Geographies of production I: Global production and uneven development

I Introduction

Uneven development is back on the agenda of economic geography after several decades of relative neglect. Since the 1990s, the sub-field has been primarily animated by its explorations of ‘the nodal, the near and the networked’, taking the field’s focus on ‘the sociospatial, the systemic and the structural’ as a largely taken-for-granted platform for the cultivation of various heterodox concepts and conversations (Peck, 2016: 307). Among the most prevalent of the latter is the global production network (GPN) framework, associated with the Manchester-Singapore school. GPN scholars investigate the value, power and embeddedness of transactionally linked but geographically dispersed production arrangements. The geography framework was developed in conversation with development sociology’s research program on global commodity chains and, later, global value chains. A detailed discussion of the differences between these heuristics can be found elsewhere (e.g. Bair, 2009). Suffice it to say that geographers have sought to foreground more robust territorial and non-firm dimensions that were given short shrift by their sociology counterparts. Despite differences in emphasis, however, GPN and global value chain studies both train their analytical gaze on the advantages that accrue to firms and regions that leverage their participation in transnational supply chains to access higher value niches in the production structure; this is what the literature calls economic upgrading (focusing on firms) or strategic coupling (with an eye to regions). Jennifer Bair and I have summarized this predilection for exploring the largely beneficial effects of linking supplier firms and regions to contracting chains coordinated by multinational corporations as an ‘inclusionary bias’ (Bair, 2011): this bias clouds our understanding of the extent and significance of the reproduction of low-value positions as well as the constitutive exclusions of firms, regions and workers that remain part and parcel of network formation and restructuring. To address this weakness, Bair and Werner (2011), Bair et al. (2013), and McGrath (2017) have developed the notion of ‘disarticulations’ to signal the re-working of the GPN framework to center on uneven development. These authors frame global production networks as both temporarily stabilized outcomes of inherited patterns of uneven development and contributors to their ongoing formulation, especially as firms and firm networks reorganize to adjust to various pressures.

Current calls in the discipline to focus on new geographies of uneven development have emerged from multiple corners beyond GPN studies, of course. They have been spurred, or at least emboldened, by serial crises in the global economy and a concomitant surge in related debates in the social sciences, including geographical political economy (e.g. Sheppard, 2016; Hudson, 2016; Peck, 2016; Dunford and Liu, 2017), urban and regional studies (Hadjimichalis, 2011; Hadjimichalis and Hudson, 2013; Phelps et al., 2018), international relations (e.g. Callinicos and Rosenberg, 2008; Kiely, 2012; Rosenberg, 2010; Desai, 2015), and development studies (e.g. Makki, 2015). Not surprisingly then, uneven development is also being take up more significantly in the analysis of production networks beyond the initial critiques. As I discuss in more detail below, a growing number of studies in sociology and geography pay increasing attention to the dynamics of uneven development, suggesting a reformulation of the hitherto dominant assumption of felicitous inclusion, and an analytical centering of workers, firms, and regions that remain trapped in low-value positions or excluded by heightened competition. How do we make sense of this emerging trend in the study of global production and how do we parse the differences among the diverse range of contributions?

In the rest of this report, I examine three approaches to uneven development in the global production network literature in economic geography and cognate fields. The first, ‘mainstream’ position examines uneven development as a contingent empirical outcome – a ‘dark side’ – of the process of global market integration. The second and third positions that I review share a broadly Marxist or neo-Marxist orientation, including a theoretical commitment to formulate uneven development in dialectical terms and an embrace of a dynamic, integrated, but nonetheless structured global economy. The second position – global production networks as conduits of unequal value distribution – theorizes the mechanisms through which global production arrangements reproduce global inequality. The third position – global production ‘without guarantees’ – adopts Gramscian and/or feminist frameworks to historicize sub-national regional experiences of global production network incorporation. All three positions offer rich accounts of actually existing processes of restructuring of the global economy, albeit differing significantly in their theoretical commitments and normative assumptions. Despite these differences, more dialogue across these positions is warranted if we are to leverage the relatively powerful perspective that production network studies afford to grasp the vicissitudes upending received patterns of uneven development and portending uncertain futures.

II Exploring the ‘dark side’ of global production

In their recent book, Coe and Yeung seek to transform GPN from a heuristic, explanatory framework to a dynamic theory (dubbed GPN 2.0) that will ‘enhance the ability of GPN thinking to contribute to explanations of patterns of uneven territorial development in the global economy’ (2015: 22). Keeping with a firm-centric focus, the authors identify firm-level organizational strategies (e.g. intra-firm coordination versus inter-firm control) as the proximate dependent variable of what they call competitive dynamics (e.g. financial discipline and downward cost pressures). Their principal goal is to determine how the interaction between competitive dynamics and firm-level strategies leads to given outcomes for ‘strategic coupling’, a term that signals the prospects for regions plugged into GPNs via goods or services suppliers to improve their economic condition. In thinking about uneven development, the shift from a static equilibrium view of the firm-GPN nexus to what the authors call ‘value capture trajectories’ is moderately useful if one accepts the overall premise. ‘Trajectories’ refer to the wide range of methods that firms employ to ‘capture’ more or less value (as surplus above costs and normal profits, see below) over time. Built into this notion of trajectories are the vagaries of growth and decline, indicative of a wide range of possible outcomes of a firm’s participation in global production networks. Coe and Yeung scale up these trajectories to sub-national regions to identify three modes of strategic coupling and, drawing on evolutionary economic geography, to emphasize territorial change over time. Extending the timescale allows the authors to account better for variation not only in the traditional remit of GPN studies (i.e. ‘inclusion’, or how regions initially link to GPNs), but also in how these modes of linkage can shape outcomes, including negative ones such as heightened intra-regional inequality and lock-in to low-value, dependent positions in networks (see also Coe and Hess, 2011).

Coe and Yeung’s revision forms part of a broader effort within mainstream GPN and cognate studies in sociology to address the dark side of participation in global supply chains. Sociologists, for example, have developed a parallel dimension to economic upgrading that they call social upgrading, which examines how the changes in what a firm does or the process of doing it affects the rights and entitlements of workers (Barrientos et al., 2011). This shift towards including worker-based dimensions is informed in part by large-scale transformations of supply chains in the last decade that have made supplier participation generally more complex and ultimately more restrictive (Gereffi, 2014). In studying how suppliers are adapting and learning under these highly competitive conditions, a number of empirical studies conclude that economic upgrading, which secures suppliers’ continued participation in a restructured production network, often does not translate into better outcomes for workers. Increasing evidence suggests that shifts to more technologically-intensive export production is associated with reductions in the proportion of female workers (e.g. Barrientos, 2014; Kucera and Tajani, 2014; Greenstein and Anderson, 2017), for example. As firms respond to pressures to undertake more complex functions, processes or higher value products, and to meet intensified cost-capability pressures (that is, to do more for less), they adjust by creating more fine-grained stratifications among their workforces, between permanent and temporary employment in garment manufacturing (Plank et al., 2012; Rossi, 2013) or between permanent and seasonal workers in export horticulture (Alford et al., 2017). As I have discussed at length elsewhere (Werner, 2016a), studies of dark sides of economic upgrading and strategic coupling can offer an empirically-grounded perspective on new geographies of uneven development if read with an eye to them.

III Global production networks as conduits of unequal value distribution

Firm-centered approaches to global production – and the dark sides that they might reveal – nonetheless suffer from key limitations that hinder the analytical project of understanding long-standing and emerging geographies of uneven development through the study of global production. Uneven development framed as a ‘dark side’ generally presumes that these arrangements are outcomes of firms or firms in regional contexts. Leaving to one side other limitations that follow from firm-centrism (e.g. limited attention to labor, the state and geopolitics), the principal problem with this approach for my purposes here is a failure to engage with macro-scale questions of global inequality. Coe and Yeung, for example, reject hierarchy in the global economy even as they acknowledge that structural inequality exists (2015: 179). In so doing, the possibility to marshal the rich empirics and analysis of GPNs into a dynamic, spatiotemporal understanding of a globally integrated and dynamically structured global economy is lost. Mainstream production network studies tend to dismiss macro-geographies of uneven development out of hand as a form of rigid structuralism that necessarily reduces actors and places to their functions in the global hierarchy. As a result, the mainstream scholarship pays little heed to the ways that structural hierarchy conditions the possibilities for firms and regions. Empirical questions that are sidelined as a result include, for example, how might regions or firms that benefit from higher value functions and more autonomy in the network affect the possibilities for other regions or firms to occupy the same positionality? Or, the converse: do firms or regions locked in to ‘low-value’ (i.e. highly competitive) functions make it possible for other firms or regions to be linked in more ‘value adding’ ways?

In this section, I turn to literature on global production that places macro-scale geographies of uneven development at the center of its inquiry. Much of this work emerges out of, or in close conversation with, the world-systems tradition. World-systems scholars Hopkins and Wallerstein (1977, 1986) coined the term ‘commodity chain’ to disrupt nation-centered accounts of development and, in particular, modernization theory and its attendant ideas of sequentialist stages of development. Instead, they proposed the study of linked sets of activities – from inputs to production and consumption – organized transnationally. In this tradition, returns are created (and destroyed) and distributed via these chains or networks. The value characteristics of the economic activities that make up the chain are not static, moreover  (Arrighi and Drangel, 1986). They stem from capitalist strategies, supported by states, to distance themselves from highly competitive functions – thus accruing above average profits from (semi)monopolized ones – through organizational (e.g. outsourcing), technological (e.g. innovation) and, with the support of states, legal (e.g. intellectual property rights, trade regulations) strategies. The main point is that core and peripheral positions in the world-system concentrate core-like and peripheral-like functions. These attributes do not correspond to particular activities (e.g. assembly or resource extraction), but instead reflect the degree of monopolization of the activity and thus the possibility to earn above-average returns. Arrighi and Drangel’s classic work (1986) identified a  resulting tri-modal hierarchy to the global economy (core, semi-periphery, and periphery) as a relatively stable, if dynamically configured, structure. In an early dialogue with geography, Peter Taylor (1988) offered a ‘supportive note’ to this approach by demonstrating that the tri-modal distribution was even more robust if one used finer-grained spatial units (i.e., using a subnational rather than a national  areal base). Obviously, this ground is well-worn, but I repeat it here because a new generation of commodity chain studies seeks to recuperate the basic project in order to grasp the reproduction of spatial hierarchy in today’s global economy via the study of production networks (Bair, 2005, 2009, 2014).

Let me briefly highlight a few key contributions. Much of this literature seeks to marshal the study of global production to better understand observed patterns of global inequality. Despite the championing of network extension to new places and actors and subsequent ‘upgrading’ in the mainstream literature, dynamic but durable inequalities persist. Thus, the aspirational, catch-up type underpinnings of the mainstream literature are discarded for a more sobering look at how production network restructuring relates to the reproduction of global income inequality. In an early contribution, Arrighi et al. (2003) demonstrated that the convergence in industrialization in the late 20th century had not yielded an attendant convergence in incomes; instead, the relative value of manufacturing had decreased as core regions had benefited from more monopolized activities (such as R&D) and accumulated resources, especially financial and political ones, that had shielded these areas, relative to others, from the destructive phase of capitalist restructuring. Brewer (2011) extends these basic insights to critique the clear disconnect between the notion of generalizable upgrading and durable systemic patterns of inequality, parsing the complex debates on income polarization and connecting these to commodity chain studies. Mahutga (2014) uses trade data to compare the positional power of countries in the well-studied auto and garment commodity chains and finds that participation in these chains increases polarization within the world-system, albeit not in the same way. He argues that stratification in the global economy linked to production network participation thus requires sector-specific analysis. The question of the actual morphology of global hierarchy is again being hotly debated in the world-systems tradition and extending from it. Karataşli (2017), squarely in that tradition, has recently extended the Arrighi and Drangel contribution to posit a contemporary ‘quadri-modal’ distribution. Collections by Bond and Garcia (2015) and Wallerstein et al. (2015) consider this question directly in the context of the rise of the so-called BRICS. And Hudson (2016) has offered the provocation of a ‘new’ New International Division of Labor, reflecting similar upheavals in the structure of the global economy.

At the heart of Marxist debates over the world-systems approach has long been the ontological status of value posited as a form of Schumpeterian rent (i.e. above-average profits due to monopoly over a relatively scarce asset) in the latter, in contrast to Marx’s original notion of surplus value, which reflects relations of production between capital and labor. The distinction reveals both theoretical commitments and empirical emphasis. Neil Smith and David Harvey both sided with the orthodox position in early work on uneven development; Smith (2008 [1984]: 289), in particular, dismissed world-systems theory for its position on value. Beyond the particularities of the value definition (see Arrighi, 2007, for an excellent synthesis), however, what has unified Marxist global production studies within and beyond the world-systems tradition is an understanding of the global economy – and the states and production networks that constitute it – as a (variously) complex totality. Thus, for Starosta, in perhaps the most orthodox proposal, the forms of production networks are expressions of the unfolding of the law of value wherein all capitals are, in principle, subject to the general rate of profit: deviation from that rate – in the form of either export suppliers, national industries, or multinational corporations – is not explained by monopoly control (or lack thereof), but rather place-specific social relations that permit capitals (i.e. firms) to survive at a lower rate of profit (such as non-market subsidies to social reproduction) (Starosta, 2010; Fitzsimons and Starosta, 2017; see also Purcell et al., 2016). Selwyn (2015) offers a mixed approach that hinges Schumpeterian rents to class relations: after all, the impetus to innovate and the possibility to implement such changes is determined in some part by the power of the working class. Building on the insights of agrarian and world-systems scholar Farshad Araghi (2003), Selwyn argues that the goal of production network analysis should be to understand how ‘the form and intensity of the exploitation of labour in different parts of the globe, and the systems of development that are based upon them, are dialectically inter-related’ (2015: 258).

As McGrath has recently reminded the readers of Progress, following in the tradition of Elson’s (1979) classic intervention, the conceptualization of macro-scale geographies of inequality benefit from a critical, heterodox, politically-attuned approach to the question of value (McGrath, 2017). Towards such an endeavor, Werner (2016a), Argent (2017), McGrath (2017) and Ouma (2015) have brought questions of value transfer, dispossession and appropriation more centrally into the geography literature on production networks. Recent work by Quentin and Campling adds an additional valence to this effort by exploring what they call ‘global inequality chains’ that reflect the interaction between global production networks, on the one hand, and ‘global wealth chains’, on the other. The latter refers to the increasingly complex mechanisms marshaled by global corporations to capture surplus by avoiding tax liabilities using strategies of offshore incorporation and transfer pricing (Quentin and Campling, 2018; see also Seabrooke and Wigan, 2017; Bassens and Van Meeteren, 2015). World cities scholars are also uncovering how financial nodes and the producer services that concentrate within them contribute to the unequal transfer of value. Parnreiter, for example, looks at the direct role of law firms in this process via intellectual property rights claims and labor regulation arbitrage (2017; see also Brown et al., 2010). Although sharing similar objects of analysis with the mainstream GPN literature here (cf. Coe et al., 2014), these latter studies insist that the specifics of financial, fiscal, or extra-economic forms of appropriation and dispossession are understood within the context of global hierarchy as a concrete abstraction.

IV Global production ‘without guarantees’

There are of course significant dangers to macro-scale analyses of uneven development. An immediate drawback of global inequality chains can be their reliance on income inequality as a measure (and attendant data problems plus the invisibility of unpaid labor and other measures of well-being). Beyond the empirics, scholarship on uneven development must avoid the pitfalls of ‘encompassing comparison’, an approach that presumes that the ‘whole’ or totality governs its parts (McMichael, 1990). If we understand global hierarchy to be a complex totality, the latter must not be mistaken for a totalizing account wherein the multiple, particular arrangements of capitalist accumulation are reduced to their parts in the system. Nor should this critique, however, be grounds for dismissal of totality altogether in favor of unprincipled appeals to geographic polycentricity and heterogeneous causality as is so often the case in the mainstream literature. We can instead seize upon open non-teleological understandings of dialectical relations (Hart, 2016) wherein ‘totality is a conceptual procedure, rather than an empirical or conceptual premise…in which the whole is discovered through the analysis of the mutual conditioning parts’ (McMichael, 1990: 391; see also Friedman, 2016).

In this final section, then, I focus on studies of global production that contribute to our understanding of uneven development as a complex totality through the study of regional conjunctures. Variously inspired by Gramsci, Hall, and Massey, this work sees ‘the region’ as a process, ‘a product of interrelations…constituted through interactions’ (Massey, 1999: 2), as an indeterminate multiplicity made through and in turn remaking social relations at multiple scales. The most explicit attempts to redefine the global production literature through this lens in geography include Pickles and Smith (with multiple co-authors) (2015) and Werner (2016a). In their study of the post-Soviet apparel economy, Pickles and Smith discuss the divergent outcomes of transforming textile and garment firms in post-Soviet states, formerly state-protected industries, into export platforms for the European Union (EU) market. Combined with EU accession and sectoral trade liberalization, the authors trace related processes of firm restructuring and geographical shifts in the context of serial crises. The result is neither a uniform race to the bottom nor robust and generalized upgrading (indeed the term does not contemplate the initial ‘downgrading’ of state-sponsored industry), but rather significant variegation, including novel East–East geographies of uneven development as Slovak garment firms sub-contract ‘unskilled’ work to Ukrainian sub-contractors. Articulation, rather than evolution, signals the contested and contingent, yet also structured and conditioned, outcomes of this process. The parallels with my research on restructuring and regional disinvestment in the Dominican garment industry are striking despite the distinct contexts. My work traces how the legacies of colonialism – or what I call coloniality (following the work of Anibal Quijano and Silvia Federici) – shape contemporary outcomes in overdetermined ways. In particular, domestic firms adjust to global pressures through large-scale layoffs, while marshaling a particular regional culture tied to small-scale export tobacco in order to minimize class conflicts. There are important differences with the Eastern European case, but also some fascinating similarities: in particular, the emergence of South–South subcontracting networks between the Dominican Republic and Haiti that serve a similar function to the East–East networks described by Pickles and Smith.

The region as ‘conjuncture’ offers considerable promise for reconstructing the complex totality of uneven development. This approach is often mobilized in the interdisciplinary tradition of agrarian studies, a tradition that has long served as a sympathetic but critical fellow traveler of the world-systems project (see e.g. Mintz, 1977; cf. Coronil, 1996). Agrarian studies has much to offer the study of production network restructuring in the Global South since the debates over agrarian change have long wrestled with the constitutive complexities of transition (from the [in]famous ‘articulation of modes of production’ debates onward) and disposed of any Eurocentric attachment to the inevitability of expanded wage relations. Gill Hart’s Disabling Globalization (2002), which grounded itself in studies of agrarian change to understand divergent but related outcomes of Taiwanese FDI in the South African garment industry, is an excellent example of this approach, foreshadowing a number of the insights in Pickles and Smith. Tania Li’s (2014) book on the articulation of forest dwellers into the global cacao chain offers a recent sustained engagement with conjunctural approaches to agrarian change. Her revisit ethnography highlights a recurring theme in Southeast Asian studies of export commodity production: small suppliers (be they peasants or swidden agriculturalists, indigenous or migrant) may accede to capitalist relations of labor and property even in the absence of corporate-driven or state-driven dispossession (see also Hall, 2012).

Conjunctural analysis of global production feels particularly urgent in the current period. As Gramsci wrote (of his place and time but easily applied to ours), the ruling class has ‘lost its consensus’; it dominates but no longer leads. ‘The crisis consists,’ he continued, ‘precisely in the fact that the old is dying and the new cannot be born; in this interregnum, a great variety of morbid symptoms appear’ (1971: 275–6). The dogged insistence that the globalization of production would be a rising tide that lifts all boats has proven to be patently false; instead some boats are lifted while others sink (and, if I may, rising tides as a positive metaphor seems particularly ill-suited to the anthropo/capitalocene). What conjunctural analysis offers in this context is an attention to cultural formations and an attendant openness to political outcomes, not as unbridled contingency but rather as a mix of both conditioned and indeterminate forces. The study of outcomes will only be strengthened if we turn our lens as much to the new geographies of restructuring in the Global South as to the legacies of serial disinvestment in the historically peripheralized regions of the Global North. I am not suggesting that the same studies incorporate all of these moving parts, but rather, expanding upon Ramamurthy’s (2004) call for a feminist commodity chain analysis, that we continue to trouble the false binary that global production networks link producers in the Global South to consumers in the Global North. For Ramamurthy, and many others, this has meant interrogating the subjectivities of Global South workers in and beyond capitalist production as they experience ‘adverse incorporation’ (Phillips, 2011) and iterative or permanent exclusion (recent work includes, e.g., Dutta, 2016; Prentice, 2016; Sum, 2017; Gago, 2017). Future work might take these insights from the Global South and trouble the class- and gender-based assumptions of the ‘consuming’ Global North in order to develop stronger dialogue with those who study the seemingly interminable tide of disinvestment, and associated social malaise, that continues to grip rustbelts 40 years on from the introduction of the ‘first’ New International Division of Labor (e.g. Smith, 2015; Smith and Winders, 2017).

V Conclusion

My intention in this report has not been to offer a comprehensive review of the literature, but rather to parse different approaches to understanding the multi-scaled, relational process of uneven development through the lens of global production. Global production ‘without guarantees’ – a conjunctural analysis that historicizes regional change and centers questions of social relations without presuming what these will look like – offers a useful corrective to macro-scale approaches, but neither should the latter be rejected. Moreover, the boundaries between all three approaches sketched out here – the (still relatively heterodox) mainstream literature, and Marxist and neo-Marxist approaches – are not hard and fast. Clear epistemological differences remain; and while we are well served by recognizing that these different approaches, and the distinct priorities that they reflect, are incommensurable, they are also not zero sum. In the best scenario of engaged pluralism (Barnes and Sheppard, 2010), much can be learned by navigating between these approaches, and holding them in productive tension, if we are to gain a handle on the relatively fast-moving, rapidly transforming geographies of uneven development in our time.