Copyright Sociological Research Online, 2003


Alasdair Crockett and David Voas (2003) 'A Divergence of Views: Attitude change and the religious crisis over homosexuality'
Sociological Research Online, vol. 8, no. 4, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 31/10/2003      Accepted: 28/11/2003      Published: 28/11/2003


British attitudes towards homosexuality have changed with astonishing rapidity over recent decades. Society has managed to assimilate these shifts with relative ease. The Christian churches, however, as repositories of tradition and defenders of inherited values, have been finding it increasingly difficult to adjust to the new environment. The Church of England is internally divided in the face of an external crisis: the Archbishop of Canterbury acknowledges that the global Anglican Communion could split over the issue, and the church faces similar pressures domestically. These events raise important questions about how religious institutions come to terms with modernity. The rapidity of social change, the decline in deference to authority, the increase in tolerance of anything that seems a private matter, and the sense that sexuality is fundamental to the free expression of personal identity, all make it difficult for a church to declare that sexual orientation might disqualify one from ministry or even membership.This paper analyses empirical evidence covering two decades from the British Social Attitudes and British Household Panel surveys. It is apparent that no real consensus yet exists on basic issues of sexual morality. Society as a whole is highly polarised over the question of whether same-sex unions are wrong, with significant and increasing divisions between young and old, women and men, and religious and non-religious. Far from being better placed than others to avoid disputes, Christian churches suffer from compounded problems. The attitudes of lay Christians are starkly and increasingly polarised along the dimensions of ideology and religious practice. This gulf presents a particular problem for churches with both liberal and evangelical wings, notably the Church of England.

Attitudes; Christianity; Church Of England; Gender; Generation Generation; Homosexuality; Religion; Secularization


British attitudes towards homosexuality have changed with astonishing rapidity over recent decades. Acts that were illegal forty years ago are now widely regarded as wholly acceptable. Society has managed to assimilate these rapid shifts with relative ease; the Christian churches, however, have been finding it increasingly difficult to adjust to the new environment. The Church of England is internally divided in the face of an external crisis: the Archbishop of Canterbury acknowledges that the global Anglican Communion could split over the issue, and the church faces similar pressures domestically.

These events raise important questions about how religious institutions come to terms with modernity. Organised religion is a repository of tradition and a defender of inherited values; it acts as a brake on innovation, even if it cannot stand in the way of the new indefinitely. The question is whether that role is still viable. The rapidity of social change, the decline in deference to authority, the increase in tolerance of anything that seems a private matter, and the sense that sexuality is fundamental to the free expression of personal identity, all make it difficult for a church to declare that sexual orientation might disqualify one from ministry or even membership.

Before entering into matters of social theory, it would help to answer a few basic questions about what is going on in Britain. Why has the crisis come now, rather than 10, 20 or 30 years ago? Why have disagreements over the issue been increasing rather than decreasing? And why have the churches associations based on affinity and a common cause been especially prone to division? We have examined the British Social Attitudes and British Household Panel surveys from the past two decades in an effort to find out.

* Why now?

About half the British population has no memory of the time when it was criminal to be a practicing (male) homosexual.[1] Campaigns against discrimination have been active for decades; stereotypes and prejudice have been steadily eroded; gay characters kiss in the early evening soaps, and for 20 years there have been openly gay MPs and even government ministers. If churches were going to break apart over this issue, one might suppose that it would have already happened.

Part of the reason the crisis was so long delayed lies in the nature of organised religion. Some sectors of society are designed to be innovative: the creative arts and entertainment are obvious examples. Others are intended to safeguard tradition, e.g. the military and the church. Different institutions require different degrees of external pressure in order to change. Where experiment is valued, new things will be tried unless there is pressure to suppress them. In religion, creeds and values may be preserved until the status quo is no longer defensible. The level of public sympathy that enforces gay rights in most spheres may not suffice in the religious arena.

The ordination of women by the Church of England provides an example: Britain had a female Prime Minister over a decade before the parish of Nether Wallop could have a female priest. Similarly, the high proportion of marriages ending in divorce, in conjunction with the general acceptance of marital dissolution as a normal event, is only now beginning to affect the Church's treatment of remarriage. Religious institutions can resist pressure to change long after secular society has already adapted. Indeed, the pressure is hardly even felt until rejection of the old is well advanced in the secular sphere.

One might, therefore, conjecture that popular acceptance of homosexuality is only just reaching a level that makes defence of the status quo awkward for mainstream religious groups. In the 1983 British Social Attitudes Survey (the first carried out), just over half the respondents gave the strongest censure allowed, and declared that 'sexual relations between same- sex adults' were 'always wrong'; only one in five said that they were 'not at all wrong' or 'rarely wrong'.[2] By 2000, however, the proportion of respondents who were broadly accepting had doubled, while just over a third still condemned the practice without exception. The extent of the increase in public acceptance of homosexuality over recent decades is now too large to be ignored, but it is not yet quite so universal that the churches have been obliged to abandon their previous positions.

One might have expected a progressive growth of toleration, first in secular society, then among lay Christians and finally in the religious institutions themselves. Although the proportion of people in Britain saying that gay sex is not at all wrong has grown substantially, and at a roughly equal proportional rate among the religious and non-religious (albeit from different starting points), the general trend masks a growing gulf between liberal and conservative. Among the general population and more especially among Christians, ideological divisions have sharpened. As we shall show, there is a large and growing generation gap, a large and growing gender gap, and most importantly for the churches a large and growing gap between liberal and conservative Christians. The dimensions of age, gender and religiosity interact, with important implications for religious institutions: for example, the attitudes towards homosexuality of a young female Christian and an elderly male Christian are likely to be at opposite extremes even if they belong to the same denomination. In what follows we examine first the distribution of views in society at large, and then specifically the features of self- identified and practicing Christians.

* How have attitudes changed?

Table 1 shows the breakdown of responses to the question about sexual relations between same-sex adults, one that has been asked in the British Social Attitudes (BSA) survey in almost identical form for most years between 1983 and 2000. Although by 2000 over a third of the adult British population declared that sexual relations between people of the same sex are 'not wrong at all', slightly more still said that they are 'always wrong'. More people strongly censured than completely accepted same-sex unions. Even when one extends the 'anti' and 'pro' categories to include 'mostly wrong' on the one hand and 'rarely wrong' on the other, the 'antis' still predominated (46 percent versus 41 percent).

Table 1. Attitudes towards homosexual practice (%)
Source: BSA surveys, 1983-2000. The BSA was not carried out in 1988 or 1992 and no comparable question was asked in 1986, 1996, 1997 or 2001.

a The 1991 and 1994 data appear to have been coded on a four point scale without a 'rarely wrong' and 'don't know' options but with a 'can't choose' option that many (c. 11%) respondents selected. The data for these years are shown in grey to indicate a potential lack of comparability.

Note: These and all subsequent figures are based on data weighted to adjust for the different probability of selection for individuals from small and large households, and thereby make the BSA figures representative of the population of Great Britain. Responses of 'it depends' and 'don't know' are not shown and hence values do not sum to 100%.

As Scott (1998) and Kelley (2001) have shown with respect to Britain and other advanced industrial societies, these figures should disabuse us of the liberal temptation to suppose that homosexuality is now condemned by only a small minority. That said, as Evans (2002) has noted more recently, and with the benefit of the more recent BSA data analysed here, the speed of change in attitudes on this subject in recent years is remarkable: the 13 years from 1987 to 2000 saw a shift from a substantial majority (64 percent) believing that gay sex is always wrong and only a few (11 percent) having no objection, to the numerically balanced though starkly polarised breakdown of responses in 2000.

While it seems likely that the liberalising trend of the 1990s will continue, it is worth noting that attitudes became less tolerant during most of the 1980s. This shift presumably resulted from the AIDS panic, at the height of which 60 percent of BSA respondents in 1987 thought that the statement 'within five years AIDS will cause more deaths in Britain than any other single disease' was 'more or less true' (the most positive evaluation allowed), and only 11 percent believed that this claim was 'highly exaggerated'. It has been documented elsewhere that homophobia increased when AIDS related fears were at their height in the mid to late 1980s (Scott 1998: 820). As worries over AIDS diminished in the 1990s, acceptance of same-sex relationships increased steadily and rapidly.

This aggregate trend of rapidly increasing acceptance of homosexuality masks substantial and growing differences within the British population, most notably by age and sex, as figures 1 and 2 show.[3]

Figure 1. Censure of homosexuality by age, 1983 to 2000
Note: the best-fit lines are locally weighted and of variable bandwidth (SPSS 'Lowess' fit criteria are 50% of points and 5 iterations).

Figure 2. Censure of homosexuality by sex, 1983 to 2000
Note: the best-fit lines were plotted using the same criteria as Figure 1.

The black line in Figure 1 makes visual the figures in Table 1: the gradual rise in the proportion of the population condemning homosexuality in the early 1980s, followed by a steady decline though the 1990s. The red line in Figure 1 shows the trend by the youngest cohort (aged 18 to 24 years at the time of the survey). The blue line shows the trends by the oldest cohorts (aged 65 plus). There is a large and growing gap between the young and the old. The vertical axis is plotted on a log scale to make clear that the proportional difference between the attitudes of the youngest and oldest has increased substantially over the period.[4] To illustrate, in 1983-4 (both years pooled together) those aged 65 and above were 80 percent more likely to believe that homosexual relations are always wrong than those aged 18-24. By 1999-2000 (both years pooled) they were 221 percent more likely.

Increasing acceptance is only partly a period effect, with everyone becoming more liberal; a strong cohort effect is just as evident, in which more tolerant younger generations displace older and more judgmental ones. Interestingly, when one examines attitudes by age, the proportion in 2000 saying that homosexuality is wrong is little more than a quarter in all age groups up to 55, beyond which it increases sharply. By 2000, the baby boomers and their successors are considerably more liberal than the previous generations. They have not always been so liberal, though. The 1983-4 figures reveal that 38 percent of people aged less than 35 considered that homosexuality was always wrong.

There has also been a growing gender gap, with women (red line) finding homosexuality increasingly easier to accept than men (blue line), as shown in Figure 2. The vertical axis in Figure 2 is also a logarithmic scale to make clear how the proportional difference between the sexes increased throughout the 1990s. In 1983-4 men were only 16 percent more likely than women to believe that homosexual relations are always wrong. This gap remained about the same throughout the 1980s,[5] but increased throughout the 1990s, and by 1999-2000 men were 57 percent more likely than women to believe homosexual relations are always wrong.

The gender differences are even more apparent when considered jointly with age. In 1999-2000, men aged 18-24 were three and a half times as likely as women of the same age to condemn homosexuality; the gap declines steadily with age and is only 24 percent by age 65. The disparity by age and sex in the liberalising trend has greatly increased social divisions. In 1983-4 men aged 65 and over were little more than twice as likely as women aged 18-24 to believe that homosexual relations are always wrong. By 1999-2000 the corresponding difference exceeded a factor of 8, a staggering disparity.

* Why the church crisis?

The fact that society as a whole is increasingly divided not least by gender and generation does not necessarily explain why religious organisations should be the same. Many associations would expect their members to be sufficiently like-minded to be able to reach consensus on major public controversies. It would not be surprising if affiliates or churchgoers of a particular denomination tended to share a worldview that was either broadly accepting or rejecting of homosexuality. At a minimum one would suppose that differences within a religious group would be smaller or at worst no bigger than those found in society as a whole.

Perhaps the most astonishing finding of our study, however, is that the split within the religious subpopulation over homosexuality is even more severe than in society at large. As will be shown, liberal Christians (particularly young liberal Christians) are more than usually accepting of homosexuality; conservative Christians (including the young) are more than usually disapproving.

As a whole, Christians have higher than average levels of disapproval of homosexuality. Over the course of the last two decades there has been a persistent difference in attitudes between the religious and the non-religious. In 1983-84, 57 percent of self-identified Christians responded that homosexual relations were 'always wrong' compared with 43 percent of non-religious respondents; in 1999-2000 the gap was slightly wider, at 44 vs. 28 percent. As Table 2 shows, the contrast holds whether one looks at self- identification (affiliation) or regular attendance at services.

Table 2. Attitudes towards homosexual practice (%): religious and non- religious
Note: This table excludes the small number of respondents who failed to state their religious affiliation or their frequency of church attendance (or who stated 'varies too much to say').

a Non-attendance is classed as never or almost never (the questionnaire asks respondents to exclude attendance at special occasions such as weddings, funerals and baptism).
b Occasional attendance is classed as less than monthly.
c Regular attendance is classed as at least monthly.
d See note a to Table
1 about these years.

The first two columns of Table 2 indicate that there has been only a modest increase in the proportional difference between the attitudes of the religious and non-religious population. In 1983-4 religious affiliates were 33 percent more likely to believe homosexual relations were always wrong, in 1999-2000 they were 52 percent more likely. More significantly, there has been an important internal shift in the attitudes within the religious population between 1983 and 2000. As indicated by the rightmost three columns in Table 2, attitudes of occasional church attenders are no longer about half way between those of frequent attenders and non- attenders; they now simply resemble the latter, as Figure 3 helps to illustrate.

Figure 3. Censure of homosexuality by regularity of church attendance, 1983 to 2000
Note: the best-fit lines were plotted using the same criteria as Figure 1.

More systematic multivariate statistical analysis highlights two further important differential shifts in attitudes within the religious population. First, there are substantial generational differences (over and above the general cohort effects already noted). Secondly, there appears to be a widening gap between religiously liberal and evangelical positions.

Table 3. Logistic Regression of 'anti-gay' attitudes

Dependent variable: sexual relations between same-sex adults 'always wrong' (1), not always wrong (0)

Note: (1/0) indicates a dichotomous variable, where 1 indicates the presence of condition and 0 indicates absence. For example, Anglican (1/0) is a variable taking the value 1 if the respondent is Anglican and 0 otherwise.

The variables in the table are those that were selected for analysis using a forward stepwise selection criterion (p < 0.05). For each major denominational grouping (Anglican, Catholic, Baptist, Presbyterian, Methodist and non-denominational Christian), a dichotomous variable indicating affiliation to that denomination, and interaction terms between denominational affiliation and age, male gender, female gender and regular church attendance were added. All variables not reported in the table failed to meet the significance criteria for entry into the analysis.

Table 3 shows the output from a logistic regression analysis of anti-homosexual attitudes.[6] The results reinforce the interpretation of the previous graphs and tables. The odds of having 'anti-gay' attitudes increase for each year of age (above 18) and are substantially higher for men than for women and, as the interaction of age and sex shows, the gender gap is largest among the young. Turning to the religious variables, one finds that young Anglicans are less likely to hold 'anti- gay' attitudes than non-Anglicans at age 18, but there is a positive interaction between Anglicanism and age, such that for each additional year of age the odds of an Anglican having anti-gay attitudes increases by more than would be expected for non- Anglicans. The results indicate that the odds of an 18 year-old Anglican having 'anti- gay' attitudes are 48 percent lower than for non-Anglicans, for a 51 year-old Anglican the odds are equal, and for a 65 year-old Anglican they are 32 percent higher.

There is also a positive interaction between Presbyterianism and age, with the odds of a Presbyterian having 'anti-gay' attitudes increasing by more than the usual degree with age. Since Presbyterian affiliation was not a significant variable, the results suggest that while young Presbyterians are no more likely than others to disapprove of homosexuality, the same cannot be said for older Presbyterians (who, at age 65, are 75 percent more likely to disapprove than non-Presbyterians). Lastly, for Catholics, Baptists and non- denominational Christians, while mere identification (affiliation) with the group does not significantly alter the odds of having 'anti-gay' attitudes, regular church attendance does. The odds of a churchgoing Catholic having 'anti-gay' attitudes are 60 percent higher than for other people, for non-denominational Christians the odds are nearly three times higher, and for Baptists the probability of disapproval is even greater.

The emerging picture is that the attitudes of the younger affiliates of certain denominations (in particular the Church of England and the Church of Scotland) are far closer to those of the general population than to the ones held by their older or more conservative co-religionists. Young Anglicans, for example, are not only less 'anti-gay' than the general non-Anglican population (as evidenced by Table 3) but also than their non- religious contemporaries. Older affiliates of all denominations (even the Church of England) are less liberal, both in absolute terms and relative to their non-religious peers. To illustrate, in 1999-2000 only 17 percent of Anglicans aged less than 25 held 'anti-gay' attitudes compared with 21 percent of non-religious young adults. In contrast, 68 percent of Anglicans aged 65 and over held 'anti-gay' attitudes, compared with 56 percent of the non-religious of the same age.

Society as a whole is split over whether homosexuality is wrong, and Christians are even more divided than others. While it is clear that younger members of the Church of England have become more accepting of homosexuality relative to the general population, there is some evidence that members of more conservative denominations (in particular Baptists) have become less accepting, as shown in Figure 4. Although the chart should be treated with caution (as the underlying numbers are quite small for the smaller denominations), the pattern would appear to be clear: new generations of Baptists (and perhaps other evangelicals?) are not necessarily more tolerant of homosexuality than their elders (indeed they appear to be less so), while Anglicans, Catholics, Methodists, Presbyterians and non-denominational Christians show major generational differences. As Finlay and Walther (2003) have shown in analysis of American students, the differences in attitudes towards homosexuality hidden within an aggregate 'Protestant' label can be substantial.

Figure 4. Censure of homosexuality by denomination and age, 1999-2000

* How are attitudes being passed on?

A good source for investigating the inter-generational transmission of attitudes is the British Household Panel Survey (BHPS)[7], which included a question on homosexuality in Waves 8 (1998-1999) and 10 (2000-2001). While the question, which asks for level of agreement/disagreement with the statement that 'homosexual relationships are always wrong', is a more ambiguous instrument than the BSA question,[8] the longitudinal and household characteristics of the BHPS allow us to compare the attitudes of young adults with those of their parents. Such a comparison is made in Table 4, which presents results for two types of household: those where both parents are accepting of homosexuality (upper panel) and those where both parents censure it (lower panel).[9] The results are further broken down according to whether neither, one, or both parents report a religious affiliation (left-hand section) and whether neither, one, or both parents attend church regularly (right-hand section). To avoid contamination from cohort effects, we restrict our analysis to children aged between 16 and 34 (at Wave 8 of the BHPS).

Table 4. 'Pro-gay' and 'anti-gay' attitudes of children by parental attitude and religiosity (Wave 8 of the BHPS)

Note: Parental religious characteristics are taken from Wave 1 of the BHPS (or Wave 9 if Wave 1 had missing data), to better reflect parental religiosity during the child's formative years. Regular church attendance is defined as monthly or more frequent.

The results are very interesting. The transmission of acceptance of homosexuality ('pro-gay' attitudes) to the next generation does not seem to be affected by religious affiliation (see the upper panel of Table 4). About three-quarters of the children of 'pro-gay' parents are themselves 'pro-gay', regardless of parental religious affiliation or regularity of church attendance. In contrast, 'anti-gay' attitudes are far more successfully transmitted by religious parents, as the lower panel of Table 4 makes clear. While only a quarter of the children with non-religious 'anti-gay' parents are themselves 'anti-gay', the risk increases markedly to 42 percent if both parents are religious. The results by parental attendance are even stronger: 29 percent of children of two 'anti-gay', non-attending parents are 'anti-gay', rising to 37 percent when one parent attended regularly, and to 50 percent where both parents attended regularly.

At face value, these findings from the BHPS suggest that anti-homosexual attitudes among the British population may become increasingly confined to Christians, and in particular to the minority of observant Christians. However, one should note that the proportion of Christians in the British population is itself shrinking, so in each generation fewer children are being raised by religious parents, which more than cancels out the differential transmission of anti-homosexual attitudes. Indeed, the proportion of the 'anti-gay' population that is Christian has actually decreased over the last two decades.[10] Table 5 illustrates this point by showing the frequency of 'anti-gay' attitudes among two generations of BHPS respondents.

Table 5. 'Anti-gay' attitudes by sex, generation and religiosity
Source: British Household Panel Survey

Note: Sons and Daughters comprise all individuals whose natural mother is also a BHPS panel member. Mothers and fathers are the parents of these children ('fathers' include current male partners of the mother). Persons in each generation are categorised separately (e.g. religious sons are not necessarily sons of religious fathers and/or mothers). 'Anti-gay' and religious characteristics of parents are defined as for Table 4. Religious characteristics of children were taken from Wave 9 (or Wave 7 then Wave 1 if the later wave had missing data).

Non-churchgoing young adults (the first two columns in Table 5) are substantially less likely to have 'anti-gay' attitudes than their older counterparts, with inter-generational declines in disapproval between 20 and 50 percent. Among the young generation, those who are religiously affiliated (but non-practicing) are now extremely close to the non- religious in their attitudes towards homosexuality, while there is a clear contrast in the parental generation.

Among the churchgoing population of the two generations, however, there is very little inter-generational difference in 'anti- gay' attitudes: about half the churchgoing men and a third of churchgoing women have 'anti-gay' attitudes in both generations.

Both the BSA and BHPS data show how the differential concentration of 'anti-gay' attitudes within the Christian population (particular the conservative and observant Christian population), while increasing the polarisation of attitudes within the Christian population, does not equate to Christians forming a greater proportion 'anti-gays' among the British population. In the parental generation, 85 percent of 'anti-gay' women and 71 percent of 'anti-gay' men were Christian (by affiliation). In the children's generation, the corresponding proportions had shrunk to 61 percent and 41 percent respectively. In a secularising country like Britain, in contrast to the United States where the religiously active population is relatively stable, the relative success of churchgoing parents in transmitting disapproval of homosexuality to their children is not an effective brake on increased acceptance of homosexuality among the population as a whole, but it does serve to ensure that the religious community will remain starkly polarised on the issue for the foreseeable future.

* Conclusion

Recent changes in attitudes towards homosexuality have been among the most rapid observed on any issue. Although this shift has undoubtedly affected the social climate, no real consensus on the basics of sexual morality yet exists. Society as a whole is highly polarised over the question of whether same-sex unions are wrong, with significant divisions between young and old, women and men, religious and non-religious. Far from being better placed than others to avoid disputes, Christian churches suffer from compounded problems. Whether framed in theological terms as disagreements over scriptural authority or divine love, or described in more secular fashion as clashes between traditional and reforming ideologies, the internal strife has been serious. The analysis presented here has shown that the attitudes of lay Christians are also starkly polarised, presenting a particular problem for churches with both liberal and evangelical wings. Anglicans are famously adept at pragmatic compromise, but the current situation presents a challenge to the Church of England in particular: it is split down the middle and the opposing opinions are unlikely to moderate in strength. If present trends continue then a liberal consensus may eventually be reached. The evidence from major social surveys considered here, however, suggests that the present divergence of views is likely to be a feature of the religious scene for some time to come.


1 Lesbianism was never specifically legislated against. Male homosexual acts were decriminalised in England and Wales in 1967, though the age of consent was fixed at 21 rather than 16 as for heterosexuals. In Scotland, analogous decriminalisation did not occur until 1980. In 1994, the homosexual age of consent was reduced to 18 years in England and Wales.

2 The source for this and all other BSA data analysed in this paper is: National Centre for Social Research, British Social Attitudes Survey, 1983 to 2000 [computer files]. Colchester, Essex: UK Data Archive [distributor]. SN 1935, 2035, 2096, 2315, 2567, 2723, 2840, 2952, 3439, 3572, 3764, 3921, 4072, 4131, 4318, 4486.

3 Since the focus is on Christianity, these and all subsequent tables and figures exclude the small but growing number of people with a different religion (4.7 percent of the weighted sample in 2000). In 1983-4, the attitudes of non-Christian religious adherents were similar to the sample as a whole: 55.3 percent thought homosexuality was always wrong, compared with 52.3 percent of the total sample. In 1999-2000 the attitudes of the non-Christian group were no more tolerant, opening up a gulf between non-Christian adherents and the general population: 56.3 percent of non-Christians thought that homosexuality was always wrong compared with just 37.3 percent of the total sample.

4 On a log scaled vertical axis (as in Figures 1 and 2), a constant vertical distance indicates a constant proportional difference (e.g. 100 is 10 times greater than 10 and 10 is 10 times greater than 1, so that the distance between 1 and 10 will be the same as that between 10 and 100).

5 As Scott (1998: 839) suggested, the gender gap may have closed in the late 1980s as a result of the AIDS shock, which may have differentially increased anti-homosexual views among women. Women were certainly more afraid of AIDS: in the 1987 BSA, 64 percent of women and 56 percent of men thought that AIDS would be killing more people in Britain than any other single disease in the next five years.

6 The odds ratio measures the probability of the characteristic (being 'anti-gay') in one group relative to another. An odds ratio greater than 1 implies that the characteristic is more likely in that group (e.g. men) than among the contrasting group (e.g. women). For an odds ratio less than 1, the reverse is true. Column 1 of Table 3 shows the log of the odds ratio; the odds ratio itself is in column 3. Each unit increase in the variable (e.g. each additional year of age) results in the odds of the characteristic (being 'anti gay') being multiplied by the amount shown.

7 The source for this and all other BHPS data analysed in this paper is: University of Essex. Institute for Social and Economic Research, British Household Panel Survey; Waves 1-11, 1991-2002 [computer file]. Colchester, Essex: UK Data Archive [distributor], May 2003. SN: 4651.

8 A similar question was asked in the 1996 and 1997 BSA, and the breakdown of responses is very similar to those obtained in Waves 8 and 10 of the BHPS.

9 Acceptance is defined as either disagreeing or strongly disagreeing with the statement that 'homosexual relationships are wrong', censure is defined as agreeing or strongly agreeing with the statement.

10 In the 1983-4 BSA data, 73.4 percent of those who thought that homosexuality was always wrong were Christians (and 68.1 percent of the sample was Christian). In 1999-2000, only 66.8 percent of those who thought that homosexuality was always wrong were Christian (but only 58.1 percent of the sample was Christian).


EVANS, G. (2002) 'In search of tolerance', in A. Park, J. Curtice, K. Thomson, L. Jarvis, and C. Bromley (eds.), British Social Attitudes: the 19th Report. London: Sage.

FINLAY B. and Walther, C.S. (2003) 'The relation of religious affiliation, service attendance, and other factors to homophobic attitudes among university students', Review of Religious Research, Vol. 44, No. 4, pp. 370-393.

KELLEY, J. (2001) 'Attitudes towards homosexuality in 29 nations', Australian Social Monitor, Vol. 4, No. 1, pp. 15-22.

SCOTT, J. (1998) 'Changing attitudes to sexual morality: A cross-national comparison', Sociology, Vol. 32, No. 4, pp. 815845.

Copyright Sociological Research Online, 2003