The Ministry of Everyone Else

Guest essay by Tze Ming Mok

Despite what you might have been told, the main purpose of collecting and analysing data on ethnic identity in Aotearoa New Zealand is to lay bare structural racism and inequality so that it can be eliminated.¹ But ethnic data collection and categorisation in our particular nation-state is unavoidably awkward, often stupid-sounding, and usually insults somebody, because it is an attempt to redeem a historically racist practice by continuing it.

Originally birthed alongside slavery in the service of European imperialist expansion, our modern-day incarnation of counting up ethnic groups out here in the colonies, tries to refashion the tools of administrative oppression for the purposes of something more liberatory, or at least, liberal. In service of this often shaky project of redemption, Western governments generally try to collect “information about identities that mark out difference; that are associated with disadvantage; and that can be agreed upon” (Mok, 2019, p. 37).

Agreed upon! Ambitious in itself. In the mid-2000s, unrepresented and Ministryless, myself and another ‘Asian’ pondered whether we should be campaigning for a ‘Ministry of Asian Affairs’, a ‘Ministry of Ethnic Affairs’, or neither, and what labels or categories made sense to describe all the people in Aotearoa who were neither the white majority, nor Indigenous to Oceania (Mok & Rasanathan, 2006). Let us briefly review some terrible options we examined in the Aotearoa Ethnic Network Journal.²

At the time, Aotearoa had a rapidly burgeoning population sitting under an ‘Asian’ label that left many of us ill-at-ease, due to its erasure of wildly diverse identities and social outcomes (Mok, 2022). Also emerging into public consciousness was the even more conceptually disconnected Statistics NZ category covering Middle Eastern, Latin American and African communities – ‘MELAA’. At the time, I thought the regions had been ordered contrary to geography in this acronym because otherwise it would have been LAAME. Meanwhile, Australia had come up with the term CALD – Culturally And Linguistically Diverse. That’s right, people like us really got called CALD. Maybe we should have been named NAMD, just so the government could really hammer home how much of an afterthought we are. It could have stood for ‘Not Another Minority Department’. I suggested at the time, that for the explicit purposes of politicised strategic essentialism, we may as well call ourselves NWWOM – Non-White Without Own Ministry, with the upside of sounding like an Orientalist gong sound-effect: NWOMMMMM.

All this pioneering thought-leadership went to waste, because years later in 2021 we ended up with a ‘Ministry for Ethnic Communities’ (MEC), so we were no longer Without Own Ministry, and even if we fell in a forest, we could not make a NWOMM. But we were now expected to bear the national burden of being Ethnic – unlike all other ethnic groups who were, by implication, lacking ethnicity. We were the leftovers, the residual categories, represented by the Ministry of Others. We found that some white people were allowed to sit with us too, even if they were fond of blackface. Appropriately, MEC is French slang for ‘bro’ or ‘man’ or ‘dude’ – in other words, MEC sounds like an especially European person trying to sound cool so you will let them be your friend. Jokes aside (and sorry that there were so many), we are not together because we are the same. We are together – ‘lumped’ together is a common phrase you will hear – because we are not big enough on our own to create enough space that matters politically.

But once we get it, what should we be using that space for? Asian communities (as the biggest Ethnic of the Ethnics) have elaborated extensively, especially in public health, on the need for government to understand that when it uses these umbrella categories for any real purpose, those uses must be premised upon difference. Great stats for some Asian communities who had privileged migration pathways into Aotearoa (like my own 1970s native-English-speaking Commonwealth Doctor Skills Shortage generation), masks bad stats for other Asian communities who may have fled war, coups and genocides. So the data needs to be split up – disaggregated. Expert after expert has demonstrated that if the resonance and recognisability of the ‘Asian’ brand in public policy is not used for understanding the disaggregated outcomes and diverse needs under the label, that it is worse than useless. (Ho, 2015; Horner & Ameratunga, 2012; Liao, 2019; Parackal et al., 2021; Patel et al., 2022; Peiris-John et al., 2021).

The same goes triple for ‘MELAA’. The provenance of this unfortunate and counterintuitively-ordered acronym is actually not avoidance of being called ‘LAAME’ (socially unacceptable) or ‘MEALA’ (a food delivery app?) but size order. They are literally the next biggest three cabs off the rank in order of size, that got pulled out of the ‘Other’ category, granted the privilege of at least being specifically named (if not NAMD) – with the alternative being no-one knowing they existed at all. Defensible maybe from a statistical perspective if you want to make as few people ‘Others’ as possible; but indefensible as an umbrella category on which we would base any real analysis, public policy, or even public communication (Armah, 2022).

Labels and names matter – we wear them, under sufferance, all our lives. They should at least make some kind of sense. Thus: Data disaggregation should be the priority for all of us huddling under the ‘ethnic’ umbrella. Here, happily, Figure.nz has hit the motherlode. There is a common misconception that Statistics NZ does not ‘have’ detailed data about individual ethnic groups in the Census because the main statistics we see reported in the news or on Statistics NZ summary pages fall under the Big Four ‘Level 1’ categories of NZ European, Māori, Pasifika, and Asian, with MELAA and ‘Other’ sort of tacked on the end. When people at my office tell me that basic forms of ethnically granular data are ‘not there’, I go to NZ dot stat dot stats dot govt dot nz and point at my screen to show them how many self-identified Malaysian and Singaporean Chinese (Level 4 categories) were left in Mt Roskill in 2013 when I was out of the country, while they sit there and say with incredulity, ‘how old is this website?’

A wide range of 2018 Census data, such as on income, occupation, education, language, and religion is available at NZ.Stat (soon to be getting an interface refresh) by Level 3 ethnic breakdown within geographic areas, which provides data for ‘Middle Eastern’, ‘Latin American’ and ‘African’ separately; and names the biggest specific Asian ethnic groups: Chinese, Indian, Filipino, Korean, Vietnamese, Cambodian, and Japanese. A relatively simple custom request for data to Statistics NZ would net you detailed breakdowns of individual Level 4 ethnic group, which is where you name absolutely everyone possible, as long as the numbers are not too small for protecting anonymity. But those barriers (and the formatting challenges of NZ.Stat downloads) are often a bridge too far for the ordinary user who needs some ‘stats’ stat.

This is why this Figure.nz microsite is such an important resource, as they have uploaded a plethora of their trademark simple chart breakdowns, disaggregated at the finest ethnic level possible, using a custom 2018 Census Level 4 dataset that Statistics NZ provided to our bros, our dudes, our MEC - the Ministry of Ethnic Communities. The main dataset has been cleaned up by Figure.nz and is downloadable here as an Excel file, for the ever-so-slightly more advanced, but not really all that advanced user. This microsite charts the breakdowns of key Census variables such as occupation, education, industry, and languages spoken, for the Level 4 ethnic groups within ‘Asian’, ‘Middle Eastern’, ‘Latin American’ and ‘African’ categories, as well as data from other sources that we might feature in, although mainly at less detailed levels.

Looking at the Level 4 data is a tidy lesson in the untidyness of what is considered an ‘ethnic group’. We can see that the labels ordinarily considered a ‘real’ ethnic group, such as ‘Indian’ or ‘Chinese’ are colonial and postcolonial nation-state constructs that come unmoored through centuries of diaspora. For example, here are all the counts of detailed ethnic group of all the groups with connection back to the historical territory known variously as ‘India’ or ‘Hindustan’ or ‘Bharat’ or ‘the British Raj’. This includes (as you will see by the code numbers) specific ‘write-ins’ that get likely get back-coded to just ‘Indian’ in the higher level counts, and some that aren’t. I have not included ethnic groups specifying Sri Lanka or the Maldives as I’m probably going to get in enough trouble for this as it is.

Code	Label	Number
43100	Indian not further defined	221916
43111	Bengali	201
43112	Fijian Indian	15132
43114	Indian Tamil	315
43115	Punjabi	459
43116	Sikh	192
43117	Anglo Indian	381
43118	Malaysian Indian	474
43119	South African Indian	1632
43199	Indian not elsewhere classified	348
44412	Bangladeshi	2337
44414	Pakistani	6135

In India, the term ‘Indian’ is a nationality, not an ethnic group; and the same goes for Bangladeshi and Pakistani. All three countries have Bengalis and Punjabis, and also have Sikhs, which is a religion, not an ethnicity, and is counted separately in the religion question (see my graphs and maps of this data here), although is associated with an independence movement (which may see you assassinated by the Indian secret service. That’s the country of India’s secret service, not a secret service for ‘ethnic Indians’, which aren’t an ethnic group in the country of India). Ultimately, some people are just really keen on writing ‘Sikh’ in as their ethnic group. The diaspora is where homelands become imaginary, and what could be a greater work of imagination than creating your official identity through acts of administrative form-filling? Meanwhile, South African Indians may have originally arrived in Africa in the 19th century from a place that is now Bangladesh, but who knows, so they get backcoded as Indian at Level 3. Write-in Punjabis of any kind may have migrated from the current country of Pakistan, who knows, yet they are also backcoded as Indian at Level 3.

You see how working for Statistics NZ ethnic classification department is probably a huge pain in the ass? And that’s just working with the data, not dealing with people like me who complain a lot. The important thing to realise is that these types of people with historical ethnic connections to the ‘Indian subcontinent’ have different types of outcomes, and to ask yourself why. For example, you can see in this chart that the small number of Indians who described themselves as ‘Indian Tamil’ were the most tertiary-educated of all of the different kinds of Asians full stop, followed closely by Pakistanis; while those who identified themselves as Fijian Indians had the lowest level of tertiary education of all types of ‘Indians’, at a similar level to Southeast Asian communities associated with forced migration and refugee pathways into Aotearoa.

We see the same kind of complexity for different types of ‘Chinese’ people listed below. I have assigned various kinds of trouble I might be in here, to a footnote.³

Code	Label	Number
42100	Chinese not further defined	231387
42111	Hong Kong Chinese	3177
42112	Cambodian Chinese	1413
42113	Malaysian Chinese	4866
42114	Singaporean Chinese	675
42115	Vietnamese Chinese	609
42116	Taiwanese	6570
42199	Chinese not elsewhere classified	222

Debates about ‘what is Chinese’ aside, all those who at least said that they were Chinese all have different outcomes – 46% of those declaring themselves Singaporean Chinese are professionals, compared with 14.6% of those who listed themselves as Cambodian Chinese, as you can see in this chart. I wrote a more detailed paper on this with regard to the UK Chinese population, disaggregating Chinese communities by specific countries of birth, and called it ‘All look the same?’ (We do not.) The MEC data request would have been usefully enhanced by including detailed Census country of birth by ethnic group as well (as far as possible), to get a more explicit understanding of the size and proportions of our culturally, socioeconomically, and politically divided communities. For example, this would really help me improve the maps I made to figure out the least awkward place for me to do my Asian grocery shopping if/when China invades Taiwan.

Any future requests of Level 4 data should also make sure to include income data as per the Level 3 tables available on NZ.Stat, if we are really to get down to the nitty-gritty. But what we have here is a good start. It is particularly good for easy public access and use, and is fascinating to browse if you ever wanted to know, for example, what proportion of Burmese here identify as Muslim, or which Middle Eastern group is the most likely to be NZ-born (it’s the Lebanese). Interestingly, despite the strong dislike of the MELAA category itself, for the counts of each of its subgroups, the generic categories of ‘Middle Eastern’, ‘Latin American’ and ‘African’ appear to be the most popular choices as ethnic groups (not counting white South Africans that is) despite not really being individual ethnic groups – as if these communities are adopting the categories that give them the most social use value for their context. The generic ‘Asian’ however, is not so popular among ‘Asians’. Just as non-Asian New Zealanders have been complaining about ‘Asians’ for decades, so have Asian New Zealanders been complaining about the word ‘Asians’ for decades.

Each chart has a link to the full MEC dataset in the metadata scrolldown, which includes counts as well as percentages in one place, for all detailed Level 4 ethnic groups for the whole country (not just for Asians and ‘MELAA’). I would recommend for those interested in social outcome data, that the full dataset be used for the national and Pākehā rates as a comparison point, so we actually have an idea of what we should be aiming for.

The real battle ahead for disaggregation is not with Census data, nor with survey data which is increasingly able to disaggregate broad Asian subgroups (Peiris-John et al., 2021), given the large size of our overall population, especially in Auckland. The main event should be improving disaggregated reporting from administrative data. By ‘administrative data’, I mean the vast troves of information generated by every contact we have with public services, across health, education, the tax and property system, housing, benefits, immigration, and all other government services, that track how we are doing in our real lives. This is where we can learn far more about social disparities and extra needs in detail, than from the clutch of 5-yearly data points that the Census provides.

When it comes to public services, policy standards around data collection mean that granular ethnic data is there, and if adequately resourced, able to be analysed and reported on crucial issues. Currently, it is only sporadically reported for the public in a detailed way – for example, one of the most useful disaggregated reports using administrative data to sum up the health of subgroups trapped inside the ‘MELAA’ category was more than a decade ago (Perumal, 2010). The Level 2 ethnic breakdowns provided by the Ministry of Health while we chased our Covid-19 nationwide vaccination targets in 2021, showed how valuable it was to understand the high and low rates by age bands of even slightly more specific ethnic groups beyond ‘Asian’. For example, we found out that despite very high vaccination rates among working-age Chinese, elderly Chinese people had very low rates – possibly due to factors such as relying on masking and isolation instead of vaccination, or not being adequately reached by health services with limited cultural outreach capacity. It also showed how fast this data can actually be made public when it becomes a priority.

All public sector data is linkable through the Integrated Data Infrastrastructure, but access to and usage of that data is a story that would take a whole other website. For now, I would note that it is an ongoing struggle to persuade agencies that are meant to be here for our welfare, to tell us what they know about us, or to look for the first time at our data to figure out what information they hold on us, including what we need and where we are being failed, and in a way that is understood to be not a matter of pitting ‘Asian’ or ‘MELAA’ interests against Māori and Pasifika prioritisation. Depending on the direction of government, the use of the IDI for either ‘social investment’ or ‘social wellbeing’ may carry increasing risks around transparency, social license and Indigenous data sovereignty even as analysis could continue to fail us on ethnic disaggregation. There is both a mountain of potential and a cliff-edge of risk.

From down here in the weeds, one of the things we can do is normalise the public use and visibility of detailed and disaggregated ethnic data, to show it is something – maybe the one single thing – that ‘ethnic people’ do agree on, and want more of.

Tze Ming Mok is a Mt Roskill-born Malaysian-Singaporean Chinese writer and researcher. Her PhD thesis on ethnic classification and enumeration in the UK was awarded best Social Policy PhD by the London School of Economics in 2020, but she obviously wasn’t trying quite so hard with this article.

References

• Armah, M. (2022). Ethnic minorities want “crude” MELAA classification changed for Census 2023. Stuff.
• Ho, E. (2015). The Changing Face of Asian Peoples in New Zealand. New Zealand Population Review, 41, 95–118.
• Horner, J., & Ameratunga, S. N. (2012). Monitoring immigrant health and wellbeing in New Zealand: Addressing the tyranny of misleading averages. Australian Health Review, 36(4), 390–393.
• Liao, R. (2019). In the shadow of exclusion: The state of New Zealand Asian health. New Zealand Medical Student Journal, 29, 32–36.
• Mok, T. M. (2019). Inside the box: Ethnic choice and ethnic change for Mixed people in the United Kingdom (Issue December). London School of Economics and Political Science.
• Mok, T. M. (2022). Black Asian, White Asian: Racial Histories and East Asian Choices in the White Settler State. In Towards a Grammar of Race in Aotearoa New Zealand (1st ed., pp. 126–137). Bridget Williams Books.
• Mok, T. M., & Rasanathan, K. (2006). Should we be pushing for a Ministry of Asian Affairs, a Ministry of Ethnic Affairs, or neither? A ‘Socratic’ ‘dialogue’ between two ‘Asians.’ Aotearoa Ethnic Network Journal, 1(1), 41–45.
• Parackal, S., Coppell, K., Yang, C., Sullivan, T., & Subramaniam, R. (2021). Hidden figures and misnomers: A case for disaggregated Asian health statistics in Aotearoa New Zealand to improve health outcomes. The New Zealand Medical Journal, 134, 109–116.

Footnotes