A U ST R A L IA N P O P U L AT IO N ST U D IE S 2018 | Volume 2 | Issue 1 | pages 26–38 © Wilson and Shalley 2018. Published under the Creative Commons Attribution-NonCommercial licence 3.0 Australia (CC BY-NC 3.0 AU). Journal website: www.australianpopulationstudies.org Estimates of Australia’s non-heterosexual population Tom Wilson* Charles Darwin University Fiona Shalley Charles Darwin University * Corresponding author. Email: tom.wilson@cdu.edu.au. Address: Northern Institute, Charles Darwin University, Ellengowan Drive, Darwin, NT 0909 Paper received 14 February 2018; accepted 28 April 2018; published 28 May 2018 Abstract Background Demographers have studied minority populations for many years, but relatively little attention has been paid to sexual minority groups. Population estimates for sexual minorities would be useful as denominators for a range of health and socioeconomic indicators, to monitor representation in employment, assist budget planning and inform the marketing of goods and services. Aim The aim of this paper is to present some approximate estimates of the non-heterosexual adult population of Australia in mid-2016 by sex, broad age group and state and territory. Data and methods Data on sexual identity were sourced from three nationally representative surveys: the Household Income and Labour Dynamics in Australia survey, the second Australian Study of Health and Relationships and the ABS General Social Survey. Use was made also of 2016 ABS Census of Population and Housing (Census) data and Estimated Resident Populations. Prevalence rates of the non-heterosexual population aged 18+ were averaged over the three surveys and multiplied by ERP to obtain national population estimates. Census data on same-sex couples were used to distribute the national estimates by state and territory. Results Australia’s non-heterosexual population aged 18+ in 2016 is estimated to have been 592,000, representing about 3.2% of the adult population. New South Wales is home to the largest non- heterosexual population (about 204,000) and the Northern Territory the smallest (4,700), while the highest prevalence is in the Australian Capital Territory (5.1%). Conclusions Australia’s non-heterosexual population is a relatively small population, but its prevalence varies considerably by age and sex and between states and territories. Estimates of this population should prove useful for monitoring health and wellbeing and for a variety of planning and policy purposes. Key words Population estimates; non-heterosexual; lesbian, gay, bisexual; states and territories; Australia. http://www.australianpopulationstudies.org/ mailto:tom.wilson@cdu.edu.au Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 27 1. Introduction Minority populations have been the subject of demographic research and policy development for many decades, particularly ethnic minority populations (e.g. Coleman and Salt 1996; Jivraj and Simpson 2015) and Indigenous peoples (e.g. Smith 1980; Taylor 2011). Attention has also been paid to groups such as foreign-born populations (e.g. Edmonston 2016) and minority religious populations (e.g. Peach 2006; Pew Research Center 2017). Over the last two decades, and especially in recent years, demographers and statisticians have started to examine sexual and gender minority populations in western countries (e.g. Aspinall 2009; Baumle, Compton and Poston 2009; Baumle 2013; Black et al. 2000; ONS 2010; Meier and Labuski 2013; ONS 2017). In Australia, the non-heterosexual1 population was subject to increased attention in the lead-up to the Australian Marriage Law Postal Survey in 2017. Following public debate about same-sex marriage over many years, the Australian Government decided to put the issue to the electorate in late 2017. Effectively a plebiscite, the survey asked: ‘Should the law be changed to allow same-sex couples to marry?’ Of those who returned their forms (and excluding unclear responses), 62 per cent voted ‘yes’ (ABS 2017a). Shortly afterwards Parliament voted overwhelmingly to amend the Commonwealth Marriage Act 1961, and the updated law came into force in December 2017. Yet while progress such as this is made towards equality for non-heterosexual people in Australia, relatively little is known about the demography of this population. More demographic research could increase knowledge and understanding of this segment of society, leaving less room for stereotypes and myths (Gates 2012). Better demographic evidence could highlight where disadvantage and discrimination remain, and where policy and legislation need revision. Sexual minority population estimates specifically could prove valuable for a variety of health, social justice and economic purposes. For example, estimates could be used to monitor representation in employment, assist budget planning for specific services for sexual minorities, form denominators for a range of health and socioeconomic indicators, and inform the marketing of goods and services (e.g. same-sex weddings). In addition, sexual minority demography can prove useful in broader analyses of demographic trends and theory, such as the influence of gender roles within couples on the labour market consequences of migration (e.g. Cooke 2005). The bulk of existing research on sexual minority demography focuses on the United States and, to a lesser extent, the United Kingdom (e.g. Black et al. 2000; Black et al. 2002; Cooke and Rapino 2007; Flores et al. 2016; Gates 2014; Geary et al. 2018; ONS 2017; van Kampen et al. 2017). Australian research in the area is limited, though a small number of contributions have been made over the last two decades (e.g. Dempsey 2015; Grulich et al. 2003; Madeddu et al. 2006; Perales and Baxter 2018; Prestage et al. 2008; Richters, Altman et al. 2014; Smith et al. 2003; Wilson 2004). Many Australian- based papers are concerned primarily with important sexual health issues, and only a few present population estimates. Prestage et al. (2008) estimated the population of gay and bisexual men in Australia in each of the states and territories in 2001, while Madeddu et al. (2006) created estimates of gay and bisexual men across inner Sydney postcodes in 2001. Both multiplied gay and bisexual prevalence rates by census counts, using data from the Australian Bureau of Statistics (ABS) Census 1 We use the term ‘non-heterosexual’ here, but acknowledge there is no standard terminology for sexual and gender minority groups. We discuss this issue in section 2 of the paper. 28 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 of Population and Housing (Census), to obtain estimates. Several other studies included estimates of the prevalence of the non-heterosexual population or parts of it (primarily gay and bisexual men). We argue that there is a need for sexual minority population estimates for Australia. Ideally, these would be current, available annually, contain age, sex and geographic detail, cover the whole non- heterosexual population but also include separate estimates for lesbian, gay, bisexual and other minority sexual orientations. Preferably, the estimates would use Estimated Resident Populations (ERPs) in their calculation rather than census counts, which suffer from undercount (particularly in the young adult ages). But creating sexual minority estimates is far from straightforward. There are so few demographic data sources which include sexual orientation (Wooden 2014), and those that do so have several limitations. Defining who should be included in the non-heterosexual population is also complicated because of the complex nature of sexual orientation. The aim of this paper is to present some approximate estimates of the non-heterosexual adult population of Australia in mid-2016 by sex, broad age group and state and territory. The estimates are approximate in the sense that they are constructed from several imperfect data sources and necessarily involve several assumptions. At this stage we present state and territory and national estimates only, as well as those for the non-heterosexual population as a whole. Finer geographical detail, annual numbers, alternative definitions (see section 2), and separate estimates for lesbian, gay, bisexual and other minority sub-populations are planned in subsequent work. Following this introduction we consider how sexual orientation is conceptualised and the issue of appropriate definitions and terminology for sexual minority populations. Section 3 describes the data sources and population estimation methods, while section 4 presents the results. A final section includes a short discussion and conclusion. 2. Definitions and terminology Sexual orientation has been defined as: an enduring pattern of emotional, romantic, and/or sexual attractions to men, women, or both sexes … [and] a person’s sense of identity based on those attractions, related behaviors, and membership in a community of others who share those attractions. (American Psychological Association 2008 p. 1) It is generally considered to comprise three elements (Durso and Gates 2013): (i) sexual attraction – the feeling of sexual desire towards others, considered by some to be the fundamental basis of sexual orientation (ii) sexual behaviour – sexual activity (iii) sexual identity – how a person describes their sexual orientation (e.g. heterosexual, gay, lesbian, bisexual). The non-heterosexual population referred to in this paper includes all those who identify as gay, homosexual, lesbian or bisexual, or construct their sexuality in other ways using non-heterosexual terminology (e.g. queer). Our focus here is on sexual identity rather than behaviour or attraction. The sexual identity, behaviour and attraction of individuals are not necessarily coterminous. The Venn diagram in Figure 1 illustrates the relationship between the three dimensions of sexual Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 29 orientation for the non-heterosexual population. Several surveys have found that a relatively large group of people have felt some same-sex sexual desire at some point in their adult lives; a smaller but not wholly overlapping group have engaged in same-sex sexual activity’ and a smaller group still identifies as being non-heterosexual (e.g. Geary et al. 2018; Richters, Altman et al. 2014; Smith et al. 2003). Identity is shown in the diagram as ‘current’ identity because some people change the way they describe themselves over time. Figure 1: The relationship between the three dimensions of sexual orientation Source: Loosely based on Figure 1 in Richters, Altman et al. (2014) and Figure 2 in Geary et al. (2018). It is important to recognise that people counted as non-heterosexual in data sources are only those willing to identify as such in statistical collections. Strictly, this is the revealed non-heterosexual population. Others may consider themselves to be gay, lesbian, bisexual or other in private, but choose not to disclose their sexuality in public. This is the hidden (or ‘closeted’) population, and is not included in our population estimates (see Gates 2012 for a discussion of the issue). In addition, it is important to note that sexual orientation is distinct from gender identity and intersex issues, although they are often considered together – for example, in the term LGBTQI (Lesbian, Gay, Bisexual, Transgender, Queer and Intersex). Gender identity refers to ‘a person’s innate, deeply-felt psychological identification as a man, woman, or something else’ (Fenway Health 2010 p. 3). The transgender population consists of those whose gender identity or expression differs from their birth sex, while intersex refers to ‘a spectrum of conditions involving anomalies of the sex chromosomes, gonads, reproductive ducts, and/or genitalia’ (Fenway Health 2010 p. 9). Transgender and intersex populations may be of any sexual orientation. Finally, on the issue of terminology, we acknowledge that many different terms are used in the literature covering different population sub-groups, with no single term being standard (Dempsey 2015). These include LGB (lesbian, gay and bisexual), LGBT (lesbian, gay, bisexual and transgender), LGBTIQ (lesbian, gay, bisexual, transgender, intersex and queer), and LGBTQQIP2SAA (lesbian, gay, bisexual, transgender, queer, questioning, intersex, pansexual, 2-spirited, asexual and allies). We choose the shorthand term ‘non-heterosexual’ to refer to gay, lesbian, bisexual and related sexual Any lifetime same- sex behaviour Any lifetime same- sex attraction Current non- heterosexual identity 30 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 minorities. Gender minorities such as the transgender population, and sex minorities such as intersex people, are included in our population estimates only if they have identified as non-heterosexual. 3. Data and methods 3.1 Data sources Data on people identifying as non-heterosexual were acquired from three representative national surveys: the Household Income and Labour Dynamics in Australia (HILDA) survey, the second Australian Study of Health and Relationships (ASHR2), and the ABS General Social Survey (GSS). Unpublished data on the population identifying as non-heterosexual by sex and broad age group (18–24, 25–34, 35–44, 45–54, 55–64 and 65+) were obtained and non-heterosexual proportions of the population (or prevalence) calculated for each dataset. The HILDA survey is a nationally representative household-based longitudinal survey on the lives of Australians (Wilkins 2017). It covers a wide range of topics including family relationships, employment, education, income, health and wellbeing, life events and personal attitudes. It is conducted annually by interview and self-completion questionnaire with all those aged 15 years and over in the participating household. A question on sexual identity was included in Wave 12 (2012) as part of the self-completion questionnaire (Wilkins 2015; Wooden 2014). In this wave interviews were conducted with 17,476 respondents and self-completion questionnaires were received from 88 per cent of the interviewed sample. The sexual identity question asked: ‘Which of the following categories best describes how you think of yourself? Heterosexual or Straight; Gay or Lesbian; Bisexual; Other; Unsure/Don’t know; and Prefer not to say’ (Melbourne Institute 2012). The ASHR2 was undertaken in 2012 and 2013 to study the sexual and reproductive health, sexual practices and wellbeing of the Australian population (Richters, Badcock et al. 2014). The survey was conducted via telephone interview using random digit dialling of landline and mobile phones. The survey covered Australian residents aged 16–69 and obtained responses from 20,094 participants, achieving a response rate of 66 per cent of eligible people contacted. The sexual identity question asked: ‘Do you think of yourself as: 1. Heterosexual or straight; 2. Homosexual (gay [asked of males]; lesbian [asked of females]); 3. Bisexual’. Other responses were coded as: 4. Queer; 5. Not sure; undecided; 6. Something else/other (Richters, Altman et al. 2014). The GSS is a household-based survey run by the ABS every four years. The aim of the survey is to provide a broad range of information on an individual’s social circumstances and their relative levels of advantage and disadvantage (ABS 2015a). The survey covers all people aged 15 years and over who are usual residents of private dwellings and is conducted by face-to-face interview. In the 2014 survey 12,932 out of 16,145 eligible persons responded to the survey, representing a response rate of 80 per cent. A question on sexual identity was introduced for the first time in 2014 and was asked of all participants aged 18 years and over. It asked: ‘Which of the following options best describes how you think of yourself? 1. Straight (Heterosexual); 2. Gay or Lesbian; 3. Bisexual; 4. Other; 5. Don’t Know’ (ABS 2015b). In addition, our study made use of some data from the 2016 Census. Counts of persons in same-sex couples by broad age group, sex and state and territory were extracted from the ABS TableBuilder Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 31 Pro online data service (ABS 2017b). A direct question on sexual orientation is not asked in the Australian Census. Instead, a same-sex couple variable is created by combining information from two Census questions: the question on the relationship of an individual in the household to person 1 on the Census form; and the question ‘Is this person male or female?’. If the relationship of someone to person 1 is recorded as husband or wife or de facto partner, and both individuals record the same answer to the question on gender, then a same-sex couple is identified (ABS 2018). Nationally, 2016 census counts of individuals aged 18 and over living together as part of a same-sex couple totalled about 90,000 (only 0.5 per cent of adults counted in the Census). This is only a subset of all non-heterosexual people, of course. Aside from census undercount, non-heterosexual people who are single or in living- apart-together relationships are excluded from the census count of individuals in same-sex couples. HILDA survey findings from 2012 show that, amongst those aged 25–59, only 55 per cent of non- heterosexual men and 59 per cent of women are partnered, compared to 74 per cent of heterosexual men and women in that age range (Wilkins 2015). In addition, people in cohabiting same-sex couples are not always identified in the Census as being in a same-sex relationship: for example, when neither is person 1 on the Census form, or if they do not wish to reveal their relationship. Furthermore, there is probably some error in the same-sex couple counts due to errors in responses to the question on gender. Research in the United States has revealed this to be a problem with their same-sex couple census data (e.g. Gates 2010; DiBennardo and Gates 2014), but there is no information as to whether equivalent response errors have affected Australian Census data. Finally, mid-2016 ERPs by broad age group, sex and state and territory were obtained from the ABS.Stat online data extraction tool (ABS 2017c). 3.2 Methods Estimates of the adult non-heterosexual population of Australia in 2016 were created by multiplying non-heterosexual proportions averaged across the three surveys (HILDA, ASHR2, GSS) by ERPs published by the ABS for 30 June 2016. The estimates were prepared by sex and broad age group (18–24, 25–34, 35–44, 45–54, 55–64 and 65+) for Australia and each of the states and territories. The decision to use proportions averaged across the three surveys was taken because of the non-trivial variations in proportions between surveys and no clear evidence that one survey was superior to another. National estimates were calculated by multiplying averaged proportions by ERPs: 𝑃𝐴𝑢𝑠,𝑠,𝐴 𝑁𝐻 = 𝑝𝑟𝑜𝑝𝐴𝑢𝑠,𝑠,𝐴 𝑁𝐻 𝐸𝑅𝑃𝐴𝑢𝑠,𝑠,𝐴 (1) where 𝑃 denotes population estimate, 𝐴𝑢𝑠 Australia, 𝑝𝑟𝑜𝑝 the proportion of the population, 𝑁𝐻 non-heterosexual, 𝐸𝑅𝑃 Estimated Resident Population, 𝑠 sex and 𝐴 broad age group. State and territory estimates were created by distributing national estimates according to the number of individuals in same-sex relationships identified in the 2016 Census: 𝑃𝑖,𝑠,𝐴 𝑁𝐻 = 𝑃𝐴𝑢𝑠,𝑠,𝐴 𝑁𝐻 𝐶𝑖,𝑠,𝐴 𝑆𝑆𝑅 𝐶𝐴𝑢𝑠,𝑠,𝐴 𝑆𝑆𝑅 (2) where 𝑖 denotes state or territory, 𝐶 census counts, and 𝑆𝑆𝑅 the population in same-sex relationships. While the Census same-sex couple data provide an imperfect geographical distribution proxy, they do not suffer from small sample size. Survey estimates of state and territory non-heterosexual 32 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 populations were based on samples too small to be of use (e.g. there were zero non-heterosexual males in the Northern Territory in the HILDA data). Clearly, the above methods incorporate a number of assumptions which are unlikely to hold precisely. No adjustments have been made for net undercount in the Census. It is assumed that the average proportions across the three surveys provide an accurate measure of the identified non- heterosexual population of Australia. It is also presumed that the proportion of the population identifying with non-heterosexual identities remains unchanged between the time of the surveys in 2012 to 2014 and mid-2016. For the state and territory population estimates, the distribution of people in same-sex relationships recorded by the 2016 Census is assumed to match the distribution of the wider non-heterosexual population. This last assumption is probably the most approximate. It effectively assumes that, for each age-sex group, the proportion of the non-heterosexual population in a same-sex relationship as identified by the Census is the same in each state and territory. This is unlikely to be exactly the case, though we note that the Census distribution of all persons in couples (irrespective of gender composition) by state and territory and broad age group is extremely close to the equivalent ERP distribution. However, variations from our assumption are likely to occur due to educational composition. Using survey data from California, Carpenter and Gates (2008) found that highly educated gay men and lesbian women were more likely to be partnered than those who were less educated. 4. Non-heterosexual population estimates 4.1 New estimates The extent to which the Australian population identifies as non-heterosexual, using the averaged findings from the three surveys, is shown in Table 1 below. Amongst the adult population of Australia, 3.1 per cent of males and 3.4 per cent of females describe themselves as non-heterosexual. These percentages exclude survey participants who refused to answer the sexual identity question or who replied ‘don’t know’. The percentages are higher at younger ages and lower at older ages, with the age gradient being more pronounced for females. Table 1: Percentage of the Australian adult population identifying as non-heterosexual averaged across three surveys, 2012–2014 Age group Males Females 18–24 4.0 6.4 25–34 4.0 4.8 35–44 3.0 3.8 45–54 2.6 2.6 55–64 2.4 2.0 65+ 2.3 1.1 18+ 3.1 3.4 Sources: GSS 2014; ASHR2; HILDA Wave 12. Estimates of Australia’s non-heterosexual population by age group and sex are presented in Table 2. They suggest that the national non-heterosexual population aged 18+ in mid-2016 was a little under Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 33 600,000, representing 3.2 per cent of the total adult population. The figures indicate there were more non-heterosexual females than males in the younger adult ages, with the situation reversed in the older age groups. Overall, at ages 18 and above the non-heterosexual population is younger than the Australian population as a whole. Table 2: Estimates of the Australian adult population identifying as non-heterosexual, 2016 Age group Males Females Persons 18–24 47,098 71,839 118,937 25–34 71,804 86,551 158,355 35–44 48,874 61,770 110,644 45–54 40,848 41,009 81,857 55–64 32,579 28,684 61,263 65+ 39,848 21,011 60,859 18+ 281,052 310,863 591,915 Source: Authors’ estimates. Population estimates for the states and territories are shown in Table 3. Not surprisingly the most populous states, New South Wales, Victoria and Queensland, are home to the largest non-heterosexual populations, while the Australian Capital Territory (ACT), Tasmania and the Northern Territory have the smallest populations. All jurisdictions have more females than males in their non-heterosexual populations with the one exception of New South Wales. As a fraction of the total population, South Australia, Western Australia and the Northern Territory have the smallest percentages identifying as non-heterosexual. The ACT has a relatively large non-heterosexual population at 5.1 per cent of its total adult population, a function to some extent of its comparatively young population (Wilson 2016). Table 3: Estimates of the adult population identifying as non-heterosexual by state and territory, 2016 State/territory Population aged 18+ % of ERP Males Females Persons Persons NSW 106,400 98,023 204,423 3.4 Vic 76,267 80,790 157,057 3.3 Qld 48,996 63,596 112,592 3.0 SA 14,265 20,818 35,083 2.6 WA 21,280 29,828 51,108 2.6 Tas 4,992 5,953 10,945 2.7 NT 2,147 2,596 4,743 2.6 ACT 6,705 9,258 15,964 5.1 Australia 281,052 310,863 591,915 3.2 Source: Authors’ estimates. Notes: NSW = New South Wales; Vic = Victoria; Qld = Queensland; SA = South Australia; WA = Western Australia; Tas = Tasmania; NT = Northern Territory; ACT = Australian Capital Territory. 4.2 Comparisons with other studies Comparing our non-heterosexual population estimates with estimates produced by others is not straightforward because of variations in population coverage, reference dates, age ranges, data sources and social and cultural contexts. Nonetheless, it is instructive to compare our figures with the few earlier estimates prepared for Australia. 34 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 The gay and bisexual male population aged 16+ was estimated by Prestage et al. (2008) to be 182,624 in 2001, representing 2.5 per cent of the male population. The prevalence estimates across states and territories varied from 0.8 per cent for Tasmania and 0.9 per cent for the Northern Territory to 2.9 per cent for the ACT and 3.0 per cent for Sydney. Those numbers were based on the first Australian Survey of Health and Relationships (ASHR). Our non-heterosexual population prevalence rates for just the male population aged 18+ varied from a low of 2.2 per cent in South Australia and Western Australia and 2.3 per cent in the Northern Territory up to 3.6 per cent in New South Wales and 4.4 per cent in the ACT. It was 3.1 per cent for Australia as a whole. How do our estimates compare to those produced for other countries? Gates (2011) estimated that about 8 million adults identify as lesbian, gay or bisexual in the United States, representing about 3.5 per cent of the adult population. Geary et al. (2018) estimated 1.2 million people aged 16–74 in Britain identified as lesbian, gay, bisexual or other in 2011, or 2.7 per cent of the population. A recent estimate by the Office for National Statistics put the population aged 16 and over describing themselves as lesbian, gay, bisexual or other in the United Kingdom in 2016 to be 1.3 million, representing 2.5 per cent of the population (ONS 2017). In Canada, the Canadian Community Health Survey found that 3.0 per cent of adults aged 18–59 reported themselves to be lesbian, gay or bisexual in 2014 (Statistics Canada 2017). In terms of age distributions, our data are consistent with most other studies, which have found a higher prevalence of non-heterosexual identities at younger ages (e.g. Gates 2014; ONS 2017). However, this is not a universal finding, with the prevalence estimates for Britain presented by Geary et al. (2018) showing little difference for ages between 16 and 64. Overall, our non-heterosexual population estimates are reasonably consistent with other respectable studies, and consist of plausible and sensible numbers. 5. Conclusions This paper has presented a novel set of population estimates for Australia’s non-heterosexual population, with disaggregation by sex, broad age group and state and territory. The estimates should be regarded as approximate given the limitations of the data sources and the assumptions inherent in our methods. Nonetheless, they provide a useful overview of the current non- heterosexual population of Australia which was not previously available. It is emphasised that, conceptually, these estimates refer to the population identifying as gay, lesbian, bisexual and in other non-heterosexual ways, and not those who engage in same-sex sexual behaviour or who have ever experienced same-sex sexual attraction (Figure 1). Population estimates based on these alternative definitions should be useful for health surveillance and will be the subject of subsequent research. Ideally, state and territory and sub-state, non-heterosexual population estimates would be based on Census or large-scale survey data. The Office for National Statistics in the United Kingdom asks a sexual identity question in its Annual Population Survey, which is sufficiently large (covering 41,000 households each quarter) to enable sexual identity population estimates to be produced for local authorities (ONS 2017). Perhaps at some point the ABS will follow suit and include a direct sexual identity question in a large-scale survey (or even the Census). In the meantime it would be useful to investigate alternative data sources for estimating the state and territory distributions because the use of same-sex couple data from the Census is imperfect. Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 35 It is possible that our ACT prevalence is marginally over-estimated due to its highly educated population. There is likely to be a higher proportion of non-heterosexual people in the ACT in same- sex cohabiting relationships than in other jurisdictions (Carpenter and Gates 2008), which violates the assumption of a fixed ratio between the number of people in same-sex couples in the Census and the non-heterosexual population across jurisdictions. Data on the proportion of non-heterosexual adults in same-sex relationships would be ideal as it would enable census counts of persons in same- sex couples to be more accurately scaled up to the total non-heterosexual population. It would also be useful to investigate the potential of new web-based data sources, such as Google search data and social media profiles. For example, some Facebook users reveal sexual orientation on their profiles, but even when they do not, Facebook ‘likes’ can be used to accurately predict sexual orientation in 88 per cent of cases (Kosinski, Stillwell and Graepel 2013). In a more controversial paper, Wang and Kosinski (2018) applied neural networks to detect sexual orientation indirectly from photographs of faces on Facebook profiles. Clearly, approaches such as these raise some challenging ethical questions, and also questions about data reliability and representativeness. Future research should explore these new data sources while also continuing to rely on the strengths of existing survey and Census data. Key messages • Sexual minority population estimates are useful for a variety of health, social justice and economic purposes, and for broader analyses of demographic trends and theory. • Australia’s non-heterosexual population aged 18+ in 2016 is estimated to have been 592,000, representing about 3.2 per cent of the adult population. • The percentages identifying as non-heterosexual are higher at younger ages and lower at older ages, with the age gradient being more pronounced for females. • As would be expected, the most populous states, New South Wales, Victoria and Queensland, are home to the largest non-heterosexual populations, while the ACT, Tasmania and the Northern Territory have the smallest populations. • As a fraction of the total population, South Australia, Western Australia and the Northern Territory have the smallest percentages identifying as non-heterosexual (2.6%), while the ACT has the highest (5.1%). Ethics approval Approval for this study was received from the Charles Darwin University Human Research Ethics Committee (Approval H17122). Contact: ethics@cdu.edu.au. Acknowledgements We gratefully acknowledge helpful advice received from Emeritus Professor Gary Gates during the course of this study, and the helpful comments of the anonymous reviewers on an earlier version of this paper. Our thanks and appreciation is also extended to Hamish McManus from the Kirby Institute, UNSW, who extracted data from the ASHR2, and Laura Bennetts-Kneebone from the Commonwealth Department of Social Services who extracted data from the HILDA survey. mailto:ethics@cdu.edu.au 36 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 References ABS (Australian Bureau of Statistics) (2015a) General Social Survey, 2014. Cat. No. 4159.0. Canberra: ABS. ABS (2015b) General Social Survey household questionnaire. Cat. No. 4159.0. Canberra: ABS. http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/4159.02014?OpenDocument. Accessed on 8 February 2018. ABS (2017a) Australian Marriage Law Postal Survey, 2017. Cat. No. 1800.0. Canberra: ABS. ABS (2017b) TableBuilder Pro. Canberra: ABS. http://www.abs.gov.au/websitedbs/censushome.nsf/home /tablebuilder. Accessed on 8 February 2018. ABS (2017c) ABS.Stat. Canberra: ABS. Available at: http://stat.data.abs.gov.au/. Accessed on 8 February 2018. ABS (2018) Census of Population and Housing: reflecting Australia – stories from the Census, 2016. Cat. No. 2071.0. Canberra: ABS. http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/2071.0Main+Featur es852016?OpenDocument. Accessed on 8 February 2018. American Psychological Association (2008) Answers to your questions: for a better understanding of sexual orientation and homosexuality. Washington DC: American Psychological Association. http://www.apa.org/topics/lgbt/orientation.aspx. Accessed on 8 February 2018. Aspinall P (2009) Estimating the size and composition of the lesbian, gay, and bisexual population in Britain. Equality and Human Rights Commission Research Report 37. Manchester: Equality and Human Rights Commission. Baumle A K (ed.) (2013) International Handbook on the Demography of Sexuality. Dordrecht: Springer. Baumle A K, Compton D and Poston D L (2009) Same-Sex Partners: The Social Demography of Sexual Orientation. Albany NY: SUNY Press. Black D, Gates G J, Sanders S and Taylor L (2000) Demographics of the gay and lesbian population in the United States: evidence from available systematic data sources. Demography 37(2): 139–154. Black D, Gates G J, Sanders S and Taylor L (2002) Why do gay men live in San Francisco? Journal of Urban Economics 51(1): 54–76. Carpenter C and Gates G J (2008) Gay and lesbian partnership: evidence from California. Demography 45(3): 573–590. Coleman D A and Salt J (1996) Ethnicity in the 1991 Census. Vol. 1. Demographic Characteristics of the Ethnic Minority Population. London: HMSO. Cooke T (2005) Migration of same-sex couples. Population, Space and Place 11(5): 401–409. Cooke T and Rapino M (2007) The migration of partnered gays and lesbians between 1995 and 2000. The Professional Geographer 59(3): 285–297. Dempsey D (2015) Familiarly queer? Same-sex relationships and family formation. In: Heard G and Arunachalam D (eds) Family Formation in 21st Century Australia. Dordrecht: Springer; 225–240. DiBennardo R and Gates G J (2014) US Census same-sex couple data: adjustments to reduce measurement error and empirical implications. Population Research and Policy Review 33(4): 603–614. Durso L E and Gates G J (2013) Best practices: collecting and analyzing data on sexual minorities. In: Baumle A K (ed.) International Handbook on the Demography of Sexuality. Dordrecht: Springer; 21–42. Edmonston B (2016) Canada’s immigration trends and patterns. Canadian Studies in Population 43(1–2): 78–116. Fenway Health (2010) Glossary of Gender and Transgender Terms. Boston, MA: Fenway Health. http://fenwayhealth.org/documents/the-fenway-institute/handouts/Handout_7- C_Glossary_of_Gender_and_Transgender_Terms__fi.pdf. Accessed on 8 February 2018. Flores A R, Herman J L, Gates G J and Brown T N T (2016) How many adults identify as transgender in the United States? The Williams Institute, UCLA School of Law, Los Angeles. https://williamsinstitute.law.ucla.edu/wp-content/uploads/How-Many-Adults-Identify-as- Transgender-in-the-United-States.pdf. Accessed on 8 February 2018. http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/4159.02014?OpenDocument http://www.abs.gov.au/websitedbs/censushome.nsf/home/tablebuilder http://www.abs.gov.au/websitedbs/censushome.nsf/home/tablebuilder http://stat.data.abs.gov.au/ http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/2071.0Main+Features852016?OpenDocument http://www.abs.gov.au/AUSSTATS/abs@.nsf/Lookup/2071.0Main+Features852016?OpenDocument http://www.apa.org/topics/lgbt/orientation.aspx http://fenwayhealth.org/documents/the-fenway-institute/handouts/Handout_7-C_Glossary_of_Gender_and_Transgender_Terms__fi.pdf http://fenwayhealth.org/documents/the-fenway-institute/handouts/Handout_7-C_Glossary_of_Gender_and_Transgender_Terms__fi.pdf https://williamsinstitute.law.ucla.edu/wp-content/uploads/How-Many-Adults-Identify-as-Transgender-in-the-United-States.pdf https://williamsinstitute.law.ucla.edu/wp-content/uploads/How-Many-Adults-Identify-as-Transgender-in-the-United-States.pdf Australian Population Studies 2 (1) 2018 Wilson T and Shalley F 37 Gates G J (2010) Same-sex couples in US Census Bureau data: who gets counted and why. The Williams Institute, UCLA School of Law, Los Angeles. https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/same-sex- couples-in-us-census-bureau-data-who-gets-counted-and-why/. Accessed on 8 April 2010. Gates G J (2011) How many people are lesbian, gay, bisexual and transgender? The Williams Institute, UCLA School of Law, Los Angeles. https://williamsinstitute.law.ucla.edu/research/census-lgbt- demographics-studies/how-many-people-are-lesbian-gay-bisexual-and-transgender/. Accessed on 21 August 2016. Gates G J (2012) LGBT identity: a demographer’s perspective. Loyola Law Review Los Angeles 42(3): 693–714. Gates G J (2014) LGBT demographics: comparisons among population-based surveys. The Williams Institute, UCLA School of Law, Los Angeles. https://williamsinstitute.law.ucla.edu/wp- content/uploads/lgbt-demogs-sep-2014.pdf. Accessed on 8 February 2018. Geary R S, Tanta C, Erens B, Clifton S, Frah P, Wellings K, Mitchell K R, Datta J, Gravningen K, Fuller E, Johnson A M, Sonnenberg P and Mercer C H (2018) Sexual identity, attraction and behaviour in Britain: the implications of using difference dimensions of sexual orientation to estimate the size of sexual minority populations and inform public health interventions. PLoS ONE 13(1): e0189607. Grulich A E, de Visser R O, Smith M A, Rissel C E and Richters J (2003) Sex in Australia: Homosexual experience and recent homosexual encounters. Australian and New Zealand Journal of Public Health 27(2): 155–163. Jivraj S and Simpson L (2015) (eds) Ethnic Identity and Inequalities in Britain: The Dynamics of Diversity. Bristol: Policy Press. Kosinski M, Stillwell D and Graepel T (2013) Private traits and attributes are predictable from digital records of human behaviour. PNAS 110(15): 5802–5805. Madeddu D, Grulich A, Richters J, Ferris J, Grierson J, Smith A, Allan B and Prestage G (2006) Estimating population distribution and HIV prevalence among homosexual and bisexual men. Sexual Health 3(1): 37–43. Marriage Act 1961 (Cwth). Compilation prepared on 9 December 2017 taking into account alterations up to Act No. 129 of 2017. https://www.legislation.gov.au/Details/C2017C00382. Accessed on 28 January 2017. Meier S C and Labuski C M (2013) The demographics of the transgender population. In: Baumle A K (ed.) International Handbook on the Demography of Sexuality. Dordrecht: Springer; 289–327. Melbourne Institute (2012) HILDA self-completion questionnaire. http://melbourneinstitute.unimelb.edu.au/assets/documents/hilda- questionnaires/SelfCompletionQuestionnaireW12M.pdf. Accessed on 8 February 2018. ONS (Office for National Statistics) (2010) Measuring sexual identity: an evaluation report. London: ONS. www.ons.gov.uk/ons/guidemethod/measuring-equality/equality/sexual-identity- project/measuring-sexualidentity–an-evaluation-report.pdf. Accessed on 8 February 2018. ONS (Office for National Statistics) (2017) Subnational sexual identity estimates, UK: 2013 to 2015. https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/sexuality/articles/sub nationalsexualidentityestimates/uk2013to2015. Accessed on 8 February 2018. Peach C (2006) Islam, ethnicity and South Asian religions in the London 2001 Census. Transactions of the Institute of British Geographers 31(3): 353–370. Perales F and Baxter J (2018) Sexual identity and relationship quality in Australia and the United Kingdom. Family Relations 67(1): 55–69. Pew Research Center (2017) Europe’s growing Muslim population. Washington DC: Pew Research Center. http://www.pewforum.org/2017/11/29/europes-growing-muslim-population/. Accessed on 8 February 2018. Prestage G, Ferris J, Grierson J, Thorpe R, Zablotska I, Imrie J, Smith A and Grulich A E (2008) Homosexual men in Australia: population, distribution and HIV prevalence. Sexual Health 5(2): 97–102. https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/same-sex-couples-in-us-census-bureau-data-who-gets-counted-and-why/ https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/same-sex-couples-in-us-census-bureau-data-who-gets-counted-and-why/ https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/how-many-people-are-lesbian-gay-bisexual-and-transgender/ https://williamsinstitute.law.ucla.edu/research/census-lgbt-demographics-studies/how-many-people-are-lesbian-gay-bisexual-and-transgender/ https://williamsinstitute.law.ucla.edu/wp-content/uploads/lgbt-demogs-sep-2014.pdf https://williamsinstitute.law.ucla.edu/wp-content/uploads/lgbt-demogs-sep-2014.pdf https://www.legislation.gov.au/Details/C2017C00382 http://melbourneinstitute.unimelb.edu.au/assets/documents/hilda-questionnaires/SelfCompletionQuestionnaireW12M.pdf http://melbourneinstitute.unimelb.edu.au/assets/documents/hilda-questionnaires/SelfCompletionQuestionnaireW12M.pdf http://www.ons.gov.uk/ons/guidemethod/measuring-equality/equality/sexual-identity-project/measuring-sexualidentity–an-evaluation-report.pdf http://www.ons.gov.uk/ons/guidemethod/measuring-equality/equality/sexual-identity-project/measuring-sexualidentity–an-evaluation-report.pdf https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/sexuality/articles/subnationalsexualidentityestimates/uk2013to2015 https://www.ons.gov.uk/peoplepopulationandcommunity/culturalidentity/sexuality/articles/subnationalsexualidentityestimates/uk2013to2015 http://www.pewforum.org/2017/11/29/europes-growing-muslim-population/ 38 Wilson T and Shalley F Australian Population Studies 2 (1) 2018 Richters J, Altman D, Badcock P B, Smith A M A, de Visser R O, Grulich A E, Rissel C and Simpson J M (2014) Sexual identity, sexual attraction and sexual experience: the Second Australian Study of Health and Relationships. Sexual Health 11(5): 451–460. Richters J, Badcock P B, Simpson J M, Shellard D, Rissel C, de Visser R O, Grulich A E and Smith A M A (2014) Design and methods of the Second Australian Study of Health and Relationships. Sexual Health 11(5): 383–396. Smith A M A, Rissel C R, Richters J, Grulich A E and de Visser R O (2003) Sex in Australia: sexual identity, sexual attraction and sexual experience among a representative sample of adults. Australian and New Zealand Journal of Public Health 27(2): 138–145. Smith L R (1980) The Aboriginal Population of Australia. Canberra: ANU Press. Statistics Canada (2017) Sexual orientation. http://www.statcan.gc.ca/eng/dai/smr08/2015/smr08_203_2 015#a3. Accessed on 8 February 2018. Taylor J (2011) Postcolonial transformation of the Australian Indigenous population. Geographical Research 49(3): 286–300. Van Kampen S, Fornasiero M, Lee W and Husk K (2017) Producing modelled estimates of the size of the lesbian, gay and bisexual (LGB) population of England: final report. London: Public Health England. Wang Y and Kosinski M (2018) Deep neural networks are more accurate than humans at detecting sexual orientation from facial images. Journal of Personality and Social Psychology 114(2): 246–257. Wilkins R (2015) The Household, Income and Labour Dynamics in Australia Survey: selected findings from Waves 1 to 14. 10th Annual Statistical Report of the HILDA Survey. Melbourne Institute, The University of Melbourne. Wilkins R (2017) The Household, Income and Labour Dynamics in Australia Survey: selected findings from Waves 1 to 15. 12th Annual Statistical Report of the HILDA Survey. Melbourne Institute, The University of Melbourne. Wilson S (2004) Gay, lesbian, bisexual and transgender identification and attitudes to same-sex relationships in Australia and the United States. People and Place 12(4): 12–21. Wilson T (2016) Visualising the demographic components of change shaping the age structures of Australian state and territory populations. Poster presented at the Australian Population Conference, Sydney, December 2016. https://tinyurl.com/y7622g4z. Accessed on 8 February 2018. Wooden M (2014) The measurement of sexual identity in Wave 12 of the HILDA Survey (and associations with mental health and earnings). HILDA Project Discussion Paper Series No. 1/14, Faculty of Business and Economics, The University of Melbourne. http://www.statcan.gc.ca/eng/dai/smr08/2015/smr08_203_2015#a3 http://www.statcan.gc.ca/eng/dai/smr08/2015/smr08_203_2015#a3 https://tinyurl.com/y7622g4z