New questions in the 2006 population census: some initial findings.

Hugo, Graeme

The 2006 Australian census covered several new topics. These include questions that enable the author to estimate the numbers and characteristics of persons with a profound or serious disability and their carers. Questions were also asked about the level of unpaid work. The study highlights the extent to which females engage in such work. A key finding is that women aged 35-44 bear a heavy unpaid workload, relative to males, despite a high rate of labour force participation. The article also reports the results of 2006 census questions on volunteering, on internet connections and numbers of children ever born.

INTRODUCTION

Much demographic change is incremental rather than sudden and hence can creep up on policy makers. It is important therefore to regularly take a snapshot of the population and society through the census of population and housing to take stock of such change. However, in order to be able to detect and measure these changes the census must include the right questions which address not only contemporary but also emerging, social and population issues. If the census is to provide policy makers and society generally with timely, relevant and accurate information it needs to keep up, or even be ahead of, changes in Australian society. Accordingly the Australian Bureau of Statistics (ABS) canvasses widely regarding the topics that should be incorporated in each census. At the 2006 census four topics were included which were not in the 2001 census questionnaire although one of the questions has been asked before and is currently in a cycle of being included in every second census. However there are three totally new topics that reflect emerging issues in Australian society of increasing significance. This paper discusses some of the initial results from each of the four questions.

NEED FOR ASSISTANCE

The Australian population aged 65 years and over will double over the next 30 years but the numbers aged 75 and over will treble and those aged 85 plus will quadruple as is shown in Table 1. A recent paper (1) used the ABS Survey of Disability, Aging and Carers to project that the numbers of Australians requiring assistance will increase from 956,635 in 2003 to 1,215,407 in 2011, 1,713,732 in 2021 and 2,215,041 in 2031. (2) However, our understanding of the changing level and patterns of need for assistance in day-to-day activities remains limited. Accordingly it was decided to include a suite of questions regarding the need for assistance in the 2006 census and these questions are shown in Figure 1. The responses to questions 20, 21 and 22 have been used by the ABS to develop a Core Activity Need for Assistance variable to establish the number of people with a profound or severe disability who need assistance in day-to-day activity. People with a profound or severe disability are defined as needing help or assistance in one or more of the three core activity areas of self-care, mobility and communication because of a disability, long-term health condition (lasting six months or more), or old age. All told, some 821,649 Australians or 4.4 per cent of the total population indicated that they needed someone to help them and had a profound or severe disability. The non-response rate was high at 6.4 per cent however. As would be expected there is considerable variation with age. Figure 2 shows that for both females and males there is a regular increase in the percentage needing care after age 20 and a steepening after age 70. In the younger ages the rate is higher for males than females but there is a crossover at age 65 and, for the oldest groups, the need for assistance is greater among women. For males, eight per cent of 65 to 74 year olds have a severe disability compared with 8.3 per cent of women while for those aged 75 to 84 the proportions are 18.9 and 24.1 percent and those aged 85 and over 43.5 and 57.5 per cent. This seems to reflect the fact that most older men live in couple households and many may not report the care provided by their partner while a much greater proportion of older women live alone and do not have a partner to care for them. Overall, 4.1 per cent of Australian males have a profound disability and 4.8 per cent of women, reflecting the older age structure among women (ABS defines profound as including those indicating a profound or severe disability).

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Much of the interest in the disability variable will be in analysis of differentials between different socioeconomic, geographical, ethnic and other subgroups. Some of the potential for such analyses is evident in Figure 3 that compares the age-specific incidence of disability for the Indigenous and total populations. The rates are higher for Indigenous men (4.9 per cent) than for all men but a little lower for Indigenous women (4.4 per cent compared with 4.8 per cent). However, it will be noted that the rates are higher for all age groups for Indigenous people reflecting their disadvantaged situation. Moreover, it seems that cultural factors may lead to significant underrepresenting of disability among the Indigenous population.

As Figure 1 indicates, a question was included in the census on persons providing unpaid assistance to a person with a disability. The ABS has used the responses to this question to develop an Unpaid Assistance to a Person with a Disability variable which records people who, in the two weeks prior to census night, spent time providing unpaid care, help or assistance to family members or others because of a disability, a long-term illness or problems related to old age. The extent of unpaid care given by Australians is considerable with 8.8 per cent of males aged over 15 and 13.5 per cent of females providing such care, although the rate of non-response was quite high at 10.1 per cent. Nevertheless, 1.6 million adult Australians indicated they provided unpaid assistance to people with a disability.

[FIGURE 4 OMITTED]

Figure 4 shows the age-sex specific rates of providing this assistance and some very interesting patterns are evident. As would be anticipated women outnumber men as carers in all but the very oldest groups. The gender differential peaks in the early baby boom (55 to 64) age group at 21.3 per cent of all women providing care and 12.6 per cent of males. Interestingly, at older ages there is a convergence in male and female carer rates and, in the very oldest ages, there is a higher rate of unpaid caring among men.

The census data underlines the fact that in contemporary Australia the majority of care given to older persons, children and other persons with a disability is provided by the family. (3) Caring in Australia has been predominantly an unpaid, and family, role. This has increased with the success of federal, state and local government initiatives to keep older Australians with disabilities 'at home not in a home'. (4) Indeed the ABS Surveys of Disability, Ageing and Carers (5) have shown that the proportion of older people with disabilities who live in cared accommodation fell from 13.2 per cent in 1998 to 11.5 per cent in 2003 while the reduction for those with severe disability fell from around a third to 27.4 per cent. Despite the increase of over 150,000 in the number of older disabled persons there were absolute declines in the numbers with acute disability living in cared accommodation (from 159,946 to 153,906) as there was for all reported disability (161,726 to 160,022). While the ageing of the population will increase the overall numbers of Australians with a disability over the next three decades, the ability of the family to provide that care may be reduced over this period because:

* lower fertility among baby boom mothers compared with that of their mothers will mean that they will have fewer children available to care for them

* increased mobility means that there is a greater likelihood that the smaller number of children will not be living close to their older parents needing care

* increased incidence of divorce and separation means that a smaller proportion can be expected to have a partner who is co-resident and able to provide support than has been the case for the present generation.

If policies designed to increase participation rates in the formal workforce among women, older Australians and other groups are successful it will mean a smaller proportion will be available to provide care. (6) Indeed the need for Australians in their 50s and 60s to provide care for older family members may militate against contemporary policy imperatives to increase workforce participation, especially among older people.

DWELLING INTERNET CONNECTION

The 2001 Australian census was the first to include questions on the use of information technology. It asked whether persons had used a personal computer at home and also whether they had used the internet in the last week. The results from these questions were analysed in an Australian Census Analysis Project monograph (7) which indicated that 43 per cent of the population had used a computer at home and 38 per cent the internet. At the 2006 census there was only one question included relating to information technology and this sought to measure how widespread household access is to the internet, both dial-up and broadband. The level of non-response to the question was moderate at 3.3 per cent but of those who answered the question some 63.4 per cent indicated they had an internet connection--40.5 per cent have broadband and 22.9 per cent dial-up. This represents an increase of around two-thirds over the five-year period and reflects the rapid spread of the internet in Australia.

There will be much interest in establishing the variations between groups and areas in access to the internet. The 2001 census showed considerable variations by age, education level and ethnicity. (8) One of the strongest contrasts in 2006 was between capital cities and non-metropolitan areas. While 66.8 per cent of metropolitan households had the internet, 57.5 per cent of households in non-metropolitan areas had it. Moreover there are even greater differences in access to broadband with 45.9 per cent of metropolitan households contrasting to only 31.3 per cent of their non-metropolitan counterparts. On the other hand a larger proportion of non-metropolitan households (25.6 per cent) had dial-up access than metropolitan (20.2 per cent). These data undoubtedly will add fuel to the debate in Australia regarding urban-rural inequalities in access to broadband and its implications for diversifying economies in regional areas.

There were some differences between states with respect to access to the internet with the proportions of households ranging from 75.2 per cent in the ACT to 54.9 per cent in Tasmania. Indeed the growing inequality gap between the eastern mainland states and Western Australia on one hand and the Northern Territory, South Australia and Tasmania on the other identified by Harding, Lloyd and Greenwell (2001) seems to be underlined by these data. In Queensland the connection rate was 64.8 per cent, NSW 63.6, Victoria 63.1 and Western Australia 65.2, while for the Northern Territory it was 60.2 and South Australia 58.2 per cent.

[FIGURE 5 OMITTED]

The proportion of households with internet is higher in separate houses (65.9 per cent) than it is in either semi-detached houses (41.9), flats (44.9) and other dwellings (63.2 per cent). These differences reflect both a socioeconomic and age gradient in access to internet. Figure 5 shows how the proportion of households with internet connections increases with gross household income. There are exceptions in the lowest income categories that perhaps equate with a significant number of student households having computers and internet access. Socioeconomic differences are evident too in the fact that, while 83.7 per cent of couple households with children had internet access, this was the case for only 61.5 per cent of single-parent households. There is then some evidence of a digital divide in Australian society. The fact that internet connection is inversely associated with age is evident in the fact that 62 per cent of couple households without children have internet access while only 33 per cent of single-person households had internet access. Clearly a majority of households with older people in them in Australia do not have internet connections. This has important implications for any policy interventions that seek to break down loneliness and isolation among older Australians and improve their access to services and information through using the great potential of the new information technology.

NUMBER OF CHILDREN EVER BORN

With the realisation that no demographic variable is more influential in shaping the future trajectory of ageing than fertility, the level of fertility is increasingly the focus of attention in OECD countries. However, after the 1986 census the ABS decided that it would ask a fertility question on a ten-year cycle in every second Australian population census. Part of the reasoning for this was the fact that birth registration and hospital-based midwives' data on births are comprehensive. Nevertheless the utility of that data for analysing differences in fertility between subgroups in Australian society remains limited and the census remains the main source for examining trends in differences in fertility between socioeconomic, ethnic, birthplace, educational and other groupings in Australian society. Census fertility data are also critical for the examination of lifetime fertility and childlessness. Moreover the timing of the 2006 census with respect to fertility is important. There has been a well documented upturn in the total fertility rate since 2001. (9) The 2006 census question asked all women to indicate the number of babies they had given birth to including all live births but excluding adopted, foster and stepchildren.

The census data indicate that the extent of childlessness among Australian women aged 15 years and over increased only slightly from 31.4 per cent in 1996 to 32 per cent in 2006. The age-specific pattern of childlessness is shown in Table 2 and indicates that there are greater intercensal differences at specific ages than there is for the population overall. In fact the incidence of childlessness increased between 1996 and 2006 up to age 64 but reduced in older ages compared with 1996. The increases are especially great among younger women with differences of 6.8 and 7.5 percentage points in the 25 to 29 and 30 to 34 age groups respectively. This reflects the well-documented pattern of young Australian women delaying the commencement of childbearing. However, the higher rates of childlessness among women in their 40s and 50s indicates that there has been an increase in the proportion of women with no children when they leave the childbearing ages.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

Turning to the average number of children for all woman aged over 15, there was a small overall decrease between 1996 and 2006 from 1.78 to 1.72 but, as Figure 6 shows, there was some variation with age. As is the case with childlessness, there is a crossover around age 64 with younger women in 2006 having a smaller number of children than women at the equivalent ages in 1996 but older women having more children. This obviously reflects the replacement of the lower fertility mothers of the 1930s and early 1940s low fertility years by the higher fertility baby boom mothers in the oldest age groups. It is interesting, however, that even in the 'catch up' older childbearing age groups of 30 to 34, 35 to 39 and 40 to 44 there was a reduction of 0.24, 0.21 and 0.16 of a child in the average number of children. It will be interesting to see if the increase in the total fertility rate observed in recent years makes up any of this difference in average number of children by the time the current generation of women in their 30s leave the childbearing ages.

The average number of children of Indigenous women (2.06) was higher than for non-Indigenous women despite the former being a much younger population with a higher proportion not yet beyond childbearing ages.

Figure 7 compares the age-specific average number of children for Indigenous and non-Indigenous women and it will be noted that they are higher for Indigenous women in each age category. The figure shows that the largest difference (1.07 children) was in the 25 to 34 peak childbearing age grouping.

Table 3 compares the average number of children that were born to women from different birthplace groups. Only those in the 35 to 44 and 45 to 54 age groupings who have nearly or already completed fertility are shown but wide variations are in evidence. Very few birthplace groups have more children on average than the Australia-born. These were predominantly from mainly Islamic nations although women from other Islamic nations like Indonesia have fewer children on average than the total Australian population. The noticeable feature, however, is the low levels of fertility among Asia-born groups. This reflects the high degree of selectivity of the migration process which focuses on skill and higher education and favours low fertility groups but also is a reflection of the dramatic falls in overall national fertility levels in most Asian countries.

It will be interesting to examine any socioeconomic differentials in fertility once the requisite tabulations are available. There is considerable debate about the nature of such differentials and whether or not they are widening. However, in this context it is worth noting a recent ABS (10) analysis of total fertility rates (TFRs) for statistical local areas against socio-economic-indices-for-areas scores. This indicated that a disproportionate amount of recent increases in the TFR was recorded by women living in higher socioeconomic areas.

[FIGURE 8 OMITTED]

UNPAID WORK

It has long been recognised that much of the work carried out in Australia is unpaid although official data on work have only related to paid work. In the consultations carried out by the ABS prior to censuses over the last three decades there have been strong representations to measure unpaid work. As a result the ABS included four questions on unpaid work in the 2006 census and these are reproduced in Figure 8. They cover four areas of unpaid work and were asked of all people 15 years and older. The first question asked for information on the time spent doing household, domestic work. As would be expected, the proportion of males reporting they did no domestic work (28.1 per cent) was significantly higher than for women (17.1 per cent). Moreover the median number of hours spent per week on these tasks was higher for women (11.0 hours) than men (3.8 hours). However Figure 9 shows that there are wide variations between age groups in both the amount of household work done and in the gender differential. The highest level of participation for both men and women is in the 35 to 44 years age group. However the gender differential is also greatest in these peak family rearing age groups as well. Given increasing levels of workforce participation among women in these age groups it seems that many Australian women experience the double load of full-time paid work as well as substantial unpaid workloads as well.

If we focus on the median number of hours spent on housework, Figure 9 shows some significant differences between men and women. The number of hours put in by women are more than double those of males in the workforce age groups of 20 to 64. There is an increase in the number of hours spent on housework both among males and females in the early retirement years. Indeed for men the peak average number of hours spent on housework occurs in the 65 to 74 age group. For older women, however, the peak number of hours compares with the 35 to 44 age group, although there is a small increase between the 45 to 54 age category and the 55 to 64 and 65 to 74 age categories.

[FIGURE 9 OMITTED]

The second unpaid work question related to the provision of unpaid care to family members and others who are disabled. This was discussed earlier in conjunction with the discussion on the incidence of disability. The third question relates to unpaid childcare and again a significant gender differential is in evidence with 31.5 per cent of women reporting that they provided unpaid childcare in the fortnight before the census compared with 23.1 per cent of men. A third of the women provided some care to children other than their own compared with 23.2 per cent of the men. There are strong variations between different age groups as is evident in Figure 11. As would be anticipated the peak childcare age group is 35 to 44 when 61.8 per cent of women and 49.5 per cent of men reported providing such care. Nevertheless the importance of grandparents playing a significant childcare role is strongly in evidence with high rates of childcare provision in the 55 to 64 category (22.6 per cent for women and 12.8 per cent for men) and 65 to 74 (16.9, 12.1) category.

[FIGURE 10 OMITTED]

[FIGURE 11 OMITTED]

There is then a distinctive pattern emerging for women aged 35 to 44 of heavy workloads in terms of unpaid domestic work as well as childcare and care for family members with disabilities. When this is added to a high level of participation in the paid workforce the pressure on many Australian women in this age group is evident.

The final question relates to volunteering and asks if adults did any voluntary work through an organisation or group in the twelve months preceding the census enumeration. It has long been recognised that volunteers play a critical role in providing many services to the community which otherwise would have to be paid for. It will be interesting to analyse these data from a number of perspectives, especially variations between communities and groups in the rate of volunteering. As in the case of the other three questions on unpaid work a higher proportion of women are volunteers (20 per cent) than men (15.7 per cent). This is the case for all age groups, as Figure 12 indicates, although the differential is least in the older age groups. Again one is struck that it is women in the 35 to 44 age group who have the highest level of participation in the voluntary sector. There is less variation by age in volunteering than there is in other areas of unpaid work.

[FIGURE 12 OMITTED]

CONCLUSION

The population census has the important role of influencing policy makers and the community generally by providing information about the pace and nature of social, economic and demographic change in Australia. As a result censuses need to be constantly changing in terms of the questions they ask and the concepts and classifications they employ to present their findings. The 2006 census had more new questions than most recent enumerations and each new question reflects a dimension of Australian society which is emerging as being of greater significance. It is crucially important that Australia's census keeps attuned to, and even slightly ahead of, change in the community so that it can capture the scale and nature of that change.

References

(1) G. J. Hugo, 'Contextualising the "Crisis in Aged Care" in Australia: A demographic perspective', Australian Journal of Social Issues, vol. 42, no. 2, 2007, pp. 169-182

(2) Need for assistance is defined by the ABS as 'a person with one or more disabilities or aged 60 years and over, is identified as having a need for assistance with an activity if, because of their disability or age, they report that they need help or supervision with at least one of the specified tasks constituting that activity'.

(3) See Australian Institute of Health and Welfare (AIHW), The Future Supply of Informal Care 2003-2013, AIHW, Canberra, 2003; Carers in Australia: Assisting Fail Older People and People with a Disability, AIHW, Canberra, 2004; Australia's Welfare 2005, AIHW, Canberra, 2005.

(4) Parliament of the Commonwealth of Australia, In a home or at home: accommodation and home care for the aged, Report from the House of Representatives Standing Committee on Expenditure, Australian Government Publishing Service, Canberra, 1982

(5) Disability, Ageing and Carers: Summary of Findings, Australia, Catalogue No. 4430.0, ABS, Canberra, 2004

(6) G. J. Hugo, South Australia's ageing workforce in an international context, Presentation to Seminar Series of ARC Linkage Project on Demographic Change, Ageing and the Workforce: An Integrated Model to Inform Workforce Development and Planning in Australia, Adelaide, 17 March 2006

(7) J. Primrose, 2001 Census: Computer and Internet Use, census Paper No. 03/03, ABS, Canberra, 2003

(8) ibid.

(9) P. McDonald, 'Has the Australian fertility rate stopped falling'?, People and Place, vol. 13, no. 3, 2005, pp. 1-5; Australian Social Trends, 2007, Catalogue No. 4102.0, ABS, Canberra, 2007, p. 9

(10) ibid.

Table 1: Projected growth of the population aged 75 to 84 and 85 plus,
Australia

 65 to 74 75 to 84 85 plus
 per cent per cent per cent
Year number growth number growth number growth

2001 1,321,708 905,742 288,846
2011 1,709,849 2.6 1,029,943 1.6 431,806 5.2
2021 2,469,548 3.7 1,418,022 3.2 584,403 3.1
2031 2,841,728 1.4 2,082,350 3.0 856,061 3.9

Source: ABS, Estimated Resident Population data and 2005 Projections,
Series B

Table 2: Per cent women childless by age, Australia, 2006 and 1996

 1996 2006

15 to 19 years 96.5 97.6
20 to 24 years 81.5 85.5
25 to 29 years 56.3 63.5
30 to 34 years 29.0 36.5
35 to 39 years 16.8 21.2
40 to 44 years 12.8 15.9
45 to 49 years 10.7 14.4
50 to 54 years 9.7 13.1
55 to 59 years 8.9 11.3
60 to 64 years 8.8 10.2
65 to 69 years 9.7 9.4
70 to 74 years 11.0 9.2
75 to 79 years 12.4 10.1
80 to 84 years 14.6 11.1
85 years 18.0 13.0
Total 31.4 32.0

Source: ABS population censuses of 1996 and 2006

Table 3: Average number of children per women aged 35 to 44 and 45 to 54
by birthplace, Australia, 2006

 35 to 44 45 to 54

Australia 1.94 2.20
Lebanon 3.10 3.51
Iraq 2.59 2.82
Turkey 2.23 2.44
Egypt 2.19 2.18
Malta 2.10 2.39
Bosnia-Herzegovina 1.93 2.01
Croatia 1.93 2.21
New Zealand 1.91 2.12
Fiji 1.88 2.13
Greece 1.88 2.30
Papua-New Guinea 1.88 2.32
South Africa 1.88 2.12
United Kingdom 1.87 2.09
Italy 1.85 2.32
Vietnam 1.84 2.19
Netherlands 1.81 2.29
Sri Lanka 1.76 1.85
India 1.75 1.92
Ireland 1.75 2.13
Philippines 1.72 1.78
Indonesia 1.63 1.99
United States 1.61 1.85
Poland 1.60 1.71
South Korea 1.59 1.78
Malaysia 1.59 1.81
Singapore 1.59 1.78
Canada 1.56 1.75
Germany 1.53 1.82
China 1.41 1.38
Thailand 1.31 1.53
Japan 1.16 1.45

Source: ABS 2006 census