Church teaching and LGB mental health

Peter Ould writes: The Oasis report, “In the Name of Love”, has received lots of attention since its release on Friday.  The Oasis paper makes three claims, two of which are relatively uncontroversial. The first is that “LGB people are significantly more likely to experience mental health problems than heterosexuals“. Several papers are cited to support this contention and there is little academic dispute of this key point. Second, the paper states, “These problems with mental health are as a result of discrimination and a sense of societal inferiority”. Again, the link between what is known as “minority stress” and mental health is a well established observation and we find it in a number of other minorities (BAME etc).

In the LGB community there is a particular link between the level of non-affirmation of homosexuality and increased mental health issues, but the Oasis paper does not actually cite any quantitative research to support the direct link between between LGB minority stress and mental health outcomes when compared with other possible causes for depression etc. The paper also makes the claim that “There is a growing consensus that these vulnerabilities arise as a result of the treatment of LGB people in society” but again no direct quantitative research is presented to support this. Indeed, some of the more recent quantitative research in countries that have seen dramatic improvements in societal attitudes towards LGB people evidence very little change in mental health outcomes when compared to other surveys a decade or more ago (for example R Bränsträm et al 2015 when compared with earlier Swedish studies).

Of course, none of this means that we should neglect issues of LGB minority stress and mental health outcomes, but we should be cautious when ascribing particular blame for it to a specific cause. In this light of this, the third and key claim of the report that has been heavily publicised is the assertion that conservative teaching in churches specifically contributes to negative mental health outcomes in LGB people. The report claims (p14),

The research presented in this report has made stark observations about the role of UK churches in contributing to poor mental health in lesbian, gay and bisexual people. The consequences of this can lead to anxiety, depression, low self-esteem, self-harm and even suicide.

That is quite a bold claim to make because it has tremendous pastoral consequences, particularly for the Church of England which plays a unique role in the life of this country. It is of course right to highlight things that the church is doing if they are harmful and to encourage improvements. In the past two weeks we have heard a number of allegations about abuse conducted by those associated with the Church of England and also others within the Church who did not appear to act appropriately when they found out about it. We expect our leaders to take their strategic vision and culture setting role very seriously, and if the teaching of the Church is harming people we need to respond to that.

But the problem with the Oasis paper is that it provides not one shred of research evidence that demonstrates a link between conservative Christian teaching and mental health outcomes for LGB people. That’s curious, because the research is available, it is of a level that is statistically rigorous and it is easily accessible. Why is it that Oasis report doesn’t engage with it?

Barnes and Meyer (2013)

The answer may be in what the research shows. Take for example Barnes and Meyer (2013) on “Religious Affiliation, Internalized Homophobia, and Mental Health in Lesbians, Gay Men, and Bisexuals“. (‘Internalized homophobia’ is a technical used in research that refers to negative self-perceptions of gay and lesbian people in relation to their sexuality.) In this study the researchers interviewed almost 400 LGB men and women and explored whether their exposure to “non-affirming” religious settings (i.e. conservative churches) increased their depressive symptoms and decreased psychological well-being. This is exactly the claim being made in the Oasis paper so we should have expected them to review research like this that directly explores the claim they are making.

There are two key outcome observations from Barnes and Meyer (2013) and they are found in the tables below. Let me walk you through them. Table 3 is a series of mathematical models that looks at the correlations between the items on the left (Black, Latino and being affiliated to a “non-affirming” church) and four different other markers (one of which is repeated). Let me explain what each model shows us.

  • The first model tells us that in this sample, if you are Black or Latino LGB, you have a higher likelihood of affiliating with a “non-affirming” church. This is an observation in line with other social research which tells us that the same correlation exists in the wider population, namely that those from ethnic minorities have higher church attendance and affiliation than white Caucasians.
  • The second model tells us that if you are affiliated with a non-affirming church that this affiliation is a far higher predictor of you having “internalized homophobia” than racial identity. Indeed, once you include the affiliation with a non-affirming church your race doesn’t have a statistical impact on your internalized homophobia (evidenced by the p values of the B co-efficients being greater than 0.05). This means that the first hypothesis the researchers were testing, that Blacks and Latinos find their internalized homophobia increasing because of affiliation to a conservative church is not proven by the actual research data.
  • The third model shows us that Black and Latino LGB people report much higher attendance at a conservative church than those who are white. Again, this is to be expected from the wider population.
  • The fourth model, simply repeats the findings of the second model in a different way.

Remember, in all of the above the researchers are not demonstrating a cause and effect relationship, they are simply demonstrating mathematically where relationships do and don’t exist.

Now we move onto Table 4 and here the results are staggering. Let me explain the models.

  • Model 1 looks at the relationship between being LGB and having depressive symptoms and affiliating with a conservative church. The results show that affiliating with a conservative church reduces depressive symptoms in someone who is GLB, but the result is not statistically significant.
  • Model 2 adds in a measure of internalized homophobia in the individual alongside their affiliation to a conservative church. This time we see the clear relationship between internalized homophobia and depression, but now the effect of affiliating to a conservative church becomes much more powerful in ameliorating that depression. This effect is still statistically insignificant (p = 0.1) but it is more significant than before and is in the region of significance where we might want to explore it more. What this model means we can say, with confidence, is that “if affiliating with a conservative church has an effect on depressive symptoms for someone who is LGB, that effect is to reduce the depressive symptoms”.
  • Model 3 now looks at generalised psychological wellbeing. Again, affiliating with a conservative church appears to improve psychological wellbeing amongst those who are LGB, but at a statistical level that is not very meaningful.
  • Model 4 looks at measuring internalized homophobia as a factor in psychological wellbeing and then adding in affiliation with a conservative church. As you would expect there is a clear and obvious inverted link between internalized homophobia and psychological welfare, but once again being affiliated with a conservative church ameliorates that effect (in that those who are affiliated with a conservative church have better psychological well-being than those who don’t, including those who affiliate with a liberal church). The effect is still statistically insignificant (p=0.1) but again it is at a level that demands greater research.

What the authors have demonstrated here is that their research shows that, contrary to the claims that Oasis make with no evidence to support them, LGB people do not have worse mental health outcomes when in “non-affirming” church environments and indeed there is some evidence to suggest that attending a conservative church actually improves mental health for LGB people, even when they have significant internalized homophobia.

The paper does not look at whether attendance at a conservative church has a statistically significant improvement on LGB mental health (it only explores affiliation) and this would have been a welcome extra part. I will explore whether I can access the underlying data set and test this hypothesis.

Lease et al (2005)

There is one other paper that looks at this issue (Lease SH, Horne SG, Noffsinger-Frazier N. Affirming faith experiences and psychological health for Caucasian lesbian, gay, and bisexual individuals. Journal of Counseling Psychology. 2005;52:378–388). This paper uses a different mathematical approach (partially mediated model) but comes up with the following results:

The paths between Affirming Faith Experiences and Psychological Health and between Internalized Homonegativity and Spirituality were not significant, indicating that Affirming Faith Experiences did not have a direct effect on Psychological Health.

To be clear what Lease et al (2005) tells us – affirming (i.e. liberal churches) do not produce improvements in the mental health outcomes of those attending. The research did find (like Barnes and Meyer 2013) that internalised homophobia was a direct link to mental health outcomes, but this was not improved by attending a liberal church.


It is disappointing to find Oasis making such sweeping claims for the impact that conservative churches have on LGB people without actually engaging with the research that explicitly explores this issue. Neither of these two academic papers that have quantitative models assessing the claims are referred to or cited in “In the Name of Love” and instead they rely on generalised research on minority stress and then try to blame this on conservative churches. The research evidence suggests this is an incorrect association to make, and indeed conservative churches may actually help reduce depression and other mental health issues in those LGB members and affiliates. Other research indicates (for eg Kubicek et al 2009, Jones and Yarhouse, 2011) that it is not necessarily conservative or liberal theology per se that helps to mediate internalised homophobia and other LGB minority stress, but rather the very engagement in a religious community and establishing a coherent sexual/spiritual understanding of one’s self (whether liberal or conservative) that reduces mental health issues.

It is clear that LGB folk do, on average, suffer from minority stress (for eg internalized homophobia) like many other minorities do and we should take that very seriously in our pastoral response. Barnes and Meyer (2013) would seem to suggest that rather than being at blame for that minority stress, non-affirming churches may actually help to relieve it.

The Church of England has placed great emphasis on engaging with good scientific literature on issues around LGB pastoral concerns, as evidenced in both the Pilling and Shared Conversations processes. As Synod meets this week to discuss where we move forward on wider LGB issues, it’s important to be open and honest about the research in key areas.

Revd Peter Ould is a Church of England priest based in Canterbury and a writer and broadcaster on issues to do with the Church and sexuality. He has two decades of commercial and academic experience interpreting and communicating mathematical models within the banking, opinion polling and social sciences sectors.

Follow me on Twitter @psephizoLike my page on Facebook.

Much of my work is done on a freelance basis. If you have valued this post, would you consider donating £1.20 a month to support the production of this blog?

Signup to get email updates of new posts
We promise not to spam you. Unsubscribe at any time.
Invalid email address

If you enjoyed this, do share it on social media (Facebook or Twitter) using the buttons on the left. Follow me on Twitter @psephizo. Like my page on Facebook.

Much of my work is done on a freelance basis. If you have valued this post, you can make a single or repeat donation through PayPal:

For other ways to support this ministry, visit my Support page.

Comments policy: Do engage with the subject. Please don't turn this into a private discussion board. Do challenge others in the debate; please don't attack them personally. I no longer allow anonymous comments; if there are very good reasons, you may publish under a pseudonym; otherwise please include your full name, both first and surnames.

47 thoughts on “Church teaching and LGB mental health”

  1. I must say that I’m completely and utterly unsurprised by the findings of the research Peter presents here. My anecdotal experience of being involved in conservative churches and seeing many who are involved in revisionist/”affirming” churches is that, generally speaking, the longer an individual engages with a conservative church the better their mental health becomes, irrespective of the underlying cause of mental health issues.

    It would seem that Oasis have been presenting alternative facts.

  2. Pete – please help me I’m being thick.

    You write:
    ‘The second model tells us that if you are affiliated with a non-affirming church that this affiliation is a far higher predictor of you having “internalized homophobia” than racial identity. Indeed, once you include the affiliation with a non-affirming church your race doesn’t have a statistical impact on your internalized homophobia.’

    To me this says that affiliation with a non-affirming church is a strong predictor of ‘internalised homophobia’. And thus that such affiliation may be a causative factor in ‘internalised homophobia’. But you conclude: ‘This means that the first hypothesis the researchers were testing, that Blacks and Latinos find their internalized homophobia increasing because of affiliation to a conservative church is not proven by the actual research data.’ This to me seems the opposite conclusion. What am I missing?

    The other studies seem to show that, given ‘internalised homophobia’, affiliation with a non-affirming church predicts (slightly) better mental health. But given the above, that ‘internalised homophobia’ itself is strongly predicted by affiliation with a non-affirming church, and given also the ‘clear and obvious inverted link between internalized homophobia and psychological welfare’, are we not justified in concluding that non-affirming churches predict ‘internalised homophobia’, and ‘internalised homophobia’ predicts poorer psychological welfare, and thus there may be an indirect causative link between being a non-affirming church and poor psychological welfare, via increased ‘internalised homophobia’ in a non-affirming environment? Is that not what critics are really getting at – not that we can help those who have ‘internalised homophobia’ but that we cause it in the first place?

    Having said this, some of the other studies do seem to make the key point. As you say:
    ‘Indeed, some of the more recent quantitative research in countries that have seen dramatic improvements in societal attitudes towards LGB people evidence very little change in mental health outcomes when compared to other surveys a decade or more ago (for example R Bränsträm et al 2015 when compared with earlier Swedish studies).’ And: ‘Affirming (i.e. liberal churches) do not produce improvements in the mental health outcomes of those attending. The research did find (like Barnes and Meyer 2013) that internalised homophobia was a direct link to mental health outcomes, but this was not improved by attending a liberal church.’

    So maybe I’m just missing something. Can you help?

    • “And thus that such affiliation may be a causative factor in ‘internalised homophobia'”

      I don’t say that.

      The first set of models tell us that those GLB people who attend a conservative church are more likely to have minority stress. The key point though is that we do NOT know from the study where there is a causative chain here at all (i,e. the attendance causes the minority stress OR the higher minority stress makes people attend conservative churches). What we CAN do is hypothesise why the correlation exists and then see if any research exists to test that hypothesis.

      Here is one hypothesis that might explain it (the association between higher minority stress and conservative church affiliation) – those GLB people who grow up in conservative religious environments will, by their nature, have higher GLB minority stress than those who don’t. This may simply be the internal struggle of trying to reconcile their emotions and desires with the religious teaching they have heard, even if that religious teaching is conservative but compassionate. These folk naturally gravitate to conservative churches and so we see the association of higher levels of minority stress with conservative church affiliation as a young adult. The conservative church the young adult attends hasn’t caused the minority stress itself – the minority stress already existed.

      Once this GLB young adult is at a conservative church we then look at their mental health (compared to those not at a conservative church). Table 4 models 1 and 3 tell us that being GLB at a Conservative Church is NOT in any way a predictor of worse mental health compared to those GLB who aren’t at a conservative church. However, when we include minority stress in the model we see that that IS a clear predictor of mental health and then after that the attendance of a conservative church might actually help IMPROVE the mental health position of a GLB young person compared with someone who doesn’t attend a conservative church but has the same level of minority stress.

      Does that make sense?

      The key issue is not jumping to conclusions – you write, “thus there may be an indirect causative link between being a non-affirming church and poor psychological welfare” but that is a hypothesis that you would need to test. Models 2 and 4 on table 4 seem to suggest the opposite, that actually being in a non-affirming church leads to better mental health outcomes for someone with the same level of minority stress who isn’t attending a non-affirming church.

      The models tell us the associations – we need to be very cautious about presenting any hypothetical causations as “fact”, and this is one of the major errors of the Oasis paper.

    • Hi Will,

      Peter can explain for himself, but I see a difference between a strong correlation and causation.

      As a comparison, women who underwent combined HRT also exhibited a lower-than-average incidence of coronary heart disease.

      Although, the correlation makes the former a fairly accurate predictor of the latter, the inference that HRT reduces CHD (it actually increases it) was disproven by randomized drug trials.

      In fact, the reduced incidence of CHD was eventually attributed to a coincident factor: better-than-average diets/exercise regimes of women with higher socio-economic status who predominate among HRT recipients.

      Similarly, a strong correlation between affiliation to a ‘non-affirming’ church and ‘internalized homophobia’ does not prove that the former is the cause of increase in the latter.

      The Oasis study (without reference to quantitative research) crudely condemned ‘the role of UK churches in contributing to poor mental health in lesbian, gay and bisexual people’.

      Yet, Oasis have not eliminated other co-incident factors, which may be more relevant in determining the root cause.

  3. Hi Peter,

    Thanks for going to the trouble of actually looking at the papers, and for explaining them.

    I feel you are misrepresenting the second paper. As you say, no direct significant link was found between affirming churches and psychological well-being. However, an indirect link (0.19 effect, p<0.01) was found. The indirect link works because affirming churches appear to both increase measures of spirituality (which is linked to psychological well-being), but also crucially because there is a direct link between attending affirming churches and decreased internalised homo-negativity (effect size -0.33, p<0.01).

    In other words, attending affirming churches is significantly linked (both statistically and in effect size) with people feeling less negative about their own homosexuality, which in turn means that they have greater psychological health. Quotation: "Interaction with an affirming faith group is clearly beneficial." (p.365).

    This is actually borne out by the conclusions the authors drew in the first paper. Being lesbian or gay in such an environment is bad for health (because it is associated with significantly higher rates of internalised homonegativity). However, this is balanced by the general positive effects that being associated with a faith community has in other areas. Note also that the group who affiliated were self-selecting; as the authors note, many who had bad or damaging experiences may have ended up in the non-affiliated group (other research suggests that lesbians and gays leave church at a much faster rate than average).

    To summarise: going to church is good for you. Being gay or lesbian in a conservative church is bad for you (but the generally positive effects of church may balance it out for some). Being gay or lesbian in an affirming church is good for you.

    • You write:

      “To summarise: going to church is good for you.”


      “Being gay or lesbian in a conservative church is bad for you”

      No – the research does NOT tell us this. The research tells us that those who attend a conservative church tend to have higher levels of minority stress than those who don’t. That is an associative link and you have turned it into a causative link. By all means propose a hypothesis as to why there is that associative link, but until it is tested quantitatively it is just a hypothesis.

      You are making exactly the same failure as the Oasis paper – turning a hypothesis into a fact and relying on anecdotal evidence to support it.

      ” (but the generally positive effects of church may balance it out for some).”

      Let’s be very clear what the 2013 paper tells us. It shows that for two people with the same level of minority stress, the one who attends / is affiliated to a conservative church does not have worse mental health outcomes and might even have better mental health outcomes.

      ” Being gay or lesbian in an affirming church is good for you.”

      Again, you’re not understanding what the 2005 paper is and isn’t saying. That those who attend an affirming church have lower levels of minority stress is not in dispute – this is the same as the finding in the 2013 paper. The issue is the chain from minority stress / internalised homophobia to mental health issues. It is at this point in the model that the researchers can find no link between attendance at an affirming church and mental health outcomes.

      Lease et al write:

      “The indirect nature of the effect of faith group affirmation on psychological health speaks to the complexity of examining religion and health. Although the common discourse suggests a negative association between religion and psychological health for LGB individuals, this relationship might be misunderstood without the inclusion of positive faith experiences and intervening variables more proximal to psychological health.”

      Lease et al essentially pushes us to do more research as does the 2013 paper. Lease et al recognise that “spirituality” in general appears to be the link to better mental health and that’s why I included the two qualitative papers that explore this from a liberal and conservative perspective, both of which provide good evidence to support this hypothesis.

      • OK, I should have written that there is a clear association between going to church and good mental health, and a clear association (if you are lesbian or gay) between affiliating to a conservative church and greater internalised homophobia. And there is a clear association (if you are lesbian or gay) between going to an affirming church and good mental health. And there is an association between internalised homophobia and poorer mental health.

        I think I know what my hypothesis would be.

        You say:
        “Let’s be very clear what the 2013 paper tells us. It shows that for two people with the same level of minority stress, the one who attends / is affiliated to a conservative church does not have worse mental health outcomes and might even have better mental health outcomes.”

        However, the same paper also tells us that you are more likely to have minority stress in the first place if you affiliate to a conservative church. To quote from Barnes et al:
        “participants who attended in non-affirming religious settings had significantly higher internalized homophobia than those who attended in affirming settings and those who never attended”.

        I stand by my understanding of the Lease paper. They propose a clear model, whereby attending an affirming church is directly linked both with spirituality and also with lower internalised homophobia, both of which in turn are directly linked with better psychological well-being. Only highlighting that there is no direct link is misleading. In terms of effect size, the larger link is between affirming church and lower internalised homophobia, and in turn between internalised homophobia and psychological health.

        You write:
        “It is at this point in the model that the researchers can find no link between attendance at an affirming church and mental health outcomes.”

        On my reading of the paper, you are flat-out wrong. They do find a link – (effect size 0.19), but the link happens because going to an affirming church is associated with greater spirituality, and lower internalised homophobia. Which is the same model that Barnes et al end up suggesting in reverse for non-affirming churches.

        • Jonathan,

          Let me give you the full quote for your 0.19 effect:

          “Although the Affirming Faith Experiences factor had no direct effect on Psychological Health, it did have an indirect effect on Psychological Health through its combined positive relation with Spirituality and negative relation with Internalized Homonegativity. The standardized parameter for this indirect effect was .19 ( p < .01)."

          This is simply confirming what I said – affirming church environments by themselves do NOT improve mental health. It is the inter-relation of church environment, resolution of conflict between religion and sexuality and the interaction of that with IH that is the key. This is a multi-dimensional model and you cannot therefore say that f(x)=y – the model is f(x,y,z,) = A.

          • As far as I can see, the quote is saying what I said (no surprise – I was partly basing it on this quote). I repeat – to say there is no link, or to say there is no direct link whilst failing to measure the indirect link, is misleading.

          • With the greatest of respect, I don’t think you understand what the multivariate regressions are telling us. You’re relying on one sentence which I’ve already shown has to be understood in relation to the regressions – you are implying a univariate relationship and the models don’t support that.

          • I admit it’s some time since I regularly looked at multivariate regressions. But I am not assuming a univariate relationship, do not think it’s a simple relationship, and still think your presentation misses out an important part of the paper.

  4. I would also note that the Oasis report is not primarily concerned with ‘conservative teaching in churches’ – but rather the conservative teaching outside of the churches in the broader media, with churches acting as the ‘biggest organised discriminator'(p.13). This was the third claim that the report makes. The research that Peter cites is only indirectly related to this issue.

    • You are in some parts correct, but the two issues are not unrelated. I have provided some very good quantitative research that shows that conservative church affiliation does not harm GLB mental health and in fact may improve it.

      If you want to suggest that conservative church teaching causes minority stress in the first place (and that is a good hypothesis) you need to point me to some good quantitative or qualitative research that shows that. The problem with every single paper that the Oasis report references is that most of the “evidence” presented for this causative link is anecdotal and where there is an attempt at a qualitative survey the kind of proper analysis that you see in the papers I cite (proper psychological measures, peer reviewed methods of rating religious involvement or internalised homophobia) haven’t been utilised. To be frank, lots of these surveys are of the level of:

      Interviewer – “Has a conservative church caused you mental health issues”
      Responder – “Yes”
      Interviewer – “Then it must be true”

      That is not the correct way to build a qualitative dataset and is in marked contrast to proper research. You’ll note that the IGLE papers that are cited by Oasis are not peer reviewed studies and provide no references themselves to the key claims that conservative religious experience causes GLB mental health issues.

      I am of no doubt that some GLB people report mental health issues related to conservative church association. I have heard many of their stories. But what we actually need to do with those claims is quantify and qualify them and make sure that what people are reporting is actually what they have experienced, and to help see whether the causal links they claim in their anecdotes are actually found in the wider population.

  5. Steve Chalke is the theological genius who informed us a few years back that the doctrine of penal substitutionary atonement taught that God was a cosmic child-abuser.

    I look forward to his next piece of Scholarship showing that this teaching is responsible for child abuse around the world.

    Of course his report is a sloppy piece of work with more holes that Rab C. Nesbitt’s vest.
    But the same could be said fro a lot of sociologically-based campaigning (which is what this is).

    That people with same-sex attraction have significantly poorer mental health than the population at large has been known for many years. What this report didn’t mention is the correlation between SSA and smoking, alcohol and drug use (including ‘poppers’ and crystal meth), and STI. Which one feeds the other is the question.

  6. Brian is correct re these correlations, and another very significant one he does not mention is the correlation with promiscuity. Promiscuity will play havoc with people’s sense of identity and integrity. Promiscuity is not, however, the result of persecution or unpopularity as a group (otherwise it would be everywhere among JWs) – rather, people embark on it because the prevailing narrative tells them it will be fun, whatever the reality.

  7. I read an article about the report and was struck at how lazy the assumptions were. Data was presented with absolutely no connection with the argument being made. There’s no need to make it up!

  8. Christian churches are often blamed for men who are same-sex attracted having feelings of self-loathing that can make them suicidal. May it not rather be that we have an inbuilt sense which is God-given, of what is right and wrong in human relationships , and this is what produces these feelings, even for those who have had no church connections? Then, trying to fight the feelings on their own produces despair. Hence belonging to a non-affirming church thereafter can help them. I don’t think I have ever seen any research on this aspect.

  9. This analysis is wholly wrong. For so many reasons they won’t all fit in this comments box. Here’s a few of the most fundamental:

    – the Oasis report does not claim what you say it claims. It does not say that going to a conservative church worsens your mental health. It claims that the church is a major source of discrimination and homophobia within society (almost all churches of all flavours), and this discrimination and homophobia in society is the source of worse mental health outcomes for LGB people. In other words, church teaching has negative impact on everyone, not just those who attend the church, by creating a negative climate for LGB people everywhere. I think you accept that minority stress through victimisation and homophobia is a fact. The only extra thing the Oasis report adds is that the church contributes to this.
    – Oasis do cite evidence for this – there are numerous sources, including the ILGA report, which in turn draws on over 30 additional sources, including this This report (of primary research) shows how LGB people in Europe feel religious institutions are a source of social exclusion. A third of LGB people in that survey experienced direct discrimination within their religious community and states, “Church institutions were often described as inherently homophobic, which in some cases led to the development of internalised homophobia at the intrapersonal level”. However, you wont find a table saying that going to a conservative church is worst for your mental health, because that claim is not made in the Oasis report.
    – The study you cite above serves a purpose, but not the one you think. It is a very small study of LGB people in New York. Of those it sampled, just 43 of them went to church once a month or more (assuming they were honest about this, and in most surveys people exaggerate their church attendance). Almost half of them were Latinos and just 4 regular church goers were white protestants. Reading across anything to the UK is a herculean leap of faith, and claiming that their mental health outcomes were affected by the teaching they heard in these churches is another jump not contained in the evidence.
    – Even if you do take the evidence you cite at face value the conclusion of that report is that exposure to conservative churches does damage mental health outcomes through increased internalised homophobia. However, other LGB people had improved mental health outcomes, so the effect in aggregate was offset. So are you seriously saying that it is to traumatise some LGB people, as long as others see improved mental health outcomes and the overall effect is neutral? Not cool. Your conclusion is almost diametrically opposite to that of the authors of the study.

    However, outside of this, I find it astonishing that the real experience of mental health and suicide amongst LGB people can be treated in this way. As something to be batted away as ‘nothing to do with us’. This stuff must be taken much more seriously than this. You have often argued that this issue is not handled well by many conservative churches and I know you are looking for a different response and new narrative through the Living Out initiative. You must realise that too often churches have got this wrong and had a devastating impact on people. The Oasis report, says nothing more or less that that. I would expect you to back that.

    I will conclude with the final paragraph from the study you cite above, which is almost diametrically opposed to the conclusions you managed to extract from it.

    “Our results contribute to the increasing evidence that clinicians working with LGBs need to be attuned to their clients’ religious backgrounds and current religious commitments (Bartoli & Gillem, 2008; Haldeman, 2004; Morrow, 2003). Clients’ exposures to homophobic religious environments should be plumbed, as well as how clients have responded to the strain that engagement in these environments may have caused them. To the extent clients were slow to extract themselves from non-affirming environments, or continue to expose themselves to such environments, clinicians need to be sensitive to competing forces that keep LGBs there (Bartoli & Gillem; Haldeman, 2004). Additionally, affirming environments perhaps need to pay attention to the extent to which they are potentially a refuge for a large number of LGB individuals coming from diverse religious, cultural, and social backgrounds. Increased sensitivity to this diversity could help meet some currently unmet demand for affirming settings. A profitable avenue of future research would be to compare mental health outcomes longitudinally of those who stay in non-affirming settings with those who traverse to affirming settings. Presumably, given a fitting affirming environment, those who make this change continue to reap the mental health benefits often afforded by religious communities, while avoiding the competing costs imposed by non-affirming environments.”

    • So, let’s see.

      You assert that that the Oasis report ‘claims that the church is a major source of discrimination and homophobia within society (almost all churches of all flavours), and this discrimination and homophobia in society is the source of worse mental health outcomes for LGB people. [emphasis mine]

      Report actually goes further: In the pages that follow, we will establish that there is watertight evidence that lesbian, gay and bisexual people are more likely than heterosexuals to experience poor mental health. We will demonstrate that negativity within society is a major cause of this. We will show, beyond reasonable doubt, that it is the Church and local churches who are fuelling this negativity

      The Report also concludes: In the UK, local churches are one of the biggest ‘organized discriminators’ of LGB people. Christians are also the biggest grouping of people who fuel negative attitudes about same-sex relationships in media and society.

      Oasis’ purported evidence of this is to be found on page 12 of the Report:

      An analysis by the Oasis Foundation found that there is significant evidence to show that the Church is the biggest negative voice in the public, political and media discussion around the
      legitimacy of same-sex relationships.

      Of the signatories listed on the website of the Coalition for Marriage (the campaign against same-sex marriage), 74% can be publicly identified as Christian.
      Of the MPs who voted against the introduction of same-sex marriage in 2013, 54% self-identify as Christian and many others may privately consider themselves people of faith.
      An analysis of 100 national media articles on the topic of ‘same-sex marriage’ found that 47%
      contained a negative comment, and of those negative comments 91% are from a Christian
      leader or commentator or politician who can be identified as Christian.

      So, let’s get this clear: Oasis’ claim is that there is a causal link between LGBT mental health problems and negativity within society. And the proof that Christians form the largest group fuelling this negativity is their predominance among the following:
      (i) affiliation with campaigns against SSM
      (ii) MPs who exercised their democratic right to vote against the SSM Bill (for whatever reason)
      (iii) Commenters who do not write positively in response to national media articles about SSM.

      Of course, the problem here is that Oasis is equating these examples of non-affirmation to the fueling negativity. It’s a false dichotomy.

      I also don’t affirm Islam, but that neither means I’m fueling negativity towards Muslims, nor encouraging Islamophobia, whether internalised or otherwise.

      So the report goes beyond the claim that the C

    • If “almost all churches” believe homosexual practice to be wrong, which I get tired of repeating, is NOT homophobia, but the belief of the worldwide church since the time of Christ and the Jewish nation long before that, may it not simply be that their belief is right? What gives this generation the right to declare that we know better than any generation that has gone before us?

    • Briefly,

      i) Yes it does – the quote I use makes that direct connection
      ii) Most of the cited research (ILGA etc) is anecdoctal. Indeed, the ILGA youth suicidation claim has NO references for the link it asserts (so Oasis is relying on an assertion in a non peer reviewed document)
      iii) I have said very clearly that the papers I have communicated call for us to do further research. We can do as much work with them as we can, recognising their limitations. Again, this “small sample size” argument is often used by revisionists to reject research that they don’t like but ignored with research that fits their biases (for example Shidlo and Shroeder 2001 on harm of ex-gay therapy which has a much smaller sample, suffers from real sample bias and has no qualification around the claims of mental health outcomes)
      iv) We know there is a link between internalised homophobia / minority stress and religious upbringing but that link vanishes when it is compared between religious upbringing and suicidal ideation / mental health deterioration (for eg Gibbs 2015 which we have discussed on Twitter over the past 24 hours – We DO need to do more work on finding the causative links with internalised homophobia but the **associative** links we have at the moment are not sufficient to blame conservative church teaching for IH.
      v) Your final paragraph is suggestions from the researchers for more research – exactly what I am asking for. They are actually cautious in making any causative link and so should we.

    • I find several interesting paradoxes in this conversation between Peter Ould and The Church Mouse. I have not had an opportunity to check behind the claims of either of you, so I take them at face value for the present.

      Peter seems to me to be saying that whilst more research is needed, the research we have suggests that, paradoxically, same-sex attracted people in conservative churches experience both increased homonegativity and increased wellbeing. i.e. whilst homonegativity is normally correlated with decreased wellbeing, the opposite is (generally) the case in conservative churches. (This is where the multivariate analysis is so important.)

      The Church Mouse seems to be saying that the Oasis report is not really about same-sex attracted church members, but same-sex attracted people in society as a whole and that the message of conservative churches is increasing homonegativity in society as a whole where the wellbeing inversion paradox does not occur. This also seems paradoxical given how little we often feel society listens to the church.

      I hope I have correctly represented what both are saying. (And I recognise that I have just picked two items out of many things being discussed.)

      These claims do not appear to me to be necessarily contradictory. And, if they’re both true, this leads to a third paradox: that conservative churches are both increasing wellbeing of same-sex attracted people (those who become members) and also decreasing wellbeing of same-sex attracted people (those who do not become members).

      This is so important because same-sex attracted people are people, i.e. they are us! Or, more correctly, there is no legitimate concept of “they”.

      I would certainly like to know more about whether each of these claims is true. If the Church Mouse/Oasis claim is correct, then I want to know how I can avoid fuelling homonegativity in society. If Peter’s claim is correct, I want to know more about how my church can be life-giving. And, particularly, I want to know how I can do both together.

      • Tom, that is really helpful. But there needs to be another question: who is it that is broadcasting that the Church is so negative about same-sex attracted people? No ‘conservative’ church I know of even mentions it. The loudest voices are those of people like Steve Chalke.

        So the logic of your observation is that Chalke and others ought to stop criticising other Christians in the media…isn’t it?

  10. This is a penetrating analysis and critique of the Oasis report. Rev Chalke’s claims that conservative teaching leads to suicides is a disgraceful slur on this section of the Anglican communion.

    • We discussed this yesterday.

      Table 4 tells us that it is only unresolved religious upbringing conflict that leads to negative mental health outcomes. Furthermore Table 5 tells us that it is the parental relationship that is the key driver and NOT the religious background.

      This would lead us to hypothesis that a general conservative religious background is not itself the driver of mental health problems, rather the personal relationships within which is religious conflict are the key indicators. We need more work to discern the causes of IH in order to explore this.

      And Gibbs 2015 also tells us that resolved religious conflict leads to no increase in mental health problems.

  11. I’m a great believer in starting with common sense. I’d also be the first to say that common sense is not infallible and that counter-intuitive reality does sometimes show common sense to be misleading. So the obvious place when considering an issue is to start with common sense; but it is also reasonable to search out evidence and, having done so, it is the evidence which you must accept, whether it agrees with your assumed common sense or not. Of course evidence may be disputed, and it may be incomplete (we may not yet be in a position to form a fixed view); so great care is needed on all sides to ensure objectivity and also to be honest about uncertainty.

    The issue of mental health in the LGB community is an obvious example of this. You would expect (common sense) that human beings whose psychological gender is at odds with the design of their bodies (their sex) may well experience some degree of related internal stress. Added to that, LGB people cannot fail to be aware of their deviation from a very obvious natural order surrounding sexual behaviour. Add to that their occupancy of a very obvious social minority group, which is likely to be stressful. Add to that the particular nature of being a part of the LGB community. And, for our current times, add to that the emotionally charged campaigns which have constructed a highly assertive biological and social victim narrative from which no LGB people are allowed to deviate or to attempt to escape. It would be extraordinary if people in such a group were untroubled by any mental discomfort. None of this is judgemental either way– no ‘blame’ is attributable either to LGB people or to non LGB people, it’s simply what your would expect.

    I don’t like labelling churches which are mainstream orthodox churches regarding the CofE doctrine on sex and sexuality as ‘conservative’ churches; that is to concede that liberal attitudes are now the mainstream – as things stand they are not mainstream, and will not be unless the ‘one man and one woman’ doctrine is discarded.

    However Peter Ould’s article here clearly shows that there is at least no significant evidence of orthodox churches being more harmful to the mental health of LGB people than any other type of church. It is a case of the evidence being in line with what a reasonable person might expect, but not if you are someone whose understanding has been captivated by the ‘victim’ narrative previously mentioned. And so this evidence calls into question the sustainability of one of the main pragmatic arguments (causation of mental harm) against the current CofE position. If that argument cannot be sustained, and from this evidence it cannot, then the case for changing the church’s position is reduced to the socio-political one of ‘equality’ – more a case of virtue signalling rather than the elimination of actual harm. Since when has virtue signalling carried more weight for a Christian church than being faithful to God’s commandments?

    But most fundamental of all for Christians, whatever our evidence may reveal, is the revelation of God’s intention for how we should live. If we truly believe that God intends our flourishing and does not take pleasure from making us suffer, then it is not his nature or his word or his people that we need to challenge, it is our own willingness to submit ourselves to his will. No one ever promised that there’s never any cost to this but there’s a whole lot of common sense in accepting that God has the last word.

    • “You would expect (common sense) that human beings whose psychological gender is at odds with the design of their bodies (their sex) may well experience some degree of related internal stress.”

      Well, that may be common sense, but it has no relevance to the great majority of LGB people. There are, of course, exceptions (as there are among heterosexual people), but as a rule they do not have a psychological gender which is at odds with their physical sex.

      • Gender is a variable social construction of the different preferences and roles of men and women; it is not a scientific definition. It need have no direct relationship with having a biologically distinguishable ‘male or female brain’, for which there appear to be no simple differences anyway. Physical sex differences are binary except in a very few unfortunate cases and preference for sexual relations with the opposite sex are in line with that binary difference for the great majority of people, to the extent that it is reasonable to associate it with gender, as most people do.

        The psychology which operates in this one area of behaviour (sexual preference) does not mean that it necessarily operates in any other part, which is why homosexual people do not wish to change their sex. It is therefore reasonable to talk about having a psychological gender in the context of a particular area of preference, unless you are trying to say that behaviour has nothing to do with psychology.

        • Yes, physical sex differences are binary (with rare exceptions). We associate preference for sexual relations with the opposite sex with gender or sex (and it is not clear to me that there is any objective difference between sex and gender in this context) simply because it is an everyday fact of life that that is the preference of the overwhelming majority, and as far as we know always has been and presumably always will be. Of course, the two are related, in that we couldn’t have a sexual preference if we didn’t have a sex. The sexual preference of a small minority of people is a different one.

          As you rightly observe, homosexual people do not generally wish to change their sex (although a small number do, just as a small number of heterosexual people do). That is because they do not have a psychological gender or sex, socially constructed or otherwise, which is at odds with their physical sex. Being homosexual is not a gender (or sex) any more than being heterosexual is.

    • We should be clear that internalised homophobia (IH) is a real thing and is directly related to poor mental health (including suicide). The issue is what causes IH and we need more research in this area. Lease et al (2005) which I refer to in my piece suggests that a conservative religious background does NOT mediate through to be a statistically significant cause of the mental health outcomes – rather it’s the way an individual integrates (or fails to do so) their religious background and sexual identity that is the driving factor. This coheres with what we find in Gibbs 2015 and the more qualitative based research I have also linked to.

      • ‘rather it’s the way an individual integrates (or fails to do so) their religious background and sexual identity that is the driving factor’

        Good point, Peter. This also coheres with key APA findings and recommendations (
        ‘Same-sex sexual attractions and behavior occur in the context of a variety of sexual orientations and sexual orientation identities, and for some, sexual orientation identity (i.e., individual or group membership and affiliation, self-labeling) is fluid or has an indefinite outcome.’

        ‘Some individuals choose to live their lives in accordance with personal or religious values (e.g., telic congruence). however, telic congruence based on stigma and shame is unlikely to result in psychological well-being.’

        ‘The available evidence, from both early and recent studies, suggests that although sexual orientation is unlikely to change, some individuals modified their sexual orientation identity (i.e., individual or group membership and affiliation, self-labeling) and other aspects of sexuality (i.e., values and behavior). They did so in a variety of ways and with varied and unpredictable outcomes, some of which were temporary.’

        ‘Sexual orientation identity exploration can help clients create a valued personal and social identity that provides self-esteem, belonging, meaning, direction, and future purpose, including the redefining of religious beliefs, identity, and motivations and the redefining of sexual values, norms, and behaviors (Beckstead & Israel, 2007; Glassgold, 2008; Haldeman, 2004; Mark, 2008; Tan, 2008; Yarhouse, 2008).’

        ‘We encourage LMHP (Licensed Mental Health Professionals) to support clients in determining their own:
        (a) goals for their identity process;
        (b) behavioral expression of sexual orientation;
        (c) public and private social roles;
        (d) gender role, identity, and expression;
        (e) sex and gender of partner; and
        (f) form of relationship(s).’

        What’s important is where Church lays emphasis in teaching on and engaging pastorally with human relationships.

  12. The other day I read a lovely story on the Living Out website: I am not disgusting. I am not alone. I am redeemed.

    We are, of course, dealing with anecdotes here – and as someone once said the plural of anecdote is not anecdata. But when there are conflicting stories – and this one is far from the only one – what do you do? How is it some people seem to report such positive outcomes from within conservative theology, and other people apparently do not?

    I believe Oasis have simply assumed the secular narrative – that being ‘non-affirming’ is damaging to GLB people – and have not spent the time considering other options. I think there are many options for why GLB people might have significant mental health issues etc. Simply to assume the popular one is not good science or truth-seeking.

    • Wow, how can you read this as a success story and evidence that conservative Christian teaching is not harmful?

      “I had grown up in a Christian environment where gay people were dehumanized and called “zombies.” Christians had always given me the impression that it was impossible to struggle with homosexuality and to love Christ at the same time. I was taught to fear people who weren’t straight, and so my own sense of self-worth plummeted when I realized the portrait of homosexuality they painted was actually a mirror I was looking into. I felt so disgusted with myself, so isolated from other people, so alone …. I would wake up each morning in distress; it seemed that ending it all and just going to heaven would be so much easier.”

      Then I found Living Out and realised that I was not alone, they’ve helped me deal with the intense shame my sexuality causes by staying celibate. If I’d grown up like that, I too would find celibacy a relief.

      • The point is that, on a superficial reading, ‘conservative’ theology seems to be part of the solution as well as part of the problem. This in itself should caution us against making hasty assumptions. The experience of Living Out folk should not be ignored.

      • Because this narrative fits in with the research above – an initial very negative experience of religion in conflict with sexual identity that is resolved within a conservative context. The research above (and Gibbs 2015 – indicates that whilst the religious background is highly associated (which is not the same as causative) with internal homophobia (and this completely to be expected) it is the resolution (or lack of it) in the adult that is the key indicator of mental health outcomes. That resolution can happen in a conservative rather than a liberal context.

  13. “It is disappointing to find Oasis making such sweeping claims for the impact that conservative churches have on LGB people without actually engaging with the research that explicitly explores this issue.”

    LOL. You have a high view of Rev Chalke if you expected anything else. Steve has made a career out of blagging it.

  14. Wow. What a dodgy, dodgy and possibly deliberately wonky analysis? First rule in statistics, if it’s not significant, it’s not significant. We don’t pretend it is because we’d like it to be… 0.05 is the usual level.
    The analysis makes no reference to the obvious methodological issue of sampling bias, nor the problem of power (only n=34 in this sample attend non affirming churches more than once a month). You may know therefore that this would totally cock up a regression model (yet you make no mention of it). And you make no mention then if the issue of power… but I’m beginning to think there may be a reason for that.
    The paper found no negative effect of attending non affirming churches. It did not find – as you incorrectly assert, that attending such churches was positive.
    It’s a shame you have stated this. I would query your independence as a ‘researcher’ based on this nonsense!

    It seems there are a few people commenting who say the same thing. Yet you don’t listen.

    • ” First rule in statistics, if it’s not significant, it’s not significant. ”

      Not quite. If it’s significant then it’s significant. If it’s not quite as significant then it raises questions that need addressing.

      So look again at what I said. I recognised very clearly that that link between resolution and affiliation with a conservative religious environment was not significant at the 95% level but was at the 90% level. That leads us to accept the null hypothesis (religious affiliation does not affect mental health outcomes) and to prompt more questions around this observation, especially since this 90% significance occurs as an interaction with other variables where by itself it is completely insignificant.

      I completely accept your point on the sub-sample size. I have not ignored that at all, but I would ask for some consistency in your approach to that. I you don’t think that size of sub-sample is acceptable in this kind of research I wouldn’t expect to find you anywhere supporting the idea that we “know” that sexual orientation change efforts cause mental harm, because those surveys tend to suffer from extreme sample bias and sample size issues. Perhaps a bit of self reflection is in order.

      • If something is not significant we don’t pretend it is. We don’t say something ‘reduces’ if it doesn’t (if it’s not significant) we don’t say it ‘gets better’ if it doesn’t (if it’s not significant). We also don’t say it’s not ‘quite as significant’ (when it isn’t significant) unless we really want something to be significant when it isn’t. As researchers we might feel that – but it’s called researcher bias and any researcher guards hard against that! The peer review process protects others from it usually – which is what is incredibly irresponsible about your ‘article’. You published it on a blog, trying to sound clever by ‘walking people through’ a regression model that you totally misinterpreted and made false claims about.

        This is absolutely irresponsible- particularly in research. You would be totally discredited if you were a proper academic – which is why I guess you post on blogs, where peer review is all but absent!

        I feel for people who listen to your analysis and quote or cite it blindly… I feel more for the young people who believe what you say and continue attending environments that increase internalised homophobia because you have lied and told them it is good for them.

        Why would someone, anyone do that?

        • With the greatest of respect Nicola, you’re not a statistician and you’re just parroting stuff you go told on a quantitative methods course.

          95% is not a magic marker – Something that is 95% significant is still likely to be wrong once in twenty goes (as it were). We typically accept 95% as the marker of a strong relationship, but being just outside that doesn’t mean that a relationship is spurious. Rather, it simply means the relationship is not quite strong enough for us to treat it as “good as certain” but it can often prompt further study to refine hypotheses.

  15. I wonder if this rambling has any relevance to the apparently unrelated, but maybe in fact very related confusion about whether or not “God as Father violently punishes his son for the salvation of the human race”, and confusion about our human relationships, whether sexual or otherwise?

    A third of Britons believe they have a guardian angel, according to a poll conducted by ICM for the Bible Society. in 2016, 2000 people were interviewed from across the country, 39% of women and 26% of men believe in angels, 39% of over 75 year olds. Philip Pullman’s daemons have angelic properties; in a BBC webchat he describes the attractiveness of ‘the many worlds theory’ as ‘making sense in terms of the laws of physics’, allowing maybe a little wriggle room for the numinous? In St Matthew’s Gospel, 18:10, Jesus is recorded as saying that every child has an angel in heaven. Could Jesus’ angel be – God?

    Was the Quranic denial of the Trinity (not a Bible word) at least partly a result of Catholic and Orthodox Christians not being able to agree, how to describe the Trinity? In an attempt to sort the problem once and for all, I once tried to engage a Grand Mufti in discussion through an interpreting Bishop, about ‘whether the space and time in which God exists is bigger than ours, whether there is a fourth dimension outside and around our world’ which would explain our human inability to better understand the relationship between the human Jesus and the One he taught us to pray to, as “Our Father.” “Avoid metaphysical speculation” was (I think, through the Bishop) the Grand Mufti’s advice.

    But I won’t give in. Moving from a 2-dimensional triangle as a Trinitarian symbol, I have been making models of a polyhedra, a three sided pyramid with a fourth side, or base. Might that not be a better ‘symbol’ for God? Can the relation between the human Jesus and ‘Our Heavenly Father’ to whom he prayed, be defined more adequately for the third millennium?

    And, as we struggle over human sexual relationships, could we not learn from Abraham and his children from Isaac, Ishmael and his children through faith, acknowledge (with him) the mess he got himself into by not trusting God, but ourselves resolve to trust God more?

    I could try to dig myself out of the many holes in this comment, but will stop before I dig deeper. Unless someone picks up on any of them, and Ian permits the comment to run on!

    • Was the Quranic denial of the Trinity (not a Bible word) at least partly a result of Catholic and Orthodox Christians not being able to agree, how to describe the Trinity?

      …but I thought Islam is held to have begun in the 7th century, whereas the split between what became Eastern Orthodoxy and Catholicism didn’t fully occur til 1054, so chronologically is that likely?

      in friendship, Blair

      • Attempting to dig myself out of that one: maybe if I had used lower case catholic and orthodox? Maybe better, the Western church v. the Nestorians?


Leave a Reply to Ian Paul Cancel reply