GaWC Research Bulletin 54

This Research Bulletin has been published in WA Dunaway (ed) (2003) Emerging Issues in the 21st Century World-System Vol II New Theoretical Directions for the 21st Century World-System, Westport, CN: Praeger, 111-29.

Please refer to the published version when quoting the paper.

(Z)

Rediscovering Cities & Urbanization in the 21^st Century World-System

D.A. Smith^*

It is no exaggeration to say that we live in an "urban world." Today over three-quarters of the populations of the United States, Europe, Japan and other "core" nations live in cities (see United Nations 1996). Indeed, the great historic shift from the countryside to the city in these societies over the past two hundred or so years provided a major impetus for the development of the social sciences, in general, and sociology, in particular. Arguably, this epochal demographic shift (and the array of social, political and economic changes that accompanied it) provided the central context for the writings of "classical" European social theorists like Emile Durkheim, Max Weber and Karl Marx. Similarly immigration and rapid urbanization in late nineteenth and early twentieth century North America provided the impetus for the institutionalization of early US sociology epitomized by "the Chicago School" (e.g. Park and Burgess 1925; see Smith 1995 for a discussion).

Yet, in retrospect, the early growth of primarily "western" urbanization, however momentous in appearance and consequence it may have seemed to contemporary observers, is only part of the story - and, in terms, of sheer scale and volume, a relatively small portion at that. In the final half of the twentieth century, urbanization "went global." While only about three percent of the world's people lived in places that were classified as cities in 1900, by 2000 estimates suggest that close to half were urban dwellers (United Nations 1996). There are now over two billion people living in cities around the world. This means that urban migration in the past three or four decades is probably the most massive shift of human population ever. The vast majority of that very rapid city growth did not occur in the so-called "advanced" or "industrial" nations, but in the less developed or underdeveloped countries of "the south" (Roberts 1978). This is where the fastest growing "megacities" are located (Dogan and Kasarda 1988; Lo and Yeung 1998) - and these places are home to rapidly rising populations of some of the poorest people on earth who are facing a rapidly escalating array of social, economic, political and environmental crises. These giant cities in Africa, Latin America and Southeast Asia present pressing and possibly intractable policy dilemmas - if the contemporary world-system is on the brink of crisis (Wallerstein 1999a) it would not be surprising if some of the impetus for its destabilization bubbled up from these massive metropoli, where so many of global capitalism's starkest contradictions are most clearly visible.

While the sheer scale, striking unevenness, and harsh human consequences of urbanization in the underdeveloped countries demand our attention, there is another aspect of urbanization that may be equally salient in today's "globalizing" world. By the mid-1980s - long before the current buzz about "globalization" - it was becoming obvious that some urban nodes were playing increasingly critical roles in the world-economy and that networks of relationships between cities were intensifying and reconfiguring. Major changes in the organization of production and finance, which were various described at the twilight of "global Fordism" (Lipietz 1987), the emergence of "flexible production' (Scott 1988), or the rise of a "new Leviathan' (Ross and Tracte 1990), focused attention on "world cities" (Friedmann and Wolff 1982; Friedmann 1986) or "global cities" (King 1990; Sassen 1991). These places were seen as the crucial nodes in the rapidly multiplying and lengthening strands of the global economic webs of trade, travel, communication, and finance that defined the world-economy. The rise of "the global factory " (Fuentes and Ehrenreich 1983) in manufacturing was the beginning of a move toward more thoroughly globalized production of an array of goods and services, and more and more rapid flows of capital, people, commodities and information around the planet. The sheer velocity and complexity of these exchanges can be quite dizzying and challenge attempts to control it (Wade 1996; Sassen 1996). But there is an emerging consensus that "global cities' become prominent in these networks as "command and control centers" (Sassen 1991), where the headquarters of multinational corporations, giant banks, and new supranational economic institutions (trade organizations, "development banks,' etc) are located. These world cities become the increasingly dominant centers of progressively more integrated, hierarchical world city system (Knox and Taylor 1995; Smith and Timberlake 1995). Perhaps paradoxically, though, instead of leading to the general prosperity of its citizens, this particular urban dynamic seems to generate social polarization within the global city (Sassen 1991).

The dramatic transformations of and in cities, the epochal shift of world urbanization numbers toward the previously un- and under-developed countries, and the new appreciation of global city networks, all provided a very daunting challenge to conventional urban sociology. Elsewhere I have argued that older approaches, grounded in human ecology and modernization theory, was not up to the task (Smith 1995). Instead, these approaches were supplanted in a fundamental "paradigm shift" that involved putting cities in global perspective and prompted the rise of a new "urban political economy" approach (see Walton 1979 for a pioneering statement, or Smith and Timberlake 1997 for a more comprehensive overview). While this approach continues to be anchored in broad currents of urban sociology, much of the impetus for this new perspective on cities and urbanization came from scholars deeply engaged in the world-system perspective (for greater detail, see Smith 2000).

But something odd has happened in the past decade or so. Interest in the urbanization in the world-system, which had blossomed in the 1980s, waned in the 1990s. Despite the fact that the study of "global cities" remains vibrant, and there continues to be great attention on problems of "megacities" in the periphery, these topics do not seem to interest world-system analysts these days (and, perhaps not so coincidentally, much of the research now seems to be done by scholars who are not sociologists). Peter Taylor (this volume) correctly notes that "cities, in any major contemporary form - mega-cities, metropolitan centres, world cities, global cities - are conspicious by their absence from Wallerstein's essays and indeed other arguments about the future from a world-system perspective."

While I admire Taylor's innovative and speculative attempt to fundamentally "recast" world-system analysis by putting cities into a "starring role," my goal is this essay is much more modest. My point, really, is more akin to making sure that urbanization at least is part of the casting call and gets some sort of a role in our analyses, even if that is only in some sort of "supporting" capacity. Taylor's effort to re-center world-systems analysis onto city networks is iconoclastic and intriguing, and he presents a hopeful community-based vision of an alternative global future. Its interesting - and appealing! But we do not have to believe that urban communities will become the heroic vanguard of a coming global transformation to argue that world-system analysts need to pay more attention to cities. Much as I appreciate Taylor's analysis I am, in fact, less sanguine than him (and many others) that the current moment of "globalization" really means that nation states are destined for replacement (see Smith, Solinger, and Topik 1999). So my goal in this essay is less lofty and more mundane: In many poorer regions cities are currently the (relatively new) home to much of humanity (including a large portion that is not doing very well at all), at the same time that the leading urban centers of the core seem to be increasingly taking on new roles as "world cities." In both these case, peripheral and core, the underlying dynamic of urbanization (in stark contradistinction to old "modernizationist/developmentalist" myths) seems to be leading to more socioeconomic inequality and polarization. Therefore my simple plea is that we need to bring cities and urbanization back in to world-system studies.

In this essay, I would like to promote this "rediscovery" of cities and urbanization by first rather briefly reprising the short history of the "urbanization in the world-economy" perspective that seemed to take off in the 1980s (but then stalled out in the past decade or so). This will be done rather quickly, highlighting particularly pregnant aspects of this work and/or foreshadow some portents of the current fallow period (for a fuller discussion see Smith 2000). Then I will attempt to come to grips with why world-system oriented studies of cities and urbanization waned, with an eye toward how this might help us understand how to revitalize this area of scholarship. Part of the explanation (which also may account for less enthusiasm for the political economy of the world-system perspective more generically) is that this urbanization in the world-economy approach was a victim of its own success - today global perspectives on cities are so widely accepted that there is little novelty in this view. The need to understand urban dynamics in their world-economic context is not "news," and gets taken for granted. But I will also discuss some currents in both mainstream sociology and world-system analysis itself that worked to undermined this type of research and discouraged scholars from pursuing it. Finally, I will point to some potentially fruitful future directions for research and theorizing which could renew and reinvigorate this dormant field of study. One of these involves using the current, vibrant, scholarship on "global cities" and world urban hierarchies as a platform: in addition to "bringing cities back in" to world-system analysis, it is crucial to reintegrate this urban research to the core conceptual debates of political economy of world-system scholars (Peter Taylor's work clearly suggests a path, doubtlessly there are others). Another area that may be just as important - but tends to be grossly under-theorized and frequently ignored is the issue of peripheral urbanization, "megacities," and urbanization/underdevelopment dynamic. While these massive (and growing!) cities of the poor may have slipped beyond many sociologists scholarly gazes - and they doubtlessly are not as fashionable to study (or visit) as the leading global cities - they present a very stark reality, and a theoretical and practical challenge, that world-system analysts are uniquely equipped to confront.

A SHORT HISTORY OF A PROMISING PERSPECTIVE

The genealogy of a world-system approach to urbanization can be traced back at least to Andre Gunder Frank's provocative early writings in which he explains exploitation and underdevelopment via "a chain of constellations of metropoles and satellites (which) relates all parts of the whole system from its metropolitan center in Europe or in the United States to the farthest outpost in the Latin American countryside" (Frank 1969:6). David Harvey (1973:232) conjures up similar images: "Within countries functioning hierarchies of city types provide circulation channels of surplus value... (there is) a massive circulation of surplus value in which contemporary metropolitanism is embedded." By the mid- to late-1970s several comparative urban scholars working in different parts of the world were using the idea of "dependent urbanization" (McGee 1976; Castells 1977; Roberts 1978). In 1977 world-system oriented sociologist John Walton published a brief but seminal article that sketched out comparisons between Latin American and African urban systems and persuasively argued that the dynamics of city growth are shaped by the historical context of a region's initial incorporation into the world-system, as well as subsequent changes in its status and role in that global system. He later contributed a string of synthetic articles outline "urban political economy" (Walton 1979) or "the new urban sociology" (Walton 1981). In the early 1980s both Walton (1982) and Chase-Dunn (1984) published essays that systematically and explicitly laid out the "urbanization in the world-system" approach and challenged other scholars to pursue the new directions for research that it implied.

Michael Timberlake's 1985 edited volume, Urbanization in the World-Economy, provided the most comprehensive effort to develop the "urbanization in the world-economy" perspective, as well as a crucial benchmark for the potential of this approach. Timberlake lays out the basic premise of this perspective:

The chapters in his book reflected the range of methodologies, regional foci, and coverage of historical versus contemporary cities that characterized the best of this "urbanization in the world-system" approach during this period of its flowering. They ranged from a theoretical paper by Alejandro Portes on the role of the urban informal sector in the world-economy, a series of chapters about various national and regional dynamics, an historical image of the changing world urban hierarchy over the past 1,200 years by Christopher Chase-Dunn, a chapter on women and cities in the world-system by Kathryn Ward, an early essay on capital mobility and labor migration by Saskia Sassen that prefigured her later work on global cities, to various discussions of quantitative methodology and analysis, including one on core-periphery urban patterns by Glenn Firebaugh (whose later high-profile attacks against this very genera of cross-national world-system research will be discussed below). This was an impressive volume, in both the quality and range of analysis.

While the publication of the Timberlake (1985) volume may have represented a high-water mark for world-system analysis of urbanization, it was fairly representative of related research that was being published in the 1980s in books and refereed journals. A plethora of case studies appeared. Most were in books, with Armstrong and McGee's 1986 volume comparing Latin America and Asia, David Drakakis-Smith's 1986 collection on urbanization in "the developing world," and Feagin and Smith's 1987 edited collection on "global restructuring and community politics" (mainly in core cities), as particularly noteworthy examples. But, despite some resistance to this genera in mainstream sociology, some case studies even made it into major sociological journals (Feagin 1985; Smith 1987). A paper on colonialism and the urban hierarchy in Burma appeared in the Braudel Centers' Review (Saueressig-Schreuder 1987) at about this time, though it was very descriptive and data-oriented and would have benefitted from more engagement with the rich literature on urbanization in the world-economy. While these case-oriented studies illustrated the diversity of peripheral or dependent urbanization, they clearly showed that a global political economy approach, reflecting the basic orientation of the Timberlake quote above, is critical to understand cities and city growth in the capitalist world-system.

Contemporaneously, a series of cross-national comparative quantitative studies appeared in the major sociology journals and refereed outlets in development and urban studies. Several of these articles appeared in the high-profile American Sociological Review: one by Kentor (1981) on peripheral urbanization, another authored by Bradshaw (1987) on urbanization and underdevelopment, an article by London (1987) on "the structural determinants of Third World urban change, one authored by London and Smith (1988) on "urban bias, dependence, and economic stagnation in non-core nations," etc. An statistically sophisticated article detailing asymmetrical relationships iin "the world system of cities" by Meyer (1986) came out at about the same time in Social Forces. Although these attempts to "test" various propositions derived from the world-system perspective and competing approaches didn't always arrive at completely consistent results - and some of the articles, rather unsatisfyingly, point to "mixed results" with no clear theoretical "winner" (see a critique of this slide toward eclecticism in Smith 1991: 51) - the preponderance of the evidence supported the claim that dependent urbanization was very "uneven" and showed that world-economic structural variables (like trade dependence or foreign capital penetration) promoted a distinct urban dynamic in the periphery.

To this point my story has been one of success - in fact, a veritable "Kuhnian revolution" in which the urbanization in the world-economy perspective, as part of new urban political economy approach, supplanted the older ecological/modernizationist paradigm (Smith 1995). In an essay a decade ago I wrote, "By 1990 serious sociological studies of comparative urbanization which ignore the international political economy approach are extremely rare" (Smith 1991: 50). To be perfectly blunt: the new urban political economy succeeded on a number of fronts where the older urban sociology had failed: bringing in conceptual apparatus to deal with the global system, states and other political actors, capitalists and capitalism, historical contingencies, etc. On its merit the field was won. And in the halcyon years of the late 1980s it did seem that this "new urban sociology" was rolling along, with a number of researchers contributing to a rather vibrant and growing literature on comparative urbanization. Major books were published detailing rich historical-structural case studies of cities all around the globe; articles regularly appeared on peripheral urbanization and world city systems, even in the most staid mainstream sociology journals.

This period did not last. Two anthologies that emerged at about the same time (1990-1991) foreshadowed the loss of momentum. The first, based on a meeting of the International Geographic Union's working group on "Urbanization in Developing Countries," in 1986 was edited by geographer Drakakis-Smith and provisionally titled, The Urban Totem: Economic Development and Urbanization in the Periphery and Semiperiphery. By the time that the volume actually appeared in 1990, it was published as Economic Growth and Urbanization in Developing Areas. This rather banal (not to mention overly sanguine!) title seems to deliberate "distance" the book from "world-system theory." The second telling volume was the Political Economy of the World-System (PEWS) annual (Kasaba 1991), Cities in the World-System, based the previous year's conference. It seemed to be propitious time for this theme, building one the success of the urbanization in the world-system approach. But the conference was organized by a scholar expert in Ottoman history, the (arguably) most influential chapter was about Baltic sound tolls between 1550 and 1750 (Boswell, Misra and Brueggeman 1991), and the twelve of the fourteen authors brief biographies make no mention of any specialization in urban topics. The 1991 volume contains interesting essays, but the bold attempts at conceptual advance that characterized the Timberlake collection a few years early are missing, replaced by a theme of implicit "particularlism." In the introduction the editor envisions that the collection will allow readers to "attain a more accurate appreciation of the historical significance of specific cities" (Kasaba 1991: xii). This is an admirable goal - but quite a retreat from the programmatic visions of "new urban sociology."

By the early 1990s, articles in major sociology journals on dependent or peripheral cities, or even global cities and networks, no longer appeared. The quantitative cross-national research on comparative urbanization that had "broken through" into the major sociology journals seemed to be "played out." The most compelling case studies about "the urban South" moved away from holistic attempts to understand city growth and urban patterns in terms of global dynamics and toward more focused research on specific institutional changes (gender roles/inequality, industrialization, political dynamics) (for examples, see Tiano 1994; Walton and Seddon 1994; Davis 1994). While these studies typical emerged from concerns that the authors developed in dialogue with world-system analysis, they moved in different, specialized directions and, thus, had little cumulative impact on a shared perspective on cities in the world-system. Similarly, the rapidly growing scholarship on "global cities"/ "world cities" took its initial inspiration from a Wallersteinian world view (Friedmann and Wolff 1982) and the most comprehensive statement came from a sociologists directly tied to the original "urbanization in the world economy" framework (Sassen 1991). Yet, curiously, this fertile area also lay fallow in both sociology and world-system analysis, as geographers and urban planners became the main tillers of this field (Knox and Taylor 1995).

WHY DID INTEREST IN "URBANIZATION IN THE WORLD-SYSTEM" WANE?

This raises the logical question: what happened to this perspective that promised so much in the 1980s, but faded in the folowing decade? It is unlikely that there is a single answer, and there is little doubt that multiple forces were at work. I would like to suggest three broad factors:

The Pervasiveness of a Global Perspective

The first is important but may be the least interesting: today virtually everyone accepts the notion that various social dynamics must be placed in their global context. In a way this may be a case of the world-system perspective being too successful and convincing! At any rate, in this era of "globalization," the importance of a global perspective is widely, or perhaps universally accepted. For example, Kasarda and Crenshaw (1991), while vociferously disavowing the utility of "dependency/world systems" approaches to cities (see below), freely acknowledge that "clearly international relations and global flows of capital, technology and population do affect the process of urbanization" (492). The pervasiveness of this sort of view means that the basic premise that cities must be interpreted in terms of worldwide forces and contexts no longer seems novel and innovative (a crucial consideration for scholars who want to ride a "cutting edge" wave). Today virtually all serious students of cities (even those who tenaciously cling to the old "modernization/ecology" framework) accept the notion that these places are enmeshed in various global networks. This may seem like a significant theoretical advance - and it does represent great progress over dated "modernization" images of exclusively national urban systems that unfold according to their own "internal" dynamics. But the current buzzword of "globalization" itself means so many different things to different people, and is so loosely conceptualized, that this new "global perspective" falls far short of the coherent conceptual framework of the "urbanization in the world-economy" view.

The other pressures that worked against the urbanization in the world-system approach were both more specific and part of broader intellectual currents within our discipline and subfield. One was centered in the institutions and debates of "mainstream sociology"; the other reflected recent "turns" in world-system thinking.

Marginalization in Mainstream Sociology

The first of these pressures was a critical reaction from mainstream sociology. World-system analysis (of cities or any other substantive topics) has always existed rather uneasily within US sociology (cf. the detailed discussions in Wallerstein 1991). Scholars like to think of academia in general, and sociology in particular, as very "open" to divergent views and theoretical perspectives. But as both an author and a reviewer for major US-based sociology journals, I have seen far too many written evaluations of manuscripts recommend rejection based on vague notions that the world-systemic views are "too Marxist" or "politically biased" (the assumption here seems to be that conservative approaches that don't challenge the status quo are "apolitical"). Although wise editors know how to interpret (or discount) this sort of review, the existence of this type of mentality among journal (and grant proposal) referees, presents a serious obstacle to publication (or funding) for world-system scholars attempting to negotiate careers in academic sociology.

In a context where some sociologists were never convinced of the legitimacy of any of this work, a particularly influential critique of the quantitative cross-national world-system studies appeared in the early 1990s. In an article in the American Sociological Review (1992), Firebaugh claimed that most of the prominent in this genera were fatally flawed based on a misinterpretation of the meaning of a key variable ("foreign capital penetration"). A technical discussion of this controversy is not warranted here (for that, see Crowly, Rauch, Seagrave and Smith 1998: 38-39). In brief, Firebaugh used his considerable statistical acumen to show that by replacing measures of "stocks" versus "flows" of external capital with an overall rate of foreign investment, it positively correlates with overall economic growth (Firebaugh 1992). While this article was narrowly based on a mathematical argument about "denominator effects" of the measure, Firebaugh (1992) made sweeping claims about the "illogic" of the world-system oriented quantitative studies. His paper was widely read and cited - and many sociologists believed that it dealt a decisive blow and debunked all quantitative cross-national world-system studies of the 1980s (the editor of one of the "big three" sociology journal told me, candidly, that he doubted that he would publish any more of this genera, in light of the Firebaugh critique). In fact, an illuminating exchange in the American Journal of Sociology between Firebaugh and his critics in 1996 decisively demolished most of his critique. Boswell and Dixon (1996) highlighted both technical and logical holes in the Firebaugh argument (the most basic: that he seemed to simply not understand that the key conceptual issue was to show that foreign investment is "less productive" than local capital and led to various "negative externalities" ). So, in the end, the "Firebaugh critique" turned out to be much less substantial that it originally seemed. However, by the time the 1996 rebuttal appeared in print, a great deal of damage had already been done - an entire genera of research was delegitimated.

While the methodological controversy cast aspersions on some of the most prominent cross-national studies of urbanization in the world-economy, there were also more specific arguments raging within comparative urban sociology in the 1990s. The paradigm shift I described above elicited various reactions from established urban sociologists. While some distinguished urban ecologists tried to understand and enter into a dialogue with the new urban political economy approach (see Hawley 1984), some rose up to vehemently defend the old orthodoxy, and still others tried to ignore the theoretical shift entirely.

An essay by Kasarda and Crenshaw (1991) in the Annual Review of Sociology exemplifies the defensive mode, proclaiming that the modernization/ecology perspective remained dominant in the study of "third world urbanization," and counterattacking the urbanization in the world-system approach as ideological and muddled. Despite a "matter-of-fact" tone and a stated goal of reviewing and evaluating recent debates in this area, the authors proceeded to blatantly distorted (or, charitably, totally misunderstood!) world-system and other "critical" approaches to urbanization and development. They claimed "to have trouble understanding the phrase 'uneven development'" (485), suggested that there is no reason for urbanists to consider "class relations" in their analyses, scoffed at the notion of a hierarchy of "world cities," and expressed particular nervousness about the use of notions of "exploitation" and "capitalism" (particular if it is suggested that the two be conjoined!). Showcased in an annual volume that purports to present "state-of-the-art" summaries, this essay provided a very twisted interpretation of the urbanization in the world-economy scholarship. While some sort of constructive critique of the urbanization in the world-system literature was warranted at the time (and might have opened new channels for dialogue), the Kasarda and Crenshaw essay was totally dismissive. It presented a jaundiced view of view of the global political economy approach, misleading those who were unfamiliar with comparative urban research, while providing further justification to unsympathetic mainstream scholars to ignore this work.

By the mid- to late-1990s, the mode of marginalization for the world-system approach to cities took a rather different form. At this point, the theoretical excitement about urban political economy's insights had faded and a conceptual malaise had descended on comparative studies of cities and development. Increasingly, literature reflected a full-fledged retreat into a rather intellectually barren form of eclecticism. Exemplifying this trend are two anthologies edited by Josef Gugler (1996, 1997). Interestingly, Gugler helped to undermine the old modernization/ ecology paradigm with his studies stressing the need to place West African city growth in an international political economy context (Gugler and Flanagan 1977). But two decades later, as the editor of both these collections (optimistically referring to urbanization in "the developing world"), Gugler praises the "comprehensive review" of Kasarda and Crenshaw (1991) and advocates using multiple theoretical approaches while "privileging" none, with a heavy focus on distinct "regional trajectories." The various chapters tend to be rather descriptive (but useful and, at times, insightful) - and some of the original essays on regional patterns (for example, Abu-Lughod 1996; and Oliveira and Roberts 1996) are fine contributions, very much in the tradition of the urbanization in the world-economy perspective. But the attempt to cover "the entire range of social science approaches to urbanization," results in a theoretical melange, with no underlying or unifying conceptual framework. Readers find lots of data on trends and patterns but little in the way of any comprehensive or cumulative argument about "the urban transformation of the developing world." It's hard to argue that the road from Timberlake (1985) to Gugler (1996, 1997) represents intellectual progress.

Urban Neglect in World-System Analysis?

It is much more difficult to discern any intentional effort to marginalize urban studies among PEWS-oriented scholars. But Taylor's observation about the relative neglect of all things urban in recent world-system writings and research does ring true. It's particularly perplexing that the burgeoning literature on "global cities" in recent years seems to have moved away from dialogue with the theoretical framework that helped to launch its pioneering studies.

In part, this may simply reflect the personal intellectual trajectories of the leading scholars: some of those most identified with "world-system analysis" (Arrighi, Wallerstein) focused on larger world-historical structures than cities, others who studied world urban systems moved toward other substantive foci (Chase-Dunn), and some major comparative urbanists who were key contributors to the PEWS tradition later were pulled away by other interests and commitments (Portes, Sassen). The selection of a non-urbanist to organize that PEWS-sponsored "Cities in the World-System" conference at a critical time in the early 1990s might be the same sort of "historical accident." Perhaps a specifically "urban" focus is less attractive to world-system scholars these days because they accept the basic assumption that cities, like other social structures and processes, must be studied "holistically" (see Timberlake quote, above), and this makes it much less easy to delineate a clear object of "urban" object of study (for the classic formulation of this claim, see Castells 1977). This might suggest that more complex and richly contextualized case studies that involve cities - but understand them in their larger socio-historical contexts - are more valuable.

But I also think that a (mis)interpretation of the "epistemological turn" in Wallerstein's recent work may have contributed to the lost momentum for the urbanization in the world-economy approach (and maybe to declining influence of PEWS-oriented research on some other topics, as well). It is abundantly clear that "mainstream sociology" has not been tolerant of world-systemic (and other critical) perspectives; and Wallerstein (1991, 1997) and the Gulbenkian Commission (Wallerstein, et. al. 1996) have elegantly and persuasively argued for fundamental reforms to "open the social sciences." They envision the dissolution of disciplinary boundaries, the end of various "antinomies" that divide the "two cultures" of "the humanities" versus "the sciences," etc. This is visionary stuff - most of us would fully support the rise of a new multidisciplinary world-historical "interscience" (Wallerstein 1999b). However, I think that this "turn" is open to misunderstanding. Some folks working within the world-system tradition seem to believe that the message is to reject "nomothetic" approaches entirely, in favor of long-term qualitative historical analysis. Arguably some of the most important - and certainly the most visible -- world-system research in the past two decades were statistical studies (on urbanization, development and some related topics) that appeared in the major sociology journals. Virtually all of this work was open to various critiques - but to reject it out of hand because it is "too positivistic" to "fit" into a world-historical view of the world is short-sighted. It also distorts what I see at the basic Wallersteinian/Braudelian vision. In an interesting recent essay on "interscience" Wallerstein (1999b: 2) points out that Braudel view actually "calls for mathematization" as well as focus on "local specificity" and "the longue duree." In particular, in a 1969 essay Braudel notes the diversity of the human sciences, praises an (very positivistic) icon of US survey research (Paul Lazarfeld), and says, "all these new methods are within arm's reach... Let us not fail to take advantage of this doubling or tripling of the odds" (Braudel 1969: 96 quoted/translated by Wallerstein 1999b:2). In his essay, Wallerstein concludes that it is important to try to arrive at reformulated "common epistemology" for the "interscience" he and Braudel advocate.

It is true - and unfortunate - that many of our sociological colleagues have unfairly judged world-system research; we need to guard against a equally narrow parochialism in regard to developments within quantitative sociology. To revitalize the urbanization in the world-economy paradigm fully, it is important to "open world-system analysis" to a variety of methodological approaches that allow us to fully study urban dwellers, cities, and urban networks, today and in long-term historical perspective.

URBANIZATION IN THE WORLD-SYSTEM: AN AGENDA FOR THE 21^ST CENTURY

How can be "bring cities back in" to world-system analysis and reinvigorate "urbanization in the world-economy" as a critical research framework in urban sociology? These final pages offer some suggestions for renewal. It makes sense to begin with the most successful scion of the 1980s approach that continues to blossom today: the "world cities/global urban hierarchies" research. But, while it emerged as a result of intellectual cross-fertilization between urban studies and world-system analysis, today studies of global cities tend to be carried out in relative isolation from (and far off the cognitive "radar screen") of most PEWS-oriented scholars. Closer dialogue between world-system sociologists and this vibrant and developing body of research would be mutually beneficial (and essays like that of Taylor, this volume, contribute to this effort). Here, I will sketch out some large analytical questions about world cities in a world-system that demand further attention, could spur new research, and could lead to new theoretical understandings. But I also intend to "push" in a different direction. While the notion of a "global city system" may imply a return to a focus on those "chain(s) of metropoles and satellites" reaching deep into the peripheral regions of the world-economy that A.G. Frank described many years ago, in fact, the swelling cities of the poor in "the South" tend to be both under-emphasized and under-theorized. For a variety of reasons, the focus of much of the "global cities" research is on the upper echelon cities near the apex of the urban hierarchy: those dominant command centers; attention falls off as we move toward urban areas near the bottom.

The unfortunate result is that the massive changes and growing problems of these "megacities" are often addressed in descriptive studies based on theoretically eclectic, if not conceptual muddled, theoretical frames (see the discussion of Gugler 1996 and 1997, above). The careful attention on the urbanization/underdevelopment dynamic that was stressed by the urbanization in the world-economy's notion of "dependent urbanization" is missing. The challenge to revitalize this neglected intellectual space is particular pressing.

World Cities in a World-System

The new focus of interdisciplinary urban scholars on "global cities" in the 1980s was an example of the influence of the political economy of the world-system view across the social sciences: the progenitors of the approach explicitly acknowledged the influence of the Wallersteinian image of global political economy (cf. Friedmann and Wolff 1982; King 1990); the incipient ideas of Sassen's "global cities" framework initially appeared in Timberlake's 1985 volume. In subsequent years, the provocative "hypotheses" these scholars laid out (see particularly Friedmann 1986) gave rise to a plethora of distinct strands of research. While they work on an array of topics, these scholars all begin with the premise that we can look at the relationship of cities in global networks, these urban systems of connections lead to the emergence of hierarchy and dominance, and that the leading "world cities" are the places "through which regional, national, and international economies are articulated with the global capitalist system of accumulation (Friedmann 1995:25-26)." This is a dynamic system: urban centers can rise and fall, they can "articulate" with other cities in particular ways, and the role that places play in these networks will have a great impact on urban social, economic and political life within these nodes. The great metropolises at the very pinnacle of the hierarchy play particularly important roles. In her ambitious treatise, The Global City: New York, Tokyo and London, Saskia Sassen (1991) explains how these places serve as "command and control centers" for today's global economy, coordinating increasingly geographically far-flung business empires of finance, manufacturing, retailing, etc. These "command posts" hold great global power, are home to specialized business service providers, centers for innovation, and crucial markets. But Sassen and others (Friedmann and Wolff 1982; King 1990; Ross and Trachte 1990) also argue that these same global cities are also sites for growing social and economy polarization. We're left with an intriguing paradox: these "command posts" for the world economy, home to the earth's wealthiest and most powerful people, are also places with growing low-wage labor forces, isolated and politically marginalized immigrants, and dispossessed racial and ethnic minorities.

This is a powerful vision, which (with some urgency!) raises a series of intriguing research questions. Three stand out. The first can prosaically be labeled a "measurement" issue. Is there a global city hierarchy that can be clearly measured and "mapped" with empirical data? While the early conceptions of the global city hierarchy (ala Friedmann 1986) had great "face validity," the networks and rankings that resulted were theoretical "guesstimates" made by extremely well-informed urban scholars. Moving to a more rigorous image of the morphology of the global urban hierarchy has proven to be nettlesome. Meyer (1986) made an early attempt, examining international banking headquarters in various cities in Latin American, North America and Europe: Sassen (1991) provides volumes of both national- and city-level data to make the case for the dominance of New York, Tokyo and London. Admirably as these efforts were, they still failed to directly tap the relationalaspect of global cities. My own recent work with Timberlake explicitly addresses this problem by using formal network analysis to examine inter-urban linkages. There is a paucity of data on city-to-city flows and connections, so we have used information on international air travel (which is available from airport-to-airport) (Smith and Timberlake 1995, 2001). We derive a hierarchical network of urban centers (with London, Paris, Frankfurt, and New York at the top) that makes theoretical sense. But we are painfully aware of the limitations that our data impose (we doubt, for example, that Frankfurt would rank as high on most other measures). Concurrently, the "Globalization and World Cities Study Group and Network" at Loughborough University developed alternative approaches to the measurement problem: They also note the paucity of relational data and analysis. There approach eschews formal network analysis of existing data in favor of developing tailor-made measures of information they have collected on city-to-city ties based on 1) a content analysis of business news, 2) survey data on the movement of skilled labor within firms, and 3) the concentration of "producer services" in cities (which is analogous to the data Meyer (1986) utilized). The result is somewhat piecemeal -but nevertheless very interesting - patterns with clear New York/London dominance and rich linkages between these places and other global cities (Beaverstock, Smith, Taylor, Walker and Lorimer 2000). Even more recently Taylor, Catalano and Walker collected data on producer services for 100 firms and 316 cities and calculated a measure of "global connectivity" (this generated the cartogram in Taylor, this volume). The Loughborough studies and our own research represent progress. But clearly identifying good network data -- or collecting it -- are an urgent priority. More help is needed!

The second analytical issue - and one that really requires deep engagement with the PEWS theoretical tradition - is, How does the world city system articulate with the wider world-economy. The title of this section (which mirrors Knox and Taylor's 1995 volume) implies that there is an articulation or "nesting" between the global urban network and the capitalist world-economy. These two hierarchic systems clearly overlap and mutually reinforce one another (the global cities are the "command and control" centers for the far-flung transnational economic enterprises with "global reach"). Over twenty years ago, Walton (1977) suggested that the world city hierarchy is "nested" into the broader world system; today Sassen (forthcoming) claims that global cities are key "pivots" on "global circuits." Contemporary global restructuring may bring about changing relationships between cities, regions, states and the global system. Taylor's provocative argument in this volume that cities should be cast in "a starring role" in today's world-system is one logical extreme. More likely it makes more sense to see envision a more dynamic and complex world-system in the twenty-first century, in which nation states, corporate actors, and transnational classes remain key players, but the importance of key global cities fulfilling their "command post" functions becomes hard to ignore, and global urban networks become more obvious in defining the contours of the world-economy. This is an area that demands empirical investigation. Using network imagery - if not the formal statistical methodology - should help us unravel the nature of the articulation between the global urban network and the modern world-system.

The final issue swirling around "the world city hypothesis" involves the notion the global city status has ramification for social dynamics within these urban areas. The literature on world cities suggests that these most dominant places are also centers of some of the global economy's most severe and in tractable contradictions. The great metropolises at the lofty heights of the urban dominance structure are home, simultaneously to enormous wealth and power polarization, marginalization and poverty. Various case studies provide qualitative and quantitative evidence of growing inequality in world cities (cf. King 1990; Sassen 1991: Ch. 8 & 9), and Sassen attempts to explain how the concentration of upper-income business professionals in global cities creates pressure for "gentrification" forcing poor people from affordable housing, while simultaneously generating demand for low-wage, sometimes "informalized" service jobs that attend to the needs of these wealthy residents. While the link between global cities and polarization seems to be borne out by some data - and the fact that contemporary cities like New York, London, or Tokyo empirically "fit" effectively debunks that fallacious image associated with urban boosterism that moving to world city status is an unalloyed benefit to all city residents. However, recent evidence suggests that socio-economic polarization is increasing world-wide promoted by policies promoted by "global neo-liberalism." This raises some important questions: Is the increased inequality reported in various urban case studies really intrinsically linked to global cities? What particular mechanism might be at work in these places that do not operate elsewhere? Though I find the global city-polarization thesis intriguing - among other things it suggests that these places might be venues that could become political powerkegs (a contrast to Taylor's image of "cities as actual communities"). But there is reason for skepticism - and yet another priority for theoretically informed research.

"Dependent Cities"? Urbanization and Underdevelopment

If we take the notion of a global urban network seriously, that web reaches out to virtually all points on the planet today. While they are relatively low on the world hierarchy the giant lead cities of most of the earth's poor countries are linked into this system in structurally peripheral positions, at the fringes of these networks. These urban conurbations are also "peripheral" in a world-system sense (cf. Walton 1982) and they are examples of "dependent urbanization" (Castells 1977: Ch. 3).

These concepts, of course, were central to the 1980s version of "the urbanization in the world-economy," but are little used now. It is clearly true that an oversimplified "theory:" of "dependent urbanization," positing bifurcating paths of core and peripheral city growth, or even one that used a Wallersteinian tripartite division (arguing for distinctive "semiperipheral urbanization" patterns) was bound to quickly run into empirical problems. There really are many variations on the general urban themes across the face of the planet - and they do reflect local historical contexts and regional trajectories, as well as the structural positions that cities and their hinterlands occupy in global hierarchies. No simple categorization will "explain" this diversity - the type of synthetic urban theory building that a new approach to cities in the world-system should attempt would be much more complicated (Smith 1995). The idea (to return to a quote from Timberlake 1995, above) is to use the idea of "dependency" or "peripherality" as an conceptual entry point for global historical studies of cities and as a tool to "pry into" the political economy of urbanization in varied circumstances. This was a productive theme of many of the case studies of "Third World urbanization" of the 1980s.

As we enter the new century, cities in the poor and underdeveloped regions of the world are growing just as rapidly and are sites of immense inequality, poverty and deprivation, and human suffering. But comparative urban studies seems to be at an impasse: the favored terminology to describe these places is "megacity" (Dogan and Kasarda 1988). But this is truly theoretically empty verbiage: if it implies anything beyond a description of the great scale of these urban areas, it can only be conceptually confounding, by implying similarities based on sheer size. Alternative approaches that emphasize distinctive regional patterns of urbanization (cf. Gugler 1996) are better at capturing historical nuances - but this also ends up offering little overarching theoretical direction.

Instead, I think we need to revitalize a global political economy approach that can directly address the urbanization/underdevelopment dynamic that is so pervasive in "the South" today. An interesting - if somewhat flawed - attempt to do this in a particular region is found in a collection of recent essays on "the urban Caribbean" (Portes, Dore-Cabral, and Landolt 1997). The authors attempt to collect comparative data on demographic patterns, economic changes, urban inequality and poverty, residential segregation, political participation, etc. While the resulting case studies are empirically uneven, they are motivated by common concerns about the how the interaction between global dynamics and historically condition local and national conditions effect urban outcomes. This combines an appreciation of national/regional distinctiveness in the face of increasingly strong integrating and subordinating pressure from "the new global economy" (which may weigh particularly heavily on this archipelago in the shadow of the United States).

The larger point is theoretically informed research on cities and urban systems in "the South" (or "the periphery") still holds great promise. These studies must begin by situating cities and urban systems in the hierarchic global system (of world cities and world-system), understand the underlying capitalist dynamic of the process, appreciate that national and regional contextual variations will come into play, and focus on the essential political nature of the processes and role of state and corporate actors. Furthermore, this research should be multimethodological, and it must be attuned to the changing rhythms of capital accumulation and how that will change the nature of urban development. This sounds like the kind of "historical social science" or "interscience" that Immanuel Wallerstein and Fernand Braudel call for - and its time for world-system analysts to get to work carrying out these studies of cities and urban systems.

REFERENCES

Abu-Lughod, Janet. 1996. Urbanization in the Arab World and the International System. In The Urban Transformation of the Developing World, ed. Josef Gugler, 185-208, New York: Oxford University.

Armstrong, Warwick and Terence McGee. 1986 Theatres of Accumulation: Stuides in Asian and Latin American Urbanization. New York: Methuen.

Beaverstock, J.V., R.G. Smith, Peter Taylor, David Walker, and H. Lorimer. 2000. Globalization and World Cities: Some Measurement Methodologies. Applied Geography 20(1)43-63.

Boswell, Terry and William Dixon. 1996. Dependency, Disarticulation, and Denominator Effects: Another Look at Foreign Capital Penetration. American Journal of Sociology 102(2)543-562.

Boswell, Terry, Joya Misra, and John Brueggeman. 1991. The Rise and Fall of Amsterdam and Dutch Hegemony: Evidence from the Baltic Sound Tolls, 1550-1750. In Cities in the World-System, ed Resat Kasaba, 55-80.

Bradshaw, York. 1987. Urbanization and Underdevelopment: A Global Study of Modernization, Urban Bias, and Economic Dependency. American Sociological Review 52(2)224-239.

Braudel, Fernand. 1969 Unite et deversite des seicence de l'homme. In Ecrits sure l'histoire, 85-96. Paris: Falmmarion.

Castells, Manuel. 1977. The Urban Question: A Marxist Approach. Cambridge, MA: MIT Press.

Chase-Dunn, Christopher. Urbanization in the World-System: New Directions for Research. In Cities in Transformation, ed. Michael P. Smith, 111-120. Beverly Hills: Sage.

David, Diane. 1994. Urban Leviathan: Mexico City in the Twentieth Century. Philadelphia: Temple University Press.

Dogan, Matthei and John Kasarda. 1988. The Metropolis Era: A World of Giant Cities & Mega-Cities (two volumes). Newbury Park, CA: Sage.

Drakakis-Smith, David (ed). 1987. Urbanisation in the Developing World. London: Croom Helm.

Drakakis-Smith, David. 1990. Economic Growth and Urbanization in Developing Areas. London: Routledge.

Feagin, Joe. 1985. The Global Context of Metropolitan Growth: Houston and the Oil Industry. American Journal of Sociology 90(6)1204-1230.

Feagin, Joe and Michael P. Smith (eds). 1987. The Capitalist City. London: Basil Blackwell.

Friedmann, John 1986. "The world city hypothesis." Development and Change 4:12-50.

Friedmann, John 1995. "Where we stand: a decade of world city research," pp. 21-47 in Paul Knox and Peter Taylor (eds.), World Cities in a World-System, New York: Cambridge University Press.

Friedmann, John and Goetz Wolff. 1982. "World city formation: an agenda for research and action," International Journal of Urban and Regional Research 6:309-344.

Feuntes, Annette and Barbara Ehrenreich. 1984. Women in the Global Factory. Boston, MA: South End Press.

Firebaugh, Glenn. 1992. Growth Effects of Foreign and Domestic Investment, American Journal of Sociology 98(1)105-130.

Frank, Andre Gunder. 1969. Latin America: Underdevelopment or Revolution? New York: Monthly Review Press.

Gugler, Josef and William Flanagan. 1977. On the Political Economy of Urbanization in the Third World: The Case of Africa. International Journal of Urban and Regional Research 1(2)272-292.

Gugler, Josef, ed. 1996. The Urban Transformation of the Developing World. New York: Oxford University Press.

Gugler, Josef, ed. 1997. Cities in the Developing World: Issues, Theory, and Policy. New York: Oxford University Press.

Harvey, David. 1973. Social Justice and the City. Baltimore: Johns Hopkins University Press.

Hawley, Amos. 1984. Human Ecological and Marxian Theories. American Journal of Sociology 89(4)904-917.

Kasaba, Resat (ed). 1991. Cities in the World-System. Westport, CT: Greenwood.

Kasarda, John and Edward Crenshaw. 1991. Third World Urbanization: Dimensions, Theories, Determinants. Annual Review of Sociology 17:467-501.

Kentor, Jeffrey. 1981. Structural Determinants of Peripheral Urbanization: The Effects of International Dependence. American Sociological Review 46(2)201-211.

King, Anthony. 1990. Global Cities: Post-Imperialism and the Internationalization of London. London: Routledge.

Knox, Paul and Peter Taylor (eds). 1995. World Cities in a World-System, New York: Cambridge University Press.

Lipietz, Alain. 1987. Mirages and Miracles: The Crisis of Global Fordism. London: Verso.

Lo, Fu-chen and Yue-man Yeung. 1998. Introduction. In Globalization of the World of Large Cities, eds Fu-chen Lo and Yue-man Yeung, 1-13. Tokyo: United Nations University Press.

London, Bruce. 1987. The Structural Determinants of Third World Urban Change"An Ecological and Political Economic Analysis. American Sociological Review 52(1)28-43.

London, Bruce and David A. Smith. 1988. Urban Bias, Dependence, and Economic Stagnation in Non-Core Nations. American Sociological Review 53 (3)454-463.

McGee, Terence G. 1996. The Persistence of the Proto-Proletariat: Occupational Structures and Planning the Future of Third World Cities. Progress in Geography 9:3-38.

Meyer, David R. 1986. World System of Cities: Relations between International Financial Metropolises and South American Cities. Social Forces 64(3):553-581.

Portes, Alejandro, Carlos Dore-Cabral and Patricia Landolt., eds. 1997. The Urban Caribbean: Transition to the New Global Economy. Baltimore: Johns Hopkins University Press.

Oliveira, Orlandina and Bryan Roberts. 1996. Urban Development and Social Inequality in Latin America. In The Urban Transformation of the Developing World, ed. Josef Gugler, 253-314, New York: Oxford University

Roberts, Bryan. 1978. Cities of Peasants: The Political Economy of Urbanization in the Third World. Beverly Hills: Sage.

Ross, Robert and Kent Trachte. 1990. Global Capitalism: The New Leviathan. Albany: State University of New York Press.

Sassen, Saskia. 1991. The Global City: New York, London, Tokyo. Princeton University Press.

Sassen, Saskia. 1996. Losing Control? Sovereignty in an Age of Globalization. New York: Columbia University Press.

Saueressig-Schreuder, Yda. The Impact of British Colonial Rule on the Urban Hierarchy of Burma. Review 10:245-277.

Scott, Allan. 1988. "Flexible Production Systems and Regional Development." International Journal of Urban and Regional Research 12: 171-186.

Smith, David A. 1987. "Dependent Urbanization in Colonial America: The Case of Charleston, South Carolina. Social Forces 66(1)1-28.

Smith, David A. 1991. Method and Theory in Comparative Urban Studies. International Journal of Comparative Sociology 32(1-2)39-58.

Smith, David A 1995. "The New Urban Sociology Meets the Old: Rereading Some Classical Human Ecology." Urban Affairs Review 30(3)431-456.

Smith, David A. 1996. Third World Cities in Global Perspective: The Political Economy of Urbanization. Boulder, CO: Westview.

Smith, David A. 2000. Urbanization in the World-Economy: A Retrospective & Prospective. In A World-System Reader, ed. Thomas Hall, 143-168. Boulder, CO: Rowman and Littlefield

Smith, David A. and Michael Timberlake.1995. Conceptualising and Mapping the Structure of the World System's City System. Urban Studies 32: 287-302.

Smith, David A. and Michael Timberlake. 1997. Urban Political Economy. In The Urban World, J. John Palen, Chapter 5, 109-128. New York: McGraw-Hill.

Smith, David A. and Michael Timberlake. 2001. World City Networks and Hierarchies, 1977-1997: An Empirical Analysis of Global Air Travel Links. American Behavioral Scientist 44(10)1656-1678.

Smith, David A., Dorothy Solinger and Steven Topik (eds). 1999. States and Sovereignty in the Global Economy. London, UK: Routledge.

Taylor, Peter. Forthcoming. Recasting World-Systems Analysis: City Networks for Nation-States. This volume! Westport, CT: Greenwood.

Taylor, Peter, G. Catalano, and David Walker. 2001. Measurement of the World City Network. Unpublished paper of the "Globalization and World Cities Studies Group and Network" available at http://www.lboro.ac.uk/gawc/rb/rb43.html.

Tiano, Susan. 1994. Patriarchy on the Line: Labor, Gender, and Ideology in the Mexican Maquila Industry. Philadelphia: Temple University Press.

Tilly, Charles. 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russell Sage Foundation.

Timberlake, Michael (ed). 1985. Urbanization in the World-Economy. New York: Academic Press.

United Nations. 1995. 1993 Demographic Yearbook. New York: United Nations.

Wade, Robert. 1996 Globalization and its Limits: Reports of the Death of the National Economy are Greatly Exaggerated. In National Diversity and Global Capitalism, ed. Suzanne Berger and Robert Dore. Ithaca, NY: Cornell University Press.

Wallerstein, Immanuel. 1991. Unthinking Social Science. Cambridge, UK: Polity Press.

Wallerstein, Immanuel, et. al. 1996. Open the Social Sciences. Stanford, CA: Stanford University Press.

Wallerstein, Immanuel. 1997a. Social Science and the Quest for a Just Society. American Journal of Sociology 102(5)1241-1257.

Wallerstein, Immanuel. 1999a. The End of the World as We Know It: Social Science for the Twenty-First Century. Minneapolis: University of Minnesota Press.

Wallerstein, Immanuel. 1999b. Braudel and Interscience: A Preacher to Empty Pews? Unpublished paper available at http://fbc.binghamton.edu/iwjb.htm.

Walton, John. 1977. Accumulation and Comparative Urban Systems: Theory and Some Tentative Contrasts of Latin America and Africa. Comparative Urban Research 5(1)5-18.

Walton, John. 1979. Urban Political Economy: A New Paradigm. Comparative Urban Research 7(1)5-17.

Walton, John. 1981. The New Urban Sociology. International Social Science Journal 33(2)374-390.

Walton, John. 1982. The International Economy and Peripheral Urbanization. In Urban Policy Under Capitalism, eds Norman and Susan Fainstein, 119-135. Newbury Park, CA: Sage.

Walton, John and David Seldon. 1994. Free Markets and Food Riots. Oxford, UK: Blackwell.

NOTES

^* David A. Smith, Department of Sociology, University of California, Irvine. An early version of this paper was originally presented at the 25^th annual Political Economy of the World-System (PEWS XXV) Conference, Virginia Polytechnic Institute and State University, Blacksburg, VA, April 2001.

Edited and posted on the web on 30^th August 2001

Note: This Research Bulletin has been published in WA Dunaway (ed) (2003) Emerging Issues in the 21st Century World-System Vol II New Theoretical Directions for the 21st Century World-System, Westport, CN: Praeger, 111-29