Geographies of production III

I Introduction

In this report I review recent economic-geographical work on labour. ¹ Time is arguably ripe for such a review. While two of my predecessor’s reports contained subsections on ‘low-paid work and migrant labour’ (Reimer, 2009) and ‘work and employment’ (Reimer, 2010), respectively, not since Yeung (2002) has there been a report looking exclusively at labour issues. If the number of sympathetic reviews published in recent times is a measure of the vitality of a research field, then it would appear that ‘labour geography’ is in very rude health indeed (see, for example, Castree, 2007, 2010; Coe and Jordhus-Lier, 2011; Herod, 2010; Lier, 2007; McDowell, 2008; Rutherford, 2009, 2010; Tufts and Savage, 2009). Two recent overview collections – one almost entirely produced by geographers (Bergene et al., 2010) and one with a strong contingent of geographical contributors (McGrath-Champ et al., 2010) – serve to reinforce the impression that labour geography has reached a certain level of maturity and coalescence around a core series of conceptual categories and theoretical concerns, with the lexicon of place, space and scale being well established at the heart of the project to build an avowedly geographical approach to labour questions (Castree et al., 2004). ² The common refrains across these various reviews and introspectives are to celebrate labour geography’s achievements hitherto – most notably in terms of foregrounding workers as active geographic agents of change and in revealing the spatial variability of labour market processes – at the same time as pointing to an ongoing series of challenges pertaining to both the theoretical and empirical coverage of work in this area. ³

As several of these reviews have also noted, labour geography remains a broad church, encompassing relatively distinct strands of work on the multiscalar constitution and institutional embeddedness of labour markets, the agency of workers and their collective organizations, issues of worker identity and migration, and the importance of material landscapes in shaping labour struggles. This review will largely focus on the second and third of these strands, partly for reasons of brevity but also to reflect the relative balance of recent research activity across labour geography. Whether this recent work constitutes a new ‘fourth phase’ of labour geography research – as suggested tentatively by Coe and Jordhus-Lier (2011) – remains to be seen, but since the early 2000s there does seem to have been a broadening of labour geography’s remit in both theoretical and empirical terms. I explore these developments in three stages. First, notions of worker agency – for many the central leitmotif of labour geography – are increasingly being debated and retheorized as part of less sanguine assessments of the relative position of labour within contemporary capitalism. Second, I profile the constellation of research on precarious forms of work, shifting worker identities and migration, with labour market intermediaries starting to emerge as an important element shaping these dynamics. Third, there remains a vibrant strand of research looking at unionism and its responses, successful or otherwise, to the changing landscapes of the global economy. This is a body of work which has broadened its scope to include an ever wider range of organizational, sectoral and geographical contexts.

II (Re)theorizing the agency of labour

Much recent work in labour geography has sought to highlight the potential for labour agency, and the field has now assembled a rich body of case studies to support that core claim. There are growing concerns in the literature, however, about the validity of simply asserting the potential of labour agency in general. As Siemiatycki (2012: 473) argues, ‘the viability of labour geography does not depend on arguing that labour is powerful. It hinges on arguing that labour can be a distinct, autonomous force in particular temporal and spatial circumstances’. Accordingly, a series of recent contributions have started to developed a more ‘embedded’ or constrained notion of worker agency (Coe and Jordhus-Lier, 2011). In analytical terms, an unpacking of the notion of agency needs to be combined with reconnecting agency to the wider societal structures in which is it embedded.

With respect to the former, labour agency can usefully be determined along several key coordinates. Four dimensions have been highlighted in recent research. First, there needs to be more clarity as to what actually counts as meaningful labour agency. Cumbers et al. (2010), for instance, convincingly argue that Katz’s (2004) distinction between strategies of resilience (everyday coping strategies), reworking (efforts to materially improve the conditions of existence) and resistance (direct challenges to capitalist social relations) offers a useful way forward. This typology reflects both the level of intervention and the difference between trying to change the prevailing system from within or without. It also suggests that much of what labour geography has hitherto identified as labour agency relates to reworking strategies, reflecting the dominance of strategies to alter the relations of production from within, for example through wage disputes (Mann, 2007). Clearly, however, having an appropriate language for describing the nature of the agency under consideration is an important first step. Second, researchers need to be precise about the geographies of the agency under consideration. Both Lier (2007) and Munck and Waterman (2010) provide useful schemas for thinking about the full gamut of worker spatial strategies. Lier intersects the scale of worker action (local or translocal) with the targets of such action (local or translocal) to delimit four different spatial strategies, while Munck and Waterman map out labour strategies from the local to the global scales, distinguishing between those that are more embedded in the market from those more embedded in society (e.g. business unionism versus social movement unionism). Agency always has a geographical dimension in another sense too, in that it emerges in and from particular places. These various geographies need to be carefully distilled in work that foregrounds worker agency.

Third, there are important temporalities to labour agency. Labour geography has arguably tended to highlight moments of success in labour struggles without placing them in historical context. The timing of agency, in relation both to political context and to other worker actions, successful or otherwise, has a non-trivial influence on the likelihood of success. There are also issues relating to the temporality of gains from worker action, for example in terms of labour migrants and short-term employees who may have fewer vested interests in long-term and institutionalized gains. Time is also important in terms of the learning processes for workers, unions, employers and the state that individual moments of agency are part of (e.g. Meyer and Fuchs, 2010). Fourth, there needs to be clarity about the balance of individual versus collective agency in the worker action under consideration. While labour geography has tended to focus on collective action for obvious reasons, there is growing awareness of the individual agency exercised by workers on an almost daily basis (see, for example, Rogaly, 2009, on unorganized migrant workers) and that agency is shaped not only by worker identities but also by the wider intersections of class, gender and race. In short, we need to recognize the other subject positions – beyond those of worker and union member – on which workers act. Labour agency then, needs to be deconstructed as an analytical category if it is to have any meaningful traction. As Bezuidenhout and Buhlungu (2011: 257) summarize, ‘worker agency can be informal or formal, individual or collective, spontaneous or goal directed, sporadic or sustained, and it can operate on different scales’.

This dissection needs to be accompanied by actively reconnecting agency to the wider societal structures in which it is always and incontrovertibly embedded (Coe and Jordhus-Lier, 2011). As Mitchell (2011: 586) alludes, it is important ‘to reassert the need to situate the analysis of workers’ struggles within their structural context – that is, to take seriously the already existing geography of production and reproduction’. For Lambert and Gillan (2007), this involves questioning the ‘optimism of agency’ and the assumptions of voluntarism inherent to labour geography, and instead developing a more embedded conception of agency that reconnects with the continued structural domination of capital in many contexts. ⁴ This suggests that more intense engagement is required between labour geography and the more traditional concerns of economic geography with respect to corporate strategies and restructuring. The multidisciplinary and overlapping literatures on global commodity/value chains (GCC/GVC) and global production networks (GPN) ⁵ – in which geographers are heavily involved – is one area where such a dialogue is starting to emerge.

On a theoretical level, this interaction makes sense given that, as Cumbers et al. (2008) and Rainnie et al. (2011) have pointed out, GPNs are as much systems of embodied labour as they are interlinked systems of firms: ‘labour needs to be conceptualized a priori as a value producer in GVCs and thus as a social actor with its own interests regarding the organizational, spatial and political structure of a value chain’ (Riisgaard and Hammer, 2011: 186, emphasis in original). There is by now a well-established literature that explores the impacts of GPNs on workers in developing countries and the potential (or not) for private standards schemes and ethical trading initiatives to ameliorate working conditions (e.g. Posthuma and Nathan, 2010). The challenge, however, as laid out by Selwyn (2012: 217, emphasis in original), ‘is that it is not enough for GCC analysis to illustrate how integration into GCCs by supplier firms in less developed regions impacts on labour. Rather, we are concerned with investigating how action by labour may co-determine processes of local capitalist development’. Recent work that tries to explore the social upgrading potential of incorporation into GPNs has started to make space for the variable bargaining power of labour across different sectors and contexts (Barrientos et al., 2010). Increasingly, however, the agency of workers is being seen as a constitutive element of GPNs dynamics. The key for Selwyn (2012) is to explore the conditions under which worker’s structural power – derived from their position in the production process – can be transformed into associational power ⁶ – based upon collective organization – that is able to extract concessions from the state and/or capital (see also Webster et al., 2008). For example, the establishment of a well-organized and militant rural trade union in northeast Brazil’s grape export sector has enabled gains for workers to be accrued within a wider GPN that is highly sensitive to interruptions in production (Selwyn, 2011).

As Riisgaard and Hammer (2011) point out, the ability to construct associational power is ultimately driven as much by local and national institutional context as it is by GPN structures. Their research on the cut flower and banana sectors demonstrates that ‘the governance of inter-firm linkages, while crucial in structuring the terrain for labour, is always mediated by the specific social relations of local production and labour control regimes as well as the histories and orientations of the respective actors’ (p. 183). They argue that the potential for worker agency is shaped by the intersection of three factors: the overall ‘drivenness’ of the network, whether it is a producer- or buyer-driven system, and the nature of the local labour control regime. In a study of aquaculture unionism in Chile, Oseland et al. (2012) similarly argue that worker agency is heavily conditioned by national factors, with three domains being scaled heavily at the national level: the organization of labour networks, the regulation of economic and employment affairs, and the sphere of public discourse about work and its regulation. Pickles and Smith (2010), in turn, demonstrate how the variable ‘positional power’ of workers in the clothing industries of central and eastern Europe is determined not only by industry dynamics but also the institutional cultures and variable forms of transition experience in different countries, while Bezuidenhout and Buhlungu (2011) discuss how the state has played a hugely significant role in shaping the potential for labour agency within the mining compounds of South Africa. Collectively, these studies reiterate the importance of how labour is produced, reproduced and regulated in different local and national contexts, suggesting that labour agency needs to be re-embedded in state formations as much as it does in the global structures of capital. ⁷

III Precarity, new migrant divisions of labour, and intermediaries

If a concern with agency has been one running thread in labour geography research, another has been to reveal the unequal and highly segmented nature of contemporary labour markets, especially in advanced service economies. ⁸ In particular, research has charted the continued expansion of insecure forms of work across a range of construction, manufacturing and consumer service sectors characterized by low-pay, part-time or temporary contracts, weak health and pension benefits, and low levels of union coverage (McDowell and Christopherson, 2009; Peck and Theodore, 2010). Variously called insecure, contingent or precarious work, Waite (2009) makes the case that precarity is a useful term that implies both a generalized condition of work and a possible point of mobilization, but cautions that we should stop short of identifying a ‘precariat’ that risks obscuring the heterogeneity and complex identities of the different groups of workers involved (cf. Standing, 2011). While there are many drivers of the growth in precarious work – for example, labour market deregulation, wage squeezes resulting from global competition, shrinking public-sector budgets, and the increased use of temporary and contractual employees – Wills (2009) convincingly diagnoses subcontracting as the central mechanism involved, which both brings workers into downward competition with each other and can make it hard to discern the ultimate employer of precarious workers. As she argues, ‘subcontracted employment is becoming paradigmatic. Such employment is now commonplace across the public and private services, distribution, and manufacturing, and it has serious implications for labour in terms of wages and conditions, employment experience, and power relations’ (Wills, 2009: 456).

Increasingly, labour geographers are concerned with the various interconnections between precarity and migration. Two incredibly rich and productive strands of work have explored such intersections in the context of the Greater London economy. One derives from the work of Jane Wills and her colleagues on what they term the ‘new migrant divisions of labour’ (NMDL) (see Wills et al., 2009b, 2010, for overviews; see also Wills, 2008). The NMDL concept captures the intensification of immigration to the UK over the past two decades, the diversification of the countries of origin, the proliferation of migration routes and channels, and the ways in which migrants have largely been integrated into low-paid segments of the workforce. State migration policies are seen as a central driver of the NMDL. As May et al. (2007) argue:

far from acting to protect workers from the worst excesses of low-paid work … policies of labour market deregulation, welfare ‘reform’ and ‘managed migration’ have helped create a new ‘reserve army of labour’ in London whose ranks are filled with a disproportionate number of migrant workers. (May et al., 2007: 152)

The intersections of migration and precarity in the London economy have also been explored in a similarly impressive manner in a cognate programme of research conducted by Linda McDowell and colleagues. In a series of papers based upon detailed qualitative workplace research in the hotel and hospital sectors, they have explored the ways in which identity and ethnic difference is constructed for economic migrants. In the context of new patterns of migration into the UK since EU expansion in 2004, they demonstrate how migrants with varying social characteristics and residence/citizenship entitlements are differentially incorporated into precarious and low-paid occupations within the Greater London economy (McDowell et al., 2009). Staffing agency employees – ‘perhaps the most significant category of precarious work and certainly the most exploitative’ (p. 19) – are central to this story, with agencies assembling an inherently local workforce undertaking unskilled, embodied service work from a wide and highly differentiated pool of economic migrants (see also McDowell et al., 2008). Workplace analysis is mobilized to show how, in the hotel sector, managers construct a segmented migrant workforce in which stereotyped assumptions about the suitability of migrants for certain kinds of work are reproduced through the ongoing interactions between managers, workers and guests (McDowell et al., 2007), while in hospitals low-paid migrants workers are enrolled into highly embodied and intimate forms of caring work (Dyer et al., 2008). ¹¹ A final important contribution reveals how migrant identities – in this case Indian middle-class male identities in the hospitality sector – are produced through a transnational field that mobilizes different notions of the desirability of forms of work in home and host contexts (Batnitzky et al., 2008; cf. Kelly, 2012; Lusis and Bauder, 2010). These various findings chime closely with ongoing research into the labour market segmentation of Filipino immigrants within the Toronto labour market (Kelly, 2010; Kelly et al., 2009).

The role of labour market intermediaries – and temporary staffing agencies in particular – in producing precarious work and labour market segmentation is an increasingly common refrain in this line of research. While a previous tranche of research by economic geographers served to reveal the systemic role of staffing agencies in flexible labour markets such as the USA and the UK (e.g. Peck and Theodore, 2007), recent work on staffing has been more global in scope and is underpinned by three themes. First, and following in the footsteps of the path-breaking work of Peck and Theodore (1998), the active labour market role of staffing agencies as profit-seeking intermediaries has been emphasized. This has shed light on the expansionary strategies and lobbying activities of staffing firms and their trade associations as they seek to expand the market spaces in which they operate as well as enhancing the wider social acceptability of their activities (Coe et al., 2010; see also Barrientos, 2008, on labour contractors). Agencies arguably also play an important regulatory role in mediating forces of supply and demand (as argued in Stenning and Dawley’s 2009 study of the labour market incorporation of Polish immigrants to northeast England). Second, work has profiled the emergence of a cadre of global staffing agencies that, since the 1990s, have expanded far beyond their core regions of North America and western Europe. The localized nature of temporary staffing activity, however, has meant that they have had to pursue highly spatially variable and territorially embedded corporate strategies (Coe et al., 2007, 2011b). Third, research has demonstrated that, despite these ongoing processes of globalization, the persistence of distinctive national markets for temporary staffing is the hallmark of the global industry. National case studies of Australia (Coe et al., 2009b), Japan (Coe et al., 2011a, 2012), central and eastern Europe (Coe et al., 2008) and Sweden (Coe et al., 2009a) have shown how temporary staffing is heavily shaped by national institutional and regulatory formations. Overall, while economic geographers have made important strides in exploring the nature of the staffing industry, much remains to be done. In particular, the role of agencies and other intermediaries in both mediating labour migration itself and shaping processes of segmentation in host economies merits urgent further attention.

IV New modes of organizing

The third strand I will review brings together the twin concerns of labour agency and the structural realities of contemporary labour markets by exploring the evolving spatial strategies of collectively organized labour. The challenges for organizing posed by the growth in precarity described above are considerable. Lier (2009) couches these challenges in terms of the multidimensional fragmentation of workers caused by extensive outsourcing and subcontracting practices. Not only are workers fragmented across space in smaller and more isolated units, but they are fragmented administratively by the complex networks of direct employers and subcontractors, contractually through the individualization of the employment relation, and temporally by the intensification of part-time and shift work. These processes all serve to pose problems to the traditional modes of workplace unionism and associated methods of recruitment and collective bargaining (see, for example, Rhee and Zabin, 2009, on the care sector). Hence, as Anderson et al. (2010: 394) evocatively describe, ‘in order to navigate the increasingly complex catacombs of employer power, unions need creative new strategies that fundamentally reconstitute the geography of their practices’.

Here I will briefly consider two recurring and intersecting themes running through recent work in this area. The first relates to the appropriate scalar configuration of worker action in the contemporary global economy. On the one hand, evidence continues to accrue that certain kinds of local action can achieve meaningful purchase within GPNs. As Padmanabhan’s (2011: 21) study of garment workers in Kerala, India, demonstrates, ‘organizing locally can, in fact, be an effective strategy for use in case of confrontation with social actors who are organized at the global and other extra-local scales’. On the other hand, work such as that of Cumbers et al. (2008; see also Cumbers and Routledge, 2010) on ICEM (the International Federation for Chemical Energy Mine and General Workers) demonstrates that building meaningful transnational union action is extremely problematic. ‘Convergence spaces’ do not simply emerge, but need to be created through ongoing struggles that serve to reveal the tensions between, and uneven power bases of, the constitutive national unions. These ‘geographical entanglements’ are heavily shaped by personalities, access to resources, and other contingent factors (cf. Ford and Dibley’s 2012 study of the involvement of international labour organizations in forging a labour movement in post-tsunami Aceh, Indonesia). Similarly, Aguiar and Ryan (2009) argue that the spread of the Justice for Janitor’s campaign through the international networks of the Service Employees International Union (SEIU) is spatially selective and heavily shaped by how it is embedded in the union structures of different national territories. They conclude that Justice for Janitors is in reality a complex, multiscalar campaign which can be adapted to suit diverse institutional contexts.

Increasingly, the literature seems to be adopting this language of multiscalarity as opposed to simply identifying the necessity of upscaling. As Munck and Waterman (2010: 283) describe, ‘there is now a growing tendency to accept that we might think globally but act locally … labour’s contestation of capitalism and its effects will more often than not be multiscalar’. Tufts’ (2007) notion of the spatial circuit of union renewal is useful in this regard. Building upon research into UNITE HERE Local 75 in Toronto’s hotel sector, he adopts a fluid multiscalar approach that emphasizes the non-deterministic and reinforcing interconnections between interventions at different spatial scales: ‘from workplace action to trans-national organization, each scale of labour organization contributes information and resources that drive various labour actions ranging from local community coalition building to the global administration of international agreements’ (Tufts, 2007: 2387). Another important intervention in debates around multiscalar union strategies has come from Anderson et al. (2010; see also Anderson, 2009), who advocate strategic interventions that are alive to the delicate geographies of transnational corporations and their wider production networks. There are three elements to this. First, unions can engage in ‘power analyses’ to identify the most effective locations for strategic action, for example in targeting the buildings where cleaners work as opposed to the contracting company themselves. Second, effective action may require the identification of key zones – such as urban districts, whole cities or transportation nodes – that can be carved out as effective arenas for action. Third, the upscaling of union action with respect to transnational capital may allow power to be ‘projected’ from relatively powerful to weaker nodes within the network. As Anderson (2009) notes, however, ‘unlocking’ powers that are embedded in particular local and national contexts is fraught with difficulty.

The second theme concerns the extent to which effective union interventions will require a shift from traditional business unionism to community unionism. There is by now a rich multidisciplinary literature on community unionism, to which economic geographers have made a healthy contribution (for an overview, see Lier, 2009). This literature has shown how, in addition to pursuing the new geographical strategies described above, unions have increasingly extended their activities into new community sites of recruitment and new domains of mobilization in partnership with other civil society actors. Again, Wills (2009) has been a powerful contributor to debates, arguing that:

the London living wage campaign demonstrates the way that union-community alliances are necessary to win traditional trade union demands in the context of subcontracted employment, and unless they face up to the implications of this new employment paradigm, trade unions are unlikely to realize the promise of organizing campaigns. (Wills, 2009: 457)

She notes how ‘relatively shallow relationships’ (p. 457) across London Citizens – a network of educational, community, faith and labour organizations – were able to lever impressive gains for workers whether they were in a trade union or not (see also Wills et al., 2009a). As such, broad-based community coalitions can be central to the drive for what Soja (2010) terms ‘spatial justice’. More recently Wills (2012: 114) has argued that this model reflects the construction of a new ‘geo-political architecture’ that is able to overcome some of the barriers posed to organizing by modern urban society, such as rapid population churn and ‘stretched’ social networks, by building connections between ‘institutional islands of social solidarity’. Such initiatives may also be the only really effective way for the labour movement to reach out to groups excluded from traditional union activity, as argued by Mills and Clarke (2009) in the case of Aboriginal workers in Canada, and Fitzgerald and Hardy (2010) with respect to Polish migrants in the UK. Echoes are also to be found in Hardy et al.’s (2008) account of how women have created ‘new political spaces’ both in and beyond the workplace in post-socialist Poland. Clearly, community unionism is not the only way forward for contemporary unions, however. In two important interventions, and again based on work in Toronto’s hotel sector, Tufts (2009, 2010) has proposed a model of ‘Schumpeterian’ unionism that falls somewhere between the ideal-types of business unionism and social movement unionism. In effect, he diagnoses a form of unionism that reflects how the traditional union model has been transformed into a grounded, pragmatic approach that sees the union operate as strategically as possible within the realities of contemporary deregulated, flexible labour markets.

V Conclusion

Recent reviews have diagnosed a series of ongoing theoretical and empirical challenges facing the subfield of labour geography. The rich body of contemporary work profiled here attests to the fact that labour geography is a vibrant, multistranded and innately interdisciplinary project that is making significant headway in tackling the challenges relating, for example, to theorizing agency, and pushing empirical research into new sectors, new geographical regions and new domains and modes of collective worker organization. In other areas, such as incorporating more fully theories of the state and national labour systems, and recognizing the important role played by labour market intermediaries, much more remains to be done. Moreover, there is still a sense that at times the different strands of labour geography should be brought more effectively together; theoretical work on agency, for example, could usefully be connected to research on economic migration of different kinds.

Wider questions also persist about the normative and political ambitions of labour geography (cf. Castree, 2010; Tufts and Savage, 2009). It remains something of a paradox that labour geography has blossomed as a field of academic endeavour over a timeframe that has seen the continued ascent of neoliberal globalization and worsening conditions and relative rewards for the majority of workers worldwide. The extent to which labour geographers are genuinely highlighting the agency of workers as opposed to charting the continued power of capital to shape global production systems is a moot point, and clearly has implications for the interventions that labour geographers seek to make. Ultimately, perhaps, it is this desire both to profile the structural (albeit geographically variable) position of labour within capitalism and to highlight the instances of progressive change achieved by workers that defines labour geography. To my mind at least, more effectively connecting these twin impulses brings us back to the need to continue developing refined and geographically nuanced notions of worker agency.