Sexual identity workshop: summary of discussions

Our 2021 Census topic consultation identified a clear need among data users for improved information on sexual identity, especially in relation to policy development, service provision and fulfilling duties under the Equality Act 2010. As a result, we have published a Sexual identity research and testing plan, which sets out the work we will do to help us determine whether, and how best, to meet user needs for information on this topic. These plans build on the previous work of the Sexual Identity Project between 2006 and 2009. The plans include engaging with relevant stakeholders, including representatives of lesbian, gay, bisexual plus (LGB+) groups.

A summary of the workshop on gender identity held on the same day is also available.

Purpose of the sexual identity workshop

The workshop was designed to provide an opportunity for (potential) sexual identity data users and our own staff to explore data needs together. The workshop also provided the opportunity for participants to understand and reflect on our research and testing plans for the sexual identity topic.

Participants were asked to discuss in groups their data needs and the impact and benefits that such data could have. They also considered factors that may affect data quality and their expectations of data quality in relation to the potential sensitivity of the sexual identity topic. We presented our Sexual identity research and testing plan and asked for feedback on this.

2.1 Participants

Workshop participants were primarily identified through responses to the public 2021 Census topic consultation, which anyone could take part in. We aimed to include a good range of relevant parties. The workshop was attended by representatives from:

• national health organisations

• special interest groups representing LGB+ identities (for example, Stonewall and AVEN UK National Asexuality Network)

• organisations with an equalities focus, including the Equalities and Human Rights Commission (EHRC) and Government Equalities Office (GEO)

• academics and researchers

• the UK Statistics Offices: Office for National Statistics (ONS), National Records of Scotland (NRS) and the Northern Ireland Statistics and Research Agency (NISRA)

Discussions confirmed that data are required for service provision and equalities monitoring, particularly given provisions in the Equality Act 2010. There is an overarching need for a reliable estimate of the population identifying as a sexual minority, as well as a need for local level estimates and multivariate analysis.

It was felt local level data would:

• increase understanding of who lives in a local area and their characteristics of sexual identity, specifically for local authority and clinical commissioning group area

• inform local needs assessments and planning for delivery of appropriate services, which can be targeted more effectively (for example, sexual health services)

• provide evidence to support the need for provision of services (for example, local authority youth services) and provide visibility of LGB+ population to service providers and policy-makers

A reliable estimate would:

• provide quantitative evidence, for example, of differential outcomes for minority groups, which engages policy and decision-makers and helps ensure alignment with the Equality Act 2010

• improve ability to more accurately cost services based on numbers requiring them

• help manage programmes and save resources in ways that do not disproportionately impact on LGB+ groups

Ability to carry out multivariate analysis would:

• provide evidence of any correlation between sexual identity and other characteristics; for example, knowing about the interaction of identity with ethnicity is important for healthcare provision

The discussions highlighted a strong need for a reliable estimate of the non-heterosexual population, with some demand for further breakdown of identities within the non-heterosexual population beyond the existing “Gay or Lesbian”, “Bisexual” and “Other” categories in the National harmonised standard, some of which could be achieved by including a question on gender identity.

Reasons for further breakdown included improving visibility of different identities and identifying sub-groups to facilitate service provision. Examples of identity categories suggested by participants include:

• asexual

• distinguishing gay men, gay women, and lesbian (women) as different identities

• response options that do not assume gender; that is, options for individuals who do not identify as male or female to provide their sexual identity, some of which could be achieved by including a question on gender identity

• broadening the existing response options to be more encompassing of the wider spectrum of identities

Discussions often focused on the ability of the current National harmonised standard to meet needs. There was a general feeling that the current estimates available from ONS household surveys, determined using the harmonised question, do not adequately represent the size of the LGB+ population. It was acknowledged that providing individual response options for detailed and evolving sexual identities is not appropriate.

Generally, participants were supportive of the self-response version of the harmonised standard with an additional write-in “Other” response option, although there was some concern over the difficulty in analysing write-in responses. Some workshop participants set out an argument that the harmonised question would benefit from a more nuanced set of response options, for example, allowing respondents to identify as “mostly heterosexual or straight” or “completely heterosexual or straight”.

It was clear that not all data needs can be met using one measure. A measure of sexual identity is most relevant for meeting discrimination and equalities-based needs. Some organisations expressed a need for information about sexual behaviour (for example, for targeting sexual health services) and sexual attraction. As the census is a mandatory household survey for the general population, it is not considered a suitable vehicle for collecting information about sexual behaviour because all questions included must be publicly acceptable.

Discussions on data quality helped participants to identify and appreciate factors which may affect the quality of any data collected on sexual identity, as well as the complexity of question and output design to provide meaningful information to users.

Of high importance to participants was the effectiveness of the question and, in particular, whether the question captured the concept to be measured. Privacy and confidentiality were also considered important in terms of achieving suitable data quality. Given the census can be completed by 1 householder on behalf of others within a household, participants felt it was important that respondents have suitable options for responding privately and have confidence their response will remain confidential.

Other factors participants recognised as affecting data quality, but considered of less importance, included:

• response rates achieved

• location of the question within the questionnaire

• ordering of response categories

• contextual or background information available to respondents

• extent of outputs produced from data collected

Participants were given the opportunity to reflect on some of the complexities around the potential sensitivity of asking a sexual identity question in a household survey context, by completing a pilot public acceptability questionnaire.

Research and testing plans

We presented the Sexual identity research and testing plan, which includes:

• continued production of sexual identity estimates from ONS social surveys and work to examine the feasibility of producing local authority level estimates from these survey results

• a large scale split-sample test in 2017 covering about 100,000 households; the main aim of the test is to assess the impact on overall response of including a sexual identity question

• qualitative research (focus groups and cognitive interviews) to inform design (wording and administrative design) for a sexual identity question to be used in the 2017 Test; this research includes reviewing the national harmonised standard for self-completion

• further understanding the public acceptability of asking a question on sexual identity in the census by commissioning an independent quantitative survey to complement results from the 2017 Test

• review of existing approaches of collection of data about sexual identity in the UK and internationally

Participants provided positive feedback on the research and testing plans. There was support for all aspects of the plans, particularly the need to understand the effect on overall non-response. Feedback and suggestions included:

• the value in ensuring data are comparable and harmonised standards are used

• suggestions to emphasise research on sexual minorities beyond those currently included in the harmonised question to better understand the response option needs of these groups

• exploring potentially useful partnerships; for example, linking up with NHS Sexual Orientation Monitoring work and getting LGBT advocacy groups on-board to help with publicity around fieldwork

• our efforts to develop and improve sexual identity information available from social surveys were supported – since the workshops, we have published the 2015 Experimental Official Statistics on sexual identity in the UK – sexual identity estimates are based on social survey data from the Annual Population Survey (APS); the Sexual Identity Quality and Methodology Information document contains important information on the quality of this output

• widespread support for exploring the identities of respondents identifying with the “Other” response category and suggestions to further explore terminology and response categories

• understanding non-response through follow-up surveys

The main outcomes from the workshop were:

• sexual identity information needs were further clarified, building on our understanding of user needs following the 2021 Census topic consultation; it was recognised that the census was not a suitable vehicle for meeting needs for information about sexual behaviour

• we received valuable feedback indicating there are no significant gaps in the sexual identity research and testing plans beyond those mentioned above

• there was generally support for the self-response version of the harmonised question with an additional write-in “Other” response option or with the inclusion of a more nuanced set of response options

• participants have a better understanding of the factors we are considering as we determine whether and how to meet sexual identity information needs through the 2021 Census

• we identified opportunities for collaboration as we prepare for the 2017 Test and our ongoing work

We received positive feedback from workshop participants. Participants felt the workshop enabled them to share views and requirements for sexual identity data and that they were better informed about our research and testing plans on the sexual identity topic. In light of comments from participants, a list of those who attended the workshop has been included as part of this note.

Next steps

Results from the 2017 Test, public acceptability research and other testing will inform a recommendation on whether or not to include a question on sexual identity in the 2021 Census. Our research timetable is determined by the legislative framework that governs the census. Preparation for the Census White Paper in 2017 is the first main milestone in the legislative process. It provides the opportunity for informed public and Parliamentary debate before further legislative arrangements are taken forward.

We plan to keep stakeholders, including all participants who attended the August 2016 workshop, informed by sharing findings from research and testing. Depending on whether sexual identity is recommended for inclusion in the 2021 Census, further testing and development of the question and the administrative context will be undertaken.

We will also follow up with individual organisations on specific issues, particularly where opportunities for collaboration have been identified.