Australia's LGBTIQ Research Data Landscape
The Committee on the Status of LGBTQ + Individuals in the Economics Profession (CSQIEP) of the American Economics Association (AEA) awarded Saxby, De New and Petrie ( 2020) the inaugural 2020/2021 AEA CSQIEP @LGBTQ_Econ research award for outstanding contribution to the economics of LGBTQ + people.
Abstract
The past decades have seen rapid changes in societal attitudes and policy surrounding members of the LGBTIQ community, however this community continues to experience adverse social and economic outcomes relative to their cisgendered heterosexual counterparts. Understanding the LGBTIQ data landscape is essential to promote scholarship in this space and highlight avenues for future data collection. Here I provide an overview of the 17 national data surveys in Australia that capture information on either sexual or gender diversity and are either repeated or longitudinal samples. For each dataset, I provide details on questions pertaining to sexual and gender diversity, indicative sample sizes and age ranges, scope of survey, and data custodians. Altogether, I find thatcollection of information on sexual and gender diverse (SGD) populations in Australia is poor and terminology varies widely. Only seven datasets capture information on both sexual and gender diversity and, of these, none are nationally representative, and only one is a longitudinal sample. Altogether, this review highlights the need for large-scale, representative and longitudinal data capture of SGD populations in Australia.
1 Introduction
While the past decades have seen tremendous improvements in the rights and visibility of sexual and gender diverse (SGD) populations in the developed world, these populations continue to experience adverse social, economic and health outcomes relative to their cisgenderedheterosexual counterparts (Hillier et al. 2003; Mayer et al. 2008; Badgett 2009; Mcnair, Szalacha and Hughes et al. 2011; Grulich et al. 2014; Denier and Waite 2016; Dempsey et al. 2020). To provide ongoing monitoring of these inequalities, there is a pressing need for robust and high-quality data that captures information on SGD populations alongside key demographic, health, social and economic dimensions. Understanding the extent of data capture across different domains and subgroups within the lesbian, gay, bisexual, trans and gender diverse, intersex or queer (LGBTIQ) community is vital to inform future scholarship and data collection. To this end, here I provide an overview of Australia's national LGBTIQ data landscape, including information on the approaches taken to capture gender and/or sexual diversity and indicative sample sizes for each dataset, and provide recommendations for advancing scholarship in this space.
1.1 Key Definitions and Considerations
Language around sexual and gender diversity has been rapidly changing over the last several decades. First, it is important to note that until recently the term gender was often, and still sometimes is, used interchangeably with sex. However, they are distinct concepts (Diamond 2002). Sex is defined as biological sex assigned at birth (that is, male, female, or intersex) and is based on the physical features or sex characteristics (including genitals, gonads and chromosome patterns) one is born with (Hill et al. 2021a). The term intersex is used to describe people born with sex characteristics that do not fit the ‘typical’ binary notions of male or female (Hill et al. 2021a). Gender can be thought of as a social and cultural construct used to classify individuals as male or female (Diamond 2002). Gender identity therefore recognises an individual's conception of their gender and how they express this to others; for example, being assigned female at birth, but identifying as a male, or, not identifying with any gender. Cisgendered individuals are those whose gender identity matches their sex assigned at birth. Individuals may also use terms that capture a range of gender identities such as gender non-conforming, gender non-binary, genderqueer, or gender diverse. These individuals may identify more closely with masculinity or feminity, with both, or with neither. As umbrella terms, transgender or trans are alsoused to describe people who have gender identities, expressions, or behaviours that differ from their birth sex. Gender identity does not predicate any specific sexual orientation. Conceptualisation of sexuality is generally captured through three different domains: behaviour, attraction, and identity (Mishel 2019; Laumann et al. 2000). Sexual behaviour corresponds to the gender of individuals' sexual partners, sexual attraction describes who individuals are attracted to. Sexual identity describes how individuals expresses themselves sexually or romantically (Mishel 2019; Lyons et al. 2021) and includes diverse labels such as lesbian, gay, bisexual, queer, pansexual and asexual. These terms will be used throughout the review.
2 Methods
In order to catalogue datasets that include information on SGD individuals, I first conducted a ‘grey’ literature review.1 I reviewed recent reports and publications (including by government agencies, research centres, and researchers), which had sought to quantify the number of people who identify as LGBTIQ, and their associated reference lists (Qu, Knight and Higgins 2016; ABS 2018; AIHW 2018; Wilson and Shalley 2018; Callander et al. 2020; Carman et al. 2020; Wilson et al. 2020; Wilson, Temple and Lyons 2021). I next reviewed datasets operated and/or funded by ACON, Australia's largest not-for-profit LGBTIQ community health organisation, together with their research partners: the Centre for Social Research in Health, University of New South Wales (UNSW); the Kirby Institute, UNSW; the Australian Research Centre in Sex, Health and Society, La Trobe University; the Centre for Values, Ethics and the Law in Medicine, University of Sydney; the National Drug and Alcohol Research Centre; and Gay and Lesbian Health Victoria (ACON 2021). Finally, I reviewed all datasets currently provided, or available upon request, through government agencies. This included the Australian Bureau of Statistics (via Microdata download or the DataLab), the Australian Institute of Health and Welfare, the Australian Institute of Family Studies, and the Australian Government Department of Social Services (via the National Centre for Longitudinal Data). Datasets that were non-repeated (that is, datasets that were neither longitudinal nor repeated cross-sectional samples), did not recruit participants from across all of Australia, as well as specific subpopulation datasets (for example, rural only respondents) were not included. I then reviewed all of the data dictionaries and/or original survey questionnaires for each datasets' inclusion of questions pertaining to either sexual or gender diversity; including same-sex relationship status, sexual identity, sexual attraction, sexual experiences, intersex status, and gender identity. Where data dictionaries or survey questionnaires were not made available, I sourced information from publications that had recently used the data, or other reports that provided summary statistics by sexuality and/or gender identity. I then summarised each of the included datasets in two formats. First, I summarised the stated intent of the survey, whether it was representative or longitudinal, and whether it included questions on sexual and/or gender diversity. I then provided details on sample size, age range, and data custodian/s, as well as the specific response options available for questions on sexual and/or gender diversity.
2.1 Overview of Datasets Containing Information on SGD Populations
The review process identified 17 national datasets that included at least one question relating to either sexual or gender diversity. An overview of the stated intent, representativeness, longitudinal status, and domain of sexual and/or gender diversity captured for each dataset is provided in Table 1. Detail on the dataset sample size, age range, data custodian/s, and specific questions posed relating to sexual and/or gender diversity is provided in Table 2.
| Dataset | Frequency of collection and stated intent | Representative | Longitudinal | Sexual identity | Sexual attraction | Sexual behavioura | Gender identity |
|---|---|---|---|---|---|---|---|
| Australian Longitudinal Study on Women's Health | Collected every 1–4 years, depending on the cohort, the longitudinal cohort study aims to understand factors contributing to the health and wellbeing of women over life course | ✓ | ✓ | ✓ | ✓ | ✓ | – |
| Australian Study of Health and Relationships | Collected every 10 years since 2001–2002, this survey aims to provide a snapshot of sexual health and wellbeing of the Australian population | ✓ | – | ✓ | ✓ | ✓ | – |
| Australian Trans and Gender Diverse Sexual Health Survey | Conducted in 2018, the aim of the survey is to collect information on sexual health and wellbeing of Australia's trans and gender diverse communities. It is uncertain when this will be repeated. | – | – | ✓ | – | ✓ | ✓ |
| Australian Workplace Equality Index Employee Survey | Collected annually since 2011, the surveys interview employees at different companies and provide national benchmarking on LGBTQ workplace inclusion | – | –b | ✓ | – | – | ✓ |
| Census | Carried out every 6 years, the Census provides the official count of population and dwellings and collects details of age, sex and other characteristics of the population | ✓ | –c | ✓* | – | – | – |
| Following Lives Undergoing Change | Biannual questionnaires since 2014 with the aim to monitor sexual behaviour, drug use, health-seeking and social connectedness among gay and bisexual men | – | ✓ | ✓ | – | – | ✓ |
| Gay Community Periodic Surveys | Annual surveys that serve as a behavioural surveillance system for HIV, monitoring sexual practices, drug use and patterns of testing for HIV and other STIs among gay and bisexual men | – | – | ✓ | – | – | ✓ |
| General Social Survey | Carried out every 4 years since 2002 and annually since 2019, the surveys aim to collect information on social characteristics, wellbeing and experiences of people in Australia | ✓ | – | ✓ | – | – | – |
| Household, Income and Labour Dynamics in Australia | Conducted annually since 2001, these household surveys explore economic and personal wellbeing, labour market dynamics and family life over the life course | ✓ | ✓ | ✓ | – | – | –d |
| Journeys Home | Conducted biannually between 2011 and 2014, this cohort study aimed to explore factors affecting housing stability | ✓ | ✓ | ✓ | – | – | – |
| Longitudinal Study of Australian Children | Biennial surveys conducted since 2003 that explore children's social, economic and cultural environments on their wellbeing over the life course | ✓ | ✓ | – | ✓ | – | – |
| National Drug Strategy Household Survey | Conducted every 3 years to provide a snapshot of alcohol and tobacco consumption, and illicit drug use among the general population | ✓ | – | ✓ | – | – | – |
National Survey of Mental Health and Well-being |
These have been conducted in 1997 and 2007 and 2020. These targeted mental health epidemiological surveys aim to estimate the prevalence of mental disorders | ✓ | – | ✓ | – | – | – |
| National Secondary Students and Sexual Health survey | Surveys are conducted every 5 years since 1992 with the aim to gather information on HIV and sexually transmissible infection knowledge, as well as sexual risk behaviour among secondary school students | –e | – | ✓ | ✓ | ✓ | ✓ |
| Private Lives | Conducted in 2005, 2011 and 2019, private lives aims to understand the health and wellbeing of LGBTIQ Australians | – | – | ✓ | ✓ | – | ✓ |
| Ten to Men | Ten to Men has collects information about male health and its key determinants and has conducted the cohort surveys in 2013–2014, 2015–2016 and 2020. The timing for the next wave of data collection is yet to be confirmed | ✓ | ✓ | ✓ | ✓ | ✓ | – |
| Writing Themselves In | Writing Themselves In surveys have been conducted in 1998, 2004, 2010 and 2019 and collect information on experiences with education, homelessness, harassment, assault, mental health and community connections among LGBTIQ young people | – | – | ✓ | ✓ | ✓ | ✓ |
- a This refers to sexual behaviour as a mechanism through which to capture sexual diversity and refers to questions pertaining to gender of sexual partners.
- b These are repeated cross-sectional surveys, but are longitudinal at the employer/company level.
- c The Australian Census Longitudinal Dataset combines a 5 per cent sample of data from the 2006, 2011 and 2016 Censuses (Chipperfield, Brown and Watson 2017).
- d Information on gender identity is likely to be available in upcoming waves.
- e While the custodians do not explicitly note that the sample is representative of all high school students in Australia, they note the samples closely aligns with the Census (Fisher et al. 2019).
| Dataset, sample size (year), age range, custodian | Sexual diversity | Gender diversity | SGD sample size |
|---|---|---|---|
Australian Longitudinal Study on Women's Health (Hillier et al. 2003; Loxton et al. 2018; Australian Longitudinal Study on Women's Health 2020; Campbell, Perales and Baxter 2020; Perales et al. 2021b) n = 13,714 for 1946–1951 cohort (1996) n = 14,247 for 1973–1978 cohort (1996) n = 17,012 for 1989–1995 cohort (2012) 18 years and above Jointly managed by the University of Queensland and the University of Newcastle |
Sexual identity: • exclusively heterosexual • mainly heterosexual • bisexual • mainly homosexual (lesbian) • exclusively homosexual (lesbian) • other NB: 1946–1951 cohort asked in 2001, 1973–1978 cohort asked in 2000, 2003 and 2012, 1989–1995 cohort asked in 2013, 2014, 2015, 2017. Response option ‘other’ included in 2017 for 1989–1995 cohort only. Information based on most recent publication (Perales et al. 2021b) Sexual attraction: • only to females • more often to females; • equally to both • more often to males • only to males • never to anyone NB: asked of 1989–1995 cohort only. Sexual experiences: • only with females • more often with females • equally with both • more often with males • only with males • no experience NB: asked of 1989–1995 cohort only. |
NR | In Wave 3 (2003) for the 1989–1995 cohort, 8,850 answered the sexual identity question. 568 (6.4%) identified as mainly heterosexual, 100 (1.1%) identified as bisexual, and 99 (1.1%) identified as mainly lesbian or exclusively lesbian (Szalacha et al. 2017) |
Australian Study of Health and Relationships (Richters et al. 2014) n = 20,094 (2012) 16–69 years Partnership of custodians including: La Trobe University; the University of New South Wales; the University of Sydney; and the University of Sussex. |
Sexual identity: • heterosexual or straight • homosexual (gay) • bisexual • queer • not sure; undecided • something else/other Sexual attraction: I have felt sexually attracted only to… • females, never to males• more often to females, and at least once to a male• about equally often to females and to males• more often to males, and at least once to a female• only to males, never to females• I have never felt sexually attracted to anyone at all Sexual experience: • I have had sexual experiences only with females, never with males• more often with females, and at least once with a male• about equally often with females and with males • more often with males, and at least once with a female • only with males, never with females• I have never had any sexual experience with anyone at all. |
NR | Among men and women respectively, 1.9% and 1.2% identified as homosexual while 1.3% and 2.2% of the sample identified as bisexual (Richters et al. 2014) |
Australian Trans and Gender Diverse Sexual Health Survey (Callander et al. 2021; Rosenberg et al. 2021) n = 1,613 (2018) 16–80 years The Kirby Institute, University of New South Wales |
Sexual identity: Participants could describe their sexual orientation with free text. Sexual orientation groups were aggregated in the survey report as: • queer • pansexual • bisexual • homosexual • asexual • heterosexual • gynosexual • androsexual • no label Sexual experience: Participants can report gender of previous sexual and/or romantic relationships. Categorises are non-exclusive and these are aggregated in the report as: • trans men • trans women • cisgender men • cisgender women • non-binary people |
Gender identity: • trans man • trans woman • non-binary, assigned male at birth • non-binary, assigned female at birth |
Whole dataset: 1613 trans or gender diverse individuals in 2018 |
Australian Workplace Equality Index Employee Survey (ACON 2021; Perales et al. 2021a) n = 26,700 (2020) All ages of employed individuals NB: provided in six categories from ‘< 24 years’ to ‘65 + years’ ACON |
Sexual/gender identity: • cisgender heterosexual man • cisgender heterosexual woman • cisgender non-heterosexual man • cisgender non-heterosexual woman • trans man • trans woman • trans non-binary (assigned male) • trans non-binary (assigned female) |
Sexual/gender identity: • cisgender heterosexual man • cisgender heterosexual woman • cisgender non-heterosexual man • cisgender non-heterosexual woman • trans man • trans woman • trans non-binary (assigned male) • trans non-binary (assigned female) |
In 2020, the sample included 2606 (9.7%) cisgender non-heterosexual men 2238 (8.4%) cisgender non-heterosexual women, 90 (0.3%) trans men, 114 (0.4%) trans woman, 73 (0.3%) trans non-binary (assigned male at birth), 139 (0.5%) trans non-binary (assigned female at birth) (Perales et al. 2021a) NB: Sample with non-missing responses to the sexual/gender identity question. |
Census (ABS 2017, 2022;Saxby, De New and Petrie 2020) 24 million (2016) All ages Australian Bureau of Statistics NB: Other Census-linked products available through the Multi-Agency Data Linkage Product (MADIP) have a partnership of custodians including with the Australian Taxation Office; the Department of Education, Skills and Employment; the Department of Health; the Department of Social Services; Services Australia |
No explicit question but individuals in ‘same-sex relationships’ can be identified (if they living together at time of Census) based on their responses to questions on relationship and sex | In the 2016 Census, the ABS first included a response option of ‘other’ for the question on sex in the online form and a descriptive term on the paper form (ABS 2018). Although this information was collected, no studies have been published using this 2016 data and it remains unclear whether this information will be made available in the future. In 2021, a similar option was made available such that respondents could select ‘non-binary sex’ | In the 2016 Census, there were 41,265 in same-sex female relationships and 42,254 in same-sex male relationships (Saxby, De New and Petrie 2020). In the 2016 Census, 1260 people responded as ‘other’ or otherwise provided a descriptive term indicating they were sex and/or gender diverse (ABS 2018) |
Following Lives Undergoing Change (Hammoud et al. 2017; Clackett et al. 2018) n = 2,250 16 years and above The Kirby Institute, University of New South Wales |
Sexual identity: • gay • bisexual • other |
Gender identity/intersex variation: • male • transgender male • intersex |
Whole sample. Flux initially recruited 2250 GBM in 2014–2015, with 17 individuals identifying as either transgender or intersex (Hammoud et al. 2017) |
Gay Community Periodic Surveys (Zablotska et al. 2011; Broady et al. 2019; Saxby, Chan and Bavinton 2021) n = 9019 (2019) 16 years and above The Kirby Institute, University of New South Wales |
Sexual identity: • gay • bisexual • heterosexual • other |
Gender identity: • male • female • non-binary • other Gender assigned at birth: • male • female |
Whole sample. In 2019, there were 9019 respondents (Saxby, Chan and Bavinton 2021) |
General Social Survey (Bastos, Harnois and Paradies 2018; Vandenbroek 2018; ABS 2020) n = 5,300 households (2020) 15 years and above Australian Bureau of Statistics |
Sexual identity: • straight (heterosexual) • gay or lesbian • bisexual • other NB: included since 2014 |
NR | In 2020, 4% of GSS respondents described themselves as gay, lesbian or bisexual (ABS 2020) |
Household, Income and Labour Dynamics in Australia(Department of Social Services 2018; Perales and Todd 2018) n = 17,694 (2016) All ages The Department of Social Services and the Melbourne Institute of Applied Economic and Social Research, University of Melbourne |
Sexual identity: • heterosexual or straight • gay or lesbian • bisexual • other • unsure/don't know NB: Information on sexual identity was first collected in Wave 12 and has been asked again in Wave 16 |
Gender identity: NR, however the survey is likely to update their questions pertaining to gender diversity in upcoming waves Intersex variation: Although not specific to intersex variation, in Wave 20, the question on ‘sex’ allowed for response options: • male • female • other |
In Wave 16, 2016, there were 554 (3.1%) respondents who self-identified as gay, lesbian or bisexual (Perales and Todd 2018) |
Journeys Home(Andrew Bevitt et al. 2014; Dempsey et al. 2020; Department of Social Services 2018) n = 1,659 (2011) 15 years and above The Department of Social Services and the Melbourne Institute of Applied Economic and Social Research, University of Melbourne |
Sexual identity: • heterosexual • homosexual • bisexual |
Intersex variation: • male • female • intersex |
In Wave 1, 2011, 54 and 74 individuals self-identified as homosexual and bisexual respectively. There were no intersex respondents (Andrew Bevitt et al. 2014; Dempsey et al. 2020) |
Longitudinal Study of Australian Children (Perales and Campbell 2019; Perales, Campbell and Olfaherty et al. 2020; Edwards 2014; Arnup, Black and Johnston 2021) n = 5,107 for C cohort (2004) n = 4,983 for K cohort (2004) 0–1 years in 2004 (B cohort), 4–5 years in 2004 (K cohort) followed longitudinally The Department of Social Services, the Australian Bureau of Statistics, and the Australian Institute of Family Studies |
Sexual attraction: • I'm attracted only to girls • I'm attracted only to boys • I'm attracted to girls and boys • I'm not sure who I am attracted to • I do not feel any attraction to others NB: question on sexual attraction was asked to respondents in K cohort at the age 14–15 |
NR | Within the K cohort in Wave 6, 2014, 24 individuals reported they were exclusively attracted to individuals of their same sex and 93 individuals reported they were attracted to individuals of both sexes (Perales, Campbell and Olfaherty 2020) |
National Drug Strategy Household Survey (White et al. 2013; Tait 2015; Praeger et al. 2019; AIHW 2021) n = 23,855 (2013) 12 years and above or 14 years and above, depending on the year of the survey Australian Institute of Health and Welfare |
Sexual identity: • heterosexual or straight • homosexual (gay or lesbian) • bisexual • not sure, undecided • something else, other NB: Collected since 2007 |
NR | In the 2013 NDSH, there were 277 (1.2%) men and 302 (1.3%) women that identified as homosexual or bisexual, respectively (Roxburgh et al. 2016) |
National Survey of Mental Health and Well-being n = 8,841 households (2007) 16–85 years Australian Bureau of Statistics |
Sexual identity: • heterosexual • homosexual • bisexual NB: collected since 2007 |
NR | Not able to be determined. ABS publicly available downloads only report prevalence of mental health disorders by sexual orientation and no published reports were found |
National Secondary Students and Sexual Health survey (Fisher et al. 2019) n = 8,400 (2018) 14–18 years Australian Research Centre in Sex, Health and Society, La Trobe University |
Sexual identity: • heterosexual or straight • gay or lesbian • bisexual • not sure Sexual attraction: • only attracted to females • mostly attracted to females • equally attracted to females and males • mostly attracted to males • only attracted to males • not sure Sexual behaviour: Gender of partner in last sexual encounter: • male • female • other |
Gender identity: • female • male • transmale • transfemale • genderqueer/gender non-conforming • another identity |
In 2018, 201 (2.4%) of the sample responded they had a gender other than male or female; 208 (2.5%) identified as having a gender identity other than male or female; and, 419 (5.1%), 1317 (16.1%), and 421 (5.2%) (individuals self-identified as gay or lesbian, bisexual and ‘not sure’ respectively. 35% of respondents indicated an attraction to the same gender or to multiple genders male (87.4%) and female (97.4%) students indicated their last sexual experience occurred with someone of the opposite gender (Fisher et al. 2019) |
Private Lives (Leonard et al. 2012; Hill et al. 2021a) n = 6,835 (2019) 18 years and above Managed jointly by Rainbow Health Victoria and the Australian Research Centre in Sex, Health and Society, La Trobe University |
Sexual identity: • gay man • lesbian • bisexual • gay woman • don't use label • queer • dyke • heterosexual/straight • not sure • other Sexual attraction: • I have felt attracted only to females, never to males; • more often to females and at least once to a male • about equally often to females and to males • more often to males and at least once to a female • I have felt sexually attracted only to males, never to females • I have never felt sexually attracted to anyone at all |
Gender identity: • male • female • non-binary • I use a different term NB: multiple response options available and also a question about gender at birth. Intersex variation: Were you born with a variation in your sex characteristics? • yes • no |
Of the participants from the third Private Lives Survey, carried out in 2019, 300 (4.4%) were trans men, 285 (4.2%) were trans women, and 921 (13.6%) were non-binary. 1394 (20.5%) self-identified as lesbian, 1958 (28.7%) as gay, 1387 (20.4%) as bisexual, 503 (7.4%) as pansexual, 833 (12.2%) as queer, 215 (3.2%) as asexual and 525 (7.7%) as ‘something else’. There were 47 (<1%) participants with an intersex variation/s (Hill et al. 2021a) |
Ten to Men (Pirkis, Macdonald and English 2016; AIFS 2019; Bollier et al. 2020) n = 16,021 (2013) NB: own calculations 10–55 years in Wave 1 (2013) followed longitudinally Australian Institute of Family Studies |
Sexual identity: • heterosexual • bisexual • homosexual • not sure • other Sexual attraction: • I am attracted only to girls/females • I am attracted only to guys/males • I am attracted to girls/females and guys/males • I am not sure who I am attracted to • I do not feel any attraction to others Sexual behaviour: Continuous response options to ‘total number of lifetime female and male partners’ |
NR | Of the 16,021 male respondents in Wave 1, carried out in 2013 (that is, as there is no gender identity question), 209 (1.3%) self-identified as bisexual, 229 (1.4%) as homosexual, 322 (2.0%) as ‘not sure’ and 225 (1.4%) as ‘other’. In terms of sexual attraction, 244 (1.5%) were attracted to girls/females and guys/males, 202 (1.3%) were attracted to only guys/males, 77 (<1%) were not sure who they were attracted to and 146 (<1%) did not feel any attraction to others NB: own calculations |
Writing Themselves In (Hillier et al. 1998; Hillier, Turner and Mitchell 2005; Hill et al. 2021b) n = 6,418 (2019) 14–21 years Australian Research Centre in Sex, Health and Society in La Trobe University |
Sexual identity: • lesbian • gay • bisexual • pansexual • queer • asexual • something different NB: more response options are available in the raw survey data. In 2010 and 2019 options were expanded Sexual attraction: • people of the same-sex only • both sexes • opposite sex only • unsure Sexual behaviour: Respondents are asked about gender of their sexual partners |
Gender identity: • cisgender man • cisgender woman • trans man • trans woman • nonbinary NB: more response options are available in the raw survey data. Intersex variation: Were you born with a variation in your sex characteristics? • yes • no |
Whole sample is SGD population. In the fourth Writing Themselves survey, carried out in 2019, 50.6% of participants identified as cisgender women, 22.3% cisgender men, 19.5% non-binary, 6.5% trans men and 1.2% trans women. There were 20 (<1%) participants with an intersex variation/s (Hill et al. 2021b) |
- Abbreviation: NR, not reported.
The majority of the datasets containing information on sexual and/or gender diversity were managed by academic institutions (n = 10), followed by government organisations (n = 6) and one not-for-profit organisation. Of the datasets capturing some information on sexual diversity, sexual identity was most commonly reported (n = 15), followed by sexual attraction (n = 5) and sexual behaviour (n = 6). All of the datasets capturing information on gender identity (n = 7) also collected information on some aspect of sexual diversity. Only one longitudinal dataset, ‘Following Lives Undergoing Change’, captured information on both sexual and gender diversity. No government-ran surveys captured information on gender diversity.
Terminology and questions used to capture information on sexual and gender diversity were applied differently across the datasets. Only four surveys asked about intersex variation and the Census was distinct in allowing a response option of ‘other’ for the question on sex.2 In the ‘Australian Workplace Equality Index Employee Survey’, sexual and gender diversity were captured simultaneously (that is, within the same question), whereas in ‘Following Lives Undergoing Change’, gender identity and intersex status were captured simultaneously. Sexual identity questions were generally shown to be more consistent across different surveys.
Surveys that were representative of the Australian population and contained information on either sexual or gender identity had sample sizes with, on aggregate (for example, combining all LGB respondents), less than 1,000 SGD respondents. Larger samples focused on specific subgroups of the SGD community, for example: ‘Following Lives Undergoing Change’ and the ‘Gay Community Periodic Surveys’ recruits cis and trans, gay, bisexual, and queer men; the ‘Australian Trans and Gender Diverse Sexual Health Survey’ focused on trans and gender diverse communities; and boththe ‘National Secondary Students and Sexual Health survey’ and ‘Writing Themselves In’ surveysfocused on young LGBTIQ people. These largerdatasets of specific SGD populations have applied convenience sampling through online and community-based recruitment (Crouch et al. 2012; Fisher et al. 2019; Hill et al. 2021b) and, with the exception of ‘Following Lives Undergoing Change’, were all repeated cross-sectional samples.
3 Discussion
This work has provided an overview of the 17 national datasets that currently collect information on SGD populations in Australia. This overview provides information on the frequency and representability of data capture and indications of sample sizes, as well as specific detail on how questions pertaining to sexual and gender diversity are captured.
It was found that nationally representative samples do not adequately capture information on SGD populations or, if they do, are not sufficiently powered to allow for meaningful subgroup or intersectional analyses. Larger datasets of SGD populations have generally applied convenience sampling and, to this end, there may be larger sample sizes in jurisdictions with greater levels of engagement from local community groups or stakeholders. Importantly, SGD-specific datasets do not allow for comparison to non- LGBTIQ populations, which hampers research into inequalities and causal inference associated with SGD-specific policy. Moreover, whether these SGD-specific dataset samples are indeed ‘representative’ of the SGD population in Australia is ultimately not able to be determined without larger scale, probabilistic sampling methods. It is, therefore, vital to advocate for government agencies to include questions on both sexual and gender diversity in large-scale, representative, national datasets, such as through the Census. Indeed, other countries have started to collect this information through their Census to try and address these gaps (Cooley 2020).
Terminology and the measurement of gender and sexual diversity was shown to be highly variable across different surveys. Capturing sexual diversity through sexual identity was more common than through sexual attraction or sexual behaviour and surveys were fairly consistent in their response options for sexual identity. Indeed, specific dimensions of sexuality may be more appropriate for different surveys. For example, information on sexual behaviour might be particularly important to surveys focusing on provision of sexual health services (Cooley 2020), while sexual identity might be useful for monitoring broader health or socioeconomic inequalities. Questions that aimed to capture aspects of gender diversity were less consistent across the datasets. Moving forward, attempting to provide some level of harmonisation will be critical to facilitate comparison across different datasets over time. This harmonisation is particularly important as SGD population estimates will vary widely depending on how questions are presented and what dimensions are captured (Richters et al. 2014; Bauer et al. 2017; Urwin, Mason and Whittaker 2021). Ongoing community consultation is recommended to ensure that language around sexuality and gender aligns with community expectations and adequately captures SGD identities.
Finally, I showed that samples that had included information about sexual and/or gender diversity in longitudinal samples were also lacking. This is particularly important as both sexuality and gender are fluid over individuals' life course. Incorporating questions on both sexual and gender identity into longitudinal data collections is therefore encouraged and poses significant opportunity to explore important transitions in SGD Australians.
It is important to note several limitations to this review. As I focus on datasets that operate at a national level, and are chiefly provided by government organisations, as well as those that have been used to provide LBGTIQ specific population-based estimates, this is not intended to be an all-encompassing review. Many organisations and researchers have developed their own cohort studies or community samples to include specific SGD populations (examples include, but are not limited to: Pitts et al. 2009; Crouch et al. 2014; Prestage et al. 2014; Mooney-Somers et al. 2019; Tollit et al. 2019; Ferlatte et al. 2020; ACON 2021). This work is to be commended and has undoubtedly developed important insights into SGD populations. This review also does not provide detailed information about the full breadth of outcomes and characteristics included within each of the different surveys (for example, sociodemographic characteristics, labour market outcomes, linkage to administrative data, and so forth). Information about the relevant data custodians and references are therefore provided to assist readers should they wish to know more about the specific datasets.
Despite these limitations, this review provides important information about the current state of national data collection for SGD populations in Australia. As always, when attempting to research minority populations, there will be a trade-off between the granularity of information able to be provided through traditional national surveys versus the potential for disclosure risks and limitations in statistical power. These are particular challenges in Australia; a country with a relatively small and geographically disparate population. Nevertheless, this review does elucidate the need for large-scale, representative and longitudinal data capture of SGD populations in Australia. Future data collections need to apply appropriate and consistent language in order to ensure that SGD identities are appropriately reflected in our population.
4 Best Practice
Turning specifically to economic research, at a minimum, it is recommended that longitudinal surveys aim to capture both sexual and gender identity over multiple waves. This will be essential to estimate the causal impact of different policies, as well as to minimise potential bias associated with extrapolation of these measures across multiple waves (Urwin, Mason and Whittaker 2021). Additionally, including these questions in the Census is strongly recommended to ultimately capture a ‘true’ national sample of SGD populations and facilitate comparison and weighting relative to other datasets. The importance of having a better understanding of SGD populations across regions in Australia to inform targeted policy cannot be understated. Specific questions could consider standardised terminology as used by the Gay Community Periodic Surveys as well as other surveys developed in partnership with ACON, which involve extensive consultation with the LGBTIQ community. More broadly, data custodians should consider the development of tools and resources that can support ongoing scholarship in this space. For instance, the Household, Income and Labour Dynamics in Australia survey has a dedicated online data dictionary, with search functionality, which provides information on all survey questionsand response options, across which waves thequestions were asked and wave-specific sample sizes for select variables (albeit, at this stage not sexual identity, presumably due to sample size limitations) (Melbourne Institute of Applied Economic and Social Research 2022). This information and similar‘user friendly’ interfaces would be extremely valuable to enable researchers to plan what analyses are possible across different datasets. Moreover, more aggregate level statistics could be provided to minimise disclosure risk (for example, ‘proportion of sample not heterosexual’). In the interim, I hope this review can assist researchers in planning their analyses, as well as highlight to data custodians and decision makers where improvements could be made to data collection toensure that evidence-based policy and a rich literature on the Australian LGBTIQ community is realised.
Acknowledgment
Open access publishing facilitated by Monash University, as part of the Wiley - Monash University agreement via the Council of Australian University Librarians.
