Economic advantage and disadvantage among older Australians: producing national and small area profiles.
Gong, Cathy ; Kendig, Hal ; Harding, Ann 等
1. INTRODUCTION
Theories of 'cumulative advantage/disadvantage' suggest
that inequalities across the life course underlie the increasing gulf
between the well-off and the financially disadvantaged in later life
(Dannefer, 2003). The social and economic resources that enable people
to buy into housing and neighbourhoods and pay for transport reflect
life-long inequalities (Dannefer and Kelly-Moore, 2009). Quality of life
among older people is not only influenced by education, labour force
participation and health during their early and working years, but also
by the housing market and processes of residential location and change
(O'Rand, 2006; Phillipson, 2010; Kendig et al, 2012).
Among developed countries Australia is notable for having an older
population with high levels of home ownership notwithstanding relative
low incomes with most on a moderate government pension (Pynoos et al.,
2007). The locations of older people, especially those who are home
owners and some public tenants, typically reflect household incomes and
housing markets applying many decades earlier. Mobility rates and
housing outlays are generally low for older people, irrespective of
local housing markets, while younger buyers and private tenants must
have the resources to meet current market prices (Kendig et al., 2012;
Warnes, 2010). These life span developments and housing market processes
can result in major disjunctions in the economic resources of younger
and older people within small areas.
Local environments are especially important for older people who
spend most of their time at home and in their neighbourhood; many have
low incomes and mobility limitations and do not drive (Marmot et al,
2010; Kendig et al., 2012; Kendig and Phillipson, in press). For
example, Kendig et al. (2012) report that the home and neighbourhood are
important for the independence, social participation, and well-being of
older people in Australia and other developed countries. However,
previous locational analyses of older Australians have not taken much
account of the ways in which socio-economic resources are associated
with spatial heterogeneity among older Australians (Miranti et al.,
2010). Patterns of residential development and urban infrastructure are
important policy concerns particularly for cohorts entering later life
with rising car ownership levels and aspirations for continuing mobility
in daily life (Berry, 2007).
In an earlier paper Gong et al. (2012) presented a spatial
microsimulation model that created small area synthetic data to study
advantage and disadvantage among older Australians. It mainly focused on
the development of the model including procedures for data linking,
benchmarking, reweighting and validation. The present paper applies this
model in order to identify and interpret spatial concentrations and
clusters of deep economic disadvantage and relative economic advantage
among older Australians. The use of multiple dimensions of advantage and
disadvantage, specific to older populations reveals spatial
distributions that are not apparent in the overall socio-economic
wellbeing of an area for the wider population as measured by the
Socio-Economic Indexes for Areas (SEIFA) 2006 (ABS, 2008). The spatial
microsimulation model draws on the national ABS Survey of Income and
Housing (SIH) and the Census. The spatial units of our analysis are
based on the Statistical Local Areas (SLAs), with some of them being
aggregated up to Local Council Electoral Wards for Brisbane and
Statistical Subdivisions (SSD) for Canberra. Economic disadvantage and
advantage are defined by drawing on the concept of social exclusion,
with its emphasis on multiple sources of disadvantage, rather than any
single measures such as income poverty (Hayes et al., 2008). The methods
and findings on locational patterns in economic well-being among older
Australians can provide analytical tools to inform policy development
and decision-making concerning housing and public and private service
provision.
2. THE CONCEPT OF WELLBEING FOR OLDER PEOPLE
The extensive debate on the concept of social exclusion has tended
to focus more on children and working aged people with less attention to
those in later life. There are numerous studies of spatial differences
in a range of socioeconomic indicators for the total population and for
children (see, for example, ABS, 2008; Daly et al, 2008; Lewis and
Corliss, 2009; Abello et al, 2012). Yet despite the substantial risks
faced by many older Australians - in terms of cumulative
material/financial disadvantage, losing partner or living alone, lack of
access to services and community activities, and social isolation and
age discrimination, little work has been done on socio-economic and
spatial dimensions of differences in ageing experience (Davies, 2005;
Tanton et al., 2009; Miranti et al, 2010; Gong et al., 2012; Lui et al,
2011).
Broad measures of economic well-being suggest that differences in
incomes between older and younger Australians may be more marked than in
many other affluent nations. OECD data show that Australia ranked as one
of the lowest countries in terms of the ratio of average equivalised
disposable income of people aged 65 and over to that of people aged 18
to 64 years. According to OECD Statistics, the income ratio of
retirement age (65+) to working age (18-64) in Australia was 0.63 in
2000, ranking second lowest among 26 OECD countries (only higher than
that of United Kingdom). Although this ratio slightly increased to 0.64
by the mid-2000s, Australia still ranked fourth lowest among 30 OECD
countries (higher than Ireland, New Zealand and Korea). Similar findings
are reported by the Global Age Watch Index 2013 for old populations aged
60 and over. Among 91 countries, Australia is rated at 57 for economic
security (indicated by income alone) while overall quality of life is
rated at 14 after combining income, health, employment, education and
environment.
The cross-national differences may be due in part to a less
generous social security system for older people in Australia than in
many other OECD countries. The consequences for standards of living in
later life may be mitigated, however, by Australia's high rates of
outright home ownership in later life and consequent low housing outlays
(Kendig and Bridge, 2007; Kelly, 2009). While these international
comparisons do not provide data about differences between particular
groups of older adults, they do suggest the importance of developing
further knowledge about the economic well-being of older Australians.
Although Australia has experienced two decades of economic growth and
rising average incomes, some people in some communities continue to be
'left behind'. Further older people in the most disadvantaged
neighbourhoods are more likely to be socially excluded and hence have a
very low quality of life (Scharf et al, 2005).
Research has identified social exclusion as a major issue for the
ageing population (Barnes et al., 2006; Naughtin, 2008; Miranti and Yu,
2011; Lui et al. 2011). Naughtin (2008), drawing on national survey data
from the English Longitudinal Study of Ageing (ELSA) and policy work
done by the UK Social Exclusion Unit Office of the Deputy Prime Minister
(2006) and Barnes et al. (2006), has adapted to an Australian context
seven dimensions of social exclusion applicable to later life. These are
social relationships, cultural activities, civic activities, access to
basic services, financial products and material consumption. Naughtin
(2008) also notes that the risk of social exclusion for older people
increases with various factors--including age (with those 80 years and
above being more prone to exclusion), living alone or having no
children, poor mental or physical health, no access to a private car or
lack of access to public transport, living in rental accommodation,
having low income and/or being reliant on welfare and having no access
to a telephone. In Australia, Miranti and Yu (2011) and Mclachlan et al.
(2013) have examined the persistence of social exclusion for older
people. They found higher risks of persistent social exclusion for older
people with poor educational attainment, older people living in the most
disadvantaged areas, older people who have less engagement in employment
since finishing fulltime education, and older people with caring
responsibilities, or who have a disability themselves.
3. MEASUREMENT, DATA AND VALIDATION
There is no consensus on where older adulthood begins and, in many
cases, the choice of age groups to represent later life is a contextual
one. We define older people as those aged 55 years and above, including
Australians who are already retired as well as substantial numbers who
are moving towards retirement. Based on 2006 Census data, only 53
percent of Australians aged 55-64 are still in the labour force and, of
those who are in paid work, 30.8 percent work part-time. As discussed
above, older Australians have relatively low income but high home
ownership. It is important to move our economic wellbeing analysis for
older people beyond income alone by incorporating homeownership and
other crucial variables, such as dependence on government benefits. We
investigate multiple dimensions and disparities of economic advantage
and disadvantage by contrasting two relatively extreme groups defined as
follows (see also Gong et al, 2012):
* Deep economic disadvantage (6.6 percent of the older population).
These individuals are in the bottom quintile of the equivalised national
household disposable income distribution, paying public or private rent,
and relying mainly on government income benefits (more than half of
their household income is from government benefits);
* Relative economic advantage (16.5 percent of the older
population). These individuals are in the top two quintiles of the
equivalised national household disposable income distribution, paying no
rent or mortgage, and relying mainly on private household income (more
than half of their household income is from private income, including
superannuation).
We include people renting in both the public and private rental
markets in our 'deep economic disadvantage' group. Neither
group has the wealth in their residences that can be central to the
economic well-being of older home owners. Although public housing
tenants pay lower rents than those in the private market, they
nevertheless generally have very low income, while older Australians
renting in the private market are widely acknowledged as a group likely
to experience substantial housing stress (Tanton and Phillips, 2013). In
addition, their housing tenure may also be associated with other
disadvantages including poor housing quality and insecurity of occupancy
(Kendig and Bridge, 2007).
It is important to note that our two groups purposefully contrast
extremes of economic disadvantage and advantage to better understand
spatial inequalities. In fact, the majority of older Australians fall
into neither of our groups. The most common economic situation for older
Australians (especially those past the usual retirement age of 65 years)
is to have outright ownership of their home (thus excluding them from
our deep economic disadvantage group) combined with low income (thus
excluding them from our relative economic advantage group). In addition,
our measure of disadvantage is a narrow one, and it is important to
understand that many economically vulnerable older people (e.g. home
owners reliant only on the age pension) are not included in our
'deep disadvantage' group.
We present results at both a national and small area level, drawing
on a range of data sources and multiple measures of advantage and
disadvantage. First, the Survey of Income and Housing (SIH) 2005/06 is
used to draw a national picture. Second, synthetic small area estimates
created from a spatial microsimulation model, by combining the
information in SIH 2003/04, SIH 2005/06 and Census 2006, are used for
small area analysis. The Census data 2006 has also been used to validate
our synthetic estimates at a small area level. Both SIH and Census data
are collected by the Australian Bureau of Statistics (ABS). This
analytical approach yields detailed socio-economic information otherwise
unavailable on a small area basis.
We use the Statistical Local Area (SLA) as our base spatial unit of
analysis (see ABS, 2007 for a full description). To allow comparison of
regional characteristics across Australia, SLAs in Brisbane and Canberra
were aggregated into larger geographic units, so that they were more
similar in population size to SLAs in other areas--Local Council
Electoral Wards for Brisbane and Statistical Subdivisions (SSD) for
Canberra. This aggregation methodology follows that developed by Baum et
al. (2005) and used in Daly et al. (2008) and Gong et al. (2012). After
aggregation, the population sizes of older people aged 55 years and over
in the spatial units on which this study is based ranged from 100 to 31
235. The spatial units therefore vary from small neighbourhoods to
relatively large communities.
Census enumerations could have been used to identify broad groups
of economically vulnerable and relatively advantaged elders, using
variables available in the Census such as gross income and tenure type
However, this approach has significant limitations in terms of our
research purpose. First, the income collected in the Census is household
gross income by income ranges, which cannot be used to measure the
actual household living standard after paying income tax and adjusting
for household size (Saunders and Bradbury, 2006). Second, income source
information is not available in the Census, which makes it impossible
for us to combine the multiple measures of income, housing cost and
welfare dependence into one complex indicator.
The spatial microsimulation model used in this study provides the
methodology to generate small area estimates otherwise unavailable for
older Australians experiencing multiple economic disadvantage or
advantage. The spatial microsimulation model uses reweighting technology
to create a synthetic household data set by combining the superior
detailed data available from a national survey (e.g. SIH) with Census
data which covers almost all households in private dwellings, including
retirement villages (Tanton et al., 2011). The Census data 2006 was used
to set up the benchmarks (same as in Gong et al., 2012) for reweighting
as well as to validate our synthetic estimates at a small area level.
This approach yields synthetic household weights for each small area in
Australia; it replicates, as closely as possible, the characteristics of
the actual households within each small area in Australia (Chin et al,
2005; Chin and Harding, 2007; Lymer et al., 2008; Vidyattama and Tanton,
2010; Harding et al, 2011). A full discussion on how to use this
methodology to generate the small area estimates for this study is
described in Gong et al. (2012).
Spatial microsimulation allows us to produce accurate synthetic
estimates of household characteristics for the vast majority of small
areas in Australia. However, the analysis excludes some small areas
(166) for which sufficiently accurate weights could not be obtained.
Most of these areas have very small populations and unusual
characteristics (e.g. industrial areas). In the Northern Territory,
these limitations required the exclusion of almost half of the
Territory's population and thus the estimates for the Territory
should be treated cautiously. We also excluded additional small areas
with small sample sizes for the key variables. This leaves us with a
final sample of 908 small areas, containing around 99 percent of the
in-scope population aged 55 and above. In addition, because the SIH
includes only records and the Census only collects income information
for older people living in occupied private dwellings (including
retirement villages), our synthetic estimates exclude people living in
non- private dwellings (such as residential care centres). The inability
to estimate the small area data of older people in non-private dwellings
and very remote areas may reflect the weakness of small area estimation
by combining SIH with Census data as discussed by Vidyattama et al.
(2013). Households with zero and negative incomes are also removed from
our sample.
The validation of synthetic small area estimates created by a
spatial microsimulation model is extremely important (Ballas and Clarke,
2001). Before the data was used for this study, Gong et al. (2012) had
conducted some validation for the key variables: (1) household income
and housing tenure type at small area level with similar variables from
the Census (see Figures 4 and 5 in Gong et al, 2012); (2) dependence on
income from government sources at a national and state level using a
similar variable from ABS Survey of Income and Housing (SIH) (see Table
2 in Gong et al, 2012). It was found that the model estimates are robust
for a large set of small areas and the ranks of the states are fairly
closely aligned with SIH estimates, although, the synthetic estimates
are slightly different from the SIH estimates.
Data from the model was further validated for this study against
data of 2006 Census and 2008 welfare recipients. As the welfare
dependence data is not available in Census 2006, the 2008 small area
data of the number of welfare recipients aged 16 and over from the
Social Health Atlas of Australia 2010 is used. Both the Standard Error
about Identity (SEI) and Pearson's correlation are used to measure
the accuracy of our small area synthetic estimates against the existing
small area data which were found to be similar. The SEI measures the
total difference between two estimates, while the Pearson's
correlation measures whether two estimates have a similar ranking across
areas.
Figure 1 compares our synthetic estimates of the number of people
aged 55 years and over with main income from government benefits against
the number of welfare recipients aged 16 plus at a small area level. The
R-square (0.96) of the regression line in the middle indicates a very
high correlation between these two estimates, which is further evidenced
by the very high value of Pearson's correlation (0.98). It is
expected that the SEI value here is relatively low (0.62) because the
two data sets are measuring different populations though their rankings
are comparable.
[FIGURE 1 OMITTED]
Table 1 presents further results of validations for our small area
synthetic data for each of the variables for which comparable data is
available in the Census and their combinations against similar estimates
derived from Census data. The accuracy is very high for all the
variables and combined measures. The SEI and Pearson's correlation
of the proportion and number of older people aged 55+ in each of the
disadvantaged and advantaged groups are higher than 0.90 except for a
relatively low but still acceptable SEI value (0.84) for the group
paying private rent. The SEI values in this table are higher than in
figure 1 because the Census estimates and our estimates of household
income and tenure type are much closer.
4. RESEARCH FINDINGS
National Findings
We first present a national picture of economic disadvantage and
advantage among older people in Australia using the Australian Bureau of
Statistics (ABS) survey of Income and Housing (SIH) 2005/06. Table 2
shows the summary 'deep disadvantage' estimates as well as
those for its components defined, as noted above, as being in the bottom
quintile of the income distribution, paying rent, and relying mainly on
government income. It shows that 36 percent of people aged 55 years and
over are in the national bottom quintile of household equivalised
income, 12.5 percent are still paying rent (of which 8.3 percent are
private renters and 4.2 percent public renters), and 45.6 percent have
their main household income (more than half) from government benefits.
Only 6.6 percent of older adults fall into our definition of deep
economic disadvantage, largely due to the small proportion of older
people who are still paying rent.
There is considerable variation on the distribution of
disadvantage. Among those living alone, 15.7 percent fall into our
'deep disadvantage' group; they are more likely than other
groups to be in the bottom income quintile, reliant on government
benefits, and to be paying rent. In sharp contrast, only 0.8 percent of
older people living in a household with at least one person employed
fall into our deep disadvantage definition. There are some gender
differences. On average, women are more likely than men to be in the
bottom income quintile, reliant on government benefits or paying public
rent. Some modest differences are evident in this set of variables
between capital city and balance of state areas. The proportion of older
adults reliant on government benefits and in the bottom income quintile
is slightly higher in non-capital city areas. Capital city areas have a
higher proportion of older people paying public rent, reflecting the
historical location of these public investments.
Table 3 shows that the proportion of older adults falling into the
advantaged group is 16.5 percent, somewhat larger than that falling into
our multiple economic disadvantage group (6.6 percent). This group
includes, as noted above, those in the top two quintiles of the income
distribution (26.2 percent), paying no rent or mortgage (69.4 percent),
and relying mainly on private household income (54.4 percent).
Differences between capital cities and the balance of Australia are more
evident for the advantaged group than for the deep disadvantage
variable, with 18.5 percent of older people living in capital cities
falling into this advantaged group, compared with only 12.8 percent of
those living outside capital cities. Men are also more likely to fall
into our relative advantage group than women, and the presence of at
least one working person in the household is also strongly associated
with relative economic advantage.
Small Area Analysis Using Spatial Microsimulation Data
We first calculate the proportions of older people aged 55 years
and over who fall into our deeply economically disadvantaged group as
well as the relatively economically advantaged group for each small
area. Our results show that the proportions of elder disadvantage at
small area level range from 0.40 percent to 36.80 percent and from 3.40
percent to 63.10 percent for the proportions of elder advantage. This
demonstrates the much larger spatial disparity within small areas in
contrast to the national averages of 6.6 percent for the disadvantaged
group and 16.5 percent for the advantaged group.
Using a natural breaks classification (a common statistical method
used to display geographic data into natural groups), we have divided
all the small areas into 5 groups based on their concentration rate of
elder disadvantage or advantage. It should be noted that all the small
areas have both disadvantaged and advantaged older people. The most
disadvantaged (or advantaged) groups simply include areas where the
proportion of disadvantaged (or advantaged) older people is highest.
Table 4 presents the numbers of small areas in each group divided
by the range of the concentration rate of deep economic disadvantage
using natural breaks classification. Groups 4 and 5 shown in the last
results column have the highest concentration of elder disadvantage,
ranging from 11.61 percent to 21.00 percent and 21.01 percent to 36.80
percent, with an average rate of 14.87 percent. 64 small areas in these
two groups have been identified as our most disadvantaged areas which
cover 7.05 percent of small areas, 6.11 percent of older people, 14.99
percent of disadvantaged older people and 4.46 percent of advantaged
older people.
The concentrations of disadvantaged older adults are present in
both capital and non-capital city areas, but mainly appearing in the
centre of capital cities (except for Hobart) and some remote areas along
the western coast and in north western NSW. Among the 64 most
disadvantaged small areas, 26 are outside of capital cities and 38 in
capital cities (13 are in Adelaide, 8 in Sydney and 7 in Darwin). These
findings are not unexpected. For example, the outer northern and outer
southern suburbs in Adelaide (such as Hackham and Elizabeth), the
western suburban areas in Sydney (such as Blacktown and Parramatta), the
outer northern and eastern suburbs in Darwin (such as Coconut Grove and
Karama) are the areas with relatively low income.
Table 5 shows the corresponding numbers and proportions in Table 4
but in terms of elder advantage instead of disadvantage. The groups 4
and 5 in Table 5 have the highest concentration of elder advantage,
ranging from 24.11 percent to 33.20 percent and 33.21 percent to 63.10
percent with an average rate of 27.88 percent. 149 small areas in these
two groups have been identified as the most advantaged areas which cover
16.41 percent of small areas, 20.30 percent of older people, 13.03
percent of disadvantaged older people and 33.88 percent of advantaged
older people.
The concentrations of advantage among older people are more likely
to happen in the capital cities but less common outside the capital
cities. Among the 149 most advantaged small areas, 98 are in capital
cities and 51 are outside of capital cities. The advantage
concentrations mainly reflect the overall socioeconomic status of these
areas: for example, Sydney and Melbourne have the largest number of
advantaged small areas (25 in Sydney and 19 in Melbourne), in which the
corridors of suburbs in Melbourne's east and Sydney's northern
and eastern suburbs are areas of generally high income.
How do the concentrations of deep economic disadvantage and
relative economic advantage among older people play out together at a
small area level in Australia? In Table 6, the second and third columns
present the numbers of small areas which fall into either the most
disadvantaged or most advantaged groups, but not both. The fourth column
provides the number of small areas with the highest concentration of
both elder disadvantage and advantage. The fifth column gives the number
of small areas which fall into neither the most disadvantaged nor the
most advantaged groups. Though the concentrations of deep economic
disadvantage and relative economic advantage are both spread across
capital cities and the balance of state, elder disadvantage is more
likely to concentrate in the balance of Australia while elder advantage
is more likely in capital cities. The areas of concentrated elder
advantage are generally different from those of concentrated elder
disadvantage. There are a significant number of small areas falling into
the groups of either the highest disadvantage only or the highest
advantage only (except for Canberra and Hobart where there is no small
area falling into the group with the highest concentration of
disadvantage only). Only two small areas have both the highest
concentrations of elder disadvantage and advantage. These two areas are
North Canberra and Perth Remainder, where 41.2 and 46.8 percent of older
people, respectively, have income in the top two quintiles and also 17
percent and 10.9 percent of older people, respectively, are living in
public housing.
This finding has evidenced a strong clustering of elder
disadvantage and advantage, in both the balance of states and the
majority of capital cities. For example, in Sydney, there are 8 small
areas with the highest concentration of disadvantage only, 25 small
areas with the highest concentration of advantage only, 30 small areas
with neither high concentration of disadvantage nor high concentration
of advantage, and zero small areas with high concentration of both
disadvantage and advantage.
Further assessing the co-occurrence of elder disadvantage and
advantage concentrations is critical for understanding the extent of
homogeneity or diversity within an area, and for planning both public
and private service provision. Figure 2 presents a national map with the
eight capital cities as insets using the same four categories as in
Table 6. On the maps, the dark blue indicates the areas with the highest
concentration of disadvantage only, the yellow shows the areas with the
highest concentration of advantage only, the light blue displays the
areas with the highest concentration of both disadvantage and advantage,
and the green is for other areas which fall into neither the most
disadvantaged or the most advantaged groups.
[FIGURE 2 OMITTED]
This map provides further evidence of the observed clustering of
elder disadvantage and advantage in Table 6. When looking at the map,
the clustering of elder disadvantage and advantage is clear and present
in both capital city and non-capital city areas, but it is more marked
in Sydney and Adelaide. There are certain areas falling into the groups
of high disadvantage only (dark blue) and high advantage only (yellow),
with very few areas in the group with both high disadvantage and
advantage (lighter blue). The majority of areas are in the groups with
neither high disadvantage nor high advantage (green). The small areas
with the highest concentration of deep economic disadvantage only, are
mainly located in the inner western suburbs of Sydney and the western
suburbs of Adelaide, as well as a few areas in the centre of Melbourne,
Perth and Brisbane. The small areas with the highest concentration of
relative economic advantage only, are pronounced in Canberra and Sydney,
and are also evident in the east of Melbourne and Adelaide as well as
the west of Brisbane and Perth. The only two areas with both the highest
concentration of disadvantage and advantage are in Canberra and Perth,
as mentioned above.
Contrasting Our Measures with a General Measure of Area-level
Socio-economic Wellbeing
Spatial concentrations of disadvantage are generally different for
children, working age people and older people (Tanton et al, 2012). In
order to check whether the spatial distribution of deep economic
disadvantage and relative economic advantage among older Australians
that we estimated mirror the overall socio-economic wellbeing of an area
for the wider population, we compare our measures with the
Socio-Economic Indexes for Areas (SEIFA) 2006 (ABS 2008). The SEIFA
index was derived by the ABS based on the characteristics of the
residential areas of 2006 Census respondents, in which, the Index of
Relative Socio-Economic Advantage/Disadvantage (IRSEAD) was chosen for
this comparison as it includes measures of both relative advantage and
disadvantage (Australian Bureau of Statistics 2008).
It is found that our deep disadvantage variable has a negative
correlation (-0.23) while the advantage variable has a positive
correlation (0.68) with the SEIFA IRSEAD index. The sign of the
correlations indicates that more advantaged areas (with higher SEIFA
scores) have lower proportions of deeply disadvantaged and higher
proportions of relatively advantaged older people on average as would be
expected. The moderately strong correlation between our relative
advantage variable and the IRSEAD index shows that areas which fall into
our most 'relatively advantaged' group are quite likely to
also fall into the most advantaged small areas as measured by the SEIFA
IRSEAD index. On the other hand, the weaker correlation coefficient
between our relative disadvantage variable and the IRSEAD index shows
that those areas which have the highest proportion of older people who
are 'deeply economically disadvantaged' are more likely to be
spread out across both advantaged and disadvantaged areas using the
IRSEAD index. This is further evidenced by the transition matrix in
Table 7 which shows how the percentage of deeply disadvantaged older
people and the IRSEAD index match up at each quintile level. In total,
there are only 31.5 percent of older people falling into the same
area-level quintile shown by the IRSEAD index. The differences are
spread right across the distribution of small areas - although, once
again, the greatest agreement (9 percent) between the two measures
occurs in the most advantaged/least disadvantaged quintile. However,
there are 2.4 percent of older people living in 16 small areas which
appear the most advantaged according to IRSEAD index while have the
highest concentrations of deep economic disadvantage as measured by our
variable. Among these 16 areas, 4 are in Sydney, 6 in Melbourne, 2 in
Adelaide, 3 in Perth and 1 in Queensland. These areas have relatively
higher proportions of older people with income in the top two quintiles
versus higher proportions of older people paying public or private rent.
While our measures of advantage and disadvantage are constructed
differently from the SEIFA index, the findings above indicate that the
widely used SEIFA index may not always be able to provide an indication
of the circumstances of older individuals or other specific
subpopulations within small areas.
5. CONCLUSION AND FUTURE DIRECTIONS
The spatial microsimulation approach makes it possible to address
spatial questions about variables that are not available in the Census
but are available in their 'regionalised' sample survey data -
with such questions often being of great interest to policy makers in
terms of spatial inequalities and targeting services. In this article,
the spatial microsimulation data allow us to more accurately estimate
for small areas the populations of highly vulnerable older people as
defined by multiple indicators of advantage and disadvantage in terms of
income levels, income source and housing costs. This information is of
crucial importance in identifying target groups for addressing spatial
aspects of age, economic wellbeing, and social exclusion. The estimates
on advantaged older populations can be used for a wide variety of
purposes such as marketing strategies for retirement communities and
later life leisure and financial products. For example, the increasing
number of economically advantaged baby boomers might drive higher demand
and different expectations on retirement villages.
Our findings reveal substantial heterogeneity and strong clustering
of multiple economic disadvantage and advantage nationally and even more
so at a small area level. Although capital city areas were more likely
to contain higher proportions of relatively economically advantaged
older people, the picture of both elder disadvantage and advantage was
mixed in both capital cities and the balance of states. The presence of
substantial concentrations of older, low-income rent payers in some
capital city areas is particularly concerning due to the high and
increasing rents in many of Australia's urban areas. The
disparities between the most and least affluent older Australians are
expected to accelerate in the future as increasing proportions of baby
boomers bring to later life more superannuation benefits, while
conversely, those poorer baby boomers who do not own homes also are
expected to increase (Yates et al, 2008). The impact of the global
financial crisis has raised further questions on older people's
economic security (Kendig et al., 2013) while health and aged care
reforms now underway raise important matters concerning the regional
organisation and delivery of care. These changes underscore the
importance of understanding spatial heterogeneities and inequalities
among older people across Australia.
More fundamentally, this small area analysis can shed light on how
housing markets and urban development may underlie spatial dimensions to
socio-economic disadvantage in later life. It would also be possible to
examine how urban and rural changes are influencing opportunities for
different aspects of social inclusion. The spatial disparities are
especially important for vulnerable older people, who can be strongly
affected by local environments and social exclusion yet have limited
options for moving to better locations. The findings potentially can
inform urban planning, service allocations, and social inclusion
policies that could ameliorate the economic and social inequalities
faced by vulnerable people across their life span (Mahjabeen et al.,
2009).
In this paper, we have been focusing on the geographic analysis of
the proportions rather than the absolute numbers. Nevertheless, it
should be noted that both metrics should be considered in terms of using
these data to analyse needs or plan services as they might provide
different pictures. For example, when proportions are used to examine
the geographic distribution, the concentrations of disadvantage can be
more clearly seen in areas outside the capital cities. In contrast the
use of absolute numbers shows that greater numbers of deeply
disadvantaged older people are more likely to live in the capital cities
rather than in the remainder of Australia.
In future work, we can include more domains into our definition of
disadvantage and advantage for older people at small area level, such as
health and health services, productive participation, social activities
and connections and neighbourhood environment. We also could replicate
the microsimulations for 2011 and later Census years in order to
identify and monitor patterns and predictors of change. These
developments would provide the opportunity to further develop an
age-specific measure of small area advantage and disadvantage beyond
economic well-being for older Australians.
ACKNOWLEDGEMENTS: This paper was funded by an Australian Research
Council Discovery Grant DP664429. The authors would like to thank the
fellow Chief Investigators on this grant, Professor Fiona Stanley,
Professor Bob Stimson, Dr Sharon Goldfeld, as well as the Australian
Bureau of Statistics for their inputs to the broader projects under this
grant. The small area data were produced from NATSEM Spatial
Microsimulation model SPATIALMSM/09D based on Census 2006 data. A new
model based on Census 2011 data has not as yet been constructed. This
paper was completed with support from the ARC Centre of Excellence in
Population Ageing Research (CEPAR). A 2001 concordance between SLAs and
Brisbane Local Council Electoral Wards was kindly supplied by the Centre
for Research into Sustainable Urban and Regional Futures (CR-SURF) at
the University of Queensland. This was modified by the authors for use
with 2006 SLAs.
REFERENCES
Abello, A., Gong, C. H., Daly, A. and McNamara, J. (2012). Spatial
Dimensions of Child social Exclusion Risk in Australia: Widening
the Scope. Child Indicator Research, 5(4), pp. 685-703.
ABS (2007). Australian Standard Geographical Classification (ASGC).
ABS catalogue No. 1216.0, Australian Bureau of Statistics.
ABS (2008). Information Paper: An Introduction to Socio-Economic
Indexes for Areas (SEIFA) 2006, ABS catalogue No. 2039.0, Australian
Bureau of Statistics.
Ballas, D. and Clarke, G.P. (2001). Modelling the Local Impacts of
National Social Policies: a Spatial Microsimulation Approach.
Environment and Planning C: Government and Policy, 19, pp. 587-606.
Barnes, M., Blom, A., Cox, K. and Lessof, C. (2006). The Social
Exclusion of Older People: Evidence from the First Wave of the English
Longitudinal Study of Aging (ELSA). United Kingdom, Office of the Deputy
Prime Minister, London.
Baum, S., O'Connor, K. and Stimson, R. (2005). Fault Lines
Exposed: Advantage and Disadvantage across Australia's Settlement
System. Monash University ePress.
Berry, M. (2007) Ageing in Space: Transport, Access and Urban Form.
In A. Borowski, S. Encel, and E. Ozanne (Eds) Longevity and Social
Change in Australia, UNSW Press, Sydney, pp. 239-264.
Chin, S.F. and Harding, A. (2007). SpatialMSM--NATSEM"s Small
Area Household Model of Australia. In A. Harding and A. Gupta (Eds)
Modelling Our Future: Population Ageing, Health and Aged Care,
International Symposia in Economic Theory and Econometrics. North
Holland: Amsterdam.
Chin, S-F., Harding, A., Lloyd, R., McNamara, J., Phillips, B. and
Vu, Q.N. (2005). Spatial Microsimulation Using Synthetic Small-area
Estimates of Income, Tax and Social Security Benefits. Australasian
Journal of Regional Studies, 11(3), pp. 303-344.
Daly, A., McNamara, J., Tanton, R., Harding, A. and Yap, M. (2008).
Indicators of Social Exclusion for Australia's Children: An
Analysis by State and Age Group. Australasian Journal of Regional
Studies, 14(2), pp. 133-154.
Dannefer, D. (2003). Cumulative Advantage/Disadvantage and the Life
Course: Cross-fertilizing and Social Science Theory. Journal of
Gerontology, 58(6), pp. 327-337.
Dannefer, D. and Kelly-Moore, J. (2009). Theorizing the Life
Course: New Twists in the Paths. In V. Bengtson, D. Gans, N. Putney and
M. Silverstein (Eds) Handbook of Theories of Aging, Second Edition,
Springer, New York, pp. 389-412.
Davies, J.S. (2005). The Social Exclusion Debate: Strategies,
Controversies and Dilemmas. Policy Studies, 26(1), pp. 3-27.
Gong, H., McNamara, J., Vidyattama, Y., Miranti, R., Tanton, R.,
Harding, A. and Kendig, H. (2012). Developing Spatial Microsimulation
Estimates of Small Area Advantage and Disadvantage among Older
Australians. Population, Space and Place, 18(5), pp. 551-565, DOI:
10.1002/psp.692.
Harding, A., Vidyattama, Y. and Tanton, R. (2011). Demographic
Change and the Needs-Based Planning of Government Services: Projecting
Small Area Populations Using Spatial Microsimulation. The Journal of
Population Research, 28(2-3), pp. 203-224.
Hayes, A., Gray, M. and Edwards, B. (2008). Social Inclusion:
Origins, Concepts and Key Themes. Paper prepared by the Australian
Institute of Family Studies for the Social Inclusion Unit, Department of
Prime Minister and Cabinet, Online version accessed 18 October 2008,
http://www.socialinclusion. gov.au/publications.htm.
Kelly, S. (2009). Reform of the Australian Retirement Income
System, NATSEM Research Report prepared for The Brotherhood of St.
Laurence.
Kendig, H. and Bridge, C. (2007). Housing Policy for a Long-lived
Society. In A. Borowski, S. Encel and E. Ozanne(Eds) Longevity and
Social Change in Australia. University of New South Wales Press: Sydney,
pp. 219-238.
Kendig, H., Clemson, L. and Mackenzie, L. (2012). Housing,
Neighbourhoods and Ageing in Place. In S. Smith , M. Elsinga , L. Fox,
S. E. Ong and S. Wachter (Eds) International Encyclopaedia of Housing
and Home. Oxford: pp. 150-155.
Kendig, H., Wells, Y. and O'Loughlin, K. (2013). Australian
Baby Boomers Face Retirement during the Global Financial Crisis. Journal
of Aging & Social Policy, 25, pp. 264-280. DOI:
10.1080/08959420.2013.795382.
Kendig, H., and Phillipson, C. (in press). Building Age Friendly
Cities: New Approaches and Evidence. In N. Denison and L. Newby (Eds)
Health Inequalities: Social Determinants at a Local Level. British
Academy Policy Centre.
Lewis, P. and Corliss, M. (2009). Social Exclusion of Children: The
Regional Dimention. Australasian Journal of Regional Studies, 15(2), pp.
195-211.
Lui, C., Warburton, J., Winterton, R. and Bartlett, H. (2011).
Critical Reflections on a Social Inclusion Approach for an Ageing
Australia. Australian Social Work, 64(3), pp. 266-282.
Lymer, S., Brown, L., Yap, M. and Harding, A. (2008). 2001 Regional
Disability Estimates for New South Wales, Australia, Using Spatial
Microsimulation. Applied Spatial Analysis and Policy, 1(2), pp. 99116.
Mahjabeen, Z., Shrestha, K. and Dee, J. (2009). Rethinking
Community Participation in Urban Planning: The Role of Disadvantaged
Groups in Sydney Metropolitan Strategy. Australasian Journal of Regional
Studies, 15(1), pp. 45-63.
Marmot, M., Allen, J., Goldblatt, P., Boyce T., McNeish, D., Grady,
M. and Geddes, I. (2010). Fair Society, Healthy Lives. The Marmot Review
Report. UK.
McLachlan, R., Gilfillan, G. and Gordon, J. (2013). Deep and
Persistent Disadvantage in Australia. Productivity Commission Staff
Working Paper, Canberra.
Miranti, R., Payne, A. and McNamara, J. (2010). Towards Small Area
Indicators of Well-being for Older Australians: Concepts and Issues.
NATSEM Online Discussion Paper, DP67.
Miranti, R. and Yu, P. (2011). Persistence of Social Exclusion
among Older People in Australia: What Are the Protecting Factors? NATSEM
Working Paper, WP11, presented at HILDA Survey 2011 Conference.
Melbourne, 14-15 July 2011.
Naughtin, G. (2008). Social Inclusion and Older People. Brotherhood
of St Laurence, Proceedings of Social Inclusion Down Under Symposium.
O'Rand, A. (2006). Stratification and the Life Course: Life
Course Capital, Life Course Risks, and Social Inequality. In R. H.
Binstock and L. K. George (Eds) Handbook of Aging and the Social
Sciences (sixth edition).
Phillipson, C. (2010). Ageing and Urban Society: Growing Old in the
'Century of the City. In D. Dannefer and C. Phillipson (Eds) The
SAGE Handbook of Social Gerontology, London, Sage Publications: pp.
597-606.
Pynoos, J., Kendig, H. and Nishita, C. (2007). Housing. In J.
Birren (Eds) Encyclopedia of Gerontology (second edition), Elsevier, pp.
70919.
Saunders, P. and Bradbury, B. (2006). Monitoring Trends in Poverty
and Income Distribution: Data, Methodology and Measurement. The Economic
Record, 82, pp.341-364.
Scharf, T., Phillipson, C. and Smith A.E. (2005). Multiple
Exclusion and Quality of Life amongst Excluded Older People in
Disadvantaged Neighbourhoods. Social Exclusion Unit, Office of the
Deputy Prime Minister: London. March 2005.
Tanton, R., Vidyattama, Y., McNamara, J., Vu, Q.N. and Harding, A.
(2009). Old, Single and Poor: Using Microsimulation and Microdata to
Analyse Poverty and the Impact of Policy Change among Older Australians.
Economic Papers, 28, pp. 102-120.
Tanton, R., Vidyattama, Y., Nepal, B. and McNamara, J. (2011).
Small Area Estimation Using a Reweighting Algorithm. Journal of the
Royal Statistical Society: Series A, 174(4). DOI:
10.1111/j.1467-985X.2011.00690.x.
Tanton, R., Gong, C.H. and Harding, A. (2012). Evidence of Multiple
Life Stage Disadvantage for Small Areas in Australia. Economic Papers,
31(1), pp. 76-86.
Tanton, R. and Phillips, B. (2013). A Measure of the Depth of
Housing Stress and its Application in Australia. Economic Papers, 32(1),
pp. 99-109.
UK Social Exclusion Unit Office of the Deputy Prime Minister
(2006). A Sure Start to Later Life Ending Inequalities for Older People.
A Social Exclusion Unit Final Report. Office of the Deputy Prime
Minister: London.
Vidyattama, Y., Miranti, R., McNamara, J.,Tanton, R. and Harding,
A. (2013). The Challenges of Combining Two Databases in Small Area
Estimation: an Example Using Spatial Microsimulation of Child Poverty,
Environment and Planning A, 45(2), 344-361.
Vidyattama, Y. and Tanton, R. (2010). Projecting Small Areas
Statistics with Australia Spatial Microsimulation Model (SPATIALMSM).
Australasian Journal of Regional Studies, 16(1), pp. 99-126.
Warnes, T. (2010). Migration and Age. In D. Dannefer and C.
Phillipson (Eds) The SAGE Handbook of Social Gerontology, London, SAGE
Publications: pp. 389-404.
Yates, J., Kendig, H., Phillips, B., Milligan, V. and Tanton, R.
(2008). Sustaining Fair Shares: The Australian Housing System and
Intergenerational Sustainability. Final Report for the Australian
Housing and Urban Research Institute, Sydney Research Centre,
RMIT-NATSEM Research Centre. AHURI Final Report, No. 111.
Cathy Gong
Research Fellow, Centre for Research on Ageing, Health and
Wellbeing (CRAHW) and ARC Centre of Excellence in Population Ageing
Research (CEPAR), Australian National University (ANU), Canberra, ACT,
0200, Australia. Email: cathy.gong@anu.edu.au.
Hal Kendig
Professor, Centre for Research on Ageing, Health and Wellbeing
(CRAHW) and ARC Centre of Excellence of Population Ageing Research
(CEPAR), Australian National University (ANU), Canberra, ACT, 0200,
Australia.
Ann Harding
Professor, National Centre for Social and Economic Modelling
(NATSEM), Institute for Governance and Policy Analysis (IGPA),
University of Canberra (UC), Canberra, ACT, 2601, Australia.
Riyana Miranti
Senior Research Fellow, National Centre for Social and Economic
Modelling (NATSEM), Institute for Governance and Policy Analysis (IGPA),
University of Canberra (UC), Canberra, ACT, 2601, Australia.
Justine McNamara
Adjunct Associate Professor, National Centre for Social and
Economic Modelling (NATSEM), Institute for Governance and Policy
Analysis (IGPA), University of Canberra (UC), Canberra, ACT, 2601,
Australia.
Table 1. Validation for Spatial Microsimulation Estimates against
ABS Census Data.
SEI for Pearson R SEI for Pearson
proportion for numbers for numbers
proportion
(1) In bottom
income quintile 0.99 0.98 0.99 1.00
(2) In top two
income quintiles 0.97 0.99 0.99 1.00
(3) Public renters 0.95 0.98 0.98 0.99
(4) Private renters 0.84 0.90 0.98 0.99
(5) Paying no rent
or mortgage 0.95 0.93 0.98 1.00
(6) Paying
public/private rent
and in bottom
quintile 0.92 0.95 0.99 0.99
(7) Not paying
rent/mortgage and
in top two quintiles 0.97 0.98 1.00 1.00
Note: The income quintiles used in this table are national quintiles
of household equivalised income. Data source: SPATIALMSM/09D; Census
2006.
Table 2. Economic Disadvantage Variables for Persons Aged 55+,
Distribution by Population Characteristics, Australia 2006.
Bottom Main
Deep equivalised Paying Paying income
economic income private public (govern
disadvantage quintile rent rent ment)
Characteri
stics % % % % %
(1) All
persons
55+ 6.6 36.0 8.3 4.2 45.6
(2) Female 7.5 40.2 8.0 4.7 50.9
(3) Males 5.7 31.5 8.7 3.6 39.9
(4) Living
alone 15.7 67.3 11.2 7.8 77.8
(5) Living
in
household
where
anyone
working 0.8 9.0 8.3 1.9 8.8
(6) Living
in a capital
city 7.0 33.8 8.4 4.7 41.0
(7) Not
living in a
capital
city 6.0 40.2 8.4 3.1 53.8
Notes: (1) The 'deep economic disadvantage' variable refers to people
aged 55 and over, in the national bottom quintile of equivalised
disposable household income, with main income source from government
benefits, and paying private/public rent. (2) As our definition of
older people is person- based, some of the older people included in
our analysis will be living in households with people younger than
55. Data source: ABS Survey of Income and Housing 2005/06.
Table 3. Economic Advantage Variables for Persons Aged 55+,
Distribution by Population Characteristics, Australia 2006.
Relative Top two Not paying Main
economic income rent or income
advantage quintiles mortgage (private)
Characteristics % % % %
(1) All persons 55+ 16.5 26.2 69.4 54.4
Females 55+ 14.7 22.5 70.7 49.1
Males 55+ 18.4 30.2 67.9 60.1
(2) 55+ living
alone 6.5 7.3 75.9 22.2
(3) 55+ living in
household where
anyone working 28.8 49.7 57.5 92.0
(4) 55+ living in
a capital city 18.5 29.5 67.6 59.0
(5) 55+ not living
in a capital city 12.8 19.9 72.5 46.2
Note: The 'relative economic advantage' variable refers to people
aged 55 and over, in the national top two quintiles of equivalised
disposable household income, with main income source from private
income, and not paying private/public rent or mortgage. Data source:
ABS Survey of Income and Housing 2005/06.
Table 4. Proportions of Small Areas and Older Population by the
Concentration Rate of Disadvantaged People Aged 55+.
The range of concentration rate of disadvantaged
older people
11.61%-
0.40%- 3.81%- 7.11%- 21.00% and
3.80% 7.10% 11.60% 21.01%-
36.80%
Groups 4 & 5
Group 1 Group 2 Group 3 (Most
disadvantaged)
(1) Average
concentration
rate of
disadvantaged
older people
(%) 2.53 5.41 8.89 14.87
(2) #of SLAs 334 325 185 64
Note: "Disadvantaged older people" are defined as older people in the
bottom income quintile, paying rent and mainly relying on government
income. Data source: SPATIALMSM/09D.
Table 5. Proportion of Small Areas and Older Population by the
Concentration Rate of Advantaged People Aged 55+.
The range of concentration rate of advantaged older
people
3.40%- 13.31%- 18.01%- 24.11%-33.20%
13.30% 18.00% 24.10% & 33.21%-63.10%
Groups 4 & 5
Group 1 Group 2 Group 3 (most advantaged)
(1)Average
concentration
rate of 11.39 15.38 20.46 27.88
advantaged
older people
(%)
(2)# of SLAs 190 343 226 149
Note: "Advantaged older people" are defined as older people in the
top two income quintiles, not paying rent or mortgage and mainly
relying on private income. Data source: SPATIALMSM/09D.
Table 6. Number of Small Areas with the Highest Concentration of
Disadvantage and Advantage among People 55+.
Small Small areas
Small areas areas with with high
with high high disadvantage Other Total
disadvantage advantage and small small
only only advantage areas areas
Balance 26 51 0 535 612
Capital
cities 36 96 2 162 296
Sydney 8 25 0 30 63
Melbourne 4 19 0 54 77
Brisbane 2 10 0 21 33
Adelaide 13 12 0 29 54
Perth 2 13 1 19 35
Hobart 0 1 0 5 6
Darwin 7 10 0 4 21
Canberra 0 6 1 0 7
All 62 147 2 697 908
Notes: (1) "High disadvantage" is defined as having the highest
proportion of older people in the bottom income quintile, paying rent
and mainly relying on government income, as shown in the last column
in Table 4. (2) "High advantage" is defined as having the highest
proportion of older people in the top two income quintiles, not
paying rent or mortgage and mainly relying on private income, as
shown in the last column in Table 5. Data source: SPATIALMSM/09D.
Table 7. Quintiles for Aged 55+ at SLAs by the Percent of Deeply
Disadvantaged Older People and SEIFA IRSEAD Index.
Quintile for Quintile for aged 55+ at SLAs by percent of
aged 55+ by disadvantaged older people
SEIFA 5 (Most 4 3 2 1 (Least
IRSEAD index disadvantaged) disadvantaged)
1 6.8 3.5 5.4 2.4 1.9
(Disadvantage)
2 5.00 5.8 3.7 3.7 1.8
3 3.9 6.7 4.2 2.9 2.3
4 2.1 2.8 4.8 5.7 4.7
5 (Advantage) 2.4 1.2 1.9 5.5 9.0
Note: "Disadvantaged older people" are defined as older people in the
bottom income quintile, paying rent and mainly relying on government
income. Data source: SPATIALMSM/09D.
