Changing cartographies of health in a globalizing world

Anthropologists have described, often in eloquent detail, local destruction of opportunities to lead a healthy life (the social determinants of health) attendant on the macroscale economic processes conveniently described as ‘globalization’. Recent reorganizations of production and finance redraw maps both literal and metaphorical of the inequalities that affect health. I argue that it is essential to focus attention on the common origins of such local destructions in new modalities and power structures of global capitalism, and in doing so to focus on what William Robinson has described as a shift from ‘territorial’ to ‘social cartographies’. These include a number of cross-border ‘emerging markets’ or bidding wars that are relevant to health and its social determinants. The article sets out three propositions about how the social science of health disparities should respond to globalization, emphasizing possibilities for research on globalization and health that draw on the complementary perspectives of anthropology and political economy.


Introduction
Over the past decade, my research has focused on the implications of globalization for health disparities. Definitional controversies persist as they will in the academic world, but the description of globalization as a 'pattern of transnational economic integration animated by the ideal of creating self-regulating global markets for goods, services, capital, technology, and skills' (Eyoh and Sandbrook 2003, 252) suffices here. Although this definition refers to 'self-regulating' markets, these class struggle as the only resolution to urban problems'. She claims that '[b]y studying situated phenomena through a lens that understands them as singular moments in a unified and integrated global process, analysts lose sight of complex urban situations as particular engagements with the global' (Ong 2011, 2). Now, the Marxist models she describes in stylized form are something of a caricature, at least as such models are deployed by serious social scientists. The point is important because much analytical clarity, and many opportunities for engagement and resistance, can be lost if we neglect the common origins of numerous metropolitan situations in a 'unified and integrated global process'. For example, an important analysis pointed out that contests over the occupation and development of metropolitan space in countries rich and poor alike reflect an economic dynamic in which 'the prime resources of the city are increasingly appropriated by the affluent. And globalization is inflationary as the new rich are able to pay more for a range of key goods, especially land' (United Nations Human Settlements Programme 2003, 52). Another illustration of the value of a focus on political economy, discussed further below, is provided by the concept of labour arbitrage -adopted from a trade journal by Ong (2006, 161) herself to describe, with remarkable succinctness and clarity, a key driver of global reorganizations of production.
The second part of my core argument is captured by political sociologist William Robinson (2002), a leader of the emerging transdiscipline of critical globalization studies (see Appelbaum and Robinson 2005), in his insistence that development now must be understood with reference to 'social' rather than 'territorial cartographies'. These are defined by the relations of people living within a specific set of political boundaries, and often in close proximity, to 'new globalised circuits' of production and capital accumulation (Robinson 2002(Robinson , 1060. Urbanist Saskia Sassen (2012, 323-329) similarly argues the need for a new geography of centres and margins. Whatever the chosen terminology, this perspective expands the first part of my argument with more specific reference to the conditions of life and work that affect health, and is elaborated upon in the sections of the article that follow. I first describe global (or at least transnational) reorganizations of production and finance that are now rather familiar. In cartography, 'isolines' connect points of equal value on a map, such as points of equal elevation on contour maps or barometric pressure on weather maps. In the twenty-first century, it is useful to think in terms of literal and metaphorical maps in which isolines connect individuals, households, and communities based on the terms -analogous to the terms of trade between countries -on which they are connected (or not) to global flows and transactions. Those isolines may cross multiple national borders, even as they define and describe a particular set of opportunity structures that are only loosely, if at all, dependent on national location. I then describe a number of cross-border 'emerging markets', 2 including the competition for metropolitan space, which are relevant to health and its social determinants. I conclude with three propositions about how the social science of health disparities should respond to globalization, in a way that I hope will encourage reflection on conceptual synergies and transdisciplinary topical collaborations between anthropology and political economy.

Global reorganization of production transforms labour markets
A key driver of, and motivation for, global reorganization of production is labour arbitrage: 'the ability to pay one labor pool less than another labor pool for accomplishing the same work, typically by substituting labor in one geography for labor in a different locale' (trade journal article quoted in Ong 2006, 161;see also Roach 2006). This is by now a relatively familiar story, since documentation of a new international division of labour dates from the late 1970s (Fröbel et al. 1980;Vaughan 1978). As Ong (2006, 160) correctly observes, '[t]he global search for cheap labor in manufacturing … can be called a kind of industrial labor arbitrage that operates according to the logic of same skills, different pricing.' Opportunities for labour arbitrage multiplied as China, India, and the countries of the former Soviet bloc integrated into the global economy, roughly doubling the size of the global labour force (Freeman 2007). Those opportunities were and are underpinned by a legal infrastructure of bilateral, regional, and multilateral trade and investment agreements. They are multiplied and enhanced by technological and managerial innovations that enable a transition from trade in goods to trade in tasks (Grossman and Rossi-Hansberg 2006), many of which can be performed in a variety of locations and quickly relocated. Critically, the transition from trade in goods to trade in tasks is no longer confined to manufacturing, but extends far outside it as a consequence of '[t]he new susceptibility of cognitive functions to a computerized division of labor' (Ong 2006, 161) . 2 The term 'emerging markets' was coined in 1981 by a senior official of the International Finance Corporation, the commercial arm of the World Bank, 'as a way to change the financial world's perception of developing countries' and increase the flow of foreign direct investment to such economies (International Finance Corporation 2011); it is used here in quite a different sense.
Workers in low-and middle-income countries (LMICs) today are routinely exposed to hazards similar to those that were eliminated from most workplaces in high-income countries generations ago. This point temporarily came to broader public attention early in 2012 after the appearance of a New York Times series (Duhigg and Barboza 2012; Pouille 2012) on the plants in China where Apple products are produced under contract by the Taiwanese-owned company Foxconn Technology (although the companies reported upon are not the only or the worst offenders). An evocative parallel can be found in the garment industry. When 148 workers died in the Triangle Shirtwaist Factory fire in New York in 1911, the resulting public outcry and impetus for labour organizing eventually led to improved safety standards. Today, 'middle-class' consumers still buy garments made by workers at risk of dying when they are locked into factories, but a century later the factories and the fires are half a world away in locations like Bangladesh, which is now (after China) the world's second-largest exporter of clothing (Brooks 2007, 85;Yardley 2012). In one recent case US retailer Wal-Mart, a major customer of a factory where more than a hundred workers died in a fire, had apparently resisted initiatives to improve worker safety because of their cost (Greenhouse 2012); it was able to do this because of the market power it exercises over suppliers from its position at the top of commodity chains that cross multiple national borders. Indeed, the operations of Wal-Mart perhaps more than those of any other single firm illustrate how globalization 'connects sales clerks in Bentonville to truck drivers at the U.S.-Mexico border and line workers in Guangdong' (Gereffi and Christian 2009, 580-81; see also Bair and Bernstein 2006;Vidal 2012), with far-reaching consequences for work conditions and living standards.
Such restructuring of labour markets and associated economic opportunities is central to Robinson's call for studies of development to transition from territorial to social cartographies. He writes: 'A global division of labour suggests differential participation in global production according to social standing and not necessarily geographic location, and accounts for sweatshops in East Los Angeles and Northern Honduras, as well as gated communities in Hollywood and São Paulo. Social order is increasingly organised globally, not nationally' (Robinson 2002(Robinson , 1067. Thus, globalization has increased the incomes of the richest one percent of the world's population, but also of those in the middle strata of China's income distribution, while generating 'stagnation of incomes among the poor people or lower middle classes in rich countries' (Milanovic 2014, 80). The changing employment relations that have contributed to this effect in rich countries are exemplified by substitution of rural home-based piecework for unionized assembly line work in the US auto industry (Gringeri 1994), and what DiTomaso (2001) has described as 'the subcontracting of everyone'. Further support for Robinson's analysis is provided by the emergence of stereotypically 'third-world' labour market arrangements and working conditions in countries like the United States, as Ong (2006, 123) points out with reference to 'the return of post-Fordist sweatshops to North America'. Her concern here is with southeast Asian immigrants in northern California, where 'Asianmanaged high-tech networks depend on ethnic enclaves where carceral labor practices shape a unified Asian-American ethnicity,' but she notes that '[w]hile the 'face and practices of electronics manufacturing are Asian, the capital that drives their location and relocations is mainly Wall Street' (Ong 2006, 135). Like the opportunities for labour arbitrage presented by offshoring, the domestic availability of this population arguably depresses the wages of a substantial portion of the documented workforce and limits the prospects for collective action.
Living conditions as well as working conditions can be stereotypically third world. An interviewbased study of undocumented Mexican immigrants living in New York City, which in many respects constituted a reprise of a New York Times series published more than a decade earlier (Alvarez 1996;Bruni 1996;Bruni and Sontag 1996;Sontag 1996), described a 'basement shared by eighteen single men, an unheated garage that is home to two Mexican families, an abandoned tractor trailer … and an abandoned tugboat and barge, both vessels half in and half out of the water, and each providing housing to another two or three men' (Hellman 2008, 159). Instead of thinking in terms of national labour markets, it is more useful to think in terms of nonterritorial divisions among integrated, precarious, and excluded forms of employment (Cox 1999) that exist in countries rich and poor alike, with important consequences for disparities in health status. The recent tendency of some firms to return work to high-income countries, or to territories closer to them (Booth 2013), as labour costs rise (because of localized shortages rather than collective action, which often remains hazardous to those involved) and as the difficulties of long-distance management become apparent, actually underscores this point by reflecting the locational flexibility and fluidity of contemporary production against the background of downward wage pressure in the high-income world. In the words of a prescient Economist survey that appeared before the 2008 financial crisis, 'the usual argument in favour of globalisation -that it will make most workers better off, with only a few lowskilled ones losing out -has not so far been borne out by the facts. Most workers are being squeezed' (Woodall 2006, 6).

Finance and new contours of marginalization
Since the start of the 1980s, financial markets as well as production have been reorganized across borders as a consequence of four interacting processes. First, because of advances in information processing and telecommunications, vast sums of money can be moved around the world almost instantaneously. Second, the United States and the United Kingdom competed to deregulate the domestic and overseas activities of their financial services industries (Girón and Correa 2002;Helleiner 1994). These two processes were essential to a third trend, which has been characterized as financialization: 'a pattern of accumulation in which profit making occurs increasingly through financial channels rather than through trade and commodity production' (Krippner 2005, 174).
Fourth, by the start of the 1980s the precarious state of many LMIC economies threatened the financial health of the institutions that had lent heavily to them (Makin 1984 (1994)(1995), south Asia (1997-1998), and Russia (1998, where the resulting economic devastation exposed millions of people to poverty and economic insecurity. In the aftermath of the Mexican financial crisis, the managing director of the IMF at the time noted the importance of 'market perceptions: whether the country's policies are deemed basically sound and its economic future, promising', conceding (or perhaps threatening) that 'shifts in the market's perception of these underlying fundamentals can be quite swift, brutal, and destabilizing' (Camdessus 1995).
Capital markets therefore have 'the power to discipline national governments. … These markets can now exercise the accountability functions associated with citizenship: they can vote governments' economic policies in or out, they can force governments to take certain measures and not others' (Sassen 2003, 70). The financial crises that result from such exercises, which have become increasingly frequent and severe since 1980, albeit not in a linear fashion (Laeven and Valencia 2013), magnify economic inequality in several ways. The effects of initial job losses and wage declines (Diwan 2001;World Bank 2000, chapter 2) are compounded by the fact that employment consistently recovers more slowly than economic output in the aftermath of financial crises crises (van der Hoeven and Lübker 2006; Walton 2009). Public bailouts of financial institutions that commonly follow financial crises transfer wealth from a broad segment of the population (the national tax base financing the bailout) to a minority of households with substantial deposits, debtors whose loans are subsequently forgiven, and financial institution shareholders (Halac and Schmukler 2004). The small minority of the population owning mobile assets can take advantage of opportunities for capital flight, transferring assets abroad 'to evade domestic social control' in the form of taxation, regulation, or devaluation (Ndikumana and Boyce 1998, 199). In a world in which private banking and tax havens flourish ('How BCCI Grew and Grew' 1990;'Private Banking: Rich Pickings' 2006;Shaxson 2011), 'high net-worth individuals' 3 enumerated in the annual World Wealth reports (Capgemini and RBC Wealth Management 2013) who live in LMICs enjoy broadly the same portfolio choices and consumption options -albeit with a need for greater attention to personal security -as their counterparts in London or Los Angeles. This trend is not restricted to financial instruments, with wealthy Brazilians for instance seeking out expensive real estate in New York (Barrionuevo 2012). Finally, a recent study of seventeen high-income economies found that the 'fiscal consolidation' (austerity) that follows financial crises decreases the share of national income paid out in wages and salaries, while increasing that of capital, and increases long-term unemployment (Ball et al. 2013).
The financial crisis that spread across the world in 2008, as the collapse of the market for securitized, high-risk mortgages revealed 'underappreciated, but potent, interconnections between firms in the global financial system' (Bank of England 2008, 9), further transformed the social cartography of a financially globalized world. Financial institutions in the United States (especially) and the United Kingdom had aggressively fought back against domestic regulation in order to create new financial products that extracted resources from working-class households (Immergluck 2011;Sassen 2009). When the markets for these products collapsed, trillions of dollars in cash and credit guarantees were mobilized to bail out financial institutions (Bank of England 2009), while the costs were borne first and worst by people who had no role in precipitating the crisis and no control over its outcome. In LMICs, the initial response of social protection systems was described as 'limited' and inadequately directed toward the most vulnerable (McCord 2010), and a 2011 UNICEF review found that after an initial period of expansionary fiscal policy many LMICs were considering such austerity measures as limiting or rolling back wages for teachers and health workers, retrenching existing social protection transfers, and increasing consumption taxes on basic goods like food (Ortiz et al. 2011a).
3 Individuals with more than US$1 million in liquid, investable, financial assets.
Elsewhere, the combination of reduced government revenues with heavy borrowing to finance stimulus spending contributed to subsequent sovereign debt crises that as of mid-2012 threatened several Eurozone economies; for some high-income countries, the accumulation of debt is likely to create long-term fiscal policy constraints that limit the space for social protection (Schäfer and Streeck 2013). By early 2012, officially reported unemployment had risen to 14 percent in Ireland, the former Celtic Tiger, and to more than 20 percent in Spain (50 percent  At the epicentre of the crisis in the United States, by late 2011 unemployment had more than doubled from its pre-crisis low, and official figures may represent a serious underestimate (Doyle 2011;Western and Beckett 1999). A largely invisible cohort of as many as fourteen million US households faced dispossession by foreclosure, a substantial proportion of them consisting of renters (Sassen 2011). By mid-2013, more than forty-seven million people -roughly one out of every seven US residents -were receiving federal government food vouchers (Supplemental Nutrition Assistance, or food stamps) and millions more were eligible (Food Research and Action Center 2013). Meanwhile, and critically, in 2011 there were more high net-worth individuals in nine out of the ten largest US metropolitan areas than in 2007; the exception was deindustrialized Detroit, which saw a decline but only a slight one (Capgemini 2012). In short, the financial crisis magnified the ratcheting-up of intranational economic inequality associated with the cross-border reorganization of production and finance. As the editor of Le Monde Diplomatique has written: 'The inequality machine is reshaping the whole planet' (Halimi 2013).

'Emerging markets' of a different kind
The operations of the inequality machine juxtapose spreading economic insecurity with accumulations of wealth that isolate their owners from reliance on all but the most basic institutions of collective or social provision, and sometimes even from those. Such juxtapositions are especially visible in so-called world cities, where the command and control functions of the global economy generate a parallel demand for legions of low-wage service workers who clean the buildings, drive the taxis, cook and deliver the restaurant meals, and often provide child care (see, for example, Abu-Lughod 1999, 285-320;Sassen 2012, 241-272); exploiting the workers who provide these services can in turn offer new opportunities for capital accumulation, for example in contracted-out service provision (Winerip 1998). The inequality machine is also associated with multiple cross-border markets in which the rich now outbid others for resources of various kinds or take advantage of opportunities presented by differences in income and wealth.

Gentrification
The first of these emerging markets involves contests over urban or metropolitan space: gentrification, broadly defined (Sassen 1991, 255;Slater 2011). We associate gentrification in the first instance with high-income cities, but it has now spread to numerous LMIC metropolitan areas, where the largest-scale variants are referred to by one author as 'megagentrification' (Lees 2012) and by another, in the case of Mumbai, as 'class cleansing' (Whitehead and More 2007). After economic liberalization and a dramatic increase in foreign direct investment, in the mid-1990s portions of Mumbai's central business district temporarily became 'the most expensive in the world'; residential real estate prices, and therefore opportunities for profit, likewise increased (Grant and Nijman 2004; see also Nijman 2000) -in a city where it was estimated contemporaneously that '[m]ore than half' of the population lived 'in slums and on pavements or under bridges and near railway tracks' (Patel 2007, 76).
The increased wealth of the winners from globalization, those well placed for a variety of reasons to benefit from new economic opportunities, has led to the emergence of real estate investment and provision of associated infrastructure and financing as a major theatre of capital accumulation for both domestic and transnational investors outside the high-income world, where it had earlier become well established. In cities as diverse as Accra, Beirut, Santiago, Johannesburg, and Bangalore, the prospect of real estate profits has attracted capital to residential and commercial projects oriented toward corporate clients and rich individuals (Goldman 2011;Grant and Yang 2009, 18-89;López-Morales 2010;Makdisi 1997;Murray 2008, 191-209). Of special note are the profits to be made in many metropolitan areas from major entertainment and sporting mega-events such as the Olympic Games and Grand Prix auto races, which can be highly profitable (Lefebvre and Roult 2011;Schausteck de Almeida et al. 2013;Scherer 2011). Along with certain kinds of infrastructure and amenities associated with such events or oriented to the needs of the rich, such spectacles are viewed, or at least justified for public consumption and for purposes of legitimizing public subsidies, as broadly beneficial because they brand a city as a world-class location suitable for foreign investment and attracting wealthy tourists.

Health workers and health care
Another emerging market involves the mirror-image processes of health worker migration, to either higher income countries or the well-financed private sector within their own countries, and medical tourism, or the purchase of health care outside one's own country. 4 A 2004 study concluded that one million more health professionals would be needed in sub-Saharan Africa alone by 2015, explaining the deficit with reference to AIDS, labour migration, and underinvestment in health systems over '[t]wo decades of economic and sectoral reform' (Chen et al. 2004; see also Connell et al. 2007). As  (Korir and Kioko 2009, 2). The attraction of higher remuneration and better working conditions is understandable, as is the desire of health professionals' to escape low and precarious pay, working conditions that are unimaginable in highincome countries, and pervasive threats to personal safety (Connell et al. 2007;Dugger 2004;Sachs and Sachs 2004;Schrecker and Labonté 2004). Nevertheless, some observers regard the recruitment of health-care workers from sub-Saharan Africa, the world's poorest region, as 'an international crime' because of the debilitating effects on national health systems there (Mills et al. 2008).
While health professionals leave LMICs, growing numbers of patients seek medical care outside their countries of residence. They are motivated variously by long waiting times, costs, and -in the case of wealthy LMIC nationals -quality of care (see, for example, Alsharif et al. 2010;Connell 2011). The territorial complexity of this process is suggested by the fact that, while US hospitals seek wealthy foreign nationals able to pay full market prices in order to boost their bottom lines (Freudenheim 1996;Zhao 2001), under-or uninsured US residents who cannot afford treatment are finding options in Mexico or more remote locations such as India, where prices may be a fraction of those in the United States (Cortez 2008;Dalstrom 2012;Gupta and Das 2012;Turner 2007).
Whatever their country of residence, patients with world-scale incomes and wealth can and do comparison-shop for health care among multiple jurisdictions (Herrick 2007;Turner 2007). This practice is actively promoted by private hospitals in some destination countries using strategies that include international accreditation, partnerships with globally recognized institutions in high-income countries, and offering a variety of 'concierge' services and hotel-style amenities. Simultaneously, medical brokerages in countries like the United States 'bring the search for healing into the era of global capitalism' (Turner 2007, 307).

Organ trade and clinical research
Related to medical tourism is the market for human organs -the topic of urban legends, but now documented beyond serious dispute by anthropologist Nancy Scheper-Hughes (2003, whose extensive fieldwork is foundational. She describes kidney sellers as 'stranded Moldovan and  (Shimazono 2007). It is also complex and inherently globalized, involving recruiters, brokers, and transplant surgeons in multiple countries and requiring at minimum a certain level of official indulgence.
Clinical research has likewise become globalized, with the rise of private, for-profit contract research organizations (CROs) and the emergence of sites outside high-income countries as preferred locations for clinical trials (Glickman et al. 2009;Petryna 2009). The attractions of such locations start with lower cost: a 2005 article noted that running a trial in Romania costs one-tenth as much per patient as in the United States, and quoted a pharmaceutical CEO as saying that '[g]lobalization is the ultimate arbitrage for companies' like his (Lustgarten 2005, n.p.); note the parallel with the concept of labour arbitrage. Developing and transition countries offer more patients, and sicker patients, than are available in the high-income world -an attraction that will be magnified by the rapidly growing prevalence of noncommunicable diseases in such economies (World Health Organization 2010): the so-called double burden of disease. Recruitment is facilitated by lack of access to care, for all but the rich, outside the environment of the clinical trial -problematizing, to put it mildly, the nature of informed consent when the usual standard of care is no care at all (Angell 1997;Kent et al. 2004).

Land grabs
Especially after 2007, agricultural land became the object of cross-border bidding wars as foreign investors and the governments of food-importing countries engaged in large-scale purchases or long-term leases of agricultural land ('land grabs', a widespread term used here for convenience; for overviews see White et al. 2012;Cotula et al. 2011;De Schutter 2011). These transactions occurred against a background of, and partly in response to, rapid food price increases in 2007-2008 that threatened to undo much of the modest progress in reducing undernutrition and food security in Four points about land grabs are especially important. First, many land grabs are also not only about land, but also about the accompanying access to water resources (Mehta et al. eds. 2012;Rulli et al. 2013); this implies an expansion of a 'virtual water trade' Tamea et al. 2013) that is already quite substantial. Second, whether as a reflection of 'unwavering faith in the role of foreign investment in national economic development' (German et al. 2011, 2) or because of potential gains for politically connected local elites (Ansoms 2011;Borras and Franco 2012), governments in countries where land is acquired are routinely involved as facilitators and promoters of the transactions. Such transactions are often negotiated directly between high-level protagonists, completely bypassing local actors (Cotula 2011;Sadeque 2009;Vermeulen and Cotula 2010), even though they may be displaced in large numbers. Third, the dynamic cannot be understood in terms of conventional North-South distinctions, because of what has been described as an 'emerging 'North-South-South' dynamic' (Land Deal Politics Initiative 2012; see also Rulli et al. 2013).
Governments of many non-Northern countries (including China, South Africa, and several countries in the Middle East) or companies based in those countries are actively involved in land acquisitions; companies based in Brazil and India, among other countries, are originating acquisitions while at the same time land in those countries is being acquired by foreign actors. 5 Fourth, and most fundamentally, the underlying political dynamic is one of 'wide-ranging global "land reform" -in this case, a regressive land reform where governments take land from the poor and give (or sell or lease) it to the rich' (White et al. 2012, 620). The irony is especially bitter for those old enough to remember earlier prescriptions for land reform as a partial remedy for catastrophic levels of domestic inequality in economies where agricultural production still supported a substantial proportion of the population and land ownership was highly concentrated.

Thoughts on research directions: Bringing class back in
The preceding observations suggest three more general propositions about the social science of health disparities and their underlying causes. First, at the risk of being repetitive, globalization restructures social relations in a way that breaks down familiar territorial divisions and rescales the politics of issues like the distribution of wealth and poverty and the locus of control over economic activity. Convergence of national per capita incomes, a result of the strong performance of a number of LMIC economies after the 2008 crisis, in many cases distracts attention from growing structural inequalities within national borders in access to the various prerequisites for leading a healthy life, and the local manifestations of those inequalities. Worldwide, it is estimated that circa 2008 roughly three-quarters of the world's poorest people (based on the World Bank's $1.25/day threshold of extreme poverty) lived in countries where per capita incomes were above the threshold conventionally used to define a low-income country (Sumner 2012). In an extreme example, India's more than 400 million extremely poor people coexist with 153,000 high net-worth individuals (Capgemini and RBC Wealth Management 2013), in a class structure described by Arundhati Roy (2012, n.p.): 'India's 100 richest people own assets equivalent to one-fourth of the GDP. … The 300 million of us who belong to the new, post-IMF "reforms" middle class -the market -live side by side with … the ghosts of 250,000 debt-ridden farmers who have killed themselves, and of the 800 million who have been impoverished and dispossessed to make way for us.' One cannot of course generalize from this example, and in some middle-income countries economic inequality has been declining, albeit from extremely high levels (Lustig et al. 2012). Nevertheless, profound scepticism is in order about the ability of the rising tide of growth through integration into the global marketplace to lift all boats, even in countries where growth is rapid and sustained. This point has been conceded even by the World Bank with respect to the operation of labour markets (World Bank 2007, 67-100).
This leads to a second point: class -in the Marxist sense 6 -as an analytical category remains relevant, even indispensable across multiple scales and contexts: 'Marxism would be abandoned at too high a cost, given the contribution of its method of abstraction to identifying the generative mechanisms and emergent causal powers associated with capital as a social relation and commodification, which can be viewed as the central articulating principle of the contemporary global political economy' (Ryner 2006, 63). Jon Elster (1986), Leslie Sklair (2005), and Erik Olin Wright (1985), among others, working from diverse methodological starting points, show that classes are not abstractions isolated from individual agency, but (conversely) that agency does not operate in isolation from material situations. Contemporary globalization means that those situations are increasingly defined by economic processes that operate across multiple national borders. This holds whether we are discussing the owners and managers of financial assets whose portfolio choices collectively constitute the verdict of 'the markets'; gentrifying households and real estate capitalists who use a combination of private and state power to displace those who stand in the way of 'higher value uses' of urban locations; city dwellers who claim that the poor have 'infested everything' and favour 'a bottle of kerosene and a match' for the favelas; 7 subsistence producers who are displaced by land grabs; or the expanding proportion of the working class in high-income countries who are trapped in low-wage, insecure, service-sector jobs, going deeper into debt and facing long-term post-crisis unemployment and a gradual drift into unemployability, perhaps compounded by subprime loans and foreclosure (see, for example, Ehrenreich 2001; Hodson et al. 2014;Leicht and Fitzgerald 2014). These idiosyncratically chosen examples suffice to show the 6 There are, of course, multiple Marxist definitions of class; the distinctions among them are outside the scope of this article, but they share a focus on the productive process and on property rights. Elster's (1986, 147) definition of a class as 'a group of people who by virtue of what they possess are compelled to engage in the same activities if they want to make the best use of their endowments' will suffice. The key point is that Marxist conceptions of class must be distinguished from conceptions of class rooted in the work of theorists like Pierre Bourdieu, which are 'linked … to the interplay between economic, social and cultural capital', as exemplified by the British Broadcasting Corporation's 'Great British Class Survey' (Savage et al. 2013). 7 These quotations are taken from a long interview with the wife of a Brazilian real estate agent that opens Teresa Caldeira's (2000) superb study of São Paulo, City of Walls. degree of nuance that is compatible with Marxist conceptions of class, and indeed essential to their contemporary application to social scientific explanation. Political commitments and allegiances cannot be read off class position in a mechanistic way, but at the same time class position cannot be ignored in understanding those commitments and allegiances.
Third, the role of the state and of public policy must be (re)considered in light of these first two propositions. Observations about the distinctive nature of the power of capital (or in some formulations business; see Lindblom 1977) are hardly new, but it cannot be emphasized too strongly that globalization fundamentally strengthens the position of capital because of the opportunities for exit (Hirschman 1970) associated with the reorganization of production across borders and the shift of power from governments to financial markets. The extent of the policy space that remains available to particular national governments is an ongoing debate in political science and international relations, and obviously depends on the policy context and the position of the country in question vis-à-vis other actors in the global environment. Importantly, the state and elites who are defined by their connection to state power have often advanced globalization, rather than simply responding to it, even at the cost of shrinking policy space. Sandra Halperin (2013, 224) states the point: 'The discourse of globalization emphasizes the necessity of governments to adapt to newness and difference, a necessity that forecloses choice. But government policies are designed, not to adapt to new circumstances, but to promote them.' Alice Amsden (1990, 18) memorably argued that the 'discipline of labour by the state lies at the heart of all late industrialization'. Although that discipline remains an imperfect process, it is often used proactively to sustain low labour costs and flexible employment regimes that are attractive to foreign direct investors and managers at the top of global commodity chains in firms like Wal-Mart. It should not be forgotten that China's remarkable industrial revolution has been achieved under conditions of continuing authoritarian rule. It is also possible to interpret retrenchment in social provision, especially when combined with intrusive workfare regimes -a combination exemplified by 1996 welfare reforms in the United States (Wacquant 2009, 76-109) -as an effort to discipline labour in the high-income world and part of a larger neoliberal project. In other manifestations of the state's proactive role (roll-out neoliberalism), land grabs could not proceed without acquiescence from the states in which they occur, and often such states actively promote the acquisitions in question. State strategies to support gentrification and real estate capitalism that involve varying degrees of coercion have been deployed in jurisdictions across the globe and the income spectrum.
For example, Indian governments have been actively clearing out low-income people who are in the way of 'higher value' uses like commercial offices and technoparks. An official of the Maharashtra state government has been quoted as saying that: 'With the [Mumbai] slum demolitions, we showed political courage for the first time and sent a strong signal that you cannot expect free space in this city anymore' (Lakshmi 2005). Large-scale sports events, in particular, frequently require major relocations of populations (Burke 2010;Centre on Housing Rights and Evictions 2007;Fowler 2008;Romero 2012). Beijing is reckoned to have displaced more than a million people for the 2008 Olympics, and large-scale displacement of residents in spite of public protests was associated with Brazil's commitment of state resources in preparation for 2014 World Cup and the 2016 Olympics (Broudehoux 2013;Romero 2012;Saad-Filho and Morais 2014;Watts 2013). This is far from a complete inventory of state strategies to facilitate and accelerate globalization -it leaves out, for example, the entire field of trade policy -but it is enough to make the point.
Here, again, an important analytical contribution comes from an anthropologist. In a discussion of Indian state policy with respect to large dam and urban transport projects funded by the World Bank, Shalini Randeria (2007; see also Randeria and Grunder 2011) offers the concept of 'cunning states' that 'lack the political will, rather than the space, for autonomous policy making' (Randeria 2007, 6). Cunning in this case was deployed in order to limit accountability both to international institutions like the World Bank and to India's own citizens, while forced displacement, internal resettlement, and expropriation compensated at artificially low values continued. Randeria's failure, however, to connect state actions with specific class interests -in contrast (for example) to Goldman's (2011) exploration of high-end real estate developments and the construction of the requisite subsidised airports and highways in Bangalore -limits the value of her analysis despite its descriptive richness and the importance of the cunning state concept. In other words, cunning and the absence of political will that leads to its exercise represent a phenomenon that demands explanation (explanandum), rather than a source of explanations (explanans).
Post-crisis austerity programs foreground this point. The long-term arithmetic of failing to balance budgets and reduce sovereign debt cannot be ignored. At the same time many governments arguably retain considerable room to select among options for addressing deficits by way of expansionary vs.
contractionary fiscal policy, in other words, by raising revenues instead of reducing expenditures, and within the latter category choosing among expenditure reductions with different distributional impacts. In the past, cunning appears often to have been deployed as cover for the existing preference of state elites, as shown by archival research to have happened in the earlier case of British responses to the sterling crisis of 1976 (Rogers 2009). Further research will be valuable in generating understandings of the politics of health inequalities by revealing the extent to which the financial crisis in effect provided valuable cover for the neoliberal project, cloaking it (and the associated negative health effects; see Marmot et al. 2013;Stuckler and Basu 2013) in an aura of inevitability. To understand how and why these choices are made, situationally specific alliances between state elites and particular classes, within the context provided by globalization, assume special importance.
Social scientists today tend to be sceptical about so-called grand narratives. Good reasons may exist for this scepticism, yet analytical approaches organized around class and class conflict can offer considerable explanatory power even in the absence of commitment to such narratives. The evidence presented here suggests that recognizing the common origins in macroscale social processes of various local situations that potentially destroy health is essential if the contemporary social science of health is to reflect the importance of the 'central engines' referred to by Diderichsen and colleagues (2001), and to speak truth about their operation. The time is therefore ripe not only for more cross-fertilization between the work of anthropologists and those in other disciplines like political economy, but also for more formal transdisciplinary research collaborations that address the challenges to health in a globalizing world.

Provenance and acknowledgements
The first version of this article was presented at the symposium Beyond Biosocialities in Medical