Spatial and Other “Fixes” of Historical Capitalism1


Giovanni Arrighi



Capitalism is the first and only historical social system that has become truly global in scale and scope. Mapping this transformation over time is a particularly challenging task. I have no disagreement with Christopher Chase Dunn’s and Thomas Hall’s contention that the world capitalist system–like other world-systems–can be described by means of four kinds of social interaction networks, each operating at a different spatial scale: bulk goods networks at the smallest scale, prestige goods and information networks at the largest scale, and political-military networks at an intermediate scale (1997: 52-55). This is a useful and illuminating proposition, and there can be little doubt that a mapping of these networks over time for the world capitalist system would provide compelling evidence of its peculiar expansionary character in comparison with all other world systems.

            Granted this, the resulting spatial-temporal map would provide little information concerning the inner dynamic of historical capitalism. It may even obscure the processes that have been associated with its globalization over the last half millennium. Worse still, this globalization has occurred through a tremendous increase in the number and variety of each kind of network, as well as an increase in the scale of bulk goods and military-political networks relative to prestige goods and information networks. Without some theoretical guidance in the selection of the networks to be mapped, there is a real risk of producing maps that are so confusing as to be worthless.

            The purpose of this paper is to propose a conceptual map focused specifically on the processes associated with the globalization of historical capitalism. This is not the kind of map that this conference advocates. It is nonetheless a first necessary step in the identification of the kind of geographic and historical information that is needed in order to represent graphically the spatial-temporal dynamic of historical capitalism. I shall begin with a brief discussion of David Harvey’s (2003) concepts of “spatial-temporal fix,” “switching crisis,” and “accumulation by dispossession.” I will then show that these concepts find a close correspondence in the evolutionary pattern of world capitalism identified in The Long Twentieth Century (Arrighi 1994) and developed further in Chaos and Governance in the Modern World System (Arrighi and Silver 1999). I will conclude by pointing to the kind of geographic and historical information that would be most useful to represent graphically this evolutionary pattern and the resulting globalization of historical capitalism.


I. Spatial Fixes, Switching Crises, and Accumulation by Dispossession

In seeking a connection between processes of capital accumulation and expansionist political-military projects–such as the Project for the New American Century that has inspired the US War on Terrorism and the invasion of Iraq–Harvey has deployed a complex conceptual apparatus, the center-piece of which is the notion of spatio-temporal fix. In his argument, the term “fix” has a double meaning.

A certain portion of the total capital is literally fixed in and on the land in some physical form for a relatively long period of time (depending on its economic and physical lifetime). Some social expenditures (such as public education or a health-care system) also become territorialized and rendered geographically immobile through state commitments. The spatio-temporal ‘fix’, on the other hand, is a metaphor for a particular kind of solution to capitalist crises through temporal deferral and geographical expansion. (2003: 115)


            Temporal deferral and geographical expansion “fix” the overaccumulation crises that arise from the chronic tendency of capital to accumulate over and above what can be reinvested profitably in the production and exchange of commodities. As a result of this tendency,  surpluses of capital and labor are left unutilized or underutilized. The incorporation of new space into the system of accumulation absorbs these surpluses in two ways. At first, it promotes their utilization in the activities involved in opening up the new space and endowing it with the necessary infrastructure, both physical and social. And then, once the new space has been adequately “produced,” the surpluses of labor and capital can be absorbed in the new productive combinations that have been made profitable by the spatial enlargement of the system of accumulation (Harvey 2003: 109-112).

            As Harvey notes, this metaphorical meaning of spatial-temporal fix as solution to capitalist crises can and recurrently does enter into contradiction with the material meaning of the expression. For the geographical expansion, reorganization, and reconstruction that absorb surplus capital and labor “threaten... the values already fixed in place (embedded in the land) but not yet realized.” Hence,

The vast quantities of capital fixed in place act as a drag upon the capacity to realize a spatial fix elsewhere.... If capital does move out, then it leaves behind a trail of devastation and devaluation; the deindustrializations experienced in the heartlands of capitalism... in the 1970s and 1980s are cases in point. If capital does not or cannot move... then overaccumulated capital stands to be devalued directly through the onset of a deflationary recession or depression. (Harvey 2003: 116)


            Either way, spatial fixes can be expected to be associated with interregional volatility and the redirection of capital flows from one space to another. The redirection may occur smoothly, or it may involve what Harvey calls “switching crises” (2003: 121-23; 1982: 428-29). Harvey does not spell out the relationship between overaccumulation crises, spatial-temporal fixes, and switching crises. But the drift of his argument seems to be that, while overaccumulation crises are the cause, switching crises are a possible effect of the spatial-temporal fixes that recurrently revolutionize the historical geography of capitalism. They stem from resistance to the relocations involved in spatial fixes–a resistance that at least in part originates from the contradictory logic of capital accumulation itself. Indeed,“the more capitalism develops,” argues Harvey, “the more it tends to succumb to the forces making for geographical inertia.”

The circulation of capital is increasingly imprisoned within immobile physical and social infrastructures which are crafted to support certain kinds of production... labor processes, distributional arrangements, consumption patterns, and so on. Increasing quantities of fixed capital... check uninhibited mobility.... Territorial alliances, which often become increasingly powerful and more deeply entrenched, arise.... to conserve privileges already won, to sustain investments already made, to keep a local compromise intact, and to protect itself from the chill winds of spatial competition.... New spatial configurations cannot be achieved because regional devaluations are not allowed to run their course. The uneven geographical development of capitalism then assumes a form that is totally inconsistent with sustained accumulation either within the region or on a global scale. (1982: 428-9)


            In discussing the spatial fix that in his view is most prominent in the present conjuncture (the emergence of China as the main absorber of surplus capital), Harvey adds a new element to the forces of geographical inertia that may prevent new spatial configurations from being achieved: resistance to hegemonic change. For this “remarkable version” of spatial-temporal fix “has global implications not only for absorbing overaccumulated capital, but also for shifting the balance of economic and political power to China as the regional hegemon and perhaps placing the Asian region, under Chinese leadership, in a much more competitive position vis-a-vis the United States.” This possibility makes US resistance to a smooth spatial fix all the more likely, despite the fact that such a fix holds out the best prospect for a solution of the underlying overaccumulation crisis (Harvey 2003: 123-4).

            The association between spatial fixes and hegemonic shifts thus strengthens the “catch-22" that always confronts previously leading centers of capitalist development. The unconstrained development of capitalism in new regions brings devaluation to these centers through intensified international competition. Constrained development abroad limits international competition but blocks off opportunities for the profitable investment of surplus capital and so sparks internally generated devaluations (Harvey 1982: 435). If the competitively challenged center is also a hegemonic center, either outcome threatens to deflate not just its assets but its power as well.

            Harvey envisages two possible ways out of this catch-22. One is the use of financial means “to rid the system of overaccumulation by the visitation of crises of devaluation upon vulnerable territories” (2003: 134). And the other is the use of political and military means to turn international competition to the advantage of the more powerful states. The deployment of these means constitutes the “sinister and destructive side of spatial-temporal fixes to the overaccumulation problem.”

Like war in relation to diplomacy, finance capital intervention backed by state power frequently amounts to accumulation by other means. An unholy alliance between state powers and the predatory aspects of finance capital forms the cutting edge of a “vulture capitalism” that is as much about cannibalistic practices and forced devaluations as it is about achieving harmonious global development. (2003: 135-6)


            Harvey goes on to note that these “other means” are what Karl Marx, following Adam Smith, referred to as the means of “primitive” or “original” accumulation. He quotes approvingly Hannah Arendt’s observation that “the emergence of ‘superfluous’ money... which could no longer find productive investment within the national borders,” created a situation in the late 19th and early 20th centuries whereby Marx’s “original sin of simple robbery... had eventually to be repeated lest the motor of accumulation suddenly die down” (Harvey 2003: 142). Since a similar situation appears to have emerged again in the late 20th and early 21st centuries, Harvey advocates a “general reevaluation of the continuous role and persistence of the predatory practices of ‘primitive’ or ‘original’ accumulation within the long historical geography of capital accumulation.” And since he finds it peculiar to call an ongoing process “primitive” or “original,” he proposes to replace these terms with the concept of “accumulation by dispossession.”

            Historically, accumulation by dispossession has taken many different forms, including the conversion of various forms of property rights (common, collective, state, etc.) into exclusive property rights; colonial, semi-colonial, neo-colonial, and imperial appropriations of assets and natural resources; and the suppression of alternatives to the capitalistic use of human and natural resources. Although much has been contingent and haphazard in the modus operandi of these processes, finance capital and the credit system have been major levers of dispossession, while the states, with their monopolies of violence and definitions of legality have been crucial protagonists (Harvey 2003: 145-9). But whatever its manifestations, agencies, and instruments,

What accumulation by dispossession does is to release a set of assets (including labor power) at very low (and in some instances zero) cost. Overaccumulated capital can seize hold of such assets and immediately turn them to profitable use. (Harvey 2003: 149)


            Accumulation by dispossession can take place both at home and abroad. The more developed capitalistically a state is, however, the greater the difficulties involved in practicing it at home, and the greater the incentives and the capabilities to practice it abroad. It follows that accumulation by dispossession is only in part a substitute for spatial fixes to overaccumulation crises. To an extent that increases with the development of capitalism in the states or regions facing overaccumulation problems, it involves a spatial fix of its own–a spatial fix, that is, that expands the geographical scope of the system of accumulation through the forcible or fraudulent appropriation of something for nothing, rather than through the exchange of nominally equivalent values.


II. A Conceptual Map of Historical Capitalism

The concepts reviewed in the preceding section can be used, as Harvey does, to interpret current US dispositions to remake the map of the world to suit US interests and values, in comparison with the dispositions that drove the territorial expansion of capitalist states in the late 19th and early 20th century. But they can also be used to interpret the peculiar expansionary tendencies of historical capitalism over a much longer time horizon than that encompassed by Harvey’s observations. This much longer horizon stretches as far back in time as we can detect overaccumulation crises that are in key respects comparable to the present one.

            As I have argued in The Long Twentieth Century, persistent systemwide overaccumulation crises have characterized historical capitalism long before it became a mode of production in the late 18th and early 19th centuries. Taking long periods of “financialization” across political jurisdictions as the most valid and reliable indicator of an underlying overaccumulation crisis, I identified four partly overlapping “systemic cycles of accumulation” of increasing scale and decreasing duration, each consisting of a phase of material expansion–in the course of which capital accumulates primarily through investment in trade and production–and a phase of financial expansion, in the course of which capital accumulates primarily through investment in property titles and other claims on future incomes. Contrary to the reading of some critics, the identification of these cycles does not portray the history of capitalism as “the eternal return of the same,” as Michael Hardt and Antonio Negri put it (2000: 239). Rather, they show that, precisely when the “same” (in the form of recurrent system-wide financial expansions) appears to return, new spatial-temporal fixes, major switching crises, and long periods of accumulation by dispossession have revolutionized the historical geography of capitalism. Integral to these “revolutions” was the emergence of a new leading agency and a new organization of the system of accumulation.

            A comparison of these distinct agencies and organizations reveals, not only that they are different, but also that the sequence of these differences describes an evolutionary pattern towards regimes of increasing size, scope and complexity. This evolutionary pattern is summed up in figure 1 (the figure and much of what follows in this section are taken from Arrighi and Silver 2001: 264-68). The first column of the figure focuses on the "contain­ers of power"—as Anthony Giddens (1987) has aptly characterized states—that have housed the "headquarters" of the leading capitalist agencies of the succes­sive regimes: the Republic of Genoa, the United Provinces, the United Kingdom, and the United States.



 Leading                                Regime Type/Cycle                                                          Costs Internalized



                                    Extensive                     Intensive              Protection       Production     Transaction     Reproduction   







                                                                         US                              Yes                 Yes                    Yes                  No                                    





                                     British                                                             Yes                 Yes                    No                   No






                                                                          Dutch                         Yes                 No                      No                   No






                                        Genoese                                                         No                  No                     No                  No






Figure 1: Evolutionary Patterns of World Capitalism



Source: Arrighi and Silver (2001: 265)



            At the time of the rise and full expansion of the Genoese regime, the Republic of Genoa was a city-state small in size and simple in organization, which contained very little power indeed. Yet, thanks to its far-flung commercial and financial networks the Genoese capitalist class, organized in a cosmopolitan diaspora, could deal on a par with the most powerful territorialist rulers of Europe, and turn the relentless competition for mobile capital among these rulers into a powerful engine for the self-expansion of its own capital. At the time of the rise and full expansion of the Dutch regime of accumulation, the United Provinces was a hybrid kind of organization that combined some of the features of the disappearing city-states with some of the features of the rising nation-states. The greater power of the Dutch state relative to the Genoese enabled the Dutch capitalist class to do what the Genoese had already been doing—turn interstate competition for mobile capital into an engine for the self-expansion of its own capital—but without having to "buy" protection from territorialist states, as the Genoese had done through a relationship of political exchange with Iberian rulers. The Dutch regime, in other words, “internalized” the protection costs that the Genoese had “externalized” (see Figure 1, column 4).

            At the time of the rise and full expansion of the British regime of accumulation, the United Kingdom was not only a fully developed nation-state. ­It was also in the process of conquering a world-encompassing commercial and territorial empire that gave its ruling groups and its capitalist class a command over the world's human and natural resources without parallel or precedent.  This command enabled the British capitalist class to do what the Dutch had already been able to do—turn to its own advantage interstate competition for mobile capital and "produce" all the protection required by the self-expansion of its capital—but without having to rely on foreign and often hostile territorialist organizations for most of the agro-industrial production on which the profita­bility of its commercial activities rested. If the Dutch regime relative to the Genoese had internalized protection costs, the British regime relative to the Dutch internalized production costs as well (see Figure 1, column 5). As a consequence of this internalization, world capitalism continued to be a mode of accumulation and rule but became also a mode of production.

            Finally, at the time of the rise and full expansion of the US regime of accumulation, the US was already something more than a fully developed nation-state.  It was a continental military-industrial complex with sufficient power to provide a wide range of subordinate and allied governments with effective protection and to make credible threats of economic strangulation or mili­tary annihilation towards unfriendly governments anywhere in the world.  Combined with the size, insularity, and natural wealth of its domestic territory, this power enabled the US capitalist class to internalize not just protection and production costs—as the British capitalist class had already done—but transaction costs as well, that is to say, the markets on which the self-expansion of its capital depended (see Figure 1, column 6).

            This steady increase in the geographical size and functional scope of successive regimes of capital accumulation on a world scale is somewhat obscured by another feature of the temporal sequence of such regimes. This feature is a double movement, forward and backward at the same time. For each step forward in the process of internalization of costs by a new regime of accumula­tion has involved a revival of governmental and business strate­gies and structures that had been superseded by the preceding regime. Thus, the internalization of protection costs by the Dutch regime in comparison with the Genoese regime occurred through a revival of the strategies and structures of Venetian state monopoly capitalism that the Genoese regime had superseded.  Similarly, the internalization of production costs by the British regime in comparison with the Dutch regime occurred through a revival in new and more complex forms of the strategies and structures of Genoese cosmo­politan capitalism and Iberian global territorialism.  And the same pattern recurred once again with the rise and full expansion of the US regime, which internalized transaction costs by reviv­ing in new and more complex forms the strategies and structures of Dutch corporate capitalism (see Figure 1, columns 1 & 2).

            This recurrent revival of previously superseded strategies and structures of accumulation generates a pendulum-like movement back and forth between "cosmopolitan-imperial" and "corporate-national" organizational structures, the first being typical of "extensive" regimes—as the Genoese-Iberian and the British were—and the second of "intensive" regimes—as the Dutch and the US were.  The Genoese-Iberian and British "cosmopolitan-imperial" regimes were exten­sive in the sense that they have been responsible for most of the geographical expansion of world capitalism. Under the Genoese regime, the world was "discovered," and under the British it was "conquered." The Dutch and the US "corporate-national" regimes, in con­trast, were intensive in the sense that they have been responsi­ble for the geographical consolidation rather than expansion of the historical capitalism. Under the Dutch regime, the "discovery" of the world realized primarily by the Iberian part­ners of the Genoese was consolidated into an Amsterdam-centered system of commercial entrepots and joint-stock chartered compa­nies.  And under the US regime, the "conquest" of the world realized primarily by the British themselves was consolidated into a US-centered system of national states and transnational corporations.

            This alternation of extensive and intensive regimes ­blurs our perception of the underlying, truly long-term, tendency towards the formation of regimes of increasing geographical scope. When the pendulum swings in the direction of extensive regimes, the underlying trend is magnified, and when it swings in the direction of intensive regimes, the underlying trend appears to have been less significant than it really was. Nevertheless, once we control for these swings ­by comparing the two intensive and the two extensive regimes with one another—the Genoese-Iberian with the British, and the Dutch with the US—the underlying trend becomes unmistakable.

            The globalization of historical capitalism has thus been based on the formation of ever more powerful cosmopolitan-imperi­al (or corporate-national) blocs of governmental and business organizations endowed with the capacity to widen (or deepen­) the functional and spatial scope of the system of accumulation. And yet, the more powerful these blocs have become, the shorter the life-cycle of the regimes of accumulation that they have brought into being—the shorter, that is, the time that it has taken for these regimes to emerge out of the overaccumulation crisis of the preceding dominant regime, to become themselves dominant, and to attain their limits as signaled by the beginning of a new ­overaccumulation crisis.  Relying on Braudel’s dating of the beginning of financial expansions, I have calculated that this time was less than half both in the case of the British regime relative to the Genoese and in the case of the US regime relative to the Dutch (Arrighi 1994: 216-17).

            This pattern of capitalist development whereby an increase in the power of regimes of accumulation is associated with a decrease in their duration, calls to mind Marx's contention that "the real barrier of capitalist production is capital it­self" and that capitalist production continually overcomes its immanent barriers "only by means which again place these barriers in its way on a more formidable scale" (1962: 244-5). But the contradiction between the self-expansion of capital on the one side, and the development of the material forces of production and of an appropriate world market on the other, can in fact be reformulated in even more general terms than Marx did.  For capitalism as historical social system ­became a "mode of production"—that is, it internalized production costs—only in its third (British) stage of develop­ment.  And yet, the principle that the real barrier of capitalist development is capital itself, that the self-expansion of exist­ing capital is in constant tension, and recurrently enters in open contradiction, with the expansion of world trade and production and the creation of an appropriate world market—all this was clearly at work already in the Genoese and Dutch stages of develop­ment, notwithstanding the continuing externalization of agro-industrial production by their leading agencies. In all instances the contradiction is that the expansion of trade and production was mere means in endeavors aimed primarily at increasing the value of capital. And yet, over time it tended to generate more capital than could be absorbed profitably within the confines of the extant spatial-temporal fix (in the material meaning of the expression), thereby threatening to drive down overall returns to capital and thus deflate its ­value.

            The resolution of the ensuing overaccumulation crises through a new spatial-temporal fix (in both meanings of the expression) has taken relatively long periods of time–as a rule more than half a century. In all instances, the resolutions have been punctuated by major switching crises and have involved processes typical of accumulation by dispossession. Although much in the modus operandi of these processes has indeed been contingent and haphazard as Harvey suggests, in Chaos and Governance my co-authors and I have nonetheless detected some regularities, three of which are especially germane to our present concerns.

            First, one kind or another of financialization has always been the predominant response to the overaccumulation problem of the established organizing centers of the system of accumulation. Thanks to their continuing centrality in networks of high finance, these centers have been best positioned to turn the intensifying competition for mobile capital to their advantage, and thereby reflate their profits and power at the expense of the rest of the system. Over time, however, financial expansions have promoted the geographical relocation of the centers of capital accumulation by rerouting surplus capital to states and regions capable of ensuring a more secure and profitable spatial-temporal fix to the overaccumulation crisis. Previously dominant centers have thus been faced with the Sisyphean task of containing forces that keep rolling forth with ever renewed strength. Sooner or later, even a small disturbance can tilt (and historically have invariably tilted) the balance in favor of the forces that wittingly or unwittingly are undermining the already precarious stability of existing structures, thereby provoking a breakdown of the system of accumulation (Arrighi and Silver 1999: 258-264).

            Second, the states have been key protagonists of the struggles through which old spatial-temporal fixes are destroyed and fixes of greater geographical scope are attained. In the past, switches to fixes of greater geographical scope were premised on the interstitial emergence of governmental-business complexes that were (or could plausibly become) more powerful both militarily and financially than the still dominant governmental-business complex--as the US complex was relative to the British in the early twentieth century, the British complex relative to the Dutch in the early eighteenth century, and the Dutch relative to the Genoese in the late sixteenth century. In the present transition, it is not yet clear whether and how a governmental-business complex more powerful than the US complex can emerge and eventually provide a solution to the ongoing overaccumulation crisis. But in so far as the past dynamic of historical capitalism is concerned, this tendency towards the formation of ever more powerful governmental-business complexes is one of its most important features (Arrighi and Silver 1999: 88-96, 263-70, 275-8, 286-89).

            Finally, in each transition accumulation by dispossession has provoked movements of resistance and rebellion by subordinate groups and strata whose established ways of life were coming under attack. Interacting with the interstate power struggle, these movements eventually forced the dominant groups to form new hegemonic social blocs that selectively included previously excluded groups and strata. This increasing “democratization” of historical capitalism has been accompanied by a speedup in the impact of social conflict on overaccumulation crises. Thus, while the overaccumulation crisis of the Dutch regime of accumulation was a long drawn out process in which systemwide social conflict came much later than the systemwide financial expansion, in the overaccumulation crisis of the British regime the systemwide financial expansion gave rise almost immediately to systemwide social conflict. This speedup in the social history of capitalism has culminated in the explosion of social conflict of the late 1960s and early 1970s, which preceded and thoroughly shaped the crisis of the US regime of accumulation (Arrighi and Silver 1999: 153-216; 282-6; Silver 2003).


III. Toward a Geographical Representation of Historical Capitalism

The foregoing analysis suggests five basic rules that in my view are essential to a minimally accurate geographical representation of the processes that underlie the globalization of historical capitalism over the past half millennium. Having run out of time, I can only list these rules and leave their further specification for the presentation of the paper at the conference.

            Rule # 1. The idea still dominant in world-system analysis of a quantitatively expanding but structurally invariant world capitalist systems must be abandoned, including and especially the notion of Kondratieff cycles, hegemonic cycles, and logistics as empirical manifestations of such a structural invariance. The globalization of historical capitalism must instead be represented as involving fundamental structural transformations of the spatial networks in which the system of accumulation has been embedded.

            Rule # 2. In this kind of representation, priority should be given to the networks of each regime’s leading business and governmental organizations. Due attention should be paid to the fact that the spatial organization of “cosmopolitan-imperial” (extensive) regimes is quite different from that of “corporate-national” (intensive) regimes. Comparisons across time of the degree of globalization attained by historical capitalism must take into account and highlight the most important differences between these two kinds of regime.

            Rule # 3. Representaions should focus on those points in time that enable us to highlight not just cyclical but structural transformations as well. Points in time close to the change of phase from material to financial expansion are the most important from both points of view. By comparing a succession of representations at such times, we would highlight structural transformations. And by comparing each of these representations with analogous representations at later points within the life-time of the same regime, we would highlight the cyclical transformations involved in the recurrence of overaccumulation crises.

            Rule # 4. Representations of any particular regime of accumulation at a late stage of its development, should depict not just the dominant spatial fix but also the interstitial emergence of agencies and networks that subsequently provided a solution to the underlying overaccumulation crisis. If feasible, we should try to represent also the interstitial emergence of other agencies and networks that never became dominant but constituted plausible historical alternatives to those that did.

            Rule # 5. To the extent that social conflicts are included in the representations, account should be taken both of their concentration in periods of accumulation by dispossession, and of their transformation from being a “dependent variable” to being an “independent variable” in relation to overaccumulation crises. In any event, the geographical mapping of social conflict requires concepts and techniques that fall beyond the scope of this paper.




Arrighi, Giovanni (1994). The Long Twentieth Century. Money, Power, and the Origins of Our Time. London: Verso.

Arrighi, Giovanni and Beverly J. Silver (1999). Chaos and Governance in the Modern World System. Minneapolis, MN: University of Minnesota Press.

Arrighi, Giovanni and Beverly J. Silver (2001). “Capitalism and World (Dis)order.” Review of International Studies 27, 257-279.

Chase-Dunn, Christopher and Thomas Hall. (1997).     Rise and Demise : Comparing World-systems. Boulder, CO : Westview Press

Hardt, Michael and Antonio Negri. (2000).  Empire.  Cambridge, MA: Harvard University Press.

Giddens, Anthony (1987). The Nation-State and Violence. Berkeley: University of California Press.

Harvey, David. (1982). The Limits to Capital.

Harvey, David. (2003). The New Imperialism. Oxford: Oxford University Press

Marx, Karl (1962). Capital, Vol. III. Moscow: Foreign Languages Publishing House.

Silver, Beverly J. 2003. Forces of Labor. Workers’ Movements and Globalization Since 1870. Cambridge and New York: Cambridge University Press.

            1 Paper to be presented at the Conference on Globalization in the World-System: Mapping Change over Time. University of California, Riverside, February 7-8, 2003.