Peter Hall
Special Advisor to the Secreatary of State
for the Environment and Strategic Planning, Britain,
and Professor of Planning at the University College London
CLICK FOR BIO

Madrid, Spain
Monday, 3 May 1999

I am deeply honoured to have this opportunity to address the conference. There could be no better opportunity than now to discuss the problems of global-level cities and the super projects that have come to be associated with them. We are coming to the end of a decade that has been marked by the impacts of globalization upon cities, accompanied by urban super projects on a scale never before seen. We have seen some of these projects truncated and threatened by the global financial crisis of the late 1990s, which has been both triggered by urban real estate investment and has, rather remarkably, in turn been a major precipitating cause. Now, with the worst of the crisis apparently behind us but with cities and financial institutions undergoing sober reappraisals of their position, is a time for analysis and reflection, which will help us face the challenges that the growth of these cities will present in the coming century. I would like to help make a modest contribution to that process.

The Driving Forces: Globalization and "Informationalization"

Underlying the entire process is a profound set of changes in the operation and geography of the world economy. We call it by the shorthand word "globalization", but it is more than that. It consists of two separate processes:

First, there is globalization: this is itself a complex process, but it has at least two components that are critical for the economies of cities. One is the breaking down of trade barriers; the other, associated with it, is the increasing global reach of large corporations, in both the manufacturing and in the service economy. Together, these produce a world in which the location of production, again in both manufacturing and services, can be shifted to the location or locations where the return on investment is highest. This creates a new worldwide calculus of lowest-cost, highest-return location. In many cases the rational location will be the lowest-cost one, but this must always be the lowest-cost location where production can be efficiently maintained, especially in manufacturing. In other cases, especially advanced services, the rational location is the one that will effectively deliver the product, which often is not the low-cost location.

Second, there is what we can only call tertiarization or, an even uglier word, informationalization: a progressive shift in advanced economies away from manufacturing and goods-handling, and towards employment and production in services. This is partly a result of globalization, which shifts manufacturing to lower-cost locations, but it is also of the search for increasing efficiency in the remaining production facilities, whereby firms substitute capital for high-cost labour. Because until now productivity gains have been easier to obtain in goods-handling than in services, this results in rapid contraction in employment and a slower fall in value of production. It appears however that the same effects may be beginning to occur in service production also, as more routine functions are transferred from larger higher-order cities to smaller lower-order ones, and the remaining non-routine ones have very high value added.

As a result of these two processes, we see the emergence of a new worldwide system of cities, based not so much on what is produced but on the level at which it is produced. Underlying this at a global scale is still the division between advanced cities providing advanced higher-order services, and developing cities still engaged in manufacturing and in more routine service delivery. But the distinctions are more subtle than that: in the largest and most advanced cities we find an increasing concentration on a relatively small group of information-handling services, which consist of finance and business services; command and control functions (including government and corporate headquarters); tourism, both of the leisure and the business varieties; what can be called the creative and cultural industries, including the performing arts and the print and electronic media; and specialized services such as advanced healthcare, higher education, and research and development. All these become essentially the economic bases of global super-cities.

A New Urban Hierarchy

There is thus a global hierarchy of cities. At the very top is a very small group of acknowledged global cities, certainly including London, New York and Tokyo in a 24-hour system of financial exchanges. But immediately below this level is a much larger group of contending cities, which are variously called sub-global or alpha locations, and which essentially compete to perform more specialized niche functions such as serving specific geographical markets (thus Singapore, Hong Kong, Shanghai and Sydney for the western Pacific Rim) or specific specialized types of service (Milan for fashion, Paris for artistic tourism, Los Angeles for the media). In some of these latter cases one city has absolute domination; in many others, several cities vigorously compete, especially within the same geographic zone (Milan and Paris for fashion; San Francisco and Los Angeles for Pacific Rim banking).

These cities survive, and even enhance their position, because despite the vast growth of telecommunications traffic and the birth of informatics, very much information is still exchanged face to face through personal contact. Three obvious measures of this fact are the growth of business air and high-speed train traffic, the growth of the hotel industry and the burgeoning of the conference and convention industry of which this is a particularly notable example. In fact, research shows fairly clearly that over a long period the growth of telecommunications traffic and personal inter-city traffic have marched hand in hand. This should not be surprising: as the British researcher John Goddard discovered years ago, telecommunications are used much of the time for preliminary routine encounters, leading to a positive need for the more complex kind of exchange for which face-to-face contact is deemed essential.

Because of this, cities competing for a position at the summit of the global hierarchy of cities need not only top-quality telecommunications, but also fast and frequent and direct linkages to other cities, by air for longer-distance linkages, or - an increasingly significant factor in Europe and in Japan - by high-speed train for medium-distance journeys of up to about 800 kilometres. Of course, there is a circular and cumulative process here: a high position in the hierarchy naturally expands the range and frequency of services from the local airport as well as the frequency of train services. But there is an element of serendipity and luck in this process. Within Europe, a small capital city like Amsterdam or Brussels or Copenhagen will benefit from the fact that its airport serves the national hub airline. A intermediate city on a high-speed rail line built to connect distant cities, such as Hannover or Stuttgart in Germany, will benefit from the fact.

As cities expand their range and volume of high-end service functions, these naturally concentrate in the traditional central business district. The principle of agglomeration for face-to-face contact still applies as powerfully today as when the American economist Robert Murray Haig enunciated it in a classic paper of 1926. And it is not just the fact of short distances, but the huge supporting network of ancillary services, ranging from law and accountancy and advertising to sandwich shops and bars. Further, the different specialized functions themselves have a powerful degree of synergy: cultural and creative activities serve tourism, finance and business services support business tourism, and command and control functions support advertising and public relations which are linked to the media industries.

A New Urban Form: The Rise of the Polycentric City Region

Typically, since early times, these occupied specialized quarters or sections of the central area: the City of London, the Inns of Court next door to it, Whitehall, were all well established by the 16th century. The same principle can be seen in the division between Downtown and Midtown in New York, or between Otemachi, Marunouchi and Kasumigaseki in Tokyo. However, since World War Two the growth of service functions has brought about a much more complex and dynamic spatial form. In many cities we can distinguish between an original commercial core, generally several centuries old, and a newer core that typically developed in the former high-class residential area, either shortly before or shortly after World War II. But in addition, there has been the deliberate development of major commercial sub-centres outside either of these traditional cores: La Défense in Paris, Shinjuku in Tokyo, Canary Wharf in London and most recently Potsdamer Platz in Berlin.

New high-speed train stations may sometimes serve as the nodes for such developments, as at Part-Dieu in Lyon or Shin-Osaka in Osaka. Beyond these, specialized sub-centres may develop to serve particular functions such as tourism and entertainment, such as the Tokyo or Osaka waterfronts. And, of course, major airports develop as powerful concentrations of activity in their own right, not only for travel but also for associated commercial functions, as can be seen around London Heathrow or Paris-Charles de Gaulle or Stockholm Arlanda.

Thus the activity pattern of almost every city tends to become steadily more polycentric. The old monocentric pattern, where all transport infrastructure converged on a single tight centre, is replaced by a more complex network with multiple nodes. Tokyo, where the original multiple core around Tokyo station has now been powerfully supplemented by the ring of centres around the Yamanote ring rail line - Shinagawa, Shibuya, Shinjuku, Ikebukuro and Ueno - is the classic case: Shinjuku alone has the same daytime employment, one million, as the whole of central London. But all cities are going the same way, and some of the larger cities provide very dramatic examples: the Paulista in Sao Paulo, the Kowloon Waterfront or Pudong in Shanghai.

In the largest global city, however, this polycentric structure goes even further. In effect, the city becomes the core of a huge city region, which some Asian urban experts have termed a mega-city. This, I should point out, is one of those difficult terms because different experts use it to mean different concepts. The mega-city in the Asian sense is a cluster of cities and towns of different sizes, intensely networked, stretching over a vast territory of several thousand square kilometres, and collectively containing up to 30 million people. The classic examples are the Pearl River delta of China, with a total land area of 45,026 km², containing Hong Kong, Guangzhou, seven medium-sized cities and 13 xians or counties, plus two out of four of China's special zones, Shenzhen and Zhuhai, and with 22.1 million people in 1993; the lower Yangtze region in China around Shanghai; the Tokaido megalopolis in Japan, stretching 500 kilometres from Tokyo to Osaka and containing some 60 million people; and the emerging Singapore-Johor Baharu axis between Singapore and Malaysia.

Such complexes occur elsewhere in the world, as in South East England, Randstad Holland, New York-New Jersey and Southern California. In every case, different cities or districts within the complex perform distinct functions, with high-level services in the traditional core (though more weakly so in Los Angeles), more routine services and manufacturing in more distant centres. However, the most significant, represented by Singapore-Johor Baharu and by the Pearl River Delta before 1997, are those which straddle international frontiers and thus embrace separate regions or countries at different stages of economic development. For, as well documented for the Pearl River delta, such mega-cities develop an internal division of labour, with the core city increasingly concentrating on the high-level service functions which connect the entire region with the wider world, and the other parts concentrating on the manufacturing and goods-handling functions which the core city has given up. By the early 1990s, for instance, manufacturing employed less than 20 per cent of the Hong Kong workforce. Further, the entire complex is highly networked through flows of capital and expertise from the core city, which in effect provides a set of triggering mechanisms for the smooth operation of the entire complex. The skills in manufacturing and marketing, which were formerly monopolized in the core city, are effectively exported through a kind of internal colonization. It is an extremely effective form of regional economic development, as exemplified by the dramatic growth of the entire Pearl River delta region in the 1980s and 1990s.

The Urban Megaproject: An Asian Phenomenon?

One result, most dramatically illustrated in the Eastern Asian cases, has been the planned promotion of entirely new cities or city districts, often in the form of "urban megaprojects" involving huge flows of foreign direct investment. These megaprojects, as the geographer Kris Olds defines them, are

      ... large-scale (re-)development projects composed of a mix of commercial, residential, retial, industrial, leisure, and infrastructure uses. They are developed primarily in the inner city, on large tracts of former port, railway, industrial, military, or racetrack lands, or on `underutilized' suburban, agricultural, swamp, or island land within the extended metropolitan region (Olds 1995, 1713).

Olds identified over three dozen such megaprojects under development worldwide in the mid-1990s. They tend to resemble each other, for they are invariably developed by the same professional groups in pursuit of the same internationalization strategies; marketed to overseas firms and rich individuals for lease or purchase, they are designed as 21st-century global `utopias'. They became especially common in the Asia Pacific Rim because of the economic boom there in the early 1990s. Here, as in North America and Australia, they include major development of seaports and airports; high-technology office districts including teleports, often acting as nodes in huge development corridors for the exchange of goods, people and information; and linked luxury residential districts for the elite.

No less than half her examples, eighteen of them, are in the Pacific Rim: two in Canada (Vancouver), one in the USA (Mission Bay), four in Australia (Sydney, two in Melbourne, one in Adelaide), one in Singapore, two in Malaysia (Johor Baru and Kuala Lumpur), one in the Philippines (Manila), one in Thailand (Bangkok), one in China (Lujiazui), and five in Japan (Tokyo, Yokohama and three in Osaka); oddly, she excludes the SEZs. Of course, as everyone in this conference knows, many of these developments ran into more or less severe difficulties at just about the time she was writing.

Anatomy of an Urban Real Estate Boom - and Crash

The World Bank's Bertrand Renaud argues that the late 1980s saw the first "global real estate cycle"; after it peaked in 1990, the succeeding crash brought big impacts for the real estate industry, the financial system, the household sector, and the overall economy, which were still being felt in many countries in the late 1990s. Nowhere was this more so than in Japan. For a dominant factor in the saga was Japanese foreign investment: Japan's share of net long-term capital outflows among the world's leading capital exporters (the G-7 countries plus traditional capital exporters Denmark, The Netherlands, Switzerland and Saudi Arabia) rose from one quarter in 1982 to nearly 90 per cent in 1987, and Japan had become the biggest net creditor nation on record.

These massive capital flows, Renaud shows, were directly linked to the land bubble. The trigger was the Plaza Accord and accompanying Louvre Agreement of 1985, which realigned dollar-yen exchange rates and brought Japanese central bank discount rates as low as 2.5 per cent. This led to a huge flow from banks and other institutions to real estate companies, raising their share from a stable 7 per cent (maintained over two decades 1965-85) to 17 per cent in 1990. Renaud concludes, and I quote: "The combination of low interest rates and abundant liquidity favored real estate investments and impacted most sharply on the inelastic land supply to generate accelerating land price acceleration" (Renaud 1997, 16).

But this syndrome was soon exported from Japan to neighbouring countries. Within Japan itself, the supply of land was restricted not only by the facts of geography but by urban development regulations and fiscal arrangements that encouraged land hoarding. To complicate the situation further, Japanese tax treatment of real estate gains was very favourable. In Japan, increases in the price of land held by corporations meant a rise in the value of this asset on their balance sheet, and these unrealised capital gains lead to easier access to bank loans, which could be used not only for new productive investment, but for more speculative real estate investment. And this real estate appreciation then played a critical role in Japanese net capital outflows. First, banks lent directly or indirectly to real estate companies. Second, Japanese banks traditionally relied on real estate collateral rather than project quality or cash flow to make loans; Japanese firms used this to invest abroad, not only in portfolio investments, but also in direct real estate investments.

But all this ended abruptly in 1990 when the new governor of the Bank of Japan, Yasushi Mieno, raised the discount rate five times, from 2.5 to 6.0 per cent. Japanese real estate prices declined sharply, and Japanese banks faced an unprecedented increase in nonconforming loans. Renaud comments that "nonperforming real estate loans have been at the epicenter of banking problems in every country since 1990" (Renaud 1997, 30). But the problems essentially originated in Japan, and have returned there in the form of the bank failures of the mid-1990s.

This I think is a cautionary story. Economic history shows that real estate development follows the business cycle but also massively amplifies it. Sixty years ago Joseph Schumpeter suggested in his classic study of business cycles that every 55 years or so a massive speculative real estate bubble presages a crash in the stock market; later, Simon Kuznets suggested that the real estate cycle was shorter than this, 20-25 years, and still more recently Brian Berry has proposed that two Kuznets cycles nest within each of Schumpeter's long waves, with a boom-crash cycle occurring at the end of the upswing phase and towards the base of the downswing. Whatever the truth, history provides plenty of evidence of real estate crashes, as in America during the 1930s, when the Empire State Building was locally known as the Empty State, and when multiple bank failures took place. What is almost certainly new about the 1990s is the scale of the global relationships and the way that massive capital outflows into speculative real estate ventures can destabilize banking systems in one of the leading economies of the world.

The Challenge: Urban Organization and Urban Finance

The pace of growth in these mega-cities, almost unprecedented in world history, brings huge problems. The new global advanced-service functions lead to escalating rent levels, which squeeze out other activities and put pressure on housing land, in turn pushing new housing construction either higher and higher, denser and denser near the centre, as in Hong Kong and to some degree in Singapore; and farther and farther out, as in Tokyo. And in turn, the resulting commuting times - up to 2 hours each way each day in the case of Tokyo - provide a classic case of negative urban externalities. Another consequence is growing and generalized traffic congestion across the entire metropolitan region, only partially countered by mass transit construction and traffic restraint.

The Asia Development Bank's 1996 report reviews these problems of the late 20th-century megacities: affordable housing, traffic congestion, waste disposal, pollution, social dislocation. It ends with a warning:

Without better intervention to address these problems, megacities will become even more congested, polluted, unhealthy, expensive and socially divided. A downward spiral will follow if inward investment is deterred: the megacity's comparative disadvantage will diminish, with correspondingly fewer resources to manage the growing problems (Asian Development Bank 1997, 35).

The problem is that in too many cities, the delivery mechanisms are failing because resources are lacking. Between $20 bn. and $40 bn. would be needed annually to provide services sufficient to sustain productivity and achieve improvements in the quality of life, the Bank believes. The ADB's answer is for local governments to move from directly providing services such as water and public transport to facilitating provision by others, by means of public-private partnerships, and phased provision of infrastructure with well-conceived strategic plans which can then be executed with the aid of private capital.

In achieving such results, cities in different parts of the rapidly-developing world have different bundles of assets and liabilities. Asian cities, perhaps because some belong to colonial and post-colonial regimes, perhaps because of a great sense of collective welfare, have often been in the forefront worldwide in developing solutions such as traffic restraint; Singapore's electronic road pricing system, introduced in 1998, was the world's first. But these solutions work better in closely-bounded high-density cities, like Singapore and Hong Kong, than in sprawling polycentric edge cities such as those characterizing much of Latin America. And they depend on a willingness to suppress individual preferences to the wider collective interest, which may be distinctively Asian or Confucian. Latin American urban societies are more individualistic; their great strength in the last two decades, related to the triumph of democracy in that continent, has been the decentralization of power from the national centre to the great cities. But sometimes this has not been accompanied by equivalent transfers of resources, and in any case the transfer is compounded by the complexity of local government structures in and around many of these cities, which leaves many small and ill-equipped peripheral municipalities facing the consequences of rapid urbanization.

The danger is that in the scramble for growth, these newest members of the world great city club will neglect their responsibility to protect and enhance the quality of life. If so, they will surely regret it. For, faced with rising costs, not least the costs of land, global investors can all too easily remove to the less-developed city down the road. They will be most unlikely to pay for the social costs they have imposed. And, particularly in the early stages of development - as in the Pearl River delta at the present time - local governments have neither the resources nor the professional competence to do much about the problem. This is the central dilemma for the new mega-cities of the rapidly developing, rapidly urbanizing world.

REFERENCES

Asian Development Bank (1997) Annual Report 1996. Manila: ADB.

Olds, K. (1995) Globalization and the Production of New Urban Spaces: Pacific Rim Megaprojects
     in the Late 20th Century. Environment and Planning A, 27, 1713-1743.

Renaud, B. (1997) The 1985 to 1994 Global Real Estate Cycle: An Overview. Journal of Real
     Estate Literature, 5, 13-44.