Mondialisation : opportunités urbaines ?*
Article Index

The contemporary globalized economy is based on permanent innovation and ever-growing flows of goods, investments and talented people. Some large cities offer an urban fabric perfect for globalization's imperatives, and their built environment and resources allow them to attract international investment and grow richer. In this globalized game, new large metropolises are being born - and older ones are revitalized - while smaller cities with less to offer find themselves marginalized.

* This paper is an updated version of a keynote address given by the author at the Leverhulme International Symposium at the London School of Economics in 2004.


In his description of the nineteenth-century origins of modernity, the French historian Fernand Braudel described a great rivalry between the city and the state. Three centuries later, the state's victory is undeniable: except for some city-states such as Singapore and Dubai, nation-states are still the primary global actors, despite the effects of globalization. However, nation-states' social, political and geographic primacy could change, due to the undoubted trend towards an ever-stronger urban concentration of wealth and power (as well as poverty and despair), especially in the largest cities. One of the lessons of the nineteenth century, affirmed for the most part by the last decades of the twentieth, is that decreasing the costs of communication does not lead to a wider distribution of wealth and power. On the contrary, it leads to their polarisation. In a world where goods, information and, to a certain extent, people circulate ever faster and more freely, the concentration of spatial inequalities increases. This is a worldwide trend. In Europe and in America, cities' economies have long been integrated into State-managed structures, particularly in Europe's omnipresent welfare states. In Africa and Asia, rapid urban growth is creating massive territorial inequalities that will deeply affect political circumstances. By 2015, twelve of the twenty largest cities in the world will be in Asia, and the majority of the two billion additions to the world's population will live in Asian cities. It is fascinating to compare the economic weight of cities and nations: the GDP of Tokyo is twice that of Brazil, and the GDP of the Kansai region in Japan is higher than that of Spain. Very different rationales produce waves of urbanization in Lagos or London, Shanghai or Tokyo, Mumbai or Paris, so our observations will be about cities in the developed world that are asserting themselves as suitable ecosystems for the economy and advanced technology.


The polarisation of wealth and power can be explained simply. When communication costs are high or very high, as was the case before the arrival of railroads, the world was composed of separate, closed compartments that reduced competition between firms. Spatial separation created monopoly-privilege rent-seeking and prevented economies of both scale and agglomeration for consumers and producers. Conversely, when the flow of goods and information improves, the positive effects of economies of scale and agglomeration appear in higher returns and agglomeration externalities. Cities' growth processes show the considerable and most certainly underestimated power of technological and pecuniary (i.e. tied to price mechanisms) agglomeration externalities.


These basic agglomeration effects are relevant throughout many periods in history. This system is strongly reinforced by the specific characteristics of global competition and by the transition from price-based competition to a more complex trade pattern of non-price-based competition - one where quality, diversity and innovation in goods and services become the key factors of economic survival. Traditional distinctions between price-based competition and non-priced-based competition prove less and less evident in all but niche markets. Organizations that compete globally generally do not have a choice about competing for on price or on product/service differentiation, because they must compete for on both simultaneously. These changes in competitive profile result from the breaking down of traditional barriers that protected national oligopolies. The chief discovery, explicit or more frequently implicit, that organizations make concerns the strong feedback loop between quality and innovation-based competition on one hand, and agglomeration externalities on the other hand. Three driving forces reinforce the relationship: the increased mobility of differentiation; innovation and quality-based competition; and urban network effects.


Traditional competition, where natural and national barriers played a central role, is space-based by nature. The new world, where cities become leading economic actors, is a world of time-based competition. If an organization is less protected by space, it has to be faster and more responsive; it has to learn the rules of a new, more open game, and do so better and more quickly than in the past. The rules governing the distribution of innovations in production processes or goods illustrate this trend. There is no longer any place for the international product life cycle that the American economist, Raymond Vernon, described in the 1960s, where new goods and processes slowly migrated from the developed country where the product or process was invented (the United States) to Europe and then to developing countries as their level of standardisation increased. Today, new goods and production processes appear nearly simultaneously, although unequally, throughout the world.


These observations call forth two comments. The first addresses the famous controversy between economists about the effects of globalization and of technology on the changes that are underway in our societies, particularly labour markets. On one side, the American economist Paul Krugman and the majority of his colleagues criticize the habitual over-estimation of the effects of internationalization. On the other side are those who focus on international trade and worldwide restructuring. Observing the phenomenon at work in organizations shows that there is a very tightly linked interaction between globalized competition and technological innovation - and not only for so-called "defensive" innovation in trade between developed and developing countries (Thoenig and Verdier 2003, 19-32; Wood 2004). Large cities' economies play essential roles in this interaction, as places where experiments in both production and consumption can take place at the same time.


The second comment is that the classic distinction between new, emerging industries and mature industries should be reduced. The current interdependence between globalization and innovation affects nearly all activities. Old industries such as steel and automobile manufacturing have to innovate constantly, just as do new industries such as multimedia and biotechnology. Economic history shows the crucial impact of the territorial matrix during a new industry's initial development phases, due to tacit knowledge and other analogous externalities. That remains true, but very large cities or "megacities" are not simply enormous clusters of technology parks, even if they often include them. The relationship between the metropolitan environment, quality and/or innovation-based competition and globalization defines a larger framework that includes manufacturing activities and mature service industries, inasmuch as they are involved in international competition.


Cities' Economies as "Schumpeterian hubs"


It is not possible to review the details of the economic mechanisms that underpin the growth of very large or even second-tier cities now expanding rapidly in Europe. Nor is it possible to list the many externalities related to these mechanisms: excellent work has been done in that field elsewhere (Fujita and Thisse 2002). There are many familiar "mantras" that do not really explain what is happening today and can therefore get in the way of a deeper analysis. Such is the case, for example, with the well-known concept of the "knowledge-based city." What does the term mean? Wasn't seventeenth- and eighteenth-century Paris, with its dense network of highly qualified luxury goods manufacturers and trades, already an urban economy based on knowledge and apprenticeship?


In today's framework, modern developed cities reflect the Schumpeterian dimension of advanced economies and the systematic and relational nature of efficiency. To understand the current context, look at the dynamic aspects of agglomeration economies. Cities offer not only complementary assets, such as the inputs and outputs of a local production system's structure and constant coordination between economic actors, but also the possibility of quickly and effectively reorganizing networks and value chains. Cities are powerful hubs that allow chains between producers, consumers and other social participants (such as universities) to be created and reconfigured constantly.


Cities accelerate the process of exploring possibilities and putting previously unconnected actors in touch with each other: this is the basis of growth in a Schumpeterian environment. In fact, megacities are laboratories for new products, services and lifestyles for consumer markets. In this sense, today's physical agglomeration is probably less pertinent for the supply of diversified goods and services than during earlier phases of development: the Internet and modern commercial logistics provide access nearly everywhere to a highly diverse level of goods and services. However, megacities still provide an essential framework for the process of creating the most advanced forms of consumption, before goods and services enter into production and mass distribution.


Next, the size of labour markets is a competitive advantage for cities. Recent data indicate that relatively weak variations in employment levels hide very high degrees of job creation and destruction within a constant and relatively broad process of employment turnover. From annual increases or decreases in total job supply that vary between 1% or 2%, it is estimated that about 15% of the jobs are created or destroyed during the same period: this figure remains fairly constant in all developed countries, surprisingly so in fact (Cahuc and Zylberberg 2004; Davis and Haltiwanger 1999). The key variables in labour market efficiency are the quality of the adjustments made and the search process between job seekers and employers. Thus, it is easy to understand why large, accessible labour markets, where one does not have to move house for a new job, facilitate the search process. A readily accessible job pool for urban residents depends crucially on the quality of the city's transportation system, a key factor in urban dynamism.


Finally, we note that competitiveness is no longer simply a matter of intensifying traditional work and productivity effects (Veltz 1996, 2000). Efficiency depends less and less on the simple division of labour, which is and has been the main engine of productivity for ages, and more on the quality of cooperation processes. These processes involve several actors who share common goals and who depend on information exchanges and the ability to synchronize their tasks. Only a small portion of these cooperation processes can be standardized or mechanized. Criteria such as innovation or the quality of goods and services, or the reliability of sophisticated means of production - an essential factor for the productivity of capital - fundamentally depend on the quality of formal and informal communication between various actors in the value chain. The various components of an organization require high-quality cooperation: between the organization and its suppliers and clients; between product or service engineering, production, marketing and other functions; and between the organization and its overall environment. From this perspective, cities are the principal source of networked relationships feeding these open processes of coordination. The essential point is central command or even market forces do not suffice to put these open cooperation processes into place. Cities provide numerous kinds of network effects. In this context, producers progressively moving from an organizational pyramid based on a hierarchy to flatter structures shaped by networks find a direct echo in the urban context: megacities provide a suitable ecosystem for the continuous restructuring of such networks.


These three dynamics - city-laboratory, city-labour market and city-network - are the products of short-term adjustments, even as they rely on lasting mechanisms - in particular, social capital, confidence, shared culture and tacit knowledge. Cities, especially the largest ones, also provide another kind of flexibility, a source of insurance. From a chief executive's point of view, choosing an urban location for his or her company requires the least risk with the most potential for opportunities. Compared to a small town, a big city provides easier and faster access to needed skills - including ones that are yet unknown - and probably makes it easier to leave the area without paying high exit costs, whether financial, social or political. The same is equally true for individuals and households.


Let us look at the issue of face-to-face contacts. There is no automatic link between such contacts and the resurgence of cities. People do not congregate in cities simply to facilitate face-to-face contacts. In-person discussions are easier in an urban setting, but many other factors combine to create the interlinked, "hub economy" described earlier. The ability to reconfigure value chains constantly and flexibly, to improve efficiencies in all kinds of research processes, and to reduce the number of irreversible choices in the setting up of new offices or plants, are all powerful vectors of urban growth that do not depend on face-to-face contact. In fact, mobile phones and email are certainly an urban economy's most powerful technological tools.


The "Archipelago Economy"


Concentrations that benefit large cities are not only a local phenomenon, nor a combination of local phenomena. The trend is part of a global reworking of the spatial framework of developed societies. A succinct metaphor for this new framework is the archipelago (Veltz 1996). This view resembles the description proposed by Allen J. Scott (1997) and Michael Storper (1997). It differs, however, from Manuel Castells' view (1996), which describes the "rise of the networked society" and puts a greater emphasis on technology as a vector for change.


The emerging structure breaks even further from the traditional organizational tree hierarchy (where successively higher levels oversee successively lower levels), still at the heart of most institutional and political organizations. Relationships between actors, whether commercial, social or political, are strongest when the distances are shortest. Braudel's world-view posits an enormous base at one extreme, made up of the proximate community's economy with its large share of subsistence activity. In the middle, one finds the local market economy, regionally-based but becoming increasingly national. At the other end, there is long-distance trade, which has structural importance but very thin flows. These different levels are organized entirely according to a hierarchy based on distance.


However, Braudel's world no longer exists! The local level and the global level are mostly interconnected everywhere. Distance is no longer relevant to the strength or frequency of contacts. It is difficult to determine "natural" levels of subordinates in organizational structures and in policy formulations - even if the European Union has attempted to make "subordination" a key concept in its own structure.


The global economy's worldwide network is increasingly a horizontal network, one linking its principal hubs and nodes. Vertical and hierarchical links become less important. According to the numbers we have on varied flow types, flows between the largest core centres are increasing more rapidly than other types. Broadly speaking, two major models of spatial organization can be distinguished. The first is the network of cities that played a major role in Florence and Flanders during the Middle Ages, as well as in many other regions and periods of economic history. For example, colonial trading posts or the boating and shipping centres of ancient Greek cities follow this model, in which total territorial control was not essential. The second model is one of "territorial economies," which seek extensive and complete control over vast continental areas. This model is central to the difficult birth of unified states, but not necessarily of nation-states, such as those created in Europe by the French and Spanish monarchies. In France, the Ancien Régime appears to be a pure example of a "territorial economy" because it deployed centralized power over a vast area where transportation was difficult, but in fact, the situation was more complex (Fox 1971).


Is the contemporary trend a revenge of the first model, of the network of cities? It is, in a sense, but it is worth emphasising that today's networks of cities differ fundamentally from ancient ones. It is not possible to understand them without taking into account the major role played by states. Despite increasing cross-border flows and despite stronger and denser transnational relationships, such networks are still very much embedded in state-based structures and regulations. Networks of cities are not going to replace the mosaic of nation-states forming our basic architecture in the near future. They will mingle increasingly with former structures, disturbing the states-based organization at the same time they reinforce it.


Peripheries Become an Increasing Burden, Even in Europe


A crucial phenomenon is the weakening of links between the urban hubs and the hinterlands, between the cores and the peripheries. This hypothesis seems paradoxical. Modern "new economic geography" pushes core-periphery models. Yet the needs of the formal theory and those of the real world can differ. It is easy to understand that in many cases, the traditional functions of the peripheries and the traditional links between core and non-core spaces have become obsolete. The provisioning of large cities, once so important, is no longer a relevant territorial issue, and more a problem solved by global logistics networks. The unskilled labour that used to be the basic commodity of peripheries is not needed in advanced economies. On the contrary, the poor areas surrounding large cities appear as a burden for the rich regions that constitute urban cores. In short, the rich no longer need poor people.


Some authors, such as the Japanese corporate strategist Kenishi Ohmae, have made that this theoretical argument. He emphasizes the success of region-states or city-states such as Singapore, a case that fascinates him: Singapore's competitive advantage comes from its freedom from the weight of subsidized agriculture and a corrupt state bureaucracy (Ohmae 1996). Likewise, the economic performance of small countries is remarkable - for example, of Ireland and Denmark in Europe. They have powerful advantages: strong social and institutional cohesion, low transaction costs and a small-scale, transparent redistribution system. Their macro-economic tools are also efficient. For example, if these countries opt for a fiscal policy designed to attract foreign investment, the policy will have a far greater impact than in larger countries, where foreign investment effects must coexist within a much larger domestic market.


On the other hand, several economists draw attention to what is called "the curse of mid-sized countries" that have neither the advantages of a small landmass nor the advantages of a large, structured domestic market, such as that of the United States. Regarding large but structurally weak empires, one can wonder about the value (other than natural resources) added by the vast hinterlands that threaten the booming cities of Moscow or Shanghai. Peripheries are an increasing burden even in Europe, and not only in Italy. Ethical questions aside, it is difficult to predict whether the increasing selfishness of core regions will be politically tenable or not.


Today's networks of cities differ greatly from those of ancient times in that they do not connect clearly identified collectives or interest groups, but consist instead of interconnected and complex webs, with cross-functional supply chains and social and intellectual communities spread across the world. They form nodes in distributed networks rather than high points on a graph. Socrates compared Greek cities to a gathering of frogs around a pond. However, each frog was an individual and could fully compete or cooperate with the other frogs. In the networks of the Middle Ages, the dominant players were merchant groups who mixed competitiveness with cooperation in a sort of "co-opetition." They glorified local citizenship and created wealth via "superprofits" or extra surplus-value, extracted from maintaining power over space, through high communication costs and risks linked to their long-distance operations.


Today, the main players in consolidation are multinationals, business groups and transnational scientific groups who operate in stable and homogenous environments, or who endeavour to create such environments across the globe. They do not extract wealth and power by capitalizing on large differentials in capacity between various parts of the world, but rather from their ability to create controlled networks composed of homogenous places. Such places often seem to be islands within their local environment, hosting standardized operations as well as open, creative processes. The academic archipelago of university campuses provides an apt illustration. In the world of manufacturing, differences in productivity occur not so much between countries as between the components of leading supply chains and other supply chains, wherever they might be located. Naturally, this is only a trend: transnational organizations and communities continue to have local touch points and preferred local partners, although transversal systems detached from purely local ones are growing rapidly.


Regardless, the cross-functional logic, whether formal or informal, has rapidly gained ground. Consequently, it is more difficult to identify local collectives comparable to the "urban bourgeoisie" of the past, especially in the biggest cities. Even in "world cities," or those deemed to be important nodes in the global economic system, such as New York, Tokyo, London and others, there is a growing gap between ordinary people attached to a local identity and the main engines of a cosmopolitan society.


Cities and Energy, and Climate Challenges


For the last several years, questions about energy, climate and global public goods

Global public goods are non-excludable and non-rivalrous goods, which means, respectively, that consumption of the good by one individual does not reduce availability of the good for consumption by others; and that no one can be effectively excluded from using the good, whose benefits reach across borders, generations and population groups., Examples include clean air, health, and the absence of armed conflict and pandemics.

have invaded debates about the future of cities and urban public policy. The spectacular growth of Asian cities, especially Chinese and Indian ones where most of the next few decades' population growth will live, appears as a central issue in the fight against climate change. As a matter of fact, the consequences for the planet will be very different depending on whether these cities choose a development model that resembles Europe's, with compact cities that use relatively little energy, or the North American model of energy-intensive sprawling. That said, two other truths bear remembering. First, the urban built environment is more energy-efficient than non-urban environments that are spatially dispersed. If cities have a high concentration of greenhouse gas emissions, it is very simply because they have a high concentration of human activities. They allow for economies of scale and energy concentration that non-urban environments cannot duplicate. Even in the United States' most sprawling cities, per-person emissions levels remain lower than those of the entire country on average. Second, city politicians in developing countries will remain under pressure to propose new solutions for basic service access, e.g. water, transportation, housing and sanitation, for reasons that will primarily be social and political rather than environmental. Fortunately, in most cases, solutions that lean toward social sustainability also carry positive environmental effects, although this does not hold in every case. Thus in developed cities, many beneficial improvements in energy use or climate concerns may bring high extra costs, which makes them acceptable or even fashionable for the middle and upper classes. However, such improvements may penalize the working class who are relegated to poorly equipped and poorly served suburbs. Managing these incongruities will probably often prove difficult. Finally, in the urban field we note the lack of motivational mechanisms such as carbon offset markets. An industrialist in a developing country who improves his processes could benefit from such encouragements, but that will not be the case for a mayor who invests in reducing automobile use in favour of mass transit (Giraud and Lefevre 2006).


City versus State and Public Policy Issues


Let us conclude this brief analysis with a few remarks directed at European cities and public policy. As the French sociologist and political scientist Patrick Le Galès (2003) has clearly stated, the dynamics of European cities remain closely linked to state mechanisms, especially the welfare state. It is stupid to pit nation states and cities (or city regions) against one another, because they share a history and destiny. However legitimate city governments' demands for greater autonomy, their boastful declarations of independence are absurd, inasmuch as they ignore the high number of industries and jobs directly or indirectly depending on the national public sector and/or national transfer payments. The number of state-dependent industries and jobs is substantial in all cities, especially second-tier cities. The French economist Laurent Davezies (2008) has compiled figures showing that an enormous share of the activity in French provincial county seats does not come from private market mechanisms.


On the other hand, Paris is the most "private" city in France, despite the renowned centralization of its public administration. Paris' economy thus depends more heavily on the international situation. However, even in the Paris metropolitan area, a large share of non-commercial revenues and activities are linked to the French state. These state-related activities and revenues serve as a kind of shock absorber that hardly exists in America or Asia. Like Janus, European cities have two faces. On one side, they have been driven historically by an economy based on flexibility; on the other side, they remain deeply anchored in a shock-absorbing state welfare system. The latter probably gives them a real competitive advantage compared to cities that rely only on flexibility, as long as there is a balanced mix of the two forces. European cities are - or could be, or should be - the preferred laboratories for experimenting with needed public-sector reforms, as well as for new alliances between public and private actors. They already offer a diversified and unique range of public-private partnership experiments, rich with lessons for urban and mass services. This component of their makeup should be recognized for the huge competitive advantage it is, rather than something to dismantle for ideological reasons.




Let us look for a moment at the policies that will ensure the economic and social success of our cities, even if that is not the main subject here. As we said earlier, prosperous cities are those where market-oriented dynamism best captures social externalities and effects linked to non-commercial interdependencies, such as "milieu effects" or those of the immediate environment. In these cases, the fluidity of market relationships integrates with enduring social structures, providing non-physical goods such as mutual confidence, tacit knowledge and an acceleration in the collective apprenticeship process. Relevant polices will aim fundamentally to consolidate this non-physical infrastructure. Physical infrastructure remains important as the sine qua non of development, but these non-physical assets are more crucial still. The determining factors for success include the ability to foster cooperation between organizations; improving the quality of coordination between institutions; production values, expectations and shared projects; and, at the end of the day, increasing the quality of public and private governance. Resources essential for development are socially constructed, rather than conferred by nature or geography. Decreasing communication costs will relativize traditional parameters for geographic location. As social and political actors, cities are responsible for their own success or their failure.


Another important consideration is that territorial marketing should not become the be-all and end-all of development policy. Many cities, especially those facing financial difficulties, are obsessed with finding outside investors and are ready to do whatever it takes to attract them, usually using tax incentives. However, such cities tend to forget three basic rules. First, success or failure in creating jobs depends most on the health of local businesses (Chesire and Gordon 1998). Consequently, it is often more profitable to monitor trends closely and to seize risks and opportunities within the local economy than to try to attract new investors. Second, cities strangely persist in making costs their central strategic tenet, while competition in our countries' economies now depends not only on price, but rather on quality and innovation. Cities' decisive competitive advantages are far more sophisticated and complex. What is more, they must differentiate their objectives and strategies, resolutely refusing to opt for narrow specialization. Third, as the British scholars Ian Gordon and Paul Cheshire (1998) also stress, it is imperative to find the right scale for development policies: unfortunately, local policies too often limit themselves to a zero sum game between different areas, different actors and different interests within the same city.


Finally, it is worth strongly stressing a profound change that may prove decisive for new trends in the worldwide geographic-economy. Increasingly, personal mobility prevails over the mobility of capital. In the short term, the choice of location appears to be a structural issue. However, in the medium term, individuals' choice of residence at the national and international scales - especially those workers with the most resources, skills and entrepreneurial capacity - ultimately determines the geography of work. Developments in communication technologies permit an even greater choice of location and accentuate this phenomenon. International communities of scientists or engineers become conduits for the distribution of technologies and development - as important in this respect, if not more important, than multinational firms (Saxenian 2006). Thus, it becomes more strategic for cities to attract talent than to attract capital. Policies to develop urban amenities, quality of life and cultural interests therefore directly affect urban economic policies. Cities always need companies to grow. Yet increasingly, companies need cities to attract and retain the best-performing employees.