Producing modelled estimates of the size of the lesbian, gay and bisexual (LGB) population of England
Public Health England (PHE) is committed to understanding the health and wellbeing needs of different population groups and reducing health inequalities. In addition, PHE has a duty, as part of the Equality Duty, to consider the needs of all individuals in their day to day work in shaping policy and in delivering services. The Duty requires PHE to have due regard for the need to reduce discrimination, advance equality of opportunity and foster good relations between those who share protected characteristics and those who do not. Sexual orientation is a protected characteristic under the Equality Duty.
There is evidence to suggest that minority sexual orientation groups can experience high prevalence of poor mental health and low wellbeing. In order to assess health outcomes by sexual orientation, there is a need for a widely accepted national estimate of the size of the lesbian, gay and bisexual (LGB) population in England. This is currently unavailable. Data on sexual orientation are collected as part of a number of national general social surveys. However, obtaining robust estimates is problematic. The proportion of people self-identifying as LGB is influenced not only (substantively) by underlying cultural factors and stigma associated with responses and legality, but also methodologically by survey population, sample size and survey methodology. So far, no studies have used a systematic approach to identify and synthesise all relevant existing surveys and measurement tools to produce an aggregated estimate of the LGB population for England.
The objectives of the project were to locate and critically review existing research, surveys and measurement tools that estimated the size of the LGB population of England. A methodology is proposed for accurately and robustly estimating the size of this population, seeking to reduce error inherent in the reviewed surveys. The weighted average of the national survey data results in a synthesized estimate of the LGB population of England, broken down by socio-demographic and geographic variables. This report is intended for all organisations that have an interest in monitoring outcomes for LGB populations and require a population denominator. It is specifically focused on the LGB population and further work is needed to include transgender and intersex people.
Sexual orientation is a combination of a person’s sexual attraction, behaviour and self-identity. The ways in which these factors are measured in social surveys differs greatly depending on the purpose of the survey. The three concepts may refer to distinct but overlapping groups about which estimates can be derived separately, but combinations are inherently complex. This project limits its estimates to self-reported sexual identity as this is considered to provide the most robust estimate of the LGB population, thereby recognising that this may potentially underestimate the size of this particular population. We were guided on both substantive issues and project methodology and outcomes by an Expert advisory group established for this project.
Informed by formal systematic review methods, a search was performed using standardised search terms to trawl relevant databases and grey sources. A total of 22 relevant national surveys were identified which met the inclusion criteria and included a sexual orientation question (no date restriction was applied). Individual survey estimates for people that identified themselves as lesbian, gay, bisexual or ‘other’ ranged from 0.90% [95%CI 0.40, 1.83] to 5.52% [95%CI 4.63, 6.56]. An exploration of methodological approaches led to the conclusion that surveys with study populations likely to be biased in terms of LGB prevalence could not be included into a general pooled estimate and that weights should be applied in the final synthesis based on sample size, response rate and proportion of missing data.
Thus the final synthesis includes 15 surveys that represent the general adult population of England. The weighted estimate of people identifying as LGB or ‘other’ is 2.5%. If this were applied to the census-based mid-2014 population estimate, this suggests that there would be 1,358,848 people identifying as LGB or ‘other’ in England. If people who responded ‘prefer not to say’, ‘don’t know’ or gave no answer (ranging from 0-10% for individual surveys) are assumed to all be LGB, then the upper limit becomes 5.89%. Stratification by age, gender and ethnicity (based on the distribution of the broadest survey; the GP Patient Survey 2015), indicates that the proportion of self-identified LGB is highest in men, younger age groups, and mixed/multiple or other ethnic groups. The GP Patient Survey provides local authority data for Greater London, Greater Manchester and Brighton and Hove which had an overall LGB prevalence of 5.1%, 3.6% and 9.9% respectively.
This study is the first to systematically review existing sources and synthesize data into a new robust estimate of the size of the LGB population of England. However, these results should be interpreted with caution. Not only is the synthesised result sensitive to error in each survey, but 2.5% is likely to be an underestimate as it is derived from general social surveys that did not have the specific aim of enumerating the LGB population. The upper limit of 5.89%, on the other hand, is almost certainly an overestimate of the included sources as it is unlikely that all non-responders are LGB. This project did not aim to qualitatively assess the propensity to under or misreport sexual identity.
Furthermore, it should be noted that any national estimate is likely to conceal significant sub-national variation. As this study has shown, LGB prevalence in the areas of Greater London, Greater Manchester and Brighton and Hove are higher than the mean national estimate found in this study. Future studies should explore more local surveys using the methodology proposed in this study to synthesise results. In addition, future research could explore the impact of the mode of administration and context in which the sexual orientation question is posed on LGB responses. Finally, more work could be done to clarify who constitutes the group of ‘others’.