Saturday, December 9

How I answered the census ‘gender’ question

I don’t have a gender identity – and you probably don’t have one either.

New Zealand’s 2018 census was widely regarded as a failure, leading to the resignation of the Chief Executive of Stats NZ. This year’s census has also become mired in controversy. This time, the controversy is due to a poorly defined and confusing question about ‘gender’.

Outside of those who are well versed in the culture wars, few New Zealanders will have fully understood the meaning and implications of this question. But many people know instinctively that it doesn’t feel quite right. For those who care about responding honestly and accurately, answering this question is a minefield.

The information collected by the census is important, and I’d encourage everyone to complete it. I also support Stats NZ’s goal of collecting more data that can help agencies to serve New Zealand’s transgender population. However, Stats NZ needs to redesign how it achieves this goal.

What the census gets right

Unlike some surveys, the 2023 census asks respondents about their sex. Moreover, there are only two options – male and female. The census correctly asks about differences of sex development (also known as ‘intersex’ conditions) in a separate question. The census design thus avoids promoting the false idea that sex is a spectrum.

However, the census also asks people about their ‘gender’. And there are significant problems with the gender question.

The Stats NZ definition of gender is incomprehensible

The census defines gender as a “social and personal identity as male, female, or another gender”. But what does it mean to identify as “female” (for example) in this context? Clearly, the definition can’t be referring to identifying as biologically female, or else gender would just be a synonym for biological sex (which Stats NZ insists it is not).

Perhaps Stats NZ has defined gender more clearly elsewhere? If you Google, you can find the official Stats NZ definition of gender, which is:

Gender refers to a person’s social and personal identity as male, female, or another gender or genders that may be non-binary. Gender may include gender identity and/or gender expression.

Since this definition refers to ‘gender identity’, I looked up this definition as well. The Stats NZ definition of gender identity is:

Gender identity refers to a person’s internal and individual experience of gender.

These definitions are a complete mess. Gender is defined as a “personal identity” that is somehow distinct from gender identity. Gender might include gender identity, or it might not (as implied by the use of “and/or” in the Stats NZ definition of gender). And the Stats NZ definition of gender identity refers circularly back to gender.

The Stats NZ definition of gender also states that it might be based on your ‘gender expression’. The Stats NZ definition of gender expression is:

Gender expression refers to a person’s presentation of gender through physical appearance – including their dress, hairstyles, accessories, cosmetics, mannerisms, speech, behavioural patterns, names, and personal references. Gender expression may or may not conform to a person’s gender identity.

Read literally, this maze of definitions leaves it unclear whether the census gender question is asking about your “internal and individual experience”, or about your hairstyle. If a woman cuts her hair short, does that make her transgender? The Stats NZ definition of gender suggests that it might.

We should be able to read official statistical definitions literally, and have them make sense. As Stats NZ’s own survey design manual notes, “using language that is hard to understand, or overly technical and full of abbreviations and words that are not defined, makes it difficult for respondents to answer as intended”. Confused definitions lead to confused data. 

Unfortunately, Stats NZ’s confused definitions mean that census results relating to gender will be difficult to interpret. This is poor practice, and Stats NZ should be asked to do better.

The 2021 census of England and Wales should offer a warning to Stats NZ. This census was the first in the world to include a question about gender identity. It produced results that seem unreliable and difficult to believe. For example, it found that “one in every 67 Muslims is transgender”.

This and similar findings were likely due to widespread confusion about what the gender identity question meant. One of the strongest predictors of reporting a trans identity was having English as a second language.

What is gender (identity)?

While Stats NZ doesn’t provide a coherent definition of gender (other than suggesting that it might depend on your hairstyle), we can gain a bit of clarity by referencing other official sources.

For example, the World Health Organisation defines gender as the “norms, behaviours and roles associated with being a woman, man, girl or boy”. Similarly, Merriam Webster defines gender as “the behavioural, cultural, or psychological traits typically associated with one sex”—in other words, sex stereotypes.

Gender identity refers to the gender role that someone identifies with. For example, having defined “gender” to mean “gender role”, the World Health Organisation proceeds to define gender identity as “a person’s deeply felt, internal and individual experience of gender”. In other words, gender identity refers to an affinity with male or female stereotypes.

Confusingly, gender identity is often referred to as “gender” for short. In keeping with this, the census gender question appears to be asking about gender identity (and it would be a lot clearer if it was written accordingly).

It’s important to understand that adopting a gender identity does not mean identifying with your biological sex (or the opposite sex). Rather, as the World Health Organisation definitions imply, it means embracing a set of sex stereotypes.

If you’re not yet convinced of this, I highly recommend reading philosopher Kathleen Stock’s book Material Girls, which explains the origins of the concepts of gender and gender identity very clearly. Similarly, this article surveys the definitions of gender provided by a range of official sources, and my widely-shared article in Reality’s Last Stand shows how these concepts are explained to children in schools.

My experience has been that any definition of gender identity that claims not to be based on sex stereotypes quickly collapses under close examination.

Do I have a female gender (identity)?

I’m a woman, and I don’t consider myself to be transgender. Given this, I think it’s safe to say that Stats NZ expects me to happily tick the “female” gender box in the census.

Yet, like most women in New Zealand, I don’t embrace rigid female sex stereotypes. I grew up in Latin America, where machismo is still rife. The traditional female role involves sole responsibility for housework and caregiving, and a subservient position to men. I don’t want to define myself by this role.

Nor do I feel any deep inner sense of being female, other than being aware of my female body (i.e. my biological sex). Since I’ve never been a man, I don’t know how being male would feel. And if I was able to magically transform myself into a man, it seems logical that any different feelings I experienced would be caused by having male biology, or by how I was treated by society due to my biology.

The non-binary dilemma

So perhaps I don’t have a female gender identity – perhaps I’m non-binary? After all, I can relate to some female stereotypes and some male ones. And I’d like to think that I have some unique aspects to my personality that don’t neatly fit either male or female stereotypes.

It’s here that one of the central contradictions of gender ideology kicks in. While gender identity is supposedly a purely internal experience, it is also closely tied to real-world physical changes.

If I identified as non-binary, then I’d be expected to adopt non-standard pronouns (they/them, or perhaps zhe/zher). I would also be describing myself as a potential candidate for medical procedures to align my body with my new identity. 

These body modifications could involve binding my breasts so tightly I could no longer breathe properly, and eventually having them surgically removed. Or they could involve having my genitals excised. Without these body modifications, it would be (falsely) supposed that I could never be happy, and might even commit suicide.

In identifying as non-binary, I would also be endorsing the idea that everyone else (who doesn’t identify as non-binary) is binary. That is, I would be implying that I expect them to fit rigid gender stereotypes.

None of this appeals to me. This is why I refuse to adopt any gender identity at all.

Am I agender?

So, since we’ve established that I have no gender identity, I can just answer the census gender question by choosing “Other” and writing “No gender”, right?

Unfortunately, if I did this, it could cause some confusion. In gender activist circles, having no gender identity makes you “agender” – neither a man nor a woman. 

Like non-binary people, agender people are expected to use they/them pronouns. They are also considered candidates for genital nullification surgery, to “affirm” their lack of a gender identity. This is not really a category I want to put myself into.

When people write “none” in the census, Stats NZ will not know whether they consider themselves to be agender, or whether they reject the concept of gender identity altogether. The failure to distinguish between these two very different groups is a significant flaw in the design of the census.

The census gender question reflects a toxic belief system

The census gender question sits in the context of a broader political push by gender activists to replace biological sex with self-declared gender identity in law and society. One example is the incredibly unpopular drive to imprison male sex offenders in women’s prisons. Another is the offensive practice of forcing women to compete against males in sports. But the most damaging aspect of this political movement has been the needless medicalisation of children who don’t conform to sex stereotypes, with serious and lifelong health consequences. The damage to young New Zealanders has been especially profound.

Gender activism is animated by a toxic belief system known as gender identity theory, or gender ideology. According to this theory, what makes you a man or a woman is not your biology, but your gender identity (i.e. your subjective feeling of maleness or femaleness).

Belief in gender ideology is distinct from being transgender. Most advocates of gender ideology are not trans. And people can choose to medically transition despite rejecting gender ideology. In fact, people started medically transitioning long before gender ideology emerged. So we can (and should) accept trans people, while rejecting gender ideology as a belief system.


There are supernatural elements to some variants of gender ideology. Many adherents believe that gender identity is an ineffable essence or soul that everyone has, but only transgender people are truly in touch with (similar to the religious concept of being touched by the Holy Spirit). Some adherents even believe that a gender identity can transform human flesh, changing men into literal biological women (this is analogous to the Catholic doctrine of transubstantiation). And true believers seek to damn those who question these doctrines as modern-day heretics.

Because it contains supernatural elements, gender ideology is essentially a religion. No one should be forced to endorse this belief system, in the same way that no one should be forced to endorse the beliefs of any other religion.

Gender ideology was the driving force behind the recent violence against women at the Let Women Speak event in Auckland. Similar incidents have occurred overseas. Many New Zealanders now recognise the alarmingly intolerant nature of gender ideology, and want nothing to do with it.

Unfortunately, the fingerprints of gender ideology can be found all over the census. For example, within gender ideology, biological sex is falsely characterised as arbitrary and changeable. We can see this reflected in the Stats NZ definition of sex, which claims that “a person’s sex can change over the course of their lifetime and may differ from their sex recorded at birth”. Similarly, the 2023 census asks for your “sex at birth”, implying that this might somehow be different from your sex now.

The signal sent by Stats NZ’s ‘Gender by default’ policy is also obvious. Under this policy, statistical reporting is based on people’s gender identities, and information about their sex is hidden most of the time. This policy thus implies that your gender identity is more important than your sex. I disagree with this premise, and I don’t want the data I provide to Stats NZ to be abused in this way.

Meekly completing the census in line with gender identity theory feels like endorsing gender ideology, and the political movement it has inspired. I refuse to do it.

A layer of deception

Whether accidentally or deliberately, the incoherent language and definitions used by Stats NZ serve a purpose.

For example, when the census asks for your ‘gender’, and not for your ‘gender identity’, the word gender feels reassuringly familiar. This is because it traditionally referred to biological sex (and is still understood this way by many people). Gender activists have redefined this word by stealth.

It would be easy for Stats NZ to avoid the ambiguous word ‘gender’, and say either ‘gender role’ or ‘gender identity’ depending on their intended meaning. But that would make the absurdity and radicalism of gender identity theory obvious.

New Zealanders deserve a higher level of transparency and honesty from Stats NZ.

Forcing compliance with gender ideology 

The census provides no guidance for people who don’t believe that the concept of gender identity applies to them. I wrote to ask what I should do if I objected to the gender question:


How can I respond to the question ‘What is your gender?’ In order for it to reflect that I don’t have a gender? I don’t believe in gender identity, which is the definition you are using for the term gender. 

I object to this question, as there’s no option to truthfully answer it when you don’t uphold gender identity as a belief system. 



I received the following response:

Hi Laura,

All census responses are recorded, stored, and output securely, following our confidentiality rules. We would encourage you to respond accurately to all census questions to best inform decision making for New Zealand. However if you prefer not to disclose your gender, we recommend giving a response only to gender, but not sex at birth.

Please do not hesitate to contact us, either via the online General Enquiries form or on the phone number below, if we can be of further assistance. 

Kind regards

[Name redacted]

Customer Service Specialist

Toll free helpline 0800 CENSUS (0800 236 787)

This response is obvious nonsense. In my email, I clearly stated that I disagree with gender identity as a belief system. Yet the Stats NZ response presupposes that I do have a gender identity (which they imply that I want to hide). And then for inexplicable reasons, they suggested that I answer the question about gender and not the question about sex!

The census needs to change

Because the gender question in the 2023 census is confusing and poorly defined, data resulting from this question is likely to be misleading. It should be discarded, or at least interpreted with extreme caution. Data from the question about sex is likely to be much more meaningful and reliable, and it is this data that should guide important public policy decisions.

In future surveys and censuses, Stats NZ should avoid the use of the ambiguous term ‘gender’, replacing it with ‘gender identity’. It also needs to provide a clear, non-circular definition of what it means by gender identity. And it needs to provide meaningful advice on how people can respond to questions about gender identity if they disagree with gender ideology.

Given already low census response rates, Stats NZ cannot afford to alienate large swaths of the population by forcing them to endorse an ideology they object to. To avoid this problem, ideally, future censuses would include a yes/no question asking whether people consider themselves to have a gender identity. Only those who say ‘yes’ should be asked to describe what that gender identity is. At the very least, Stats NZ needs to provide an explicit option for people to select if they don’t buy into gender ideology (e.g. “I don’t believe the concept of gender identity applies to me” or “I refuse to answer”).

Stats NZ has also said that if people leave the gender question blank, then they will fill in a response for them without their permission. This practice violates people’s right to ensure that data held about them is accurate. For people who reject the concept of gender identity, a blank response may be the most accurate answer. Stats NZ needs to find a different way around this problem (e.g. by classifying the person as non-transgender, without making any assumptions about their gender identity).

Future surveys can also obtain more useful information about the transgender population by asking directly about past and current use of cross-sex hormones – not just about gender identity. Asking about medical treatments received will provide data that is far more helpful for healthcare planning than subjective and ambiguous questions about identity (especially given the numerous health problems associated with taking cross-sex hormones).

Finally, the ‘gender by default’ policy needs to end, so that census data can easily be analysed by sex. We cannot just assume that gender identity is more important for outcomes than sex – only data can tell us that. To hide this data from researchers is enforcing ignorance in the name of protecting a sacred ideology. Such a policy has no place in a free, open, and secular society.

How I responded to the gender question 

My response to the census gender question was “The concept of gender identity is not relevant to me”.

It is my hope that Stats NZ will not violate my rights, and corrupt the census data, by changing my response to mean something different. I believe the census is important, so I’ve put my faith in the system and completed it as best I can, despite my misgivings. Time will tell whether I’ve made the right decision.

  • Laura López is the mother of two girls and holds a graduate degree in psychology. She writes Arguments With Friends on Substack.
  • This article was first published in The Standard, where there’s a lively discussion that you can join here.