Research and analysis

The prevalence of conversion therapy in the UK

Published 29 October 2021

1. Executive summary

1.1 Introduction

This report describes an approach to get a more accurate picture of the prevalence of conversion therapy in the UK. So-called ‘conversion therapies’ are:

techniques intended to change someone’s sexual orientation or gender identity. These techniques can take many forms and commonly range from pseudo-psychological treatments to spiritual counselling

(Government Equalities Office, 2018, page 83)

The best evidence available on the extent of conversion therapy in the UK comes from the National LGBT Survey from 2017.

It shows that:

  • 5% of respondents said they had been offered conversion in an attempt to “cure” them of being lesbian, gay, bisexual, and transgender (LGBT) in their lifetime
  • a further 2% said they had undergone conversion therapy
  • 4% of transgender respondents said they had undergone conversion therapy, and 8% reported having been offered it

1.2 Approach

The National LGBT Survey is a self-report survey of over 108,000 people who identify as LGBT and live in the UK.

The survey over-represents certain groups, such as younger people and those living in the South East of England, which means the findings may be subject to sample bias. Through this exploratory research, in an attempt to correct for this and to ensure the data collected would more accurately reflect the UK population as a whole, weighting was carried out on the conversion therapy variables.

1.3 Main findings

The results showed that the headline weighted percentages for these questions were not markedly different from the unweighted results.

  • 2.9% of people said they had received conversion therapy in their lifetime, compared with 2.4% for the unweighted responses – an increase of 0.5 percentage points, but also a roughly 20% higher estimate
  • 5.0% of people were offered conversion therapy in their lifetime, compared with 5.4% for the unweighted responses – a decrease of 0.4 percentage points, but also a 7% lower estimate

However, as the analysis became more detailed, for some measures or sub-samples – for example, people who had experienced conversion therapy by healthcare providers – the effect of the weighting was more marked, with slightly larger differences between the weighted and unweighted results.

1.4 Conclusion

The results showed that the weighted results for the survey questions on conversion therapy were not markedly different from the unweighted results. This is likely due to a combination of factors related to the sample size and composition, and the associated approach to weighting.

Surveying the LGBT population is a fundamentally challenging thing to do, which is why researchers often have to rely on self-selecting samples. Also, people who experience conversion therapy represent a small proportion of the overall population. This creates a substantial challenge in generating any robust estimates of prevalence, as any survey will need to cover a very large population in order to identify enough people who have experienced conversion therapy.

2. Introduction

In 2018, the government said it would explore options for ending conversion therapy in the UK. To support this, the Government Equalities Office (GEO) commissioned research, including a rapid evidence review, to improve understanding of the practice, experience and effect of conversion therapies.

Prevalence estimates are useful for identifying the presence of a characteristic or issue, to potentially to understand its extent, and whether action to mitigate it is improving or not. There are 3 measures of prevalence:

  • point prevalence – the proportion of the population experiencing the characteristic at a specific point in time
  • period prevalence – the proportion experiencing it at a given period (for example, in the last year)
  • lifetime prevalence – the proportion experiencing it over their lifetime up to the point of measurement – ‘prevalence’ in this report refers to this

There are challenges to measuring the prevalence of conversion therapy in the UK, largely due to its hidden and complex nature. An estimate of the full picture is complicated by the need for accurate and up-to-date information. Incidents of conversion therapy are not currently captured in any official statistics or through administrative data sources. Part of the reason for this is that it is not the responsibility of any agency or body to do so. For this reason, people reporting their experience of conversion therapy in a survey is likely to be the only source of quantifying its prevalence.

Conversion therapy generally targets LGBT people. There are also issues with surveying, as it is difficult to get a representative sample – there is currently no sample frame (a list of all LGBT people) to take a sample from. This is why researchers often have to rely on self-selecting samples.

People who experience conversion therapy also represent a small proportion of the LGBT population, which itself is a small proportion of the overall population. This creates a substantial challenge in generating any robust estimate of prevalence of conversion therapy, as any survey will need to cover a very large population in order to identify enough people who have experienced it.

2.1 Overview of the evidence

The review of international literature by Jowett and others (2021) found that “(t)here is little representative evidence regarding the prevalence of conversion therapies due to the hidden nature of the population and a reliance on self-selected samples” (p31).

As a result, it is difficult to estimate the true prevalence of conversion therapy among LGBT people in the general population. The evidence review by Jowett and others reported estimated levels of conversion therapy in 3 countries outside the UK.

Canada

  • 3.5% of over 8,000 male respondents, self-identifying as a sexual minority, said they had undergone conversion therapy over their lifetime (Salway and others, 2020)

China

  • 6% of over 15,500 LGB respondents using health services had been “recommended, coaxed into, or provided conversion therapy for sexual orientation, gender identity or gender expression” (Suen and Chan, 2020)

USA

3 different studies on the prevalence of conversion therapy among LGBT people found that:

  • 7% of around 1,500 male respondents self-identifying as a sexual minority said they had undergone conversion therapy (Blosnich and others, 2020) – this survey used random digit dialling and the estimate is more reliable than those used for self-report surveys
  • 18% of around 1,000 male respondents self-identifying as a sexual minority reported lifetime experiences of conversion therapy (Meanley and others, 2020)
  • 14% of 27,500 transgender people living in the USA reported receiving “attempts to change a person’s gender identity from transgender to cisgender” – 5% reported having received the intervention in the 5 years between 2010 and 2015 (Turban and others, 2019)

This evidence suggests that the proportion of people from LGBT populations outside the UK receiving conversion therapy can vary by country, with the figures ranging between 3.5% and 15%.

2.2 The prevalence of conversion therapy in the UK

There is no representative data on the extent of conversion therapy in the UK. The National LGBT Survey provides our best evidence of the proportion of LGBT people having been offered or receiving conversion therapy over their lifetime. Like the majority of the international studies, it uses a self-selecting sample.

The survey found that 2% of all respondents reported having undergone conversion therapy in an attempt to “cure” them of being LGBT in their lifetime, with 5% having been offered it.

2% of gay and lesbian respondents said they undergone conversion therapy, with 5% reporting they had been offered it. In comparison, 4% of transgender respondents said they had received it, and 8% said they had been offered it.

These findings, along with some of the available international evidence, suggest that transgender people are more likely to be offered and receive conversion therapy than non-transgender people. However, it is not possible to infer why this is the case. None of the surveys with transgender respondents asked whether respondents were offered or received conversion therapy because of their sexual orientation or gender identity. This can further obscure a fuller understanding of the true picture of the nature of conversion therapy.

3. Approach

3.1 Weighting the data from the National LGBT Survey

When surveying the LGBT population it can be difficult to find a nationally-representative sample. This means that surveys can be subject to sample bias. Large sample sizes can often help to mitigate this problem, as can weighting the sample to make it more representative of the population.

Younger people and those living in the South East of England are over-represented in the National LGBT Survey 2017 dataset. Weighting was carried out to help ensure the data would more accurately reflect the UK population as a whole.

A design weight for each respondent was calculated using the population totals and observed number of people in each region. This provided a measure of each respondent’s likelihood of being sampled. From this, a weighting figure was calculated using the nationally representative Annual Population Survey to calibrate the survey data.

By applying the weighting to the survey dataset, a revised estimate of the proportion of respondents who were either offered, or received conversion therapy was calculated. This weighted estimate provides a more accurate assessment of the extent of conversion therapy with a self-selecting sample of LGBT respondents than the previous figure.

4. Results

The new weighted percentages, original unweighted percentages, and percentage point difference between the 2 figures are shown below in the tables below.

4.1 People offered or who had received conversion therapy

Table 1 shows the percentage of respondents who said they had been offered or received conversion therapy.

Table 1: Whether respondents had been offered, or had received, so-called ‘conversion’ or ‘reparative’ therapy in attempt to ‘cure’ them of being LGBT

Respondents who said they had been offered or received conversion therapy Weighted
%
Unweighted % Percentage point change
Had it 2.9% 2.4% +0.5
Offered it 5.0% 5.4% -0.4
Neither had or been offered it 90.9% 90.7% +0.2
Don’t know if they had or been offered it 1.2% 1.4% -0.2
Respondents (rounded) 107,850 1,137,564  

Note: Excludes people who answered ‘Prefer not to say’. Respondents: All respondents.

The weighting did not markedly alter the headline statistics on the percentage of respondents saying they had been offered or received conversion therapy.

Table 1 shows that:

  • 2.9% of people said they had received conversion therapy in their lifetime, compared with 2.4% for the unweighted responses – an increase of 0.5 percentage points, but also a roughly 20% higher estimate
  • 5.0% of people were offered conversion therapy in their lifetime, compared with 5.4% for the unweighted responses – a decrease of 0.4 percentage points, but also a 7% lower estimate

4.2 Who offered and carried out conversion therapy

Table 2 shows the percentage of respondents saying who had offered them conversion therapy.

Table 2: Who offered the so-called ‘conversion’ or ‘reparative’ therapy?

Who offered the therapy Weighted
%
Unweighted % Percentage point change
Faith organisation or group 52.2% 53.4% -1.2
Parent, guardian or other family member 28.5% 29.9% -1.4
Person from my community 13.5% 15.9% -2.4
Healthcare provider or medical professional 11.9% 8.4% +3.5
Any other individual or organisation not listed above 18.5% 18.4% +0.1
Prefer not to say 2.2% 2.7% -0.5
Respondents (rounded) 5,810 57,106  

Note: Respondents could select as many people or organisations as applicable. Respondents: All respondents who had ever been offered ‘conversion’ or ‘reparative’ therapy but had not had it

Table 3 shows the proportion of respondents saying who had conducted conversion therapy on them.

Table 3: Who carried out the so-called ‘conversion’ or ‘reparative’ therapy?

Who offered the therapy Weighted
%
Unweighted % Percentage point change
Faith organisation or group 50.4% 50.7% -0.3
Healthcare provider or medical professional 28.7% 19.0% +9.7
Parent, guardian or other family member 12.6% 15.8% -3.2
Person from my community 8.6% 9.4% -0.8
Any other individual or organisation not listed above 12.0% 13.7% -1.7
Prefer not to say 8.0% 11.2% -3.2
Respondents (rounded) 2,640 32,573  

Note: Respondents could select as many people or organisations as applicable. Respondents: All respondents who had ever received ‘conversion’ or ‘reparative’ therapy

The weighted responses also showed that faith organisations or groups were still most likely to offer and carry out conversion therapy.

  • Table 2 shows that 52.2% of people who were offered conversion therapy in their lifetime said a faith organisation or group had offered it to them, compared with 53.4% for the unweighted responses – an decrease of 1.2 percentage points and relative difference of under 1%

  • Table 3 shows that 50.4% of people who received conversion therapy in their lifetime said a faith organisation or group had carried it out, compared with 50.7% for the unweighted responses – an decrease of 0.3 percentage points and relative decrease of around 2%

There were more marked differences between the weighted and unweighted responses for respondents who were offered and received conversion therapy by a healthcare provider or medical professional.

  • Table 2 shows that 11.9% of people who were offered conversion therapy in their lifetime said a healthcare provider or medical professional had offered it to them, compared with 8.4% for the unweighted responses – an increase of 3.5 percentage points and relative difference of roughly 40%

  • Table 3 shows that 28.7% of people who received conversion therapy in their lifetime said a healthcare provider or medical professional had carried it out, compared with 19% for the unweighted responses – an increase of 9.7 percentage points and relative difference of roughly 50%

5. Conclusion

This report has described an approach to weighting the data from the National LGBT Survey using the conversion therapy questions in an attempt to correct for potential sample bias. The results showed that the headline weighted percentages for these questions were not markedly different from the unweighted results. However as the analysis became more detailed, for some measures or sub-samples – for example, people who had experienced conversion therapy by healthcare providers – the effect of the weighting was more marked.

This is likely to result from:

  • the original composition of the sample
  • the weighting approach used, which attempted to correct for age and region
  • the smaller numbers of respondents involved, which are more subject to variation resulting from the weighting

6. Issues for consideration

While the weighted figures are more reflective of the UK population as a whole, the approach is limited by the variables which could be matched reliably to those available (in this case, only age and region). Without population totals for all demographics or categories it becomes harder to maintain control over the distribution.

This is exploratory analysis based on a single issue – conversion therapy – the weighting has not been applied to any other questions in the survey.

The National LGBT Survey 2017 asks the question “Have you ever had so-called ‘conversion’ or ‘reparative’ therapy in an attempt to ‘cure’ you of being LGBT?” and so respondents will have responded with their own view of what they considered to be conversion therapy.

The survey did not ask respondents when their experiences took place.

7. References

Blosnich JR, Henderson ER, Coulter RW, Goldbach JT, and Meyer IH (2020). ‘Sexual orientation change efforts, adverse childhood experiences, and suicide ideation and attempts among sexual minority adults’, United States, 2016–2018. American journal of public health, (0), e1-e7

Government Equalities Office (2018). ‘National LGBT Survey: Research report’. Government Equalities Office

Jowett A, Brady G, Goodman S, Pillinger C and Bradley (2021), ‘Conversion Therapy: An evidence assessment and qualitative study’. Government Equalities Office

Meanley S, Haberlen SA, Okafor CN, Brown A, Brennan-Ing M, Ware D, and Plankey MW (2020). ‘Lifetime exposure to conversion therapy and psychosocial health among midlife and older adult men who have sex with men’. The Gerontologist

Salway T, Ferlatte O, Gesink D, and Lachowsky NJ (2020). ‘Prevalence of exposure to sexual orientation change efforts and associated sociodemographic characteristics and psychosocial health outcomes among Canadian sexual minority men’. The Canadian Journal of Psychiatry, 0706743720902629

Suen, YT and Chan RCH (2020) ‘A nationwide cross-sectional study of 15,611 lesbian, gay and bisexual people in China: disclosure of sexual orientation and experiences of negative treatment in health care’. International Journal for Equity in Health, 19, 1-12

Turnban JL, King D, Reisner SL, and Keuroghlian AS (2019) ‘Psychological attempts to change a person’s gender identity from transgender to cisgender: estimated prevalence across US States, 2015’. American journal of public health, 109(10), 1452-1454