Guest post; Implementing the Statistical Standard for Gender Identity: How should Stats NZ ask the question?

Gloria Fraser, Victoria University of Wellington

Statistics New Zealand’s decision to include categories beyond ‘male’ and ‘female’ in a new gender identity statistical standard has been hailed as a “world first”. And, unfortunately – it is. We live at a time when an exploding body of research documents the alarmingly high levels of violence and discrimination experienced by trans people, when trans women and girls appear on the cover of Time and National Geographic magazines, and when we are having more conversations around gender neutral bathrooms than ever before. In spite of this, the overwhelming majority of gender identity questions on surveys, healthcare intake forms, and censuses around the world continue to offer just two response options. With this standard (and restrictive) measure of gender identity, it is impossible for trans people to be counted. Nonbinary people are rendered entirely invisible; they cannot select either gender item, so are excluded from reporting their gender at all.

New Zealand Census questions on sex, taken from the 1916, 1986, and 2006 questionnaires. The 2018 Census will likely be the first ever to ask New Zealanders about their gender identity.

The use of a ‘male’/’female’ tick box to measure gender identity is more than just poor methodology; this has serious consequences for the health, wellbeing, and social inclusion of trans people. Without accurate gender identity data it is impossible to establish the size of the trans population in New Zealand. International research estimating the proportion of the population who are trans produces wildly varying results; data from presentations to overseas gender clinics give estimates as low as 4.6 people in 100,000 (perhaps because they ignore that not all trans people present to specialist clinics for gender-affirming healthcare). Another study claims that the number of people falling under the trans ‘umbrella’ may be as many as 1% of the population.

Population size partially determines the amount of funding that is allocated to serving the needs of a particular group. Because most official records do not capture data on trans people, trans New Zealanders are, most likely, receiving far less than their fair share of medical and mental health care. On top of this, with every research survey, census, and demographic form that fails to acknowledge the fluid and non-binary nature of gender identity, the common cultural gender binary is legitimised and reinforced. We need gender identity measures that challenge this binary, ensure all New Zealanders are counted, and give people the opportunity to correctly identify their gender.

It is wonderful that Statistics New Zealand has decided to collect gender identity data, and to ensure that their measures are inclusive of trans and nonbinary people. For the reasons outlined above, the importance of this data cannot be understated. If Statistics New Zealand do this right, New Zealand would be in an absolutely unique position. We would have population-level data about the needs of trans people, which could be linked with health data (e.g., cancer registrations) to generate urgently needed transgender cohort studies.

Statistics New Zealand has not announced how the gender identity question in the 2018 census will be worded, and when I contacted them recently I was told that this decision has not yet been made. The classification of gender identity that has already been released, however, suggests that the direction in which Statistics New Zealand is heading may not be quite what we hoped.

stats nz.png
Statistics New Zealands classification and coding process, released July 2015.

Upon release of the statistical standard, trans community members pointed out that this method of classifiying gender identity is othering – it separates trans and nonbinary people from the cis population. Statistics NZ has also used terminology which is frustrating for many in the community, such as “transgender male to female” and “transgender female to male”. A recently published paper by leading scholars in the field reflects these concerns: Pega and colleagues note that the standard may “obscure some of the complexity within the broader transgender population”, is not intersex inclusive, and does not guarantee that all trans people will be counted. What is to stop a trans woman from identifying as female or wahine, either to avoid othering herself, or because this is a more accurate reflection of her identity than ‘trans’ or ‘gender diverse’?

I do not envy the job of Statistics NZ. Somehow, they must (1) collect data that identifies trans New Zealanders, while (2) not othering trans people. It seems that it is easy to meet one goal, but is it possible to do both? Let’s think through some options.

stats nz 2.png
Statistics New Zealand’s suggested examples for phrasing gender identity questions.

Personally, I’m a fan of the open-ended box. It allows people to freely self-identify so it isn’t othering, it doesn’t require people to choose between identity labels, and it doesn’t ask people to take on identity labels that might not feel right for them. This question alone, however, doesn’t get around the problem mentioned earlier: if someone writes in “woman”, how do we know if she is cis or trans? We need to face the fact that this matters – if we don’t know this, we can’t count trans people, and we need good data to fight for policy change and funding increases.

The other two options also face this problem – with just one item, we cannot guarantee that we have identified all trans people. On top of this, I seriously doubt there are many people out there who individually identify as “gender diverse” – while this may be a suitable umbrella term, individual identity labels tend to be slightly more specific.

One alternative option, that may meet both goals of (1) collecting data that identifies trans New Zealanders, while (2) not othering trans people, is to ask the question in two steps, by firstly asking about gender identity, and secondly asking about sex documented on a person’s original birth certificate. This way, trans women and men can identify as women and men, while being identifiable as trans because their assigned sex at birth differs from their gender identity. This option is recommended by the Centre of Excellence for Transgender Health.

Because I doubt that Statistics New Zealand has the resources to code open-ended responses from the entire population, the gender identity question would probably be best answered by selecting options from a categorical list. Respondents should be able to select multiple options, so they do not have to choose between, for example, identifying as a woman and identifying as trans. Empirical evidence suggests that these questions are easily understood by the general population, even if some don’t understand what it means to be trans, genderfluid, nonbinary, or agender. An example follows:

  1. How do you identify your gender? Please tick as many as apply.
  • Male
  • Female
  • Transgender
  • Genderfluid
  • Nonbinary
  • Agender
  • Different identity (please state) _________
  1. What sex were you assigned at birth, on your original birth certificate?
  • Male
  • Female
  • Indeterminate
  • No sex listed


As with any measure of gender identity, this measure is not perfect. Perhaps most worryingly, it asks people to disclose their sex assigned at birth, which has potential to be uncomfortable or distressing. Because of this, it will be important for Statistics New Zealand to explain why this information is so crucial to collect. On top of this, it is impossible to construct a complete list of gender identities, meaning that some identities will always be excluded from listed options. Hopefully, the inclusion of an open-ended box will ensure that people of all genders can accurately describe how they identify, and could be an important space for culturally specific identities such as takatāpui, whakawahine, tangata ira tane, fa’afafine, and akava’ine.

Statistics New Zealand may object to these suggestions on the basis that this kind of information is too complex to collect, code, and analyse. In response to this, I argue that it is unacceptable to lump such a diverse group into one umbrella category, as this leaves unexamined the needs within this group. The time has come for the collection of high quality gender identity data, where people of all genders can identify as they wish and be counted. Statistics New Zealand needs to ensure that no New Zealander remains invisible.

Gloria Fraser is a doctoral candidate at the School of Psychology at Victoria University of Wellington. She is particularly interested in the intersection of sex-sexuality-gender diversity and clinical psychology. Her doctoral research focusses on queer experiences of mental health support in Aotearoa. Gloria is also a research coordinator for the New Zealand Attitudes and Values Study (NZAVS). Gloria has used NZAVS data to develop a gender identity statistical standard for coding open-ended responses. This can be accessed as a technical document via the NZAVS website for use by other researchers.


The Youth’12 Reports on Queer and Trans Students and Issues with Queer Academia

There’s been a lot of work done and released this year on the status of secondary school environments for queer and trans students: primarily the Green Party’s Rainbow Report (talked about here), the University of Auckland-led Youth’12 survey – specifically the reports on trans students and same-sex/”both”-sex attracted students, launched last night, and Dr. Simon Denny’s paper on the impact of supportive school environments on sexual minority students (paper to be published this week in the Journal of Clinical child & Adolescent Psychology). All of these reports provide a hard basis for change in our schools, and will be undoubtedly useful in affecting change over the next few years.

But reading through them, there are issues that arise – granted, they are more symptomatic of academia in general rather than the fault of the authors, who I know are nothing if not well-intentioned.

One of the first issues I have is the way they approach their statistics. In both Youth’12 releases, it is obvious that both queer and trans students suffer more than their cis and straight peers, in multiple areas – most worryingly, areas of mental health, self-harm, and suicide. And while the report on attraction does look at the intersections of ethnicity and deprivation, there is never any report on the intersection of queer and trans experiences – the report fails to confirm whether trans students who are also attracted to the same or multiple genders experience a more severe occurrence of these rates. The report is also somewhat foggy on its methodology when it comes to sex and gender – if I understand correctly, the survey asked a question on the students’ sex and one on their gender, but the report does not clarify which is used to sort students into the “male” and “female” categories.

The second big issue I have actually comes from the authors solving a previous issue – the 2007 survey used the terms “heterosexual” and “non-heterosexual” instead of “opposite-sex attracted” and “same/both-sex attracted”. This was changed between surveys to prevent the categorising of queer students into some “Other” section – the “non” to the heterosexual ‘norm’. While this was definitely a concern that needed to be addressed, the way it was done only excludes people further – as we know, a) sex is not a binary and thus this question excludes intersex students and b) attraction is increasingly less thought of in terms of sex and rather in terms of gender. It is worth noting at this point that during the launch presentation, Dr. Lucassen did occasionally use the term ‘same-gender attracted’ rather than sex, so perhaps this is an issue that the authors are already aware of. Regardless, considering people with intersex conditions are more common than redheads, the exclusion of this group is something that needs to be addressed – though this issue can likely be marked down to the extremely slow nature of academia as a whole.

But sadly, these two issues mean that this data isn’t being used to its full value. Our most valuable reports and data are letting down those who need it most, who are at the intersections of multiple issues.

An interesting point that came up in Dr. Denny’s presentation and paper is that same-sex attracted women in schools do not benefit from supportive school environments in the same way as same-sex attracted men, for whom a supportive school environment resulted in lower rates of depressive symptoms and attempted suicides. While it’s important to note that the sample size of same-sex attracted males was too small to be totally confident, it is an interesting result that I have an anecdotal theory on.

A “supportive school environment” in Denny’s paper includes factors such as whether students felt like adults at school cared about them and whether they felt safe at school, but also included questions on their attitudes of their peers – which I believe is the big factor in the disparity between the results for women and men. The way masculinity is valued in society and the way queer behaviour in hegemonically masculine men is stigmatised and associated with femininity or a lack of masculinity means that homophobia has a heightened affect on young queer men and their social groups compared to young women. This is not to say that women do not suffer from homophobia, to claim such would be ridiculous – but I believe there are heightened amounts of peer pressure and casual homophobia within social groups of young men than there are among women – and as a result young same-sex attracted men are going to be seen to benefit more from a supportive school environment that includes supportive peers.

Finally, the 2007 version of this data, used in Denny’s paper, included a survey question directed at teachers, asking whether they thought their school environment was supportive. The data reflected the results from the same question asked of same-sex attracted males – that is that schools that rated more supportive from both sides, student and teacher, resulted in lowered depressive symptoms for same-sex attracted males. This is an interesting point, considering that the Greens’ report earlier this year showed us that principals and other school staff, along with ERO, are not necessarily reliable when it comes to reporting on queer issues within their school – ERO do not even have a measure for the safety and well-being of queer students within schools. When we see multiple principals saying they simply do not have queer students within their schools, or that their bullying policies already cover and protect all their students, that their students don’t have those levels of problems, it can be hard to trust the reports of staff.

That is where these reports become extremely useful – which was brought up by Dr. Lucassen at the launch. When a principal says “we don’t have those students here,” we now have data proving that queer students are prevalent across the nation, in both low- and high-decile schools, in areas of both high and low deprivation, across ethnicities, and in both rural and urban areas. When a principal says their bullying policies already work, we have data proving that over 60% of our queer youth are afraid that someone will hurt or bother them at school, that around 10% have not gone to school because of this, and that 20% are bullied weekly. When a principal says that things aren’t bad for their students, we have data proving that almost half of queer students have seriously thought about suicide compared to 15% of straight students. And that’s just the “same/both-sex attracted” data – things are the same or worse for trans students.

There are problems with these reports, as there are with essentially every piece of queer academia published. But the latest report shows that things can be changed to fix issues, as with the “het/non-het” question (as misguided as that ‘improvement’ was), meaning that we can aim to do better. And in the meantime, this data truly will be invaluable to bring the hard line to schools, principals, and the Ministry of Education. Things may be getting better, but our youth are still dying, and that’s not good enough.