Key words for a political geography of capitalism

Kevin R. Cox

Ohio State University

Texto de referência para o debate na mesa redonda "Os desafios e os novos debates na Geografia Política Contemporânea" do dia 10/12/2020 no Ciclo de Debates Virtuais do IV CONGEO

ABSTRACT: In political geography, there are certain keywords. They presuppose everything else that political geographers study. My personal list: territory; state; country; and geopolitics. All these take on a particular form in a world where production is organized along capitalist lines; in fact they can be said to express capital’s contradictions. I organize my discussion in order of levels of abstraction. The state presupposes territory; ‘country’ presupposes the state. And so on. I then discuss the implications of the discussion for research.


Key words are an idea originally popularized by the Marxist literary critic Raymond Williams. He wrote a book with the title Keywords in which he examined words in common contemporary use, particularly among academics, and tracing the development of their – often disputed – meanings over time with some indication of why they changed: so ‘capitalism’, ‘consumer’, ‘country’, ‘individual’, ‘interest.’ Harvey (2006) then took up the idea in a festschrift edited in his honor, pointing out that a key word that Williams omitted was ‘space’: so ‘Space as a Keyword’ though without much in the way of delving back into the Middle Ages and earlier as Williams did with many of his keywords, but emphasizing space under capitalism.

I take my cue from Harvey in foregrounding the significance of capitalism for the shared vocabulary of contemporary human geography and in particular, for understanding political geography. The very idea of political geography expresses the separations that emerge as common discourse under capitalism. It is not political geography that is my focus, though, as what I take to be keywords in the sub-field, which again take on a very particular meaning under capitalism. Four key words, therefore: territory; state; country; and geopolitics, and to be treated in that order. This is an order according to levels of abstraction in which lower levels presuppose the existence of higher levels: how higher level abstractions are a pre-requisite for abstracting those at lower levels. So according to Marx, you cannot abstract ‘labor as such’ until you have abstracted capitalism; this is because its meaning as indifference to concrete forms of labor subsequent to the de-skilling of tasks, only emerges subsequent to the capitalist form of development (1973: 104-105.) One can make the same argument for the word ‘location’: it makes no sense except in a society where locations are substitutable one for another (Cox 2014: 173). My focus then, is how to understand these keywords in political geography as they emerge under capital and reflect its contradictions.


For political geography, this word is utterly foundational. Again, it is not one of Williams’ keywords, though there is clearly a potential for an historical treatment as pre-capitalist meanings were significantly different (Elden 2013.) We should also note what a subterfuge existence it has enjoyed. Only after the 1970s did political geographers start to explicitly talk about it though Hägerstrand (1973: 85) sensed the possibilities quite early on with his ideas about projects and how to fit them into a time-space that was already significantly occupied.[1] Classical geopolitics, most notably Mackinder and Bowman, was all about territory, but you will struggle to find them actually using the term. To scramble things further, the economic geographers of the 1970s on, would use the term to indicate something localized, as in territorialized development.[2]

In The Dictionary of Human Geography (2009: 745), John Agnew defines territory thus: “A unit of contiguous space that is used, organized and managed by a social group, individual person or institution to restrict and control access to people and places.” He then remarks that the dominant emphasis in discussions of territory has been political, as in the power to limit access, or ethological, as in the territoriality of birds and other animals. This seems to me to be very fair and avoids the sort of statist version found in dictionaries where territory is associated only with states; after all, neighborhood organizations in the United States and in Western Europe organize to keep out certain sorts of development that they regard as undesirable. Territory is indeed about boundaries, but not necessarily formal ones recognized by a state. It is about insides and outsides, excluding and including. These are the sorts of general properties that one can infer from the variety of historical usages. But general definitions are useful only up to a point. While they define certain common properties, they lift them out of the particular historic-social relations that energize them and that give them very particular meanings. Territory is different under capitalism than what it was under pre-capitalist forms of production.

As David Harvey (1985) argued in his pathbreaking paper on the geopolitics of capitalism, in the development of the capitalist space economy there are both localizing and de-localizing moments and it is in their relation one to another that lies the key to capitalist territoriality. Under capitalism, production is socialized. This occurs through the development of the division of labor both within the factory, between different units of capitalist production and between town and country. It also depends on the creation of shared means of production: again within the workplace itself, as with assembly lines or oil refineries; but also those means of production that are shared among firms like means of transport and provision of water or today, state subsidized training facilities. In addition, there are shared means of the reproduction of labor power: a housing stock that changes hands, schools, means of getting to work.

All this occurs within a space through which capital can circulate profitably, given the existing means of transport and communication. Housing for workers has to be relatively close to places of work. Suppliers of raw-materials cannot be located so far away that transport costs cannot be recuperated through sales. This obviously has scalar elements; the space of a housing market is different from that of a spatial division of labor, but the scales are complementary. The daily journey to work is limited in its range, but that does not mean that labor cannot make longer, more permanent moves over a larger area that corresponds more to one within which the other forms of commodity exchange are taking place: what will be called here a space of accumulation.

This, though, is to abstract from the class relation fundamental to capitalist development. The reason for shared means of production and refinement of the division of labor, including in its spatial aspects, is to facilitate the extraction of surplus value; to speed up the rate at which capital circulates so that it can absorb more surplus labor; to expel workers from the labor process as more can be produced with less; and so to create a pool of unemployed to be drawn on as over-accumulated capital develops new products and so needs the labor power to produce them. Class tension is always hovering in the background. Unemployment, the decanting of production to branch plants elsewhere, exercise continual downward pressure on wages and conditions of work so that resistance is never far away and workers organize and go on strike to negotiate better terms.

Accordingly class conflict threatens to disrupt the process of circulating capital through each of its necessary, and successive forms, so that it can absorb more and more surplus labor. Some compromise, some social settlement, some set of norms and values that go on behind the backs of everyone but which will work to capital’s advantage, has to be achieved. This compromise, this uneasy state of social peace, takes the form of what Harvey (1985) has called a ‘structured coherence’: a class relation shored up by some sense of shared institutions, shared values, understandings and practices, gender and race relations, a particular division of labor’s consumption fund that can allow capital to divide and rule. This builds on the illusions of the wage relation that serve to conceal the exploitative relation that it is. It is solidified through the state, both through what the state does, as in its assertion of equal rights and the provision of schools that instill a sense of national belonging; and through what is more implicit, as in the way that governments of the day provide a shared topic of interest regardless of social position and the manner in which state boundaries provide a frame for other shared activities like football competitions.

In their details though – and this is crucial to the argument – structured coherences vary. They will inevitably serve the capitalist class, though more in some instances than in others; but in their concreteness they reflect different histories, different juxtapositions of influences from elsewhere, different working class divisions, different traditions outside the workplace. This is obvious when we consider countries as spaces of accumulation: Germany is not France; the countries of Western Europe share some things that they do not share with the United States, not to mention that the US is of a continental scale and this has left its mark on the national economy. It also applies within countries. The Scottish Clydeside centered on Glasgow, was once distinct for not just its shipbuilding, iron and steel and coal mining, not to mention its colonial trade, but also its large immigration from Ireland and then its harsh sectarian divisions that extended into the local passion, football; so a very male world too.

This is capital’s localizing moment: its creation of spaces of accumulation along with a structured coherence to facilitate the continuity of what is a harsh, exploitative regime. Crucially it is a base for capitalist expansion; for the exploitation of more distant markets, to the extent that transport costs allow; for expansion through pulling in new labor power from elsewhere, new sources of raw-materials and of necessary consumer goods; for the development of a more intense division of labor to facilitate productivity and hence the ability to invade new markets and drive out competitors.

This means, though, that capital also has a delocalizing moment. It is as a result of its contradictions, that capital spills out from an existing space of accumulation. Its failure, in virtue of its very nature, to create a market sufficient to absorb its increasing stream of products, pushes it in search of consumers elsewhere. The continual drive to lower costs in order to steal a competitive march then results in attempts to substitute cheaper raw-materials to be found at a greater distance. New technologies mean a new geography: the transformations subsequent to the replacement of the steam engine by electricity are classic – radical change at all geographic scales.[3] Meanwhile tendencies to over-accumulation signaled by crisis, can result in the emergence of entirely new sectors with more expansive geographies: cotton textiles, the auto industry, and lots more. This is predicated on the continual lowering of transport costs and improvements in communication, both pushed by demand from other capitalists, but also by the transport industry’s own desire to attract customers through a continual lowering of the real costs of movement.

This emergent capitalist geography is then complemented by other changes that help to cement a new space of accumulation. State regulation has to occur over a wider geographic area and more local forms abandoned. As capitals spread out, so their cost competition becomes an issue for labor. Unions have to be become more extensive in their organization if wages and conditions of work are to be taken out of the competitive calculations of capital. Labor unions that are more local in their membership, federate to protect themselves against the competitive wage cutting of employers.

How earlier growth points are affected can vary. Some will remain relatively undisturbed by it all, operating in sectors that are unchallenged: the rise of the automobile industry did not threaten the mill towns of Lancashire, Yorkshire or New England. Some will reinvent themselves: Coventry went relatively seamlessly in the course of the nineteenth century from silk weaving to clock and watch manufacturing to bicycles and then in the twentieth century, to automobile production. In these cases, it is growth as usual, pulling in new workers, expanding markets and so on. But other places will struggle. The creation of larger spaces of accumulation can threaten existing ones in complex ways. New products from elsewhere, or simply cheaper ones, can threaten the economic base of a particular region. As national markets replace regional ones, firms may create branch plants elsewhere undermining old employment centers. There is a long history of this sort of rise and fall: many of the branch plants that once provided employment for numerous small towns across the rural US have now gone to Mexico or China. Declining demand for coal in the United Kingdom as steam trains were replaced by diesel, as home heating shifted to oil or electricity, and as the steel companies started importing coal over very long distances indeed – like from the US, Russia and believe it or not, Australia – created regional crises in major coalfield areas, most notably South Wales and Northeast England.

The problem is one of fixity: the mobility of capital enters into contradiction with its fixity (Harvey 1985.) Space of accumulation are based on a socialization of production which in turn depends on massive fixed investments: investments in the common physical infrastructure like rail and docks, in the production facilities themselves – mines, steel mills, auto assembly plants, heavy engineering plants – and in housing for workers. In a relatively stable space economy, this creates the social and physical conditions for a structured coherence that can provide the basis for continuing growth; but in an unstable one, which is the typical case, it can be profoundly problematic.

Most of the physical investment is of long life, which means that it may well not have been amortized. This means in turn that banks with loans yet to be paid off will feel the draught. On top of fixed investments there is a tissue of social relations that facilitates production in particular places or regions: relations of trust between firms, a bush telegraph through which workers hear about jobs and which makes labor recruitment that much easier, not to mention the ties within extended families. The complex of highly particular social practices and customs holds people back from moving: not just workers but small businesses. In short, as some production at least, declines and moves elsewhere, it is not easy liquidating existing fixed capital and those relations between firms and people built up over a long period of time: relocation can come with heavy costs.

This will be the occasion for a variety of policy initiatives to protect the future of the area: in short, policies of a territorial nature to revive the local economy; to keep out foreign products undermining the local economic base; to restructure so as to make it more competitive. Or, alternatively, calls for the government to implement some sort of regional policy which will redistribute employment to the advantage of areas now feeling the shock of industrial change; what are now known as ‘rust belts.’ This can occur at all manner of geographic scales and may then affect them all as the literature on globalization has affirmed. It can even occur within the same sector of the economy. In the US, policies targeted at reducing acid rain meant growth in the low sulphur coal areas of Colorado, Montana and Wyoming, but doom for the high sulphur coal mining regions of the Midwest and triggering off a politics of uneven development (Cox 2016 : 47-49.) But any attempt to protect the future of local or regional economies in a context of dislocating change stands to be met by opposition from those gaining from it. Localities, regions and, indeed, countries, are both proactive about growth and to the extent that the geography of development turns against them, resistant to those particular changes.

How this politics develops can be hugely variable. Fundamentally, it is the way in which class relations intersect with geographic change and this varies from country to country and the particular constraints and opportunities created by the structure of the state. What happened in the course of the controversy around acid rain would have been impossible in France or the United Kingdom. It is to the questions of state and country that we turn next.


States always have boundaries but in the case of capitalist states, their territorial form is quite distinct. In feudal Europe where territory from the standpoint of the ruling class meant rent from land [4], principalities, dukedoms, bishoprics, or whatever, could actually be dispersed among each other – a bit here and a bit there (Figure 1.) This was often an outcome of an inter-dynastic marriage, which was always seen as a way of enhancing feudal revenues. In contrast, state territory under capitalism tends to be quite compact, unpunctured [5] and apart from island states, like Indonesia or New Zealand, unfragmented.[6] This is also true of the metropolitan areas, regions, the provinces into which states are divided or, if lacking any institutional form, those areas that are mobilized to do battle with others. This in turn testifies to the very particular material base of territory under capital: a reflection of spaces of accumulation, some still relevant, others long incorporated into other territorial formations, which are in turn structured geographically by considerations of the relative costs of making capital circulate within them. As spaces of accumulation have been mobilized by capital, either as a means of defending a structured coherence in place or as a springboard for expanding elsewhere – a sort of defense through attack – so preexisting territorial forms have been mobilized to that purpose, or new ones created through acts, say of national unification.

In short, while states are capitalist, we cannot envision them in their current functions without taking into account the fact that they are also territorial in their form, and territorial interests are a necessary condition, rather than the state as condition for the territorial. They pursue territorial struggle with other states: struggles over trade, investment, the changing international division of labor and, at one time, empire. They also though, quite crucially, mediate territorial struggle within their territorial jurisdiction. This is apparent in the way in which states function internally, but also externally, since it is through the resolution of internal struggles that states position themselves for pursuing territorial interests on a global scale. The dominance of London and the Southeast in British trade and public investment policy is a nice instance of this.

Figure 1: The Palatinate of the Rhine

Jessop (2000: Chapter 12) has referred to three dimensions of the contemporary state: what he calls inputs or the way in which demands are made on it; policy outputs; and the way in which the state organizes its division of labor or ‘through-puts.’[7] All of these have territorial moments (see Table 1.) Governments are typically elected from constituencies, congressional districts, communes; the territorial aspect is particularly prominent where there are regional or separatist parties. Policy always has a geographically differentiating aspect, sometimes implicit, as in the way taxation falls differently according to the uneven development of a national economy; or explicitly as in policies that are geographically targeted, like urban renewal or policies aimed at limiting the growth of large cities. The state then organizes its division of labor territorially: not just the distinction between local and national government and the variable distribution of powers, but also the judiciary with its various circuits.

Table 1 Territorial aspects of the state.

For reasons having to do with the structured coherence in which they are embedded, states then vary in their form. The American state is extraordinarily decentralized, according powers and responsibilities downwards in a way that is in sharp contrast to more centralized ones like France and the United Kingdom. Representation is also more territorialized through the institution of the primary election and the committee system where legislators sit on committees whose business is of particular interest to respective legislative districts (Cox 2016: Chapter 8.) Likewise, and pursuing the American case more, explicitly territorialized policies have had a history that again differs from that of France or the United Kingdom. The latter have one of formal regional policy, whereas this has been almost entirely lacking in the US (ibid.)

States are conditioned by the particularities of respective structured coherences, while then serving to reinforce them. The American state is extraordinarily decentralized. In this regard, it expresses central values of the social formation. These include a suspicion of the activist state, an emphasis on the merits of the market and the virtues of individual effort rather than collective action of the sort represented by the labor movement. This decentralization sets in motion a competition between local governments and between state governments which mimics the market and which serves to shift the balance from collective struggle through the labor movement to the merits of individual effort. States compete in their business climates for new investment; exalting the values of trickledown for raising individual incomes is part of the story then told. Local governments compete for ‘top notch’ citizens[8] who will share out their property wealth in the form of enhanced school revenues and therefore increased spending per pupil. All this is in sharp contrast to the much more centralized states of France and the United Kingdom and their different balance of class forces: countries with long histories of class militancy and where the idea of a welfare state in which all share has much deeper resonance.

These contrasts then help to explain quite distinct territorial politics. In the US an implicitly cross-class form of territorial alliance tends to hold sway. The decentralized form of the American state has facilitated the formation of growth coalitions in urban areas (Cox 2017.) Local governments anxious about their tax revenues and developers worried about their rents take the lead, but the whole developmental impetus has to be legitimated with the masses through reference to a politics of the collective. This can take the form of how it is all about building a bigger and better Chicago or claims of how everyone benefits from enhanced development through a (very dubious) process of trickledown. The coalition behind Trump’s ‘Making America Great’ is another instance. It brings together a diverse set of forces, but judging from the voting statistics it not only includes small businesses everywhere but also support from de-industrialized areas and small towns reeling from the closure of branch plants: so local businesses and the local – now lumpen – proletariat too.

The form assumed by territorial politics in France and the United Kingdom has been markedly different. There is a history of regional policy in both and despite recent regional devolution of responsibilities in France, the central state still holds the purse strings and makes release of money contingent on the consistency of regional plans with the national plan. In the United Kingdom local governments plan and regulate land use, but the central government has ultimate decision making power in cases regarded as of national interest. Regardless, everywhere local or regional interests and movements adapt to the particularities of the structured coherence in question, including its state form; or, in the case of the United Kingdom and the national movements in Scotland and Wales, seek to change it.


This is the one term in our set of four that is actually addressed by Raymond Williams. He defines its general use as ‘native land.’ It has close associations with ‘state’ and also with ‘nation’ but its sense is more positive: one does things for one’s country and rarely for one’s nation or for one’s state. National anthems, particularly in their alternative versions, often recognize that as in ‘America the beautiful’ or the British Labour party anthem, ‘Jerusalem’, the Afrikaner ‘Die Stem’, Canada’s ‘O Canada’ or Russia’s ‘Patriotic Song.’ References to ‘land’ or ‘-land’ are copious and indicative of where the emphasis lies.[9]

‘Native land’ has a territorial ring to it: a sense of attachment and fixity. We should also note that while Williams does not date this use, it almost certainly emerges, as does the idea of nation, with capitalism and it is the capitalist form of development that I would prioritize in understanding the concept of country.[10] This is because of the way in which the socialization of production unleashed by capitalist development results in a wider and shared intercourse: a spatial division of labor, shared means of production like the railway, and eventually shared print media. The latter were initially regional in Europe, but the railway then allowed them to become national. This shared intercourse, greatly enhanced through the migrations induced by capitalist development, would then provide the basis for common recreational forms: the creation of country-wide competitions and the creation of codes to regulate them. Country-wide interest in results would then be enhanced by the development of newspapers with a more than regional distribution.

This could, of course, be something shared in duress rather than viewed as a positive experience, and as such, one to be resisted. How is it that even if some recognize that the country works for some people, white men, for example, better than for others, nevertheless ‘country’ still emerges as something with positive valence? In part, it has to do with a familiarity that is mutual. ‘Country’ is a highly concrete experience: particular sorts of practice, particular traditions, particular landscapes. All give purpose to life so that when they are stripped away, there is a sense of loss (Marris 1975.) In part it has to do with the way in which economic life is experienced. In his discussions of the labor contract Marx emphasized this sense, albeit a deceptive sense: an equal relationship in which the worker exchanged his labor power for a wage that would be equivalent to it; and in which people related freely and without force. The worker could always refuse the contract offered. You might not be attached to an employment relation in the way in which you are attached to Match of the Day,[11] but it is not something that is immediately experienced as alienating from others.

The capitalist state then plays its own role in the construction of country and the sense of something shared and not necessarily in an oppressive sense. A public school system, the propagation of a national language, the establishment of civil rights for all, shared experiences like voting and government policies, a shared national culture as in literature and history, or the shared media and movie icons, all deepen the sense of country-hood. The state selects in certain symbols of the country, most notably people to be memorialized in the names of buildings, statues and street names or in national honors lists; but ones that have to be accepted as national heroes or unifying figures. The sense of country, equalizing as it is, is never devoid of hierarchy but it has to be one that is broadly acceptable.

The state deepens not just the sense but the reality of a country as an exclusive space: a territorial frame, therefore. Through its bounding and excluding activities it makes country a container within which circulate particular practices, particular cultural forms, including quite crucially, the landscape forms that are experienced daily and give a sense of being home. France is different: concrete poles for the electricity lines, green crosses for pharmacies, houses surrounded by walls with big metal doors giving entry to the compound (Figures 2.1 and 2.2.) In the United Kingdom, on the other hand, one has to be impressed by the prevalence of the semi-detached house, a landscape of hedgerows or walls of roughly hewed native rock, and a chequer board of often, still small fields.

Figure 2.1 French electricity poles

Figure 2.2: French pharmacy sign

Emphatically, it is a reified concept that then, in its normative associations, reflects the deeply alienated character of life in a capitalist society. To the extent that ‘country’ is seen as a product of geohistory, rather than something that has always been there, it is a purified one: a geohistory that makes no reference to the turmoils and struggles of capitalist development that led to its construction. Rather the emphasis will be on seamless unity, of a shared and undisputed heritage, without class tensions and contestations, and occluding the fate of those who just happened to be in the way like indigenous peoples and peasants or animal species long since dispersed.

This is a country in which its ‘natives’ – a term evoking the territorial base of the idea as in Williams’ ‘native land’ – can take pride. There has long been an implicit benchmarking of countries, whether it be Nobel Prizes, Olympic gold medals, and, not least, military prowess. It is a matter of a vicarious pleasure that, in virtue of the narrowness of one’s own life, frustrations at an inability to achieve the values of a capitalist society, always denied to the vast majority, allows at least some self-respect. We might be unemployed and living a hand to mouth existence, but at least we can join with others celebrating a national victory in international sport.

The idea of country, its reified character, can then be played out at smaller, more regional, scales. This is clearer in some cases than in others: so the French Midi rather than say, Beauce – not just the relative exoticism of its climate but its cultural icons like Pagnol and Fernandel, and its literary celebration in the work of Daudet and Giono. Likewise in England, the Northeast has a much stronger resonance in terms of a shared geohistory, than East Anglia; and the Black Country or the Five Towns celebrated in the novels of Arnold Bennett, more than South Yorkshire or South Lancashire.


The idea of geopolitics depends on those of territory and territorial interests as something to be defended and advanced; but also on the state, whose agents are called on in this process of defense and promotion, and on country, which legitimates engagement with those endangering it. It is, it should be emphasized, and like both territory and state, something that can apply not just between states but also within them as those embedded in more localized structures of coherence struggle for dominance. In what follows, I confine myself to geopolitics on an international scale, though the arguments are applicable equally to what goes on within countries and indeed which conjoins the intra-national with the inter-national.

Harvey’s (1985) geopolitics of capitalism is foundational here, but, as such, something to be built on. He puts territory at the center and those structured coherences which form in the context of capital’s localizing tendencies, as outlined earlier; and then how their continued existence is put in question by contradictory tendencies of delocalization. The upshot will be various attempts to ensure that continuity, by restructuring, perhaps, even by protectionist measures but always in class contested terms as the compromise made concrete in a particular structured coherence is threatened. If capital is to prevail it needs alliances with at least some fractions of the working class, usually the more privileged strata. So there will be talk of reviving ‘the national economy’, of the ‘Chinese threat’, ‘throwing off the albatross of the EU’, and of ‘Making America Great Again.’ We are so used to this sort of talk that we often fail to notice its ideological content. Its appeal is territorial. Class divisions are overridden. The message is that we, employers and employed, are in this altogether doing battle for markets with the foreigners, and all for the country we love. As Harvey wrote, “Global processes of class struggle appear to dissolve before our eyes into a variety of interterritorial conflicts” (1985: 152.)

The state plays a quite massive role in this process, most notably in constructing class alliances and then in formulating policy. Struggle on the stage of the global economy becomes a major part of its raison d’être. Everything else – the welfare state, educational policy, investment in national infrastructure – gets passed through this optic. And there is of course, good reason for this since of all the different agents in capitalist society, it is the most immobile of all. The English, workers and firms alike, can, if absolutely necessary, move somewhere else but this is a privilege denied to states: they are territorially fixed.

This is an important argument. It changes how we view territorialization and the capitalist state in quite radical ways and dispensing with all the superficialities of classical international relations or the textbooks of ‘international economics.’ It is a conception that can be developed: amended entirely in the spirit of Harvey’s thesis. It also stands for some critical revision without challenging its underlying assumptions.

In the first place, there is a scalar question. National governments regulate respective capitalist economies and that regulation is itself subject to the sorts of global competition at the heart of Harvey’s conception. They regulate all manner of things that bear directly and indirectly on the accumulation process and keeping capital in business: the money supply, capital-labor relations, property relations: in other words, all those things that enter into calculations of ‘business climate’ and how a country ranks on the (misleadingly labeled) ‘Index of Freedom.’[12] What can easily be overlooked is the fact of regulation beyond a country’s own boundaries. How is the global economy regulated? In part, it is self-regulation as in anxieties about ‘business climate.’ There are also more formal institutions set up to ensure that capital can function on a global scale and there have been a succession of these: the gold standard of the nineteenth century; the Bretton Woods agreements around a modified gold standard; the International Agreement on Tariffs and Trade, now transformed into the World Trade Organization; and the International Monetary Fund. These all have the function of lubricating trade and investment on a global scale: arbitrating trade disputes, and keeping countries within the orbit of the global economy. They are and always have been highly contested because their design always works for some more than for others. Control of those institutions, particularly their design, then becomes a political stake. They can become the means through which those who designed them in the first place can use them to advantage their own capitals on the global stage.

These relations were dramatically on display in the way the economic crisis in East Asian countries in the late 90s played out. These had long been in the hostile eye of the American government because of the way national government policy worked to keep out the multinationals, including, of course, the American ones. The goal of government policy had been national development through close relations between the government and major domestic firms. Multinationals were frowned on because of agendas that could work contrary to those of a developmental state. But what the US government saw, egged on by the financial press, was ‘crony capitalism.’ The crisis was its opportunity. The ‘crony capitalisms’ needed short term financing of negative trade balances and turned to the IMF. The IMF, though, had always reflected American priorities for no better reason than the Americans, along with their West European allies, equally hostile to the developmental state, monopolized its voting rights. So yes, countries like Indonesia, South Korea and Thailand would get loans but on conditions; the investment and takeover rules governing foreign firms had to be relaxed (Wade and Veneroso 1998.)

The effort to monopolize an extra-territorial space recalls the earlier formal empires which had in mind imposing an equally oppressive spatial division of labor: colonies, formal or otherwise, would serve as sources of cheap raw-materials and as outlets for manufactures, both capital and consumer goods. The formality of this type of subordination would give way to more informal methods of control as colonies gained their independence, but not before a global struggle over the division of imperial space. This reached its climax in World War II: if it had its way, Germany would dominate a New Europe, Japan was bent on creating its so-called Co-Prosperity Sphere in East Asia while Mussolini dreamed of a new Roman Empire.

There is a second point about Harvey’s thesis. While class struggle is at the center of his arguments about structured coherences and the development of cross-class alliances, it needs to be taken further. Geopolitical struggle is not only about the struggle of one nationally based capitalist class alliance against another. It is also about the global capitalist class in its entirety against a working class that does not necessarily confine its struggle to within national borders, and in particular, therefore, against the international labor movement. Oddly, the political geographer Halford Mackinder who was far from progressive in his political views, understood this clearly, as he demonstrated in his well-known book, Democratic Ideals and Reality (Cox 2019.) In this regard, and more concretely, the capitalism of geopolitics needs to be written to take into account a struggle that goes back to 1917, and the Russian Revolution. This was followed by abortive revolutions in Germany and Hungary and by a wave of militant strike activity across the capitalist world. These were seminal events that sent shockwaves across national ruling classes. The problem became: How to turn back the red tide?

This was the background to what has been called ‘Europe’s civil war’ (Graham 2005; Traverso 2016): a struggle by the ruling class to reimpose its dominance and played out most vigorously in Franco’s triumph and subsequent bloodletting in Spain and in the repression of the labor movement by other fascist countries. Meanwhile, the ruling class in France and Great Britain was quite happy to allow this to happen. Working class resistance would continue during the war through the communist-dominated resistance movements. It would only be finally settled, at least for the time being, with the end of the civil war in Greece in 1949: again we should have no doubt which side ruling classes elsewhere, including the British and the Americans actively supported.

This is not to argue that the Second World War was not also about global hegemony: not just countering the threat of Germany and Japan to the US’s pretensions but also the US putting a nail in the coffin of the British Empire through the way it negotiated its war aid deals: no shared anglophone sentimentality there! Germany led the charge against the USSR but Britain and the US were fine with that; it was the long term implications of German victory for domination of the global economy that worried the Americans.

Once the war was over, the struggle against the working class of the world would continue in the form of the Cold War: the attempt to push back the so-called ‘evil empire’ and save capitalism for the world. Countries like Cuba, El Salvador and South Africa would now be accorded heightened importance in geopolitical calculations. Why capitalism prevailed in this particular struggle, aside from the huge advantage in material resources that it was able to deploy, is complicated. While globalization and the roll-back of the state everywhere would play a role (Clarke 1990), the working class in the developed world had already been softened up by changes in family form (Cox 2020), the collapse of the industries that had been at the heart of the labor movement, and the great shifting right show of their parliamentary representatives.


Since this was a paper written for a conference with considerable student presence, I am concluding it with what the various discussions imply for concrete research possibilities. What in my view, might be some priorities?

First, countries matter. They matter, as we have seen, in understanding the geopolitics of capitalism. They also matter through respective structured coherences, which aside from their capitalist nature, are always different. There is major scope here for comparative study. I am impressed by some of the similarities between South Africa and Bolivia: both resource dependent, both intensely racist pitting Europeans against indigenous peoples, both with a strong presence of migratory labor, and both with a history of territorial struggle that tends to occlude class interests. So what might one learn from looking at Bolivia through a South African lens and vice versa? What research questions of a more specific nature might they raise? Alternatively what difference does a structured coherence make to the way in which migratory labor or land reform struggles work?

Second, even while countries are a dominant presence in world political geography, I am impressed by the scale indifference of the processes that I have been discussing. The competing poles of territory and class are as apparent within metropolitan areas as at the national level. I remarked on the scale indifference of geopolitical processes. One talks about global hegemony, but there are more sub-continental sorts of hegemony: think South Africa in the context of Southern Africa and Mexico in Central America. There are regional hegemonies within countries to be examined: the predominance of London and Southeast England in the United Kingdom is currently highly controversial. The geopolitics of metropolitan areas, particularly where they are jurisdictionally fragmented, as in the US, then provides a focus rich in its implications.

Finally, of course, there is Brazil. To refer back to the question of regional hegemony, what difference does it make that Brazil is a lusophone country in a hispanophone sea? There is a geopolitics of climate change, so what sense can one make of the country’s insistence on the clearing of the forests of Amazonia? What does this say about the strength of the Brazilian state, or otherwise, to organize its space? What does it say about the Brazilian state and its relation to those other keywords of political geography, territory, country and geopolitics?


[1] “What is in fact called for is a new type of political geography, dealing with power in space-time terms of considerable conceptual precision. Power relations ae of such immense importance for the understanding of how projects compete in available budget-space that a well-conceived political geography could well develop into the core of human geography. [2] For example, the title of the book edited by Allen Scott and Michael Storper (1986) Production, Work, Territory: The Geographical Anatomy of Industrial Capitalism. [3] Not least: an emancipation of industry from coalfield locations; and within metropolitan areas, the ability to have a factory on one floor only, which meant the possibility of locations on the edge of the city where land was/is cheaper. [4] Albeit, often in kind. [5] As for instance the way in which South Africa surrounds Lesotho. [6] The fact of Alaska and Hawaii make the US a bit peculiar in that regard. [7] See Cox (2016: 300-318) for a more developed discussion. [8] “After the mid-1960s, then, the metropolitan area seemed to be composed of a mélange of service and economic areas, varied in size, which existed to foster pursuit by individuals of personal goals and objectives. Each of these areas competed for economic resources, power and ‘top-notch’ citizens, and the competition not only pitted Cincinnati against its suburbs but also big city neighborhood against big city neighborhood, suburb against suburb, and neighborhoods within a particular suburb against one another. Each of these communities, in other words, comprised a community of advocacy. And the larger units, such as Forest Park or Cincinnati, constituted a community of advocacy made up of smaller communities of advocacy” (Miller 1981: 239.) [9] The French national anthem, the Marseillaise, is different, composed during the revolution and reflecting the tensions of the time; even so, ‘patrie’ or fatherland does put in an early appearance. [10] Compare Hobsbawm (1990:3): “Nations, we now know … are not, as Bagehot thought, ‘as old as history.’ The modern sense of the word is no older than the eighteenth century, give or take the odd predecessor.” [11] A widely followed BBC football program. [12]


Clarke S (2020) Crisis of socialism or crisis of the state? Capital and Class No.42, 19-29.

Cox K R (2014) Making Human Geography. New York: Guilford.

Cox K R (2016) The Politics of Urban and Regional Development and the American Exception.

Syracuse NY: Syracuse University Press.

Cox K R (2017) Revisiting the city as a growth machine. Cambridge Journal of Regions,

Economy and Society 10, 391-405.

Cox K R (2019) A Mackinder centenary and a revisionist note. Available at Unfashionable


Cox K R (2020) Geographies of decline and the labor movement. Available at Unfashionable


Elden S (2013) The Birth of Territory. Chicago: Chicago University Press.

Graham H (2005) The Spanish Civil War: A Very Short Introduction. Oxford: Oxford University


Hägerstrand T (1973) The domain of human geography. Chapter 4 in R J Chorley (ed.)

Directions in Geography. London: Methuen.

Harvey D (1985) The geopolitics of capitalism. Chapter 7 in D Gregory and J Urry (eds.),

Social Relations and Spatial Structures. London: Macmillan.

Harvey D (2006) Space as a keyword. Chapter 14 in N Castree and D Gregory (eds.), David

Harvey: A Critical Reader. Oxford: Blackwell.

Hobsbawm E (1990) Nations and Nationalism Since 1780. Cambridge: Cambridge University


Jessop, B (1990) State Theory. Cambridge: Polity Press.

Marris P (1975) Loss and Change. Garden City NY: Doubleday Anchor.

Marx K (1857-58. Republished 1973) Grundrisse Harmondsworth, Middlesex: Penguin Books.

Miller Z (1981) Suburb: Neighborhood and Community in Forest Park, Ohio, 1935–1976.

Knoxville: University of Tennessee Press.

Traverso E (2016) Fire and Blood: The European Civil War, 1914–1945. London: Verso.

Wade R and Veneroso F (1998) The Asian crisis: the high debt model versus the Wall

Street-Treasury-IMF Complex. New Left Review No.228, 3-24.