Demarcating functional economic regions across Australia differentiated by work participation categories.
Stimson, Robert ; Mitchell, William ; Flanagan, Michael 等
1. INTRODUCTION
For a long time regional scientists have been investigating spatial
variations in regional economic development/performance using spatial
econometric modelling to help identify factors that might explain that
variability (for example, Molho, 1995; Niebuhr, 2003; Patacchini and
Zenou, 2007; Mitchell and Flanagan, 2016). Invariably such
investigations are dependent on using aggregated data that is usually
readily available for de jure regions, that are usually designed for
administrative purposes, to cater to local authorities in historically
defined regions, for example Local Government Areas (for example,
Randolph and Holloway, 2005). This extends from the early regional
science literature where one of the factors considered when delineating
regions was that they be designed to match political or administrative
boundaries (see, for example, Berry, 1968; and Richardson, 1973). These
administrative regions though, while useful to the local authorities, do
not reflect the underlying processes that generate economic data and, as
a result, we encounter the modifiable areal unit problem (MAUP)
(Openshaw, 1984). The MAUP proposes there are literally thousands of
ways small areal units can be aggregated to regions to provide different
regionalisations and hence different data, varying in terms of the size
of regions (scale problem) and the grounds on which areal units are
joined (aggregation problem). A manifestation of the MAUP is the
occurrence of spatial autocorrelation, where regions are spatially
dependent on regions they are near to. Positive (negative) spatial
autocorrelation is where regions with similar (dissimilar) attributes
are spatially proximate, which invalidates the OLS assumptions and
requires the analyst to employ spatial econometric tools (Anselin, 1988,
LeSage and Pace, 2009).
Ideally such modelling would use functional regions as the spatial
base which, theoretically, should overcome this problem. In their
investigation of spatial variations in regional endogenous employment
performance over the decade 1996-2006, Stimson et al. (2011) showed that
when spatial econometric modelling is conducted using functional
economic regions (FERs) rather than de jure regions (Local Government
Areas) as the spatial base for modelling, then the spatial
autocorrelation encountered when using de jure regions might be
overcome.
In contrast, Patacchini and Zenou (2007: 170) found that
Travel-To-Work areas (TTWAs) in the UK, regions "which are by
definition self-contained labour markets", do not eradicate the
spatial dependence of unemployment. They conclude though that the
spatial dependence that occurs is mainly due to spatial spillovers,
where workers can search for and work in different TTWAs, and therefore
was due to the commuting that occurs between different TTWAs because
they are in fact not completely self-contained. TTWAs are designed using
a variation of the algorithm first proposed by Coombes et al. (1986) and
used widely throughout the literature (for example, Andersen, 2002;
Watts et al., 2006; and Casado-Diaz et al., 2010). In the new
statistical geography introduced by the ABS at the 2011 Census, labour
markets were a key consideration to the design of the Statistical Areas
Level 4 (SA4s), in an attempt to incorporate functional regions (ABS,
2010). As these are the units of dissemination for the Labour Force
Survey, their self-imposed constraints (for example minimum populations)
somewhat inhibit the regions design (Watts, 2013).
Thus, we are now focusing our modelling of regional economic
performance in Australia on using FERs as the spatial base, and we are
deriving those FERs using joumey-to-work (JTW) commuting flows data that
is available in the Australian census. In doing so we employ the
Intramax procedure developed by Masser and Brown (1975).
In this paper we report on how, through analysis being conducted at
the Centre of Full Employment and Equity (CofFEE) at the University of
Newcastle in Australia, FERs have been derived using the 2011 census JTW
data. Our intent is to use FERs as the spatial units for modelling the
determinants of spatial variations in the performance of regional labour
markets over the decade 2001 to 2011, particularly with regard to
unemployment and employment growth. Those FERs are designated by us as
CofFEE Functional Economic Regions (CFERs).
But we go further than deriving FERs that just relate to aggregate
employment across all industry sectors as it is well known that spatial
patterns in the degree of spatial concentration or dispersal of jobs
differ between occupation categories (Bill et al., 2008; Sang et al.,
2011). In addition, it is likely that the spatial locations of male and
female jobs may also differ, as might the spatial patterns of commuting
to jobs according to the mode of transport for the JTW (Crane, 2007;
BITRE, 2015). To address those issues we have thus developed a series of
10 regionalisations of CFERs across Australia for 2011 as specified in
Table 1. Each of those regionalisations is designed using the JTW
commuting flows of their respective cohort of worker categories as shown
in the table. The implied homogeneity of functional regions was a
shortcoming raised by Morrison (2005), and in this was we are
recognising a differentiated labour force.
In this paper we first outline the methodology and the data used to
derive those ten regionalisations of CFERs. We then proceed to discuss
the number of CFERs across Australia that are thus derived, providing a
comparison with Labour Force Regions (LFRs) used by the Australian
Bureau of Statistics (ABS). We then proceed to briefly discuss the
spatial characteristics of the CFERS that have been derived for the 10
regionalisations listed in Table 1.
2. METHOD
The Intramax Procedure
The Intramax procedure (after Masser and Brown, 1975) is used to
derive CFERs for all 10 the CFER regionalisations. The procedure
considers the size of the interactions in the JTW commuting flows matrix
that are in the form of a contingency table. It then formulates the
"objective function in terms of the differences between the
observed and the expected probabilities that are associated with these
marginal totals" (p. 510). A schematic representation of the square
JTW flows matrix is shown in Table 2, where the rows are designated as
origins and the columns are destinations.
If we view Table 2 as a contingency table, then the expected values
of each element are derived as the product of the relevant column sum
(Equation 3 below) times the ratio of the row sum (Equation 2) to total
interaction (Equation 4). For example, the expected flow out of region 2
into region 1, [a.sub.21] in Table 2, where [a.sub.ij] is the element in
row i and column j of the contingency table (JTW matrix), is given as:
(1) [a.sup.*.sub.21] =
[[summation].sub.i][a.sub.i1]([[summation].sub.j][a.sub.2j]/
[[summation].sub.i][[summation].sub.j][a.sub.ij]) =
[[summation].sub.i][a.sub.i1]([[summation].sub.j][a.sub.2j]/n)
This is the "flow that would have been expected simply on the
basis of the size of the row and column marginal totals" (Masser
and Brown, 1975: p. 512).
The row sum of the JTW matrix is:
(2) [a.sub.i*] = [[summation].sub.j][a.sub.ij]
The column sum of the JTW matrix is:
(3) [a.sub.j*] = [[summation].sub.i][a.sub.ij]
The total interaction, n, is the sum of the row sums:
(4) n = [[summation].sub.i][[summation].sub.j][a.sub.ij]
The null hypothesis for independence between the row and column
marginal totals of a contingency table is defined as:
(5) [H.sub.o]: [a.sup.*.sub.ij] =
([[summation].sub.j][a.sub.ij][[summation].sub.i][a.sub.ij])/n =
([a.sub.i*][a.sub.j*])/n
The objective function of the hierarchical clustering algorithm,
using a non-symmetrical JTW matrix, is defined as:
(6) max I = ([a.sub.ij] - [a.sup.*.sub.ij]) + ([a.sub.ji] -
[a.sup.*.sub.ji]), i [not equal to] j
In the Flowmap software, which we used to perform the Intramax
procedure for the CFERs, Equation (6) is modified as follows
(Breukelman, et al., 2009):
(7) max I = [[T.sub.ij]/[O.sub.i][D.sub.j]] +
[[T.sub.ji]/[D.sub.j][O.sub.i]], i [not equal to] j
where:
[T.sub.ij] is the interaction between the origin SA2 i and
destination SA2 j;
[O.sub.i] is the sum of all flows starting from origin i; and
[D.sub.j] is the sum of all flows ending at destination j.
This alters the focus from the absolute difference between the
observed and expected flows to the proportional difference.
At each stage of the clustering process, fusion occurs between the
regions that have the strongest commuting ties (interaction), as
represented by Equation (7). The stepwise procedure then combines the
clustered interaction, and the matrix is reduced by a column and a row.
The remaining actual and expected commuting flows are re-calculated and
the i,j combination of regions maximising (Equation 7) is again
calculated, and so-on. If there is a continuous network of flows across
the study area, with N regions, after N-1 steps, all regions would be
clustered into a single area, and by construction, all interaction would
be intrazonal with one matrix element remaining.
To render the concept of functional regions operational, some level
of clustering (number of steps) has to be chosen and the resulting
regionalisation defined. There are two main approaches to deciding how
and when to stop the clustering process:
1. The first is by reference to intra-regional flows, where the
user may stop the clustering process when a certain percentage of flows
are intra-regional, or where there is a large increase in the
intra-regional flows.
2. Alternatively, the user may want to stop the Intramax method
when a pre-determined number of regions has been formed. We stop the
clustering for the regions around the 75 percent mark.
The Data Used
Data from the ABS's 2011 census was used to design the CFERs
employing the ABS TableBuilder product. The spatial area building blocks
we use to derive the CFERs are the SA2 units within the hierarchy of the
new Australian Statistical Geography Standard (ASGS) that was introduced
for the first time in the 2011 census. Those SA2s tend to equate to
suburbs within the metropolitan cities and larger regional cities and to
towns and surrounding areas in regional Australia. It is the SA2s that
are used by the ABS as the origin and destination zones for reporting
commuting flows for JTW data in the 2011 census.
In the case of all of the CFER regionalisations we have derived, a
commuting flow matrix was designed listing the flows between all
possible SA2s, which are local areas that basically equate to suburbs
and towns.
The JTW data from the 2011 census has two notable limitations:
1. First, the ABS has strict rules on confidentialising its data
for the purpose of making it impossible to identify a particular person,
which does provide some limitation to the data's accuracy at small
numbers. For small numbers the ABS randomises the data and the smallest
flow is a value of 3.
2. The second limitation is a result of the different reference
periods for the questions asked in the Census. While our origin SA2 is
the usual address of workers (at which they will have lived for 6 months
or more in 2011), the workplace address is taken for the main job held
in the week prior to the date of the census count. To address this
limitation we enforce a one-way threshold commuting distance of 300km,
above which the flow is excluded from the dataset, so as to exclude
flows where it is obvious a person was not carrying out a daily commute.
The distance of a commute was taken as the distance between the
population-weighted centroids of the origin and destination SA2s.
Using the Flowmap Software
In using the Flowmap software to run the Intramax procedure, there
is a requirement that all spatial areas (that is, the 2011 census SA2s)
used in the calculation be interactive. That interactivity is defined as
an SA2 being required to have both resident workers and workplaces, and
at least one of these must interact with another SA2. Hence, prior to
running the Intramax, we needed to remove SA2s that were
non-interactive.
When we included flows from all workers there were 25 SA2s across
Australia with no flows, plus another 11 SA2s with only an intra-zonal
flow. In addition there were 38 SA2s that had inflows but no outflows,
and there were two SA2s with outflows but no inflows. SA2s with only an
intra-zonal flow represent self-contained labour markets, and are given
the same status as regions that are formed through the Intramax process.
SA2s with only one direction flow were placed into regions after the
Intramax procedure completed. SA2s with no flows were removed and are
classified separately.
For the JTW 'mode of transport' regionalisations there
were many SA2s with flows in just one direction. As these flows were
important in the design of the CFERs (as opposed to the others where
their number was very small), an intra-zonal flow of 1 was added to
those SA2s so they became interactive and remained part of the flow
dataset utilised in the Intramax procedure.
Dividing Australia Into Large Regions
There was substantial commuting between towns on either side of
State and Territory boundaries in Queensland and NSW, NSW and ACT, NSW
and Victoria, and Victoria and SA. As there were negligible JTW
commuting flows between the other States and Territories, we divided up
Australia into the following four large regions:
1. East Coast plus South Australia (EC+SA), consisting of these
states/territories:
* New South Wales
* Australian Capital Territory.
* Victoria
* Queensland
* South Australia.
2. Western Australia (WA).
3. Tasmania (TAS).
4. Northern Territory (NT).
The Intramax procedure was run separately for each these large
regions.
3. OVERVIEW OF RESULTS DERIVED FROM THE INTRAMAX PROCEDURE TO
PRODUCE CFERS
Australia is a very large country and its relatively small
population of around 24 million is highly concentrated geographically,
with almost seven out of ten people living in just five large capital
city metropolitan regions (Sydney, Melbourne, Brisbane, Perth and
Adelaide) whose populations range from a little over one to almost five
million. Those capital city metropolitan areas are 'primate
cities' in their respective States, and there is certainly not a
well-developed urban hierarchy--as per Zipfs (1949) 'rank size
rule'--across Australia's urban settlement system. The vast
bulk of the nation's settlement is located within a few hundred
kilometres of the east, south-east and south-west coasts of the
Australian continent. The interior of the country is very sparsely
settled with extremely low population densities, with much of that
settlement occurring in remote small indigenous communities. Outside the
large state capital city metropolitan regions there are just a handful
of urban centres with populations over 100 000, and only one with more
than 500 000. The large majority of urban centres outside the
metropolitan city areas (in what is commonly referred to as rural and
regional Australia) tend to be small.
It might be expected that within the large capital city
metropolitan areas there would be a number (probably relatively small)
of FERs. And it might be expected that there would be a relatively large
number of FERs beyond the capital cities across the vast expanses of
rural and regional Australia, with a number of FERs focusing on the
larger regional cities and towns often incorporating smaller urban
centres in the surrounding hinterlands, and with the FERS in the more
sparsely settled interior areas being very large geographically.
The CFERs and Regions for Various Aggregations of the Australian
Statistical Geography Standard (ASGS)
Table 3 shows the number of regions across the four large regions
into which we have divided Australia that are produced by the Intramax
procedure for the Original CFERs (that is, based on the JTW commuting
flows data for all workers across all industry categories). The table
also indicates the number of areas in the various ASGS classifications
that are included in the CFERs.
Across the five states/territories comprising the EC + SA large
region, the Intramax procedure produced 79 interactive CFERs, with 5
non-interactive SA2s that are self-contained labour markets (SCLMs).
Associated with those CFERs there are 72 SA4s located across the same EC
+ SA large region at the 2011 census. For the Western Australia large
region there are 18 CFERs, with 4 SCLMs, and 9 SA4s; for Tasmania there
are 12 CFERs, with no SCLMs and 4 SA4s; and for the Northern Territory
there are 14 CFERs, with 2 SCLMs, and 2 SA4s.
Of particular interest is the comparison between the number of
CFERs and the regions at which the ABS disseminates data collected
through its Labour Force Survey. Previously, under the old Australian
Standard Geographical Classification (ASGC) used prior to the 2011
census, this data was made available for the ABS Labour Force Regions
(LFRs). However, since the introduction of the new national
geography--the Australian Statistical Geography Standard (ASGS)--at the
2011 census, this is now at the SA4 level of the national geography.
As shown in Table 3, there are more of the Original CFERs than
there are SA4s; however, there are more SA4s than there were ABS LFRs,
partly due to the fact that "labour markets were a key
consideration in (their) design" (ABS 2010, p. 27). The SA4s must
be large enough to accommodate the ABS sample sizes for its surveys
without giving results with standard errors that are too large to make
the data meaningful; and the SA4s must also aggregate to capital
city/rest of state and state/territory borders. Importantly, those
requirements are not placed on our CFERs, whose boundaries are permitted
to cross those borders if the JTW commuting flows data are such that
they in fact cross those administrative boundaries. This is the big
advantage in using functional regions as against de jure regions as the
spatial data base for the analysis of regional labour markets.
Regions for the Various CFER Aggregations
Turning now to consider the 10 regionalisations for which CFERs
have been derived using the Intramax procedure, Table 4 lists the number
of regions corresponding to each of the CFER aggregations. In each case
we began by removing the Other Territories and Lord Howe Island, which
left us with 2 192 SA2s across Australia.
In Table 4 there are three columns for all four of the large
regions into which we divided Australia:
* Column (a) is the number of interactive CFERs produced using the
Intramax procedure. There are 123 such interactive Original CFERs across
Australia.
* Column (b) is the number of SA2s that are non-interactive, which
we call Self-Contained Labour Markets (SCLMs). There are 11 SCLMs across
Australia, almost all being located in the EC + SA and the WA large
regions.
* Column (c) is the number of SA2s that have no flows, for all
workers, or for a particular gender, broad occupation class, or mode of
commuting. There are 25 of those across Australia, predominately in the
EC + SA large region with a few in the WA large region.
The SA2s that are SCLMs are considered analogous to interactive
CFERs as they represent a labour market and therefore should be included
in any analysis. However, SA2s with no JTW commute flows should be
excluded from any analysis.
In total, across Australia there are 159 Original CFERs, of which
123 are interactive CFERs, 11 are SA2s that are SCLMs, and 25 are SA2s
that have no JTW commuting flows.
From Table 4 it is evident that there is a substantially larger
number of Female CFERs than there are Male CFERs across Australia, and
this is also the case for Less Skilled CFERs and for Trades CFERs as
against Skilled CFERs, while the number of Less Skilled CFERs and Trades
CFERs are about the same.
With relation to JTW mode of travel, because the JTW is dominated
by road transport there is a large number of Road JTW CFERs across
Australia, with that number being similar to the number of Original
CFERs. The number of Multiple Transport Mode JTW CFERs is somewhat less
numerous. Not unexpectedly there are many more Bicycle JTW CFERs than
there are for the other modes of commuting. And the number of Rail JTW
CFERs is very small. But with respect to the latter, it needs to be
noted that Rail JTW CFERs exist only in and around the capital cities of
Sydney, Melbourne, Brisbane, Adelaide and Perth. In some cases the Rail
JTW CFERs do extend outside the ABS defined area of the capital cities,
however, in general there were very few workers from outside these areas
that indicated they used the rail network for their mode of commuting.
Hence, we excluded any flows from outside the areas that make up the
Rail JTW CFERs.
As may also be seen from Table 4, some of the aggregations have
many SA2s that are SCLMs, and also many SA2s with no flows. This
reflects the type travel mode that is being used for the JTW commute.
For example, in the case of the Bicycle JTW CFERs there are naturally
many non-interactive SA2s as, generally, the distance someone would ride
a bike is small and it is quite likely that bike riders do not leave
their SA2 of origin. In addition, the Bicycle JTW CFERs also have many
SA2s with no flows, reflecting the lack of popularity of using a bike to
commute to work in those areas.
4. CONVENTIONS FOR NAMING THE CFERS
For all of the 10 regionalisations we followed the same naming
conventions. Each unique area has an area name, whether it is an
interactive CFER, an SA2 with only an intra-zonal flow (a SCLM), or an
SA2 without any flows. For those that were classified as CFERs their
name attempts to explain where they are placed in Australia. If a CFER
crossed a state/territory boundary we included the name of at least one
area from each state in the CFER name to indicate this, except in the
case of the ACT which in most cases was part of a CFER that included
surrounding towns in NSW, where the name for the CFER is ACT and
surrounds. If a CFER was a single SA2 it took on the name of the SA2.
SCLMs also took on the name of their SA2, as did those without flows.
Each area also has a corresponding area code. CFER codes are
four-digit numbers:
* The first digit aligning with the state/territory the most (or
all) of the CFER (or self-contained or no flows SA2) is in: NSW = 1;
Victoria = 2; Queensland = 3; SA = 4; WA = 5; Tasmania = 6; NT = 7; and
ACT = 8.
* The second digit indicates the type of region it is: 1 = the
region is an interactive CFER, formed through the Intramax procedure; 2
= the region is a single SA2 that only has an intra-zonal flow (that is,
a SCLM); and 3 = indicates the region is an SA2 that had no commuting
flows and as such was excluded from the analysis.
* The final two digits then start at 01 for the region
incorporating the capital city CBD, and increase as the regions fan out.
While SCLMs should be treated as interactive CFERs in most analyses,
this coding structure allows analysts to consider the difference between
these types of regions.
The SA2s that were not part of interactive Rail CFERs were
combined, given the name Not Included and the code 9000. There were more
than 100 SCLMs in NSW for the Bicycle CFERs, hence these begin at 1195
and continue to 1299.
5. MAPPING THE CFERS
We have mapped the 10 regionalisations of CFERs to show the spatial
pattern of these functional economic regions across Australia, with map
inserts focusing on the Greater Capital City Statistical Areas (GCCSAs)
of Sydney, Melbourne, Brisbane, Perth and Adelaide. Those maps are
provided in Figures 1 through 10. Alternatively, an interactive map can
be found at http://e1.newcastle.edu.au/coffee/maps/CFER2011/AusByCFER2011.html. Note that in these maps the boundaries of all of the CFERs
derived from the Intramax procedure are shown, not just for the
interactive CFERs.
The discussion that follows draws attention to some of the
significant features of those maps for the 10 regionalisations of the
CFERs.
Original CFERs
The boundaries of the Original CFERs derived from the 2011 census
JTW commuting flows data for all workers across all industry categories
are shown in Figure 1. There are 159 Original CFERs across Australia.
[FIGURE 1 OMITTED]
There are multiple Original CFERs in Australia's five
GCCSAs--with the boundaries often extending beyond the de jure defined
GCCSAs--indicating that those large metropolitan concentrations of
people are characterised by a poli-centric structure in which distinct
functional labour market regions have emerged. The spatial shape of
these Original CFERs tend to be elongated stretching out along major
transport routes
For the Sydney GCCSA there are 7 Original CFERs, with an additional
three adjoining to encompass the Newcastle region to the north and the
Wollongong region to the south. For Melbourne there are 6 CFERs with a
further one adjoining encompassing the Geelong-Surf Coast. Brisbane has
4 CFERs plus the adjoining Gold Coast-Tweed to the south and the
Sunshine Coast to the north to encompass what is known as the
Brisbane-South East Queensland region. For Perth there are 4 CFERs.
However, for Adelaide, the smallest of the GCCSAs, there is only one
CFER which encompasses Greater Adelaide and the Barossa, with an
adjoining CFER to the east that includes the Adelaide Hills-Murray
Bridge-Fleurieu Peninsula.
In and around the National Capital area of Canberra there is only 1
large interactive CFER. And there are 4 Original CFERs in and around
Hobart, and 3 in and around Darwin.
Outside the GCCSAs, across regional Australia the Original CFERs
tend to focus largely on the larger regional cities and towns and
encompass surrounding hinterland areas that may include a number of
smaller urban centres, with the shape of those CFERs tending to be
elongated (linear) along major transport routes. It is significant (but
unsurprising) that the Original CFERS in regional Australia tend to
cross over the State borders in the EC + SA large region, especially
along the Murray River which forms the border between NSW and Victoria,
along the eastern part of the NSW-Queensland border, and in what is
often referred to as the Green Triangle section of the Victoria-SA
border.
The Original CFERs tend to become less numerous and much larger in
size with increasing distance inland from the coastal fringes of
Australia, reflecting the rapid decrease in population density and the
lack of larger urban centres in the inland and more remote areas of
Australia. In some of the remote inland areas--especially in outback
Queensland, in the Northern Territory, and in the inland and north-west
Western Australia--there are some more self-contained CFERs. These are
associated with mining settlements or Indigenous communities and many of
these regions, particularly the Indigenous communities, have very small
economies with very little commuting, and as such, maintain the default
boundaries applied by the ABS for their SA2.
Gender-differentiated CFERs
The boundaries of the Male (MCFERs) and the Female (FCFERs) regions
are shown in Figures 2 and 3 respectively.
Across Australia there are 159 Male CFERs (the same as the number
of the Original CFERs), but the number of Female CFERs is considerable
greater at 191. This could reflect the gender differences in the
incidence of male and female employment in different industry and
occupation categories and the patterns of spatial concentration and
dispersal of male and female jobs in those sectors of economic activity.
But within and around the GCCSAs there is not a lot of difference. For
the area in and around Sydney there are 12 Male CFERs and 13 female
CFERs; for Melbourne it is 7 and 10; for Brisbane it is 6 and 6; for
Perth it is 4 and 4; and for Adelaide it is 2 and 3. The ACT has 5 Male
CFERs and 7 Female CFERs. Hobart has 4 CFERs for both Males and Females.
Thus it is beyond the large metropolitan city regions into regional
Australia where there are a substantially larger number of Female CFERs
than Male CFERs with a tendency for the Female CFERs to be more confined
to focusing on country towns and lesser inclined to encompass the
hinterland areas surrounding the larger regional cities, and that is the
case across all four of the large regions into which we divided
Australia.
In the remote areas of NT and WA there are quite a large number of
both Male and Female CFERs that are confined largely to Indigenous
settlements and to mining settlements.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Occupation and Skills differentiated CFERs
The boundaries of the Skilled CFERs, the Less Skilled CFERs, and
the Trades CFERs are shown in Figures 4, 5 and 6.
Skilled CFERs
From Figure 4 we see there are a total of 163 Skilled CFERs across
Australia. Focussing on the Sydney GCCSA, there are 6 Skilled CFERs,
plus another 3 taking in the Central Coast and Newcastle-Hunter area to
the north and Wollongong-Illawarra-Batemans Bay area to the south, and
another 2 encompassing the Blue Mountains to Sydney's west. In and
around Melbourne there are 6 Skilled CFERs, including the area around
Geelong. However, in and around Brisbane there is just 1 Skilled CFER
covering Greater Brisbane, plus 3 further Skilled CFERs taking in the
Gold Coast-Tweed to the south, Ipswich-Toowoomba to the west, and
Sunshine Coast to the north. A single large Skilled CFER encompasses
Adelaide-Barossa-The Coorong. In and around Perth there are 5 Skilled
CFERs. It is noteworthy again that in and around the ACT there is only 1
interactive Skilled CFER which, while there are 4 in Hobart, and Darwin
has 4.
Beyond the areas within and surrounding the GCCSAs the Skilled
CFERs tend to take on somewhat similar forms to the previously-discussed
original and Male and Female CFERs, except that there are a slightly
larger number of them compared to the Male CFERs but fewer than the
Female CFERS. The Skilled CFERs in regional Australia are certainly
focused largely on the economic functions in those larger regional urban
centres that depend on skilled workers draw from large hinterlands,
indicating the smaller regional urban centres do not have local skilled
worker labour markets.
In remote areas--especially in WA--there are distinct Skilled CFERS
focused on mining settlements, and also on remote indigenous
settlements. But across the inland remote areas of Australia there are
few very large Skilled CFERs.
[FIGURE 4 OMITTED]
Less Skilled CFERs
As shown in Figure 5 there are substantially more Less Skilled
CFERs across Australia at 188 compared to the 163 Skilled CFERs.
The Less Skilled CFERs are also more numerous in and around the
GCCSAs. There are 14 across the Sydney-Newcastle-Wollongong areas; 12 in
and around Melbourne-Geelong; 7 across the Brisbane-SEQ region; 3 across
the Adelaide area; and 6 across the Perth area. Again in and around the
ACT there is just 1 interactive Less Skilled CFER. Hobart has 4 Less
Skilled CFERs, and Darwin has 3.
Beyond the GCCSAs and their surrounds into regional Australia the
Less Skilled CFERS are considerably more numerous than is the case for
the Skilled CFERs. They tend to focus not only on the larger regional
cities and towns, but also on some of the smaller regional urban
centres, which indicates that many urban centres in regional Australia
can sustain local labour markets for Less Skilled workers.
Across the inland remote areas of Australia there are few in number
but large Less Skilled CFERs. But there are a considerable number of
Less Skilled CFERS in the remote areas of NT and WA focusing on mining
settlements and indigenous settlements.
[FIGURE 5 OMITTED]
Trades CFERs
Figure 6 shows there are even more Trades CFERs across Australia at
203 in total.
In and around the GCCAs, there are 15 across the
Sydney-Newcastle-Wollongong area, and 11 across Melbourne-Geelong. There
are 5 Trades CFERs across Brisbane-SEQ; 4 across the greater Adelaide
area; and 4 across Perth. Once more in and around the ACT there is just
1 interactive Trades CFER, while Hobart has 4, and Darwin has 3.
Across regional Australia again the Trades CFERs tend to focus
predominantly on the larger regional cities and towns, but a few of the
less large urban centres do seem to support Trades CFERs.
Again in the remote parts of NT and WA there are Trades CFERs that
focus on mining settlements and indigenous settlements.
[FIGURE 6 OMITTED]
JTW Mode of Transport differentiated CFERS
The boundaries of the Road JTW CFERs, Rail JTW CFERs, Bicycle JTW
CFERs, and Multiple Transport JTW CFERs are shown in Figures 7, 8, 9 and
10. What these maps represent are approximations of largely
self-contained commute sheds for specific JTW travel modes for workers
at the time of the 2011 census, in which there is a preponderance of
that mode of commuters who both live and work within a designated CFER.
Road JTW CFERs
As shown in Figure 7 across Australia there are 168 Road JTW CFERs,
which is slightly more than the number of Original CFERs. It is
important to stress the overall high incidence of the private motor
vehicle as the predominant model of travel to work in Australia, and the
almost total reliance on that mode of travel across regional Australia.
For the GCCSAs, we see in and around Sydney 7 Road JTW CFERs, with
another 6 to the north, west and south embracing Newcastle, Wollongong
and the Blue Mountains areas. There are 6 Road JTW CFERs encompassing
the Melbourne GCCSA and Geelong-Surf Coast; 4 across the Brisbane GCCSA
plus Gold Coast-Tweed and Sunshine Coast; 3 across the greater Adelaide
area; and 6 across the Perth GCCSA. The ACT has only 1 interactive Road
JTW CFER, while there are 4 in Hobart and 3 in Darwin.
Across regional Australia the Road JTW CFERs tend to focus on the
larger urban centres and to take in smaller urban centres in their
hinterlands, often covering quite large areas spread out along the main
roads.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
Rail JTW CFERs
Figure 8 shows that it is only the largest 5 of the GCCSAs with
commuter rail networks that have Rail JTW CFERs.
There are only 19 such CFERs in total--just 3 across Sydney and
extending south to Wollongong. Across greater Melbourne there are 3 Rail
JTW CFERs which extend north-west to Ballarat and south-west to Geelong,
and east to Traralgon, plus an additional one that encompasses
Bendigo-Castlemaine along a regional rail service. For the Brisbane
region there are 3 Rail JTW CFERs that extend south to the Gold Coast
and west beyond Ipswich. Adelaide has 3 Rail JTW CFERS, while Perth has
4. As one would expect, these Rail JTW CFERs are extensive in size and
elongated in shape which reflects the radial commuter rail networks that
radiate out from the capital CBDs.
Bicycle JTW CFERs
There has been a considerable public policy push in Australia to
encourage cycling as a mode of travel and while the incidence of cycling
for the JTW is increasing it still represents a minute proportion of the
JTW. As shown in Figure 9, these Bicycle CFERs tend to be small
geographically and there is a large number of them--a total of 817
across Australia. This is not surprising as it is unlikely that workers
who cycle to work would be prepared to travel a long distance.
For the GCCSAs and their surrounding areas, there are about 25
Bicycle CFERs across Sydney, Newcastle, Wollongong and the Blue
Mountains area; about 20 across greater Melbourne; almost 30 across
Brisbane-SEQ; 17 across greater Adelaide; and 10 across Perth. There is
just 1 large interactive Bicycle JTW CFER across the ACT. There are 4
across Hobart and 5 across Darwin.
[FIGURE 9 OMITTED]
There are a large number of Bicycle JTW CFERs across regional
Australia--literally numbering in the hundreds--with them tending to
focus on both the larger and the smaller urban centres.
Multiple Transport Mode JTW CFERs
Figure 10 shows the Multiple Transport Mode JTW CFERs which number
291 across Australia.
For the Sydney GCCSAs there are 3 Multiple Transport Mode CFERs,
and there are additional ones to the north and south which extend beyond
Newcastle and Wollongong and as well into the Blue Mountains. Greater
Melbourne has 4 Multiple Transport Mode CFERS which extend well beyond
that GCCSA along the regional commuter rail links east into the La Trobe
Valley, west to Bacchus Marsh and north to Seymour and west and
north-west to Ballarat and Bendigo. It is interesting that there is just
a single Multiple Transport Mode CFER covering a large area that
encompasses Brisbane-Gold Coast-Toowoomba, with another covering the
Sunshine Coast. A single Multiple Transport Mode CFER covers the whole
of Greater Adelaide and surrounds. And there are several Multiple
Transport Mode CFERs across the greater Perth region. In and around the
ACT there is just 1 large interactive Multiple Transport Mode CFER.
There are 2 in Hobart and 2 across Darwin.
In the regional areas of Australia the Multiple Transport Mode
CFERs are focused largely on the larger urban centres and typically
encompass a number of urban centres surrounding them.
[FIGURE 10 OMITTED]
6. CONCLUSION
This paper has outlined how we have been developing a new spatial
base for investigating regional performance across Australia employing
an approach that seeks to derive functional regions using the Intramax
procedure and JTW data available in the 2011 census. In addition to
deriving FERS that relate to aggregate employment across all industry
categories (the Original CFERS), we have also derived regionalisations
that segment workers into gender, occupation/skills categories, and
different transport modes for the JTW commute.
The paper discusses the outcomes of the 10 regionalisations derived
from the JTW commuting data using the Intramax procedure and highlights
some of the spatial patterns that result both across Australia's
major capital city areas and across regional Australia. Not surprisingly
there are considerable variations in both the number of CFERs that are
derived for the 10 regionalisations used for this paper as well as the
spatial characteristics of some of the patterns for those employment
segmentations.
This research adds further evidence to demonstrate that labour
markets are not homogenous across a space economy. The regional
demarcations based on gender and occupations/skills of the labour force
certainly show that for Australia's large capital city regions
there are distinct local labour markets as a result of differences in
the emerging patterns of spatial diffusion and concentration of
employment that have subtle differences for jobs dominated by male and
female work and by levels of skill and occupation. And it is also
evident that the mode of travel to work chosen by commuters results in
very substantial differences in the incidence and patterns of functional
regions.
The research presented in this paper will now be used as the basis
for much more detailed interrogation using spatial econometric
analytical tools to investigate possible determinant of spatial
differentials in the economic performance of the CFERs derived through
the Intramax procedure. In that we will concentrate on the distribution
of unemployment and employment growth across the regions and investigate
the regional disparities that exist. Further, we will endeavour to
explore in depth the characteristics of the gender and occupation/
skills segmented CFERs on a region-by-region basis for the major capital
city areas and across parts of regional Australia. It may also be
interesting to compare the Australian case with the situation in Europe
and/or the US.
ACKNOWLEDGEMENT: The research on which this paper is based is
funded by the Australian Research Council (ARC) Discovery Project Grant
#DP150103437.
REFERENCES
Australian Bureau of Statistics (ABS) (2010). Cat.
1270.0.55.001--Australian Statistical Geography Standard (ASGS): Volume
1--Main Structure and Greater Capital City Statistical Areas, July 2011,
Australian Bureau of Statistics, Canberra.
Andersen, A.K. (2002). Are Commuting Areas Relevant for the
Delimitation of Administrative Regions in Denmark? Regional Studies,
36(8), pp. 833-844.
Anselin, L. (1988). Spatial Econometrics: Methods and models.
Kluwer Academic Publishers, Dordrecht, The Netherlands.
Berry, B.J.L. (1968). A Synthesis of Formal and Functional Regions
Using General Field Theory of Spatial Behaviour. In B.J.L. Berry and
D.F. Marble (Eds.) Spatial Analysis. Prentice-Hall, Englewood Cliffs,
New Jersey.
Bill, A., Mitchell, W. and Watts, M. (2008). The Occupational
Dimensions of Local Labour Markets in Australian Cities. Built
Environment, 34(3), pp. 291-306.
BITRE (2015). Australia's commuting distance: cities and
regions. Bureau of Infrastructure, Transport and Regional Economics,
Canberra.
Breukelman, L., Brink, G., de Jong, T. and Floor, H. (2009).
Flowmap 7.3 Manual. Faculty of Geographical Sciences, Utrecht
University, The Netherlands, http://flowmap.geo.uu.nl.
Casado-Diaz, J.M., Martinez, L. and Florez-Revuelta, F. (2010).
Spanish Local Labour Markets. An Application of the New British
Procedure. In J.M. Albertos and J.M. Feria (Eds.) La ciudad
Metropolitana en Espana: Procesos Urbanos en los Inicios del SigloXXI,
Madrid, Thomson-Civitas, pp. 275-313.
Coombes, M.G., Green A.E. and Openshaw S. (1986). An Efficient
Algorithm to Generate Official Statistical Reporting Areas: the Case of
the 1984 Travel-to-work Areas Revision in Britain. Journal of the
Operational Research Society, 37, pp. 943-953.
Crane, R. (2007). Is There a Quiet Revolution in Women's
Travel? Revisiting the Gender Gap in Commuting. Journal of the American
Planning Association, 73(3), pp. 298-316.
LeSage, J.P. and Pace, R.K. (2009). Introduction to Spatial
Econometrics, Taylor & Francis, Boca Raton.
Masser, I. and Brown, P.J.B. (1975). Hierarchical Aggregation
Procedures for Interaction Data, Environment and Planning A, 7, pp.
509-523.
Mitchell, W.F. and Flanagan, M. (2016, forthcoming). The Changing
Patterns of European Unemployment in the Face of the Financial Crisis
and Policy Austerity. Urban Studies--Special Issue on
"Employability and labour market policy in an era of crisis".
Molho, I. (1995). Spatial Autocorrelation in British Unemployment.
Journal of Regional Science, 35(4), pp. 641-658.
Morrison, P.S. (2005). Unemployment and Labour Markets. Urban
Studies, 42(12), pp. 2261-2288.
Niebuhr, A. (2003). Spatial Interactions and Regional Unemployment
in Europe. European Journal of Spatial Development, 5, pp. 1-26.
Openshaw, S. (1984). The modifiable areal unit problem. Concepts
and Techniques in Modern Geography, 38(41), GeoBooks, Norwich, England.
Patacchini, E. and Zenou, Y. (2007). Spatial Dependence in Local
Unemployment Rates. Journal of Economic Geography, 7, pp. 169-191.
Randolph, B. and Holloway, D. (2005). The Suburbanization of
Disadvantage in Sydney: New Problems, New Policies. Opolis, 1(1), pp.
49-65.
Richardson, H. (1973). Regional Growth Theory, Macmillan, London.
Sang, S., O'Kelly, M. and Kwan, M. (2011). Examining Commuting
Patterns: Results from a Journey-to-work Model Disaggregated by Gender
and Occupation. Urban Studies, 48(5), pp. 891-909.
Stimson, R. J., Mitchell, W., Rohde, D. and Shyy, T-K. (2011).
Using Functional Economic Regions to Model Endogenous Regional
Performance in Australia: Implications for Addressing the Spatial
Autocorrelation Problem. Regional Science Policy and Practice, 3 (3),
pp. 131-144.
Watts, M. (2013). Assessing Different Spatial Grouping Algorithms:
An Application to the Design of Australia's New Statistical
Geography. Spatial Economic Analysis, 8(1), pp. 92-112.
Watts, M.J., Baum, S., Mitchell, W.F. and Bill, A. (2006).
Identifying Local Labour Markets and Their Spatial Properties. Paper
presented to ARCRNSISS Annual Conference, Melbourne, May.
Zipf, G. (1949). Human Behaviour and the Principle of Least Effort,
Addison-Wesley, Cambridge.
Robert Stimson
Professor, Department of Resource Management and Geography,
University of Melbourne, Parkville, Vic, 3010, Australia. Email:
[email protected]
William Mitchell
Emeritus Professor, Research and Innovation Division, University of
Newcastle, Newcastle, NSW, 2308, Australia. Email:
[email protected]
Michael Flanagan
Senior Research Assistant, Research and Innovation Division,
University of Newcastle, Newcastle, NSW, 2308, Australia. Email:
[email protected]
Scott Baum
Professor, Griffith School of Environment, Griffith University,
Brisbane, Qld, 4111, Australia. Email:
[email protected]
Tung-Kai Shyy
Research Fellow, School of Information Technology and Engineering,
University of Queensland, Brisbane, Qld, 4072, Australia. Email:
[email protected]
Table 1. The ten regionalisations of functional economic regions
across Australia derived from JTW data available in the 2011
census.
All workers:
1. Original CFERs (CFER2011)
Gender-based:
1. Male CFERs (MCFER2011)
2. Female CFERs (FCFER2011)
Occupation-based:
2. Skilled CFERs (SCFER2011)--workers in ANZSCO categories:
* Managers
* Professionals
3. Less Skilled CFERs (LSCFER2011)--workers in ANZSCO categories:
* Community and Personal Service Workers
* Clerical and Administrative Workers
* Sales Workers
* Machinery Operators and Drivers
* Labourers
4. Trades CFERs (TCFER2011)--workers in ANZSCO categories:
* Technicians and Trades Workers
JTW Mode of Transport-based:
5. Road JTW CFERs (RoCFER2011)--workers who used one (and only one)
of the following modes of transport to travel to work:
* Car as driver
* Car as passenger
* Bus
* Motorbike
* Taxi
* Tram
* Truck
6. Rail JTW CFERs (RaCFER2011)--workers who travelled to work by
train (only)
7. Bicycle JTW CFERs (BCFER2011)--workers who travelled to work
by bicycle (only)
8. Multiple Transport Mode JTW CFERs (MTCFER2011)--workers who
used more than one mode of transport (those above as well as a
classification of 'Other')
Source: the authors
Table 2. JTW Flow Matrix With j Regions
Origin\ Region 1 Region 2 ... Region j Total
Destination
Region 1 1 to 1 1 to 2 ... 1 to j [summation
over (j)]
[a.sub.1j]
Sum of
flows out
of Region 1
Region 2 2 to 1 2 to 2 2 to j [summation
over (j)]
[a.sub.2j]
... ... ... ... ... ...
Region j j to 1 j to 2 ... j to j [summation
over (j)]
[a.sub.jj]
Total [summation [summation ... [summation n =
over (i)] over (i)] over (i)] [summation
[a.sub.i1] [a.sub.i2] [a.sub.ij] over (i)]
[summation
Sum of over (j)]
flows into [a.sub.ij]
Region 1
Total
Interaction
Source: the Authors, after Masser and Brown, 1975.
Table 3. The Number of Original CFERs Across Australia's Large
Regions, and the Various Aggregations of the ASGS Statistical
Areas in 2011.
NSW Vic QLD SA ACT
States/Territories 1 1 1 1 1
Greater Capital City 2 2 2 2 1
ASGS:
Statistical Area 17891 13335 11039 4087 918
Level 1
Statistical Area 538 433 526 170 110
Level 2
Statistical Area 91 65 80 28 9
Level 3
Statistical Area 28 17 19 7 1
Level 4
CofFEE FERs 79
(the Original (5)
CFERs)
WA Tas NT Australia *
States/Territories 1 1 1 8
Greater Capital City 2 2 2 15
ASGS:
Statistical Area 5508 1446 537 54761
Level 1
Statistical Area 250 98 68 2193
Level 2
Statistical Area 33 15 9 330
Level 3
Statistical Area 9 4 2 87
Level 4
CofFEE FERs 18 12 14 123
(the Original (4) (2) (11)
CFERs)
Source: ABS, 2010; authors' calculations. Note: * Does not
include Other Territories.
Table 4. Regions for the Various CFER Aggregations.
EC + SA WA
a b c a b c
Original CFERSs 79 5 15 18 4 6
Male CFERS 72 7 18 16 4 6
Female CFERs 95 9 21 19 4 6
Skilled CFERS 65 7 23 16 4 10
Less Skilled CFERs 95 11 16 20 4 6
Trades CFERs 95 12 28 19 4 6
Road JTW CFERs 87 6 16 17 4 5
Rail JTW CFERS * 13 -- -- 4 -- --
Bicycle JTW CFERs 131 316 185 18 53 28
Multiple Transport 70 68 52 21 18 13
Mode JTW CFERS
TAS NT
a b c a b c
Original CFERSs 12 0 3 14 2 1
Male CFERS 12 0 3 16 3 2
Female CFERs 12 1 3 15 3 3
Skilled CFERS 12 1 3 15 3 4
Less Skilled CFERs 13 0 3 16 2 2
Trades CFERs 13 1 3 13 5 4
Road JTW CFERs 11 1 3 14 2 2
Rail JTW CFERS * -- -- -- -- -- --
Bicycle JTW CFERs 10 21 27 15 14 11
Multiple Transport 12 7 5 5 14 6
Mode JTW CFERS
Australia
a b c
Original CFERSs 123 11 25
Male CFERS 116 14 29
Female CFERs 141 17 33
Skilled CFERS 108 15 40
Less Skilled CFERs 144 17 27
Trades CFERs 140 22 41
Road JTW CFERs 129 13 26
Rail JTW CFERS * 17 -- --
Bicycle JTW CFERs 162 404 251
Multiple Transport 108 107 76
Mode JTW CFERS
Notes: a = interactive CFERs after the Intramax
procedure, b = non-interactive SA2s (self-contained
labour markets), c = SA2s with no flows.
* Only SA2s that had sufficient rail commuting were
included in any analysis. Source: the Author's
calculations.
