The roots of Sydney’s crisis lie in the rejection of office decentralisation back in the 1950s-1970s

 

John Muscat / The New City

The New South Wales Government’s historic decision to lift Sydney CBD’s 150-foot building height restriction in 1957 seems inevitable in retrospect. But it was controversial at the time and attracted strong criticism from technical urban planners into the 1980s. Debates about height limits were tied up with broader discussions about whether to embrace decentralising trends induced by mass motorisation and aspirations for decent home ownership or reinforce growth in the traditional city centre.

Approaches to metropolitan planning were guided, in theory, by the County of Cumberland Planning Scheme, passed into law by the NSW Parliament in 1951. Generally, the County’s boundaries encompassed the greater Sydney metropolitan region of the time. Among key problems, the plan nominated “over-centralisation and congestion of industry; congested and confused traffic; slum housing.” These were expected to be made worse by a predicted rise in the County’s population from 1,702,000 in 1948 to 2,227,000 in 1972. As it turned out, that forecast was only around half the actual growth rate. This unforeseen surge put paid to some of the much-touted features of the Scheme. The encircling green-belt, for instance, died a death of a thousand cuts. 

Under the plan, people and jobs were to be distributed between a central zone and some twenty dispersed urban districts. Central zone included the CBD and adjoining areas focusing on government and administration, higher education, commerce and industries associated with the port and transport terminals. Urban districts, also referred to as ‘dispersal centres’, ‘growth areas’, and ‘satellite towns’, comprised communities largely self-sufficient in shopping, entertainment, education, culture and amenities, with local industrial areas.

“Much of the over-centralisation and congestion in the inner areas of the County”, said the Scheme’s 1948 report, “are due to [the] radial pattern” of the existing transport system. It was envisaged that central and outer districts would be interconnected by the extension of rail lines and an upgraded road network made up of various classifications of road-links, such as new long distance ‘motorways’ or ‘expressways’. For one writer the County of Cumberland Scheme was “Sydney’s first notion towards decentralisation and the formation of a multi-centred city.”

Denis Winston’s Sydney’s Great Experiment: The Progress of the Cumberland County Plan (1957) remains one of the best accounts of the Scheme. Then Professor of Town and Country Planning at the University of Sydney and former president of the Australian Planning Institute, Winston was an influential voice in town planning circles and fervent advocate of decentralisation. The word was used in a dual sense as intra and extra metropolitan, dispersal to outer metropolitan zones as well as rural centres. He welcomed the plan as a worthy attempt to manage Sydney’s explosive post-war growth. “An important consideration in the County Scheme,” writes Winston, “is that the proposals follow broadly a natural trend … that is to say, the general trend towards decentralization … planned dispersal is therefore merely strengthening a natural and useful trend …”

Only decentralisation, he argued, offered genuine solutions to mounting problems like traffic congestion. “All surveys show that it is the centralization of employment within the County that is the essence of the traffic problem.” The answer was to “provide alternative centres of employment more closely related to the new suburban living areas, and ease congestion at the centre by dispersing the destination points.” Since Australians had taken to motor vehicles with gusto, there were few alternatives: “the number of motor vehicles per thousand of population is now over twice that of the United Kingdom and … commercial vehicles per thousand of population now actually exceed those in the United States.” Peter Spearritt and Christina DeMarco describe the growth of car use in Sydney after 1947 as “spectacular”, rising from 13 per cent of all trips in that year to 47 per cent in 1960 and 72 per cent in 1971.

Dispersal also offered an escape from “the slum areas, chiefly near the centre of Sydney: these included more than 40,000 dwellings which could only be described as unfit for human habitation and a further 80,000 homes which were structurally in bad condition …”

Winston’s views on CBD development followed accordingly:

Generally speaking the higher the building, or rather the greater the total floor space of the building, the greater is its capacity for goods and people and therefore the greater its influence on the traffic in the streets around it. The provision of wider roads and better transport services alone would only enable more people to reach the already overcrowded centre.

The only really effective means for improving traffic conditions is the limitation of floor space … This is the purpose of the device known as the “floor space index” … Action on these lines in Sydney, accompanied by a real decentralization movement … is even more necessary than improved roads and more bridges … The traffic problem of Sydney, as of most great cities, is the problem of too many people and too many goods trying to get in and out of a restricted area … the solution lies in … the careful relation of building bulk to the size and arrangement of the streets and transport services.

This was the standard view of professionals like town planners and city engineers at the time. As Spearritt and DeMarco note of Winston, “many of Australia’s leading town planners and economists agreed with him.” Likewise geographers, including K W Robinson of Sydney University, who observed of the CBD in 1952:

Nowadays the streets are too narrow and activity is too confined to be efficient. The access advantages derived from the peninsular character of the city and its wharfage facilities, are offset by the inadequacy of space for building and transportation. For this reason, the site which was adequate for a gaol in 1788 has become a veritable confusion in modern times. The streets existing in 1821 provided adequate coverage for the area, in terms of the actual space which they occupied. Yet for the same area at present there are many more streets, carrying a volume of traffic out of all proportion to their size. The original pattern remains, but it has been augmented and strained far beyond its capacity.

Winston was considered a pupil of British town planner Sir Patrick Abercrombie and his New Towns movement, which descended from Ebenezer Howard’s Garden City movement. At Winston’s invitation Abercrombie toured Australia in 1948 and the County of Cumberland Scheme was in some limited ways informed by the Greater London Plan of 1944, which he shaped. Influence of the Garden City movement is noticeable in Winston’s claim that decentralisation “can save the County of Cumberland from becoming just another Manchester or Chicago.”

Sydney’s Great Experiment was completed before the decision to lift CBD building heights, which Winston is unlikely to have welcomed. Over time, the NSW Government walked away from features of the County of Cumberland Scheme culminating in abolition of its administrative apparatus in 1963. Urban planners and commentators who came to prominence in the 1960s and 1970s like Hugh Stretton, Peter Harrison, Leonie Sandercock, Max Neutze, Ian Alexander and Patrick Troy were critical of the course Sydney pursued following landmark events of the 1950s. Like Winston, they saw decentralisation as the only sound response to metropolitan problems which began to seem intractable.

From the perspective of today’s urban land value crisis, their warnings were remarkably farsighted. They accused successive NSW governments of failing to remedy over-centralisation and fixating on the CBD while manipulating City of Sydney Council for questionable purposes. 

Australian National University academic Hugh Stretton was voted one of Australia’s ten most influential public intellectuals. First published in 1971, his Ideas for Australian Cities became a widely-read touchstone for urban policy professionals. Stretton opens the chapter about Sydney declaring, with reference to the CBD, “Sydney is a government town.” Along with the “wharves, rails, quay, domain, bridge, opera, town hall”, enough of the “commonwealth and state skyscrapers” are publicly owned “to shape any private structures and movements between them.” Stretton was writing half way through the second phase of Sydney CBD’s unparalleled office development boom of 1968-1974. He claimed “the state concentrates its own offices in the city centre, and does nothing effective to deter others from doing the same.”

The County of Cumberland Scheme, “which had very civilized purposes”, made only modest progress before it was abandoned:

… many of the planners’ ambitions were frustrated. Many of their new roads and most of their railways were not built. The old city centre continues to grow without limit. Little except retailing has gone out to the suburban centres … Sydney’s traffic problems are worse than ever. As an inevitable result of its continuing growth and central concentration, most of its land is dearer and its living and productive costs are probably higher than ever.

Little was done to reorient the city’s transport system beyond radial CBD-centric routes:

… it is into the radial routes – and their further boosting and congestion of the centre – that much of Sydney’s transport investment still goes … it probably helps the [urban districts] to attract more … retailing … But they have little chance of sharing out any of the other central activities of the metropolis. Government might still help to make a cluster work if it distributed plenty of its own offices and the other activities it controls to the [urban districts]. Whatever advice it may have had from its planners, the New South Wales government so far concentrates jobs and customers’ traffic into the single old centre more resolutely than any other Australian or American state government does [emphasis added].

In 1966 property economist T F M Whipple had found that 40 per cent of CBD office space in privately owned buildings was occupied by government agencies. Ten years later the Sydney CBD Survey (1976) found that this was the case for 38 per cent of total office space. Even so, the NSW Government had established a stand-alone Department of Decentralisation in 1966 and its State Planning Authority issued a successor plan to the County of Cumberland Scheme in 1968, the Sydney Region Outline Plan 1970-2000. Both made the right noises but were met with official indifference or resistance, as Stretton explained:

… the planners agonize often and perceptively about the problems and costs of over-centralisation. But with one problem after another, that is all they can do. They condemn the unplanned, uncontrolled growth of the city centre. More central employments ‘ought’ to disperse to suburban centres or to other cities. Among devices for achieving this the government might impose building limits or punitive taxes on the existing centre, move some of its own activities out … or even push some growth towards other states. The last is of course unthinkable.

“In [Max] Neutze’s words,” writes Stretton, “decentralisation was everyone’s policy but no one’s program.” Since “nothing new [was] proposed to beat the old centre’s gravity”, Sydney could “merely expect pressures, congestions, rocketing land prices.”

In her accounts of Australia’s erratic forays into metropolitan planning, Cities for Sale (1975) and Property, Politics and Urban Planning (1990), Leonie Sandercock observed that for NSW politics in the late 1940s, “decentralization had been an essential part of the vision of a better post-war world.” Her highest concerns included “the failure of efforts to decentralize and to control land prices”. Sandercock too attributed collapse of the County of Cumberland Scheme to poor follow-through on transport projects, but also to “most local councils [which] were reluctant to complete local planning schemes, preferring to exercise … discretionary control over private developments …”

Prominent amongst these was Sydney City Council, under which “the old city centre continued to grow without limit, increasing traffic problems and land prices.” Sandercock called it “Manhattan Syndrome”. Overall, “there was some improvement in the living and working relationship in regard to manufacturing and retailing employment, [but] there was no restraint on the build-up of employment in central and inner areas, which accounted for 37 per cent and 18 per cent of metropolitan employment in 1966.”

Present-day urbanists maintain that action to decentralise office work would have been futile in any event. They argue it is in the nature of office-based occupations to concentrate in the CBD, invoking what may be called ‘the CBD agglomeration myth’. Simon Kuestenmacher, demographer and writer for The Australian newspaper, restated the myth in a recent column:

Why did our cities feature dominant CBDs in the first place? As our economy transitioned from manufacturing towards knowledge work, office jobs made an ever-growing share of the workforce. But why did these jobs cluster in the CBD? Why weren’t they spread out evenly across the city? To some degree zoning practices are responsible but mostly it was the nature of these new office jobs that led to the clustering effect. A knowledge job benefits from being co-located with other knowledge jobs.

In contrast, geographers have traced the origins of CBD locations to landforms, physical features and associated convergence points in the era of fixed-route maritime and railway transportation. As George W Hartman put it in “The Central Business District: A Study in Urban Geography” (1950):

The first substantial developments of “towns and cities” in the United States generally were made at locations most advantageous to trade and commerce … at a junction of overland trade routes, along a navigable stream or water body, at some other inland break-of-bulk or change-of-ownership transportation site, or at some strategic point favourable for mining, manufacturing, or resort activities …

Raymond Vernon explained in Metropolis 1985 (1963), his influential interpretation of the 1956-59 New York Metropolitan Region Study, that:

We have already indicated the historical process by which the great cluster of wholesalers, central offices, banks, and related activities come to be located in the New York Metropolitan Region. The Port begat the wholesalers; these strengthened the pull upon the financial institutions; finance attracted central offices; the latter drew the advertising agencies and others; and each new accretion reinforced the old …

This was certainly true of the harbour city of Sydney. “As the major port in the State”, writes geographer Maurice Daly, author of the classic Sydney boom, Sydney bust (1982), “Sydney provided the transport, storage, financial, and commercial systems whereby the products of the interior could be delivered to the world, and the conjunction of the port with the central business district implied a tight clustering of such activities in the CBD.” In urban economics textbooks like that of Arthur Sullivan, explanations for the development of office districts at such sites emphasised transportation costs. Office-based activity “has higher transportation costs because it uses people – with high opportunity costs – to transmit output … the office industry, with its higher transportation costs, occupies land closest to the city centre.” CBD researchers had found that, allowing for the retail core, the office core was traditionally as close as possible to the most accessible point inside the CBD (technically known as the Peak Land Value Intersection).

But things changed, as Daly points out, when “better transport systems removed the need to be adjacent to the port for warehousing and wholesaling [and] the motor vehicle took population, manufacturing and retailing into the suburbs.” Mention can also be made of containerised freight and air transport. Geographer Jean Gottmann, ironically a progenitor of the CBD agglomeration myth, identified the emergence of urban areas dominated by office work untethered to fixed-location landforms, infrastructure or resources. In The Coming of the Transactional City (1983), Gottmann wrote that

… the expanding activities in the cities have shifted from manufacturing work and distribution to transactional work and related services. That means, in terms of buildings, a transfer of emphasis from industrial plants to offices … I believe that it has announced the liberation from constraining work on the land, at the machines … In these new ways of life the individual expects to obtain greater freedom in arranging his own conditions of living, the use of time, and the use of space.

In these circumstances the concept of “typical central business uses” was no longer as meaningful as it seemed to American CBD researchers Raymond Murphy and James Vance in the 1950s. Hans Carol of the University of Cincinnati argued in 1960 that their handling of the concept was, in any event, undeveloped and failed to “follow [the] line of thought further as a theoretical or practical basis for the decision as to what are and are not CBD functions.” Carol came up with the more rigorous concept of “absolute centrality”, applying to urban functions or uses which “serve the city as a whole” (entire metropolitan region) and only exist for that specific purpose. By the late 1970s, however, very few, if any, uses met this definition. 

One flaw of the CBD agglomeration myth is that the location of office supply is superficially equated with the location of office demand. One of the urban policy academics who came to prominence in the 1970s, Ian Alexander of the University of Western Australia, wrote extensively on Sydney’s pattern of office location. His article “Office Decentralisation in Sydney” (1978) recognises that “while other economic activities have shown an increasing degree of dispersal away from their traditional central locations, offices have remained highly centralised.” Unlike in the United States, “where the majority of new office development in many urban areas [was] occurring in the suburbs”, explains Alexander, “in Sydney, the central area accounted for more than 75 per cent of the value of new office buildings over the 1960-75 period.” These years correspond to both stages of the CBD’s pumped-up redevelopment boom. 

Alexander did detect “some signs that this traditional centralisation is beginning to break down.” Around 100 firms moved to suburban centres between 1965 and 1975 while a further 120 or so commenced operations in suburban centres. This represented a very modest trend, however, since these relocating firms accounted for only around 2,000 jobs or 1 per cent of the central area’s 1971 office employment. By comparison, London diverted 115,000 jobs away from the central area over the decade to 1975, equivalent to 15 per cent of central London’s office employment in 1971.

Based on a survey of factors causing the relocation of Sydney office firms in 1975, Alexander found that the decision was often “heavily influenced by internal factors” and “factors relating to the evolution of the firm [rather] than from any revaluation of locational needs.” Highly ranked factors leading to relocation included “no room for expansion”, “high cost of existing location” and “reorganisation of operations”. On professional and business service firms, commonly assumed to need a CBD location, “interviews indicate that the actual intensity of contact with [their] clients is not necessarily strong” and “most professional firms in outer suburban centres thought such [a CBD] advantage was largely mythical.” Alexander stresses “the importance of non-quantifiable factors in the office location process.”

A second survey ranking location factors for offices in CBD and suburban locations found that “availability of premises” ranked highest for CBD offices, well above “access to associated business” and “access to business services”, which ranked eighth and ninth respectively. Alexander comments that

[t]his points to the process of office supply leading demand rather than simply responding to it. While office developers claim to be doing the latter, it is clear that the pattern of vacancies at any one time exerts a strong influence on the pattern of office dispersal. This influence is accentuated in times of oversupply like that prevailing at the time of writing. There is little doubt that in Sydney office developers are playing a greater part in the office location process than are the planners … the developers have virtually had a free hand in city redevelopment, and office building in the centre proceeded at boom rates for several years from the mid-1960s. This is one reason for the vast imbalance between supply and demand evident at the present time.

Offices weren’t necessarily built where firms wanted to locate but firms necessarily located where offices were built. Max Neutze, an urban policy academic at the Australian National University in the 1970s, had similar views on the agenda-setting role of property investors and developers:

While few developers have much influence on the pattern of new development the options in the case of redevelopment are a good deal wider … Central city office redevelopment has frequently been initiated by investors and developers. Since the office space is so highly localised these redevelopments have consolidated the pattern of development rather than changing it.

This was a serious problem in the case of “Sydney [which] stands at one extreme in the urban pattern because … its [CBD] site is more confined than that of any other mainland capital … [and] … its city centre is far from the geographical centre of the urban [metropolitan] area”.

Alexander maintains “there is evidence to suggest that the pattern of relocation to inner centres may be as much a product of office space availability and imperfect market knowledge as of any real inability to function in outer centres.” He thought government offices should have been deconcentrated and “developers need[ed] to be encouraged to invest less in the central area and more in the suburbs, so as to improve office availability in non-central locations”. That was not an unusual proposal at the time. Town planner R W Archer found in 1969 that the “efficiency” of Sydney CBD was severely compromised by “failure to prevent overdevelopment of office sites.” But Alexander felt that the types of measures taken in the United Kingdom, including office development permits, would “probably never be politically acceptable given the powerful financial interests that are involved in office development.” 

Later, Alexander explored the problems of office location across English-speaking countries in a full-length book, Office Location and Public Policy (1979). He confronts the CBD agglomeration myth head on. Noting that “the classic explanation for the clustering of office activities in city centres” has been “need for direct personal contact in transacting business”, he continues:

Although such beliefs still persist strongly, it will be shown … that there is now considerable evidence to show that, given information flows are an important locational consideration for office activities, they are by no means an absolute constraint upon location.

Alexander proceeds to cite a wave of research from Sweden and Britain on “developments in communications technology that allow the replacement of face-to-face meetings by other means” and the classification of business contacts as routine “programmed” contacts or more complex “orientation” contacts. The former could be transacted as easily from a decentralised location while the latter “form only a small proportion of a typical office firm’s contact patterns” and “only involve a minority of office personnel.” Orientation contacts could also be conducted from a non-central base even if extra travel is involved. Meetings “involving discussion, problem solving, and information seeking could be carried out using relatively simple telecommunications.”

Suburbanisation trends over the 1960s and 1970s demonstrated that this applies as much to “detached” head offices as it does to branch offices or plant-attached offices. “[E]ven the largest headquarters can fulfil their need for personal contacts from locations well outside the CBD.” Silicon Valley is a case in point. Such findings would have been confirmed by arrival of the internet and more advanced mobile communications technologies, not to mention work practices mainstreamed during the covid pandemic (in Work From Home highly qualified and ambitious corporate ‘knowledge workers’ felt they could be just as productive away from office centres, debunking the CBD agglomeration myth).

At one point Alexander revisits Murphy and Vance, quoting a researcher who thought their idea of central business uses “rested primarily on inference … [their] work concentrated too much on pattern and too little on process.” Proponents of the centralisation hypothesis relied on scant “proof other than revealed preference”, which was “overtaken by events, since the days when all office activities were automatically drawn to a central location are long since gone.” If decentralisation was not more pronounced, it was because “locational inertia [kept] them in a central location” and, importantly, “some firms – particularly those in the insurance and banking sector – have invested heavily in central office building.” State planning and transportation policy was a factor in locational inertia:  

Australian cities have shown a high degree of suburbanisation of population and economic activity … However, office developers and detached office activities have shown a particular propensity to locate in the central areas. Part of the explanation for this difference is that the comprehensive freeway networks in US cities, that have been a major agent encouraging office relocation, are absent from Australian cities.

In the book, Alexander expands on the office-occupier survey results presented in his 1978 article:

As is the case elsewhere, most of Sydney firms surveyed rent rather than own premises. This makes them highly dependent on the availability of premises to rent … 74 per cent of offices regard this as an important location factor … the particular location chosen is often affected by where suitable premises in terms of costs and size can be found.

Developers supposedly build on the basis of expected demand. But much development is highly speculative, and is often only based on sketchy analysis of the likely market for space. The all too frequent emergence of chronic oversupplies of office space in central areas is evidence of this.

Stretton was in no doubt that Sydney’s office centralisation was an orchestrated phenomenon:

… rising rates of destruction and renewal are made profitable by the land values of overbuilt office centres. Nor are such policies adapting democratically to ‘the peoples’ locational decisions: they work in favour of the interests of tiny minorities of central investors and developers … Plenty of Sydney people – planners, politicians and others – like to regard their centralized over-growth as natural, unavoidable, an international phenomenon. It is nothing of the sort. Sensible schemes to limit and disperse central office building have been proposed since 1953 – and regularly murdered by politicians and central investors. For nearly half a century London has limited its central office development to less than half the densities Sydney was building throughout the 1960s.

After discussing features of the 1970s office over-supply, later examined in more detail by Daly, Stretton repeats:

So the celebrated office boom at Sydney Cove was not a response to strong office demand from people who needed to do business there. It was a calculated risk by the office builders, gambling on inflation and rising building costs and especially on the extreme centralist policy which was visible in the government’s own office activity … [emphasis added]

Looking back at the County of Cumberland Scheme and the early 1950s, Daly observed that “the falling number of workers in the CBD seemed natural and inevitable, and the CBD was viewed increasingly as simply the largest in the hierarchy of centres in the metropolitan area, and one which would continue to decline in importance.” That it followed a different course to renewed metropolitan primacy can be ascribed to political choices which made central office development highly profitable for the property-finance industry. Since the 1970s, of course, the pace of office dispersal to suburban locations has picked up substantially, but the legacy of over-centralisation in metropolitan planning persists. Hoisted on fabricated props, the CBD would come to need an ongoing program of preferential treatment, with adverse consequences for congestion and housing affordability.

This is an edited version of Chapter 10 from our feature report Rise of Luxury Urbanity as a System: Sydney CBD. Sources for quotations in the above article are footnoted in the report.

Our main website: The New City