The Australian Bureau of Statistics (ABS) is exploring potential new topics to include in the 2021 Census of Population and Housing (‘Census’). They are also looking to enhance existing topics to make them more meaningful and reviewing the overall content to ensure it remains relevant. As part of this process, the ABS conducted a public consultation earlier this year. From the 445 submissions received, the ABS has created a short list of topics considered to have a strong or moderate case for change. While the options are still being explored, the inclusion of a question on sexual orientation would provide a depth of information not currently available at the population level. Questions do however, remain over the practicalities of including such a potentially divisive topic in the Census, as well as whether it is the best way to collect this information.

How has the ABS previously collected information on sexual orientation?

The ABS included a question on sexual orientation in the General Social Survey (GSS) for the first time in 2014. It is likely this information will also be collected in subsequent surveys.

The GSS was conducted throughout Australia from March to June 2014, with participants randomly selected from private households that contained people aged 15 years and over (excluding very remote areas and people living in discrete Aboriginal and Torres Strait Islander communities). The final count of dwellings was 12,932, representing an average response rate of about 80%.

How did the ABS collect information about a person’s sexual orientation?

From the selected dwellings, the ABS randomly selected one person aged 18 years or older to be asked the following question:

Q. Which of the following options best describes how you think of yourself?

A. Straight (Heterosexual); Gay or Lesbian; Bisexual; Other; or Don’t Know. The category of ‘Other’ includes people identifying as a sexual orientation other than heterosexual, gay, lesbian or bisexual.

The question was accompanied by a list of possible response categories (a ‘prompt card’), or the selected person could choose not to answer the question (i.e. ‘not stated’). The data is representative of approximately 17.6 million people. Of these, 160,500 people (almost one per cent) answered don’t know or did not provide a response.

What type of data is available from the GSS?

Table 18 of the survey release (via the ‘Downloads’ tab) provides broad level data on sexual orientation cross-referenced by selected other characteristics (for example, self-assessed health status, family support, community activity, discrimination, broad age group). The full suite of answers about sexual orientation has been condensed to: heterosexual; gay or lesbian; and other. In the published data, people identifying as bisexual have been combined into the ‘other’ category, due to the small number of people in this category and the quality of the data once it is split in various ways. Clients of the Parliamentary Library may request additional data; however, the release of very detailed data on sub-population groups is governed by the ABS and their guidelines to meet confidentiality requirements and benchmarks of data quality.

What can this existing information tell us?

In 2014, there were approximately 267,900 people who identified as gay or lesbian and 254,900 who were classified as other. Together, these people represented about three per cent of the population (excluding people whose response was unknown). Unpublished survey data provides a breakdown for the people grouped as ‘other’—157,500 identified as bisexual and 98,700 identified with some other sexual orientation.

The survey article, ‘How do Australians feel about their life as a whole?’ includes summary data from Table 18. The chart below provides selected indicators from the published data.

Selected social and economic indicators by sexual orientation, 2014

Source: ABS, General Social Survey, 2014, cat. no. 4159.0 (Table 18)

What would be different about the collection of data through the Census?

A major drawback of the GSS data is the unreliability of sub-state level data, as the survey is not designed for this type of analysis. The Census data set is large and provides information on people across the country that can be analysed for smaller regions.

The Census provides a rich picture of the Australian population and, because everyone is counted, data is able to be released on small populations within the community. The Census collects a range of personal and household characteristics that can be combined with other characteristics for a range of analyses. See examples in the above chart.

Some past Censuses have provided data on couples who lived together in households, but they have not provided an overall picture of sexual identity, missing people who were a couple but lived apart, those who were single and people living in group households. The existing data (‘same-sex couple indicator’) is created by using people’s relationships within a household and their selection of male/female in the sex question. The changing nature of families and relationships could be better captured by the Census with the inclusion of a question on sexual orientation.

What are the possible drawbacks of using the Census?

The ABS is aware of the potentially sensitive nature of including a question on sexual orientation in the next Census. They are working through how this information could be discreetly collected from each person in a household, and what limitations would need to be included (for example, only asked of people aged 18 years and over). For example, if one household member had not made their sexual orientation known to other household members, this may result in a higher ‘not stated’/’don’t know’ count (the member doesn’t want others to know so they don’t state anything) or inaccurate recording (the person completing the form assumes they know the other household members’ sexual orientations). The ABS also expects the question may raise the opportunity for people to deliberately record false answers, for fun, such as has happened with ‘Jedi’ being recorded as a ‘religion’.

What’s happening next in the development of the 2021 Census?

The ABS will provide additional detail on the development of the Census in a formal publication on 14 November 2018, see: Census of Population and Housing–2021 Topic Directions.