Help with 600 word sociology paper

Global and World Cities: A View from off the Map*

JENNIFER ROBINSON

There are a large number of cities around the world which do not register on intellectual maps that chart the rise and fall of global and world cities. They don’t fall into either of these categories, and they probably never will — but many managers of these cities would like them to. Some of these cities find themselves interpreted instead through the lens of developmentalism, an approach which broadly understands these places to be lacking in the qualities of city-ness, and which is concerned to improve capacities of governance, service provision and productivity. Such an approach supports some of the more alarmist responses to mega-cities, which are more commonly identified in poorer countries. But for many smaller cities, even the category mega-city is irrelevant. My concerns in this article extend beyond the poor fit of these popular categories, though. I would like to suggest that these widely circulating approaches to contemporary urbanization — global and world cities, together with the persistent use of the category ‘third-world city’ — impose substantial limitations on imagining or planning the futures of cities around the world.

Part of the adverse worldly impact of these urban theories is, I argue, a consequence of the geographical division of urban studies between urban theory, broadly focused on the West, and development studies, focused on places that were once called ‘third-world cities’. This division might simply be an innocent acknowledgement of difference (Szelenyi, 1996). However, apart from the value-laden historical meaning of these categorical ascriptions, the persistent alignment of a ‘theory’/‘development’ dualism with the ‘West’/‘third world’ division in urban studies, suggests otherwise.1 One of the consequences of these overlapping dualisms, is that understandings of city-ness have come to rest on the (usually unstated) experiences of a relatively small group of (mostly western) cities, and cities outside of the West are assessed in terms of this pre-given

ß Joint Editors and Blackwell Publishers Ltd 2002. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

* I should like to thank people who attended various seminars where early versions of this article were presented — at Birmingham, Loughborough and Kings College London Geography Departments; at the OU- Bristol-Durham Cities Seminars, the 2000 AAG and the Urban Futures Conference in Johannesburg. All the discussions made me think harder about these points and improved the article, as did the individual comments of Richard Tomlinson, Steve Pile, Robert Beauregard and the journal’s anonymous referees. I would also like to thank students at the LSE who sat through very anxious presentations of a course I had been asked to give on ‘Third World Urbanization’, but who luckily seemed to be far more comfortable with this categorization than their teacher! Various people commented on other papers which helped a lot with the focus for this one: Kevin Cox, Edgar Pieterse, Maliq Simone, and the OU course teams for DD304 and U213. Thank you!

1 This argument would seem to hold also for other cities ‘beyond the West’, such as post-Socialist cities (Andruszet al., 1996). World city theorists now frequently assign cities in former socialist countries to a separate ‘exceptional’ category, associating their globalization with the privatization of former national services and industries (Beaverstocket al., 1999; Taylor, 2000).

International Journal of Urban and Regional Research Volume 26.3 September 2002 531–54

standardof (world) city-ness,or urban economicdynamism.This article exploresthe extent to which more recent global and world city approaches,although enthusiastic about tracking transnationalprocesses,have nonethelessreproducedthis long-standing division within urbanstudies.

As partof aneffort to promotea morecosmopolitanapproachto urbanstudies,I want to tracesomepathsacrossthis entrencheddivision betweentheoryanddevelopment.I do this by reflectingon somefashionableapproachesto cities from a positionoff their maps. Of course,the cities I am concernedwith are most emphaticallyon the map of a broad rangeof diverseglobal political, economicandcultural connections,but this is frequently discountedand certainly neverexploredwithin thesetheoreticalapproaches.Thereis a needto construct(or promote)an alternativeurbantheorywhich reflectsthe experiences of a muchwider rangeof cities. This will involve disruptingthe narrowvision of a (still) somewhatimperialist approachto cities, which has been reinforced by the strident economismin accountsof global and world cities (Smith, 2001). Elementsof urban theoryhavebecometransfixedwith the apparentsuccessanddynamismof certainstylish sectorsof the global economy,despite (and perhapsbecauseof) their circumscribed geographicalpurchaseand most unappealingconsequences.These studies have been valuable,andoffer greatinsight into the limited part of the world andeconomythat they study.My suggestion,though,is thattheseinsightscouldbeincorporatedin a broaderand lessambitiousapproachto cities aroundthe world, an approachwithout categoriesand more inclusive of the diversity of experiencein ordinary cities.

After reviewing some of the effects of global and world city approachesfrom a viewpoint off their maps,I draw attention to somealternativeapproachesto different cities. Moving beyondcomparativestudies(which are a good starting point), I suggest that recentefforts to proposean accountof ‘ordinary cities’ (Amin and Graham,1997) offer an asyet unrecognizedopportunityto developa morecosmopolitanaccountof city- ness. By this I mean to build on James Clifford’ s (1997) idea of ‘ discrepant cosmopolitanisms’,rather than a universalizingor homogenizingcosmopolitanimpulse (see Cheah and Robbins, 1998). Clifford’s interest is in delocalizing professional anthropology,partly by insistingthatculturesarenot (only) localizedandthatthevillage- based ethnographyis a limited and misleading researchtool (see also Gupta and Ferguson,1999). Connectionsand travels beyond the local are long-standing and constitutiveof local culturesall over the world: an importantconsiderationfor any urban scholar (Smith, 2001). But, importantly for this article, he also wants to encourage anthropologiststo considerthe trajectoriesof their own practicesandanalyses.Urbanists, too, could find it valuableto think aboutthe contrastbetweenthe restrictedspatialitiesof their theories— the geographyof urbantheory — and the diversecosmopolitanismsof the cities they write about.

My primaryconcern,then,is with the persistenceof a split betweenaccountsof cities in countrieswhich havebeenlabelled‘third world’ and thosein the ‘West’. Put simply, the segregati on i s between ci ti es which are captured through the rubric of ‘developmentalism’(not [yet] cities2) and cities which are thought through to produce (un/located)theory. My contention is that ‘urban theory’ is basedprimarily on the experiencesand historiesof westerncities — much as Chakrabarty(2000) suggeststhat the theoriesand categoriesof historical scholarshiphave beenrooted both in western experiencesand their intellectual traditions. And, like him, I want to suggestthat restructuringtheterrainon which differentkindsof cities arethoughtwithin urbanstudies could enhancethe understandingof cities everywhere.Many writers on cities outsidethe ‘West’ continueto complainthat urban‘theory’ hasa restrictedpurchase— with the cry that ‘Western theories of urbanization are not relevant here’ — in Cairo, or the

2 Herewe might find comparisonwith the dynamicHomi Bhabha(1994)discussesin the relationbetween colonizerand colonized,underthe rubric — not quite/notwhite.

ß Joint Editors and Blackwell PublishersLtd 2002

532 JenniferRobinson

Philippines, Malaysia or Jakarta(for example,Kelly, 1999; McGee, 1995; Dick and Rimmer,1998; Stewart,1999).This seemsto imply that thesecities are ‘different’, that they belongin a different category.Alternatively, it could confirm my point that existing ‘theory’ is locatedelsewhere— it is not developedin relation to the experiencesand trajectoriesof thesecities.3 Looked at from this perspectiveit is clear that there is a geographyto urbantheory, and one which needsaddressing.

This is not a new observation.In 1990,Anthony King wrote that:

The questionis whether the real developmentof London or Manchestercan be understood without referenceto India, Africa, and Latin America any more than can the developmentof Kingston (Jamaica)or Bombay be understoodwithout the former. Nevertheless,the real division of scholarship,as well as the ideologicalunderpinningsthat help to keepthem alive, ensurethathistoriesof ‘First’, ‘Second’and‘Third’-World citiesarestill kepttidily apart(King, 1990: 78).

In his view, and mine, urbanstudiesis deeplydivided againstitself. It is my contention that this diminishesthe vitality (andcertainly narrowsthe purchase)of urbantheoryand also, perhapsmore importantly, has certain consequencesin the field of urban policy which we oughtto be concernedabout.Anthony King’s work (1990;1995),anda rising interestin both globalizationand post-colonialismhasmeanta partial redressingof this divide, especiallywithin historical writing and a more culturally-inflectedurbanstudies (seeespeciallyRabinow,1989; Jacobs,1996; Ross,1996; Driver and Gilbert, 1998).

More hopefully, in the decadeor so sinceKing madethis observation,urbanstudies has seenthe consolidationof a significant interestin ‘world cities’ and ‘global cities’ which seeksto understandthe position and functioning of cities within the world- economy.The situation,then,may appearmorepropitiousthaneverfor an integrationof urbanstudiesacrosstheselong-standingdivisionsof Westandthird world. An analytical focus on the transnational global economy could ensure that taken-for-granted categorizationsof cities (western,third-world, African, Asian, socialist etc.) will no longer be of any relevance.Indeed,this is a claim madeby the key advocatesof these approaches(Sassen,2001;Taylor, 2001).The focusof urbantheorycould arguablyshift to understandingthe diversity of urbanexperiencesandcities within the world economy. Could this be the basisfor a more‘cosmopolitan’accountof cities,ratherthanonethat is divided, resting on partial and limited areasof the globe, and quite divergent sets of concernsor subjectmatter?

Thearticle addressesthis questionby reviewing,in turn, world cities andglobalcities approaches,the developmentalistapproacheswhich currently frame most writing on cities in poorercountries,andemergingtransnationaland‘ordinary’ city approaches.The argumentis that althoughthere is much scopein theseapproachesfor broadeningour understandingof cities and their futures,there is still considerablework to be done to produce a cosmopolitan, postcolonial urban studies. Moreover, the stakes are considerablyhigher than analytical correctnessor theoretical insight. The dearth of alternativevocabulariesandapproachescurrentlyseverelylimits imaginationsof possible futures for cities. The particular form of this limitation, makesit particularly hard to mobilize creativewaysto addressthe situationof poor andmarginalizedpeoplein cities aroundthe world.

3 One recent exampleof this is to be found in Marcuseand van Kempen (2000). Despite specifically inviting people to write in the volume about cities outside of North America/westernEurope, the introductory theoreticalmaterial makesno attempt to considerwhether literature on theseother cities might suggestalternativeapproaches.The categoryof ‘third-world city’ is usedwithout reflection — despitedecadesof suggestionsfrom peoplewho work on theseplacesthat it is an inaccuratecategory— and even the contributionswithin the volume from theseparts of the world are simply relegatedto a positionof difference.The discussionof the demiseof the welfarestate,for example,makesno attemptto considerthe relationshipbetweenUS/Europeanexperiencesand the dramatic and widely documented destructionof statecapacitiesin the nameof structuraladjustmentpolicies.

ß Joint Editors and Blackwell PublishersLtd 2002

Global and world cities: a view from off the map 533

Global and world cities

Threekey conclusionsof theinitial roundsof world cities researchcanbesummarizedas:

• World cities articulateregional,national and internationaleconomiesinto a global economy.They serveas the organizingnodesof a global economicsystem.

• World cities can be arrangedhierarchically,roughly in accordwith the economic powerthey command— competitionbetweenworld cities andthe impactof external shocksshapethe fortunesof world cities and their position in the hierarchy.Cities canriseandfall throughthehierarchy,andtheir positionis determinedby therelative balanceof global, nationaland regionalinfluence.

• Many populationsare excludedfrom the spaceof global capitalism,and thus from the field of world cities: they are ‘economicallyirrelevant’ (Knox, 1995: 41).

In his accountof cities acrossthe world, King provocativelynotedthat ‘all cities today are ‘world cities’’ (1990a:82). Unfortunately,researchand writing within the rubric of the world cities approach,or hypothesis,has generally not chosento build on this observation.Rather,in consideringthe dynamicsof the world economyin relation to cities, a structuralanalysisof a small rangeof economicprocesseswith a certain‘global’ reachhastendedto crowdout anattentivenesswithin urbanstudiesto theplaceandeffect of individual cities (King, 1995)andthe diversity of wider connectionswhich shapethem (Allen et al., 1999).Althoughstatuswithin the world city hierarchyhastraditionally been basedon a rangeof criteria, including nationalstanding,location of stateand interstate agenciesand cultural functions,the primary determinationof statusin this frameworkis economic— as Friedmann(1986: 317) notes:‘The economicvariable is likely to be decisivefor all attemptsat explanation’.This hasbecomemore,not less,apparentin the world cities literature,especiallyas the approachhasbeenclosely tied to world-systems theory,andasmorerecentresearchhasfocusedon identifying the transnationalbusiness connections which define the top rank of world cities, labelled ‘ global cities’ (Beaverstocket al., 1999; Sassen,2001).

In world-systemstheory more generally, countries acrossthe world are seen to occupya placewithin the hierarchyof the world-economy,andpossiblymaketheir way up through the categories(core, periphery, semi-periphery)embeddedin the world- economyapproach.Following this, the world cities approachassumesthat cities occupy similar placingswith similar capacityto progressup or fall down the ranks.The country categorizationsof core, periphery and semi-peripheryin world-systemstheory have thereforebeentransferredto the analysisof cities, and overlain, albeit with a slightly different geography,on an extant but outdatedvocabularyof categorizations(such as first/third world) within the field of urbanstudies.

JohnFriedmannasks,in his review of ‘World city research— 10 yearson’, whether the world city hypothesis‘is a heuristic,a way of askingquestionsaboutcities in general, or a statementabout a classof particular cities — world cities — set apart from other urban agglomerationsby specifiablecharacteristics?’(1995: 22). He suggeststhat it is both; but that the tendencyhasbeento categorizecities into a hierarchy,in which ‘world cities’ areat the top of the treeof influence.This ‘leaguetable’ approachhasshapedthe waysin which cities aroundthe world havebeenrepresented— or not representedat all — within the world cities literature.From the dizzy heightsof the diagrammer,certain significant cities are identified, labelled,processedand placedin a hierarchy,with very little attentivenessto thediverseexperiencesof that city, or evento extantliteratureabout that place.The dangerhereis that out of date,unsuitableor unreliabledata(Shortet al., 1996;althoughseeBeaverstocket al., 2000),andpossiblya lack of familiarity with some of the regionsbeing consideredcan lead to the productionof mapswhich are simply inaccurate.Theseimagesof the world (of important) cities have beenusedagain and

ß Joint Editors and Blackwell PublishersLtd 2002

534 JenniferRobinson

againto illustrate the perspectiveof world cities theorists.Pressingthe analysistowards an emphasis on the more limited range of transnational business connections characteristicof global cities, Peter Taylor (2000: 14) notes with disapproval the ‘widespreadreportingof . . . a preliminary taxonomy’ of world cities. Revisedversions, though,basedon a morerestrictedrangeof criteria (connectednessto globalizingwestern producerservicesfirms) draw remarkablysimilar conclusions,and maps.

In both the broaderandthe morenarroweconomisticapproachto identifying world cities, a view of the world of cities emergeswheremillions of peopleand hundredsof cities are dropped off the map of much researchin urban studies, to service one particularandvery restrictedview of significanceor (ir)relevanceto certainsectionsof the global economy.Perhapsmore importantly, this methodologyrevealsan analytical tensionbetweenassessingthe characteristicsand potential of cities on the basisof the processeswhich matter as viewed from within their diverse dynamic social and economicworlds (which, of course,alwaysstretchway beyondany physicaledgeto the city), or on the basisof criteria determinedby the externaltheoreticalconstructof the world or global economy(seealsoVarsanyi,2000).This is at the heartof how a world cities approachcan limit imaginationsaboutthe futuresof cities, which I will return to below.

World cities research,then,hasmovedon from the time of Friedmann’sreview and, muchinfluencedby SaskiaSassen’s(1991;2001)study,Theglobal city, hasadoptedan empirical focus on transnationalbusinessand financenetworks(e.g. Beaverstocket al., 1999; Morshidi, 2000).Nonetheless,althoughaiming to emphasizeconnectionsand not attributes(Beaverstocket al., 2000),andin the mostrecentwork to suggesta moreglobal reach for the world city networks of non-commandcentres(Taylor et al., 2001),4 a limited rangeof cities still end up categorizedin boxesor in diagrammaticmaps,and assigneda placein relation to a priori analyticalhierarchies.

In a prominentcontributionto the world cities argument,SaskiaSassen(1991;1994) hascoinedthe term ‘global cities’ to capturewhat shesuggestsis a distinctivefeatureof the current (1980s on) phase of the world economy: the global organization and increasinglytransnationalstructureof key elementsof the global economy.Her key point is that the spatially dispersedglobal economy requires locally-basedand integrated organization, and this, she suggests,takes place in global cities. Although many transnationalcompaniesno longerkeeptheir headquartersin centralareasof thesemajor cities,the specializedfirms which theyrely on to producethe capabilitiesandinnovations necessaryfor commandandcontrol of their global operationshaveremainedor chosento establishthemselvesthere.Moreover,it is no longer the large transnationalcorporations which arethecentreof thesefunctions,but small partsof a few major cities,shesuggests, which play host to and enablethe effective functioning throughproximity of a growing numberof new producerandbusinessservicesfirms (Sassen,2001).A similar argument concerningthe benefits of co-location for finance and investmentfirms, suggeststhat thesecutting-edgeactivitiesareproducedin a few major cities.Co-locationbenefitsboth thesesetsof firms as this facilitates face-to-faceinteractionand the emergenceof trust with potentialpartners,which is crucial in termsof enablinginnovationandcopingwith the risk, complexity andspeculativecharacterof many of theseactivities (Sassen,1994: 84).

While therehavebeenmany criticisms of the empirical basisfor claims that ‘global cities’ are significantly different from other major centres(Abu-Lughod,1995; Short et

4 In this working paper,the very lowestlevel of connectivityto a systemof transnationalfirms is identified ascharacterizingCairo andCasablanca,LagosandNairobi, JohannesburgandCapeTown. But the power relationsin this hierarchyof cities leavethe authorsconcludingthat ‘from this powerperspectiveSassen’s notion of global cities transcendingthe North-Southdivide seemsa trifle sanguine;globalisationbeginsto look very ‘western’ as soonas we look at direct expressionsof power’ (Taylor et al., 2001: 7).

ß Joint Editors and Blackwell PublishersLtd 2002

Global and world cities: a view from off the map 535

al., 1996; Storper, 1997; Smith, 2001), the global city approachdeploys a strong analytical emphasison process.The categoryof global city which is identified through this analysis,though,is foundedupona minor setof economicactivities basedin only a small part of thesecities. They may constitutethe more dynamicsectorsof thesecities’ economies,but Sassen’sevidenceof declininglocationquotientsfor theseactivitiesin the 1990s(e.g. 2001: 134–5)suggeststhat the concentratedgrowth spurt in this sectormay well be over. And to put thesesectorsin perspective,in London, for example,where transnational finance and business services are still most dynamic and highly concentrated,the LondonDevelopmentAgencysuggeststhat ‘about 70% of employment (in London) is in firms whosemain marketis nationalratherthan international’ (LDA, 2000: 18).

The discursive effectivenessof the global city hypothesisdependson the pithy identificationof the ‘global city’ — a categoryof cities which areclaimedto be powerful in termsof the global economy.5 Mirroring the world city emphasison a limited rangeof economicactivitieswith a certainglobal reach,aswell asits categorizingimperative,the globalcity approachhasa similar effect,droppingmostcities in theworld from its vision. If the ‘global city’ were labelled as just another example of an ‘industrial’ district (perhapsit shouldratherbe called: new industrial districts of transnationalmanagement andcontrol),it might not haveattractedthe attentionit did. But on the positiveside,some of the consequencesfor cities in poorerpartsof the world might havebeenavoided.

‘Filling in the voids’:6 off the world cities map

Accordingto Sassen,functionsof commandandcontrol of the global economyalsotake placein someformerly peripheralcities, which coordinateglobal investments,aswell as financial and businessservicesregionally.In her view, this signifies the emergenceof a new geographyto the periphery— a selectgroupof cities, somein poorercountries,are now deemedto have‘global city functions’ althoughthey fall short of being first-order global cities. ShementionsToronto, SaoPaulo,Mexico City, Miami and Sydney.This signalssomethingof the ‘end of the third world’ (Harris, 1986) as a categoryin urban studies.Nonetheless,Sassenacknowledgesthat her approach‘cannot accountfor the casesof manycities that may not haveexperiencedany of thesedevelopments’(1994:7).

Despite this, Sassenjoins others in consigning substantialareas of the globe to structural irrelevance:‘significant parts of Africa and Latin America becameunhinged from their hithertostrongties with world marketsin commoditiesandraw materials’(ibid.: 27); and: ‘Alongside thesenew global and regionalhierarchiesof cities is a vast territory that hasbecomeincreasingly peripheral,increasinglyexcludedfrom the major economic processesthat fuel economicgrowthin the newglobal economy’(ibid.: 4). Knox goeseven further to suggestthat ‘the megacities of the peripherywill fare no betterthanthe catatonic agrariansocietiesthat have fuelled their (demographic)growth, and in which both will lapsedecisivelyand irretrievably into a ‘slow’ economic time zone’ (1995: 15).

There are obviously important ways in which the changing geography of the international economyhas impactedon cities in poorer countries.As Sassen(1994: 83)

5 There are many problemsassociatedwith assigning‘power’ to the categoryof global city, when the capacitiesto control andcommandwhich arebeingidentified arelocatedin certainactorsandinstitutions within a small part of the city’s economy.‘Cities’ ascollective actors,may in fact be renderedrelatively powerlessin thesecontexts.Moreover,asJohnAllen (1999)astutelypointsout, it is unclearwherepower is produced:in the sites Sassenidentifies (the city neighbourhoods)or in the network interactionsand flows. Taylor et al. (2001)arebeginningto explorethe ideaof networkedpower.But the sloppythinking that elidescategoriesof cities with economicagentsin relationto powercontinuesto havepurchaseand, as I will arguebelow, importanteffects.

6 From Beaverstocket al. (1999: 457).

ß Joint Editors and Blackwell PublishersLtd 2002

536 JenniferRobinson

suggests,the placeof the ‘developing’ countriesas sitesof investmentfor westernbanks has declined precipitously since the 1970s oil surpluseswere recycled through poor countries;Latin America hasbeenreplacedby SouthEastAsia as the top destinationfor investmentin manufacturingby ‘highly developed’countries;andmanypoor countriesare now net exportersof capital(ibid.: 63). Nonetheless,within this approach,the international financial centresof many countriesperform important‘gatewayfunctions’ for the flow of finance and global businessservices(ibid.: 173). I certainly appreciatethat the focus of global and more recentworld cities work is on a limited setof economicactivities,which are assumingan increasinglytransnationalform, and in which relatively few cities can hopeto participate.But it is the leapfrom this very restrictedandclearly definedeconomic analysis,to claims regarding the successand power of these few cities, their overall categorizationon this restrictedbasis,and the implied broaderstructuralirrelevanceof all othercities, which is of concern.Thesetheoreticalclaimsandcategorizingmovesareboth inaccurateand harmful to the fortunesof cities defined‘off the map’.

The ‘end of the third world’ is perhapsan accurateassessmentof changesover the last threeto four decadesin placeslike Hong Kong, Singapore,Taiwan,SouthKoreaand evenMalaysia,andthe appearanceof thesecity-statesor major urbancentresin rostersof first- and second-orderglobal cities reflectsthis. But in partsof the world whereglobal cities havenot beenidentified — the ‘voids’ of world andglobal city approaches— the experienceof many countriesandcities hasbeenmuch more uneventhan the analysesI have cited suggest.For many, the 1980s and 1990s have been long decadesof little growth and growing inequality. It is, however,inaccurateto caricatureeventhe poorest regionsas excludedfrom the global economyor doomedto occupya slow zoneof the world economy.Africa, frequentlywritten off in theselarge global analyses,hashad a very unevengrowth record.As the African DevelopmentReport(2000: 1) notes:

While the continenthas,in overall terms,laggedbehind other regions,a few countrieshave produced remarkable economic results, even by world standards. . . In an encouraging development,as many as 12 countriesare estimatedto have recordedreal GDP growth rates above5% while closeto 30 countrieshad positive real GDP per capitagrowth.

It is hardto disagreethat somecountriesandcities havelost manyof the tradingand investmentlinks that characterizedan earliereraof global economicrelations.A country like Zambia,for example,now oneof the mostheavily indebtednationsin the world, and certainlyoneof the poorest,hasseenthe valueof its primary export,copper,plummeton the world market since the 1970s.Its position within an older internationaldivision of labouris no longereconomicallyviable, andit hasyet to find a successfulpathfor future economic growth (Young, 1988; Bonnick, 1997). En route it has suffered the consequencesof one of the World Bank/IMF’s most ruthless Structural Adjustment Programmes(Young, 1988; Clark, 1989). However, Zambia is also one of the most urbanizedcountrieson theAfrican continent,andits capitalcity, Lusaka,is a testimonyto the modernistdreamsof both the former colonial powers and the post-independence government (Hansen, 1997). Today, with over 70% of the population in Lusaka dependenton earningsfrom the informal sector(governmentbureaucratsare known to earnlessthan somestreettraders:Moser and Holland, 1997),the oncebright economic and social future of this city must feel itself like a dream— albeit one which was for a time very real to many people(Ferguson,1999).

Lusakais certainlynot a playerin the ‘major economicprocessesthat fuel economic growth in the new global economy’(Sassen,1994: 198). But copperis still exported,as are agricultural goods,and despitethe huge lack of foreign currency (and sometimes becauseof it) all sortsof links and connectionsto the global economypersist.From the World Bank, to aid agencies,internationalpolitical organizations,and trade in second- hand clothing and other goodsand services,Lusakais still constitutedand reproduced throughits relationswith other partsof the country, other cities, and other parts of the

ß Joint Editors and Blackwell PublishersLtd 2002

Global and world cities: a view from off the map 537

region and globe (for example,Hansen,1994; 1997). The city continuesto perform its functionsof nationalandregionalcentralityin relationto political andfinancial services, and operatesas a significant market (and occasionallyproduction site) for goodsand servicesfrom acrossthe country and the world.

It is onething, though,to agreethat global links arechanging,somearebeingcut, and that power relations, inequalities and poverty shapethe quality of those links (see,for example,Halfani, 1996). It is quite anotherto suggest that poor cities and countriesare irrelevant to the global economy.When looked at from the point of view of theseplaces which are allegedly ‘off the map’, the global economyis of enormoussignificance in shapingthe futures and fortunesof cities aroundthe world. For many poor, ‘structurally irrelevant’ cities,thesignificanceof flows of ideas,practicesandresourcesbeyondandinto the city concernedfrom around the world standsin stark contrast to these claims of irrelevance;asShatkin(1998:381)writesaboutPhnomPenh:‘In orderto arrive at a proper understanding of the processof urbanisationin LDCs, it is necessaryto examinethe ways in which countriesinterfacewith the global economy,as well as the social, cultural and historical legaciesthat eachcountry carriesinto the era of globalisation’.Thesehistorical legacies,it is clearfrom his account,arethemselvesproductsof earlier ‘global’ encounters, still very much alive in shapingthe global significanceof that placein the present.

And to pursuea morepolemicalline, mineral resourcescrucial to the global economy aredrawnfrom someof the poorestcountriesof the world (asmobile phonesdependon a mineralfound only in Zaire/DRC),wherefinanciersandtransnationalfirms negotiatewith warlords, corrupt governmentsand local armies to keep profits, production and exports flowing (Mbembe,2001). Widening the compassof analysismight help to encouragea morecritical edgeto the global andworld cities literature.Moreover,it is preciselythrough avoiding ‘risky’ investments(and pursuing vastly exploitative and violent forms of extractioninstead)in the poorestcountriesandcities in the world that the westernfinancial ‘mode of production’is ableto aim to securethe stableshareholderreturnswhich maintain post-Fordistfinance basedeconomies(Boyer, 2000). To the extent that they are absent from this aspectof the global economy,theseplacesmay well be centralto sustaining it.

On a morepositivenote,viewedfrom off the(world cities) map,in its initial versions the world city hypothesisdoessuggesta rangeof criteria by which to assessthe role and functionsof different cities — whetherthey arecentresof decision-makingandauthority in the registersof economic,cultural and political information (Friedmannand Goetz, 1982;Friedmann,1986;see,for example,Simon,1995;Hill andKim, 2000;Kelly, 2000; Tyner, 2000;Olds andYeung,2002).This meansthat the distinctiverole of quite a wide variety of cities canbe broughtinto view usingthis approach.Similarly, the spatialreach of a city’s influencecanvary, andthereis scopefor thinking aboutcities whoseprimary influencesaremoreto do with their hinterlandandnation,thanwith the global economy. So many more cities might comeinto view as significant provincial centres,political or symbolic centres,or perhapsas importanttransportand productionhubsin nationaland regional economies.Guardingagainsteconomicreductionismand moving beyondthe limitations of the global scaleof transnationalactivities would ensurethat the rangeof cities of concernto world cities theoristsis lessexclusive(Varsanyi,2000).

But there is still King’s claim that ‘all cities are world cities’, which we need to consider.And the fact that the world cities literature, even in its most nuancedform, persistsin defining somecities out of the game,as‘excludedfrom global capitalism’and thereforeas irrelevant to their theoretical reflections. Writers on cities in Africa, for example,askedto considerworld cities in their region,concludedismally that thereare no world cities on the continent— althoughthey point to Cairo and Johannesburg7 as potentials(Rakodi, 1997). Scholarsof other peripheralplaces,such as Latin America,

7 And Beaverstocket al. (1999: 455) notein their rosterof world cities that ‘Africa hasits first city in our list, Johannesburg,but thereare still no world cities found in SouthAsia or the Middle East’.

ß Joint Editors and Blackwell PublishersLtd 2002

538 JenniferRobinson

wonderaboutthe usefulnessof thesecategoriesin ‘analysingwhat is occurring’ (Gilbert, 1998:174), andthey havelittle relevanceto placesin SouthAsia or in the Middle East/ North Africa. As Stanley(2001:8) writes,‘cities in this regionarenot on the world map’.

If the categoryof world city is not applicableto a wide rangeof cities (Simon,1995), are there other ways in which the world city hypothesismight be mobilized in these ‘irrelevant’ cities? A strongerfocus on processthan categoriescould lead one to think abouthow ‘global’ economicprocessesaffect all cities — as Marcuseand van Kempen frame it, this leadsto a focus on ‘globalizing cities’, since‘globalisation. . . is a process that affectsall cities in the world, if to varying degreesandvarying ways,not only those at the top of the global hierarchy’ (2000: xvii). This formulation still leaves the enthusiasmabouthierarchiesandcategoriesin place,though,andretainsan emphasison economicactivities with a ‘global’ reach, but at least it identifies a researchagenda applicableto a wider rangeof cities.

And most importantly, perhaps, but seldom mentioned, the particular ‘global economy’which is beingusedasthe groundandfoundationfor identifying both placein hierarchy and relevant social and economicprocesses,is only one of many forms of global andtransnationaleconomicconnection.8 The criteria for global significancemight well look very different were the map-makersto relocate themselvesand review significanttransnationalnetworksin a placelike Jakarta,or Kuala Lumpur, whereties to Islamic formsof global economicandpolitical activity might resultin a very different list of powerful cities (Allen, 1999;Firman,1999;seealsoWhite, 1998on the ethnocentrism of theseapproaches).Similarly, the transnationalactivities of agencieslike the World Bank andthe IMF who drive the circulation of knowledgeandthe disciplining powerto recover old bank and continuing bi-lateral and multi-lateral debt from the poorest countriesin the world (debt, it should be pointed out, which in an earlier phasethese agenciesrecommendedto poor countries)would draw anothercrucial graph of global financial and economicconnectionsshaping(or devastating)city life.

Despiteits investmentin analysisof the world-economyandtransnationaleconomic processes— andin somewaysbecauseof it — the world cities approachcontinuesto see cities throughthe lensof categories,andto privilege the westasthe sourceof economic dynamismand globalization.In this senseit persistsin the lineage of an approachto urbanstudieswhich divided up the field of cities accordingto pre-givencriteria, suchas western,or third world. World city theoristsalso bring theseolder categoriesinto play evenas they try to capturethe ways in which the world is more complicatedthan this. Knox writes, for example, that ‘just as we can see the world city-ness of regional metropoli,so we canseethe Third-World-nessof world cities’ (1995:15). The category of world city is addedin to complicatethe pre-existingcategorizationof cities (andvice versa),but theyarestill takento bemeaningfuldescriptorsof the world of cities.Or cities beyondthe world city radar are simply not mentioned.For thesediversecities off the map, alternativeanalyticalapproachesbeckon.

Developmentalismand third-world cities

Scholarsof so-called‘structurally irrelevant’ cities find it hard to pursueresearchand policy within the frameof world cities theory,althoughsomefind creativewaysto apply it (e.g. Tyner, 2000). As Browder and Godfrey (1997: 45) point out: ‘Beyond general

8 It is perhaps appropriatehereto raisesomequestionsaboutthe datausedto classify world cities — data which in Beaverstocket al. (1999)includesno Japanesebanksandonly US/UK/AustralianandCanadian law firms. Thesefirms may be the largestin the largesteconomiesin the world, andthusarguablydriving the global economy,but this methodologyomits potentially significant dimensionsof different economic globalizationsandwill fail to capturethe regionalsignificanceof somecentres,e.g.in the contextof West Asian and North African cities (seeTaylor, 2001).

ß Joint Editors and Blackwell PublishersLtd 2002

Global and world cities: a view from off the map 539