Recognising indigenous populations in the Census: a short history
The 1967 referendum, 50 years on
27 May 2017 marks the 50th anniversary of a landmark referendum for the Aboriginal people of Australia.
Harold Holt’s government called the 1967 referendum to make two significant changes to the wording of the constitution, to recognise Aboriginal people in Australia, amending the constitution to allow the government to pass laws relating to Aboriginal people (or, more accurately, including Aboriginal people in the laws passed by the government, rather than passing separate federal laws), and counting them in the population.
(Contrary to popular belief, it didn’t confer Australian citizenship on Aboriginal people, nor did it give them the right to vote – prior to this, Aboriginal people mostly had the vote, some as far back as the late 1800s, but this was administered differently in each state. Mostly, Aboriginal people were able to vote in all states from 1949).
Counting Aboriginal people in the Census
It’s also not true to say Aboriginal people weren’t counted in the Census prior to 1967. Aboriginals have been counted, in fact, right back to colonial Censuses. In Victoria, the 1851 Census recorded 2,693 Aboriginal people! (Source: ABS, 3105.0.65.001 Australian Historical Population Statistics), but they were never included in the official population estimate (what we know now as Estimated Resident Population).
Prior to 1967, the Census asked questions on race, and in many years asked rather intrusive questions (for example about percentage Aboriginal blood), which were then used to exclude Aboriginal people from the population, for purposes such as determining electorates (which is one of the primary purposes of the Census and population counts generally).
The most popular referendum in Australia’s history
The 1967 referendum remains the most overwhelmingly supported referendum in Australia’s history. Most referenda don’t get through in this country. Of the 44 items put to the Australian people since 1901, only 8 have been carried. The 1967 referendum was carried by the greatest margin of all, a 91% “Yes” vote.
After 1967, the 1971 Census was the first to officially include Aboriginal people as part of the total population, marking the beginning of the way we record the population right up to the present day.
The modern concept of Estimated Resident Population was introduced after this Census as well, when we started counting people based on where they usually live (initially just at a state and broad regional level; now we can go right down to individual neighbourhoods by usual residence).
We still have a specific question on Aboriginality in the Census. The Census now asks “Is the person of Aboriginal or Torres Strait Islander Origin?” But this is very much a question of “identification”, rather than asking about bloodlines.
Increasing identification
Every Census, the Aboriginal and Torres Strait Islander population grows. The 1971 Census recorded 115,953 Aboriginal people (0.9% of population). Interestingly 5 years prior, before the referendum, the 1966 Census recorded 101,000 – not a lot less!
By 2011, this had grown to 548,126 – a 5-fold increase, with people identifying as Aboriginal representing 2.5% of the total population. While some of this is due to a higher fertility rate among Aboriginal people, the bulk of the change it is due to a greater propensity (or willingness) to identify with Aboriginal and Torres Strait Islander ancestry over time.
The chart below shows the growing Indigenous population of Australia by Census, from 1971 to 2011. It’s also likely that the 2016 Census will show a similar increase of around 20%.
Of course, the Indigenous population is included in the total population for all the Census years included in our profile.id tool. All our profile.id subscribers can see their Aboriginal and Torres Strait Islander population by suburb/district using the “Population summary” page, under the population heading.
If you have id’s social atlas tool, there is a map under the “Diversity” heading, which looks at the distribution of Aboriginal/Torres Strait Islander population across your area.
A continuing story of disadvantage
Unfortunately, despite continuing improvements in recognition and inclusion, Indigenous people remain disadvantaged in many areas, and in many of the measures that correlate with social disadvantage.
Due to lower life expectancy, Indigenous people are generally much younger than the total population. Furthermore, Indigenous people have, on average, lower education levels, lower incomes, and are more likely to live in single-parent families and rent in public housing.
However, this is not the case in all areas. Councils have the option to add an “Indigenous Profile” to profile.id to better understand the changing characteristics of Aboriginal people in your local area, with comparisons to the state average, and to the total population. Many councils already subscribe to the Indigenous profile tool, for example, see the City of Campbelltown’s Indigenous profile here.
Missing the Indigenous Profile for your community?
If you’re a local council interested in better understanding Aboriginal people in your area using .id’s Indigenous profile, get in touch with us and we’ll get you set up. You can subscribe to an Indigenous profile as a “Community of Interest” (a paid module), which allows users to delve deeper into the Census data.
Coming soon: 2016 Census data
New Census data will be released by the ABS later next month, on June 27th and will be updated in our community profiles very soon after that. You can read more about the release schedule for 2016 Australian Census data here.
The new census data will provide valuable insights to the stories of indigenous Australia (we would expect another significant increase in the population identifying as indigenous), as we work to build a better society for all Australians.
And for more information on the 1967 referendum, and how it effected the lives of Aboriginal people at the time, watch “Counted” on ABC TV on Friday 26th May.
.id specialises in analysing and presenting Census data in user-friendly online information tools. You can access .id’s community websites to see the demographic profile characteristics for over 250 local government areas in Australia and New Zealand.
Everybody seems to make a huge mistake in ‘reading’ the Census data since 1971 respecting the Indigenous population. Try this as an exercise: take the data in the 2011 Census and extrapolate back to 1971. Even if you build in ‘no mortality’, the population back in 1971 will turn out to be vastly more than the official figures. With mortality, even greater. I’ve estimated the1971population, on that basis, at somewhere between 300,000 and 340,000.
This has many implications:
* certainly the Indigenous population hasn’t risen anywhere near as much as we all would have liked; in fact, it has hovered around 2.5 % of the Australian total ever since 1971;
* the Indigenous birth rate has been much lower than assumed, and in fact may have been lower than the Australian average since the 1996 Census; it may not be increasing at all, and may even be very slowly declining;
* yes, Indigenous children form around 4 % of all Australian children, but after the age of about 20-25, the proportion consistently falls dramatically. Part of this may be due to increasing mortality, but given that each year, the number of migrants entering Australia amount to twenty years’ of Indigenous births, pulling back the proportion of age-groups older than 20-25 who are Indigenous.
* Re-identification has massively boosted the Indigenous count since 1971, but it is possible that a process of DE-identification may kick in as the urban, working population inter-marries and happily disperses across urban landscapes. If there appears to be a lower growth in the 2016 Census count, this may be one reason.
Just as an aside, perhaps slightly related, currently more than forty thousand (40,000) Indigenous people are university graduates. The equivalent of about 22 % of the 25-yr-old age-group graduates each year, and that is rising by about 8 % p.a. Overwhelmingly , graduates are urban, mainstream and women. Around 40-45 % go on to post-graduate study.
I’ve been following and collating data for Indigenous university participation since about 1990, and many years ago, put forward the suggestion that there could be fifty thousand graduates by 2020; this now looks like likely, a year or even two before that date. Perhaps a new target of 100,000 by 2030-2032 should be put up. The big question is: at what point will graduates form a critical mass and shat will ‘it’ do ?
Thanks for the additional info. It’s certainly the case that there would have been a much lower rate of identification back in 1971, no-one is disputing that. And your comment about migration makes sense. More migrants will tend to reduce the proportion of indigenous population (and indeed Australian-born and other birthplaces not featuring in the current migrant intake) by simple mathematics. University degrees are increasing across the whole population, not just indigenous, and I expect another big increase in the 2016 Census (see my blog about Crystal Ball Gazing for the specific prediction).
In the end the most important thing about the referendum is that it recognized the Aboriginal population as people, as Australians and as part of the population, it’s hard to believe that only 50 years ago this was not the case, but also heartening the huge level of support that this referendum received.
That last paragraph is rubbish, with respect. Of coursethey were recognised a people, whatever that’s supposed to mean. Indigenous people – my wife, for example – already had the vote. They had been formally British subjects since the beginning of settlement/invasion. Like the rest of us, they became citizens with the Citizenship Act of 1949.
My point about re-identification and the low numbers in earlier Censuses is that the people were there – they didn’t suddenly appear fully-formed out of nowhere to be suddenly counted in later Censuses. They already existed. Therefore, nobody in their right mind can use the early Census figures as if they are comparable to later ones. Just try it: work back from the 2011 Census, age-cohort by age-cohort, building in rough estimates for mortality: you will find that the earlier Census figures were grossly inadequate, but people were THERE, counted or not ! That knocks the daylights out of any suggestion that population growth has been very high. No it hasn’t. Neither has the birth-rate been remarkable: even if you compare the 1996 figures with the 2011 figures, you can see that the size of the 0-4 age-group has not risen all that much, and on figures adjusted from the 2011 Census, the growth since 1996 has been barely 1 % p.a.
In fact, I’m amazed how slowly the Indigenous population has grown since 1971, if you adjust the figures back from the 2011 Census. Perhaps barely 1.5 p.a., or less. Try it. Earn your money 🙂
Of course they were people, but the fact remains they were not counted as part of the population. And yes, most already were able to vote, and indeed voted in this referendum! I understand your point about the growth. I don’t think anyone is suggesting that the number counted in the Census in 1971 was a “true” figure. The general consensus is that the Census number is still an undercount. I wonder how long the level of identification can continue to grow each Census though.
On a somewhat different tack, your piece raises for me difficult questions about aboriginality and about how many indigenous persons there really are in Australia. The Census is not helpful on these. While it reports data separately for all respondents in Australia and, since 1967 (with no obvious justification for a separate accounting other than the repeal of s127 of the Constitution), it reports separately data on people of Aboriginal and Torres Strait ‘ origin’ whatever that might mean [‘origin’, not ‘descent’]. At the 2011 Census somewhat over half a million respondents chose to identify as having ATI ‘origin’, perhaps meaning they have some blood connection to aboriginality. I doubt that the numbers with any significant blood or cultural connection with aboriginality are anywhere like this ‘origin’ number: less than a quarter of people of ATI origin claimed aboriginal ancestry! Over 80% spoke only English at home; barely 10% claimed to speak an aboriginal language.
For me this raises also questions about ancestry as reported more generally from the Census. It seems that ‘ancestry’ must mean different things to different respondents, eg ethnic origin (Maori – which could refer to New Zealand Maori or Pasifika Maori), cultural origin (Jewish – which might or might not refer to a religion also), place origin (English), national origin (New Zealand), which means that the results reported on ancestry are pretty well useless. What should my wife report when she is a Jewish Australian but not an Australian Jew and whose ancestors came from what are now Western Europe, Lithuania and Poland via England? Or, what should I report as a New Zealand born pakeha descended from probably anglo-celtic ancestors who lived in various countries including some Australian colonies? Perhaps our children should do as most of aboriginal origin clearly do – claim Australian ancestry?
I guess that my core concern in all of this is that there are people who use Census data on aboriginality such as it is for political ends, notably for a separate treatment (including in our Constitution) of everyone descended from ATI people. That is alien to the idea that I am, we all are, Australian. To make my own political statement I do not see a case for a separate treatment in this country of anyone here on the basis of race. Rather, the case is for targetting gross social and economic injustice where-ever and to whosever it occurs.
There is an obvious interest in those people who identified as indigenous in the 2016 census and identified as non-indigenous in the 2011 census. Even if the percentage where the cross-reference is certain (exactly same name, exactly same age, exactly same address etc.), is small it should be sufficient for a study. The information must be in the data, but not the data publicly available. Do you think there would be a (political correctness) objection to the release of such data? How much would it cost to order such a custom dataset?