national and transnational urban systems€¦ · 03-sassen.qxd 3/21/2006 11:21 am page 45....

3National and

Transnational Urban Systems

The trends described in Chapter 2 point to the emergence of a new kindof urban system, one operating at the global and transnational regional

levels (Taylor 2004; Marcatullio and Io 2001). This is a system in which citiesare crucial nodes for the international coordination and servicing of firms,markets, and even whole economies that are increasingly transnational. Thisglobal map of the organizational side of the world economy needs to be dis-tinguished from the global map of the consumption of globally distributedgoods and services. And it needs to be distinguished from the global map offoreign direct investment; the fact of a few foreign direct investment projectsdoes not necessarily make a city part of the organizational map. First, the for-eign direct investment and, especially, the consumption map are far wider andmore diffuse than the organizational map, which is strategic. Second, thesecities also emerge as strategic places in an emergent transnational politicaland cultural geography. The number of cities constituting the organizationalmap and these novel political and cultural geographies grew sharply duringthe 1990s, because the global economy expanded vastly as more and morecountries, often under pressure, adopted the deregulatory and privatizingpolicies required for joining the global corporate system.

Most cities, however, including most large cities, are not part of thesenew transnational urban systems. Typically, urban systems are coterminouswith nation-states, and most cities exist within these national geographies.

45

03-Sassen.qxd 3/21/2006 11:21 AM Page 45

Correspondingly, with rare exceptions (Chase-Dunn 1984; GaWC[Globalization; and World Cities Study Group and Network] 1998; Sassen[1991] 2001; Walters 1985; Timberlake 1985), studies of city systems haveuntil recently assumed that the nation-state is the unit of analysis. Althoughthis is still the most common view, there is a growing scholarship that allowsfor the possibility that intercity networks can cross national borders directly,bypassing national states as these have reduced older gatekeeping functionson cross-border economic flows. This novel focus is partly a function ofactual changes in the international sphere, notably the formation of globaleconomic processes discussed in Chapter 2, and the accompanying deregu-lation and opening up of national systems.

In this chapter, I ask, What is the impact of economic globalization onnational urban systems? Does the globalization of major industries, fromauto manufacturing to finance, have distinct effects on different types ofnational urban systems? I focus on the effects of the shift to services and eco-nomic globalization on balanced and primate urban systems, the two majortypes of urban systems that have been identified in the research literatureon cities. Western European nations have typically been regarded as a goodexample of balanced urban systems; and Latin American nations, as a goodexample of primate systems—that is, inordinate concentrations of popula-tion and major economic activities in one city, typically the national capital.The most recent research signals some sharp changes in these two regions.Later in this chapter, I turn to the emergence of transnational urban systems.

Impacts on Primate Systems: The Caseof Latin America and the Caribbean

Many regions in the world—Latin America, the Caribbean, large parts ofAsia, and (to some extent) Africa—have long been characterized by urbanprimacy as an older scholarship has established (Abreu et al. 1989; Doganand Kasarda 1988; Feldbauer et al. 1993; Hardoy 1975; Lee 1989; Linn1983; Lozano and Duarte 1991; Stren and White 1989). Exhibit 3.1 showssome of the main cities in Latin America I discuss here; while many are pri-mate cities, which are also national capitals, some are neither. Primate citiesaccount for a disproportionate share of population, employment, and grossnational product (GNP), a fact illustrated by the figures presented in Exhibit3.2. For example, in 1970, greater São Paulo accounted for 36% of nationaldomestic product (NDP) and 48% of net industrial product in Brazil, acountry with several major economic regions. Santo Domingo accounted for56% of industrial growth and 70% of commercial and banking transactions

46——Cities in a World Economy


National and Transnational Urban Systems——47

in the Dominican Republic in 1981. And Lima accounted for 43% of grossdomestic product (GDP) in Peru in 1980.

Primacy is not simply a matter of absolute size, nor is large size a markerof primacy. Santo Domengo or Lima (Exhibit 3.2), both primate cities, arenot necessarily among the largest in the world. Primacy is a relative condi-tion that holds within a national urban system. Some of the largest urban

Exhibit 3.1 Select Cities in Latin America and the Caribbean, 2005


48

Pub

licP

ublic

U

rban

Are

aY

ear

Pop

ulat

ion

Em

ploy

men

tR

even

ues

Exp

endi

ture

sO

utpu

t M

easu

re

Bra

zil

36.0

of

ND

PG

reat

er S

ão P

aulo

1970

8.6

——

—48

.0 o

f ne

t in

dust

rial

pro

duct

Chi

naSh

angh

ai19

801.

2—

——

12.5

of

gros

s in

dust

rial

pro

duct

Dom

inic

an R

epub

licSa

nta

Dom

ingo

1981

24.0

70.0

of

com

mer

cial

and

bank

ing

tran

sact

ions

56.0

of

indu

stri

al g

row

th

Ecu

ador

Gua

yaqu

ila13

.0—

——

30.0

of

GD

P

Hai

tiA

ll ur

ban

1976

24.2

15.6

57.6

of

nati

onal

inco

me

Port

-au-

Prin

ce—

15.0

7.7

47.2

82.7

b38

.7 o

f na

tion

al in

com

eO

ther

urb

an—

9.2

7.9

——

18.9

of

nati

onal

inco

me

Indi

aA

ll ur

ban

1970

–71

19.9

17.7

c—

—38

.9 o

f N

DP

Ken

yaA

ll ur

ban

1976

11.9

——

—30

.3 o

f in

com

eN

airo

bi—

5.2

——

—20

.0 o

f in

com

eO

ther

urb

an—

6.7

——

—10

.3 o

f in

com

e

Exh

ibit

3.2

Som

e In

dica

tors

of

the

Est

imat

ed E

cono

mic

Im

port

ance

of

Sele

ct U

rban

Are

as W

orld

wid

e, V

ario

us Y

ears

, 197

0–19

81(i

n pe

rcen

tage

s)

03-Sassen.qxd 3/21/2006 11:21 AM 8

49

Pub

licP

ublic

U

rban

Are

aY

ear

Pop

ulat

ion

Em

ploy

men

tR

even

ues

Exp

endi

ture

sO

utpu

t M

easu

re

Mex

ico

All

urba

n19

7060

.0(2

9.0)

d79

.7 o

f pe

rson

al in

com

eFe

dera

l Dis

tric

t—

14.2

33.6

of

pers

onal

inco

me

Paki

stan

Kar

achi

1974

–75

6.1

16.1

of

GD

P

Peru Lim

a19

8028

.043

.0 o

f G

DP

Phili

ppin

esM

etro

Man

ila19

7012

.0—

4525

.0 o

f G

DP

Tha

iland

Met

ro B

angk

ok19

7210

.914

.0e

30.5

f37

.4 o

f G

DP

Tur

key

All

urba

n19

8147

.042

70.0

of

GN

P

Tun

isia

Tun

is19

7516

.017

.2—

—

Not

es:

a. G

uaya

s Pr

ovin

ce.

b. C

urre

nt e

xpen

ditu

res

only

.c.

Wor

kers

.d.

Fed

eral

pub

lic in

vest

men

t on

ly.

e. 1

970

data

.f.

196

9 da

ta.

Sour

ce:

Frie

dric

h K

ahne

rt,

“Im

prov

ing

Urb

an E

mpl

oym

ent

and

Lab

or P

rodu

ctiv

ity,

” M

ay 1

987,

Wor

ld B

ank

Dis

cuss

ion

Pape

r N

o. 1

0. R

epri

nted

wit

hpe

rmis

sion

.



agglomerations in the world do not necessarily entail primacy: New York,for example, is among the 20 largest cities in the world, but it is not a pri-mate city, given the multipolar nature of the urban system in the UnitedStates. Furthermore, primacy is not an exclusive trait of developing countries,even though its most extreme forms are to be found in the developing world:Tokyo and London are primate cities. Finally, the emergence of so-calledmegacities may or may not be associated with primacy. The 20 largest urbanagglomerations by 2003 (and the foreseeable future) include some cities thatare not necessarily primate, such as New York, Los Angeles, Tianjin, Osaka,Calcutta, and Shanghai, and others that can be characterized as having lowlevels of primacy, such as Paris and Buenos Aires (see Exhibit A.3).

Primacy and megacity status are clearly fed by urban population growth,a process that is expected to continue. But they combine in multiple patterns;there is no single model. The evidence worldwide points to the ongoingurbanization of the population, especially in developing countries. As in thedeveloped countries, one component of urban growth in those countries isthe suburbanization of growing sectors of the population. The figures inExhibit 3.3 show rates of urban growth in select developing countries for theperiod that saw the beginning of today’s sharp urbanization of the world’spopulation. The higher the level of development, the higher the urbaniza-tion rate is likely to be. Thus, a country like Argentina had an urbanizationrate of 90.1% by 2003, which is quite similar to that of highly developedcountries, although it is to some extent a function of the primacy of BuenosAires in the national urban system. In contrast, Algeria’s urbanization rateof 59% and Kenya’s 39% differ sharply from the urbanization level indeveloped countries. Finally, there are countries such as India and China thathave vast urban agglomerations, notwithstanding their very low rate ofurbanization; they are, clearly, among the most populous countries in theworld. As a result, the information conveyed by an indicator such as theurbanization rate in these countries differs from that of countries with moreaverage population sizes.

On the subject of primacy, the literature about Latin America showsconsiderable convergence in the identification of major patterns, along withmultiple interpretations of these patterns. Many studies in the late 1970s andearly 1980s found sharper primacy rather than the emergence of the morebalanced national urban systems forecast by modernization theory (forcritical evaluations, see Edel 1986; El-Shakhs 1972; Roberts 1976; Smith1985; Walters 1985). The disintegration of rural economies, including thedisplacement of small landholders by expanding large-scale commercialagriculture, and the continuing inequalities in the spatial distribution ofinstitutional resources are generally recognized as key factors strengthening


51

Per

Cap

ita

Size

of

Pop

ulat

ion

(in

000’

s)b

Per

cent

age

ofA

vera

ge R

ate

of G

row

thc

GN

P L

evel

2003

2030

Urb

an P

opul

atio

nbU

rban

Pop

. (%

)R

ural

Pop

. (%

)

2003

19

80–

1995

–20

00–

1980

–19

95–

2000

–C

ount

ry(U

S$)a

Urb

anR

ural

Urb

anR

ural

2003

2030

1985

2000

2005

1985

2000

2005

Arg

enti

na3,

372

34,6

423,

786

45,5

683,

043

90.1

93.7

1.88

1.39

2.00

−0.8

7−0

.88

−1.1

0

Mex

ico

6,05

2 78

,100

25,3

5711

0,77

022

,821

75.5

82.9

3.36

2.39

1.80

0.34

−0.0

70.

40

Col

ombi

a1,

779

33,8

0810

,414

51,8

608,

982

76.5

85.2

3.11

2.29

2.20

0.28

−0.0

7−0

.50

Bra

zil

2,75

9 14

8,27

030

,201

202,

686

19,3

9283

.191

.33.

712.

282.

00−1

.27

−1.0

0−2

.40

Alg

eria

2,09

2 18

,711

13,0

8932

,032

12,0

8758

.872

.63.

713.

852.

602.

511.

250.

30

Mor

occo

1,43

1 17

,564

13,0

0230

,824

11,6

8057

.572

.54.

283.

422.

801.

400.

500.

10

Mal

aysi

a4,

247

15,6

118,

814

27,3

247,

867

63.9

77.6

4.51

3.32

3.00

1.06

0.15

0.10

Sene

gal

644

5,00

85,

086

11,3

505,

577

49.6

67.1

3.34

4.26

3.90

2.11

1.52

1.00

Cot

e d’

Ivoi

re82

6 7,

464

9,16

714

,054

9,20

444

.960

.46.

635.

242.

602.

542.

260.

80

Nig

eria

471

57,9

0766

,102

134,

398

72,2

9846

.765

.06.

075.

334.

402.

222.

021.

00

Ken

ya44

9 12

,593

19,3

9425

,807

15,3

3439

.462

.78.

066.

724.

403.

172.

78−0

.40

Indi

a56

4 30

1,26

076

4,20

258

6,05

283

0,52

528

.341

.43.

913.

962.

301.

650.

931.

20

Indo

nesi

a94

6 10

0,29

411

9,85

918

7,84

689

,721

45.6

67.7

4.60

3.62

3.90

1.13

0.14

−0.9

0

Chi

nad

1,08

6 50

3,74

080

0,45

687

7,62

357

2,89

838

.660

.51.

442.

953.

201.

180.

58−0

.80

Exh

ibit

3.3

Urb

an G

row

th P

atte

rns

in S

elec

t D

evel

opin

g C

ount

ries

, Sel

ect

Peri

ods,

198

0–20

30 (

num

bers

and

per

cent

age)

Sour

ces:

a. W

orld

Ban

k (2

005)

.

b. U

nite

d N

atio

ns D

epar

tmen

t fo

r E

cono

mic

and

Soc

ial A

ffai

rs, P

olic

y A

naly

sis

(200

3).

c. U

nite

d N

atio

ns D

epar

tmen

t fo

r In

tern

atio

nal E

cono

mic

and

Soc

ial A

ffai

rs (

1988

; 200

3).

d. E

xclu

ding

Hon

g K

ong

SAR

and

Mac

ao S

AR

.


primacy (Kowarick, Campos, and de Mello 1991; PREALC [RegionalEmployment Program for Latin America and the Caribbean] 1987).

Less widely known and documented is that in the 1980s there was adeceleration in primacy in several, although not all, countries in LatinAmerica. This trend will not eliminate the growth of megacities, but it isworth discussing in some detail because it resulted in part from specificaspects of economic globalization—concrete ways in which global processesimplant themselves in particular localities. The overall shift in growth strate-gies toward export-oriented development and large-scale tourism enclavescreated growth poles that emerged as alternatives to the primate cities forrural to urban migrations (Gilbert 1996; Landell-Mills, Agarwala, andPlease 1989; Portes and Lungo 1992a, 1992b; Roberts and Portes 2006).1

This shift was substantially promoted by the expansion of world markets forcommodities and the foreign direct investments of transnational corpora-tions, both in turn often stimulated by World Bank and IMF programs.

One of the best sources of information on the emergence of these patternsin the 1980s is a large, collective, multicity study directed by Portes andLungo (1992a, 1992b) that focused on the Caribbean region, includingCentral America.2 The Caribbean has a long history of urban primacy.Portes and Lungo studied the urban systems of Costa Rica, the DominicanRepublic, Guatemala, Haiti, and Jamaica, countries that clearly reflect theimmense variety of cultures and languages in this region. These countriesrepresent a wide range of colonization patterns, ethnic compositions, eco-nomic development, and political stability. In the 1980s, export-orienteddevelopment, a cornerstone of the Caribbean Basin Initiative, and the intensepromotion of tourism began to draw workers and firms. Expanded subur-banization has also had the effect of decentralizing population in the primatecities of the Caribbean, while adding to the larger metropolitan areas ofthese cities. The effect of these trends can be seen clearly in Jamaica, forexample, where the primacy index declined from 7.2 in 1960 to 2.2 in 1990,largely as a result of the development of the tourist industry on the northerncoast of the island, the revival of bauxite production for export in the inte-rior, and the growth of satellite cities at the edges of the broader Kingstonmetropolitan area. (See McMichall 2003 generally on development.)

In some Caribbean countries, however, the new growth poles have hadthe opposite effect. Thus, in Costa Rica, a country with a far more balancedurban system, the promotion of export manufacturing and tourism hastended to concentrate activities in the metropolitan area of the primate cityof San José and its immediate surrounding cities, such as Cartago. Finally,in the case of Guatemala, export manufacturing and tourism are far lessdeveloped, largely because of the extremely violent political situation until



the 1990s (Jonas 1992). Development of export-oriented growth remainscentered in agriculture. Guatemala has one of the highest levels of urbanprimacy in Latin America because alternative growth poles have been rare.Only in the 1990s did efforts to develop export agriculture promote somegrowth in intermediate cities, with coffee and cotton centers growing morerapidly than the capital, Guatemala City.

At the same time, deregulation and the associated sharp growth of foreigndirect investment since the early 1990s (Exhibit 3.4) has further strengthenedthe role of the major Latin American business centers, particularly MexicoCity, São Paulo, and Buenos Aires. Buenos Aires has had sharp ups anddowns—a sharp downturn in 2001, due to Argentina’s massive crisis, anda resurgence in 2005. As shown in Chapter 2, privatization has been a keycomponent of this growth. Foreign direct investment, via privatization andother channels, has been associated with deregulation of financial marketsand other key economic institutions. Thus the central role played by thestock market and other financial markets in these increasingly complexinvestment processes has raised the economic importance of the major citieswhere these institutions are concentrated. Because the bulk of the value ofinvestment in privatized enterprises and other, often related, investments has


12000

10000

8000

6000

4000

2000

0Argentina Brazil Chile Mexico Venezuela

20031992–1997 (Annual Avg.)1986–1991 (Annual Avg.)

Exhibit 3.4 Foreign Direct Investment in Select Latin American Countries,1986–2003 (US$ millions)

Source: UNCTAD (1998:364–65; 2004:367–71).



been in Mexico, Argentina, and Brazil, the impact of vast capital inflowsis particularly felt in the corporate and financial sectors in their primatecities—Mexico City, Buenos Aires, and São Paulo.3 We see in these cities theemergence of conditions that resemble patterns evident in major Westerncities: highly dynamic financial markets and specialized service sectors; theovervalorization of the output, firms, and workers in these sectors; and thedevalorization of the rest of the economic system (Schiffer Ramos 2002;Parnreiter 2002; Ciccolella and Mignaqui 2002). This is a subject I return toin Chapter 4.

In brief, economic globalization has had a range of impacts on citiesand urban systems in Latin America and the Caribbean. In some cases, it hascontributed to the development of new growth poles outside the majorurban agglomerations. In others, it has actually raised the weight of primateurban agglomerations, in that the new growth poles were developed in theseareas. A third case is that represented by the major business and financialcenters in the region, several of which saw a sharp strengthening in theirlinkages with global markets and with the major international businesscenters in the developed world.

Production zones, centers for tourism, and major business and financialcenters are three types of sites for the implantation of global processes.Beyond these sites is a vast terrain containing cities, towns, and villages thatare either increasingly unhinged from this new international growth dynamicor are part of the low-profit end of long chains of production (for one of themost detailed accounts, see the larger project summarized in Beneria 1989).The character of the articulation or dissociation is not simply a questionof city size, since there exist long subcontracting chains connecting workersin small villages to the world markets. It is, rather, a question of how theseemergent transnational economic systems are articulated, how they connectspecific localities in less developed countries with markets and localitiesin highly developed countries (see, e.g., Gereffi, Humphrey, and Sturgeon2005; Gereffi and Korzeniewicz 1994; Beneria 1989; Bonacich et al. 1994;Bose and Acosta-Belen 1995; Chaney and Castro 1993; Ward 1991). Theimplantation of global processes seems to have contributed to sharpening theseparation between cities, or sectors within cities, that are articulated withthe global economy and those that are not. This is a new type of interurbaninequality, one not predicated on old hierarchies of city size. The newinequality differs from the long-standing forms of inequality present in citiesand national urban systems because of the extent to which it results from theimplantation of a global dynamic, be it the internationalization of produc-tion and finance or international tourism.



Impacts on Balanced UrbanSystems: The Case of Europe

One of the most interesting findings of a major multiyear, multicountrystudy on cities in Europe, sponsored by the European Economic Community(EEC), was the renewed demographic and economic importance of Europe’slarge cities in the 1980s. (For a summary, see European Institute of UrbanAffairs 1992; Kunzmann and Wegener 1991; see also Eurocities 1989;INURA 2003; Kazepov 2005). In the 1960s and 1970s, most, if not all, ofthese large cities had experienced declines in population and in economicactivity, whereas smaller cities experienced growth in both dimensions. Wesaw a similar pattern in the United States, where this process took the formof suburbanization.

Many analysts, both in Europe and in the United States, asserted thatcentral cities, with the exception of old historical centers with culturalimportance, had lost much of their use to people and to the economy. Thewidespread growth of small cities in Europe in those two earlier decades wasseen as a strong indication of how balanced the urban systems of WesternEuropean nations were and continue to be. And, indeed, compared withalmost any other major continental region, Western European nations hadand continue to have the most balanced urban systems in the world.Nonetheless, it is now clear that beginning in the 1980s and continuing today,major cities in Europe have gained population and experienced significanteconomic growth (Exhibit 3.5a). The exceptions were some of Europe’slarge cities in more peripheral areas: There were continuing losses in Mar-seilles, Naples, and England’s old industrial cities, Manchester and Birming-ham. But some of these cities have also seen new population and economicgains as of the late 1990s (see Kazepov 2005). At the same time, comparedwith the 1960s, the sharpest growth has occurred in the larger metropolitanareas (Exhibit 3.5b). Smaller cities slowed down, often markedly, in the1980s, and continue to do so today. Indeed, there is now an emergent fieldof research focused on “shrinking cities” in older industrial areas of Europe.Exhibit 3.6 shows some of the major cities in Europe.

These trends can be interpreted in several ways. On one hand, these couldbe mild demographic shifts that leave the characteristics of the urban systembasically unaltered, that is, urban systems remain balanced at the levels ofthe nation and of Western Europe as a whole. On the other hand, the trendscould indicate a renewed importance of major cities because the economicchanges evident in all developed countries have organizational and spatial

(Text continues on page 59)


56

1970

–197

519

75–1

980

1980

–198

519

85–1

990

1990

–199

519

95–2

000

2000

–200

5b

Cor

e C

itya

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Ham

burg

−0.7

70.

85−0

.91

0.36

−0.7

70.

060.

240.

060.

65—

0.33

—0.

13—

Fran

kfur

t—

——

—−1

.01

−0.0

41.

620.

11—

——

——

—

Dor

tmun

d−0

.41

0.08

−0.7

9−0

.27

−1.1

5−0

.56

0.54

0.37

——

——

——

Ber

lin−0

.47

−0.2

5−0

.02

−0.0

70.

210.

032.

190.

090.

18—

0.05

—0.

02—

Pari

s−1

.48

1.93

−0.6

90.

66−1

.02

0.78

1.01

2.06

0.38

—0.

38—

0.33

—

Lyo

ns−1

.79

4.25

−1.2

3−1

.18

0.07

−0.0

40.

071.

210.

74—

0.74

—0.

67—

Mar

seill

es0.

274.

47−0

.48

2.91

−1.1

01.

57−1

.10

2.84

0.39

—0.

39—

0.39

—

Mila

n−0

.14

1.06

−1.1

71.

07−2

.02

0.6

−1.0

30.

35−1

.05

—–0

.86

—−0

.86

—

Am

ster

dam

−1.8

41.

51−1

.11

0.81

−1.1

80.

570.

340.

470.

9—

0.46

—0.

53—

Rot

terd

am−1

.99

1.1

−1.3

80.

81−0

.28

0.56

0.22

0.28

0.57

—0.

29—

0.34

—

Bru

ssel

s−1

.99

0.48

−1.3

80.

15−0

.95

0.02

−0.1

70.

04−0

.04

—0.

04—

1.31

—

Lon

don

−1.8

9−0

.37

−1.6

−0.1

4−0

.38

−0.0

60.

56−0

.32

−0.0

3—

–0.0

3—

−0.0

3—

Bir

min

gham

−0.3

0.35

−1.0

1−0

.66

−0.3

30

−0.3

70.

06−0

.25

—–0

.25

—−0

.25

—

Exh

ibit

3.5

aPo

pula

tion

Cha

nge

in S

elec

t E

urop

ean

Cit

ies,

Sel

ect

Peri

ods,

197

0–20

05 (

perc

enta

ges)


57

1970

–197

519

75–1

980

1980

–198

519

85–1

990

1990

–199

519

95–2

000

2000

–200

5b

Cor

e C

itya

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Cor

eR

ing

Gla

sgow

−3.3

8−1

.47

−1.8

4−0

.11

−1.0

6−0

.17

−1.4

1–0

.32

——

——

——

Dub

lin−0

.41

—−0

.41

—−1

.61

——

—0.

65—

0.87

—0.

88—

Cop

enha

gen

−2.2

82

−1.4

70.

46−0

.59

−0.1

2−0

.72

0.14

0.3

—−4

.61

—0.

23—

The

ssal

onik

i2.

06—

1.44

—0.

930.

54—

—0.

66—

0.67

—0.

67—

Ath

ens

1.09

—−0

.16

—−1

.43

1.45

——

0.34

—0.

37—

0.37

—

Mad

rid

0.45

8.28

−0.2

8.19

–0.6

33.

160.

280.

070.

51—

0.43

—0.

43—

Bar

celo

na−0

.07

3.4

0.13

2.27

–0.5

80.

710.

04–0

.04

0.62

—0.

21—

0.21

—

Val

enci

a1.

441.

471.

111.

73–0

.41

1.26

0.6

–0.4

8—

——

——

—

Sevi

lle1.

24−0

.02

1.81

1.23

0.16

1.19

0.75

0.52

——

——

——

Not

es:

a. C

ore

Cit

y re

fers

to

citi

es in

gro

wth

or

dyna

mic

reg

ions

in W

este

rn E

urop

e.

b. E

stim

ated

.

Sour

ce:D

ata

for

1970

–199

0 ta

ken

from

Eur

opea

n In

stit

ute

of U

rban

Aff

airs

(19

92:5

6); d

ata

for

1990

–200

5 ta

ken

from

Uni

ted

Nat

ions

Dep

artm

ent

for

Inte

rnat

iona

l Eco

nom

ic a

nd S

ocia

l Aff

airs

(20

03, t

able

A.1

4).



Metropolitan

Core City Area Core Ring

Amsterdam 8.4 −17.7 34.5Antwerp 20.7 −36.7 39.2Athens 63.5 14.6 87.9Barcelona 73.1 −8.8 330.9Berlin 1.9 4.5 −9.1Birmingham 2.5 −13.4 14.0Brussels 26.6 −17.0 30.6Copenhagen 10.4 −28.9 51.6Dublin 45.5 −7.8 232.7Dusseldorf 25.3 −18.9 115.3Frankfurt 30.8 −7.8 66.4Glasgow −0.8 −35.6 41.2Hamburg 12.7 −7.0 94.8Helsinki 68.5 17.4 221.4Köln 22.1 13.2 33.1Leeds 12.5 −17.5 30.7Lille 32.1 −4.1 42.6Lisbon 73.1 −29.7 238.6Liverpool −10.1 −34.7 9.1London 7.8 −12.9 14.6Lyon 64.9 −22.6 169.7Madrid 97.6 20.0 1618.4Manchester −3.2 −38.2 7.2Mannheim 34.1 -4.6 48.9Marseille 74.3 2.6 680.4Milan 36.6 −21.6 123.8Munich 26.3 4.5 104.9Naples 78.5 −14.5 289.4Newcastle 16.9 −27.9 30.0Nuremburg 50.8 5.4 152.2Paris 39.7 −24.1 74.0Porto 38.0 −13.2 72.7Rhine-Ruhr 12.0 −18.8 17.0Rome 29.4 13.2 265.6Rotterdam 31.2 −18.2 161.2Seville 57.3 54.6 61.2Stockholm 42.7 −5.3 143.7Stuttgart 83.3 –8.6 159.5Turin 14.8 −17.0 162.1Valencia 111.8 46.1 325.8Vienna −9.9 6.6 −24.7Zurich 70.6 −22.7 220.0

Source: Demographia (2005).

Exhibit 3.5b Population Change in Select European Cities, 1965–2004(percentages)


implications for such cities. The EEC study mentioned earlier found that thesecond of these interpretations fits the data gathered for 24 cities in Europefor the 1980s. The evidence shows clearly that the period 1985 to 1990marks the crucial turnaround from negative to positive population growthin the urban core after consistent losses in the preceding periods (see Exhibit3.5; see also Eurocities 1989; Kazepov 2005; for critical accounts of thecontents of this growth, see Hitz et al. 1995; INURA 2003; Bodnar 2000).Finally, the evidence also signals that the new organizational and com-positional features of the economy of core cities can accommodate higheconomic growth with little, if any, and even negative, population growth(see also Sassen [1991] 2001, chaps. 8–9).


Exhibit 3.6 Select Cities in Europe, 2005



The organizational and spatial implications of the new economic trendsassume distinct forms in various urban systems. Some cities become part oftransnational networks, whereas others become unhinged from the main cen-ters of economic growth in their regions or countries. A review of the EECreport, as well as other major studies on cities in Europe, suggests that thereare at least three tendencies in the reconfiguration of urban systems inWestern Europe that began in the 1980s. First, several sub-European regionalsystems have emerged (CEMAT [European Conference of Ministers Respon-sible for Regional Planning ] 1988; Kunzmann and Wegener 1991; Rhine-Ruhr 2005). Second, within the territory of the EEC and several immediatelyadjacent countries (Austria, Denmark, and Greece) in the 1980s, the newEuropean Union in the 1990s, and the enlarged union as of 2004, a limitednumber of cities have strengthened their role in an emergent European urbansystem. Finally, a few of these cities are also part of an urban system thatoperates at the global level (Exhibit 3.7)

National European urban systems are also being affected by thesedevelopments. The traditional national urban networks are changing. Citiesthat were once dominant in their countries may lose that importance, whilecities in border regions or transportation hubs may gain a new importance.Furthermore, the new European global cities may capture some of the busi-ness, demands for specialized services, and investments that previously wentto national capitals or major provincial cities. Cities at the periphery willfeel the widening gap with the newly defined and positioned geography ofcentrality.

Cities in peripheral regions and old port cities began to lose ground in the1970s and 1980s in their national urban systems as a result of the new hier-archies (Castells 1989; Hausserman and Siebel 1987; Parkinson, Foley, andJudd 1989; Roncayolo 1990; Siebel 1984; van den Berg et al. 1982; Vidalet al. 1990). By the late 1980s, it had become clear that many cities wereincreasingly disconnected from the major European urban systems. Some ofthese peripheralized cities with outmoded industrial bases have reemergedwith new functions and as part of new networks in the 1990s—for example,Lille in France as a major transportation hub, including the Eurostar trans-port system, and Glasgow in the United Kingdom as a major tourism andcultural destination. Others have lost politico-economic functions and areunlikely to regain them in the foreseeable future. Yet others are becomingcenters for tourism or places for second homes; for example, a growingnumber of high-income Germans and English have bought country houses—indeed, whole “castles”—in rural Ireland, inducing other continentalEuropeans to do the same. In an ironic twist, much of the beauty andcurrent value of the Irish countryside—whole regions untouched by


61

Cit

y, C

ount

rya

2005

b19

97c

1990

c19

80c

1970

c19

60c

Tok

yo, J

apan

10 (

1)d

18 (

5)12

(2)

65

(1)

1

New

Yor

k, U

SA7

(2)

12 (

1)7

(5)

10 (

4)25

(8)

29 (

8)

Pari

s, F

ranc

e8

(1)

11 (

1)5

7 (2

)0

0

Osa

ka, J

apan

17

(3)

2 (1

)1

10

Det

roit

, USA

1 (1

)4

(2)

2 (2

)2

(2)

3 (3

)5

(2)

Lon

don,

UK

4 (1

)3

(1)

7 (2

)8

(3)

7 (3

)7

(3)

Chi

cago

, USA

13

24

(2)

56

(2)

Mun

ich,

Ger

man

y4

(1)

32

11

1

Am

ster

dam

, Net

herl

ands

1 (1

)3

00

00

Seou

l, So

uth

Kor

ea2

30

00

0

Exh

ibit

3.7

Loc

atio

n of

Top

Ban

king

, Ind

ustr

ial,

and

Com

mer

cial

Fir

ms

by C

ity,

Sel

ect

Yea

rs, 1

960–

2005

Not

es:

a. A

fter

ran

king

cit

ies

acco

rdin

g to

the

num

ber

hold

ing

the

wor

ld’s

100

lar

gest

cor

pora

tion

hea

dqua

rter

s (i

n 19

99),

the

lis

t w

as t

rim

med

to

the

top

40ci

ties

, of

whi

ch 1

0 ar

e lis

ted

in t

he t

able

abo

ve.

b. A

utho

r’s

calc

ulat

ions

bas

ed o

n “G

loba

l 500

” (2

005)

.

c. S

hort

and

Kim

(19

99:2

6).

d. T

he f

igur

e in

par

enth

eses

giv

es t

he n

umbe

r of

the

wor

ld’s

top

20

corp

orat

ions

for

tha

t ci

ty.


industrialization—is a legacy of poverty; it may now be undermined byIreland’s rapid growth, much of it centered in high-tech manufacturing. Therequirement for becoming transnational centers for tourism and secondhomes is that these sites cannot pursue industrial development and need topreserve high levels of environmental quality. Further, changes in militarydefense policies resulting from the fall of the Soviet Union and its recompo-sition into a series of smaller nation-states will cause decline in cities thatwere once crucial production centers or control centers for national securitysystems. Smaller port cities, or large ones that have not upgraded and mod-ernized their infrastructures, will be at a great disadvantage in competingwith the large, modernized port cities in Europe. Marseilles was once a greatport, strategically located on the Mediterranean; today, it has been leftbehind by Rotterdam and a few other major European ports that constitutea cluster of state-of-the-art ports. Nothing in the near future seems to securethe revitalization of old industrial centers on the basis of the industriesthat once were their economic core. The most difficult cases are small- andmedium-size cities in somewhat isolated or peripheral areas dependent oncoal and steel industries. They are likely to have degraded their environmentsand hence do not even have the option of becoming tourist centers.

The shifts that took off in the 1980s were sufficiently dramatic to engagea whole series of scholars into painting new urban scenarios for Europe.Kunzmann and Wegener (1991) asserted that the dominance of large citieswould continue in part because they would be more competitive in gettingboth European and non-European investors’ preferences for the larger high-tech industrial and service cities (see also Deecke, Kruger, and Lapple 1993).Furthermore, according to some researchers, this spatial polarization woulddeepen because of the development of high-speed transport infrastructureand communications corridors, which tend to connect major centers orhighly specialized centers essential to the advanced economic system(Castells and Hall 1994; Graham and Marvin 1996; Masser, Sviden, andWegener 1990). Much of this forecasting is turning out to be correct. We seemassive concentrations of resources in some cities (e.g., Rutherford 2004;Abrahamson 2004). But today, a growing number of small, somewhatperipheral cities in Europe are literally shrinking: Their built environment isseverely underutilized and in some cases fully abandoned (see, e.g., Giesecke2005). However, an old mining and steel city such as Lille is now one ofWestern Europe’s major transportation and communications hubs. This hasradically changed this once-dying industrial city; further, using internation-ally known architects to do much of the critical building has also made Lillea cultural destination of sorts, a wonderful turn of events for what was oncea place of factories and coal mines.



A process of recentralization may be occurring in certain cities that havebeen somewhat peripheral. Some of the smaller cities in Europe (such asAachen, Strasbourg, Nice, Liege, and Arnheim) are likely to benefit fromthe single European market insofar as they can expand their hinterland andfunction as a nexus to a broader European region. Changes in EasternEurope are likely to strengthen the role of Western European cities that usedto have extensive interregional linkages before World War II—notablyHamburg, Copenhagen, and Nuremberg—which in turn may have the effectof weakening the position of other peripheral cities in those regions. Citiesbordering Eastern Europe may assume new roles or recapture old ones;Vienna and Berlin are emerging as international business platforms for thewhole central European region.4

Finally, major Eastern European cities such as Budapest, Prague, andWarsaw may regain some of their prewar importance. Budapest is a goodexample: Toward the late 1980s, it emerged as the leading internationalbusiness center for the Eastern European region, a role illustrated by the factthat Hungary has since consistently been a major recipient of foreign directinvestment in Eastern Europe (see Exhibit 3.8) (Bodnar 2000). Although the


4500

4000

3500

3000

2500

2000

1500

1000

500

0Russian

FederationPoland Hungary Czech

RepublicRomania Ukraine

1992–1997 (Annual Average) 2003

Exhibit 3.8 Foreign Direct Investment in Select Central and Eastern EuropeanCountries, 1986–2003 (US$ millions)

Source: UNCTAD (2004:371).


absolute investment levels were lower than those in the Russian Federation(with its vastly larger territory and economy than Hungary), in relativeterms, these figures represent a sharper internationalization than in the for-mer Soviet Union. Western European and non-European firms seeking to dobusiness in Eastern Europe established offices in Budapest to launch opera-tions for a large transnational Central European region. By the early 1990s,Budapest had a rather glamorous Western-looking international businessenclave that offered the requisite comforts, hotels, restaurants, and businessservices to an extent that most other major Eastern European cities did nothave at the time.

Immigration has become a major factor in demographic and labor-forcegrowth in many European cities and, especially from the late 1990s on, anincreasingly polarizing political issue. An initial scholarship that engaged thenew migration phase that took off in the 1980s recognized that this was afeature that was there to remain (Balbo and Manconi 1990; Blaschke andGermershausen 1989; Brown 1984; Canevari 1991; Cohen 1987; Gilletteand Sayad 1984; SOPEMI [Systeme d’Observation Permanente pour lesMigrations] 1999–2005; Tribalat et al. 1991).5 Cities that function as gate-ways into Europe were expected to receive growing immigration flows fromEastern Europe, Africa, and the Middle East. This led to increasingly strongpolitical divisions as many of these cities, particularly old port cities suchas Marseilles, Palermo, and Naples, were already experiencing economicdecline and were seen as unable to absorb the additional labor and costs(for critical examinations of some of these assumptions, see Pugliese 2002;Mingione 1991). Although these cities may have functioned largely as entre-pôts, with variable shares of immigrants expected to move on to moredynamic cities, resident immigrant populations also took root. One concernin these cities has been that having their infrastructures and services over-burdened would further peripheralize these gateway cities in the emergingEuropean urban hierarchy connecting leading cities in Europe and con-tribute to sociospatial polarization. However, some of Europe’s global cities,such as Paris and Frankfurt, which are at the center of major transportationnetworks and are final destinations for many immigrants, have recognizedthe often major benefits associated with significant shares of immigrantsin their populations and workforces. In Frankfurt, for example, 28% of theworkforce is foreign born, including significant shares of top-level profes-sionals. In other cases, it is older imperial geographies that have madecertain cities key destinations, with often positive disposition towards immi-grants. Thus, Berlin, an emerging global city in a variety of highly special-ized sectors (culture, new media, software design), is also a preferreddestination of many new migrations, as is Vienna. In the past, Berlin and



Vienna were centers of vast regional migration systems, and they seem to berecapturing that old role. The enlarging of the European Union in 2004 hascreated a new perimeter for Europe with growing concerns that the Easternedge cities in that perimeter will have to take on control and gateway func-tions for which they are not fully equipped or prepared. Elsewhere (Sassen1999; 2004), I have argued that Europe has had immigration for centuries,with sharp up and down cycles of positive and negative dispositions towardimmigrants, and has always wound up with significant levels of integra-tion—as is suggested by, for example, a third of France’s native born popu-lation having a foreign-born ancestor two or three generations back, a figurethat goes up to 40% in the case of a city such as Vienna. I return to immi-gration issues in some of the later chapters.

There are, then, a multiplicity of economic and demographic geographiesof centers and margins in Europe at this time. A central urban hierarchy con-nects major cities, many of which in turn play key roles in the wider globalsystem of cities: Paris, London, Frankfurt, Amsterdam, Zurich, Madrid,and Milan. Somewhat less oriented to the global economy is a majornetwork of European financial/cultural/service capitals, such as Edinburgh,Berlin, Dublin, Rome, Stockholm, Prague, and Warsaw—some with onlyone, and others with several of these functions, which articulate theEuropean region. And then there are several geographies of margins: theEast–West divide and the North–South divide across Europe, as well as newmicro-divisions. In Eastern Europe, certain cities and regions are ratherattractive for European and non-European investment, whereas others willincreasingly fall behind (notably, those in the former Yugoslavia andAlbania). A similar differentiation exists in the south of Europe: Madrid,Barcelona, and Milan are gaining in the new European hierarchy; Naplesand Marseilles are not.

Transnational Urban Systems

A rapidly growing and highly specialized research literature began to focusin the 1980s on different types of economic linkages binding cities acrossnational borders (Friedmann and Wolff 1982; Castells 1989; Daniels 1991;GaWC 1998; Graham and Marvin 1996; Leyshon, Daniels, and Thrift1987; Noyelle and Dutka 1988; Sassen-Koob 1982; 1984; Sassen 1988).

Today, this has emerged as a major issue of interest to a variety of disci-plines (Taylor 2004), even though the data are partial and often problem-atic. Prime examples of such linkages are the multinational networks ofaffiliates and subsidiaries typical of major firms in manufacturing and




specialized services (see Exhibits 2.5 and 4.7). The internationalization andderegulation of various financial markets is yet another, very recent devel-opment that binds cities across borders (see Exhibits 2.8, 2.9, 3.9, 3.10,3.11, and several exhibits in Chapter 4).6 An increasing number of stockmarkets around the world now participate in a global equities market. Thereare also a growing number of less directly economic linkages, notable amongwhich are a variety of initiatives launched by urban governments thatamount to a type of foreign policy by and for cities. In this context, the long-standing tradition of designating sister cities (Zelinsky 1991) has recentlybeen reactivated, taking on a whole new meaning in the case of cities eagerto operate internationally without going through their national governments(Eurocities 1989; Sassen 2002; Urban Age 2005).

Some of the most detailed data on transnational linkages binding citiescome from studies on corporate service firms. These firms have developedvast multinational networks containing special geographic and institu-tional linkages that make it possible for client firms—transnational firmsand banks—to use a growing array of service offerings from the same sup-plier (Bryson and Daniels 2005; Daniels 1991; Ernst 2005). There is goodevidence that the development of transnational corporate service firms wasassociated with the needs of transnational firms for global servicing capabil-ities (Ernst 2005; Sassen [1991] 2001, chap. 5). One of the best data sets onthe global networks of affiliates of leading firms in finance, accounting, law,and advertising is the Globalization and World Cities Study Group andNetwork, usually referred to (including in this book) as GaWC. RecentGaWC research shows that the network of affiliates in banking/finance andlaw firms closely follows the relative importance of world cities in those twosectors (Exhibits 3.9a and 3.9b). The transnational banking/finance or lawfirm, therefore, can offer global finance and legal services to a specific seg-ment of potential customers worldwide. And so can the global communica-tions firms. Global integration of affiliates and markets requires making useof advanced information and telecommunications technology that can cometo account for a significant share of costs—not only operational costs butalso, and perhaps most important, research and development costs for newproducts or advances on existing products. Exhibit 3.10 shows the proba-bility that firms with offices in one of the listed cities will have a branchoffice or affiliate in another city.

The need for scale economies on all these fronts helps explain the recentincrease in mergers and acquisitions, which has consolidated the position ofa few very large firms in many of these industries and has further strength-ened cross-border linkages between the key locations that concentrate theneeded telecommunications facilities. These few firms can now control asignificant share of national and international markets. The rapid increase in



foreign direct investment in services is strongly linked with the high level ofconcentration in many of these industries and a strong tendency towardincreasing market share among the larger firms. This is particularly true

New York

Singapore

Hong Kong

Tokyo

36

30

29

29

25

23

23

21

20

Frankfurt

Paris

Zurich

Sydney

Madrid

Milan

Taipei

Mexico City

Seoul

São Paulo

Jakarta

Los Angeles

Buenos Aires

Dublin

Kuala Lumpur

Moscow

San Francisco

Toronto

Bangkok

Dubai

Geneva

Luxembourg

Manama

Manila

Mumbai

0 5 10 15 20 25 30 35 40

Number of Linkages

13

13

13

13

14

14

14

15

15

15

15

15

15

16

16

17

18

18

19

19

Exhibit 3.9a Cities with Major Levels of Banking/Finance Links to London(number of firms, 1998)

Source: Taylor, Walker, and Beaverstock (2000).


for firms servicing large corporations. At the same time, subcontractingby larger firms and a proliferation of specialized markets has meant thatsmall independent firms can also thrive in major business centers. The linkbetween international law firms and financial firms has contributed to a cen-tralization of law firms in major financial centers.

Whether these links have engendered transnational urban systems isless clear and is partly a question of theory and conceptualization. So muchof social science is profoundly rooted in the nation-state as the ultimate unitfor analysis that conceptualizing processes and systems as transnational isbound to create controversy. Much of the literature on world and globalcities does not necessarily proclaim the existence of a transnational urbansystem: In its narrowest form, this literature posits that global cities per-form central place functions at a transnational level. But that leaves openthe question of the nature of the articulation among global cities. If weaccept that they basically compete with each other for global business, thenthey do not constitute a transnational system. Studying several global citiesthen falls into the category of traditional comparative analysis. If, however,we posit that in addition to competing with each other, global cities are alsothe sites for transnational processes with multiple locations (Taylor 2004),


New YorkWashington, DC

BrusselsHong Kong

ParisLos Angeles

TokyoSingapore

MoscowFrankfurtChicago

San FranciscoWarsaw

BudapestBeijingDallas

HoustonPrague 9

9910

111212

1315

1820

2122

313535

3944

0 5 10 15 20 25 30 35 40 45 50

Number of Linkages

Exhibit 3.9b Cities with Major Levels of Law Firm Links to London (numberof firms, 1998)

Source: Taylor, Walker, and Beaverstock (2000).


69

Lin

kage

to

Lin

kage

fro

mC

hica

goFr

ankf

urt

Hon

g K

ong

Lon

don

Los

Ang

eles

Mila

nN

ew Y

ork

Par

isSi

ngap

ore

Tok

yo

Chi

cago

—89

8910

091

7910

089

8310

0

Fran

kfur

t67

—93

100

7287

100

9594

95

Hon

g K

ong

6082

—10

080

8010

085

9290

Lon

don

5977

87—

7878

9883

8386

Los

Ang

eles

6773

8910

0—

7097

8481

89

Mila

n59

8893

100

67—

100

8891

93

New

Yor

k59

7787

9877

77—

7983

85

Pari

s64

8590

100

8081

97—

9090

Sing

apor

e60

8798

100

7883

100

92—

95

Tok

yo64

8493

100

8381

100

8788

—

Exh

ibit

3.1

0In

traf

irm

Ser

vice

Net

wor

ks a

mon

g T

op B

usin

ess

Cen

ters

, 199

8 (p

roba

bilit

ies

in p

erce

ntag

es)

Not

e: T

he t

able

is a

n as

ymm

etri

c m

atri

x sh

owin

g pr

obab

iliti

es o

f co

nnec

tion

s be

twee

n ci

ties

; eac

h ce

ll co

ntai

ns t

he p

erce

ntag

e pr

obab

ility

tha

t a

firm

inC

ity

X w

ill h

ave

an o

ffic

e in

Cit

y Y

.

Sour

ce:

Tay

lor,

Wal

ker,

and

Bea

vers

tock

(20

00)


then we can begin to explore the possibility of a systemic dynamic bindingthese cities.

Elsewhere (Sassen [1991] 2001, chaps. 1 and 7; 2002), I have argued thatin addition to the central place functions performed by these cities at the globallevel as posited by Hall (1966), Friedmann and Wolff (1982), and Sassen-Koob (1982), these cities relate to one another in distinct systemic ways.For example, already in the 1980s, when the notion of global multisited sys-tems was barely developed, I found that the interactions among New York,London, and Tokyo, particularly in terms of finance, services, and investment,consisted partly of a series of processes that could be thought of as “chains ofproduction” and international divisions of labor. Thus, in the case of globalfinance in the mid-1980s, I argued that Tokyo functioned as the main exporterof the raw material we call money, while New York was the leading financialprocessing and innovation center in the world. Many of the new financialinstruments were invented in New York; and money, either in its raw form orin the form of debt, was transformed into instruments aimed at maximizingthe returns on that money. London, the world’s major banking entrepôt, hadthe network to centralize and concentrate small amounts of capital availablein a large number of smaller financial markets around the world, partly as afunction of its older network for the administration of the British Empire(Exhibit 3.11). More recently, I have replicated this type of analysis focusingon the specialized advantage of major global cities, for example, New York’s,and Chicago’s financial centers (Sassen 2002, chap. 1).

These are examples suggesting that cities do not simply compete with eachother for the same business. There is an economic system that rests on the


Insurance Rank Citya Corporations Banks Telecommunications Agencies

1 Tokyo 56 3 2 62 Paris 26 4 2 33 London 23 3 0 54 New York 22 2 1 45 Beijing 12 4 2 1

Note:

a. Cities with the most high-revenue multinational corporations.

Source: Calculations based on “Global 500” (2005).

Exhibit 3.11 Top 5 Global Command Centers Based on Corporations, Banks,Telecommunications, and Insurance Agencies, 2005


distinct types of locations and specializations each city represents.Furthermore, it seems likely that the strengthening of transnational tiesamong the leading financial and business centers is accompanied by a weak-ening of the linkages between each of these cities and its hinterland andnational urban system (Sassen [1991] 2001). Cities such as Detroit,Liverpool, Manchester, Marseilles, the cities of the Ruhr, and now Nagoyaand Osaka have been affected by the territorial decentralization of many oftheir key manufacturing industries at the domestic and international level. Butthis same process of decentralization has contributed to the growth of serviceindustries that produce the specialized inputs to run spatially dispersed pro-duction processes and global markets for inputs and outputs. In a represen-tative case, General Motors, whose main offices are in Detroit, also has aheadquarters in Manhattan, which does all the specialized national andglobal financial and public relations work this vast multinational firmrequires. Such specialized inputs—international legal and accounting services,management consulting, and financial services—are heavily concentrated inbusiness and financial centers rather than in manufacturing cities. In brief, themanufacturing jobs that Detroit began to lose in the 1970s and 1980s fed agrowing demand for specialized corporate services in New York City to coor-dinate and manage a now globally distributed auto manufacturing system.

Global Cities and Diasporic Networks

A particular type of transnational urban system is slowly emerging from avariety of networks concerned with transboundary issues such as immigra-tion, asylum, international women’s agendas, antiglobalization struggles,and many others. These types of networks have proliferated rapidly sincethe late 1980s and increasingly intensified their transactions. Althoughthese networks are not necessarily urban in their orientation or genesis,their geography of operations is partly inserted in a large number of cities.The new network technologies, especially the Internet, ironically havestrengthened the urban map of these transboundary networks. It does nothave to be that way, but at this time cities and the networks that bind themfunction as an anchor and an enabler of cross-border transactions andstruggles. Global cities especially already have multiple intercity transac-tions and immigrants from many different parts of the world. These samedevelopments and conditions also facilitate the globalizing of terrorist andtrafficking networks.

Global cities and the new strategic geographies that connect them andpartly bypass national states are becoming a factor in the development of



globalized diasporic networks. This is a development from the ground up,connecting a diaspora’s multiple groups distributed across various places. Inso doing, these networks multiply the transversal transactions among thesegroups and destabilize the exclusive orientation to the homeland typical ofthe older radial pattern (Axel 2002). Furthermore, even a partial reorienta-tion away from national homeland politics can ease these groups’ transac-tions in each city with that city’s other diasporas and nondiasporic groupsinvolved in diverse types of transnational activities and imaginaries (Bartlett2006). In such developments, in turn, lies the possibility that at least someof these networks and groups can become part of the infrastructure forglobal civil society rather than being confined to deeply nationalistic proj-ects. These dynamics can then be seen as producing a shift toward globaliz-ing diasporas by enabling transversal connections among the members ofa given worldwide diaspora, and by intensifying the transactions amongdiverse diasporic and nondiasporic groups within a given city.

Cities are thick enabling environments for these types of activities, evenwhen the networks themselves are not urban per se. In this regard, thesecities enable the experience of participation in global nonstate networks.One might say that global civil society gets enacted partly in the microspacesof daily life rather than on some putative global stage. Groups can experi-ence themselves as part of a globalized diaspora even when they are in aplace where there might be few conationals, and thus the term diasporahardly applies. In the case of global cities, there is the added dimension ofthe global corporate economy and its networks and infrastructures enablingcross-border transactions and partially denationalizing urban space.

Both globalization and the international human rights regime have con-tributed to create operational and legal openings for nonstate actors to enterinternational arenas once exclusive to national states. Various, often as yetvery minor developments signal that the state is no longer the exclusivesubject for international law or the only actor in international relations.Other actors—from NGOs (nongovernmental organizations) and first-nation peoples to immigrants and refugees who become subjects of adjudi-cation in human rights decisions—are increasingly emerging as subjects ofinternational law and actors in international relations. Therefore, these non-state actors can gain visibility as individuals and as collectivities, and theycan come out of the invisibility of aggregate membership in a nation-stateexclusively represented by the state.

The nexus in this configuration is that the weakening of the exclusive for-mal authority of states over national territory facilitates the ascendance ofsub- and transnational spaces and actors in politico-civic processes. Thenational as container of social process and power is cracked enabling the



emergence of a geography of politics and civics that links subnational spaces.Cities are foremost in this new geography. The density of political and civiccultures in large cities and their daily practices roots, implants, and localizesglobal civil society in people’s lives. Insofar as the global economic systemcan be shown to be partly embedded in specific types of places and partlyconstituted through highly specialized cross-border networks connectingtoday’s global cities, one research task to help understand how this all inter-sects with immigrants and diasporas is, then, to find out about the specificcontents and institutional locations of this multiscalar globalization, thesubject of this book. Further, it means understanding how the emergence ofglobal imaginaries changes the meaning of processes that may be much olderthan the current phase of globalization, but that today are inscribed by thelatter. Thus, immigrant and diasporic communities are much older thantoday’s globalization. But that does not mean that they are not altered byvarious specific forms of globalization today.

As discussed in preceding chapters and again later in Chapter 7, recap-turing the geography of places involved in economic political globalizationallows us to recapture people, workers, communities, and the many differ-ent political projects in and of these communities (e.g., Mele 1999; Espinoza1999). The global city can be seen as one strategic research site about theseprocesses and the many forms through which global processes become local-ized in specific arrangements. This localizing includes a broad range ofprocesses: the new, very-high-income, gentrified urban neighborhoods of thetransnational professional class and rich exiles, and the work lives of the for-eign nannies and maids in those same neighborhoods and of the poor refugeesconcentrated in asylum housing. Although the formation of the network ofglobal cities is largely driven by corporate economic globalization, multiplepolitical and cultural processes have localized in these complex, partly dena-tionalized environments. This is an old history for cities, but it has received awhole new life through the formation of today’s networks of global cities.

The next subsection briefly addresses some general issues of an emergentglobal politics centered on local struggles and actors.

A Politics of Places on Global Circuits

The space constituted by the worldwide grid of global cities, a space withnew economic and political potentialities, is perhaps one of the most strate-gic spaces for the formation of transnational identities and communities.This is a space that is both place-centered, in that it is embedded in particu-lar and strategic cities, and transterritorial because it connects sites that arenot geographically proximate yet are intensely connected to each other. It is



not only the transmigration of capital that takes place in this global grid butalso that of people, both rich (i.e., the new transnational professional work-force) and poor (i.e., most migrant workers). It is also a space for the trans-migration of cultural forms and the reterritorialization of local subcultures.

An important question is whether it is also a space for a new politics, onegoing beyond the politics of culture and identity, though likely to be embed-ded partly in these. The politics of diasporic groups may be grounded inshared identities, but these politics do not necessarily conform to the politicsof identity in our Western societies. The possibility of transnational identityformation among politicized diasporic groups is an interesting question,given a history of homeland orientation. It is one of the questions runningthrough this book, particularly because global cities are enabling environ-ments in this regard. One of the most radical forms assumed today by thelinkage of people to territory is the loosening of selfhood from traditionalsources of identity, such as the nation or the village. This unmooring in theprocess of identity formation engenders new notions of community of mem-bership and of entitlement.

Immigration is one major process through which a new transnationalpolitical economy is being constituted, largely embedded in major citiesbecause most immigrants are concentrated in major cities. It is one of theconstitutive processes of globalization today, even though not recognizedor represented as such in mainstream accounts of the global economy.Immigration becomes part of a massive demographic transition in thesecities with a growing presence of women, native minorities, and immigrantsin the population of more and more cities. Global capital and immigrants aretwo major examples, each a unified crossboder actor (or aggregate of actors)who find themselves in contestation with each other inside global cities.

Insofar as immigration is one of the forces shaping diasporas, these cur-rent features of immigration can be expected, first, to at least partly transna-tionalize diasporas, moving them away from an exclusive orientation to thehomeland; and second, to urbanize at least some of their contestatory poli-tics, moving them away from an exclusive focus on national states—eithertheir homeland state or the state that has robbed them of having a homelandstate. In the case of high-level professional diasporic groups, the dynamicsare not dissimilar, and indeed the tendency to form global networks isstrong.

These two major types of actors—global corporate capital and the mix ofdisadvantaged and minoritized people—find in the global city a strategic sitefor their economic and political operations. The leading sectors of corporatecapital are now global in their organization and operations. And many of thedisadvantaged workers in global cities are women, immigrants, and people



of color—including diasporic groups in each of these—all people whosesense of membership is not adequately captured in terms of the national,and, in the case of diasporas especially, the national as constructed in thehost country. Indeed, these groups often evince cross-border solidaritiesaround issues of substance.

There is an interesting correspondence between great concentrationsof corporate power and large concentrations of others. Large cities in boththe global South and global North are the terrain where a multiplicity ofglobalization processes assume concrete, localized forms. A focus on citiesallows us to capture, further, not only the upper but also the lower circuitsof globalization. These localized forms are, in good part, what globalizationis about, pointing to the possibility of a new politics of traditionally disad-vantaged and excluded actors operating in this new transnational economicgeography. This politics arises out of actual participation as workers in theglobal economy but under conditions of disadvantage and lack of recogni-tion—whether factory workers in export-processing zones or cleaners onWall Street.

The cross-border network of global cities is a space where we are seeingthe formation of new types of global politics of place. These vary consider-ably: They may involve contesting corporate globalization or involve home-land politics. The demonstrations by the antiglobalization network havesignaled the potential for developing a politics centered on places understoodas locations on global networks. Some of the new globalizing diasporas havebecome intensive and effective users of the Internet to engage in these globalpolitics of place. This is a place-specific politics with global span. It is a typeof political work deeply embedded in people’s actions and activities, madepossible partly by the existence of global digital linkages. Further, it is a formof political and institution-building work centered in cities and networks ofcities and in nonformal political actors. We see here the potential transfor-mation of a whole range of local conditions or institutional domains (suchas the household, community, neighborhood, school, and health care clinics)into localities situated on global networks (Sassen 2004b). From being livedor experienced as nonpolitical, or domestic, these places are transformedinto microenvironments with global span. Microenvironments with globalspan are small local entities in which technical connectivity creates a varietyof links with similar entities in other neighborhoods—whethers located inthe same city, or other cities in the same country or abroad. A communityof practice can emerge that creates multiple lateral, horizontal communica-tions, collaborations, solidarities, and supports. This can enable local politi-cal or nonpolitical actors to enter into cross-border politics (e.g., Warkentin2001; Hajnal 2002).



The space of the city is a far more concrete space for politics than thatof the national state system. It becomes a place where nonformal politicalactors can be part of the political scene in a way that is much easier than atthe national level. Nationally, politics needs to run through existing formalsystems: whether the electoral political system or the judiciary (taking stateagencies to court). Nonformal political actors are rendered invisible in thespace of national politics. The city accommodates a broad range of politicalactivities—squatting, demonstrations against police brutality, fighting forthe rights of immigrants and the homeless, the politics of culture and iden-tity, gay and lesbian politics, and the homeland politics that many diasporicgroups engage in. Much of this becomes visible on the street. Much of urbanpolitics is concrete, enacted by people rather than dependent on massivemedia technologies. Street-level politics make possible the formation of newtypes of political subjects that do not have to go through the formal politi-cal system. These conditions can be critical for highly politicized diasporicgroups and in the context of globalization and Internet access, can easilylead to the globalizing of a diaspora. The city also enables the operations ofillegal networks.

The mix of focused activism and local or global networks creates condi-tions for the emergence of transnational identities. The possibility of identi-fying with larger communities of practice or membership can bring about thepartial unmooring of identities and thereby facilitate the globalizing of adiaspora. It can weaken the radial structure (with the homeland at the cen-ter of the distribution of the groups of a given diaspora). Although this doesnot necessarily neutralize attachments to a country or national cause, it doesshift this attachment to include translocal communities of practice and/ormembership.

Beyond the impact on immigrants and diasporas, these various conditionsare a crucial building block for a global civil society that can incorporate boththe micropractices and microobjectives of people’s political passions. Thepossibility of transnational identities emerging as a consequence of micropol-itics is crucial for strengthening global civil society; the risk of nationalismand fundamentalism is, clearly, present in these dynamics as well.

Conclusion: Urban Growthand Its Multiple Meanings

Major recent developments in urban systems point to several trends. Inthe developing world, we see the continuing growth of megacities and pri-macy, as well as the emergence of new growth poles resulting from the



internationalization of production and the development of tourism. In somecases, these new growth poles emerge as new destinations for migrants andthereby contribute to a deceleration in primacy; in other cases, when they arelocated in a primate city’s area, they have the opposite effect.

In the developed world, and particularly in Western Europe, we see therenewed strength of major cities that appear to concentrate a significant andoften disproportionate share of economic activity in leading sectors. In the1970s, many of the major cities in highly developed countries were losingpopulation and economic activity. Much was said at the time about the irre-versible decline of these cities. But beginning in the mid-1980s, there hasbeen a resurgence that results in good part from the intersection of twomajor trends in all advanced economies: (1) the shift to services, includingimportantly services for firms, such as finance and corporate services, and(2) the increasing transnationalization of economic activity. This transna-tionalization can operate at the regional, continental, or global level. Thesetwo trends are interlinked. The spatial implication is a strong tendencytoward agglomeration of the pertinent activities in major cities. A fact typi-cally overlooked in much of today’s commentary about cities is that thisdynamic of urban growth is based largely on the locational needs or prefer-ences of firms and does not necessarily compensate for population losses tosuburbanization. Urban growth in less developed countries, by contrast,results largely from population growth, especially in-migration. However,beneath the megacity syndrome, we now also see, as of the 1990s, the twotrends mentioned earlier in emergent global cities; in the case of very largecities, these trends are easy to overlook, and the focus is often confined tothe megacity syndrome.

The transnationalization of economic activity has raised the intensity andvolume of transaction among cities; whether this has contributed to the for-mation of transnational urban systems is a question that requires moreresearch. The growth of global markets for finance and specialized services,the need for transnational servicing networks in response to sharp increasesin international investment, the reduced role of the government in the regu-lation of international economic activity and the corresponding ascendanceof other institutional arenas, notably global markets and corporate head-quarters—all these point to the existence of transnational economic arrange-ments with multiple urban locations in more than one country. Here is theformation, at least incipiently, of a transnational urban system.

The pronounced orientation to the world markets evident in such citiesraises questions about the articulation with their hinterlands and nation-states. Cities typically have been and still are deeply embedded in theeconomies of their region, often reflecting the characteristics of the latter.



But cities that are strategic sites in the global economy tend, in part, todisconnect from their region. This phenomenon also conflicts with a keyproposition in traditional scholarship about urban systems—namely, thatthese systems promote the territorial integration of regional and nationaleconomies.

Two tendencies contributing to new forms of inequality among cities arevisible in the geography and characteristics of urban systems. On one hand,there is growing transnational articulation among an increasing number ofcities. This is evident at both a regional transnational level and the globallevel; in some cases, there are overlapping geographies of articulation or over-lapping hierarchies that operate at more than one level; that is, there are citiessuch as Paris or London that belong to a national urban system or hierarchy,transnational European system, and global system. On the other hand, citiesand areas outside these hierarchies tend to become peripheralized.

A second major trend is for powerless groups in global cities to becomeactive in transnational activities, producing a whole series of new, and newlyinvigorated, intercity networks. Although economic transactions amongcities many have launched the formation of emergent transnational urbansystems, a proliferation of people networks began to emerge in the late1980s and have grown rapidly since then. Generally, those networks with akey basing point in cities originate from two types of conditions. One isimmigration and diasporic politics. Although these have long existed acrossthe centuries and the world, the new information technologies have made asignificant difference in the intensity and simultaneity of transactions theymake possible and in the multiplication of transversal linkages, beyond theradial pattern centered in the homeland. The second type originates from avariety of activist and information sharing networks concerned largely withlocalized politico-social struggles: It produces a kind of horizontal globalityanchored in localities. Even individuals and organizations that are notmobile—too poor or persecuted, or simply not interested in traveling—canbecome part of these new global networks. The marking condition is therecurrence of certain issues—environmental, political, social—in many local-ities across the world.

Notes

1. See also the special case of border cities such as Tijuana, which haveexploded in growth because of the internationalization of production in theMexico–U.S. border region and have become major destinations for migrants(Sanchez and Alegria 1992). Another type of case is represented by the new export



manufacturing zones in China that have drawn large numbers of migrants frommany regions of the country (Sklair 1985; Solinger 1999:277–90; Chen 2005). Forone of the best accounts of boder cities see Herzog 1990.

2. This region is here defined as consisting of the island nations between theFlorida peninsula and the north coast of South America, and the independentcountries of the Central American isthmus; it excludes the large nations borderingon the Caribbean Sea.

3. There are several new excellent global city analyses of these cities (e.g.,Ciccolella and Mignaqui 2002; Schiffer Ramos 2002; see various chapters in Gugler2006 and in Amen et al. 2006).

4. The strengthening of Berlin, both through reunification and regaining therole of capital, may alter some of the power relations among Budapest, Vienna, andBerlin. Indeed, Berlin could become a major international business center forCentral Europe after the recent 2004 enlargement, with possibly correspondingreductions in the roles of Budapest and Vienna. However, these three cities may cre-ate a regional transnational urban system for the whole region—a multinodal urbancenter of gravity, in which both competition and a division of functions have theeffect of strengthening the overall international business capability of the region.

5. This is not an exceptional situation. All developed countries in the world nowhave immigrant workers. Even Japan, a country known for its anti-immigrationstance, became a destination for migrant workers in the late 1980s, a role that hascontinued since, albeit with ups and downs and a changing nationality compositionin the flows (AMPO 1988; Asian Women’s Association 1988; Iyotani 1998; Iyotani,Sakai and de Bary 2005; Morita and Sassen 1994; Sassen 1998, chap. 4; [1991]2001, chap. 9).

6. See also the section, “Why Do We Need Financial Centers in the GlobalDigital Era?” in Chapter 5.



national and transnational urban systems€¦ · 03-sassen.qxd 3/21/2006 11:21 am page 45....

Documents