Kevin O’Connor and Robert J. Stimson

The location of population and business activity in Australia is a critical influence on a range of environmental, economic and social outcomes. This paper provides insight on recent trends, showing that the major urban areas continue to have a strong role in the settlement pattern. This outcome is compared to recent research in the US.

Shifts in the location of population and economic activity within Australia and their correspondence to recent patterns of change in the US are the concern of this paper. The results of research into the geography of national population change in the US have begun to show that the rural parts of the nation are now attracting a disproportionate share of population and investment. For example, a review of the US research, and a set of analyses on the trends in population, jobs and income between 1969 and 1994, was recently presented by Gordon and Richardson.¹ This extended their long-term research on the issue which has shown that decentralisation is the dominant force in the change in the pattern of settlement within the United States.²

The central idea of their recent paper is that breakthroughs in goods handling and in the transmission of unambiguous information have created ‘.... increasingly ubiquitous agglomeration economies’.³ This is in distinct contrast to an earlier era when narrowly circumscribed core areas were the focus of savings associated with inter-firm linkages and contacts. Gordon and Richardson’s explanation for this change rests on the availability of the interstate highway system for the movement of goods and for commuting, as well as the lower cost of telecommunications using networks of computer systems. Both of these developments mean that small firms which rely upon linkages and contacts with subcontractors for components and information no longer need to be anchored to the central zones of large cities. Such firms can now operate successfully from middle and outer suburbs, as well as from selected rural locations. Gordon and Richardson’s understanding of this change relates to Scott’s⁴ re-interpretation of the links between industrialisation and urbanisation.

Although it is important to be careful about the translation of US experience to Australia, it is challenging to review patterns of change here against that backdrop. The pattern of change revealed in Gordon and Richardson’s overview is particularly significant as there has been a decade of interest here in regional development, much of it with a non-metropolitan emphasis, including reports that carried an optimistic assessment of the fortunes of non-metropolitan Australia in a new globalised economy.⁵ This paper sets out to show what has been happening recently in terms of population growth and investment.

ANALYSING NATIONAL URBAN CHANGE

This paper reports the results of a project designed to establish the areas of Australia which have attracted the largest shares of recent population, housing and commercial development. The question of regional differences in development has attracted considerable attention in Australia over the past ten years. For example, there has been special issues of Environment and Planning A⁶ and Built Environment⁷ dedicated to the question, as well as two overview papers by Maher,⁸ a House of Representatives report⁹ and a part of an extensive Urban and Regional Development Review.¹⁰ The current project is based on an earlier paper which outlined the way in which geographic constraints upon the patterns of development in Australia would limit the spatial outcomes experienced as the national economy evolved.¹¹ That perspective was refined in a report to the Federal government on the economic role of cities,¹² and developed further in an analysis of change at the state level.¹³ This analysis of the effects of geography on national development has now been extended to the city and intra-metropolitan scale in recently completed work.¹⁴ The current contribution explores the mix of metropolitan and non-metropolitan outcomes for the first time, in order to establish whether different forces are at work at that scale.

To identify the trends in population and economic change, settlement in Australia is divided into four units: first, the metropolitan Statistical Divisions (SD), second, the fringes beyond the metropolitan SD borders, third, Canberra, and fourth the rest of non-metropolitan Australia. The first category is simply taken from Australian Bureau of Statistics definitions. The second group includes municipalities that adjoin the metropolitan SDs, in particular the coastal extensions of the Sunshine Coast and the Gold Coast around Brisbane, and some fringe areas surrounding Sydney, Melbourne, Perth and Adelaide. Some of these areas were incorporated into an initial analysis of extended metropolitan areas under the title of ‘mega cities’.¹⁵ Their definition has been refined here, following investigation of the local patterns of development surrounding the largest capital cities and the analysis of journey-to-work data, as discussed in Rapson and O’Connor.¹⁶ These fringe areas correspond to the ‘exurban areas’ used in US analysis, but in the Australian case they also include nationally significant tourist and industrial areas. When combined with their adjoining metropolitan statistical division they create a mega-metropolitan region for the present study. The third unit used is the urban region centred on Canberra, which is the Australian Capital Territory plus Queanbeyan. The population size of Canberra does not justify its inclusion in the large metropolitan class, but its functional linkages to the other metropolitan areas (in particular its road links to Sydney), as well as its basic role in public policy, mean that it acts as a key centre in the national settlement pattern. All three units comprise an aggregate called major urban areas, which encapsulates the core urban settlement of the nation. The final category is the rest of the nation, labelled here as non-metropolitan. These units have been used to assemble data on population, dwellings and non-residential construction over the period 1986 to 1995. (Non-residential construction is the building of shops, offices, factories and community facilities that provide the urban fabric for the operation of the economy and the community.)

THE GEOGRAPHY OF POPULATION, HOUSING AND CONSTRUCTION IN AUSTRALIA, 1986-1995
 

Table 1: The distribution of Australia’s population 1986-1995: shares of estimated resident population in major areas
Area	1986	1991	1994	1995
Metropolitan SDs	60.6	60.4	60.0	60.0
Metropolitan fringe	8.9	9.5	9.8	9.9
Mega Metro Regions	69.5	69.8	69.8	69.9
Canberra	1.8	1.8	1.8	1.8
Major Urban Areas	71.3	71.7	71.7	71.7
Non-metropolitan areas	28.7	28.3	28.3	28.3
Australia	100.0	100.0	100.0	100.0
Population (million)	16.017	17.283	17.843	18.051
Source: Australian Bureau of Statistics (ABS), Estimated Resident Population for Statistical Local Areas (state publications for years shown)

The relative distribution of the Australian population has changed very little over the period under review. A small increase in the share living on the fringes of the metropolitan areas, which is almost entirely due to the growth on the Sunshine and Gold Coast, has maintained the share of the nation’s population living in or near the five large mainland state capitals and Canberra at just over 70 per cent in the past decade (see Table 1). At the same time, the share of the population living in the smaller cities, regional centres and more isolated communities spread across the vast expanse of rural Australia has declined slightly. The stability in the share of the population in the major cities distinguishes Australia from the US at the outset.
 
 

Table 2: Home construction in Australia, 1986-1995: shares of new dwellings in major areas
Area	1987-88	1989-90	1991-92	1992-93	1993-94	1994-95	1995-96
Metropolitan SDs	56.7	52.1	51.2	54.2	53.9	57.5	58.1
Metropolitan fringe	16.2	15.8	14.7	14.2	15.9	14.8	14.2
Mega Metro Regions	72.9	67.9	65.9	68.5	69.8	72.4	72.3
Canberra	1.9	1.8	2.2	2.8	2.4	1.8	1.9
Major Urban Areas	74.8	69.7	68.1	71.3	72.3	72.3	74.2
Non-metropolitan areas	25.2	30.3	31.9	28.7	27.7	25.8	25.8
Australia	100.0	100.0	100.0	100.0	100.0	100.0	100.0
No. of dwellings	150,849	139,661	119,662	170,436	184,487	167,134	122,535
Source: ABS, Building Approvals by Statistical Local Areas, unpublished

The geography of new dwelling construction (shown in Table 2) differs a little from the geography of population. The fringe locations were very important places for home construction in the latter part of the 1980s as large scale construction of retirement villages, holiday homes and rural retreats pushed up the share of national home construction located on the metropolitan fringe to 16 per cent. This is more than the nine per cent of the nation’s population that live in this location. In the early 1990s, the metropolitan areas seemed to be bypassed as their share of new homes dropped to 51 per cent. This diminishing share was due especially to a big fall in home building in Melbourne.¹⁷ In 1991-92, the share of new dwellings built in non-metropolitan locations peaked at 32 per cent. However, by 1995-96, the shares in new dwelling construction located in metropolitan areas had returned to the level recorded in 1987-88, so that the major urban areas of Australia accounted for 74 per cent of the home building carried out in that year. Hence it seems that the metropolitan concentration that is evident in the patterns of population is now more accurately reflected in the operation of the housing market than it had been in the 1989-1992 period.

The location of non-residential construction does show some evidence that non-metropolitan locations are attracting economic activity, even though their housing and population shares have fallen recently. In simple terms, the share of approvals of new non-residential buildings in the metropolitan areas has fallen sharply from the peaks recorded in the boom years of the late 1980s (when these areas accounted for 70 per cent of the value of construction) to just 55 per cent of the nation’s building approvals in 1995-96, as shown in Table 3. This shift is compensated in part by an increased role for fringe locations, and by some large projects in Canberra. Thus the major urban areas’ share of non-residential construction was around 71 per cent on 1995-96, exactly equivalent to its share of population. Nevertheless, Table 3 shows that non-metropolitan locations are attracting more investment in new buildings, such that their share of construction approvals rose from 18 per cent in 1987-88 to 28 per cent in 1995-96. The question is whether this change in the location of activity is the beginning of a new trend (perhaps following the US experience) or whether it reflects some special conditions in the last few years. It is the contention of the present research that it does represent special conditions, as discussed below.
 

Table 3: Non-residential construction in Australia, 1986-1995: shares of value of building approvals in major areas
Area	1987-88	1989-90	1991-92	1992-93	1993-94	1994-95	1995-96
Metropolitan SDs	72.7	74.7	63.8	68.7	66.2	63.5	55.2
Metropolitan fringe	8.1	8.9	14.1	7.9	8.5	9.6	11.0
Mega Metro Regions	80.2	83.6	77.9	76.7	74.7	73.1	66.2
Canberra	1.6	3.9	3.1	3.0	3.7	2.0	5.1
Major Urban Areas	81.8	87.4	80.9	79.7	78.4	75.2	71.3
Non-metropolitan areas	18.2	12.6	19.1	20.3	21.6	24.8	28.7
Australia	100.0	100.0	100.0	100.0	100.0	100.0	100.0
Value ($ million)	8,412.116	9,432.157	4,191.232	4,761.476	5,371.772	5,588.356	6,587.247
Source: ABS, Building Approvals by Statistical Local Areas, unpublished

The first of these special features is that one location, Cairns, has played a significant part in the share recorded in the non-metropolitan areas. This location accounted for almost four per cent of the nation’s value of non-residential building approvals in 1994-95 and 1995-96, up from less than one per cent in the previous three years, as it attracted investment in shopping centres and hotels. In fact, 17 per cent of the value of all non-residential building taking place throughout non-metropolitan Australia was located in this one place. Closer study of the data indicates that there has also been investment in factories in non-metropolitan locations for mineral processing in Queensland and Western Australia and for new food processing factories in Victoria. Hence the shift away from the metropolitan areas implied in Table 3 reflects, for the most part, a small number of large projects in just a few places rather than dispersal to a large number of locations, which appears to be the outcome in the US.

The second special condition that needs to be taken into account here is that the very large shares of non-residential construction recorded in the metropolitan areas in the late 1980s reflected boom conditions which were unsustainable in the long term. The collapse in the metropolitan office-construction boom due to oversupply illustrates that outcome well. The boom in the late 1980s had one other important consequence. It meant that the new and restructured businesses which emerged from the recession had little difficulty finding office, factory or warehouse space to accommodate their re-organised or new activity, as vacant space was readily available in the metropolitan areas. In this way the metropolitan areas have been able to retain a stronger role in the daily business of the nation than the trends in non-residential construction would suggest. This can be seen in the shares of the nation’s firms in a range of activities as displayed in Table 4. In the overall distribution of all businesses, shown in column one, the metropolitan areas seem unrepresented, largely because the many farms, which count as separate businesses, push up the share of businesses in non-metropolitan Australia. It is in areas such as manufacturing, finance and businesses services that metropolitan areas and the mega metropolitan regions have such a significant role.
 
 

Table 4: Business locations in Australia: 1995: share of number of businesses in major areas
Area	Total businesses	Manufacturing	Wholesale	Finance and Insurance Services	Property and Business Services
Metropolitan SDs	52.0	68.7	68.4	69.2	73.5
Metropolitan fringe	8.5	8.7	8.2	7.9	8.6
Mega Metro Regions	60.5	77.4	76.6	77.1	82.1
Canberra	1.4	0.9	1.2	1.8	2.1
Major Urban Areas	61.9	78.3	77.8	78.9	84.2
Non-metropolitan areas	38.1	21.7	22.2	21.1	15.8
Australia	100.0	100.0	100.0	100.0	100.0
Number of businesses	937,849	50,143,	63,060	38,119	135,857
Source: ABS, Integrated Regional Database (IRDB), count of business locations

The critical element in the above analysis is that the major urban areas of the nation have accommodated the majority of the population and most of the construction activity associated with housing and businesses in the past few years. This outcome reflects the ability of the established urban areas of the nation to accommodate additional development, an aspect of Australian urban change which is often overlooked. This can be understood more clearly by a closer look inside these major areas.

It is apparent that the middle and outer suburban areas have been the key parts of the evolution of the settlement of Australia over the past decade. In this respect Australia has mirrored the US experience where ‘..most growth is suburban: even in the 1980s the "urban revival" was a period of strong suburban growth, when most core counties suffered a relative decline’.¹⁸
 

Table 5: Population and dwellings within mega metropolitan regions 1986-1995: share (%) of national population and dwellings, and rates of population growth
	Share of population	Rates of population change (%)		Share of population change in Australia		Share of dwelling approvals
	1995	1986-91	1991-95	1986-91	1991-95	1988-1996
Core Inner Middle Outer Fringe	4.4 6.5 21.5 27.5 9.9	-0.1 -0.1 0.6 3.0 2.9	-0.4 -0.1 0.1 2.1 2.3	-0.4 -0.6 8.4 49.7 17.0	-1.8 -0.9 2.0 51.5 20.3	3.7 2.8 13.4 35.0 15.2
Mega Metro Regions	69.9	1.6	1.1	74.1	71.2	70.1
Australia	100.0	1.5	1.1	100.0	100.0	100.0
Source: See Tables 1 and 2

THE LOCATION OF NATIONAL POPULATION AND ACTIVITY WITHIN MEGA METROPOLITAN REGIONS

This stage of the analysis utilises four areas within the metropolitan statistical divisions — core, inner, middle, outer and fringe — which are defined in Rapson and O’Connor (1997).¹⁹ The rates of population growth in outer and fringe locations were double the national average over the decade analysed here, whereas population losses were recorded in the core and inner areas, as shown in Table 5. The outer metropolitan areas (generally those parts of the suburbs that experienced rapid population growth after 1970) actually accommodated one half of the extra people added to Australia’s population in the nine years to 1995. Obviously the majority of the Australian population still prefers larger suburban residential sites to inner-area units associated with the intensification of residential land development. Even with considerable publicity (and public policy support) there has not been any reversal of losses in population in the inner areas.²⁰
 
 

Table 6: The location of non-residential construction within mega metropolitan regions and share of value of national building approvals, 1987-88 to 1995-96
	1987-88	1989-90	1991-92	1992-93	1993-94	1994-95	1995-96
Core Inner Middle Outer Fringe	39.1 4.3 14.1 14.7 8.1	31.6 6.7 16.0 20.3 8.9	21.9 3.7 17.6 20.6 14.1	14.6 7.1 21.4 25.7 7.9	17.9 3.9 21.9 22.5 8.5	19.1 4.1 13.5 26.8 9.6	11.4 6.9 17.6 19.3 11.0
Mega Metro Regions	80.2	83.5	77.9	76.7	74.7	73.1	66.2
Australia - share Australia - $’000	100.0 8,422.5	100.0 9,448.6	100.0 4,197.3	100.0 4,762.0	100.0 5,372.6	100.0 5,592.7	100.0 6,587.2
Source: ABS, Building Approvals by Statistical Local Areas, unpublished

Middle and outer metropolitan regions have accommodated most of the nation's population growth between 1986 and 1995, but this is only part of the role that these regions have played in recent urban change. Table 6 shows the shares of national non-residential construction approvals located in each part of the mega metropolitan regions. The effect of the collapse of the office building boom on the core has been dramatic. In 1993-94 core areas accounted for less than half the share evident in 1987-88, and two years later their share had fallen to 11 per cent. However, the changes do not simply signify the collapse in CBD office building. They also show a broad trend toward the location of commercial and public facilities across the breadth of the metropolitan areas, particularly in middle and outer suburbs. These two areas accounted for 43 per cent of the nation’s approved non-residential construction in 1994-95 — still less than their share of population, but a major shift from the situation recorded in 1987-88.

This outcome is the result of the construction of new hospitals, universities and other community facilities, as well as factories, shopping centres, offices and warehouses, restaurants and entertainment facilities which have been built by firms recognising the opportunities created by the re-organisation and redirection of business operations. A confirmation of the role that the middle and outer metropolitan suburbs have played in the new settlement patterns of Australia is also provided by the patterns of journey-to-work detected by research on travel in the 1991 Census.²¹ These showed increased self-containment in middle and outer areas and, in particular, strong flows from the outer (residential) areas to middle (commercial) areas. Similar outcomes have been found in the analysis of journey-to-work in US cities and in the UK.²²
 

Table 7: The location of Australian businesses in areas of mega-metropolitan regions 1995: shares of national totals in areas, by type of industry
Area	Total businesses	Manufacturing	Wholesale	Finance and Insurance Services	Property and Business Services
Core Inner Middle Outer Fringe	11.2 6.2 16.8 17.8 8.5	9.5 6.5 27.9 24.8 8.7	14.5 8.2 25.8 19.9 8.2	25.1 8.8 20.2 15.1 7.9	22.1 9.9 21.1 20.4 8.6
Mega Metro Regions	60.5	77.4	76.6	77.1	82.1
Source: IRDB, see Table 4

In overview, as Australia slowly adjusts to the global economy, it remains predominantly metropolitan, and suburban, in function and scope. This can be seen most clearly in the location of jobs and housing in its mega-metropolitan areas and within them, in the middle and outer suburbs. In a number of instances, these suburban locations provide a base for globally-focussed businesses; this is the case with firms such as NEC and Hewlett Packard in eastern Melbourne. When these globally oriented activities are added to the population-serving retail and community facilities, the complexity of the suburban framework of the economy becomes more apparent. This is best illustrated by the national shares of businesses located in parts of the mega metropolitan regions displayed in Table 7. The data indicate that 33 per cent of all Australian businesses are in the middle and outer suburbs of mega-metropolitan region; in manufacturing that share is 53 per cent and, even in the activities traditionally associated with the CBD (the finance and business services sector), one third or more of the businesses counted were in middle and outer suburban locations.

REASONS FOR URBAN CONCENTRATIONS IN AUSTRALIA

These outcomes are the results of a set of influences that are re-shaping the geography of nations and the city regions within them. These influences are what, in a different context, Hilmer²³ has called the ‘new rules of the game’. These new rules emanate from the fragmentation of the production process from single large firms into many small sub-contractors and suppliers, a fragmentation which has spread the range of inter-sectoral and inter-firm linkages across a larger number of sectors and firms out into the suburbs. This outcome, associated with outsourcing and downsizing, has created opportunities for many new firms. In the US context it may be that these linkages between sub-contractors and suppliers can reach out even beyond the exurban areas, as Gordon and Richardson²⁴ indicate that ‘the rural renaissance of the 1970s was primarily a move to smaller and mid-sized metropolitan statistical areas’. In Australia, the lack of mid-sized cities has strengthened the need for many firms to co-locate in the middle and outer suburbs.

The national scale of production and delivery following the construction of national-scale transportation and high capacity telecommunications systems put in place during the 1970s²⁵ has strengthened the attraction of metropolitan areas. At the same time, participation in global markets has usually been based in metropolitan areas, as these can provide the breadth of contacts and specialist services which firms need if they are to launch into wider markets or, alternatively, act as uncertainty minimising gateways for inbound firms and investment. The availability of international connections, via a container terminal or an airport, can also be an attraction for firms with globally organised production networks. The largest container terminals, and the busiest airports, are generally in large metropolitan areas, providing the major international port and airport access points.

One dimension left unexplored here is the unevenness of metropolitan development. In the US analyses, individual city change usually reflects Sunbelt attractions relative to Frostbelt locations. That simple distinction has only a weak parallel in Australia, but there are some significant distinctions in the patterns of recent change as illustrated in data on jobs, investment and population shown in the Australian Capital City Report.²⁶ Integration of those data with more detailed information on corporate performance and small business development in south-east Queensland²⁷ has shown that business development in that region has not been unable to match population growth rates. The national data also show that Adelaide has a weaker role than some other metropolitan areas in the national pattern, while Perth’s development has been linked closely to the role which it plays in the mining industry rather than to a broadly-based set of activities. Sydney, meanwhile, seems to be the location for the jobs filled by higher wage paying industries like finance. More detailed analysis of individual cities will establish whether there is a nation-wide trend at work, or whether the results are shaped by the performance of individual cities.

These outcomes suggest that there has been a re-definition of the links between patterns of population growth and patterns of commercial and business investment. In the past, it was common to imagine that population growth would trigger local commercial development, or alternatively that commercial investment would attract population. These simple linkages were built into regional and state planning policy in Australia. In the new context, the needs of population growth in some regions can be easily met by production facilities elsewhere in the nation (or the world), while many new job opportunities are anchored to just a few sites in the nation which can meet the needs of national and global businesses. Links which used to mean that more people generated more jobs in the same geographic region have weakened as the concentration of production has reshaped the role that large urban regions play in national and international production and consumption and, in the Australian context, they have strengthened an already highly concentrated urban settlement pattern.

This outcome differs from that found by US researchers. This is because the small size of the Australian economy, the absence of the mid-sized cities and the lack of major deterrents to development in the metropolitan areas, have directed the powerful concentration forces embedded in many new communication and transport technologies. Taken together, these changes in communications and transport maintain and enhance the role of the few large urban centres that are dotted along a part of the coast, places which have been central to the structure of Australian geography for a long period of time. It is these places that should be the centre of attention when attempts are made to understand and accommodate urban and regional growth processes.

Acknowledgments

This research was funded by the Australian Research Council grant on Agglomeration and Dispersal Processes and the Demographic and Economic Restructuring of Cities and Regions. Data assembly was carried out by Virginia Rapson. The authors acknowledge the helpful comments made by a referee and by the editors of the journal.

References

Back to People and Place Home Page