Religious change and secularisation in Scotland: An analysis of affiliation and attendance

Building on Brown (1997) and Field's (2001) research into religious decline and secularisation in Scotland in the later decades of the 20th-century, this article uses data from recurrent social sur...


Introduction
Scotland has traditionally been seen a particularly religious nation within Great Britain, partly due to its Presbyterian traditions.However, as has been recently observed, both institutionally and across the wider population, Scotland has become a more secular country (Bruce, 2014a(Bruce, , 2014b)).Building on Brown (1997) and Field's (2001) research into religious decline and secularisation in Scotland in the later decades of the 20 th -century, this analysis uses data from social surveys, nationally-representative of the Scottish adult population, to analyse affiliation and attendance in the 21 st -century.Specifically, it examines recent trends in religious affiliation and attendance in Scotland, compares key religious indicators in England and Scotland to assess areas of similarity and difference in religion and secularity, and assesses the contemporary socio-demographic basis of religious engagement in Scotland.
There are two key antecedent studies of religion in Scotland which have informed the research aims underpinning and empirical analyses undertaken within this article.Their significance needs to be acknowledged at the outset.First, Callum Brown's research on religion and society in Scotland, particularly his analysis of religious change between 1939-1997 (1997).This period was broadly characterised as 'the haemorrhage of faith' and a period of 'religious crisis' in Scotland, with the recognition that this was part of a broader geographical phenomenon of secularisation and religious decline.Brown asserted that 'religion in the late twentieth century is losing its place in Scottish society' and further claimed that 'The statistics provide a stark guide to Scottish secularisation ' (1997: 158).
Secondly, Clive Field's study (2001) made extensive use of opinion polls and social surveys to examine religious change in Scotland from the 1970s to the end of the 20 thcentury, in order to quantitatively validate Brown's 'gloomy assessment of contemporary Scottish religion ' (2001: 158).Field's wide-ranging analysis took in church attendance, religiosity, denominational affiliation, private religious practices, religious beliefs, as well as attitudes towards different religious issues.Summarising the wide array of quantitative evidence reviewed in relation to religious belonging, behaving and believing, Field concluded that: Viewed chronologically, there can be little doubt that a degree of secularisation has occurred within the past quarter-century and especially during the 1990s.
Appreciable decline has been registered in the proportion of Scottish adults describing themselves as religious; professing affiliation to a Christian denomination (especially to Protestant communions, Catholic numbers having held up reasonably well); attending church at least once a month (with half the population now habitual absentees from worship); and believing in God and an afterlife (2001: 168).
More broadly, Field arrived at two other noteworthy conclusions.Firstly, in a wider geographical context, he observed that: Viewed comparatively, Scotland no longer appears to be notably more religious than the rest of Great Britain … Levels of self-assessed religiosity, belief in god and an afterlife, and estimation of the influence and relevance of the Churches are much the same in Scotland as in Britain, with the extent of denominational affiliation, regular church attendance, and private prayer only a little higher (2001: 168).
Secondly, in a longer-term perspective and referring to Brown's thesis, he observed that: On the whole, the Scottish opinion poll evidence does tend to confirm Callum Brown's thesis of the 'the haemorrhage of faith'.Scotland is undoubtedly in the throes of secularisation and is shedding its historical reputation as a peculiarly religious nation.The Presbyterian Churches are in decline and the Presbyterian ethos which permeated so many aspects of Scottish social life is fast disappearing.Roman Catholicism is no longer the spiritual bedrock that it once was and, in any case, cannot shake of its image as an 'alien' religion (2001: 168).
While having a narrower focus in terms of the aspects of religion and society covered when compared to the wider scope of Brown and Field's research, this article builds on and extends their research in three respects.Firstly, by examining the over-time evidence bearing upon levels of religious affiliation and rates of attendance in 21 st -century Scotland.

Secondly, by providing a comparison of trends in these religious indicators in Scotland and
England in recent decades.Thirdly, by examining the contemporary socio-demographic basis of affiliation and attendance in Scotland.And as Field's analyses principally made use of data on religion from commercial opinion polls -'a source of which Brown himself has made precious little use ', as Field noted (2001: 159) -so this article makes detailed use of religion data from the Scottish Social Attitudes series and its longer-running British counterpart.

Data and method
The Scottish Social Attitudes surveys (SSA) are the principal source of data for the analysis.
The SSA surveys have been conducted every year since 1999 (with the exception of 2008), with samples of between 1,200 to 1,500 people drawn using probability sampling (based on a stratified, clustered sample).The analysis uses all of the currently available datasets (up to and including 2014).The 1992 and 1997 Scottish Election Studies (SES) are also used to furnish earlier data on religious attendance and affiliation in Scotland.The 1992 and 1997 SES feature the same question wordings on affiliation and attendance used in the SSA surveys. 1 SSA surveys provide an example of a social survey series featuring recurrent data on religion, which constitute a 'tremendous potential research resource for scholars of religious studies, modern history, and social science', allowing for the 'repurposing' of religious data (Field, forthcoming).The SES and SSA surveys are the best survey sources for the purposes of this article because of the time period they cover in combination, because they have asked about both affiliation (current and of upbringing) and attendance in every survey, and because of the consistency in the questions used.
In order to enable the careful analysis of recent religious trends in Scotland and England, one of the main aims of this article, the longer-running British Social Attitudes (BSA) series is used, in order to obtain equivalent data on religious affiliation and attendance for the adult population in England.The long-running annual BSA surveys began in 1983 (with the exceptions of 1988 and 1992) and so there is counterpart survey for each SSA survey used.Moreover, the British Election Study from 1992 -of which the SES was a key component -is used in order to provide earlier equivalent data on religious affiliation and attendance for the adult population in England.The comparison of religious data for England and Scotland is all the more robust and rigorous given that data from counterpart survey series are being used, with identical survey years, and holding constant the data collection methodologies and question wordings.
The SSA surveys have not featured recurrent questions asking about other forms of (private) religious practices or on religious beliefs, so these areas cannot be considered here (for analysis of the specialist modules on religion which featured in the 2001 and 2014, surveys, see Bruce and Glendinning, 2002, 2003, 2006;Rosie, 2002;Hinchliffe et al., 2015).
The SSA surveys have featured questions on issues which are salient for religious identities (such as education or sectarianism), but full consideration of these attitudes cannot be given here.The time period covered in the analysis overlaps with the final decade covered by Field (2001) -the 1990s -but importantly brings the analysis of affiliation and attendance in Scotland up to date.The analysis undertaken here makes an important contribution to the wider literature studying religious change and secularisation in Scotland (Brown, 1997;Field, 2001;Bruce and Glendinning, 2002, 2003, 2006;Bromley and Curtice, 2003;Bruce et al., 2004Bruce et al., , 2005;;Rosie, 2002Rosie, , 2004;;Bourque et al., 2005;Glendinning, 2006;Voas, 2006;Bruce, 2014aBruce, , 2014b)).
Throughout, the empirical analyses are based on weighted data and percentages have been rounded for presentation in the figures and tables.The empirical analysis is divided into four stages.Firstly, it examines the main trends in overall levels of religious affiliation.Secondly, it examines the overall levels of religious attendance.Thirdly, it compares these key religious indicators for Scotland and England, in order to ascertain areas of similarity and difference.Fourthly, it examines the contemporary social basis of religious affiliation and attendance in Scotland.A concluding section then reviews the main empirical findings and highlights areas for further research.

Religious affiliation
This first section looks at the data on religious affiliation (or belonging) in Scotland.The data on religious affiliation from the SES and SSA surveys are shown in Figure 1, covering the period 1992-2014.The following categories are used: Church of Scotland; Catholic; other Christian; non-Christian; no religion.The data are based on responses to the following question: 'Do you regard yourself as belonging to any particular religion?'This form of question wording, as Field has noted (2001: 160), 'puts least pressure on respondents to identify with a religion, by making no assumption that they have one and by introducing a formal concept of "belonging"', and thus tends to produce higher proportions with no affiliation.
(Figure 1 about here) In general terms, it can be seen that the proportion identifying as Church of Scotland has clearly declined over this period, from 46% in 1992 to 22% in 2014 (itself a slight increase on the figures for 2012 and 2013).Across the full range of the SSA surveys, the decline is from 34% in 1999 to 22% in 2014 (18% in 2013).The proportion identifying as Catholic has remained fairly stable in recent years -something noted by Field (2001: 161) in his analysis of earlier decades -standing at 16% in 1992 and 14% in 2014.This may be partly rooted in the stronger 'tribal' feelings of belonging within the Catholic community (Hornsby-Smith, 1987).
Of course, when assessing stability and change in data based on affiliation, it should be born in mind that many of those belonging to a particular Christian denomination may have a 'nominal' attachment (Bruce, 2014b) and may not be religiously-engaged in some other significant way -such as attending church services connected with their faith (see the following section for further discussion of attendance).Many Catholic children are still educated in denomination-specific schools.There are 369 state-funded Roman Catholic schools in Scotland, with a majority of them 'located in the west central belt of Scotland' (McKinney and Conroy, 2015: 107).It should be noted that some parts of Scotland do not have any Roman Catholic schools at all and that, particularly in the East of Scotland, such schools would have a significant proportion of children not from a Catholic background.This tradition of denominational schooling may well have contributed to the preservation of a sense of Catholic identity -evidenced in the relative stability in levels of affiliation across time -if only in a nominal sense for some.
The proportion with no religious affiliation in the SES and SSA has increased markedly over the period covered -standing at around a quarter in 1992 and a third in 1997, rising to around half in most of the recent SSA surveys (although there was a sharp drop-off between 2013 and 2014). 2 In every SSA survey since 1999, those with no religious identification have outnumbered those identifying as Church of Scotland -in more recent surveys by a margin of two-to-one or more.The proportion belonging to other, non-Christian, faiths has increased in the more recent SSA surveys, standing at 5% in 2014.The proportion affiliated with some other Christian tradition (in particular, as Anglican -Episcopal Church of Scotland) has, like Catholic identification, been reasonably stable over recent decades.The stability of Catholic affiliation over the decades was no doubt partly due to substantial in-migration from historically Catholic countries in Eastern and Central Europe, who joined the European Union in the 2004 round of enlargement.The Scottish census data also registered a sizeable increase in the proportion with no religion, increasing from 28% to 37% (National Records of Scotland, 2013: 32).In 2011, the largest share was accounted for by those with no affiliation; in 2001 it had been those identifying with the Church of Scotland (National Records of Scotland, 2013: 32).
(Figure 2 about here) Another way of charting the decline of religious identification is to look at data pertaining to being brought up within a particular religion.A question gauging this has featured on the SES and SSA surveys: 'In what religion, if any, were you brought up?' Figure 2 charts the proportions saying that they were brought up within a particular religion, or none, using the same set of affiliation categories as in Figure 1.
Particularly noteworthy is the steady decline in the proportion raised within a religious background pertaining to Church of Scotland.This has fallen from about three fifths in 1992 to around a half in 1997 and the earlier SSA surveys, and then declining further, registering at around two-fifths or lower in the most recent surveys.The proportion growing up within Catholicism has broadly remained stable over time (usually amounting to somewhat under a fifth), as was also the case with the indicator of current affiliation discussed already.
It is worth noting that recent research has shown that, for Britain as a whole, the Catholic Church's 'retention rate is the strongest of the main British [Christian] denominations', so it is relatively more effective at transmitting its identity (Bullivant, 2016: 185).Within Scotland, current evidence shows that the relative retention rate is similarly stronger amongst Catholics.That is, when current religious affiliation is examined based on the religion of upbringing, the SSA 2014 survey shows that 68% of those brought up as Catholics have retained that identity, compared to 55% of those raised as Church of Scotland and 51% of those raised within as other Christian.The overwhelming majority of those with no religious upbringing reported no current affiliation (83%).Looking back further at the equivalent data in the 1992 SES, a similar picture of relatively stronger retention amongst Catholics is evident (at 85% compared to 72% of Church of Scotland and 54% for other Christian).It is also worth noting that, based on comparing the 1992 and 2014 data, the effective retention rate of both Catholicism and the Church of Scotland has declined over time.
The proportion without a religious upbringing has approximately trebled over time (at 7% in 1992, and averaging 21% in the 2011-2014 surveys).The proportion brought up within another Christian tradition has increased to some extent over time, as has the proportion raised within a non-Christian faith (albeit still constituting a very small proportion overall).

Attendance is the most important indicator of communal religious practice (or behaviour).
There are widely-acknowledged caveats in the wider literature about using surveys to measure self-reported attendance at religious services (Chaves and Stephens, 2003), as they can be susceptible to individuals providing what they perceive to be a 'socially-respectable' response and thus overstating their attendance (Field, 2001: 159).Recent analysis has concluded, however, that over-reporting of church attendance in social surveys is more likely to be found in North America than it is in European countries, including Britain (Brenner, 2013).Population estimates and projections for church attendance in Scotland show that in 1980 17% of the population attended, falling to 15% in 1990 and declining to 11% in 2000.It fell further to 10% in 2010 and was projected to decrease to 9% in 2015 The over-time data from the SES and SSA surveys are shown in Figure 3, again covering the period from 1992 to 2014.The more detailed response options in the SES and SSA surveys have been combined into the following categories: frequent attenders (once a month or more often); infrequent attenders (less often than once a month); and never attends.The question was worded as follows: 'Apart from such special occasions as weddings, funerals and baptisms and so on, how often nowadays do you attend services or meetings connected with your religion?'.
As already noted above, Field's analysis showed that 'generally not far short of half the people are habitual absentees from religious services ' (2001: 162).Taking the analysis forward to cover the most recent period, we can see that the proportion that can be classed as 'habitual absentees' has further increased, in the region of three-fifths to two-thirds in the most recent SSA surveys.In the 1992 SES survey, in contrast, this group constituted around half.The proportion of frequent attenders (monthly or more often) has fallen from around a quarter (in the 1992 and 1997 SES surveys) to about a fifth in recent years.
(Figure 3 about here) The proportion attending on a less frequent basis (less than once a month) has also generally declined over time.While around a fifth attended infrequently in 1992 and 1997, the more recent surveys tend to show that around one in six or seven attend infrequently.
In 1992 and 1997 the proportions attending at all (whether frequently or infrequently) and never were about evenly balanced.In the most recent SSA surveys (again, 2014 shows a slightly different set of figures), the evidence suggests that, for each person saying they attend services, around two report that they do not attend (that is, beyond going for services connected to the traditional rites of passage -births, marriages and deaths).

Religion in Scotland and England: Same or different?
As mentioned at the outset, Field (2001) noted the decline in Scotland's religious distinctiveness across some common indicators over recent decades.A previous comparison of several aspects of religious engagement in Scotland and England noted that This descriptive evidence suggests that 'religion' appears to be more important in Scotland than in England.The rates of religious group membership, religious group activity and church attendance are higher and religious attitudes appear to be more favourable towards religious beliefs (Bourque et al., 2005: 265).
Due to survey limitations, this section can only focus on the religious indicators of affiliation and attendance for the two nations.A careful and robust comparison can be made over the same time period (1992 to 2014) by using data for Scotland from the SES and SSA surveys and for England from the British Social Attitudes (BSA) surveys (and one British Election Study (BES) survey from 1992, when no BSA survey was undertaken).The BSA / BES and SSA / SES surveys have used the same question wordings for the religious indicators.(Figure 4 about here) Figure 4 shows the proportions identifying with the Church of Scotland and Church of England.In general, some of the earlier surveys excepted, identification has featured at similar levels in Scotland and England, and has also followed a steady decline over time.In 1992, 35% of those in Scotland professed to be Church of Scotland, compared to 30% of those in England who said they were Church of England / Anglican.In 2014, the figures were, respectively, 22% in Scotland and 20% in England.If, historically, identification with the Church of Scotland tended to be higher than the equivalent for the Church of England, then such differences were much less evident in recent years.The proportions identifying as some other Christian identity have been traditionally higher in England, but not by particularly large margins.Of course this designation includes a rather different set of groups in both nations, but both show a perceptible increase over time.In 1999, those identifying as some other Christian amounted to 10% in Scotland and 13% in England; in 2014 the respective proportions were 15% and 16%.Recently, as historically, the other Christian grouping registers a lower level of identification than that for Church of Scotland.
In terms of the split between denominational and non-denominational affiliation amongst other Christians in Scotland, in 1992 7% identified with some other Christian denomination and 5% did not identify with any denomination; in 2014 these figures were 4% and 11%.
It is clear is that identification as Catholic has been higher in Scotland than in England, but also evident is the relative stability in levels of identification in both countries.
In 1999, 8% identified as Catholic in England; in Scotland it was 14%.These country proportions were exactly the same in 2014.Not shown here but also a recurrent feature in the data from the comparable surveys is the higher proportion in England identifying with a non-Christian faith (in 2014, this was 5% in Scotland and 9% in England).
Figure 5 reports the country data for three indicators of secularity (no affiliation, no religious upbringing, does not attend religious services).If Scotland was still, in some way and to some extent, a more religious nation than England, then we might reasonably expect to see the former lagging the latter on indicators such as these.On the whole, the trend data tend to show a very similar picture for both nations.Non-affiliation, non-attendance and being raised without a religious tradition have run at similar levels in both countries in recent years.In 1992, 25% and 31% of those in Scotland and England, respectively, declared they had no religious affiliation; in 2014 the equivalent figures were 44% and 47%.In the years immediately prior to 2014, non-affiliation was actually higher in Scotland (the drop in the level of non-affiliation in the 2014 SSA was mentioned earlier).In 1992, 7% and 8%, respectively, of adults in Scotland and England said they had been raised without a religious upbringing, increasing to 19% and 18% in 2014.The early surveys do show non-attendance at religious services to be rather higher in England than in Scotland but this difference is not really evident in more recent years.In 2014, 62% in adults in Scotland and 64% of adults in England in 2014 said they did not attend services (beyond going for special occasions).
(Figure 5 about here) Taken together, the evidence examined in Figures 4-5 underlines the broadly similar levels of indicators of secularity in Scotland and England in terms of non-affiliation and nonattendance, but also indicates some of the historically-rooted variation in the two nations' religious fabric.The latter includes the higher level of Catholic identification in Scotland and the larger proportion affiliated with non-Christian religions in England.

The sociodemographic basis of affiliation and attendance
This final section looks at the contemporary basis of religious affiliation and attendance amongst sociodemographic groups in Scotland.Sociodemographic profiles are examined for both affiliation and attendance, and are informed by Field's (2001: 161-162) findings in relation to those groups in Scottish society who were more likely to profess to having no religion, more likely to be regular churchgoers and who never or hardly ever attended church.The demographic variables examined are: • Sex: male or female.
• Educational attainment: operationalised as four categories measuring the highest qualification held (Degree / Higher Education, Highers / A-Levels, Standard Grade / GCSEs; None).
• Socio-economic classification: five categories measuring type of occupation (Employers, managers and professionals, Intermediate occupations, Small employers and own account workers, Lower supervisory and technical occupations, Semiroutine and routine occupations).
The sociodemographic basis of religious affiliation is examined first, followed by that for attendance at religious services.

Affiliation
Based on analysis of the SSA 2014 survey, the distribution of the five affiliation categories within each sociodemographic group is shown in Table 1.There is some evidence of a gap in religious affiliation in Scotland between men and women.Women are somewhat more likely than men to affiliate as Church of Scotland (24% compared to 19%) and as other Christian (17% and 14%).There is little difference in the proportion identifying as Catholic.
Men are more likely to be 'religious nones': 49% have no affiliation compared to 40% of women.
The variation in levels of affiliation based on age groups is considerable.Half of those aged 75 and over and over two-fifths of those aged 65-74 identify as Church of Scotland.In stark contrast, around one in twenty of those aged 18-24 and 25-34 report the same affiliation.In other words, identification as Church of Scotland is around ten times as likely amongst those aged 75 and over as it as amongst those aged 18-24.The age-related variation is much less marked for identification as Catholic or as some other Christian (highest amongst those aged 45-54).Identification with a non-Christian religion is most pronounced amongst those aged 25-34.
(Table 1 about here) The age-related patterning for non-affiliation is somewhat the reverse of that seen for identification with the Church of Scotland.That is, the level of non-affiliation decreases with each consecutive age group, with pronounced differences between the younger and older age groups.Nearly seven-in-ten and three-fifths, respectively, of those aged 18-24 and 25-34 profess no affiliation.Amongst those aged 35-44, it stands at half.In marked contrast, just 16% of those aged 75 and older and 25% of those aged 65-74 report having no affiliation, rising to around three-fifths of those aged 55-64 and 45-54.In the two youngest age groups (aged 18-24 and 25 to 34), a clear majority does not identify with any religion; whilst those aged 35-44 are equally split between religion and non-religion in terms of affiliation.Amongst those aged 18-24, it is less than a quarter; and for those aged 25-34, it is less than four-in-ten.Another way of expressing this marked variation is to note that while only around a third of those aged 18-24 and two-fifths of those aged 25-34 report having some form of religious affiliation, three-quarters of those aged 65-74 affiliate with a religion, as do over four-fifths of those aged 75 and older.
Based on the measure of educational attainment, the affiliation profile of those with no formal qualifications stands out.They are much more likely to identify as Church of Scotland than those with some formal qualification, and are also most likely to affiliate as Catholic (but not as other Christians).However, this higher overall level of affiliation is no doubt partly the result of underlying age-related differences: those with no qualifications being more likely to be older, and -as has been documented already -older age groups are themselves much more likely to claim some form of affiliation.Nearly three-in-ten of this group have no affiliation; which amounts to near or a bare majority in the other three groups.Based on socio-economic classification relating to occupation, there is less pronounced or consistent variation.As Field found for earlier decades (2001: 161), the 'religious nones' are more likely to be male, younger and to have some formal qualifications; but there is no clear differenced based on occupation.The most pronounced differences occur on the basis of age group.
Bivariate tests between affiliation and each of the demographic variables discussed above showed that there were statistically significant differences on the basis of sex, age group and education qualification.There was no statistical significant difference based on affiliation and socio-economic classification (Cramer's V=.079;p=.125).The strength of the association was moderate between age group and affiliation (Cramer's V=.223,p=.000),but weak between sex and affiliation (Cramer's V=.108, p=.002) and education and affiliation (Cramer's V=.120, p=.000).

Attendance
Table 2 repeats the sociodemographic profile for religious attendance (also including religious affiliation) using the categories of frequent attendance (once a month or more often), infrequent attendance (less often than once a month) and does not attend.Field found that, inter alia, regular churchgoers were more common amongst women, those aged 55 and over, and within the higher social classes (2001: 162).Based on the data from the SSA 2014 survey, there is a gap in attendance at religious services between men and women.Women are more likely to attend on a frequent or infrequent basis (amounting to 42% compared to 32% of men).Even so, 58% women say they never attend religious services (beyond attending for special occasions -the traditional rites of passage), which rises to 67% of men.The evidence for Scotland relating to sex and religious attendance (and also for affiliation, as discussed above) is therefore in accord with the well-established association between women and their greater likelihood of religious identification and involvement (Trzebiatowska and Bruce, 2012).
(Table 2 about here) The age gap is also present in relation to attendance, though the differences are not as marked as found for levels of (non)affiliation.Those aged 75 and over stand out: twofifths say they attend services on a regular basis.Regular attendance is clearly lower amongst all other age groups (in the region of 13-25%).There is less variation across age groups when it comes to infrequent attendance at services, though it is slightly higher in the older age groups.In each age group, with the clear exception of those aged 75 and older (39%), around half or even higher proportions are non-attenders -or 'habitual absentees' (Field, 2001).The variation is less clear or consistent for the measures of education and socio-economic classification.Those with no qualifications are least likely to never attend religious services.
Regular attendance at worship services can be a more important act of commitment for particular traditions, and this is clearly the case for Roman Catholicism, where canon law has obliged ordinary adherents to attend weekly Sunday mass (Horword, 2006: 13).This is reflected in the data to some extent, with Catholics (43%) much more likely to say they attend regularly than do those affiliated as Church of Scotland (23%) or other Christian (32%).Around nearly half of non-Christians report regular attendance (52%).Of all the groups, Catholics and non-Christians are least likely to say that they never attend (27% and 25%, respectively), followed by non-Christians (44%) and those identifying as Church of Scotland (48%).There is less variation in levels of infrequent attendance.Field found that 'a consistent half of Catholics reported that they went to mass on a weekly basis' in polls taken between 1976-1995 (2001: 162).In the 2014 SSA survey, just 28% of Catholics said they attended on a weekly basis -in comparison, in 1992 51% of Catholics reported attending weekly, and 41% said this in 1997.In 2014, 21% of other Christians and 10% of those identifying as Church of Scotland said they attended weekly.Amongst those belonging to other religions, weekly attendance was recorded at 35%.As might be expected, the vast majority of those with no religion said they did not attend services (91%), albeit a very small proportion reports attending infrequently (8%).
Bivariate tests for attendance and the socio-demographic variables showed statistically significant differences based on sex, age and affiliation; but not for education or socio-economic classification.The strength of association was weak for sex (Cramer's V=.100 p=.001) and age (Cramer's V=.146, p=.000) but was notably strong for affiliation (Cramer's V=.409, p=.000).

Sex and age group: Affiliation and attendance
The clear-cut findings between age groups and -to a less emphatic extent -between men and women are instructive given the recent observation that elderly women are one of the 'primary carriers of religion' in Scottish society (Bruce, 2014a: 203-204).The next step is to examine further how affiliation and attendance vary when men and women are classified by age group.This should provide further empirical insight into the claim that elderly women are one of the key 'carriers of religion'.
Figures 6 and 7 show, for men and women, the proportions within each age group with no religious affiliation and that never attend religious services.In terms of religious affiliation (Figure 6), the largest gaps when comparing age groups occur amongst those aged 55-64 and 65 to 74, followed by the 25-34 group.Men are consistently more likely to eschew an affiliation.Across the other age groups, there is a similar level of non-affiliation or the differences are of a lower magnitude; though, where there is a differential, men are always more likely to have no affiliation.In relation to professing a religious identity, older women -particularly those aged between 55-74 -are clearly more likely to be 'carriers of religion' in Scottish society than are older men.Amongst the youngest age groups, overall levels of non-affiliation are much higher for both men and women, but younger men seem to be somewhat more inclined to eschew some form of religious affiliation.

(Figures 6 and 7 about here)
Non-attendance at religious services (Figure 7) again shows some marked variation when comparing men and women by age group.In most of the older age groups, men are more likely to be 'habitual absentees' (Field 2001) than are women (with the exception of a similar level for those aged 75 and older).Amongst the other age groups (44 years and under), though, the levels of non-attendance are broadly similar for men and women.
Taking this analysis further, there is also some variation in the frequency of attendance amongst men and women when classified by age group.While both men and women aged 75 and over are most likely to say they attend monthly or more often, women are generally more likely to report monthly attendance than are men, with the exception of those aged 35-44.
The data presented in Figures 6 and 7 tend to support the assertion about the role of older women as 'carriers of religion' (Bruce, 2014a) relative to older men; not only are they generally more likely to be affiliated to a religious faith and to attend at all, they also attend on a more regular basis.These differences do not apply to those aged 75 and over, at least based on the survey data analysed in this section, so both men and women in this age group may be seen to be 'carriers of religion' in Scottish society.

Age group and religion of upbringing
The previous analysis of religious affiliation and demographic factors demonstrated a considerable amount of variance in current affiliation across age groups in two particular respects: the proportion identifying as Church of Scotland (heavily skewed towards the older age groups) and the proportion with no affiliation (heavily skewed towards the younger age groups).Are similar age-related differences also evident when over time data on religion of upbringing are examined?For each of the seven age groups, Table 3 reports the proportion saying they were brought up within each of the four religious groups and those without a religious upbringing.
(Table 3 about here) As with current affiliation, religion of upbringing also demonstrates very large agerelated differences, and once again these are of greatest magnitude -by far -for the Church of Scotland and no religion categories.An upbringing within the Church of Scotland category is much more prevalent amongst older age groups, amounting to around threefifths of those aged 65 and over.This declines to somewhat over two-fifths of those aged 45-54 and 55-64, and further to around three-in-ten of those aged 35-44.There is a further significant drop-off within the two youngest age groups, where 15% of 18-24 year olds and 13% of those aged 25-34 said they had such an upbringing.There is no clear pattern across the age groups in terms of the proportion reporting they were brought up within Catholicism or within some other Christian tradition.Albeit consisting of very small proportions, being raised within a non-Christian background is highest amongst those aged 25-34.In terms of having had no religious upbringing, while fewer than 5% of those in the two oldest age groups said they were raised outside of a religious tradition, this rises to around a tenth or more of those aged 45-54 and 55-64.Around a quarter of 35-44 year olds said they were not raised within a religious tradition, increasing to nearly three-in-ten of those aged 25-34 and nearly two-fifths of the youngest age group.

Conclusion
The secularisation of Scottish society has continued and strengthened in the early years of the 21 st -century, at least as far as affiliation and attendance are concerned.Overall, religious belonging and behaviour have further attenuated in Scotland in recent years, with a growing proportion of adults reporting that they were raised outside of any religious tradition.The 'haemorrhage of faith' in Scottish society identified by Brown in the second half of the 20 thcentury (1997) and further empirically validated by Field (2001) has clearly continued into the early part of the 21 st -century.
In relation to affiliation, the decline in identification as Church of Scotland has clearly continued since the turn of the millennium -more than halving in the period covered herewhile the proportion of Catholics and other Christians has remained broadly the same.The other notable feature has been the rise of the 'religious nones', whose overtaking of the Church of Scotland as the modal category amongst Scottish adults was evident in the early SSA surveys and has persisted since.In relation to attendance, the evidence shows that, to use Field's term (2001), around three-fifths to two-thirds can now be classified as 'habitual absentees' in terms of not going to services (beyond attending for special occasions).The proportion attending at all (whether regularly or irregularly) has declined from around half to about one-third to (somewhat less than) two-fifths.Given the well-established caveats about religious attendance measured through surveys noted earlier, the proportion of 'habitual absentees' is no doubt higher.
The trends documented here, showing the continued decline of both affiliation and attendance, provide some empirical confirmation for the latter part of Bruce's assertion that 'Scotland is now an institutionally secular society with an overwhelmingly secular population' (2014a: 205).Religiosity, then, is increasingly less widely dispersed across society in Scotland.At any rate, recent survey evidence shows that around a half of the Scottish population profess no religious affiliation and even more -at least three-fifthsnever attend religious services.The evidence also suggests that Scotland and England also seem to be relatively similar in terms of key indicators of growing secularisation and the declining social and cultural relevance of religion -that is, in terms of the increasing proportions in recent years in both nations who claim they do not have a religious affiliation, who say they were raised in a non-religious environment, and in terms of those who habitually do not attend religious services.The core religious data emerging from the early years of the 21 st -century therefore further erode Scotland's 'historical reputation as a particularly religious nation' (Field, 2001: 168).Scotland and England, on the basis of levels of affiliation and attendance, are certainly not particularly religious countries and have some secularising features in common.Scotland and England do not simply resemble each other, however: the trends documented in these two countries of course fit within a wider pattern of ongoing secularisation within many western European countries (Norris and Inglehart, 2011;Pollack, 2008;Halman and Draulans, 2006).
Finally, the contemporary socio-demographic profiles both reaffirmed in the Scottish context the long-established associations between women, older age and religiosity, and also probed and found clear confirmation for, Bruce's identification of older as one the primary 'carriers' of religion in Scotland (2014a).Given the lower levels of religious affiliation and attendance recorded amongst those in the youngest age groups -amongst both men and women -and in the absence of clear life-cycle effects promoting the adoption of religious identity and involvement in adulthood -then Scottish society seems set on a course of further secularisation in subsequent decades.
While this article has provided a detailed analysis of recent trends in religious affiliation and attendance, as well as the contemporary social basis of religious engagement, in Scotland, it could not feasibly consider several of the other areas that featured in Field's study (2001).Therefore, inevitably, further quantitatively-oriented research is both necessary and indeed worthwhile to provide empirical scrutiny of other areas relevant to religious decline and increased secularisation in Scotland in the last couple of decades.
Moreover, given the steady growth in the 'religious nones' amongst the adult population in Scotland and the generational increase in those without a religious upbringing, more serious consideration is needed of secular identities and their potential consequences for social and political attitudes.Scotland andEngland, 1992-2014 Source: Author's analysis of SES, SSA, BES and BSA surveys.Source: Author's analysis of SSA 2014.
Note: Percentages have been rounded and sum across the rows.Source: Author's analysis of SSA 2014 survey.
Note: Percentages have been rounded and sum across the rows.

Figure 4
Figure 4 reports, for Scotland and England, the proportions identifying as Church of Scotland and Church of England / Anglican, the proportions identifying as Catholic, and the proportions identifying as other Christian.Figure 5 then provides a comparison of three indicators of secularity in both countries: the proportions with no (current) religious affiliation; the proportions not raised within a particular religion; and the proportions that do not attend religious services.Based on the time-series data presented in this section, it can be established whether any particular religious indicator sets Scotland apart from England or whether, in all key respects, levels of affiliation and attendance are broadly similar.

Figures and tables Figure 1 :
Figures and tables

Figure 4 :
Figure 4: Identification as Church of England/Scotland, Catholic other Christian,Scotland and England, 1992-2014

Figure 6 :
Figure 6: Percent with no religious affiliation, age group by sex, 2014

Figure 7 :
Figure 7: Percent that never attends religious services, age group by sex, 2014

Table 3 :
Religion of upbringing, age group, 2014 Author's analysis of SSA 2014 survey.Percentages have been rounded and sum across the rows.