans, Catholics and other Christians attending church on a frequent basis. Clearly, Catholics and other Christian have consistently reported higher levels of regular churchgoing compared to Anglicans. In 1983, 55% of Catholics and 47% of other Christian reported attending church frequently. In 2016, the proportions had fallen to 43% of Catholics and 38% of other Christians. Anglicans actually show something of an increase in regular attendance, based on the full duration of the BSA data. It stood at 18% in 1983 and, perhaps somewhat surprisingly, increased to 24% in 2016 (the highest proportion recorded by the BSA), having previously stood at 18% in both 2014 and 2015. Overall, though, those belonging to non-Christian religions show the highest level of regular attendance at services. In nearly all recent surveys, a majority of this group has reported attending on a frequent basis (51% in 2016).

Figure 3: Regular attendance at religious services by Christian tradition, 1983-2016

Source: Author’s analysis of BSA surveys.

Table 1 provides a summary of religious data from the BSA surveys. The data on religion of upbringing show that the proportion saying they were raised within the Church of England has fallen from 55% in 1991 (when the question was first asked) to 28% in 2016. The proportion saying they were raised within a Catholic household was 14% in both years. The proportion raised within some other Christian tradition increased from 22% to 27%. The proportion raised within a non-Christian religion stood at 3% in 1991 and 6% in 2016. The proportion without a religious upbringing was 6% in 1991 and 25% in 2016.

Table 1: Summary of religion indicators

Affiliation	1983 (%)	2016 (%)
Church of England	40	15
Roman Catholic	10	9
Other Christian	15	17
Other religion	3	6
No religious affiliation	32	53
Religion of upbringing	1991 (%)	2016 (%)
Church of England	55	28
Roman Catholic	14	14
Other Christian	22	27
Other religion	3	6
No religion	6	25
Attendance	1983 (%)	2016 (%)
Once a month or more often	21	18
Less often than once a month	23	16
Never attends	56	66

Source: Author’s analysis of BSA 1983, 1991 and 2016 surveys.

Posted in church attendance, Measuring religion, Research note, Survey news, Uncategorized | Tagged British Social Attitudes, British Social Attitudes 2016 Survey, British Social Attitudes Surveys, Christian traditions, Religious Affiliation, religious attendance, Religious change, religious upbringing, Survey Data, Time, United Kingdom Data Service | 5 Comments

How religious groups voted in the 2016 referendum on Britain’s EU membership

Posted on May 11, 2017 by Ben Clements

Recent research has shed light on the voting preferences of the British electorate at last year’s Brexit referendum, looking at how support for remain or leave was distributed across a range of socio-demographic groups, as well as showing how it varied based on party support, policy preferences and ideological beliefs. How did religious groups in wider society vote? Were some more likely than others to have voted to leave the EU or vice versa? Data from wave 9 of the British Election Study Internet Panel Study, undertaken after the referendum (with fieldwork conducted by YouGov between 24 June-6 July 2016), allow a comparison of voting behaviour based on religious affiliation. The core sample for wave 9 is used, which enables cross-sectional analysis of the data.

The figure below shows the proportions voting remain and leave within different religious groups. Some groups showed an even split between the two options on the ballot (Methodists and Baptists) and some showed a slight preference for one side or the other (Catholics and Church of Scotland / Presbyterian for remain; Jews and other Christian for leave); but more distinct voting patterns are also evident. Those who identify themselves as Anglican or Church of England were clearly in the leave camp – 60% backed Britain leaving the EU and 40% supported staying. Muslims were clearly in the remain camp, with 69% choosing this option and 31% in favour of leaving the EU. Those with no religion (a group with a younger age profile) were in the ‘remain camp’, by 57% to 43%, as were those belonging to other non-Christian faiths (55% to 45%) and those who preferred not to disclose their religious affiliation (55% to 45%).

Source: Author’s analysis of BES Internet Panel Study 2014-2018, wave 9 (core sample).

References

Fieldhouse, E., J. Green., G. Evans., H. Schmitt, C. van der Eijk, J. Mellon & C. Prosser (2016) British Election Study Internet Panel Waves 1-9.

Posted in Religion and Politics, Religion in public debate, Research note, Survey news | Tagged Brexit, British Election Study 2015, European Union, referendum, Religious affilation, religious belonging, Survey Data, Voting Behaviour | 2 Comments

Religion and the British Social Attitudes 2015 Survey

Posted on January 9, 2017 by Ben Clements

The British Social Attitudes (BSA) 2015 survey dataset was released recently and this post provides a brief analysis of long-running data on religion available from the BSA surveys. It focuses on religious affiliation, religion of upbringing and religious attendance.

Firstly, Figure 1 reports the time-series data on affiliation for the period 1983-2015. Noteworthy features include the long-term decline in the proportion identifying as Church of England (40% in 1983, 19% in 2015), increased identification with non-Christian faiths over recent decades (3% in 1983, 8% in 2015), broad stability in levels of Catholic identification (10% in 1983, 9% in 2015), and the steady upwards climb of the proportion with no affiliation (32% in 1983, 49% in 2015). The proportion of other Christians has increased over time, from 15% in 1983 to 17% in 2015. However, the composition of this group has shifted. The proportion identifying as non-denominational Christians has risen over time, with a decreasing share professing a denominational affiliation – in particular, with the Nonconformist churches.

Figure 1: Religious affiliation in Britain, 1983-2015

Source: Author’s analysis of BSA surveys.

Secondly, Figure 2 shows the levels of religious attendance over time. For clarity of presentation, attendance has been divided up into three categories: attends once a month or more often (or frequent attendance); attends less often (infrequent attendance); does not attend. In contrast to the shifting patterns of affiliation across recent decades, attendance shows a picture of somewhat less marked change. The proportion reporting that they never attend religious services (beyond going for the traditional rites of passage) has risen from 56% in 1983 to 65% in 2015. There has been some decline levels of frequent and infrequent attending: attending once a month or more fell from 21% in 1983 to 18% in 2015 (with weekly-attending falling from 13% to 10%); the respective figures for those attending less often are 23% and 17%.

Figure 2: Religious Attendance in Britain, 1983-2015

Source: Author’s analysis of BSA surveys.

Third, Table 1 provides a summary of the three long-running indicators of religion carried in the annual BSA surveys – affiliation, religion of upbringing and attendance. The data on religion of upbringing show that the proportion saying they were raised within the Church of England has fallen from over half (55%) in 1991 (when the question was first asked) to around three-in-ten in 2015. The proportion saying they were raised within a Catholic household is the same in each survey. The proportion raised in some other Christian context has increased (from 22% to 28%). The internal composition of this category has shifted markedly over time. In 1983, the 22% was comprised of 19% who had a particular denominational affiliation and 3% who did not. In 2015, the 28% was made up of 10% with a denomination and 18% with no denominational affiliation. The proportion whose religion of upbringing was a non-Christian faith was 3% in 1983 and 9% in 2015. There has been a marked increase in the proportion saying they were not raised within any religion: from 6% in 1983 to 20% in 2015.

Table 1: Summary of religious indicators, 1983 and 2015

Affiliation	1983 (%)	2015 (%)
Church of England	40	17
Roman Catholic	10	9
Other Christian	17	17
Other religion	2	8
No religious affiliation	31	49
Religion of upbringing	1991 (%)	2015 (%)
Church of England	55	29
Roman Catholic	14	14
Other Christian	22	28
Other religion	3	9
No religion	6	20
Attendance	1983 (%)	(2015)
Once a month or more often	21	18
Less often than once a month	22	17
Never attends	56	65

Source: Author’s analysis of the BSA 1983 and 2015 surveys.

Percentages have been rounded and may not sum to 100.

Fourthly, and looking at the religious data in the BSA 2015 survey more detail, Table 2 shows the levels of non-affiliation, not having been raised within a religion and non-attendance at religious services for sex and age group.

Table 2: Indicators of secularity by demographic group

		No religious affiliation (%)	No religious upbringing(%)	Does not attend religious services (%)
*Sex*	Men	55	21	69
	Women	43	19	62
*Age group*	18-24	63	37	68
	25-34	58	31	66
	35-44	54	24	65
	45-54	51	18	65
	55-64	45	12	68
	65-74	34	7	61
	75+	24	5	59

Source: Author’s analysis of the BSA 2015 survey.

Having no affiliation is somewhat more common amongst men than women, as is not attending religious services. Across the age groups, there is considerable variation in two of the three indicators of secularity. Younger age groups are much more likely to report that they did not have a religious upbringing and do not have currently have a religious affiliation. There is much less variation across age groups in terms of the proportion not attending church or some other place of worship. Whereas amongst 18-24 year olds, 63% have no affiliation (compared to 24% of over 75s) and 37% were not raised within a religious faith (compared to just 5% of over 75%s), a majority in both groups said that they do not attend religious services (aged 18-24: 68%; aged 75 and over: 59%).

Posted in church attendance, Measuring religion, News from religious organisations, Survey news, Uncategorized | Tagged British Social Attitudes, British Social Attitudes 2015 Survey, British Social Attitudes Survey, British Social Attitudes Surveys, church attendance, Religious affilation, religious attendance, religious belonging, religious upbringing, Social Surveys, Survey Data, surveys | 1 Comment

The European Social Survey 2014: Political Attitudes of Religious Groups in Britain

Posted on June 20, 2016 by Ben Clements

This short BRIN post is the second one looking at religious data in Britain based on analysis of the European Social Survey (ESS), a cross-national survey which has so far involved seven waves conducted every two years since 2002. The first BRIN post is available here:

http://www.brin.ac.uk/2016/the-european-social-survey-and-religion-in-britain/

This second post looks at the attitudes of religious groups on a selection of political issues, using the most recent ESS survey from 2014. In each wave, the UK adult population has been sampled, but the analysis here is restricted to those living in Britain (and so excluding the small subset of respondents in Northern Ireland). The ESS country datasets for the UK can be downloaded (along with accompanying documentation) from the ESS website:

http://www.europeansocialsurvey.org/data/

Party Support

The ESS has regularly asked two questions which gauge support for political parties. Firstly, a question asking whether respondents feel close to a party and, if they do, which one. Secondly, a question asking about which party they supported in the most recent national election (if the respondent said they had voted).

Responses to these two questions in the 2014 ESS are displayed in Figure 1 (feel close to a party) and Figure 2 (party supported at most recent national election), based on five categories of religious affiliation (CofE / Anglican, Roman Catholic, other Christian, other religion, no religion).

On both measures – voting behaviour in the 2010 general election and general closeness to a party – the traditional party-denominational linkage between Anglicans and the Conservative Party is upheld. Anglicans are more likely to feel close to the Conservatives than to Labour or any other party; while around half said they voted Tory in the 2010 general election.

Figure 1: Party feel close to, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

All other groups are more likely to feel close to Labour than to the Conservatives; with this difference most pronounced for those affiliated to non-Christian religions. Similarly, all groups (except for Anglicans) reported that they were more likely to have backed Labour in 2010, with the divide again most pronounced for those from non-Christian religions. Those within group were around three times as likely to feel close to Labour compared to the Conservatives and three times as likely to have voted for Labour than the Conservatives in 2010.

It is worth noting that, for each religious group, the most common response when asked is to not feel close to any party (highest at nearly three-fifths of those from a non-Christian religion).

Figure 2: Party voted for at most recent national election, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

Left-Right Ideology

To gauge left-right ideological position, the ESS surveys have asked respondents to locate themselves on a scale ranging from 0 to 10, where 0 represents most left-wing and 10 represents most right-wing. The average score for each group on this scale is shown in Figure 3. It is evident that Anglicans position themselves more to the right than the other groups, with a mean score of 5.5, ahead of other Christians with an average of 5.2. The other three groups have somewhat lower averages of 4.8 or 4.9.

Of all the groups, then, Anglicans are most likely to express support for the Conservative Party and are more right-wing in their ideological positioning.

Figure 3: Mean scores on left-right self-placement scale, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

Note: 0-10 scale, where 0=most left-wing and 10=most right-wing.

Attitudes towards Gays and Lesbians

The ESS surveys have asked a question about gays and lesbians being able to live life as they wish. Respondents’ views are gauged by a Likert scale running from strongly agree through to strongly disagree. Here, the strongly agree and agree categories, and the disagree and strongly disagree categories, have been combined. Figure 4 reports the responses for the religious groups. For all groups a majority agrees with the statement – highest at 92% of those with no religion and lowest at 57% of those from some other religion. Very small proportions disagree, although this rises to a fifth of those belonging to some other religion (and which outweighs the proportion with a neutral viewpoint).

Figure 4: Attitudes towards gays and lesbians being free to live life as they wish, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

European Integration

Respondents’ opinion on European integration are measured by means of a self-placement scale asking about unification. On this scale, a score of 1 indicates that unification has already gone too far and a score of 10 indicates unification should go further. The mean scores are shown in Figure 5. All group scores are below the scale mid-point, but they are somewhat higher for Catholics (4.4) and those belonging to some other religion (4.6). Anglicans registered the lowest mean score (at 3.2).

Figure 5: Mean scores on a European unification self-placement scale, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

Note: 0-10 scale, where 0=unification has already gone too far and 10=unification should go further.

Immigration

Finally, the ESS has asked a series of questions on immigration – responses to three of which are shown in Figure 6. It reports the mean scores by religious group for three self-placement scales where respondents have been asked: whether immigration is generally good or bad for the country’s economy; whether the country’s cultural life is generally undermined or enriched by immigration; and whether immigrants make the country a better or worse place to live. The scales ranged from 0 to 10, with higher values representing more more positive evaluations.

Across all three questions, more positive assessments of the effects of immigration come from Catholics and those from non-Christian religions. Less favourable evaluations are evident on the part of Anglicans and those with no religion.

Figure 6: Mean scores on immigration self-placement scales, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

Note: 0-10 scales, where higher scores represent more positive evaluations of the effects of immigration and immigrants.

The 2014 ESS also included a special module of questions asking about various aspects of immigration. Responses to two such questions with a religious theme are displayed in Figure 7, which also used self-placement scales ranging from 0 to 10. Again, mean scores are presented for each religious group. Firstly, in response to a question asking how important it is for someone to have a Christian background when deciding on whether an immigrant can come and live in this country. A score of 0 equals extremely unimportant and a score of 10 equals extremely important. Secondly, in response to a question asking whether, in general, religious beliefs and practices are undermined or enriched by immigrants coming to live in this country. A score of 0 represents a view that beliefs and practices are undermined; a score of 10 that they are enriched.

Anglicans were more likely to think that having a Christian background was an important factor when making decisions to admit immigrants (at 4.1, but clearly below the midpoint of the scale). Those from other religions were least likely (at 2.0), followed by those with no religion (2.2). Those belonging to non-Christian religions were the most positive in their assessment of the impact of immigration on religious beliefs and practices (5.9); Anglicans were least positive, with a mean score of 4.0.

Figure 7: Mean scores on immigration self-placement scales, by religious affiliation

Source: Author’s analysis of ESS 2014. Weighted data. GB sample only.

Note:

[1] Christian background: 0-10 scale, where a score of 0 equals extremely unimportant and a score of 10 equals extremely important

[2] Religious beliefs and practices: 0-10 scale, where a score of 0 represents a view that beliefs and practices are undermined; a score of 10 that they are enriched

Posted in Religion and Politics, Religion in public debate, Research note, Survey news, Uncategorized | Tagged European Social Survey, European Social Surveys, European Union, immigrants, Immigration, left-right scale, lesbians/gays/bisexuals, party identification, Political Attitudes, political issues, political parties, politics, Religion and Politics, Religious Affiliation, religious belonging, Social Surveys, Survey Data, Voting Behaviour | Leave a comment

The European Social Survey: Religion in Britain

Posted on June 7, 2016 by Ben Clements

This BRIN post looks at religious data pertaining to Britain from the European Social Survey (ESS), a cross-national survey which has so far involved seven waves conducted every two years since 2002. In each wave, the UK adult population has been sampled. The most recent survey wave was conducted in 2014 – the UK country dataset has recently been released and can be downloaded (along with accompanying documentation) from the ESS website: http://www.europeansocialsurvey.org/data/.

The religious data presented here are based on analysis of the 2002 and 2014 surveys, in order to provide an over-time comparison. While each of the samples covers the UK, the small proportion of cases resident in Norther Ireland have been omitted, so that the focus it on those in living in Britain.

Four religious indicators are used here: affiliation; attendance, prayer, and personal religiosity. The analysis first examines each religious indicator in turn before looking at how different measures of religious engagement (attendance, prayer and religiosity) are associated with belonging to a particular faith or denomination. For attendance and prayer, the full set of response options provided in the ESS surveys have been collapsed into more parsimonious sets of categories. Question wordings are given underneath each table. All tables present the results from analysis of weighted data.

Religious indicators

Table 1 presents the data on religion affiliation from the 2002 and 2014 surveys. Over time, the total proportion professing some form of Christian affiliation has been broadly stable (2002: 43%; 2014: 42%). However, as an identical set of response categories was not used for Christian traditions in both surveys, this limits the observations that can be made. The proportion claiming no religious affiliation is almost identical over time (2002: 52%; 2014: 53%). The proportion recorded in 2014 is a little higher than that recorded in the 2014 British Social Attitudes survey. There has been an increase in the proportion belonging to some other religion (from 5% to 7%).

Table 1: Religious affiliation

	2002 (%)		2014 (%)
TOTAL CHRISTIAN	43	TOTAL CHRISTIAN	42
Protestant	33	Anglican	24
Catholic	8	Catholic	10
Other Christian	2	Other Christian	8
OTHER RELIGION	5	OTHER RELIGION	7
NO RELIGION	52	NO RELIGION	53

Source: Author’s analysis of ESS 2002 and 2014.

Questions: ‘Do you consider yourself as belonging to any particular religion or denomination?’ and ‘Which one?’

Note: Some of the categories included under ‘CHRISTIAN’ are not equivalent between the 2002 and 2014 surveys.

Table 2 is based on responses to a question only asked of those who said they had no religious affiliation. It gauges whether they have ever belonged to a religious faith or denomination. In both surveys around three-in-ten indicate that they have (though it is slightly higher in 2014). In both years, then, a large majority of those with no current affiliation also stated that they have never had an affiliation in the past.

Table 2: Ever belonged to a particular religion or denomination (only asked of those with no affiliation)

	2002 (%)	2014 (%)
Yes	28	32
No	72	69

Source: Author’s analysis of ESS 2002 and 2014.

Question: ‘Have you ever considered yourself as belonging to any particular religion or denomination?’

Table 3 presents data for the first of three measures of religious engagement – attendance at religious services (beyond going on special occasions). The picture is one of continuity over time – just under a fifth report that they attend services on a frequent basis (that is, once a month or more often); around three-in-ten attend less often; and about half said that they never attend services.

Table 3: Religious attendance

	2002 (%)	2014 (%)
Once a month or more	18	19
Less than once a month	32	30
Never	51	51

Source: Author’s analysis of ESS 2002 and 2014.

Question: ‘Apart from special occasions such as weddings and funerals, about how often do you attend religious services nowadays?’

Table 4 shows the responses to a question asking about prayer. It shows an increase over time in the proportion saying that they never pray, from 44% in 2002 to 50% in 2014, with small decreases in the proportions saying that either they pray at least once a week or less often.

Table 4: Prayer

	2002 (%)	2014 (%)
Once a week or more	31	29
Less often	25	22
Never	44	50

Source: Author’s analysis of ESS 2002 and 2014.

Question: ‘Apart from when you are at religious services, how often, if at all, do you pray?’

Moving beyond measures of religious practice, Table 5 shows responses to a question asking respondents to self-assess how religious they are. They are asked to locate themselves on a scale running from 0 to 10, where 0 indicates not at all religious and 10 indicates very religious. In Table 5, respondents have been categorised as to whether they have a low (scored 0-3), medium (scored 4-6) or high (scored 7-10) level of religiosity, as well as showing the overall mean score for the full scale. There has been some degree of change over time: the proportion with a low level of religiosity has increased from 40% to 48%. The proportion with a medium or high levels of religiosity have both fallen over time. In 2014, just under half have a low level of religiosity, 30% have a medium level (down from 36%) and 21% report having a high level (down from 24%). The average value underscores this movement towards lower levels of religiosity, decreasing from 5.0 to 3.8.

Table 5: Self-assessed religiosity

	2002 (%)	2014 (%)
Low (0 to 3)	40	48
Medium (4-6)	36	30
High (7-10)	24	21
Mean score	5.0	3.8

Source: Author’s analysis of ESS 2002 and 2014.

Question: ‘Regardless of whether you belong to a particular religion, how religious would you say you are?’

Religious engagement by affiliation

Table 6 takes the analysis of the religious data in the 2014 ESS survey a step further by looking at how the indicators of religious engagement are associated with the measure of religious affiliation. In other words, does religious engagement vary across different religious traditions? Table 6 provides a breakdown of attendance, prayer and self-assessed religiosity for Anglicans, Catholics, other Christians and those who belong to other religions. Data are not reported for those who do not have an affiliation.

A common finding across the three indicators of religious engagement is that Anglicans are less likely to be engaged. Anglicans are much less likely to say that they attend religious services on a regular basis (once a month or more); much less likely to report that they pray once a week or more; and are less likely to have a high level of personal religiosity.

Around a half of Catholics, other Christians and those affiliated to non-Christian religions say they attend services once a month or more. A clear majority in each group other than Anglicans also report praying once a week or more often. While 31% of Anglicans are categorised as having a high level of religiosity, this is considerable lower than the proportions for the other groups: Catholics: 47%; other Christians: 45%; other religion: 54%. Looked at another way, Anglicans’ mean score on the religiosity scale is 5.25; the average scores for the other groups are somewhat higher (highest at 6.53 for those belonging to other religions).

Finally, a summary measure of religious engagement was created based on the three indicators used already: attendance, prayer and self-assessed religiosity. Those respondents who met the following criteria of (i) attending services once a month or more, (ii) praying once a week or more and (iii) having a high level of religiosity were classed as having a high level of religious engagement. The proportion that is highly engaged – on this summary measure – within each affiliation category is shown in the bottom row of Table 6.

Within each religious group only a relatively small proportion can be identified as highly engaged on all three measures. The summary measure encapsulates what was found for each indicator when analysed in turn. That is, Anglicans somewhat stand apart from the other religious groups. Only 15% of Anglicans are classed as highly religiously engaged based on the summary measure, compared to around twice as many Catholics (33%), other Christians (31%) and those within non-Christian faiths (30%).

Table 6: Religious engagement by affiliation

	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)
Attendance
Once a month or more	25	47	48	49
Less than once a month	43	34	33	36
Never	33	20	19	15
Prayer
Once a week or more	38	61	55	68
Less often	32	22	23	21
Never	31	17	22	11
Religiosity
Low (0-3)	23	15	14	10
Medium (4-6)	46	38	41	36
High (7 to 10)	31	47	45	54
Mean score	5.3	6.0	6.2	6.5

Proportion with a high level of religious engagement*	15	33	31	30

*Based on a combined measure of: (i) attends once a month or more; (ii) prays more than once a week; and (iii) has a high level of self-assessed religiosity.

Source: Author’s analysis of ESS 2014.

Religious engagement by sociodemographic group

As a final step, Table 7 shows the incidence of different religious indicators across sociodemographic groups (based on sex, ethnicity and age). Specifically, within each group, Table 7 reports the proportion with a religious affiliation, the proportion attending services once a month or more, the proportion praying once a week or more, the proportion with a high level of religiosity, and the proportion categorised as highly religious engaged (based on the summary measure discussed already).

There are some consistent features in the data. Across all indicators, women are always more religious than men: that is, they are more likely to have some form of affiliation, more likely to practice their religion, and more likely to see themselves as being very religious. Based on the combined measure of religious engagement, 13% of women are highly religiously engaged, as against a tenth of men.

Those who belong to a minority ethnic group are much more likely to be religiously engaged those who do not. With the exception of identifying with a religion, those who belong to a minority ethnic group are more than twice as likely to be religiously-engaged. Based on the summary measure (shown in the final column), 31% of those belonging to a minority ethnic group are classed as highly religiously engaged, compared to 9% of those who do not belong to a minority ethnic group.

In terms of the evidence across age groups, those aged 65 and over are most likely to be religiously-engaged, and this finding is consistent across indicators. Those in the youngest age group are consistently least likely to be religiously engaged. Based on the summary index, those aged 65 and older are twice as likely to be highly religiously engaged compared than those aged 15-29.

Table 7: Religious engagement by sociodemographic group

	Has a religious affiliation (%)	Attends services: Once a month or more (%)	Prays: Once a week or more (%)	High level of religiosity (%)	*Religiosity: Mean score*	*High level of religious engagement (%)**
Men	44	17	24	18	3.4	10
Women	50	21	33	24	4.1	13
Belongs to a minority ethnic group	67	38	56	45	5.4	31
Does not belong to a minority ethnic group	45	16	25	19	3.6	9
Aged 15-29	33	11	18	13	2.8	8
Aged 30-49	41	19	26	20	3.4	11
Aged 50-64	49	17	28	21	3.9	11
Aged 65+	64	24	39	28	4.6	16

*Based on a combined measure of: (i) attends once a month or more; (ii) prays more than once a week; and (iii) has a high level of self-assessed religiosity.

Source: Author’s analysis of ESS 2014.

Summary

Across time, the picture is generally one of stability in terms of affiliation and attendance. There was some decline in self-assessed levels of religiosity; and a rise in the proportion who do not pray.

The examination of variation in levels of religious engagement across religious groups (defined by affiliation) and across sociodemographic groups tended to reaffirm the ‘conventional wisdom’ on which segments of wider society tend to be more (or less religious). Across faith traditions, Anglicans are least religiously engaged based on the measures used here, either separately or in combination. Sociodemographically, levels of religious engagement are higher amongst women, those aged 65 and older, and particularly so within those belonging to minority ethnic groups.

Posted in church attendance, Measuring religion, Research note, Survey news | Tagged Anglicans, Behaving, Belonging, Christians, church attendance, Church of England, European Social Surveys, Other Religions, prayer, Religiosity, Religious Practice, Roman Catholics, self-assessed religiosity, Survey Data, surveys | 1 Comment

Re-examining religious and paranormal beliefs in mid-1970s Britain

Posted on May 9, 2016 by Ben Clements

A recent YouGov survey shed interesting light on levels of religious beliefs and other forms of belief in contemporary Britain. This BRIN post takes a historical turn by analysing one of the few available surveys enabling assessment of religious and paranormal beliefs in Britain the post-war period. The analysis is based on a Gallup opinion poll undertaken in May 1975, based on a sample of adults aged 16 and older in Britain. The dataset and accompanying documentation for this survey were obtained from the United Kingdom Data Service. Taken as a whole, post-war Gallup polling in Britain provides an important resource for studying change and continuity in popular religion (for more information see Field, 2015a).

This survey dataset is particularly useful because it allows analysis of the relationship between religious and paranormal beliefs in the British public, and the nature of the association between them. For example, was the relationship between religious and paranormal beliefs tending towards a mutually-exclusive one, in which the holding of the former tended to preclude affirmation of the latter? On the other hand, did individuals subscribe to a mixture of religious or paranormal beliefs and – perhaps – were those who held religious beliefs more likely – than those who did not – to believe in paranormal phenomena? Analysis of this survey can hopefully tell us something about the incidence, patterning and overlap of the two types of belief in mid-1970s Britain.

First, Table 1 shows the distribution of responses for religious and paranormal beliefs, based on the distinction between ‘traditional religious’ and ‘non-traditional religious beliefs’ set out in Gill et al. (1998) and Gill (2003).[i] Field uses a distinction between ‘orthodox’ and ‘heterodox’ beliefs in his recent book on religious change in Britain during the ‘long 1950s’, with the former set of beliefs coming within the ‘framework of traditional Christianity’ (2015b: 74). For each belief, Table 1 shows the proportions responding ‘yes’, ‘no’ or ‘don’t know’. The survey asked about five religious beliefs and eleven paranormal beliefs.

The most prevalent religious belief was in God (71%), followed by believing in heaven (50%). Three other religious beliefs were subscribed to by varying minorities (around a third for life after death, and one fifth for belief in the devil and in hell). Overall, expressed belief in paranormal phenomena varied markedly in British society in the mid-1970s, highest at 51% for being able to forecast (and 48% for thought transference) and lowest at 12% for black magic (and 13% for exchanging messages with the dead). The proportions saying they did not know also varied across the different belief items – highest at 22% for life after death (religious beliefs) and 20% for reincarnation (paranormal beliefs).

Table 1: Overall profile of religious and paranormal beliefs

	Yes (%)	No (%)	Don’t know (%)
Traditional religious beliefs
God	71	17	12
Heaven	50	36	14
Life after death	35	43	22
Devil	20	72	9
Hell	19	72	9
Classic paranormal beliefs
Being able to forecast	51	36	13
Thought transference	48	37	15
Faith healing	43	42	15
Hypnotism	41	48	10
Horoscopes	28	66	7
Reincarnation	23	58	20
UFOs	20	66	14
Lucky charms	20	75	5
Ghosts	18	72	10
Black magic	13	79	8
Exchanging messages	12	77	11

Source: Author’s analysis of Gallup opinion poll, May 1975.

Note: Percentages rounded and sum across the rows. Distinction amongst types of belief based on that used in Gill et al. (1998) and Gill (2003).

Beliefs and sociodemographic groups

The next step in the analysis is to move beyond profiling overall levels of belief and to look at levels of belief across different social groups. Table 2 reports the proportion within each sociodemographic group saying they believe in each religious belief. Tables 3(a) and 3(b) do the same for the set of paranormal beliefs.

The more notable areas of difference for religious belief concern sex, age group and religious affiliation. Religious beliefs, in mid-1970s Britain, were always more common amongst women than men (82% and 60%, respectively, expressed belief in God; 59% and 39%, respectively, expressed belief in heaven). Some beliefs were generally more common amongst older, as compared to younger, age groups (God and heaven). Based on religious affiliation, unsurprisingly religious beliefs were much more likely to be subscribed to amongst those with some form of denominational allegiance (highest amongst Catholics). Those with no religious affiliation exhibited some level of religious belief (a quarter said they believed in God and nearly a fifth believed in in life after death). There are somewhat variant results for levels of belief based on social grade, and the marked difference in belief in God and heaven based on age completed education needs to factor in the tendency for who finished at an earlier age being drawn from the older generations in society.

Table 2: Religious beliefs by socio-demographic group (percent saying ‘yes’)

	God	Heaven	Hell	Devil	Life after death
Overall	71	50	19	20	35
Men	60	39	15	16	26
Women	82	59	23	23	44
Aged 16-24	66	41	21	21	33
Aged 25-34	60	43	16	19	31
Aged 35-44	69	47	18	20	30
Aged 45-54	74	49	17	16	33
Aged 55-64	81	58	20	23	37
Aged 65+	83	65	23	20	46
*Education: 14 or under	80	59	19	18	38
Education: 15	66	43	17	18	30
Education: 16	71	51	20	23	31
Education: 17	66	48	26	26	44
Education: 18 or over	60	35	22	20	42
Social grade: AB	67	48	27	27	38
Social grade: C1	66	45	17	18	35
Social grade: C2	71	48	19	18	34
Social grade: DE	77	56	19	20	36
Church of England	76	52	17	16	34
Church of Scotland	80	57	10	10	40
Nonconformist	84	57	23	20	56
Catholic	87	78	43	44	40
Other religion	67	50	35	36	38
No religion	22	7	6	9	18

Source: Author’s analysis of Gallup opinion poll, May 1975.

*Refers to the age at which a respondent completed their full-time education.

Belief in paranormal phenomena also varied across social groups in Britain in the mid-1970s, as shown in Tables 3(a) and 3(b). As with religious beliefs, women were more likely than men to hold paranormal beliefs (most marked in relation to horoscopes; the exception being belief in UFOs).The patterns of age groups show that younger people were more predisposed to express belief in some paranormal phenomena, and the oldest age group (65 and over) were least likely to do so for some of these beliefs. These age-related differences are particularly evident for belief in black magic, ghosts, hypnotism and UFOs, with younger people expressing higher levels of belief than older people. On the other hand, belief in faith healing tended to be higher amongst older age groups.

Based on religious belonging, there is no consistent pattern for those with no affiliation to be clearly more likely to profess belief in the paranormal. For many beliefs, they are eclipsed by those belonging to particular Christian denominations (or affiliated with some other religious group). As with religious beliefs, there were no consistent differences for social grade. Paranormal belief did tend to be higher amongst those who completed their education later on.

Table 3(a): Paranormal beliefs by sociodemographic group (percent saying ‘yes’)

	Reincarnation	Hypnotism	Black magic	Horoscopes	Thought Transference
Overall	23	41	13	28	48
Men	19	40	12	15	43
Women	26	42	14	39	52
Aged 16-24	25	50	23	31	48
Aged 25-34	20	54	17	26	53
Aged 35-44	16	48	11	23	44
Aged 45-54	21	40	12	31	50
Aged 55-64	26	33	9	28	52
Aged 65+	27	20	6	27	40
*Education: 14 or under	25	28	7	30	45
Education: 15	17	42	12	24	39
Education: 16	21	52	16	29	55
Education: 17	34	62	30	22	50
Education: 18 or over	25	59	23	28	65
Social grade: AB	25	54	17	20	59
Social grade: C1	22	48	15	22	58
Social grade: C2	22	43	13	25	44
Social grade: DE	23	31	11	38	40
Church of England	24	42	13	29	49
Church of Scotland	25	35	7	42	35
Nonconformist	25	35	9	29	57
Catholic	24	35	10	21	46
Other religion	29	47	21	35	55
No religion	9	48	17	17	39

Source: Author’s analysis of Gallup opinion poll, May 1975.

*Refers to the age at which a respondent completed their full-time education.

Table 3(b): Paranormal beliefs by sociodemographic group (percent saying ‘yes’)

	Ghosts	UFOs	Faith healing	Being able to forecast	Lucky charms	Exchanging messages
Overall	18	20	43	51	20	12
Men	16	24	38	45	14	11
Women	20	16	47	56	26	13
Aged 16-24	28	28	38	57	25	15
Aged 25-34	22	24	38	58	18	12
Aged 35-44	14	20	35	49	13	7
Aged 45-54	23	22	51	44	25	16
Aged 55-64	12	12	51	52	23	10
Aged 65+	7	10	45	45	16	10
*Education: 14 or under	12	13	46	46	23	13
Education: 15	18	20	38	52	18	10
Education: 16	19	25	39	55	17	9
Education 17	26	26	38	58	12	12
Education: 18 or over	33	31	50	58	19	20
Social grade: AB	16	23	44	50	16	12
Social grade: C1	22	23	50	57	14	13
Social grade: C2	19	20	38	49	19	10
Social grade: DE	15	17	42	49	27	13
Church of England	20	21	47	54	24	12
Church of Scotland	10	8	33	40	12	8
Nonconformist	16	13	49	56	11	9
Catholic	18	21	29	50	20	11
Other religion	21	26	53	59	26	22
No religion	15	22	32	34	7	10

Source: Author’s analysis of Gallup opinion poll, May 1975.

*Refers to the age at which a respondent completed their full-time education.

Table 4 sheds some light on the crossover of religious and paranormal beliefs amongst the British public in the mid-1970s by showing levels of paranormal belief based on responses to the religious belief questions. That is, we can compare levels of belief in the paranormal amongst those who did or who did not express belief in God, in heaven, and so on.The association between religious and paranormal beliefs

When individuals are grouped by belief in God, belief in the paranormal tended to be highest amongst those who expressed believed in God; not those who expressed disbelief in God. The same pattern is evident when individuals are grouped based on belief (or not) in heaven, in hell, in the devil and in life after death. In other words, there is no obvious basis for saying that religious and paranormal beliefs tended to be mutually exclusive belief systems amongst many individuals. In actual fact, those who held religious beliefs were more likely than those who did not to express belief in a range of paranormal phenomena. Of course, often significant proportions of those who did not subscribe to particular religious did affirm belief in paranormal phenomena, but usually they were outranked by those who did hold common religious beliefs.

Table 4: Paranormal beliefs by religious beliefs (percent saying ‘yes’)

		Hypnotism	Black magic	Horoscopes	Thought transference	Ghosts	UFOs
God	Yes	39	13	31	50	20	20
	No	45	14	18	41	13	17
Heaven	Yes	36	13	33	51	20	21
	No	44	13	21	43	15	19
Hell	Yes	45	22	36	61	32	29
	No	39	10	25	45	14	18
Devil	Yes	45	26	36	61	36	27
	No	39	9	26	44	13	17
Life after death	Yes	49	18	37	61	29	24
	No	34	10	22	37	9	15
		Faith healing	Being able to forecast	Lucky charms	Exchanging messages	Reincarnation
God	Yes	48	52	23	14	29
	No	28	50	12	5	8
Heaven	Yes	50	55	26	15	36
	No	32	47	14	8	9
Hell	Yes	55	60	25	24	44
	No	39	49	19	9	19
Devil	Yes	56	61	28	26	39
	No	39	48	19	8	19
Life after death	Yes	58	63	25	22	43
Life after death	No	31	44	18	7	10

Source: Author’s analysis of Gallup opinion poll, May 1975.

Table 5 looks at the association between these types of belief in reverse: that is, it shows levels of religious belief based on ‘yes’ and ‘no’ responses to the questions on paranormal belief. So, for example, of those who expressed a belief in hypnotism, 68% expressed belief in God, compared to 75% of those who did not believe in hypnotism (and so on).

Table 5: Religious beliefs by paranormal beliefs (percent saying ‘yes’)

		God	Heaven	Hell	Devil	Life after death
Hypnotism	Yes	68	44	21	22	42
	No	75	55	17	16	29
Black magic	Yes	68	48	32	38	48
	No	72	50	16	16	33
Horoscopes	Yes	79	60	25	25	47
	No	68	46	16	17	30
Thought transference	Yes	75	53	25	25	45
	No	70	47	13	13	25
Ghosts	Yes	79	55	34	39	56
	No	71	49	15	15	30
UFOs	Yes	72	54	28	27	43
	No	73	51	17	18	35
Faith healing	Yes	79	58	25	26	47
	No	65	45	15	14	25
Being able to forecast	Yes	73	53	23	23	43
	No	73	50	15	16	28
Lucky charms	Yes	83	64	24	28	44
	No	68	47	18	17	33
Exchanging messages	Yes	82	63	38	42	66
	No	71	49	17	17	31
Reincarnation	Yes	92	79	37	34	67
	No	64	39	13	15	23

Source: Author’s analysis of Gallup opinion poll, May 1975.

A belief typology

Given Tables 4-5 presented the finely-grained detail of how paranormal beliefs are associated with religious beliefs and vice versa, it is helpful to try and distil the essence of this messy and complex association between different types of belief. As a final step in profiling the nature of belief in the British population in mid-1970s, belief indices for religious and paranormal beliefs were computed (based on all of the different beliefs utilised above), in order to arrive at an overall typology of belief. This was undertaken based on the procedures and category labels set out in Rice (2003: 103-104).[ii] Producing two separate belief indices and then looking at the association between them produces a four-fold typology of belief (‘sceptics’; ‘classic paranormal believers’; ‘traditional religious believers’; ‘full believers’), and allows us, in broad terms, to see the proportions contained within each belief type. The results of this analysis are shown in Table 6 (for all individuals) and in Tables 7-8 (respectively, for women and men).

Table 6: Belief typology (ALL)

Tend not to believe in traditional religious phenomena

Tend to believe in traditional religious phenomena

Tend not to believe in classic paranormal

Sceptics: 52%

Traditional religious believers: 24%

Tend to believe in classic paranormal phenomena

Classic paranormal believers: 10%

Full believers: 14%

Source: Author’s analysis of Gallup opinion poll, May 1975.

Note: Based on the analytical procedures set out in Rice (2003: 103-104), and using the same category labels.

We can say that in the mid-1970s – and probably rather surprisingly – about half (52%) could be categorised as ‘sceptics’ – that is, they tended not to believe in traditional religious beliefs and paranormal beliefs, on the basis of the typology constructed based on Rice’s procedures (2003). One in ten (10%) tended to believe in only paranormal phenomena, outranked by the larger share who were ‘traditional religious believers’ (24%). When the belief typology is reproduced for women and men separately, we can see that women were more likely to have ‘full believers’ and ‘traditional religious believers’, and less likely to have been ‘sceptics’. There was no difference in the proportion of ‘classic paranormal believers’ amongst women and men. Again, it should be noted that this typology is based on a single snapshot of popular beliefs from a one post-war survey – of course the results could well be different if varied sets of religious and paranormal beliefs were used; and a more nuanced typology could be applied to the data.

Table 7: Belief typology (WOMEN)

Tend not to believe in traditional religious phenomena

Tend to believe in traditional religious phenomena

Tend not to believe in classic paranormal

Sceptics: 42%

Traditional religious believers: 30%

Tend to believe in classic paranormal phenomena

Classic paranormal believers: 10%

Full believers: 19%

Source: Author’s analysis of Gallup opinion poll, May 1975.

Note: Based on the analytical procedures set out in Rice (2003: 103-104), and using the same category labels.

Table 8: Belief typology (MEN)

Tend not to believe in traditional religious phenomena

Tend to believe in traditional religious phenomena

Tend not to believe in classic paranormal

Sceptics: 64%

Traditional religious believers: 18%

Tend to believe in classic paranormal phenomena

Classic paranormal believers: 10%

Full believers: 9%

Source: Author’s analysis of Gallup opinion poll, May 1975.

Note: Based on the analytical procedures set out in Rice (2003: 103-104), and using the same category labels.

Summary

The empirical results reported above showed that, overall, religious and paranormal beliefs were subscribed to by varying segments of the adult population in mid-1970s Britain. In terms of belief across different social groups, women were consistently more likely affirm belief in both religious tenets and paranormal phenomena. The results for the belief typology showed that, in more general terms, women were more likely to have been ‘full believers’ and ‘traditional religious believers’ than men.

Of course, analysing a single survey from the mid-1970s (and no claims are made here that this period is – or is not – of particular note for studying religious change in general or beliefs in particular in post-war Britain) provides a very limited window into the incidence, patterning and overlap of religious and paranormal beliefs within the British public in the post-war period. However, given the coverage of both types of belief in the survey – generally not available in other survey datasets available to academic researchers, certainly those held at the UKDS – the modest empirical findings presented and discussed here may offer some nuggets of interest to sociologists of religion and social historians focusing on the British context, however time-bound the analysis may be. Replicating this exercise using a contemporary sample survey of the British adult population, with suitable coverage of both religious and paranormal beliefs, may shed some light on areas of change in popular belief across the intervening decades.

References and further reading

Clements, B. (2016), Surveying Christian Beliefs and Religious Debates in Post-War Britain. Basingstoke: Palgrave Macmillan.

Field C. D. (2015a), Religion in Great Britain, 1939-99: A Compendium of Gallup Poll Data. BRIN Working Papers on Religious Statistics. Working Paper 2. February 2015. Available at: http://www.brin.ac.uk/wp-content/uploads/2011/12/Religion-in-Great-Britain-1939-99-A-Compendium-of-Gallup-Poll-Data.pdf

Field, C. D. (2015b), Britain’s Last Religious Revival? Quantifying Belonging, Behaving, and Believing in the Long 1950s. Basingstoke: Palgrave Macmillan.

Gill, R. (2003), The Empty Church Revisited. London: Ashgate.

Gill, R., Kirk Hadaway, C. and Marler, P. L. (1998), ‘Is Religious Belief Declining in Britain?’, Journal for the Scientific Study of Religion, 37(3): 507-516.

Rice, T. W. (2003), ‘Believe It Or Not: Religious and Other Paranormal Beliefs in the United States’, Journal for the Scientific Study of Religion, 42(1): 95-106.

Social Surveys (Gallup Poll) Limited. Gallup Poll, May 1975. [data collection]. UK Data Service. SN: 1330, http://dx.doi.org/10.5255/UKDA-SN-1330-1.

[i] For ‘non-traditional religious beliefs’ Gill et al (1998) noted that: ‘These items reflect a heterodox collection of ideas with decidedly cultic and “pre-Christian’ connotations. They are often disparaged as “superstitious.” Such beliefs may also connect to more recent new religious or “new age” movements with antiinstitutional, nonmaterialistic, and nonrational features’ (512).

[ii] The religious belief index was created by adding up responses towards the five religious beliefs and the paranormal belief index was created by adding up responses towards the eleven paranormal beliefs. For each individual in the sample, a value of 1 was assigned for each phenomenon they did not believe in (‘yes’ responses) and a value of 2 for each phenomenon they did believe in (‘no responses’). ‘Don’t know’ responses were assigned a value of 1.5. The indices ranged from 5-10 for the religious beliefs and 11-22 for the paranormal beliefs. Both indices were then divided into two categories. For the religious belief index, scores of 5–7.5 were given the value of 1 and scores of 8-10 were given the value of 2. For the paranormal belief index, scores of 11-16.5 were given the value of 1 and scores of 17-22 were given the value of 2. For more information, and to see this analysis undertaken for beliefs in the United States, see Rice (2003: 103-104).

Posted in Historical studies, Measuring religion, Religious beliefs, Research note, Survey news, Uncategorized | Tagged 1970s, 1975, Belief in God, believing, British society, Gallup Poll, paranormal, religious beliefs, Survey Data, surveys | Leave a comment

Religion and party choice: Evidence from the BES 2015 face-to-face post-election survey

Posted on April 19, 2016 by Ben Clements

This post presents some data from the BES 2015 face-to-face post-election survey (a cross-sectional study with a sample size of 2,987), the fieldwork for which was undertaken between May-September 2015. The dataset and accompanying documentation can be obtained here. This post looks at voting behaviour at the 2015 general election on the basis of religious affiliation and attendance.

Table 1 shows the overall profile for religious affiliation. Around a fifth (21%) of respondents identified themselves as Anglican / Church of England, 9% identified as Catholic, 6% as belonging to some other Christian denomination, 13% as Christian with no denominational affiliation, 8% as belonging to some other religion, and just over two-fifths (43%) said they did not have a religion.

Table 1: Religious affiliation

	%
Anglican / Church of England	21
Catholic	9
Other Christian denomination	6
Other Christian – no denomination	13
Other religion	8
No religion	43

Source: BES 2015 face-to-face post-election survey.

Table 2 shows the distribution of party vote share at the general election across religious groups. A clear majority of Anglicans said they voted Conservative (59%), compared to 19% of whom supported Labour, while a plurality of Catholics voted Labour (40% compared to 35% who supported the Conservatives). A plurality of other Christians with or without a denominational affiliation also reported voting Conservative. Amongst those identifying with non-Christian faiths, a large majority said they voted for Labour (65%) compared to around a quarter who supported the Conservatives (23%). There was an almost even split amongst those with no religious affiliation, 33% and 34% of whom, respectively, supported the Conservative and Labour parties. Voting for UKIP was lowest amongst those belonging to non-Christian faiths, and highest at over a tenth of Anglicans, other Christians without a denomination, and those with no affiliation.

Table 2: Party voted for at 2015 general election, by religious affiliation

	Con	Lab	Lib Dem	UKIP	Other party
Anglican	59	19	6	13	2
Catholic	35	40	5	9	11
Other Christian denomination	41	28	5	6	21
Other Christian: No denomination	44	32	8	12	5
Other religion	23	65	4	2	6
No religion	33	34	9	12	13

Source: BES 2015 face-to-face post-election survey.

Note: Percentages sum across the rows.

Table 3 shows party vote share at the 2015 general election based on religious attendance, divided into three groups (attends frequently: once a month or more; attends infrequently: less often than once a month; does not attend). Amongst frequent attenders, a plurality supported Labour (44% compared to 39% for the Conservatives) while amongst infrequent attenders there was much higher support for the Conservatives (48%) compared to Labour (28%). Amongst those who said they never attend religious services (whether they had a religious affiliation or not), 39% voted for the Conservatives compared to 31% who supported Labour. Support for UKIP was lowest amongst frequent attenders (5%), and was the same amongst infrequent attenders and non attenders (at 12%).

Table 3: Party voted for at 2015 general election, by religious attendance

	Frequent attender (%)	Infrequent attender (%)	Does not attend (%)
Con	39	48	39
Lab	44	28	31
Lib Dem	7	6	8
UKIP	5	12	12
Other party	6	7	11

Source: BES 2015 face-to-face post-election survey.

Figure 1 shows the party vote share for attendance amongst Anglicans and Catholics. That is, each of these religious denominations has been subdivided into two groups: (i) frequent attenders and (ii) infrequent attenders or non attenders. Amongst Anglicans, there was broadly similar support for the Conservatives amongst those who attend frequently and those who attend infrequently or not at all. Labour support was somewhat higher amongst those Anglicans who regularly attend church. Amongst Catholics, those who go to church services on a regular basis were more likely to say they voted Labour compared to their co-religionists who attend less often or not at all. There was less variation in support for the Conservatives at the 2015 election based on levels of attendance amongst Catholics.

Figure 1: Vote share for the Conservative and Labour parties at 2015 general election, attendance by affiliation

Source: BES 2015 face-to-face post-election survey.

Previous BRIN posts have used BES studies to look at the association between religious factors and party choice at the 2015 general election:

http://www2.brin.ac.uk/news/2015/religion-and-party-choice-at-the-2015-general-election/

http://www.brin.ac.uk/2015/the-2015-general-election-religious-affiliation-and-party-vote-share-across-constituencies/

Reference

Fieldhouse, E., J. Green., G. Evans., H. Schmitt, C. van der Eijk, J. Mellon and C. Prosser (2015) British Election Study, 2015: Face-to-Face Survey [computer file], February 2016.

Posted in Religion and Politics, Research note, Survey news, visualisation | Tagged Affiliation, attendance at religious services, British Election Study 2015, General Election 2015, Party Choice, Survey Data, Voting Behaviour | Leave a comment

Exploring Religious Data in the British Social Attitudes 2014 Survey

Posted on October 15, 2015 by Ben Clements

The latest British Social Attitudes 2014 survey has recently been released via the UKDS, which holds all of the other surveys in the BSA series (as well as the data having being added to the online British Social Attitudes Information System). This post explores religious data available in the 2014 survey.

Religious Belonging

The BSA has asked a question on (current) religious affiliation in every survey conducted since 1983. The figures from the 2014 survey are similar to those obtained in recent years. Around half said they had no affiliation (49%), while those with an affiliation divided into Anglicans (18%), Catholics (8%), other Christian (17%: consisting of 5% with a specific affiliation and 12% with no denominational affiliation), and those belonging to some other religion (8%).

The full set of data on religious affiliation, covering 1983 onwards, is shown in Figure 1, visualised in an area chart which displays the proportion within each of the five categories across time. Noticeable are the clear decline in the proportion affiliating as Anglican (40% in 1983); the growth in the proportion identifying with some other religion (3% in 1983); and the steady increase in those with professing no affiliation – often labelled ‘religious nones’ (32% in 1983). The proportion self-identifying as Catholic was 10% in 1983.

Not shown on the chart but also an important aspect of change in religious belonging is the shift within the other Christian category: those who identify as such but without any specific denominational tradition have become a greater share of this group over time, with a decreasing proportion professing a clear denominational association – in particular, with the Nonconformist churches. The proportion identifying as Christian with no specific denominational allegiance was just 3% in 1983 while those with a denominational allegiance comprised 14%. The proportion in the other Christian category has remained static over time, but there has been considerable change within. Whereas in 1983 self-identification as Anglican was more than twice the level found for other Christian (excluding Catholics), in 2014 the levels are almost identical.

Figure 1: Religious Affiliation in Britain, 1983-2014

Source: BSA surveys.

The BSA surveys have since 1991 asked about the religion (if any) that a respondent was brought up in. In 2014, 30% of respondents said they had been raised as Church of England/Anglican, 15% as Roman Catholic, 27% were raised in some other Christian tradition, with 9% brought up in some other religious faith. Nearly a fifth said they were not raised within a religious faith (19%). Back in 1991, when the question was first asked, the proportions were somewhat different. Well over half (57%) reported being raised as Church of England/Anglican, 14% as Roman Catholic, over a fifth within some other Christian tradition (22%), and 3% within a non-Christian faith. Just 6% said they were not brought up within a religious faith – a figure which has therefore tripled over recent decades. The proportion saying they were raised as Anglican / Church of England has therefore nearly halved over this period.

Religious Behaviour

The BSA surveys have also asked questions on religious behaviour, including on a regular basis about religious attendance and – in 2004 and 2014 – about membership of religious organisations and churches. Data on attendance and membership broken down by affiliation are shown, respectively, in Tables 1 and 2. Church attendance is divided into three categories: frequently (once a month or more often); infrequently (less often than once a month); never attends.

Table 1 shows that, across the categories, regular attendance is least common amongst Anglicans (18%) and much more common amongst Catholics (40%), other Christians (34%), and particularly those belonging to other religions (62%). Data from the 1983 survey show that, similarly, Anglicans were least likely to report being regular-attenders (18%), compared to other Christians (47%) and Catholics (55%). Infrequent attendance is most prevalent amongst Anglicans and they are also most likely to say they never attend services (beyond special occasions). Those belonging to other religions are least likely to report that they never attend services – less than a fifth. Splitting the other Christian category into those with and without a specific denominational affiliation, the latter group have reported higher levels of regular attendance. In 1983, 52% of those with a denomination affiliation said they attended frequently compared to 22% of those with no denomination. Three decades later, this difference is still evident: 48% of those with a denominational affiliation say they attend church regularly compared to 29% of those with no denominational allegiance.

Table 1: Attendance at religious services, by religious affiliation

	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)
Frequently	18	40	34	62
Infrequently	30	26	26	21
Never	52	34	40	17

Source: BSA 2014 survey.

Question wording: ‘Apart from such special occasions as weddings, funerals and baptisms, how often nowadays do you attend services or meetings connected with your religion?’

Overall, the proportions belonging (actively or otherwise) to a religious organisation or church fell somewhat over the decades, from 35% to 26%. Those who used to belong comprised 24% and 25% in, respectively, 2004 and 2014. The proportion saying they had never belonged to such a group rose from just over two-fifths to nearly half.

Table 2 reports data based on affiliation from the 2014 survey. Anglicans are least likely to say that they are active within a group (29%), with active participation most evident amongst Catholics (32%). Catholics are also most likely to say that they are inactive members of a group. Overall, around two-thirds of Catholics belong to a group, compared to two-fifths of Anglicans, and around half of other Christians and those from some other religion. Anglicans and those from other religions are most likely to say they have never belonged to a church or religious organisation. Dividing the other Christian category into those with and without a denominational affiliation again shows a clear difference in levels of involvement. Of those with some form of denominational affiliation, around two-third either actively or passively belong to a church or religious organisation, considerably higher than the two-fifths of non-denominational Christians who say they do so.

Table 2: Belongs to a religious organisation or church, by religious affiliation

	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)
Belong, actively participate	19	32	27	26
Belong, don’t actively participate	20	34	21	22
BELONG	39	66	48	48
Used to belong	24	21	31	15
Never belonged	36	13	21	37

Source: BSA 2014 survey.

Question wording:

‘About belonging to different kinds of groups or associations. Do you belong and actively participate; belong but don’t actively participate; used to belong but do not any more; or have never belonged to – a church or other religious organisation?’

Social Attitudes

The BSA surveys also ask a range of question probing social and political attitudes, which can be examined on the basis of religious belonging or behaviour. Two such examples are used here: same-sex equality and tolerance for religious extremists.

In overall terms, opinion in favour of marriage equality for same-sex couples has increased markedly over time: from 13% in 1989 to 49% in 2007, and settling at around three-fifths in more recent surveys. Table 3 reports the proportions agreeing that gays and lesbians should have the right to marry, based on religious affiliation. The proportions agreeing with the statement have risen considerably over time. Even so, on each occasion, those with no affiliation are most likely to agree with same-sex marriage (increasing from 20% in 1989 to 74% in 2014). While very small proportions of Anglicans, Catholics and other Christians agreed in in 1989, about a half did so in 2014. The proportions in agreement amongst those from non-Christian religions are clearly lower (33% in 2014), although it should be noted that they obviously represent small proportions of the BSA survey samples.

Table 3: Percent agreeing that gays and lesbians should have the right to marry, by religious affiliation

	1989 (%)	2007 (%)	2012 (%)	2013 (%)	2014 (%)
Anglican	10	32	43	46	49
Catholic	14	57	48	56	53
Other Christian	6	40	49	44	52
Other religion	–	33	36	23	33
No religion	20	60	70	69	74

Source: BSA surveys. Combines ‘strongly agree’ and ‘agree’.

Question wordings:

1989: ‘Do you agree or disagree that … Homosexual couples should have the right to marry one another.’

2007-2014: ‘How much do you agree or disagree that … gay or lesbian couples should have the right to marry one another if they want to.’

Finally, Table 4 reports attitudes towards a question asking about religious extremists being allowed to hold public meetings asked in the 2004 and 2014 surveys. In overall terms, attitudes on this question have hardened over time: in 2014, the proportion thinking religious extremists should (definitely or probably) be allowed to hold public meetings is 23%, lower than the 33% expressing the same view in 2004. The same is also the case for each group based on religious affiliation. In general, around a fifth to a quarter think they should be allowed in 2014, with large majorities against – highest at 78% of Anglicans. In 2004, the proportions in favour were in the range of 29-40%, with smaller – but still clear – majorities opposed (with the exception of other Christians, at 49%).

Table 4: Attitudes towards religious extremists being allowed to hold public meetings, by religious affiliation

	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)	No religion (%)
2014
Should definitely or probably be allowed	19	21	20	22	25
Should probably or definitely not be allowed	78	70	73	70	69
Don’t know	3	9	7	9	6
2004
Should definitely or probably be allowed	33	29	40	33	31
Should probably or definitely not be allowed	61	65	49	67	63
Don’t know	6	6	11	0	6

Source: BSA surveys.

Question wording: ‘There are a number of groups in society. Should religious extremists be allowed to hold public meetings?’.

BRIN readers might be interested in earlier posts which used BSA survey data to look in more detail at affiliation and attendance – both levels of change overall and variation across social groups:

http://www.brin.ac.uk/news/2015/socio-demographic-groups-and-religious-affiliation-in-britain/

http://www.brin.ac.uk/news/2014/changes-in-attendance-at-religious-services-in-britain/

Posted in Research note, Survey news, Uncategorized, visualisation | Tagged British Social Attitudes, British Social Attitudes Surveys, British society, Christianity, Christians, church attendance, Church of England, gay rights, group membership, Non-Denominational Christians, Religious Affiliation, religious attendance, religious belonging, religious extremism, Roman Catholics, same-sex marriage, social attitudes, Survey Data | 1 Comment

Socio-Demographic Groups and Religious Affiliation in Britain

Posted on June 3, 2015 by Ben Clements

The National Centre for Social Research (NatCen) issued an interesting online press release on the 31^st May, entitled ‘Church of England decline has accelerated in past decade’. The press release set out some of the main developments in religious affiliation captured in around three decades of the British Social Attitudes (BSA) surveys. The main developments identified were:

the proportion of those self-identifying as Anglican fell from 40% in 1983 to 17% in 2014;
the proportion reporting that they had no religious affiliation increased from 31% in 1983 to 49% in 2014;
the figures are broadly stable in terms of the proportions identifying as Catholic (1983: 10%; 2014: 8%) or belonging to the ‘other Christian’ category (1983: 17%; 2014: 17%);
the proportion identifying with a non-Christian faith increased from 2 per cent in 1983 to 8% in 2014.

The press release also provides, across time, GB adult population estimates for each religious affiliation category. In the adult population, the number of Anglicans declined from 16.5 million in 1983, to 14.3 million in 1994, 13.2 million in 2004, and totalling 8.6 million in 2014. The press release and accompanying data tables can be found here.

Following on from the changes in the overall religious landscape identified in the NatCen press release, this BRIN post analyses data on the distribution of religious affiliation across socio-demographic groups. It compares data taken from the 1983 and 2013 BSA surveys (the 2014 survey has not yet been released for wider usage through the UK Data Service). The main focus here is the proportions within these socio-demographic groups identifying as Anglican. Comparative data for other religious affiliation categories are provided in the tables below. The categories of religious affiliation used are: Anglican, Catholic, other Christian, other religion, and no religion.

Based on the BSA 1983 survey, Table 1 shows the distribution of religious affiliation within different socio-demographic groups, classified by sex, age, ethnicity and country. Data by English region, based on the 1983 survey, are shown in Table 2. Table 1 shows the proportion that identified as Anglican was higher amongst women (44%) compared to men (35%). There are clear differences based on age in 1983: 16% of those aged 18-24 identified as Anglican compared to 32% of those in the next oldest age group (25-34). The proportion that identified as Anglican amounted to a majority of those aged 65 and over, two-fifths of those aged 35-44, and nearly half of those aged 45-64 years of age. The proportions that identified as Anglican differed markedly based on ethnic group and country (45% in England compared to 30% in Wales and 2% in Scotland).

Table 1: Religious affiliation by socio-demographic group, 1983

	Anglican (%)	Catholic (%)	Other Christian (%)	No religion (%)

Male	35	10	14	39
Female	44	10	20	25
Aged 18-24	16	11	11	57
Aged 25-34	32	10	11	44
Aged 35-44	39	9	17	33
Aged 45-54	46	9	21	24
Aged 55-64	47	12	18	22
Aged 65-74	54	8	21	15
Aged 75+	55	7	27	11
White ethnic group	42	10	17	32
Other ethnic group	12	9	25	29
England	45	8	12	32
Wales	30	6	31	31
Scotland	2	22	47	29

Source: BSA 1983.

Note: Percentages sum across the rows and have been rounded.

Data for the other ‘religion category’ are not reported.

Based on the distribution of religious affiliation across regions in England, shown in Table 2 for the 1983 survey, identification as Anglican was highest in the North East, West Midlands and the South East. It was lowest in East Anglia / Eastern England.

Table 2: Religious affiliation by English region, 1983

Region	Anglican (%)	Catholic (%)	Other Christian (%)	No religion (%)
North East	50	15	12	17
North West	43	16	13	28
Yorkshire and Humberside	41	7	8	45
West Midlands	52	5	12	29
East Midlands	42	6	7	40
East Anglia / Eastern England	32	8	17	42
South West	45	3	17	35
South East	52	5	13	29
Greater London (inner and outer)	42	12	14	27

Source: BSA 1983.

Note: Percentages sum across the rows and have been rounded. Data for the other ‘religion category’ are not reported.

Turning to more contemporary data in Table 3, taken from the BSA 2013 survey, it is apparent that the proportions identifying as Anglican have fallen across all socio-demographic groups. The marked age differential is still present. While just 3% and 4% of those in the 18-24 and 25-34 age groups, respectively, identify as Anglican, this steadily increases to more a third of those aged 65-74 (36%) and 75 and over (35%). Women (19%) are again more likely than men (13%) to report an Anglican affiliation, as are those from a white ethnic group and those living in England (18%, compared to 13% in Wales and 1% in Scotland).

Table 3: Religious affiliation by socio-demographic group, 2013

	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)	No religion (%)
Male	13	8	14	8	57
Female	19	10	19	8	44
Aged 18-24	3	6	11	10	69
Aged 25-34	4	8	17	12	60
Aged 35-44	9	11	14	10	56
Aged 45-54	16	10	15	6	54
Aged 55-64	21	10	16	8	46
Aged 65-74	36	8	19	4	34
Aged 75+	35	9	29	3	24
White ethnic group	18	9	16	2	55
Other ethnic group	4	9	19	51	18
England	18	9	16	9	49
Wales	13	4	22	4	56
Scotland	1	13	23	1	61

Source: BSA 2013.

Note: Percentages sum across the rows and have been rounded.

Finally, Table 4 shows the distribution of religious affiliation across English regions in the 2013 survey. Anglican identification amounts to between a fifth and a quarter in the West and East Midlands, and in the South West and South East; being lowest in Greater London, at 7% (which, across regions, has the highest proportion identifying with a non-Christian religion).

Table 4: Religious affiliation by English region, 2013

Region	Anglican (%)	Catholic (%)	Other Christian (%)	Other religion (%)	No religion (%)
North East	15	8	13	2	63
North West	19	12	14	4	42
Yorkshire and Humberside	17	8	14	7	55
West Midlands	23	8	17	8	45
East Midlands	23	6	18	3	50
East Anglia / Eastern England	18	8	14	7	54
South West	22	6	16	2	55
South East	21	8	15	5	49
Greater London (inner and outer)	7	11	18	21	43

Source: BSA 2013.

Note: Percentages sum across the rows and have been rounded.

Delving into the detail of religious affiliation across socio-demographic groups in Britain shows consistency over time in terms of within which groups Anglican affiliation has been and is more common. The variation by age is particularly marked. Anglican affiliation is most common amongst those aged 65-74 and 75 and over. In the former group, however, it is nearly matched by the proportion who say they do not have any religious affiliation. Amongst the 18-24, 25-34 and 35-44 age groups, less than a tenth identify as Anglican. In each case, a clear majority has no religious affiliation (highest at around seven in ten of those aged 18-24).

Posted in Research note, Survey news | Tagged Anglicans, Belonging, British Social Attitudes Surveys, NatCen Social Research, Religious affilation, Religious change, Survey Data | 1 Comment

The 2015 General Election: Religious Affiliation and Party Vote Share Across Constituencies

Posted on May 21, 2015 by Ben Clements

As the weekend round-up of religious news on BRIN flagged up, the British Election Study (BES) 2015 has released the first version of the 2015 general election results dataset. This dataset (and the accompanying documentation) can be obtained here. Across parliamentary constituencies, the dataset includes the vote share for each party at the 2015 general election. It also includes religious affiliation data from the most recent English and Welsh census and and Scottish census (2011). The religious affiliation data are available in separate variables measuring the proportion in each of the following categories: Christian, Buddhist, Hindu, Jewish, Muslim, Sikh, other religion, no religion, not stated. Using this dataset we can perform some basic analysis of this aggregate level data – that is, across constituencies – to look at the association between religious affiliation and party vote share at the general election.

The results are reported in the table below. This reports the correlation coefficients for the bivariate associations between three of the religious affiliation indicators (% Christian, % Muslim, % No religion) and four party vote share categories (% Con, % Lab, % Lib Dem and % UKIP). These coefficients indicate the direction and strength of the association between two variables. They can range in value from -1 to +1. A positive value indicates that as one variable increases in value, the other variable also increases in value. A negative value indicates that as the value of one variable goes up, the value of the other variable goes down. The larger the (positive or negative) value, the stronger is the association between the two variables.

When undertaking and reporting this sort of analysis, two points are particularly important to bear in mind. First, as Denver et al (2012), pp. 36-37) make clear, correlation coefficients cannot tell us whether variation in one variable (here, level of religious affiliation) causes the variation in another variable (here, party vote share). They can only show the extent to which two variables are associated – that is, whether they increase or decrease in value together (they are positively correlated); or whether as one increases in value, the other decreases in value (they are negatively correlated)). Secondly, as Denver et al (2012) also caution, given that we are looking at aggregate-level data (based on information pertaining to the constituency-level) we cannot conclude from these data that the same association is present amongst individuals within constituencies (to presume this to be this case would be an ‘ecological fallacy’).

Religious affiliation and party vote share at the 2015 general election: Bivariate correlations

	Party vote share
Religious affiliation	% Conservative	% Labour	% Lib Dem	% UKIP
% Christian	.33	-.27	-.03 (n/s)	.43
% Muslim	-.29	.50	-.10	-.23
% No religion	-.16	-.29	.16	-.14
Number of constituencies	632	631	631	614

Source: British Election Study 2015 Constituency Results.

Note: ‘(n/s)’ indicates a coefficient that is not statistically significant. All other coefficients are statistically significant.

All of the correlation coefficients are statistically significant, with the exception of the association between Christian affiliation and Lib Dem vote share, but their magnitudes clear vary. The correlation coefficients for Christian affiliation indicate that it is positively-associated with Conservative and UKIP vote share and negatively-associated with Labour vote share. The correlation coefficients for Muslim affiliation show negative associations with the vote shares of Conservative, Lib Dem and UKIP, but a positive association with Labour vote share. For no religion, at the constituency level there are negative associations with the vote shares of Conservative, Labour and UKIP and a positive association with Lib Dem vote share. As the magnitudes of the coefficients show, the strongest association is between Muslim affiliation and Labour vote share, and – interestingly – followed by that between Christian affiliation and UKIP vote share (and thus higher than that obtained for Christian affiliation and Conservative vote share).

Finally, it should be noted that variation in other constituency-level indicators – such as socio-economic circumstances or ethnic group composition – is often associated with variation in party vote share.

An analysis of religious affiliation and vote choice at the general election at the individual-level will be posted when suitable BES post-election data become available.

References

British Election Study 2015 Constituency Results.

Denver, D., Carman, C. and Johns, R (2012), Elections and Voters in Britain. 3^rd edition. Basingstoke: Palgrave, pp. 36-37.

Posted in Religion and Politics, Religion in public debate, Research note, Survey news | Tagged British Election Study 2015, General Election 2015, Party Choice, Religion and Politics, Religious Affiliation, religious belonging, Survey Data | Leave a comment

How religious groups voted in the 2016 referendum on Britain’s EU membership

Re-examining religious and paranormal beliefs in mid-1970s Britain

Religion and party choice: Evidence from the BES 2015 face-to-face post-election survey

The 2015 General Election: Religious Affiliation and Party Vote Share Across Constituencies

Recent News

Recent Sources

Archives

Recent Comments

British Religion in Numbers is funded as a British Academy Research Project