A response to: ‘Conversion Therapy – A briefing note on the science by Professor Michael King (UCL) and Professor Robert Song (Durham)’

This briefing note , which has been sent round by Jayne Ozanne in support of her General Synod motion on Conversion Therapy, is designed to give scientific information in support of the proposition that members of the Church of England should reject all forms of conversion therapy for those with same-sex attraction as unethical because they cannot be shown to be effective and instead do harm.

The paper makes three basic points:

  • There are no randomised controlled trials that provide clear proof whether conversion therapies are effective or harmful (paragraphs 1-5).
  • Nevertheless the 2015 study undertaken by John Dehlin and others provides evidence that the rate of harm far outweighs the rate of benefit (paragraphs 6-11).
  • It is ‘deeply misleading’ to suggest that people are not ‘born gay’ and that ‘their sexual desires can change’ (paragraphs 12-13).

We shall look at each of these three points in turn.

Randomised Controlled Trials

The paper is correct to say that no randomised controlled trials on conversion therapies have been conducted and that it is unlikely that they will be. However, it should be noted that the reason that they are unlikely to be conducted is not simply because of fears that patients will be harmed, but because of an ideological objection within the medical establishment to the very idea of therapeutic work to change or control people’s same-sex sexual attraction. This is because the existence of such therapy would imply that such a change would be something beneficial and this in turn would imply that same sex sexual attraction is something problematic, something that the medical establishment has increasingly denied ever since the landmark APA decision in 1973 to de-list homosexuality as a mental health condition.

This ideological background to the discussion about conversion therapy is clearly revealed in the memorandum of January this year from the Royal College of General Practitioners and others that Jayne Ozanne wants the Church of England to sign up to. This states:

‘Conversion Therapy is the term for therapy that assumes certain sexual orientations or gender identities are inferior to others, and seeks to change or suppress them on that basis. Sexual orientations and gender identities are not mental health disorders, although exclusion, stigma and prejudice may precipitate mental health issues for any person subjected to these abuses. Anyone accessing therapeutic help should be able to do so without fear of judgement or the threat of being pressured to change a fundamental aspect of who they are.’ [1]

What all this means is that the claim in paragraph 4 of the paper that the reason that no randomised trials would be allowed would be because of evidence of harm is misleading because it is only partially true. There might be concern about harm, but the really big reason why such trials would not be permitted would be because of the ideological objection to the very idea of conversion therapy just noted

Evidence of harm

In the absence of any evidence of harm from randomised controlled trials the paper turns for evidence of harm to the 2015 study from John Dehlin and others entitled ‘Sexual Orientation Change Efforts Among Current or Former LDS Church Members.’

What the paper does in paragraphs 6-11 is to point out the problems with the earlier studies by Robert Spitzer and by Stanton Jones and Mark Yarhouse that claimed change with regard to people’s sexual orientation was possible and to suggest that we should accept instead the evidence of the Dehlin study of attempts at conversion therapy among the Latter Day Saints (LDS) which reported 0% elimination of same-sex attraction, 3% change in such attraction and a 40% report of harm. This evidence, the paper suggests, provides ‘good prima facie reasons for thinking there is significant potential for harm from conversion therapies’ (paragraph 11).

The problem with this argument is that the paper is misleading in its account of all three papers to which it refers.

If we turn first of all to what the paper says about the Spitzer study of 2003, we find that this is described as ‘the most notable study’ (paragraph 6). No justification is given for this claim (arguably the Jones and Yarhouse paper is actually more notable for the reasons which will be outlined below), but it is rhetorically powerful because it is linked to what is said in the rest of paragraph 6 about Spitzer’s retraction of his paper because of its methodological flaws. What is being implicitly suggested is that if the most notable paper on the subject has been retracted by its author then the case for conversion therapy must be very weak.

What the paper does not say is that Spitzer’s retraction of his paper is itself controversial. As Stanton Jones comments, there are two problems with this retraction.

‘It is important to note, first, that Spitzer has apologized for the study but his data remain. There is an enormous difference between changing one’s interpretation of scientific data, even through apology or recantation, and removing the data from the scientific canon by a retraction.

The editor of the Archives of Sexual Behavior, Kenneth Zucker, rightly noted this distinction in an interview with Psychology Today and published Spitzer’s letter under the title ‘Spitzer Reassesses His 2003 Study’ rather than retracting the study. He also made clear that Spitzer’s article underwent multiple rounds of careful, professional peer review, contrary to reports in the New York Times.

Second, the core of Spitzer’s change of heart: his loss of confidence that ‘the participants’ reports of change were credible and not self-deception or outright lying.’ Embellishing Spitzer’s words, the Times reporter Benedict Carey explains that ‘simply asking people whether they have changed is no evidence at all of real change. People lie, to themselves and others. They continually change their stories, to suit their needs and moods.’

Is self-report trustworthy? Much of psychology and medicine is premised on the validity of self-report. We depend on people to tell the truth about their depression, their headaches, their delusions, their nausea. Even some sophisticated brain-imaging studies depend on the validity of people’s self-report about what they are thinking as their brains are being imaged. It would be the foolish physician who forbade patients to speak and only looked for ‘verifiable measures.’

The critics themselves assume that self-reporting can be trusted. They accept without independent verification the self-reports of those who say they did not change their sexual orientation and those who claim they were harmed by the attempt.’[2]

It is also worth noting that several of the participants in Spitzer’s study subsequently affirmed their change of sexual orientation and vehemently protested against Spitzer’s repudiation of his own 2003 results.[3

Turning to what the paper says about the 2011 study by Jones and Yarhouse in paragraph 7, what we find is that the shortcomings of this study are emphasised. The paragraph says that it was ‘still based on self-report’ and that:

‘….once again the method of recruitment was flawed in that no participant was recruited before the therapy began and there was no control group who did not receive the therapy. Not only did the study fall far short of the standard of a randomised clinical trial but also the authors themselves say their method ‘fails to meet a number of ideal standards for longitudinal, prospective studies’ (page 408). ‘

These comments are misleading because, as we shall see, the study was not solely based on self- report and because they fail to explain why it is that, in spite of these shortcomings, the Jones and Yarhouse study may nevertheless be considered the most significant study of conversion therapy yet undertaken.

What this study did, and what no other study has done, was to undertake a ‘longitudinal study’ of people undergoing conversion therapy. That is to say, it tracked a group of ninety eight people for an extended period of time (six or even years) to see what changes, if any, occurred to them as a result of undergoing conversion therapy. Furthermore, it took great efforts not just to rely on people’s own reports but also to use proper objective psychological measures of mental health and harm at all points through their work to evaluate people’s progress through the course of therapy.

To quote Jones:

‘Most notably, instead of gathering recollections about change many years after the fact, a method regarded as vulnerable to self-deception and bias, we gathered data on individuals annually as they went through the process of attempting to change. In addition to administering the type of structured interview used by Spitzer, we repeatedly administered standardized psychological inventories and measures widely regarded as valid over the six- to seven-year period of the study, urging participants to tell the truth about their experience.’[4]

Paragraph 7 is also misleading about the findings of the study. It says the study found ‘small numbers of people who had experienced change in their sexual orientation, though the change was incremental rather than dramatic.’ This account differs markedly from Jones’ own account of the findings of the study. Jones reports:

‘About one third of the final participants abandoned their attempt to change, with many embracing a homosexual identity. About one third embraced sexual abstinence rather than a homosexual identity. About one fourth had moved away from a predominantly homosexual orientation and reported having satisfactory heterosexual relationships. One participant claimed at first to be a ‘success’ story but later repudiated his report and embraced a homosexual identity, while another had given up on change but later claimed to have successfully changed his orientation to heterosexuality.

In light of the common claim that attempts to change sexual orientation are often profoundly harmful, we administered at every assessment a standardized measure of psychological distress. Distress did not increase with continuing commitment to the process, and few subjects reported extreme levels of distress, suggesting that distress and harm are not inherent in the attempt to change sexual orientation. To be sure, harm may indeed occur when incompetent or inhumane methods are utilized, or when vulnerable minors are treated unprofessionally. On the whole, however, our evidence suggests that some people experience meaningful shifts in sexual orientation and that the attempt to change is not intrinsically or necessarily harmful.’[5]

Paragraph 7 is thus not a valid summary of the results of the Jones  and Yarhouse study and in particular it is silent about their conclusion that ‘distress and harm are not inherent in the attempt to change sexual orientation.’ The paper could have challenged the findings of the study if it wanted to, or pointed to others who had done so. What is not legitimate is its misreporting of the study’s findings. Given the emphasis in the paper on the evidence for conversion therapy doing harm its failure to report evidence to the contrary from the Jones and Yarhouse study is inexcusable.

Moving on to the Dehlin study[6] the paper acknowledges that it had some weaknesses, but argues that because of the large size of the sample (1612 participants) and the serious harm that was reported its reports of harm should be taken seriously.

What the paper does not acknowledge however, was that the large scale of the sample was achieved through the use of a one off web-based survey which relied entirely on what the participants said about their own past experiences and had none of the careful, objective, extended, real time evaluation of the participants’ condition found in the Stanton and Yarhouse study

In addition, as Peter Ould notes, this study suffers from the same problems of failure to assess evidence that were much criticised in an earlier study by Shidlo and Shroder:

‘… this study, though with a much larger sample, suffers from the same basic issue as Shidlo and Shroder in that the harm was never clinically assessed, was not compared to mental health states before therapy, attempts no quantitative analysis to detect a causal chain for ‘harm’ and also does not attempt to identify any other external mental health influencers (there is some evidence to show that the LDS population as a whole has higher mental issues such as depression than the overall population, chiefly because of the nature of LDS Church demands on people’s public and private lives – Jensen et al (1993), Idler et al (1998), Exline et al (2000) etc).’ [7]

Furthermore, as Ould also points out, the Dehlin study was based on what is known as ‘convenience sampling.’ That is to say, it basically invited anyone who wanted to report about their experiences to do so. This made for bias in the sample and, to quote Ould:

‘the authors identify this as the key limitation of their work on page 10, and indeed go as far as to say: ‘Our reliance on convenience sampling limits our ability to generalize our findings to the entire population of same-sex attracted current and former LDS church members.’’[8]

In other words, the authors of the report themselves acknowledge its inherent limitations, but this fact is totally ignored in the paper for Synod.

All these problems means that the Dehlin et al study is a much weaker study than the one by Jones and Yarhouse and yet this is the one on which the paper relies for its evidence of harm. The fundamental problem is that having a big study does not mean that the results tell you anything helpful if the methodology employed in that study is not robust.

What is more, you cannot simply say, as the paper does, that we have to take the reports of harm in the study seriously because of their ‘scale and nature’ if, as in this case, the nature of the study precluded the sort of detailed, objective, assessment of harm that the methodology employed by Jones and Yarhouse permitted.

It is true, as the paper says, that we should not dismiss the reports of harm ‘out of hand,’ but the paper should have acknowledged that it would be a mistake to give these reports the same weight as the findings of the Jones and Yarhouse study. If the result of that study is going to be challenged then much more robust evidence than that provided in the Dehlin study is required. [9]

Are people born gay and can their sexual desires change?

In paragraph 12 the paper makes five points about whether people are ‘born gay’ and whether their sexual desires can change.

Points i and ii argue that there is a strong case for some kind of biological causation for same-sex attraction and weak evidence ‘for all the main social cause theories.’ If what is being argued is that same-sex attraction is wholly biologically determined and therefore immutable, then this is an argument that is very widely rejected even by those who are supportive of same-sex relationships.

For example, the gay rights activist Peter Tatchell wrote back in 2008:

‘….an influence is not the same as a cause. Genes and hormones may predispose a person to one sexuality rather than another. But that’s all. Predisposition and determination are two different things. There is a major problem with gay gene theory, and withy all theories that posit the biological programming of sexual orientation. If heterosexuality and homosexuality, are, indeed, genetically predetermined (and therefore mutually exclusive and unchangeable), how do we explain bisexuality or people who suddenly in mid-life, switch from heterosexuality to homosexuality (or vice versa)? We can’t. The reality is that queer and straight desires are far more ambiguous, blurred and overlapping than any theory of genetic causality can allow.’ [10]

As Tatchell goes on to note, there is also the point that same-sex attraction is not uniform across cultures. In his words ‘both the incidence and expression of same-sex desire vary vastly between different societies’ whereas if such desire was biologically determined one would expect it to ‘appear in the same proportions, and in similar forms, in all cultures and epochs.’ [11]

Even if we cannot be precisely sure how biological predisposition and contingent social causation relate to each other, it seems likely that both are present in the formation of human sexual desires just as they are in all other areas of life. This means that saying someone is ‘born gay’ is misleading in the same way as it would be misleading to say that someone was ‘born alcoholic’, ‘born violent’ or ‘born altruistic.’

Points iii-v go on to argue that if it could be shown that sexual attraction was socially determined this would not mean that it is mutable, that the existence of sexual fluidity does not mean that sexual desire ‘can be manipulated by talking therapies’ and that our understandings of ‘the neural plasticity’ of sexual orientation are still in their infancy.

All these points are true, but all they mean is that you cannot determine a priori whether forms of conversion therapy are effective. They do not mean that it would necessarily be misleading to say that sexual desires can change. All they mean is that the relevant evidence would have to be produced to show that they can and that forms of conversion therapy can help in the process.

The paper of course, thinks that such evidence cannot be produced, but that is because it relies on the relatively poor quality study by Dehlin rather than the much better quality research by Jones and Yarhouse.


The paper claims that there is now ‘considerable evidence that conversion therapies are harmful.’ However, all the proof it offers for this claim is a single study the methodology of which makes the value of its findings doubtful. Meanwhile it conveniently buries the contrary evidence from another much more high quality study.

M B Davie 3.7.17

[1] Text of memorandum cited in GS 2070 A, General Synod, Private Member’s Motion, Conversion Therapy.

[2] Stanton L Jones, ‘Honest Sex Science,’ First Things, October 2012. Text at https://www.firstthings.com/article/2012/10/honest-sex-science

[3] Armelli JA, Moose EL, Paulk A, Phelan JE. A response to Spitzer’s (2012) reassessment of his 2003 study of reparative therapy of homosexuality. Archives of Sexual Behavior 2013;41:1335-1336. doi: 10.1007/s10508-012-0032-6.

[4] Jones, art.cit.

[5] Jones, art.cit.

[6] The paper notes that this paper was peer reviewed, but not that the same was also true of the Spitzer and

Jones and Yarhouse studies. Once again the paper withholds pertinent information.

[7] Peter Ould, ‘Do Sexual Orientation Change Efforts cause harm? Possibly, but….’ Psephizo, 1 July 2017, text at https://www.psephizo.com/sexuality-2/do-sexual-orientation-change-efforts-cause-harm-possibly-but/

[8] Ibid.

[9] For detailed criticisms of the Dehlin study and the way it has been used see ‘Psychotherapy for Unwanted Homosexual Attraction Among Youth,’ American College of Pediatricians, January 2016, text at https://www.acpeds.org/the-college-speaks/position-statements/sexuality-issues/psychotherapy-for-unwanted-homosexual-attraction-among-youth and Jacob Z Hezz , ‘How scientific research can becomeweaponized, ‘ at http://www.flirtingwithcuriosity.org/?p=1734

[10] Peter Tatchell, ‘Homosexuality: It isn’t natural,’ Spiked Online, 24 June 2008, text at http://www.spiked-online.com/newsite/article/5375#.WVpZ4tFK3IU

[11] Ibid.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s