A conjunctural approach to global production networks: the case of India’s software services industry

Abstract

1. Introduction

Global production networks (GPNs) are a dominant mode of industrial organization (Henderson et al. 2002; Coe 2004; Hess and Yeung 2006; Mahutga 2014; Werner 2019). Current scholarship typically focuses on established networks. Yet this does not mean that questions about their initial formation have been settled. Instead of rendering the origin stories of various production networks as descriptive context or ‘background’, this article foregrounds questions about their initial emergence. It treats historical trajectories as rich analytic terrains which support theoretical explorations into network formation and industrial restructuring. Production networks have become world-spanning, dominant forms of industrial organization. This article develops an approach to study what sets specific production networks into motion.

The arguments developed here address two limitations that linger within literatures on GPNs. GPN literature does well to prioritize actor strategies. By engaging the situated practices of key actors, it explains how general capitalist imperatives, such as cost-cutting or speed-to-market, result in different regional outcomes. This perspective values historical analysis of how new strategies emerge. What it lacks is an explicit mid-level toolkit and theory focused on the making of historical conditions. Historicizing the formation of different GPNs is not just an empirical or descriptive exercise. More can be done to theorize what actually sets network formation into motion. The shaping of actor strategies after all involves a multi-level, spatially dispersed process that unfolds over time. As such, this article contributes a theory-driven approach to historical causation, and thus strengthens the practice and strategy approach that defines GPN literature.

The second limitation is that the dynamism and instability that defines capitalist political economy tends to slip out of view. The commitment to seeing networks as inherently dynamic, as unstable relations that are changing and changeable, calls to be reaffirmed. Relentless dynamism is inherent to capitalism and this triggers crises in different contexts and fields. GPN scholars analyze the rise of production networks as being a direct outcome of the crisis of Fordism. Yet, the ‘smaller’ more situated crises spread across time and space need to be conceptually mobilized as these are profoundly impactful to specific GPN trajectories.

This article crafts new ways of handling historical causation by treating crisis as a midlevel connective and catalytic force. It looks at how crisis activates new conditions within which actor-specific strategies emerge, thus helping deliver on the promise of the GPN framework to explicate causal mechanisms that connect global structural forces and situated outcomes. Scholars have long rejected overly deterministic views of linear causation, making room for the study of multiple influences (Coe and Yeung 2015). However, approaches that show where and when specific sources of dynamism come from can be further strengthened.

This article deploys conjunctural analysis to speak to the limitations outlined above. Like GPN research, conjunctural analysis also explicitly bridges the dual registers of the global and the local, and the general and the particular (Peck 2017a, 2024; Leitner and Sheppard 2020). It is in fact surprising that there has not been greater engagement between these two synergistic frameworks. At present conjunctural analysis is well developed in certain fields (e.g. urban geography and cultural studies) but underdeveloped in others. Altogether, this paper extends ongoing efforts to bring a conjunctural approach to economic geography (Leitner and Sheppard 2020; Peck 2024).

Conjunctural analysis—as part theory and part methodology—teases out the relations between long-standing contradictions and concrete historical moments (Gramsci 1971). This paper picks up a foundational idea: crises activate a generative potential for the emergence of new social formations. It contributes an analysis of historical causation that helps distribute agency across geographical and temporal registers. It argues that the rise of new production networks and the transformation of existing ones is shaped by extremely disparate, destabilizing events that cut across social domains. Crisis-events activate new cross-regional, inter-organizational relations and alter the conditions which shape actor strategies.

This article illustrates the multi-scalar conjunctural approach by analysing an under-studied context: software and software services. Starting in the 1990s, software work came to be offshored from Global North economies to low-cost regions, specifically India (Aneesh 2006; Upadhya 2016; Peck 2017a; Narayan 2023a). This ‘offshore industry’ reflects the expansion of a transnational, tripartite relationship between corporate customers (lead firms), software suppliers, and software services suppliers. The rapid growth of this network in the early 2000s transformed cities (Goldman 2011), the nature of work (Upadhya 2009), and class structures in India (Fernandes 2006). This study returns to the origin story of this network and discovers an apt opportunity to experiment with ways of connecting abstract capitalist dynamics, actor strategies, and outcomes.

The empirical analysis presented strengthens two additional weak points within GPN literature. It furthers the ongoing diversification of analysis such that it continues to move beyond manufacturing towards services industries (Coe, Johns, and Ward 2011; Faulconbridge 2019), and also shifts the disproportionate focus on the strategies of lead firms to the strategies of supplier firms in the South. The empirical sections survey the global history of software production and identify specific moments of crisis that catalysed links between supplier firms in India and the lead firms of the Global North. The analysis investigates three critical conjunctures that operate across spatial and temporal scales:

Conjuncture I

      USA: Antitrust case against IBM (1969)

      Domain: law

Conjuncture II

      India: balance-of-payments crisis (1991)

      Domain: Economic policy

Conjuncture III

      Transnational: Y2K crisis (1999)

      Domain: computing

The rest of the paper illustrates how these historic conjunctures shaped the trajectory of this production network and consolidates the theoretical gains from doing so. The next section outlines the theoretical approach. After a discussion of the setting and methods, I analyse the three conjunctures and introduce the perspective of two interlocutors who, as industry insiders, witnessed the making of this GPN. The conclusion returns to the core theoretical problem at hand and explains the paper’s contributions to both GPN scholarship and conjunctural analysis.

2. Theoretical background

GPN literature posits an actor-centric approach to the formation and growth of production networks. Many rightly argue that cost reduction, flexibilization, and minimizing time-to-market are powerful imperatives emerging from market competition in a Post-Fordist context (Vidal 2011; Coe and Yeung 2015; Andrijasevic 2021). These general imperatives are seen to explain the strategies and practices of economic actors (Coe and Yeung 2015: 81). With regards to network formation, both state-led drivers and firm-led drivers are important. Each of these two drivers are considered below, making the gaps in analysis clear. First, the limited theoretical scaffold to explain mid-level conditions within which strategies of key actors crystalize and second, the still unfulfilled promise to place instability and dynamism at the centre of the theoretical project.

2.1 The formation of GPNs: state-level and firm-level strategies

GPN formation is said to be driven by state institutions and various political interests (Coe, Dicken, and Hess 2008; Pickles et al. 2015; Hendrikse, van Meeteren, and Bassens 2020). The misconception many attempt to correct is that the state’s role in coordinating economic activity is obsolete due to deregulation (Yeung 1998; Weiss 2018). GPN and increasingly global value chain (GVC) scholars vehemently argue that neoliberal states proactively create investment conditions, policy environments, and workforces (Pickles et al. 2015; Smith 2015; Mayer and Phillips 2017; Werner 2021). Curran, Nadvi, and Campling (2019), for example, argue that a tariff regime is a key mechanism that configures a GPN. Competing tariff regimes influence lead firms’ sourcing decisions. As such, the state’s role has been recognized and theorized in a number of ways: as a facilitator, regulator, producer, and seller (Horner 2017). The institutional environment that creates export-orientation in developing countries frequently finds mention (Baglioni et al. 2022). Mayer and Phillips (2017: 135) argue that states are active agents, that is, ‘intentional architects’ rather than ‘unwitting enablers’. An important point here is that states broker GPN formation not only at national but also at supranational scale, as seen in the role of multilateral trade agreements (e.g. NAFTA in 1994) (Mayer and Phillips 2017; Horner and Nadvi 2018). Preferential policies, developmental agendas, and trade agreements are foundational in architecting transnational linkages in various contexts (Smith 2015; Mayer and Phillips 2017). Additionally, the law plays a strong role in how GPNs are formed. Inter-organizational links and make-or-buy decisions are legally constituted (The IGLP Law and Global Production Working Group 2016). Furthermore, disparities in labour law allow for labour arbitrage and incentivize global outsourcing. However, the role of law far exceeds this. For example, legal scholars argue that ‘corporate law for the institutional form through which business is conducted, and contract law for the interface between legal entities’ (The IGLP Law and Global Production Working Group 2016) are both important. The main point here is that state processes and related actors drive the formation of GPNs using different tools and mechanisms.

Firm-level imperatives and strategies are also critical to the process. Despite ‘firm-centrism’ being criticized within GPN scholarship, firm practices matter a great deal to the process of GPN formation (Coe, Dicken, and Hess 2008). A production network is, after all, an organizational arrangement that is coordinated and governed by a lead firm (Yeung and Coe 2015). Externalization of production processes—in contrast to vertically integrated production systems—allows lead firms to cut costs and mitigate risk in the face of fluctuations in demand, disruption, and breakdown (Sayer and Walker 1992; Vidal 2011). The overarching firm-level priorities are clear with flexibility, cost-cutting, and speed being key imperatives (Coe and Yeung 2015). This explains why, but not how production networks form (Coe and Yeung 2015).

GPN formation is shaped by firm-level strategies. Lead firms create joint ventures and subsidiaries, forge partnerships, and enter into complex contractual relations to support their outsourcing exercise (Coe and Yeung 2015). Supplier strategies, although less intensively studied, are also important (Zheng 2024). Suppliers might in certain contexts, depending on what Yeung and Coe (2015) call cost–capability ratios, act as strategic collaborators with highly specialized expertise. GPN formation involves creating production and labour capacity in new regions. Firms play a strong role in how this capacity is created (and captured). Some scholars focus on knowledge and technology transfers that occur via lead firms (Grillitsch, Asheim, and Trippl 2018; Trippl, Grillitsch, and Isaksen 2018). Firm-driven strategy might also involve forging links to educational systems to support workforce creation and skills development (Gorachinova and Wolfe 2023). Importantly, firm strategy and decision-making are not shaped in isolation, but in response to competitive conditions, financial logics, labour regulation, and pressures to mitigate business risks (Coe and Yeung 2019). While firm-level agency might be complicated and contextualized, firm strategies are profoundly influential vis-à-vis GPN formation (Afewerki 2019).

This separation of state-led and firm-led drivers is an analytical exercise. Many integrate analysis as it is widely agreed that firms are embedded in broader institutional and social environments (Henderson et al. 2002). Yet there are differing perspectives on where to place the emphasis, that is, where to primarily locate agency and power. Strategic coupling, as a concept, offers a way to study GPN formation without being overly concerned about the state–capital binary. This frame shows how specific local actors and local assets come to be entwined with specific lead firms (MacKinnon 2012; Yeung 2016; Lim 2018; Dawley, MacKinnon, and Pollock 2019). Instead of starting with a search for enabling factors and drivers, ‘strategic coupling’ presents a processual approach to network formation.

The problem of how to take powerful structural drivers seriously while maintaining an indeterminate theory of historical change and actor-centric action is a stubborn one. GPN literature does well to recognize the breadth of relevant actors, even adding consumers, civil society organizations, labour associations, and universities to the roster of relevant actors (Yeung and Coe 2015). Yet, this enduring theoretical problem cannot entirely be solved by adding to a growing list of actors, strategies, and network coalitions to the mix. Hinging analysis of actor-specific behaviour on the concept of ‘network embeddedness’ can be limiting (Peck 2005; Jones 2008). Actor strategies are certainly shaped by their institutional and economic contexts. But when and where does the impetus emerge in immensely different sectoral and regional contexts? We need to better bridge these two levels of analysis: abstract dynamics on the one hand, and concrete action on part of actors on the other.

A close consideration of mid-level crises and moments of turbulence in different social domains can help bridge this gap. Training our eye towards the catalytic effects of crisis can strengthen GPN literature in a second way: by infusing in it a more dynamic and unstable view of capitalist industrial organization. The very premise of GNP 2.0 was to confront the under-theorization of the origins and dynamics of production networks (Coe and Yeung 2015). I mobilize conjunctural analysis to advance ongoing attempts to address this.

2.2 Bringing in conjunctural analysis

Conjunctural analysis centres the instability wrought by underlying social contradictions. It views new social relations as being outcomes of both chronic and escalating tensions and also seemingly “random”, unexpected crisis-events. The notion of a historical conjuncture is a foundational idea that anchors conjunctural analysis. Although it has different meanings for different thinkers, a ‘conjuncture’ can be understood as a discernible, concrete moment or crisis and a nebulous birthplace of emergent social formations. Above all, conjunctures must be identified as historically rooted moments, rather than abstract ‘isms’. For Hall, a conjuncture marks not an epochal shift but a tense moment through which shifts in regimes of capitalist accumulation occur (Hall 2006: 231).

Although it is traced to the writings of Antonio Gramsci and then Louis Althusser, this approach really came into its own with Stuart Hall’s landmark study of the rise of Thatcherism in Britain (Hall et al. 1978). For Hall, Thatcherism was a radically distinct cultural, political, and economic movement—a new rupture that then became entrenched (Chen and Morley 2006). Much like GPNs, neoliberalism is a global social phenomenon, but it takes different forms and has its own unique situated contexts in different regions.

Stuart Hall’s legacy is seen in current efforts within geography to read social forces in a fluid manner as they run across social, ideological, economic, and political registers, rather than adhere to silos prescribed by the boundaries of academic traditions (Inch and Shepherd 2020). However, in light of four decades of economic globalization, scholars have been tasked with refining conjunctural analysis such that it is more sensitive to global interconnections and relationality (Leitner and Sheppard 2020; Goldman 2023; Peck 2024; Hart 2024). For Sheppard (2022: 15), this involves ‘examining how a particular territorial conjuncture is shaped also by events elsewhere, how conjunctures concatenate across different geographical scales, and the variegated nature of conjunctural moments across space’.

Conjunctural analysis helps unspool notions of time and space in constructive and creative ways (Hart 2018, 2024). It encourages new accounts of causality, where forces that contribute to new social formations are highly distributed, both spatially and temporally (Peck 2024). Although competitive dynamics represent a general force, the actual connective thrust comes from an array of disconnected and distributed crises. Some scholars have used this approach to forge analytic connections between urban sites in very different world-regions, asserting for example that the 1997 Asian financial crisis, the 2008 global recession, and regional housing crises in the Global South aligned elite interests in different local sectors (Goldman, forthcoming). This is not to say that each crisis-event neatly causes the other, but that these create escalating conditions for new interests to coalesce. Seeing elusive connections between diverse events requires relational thinking (Hart 2024; McMichael 1990). A relational approach links together highly concrete, situated conjunctures that defy easy categorization without making direct causal claims, thus eschewing linear historiographies (Hart 2024; Pickles et al. 2015).¹ For this reason, this new branch can be termed the ‘relational-conjunctural’ approach (Goldman and Narayan 2021).

Conjunctural analysis has much to offer GPN literature. The crux of this approach is a recognition of dynamism and instability across various registers. In actively searching for crises and their effects, it presents a unique pathway to the issue of historical causation. The dispersed and distributed approach to causation speaks directly to GPN literature’s difficulty in bridging the gap between structure and practice. By borrowing from conjunctural analysis, GPN literature might better situate actor-strategies in the aftermath of mid-level crises as these unfold in different social domains and different sites. The choice on the part of GPN scholars to opt for the ‘network’ rather than ‘chain’ metaphor arose precisely from a commitment to a more sensitive treatment of agency and causality (MacKinnon 2012: 229). The rest of this paper helps follow through on that objective.

3. Research setting, approach, and method

Empirically, this study is concerned with the historical rise of India’s offshore IT industry, widely recognized as a major outpost of the global corporate computing sector (Parthasarathy 2004; Upadhya 2016; Peck 2017b; Narayan 2022). India’s IT industry does not export computing hardware. The phrase ‘IT industry’ refers to software and software related corporate services and to a lesser degree software products (Narayan 2023b). This IT industry includes both Indian and foreign IT companies with operations in India. It directly employs 5.4 million workers and has thus far has generated $254 billion in revenue from exports. Corporate customers of IT suppliers in India include the likes of Boeing, Sainsbury, Sony, Barclays, and Nestle. Almost 80 per cent of India’s IT industry revenues come from exports (Parthasarathy 2010). In 2020–21, 58 per cent of those revenues came from the USA and 28 per cent from other advanced industrialized countries (e.g. the UK, the Netherlands, Germany, Japan) (Roy 2024).

How is the story of the rise of this network best told if we eschew a purely descriptive, linear history? This is where the search for conjunctures in the form of crises begins. Those who develop conjunctural analysis are clear on the fact that it is fundamentally a creative, anti-positivist method that does not prescribe rigid practices (McMichael 1990; Peck 2024; Hart 2024). Peck (2024: 3) constructs ‘a methodological remit for conjunctural analysis in economic geography’, stating that it is ‘closer to a critical ethos or “craft” practice, rather than a codified method per se’. It is at its core, a way of historicizing the present. If there are prescriptions at all, it would be an a commitment to: (1) the granular, situated details of specific crisis-events; (2) a wariness to overwrite these in a heavy-handed fashion; (3) avoid highly abstracted, monocausal explanations; (4) a reading of conjunctures across different social domains; (5) a ‘focus on unobvious connections’ and (6) refraining from a stilted application of theory to an ‘illustrative case’ (Peck 2024). As Peck (2024: 18) notes, in lieu of ‘portable routines’, what we have are a set of epistemological obligations.

In what follows, I construct a narrative of three very distinct and extremely varied conjunctures, and then using the voice of two interlocutors, I show how these inflect lived experiences and practices. As noted, conjunctural analysis is a historical methodology. Depending on how far back in time one chooses to go, qualitative interviews, ethnographic insight, and even oral histories might not suffice. Therefore archival research or redeployment of existing historical analysis is important, then integrating it with ethnographic work and qualitative research is useful. Moreover, given that conjunctural analysis and GPN scholarship do not respect the rigid demarcations between state, economic, social, and ideological phenomena, this is fundamentally an interdisciplinary experiment, where methods and analysis are borrowed and repurposed. Taken together, this approach encourages researchers to proactively look across social domains and to highlight how conjunctures inflect actor practices (and lived experience).

I construct an analysis of three conjunctures to tell the story of the rise of this software and services GPN. The first, an antitrust case in the USA against IBM in the 1960s, relies on the work of business and computing historians. I lean on the expertise of historians via published works and also on research institutes. Specifically, the Charles Babbage Institute (University of Minnesota), which holds major archives in computing history. I was a CBI fellow (2019–20) and although I did not conduct archival work per se, my analysis of this conjuncture was improved by the institute’s expertise in the history of computing, through workshops (e.g. ‘Just Code’, 2020) and other feedback sessions with the director, Jeffery Yost, a historian of the IT industry (see Yost 2017).

To outline the second conjuncture, I primarily draw on the work of political economists who have expertise on macroeconomic and currency crises. In particular, political scientists (Kohli, 2004; Sengupta 2009) have done important work on India’s 1991 balance-of-payments crisis and how it altered the structure of the Indian economy. I forge a link between this rupture and the rise of the IT sector using the work of economists and sociologists (Balakrishnan 2006; Upadhya 2016; Goldman forthcoming).

With the third conjuncture, the Y2K computing crisis, I blend the history of computing (Sharma 2015; Yost 2017) with findings from my multi-year qualitative study on India’s IT industry, which I undertook in Bangalore, the country’s tech hub. Between 2014 and 2019, I conducted 110 formal interviews and through the ethnographic process interacted with many more people in the industry. The analysis here benefits from an extensive fieldwork effort carried out across eight firms and workers’ rights meetings, and includes in-depth interviews with managers, technology journalists, recruiters, and trade analysts. The broader research project studied the contemporary transformation of the software industry. This paper however focuses on formation (rather than transformation).

The empirical section of this paper ends with a narrative of the rise of this global network, using the voices of two Indian IT professionals, Kamala and Pavan (names changed), each with 30 years of experience in the sector. This section is vital not only because it offers a fresh analysis of the conjunctures and their effects but also because it shows how the search for conjunctures help to mediate between ‘the epochal and the everyday’ (Peck 2024: 4). While the larger research plan followed a more conventional approach to qualitative research, this paper represents an experiment in narrative-based analysis that melds oral history with the history of computing and the history of Indian political economy. The ethnographic portion puts a premium on the richness of lived experience and depth of engagement with qualitative data.

Key interlocutors are people who, given their social location or institutional position, possess unique insights. Strong relationships and frequent interaction become critical to the overall depth of analysis. Over a three-year period, I conducted multiple interviews with Pavan and Kamala that took the form of oral history and pieced together a recent history of the industry, its emergence, and then its swift and staggering expansion in the 1995–2010 period. In their biographies, life choices, and organizational experiences, I read the shifts associated with the conjunctural events I identify.

The intention is for the three conjunctures to be read together, not as context but as active moments that generate new forces, directions, and impetuses. Pavan’s and Kamala’s voices are meant to bring the effects of the conjunctures alive rather than represent the core of the paper, and to hit home the fact that lived experience, legal history, macroeconomic crisis, and computing malfunction need to be read together—no matter how unruly the narrative gets.

4. GPN formation: three conjunctural moments

4.1 Conjuncture I: the antitrust crisis (1969, USA)

The first task is to understand when and how the computing sector came to be divided between computing hardware, software products, and software services. This segmentation is essential to the structure of the global contemporary computing services sector. Delineating this segmentation asks that we pull back in time. Historians of computing offer the first clues.

The computer, as a machine, was first deployed to serve the computational needs of war (Campbell-Kelly et al. 2013). After WWII, the market for the computer had to be re-created; consultants, along with firms in the computing industry marketed the computer as a technology that rationalized information-handling bureaucratic functions. It quickly came to be regarded as a ‘business technology’. In this period that IBM came to dominate the computer industry, reigning as the sector leader from 1950 to the 1980s (Ceruzzi 2003). IBM enjoyed about 70 per cent of market share (Ceruzzi 2003). Essentially, the corporation sold computers and added the software for ‘free’.

Over several decades IBM grew to the point that it attracted the attention of the US Department of Justice for using anti-competitive tactics (Campbell-Kelly 2003; Yost 2017). During the lengthy investigation, one of the many allegations levied was that IBM ‘bundled’ very different components of the computer, selling some elements below cost and thus hindering competition. Under immense pressure, in 1969 IBM made the landmark decision to price software and hardware separately.

The mounting pressure on IBM is reflective of a deeper contradiction between monopoly and competition, which always produces an unstable industrial terrain. The regulatory domain often becomes the site of contestation, mediating the tension between competitive and monopolistic tendencies (Christophers 2016). This legal attack on IBM’s practices represented a brewing corporate crisis. The mounting regulatory pressure led to a watershed moment for the corporate computing sector. IBM’s decision to unbundle the sale of software from hardware was not a voluntary strategic decision that occurred in a vacuum, but rather was the result of a corporate crisis triggered by the pressure of regulatory action. It was under duress that IBM implemented a split between hardware and software sales.

The unbundling decision was a foundational moment, marking the birth of the modern software industry and triggering a turning point that altered the structure of the computing sector. Up to this point the instructions to the machine (‘software’) simply came along with the machine or else had to be written by the user. In Marxian terms, it possessed no exchange value in its own right and therefore did not constitute a separate part of the social division of labour. For the first time IBM put a price on software. This move created a revolutionary fissure in the labour process of computer-making (Morris-Suzuki 1984).

The effect of this development was felt immediately. Historian Campbell-Kelly notes that ‘[i]n 1969 the total revenues of software products companies had been variously estimated between $20 million and $50 million; after unbundling, analysts predicted the industry would grow much faster than the computer market as a whole, to $400 million by 1975. This prediction turned out to be substantially correct (Campbell-Kelly 2003: 64)’. The scale of growth in the software products industry reflects the fundamental changes in the production, sale, and use of software. The unbundling decision was also accompanied by reductions in the cost of computing and the widespread computerization of firms. An army of consultants had successfully established the computer as a business technology in the post-war period (Campbell-Kelly et al. 2013), and lead firms were heavily computerizing inventory management, payroll, accounting, and manufacturing processes. Computing historians closely tracked the software products market after the antitrust lawsuit. A mere three years after IBM unbundled the sale of software from hardware, a vibrant software services market was established (Campbell-Kelly et al. 2013: 187). For example, by 1972 eighty-one vendors were selling 275 software packages for the life insurance industry alone (Campbell-Kelly et al. 2013). The boom in software products in turn catalysed the expansion of another branch of the computing industry: software services (Yost 2017). Software products need to be implemented, customized, and routinely upgraded (Ensmenger 2010). This is labour-intensive software work. Recognizing a tremendous business opportunity, a software services sector grew rapidly (later as we will see was outsourced to India).

Conjunctural analysis turns our eye towards concrete moments of crisis. The antitrust case triggered an organizational crisis, catalysed through the legal and regulatory domain. The organizational crisis induced by the threat of punitive action activated the birth of a tripartite industry separation that gradually developed after IBM’s 1969 unbundling decision. This then led to the explosion of corporate software applications and programs (‘products’). As the global software products sector grew, the demand for maintaining and customizing these also grew. This transformed the nature and volume of software works and created the potentiality for a new social division of labour. IBM’s unbundling decision represents the first conjunctural event, as it transformed the structure of the computing sector. The question of how India became a vital node in this sector brings us to the next conjunctural moment.

4.2 Conjuncture II: the balance-of-payments crisis (1991, India)

The first conjunctural moment—the US antitrust case—catalysed the contemporary structure of the corporate computing sector, defined by a separation between software products and software services firms. Yet, it does not explain how and why India emerged as the main outpost site of software services. Conjunctural analysis demands we look across temporal and spatial scales at seemingly disconnected crisis events, attending to their spatial and geographical dimensions. Once again we go back in history, this time situating the crisis event not in the Global North but in the context of the Indian political economy.

After attaining freedom from colonial rule in 1947, the newly independent country ventured on a path of state-led development (Kohli 2004; Chibber 2011).This involved strict regulations, a dominant public sector, and import-substitution policies. The birth of India’s local computer industry occurred in this policy context. Computing technology was seen as important to the developmentalist agenda. Therefore, ties between Indian universities and the US and USSR, foreign visits, access to research, all played key roles in creating technical know-how in India. For example, a 10-year collaboration between nine US universities (led by MIT) and IIT-Kanpur began as early as 1961, while IIT Bombay was supported by USSR (Sharma 2015: 30), making it clear the cold war politics played a role. In 1983 import duties were reduced by 15%, marking the beginning of a series of reforms and supportive state policies that would essentially give birth the contemporary industry.

A balance-of-payments crisis in 1991 ended the era of economic protectionism as it resulted in radical economic reforms. Reserves of foreign currency in India had fallen so low, that it, seemed India would be unable to ‘repay external debt and keep investable money flowing through the public banks and into the economy […]The government was on the knife-edge of complete default of its payment commitments to the World Bank and Asian Development Bank, as well as to importers of India’s essential resources, such as petrol, chemicals, and food’ (Goldman forthcoming).

This culminated in a set of policy reforms aimed at liberalizing the economy in 1991. Market reforms ‘were in step with the IMF and Word Banks’s standard prescriptions’ (Sengupta 2009: 182). These prescriptions—conditions on loans to address the balance-of-payments crisis—contributed to export-orientation rather than import-substitution. By the 1990s, China, East Asian nations, and countries in Central America had already come to represent the new factories of the world. India set out on an unusual journey with a service-led economy rather than a manufacturing-led path.

Post-liberalization much has been written about India’s growing economy, an economy that saw a 10 per cent growth rate in 2010. With the second fastest growing economy in the world at the time, India was routinely touted as an ‘emerging superpower’. Unlike other postcolonial countries, economic growth was not linked to export-oriented manufacturing. The spike in GDP can largely be attributed to the expansion in communications, business services, and the financial sector (Aggarwal 2012). The output of the information and communication technology sector in a 12-year period (1990/91–2001/02) grew by 37 per cent. We see here a direct link between major policy reforms, shaped in part by World Bank and IMF borrowing conditions, and the rise of export-oriented software services.

The balance-of-payments crisis created an opportunity for Indian entrepreneurs in the nascent software sector to mobilize state support and take advantage of export-friendly policies. Kohli (2006) contends that the era was characterized by the state’s ‘pro-business’ stance rather than a ‘pro-market’ outlook. The post-reform era saw a state that did not so much wash its hands off the economy as ally itself with big business—specifically software firms (Upadhya 2016). In short, the export-oriented software industry expanded with the support and patronage of the state. Balakrishnan (2006) questions the notion that the industry grew because of the ‘benign neglect of the state’ and argues that an enabling policy environment played no small role; this echoes the call made by Mayer and Phillips (2017) to see the state as an active architect of GPNs. The software industry would never have flourished in India without a growing pool of skilled programmers. Highly subsidized technical education produced and continues to produce the main input in the production of software services. The state, from the time of independence, created institutions that produced ‘cheap and abundant manpower’ that gave the industry its competitiveness (Sarma and Krishna 2010: 33). Other forms of state support included tax holidays, subsidized infrastructure, and access to cheap land (Upadhya and Vasavi 2012). The software technology parks (STPs) set up by the government provided firms with the physical infrastructure and the communications infrastructure necessary to produce exportable software services. STPs represent the most important form of direct state support to the sector (Balakrishnan 2006; Sharma 2015).

Thus, the balance-of-payment crisis and the 1991 policy reform were critical conjunctures that altered state–business relationships by privileging export-orientation rather than import-substitution. These measures were seen as quick and easy shortcuts towards economic growth in response to the balance-of-payments crisis. Changes in policy were made through close consultations with (software) business associations (Upadhya 2016). In other words, the export-oriented software sector had the luxury of tailor-made policy conditions not enjoyed by other sectors. Without going deeper into all the forms of post-liberalization support for software exports, it suffices to say that 1991 was a turning point, paving the way for the rapid expansion of an industry oriented towards Global North economies. Long-standing factors such as public investment in technical education following independence and a large English-speaking population are of course relevant, but it took the balance-of-payments crisis to really set the growth of this offshore industry into motion.

4.3 Conjuncture III: the Y2K computing crisis (1999, global)

It might be tempting to entirely attribute the birth of India’s software industry to the activities of an ‘entrepreneurial state’ (Mazzucato 2018). However, the search for conjunctures reveals a third catalyst—this time in the domain of digital technology. Less than ten years after the balance-of-payments crisis, a transnational computing crisis offered itself as a tremendous business opportunity for the still fledgling software services firms in India. The first conjuncture is located in the US and the second within the domain of Indian political economy, the third is resolutely transnational in character.

Computer programs widely used by firms around the world were written in COBOL, an old programming language from the 1950s that was suddenly expected to trigger large-scale computing malfunction with the turn of the millennium. There was a creeping realization that the programming language represented years in two-digit form. With ‘99’ becoming ‘00’ on 31 December 1999, heavily computerized firms had a major problem on their hands, with their software systems being predicted to malfunction.

Conceptually, the point to stress is that the Y2K computing crisis seemingly came from nowhere—it was sudden and unanticipated. This speaks to a core strength of the conjunctural approach. It effectively demonstrates that, while change processes are shaped by broad-based structural imperatives, what really matters to the unfolding of specific relations are ‘the occasional, immediate, almost accidental’ events (Gramsci 1971: 17). The timing and nature of GPN formations remain fundamentally unpredictable and indeterminate, despite being a generalized mode of organization.

The Y2K crisis was not anticipated by the computing industry and its corporate customers, even though the problem originated in 1959 when the programming language COBOL was first created. This date field glitch lay hidden in computers around the globe and across myriad types of organizations that had been systematically computerized (Sharma 2015). In the 1990s, what emerged was a snowballing fear that a massive malfunction event was around the corner. Banking and credit card companies were the first to identify the problem as late as the mid-1990s (Sharma 2015), before it became an urgent cross-sectoral issue.

The crisis presented a debugging task so large and labour-intensive that it far exceeded what the local workforce in advanced industrialized countries could address (Sharma 2015; Mulvin forthcoming). This was an extremely welcome opportunity for Indian suppliers and was successfully used to deepen relationships with lead firms. Moreover, in India, COBOL was still a programming language that was taught in a large number of engineering colleges and departments. In other words, there was a large workforce that was able to take on the task of combing through thousands of lines of code. It was not just the macro-structural imperatives around speed, cost, and flexibility the triggered the expansion of this GPN. Ultimately it was a labour shortage that acted as a key impetus here, rather than merely the search for low-cost labour. Crisis, once again presents itself as a prism, revealing unexpected processes and factors that run the risk of being eclipsed by both macro and hyper-local dynamics, asserting the ability of conjunctural analysis to perform a mid-level, connective analytic function.

The second conceptual point is that the entrepreneurial activities of the supplier firms in India during the Y2K crisis cemented this GPN. In the next section, a key interlocutor expresses a rich worker-and-workplace perspective based on lived experience. He offers insights into the interpersonal relationships that developed, and the inherent tensions that were managed. It helps supplement this with an organizational and institutional perspective. Supplier firms adopted strategies to maximize the opportunity the Y2K crisis presented. Also, journalists found evidence that the Indian state’s Department of Electronics encouraged firms to offer courses on how to fix the ‘millennium bug’ (Sharma 2015). The state and private sector via the industry’s trade association co-led trade fairs in Japan, the USA, and Europe in the run-up to the year 2000 (Y2K). India’s export promotion board rolled out special incentive programmes and the trade association created a new group with the explicit mission to promote India as the destination for resolving the Y2K crisis (Sharma 2015). Thus, there is a direct link between this crisis-turned-opportunity and the revenues of supplier firms in India.

On the side of lead firms, between 1996 and 1999 the international corporate sector invested billions of dollars in fixing the problem (Yost 2017). Within top supplier firms in India’s sector, during the 1998–99 period, bug-fixing vis-à-vis the Y2K crisis accounted for between 20 and 30 per cent of revenues (Sharma 2015; Yost 2017). However, as we shall see in the next section, this opportunity represented a first step towards locking down a much longer-term arrangement with lead firms, the potential for which was revealed by this software crisis. The final empirical section offers a narrative through the voices of two key informants, showing how their experiences are explicitly or implicitly inflected by the above mentioned three distinct conjunctures.

4.4 An experiential narrative of GPN formation

Both Pavan and Kamala are veteran IT professionals who witnessed the inception and then boom of this supply chain. In their lived experiences and narratives as professional workers, we find the reverberating effects of conjunctural events. First, I present Kamala’s experiences. Unlike Pavan, who moved around between different kinds of firms through the course of his career, Kamala worked for a single company and saw how it was transformed by the birth of the GPN. This firm is among the top 3 Indian IT suppliers.

In Kamala’s narration of her early experiences, we see the effects of the firm-level separation between hardware, software, and services that was cemented in the 1990s, a tripartite separation that gradually developed after IBM’s 1969 unbundling decision. After the unbundling decision came the explosion of different corporate software applications. As the global software products sector grew, the demand for maintaining and customizing these also grew. This transformed the nature and volume of software work that was offshored to Indian offices and IT firms.

Kamala described how, before the establishment of export-oriented links to lead firms, in the early 1990s, she would talk to local clients, understand their processes, come back to her office, write code, test it, make changes, and wait with bated breath when the software went live. ‘There was immense learning. You were involved from requirements to implementation.’ This was the era of custom orders. Fledgling supplier companies would write applications for local clients in India from the ground up.

Kamala described her initial experiences from a time before the global economy had established a clear split between software development and software maintenance. This was a division that would only deepen with the rise of offshoring. But in the 1990s, as Kamala eloquently said, ‘there was no separation between creation and repair, this happened together. We built it and we fixed it. There was no difference between making something and fixing it or maintaining. We saw it as together—all the same process.’ In India, this unity between creation and repair changed with the rise of the new packaged software on the market: enterprise resource planning (ERP). Oracle and SAP quickly became the two dominant firms selling ERP products to Fortune 500 and other large corporations.

In the late 1990s, Kamala moved to a newly formed internal division within her company—the ERP unit. This change in occupational position was emblematic of a broader shift in the global ‘corporate IT’ sector. The rise of ERP cemented a sharp and growing bifurcation between software products players and software services players, first triggered by IBM’s unbundling decision. Product players like SAP and Oracle—unlike the company that Kamala worked at—owned proprietary software. They owned the software that they produced, turning it into a generic product that was licensed to customer firms. In contrast, Kamala said, ‘We never owned the source code of the applications we built, we handed over everything to the customer.’ These were custom orders rather than an opportunity for software services companies to monetize a generic solution to sell to other clients. This deceptively simple issue of who possesses the source code shaped a company’s workplace structure, business model, and position in the software value chain. Software became the property of the company that produced it and product companies simply permitted licence holders to use their product.

ERP is an example of expensive, generic packaged software that is industry-agnostic. Any organization with vast bureaucratic divisions can purchase and implement ERP to integrate accounting, logistics, and human resource management. Western companies had already been computerized in a piecemeal fashion in prior decades—what ERP promised was better and more efficient integration.

The growing dominance of new product players reconfigured the environment in which Kamala’s company operated. These were new conditions and the firm had to redefine itself. This new environment created new kinds of work and new opportunities for the growth of software services. As it turned out, standardized software needed to be customized and implemented. This produced a boom in back-office IT work.

Between 2000 and 2010, software services firms in India were transformed into global service providers, tasked with implementing pre-produced, ‘off-the-shelf’ software. ‘We signed agreements with product players’, Kamala continued, ‘we said to them: we will customize this for your customers’. In other words, they formed alliances with product firms, tying their fate to the increased sale of software packages. The logic was simple: more product sales would mean more services contracts. Thus, services firms initiated mass training sessions to ensure the workforce was familiar with the specific products on the market. Indian firms reorganized their internal structures around new products and went on aggressive hiring sprees, recruiting engineers to perform labour-intensive tasks associated with software maintenance.

For Kamala’s company, the change was dramatic. In 2000, it employed fewer than 8,000 people. In 2010, it had 110,000 employees and generated over 6 billion US dollars in revenue. Needless to say, the firm had not merely adjusted to a new products economy, but had also transformed itself. It had successfully claimed a spot in the international value chain as an intermediary between western product companies and their corporate customers. Meanwhile, companies like Dell, IBM, and Accenture, opened up IT services divisions in India, competing directly with Indian IT firms trying to get a piece of the growing IT services pie.

During our conversation Kamala sounded nostalgic as she relived the 1990s: ‘We never anticipated this growth.’ In the phase of rapid growth that followed, her organization became a different beast altogether. New organizational routines, rigid quality certifications, myriad hierarchies, a complex division of labour, and a new hard-nosed, aggressive CEO marked the permanent end of the slower, small-scale era of the 1990s. In 2003, Kamala chose to move out of the ERP division of her firm. She did not want her life constantly upended by the demands of international travel, which was a core aspect of the lives of senior professionals in the industry. Indian professionals who migrated to the US either successfully made the coveted transition from an H-1B visa to the green card or had temporary visas to work ‘on site’.

These were the people Kamala did not want to become, primarily because it would be disruptive to her family life in Bangalore. ‘I realized that there was an expectation that I would travel. I wasn’t comfortable with this. It was a delicately balanced situation at home. I did not want to disturb it by traveling abroad.’ Kamala decided to separate herself from the booming ERP business and moved to a more inward-looking role. Instead of working on the IT needs of client firms, she joined the internal IT division of her own firm, where she worked for the next ten years, upgrading ERP and related systems for her own organization rather than for client firms.

Kamala’s story illustrates the relevance of industry segmentation between software products and software services, and the role of the ERP product in enabling the boom in offshore service work. However, in order to really grasp the social mechanisms by which the offshore sector boomed, and its particular timing, we need to visit Pavan’s narrative, which similarly acts as a lens into network formation.

***

Pavan, an IT professional with 35 years of work experience, is also based in Bangalore and over his long career worked across different firms in this global network: lead firms (IT department), product firms, and services firms alike. Pavan said, ‘My journey with Indian software started in 1987’. He noted that ‘until then US multinationals like IBM mainly looked at India as a market to sell hardware’ rather than as a site for software and services development. Pavan witnessed India’s move from being a minor market for hardware imports to becoming the largest exporter of software services in the world. When I first met Pavan, he described his early experiences of working ‘on site’ at US companies that had begun to directly hire low-cost programming labour from India. At the time he began his career, US companies were starting to recruit Indian software employees to work in US offices based on short-term contracts. This work regime was (in)famously known as ‘body-shopping’ (for more on this see: Aneesh 2006; Xiang 2007). Indian engineers, supported by temporary work visas, physically travelled ‘on site’ to attend to client systems. Indian firms during this time, as Pavan said, ‘just provided customers with warm bodies’. He continued, ‘This was the common parlance: “How many warm bodies do you need?”’ Pavan went on to describe how body-shopping—the practice of employing low-cost IT labour to work in company locations in the Global North on hourly pay—was over time replaced by offshoring—the practice of moving work to low-cost locations in the Global South.

As discussed earlier, the second conjunctural moment saw a shift in state–business relations that privileged software exports. This offers an explanation at a national level. But what happened at the industry and firm levels in India? This brings us to the third crisis discussed.

The real thrust towards subcontracting occurred through a looming crisis. As Pavan said, ‘What really opened up the opportunity for the Indian software industry was a huge fear about Y2K’. In Pavan’s words, ‘Western societies had moved toward computerization across government, industry, and defense—everywhere you had key software programs, which would fail, and you would have a crazy scenario of destruction’. Averting this crisis would demand that millions of lines of code be corrected so as to render computers ‘Y2K-compliant’. Pavan offered a rich description of this period: ‘People suddenly began thinking: oh my god, we are looking at an incredibly dangerous situation where all our computers might malfunction and stop working because they can’t handle the turn of the century’.

The Y2K problem needed to be fixed by a large workforce willing to painstakingly sift through code written in an old programming language from the 1950s. Indian engineers, familiar with COBOL, took on what in the USA was regarded as dirty work—the mechanical, routine, low-skilled work of fixing millions of lines of code written in an old language. In Pavan’s words, ‘People in the US did not want to do Y2K changes and work in the COBOL world, so they were happy if somebody else in a Third World country [was] sitting and doing it for them at very cheap prices’.

This crisis gave Indian companies a foot in the door into international markets and they successfully over time used this opportunity to build trust and credibility with lead firms. Steadily they persuaded customer firms to offshore more complex work, convincing IT managers in the west to entrust offshore IT firms with more and more tasks. Pavan was working in the USA during this critical juncture and witnessed the shifting tides first-hand.

His narrative belies accounts of offshoring and GPN creation that focus on the eagerness of lead firms to cut costs. For example, the relocation of software work to Bangalore was not a smooth, linear, or ‘natural’ outcome of an overwhelming cost-cutting imperative and rational decision-making on part of western firms. Pavan talked about the slow, friction-filled process of persuading IT managers in the USA to relinquish control and trust an external party working in India. Conjunctural moments present corporate elites with avenues to actively persuade lead firms to subcontract in-house processes to them. As noted, this was possible because the Y2K crisis gave Indian managers some leverage. In Pavan’s words:

Over the course of doing work that was not too sexy, you [were] building relationships with key decision-makers in the US. You started angling to get more work and [said] ‘Look, you can see that I’ve got good people who can do good work. I have started looking inside your systems and I’ve looked at your code base. I can help you fix a lot of the other problems too, at a very cheap rate.’

On the part of Indian firms, Pavan’s account demonstrates the agentive process of recognizing and maximizing the potential of a conjunctural moment. In retrospect, it is tempting to see the offshoring of IT to India as a natural and inevitable outcome, the result of rational choices of US firms, aided by technological advancement that allow for remote working. If anything, these transnational networks were forged not just because of efficiency-seeking client behaviour but because of the Indian manager ‘who [was] sitting right there in the same office as their clients … constantly building relationships and looking for ways to grow their [own] business’. Pavan talked about the important function of these ‘bridge people’:

You want to show your best face to the customer. They [Americans] didn’t know if [India] is the land of snake charmers or a bunch of guys who actually know computer software. These people in key roles at US firms were helping build that confidence that says, ‘Look, I am speaking to a person who appears intelligent and smart, understands the domain, understands what goes into the building of systems. They will act on my behalf.’ You need to build this confidence.

The relocation of industrial and office work typically focuses on the agency and cost-saving drive of the US firm. However, the role of entrepreneurial activities on the part of the business elite in developing countries should not be ignored. Indian firms recognized and capitalized on the Y2K crisis as an opportunity and soon Indian IT workers became a credible threat to their US counterparts. While it is not the overriding force, the appeal of major savings in costs is important to this story. Pavan noted that US firms were essentially presented with the option of using Indian engineers for the same level of pay as someone ‘flipping burgers in McDonalds or in a retail job at Walmart’. Contracting Indian IT labour meant ‘you were getting college-educated kids who put in much more than eight hours for that much. It’s a great deal.’ Pavan talked about this growing sense of unease in the USA. US programmers and middle managers began to recognize offshoring as ‘a growing threat to their jobs’. As he observed,

It’s very disconcerting. Very scary. They started looking at their own senior management as people who were not looking out for their best interests, but only looking out to maximize the profit of their organization. So Indian companies had to deal with that [apprehension] also.

This messy transition had to be effectively managed so that the silent haemorrhaging of software jobs in the US occurred without attracting the attention of too many journalists, lawmakers, and politicians. Over a short period of time, the low-end, mechanical task of debugging code paved the way for the offshoring of high-level work that was more specialized to specific clients and specific software systems. Software services firms in India became increasingly involved with payroll, accounting, inventory, and supply chain processes and systems of management.

From the perspective of the US workplace, the insecurities and simmering resentments needed to be deftly handled. Pavan said, ‘You have to play the game of keeping local guys and remote guys apart, to keep everyone satisfied and not at each other’s throats’. To avoid attracting too much attention, Indian firms and their employees had to maintain a low profile while at the same time gaining the trust of their corporate clients. Pavan said,

It is a tricky thing. They have had to negotiate it in a variety of ways without making too many statements, or doing things that can further increase those concerns. It’s a part and parcel of this offshoring drama that has been going on for the last 20–25 years.

Today most Indian employees do not migrate physically to client sites; they ‘migrate virtually’ (Aneesh 2006). That is, they work on client systems remotely from offices in India. Although it would make financial sense from a cost perspective to have everyone work from India, IT firms configure the right balance of on-site/off-site work. IT managers need to work closely with their clients, making sure their sales and pre-sales teams are actively winning contracts that can then be executed by delivery teams in India. Now the ratio of off-site to on-site work is roughly 70:30. This means that in any given project that needs a hundred Indian employees, thirty will work in client locations and seventy in cities like Bangalore and Hyderabad. To get a sense of the scale of growth that these developments activated, consider the fact that sales from software exports rose from $164 million in 1991–92 to $6.2 billion in 2001–02 (Sharma 2015: 181).

As much as a fast-changing software sector seems to cast the 1990s and 2000s as ancient history, it is important to establish a strong analysis of the transnational underpinnings of the software sector. High-end service-based supply chains have received relatively limited attention in the supply chain literature, despite having unique characteristics. Therefore, the global software sector—its inter-organizational and cross-regional coordinates—needs to be explained at a time when it appears that ‘tech’ begins and ends with Silicon Valley. As Peck (2017a) notes, there is a corporate interest in obscuring the volume of IT work that has been offshored to firms and workforce based in India. Job displacement in the Global North through offshoring is potentially controversial. Now firms like IBM and Accenture employ more people in India than in the countries they are headquartered in. The total revenues of India’s industry, over a five-year period (2000–05), increased by 200% (Balakrishnan 2006). As noted, almost 80% of those revenues come from exports.

5. Conclusion

How do we parse the structural imperatives, enabling conditions, random windows of opportunity, and agentic practices that each operate at different scales? As demonstrated, analyses of concrete conjunctures offer a way to do this (see Table 1). An analytic treatment, rather than descriptive treatment of concrete historical moments, shows where exactly the mid-level potentiality for new social formations comes from. Thus, to contribute to GPN scholarship, I use conjunctural analysis to ‘invoke an “eventful” treatment of (always messy and mediated) causation’ (Peck 2024: 5). The wide, overly capacious gap between the grand structural forces of capitalist accumulation on the one hand, and actor-specific strategies on the other leaves much room for mid-level theorizing.

In this study, the first thrust towards the formation of this global, corporate computing network came from IBM’s decision to unbundle the sale of software and hardware. Using the work of technology and business historians I show how a boom in the software and services industry followed the unbundling decision. The second conjuncture, occurring in a completely different context with its own pre-history, gave birth to another mid-level potentiality: for India to emerge as the main site of offshore software work. Although much has been discussed about the deregulation of India’s economy and subsequent rise of its new policy environment, this is a reminder that a currency crisis preceded it. Foreign exchange reserves in India at the time would not have been able to finance more than three weeks of imports. This crisis in postcolonial India has its own independent history, its own inner contradictions and clashes, but what is important here is that it opened up the field for a new set of interests and institutional structures in support of a new GPN to emerge. This is exactly what conjunctural analysis offers—looking for unstable moments at different levels and domains that ‘open up’ existing relations. It would be preposterous to claim a linear or causal connection between IBM’s decision to unbundle the sale of software and India’s balance-of-payments crisis or even to the Y2K crisis, yet these are connected as critical conjunctural events for this specific GPN. The third crisis is one that is deeply unique to the computing sector and one that did not have a clear spatial site; in that sense, it was a transnational crisis. Here we see that the deployment of Indian supplier firms’ growth strategies really mattered. For example, we see how IT workers and managers proactively built trust and assuaged various social and political concerns, co-creating IT offshoring as a dominant business practice. This helps redress GPN’s relative neglect of supplier strategies (Zheng 2024).

This article contributes to GPN scholarship in other ways. The GPN literature has not yet fully delivered on the promise of treating networks as a set of dynamic strategic relations, always adapting and evolving (see MacKinnon 2012). A recent global pandemic made the need for economic geography to be attuned to crisis theory obvious (Vachan 2025). I show here that conjunctural analysis can help make discrete sector and regional crises more visible. The manifold effect of crises at different scales and in different spheres of social life is unmissable, if we use a lens that helps to see these.

GPN literature’s focus on actor-specific strategies is really important, but on its own the treatment of agency can seem overly rationalistic (MacKinnon 2012), giving too little importance to major events outside the domain of control and planning. Here I mobilize analysis of concrete conjunctures that complicates the domain of agency, offering a more distributed view of causality. A capacious and creative treatment of time and causation can help deliver against the mission to discard linear narratives from GPN analysis. To do so, I draw connections between a range of spaces, sites, world events, and social domains that each operate at different scales, without bracketing out the ‘region’ (i.e. the city of Bangalore) as the primary site from which to then demarcate exogenous and endogenous factors.

In utilizing conjunctural analysis, this article also contributes to its further development. When it comes to the various aftermaths of entangled crises, the empirical analysis here makes an important contribution to conjunctural analysis: it is not necessarily crisis itself but the anticipation of crisis that catalyses institutional and actor-specific action. IBM ‘chose’ to unbundle hardware and software as pressure intensified. The policy and political elites in India ‘chose’ to enact reforms before the rapidly advancing balance-of-payments crisis actually hit. The Y2K bug was removed from millions of lines of code before clocks struck midnight on 31 December 1999. This tells us that the pressures of a looming crisis are themselves enough to trigger social action. The multiple tangled temporalities invoked by conjunctural analysis (Peck 2024) might lead us to consider the powerful force of institutional and organizational anticipation, and further develop a sociological engagement with decision-making. As such, the two interlocutors introduce an oft-neglected organizational lens.

Finally, conjunctural analysis should not be separated from what Gramsci terms the organic elements of crises: the penetrating, translocal, structural contradictions. According to Gramsci the organic aspect of a crisis results from long-standing and enduring contradictions which intensify over a longer historical arc, while the conjunctural dimension of crises refers to the facets that are ‘occasional, immediate, almost accidental’ (Gramsci 1971: 177). As the term ‘conjunctural analysis’ itself indicates, the focus is on seemingly random concrete events—situated episodes in history—rather than abstract contradictions. It is however possible, indeed wise, to simultaneously hold on to the idea that GPN formation at large was triggered by the profound contradictions of Fordist capitalism and also the mid-level potentiality engendered by idiosyncratic, situated conjunctures.

Footnotes

Acknowledgements

Thanks are due to Rutvica Andrijasevic, Victoria Hattam, Jennifer Johns, Andrew Sturdy, and Jeffery Yost for their comments. I am grateful to my doctoral committee at the University of Minnesota for their unwavering support with this study. An early draft of this paper was presented at the British Universities Industrial Relations Association’s conference (2023).

Funding

None declared.

References