The image of the typical family – mum, dad, and two kids – still permeates how we define and understand the family in contemporary Australia. This ideal saturates our screens and newsfeeds and was at the centre of the marriage equality debate, underscoring the pervasiveness of the nuclear family as the dominant family form in our consciousness.
However, this conceptualisation masks the true nature of Australian families, which has changed significantly in recent decades. As sociologists and demographers have long known, the Australian family is as diverse and different as the country’s terrain.
Drawing on data from the 2016 Census, we know there are more than 6 million families in Australia. This is a significant increase from the 5 million or so families counted at the 2011 Census.
Of these 6 million families, the most-common family form (as illustrated in Figure 1) was the couple family with no children (37.76%). The next-most-common was couple families with dependent children under the age of 15 (30.64%).
These proportions confirm that the nuclear family is no longer the most common family form in Australia. One-parent families with dependent children comprise around 8% of all Australian families.
Reflecting this move away from the traditional, nuclear family and the rise of more couple families without children, is the size of families. In 2016, around 30% of all families were two-person families. A further 27% were four-person families.
Most couple families with children in Australia are so-called “intact families” (89.94%), consisting of at least one one child who is the natural or adopted child of both partners in the couple.
However, families are becoming increasingly more “blended”, as couples dissolve (due to separation, divorce or death of a partner) and new families are formed.
Blended families are a small proportion of modern Australian family forms, accounting for just over 3.7% of all families. This includes families with two or more children, at least one of whom is the natural or adopted child of both partners and at least one other child who is the step-child of one of them.
A further 6.3% of families are step-families. Here, there is at least one resident step-child, but no child who is the natural or adopted child of both partners.
Grandparent-led families are also increasingly significant.
Grandparents already play a significant role in Australian family lives through the provision of informal child care, but there are now just over 60,000 grandparent families in Australia (which a significant increase from estimates in 2004, which found around 22,500 grandparent families). Of those, 53% of grandparent families are couple families with grandchildren and 47% are lone grandparent families.
The 2016 Census gathered information on same-sex couples. Compared with opposite-sex couples, these data show that family forms differ across sexual orientation.
Overall, around 15% of same-sex couples had children. Female same-sex couples were more likely to be in couple families with dependent children (20.67%) compared to male same-sex couples (3.10%), or opposite-sex couples (37.8%).
However, same-sex couples were still more likely to be in couple families with no children than were opposite-sex couples, and they were more likely to have smaller families. Of those, around 54% of male same-sex couples with children and 51% of female same-sex couples with children had one-child families. One-third of same-sex couples had two children.
In comparison, 36% of opposite sex-couples had one child, and 42% had two children.
What these data from the 2016 Census show is just some of the diversity within the Australian family. While the idealised nuclear family of the past is no more, this does not mean that the family as a social institution is in decline, or that families in contemporary Australia are at risk.
But it does mean families are changing. Our political leaders should reflect on this diversity to ensure social policies reflect these differences, so that all families are well supported.
But if we look past these headlines, the Census gives us a unique insight into the economic outcomes of the Aboriginal and Torres Strait Islander population, single parents, those entering the labour market and those coming to the end of their working life.
What we see is geographic divergence in Indigenous employment, declines in employment for single parents and the young, and the news is mixed for those entering retirement age.
Indigenous employment outcomes
One of the targets of the Closing the Gap strategy (which is slated to undergo a “sweeping overhaul”) is to “halve the gap in employment outcomes between Indigenous and non-Indigenous Australians within a decade (by 2018).”
But the latest census data shows that there has been no demonstrable improvement in employment outcomes over that decade, or even in the five years between the 2011 and 2016 Censuses.
Specifically, 44.1% of the adult Indigenous population was employed in 2016, compared to 44.2% in 2011. Over this period, employment rates for Indigenous women increased by 1.3%, but fell for Indigenous men by a similar amount.
For the total Australian population, there has been a decline in the employment-to-population ratio over the same period (61.4% to 60.2%). On this measure, there has been a small convergence between the Indigenous community and the total Australian populations. However, this will not lead to a closing or even a halving of the gap any time soon.
This national-level stagnation also hides considerable geographic variability, as shown in the map below.
In remote Australia, employment has fallen dramatically in most regions, with employment-to-population ratios plummeting by more than 15% in some regions. Much of this decline is likely attributable to the phasing out of the Community Development Employment Projects scheme and its replacement with a work-for-the-dole program whose participants are classified as unemployed.
Clearly, current employment policies are failing to create sufficient jobs for Indigenous people in remote Australia.
In more populous areas, Indigenous employment has grown steadily, especially in NSW where employment growth has been particularly rapid and has been accompanied by substantial population increase.
There are many good arguments for rethinking the Closing the Gap targets. However, they have had the benefit of highlighting policy failure and success. Hopefully a renewed focus on strengths and positive achievements does not diminish this accountability.
Single parent families
According to the Census, there was 959,543 single parent families on the night of the Census. Of these, around 54.2% were employed, 5.9% were unemployed and the remainder not in the labour force. By comparison, in 90.6% of couple families with children, at least one parent was employed in 2016.
Through time, there are slightly fewer single parent families where the parent is employed as a percentage of all single parent families (54.7% in 2011 compared to 54.2% in 2016).
That is, despite considerable focus on childcare, significant changes to welfare policy, and a number of active labour market policies and changes to employment support, a child in a single parent household is less likely to have that parent employed than five years ago.
From a policy perspective, we should be careful about forcing single parents to work when they have caring responsibilities that cannot be substituted. However, we have to do better in providing the support, training and employment environment where those single parents who do want to work are able to.
This includes ensuring childcare is sufficiently flexible, affordable and of high quality, trialling targeted interventions that boost skills and employability, and lessening the conditionality of parenting payments that reduce the incentives for recipients to find work.
New labour market entrants and those of retirement age
The current labour market is clearly working for some, but not others. Two groups for whom employment is particularly important are the relatively young and those at the end of their working lives.
For the young, early exposure to long periods of unemployment or underemployment can be very costly throughout their lives. Early periods of unemployment predict employment outcomes across their life.
Focusing on the young, to start with, we can look at the 25-29 year old population. For the majority of this group, school and post-school education is mostly finished, and child care and other responsibilities are yet to have kicked in. Over the last five years, however, employment for this group has gone backwards.
Among the 25-29 year old group, the unemployment rate has increased from 5.9% to 7.1%, whereas the employment-to-population ratio has decreased from 78.4% to 77.1%.
The young adult labour market is not what it was pre-global financial crisis (the unemployment rate for 25-29 year olds in the 2006 Census was 5.6%), or at the peak of the mining boom. Clearly the macro-economy matters, but we also need to trial and adjust new types of active labour market support for young adults.
As a society, we are going to be increasingly reliant on those at the end of their working lives to support and mentor people entering the labour market, to maintain the tax base, and to support themselves as they prepare for retirement.
Here, the news is a bit more positive, though not completely so. It is true that the employment-to-population ratio has increased for 60-64 year olds (from 50.3% in 2011 to 53.8% in 2016). However, the unemployment rate for this group has also gone up (from 4.1% to 5.8%).
More people are working, but more people are actively looking for work. On balance, this is probably positive, but we also need to make sure that we are minimising the incidence and effects of age-related discrimination.
Delving into the employment statistics in the 2016 Census tells a more nuanced story than is evident from the monthly Labour Force Survey. In particular, remote Indigenous employment and outcomes for single parents show deeply concerning trends. Policy measures may need urgent consideration to increase employment opportunities and ameliorate poverty among these population groups.
**Correction and Editor’s note: This article was updated on 25 October to correct the first chart on Indigenous employment outcomes. The chart had previously misstated the figures on those employed, unemployed and the total labour force. The Conversation apologises for these editing errors and thanks reader John Blake for alerting us to them.
Population growth has profound impacts on Australian life, and sorting myths from facts can be difficult. This article is part of our series, Is Australia Full?, which aims to help inform a wide-ranging and often emotive debate.
The release of 2016 Census data provides a good opportunity to reflect on the future growth of Australian cities. And what better example of the future to use than Tokyo?
This 30-year growth spurt happened at a rate almost twice that predicted for Greater Melbourne, for example, from 4.4 million today to 8 million by 2050.
Tokyo’s rapid growth had a number of negative impacts. These included very significant environmental pollution. The basic approach during this period was to grow first and clean up later.
The consequence for Tokyo was disorganised patterns of urban development – sprawl. The answer involved tighter planning controls and land re-adjustment programs to improve environmental conditions and ensure infrastructure was effective.
The lesson here for Australian cities is that, in the face of rapid population growth, better forward planning is the only way to avoid or minimise negative side effects.
Lesson 2: Introduce metropolitan governance
A critical factor in Tokyo’s liveability is the role of metropolitan governance in ensuring good planning and co-ordination.
The Tokyo government also has considerable political autonomy since it generates 70% of its revenue from local taxation. In 2014, it had a budget of ¥13 trillion (A$151 billion) – on a par with Sweden’s. This makes the governor of Tokyo one of the most powerful politicians in Japan, second only to the prime minister.
Lesson 3: Commit early to world-class public transport
Public-private partnerships to develop metropolitan railways has been a standard approach in Japanese cities for most of the 20th century and continues to underpin Tokyo’s success as a global city. For example, the Mitsubishi Corporation played an instrumental role in developing the Marunouchi district around Tokyo Station. The latter was built in 1914 and connected intercity stations in a loop decades before other cities.
These public-private interventions have cemented Tokyo’s status as a transit-oriented metropolis. The city has by far the highest public transport usage in the world.
Compared to other major cities like Seoul, London, New York and Beijing, Tokyoites rely far more heavily on public transport, cycling or walking to get around. In Tokyo prefecture, rail accounts for 48% of trips, bus 3%, cycling 14% and walking 23%. Private car use accounts for only 12% of trips.
A continuous investment in rail networks above and below ground would ensure Australian cities can better accommodate predicted population growth. A fascinating map designed by Adam Mattinson shows what a subway system for Melbourne could look like based on the Tokyo model. To achieve this may require that the tram system moves underground – almost certainly a pipe dream.
Lesson 4: Decarbonise the economy as it grows
Tokyo was lucky to be able to grow rapidly in an era when climate change was not the recognised problem that it is today.
The challenge for Australian cities will be to grow their economies while reducing their greenhouse gas emissions to match the per capita levels for Tokyo, and then to cut them much further. The World Bank calculated that in 2006, per-capita emissions for Sydney were 20.3 tonnes of CO2, compared to 4.89 tonnes for Tokyo.
Tokyo is also seeking to cut its emissions by 30% by 2030 compared to 2000. In Australia, Plan Melbourne, for example, aims to achieve a target of net zero emissions by 2050 even while the population continues to grow.
While investments in low-carbon public transport will be central to meeting this target, it is also essential to pursue ambitious energy efficiency and renewable energy targets.
Tokyo is aiming for a 38% drop in energy consumption and a rise in renewable energy from 8.7% in 2014 to 30% of electricity generation in 2030. The good news is that Plan Melbourne sets a target for renewables of 40% of electricity generation by 2025.
Lesson 5: Prepare to age with dignity
Along with declining emissions intensity, Tokyo’s population is likely to start shrinking. The population of central Tokyo is expected to rise from 13.5 million today and peak in 2020 before declining to 7.1 million by 2100. The population of Greater Tokyo is expected to peak around 38.5 million about the same time.
As recently argued based on the census, a result of declining home ownership is the likelihood of couples deferring the decision to have children. A knock-on effect could therefore be a more rapidly ageing Australian population.
The Tokyo of today is certainly no utopia, due to its vulnerability to earthquakes and other natural disasters, high house prices, homelessness, rising inequality, a lack of multiculturalism and a proportion of housing as rental accommodation that dwarfs Australia’s (47.9% compared to 30.9%).
Yet the largest settlement on the planet offers useful lessons – historical, present and future – that can guide the urban policies of other countries.
You can read other articles in the Is Australia Full? series here.
The 2016 Census has revealed an increase in the number of children with disability, up nearly 40,000 since 2011. One explanation is that the census now counts disability differently, which is more in line with the way many children and families view disability.
But other children continue to miss out on support because they do not name their needs as “disability”. And services don’t yet have adequate funding for even the revealed number of children, so other children who require assistance are left out.
A census that counts people who identify as having a disability, as well as those who need support, could help resolve these problems.
Children and young people who need support related to disability has risen from 2% to 2.6% of children – or 38,309 more children than in 2011. The most striking change is boys with disability aged 5-14 years, who have increased to 4.4% of all boys their age. These rates are even higher for Aboriginal and Torres Strait Islander children – 7.4% of boys aged 5-14 years and 4.4% of all children and young people aged 0-19 years.
The census counts disability as “has need for assistance”, which it defines as “profound or severe core activity limitation”. The definition was introduced in the 2006 Census to be consistent with international measures and other national surveys, which focus on counting support needs. Before 2006, disability was not counted at all. The continued increase each census since 2006 is probably due to more Australians identifying with the definition or seeing the benefit of identifying as disabled, now that policies to support disability are changing.
Knowing who the definition covers is important. The census count of “need for assistance” is good to inform government planning about high levels of support some people need to participate equally in our communities. Estimating the number of people likely to need a National Disability Insurance Scheme (NDIS) package is a current priority. This census counted 562,629 people aged under 65 years – over 100,000 more than the NDIS planning estimates.
Equally important for children is planning access and support in school, playgrounds and other places where children participate in their families and communities. The higher 2016 Census count shows these plans need to expand.
Who isn’t counted?
The census question only counts people with high needs, not all people with disability. Unfortunately, the question is not complemented with an identity question about whether you have a disability. This means people with disability who do not need assistance – for example, some people who are blind – are not counted. The World Health Organisation estimates the larger total would be closer to 15% of all Australians, rather than the 5.1% measured in this census.
Disability advocates consistently express concern that by not asking Australians directly about their disability or impairment, the census fails to count the population of people with disability accurately – it only captures people who need assistance.
Not gathering information about this 10% of our population is a missed opportunity. It means we simply don’t know how many people with disability may benefit from, and contribute to, more accessible communities and new social and economic opportunities. For children, this is critical to having an inclusive community as a foundation.
Children and young people with disability are often positioned as passive recipients needing assistance through family, friends and services. Research with children and young people themselves, however, shows they want to be recognised for their active contribution to their families and wider networks. Their positive identity is more important to them than their support needs.
One of the interesting changes since the introduction of the NDIS is that families and service providers are now also using the “need for assistance” definition of disability, which is consistent with the inclusive vision from the UN Convention. Their advocacy with this definition means support for young children in Australia has expanded already even though the NDIS is still growing.
Children receiving disability support are now more likely to use it while they are with other children in their community, rather than in separate services. Families’ capacity to demand these inclusive services recognises the rights of their children to get the support they need to enjoy their childhood and have the same options as their peers in the future. These trends are also consistent with the insurance approach of the NDIS: that assistance now is an investment for later.
Funding and support
The increase in the numbers of children and young people with disability may reflect families’ optimism about having their children’s needs met in the new NDIS world. It certainly promises to replace long waiting lists and capped places of previous systems. The census numbers reinforce the higher number of children in the NDIS than expected, which is upsetting NDIS estimates. The NDIS has detailed data about people using the scheme. This will not resolve the question about the total number of people with disabilities though. People receiving NDIS packages are those likely to already be those identified in census data as needing support.
Data collection in schools has also recently improved with the introduction of the Nationally Consistent Collection of Data for school students with disability. Most children and young people participate in the school system, so these data will inform understanding about adjustments to support students in their education.
Bringing these large data sets together means we can understand the types of supports families need, and where there are service gaps between schools and the NDIS.
Lessons from data need to be discussed alongside the expectations and experiences of children, young people and families to ensure they’re getting the support they need. This will help children enjoy the opportunities of childhood, rather than the current disproportionate but necessary focus on dismantling barriers to belonging.
The great Australia dream of owning your own home is still alive despite the various problems plaguing housing affordability, new Census data shows. Even though the overall home ownership trend remains strong, it’s masking other issues.
The latest 2016 Census data assesses what the national home ownership and rental rates are and how these vary location. It also gives us a picture of mortgage and rental costs.
Comparing home ownership rates since the 2011 Census, there’s a slow but steady decline in home ownership rates overall – down by 2.9% from 64.9% of all Australian households in 2006, to 62.0% in 2016. However, 7.4% of households did not state their housing tenure in the 2016 Census. This accounts for some of the variation in reported rates of home ownership decline.
This contraction is nowhere of the scale of equivalent falls in home ownership in the US and UK and New Zealand over the same period.
What’s more interesting than the overall trend, is the greater decline in outright home ownership, involving no mortgage debt, from 31.0% to 29.6% between 2011 and 2016. There’s also a lesser decline in home owners who are purchasing with mortgage debt 33.3% in 2011 compared with 32.4% in 2016.
The opportunity households now have to borrow against their mortgage loans for spending undoubtedly accounts for some of this change. Also contributing to this is home purchasers are less likely to reach retirement age with no remaining mortgage debt, in the same numbers as previous eras.
Another aspect of housing affordability is masked by these numbers – the wide variation in being able to purchase a home according to age and income. Recent evidence indicates would-be-home-owners try various means including very high mortgage debt and moving to outer urban locations away from employment and into smaller dwellings, to be able to buy a house. Some even delay having kids.
Census figures show that for people wanting to purchase a home, a change in state or city location may be an option. According to the data Darwin was the most expensive city to buy in, whereas Hobart was the cheapest for home purchasers.
For households across the income spectrum, 7.2% of purchasers are paying more than 30% of their income on mortgage costs, the data shows. This figure is likely to be far higher among the lowest income (40% of households) for whom such costs place them in housing poverty.
Given the national obsession with investment in private rental, it’s no surprise that the proportion of all Australian households now renting has also increased. Census 2016 results show the private rental sector grew in size, from 20.2% in 2006, to 22.0% in 2011 and to 23.6% in 2016.
In 2016 a total of 2,089,633 Australian households rented privately, either from real estate agents or private landlords.
The growth of the private rental sector largely reflects the high costs of home purchase. Many households who rent have a relative lack of security and control over rental increases.
For those unable to pay rent in the private market, social housing is likely to provide little relief. Census data shows overall rates of social housing declining from 4.7% in 2006 to 4.0% in 2016. In this context, the growth in rates of homelessness in the last decade is perhaps not surprising.
For Indigenous Australians, the housing picture is different. Census 2016 data show among households in which at least one resident is Aboriginal and/or Torres Strait Islander, 12.2% are outright owners, 25.9% are purchaser owners, 32.4% are renting privately. Around a fifth of households, 21.5%, live in social housing, reflecting targeted social housing programs in metropolitan, rural and regional areas.
Overall, home ownership has not changed as dramatically in the last decade, as some would have anticipated. However, it’s likely with the labour market being what it is and the adaptations people are making to try and buy a home, there may be longer-term problems to be seen in future.
Excessive household debt, polarisation of cities into low and high income earning areas and deepening family housing constraints indicate these Census figures likely mask bigger problems. This may translate over time into a more costly social problem, as increasing proportions of households require housing assistance of some form. Australian society could become even more divided on the basis of housing wealth and opportunity, if these trends continue, as we expect they will.
Aboriginal and Torres Strait Islander people have a complicated history with the Australian census. Until the 1967 referendum, the question about Indigenous status was used mainly to exclude the Indigenous population from official population statistics, as required by the Constitution at that time.
Since the 1971 Census, however, the question has been used to understand Indigenous demographic and socioeconomic outcomes. This includes observing how Indigenous peoples’ situations are changing through time, and comparing them to the non-Indigenous population.
Indigenous population estimates matter for policymaking. For example, the Commonwealth Grants Commission uses estimates of the Indigenous population to advise on GST revenue allocation to the states and territories. And many Closing the Gap targets are monitored in full or in part using census data.
At the same time, many Indigenous communities and organisations argue that the way data are collected and distributed takes power out of their hands, and puts them at a disadvantage.
So, what did we learn from the most recent release of 2016 Census data?
A rapidly growing population
First, the Indigenous population is growing – rapidly. In data released on Tuesday, it was estimated there are 786,689 Australians who identify as Aboriginal or Torres Strait Islander. This is a 17.4% increase from 2011.
This estimate of the total Indigenous population is based on the 2016 Census, but also takes into account people who were missed in the count. Indigenous Australians are now estimated to represent 3.3% of the total Australian population, up from 3% in 2011.
Setting aside for the moment those who were missed, 649,171 people identified themselves as being “of Aboriginal or Torres Strait Islander origin” in the 2016 Census. This is up 18.4% from 2011.
There are several reasons for the rapidly growing visibility of Indigenous Australians in our population statistics.
The first is the natural increase of the Indigenous population. On average, Indigenous Australians have a slightly higher number of children than non-Indigenous Australians.
In addition, the children resulting from relationships between Indigenous and non-Indigenous Australians often identify as Indigenous, accelerating the growth of the Indigenous population. However, even after taking these factors into account, the Indigenous population was projected to reach no more than 746,815 by 2016 – around 40,000 people fewer than the latest census-based estimate.
While we won’t know how accurate the assumptions underlying these projections were for some time, the probable explanation for the “extra” 40,000 Indigenous people is that some individuals who chose not to identify as Indigenous in 2011 decided to identify as Indigenous in 2016.
The change in propensity to identify as Indigenous makes comparisons over time difficult. Any change in socioeconomic outcomes is likely to partly result from changing life-chances for Indigenous Australians, but is also likely to be partly attributable to the change in the group of people who are classified part of the Indigenous population.
However, because the group of newly identifying Indigenous people is around 5% of the Indigenous population, any change in socioeconomic indicator that is greater than 5% can at least be partly attributed to a change in the circumstances of the families who identified as Indigenous in 2011.
Regardless of what factors drove Indigenous population growth in the 2016 Census, it is clear that population growth was not evenly distributed.
As the map below of the change in census counts between 2011 and 2016 shows, the Indigenous population increased by the greatest amount in Brisbane, on the New South Wales central and north coast, and Sydney – Wollongong. Almost half (49%) of the growth of the Indigenous population occurred in just these three regions.
Conversely, the Indigenous population of the East Kimberley, Alice Springs, and several other remote regions appeared to decline slightly.
These initial figures should be interpreted cautiously. The proportion of records for whom we do not have an answer to the Indigenous status question on the census increased by about one-third between censuses. Nevertheless, it is clear that Indigenous population growth remains concentrated in NSW and Queensland.
The changing distribution of the Indigenous population can be better understood through a cartogram, as in the figure below. Each circle represents a location defined by the Australian Bureau of Statistics as an “Indigenous Area”. The circles representing these locations have been sized so that their areas are proportionate to their Indigenous population count in 2011.
They have been moved on the map so they do not overlap, but under the constraint that they move as little as possible, that they remain within their state boundaries (except for the ACT), and that they retain their relative position next to each other as much as possible.
We have also added a line around the greater capital city areas to help readers identify them. Each area has been coloured in a manner that indicates the change in population between 2011 and 2016.
What the cartogram shows very clearly is that in 2011, the Indigenous population was already highly concentrated in coastal NSW and Queensland. The cartogram suggests the Indigenous population has grown substantially everywhere in the greater Brisbane/Gold Coast region, in much of the NSW central and north coast, and around the fringes of Greater Sydney and Greater Melbourne.
The places where the Indigenous population declined tend to be locations that had a low population in 2011 and that are located in remote parts of Australia. However, there are plenty of exceptions to both of these general patterns.
In 2016, the number of people speaking an Australian Indigenous language at home increased by a small amount – to 63,754 persons from 60,550 in 2011.
This was an absolute increase. But it represents a decline in the proportion speaking an Australian Indigenous language – from 11.8% in 2011 to 10.5% in 2016.
This proportionate decline may result from Indigenous population growth being concentrated in areas where Indigenous languages are less commonly spoken. So, it would be hasty to jump to the conclusion that these results suggest language loss.
Indigenous people are getting into the education system earlier and staying for longer. This is likely to lead to improved socioeconomic outcomes in future.
Early childhood education rates have increased. The percentage of three-to-five-year-olds who aren’t already at primary school that were attending preschool is up from 43.5% in 2011 to 48.5% in 2016.
At the secondary school level, 59.7% of Indigenous people aged 15-18 were attending school. This is up substantially from 51.2% in 2011.
Increased school attendance has flowed through to growth in the percentage of Indigenous people aged 15 or more who have completed year 12. This rose to 34.6% in 2016 from 28% in 2011.
Finally, the proportion of Indigenous 15-24-year-olds undertaking tertiary education appears also to have grown. In 2016, 16.2% of this cohort who were not at secondary school were studying for a tertiary qualification. This is up from 14.1% in 2011.
Significantly, this growth has been driven by increased university attendance (8.5% in 2016, up from 5.8% in 2011). Attendance at technical or further educational institutions has actually been falling (7.7% in 2016, down from 8.4% in 2011). This shift from technical education to university education is suggestive of the continued growth of an Indigenous middle class.
Indigenous people are also earning more on average in 2016 than 2011.
Real median weekly personal pre-tax income is up from A$397 in 2011 to A$441 in 2016. Although incomes grew generally between censuses, the growth in incomes for Indigenous people outpaced that of the rest of the population. Nevertheless, on average, Indigenous people still receive a personal income that is only two-thirds that of the non-Indigenous population.
Similar trends were evident for household income: Indigenous household incomes are increasing more rapidly than non-Indigenous household incomes.
Similarly, home ownership rates among households with at least one Indigenous resident have increased slightly, from 37.4% to 39.6%.
Households with a resident who identifies as Indigenous are also slightly less likely to be overcrowded than in the past. The proportion of households in a house that doesn’t meet the international standard for overcrowding from 11.8% to 10.4%.
At the same time, overcrowding by the same measure among non-Indigenous households has risen between 2011 and 2016, from 3.2% to 3.6%.
The 2016 Census has mostly delivered a good news story in terms of Indigenous outcomes. But the extent to which these positive results indicate that life is getting better for Indigenous families remains unclear.
At least some of these apparent improvements are likely to be the result of people of relatively high socioeconomic status identifying as Indigenous for the first time.
In addition, there were a number of census records for which we know nothing about whether the person is Indigenous – around 6% of the total count. This is because these people didn’t fill out any census form, or they didn’t answer the Indigenous status question. Either way, we need to keep this uncertainty in mind when interpreting findings.
And while we have generally presented national figures here, socioeconomic outcomes for Indigenous people vary dramatically across Australia. Overall improvements, while welcome, may well mask less positive outcomes in specific regions.
Every five years the census asks Australians: “What is your religion?”. Ten tick-box responses are provided, along with the option to write in some other response.
In 2016 Census, the first box was for “no religion”. This was not a secularist plot, but an acknowledgement that those declaring they had “no religion” were very likely to be the most numerous category, followed by Catholics.
Alongside those declaring they have “no religion”, Australia now has – in addition to a highly diverse bloc of Christian groups that are very internally diverse – five substantial religious communities (Buddhists, Muslims, Hindus, Sikhs, and Jews).
How religion in changing in Australia
The 2016 Census showed major changes in the ranking order of religious identification groups in Australia. “No religion” leads Catholic, then Anglican, Uniting, Muslim, Buddhist, Presbyterian and Reformed, Eastern Orthodox, Hindu, Baptist, Pentecostal, Lutheran, and then Sikh.
It is important to note that to keep a stable percentage of the population from 2011 to 2016, a religious group had to grow by 8.8% – the national population growth rate.
Australia now has more Muslims and more Buddhists than Presbyterians; more Hindus than Baptists or Lutherans; and nearly as many Sikhs as Lutherans.
Among those groups attracting double-digit percentages of the population, there has been a continued rise of those declaring “no religion” to become the most numerous group (30.1%). Catholics are at 22.6%, while Anglicans – who had been nudged out of the top spot in 1986 – have now slipped from a high of 41% in 1921 to third place at 13.3%.
Among those groups attracting between 2% and 5% of the population, the Uniting church has declined to 3.7%, while Presbyterians (2.3%) are now behind Buddhists (2.4%). But both have been overtaken by Muslims (2.6%). Thanks to migration, Hindus continued their rapid rise to achieve 1.9%.
Among those groups clustered around 1%, Pentecostals remain unchanged (1.1%). Lutherans (0.7%) are down from 1.2% in 2011.
The two fastest-growing religious groups since 2011 are Sikhs, who grew by 74.1%, and the “Other Protestant” category, which grew by 79.8%. Both are now 0.5% of the population, and are more numerous than Jews (0.4%).
The Other Protestant category includes many who just wrote in “Protestant”, and those identifying with unaffiliated congregations. This category’s growth, along with that labelled “Christian not further defined” (to 2.6%), is an indication of the decline of the importance of denominational labels to Australians who prefer to just indicate they are Christian. This is also an indication of religious groups’ increasing diversity in Australia.
The Christian proportion of the population has fallen to just over 50%, down from 88% 50 years ago. More significantly, the British Protestant percentage has declined to about 20%, making it smaller than Catholics. This marks a major shift in Australian culture, which – until about 1990 – was resolutely British Protestant, with Anglicans, Presbyterians and Methodists dominant.
Not only has the proportion of Christians declined, it has become much less British Protestant. Australia’s religious life has changed beyond recognition from the 1950s and 1960s, when British Protestants comprised two-thirds of the population.
Many who were raised in those decades, including political and business leaders, still hold that form of Australia to be normal and expected. This expectation is not shared by those aged under 50.
Formerly dominant organised forms of religious life are attracting fewer participants and fewer who identify with them – but they are far from dead. While no longer dominant, they form part of a diverse array of identities and commitments that shape Australians’ lives.
When the full results become available, we will be able to see how many Australians identify with the myriad small groups – Zoroastrians, Satanists, Scientologists, witchcraft/wicca, and more.
What we can draw from these results
Neither those who would declare that Australia is a Christian country nor those who see the rise of those declaring “no religion” as the death knell of religion can take heart from this census.
Rather, the results show the diversity of Australia’s religious life. That only 9.6% refused to reply to this question – the only optional question on the census – tells us that religious identity is still of interest to Australians.
Also, declaring “no religion” does not mean that someone is anti-religious, lacking is spirituality, or an atheist. It means they just do not identify with a particular organised form of religion.
The response to the religion question provides an indication of a person’s cultural orientation and formation. Religion, culture and formation used to be overlapping and reinforcing categories. For example, Catholics were Irish, went to Catholic schools, and shared certain orientations. Presbyterians were British (Scots or Northern Irish), went to state or private schools, and shared certain orientations.
While this overlapping is no longer true, religious identity is far from meaningless. The census provides a moving series of snapshots of religious identity. But it does not tell us about religious belief, practice, or anything else about a person’s religious life.
The changing pattern of the diversity of religious identities is one indicator of a society’s degree of multiculturalism. On this measure, Australia is among the most diverse.
The types of romantic relationships Australians have, as well as the way they are recognised and measured, have changed dramatically in the last 30 years.
Much of the change in partnering has been in response to changing legal and social norms. Childbearing has been decoupled from intimate relationships by the widespread availability and use of contraception and the availability of abortion. Divorce is easier to access; women play a much greater role outside the home.
These and other forces have led to delays in marriage, increasing co-habitation (couples living together), and a larger proportion of the population who re-partner or have more than one relationship throughout their adult life.
Results from the 2016 Census, released today, allow us to track marriage and co-habitation trends for both heterosexual and same-sex couples.
In 2001 and 2016, around 40% of Australians were classified as single. By age, this pattern declines until the mid-30s, and then increases in older ages due to divorce and widowhood.
The pattern is more obvious for women – particularly in the older ages, as they are more likely to experience the death of their partner.
There has been a slight increase in co-habitation overall to 10% of Australians, and a corresponding decrease in marriage to just under 50%.
What has changed the most in these relationship patterns is that co-habitation was predominantly confined in 2001 to people in their 20s and 30s. In 2016, cohabitation is also a significant feature for people up to their mid-60s.
Same-sex couples have been identified in the census since 1996. Over each successive census, the number of couples identifying as same-sex has increased. In 2016, 46,800 couples were same-sex – an increase of 39% from 2011.
The 2011 Census showed people in same-sex couples are, on average, younger, more educated, employed in higher-status occupations, and have higher incomes.
The 2011 Census allowed same-sex couples to identify their relationship as a marriage for the first time. As would be expected, the numbers are small (1,338) – but they will rise over time, as more people travel overseas to marry legally and in the event Australia legislates for marriage equality.
What all this means
The rise of co-habitation has led to speculation that marriage is out of fashion and could disappear altogether. Our research shows the institution of marriage is not outdated. The nature of marriage is evolving, as people manage the changing role of intimate relationships in their lives.
It is also true that the marriage equality debate will lead to a re-imagining of marriage for both homosexual and heterosexual couples. Most Australians still marry, and there is no evidence that marriage will disappear – despite predictions.
However, while marriage may have lost its practical importance, its symbolic importance still seems to be high. In many ways, getting married is still seen as a marker of achievement.
Perhaps new ways of forming relationships and childbearing are not a threat to marriage: they may be a signal of the fact that more options are now available.
The distinction between same-sex and heterosexual relationships is complicated by the measurement of gender itself.
For the first time, the 2016 Census allowed non-binary gender as a response to the question of sex, although people identifying as other than male or female were required to use the paper form or to request a special online form. This would have significantly affected the overall count of people who identify as neither male nor female.
There were 1,300 validated responses that indicated a sex other than male or female. The Australian Bureau of Statistics has also estimated an additional 2,400 people responded both male and female on the paper form.
Overall, the census shows a decrease in the proportion of Australians who are married, and an increase in co-habitation of both heterosexual and same-sex relationships. We predict this will continue to rise in future censuses.
According to data released today, there were 23,401,892 people who were counted in Australia on the night of the 2016 Census who were usually resident in Australia.
After adjusting for undercount and adding back those who were overseas on census night, the Australian Bureau of Statistics (ABS) estimates that as of December 2016, Australia’s population was around 24.4 million.
Our population is growing – and fast. But can we trust the numbers?
The panel reported that the 2016 Census is “of comparable quality to 2006 and 2011 and comparable collections internationally”. It reported there was a lower net undercount (that is, people who were missed from the census) in 2016 than there was for 2006 and 2011.
This means, for the most part, we should believe the results from the 2016 Census. All data has its limitations. But it would appear that with the information we currently have the data is robust enough, at least at the national level.
But, as always, the devil is in the detail. For some individual questions, there was a high level of non-response – and that needs to be taken into account.
How we’re changing
The data tell us quite a lot about who we are as a nation, and how our characteristics are changing. The 2016 Census reveals that Australia is becoming much more diverse – in language, country of birth, Indigenous status, and religion.
In the 2011 Census, 69.8% of people reported being born in Australia. This declined over the past five years to 66.7%.
The percentage of Australia’s current population who were born in England has also declined, from 4.2% to 3.9%. Simultaneously, there was a dramatic increase in the percentage born in China – 1.5% to 2.2% – and born in India – 1.4% to 1.9%.
This increase has been driven in part by the higher rate of mortality of the Australian- and European-born populations, who are much older than more recent migrant groups. More importantly, though, most of our more recent arrivals are from Asia. Combined, those born in India, China and the Philippines made up 33% of all those who now live in Australia who arrived between 2007 and 2016.
Australia’s Indigenous population is also growing – quite rapidly.
In 2016, there were 649,171 usual residents who were identified as being Aboriginal and Torres Strait Islander. For those who answered the question, this represents an increase from 2.7% to 3% of the total Australian population.
When undercount is adjusted for, the ABS estimates that 786,689 Aboriginal and Torres Strait Islander peoples should have been counted in the census. This is a growth of 18.8%, concentrated mainly in New South Wales and Queensland, and is much faster than the growth of the non-Indigenous population.
There was also a decline in the proportion of people who spoke English as their main language at home (76.8% in 2011 to 72.7% in 2016), an increase in those reporting no religion (21.8% to 29.6%), and a very rapid increase in the number of same-sex couples (a 39% increase to 46,800 couples).
One result that needs to be treated with care is the high and increasing rate of non-responses recorded to some of these questions. There were 1,622,692 people recorded as “not stated” for the country-of-birth question, and 1,411,491 who were recorded as not stated for the Indigenous status question. This was an increase of 35.7% and 33.3% respectively from 2011, which was much faster than the growth of the total population.
That doesn’t mean that all or even most of these records are people refusing to answer the question. Rather, most are records that have been imputed due to missed households or individuals.
However, we do need to make sure we carefully exclude these records from our calculations. And more research is needed to uncover whether and why there are a number of people not answering individual items.
Implications for policy
Leaving these issues aside, there were some interesting findings that touch on ongoing policy debates.
The Gonski 2.0 school funding reforms passed parliament last week. But there was actually a decline in the proportion of infants/primary school students who were attending a non-government school since the last census, from 32.7% to 31.8% between 2011 and 2016.
There was a slight increase in the proportion for secondary school students (42.1% to 42.8% in non-government schools). This means the proportion across all schools was roughly stable (36.5% in 2016 compared to 36.7% in 2011).
There is more to the school funding debate than simply government versus non-government. But the census results show the move to the non-government sector seen over previous periods may have slowed, or even reversed.
Another important current policy debate relates to the National Disability Insurance Scheme (NDIS). While the census isn’t ideal for understanding changes in rates of disability, there is a very important question about whether people have a “core activity need for assistance”.
One particular result stands out – the rapid increase in the number of children and youth reported to have such a need. Between 2011 and 2016, the proportion of those aged 19 years or under reported as having a core activity need for assistance increased from 2.1% to 2.7%, excluding the not-stated population.
These proportions might not seem large. But it is an extra 38,209 individuals, or a 34.5% increase in children and youth with a core activity need for assistance.
Finally, even if the policy responses weren’t large, the most recent federal budget and election both had a heavy focus on housing affordability and home ownership. The census doesn’t have information on house prices, but it does show that the median mortgage payment in Australia in 2016 was A$1,755 per month. Sydney, Darwin and Canberra all have median payments of $2,000 or more.
Over the longer term, the proportion of the Australian population who own their home outright (that is, without a mortgage) has declined from 41.1% in 1991 to 31% in 2016. Much of that decline has been made up for by an increase in the proportion renting (either from government or a private landlord) from 26.9% to 30.9% over the same period.
These are just a snapshot of 2016 Census results. For the most part, we can be confident that, in the words of the Australian Statistician David Kalisch:
The 2016 Census data provides a detailed, accurate and fascinating picture of Australia and our communities.
For individual data items, we need to be more careful and circumspect. But we now have a much better idea on our nation and how it is travelling.
In a country as diverse as Australia, it is impossible to identify a set of characteristics that defines us. However, with today’s release of data from the 2016 Census, it is possible to identify some of the common characteristics, how they vary across states and territories, and how they are changing over time.
Australia undertakes a compulsory long-form census – where detailed information across several areas is required of every individual respondent – every five years.
The 2016 Census has revealed the ‘typical’ Australian is a 38-year-old female who was born in Australia, and is of English ancestry. She is married and lives in a couple family with two children and has completed Year 12. She lives in a house with three bedrooms and two motor vehicles.
Australia is getting a bit older; the typical Australian in 2011 was aged 37.
How do today’s results vary across Australia?
First, age varies by state and territory.
With variables like age, we often find the “typical” value by taking the median. In essence, we (statistically) line everyone up from youngest to oldest, and find the person who is older than half the population but younger than the other half.
In Tasmania, the median age among 2016 Census respondents was 42. But in the Northern Territory, it was 34. Those in Australian Capital Territory were also quite young (median age 35), whereas those in South Australia were relatively old (40).
The NT population’s relatively young age is influenced by the very high proportion that identify as Aboriginal and Torres Strait Islander.
While we don’t have updated estimates for that proportion (either for the NT or nationally), the data released today show that the Aboriginal and Torres Strait Islander population is quite young. The median age nationally is 23. New South Wales and Queensland have the youngest Indigenous population, with a median age of 22.
This release also tells us something about the different migrant profiles across Australia. Nationally, the most common country of birth for migrants is England. And the median age of migrants is much older than for the Australian-born population (44 compared to 38).
The most common country of birth for migrants living in Queensland was New Zealand; in Victoria it was India; in NSW it was China. There may not be too many more censuses until the most common migrant nationally was not born in England.
Ahead of the forthcoming federal budget, there has been a lot of media and policy attention on housing affordability. Today’s release of census data points to some subtle differences across Australia that may influence policy responses.
Nationally, the most common tenure type is owning a three-bedroom home with a mortgage. In Queensland, however, renters make up a roughly equal share of the population. But, in Tasmania and NSW, more people own their own home outright. And in the NT, renting is the most common tenure type.
In a finding that won’t surprise many, the typical female does a bit more unpaid work around the house than the typical male. The most common category for males is less than five hours a week. The most common for females is five to 14 hours.
We won’t know how this compares to paid work for a while yet – or whether these differences vary depending on age.
What future releases will tell us
The profiles released today offer us limited information. But the census remains one of Australia’s most important datasets.
When detailed data are released in June and then progressively throughout the rest of 2017, we will be able to dig deeper into small geographic areas or specific population groups.
We will be able to ask if there are pockets of Australia with significant socioeconomic disadvantage, and if it is worsening. We will be able to hold governments accountable for the progress we have made on the education, employment and health outcomes of the Indigenous population.
And we will be able to test whether the languages we speak, the houses we are living in, and the jobs that we are doing, are changing.
But those questions rely on a high-quality census.
The attention on the 2016 Census until now has been mostly negative. There was increased concern related to data privacy, the failure of the online data entry system on census night, and staff cuts at the ABS.
In October 2016, the ABS estimated the response rate to the 2016 Census was more than 96%, and that 58% of the household forms received were submitted online. But what matters more than how many people filled in the census and how they did it is whether the responses given were accurate. We therefore need to see a lot more interrogation of the data before taking the results at face value, but we can remain cautiously optimistic.
The ABS will be hoping that now some data is released, attention will shift to what the results tell us about Australian society. It is to be hoped the data will be robust, the insights will be newsworthy, and policy and practice will shift accordingly.
We won’t know this for sure until the first major data release of data June 27 – the data released today were just a sneak peak.