globalization, regional development, and mega-city ...ladupo.igg.unam.mx › ... ›...

Cities, Vol. 20, No. 1, p. 3–21, 2003 2002 Elsevier Science Ltd.Pergamon

All rights reserved.Printed in Great Britain

0264-2751/02 $ - see front matterwww.elsevier.com/locate/cities

doi:10.1016/S0264-2751(02)00092-6

Globalization, regionaldevelopment, and mega-cityexpansion in Latin America:Analyzing Mexico City’s peri-urban hinterland

Adria n G. AguilarInstituto de Geografıa, National Autonomous University (UNAM), Mexico City, Mexico

*Peter M. WardC.B. Smith Sr. Centennial Chair in US-Mexico Relations, Dept. of Sociology and the LBJ School of PublicAffairs, University of Texas at Austin, TX 78713-8925, USA

This paper examines the transformation of urban space in the peri-urban areas of Latin Americanmega-cities, further exacerbating the multi-jurisdictional political divisions that cover a single urbanentity. This is against the backcloth of a sharp decline in metropolitan growth rates, absolutepopulation loss in the city center, and an alleged “polarization reversal” of national urban patterns.It argues that previous approaches have failed to recognize that globally and nationally-derivedeconomic development processes are often vested in these meta-urban peripheries. Using MexicoCity as an example, the authors propose a new generic methodology that will allow for a broaderdefinition and analysis of mega-city and large metropolitan development. Data are presented withinthis new framework that help to unpack the demographic, economic and land-use changes thatare taking place in Mexico City’s broader urban area. Much of the contemporary vibrancy anddynamics of Mexico City’s metropolitan development are occurring in “hot-spots” in the extendedperiphery, which, to date, have rarely been considered an integral part of the mega-city. Yet theseareas are also some of the principal loci of contemporary globalization processes. 2002 Elsevier Science Ltd. All rights reserved.

Keywords: mega-city, Mexico City, metropolitan growth, peri-urban transformation, globalization

Introduction

Our aim in this paper is two-fold. First to open upa new line of enquiry about mega-city1 growth anddevelopment, one that focuses upon the periphery and

∗Corresponding author. Fax.: +1-512-471-1835; e-mail: [email protected] is most frequently used to describe cities of more thanten million people, although it is often used loosely for large metro-politan areas of four to ten million (Gilbert, 1993). Here wehyphenate the term; often it is un-hyphenated.

3

hinterland of very large cities since it is here thatmany of the most important leading changes associa-ted with the impact of globalization upon such citiesare taking place. Second, we wish to develop a meth-odology that will assist in opening up research intothese so-called “peri-urban” areas of mega-cities,identifying the variables and techniques that might beused to capture information and provide for a com-parative analysis between mega-cities and largemetropolitan regions. While our focus is primarilyupon Latin American metropolitan areas and we will

Globalization, regional development, and mega-city expansion in Latin America: A.G. Aguilar and P.M. Ward

be taking Mexico City as a case study in which topilot our proposed methodology, we argue that boththe theoretical and methodological points raised inthis paper may be generalized to many other largemetropolitan areas in Latin America and elsewhere.The study is important since it is one of the first tosystematically analyze the regional “penumbra” ofmega-city development—i.e. the peri-urban region—rather than just focusing upon suburban expansion orurban restructuring.

Traditionally, analysts who study large metropoli-tan areas have done so through one or more of thefollowing optics. First, there are those who haveexamined the role of the large city within the globaleconomy, usually from the point of view of its impor-tance as production, control, or financing centers.Prime examples of this work are Friedmann’s orig-inating hypothesis on “World Cities” , and then dur-ing the 1990s Sassen’s (1991; 1995) work on NewYork, London and Tokyo. A second approach hasbeen to account for the restructuring of these cities,especially the development of control functions toreplace their earlier production role (O’Neill andMoss, 1991; Vogel, 1993). Some authors have alsobegun to look at the consequences of this restructur-ing upon “new” poverty and social organizationwithin these cities (Mollenkopf and Castells, 1991;Sernau, 1997; Sassen, 2000 [especially chapters 3 and6]). A third focus has been to consider the extent towhich these cities are becoming more or less similarover time. Particularly relevant here is whether LatinAmerican metropolitan areas show evidence of con-vergence: becoming physically, culturally and econ-omically similar to cities in the US and Western Eur-ope, driven by global economic processes andinformation flows (Sassen, 1996; Mittelman, 1995).Associated arguments here are that regional inte-gration, industrial and financial decision-making areincreasingly dominated by trans-national corpora-tions, which have come to usurp—in part at least—the sovereignty of the nation state. Fortunately wehave begun to observe a counter-reaction to thesearguments, namely that national and local govern-ments continue to have considerable room formaneuver, and that one needs to focus analysis uponthe “global-local” nexus (see Dicken, 1994; Cox,1997; Parnwell and Wongsuphasawat, 1997). Asstated above, this is also a primary goal of thisresearch, namely to look at the transformations thatlarge cities are undergoing at the beginning of the 21stcentury in the light of globalization, but particularlyfocusing upon the dynamics and spatial changes thatare taking place in the urban space, and particularlyin the peri-urban areas.

Recent evidence shows that metropolitan expansionis taking a different form to that in the past. Whileurban growth rates in these large cities have generallydecelerated in the last two decades, high economicconcentration continues to persist, and metropolitanexpansion incorporates new adjacent municipalities.

4

In territorial terms, from a relatively compact metro-politan space, the contemporary mega-city presents amore polycentric expansion creating a pattern moreassociated with networks and with less pronouncedand less clear-cut borders and boundaries. This cre-ates an expansion pattern with urban dispersion trendsthat incorporate small towns and rural peripheries intoan ever wider and more complex metropolitan system.This is not to hark back to megalopolis ideas of yes-teryear; instead we are arguing that mega-cities areundergoing new dynamics and, as a consequence, arefacing new spatial and organizational challenges asthey seek to manage local urban development withina globalizing world.

Expanded metropolitan regions and theirperipheries

Research in the last decade has postulated the emerg-ence of new urban forms particularly associated withthe largest cities in developing countries. These formshave been mainly the result of what can be describedas region-based urbanization as opposed to city–based urbanization as the city influence is expanded toa wider region facilitated by advances in technology.Lower rates of metropolitan growth have coincidedwith a more intense circulation of commodities,people and capital between the city center and its hin-terland, with ever more diffuse frontiers between theurban and the rural, and a manufacturing deconcen-tration towards the metropolitan periphery, and inparticular beyond into the peri-urban spaces or pen-umbra that surround mega-cities.

Several terms and concepts have emerged tocharacterize this process. In Latin America a decon-centration of urban function and population has beenreported, with a polycentric urban form suggestive ofso-called polarization reversal with the growth ofintermediate sized cities leading to a more “balanced”national urban structure (Townroe and Keene, 1984;see also Gilbert, 1993). Others have suggested thatwhile secondary city growth is often underway, therecontinues to be a heavy concentration of productiveactivities and urban population in a “core region” thatcontains the largest city, but which also extends to amuch larger area and to subsidiary cities within theaegis of the megacity (Gwynne, 1985; UNCHS, 1996:51). In Latin America, such emerging urban forms arenow being described for the largest cities, MexicoCity, Buenos Aires, Santiago and Sao Paulo, albeitvery recently and in ways that have yet to be system-atically researched (see Aguilar, 1999a,b; Ciccolella,1999; De Mattos, 1999; Campolina, 1994; Parnreiter,2002; Ward, 1998).

For Southeast Asian cities a somewhat differentpattern emerges. Here there has been an emphasis onthe fusion or merging of urban and rural functions and


places; such that a mix of rural and urban activities inperi-urban areas of major cities appears to be takingplace—what one author calls “desakota”2—whichessentially means an extended metropolitan region. Inother cases the term mega-urban region has beenadopted, and in different parts of Asia research alongsuch lines has been underway since the early 1990s(see Gingsburg et al., 1991; McGee and Robinson,1995; Firman, 1996; Forbes, 1997).

African cities, on the other hand, do not show asignificant slowdown of urban growth, and one doesnot observe the development of a polycentric andregional urban structures. Instead city growth tends tobe firmly urban and large-city based, and is containedwithin clearly defined boundaries. There is no meta-urban or peri-urban development that is tied to, anddriven by, processes in the urban core (see Briggs andMwamfupe, 2000; Yeboah, 2000).

Our argument here is that it is within the emergingnew spatial order of mega-cities that several importantareas of analysis are being ignored. Specifically, weidentify three main arenas of neglect in the mega-cit-ies’ literature: (i) the dramatic expansion of economicactivities and urban population into regional metro-politan peripheries (i.e. beyond the suburbs); ii) theneed for new criteria and methods to provide for themeaningful delimitation of metropolitan boundariesand mega-city spheres of influence; and (iii), the mul-tiplicity of local jurisdictions and governments thatoften make for a “balkanized” administrative structureof the mega-city and its region, along with an absenceof a single tier of metropolitan government thatembraces the city as a whole.

Thus, at a prescriptive normative level, becausemega-cities display such high levels of economic cen-tralization, the policy solution traditionally has beenone of decentralization, rather than one of directingand shaping suburban and peri-urban expansion.Decentralization has been described by World Bankauthors as the “quiet revolution of the 1990s”(Campbell, 1998), and while decentralization effortsare important and always have been (Gilbert, 1976),there is a danger of misunderstanding what, precisely,is going on in these peripheral areas and theirrelations with the metropolitan core. The expandingand yet increasingly diffuse metropolitan fringesaround these cities are likely to become cruciallyimportant if we are to fully understand the changingnature of mega-cities, and if we are to develop urban-regional policies that will ensure greater sustainabilityof metropolitan areas, particularly in their intersectionwith the natural resource base of the hinterland. Inshort, the slowing of mega-city growth, active decent-ralization, and the growth of intermediate cities—important trends though these are—are potentially

2Desakota (an Indonesian term) were defined as “regions of anintense mixture of agricultural and non-agricultural activities thatoften stretch along corridors between large city cores” (McGee,1991: 7).

5

distracting attention away from what we believe to bea vibrant and new aspect of mega-city development.

Part of the problem is that because the borders andboundaries are no longer as precisely drawn and vis-ible, we have not yet developed adequate measuresand criteria for identifying the processes, nor for por-traying the extent and processes entrained in theseactive peri-urban spaces. This leads, in turn, toanother major problem—namely the disjuncturebetween these spaces and the opportunities for rep-resentative and participative democratic structures toemerge within and between them. Mega-cities usuallycomprise a raft of partially autonomous jurisdictionsand local governments (municipalities), with rela-tively limited coordination and integration betweenthem (Ward, 1996). The number of such jurisdictionsmay vary from a handful to several dozen separategovernments. All these legal and political entities aresimultaneously legitimate local governments, andform part of a single larger social and economic entity(the mega-city). Yet these units often function inde-pendently, and there is a lack of coordination betweenthem. Indeed, they may be competitive with eachother, and engage in a “drive to the bottom” in lower-ing standards as they attempt to outmaneuver eachother and attract capital investment into their areas.In short, most large metropolitan areas have a multi-jurisdictional administrative structure and rarely, ifever, possesses a single metropolitan tier of govern-ment or planning authority as the case of Mexico Citydemonstrates well (Figure 1). Yet it is not an excep-tional case; most megacities have a similar multi-plicity of government and jurisdictions.

However, creating a single metropolitan govern-ment is rarely feasible from a political standpoint,Thus if we can identify the dynamics at play through-out a metropolitan region, then it may be possible tomore effectively explore ways in which jurisdictionsmight cooperate as consortia, or to seek opportunitiesto create new levels of representation and partici-pation across the metropolitan region.

The linkages between economic growth andmega-city development

The traditional policy link towards regional decon-centration arose as a response to the explosive growthof large cities in Latin America during the import sub-stitution industrialization (ISI) period. Until the 1970scapital intensive-industrialization notably concen-trated urban and industrial activities, not to mentionpolitical decision-making, in these cities that becamethe main metropolis of each country with high pri-macy indexes. It also inhibited the growth of nearbycities, since the primary city was able to “rachet-up”its locational advantage by virtue of high rates ofeconomic growth, greater attraction of migrant labor,and its large captive market. Cities spread outwards


Figure 1 Political limits of Mexico City, the Federal District, the Metropolitan Zone and the built-up area

in an apparently uncontrolled manner, particularlythrough the creation of illegal settlements, and aggra-vated problems of the loss of fertile agricultural land,transport problems, poor infrastructure systems, etc.(Gilbert and Ward, 1985). This led to a production ofcontiguous urban space, albeit low density occupationin many cases. Growth was fuelled by in-migrationand by high rates of internal (natural) increase fromthis young adult population. Unlike their developedcity counterparts, the inner-city hub remained denselypopulated, often by working class renter populations,while the suburbs grew apace through legal middleincome residential development on the one hand, andthrough illegal “irregular” development of self-buildsettlements on the other (Ward, 1993, 1998).

Since the 1980s transformations from the globaleconomy have triggered important changes in themetropolitan dynamic of the largest cities indeveloping countries. While these transformationstend to redefine the economic base of such cities, they

6

also recast their territorial patterns in relation to thenew productive conditions. Although mega-citiesshow a decline in metropolitan growth as a whole andin levels of industrial concentration, they remain theprincipal destinations of foreign investment, and con-tinue to emerge as the main service centers, parti-cularly the more advanced services (financial, pro-fessional, high-tech) that support the productiveprocess (Garza, 2000).

As population growth slows, and even becomesnegative in some inner-city districts, three broaddemographic processes may be identified: 1) a flightof (largely) middle-class populations to other parts ofthe country or to nearby towns; 2) a further declinein inner-city population, compensated in part by agrowing densification of the existing built-up area,particularly within self-build settlements developed inearlier decades; and 3), an inflow of new migrantsboth into the urban periphery, and even more so, intothe peri-urban areas. The latter are especially likely


to come from rural areas outside of the metropolitanregion. This particular flow is not usually counted aspart of the mega-city’s growth since it is not addingdirectly to the population total within the boundedmega-city.

In territorial terms, the mega-city now presents amore polycentric spatial expansion of urban centersand sub-centers following a network pattern that tendsto sprawl along major highways and/or railroad linesradiating out from the urban core. In this patternmixed land uses are created in an expanded region,where traditional agriculture is found side by sidewith new housing projects, industrial states, largemodern factories, recreational sites, and all sorts ofsuburban developments. A new architecture and anew spatial configuration of metropolitan develop-ment has emerged.

In order to begin documenting the deconcentrationprocesses now in play, recent interpretations placegreater emphasis on flows between the metropolitancore and other middle-size or small cities; on the for-mation of transport corridors; on patterns andlocations of foreign investment; and upon regionaldispersion of population and manufacturing activities.These analyses look less at the mega-city per se, andmore at where active flows are being directed and atthe new linkages that this entrains with the traditionalcore. But we often fail to see this as a more gen-eralized mega-city space since it is only partiallyurban, and because it apparently does not form partof a contiguous built-up area.

Thus one of the fundamental issues in analyzingthis new milieu of spatially expanded urbanization isto recognize different urban scales and to define thecorresponding boundaries. A number of componentsmay be identified: the urban core which generallycorresponds to the old city limits that existed prior tothe ISI expansion phase; the intermediate ring and thesuburban built up area that represents the continuousconstructed space much of which developed as aresult of ISI and concomitant rapid city growth; themetropolitan area, that broadly bounds the wholebuilt-up area as well some of the rural hinterlandfunctionally delimited by particular criteria(commuter flows, market gardening for city consump-tion, weekend recreational areas, conservation zonesfor air quality improvement etc.); and the mega-cityregion, as defined earlier in the paper.3 Our concernhere is to focus mainly upon the “meta”-region of themetropolitan area by which we mean the corona orhalo that extends beyond the so-called hard borderand its boundaries. It is a sort of “nether” region uponwhich, and within which, the mega-city impacts andinteracts. In reality, of course, metropolitan influenceexpands by different degrees beyond that border alonga gradient (or series of gradients) that will need to be

3McGee (1995: 11) points out that in Asian cities three main unitsare important: city core, metropolitan area, and extended metropoli-tan area.

7

defined for each particular mega-city. What we arelooking for are the social, economic and territorialchanges that gradually reach out from the adjacent tothe more remote municipalities. Within that space, ofcourse, there are likely to be “hot spots” and “coldspots”, or areas that for one reason or another havejumped ahead or fallen far behind in the dynamicsarising from the ties and linkages that they enjoy withthe core.

What are the specific processes of sub-urbanizationand expanded metropolitanization that characterizethese recent phases of development? Primarily it hascomprised a metropolitan expansion outwards occu-pying the adjacent rural areas. In this territorial pat-tern a multitude of distant towns and small localitiesare integrated into the metropolitan daily sphere ofinfluence. There is an increasing functional influenceof the main city on remote municipalities, exercisedmainly through important socioeconomic and land-use transformations of its regional periphery. Muchof the growth in these municipalities derives fromcentrifugal flows from the metropolitan core, as wellas from modest direct in-migration from other peri-urban areas and, to a decreasing extent, from the moredistant provincial hinterlands.

Foreign investment has become one of the mostimportant elements in the restructuring of the urbaneconomy and in the transformation of the urban land-scape. Not surprisingly, therefore, it has also hadimportant impacts on the metropolitan periphery.These impacts include private housing developmentsoften with amenities such as golf courses or countryclubs; office development and shopping malls in thewell located and more accessible areas; and largewarehouses and productive units, although these areusually not located in prime high land value areas,but are even more peripheral and oriented to lowerland cost sites. Inevitably, the distribution of thesedifferent types of developments may lead to sharperdivisions in metropolitan space, particularly whereassociated with richer groups “walling out” lowerincome groups and other land uses.

New types of development in the peri-urbanperiphery

Two distinctive features of metropolitan expansion inthe periphery may be identified. First, urban corridorswhich are lineal developments that may concentratea predominance of different activities along the way:corporate developments, industrial parks, residentialareas, and the density varies from very compact areasto low-urban density with rural landscape in themiddle. Second, urban sub-centers in the peripheryof the mega-city that may be consolidating traditionaltowns once dominated by agricultural activities, or theresult of new (low-income) residential developmentsin metropolitan municipalities of rapid growth incor-


porated into the wider metropolitan complex for thefirst time. These sub-centers play the role of smallcities by providing cheap labor, by concentrating awide range of services, and, to a varying degree, serveas satellites or dormitory towns to the large city andto its metropolitan economy. These sub-centers canbe relatively small localities (often between 10,000and 100,000 inhabitants) located at the edge of themetropolitan frontier, at a distance between 30–60 kmfrom the urban core.4 Although apparently located in“rural’settings away from the metropolitan boundary,these sub-centers are highly capital intensive becausethey respond to the logic of globalization, foreigninvestment, as well as to new consumption patternswhich are typically associated with the mega-cityitself. Juxtaposed to this, these same peri-urban sub-centers are also predominantly working class, rep-resenting concentrations of cheap labor, with gener-ally poor standards of housing, poor quality serviceprovision, low-income consumption patterns and alow standard of living. In short, the locus of workingclass social reproduction is shifting outwards, and isnow vested within the peri-urban periphery, and notjust in the low-income suburbs of yesteryear. Liketheir former ISI working class counterparts, thesepopulations, localities, and their metropolitan munici-palities, are indirectly linked to economic globaliz-ation; yet they hardly seem to receive any benefitfrom it.

Defining the metropolitan area and region

Peripheral expansion means a process of populationredistribution and a restructuring of the metropolitanspace. Whereas the urban core declines in demo-graphic terms, it is in the periphery that metropolitanexpansion accelerates. The general statement thatmetropolitan growth of the largest cities (particularlythose in Latin America) has declined in the last twodecades in fact hides a strong growth contrastbetween the urban core and the more external periph-ery. In the latter some metropolitan municipalities aregrowing as fast as six times the average growth rateof the whole mega-city—further belying earlier argu-ments about “polarization reversal” to secondary cit-ies much further afield (Townroe and Keen, 1984;Aguilar, 1999a; Gwynne, 1985, 1999).

A key issue here is how to delimit metropolitanborders under these new growth patterns, yet this israrely if ever discussed when analyzing large cities indeveloping countries. Instead, researchers continue toadopt definitions and classifications that are drawnfrom advanced capitalist and highly urbanized coun-tries, and which may have lesser relevance to contem-porary mega-cities. In short, what criteria should beused to define metropolitan areas; what geographic

4The emergence of sub centers as a recent urban trend in BuenosAires can be seen in Ciccolella (1999: 19).

8

unit(s) should be used as the building block (basicspatial unit) for defining metropolitan areas; and howcan we build a framework for analysis and monitoringthat is relatively easy to update and revise in an iterat-ive way as circumstances change?

Traditional discussions of criteria used to definemetropolitan and non-metropolitan settlements and todelineate the Metropolitan Area Statistical Standardin the USA are a good example of these concerns,and is very similar to the dilemma exposed here. Theaccount of the spatial representation of metropolitansettlements in federal statistics takes no account ofareas outside them—i.e. non-metropolitan settlementareas.5 Indeed, as one author writes in relation to theUSA: “…the territory outside metropolitan settle-ments, with its more than 10,000 smaller cities andtowns, huge expanses of open country, and over four-fifths the nation’s land area, has never been delineatedwith geospatial units comparable to metropolitanareas” (Dahmann, 1999: 687). This means that theterritory between adjacent metropolitan areas, or justoutside them, lies almost entirely uncategorized andundifferentiated.

The points raised by Dahmann were expected toset some standards for the 2000 census, and four mainapproaches were proposed for the delineation ofmetropolitan and non-metropolitan settlements. Twoof these utilize census tracts as the basic geospatialunit, with commuting from residence to workplacedata as a clustering criterion; and two utilize(surrounding) counties as the basic geospatial unit,with one of these also using commuting as a clus-tering criterion, and the other allowing for intensityof settlement alone to serve as the defining criterion(US Office of Management and Budget, 1998; citedin Dahmann, 1999: 685). The whole idea here is notto select between competing criteria, but to stressthat geographical theory has been largely unableto incorporate the nature of evolving settlementsas it applies to settlement form and function.6Moreover, as Brenner (2002) points out, the multi-jurisdictional nature of many metropolitan areas posesnew challenges for governance, and call for a “newpolitics of scale” – a point that is equally pertinent inLatin America, as we noted earlier (see also Ward,1999). Although we have argued above for the needto develop new tools of analysis, the case of theUnited States is helpful for two reasons: first, there isa long tradition in the US of delineating metropolitancommunities, that goes back to the beginning of the20th century; and second, because the ongoing con-cerns about suburban growth and (peri-)urban

5Although the overwhelming majority of Americans today live inmetropolitan settlements as currently delineated, these settlementsaccount for less than one-fifth of the nation’s total land area(Dahmann, 1999: 687).6Adams et al (1999: 697) make this point for the case of theUnited States.


“sprawl” in American cities, closely parallel severalof the issues that concern us here.

As an example, those who defend the intensity ofsettlement propose as a criterion the relative residen-tial population density calculated at the county level.They argue that exclusive attention to the journey towork (commuting) is inadequate since it ignores themany households’ journeys that involve non-workdestinations and ignores the everyday travel behaviorof non-working households. It also ignores peoplethat work at home or out of their homes; indeed, manyworkers today are anchored to no one particularlocation (Adams et al., 1999: 708).7 On the otherhand, those defending commuting as the predominantcriterion stress the concept of functional integrationwhere work is still the dominant organizing activityin our lives. They argue that, whereas the metropoli-tan form has changed from a single principal core tomany, the basic movement of workers traveling fromwhere they live to where they work continues. More-over, that most household residential locationdecisions are still made with regard to the place ofwork (Rain, 1999: 754).

In light of this, several crucial issues may be ident-ified for particular analysis in the case of mega-citiesin developing countries. First, the newly emergingmetropolitan forms demand new approaches and cri-teria to delineate individual metropolitan areas.Second, metropolitan peripheries are expanded andcomplex territories that have to be characterizedwithin an continuum urban–rural. In principle thereare likely to be internal and external peripheries tothe metropolitan border that vary according to land-uses, urban networks, patterns of economic activityand living conditions, as well as geographical barriers.Third, residential densities, the mobilization of trans-port networks, and the relative importance of place ofwork may offer significant variables for metropolitandelineation that will allow for comparisons to bemade with other urban areas. Fourth, it is importantto analyze key variables that are shaping urbanchange in peri-urban spaces, particularly the way pub-lic policy is implemented in large and multi-jurisdic-tional metropolitan areas.

In the following section of the paper we seek tocreate a framework that might serve as a basis foranalyzing these changes in comparative perspective.By way of an example we have chosen Mexico Citysince this is the mega-city that is most familiar to us.Although every city is unique, we believe that manyof the measures that we identify, and the processesthat we depict for Mexico City are generalizable toother situations also, not least within the Americas.

7Tract-level density analysis permits, for example, the delineationof five levels of an “urban-rural continuum”: metropolitan core;metropolitan outlying; non-metropolitan, adjacent to metropolitanarea; non-metropolitan, non-adjacent with city; non-metropolitan,non-adjacent without city (Cromartie and Swanson, 1996; cited inAdams et al., 1999: 709).

9

Case study: the metropolitan growth andexpansion of Mexico City

With the aim of analyzing its regional periphery theMetropolitan Area of Mexico City (ZMCM) was div-ided into three main zones: (i) the existing built uparea that represents the central along and contiguousurban areas (zone A in Figure 2); (ii) a metropolitanperiphery or inner peri-urban space contiguous to thebuilt up area that includes a mixture of urban and ruralland uses and appears to be functionally integrated tothe city, albeit in a way that as yet remains impreciseand unanalyzed (zone B in Figure 2). We regard thisperiphery to be an integral part of the metropolitanarea; (iii) and an outer peri-urban space or expandedperiphery that is constituted by the more remotemunicipalities that are located somewhat outside themetropolitan area but are contiguous to it (zones Cand D in Figure 2). These are recently integrated, andby virtue of their geographical proximity the domi-nant metropolitan area is directly affecting them. Inreality, in other mega-cities it will often be necessaryto construct these outer spaces differently. In our case,for example, we found it helpful in the first instanceto create two separate areas in the outer-periphery(C & D), and then later to combine them into a singlecategory (Table 1).

For this division the delegaciones of the FederalDistrict and the municipalities of the State of Mexicowere adopted as the principal political jurisdictions.Not surprisingly the built-up area is not always con-tiguous with these political-administrative boundaries,so we defined as “urban” those cases in which thebuilt up area covered the majority of a unit’s territory,even if some open (rural) spaces still existed. Theoverall limits of the metropolitan zone of 1995 areas delimited by the Consejo Nacional de Poblacion(CONAPO). The outer or expanded periphery wasdelimited taking into account the fact that to the southand southeast there are natural barriers that make thecity’s urban expansion difficult in those directions.Thus, few municipalities were considered to form anactive part of the periphery in those cases. To thewest, not only are there also similar physical barriers,but given the expansion of the city of Toluca, the twometropolitan areas have begun to join together, andin effect municipalities are under the influence of bothcities. Thus, taking account of the recent rapid metro-politan expansion to the north and east, it was in thesedirections that peripheral municipalities were selectedfor close scrutiny. A territorial fringe contiguous tothe metropolitan area was delimited taking account ofthe presence of a road network, small towns and prox-imity.

In order to characterize the growth and expansionof Mexico City in the last 30 years four main pro-cesses are analyzed below; the first two are importantfrom the demographic point of view, while the third


Figure 2 The Metropolitan area of Mexico City and its expanded periphery

stands out in territorial terms, and the fourth refers tothe economic changes in the area. These are: (i) thedeclining trend of urban growth of the city as a wholesince 1970; (ii) the population redistribution trendswithin the innermost parts of the metropolitan area;(iii) a continued metropolitan expansion, consolidat-ing urban sub-centers and corridors, and the incorpor-ation of more and more distant municipalities; and(iv) a dispersion of manufacturing activities, andchanging predominance of specific sub-sectors.

Demographic changes

Mexico City’s urban growth During the ISI periodand until the early 1970s, Mexico City received alarge proportion of migrants and concentrated animportant share of productive activities. During thisperiod it experienced the highest growth rates of itshistory with over 5% per annum in the 1950s and1960s. But after 1970 its growth started to decline,and the period 1970–1990 saw a diminished rate of2.62% for the whole ZMCM (1970–80), while in thefollowing decade it dropped to 1.64%.

Several factors help explain this decline. Economiccrises and instability in the 1970s and early 1980sreduced the generation of manufacturing employmentand made the city a more expensive place to live.Also, deconcentration policies started in the 1970sbegan to take effect and promoted growth in inter-mediate cities—a situation that was strengthened withthe adoption of an export-oriented model from the

10

mid to late 1980s onwards. Intensive trade relationswith the USA favored a deconcentration of manufac-turing activities from the capital city, and in the searchfor better locational advantages (especially low-costlabor), foreign investment went increasingly intomiddle and small-sized cities. Other factors in thisdecline were the high levels of air pollution, increas-ing crime rates and declining quality of life generally.

Population redistribution Three broad trends can beidentified in the last decades: a depopulation of thehistorical city center; a greater demographic concen-tration in the city area of the State of Mexico withrespect to that in the Federal District; and a rapidgrowth of differentiated metropolitan peripheries.

By 1950 when the capital was growing rapidly thecentral part of the city covered most of the wholemetropolis, comprising 66% of the total population,all of which was within the Federal District boundary.In the following decades urban expansion and deterio-ration of the historical center began to push the popu-lation out from this area towards the (then) new per-ipheral residential developments. However, even until1970 the city center was still gaining population,although its dominance as a residential area was indecline, such that its proportion of population dimin-ished to 32% of the total city. By this time the overallpopulation was beginning to expand sharply into thesurrounding State of Mexico. This process continued,and by 1990 that central city’s proportion of the popu-


Tab

le1

Pop

ulat

ion

grow

thin

the

ZM

CM

and

its

expa

nded

peri

pher

y,19

70–2

000

MA

INZ

ON

ES

1970

%19

90%

2000

%T

otal

chan

geT

otal

chan

ge19

70–1

990

1990

–200

0

Fede

ral

Dis

tric

t6,

874,

165

72.3

08,

235,

744

51.5

68,

591,

309

45.5

30.

910.

42C

entr

alci

ty2,

902,

969

30.5

31,

930,

267

12.0

81,

688,

401

8.95

�2.

02�

1.33

Mun

icip

aliti

esSt

ate

ofM

exic

oin

the

ZM

CM

2,20

8,32

123

.23

6,96

0,76

343

.57

9,28

1,94

949

.19

5.91

2.92

Urb

anm

unic

ipal

ities

8,63

6,30

990

.83

13,7

42,0

3586

.03

15,8

58,0

7484

.05

2.35

1.44

Met

ropo

litan

peri

pher

y45

4,88

04.

781,

484,

765

9.29

2,06

1,53

410

.93

6.09

3.34

Tot

alZ

MC

Ma

9,09

1,18

995

.61

15,2

26,8

0095

.32

17,9

19,6

0894

.97

2.61

1.64

Exp

ande

dpe

riph

ery

417,

176

4.39

747,

516

4.68

948,

618

5.03

2.96

2.41

TO

TA

L9,

508,

365

100.

0015

,974

,316

100.

0018

,868

,226

100.

002.

631.

68

a Inc

lude

sth

eur

ban

and

peri

pher

alm

unic

ipal

ities

11


lation had dropped to 13%, with the central delega-ciones already showing negative growth rates, and by2000 this negative growth was further expanding intoother adjacent delegaciones (like Azcapotzalco andIztacalco) and the central “core” loss of populationwas even more pronounced (see Figures 3a and b).The proportion of population in the city center dimin-ished to 9% (see also Table 1). Rapid suburbaniz-ation, and a shift towards more strict land-use controlson the mountain slopes to the south of the FederalDistrict, led to an ever increasing proportion ofgrowth settling in the metropolitan municipalities ofthe State of Mexico. The city proportion of populationfor the Federal District fell to 54% in 1990, and in2000 this proportion, for the first time, representedless than half of the total population (48%).

In essence these data show great disparities in thepace of urban growth at the center of the ZMCM.Whereas the Federal District is growing very slowly(less than 1% in the period 1990–2000), and someof their delegaciones show negative growth rates, themetropolitan municipalities of the State of Mexico are

Figure 3 Population growth rates by Main Zones, 1970–1990 (Fig. 3a) and 1990–2000 (Fig. 3b). Population growth rates in theexpanded periphery, 1970–1990 (Fig. 3c) and 1990–2000 (Fig. 3d)

12

registering average rates of increase that are threetimes higher (between 3 and 4%) that of the FederalDistrict, and some municipalities are growing atabove 5% per annum. It is obvious that faster large-scale urbanization in terms of residential and indus-trial developments is taking place in these munici-palities, and that the metropolitan influence of the cityis rapidly expanding to the immediate peripheries inthese areas.

The distribution of population growth in the metro-politan periphery showed a differentiated pattern dur-ing the last 30 years. In the period 1970–1990, thehigher growth rates were registered in the Metropoli-tan Periphery (zone B, see Figure 2) particularly atthe edge of the built-up area and in a broad fringe tothe northeast. Here it is worth emphasizing that thehighest growth rates were located along the mainroads to Puebla (east) and to Pachuca (north-east). Inthe case of the Expanded Periphery (zone C) growthrates were more within the average of the wholeZMCM, with some exceptions or hot spots along theroads to Queretaro and Tulancingo (north-west) but


these were never within the highest range. It is clearthat metropolitan growth showed a marked tendencyto concentrate in the Metropolitan Periphery andparticularly along the main urban corridors to thenorth and east (see Figure 3a).

In the period 1990–2000, there is a centrifugaltrend of the highest growth rates into the expandedperiphery. Although the highest growth rates are stillpresent in most of the territory of the metropolitanperiphery, specifically in the north and east, nowsome of the highest population growths have spreadto the expanded periphery, particularly along the roadto Pachuca (north-east), in a larger territory contigu-ous to the metropolitan border of the ZMCM (see Fig-ure 3b). These data show the gradual expansion ofa stronger city influence into the next correspondingperipheral space.

If we analyze population growth only in the city’speriphery (both metropolitan and expanded) inrelation to the ZMCM growth rate average (which isslightly higher than that of the urban zone), there isevidence that since 1970–1990 the city’s influencewas already strong in the expanded periphery and twozones can be delineated (see Figure 3c). The first one,contiguous to the metropolitan border but external toit, receiving the more direct impacts of urban expan-sion and growing at above average rates. The secondzone comprises the more distant municipalities andthe remote fringe, which apparently experienced onlyminor impacts at this time, and grew at the averagerate or less.

In the period 1990–2000, the delineation of thesetwo external peripheries (within the expandedperiphery) remains the same, but with important dif-ferences that higher growth rates are now present ina larger number of municipalities in the first zone;and a reduced number of rapidly growing unitsappeared in the second external periphery, (which wasnot the case in the previous period, see Figure 3d).Again, a centrifugal trend of city influence is visiblewithin these differentiated peripheries.

Territorial changes

The metropolitan expansion Since the epoch of rapidurbanization the ZMCM has continued to incorporatemore and more distant municipalities into the metro-politan dynamic (see Figure 1 above). In 1970, itsmetropolitan area embraced 15 delegaciones (out of16) of the Federal District, and 11 municipalities fromthe State of Mexico. By the mid 1980s its metropoli-tan area added a single further delegacion and tenmore municipalities of the State of Mexico; while in1995 an additional 16 municipalities of the State ofMexico and one from the state of Hidalgo were incor-porated into the definition of the ZMCM. In otherwords, the expansion of its metropolitan influence andthe transformation of its immediate rural peripheries

13

has been consistent throughout these years, with adramatic quickening in the past two decades.

Other important metropolitan processes in the Cen-tral Region of Mexico have accompanied metropoli-tan expansion of the ZMCM. These are linked to theZMCM and comprise the consolidation of economiccorridors and the emergence of new urban centers atthe metropolitan level, thereby shaping the overalltrends of expansion of the mega-city in its peripheralregion. In 1970 there were only two cities with metro-politan characteristics in the Central Region of Mex-ico, these being the ZMCM itself and the city of Pue-bla (names in italics on Figure 4). At that time otherimportant cities (for example, Toluca andCuernavaca) had not expanded sufficiently to inte-grate other municipalities and thereby take on ametropolitan character of their own; and the numberof small cities (from 15,000–100,000 inhabitants) wassmall—only 14 in total—and the pattern was highlydispersed, with only a few close to the two metropoli-tan areas defined at that time (see Figure 4).

By 1995 the picture had changed substantially(Figure 4). There were seven metropolitan areas intotal, apart from those already mentioned, Queretaro,Pachuca, Toluca Cuernavaca and Tlaxcala. But whatwas even more impressive was the increase of smallcities in the region: more than 120 small urban cen-ters, most of them located either within the metropoli-tan areas or in close proximity to them. This processdisplays a much more dispersed urban patternthroughout the whole region, and the emergence of amultitude of small urban centers in the metropolitanareas tends to indicate a more multinuclear structureforming the basis of the metropolitan territories. Thenetworks of roads that connect the main metropolisare the axes that give form to this new structure ofthe mega-city expanded region (see Figure 4).

In the particular case of the ZMCM, in 1995 onecan identify around 40 small centers (larger than15,000 inhabitants), mainly to the north and eastwhere metropolitan expansion has occurred more rap-idly. These particular localities (municipalities andtowns) have apparently strengthened their productiveand labor links with the core city due to the influenceof various factors, most importantly, good road infra-structure and the availability of flat land. It is clearthat the metropolitan process has expanded its influ-ence in these directions to ever more distant locationsand that this has stimulated the growth of small cen-ters that are being rapidly transformed in socio-econ-omic terms, and that are being shaped by the dynamicof the expanded metropolitan region.

It is important to underline that while the metro-politan periphery is in constant transformation, weactually know little about it. We need to know moreabout the main socio-economic changes that theseinner- and outer- peri-metropolitan areas are experi-encing as they are incorporated, and as they enter thesphere of influence of the central city. Given thatthese small cities are now playing an important role


Figure 4 Metropolitan zones and urban sub-centers in the central region, 1995

as the new urban sub-centers in the metropolitan terri-tory, it is important to highlight some of their maincharacteristics to which we now turn.

The formation of urban sub-centers and corri-dors We were able to begin to unpack the magnitudeof these trends by examining socio-economic data forspecific small centers that have emerged in the metro-politan periphery of Mexico City. Three types of datafor the period 1970–1990 were analyzed for a sampleof 16 out of the 40 centers located in the metropolitanperiphery. These socio-economic data include econ-omic changes, services, and education level of thepopulation, each of which we discuss briefly below.The main feature of this information is that it dos notrefer to the municipality per se, but refers to the singlelocality (center), thereby offering a more precise pic-ture of the real transformations of these sub-centers.It is important to emphasize that these sub-centershave two origins. They either constitute traditionaltowns that have acquired urban elements and haveconsolidated their urban function; or they haveemerged recently as high-density residential develop-ments with a population coming mostly from urbancentral areas.

14

The data for the 16 sample sub-centers for differentlocations in the metropolitan periphery are displayedin Figure 5 (“pie”-charts). Several important featuresemerge. First, if we look at the main changes by econ-omic sector the percentage of population occupied inurban activities registered a notable increase, to thedetriment of primary activities. In the north, somesmall centers augmented their share of manufacturingdue to the influence of the large and main city’sindustrial district (see-pie charts for Coyotepec, Santi-ago Tequixquiac and San Juan Zitlaltepec). In twocenters the percentage increase was of more than 20points. The sort of industries that predominate hereare heavy, large-scale and high technology enterprisessuch as metallic and chemical industries.

To the eastern periphery, another important groupof sub-centers show a substantial increment of indus-trial activities, however, these types of activities areassociated with the emergence of illegal, large, low-income settlement swathes (see Chalco, San MartinXico and San Miguel Coatlinchan). This type ofmanufacturing more commonly forms part of theeconomic survival strategies of the poor, and is small-scale and informal. Thus it is concentrated in house-holds or workshops (especially micro-enterprises). Inthe same direction (east and southeast) one also finds


Figure 5 ZMCM and expanded periphery. Changes by economic sector in urban sub-centers, 1970–1990

15


the only (small) group of sub-centers that registereda decline in their population occupied in manufactur-ing (see Chiconcuac, Texcoco and San Rafael). Theseare centers where traditional and/or artisan industries(textiles and paper industry) lost importance, whilethe proportion employed in commerce and servicesincreased concomitantly.8

Therefore, the largest increases in tertiary activitiestend to be associated to those sub-centers that haveexperienced either important expansion of industrialactivities or large-scale residential developments.These new commercial and service centers are mainlylocated along three main urban corridors: first, theroad going to the city of Pachuca through localitiessuch as Ojo de Agua, Acozac and Tizayuca. Second,the corridor that links centers like Texcoco, Chicon-cuac and Teotihuacan to the north-east; and third, theroad that extends from Chalco to Amecameca andOzumba in the west-southwest.

Data for dwelling characteristics, specifically thepresence of drainage and water supply in the house-holds of these small centers shows a general improve-ment in the period (see Figures 6a and b). For bothof these urban services, the relative coverage in 1990is usually higher than that of 1970, notwithstandingthe high absolute increase in demand. Thus, metro-politan expansion for these centers has meant animprovement in basic services with one significantexception, namely those centers that saw the estab-lishment of large illegal settlements (Chalco and SanMartın Xico for example). In this case the situationin 1990 was much worse than in 1970 since the pro-vision of services has lagged far behind the speed ofpopulation growth.

One additional variable analyzed was the percent-age of population aged 15 years or over with post-primary education. In essence this variable serves asa surrogate for the share of the more educated popu-lation in each population center. The results are inter-esting because in contrast with the improvement ofurban services pointed out above, most of the centersshow that the proportion of a relatively educatedpopulation declined during this period (see Figure 7).Metropolitan expansion in this case has meant aworsening in education levels of the population. Thistrend of deteriorating human resources may be linkedto two main hypothesis: first, a large proportion of themost educated people have moved away from thesecenters to more central urban areas in Mexico City,or to other cities, and are being replaced by a morerural in origin and lower-educated migrants. Second,due to economic constraints some household mem-bers have been forced to leave school and to enterwork at an early age; in this way they have lost theopportunity to be more qualified. A combination ofboth processes is also possible. But the outcome is a

8For a discussion on the tertiarization of the urban economy andthe change from more to less stable jobs, particularly informal andless productive occupations, in Mexico City see Aguilar (1997).

16

deterioration in the human resources of the populationliving in these centers which increasingly appear tofunction as dormitory towns or satellite cities forimpoverished metropolitan populations in the periph-ery.

Economic changes

As part and parcel of the principal economic changesin the metropolitan periphery two main features canbe underscored: the dispersion of manufacturingactivities from central-city areas to the periphery, andchanges in the mix of dominant industrial activitieswithin the periphery itself.

Peripheral dispersion of manufacturing activitiesIn the mid-1970s the urban area (zone A on Figure2 above) concentrated the majority of manufacturingemployment with 94.5% of the overall metropolitantotal; with the metropolitan (B) and expanded periph-eries (C & D) having a rather modest proportion ofactivity with 2.1 and 3.4% of the total respectively(Table 2). Thus, although the numbers are small, themore remote periphery (of C & D) actually had aslightly larger proportion of manufacturing jobs thandid the metropolitan municipalities (B) that is alreadysuggestive of its incipient role as a regional locationof economic and urban activities. In the following twodecades two clear trends can be observed: first, thebuilt-up area started to lose relative importance interms of the concentration of manufacturing activities,with the effect that its percentage of employed popu-lation diminished to 91% in 1994 (Table 2). Second,both peripheries continued to increase their proportionof manufacturing reaching 4 and 5% for the expanded(C & D) and the metropolitan (B) peripheries respect-ively (i.e. reversing their former ranking), an equival-ent to almost 80,000 jobs. Thus, in this time periodthe share in the metropolitan periphery becameslightly more important. It is notable that between1986–1994 the expanded periphery had a very smallgain of manufacturing jobs apparently due to a sub-stantial internal losses in the dominant sub-sectors, asis shown below (see Table 2).

Clearly, in relative terms manufacturing activitiesare still highly concentrated in the urban area butappear also to be gradually dispersing to the wholeperiphery. The economic base of the latter is becom-ing dominated by urban-service functions and landuses, within which manufacturing represents animportant element of metropolitan expansion.This trend is also quite clear from Figure 5(pie charts).

Restructuring in the dominant manufacturing sub-sectorsThe growth of the manufacturing sector in the periph-ery shows important internal changes in its dominantsub-sectors for the period 1976–1994. Given that


Figure 6 ZMCM and expanded periphery. Water provision in urban sub-centers, 1970–1990 (Fig. 6a). ZMCM and expandedperiphery. Presence of drainage in urban sub-centers, 1970–1990 (Fig. 6b)

there are differences in the trajectories of performancebetween industrial sub-sectors in both peripheries, wedisaggregated them in the following way. In 1976,there were three dominant sub-sectors in the metro-politan periphery: Textiles, Mineral Products, andMetallic Products,9 which combined comprised 62%of the employed population in the manufacturing sec-tor. In 1986, two of these sub-sectors retained theirimportance (Metallic Products and Textiles) whileanother emerged, this being Food Products. In total,these three sub-sectors made up 70% of the sector,increasing marginally to 71% by 1994.

In the so-called expanded periphery (C & D) thesituation has been to some extent similar but unstable.In 1976 three main sub-sectors concentrated 74% ofthe economically active population: Metallic Pro-

9Mineral Products exclude oil and coal; Metallic Products includemachinery, equipment, and surgical and precision instruments.

17

ducts, Mineral Products, and Metallic Industries. By1986, two of the sub-sectors had lost importance, andMetallic Products, Textiles and Chemical Industry,rising to 74% of the employed population by 1994,dominated 66% of the sector. Thus, the dominant sub-sectors are very similar to those that predominate bothin the inner periphery as well as in the whole metro-politan area of Mexico City.

The internal changes in the sub-sectors of theexpanded periphery, during the period 1986–1994, arealso distinctive. On the one hand, there was animpressive increase of the textile industry that becamethe dominant sub-sector with a gain of more than7000 jobs; while on the other, the Metallic Pro-ducts’sub-sector became the second most importantbut have, instead, suffered a dramatic loss of morethan 4000 jobs. (In 1994 the total number of jobs wasalmost identical as in 1986.) Additionally, there wasa loss of a total of 6268 jobs in other less important


Figure 7 ZMCM and expanded periphery. Population older than 15 years with post-primary education, 1970–1990

sub-sectors10 (see Table 2). Viewed thus, manufactur-ing activities in the more remote periphery appear tobe unstable and subject to processes indicative ofrestructuring, job losses and gains, and an emergingdominance of a smaller number of sub-sectors. Thuswe appear to be observing important shifts takingplace within these expanded metropolitan peripheriesin response to the changing dynamics of economicgrowth and investment. In many ways, these new per-ipheries are the leading experiential edges of change,and to the extent that such processes are being drivenby Mexico’s ever tighter nexus into the global arenathese spaces are the indicators of that interaction.

Conclusions

In this paper we have argued that traditional defi-nitions and criteria for depicting and analyzing metro-politan space and mega-cities inhibit our ability toadequately understand the nuanced processes ofdemographic and economic transformation that areunderway—whether these are driven primarily byglobal or national pressures. Specifically, we haveargued for a new categorization of the peri-urbanspaces beyond the classic metropolitan and mega-cityurban fringe since it is there that many of the definingprocesses are being anchored today. While overallmetropolitan growth is often perceived to be indecline, adopting a broader definition and frame of

10These sub-sectors were Paper and Printing Products, MineralNon-Metallic Products, and Basic Metallic Industry.

18

reference for the metropolitan urban space to extendto their “extended peripheries” or hinterlands, mayrequire us to revise our understanding of populationtrends in mega-cities. We have proposed that infuture, mega-city analysis must embrace what isgoing on beyond the urban fringe, and often at a con-siderable distance from the “hard” urban boundaries,since it is here that much of the dynamics of (greater)metropolitan growth is taking place.

We have developed our arguments using the caseof Mexico City, and have sought to offer guidelinesabout how, in future, region-based analysis of mega-city development might be gauged. We have shownthat an important step is to identify the spatial areathat can meaningfully be depicted as the “extended-periphery” for any given mega-city. This will varyfrom case to case depending upon topography, thepresence or absence of other major nearby centers,principal transportation routes, etc. But our point isthat it is in this “penumbra” or peri-urban area thatimportant contemporary processes are underway,often in “hot-spot” locations that, once identified, canbe the focus of closer scrutiny and fieldwork. Indeed,our own survey of the extended periphery of MexicoCity has revealed important new growth points aswell as sectoral changes that are underway. Butbecause these are recently entrained, they may also bethe most volatile and vulnerable to economic cycles.Moreover, it is in this new peri-urban space that thereproduction of labor is most likely to be concentratedin the world’s largest cities in the 21st century. And,although our data suggest that infrastructure con-


Tab

le2

ZM

CM

and

Exp

ande

dP

erip

hery

.G

row

than

dre

stru

ctur

ing

inth

em

anuf

actu

ring

sect

or,

1976

–199

4

Em

ploy

edpo

pula

tion

Dif

fere

nce

Per

cent

age

1976

1986

1994

1986

/197

619

94/1

986

1976

1986

1994

Exp

ande

dpe

riph

ery

(C+D

)25

,578

37,1

3237

,203

11,5

5471

3.39

4.14

4.22

25,5

7837

,132

37,2

0310

0.00

100.

0010

0.00

Food

prod

ucts

1902

2105

3610

203

1505

7.44

5.67

9.70

Tex

tiles

and

leat

her

524

5997

13,2

6554

7372

682.

0516

.15

35.6

6W

ood

prod

ucts

5312

663

173

505

0.21

0.34

1.70

Pape

rm

anuf

actu

ring

and

prin

ting

1893

1595

1008

�29

8-5

877.

404.

302.

71C

hem

ical

indu

stry

774

5329

5402

4555

733.

0314

.35

14.5

2M

iner

alan

dno

n-m

etal

licpr

oduc

ts33

4750

0937

7816

62�

1231

13.0

913

.49

10.1

6B

asic

met

allic

indu

stry

3187

3485

448

298

�30

3712

.46

9.39

1.20

Met

allic

and

prec

isio

npr

oduc

ts12

,465

13,1

8987

3972

4�

4450

48.7

335

.52

23.4

9O

ther

1433

297

322

�11

3625

5.60

0.80

0.87

Met

ropo

litan

Peri

pher

y(B

)15

,934

28,6

9442

,360

12,7

6013

,666

2.11

3.20

4.81

Urb

an(A

)71

2,79

083

0,14

280

1,51

911

7,35

2�

28,6

2394

.50

92.6

590

.97

ZM

CM

728,

724

858,

836

843,

879

130,

112

�14

,957

96.6

195

.86

95.7

8

19


ditions associated with this urban growth are improv-ing relatively, the deterioration in human resources asmeasured by education levels is disturbing—not leastat a time when countries such as Mexico are urgentlyseeking to raise the human capacity of their work-force. Although these data and explanations mustremain tentative at this stage, future analyses willneed to unpack the extent to which new meta-urbaninvestment is undermining rather than enhancinghuman resource development. To the extent that theseareas have remained largely invisible to scrutiny inthe past, it would be perverse indeed were any nega-tive consequences of economic change to be ignoredbecause they were not seen to be intrinsically tied tothe mega-city core. In short, the constructs of mega-city dynamics have changed. We hope that our depic-tion of Mexico City’s contemporary region-basedurbanization, and together with our proposals for“pushing the envelope” of criteria used to measuremetropolitan space, will encourage more detailedextensive examination of mega-cities in the Americasand elsewhere.

Acknowledgements

Adrian G. Aguilar wishes to acknowledge the finan-cial support from the UNAM in the period August2000–July 2001. The authors wish to thank the col-laboration of Clemencia Santos and Irma Escamillafor the assistance with the data analysis, and prep-aration of the maps and figures.

ReferencesAdams, J S, VanDrasek, B J and Phillips, E G (1999) Metropolitan

area definition in the United States. Urban Geography 20(8),695–726.

Aguilar, A G (1997) Metropolitan growth and labor markets inMexico. Geojournal 43(4), 371–383.

Aguilar, A G (1999a) Mexico City growth and regional dispersal:the expansion of largest cities and new spatial forms. HabitatInternational 23(3), 391–412.

Aguilar, A G (1999b). La Ciudad de Mexico en la Region Centro.Nuevas Formas de la Expansion Metropolitana. In: Transi-ciones. La Nueva Formacion Territorial de la Ciudad de Mex-ico, eds. J. Delgado and B. Ramırez, pp. 147–169. Programade Investigacion Metropolitana-UAM, Plaza y Valdez, Mexico.

Brenner, N (2002) Decoding the newest ‘MetropolitanRegionalism’ in the USA: a critical overview. Cities (19)1,3–21.

Briggs, J and Mwamfupe, D (2000) Peri-Urban development in anera of structural adjustment in Africa: The city of Dar esSalaam, Tanzania. Urban Studies 37(4), 797–809.

Campolina, C (1994) Polygonized development in Brazil: neitherdecentralization nor continued polarization. International Jour-nal of Urban and Regional Research 18(2), 293–314.

Ciccolella, P (1999) Globalizacion y dualizacion en la regionmetropolitana de Buenos Aires. Grandes inversiones y reestruc-turacion socioterritorial en los Anos Noventa. Revista Eure25(76), 5–27.

Cox, K (ed.) (1997) Spaces of Globalization: Reasserting thePower of the Local. Guilford Press, New York.

Cromartie, J and Swanson, L (1996) Defining Metropolitan Areasand the Rural-Urban Continuum: a Comparison of StatisticalAreas Based on County and Sub-County Geography, Staff

20

Paper No. AGES9603, Washington D.C., Economic ResearchService, US Department of Agriculture.

Dahmann, D C (1999) New approaches to delineating metropolitanand non-metropolitan settlement: geographers drawing the line.Urban Geography 20(8), 683–694.

DeMattos, C A (1999) Santiago de Chile, globalizacion y expan-sion metropolitana: lo que existıa sigue existiendo. Revista Eure25(76), 29–56.

Dicken, P (1994) The Roepke Lecture in Economic Geography.Global-local tensions: firms and states in the global space econ-omy. Economic Geography 70, 101–128.

Firman, T (1996) Urban development in Bandung metropolitanregion: a transformation to a Desa-Kota region. Third WorldPlanning Review 18(1), 1–22.

Forbes, D (1997) Metropolis and Megaurban region in Pacific Asia.Tijdschrift voor Economische en Sociale Geografie 88(5),457–468.

Garza, G (2000) Superconcentracion, crisis y globalizacion del sec-tor industrial, 1930–1998, pp. 170–177. In: Gustavo Garza(ed.). La Ciudad de Mexico en el fin del segundo milenio. Mex-ico DF: El Colegio de Mexico and the Gobierno del DistritoFederal.

Gilbert, A (1993) Third world cities: the changing national settle-ment system. Urban Studies 30(4–5), 721–740.

Gingsburg, N, Koppel, B and McGee, T G (1991) The ExtendedMetropolis. Settlement Transition in Asia. In University ofHawaii Press, Honolulu.

Gwynne, R (1985) Industrialization and urbanization in LatinAmerica. In Croom Helm, London.

McGee, T G (1991) The emergence of Desakota regions in Asia:expanding a hypothesis. In The Extended Metropolis. SettlementTransition in Asia, (eds) N Ginsburg, B Koppel and T GMcGee, pp 3–25. University of Hawaii Press.

McGee, T G and Robinson, I M (1995) The Mega-Urban Regionsof Southeast Asia. In UBC Press, Vancouver, Canada.

Mollenkopf, J and Castells, M (eds) (1991) Dual city: the restruc-turing of New York. Russell Sage, New York.

Mittelman, J (1995) Rethinking the international division of labourin the context of globalization. Third World Quarterly 16,273–295.

O’Neill, H and Moss, M (1991) Reinventing New York: competingin the next century’s global economy. In Urban ResearchCenter, New York.

Parnreiter, C (2002) Mexico City: The Making of a Global City?In Sassen, S (ed) Gobal Networks. pp. 145–182, Routledge,New York.

Parnwell, M and Wongsuphasawat, L (1997) Between the globaland the local: Extended metropolitanisation and industriallocation decision making in Thailand. Third World PlanningReview 19(2), 119–138.

Sassen, S (1996) Losing control?: sovereignty in an age of glo-balization. In Stanford University Press.

Sassen, S (2000) Cities in a World Economy (Second Edition). PineForge Press.

Sernau, S (1997) Economies of exclusion: economic change andthe global underclass. In At the Crossroads of Development.Transnational Challenges to Developed and DevelopingSocieties, (eds) J E Behar and A G Cuzan. E. J. Brill, Leiden,The Netherlands.

Townroe, P M and Keen, D (1984) Polarization reversal in the Stateof Sao Paulo, Brazil. In Differential Urbanization. IntegratingSpatial Model, (eds) H S Geyer and T M Kontuly. RegionalStudies, 18, pp 45–54. Arnold, UK.

US Office of Management and Budget (1998). AlternativeApproaches to Defining Metropolitan and NonmetropolitanAreas, Notice, Federal Register, No. 63 (December 21),70526–70561.

UNCHS (1996) An Urbanizing World. Global Report on HumanSettlements. In Oxford University Press, Oxford, UK.

Vogel, D (1993) Tokyo and London as financial centers. In Globalcity: the economic, political and cultural influences of NewYork, (ed.) M Shefter. Russell Sage Foundation, New York.


Ward, P M (1993) The Latin American inner city: differences ofdegree or of kind? Environment and Planning A 25, 1131–1160.

Ward, P M (1998) Mexico City (Revised second edition). JohnWiley and Sons.

Ward, P M (1999) Creating a metropolitan tier of government in

21

Federal Systems: getting ‘there’ from ‘here’ in Mexico City andin other Latin American Megacities. South Texas Law ReviewJournal 40, 603–623.

Yeboah, I E A (2000) Structural adjustment and emerging urbanform in Accra, Ghana. Africa Today 47(2), 61.

globalization, regional development, and mega-city ...ladupo.igg.unam.mx › ... ›...

Documents