GaWC Research Bulletin 45

'...we need a strategy for change that can rally support, combined with clarifying alternatives....the immediate involves the struggles that go on now at the local level...that impinge on how lives are led immediately. ...all such struggles are local...and alliances are essential....in the long run, we have to engage seriously in the project of inventing the future system...We need to debate priorities and the nature of institutions that could implement them....utopistics.' - (Wallerstein 1999)

Conventional wisdom suggests that the Middle East is not performing as it should. Certainly along a range of criteria, there is no question that there is a problem. Placed in a comparative context with other regions, the difficulties appear even more dramatic. The region appears to be stuck, and the developmental train is not stopping at their station.

The causes presented for this underdevelopment remain locked into traditional frameworks of neo-modernization and located within the territorialist state-level of analysis. Likewise, policies introduced by governments or persuasively supported by international lenders to correct these problems are the usual state-directed ones. Regional-level policies are relatively few, and those which do exist suffer from a similar lack of scope and creativity. In sum, there is a crucial lack of alternative conceptualization, and few new ideas.

This paper proposes an alternative, one which is grounded in the world system perspective, and is, as a result, more historically sensitive to the reality of the longue durée in the Middle East. The argument sees Middle East underdevelopment within a global historical dynamic, focusing in particular on the interaction of forces, both "above" and "below" the state level, which have generally been ignored in understanding the dynamics and potentiality of the region. By grounding the analysis at the city-level, and stressing systems of cities as a crucial engine for progressive development, regionalism is reconceptualized.¹ Cities and systems of cities, and not states, have always been the engines for change within this region; if we return to this understanding, the policies which emerge could make a contribution to regional development and empowerment.

A HISTORY OF DEVELOPMENTAL PROBLEMS

The state-building project in the Middle East has not gone smoothly. The foundations on which the modern Middle Eastern state emerged post 1920 were fragile, and tremendous centrifugal forces consistently undermined the purposes of both the first generation of nationalist leaders and their core backers. Most states in the region quickly evolved a bureaucratic authoritarian structure, with import substitution industrialization as the state strategy for development. High levels of dependency on the core of the world system carried over from the Ottoman period, and this dependency intensified during the era of huge oil revenues.² Rentier economies and corporatist systems within "fierce states" emerged only to begin a process of disarticulation in the 1990's.³ Fundamental questions of the strength of the state remain, of its continuing legitimacy within the political culture, of the trajectory of its evolution, the artificiality of its delegative democracy,⁴ and particularly of its autonomy within the world system.

Regionalism and the regional project have likewise produced surprisingly limited results. The Arab states participated in the first wave of regionalism with the creation of the League of Arab States (1945) ostensibly to move forward their common vision and restore their "shattered" community. Subsequent extensive institutionalization and commitments under League auspices have produced few tangible outcomes at the regional level. Alternative sub-regional, state-centric institutions, often encouraged and manipulated by the core, have likewise produced limited economic or political empowerment. Although the last few years have witnessed renewed state expressions of commitment to limited regional economic projects such as an Arab Free Trade Area (AFTA), numerous factors block the way, and the hoped-for state-level return out of increased regional cooperation is still a long way off.⁵

Actors external to the region have consistently proposed alternative regional projects which serve their interests rather than those of the local community. Two such projects, serving the interests of the core, have conspired over the last five years to detach and distract local elites from autonomous intra-regional development: the US/Israeli project of a "greater Middle East" under the Madrid formula; and the Euro-Med partnership under the Barcelona Agreement. The former is currently on hold given the disastrous results of the bilateral discussions between Palestinians and Israelis. The latter continues to be slowly institutionalized around the Mediterranean through bilateral agreements between states and the EU.

For the purposes of this paper, a number of key themes concerning the underdevelopment of development in the Middle East and North Africa since the 1920s should be stressed, although they can not be discussed here in depth: the region remains a peripheral region, with only a few states moving into semi-peripheral status; the vast majority of policies implemented have only deepened the integration of the region into the world system; in the structures of credit, information, production, services and security, the Middle East is more tightly integrated globally than any other region; intra-regional linkages remain low; inter-regional connections with other peripheral regions remains low; the continuing dismantling of the corporatist state has left millions in the region with little justice, opportunity, voice or protection; institutional actors in the core are powerful players within state boundaries; local elites serve the interests of the core; and any comparative advantage currently possible for the region, primarily build around extraction of carbon-based products, will only be temporary.

AN ALTERNATIVE PARADIGM

If we desire some autonomy of action, improvement in the HDI, a decrease in the poverty gap, "true development", progressive empowerment, some delinking, how can it be achieved? How can actors within the Middle East/North Africa region shape their networks and flows in order to enhance their social welfare and power? Alternatives on the ground are few at the moment; resistance to the Washington consensus, particularly at the elite level, remains dependent upon conceptualizations of political space which privilege state sovereignty over all other spatial narratives.⁶ As such, perhaps they are bound to fail. (Magnusson 1996:9)

The alternative is to identify "more realistic and fruitful lines of observation and thought than we have tried to use so far." (Jacobs 1984:28) Drawing on world systems theory, urban sociology and social network analysis, it can be argued that the world system is organized, and always has been, around and through its cities. (Clark 1999:9) These cities, linked into networks of exchange and flows as "a system of cities", are the backcloth on which the superstructure of empires, states and other polities are based. (Modelski 1999) These city systems provide the "skeleton; the hard substance, still in evidence in so many cases, around which the flesh and blood of social organisation was to be deployed..." (Modelski 1999:385) The unit of analysis becomes the system of cities which emerges from the interconnection between a set of nodes (cities) and the sets of ties (exchange flows) which link them. (Wellman and Berkowitz 1988:16)

Under this perspective, a city can be defined as a spatial location where networks of networks "come together" over a particular period of time to create a complex, dense and dynamic structural pattern of exchange, communication and meaning. Cities are places or "theatres of articulation", and perhaps "centres of authority", within their networks. (Knox and Taylor 1995:7)

Beyond the bounded cities we identify, it is crucial to our analysis to conceptualize other network levels (scales) and thus forms or heuristic notations, important for praxis.⁷ No city, by definition, exists alone, but is embedded in a range of broader networks which can also be considered:

Taking this perspective deemphasizes the polity which is the superstructure within which city systems are embedded, and opens up the political space. Combinations of city systems may be temporarily associated together under one rubric of state, empire, kingdom or league, but cities, their networks, and the city systems which emerge remain crucial morphologies for action.⁹ City systems are crucial engines for defining identity, accumulating value, and deciding the distribution of value. In fact, we may argue that cities and systems of cities are the core ontological level for the world economy.¹⁰ Cities are the key site for the working out of politics, and through their networks constantly challenge the constructed boundaries of the current international system. In fact, cities and the city systems in which they are imbedded are crucial heuristics for the world system perspective (Knox and Taylor 1995:49), since the city is the spatial point where networks are most dense, where space becomes place, and beyond which all conceptualizations of identity are "created" in an Andersonian sense.

As a result of their interconnections, all cities have some "degree of worldness", a connectedness into and power in the global world system which can be evaluated.

Since certain cities have more connectedness across some sets of exchanges than others, it is possible to identify cities which have so much "connectedness", particularly within the global flows of capital and credit, that we need a special concept for them: "global cities". These global cities are central to the functioning of global capital networks as sites of command and control. In the meta-narrative of capital, such global cities are imperative since they "articulate larger regional, national and international economies." (Knox and Taylor 1995:22) It is this proposition that has driven the study of global cities and the hierarchy of cities found in the work of Friedmann, Sassen, Taylor, Harvey and others.

It should be noted here that the local and the global are not locked into dualistic opposition, but are simply heuristic constructs to identify the locus of network action. Thus, we must conceptualize human agency as multileveled, with actors acting locally while thinking globally, or thinking locally and acting globally, or "thinking and acting simultaneously at multiple scales." (Smith 2001:158) Actors are goal-oriented, and employ their networks at multiple levels to shape their environment.

Given recent changes in the world system, there is a general push factor driving the creation of new forms of networking and clustering. In a world system where the spatial narrative is under attack, and new threats/opportunities are arising, actors are moving to enhance the capability of their networks. For example, we can see this dynamic in the "propensity of many types of economic activity-manufacturing and service sectors alike-to gather together in dense regional clusters or agglomerations", what one author termed "collective propinquity". This drive is a "strategic response to heightened (global) economic competition" in an attempt to empower networks in a changing world system. Likewise, cities and city systems cannot afford to be passive in the face of such changes; they must be pro-active in their own interests.¹¹

Crucially, this shift of perspective privileges the level of the city system when developing policies for change. Cities, systems of cities and the nature of their networks can be developed, enhanced, empowered, or modified. The goal of "development" can, and should be, pursued in such a way as to enhance the efficacy of actors within their networks. Policies might seek to:

Any or all of these actions could be termed "going global" or "wannabe cities". Some authors have proposed the term "entrepreneurial cities" to refer to specific types of cities where a local network of elites, often using the formal municipal structures, seek to shape the networks and opportunities of their city so as to enhance its access to resources within extralocal networks. (Short and Kim 1999:117) Dubai's aggressiveness in positioning itself within the international system is an excellent example. Likewise, Istanbul was "going global" when it linked itself both into the trans-Balkan and Iranian rail networks.

However, we need a broader concept of "going global", where actors develop and implement a range of policies and actions to enhance the characteristics of their networks in the interest of empowerment. Municipal governments may pursue such policies for themselves, or other actors at other levels, such as community-based organizations, state ministries, business associations, regional organizations or international governmental organizations, may also evolve policies to pursue such goals with or for them. Developmental strategies which ignore this level of analysis are missing a crucial context in which change can be effective, since: "cities are unique in their abilities to shape and reshape the economies of other settlements, including those far removed from them geographically..." (Jacobs 1984:32) Ultimately, the argument is that there are new possibilities for multitiered reciprocity, and that cities embedded within city systems are an appropriate new/old political space for policy articulation.¹²

Since cities are embedded within the world system, the nature of the dynamic between the two is crucial to formulating policies. There is a correlation between the development of city systems and the development of the world system, although the hypotheses vary as to the relationship. Modelski suggests that the development and dynamics of city systems drive the rise and evolution of world systems. (Modelski 1999:392) Frank (1998) and Abu Lughod (1991) share this view when they suggest that city systems provided the engine within dynamic regions interlinked as a world system from 1250 - 1800 AD. Friedmann, Sassen and others, however, argue that "explanations of urban development lie in the social and economic characteristics of successive forms of capitalism." (Clark 1999:9) Taylor makes a similar claim when he suggests that "globalization has de-nationalised cities in important respects: no longer so truncated, they have transmuted into world cities."¹³ Smith offers an intermediate alternative when he says we must "study the effects that global policies and transnational networks have on cities, local power structures and ordinary people's lives, and the effects that particular cities and their people have on these networks of power, meaning and identity." (Smith 2001:183)

The Friedmann and Sassen argument privileges the economic structures of global capitalism and the "new phase" which has emerged since the 1970's. This conceptualisation suggests that we must see cities as major new sites/roles in the new world order. The spatial restructuring of the urban results from economic processes within the world system: it is within cities where global capital must be spatially located, resulting in the appearance of global cities. (Short and Kim 1999:11) Progressive development, as implied by this approach, should aim to "move up the hierarchy", to aim for global city status, but with a human face. The policies required are to facilitate the location of service sector, capital and FIRE (finance, insurance, real estate) corporations within the city, while working for a more equitable distribution of social resources.

An alternative narrative challenges this view, and suggests that an agency-oriented, grassroots perspective termed "transnational urbanism" more completely and correctly conceptualizes what is currently happening in the interaction between the world system and the urban world. The view rejects the emphasis on global economic forces and global capital as deterministic. (Smith 2001:2) The state is not dying, and global capital is not all-powerful. Instead, the dramatic changes combined under the term "globalization" must be localized on the ground; if we fall into the trap of downplaying the role of the state we will miss the continuing significance of borders, state policies and national identities. Thus we want a world system perspective which holds that global processes are shaping the options of the composite units, and encourages investigate of this; but we must also explain/understand the way actor choices and dynamics may impact the world system.¹⁴ Grounded within networks, globalization is opening up options or political space to world cities and their networks.¹⁵ The job of the analyst is to understand how such networks are both empowered and disempowered by changes in the world system.

In particular, the second view focuses attention on the space for empowement that networks may offer. It suggests that developing policies to empower grass-roots networks within cities will have important spillover effects into the empowerment of people (democracy), in limiting the negative effects of globalization, and in enhancing social welfare. In this way, we may be able to follow Wallerstein's suggestion, quoted above, to "engage seriously in the project of inventing the future system."

INSIGHTS FROM THE MENA REGION

Where are the world cities of the MENA today? Are there any global cities or degrees of worldness on which we can build? What systems of cities could be used to center and empower the MENA region? If we apply the framework of the global cities ideas of Friedmann and Sassen, a review of the MENA region shows that it has no global cities, and most of its cities and city systems are not even on the radar screen of the global capitalist system.¹⁶ This dismal showing for MENA cities can be seen in the "Roster of World Cities", produced by the Globalization and World Cities Study Group (GaWC). In reporting the GaWC findings, Beaverstock, Smith and Taylor (1999) review other studies which employ a range of criteria to evaluate world city rank. The consensus is that only Manama, Cairo, Istanbul and Tel Aviv have any of the service characteristics required to list them as "world cities." Beaverstock et.al. then go on to develop a stricter set of indicators based on the provision of corporate services. Their findings for world cities based on "different degrees of overall corporate service provision", confirms the poor linkage of MENA cities into the global capitalist command and control structure. They find that of 122 world cities, MENA cities emerge far down the list (global cities gain a top score of 12):

Perhaps one key reason the MENA region is not doing well is the weakness of its city system within the world system; this "structural hole" at its heart?¹⁷

This lack of centrality across key command and control networks for global capital may represent the current situation in relation to global capital, but misses both the historical context as well as other aspects of command and control which are articulated throughout the region. The MENA used to be central across a range of networks and flows. The current lack of centrality is only a recent phenomenon in the longue durée. Middle East history is the history of cities as the key polities: empires came and went, but city systems as the key locus for economic development of the region in the form of markets, trade, production, capital, continued. Compared to other regions, the MENA has longer experience with the urban world, was more highly urbanized earlier, had a more central role for cities in the political, spiritual and security life of the people, than anywhere else on the planet.¹⁸

The city systems of the region were "the" world system from 4000 BC to the classical era, with the largest cities and central city networks. Gradually the periphery (the Mediterranean, India and China) experienced urbanization and became interlinked, from Britain to China, through the mediation of city systems in the MENA. (Modelski 1999) Likewise, the MENA experienced the rise of Islam as part of a new polity which shifted some of its city-based networks of exchange.¹⁹

J. Abu Lughod stresses the historical continuity of MENA city systems with her focus on the eight key interlinked city systems within the world system of the 1250-1350 AD period. A.G.Frank, I. Wallerstein and F. Brudel all emphasize the relative centrality of MENA city systems for the rise or functioning of the world system in the 1400 to 1800 period. Scholars of the Ottoman period have demonstrated the interconnectedness and centrality, during the 16^th to 19^th centuries, of Aleppo, Istanbul and Izmir for trade/production across China/India, Persia, and Europe. Production and exchange included the widest possible range of goods and services, including military equipment, ship building skills, technological innovations, and administrative forms.²⁰ A quick look at the comparative population figures for MENA cities vs. those outside the region highlights the overwhelming size and continuous urban importance of these cities: Cairo dominating Europe, Africa and west Asia from 1200 until 1500; with Istanbul the largest city outside China from 1500 until 1750. (Chase-Dunn and Willard 1993)

Ironically, it is not until we get to the imperial imposition of the modern state form in the region post 1860s that our analytical bias toward cities and city systems in the MENA begins to disappear. Both analytically, as well as at the policy level, the concept of the state and of a national economy with sharp boundaries and control over its cities, makes its appearance concurrent with the western colonial regimes. Until today, the state-centric bias of data, the state narrative of boundaries, and the artificiality of the concept of a national economy hide the importance and interconnectedness of the city systems in the region.

Post-1920, the history of the state project in the MENA contains numerous stories of state/city conflict as nascent state institutions struggled to "establish control" over cities and to subject them to centralization. In many cases, city elites put up a significant struggle, and at times were successful in shaping policy for their own purposes. The relationship between Istanbul and "Turkey", for example, or that between Aleppo and "Syria", or Baghdad and the new "Iraq", are illustrative. Ultimately, all cities were "nationalized" to some degree, experiencing both political and economic truncation (Aleppo and Istanbul) while some became empowered (Amman).

After 1945, the state clearly dominates the region, with cities being bent to the will of the centralizing, authoritarian, national state project: the autonomy, exchange linkages, infrastructure and public space of cities were subverted. Occasionally, however, elites within individual cities within the polity "rebelled" and captured the heights of the state. Struggles between city networks for domination of the state often broke out into elite conflict, as in Syria. The struggle for control of Algiers between the colons and the FLN was symbolic of the struggle in Algeria. However, the best evidence for a continuation of the importance of cities can be seen both in the central role they played in the import-substitution-industrialization strategies of the state, as well as in the resistance to state policies expressed by local political movements and riots within the cities of the region.

Dramatic shifts in key characteristics and dynamics of urbanization and urban growth occur within the region post-1960. Issues such as primacy and urbanism, environmental degradation, and the informal economy become a focus of concern, policy and analysis.

It is important to note here that the vast majority of urban analysis on MENA cities during the 20^th century assumed a state-centric focus, buying into the western concept of a national economy within which cities serve the government, the artificiality of conceptualizing flows as ending at state boundaries, and an ignorance of world system effects. Often they reflect a negative view of the Middle East city as dystopia. Thus, when urban problems emerged as a focus for concern post-1960, the solutions proposed revolved around state agency: state funding, state planning, state initiative. There was little awareness of the networked nature of the city, and no power associated with any level other than that of the state.²¹

With the rise of a new paradigm on cities, emergent from changing realities and enshrined in the global cities literature post 1986, numerous studies of particular cities, of global cities, of city regions have appeared in the social science literature. This extensive collection, however, contains few references to MENA cities. Ironically, there is a structural hole concerning MENA cities within the study of world cities just as there is a lacuna within the policy and regionalism literature: cities in this region are not on the world map.

One way to begin to lift the veil covering the worldness of MENA cities and their historical continuity is to broaden our definition of 'world cities'. Keeping to the service concerns elucidated by Friedmann and Sassen, but looking beyond the narrow sectors considered in their work, an alternative sense of worldness emerges. For example, an interesting concept has recently appeared, termed the "creative industries". These industries fall within the service sector, but represent one of the fastest rising contributions to city income. They include: architectural design; fashion industry; graphics computer design; publishing; games and software; interior design; landscape design; furniture; music; cinema and cinema production. For Britain, for example, in 1999 the income from this sector was £112.5 b, employing 1.3m people and producing 5% of GDP. In the MENA, a cursory review would highlight the Egyptian film industry and its major role as a "guidance industry", as Nasir termed it, when he nationalized it in the 1960's. Since 1935 the film industry has been an important sector for the Cairo economy. In the last four years, with international investment, local initiatives and new tax laws, the industry has seen a revitalization and expects to expand during the next decade. A related but less significant money maker for Tehran would be the Iranian film industry.

Other examples of "cultural industries" in the Middle East include Gaza, Cairo, Beirut and Amman with their growth in software development; Dubai with its hosting of the Fashion Fair; Beirut's growing design and fashion industry (with products manufactured by low-cost labor in Damascus); and the music industry in Istanbul. This latter sector has established key contracts with global TNCs like PolyGram in order to deal with the changing requirements of shifting markets.²² For the music industry in Istanbul, with its $130m in sales in 1999, the pressure of the EU to deal with problems of piracy have encouraged a sustained response. In sum, there is enough evidence to investigate further the cultural industries and the global commodity chains which emerge from them. MENA cities may have greater centrality across these or other networks of exchange than they do on service indicators like law or accounting.

Another sector which might help us evaluate MENA worldness is transport. Some literature exists which evaluates the development of transport flows among cities in the world system. These studies do not focus specifically on MENA, but by using network data to examine the linked nature of these cities, they suggest interesting changes in position and flows. D. Keeling, in a 1992 study of transport linkages among world cities, found that Cairo was the only ME city registering on his global air transport network because it had a "regional center" profile.²³ In a study of "World Cities in Asia", Shin and Timberlake (2000:2273) evaluate changes in airline travel, seeking to identify "the sets of cities in which each city member is reciprocally and directly linked to each other". Their analysis indicates that Dubai is part of a clique with Bombay, Colombo and Karachi. It also shows that Dubai has been one of the top twenty cities in the world in terms of its centrality in the overall system of global cities since 1991, when it first appeared on the list. It was the first Middle East city to play such a role. (Shin and Timberlake 2000:2279) Other MENA cities considered in the analysis but "scoring" poorly due to their insignificance as nodes in the global airline system included Cairo, Damascus, Kuwait, Larnaca, Riyadh, Tel Aviv and Tunis. Seeing the MENA through the lens of global air transport clarifies the lack of centrality for the region. However, the fact that Dubai was "not on the global map" in 1992, but by 2000 was linked by 78 airlines directly with 125 other cities and handling 12m passengers/year, reminds us how temporary any rank is: city rankings can change dramatically due to city initiative, conflict, or technology. Beirut learned this to its dismay during its 15 year civil war.

A focus on representative cities in the region may allow the flavor of their historical development and degree of worldness to emerge. Rabat, for example, appears to deserve the title of a world city. It is a political and administrative center, but also a key industrial node. Significantly, to understand Rabat's position in the world system, its symbiotic dyadic linkage to Casablanca, one of the leading ports in North Africa, must be considered.

Rabat is deeply involved in official urban networking with other world cities. It is a member of numerous city-based regional and international organizations, including the United Towns Development Agency (UTDA) under the United Towns Organization (UTO); the International Union of Local Authorities (IULA); the World Association of Major Metropolises; the Arab Towns Organization; the Organization of Islamic Capitals and Cities; and the World Capitals Forum. Twining relations, for example the one with Barcelona, occupy considerable time for the municipality.²⁴ Casablanca also networks this way: the Executive Board of the World Federation of Twinned Cities held its 2001 annual meeting in Casablanca, and the city is the home of the Arab Fishery Marketing Center (INFOSAMAK).

The Rabat-Casablanca conurbation is deeply linked with the EU, and particularly with Spain. Recently, Rabat has become the site for backoffice setups for Spanish TNCs like Telefonica. The company established a new call center in Casablanca to go with ones in other Moroccan cities, and laid a new fiber optic cable to facilitate this industry. Telefonica employs Spanish speakers, and has invested $2b over the period 1998-2001.

Urban industrial growth over the last decade has been high, and Rabat-Casablanca continue to push for further FDI. The electronics components sector is the current focus with STMicroelectoronics joining Boeing and Volkswagen in the 100% growth of the sector since 1995.

In an attempt to attract further FDI, Rabat/Casablanca is working hard to upgrade both connectedness and its aura as a world city. New inter and intra regional transport routes, with Casablanca/Rabat as the Atlantic terminus, have been opened to Spain, France, Italy and across the Maghrib. The urban elites are "going global" with their bids to host the World Cup: they were unsuccessfully in 1994,1998 and 2006, but now are bidding for 2010. Likewise, the push to host the Olympics, continued exposure in sports such as track, cycling and the Davis Cup, are all part of urban promotion. Ultimately, the message is clear: Rabat/Casablanca is "going global", positioning itself as THE entropôt on the Atlantic for North Africa, and promoting its networks into North Africa, Africa, Europe and the Middle East.

Cairo has a longer history than Rabat, and is "one of the most important urban communities ever known to man" (Abu Lughod 1996) Urbanization in the area of modern Cairo goes back to 2000 BC, with the Egyptian empire of 1200 BC organized around a six-city sub-system which was part of the larger Middle East city system of 22 cities. (Modelski 1999:392) The global reach of Cairo, founded in 641 AD, varied over the centuries, reaching a peak in 1200-1500. Recent population growth has made it the largest metropolitan area in Africa and the Middle East, one of the key megacities in the world. Greater Cairo now incorporates urban areas all the way to Alexandria as well as second-tier cities built around the Cairo core. Within this Cairo/Alexandria conurbation, over 22% population growth is expected over the next few years.

Like Rabat, Cairo's municipal leadership is participating in global urban networks for support and empowerment; it is a member of all the global urban organizations to which Rabat belongs. Also like Rabat, Cairo's elites are working hard to position the city within key networks of the global economy: For example, the vision is to create a "Delta Valley", like the Silicon Valley, around Cairo. Currently there are over 300 software companies within the SMA. However, none are globally significant, with no Egyptian communications companies trading in the New York Stock Exchange, for example.

Other growth industries are ones where Cairo is central to regional command and control: tourism, pharmaceuticals, and construction. Cairo has hosted numerous investment conferences to attract investors, particularly in the areas of energy, transport, wastewater and chemicals. In sum, there are important signs of a city on the move, of a city where local elites are shaping policies to pursue a global vision for the city, but where the role of the national government remains quite heavy, and still serves to stifle development and to control local resistance.

Compared to Cairo or Rabat, Dubai has been even more aggressive at "going global". It is one of the most entrepreneurial cities in the MENA, known in the region as the "city of merchants". Essentially a city state, with an urban area running from Jebel Ali in the south to Sharjah in the north, Dubai had a population of 25,000 in 1948. By 1996 it had reached 695,000, and the projection for 2012 is 2,200,000. About 15% of this increase is expected to result from in-migration of highly skilled labor in the upper and middle-income categories.

Although the city itself traces its growth only from the mid 1800s, the harbors and people along the northeast coast of Arabia have been part of the world system from 4,000 BC. Beginning with the city systems of Sumer, a thriving caravan and maritime trade operated throughout the area. The same was during through the Hellenistic, Roman, Byzantine and Islamic eras, with settlements involved in the production and trade of incense and livestock, as well as the transit trade with Iran, the Mediterranean, Africa, and across to India.

The modern city of Dubai began to emerge around 1833 with the growth of regional trade and renewed connections into the Indian Ocean. Pearling and fishing also provided important international links. During the British imperium, Dubai became the key trading port of the Trucial States. New technology brought increased centrality (the British/Indian Steam Navigation Company, for example, shifted its port of call to Dubai in 1903) and informal linkages marked Dubai as a leading smuggling port along the Arabian coast. The progressive social "feel" of the city set it apart from other ports along the coast, and this attracted other traders and in-migration, making it the key entrepôt in the northeast.

Oil revenues began to arrive in 1972, but since their reserves and thus revenues are limited, the city leadership has been forced to plan for a non-petroleum economy based on its trading and transport base: airport construction, the biggest deep-water port in the Gulf, a free port at Jebel Ali were all quickly developed. The city has been aggressive in attracting banks, and legislation and tariff barriers to trade are low. Dubai is also diversifying into high technology, shipping and cargo, and tourism. All three depend on the strategic location of the city, brokering exchanges stretching from South Asia to Central Africa and into the former Soviet Union.

On the high tech side, the government is spending billions to create an information city for the future, the Dubai Internet City (DIC). Established as the world's first free trade zone for e-business, the goal is to house the "entire value chain" for the internet industry within its grounds as a one-stop shop for innovation, investment and development. A related development is the Dubai Media City, opened in 2000.

Other than the covered suq in Deira, Dubai has little in the way of a traditional city center or madina. As a result, the city lays its cultural claims on its modernity. Dubai has become the party capital of the Gulf, with the most nightclubs and nightlife. Additional cultural exposure comes from hosting top quality sporting events, including world-class golfing events, international tennis, rugby sevens, and the largest purse horse-racing event in the world, the Dubai World Cup. Such "show pieces", along with more traditional sports such as camel racing and hawking, are part of the active selling of the city and its cosmopolitan nature.

Dubai is the most aggressive "wannabe city" in the Middle East, working hard to position itself like its models, Singapore and Hong Kong. Consequently, Dubai faces a continuing struggle to stay competitive in the global political economy of the 21^st century. Dubai's Strategic Development Plan of 1996 laid out the city's 20-year goals, setting 6-7% annual growth targets for real non-oil GDP. The city's ability to meet this aggressive plan will depend on fluctuations in oil revenues, competition from other cities in the region, and general trends in the world economy.

Less like Dubai, and more similar to Cairo, Istanbul has a built environment and an imperial history which shapes the position of the city in the world system. Given its geography and centrality in two of the world's most important empires, one would assume that Istanbul is a "global city". However, its potential has yet to be actualized (Keydar 1999:188), although it ranks higher than other MENA cities on all global city criteria.²⁵ Like other MENA primacy cities, almost one half of the national wealth and income of the state emerges from Istanbul, creating a situation where "urban and national politics tend to merge." (Keydar 1999:196)

The proverbial conflict between "state" and city is well exemplified by Istanbul's history. Whether in the darkest days of the Crusader occupation of the Byzantine throne, or with the capture of the city by the Ottomans, the centrality and power of this imperial city to shape, define, and empower empire is clear. The city was the empire, and its capture and control, whether physically or metaphorically, was crucial to the survival of the superstructure. Even more starkly, Attaturk's national project required stripping Istanbul of its position, networks, and centrality because of the competition it represented to a new, competing narrative. Dramatically, by his death even Attaturk realized that he had failed to sublimate the city: today the state continues its struggle to constrain Istanbul's elites, image, networks and autonomy. Developments in the world system exacerbate this conflict, with:

"the struggle of the international bourgeoisie and the new professional and managerial classes of Turkey for the city's soul is in fact a struggle for the future orientation of the country as well."(Keydar 1999:196)

Within the city, class distinctions are dramatic. The global city literature stresses the divisive implications of global city status, and Istanbul is an excellent case study. In the struggle among competing narratives to define the city, all Istanbulis have bought into the globalization narrative, and although they express it differently, all utilize, and find they must utilize, the related concepts of exceptionalism and global potential to define their vision of the city within the give and take of local politics and debates over public space. (Keydar 1999:194) Islam has provided the primary logic for local organization and resistance to the forces of globalization, partly because it itself has a global narrative and a Middle East referent, thus providing a vocabulary that "credibly ranges over the same territory as a homogenized global culture." (Keydar 1999:195)

Like Rabat and Dubai, Istanbul's leadership have worked hard to present a culturally globalized face to investors. Olympic bids, conferences like Habitat II, and international level sports teams, are crucial components of this representation. Equally, Istanbul, like Rabat and Cairo, participates in global networks of cities, including the important World Association of Major Metropolis.

In a recent network analysis of European world cities, the Loughborough Study Group on Globalization and World Cities found that Istanbul was part of an outer triangle of cities which have strong "world city" characteristics in terms of certain key service sectors, but that it lacks others, making it a part of a collection of "un-European" cities linked into the core of European cities.²⁶ Constrained by its various linkages, Istanbul may be better served by maintaining its position as a broker between the MENA and European regions.

Beyond particular cities, there are few policy initiatives at the regional level to strengthen and empower city systems. The Arab Town Organization (ATO), under the League of Arab States, has produced little of substance for moving along a "MENA of cities" project. ESCWA (Economic and Social Commission for Western Asia), supposedly dedicated to enhancing regional development and empowerment under UN auspices, has responded to the Habitat II agenda by "collecting data on urbanization within member states." The World Bank with its "Mediterranean Development Forum", has facilitated discussions on urban issues, including regional linkages among urban centers, Mediterranean links with the Gulf, and policies for regional development.

Compared, however, with regional or sub-regional city system projects in other regions such as Asia and Europe, the MENA again demonstrates a "structural hole" at its heart. Other regions are using city-system strategies as a component of their developmental package. Important projects within Europe such as Eurocities, Metrex or Medcities combine with a long tradition of awareness of "city spines" and cities as engines for growth.²⁷ In Asia there are cross-border networks, "triangles" and key dyads, all implementing aggressive policy initiatives to enhance development and autonomy.²⁸ These "wannabe world cities", beginning to work together within city systems, appear to be making significant contributions to development, are influencing governance patterns, and have evolved municipal foreign policy or "city diplomacy" to a new level.²⁹

The MENA experience suggests the following:

A NEW APPROACH TO MENA DEVELOPMENT

What is to be done to create an "archipelago of MENA cities", to empower a "fertile crescent" of interlinked cities?³⁰ One set of policies could focus on enhancing the flows or exchanges among the key alpha cities (major spine) of the region (Rabat-Cairo-Beirut-Dubai-Istanbul-Tehran) in order to strengthen their interconnectedness vis their other connections.This would involve increasing transport, communication and energy exchanges through infrastructural development³¹; enhance the flows of labor, capital, expertise, services and trade among these cities through regulation adjustments, incentives, and institutional cooperation.

Second, this alpha core must be embedded, in similar ways to those argued above, within a stronger network of beta cities (minor spine), regional in conception, which would enhance flows and exchange (Baghdad-Aleppo-Aqaba-Damascus-Khartum-Casablanca-Ankara-Riyadh etc).

Thirdly, policy makers should look to enhance specific geographic, commodity chains and trade sub-systems whose transformation could empower the dynamics of the beta and alpha system: Beirut-Damascus; Cairo-Alexandria; the Red Sea city system; the Gulf city system; the Maghrib city system; the Istanbul-Tehran transport link. Such a strategy could operate on local levels as well: enhance the density and flows among the seven Bedouin Arab towns in the Negev (Lithwick 2000); as part of a renewed peace process, link Haifa, Beirut, Tyre, Sidon, Tel Aviv, Damascus, Amman, Jerusalem, Ramallah, Nazareth, Jenin, Nablus, Irbid (all within a 75 kilometer radius of each other) together into a dynamic engine for peaceful development.

Fourth, decision-makers should seek complementarity among cities, joint planning and consultation, data sharing, and linked enhancement to support already existing "natural" conurbations. Already in the Gulf there exist a number of standard metropolitan areas (Dammam-Manama; Al-Ain-Bouraimi; Dubai-Sharjah) which transcend existing political boundaries. The MENA has other such cross-boundary entities: Eliat-Aqaba; Basra-Abadan; Gaza-Askelon whose cooperative development would empower, the region but are currently divided by artificial boundaries and bureaucratic walls. Power must be de-centered from states so that creative municipal cross-boundary arrangements can emerge as required.

Other strategies might include promoting epistemic communities, action nets or policy networks among actors in cities throughout the region in order to increase dialogue, remove constraints, modify policies, or develop and share data.³²

Empowering cities and city-systems will not result only from the actions of policy makers within municipal governments, state authorities, or international donors. It also emerges from the actions of community-based organizations to defend their neighborhoods, from municipal governments developing their own municipal foreign policies or regulations as acts of resistance, from the masses of urban dwellers finding direct and indirect ways to resist the power of the world system and of the state in their lives. Transnational oppositional politics emerges from multiple acts of agency. (Smith 2001:164)

City and city system level policies will not be developed or implemented easily: elites profiting from the state structure will not devolve power and resources without a struggle. Yet the MENA city has already shown autonomy of action, and city systems exist. There is perhaps more strength and potential at this level than expected. In particular, urban inhabitants of cities like Rabat, Cairo, Istanbul and Dubai, no matter their inclusiveness into the chains of globalization, appear to be aware of a new level of possibility for cities. Local discourse, representation, arguments over public space, identity, are now reflecting an awareness of the global and of the local at the same time. People are using "transnational" linkages to further their local interests, and this both empowers them and modifies the discourse within the city/state dynamic.

The development discourse once argued that national boundaries confined the limits of social change, and that the battle was "between the state as modernizer vs. society as the locus for tradition." We now realize that this narrative was simplistic and intentional, and that our conceptual tools were wanting. More importantly, this view disenfranchised the community in the process. If we want empowerment, improved social welfare, resistance, employment, we must recenter the level of city networks and their linkage to other cities in our calculations, since it is at this level where citizens utilize their "transnational" or extra-local networks for exchange and empowerment.

We certainly do not want thoughtless deeper integration and increased dependency with the core of the world system. Current policies only promise this. By returning to a reconceptualized sense of the MENA region as a city system based on nodes embedded within networks of exchange, we may (re)invent the future system, as Wallerstein suggests. Is this such a radical idea, this utopian alternative to the distopias of the failed state, the failed region, the failed autonomous city? It is not radical, just revolutionary.