Disconnecting Labour? The Labour Process in the UK Fast Fashion Value Chain

Introduction

The rise of buyer-lead firms and their global sourcing portfolios has shaped the specific form of disaggregated production within the apparel value chain (Gereffi, 1994). Buyer-lead firms extract value through brand and marketing rents, as well as rents stemming from organising global sourcing networks, which put manufacturing locations into competition with each other over different combinations of quality, price and speed-to-market. A conundrum of global sourcing, however, results from the need to adjust product and process standards along the value chain to the requirements set by lead firms. The wide range of codes aiming to regulate minimum labour standards in apparel manufacturing – from lead firm and industry to multi-stakeholder codes – testify to the intractable nature of coordinating standards along the chain (Locke, 2013).

This article investigates how the dynamics of value chain governance are interlinked with processes at workplace level (Newsome et al., 2015), specifically focusing on the emergence of fast fashion in the UK. Fast fashion denotes business strategies that ride on a quick response to consumer preferences, supported through lean retailing and supply chain management, and are often driven by new types of buyer-lead firms such as online-only retailers (Taplin, 2014). The quick response advantage of fast fashion also relies on the geographical proximity of production and consumption and was behind the remarkable, somewhat counter-intuitive, growth in UK apparel manufacturing in the aftermath of the 2008 financial crisis. The real approximate gross value added (aGVA), for example, grew 61% UK-wide and by 58% in the East Midlands (2008–2016), with the latter growing into the largest sourcing hub for fast fashion. ¹ The fast fashion value chain offers a useful case to analyse the interlinkages between value chain governance and the labour process as it imposes a different set of requirements on supply chain actors, including the workplace: low prices, low–medium quality, very low batch production and quick response.

These interlinkages can, however, manifest as a disconnection (Thompson, 2013), for example, between the wider political economy and the labour process, or, in other words, between the standards set by lead firms and actual workplace practices. Parallel to the increase in UK apparel manufacturing there has been a growing unease about exploitative working practices in the industry, starting with insights from an undercover report in a Channel 4 programme (Fashion’s dirty secret, 2010; Undercover: Britain’s cheap clothes, 2017) to research by Hammer et al. (2015). In a more immediate sense, stakeholders have identified unauthorised subcontracting (i.e. subcontracting out of sight of lead firms’ social auditing practices) and the violation of minimum work and employment standards as key problems for the sustainability of the industry. This article offers new empirical insights into the labour process in the UK’s largest hub of fast fashion production and discusses this in the context of previous research on the labour process, ethnicity, small firms and supply chains in UK apparel manufacturing (Phizacklea, 1990; Ram et al., 2003, 2011).

This article contributes to debates on labour, global value chains (GVCs) and disconnected capitalism (Newsome et al., 2015; Thompson, 2013) by analysing the tensions across two dimensions of competition, or capital circuits, along the functional division of labour and in product markets (Pickles et al., 2016). This perspective, it is argued, can explain how the accelerated circuit of capital within fast fashion offers new opportunities at the same time as it is based on unsustainable conditions such as new patterns of segmentation, the elimination of unproductive spaces in the labour process, and further increases in the precarity of labour.

The following section of this article assesses different approaches that situate the labour process within processes of competition. The next section presents the research design and methodology that allowed investigating work and employment as embedded within GVC and product market dynamics. The subsequent sections deal with different aspects of UK apparel manufacturing: the re-segmentation of firms in the emergence of fast fashion, the role of market volatility in the labour process, changing patterns of wage formation and welfare, and the re-segmentation of the workforce. The conclusion looks at the lessons this re-emerging and thoroughly re-segmented industry offers to analyse the inter-relations between global value chains and the labour process.

Linking the labour process and value chains

Recent research has asserted labour’s central status as value producer and key agent in the functional and international division of labour and focused on ‘putting labour back into value chains’ (Newsome et al., 2015). Exploring the respective insights of GVC and labour process analyses further, we argue, requires a distinction of three different dimensions of competition or capital circuits; that is, the functional division of labour and the product market, the mechanisms that articulate these two dimensions, and their underlying interlinkages in a dynamic perspective.

Behind the first point lies a recognition that any product of a GVC is part of two different circuits of capital (e.g. Pickles et al., 2016: 28–29). One circuit takes place at the level of the value chain: it ensures the integration of different functions (types of capital, legal forms, intermediaries) and stages of production and realises profits in the sales stage. Another circuit takes place at the level of individual firms as they manage inputs in the production process and compete with others to sell their goods and services in the product market. Labour process analysis has made efforts to embed workplace studies within a broader political economy (Thompson and Smith, 2009) but has focused on firm-level capital circuits. For example, it investigated the role of technological change and product market competition for the labour process, with Brown (2008) and Edwards (1986) seeing these relations as indeterminate and interdependent. In a different formulation, Kelly (1985) argued that the analysis of the real subordination of labour needs to look beyond the capital–labour relationship: rather, it is the articulation of different parts within the circuit of capital that allows a comprehensive analysis of changes in the division of labour. What GVC analysis has brought to the discussion more recently is a focus onto the functional dimension of capital–capital relations as well as the mechanisms through which the different stages of the chain-level capital circuit are integrated.

This leads to our second point, which emphasises the articulation between firm-level and chain-level circuits of capital. GVC analysis has connected the process of production and global market dynamics through mechanisms – forms of value chain and inter-firm governance, institutions and standards (Gibbon et al., 2008) – that temporarily ‘fix’ or bridge the disaggregated production and functional fragmentation of GVCs (e.g. most of the contributions in Newsome et al., 2015 and Taylor and Bain, 2005; see Rainbird and Ramirez, 2012 for an emphasis on social institutions). The articulation of firm- and chain-level capital circuits within GVC analysis has, however, neglected the role of labour and the labour process and focused on the relations between firms (Selwyn, 2012). Research by Barrientos and Kritzinger (2004), by contrast, analysed how the dominance of buyer-lead firms in the South African fruit industry led to increased outsourcing in the form of contract labour and thereby widened the disarticulation between the GVC and the labour process. While lead firms set minimum standards for employment and working conditions, the economic and temporal parameters of GVC governance led producers to rely on contract labour, which, in turn, reduced their control over employment standards and the labour process. Thus, buyer-lead firm strategies provided a ‘fix’ for value chain pressures as far as the circuit of capital between buyers and producers was concerned. At the same time, however, the tensions and consequences of this ‘fix’ were played out in another capital circuit between competing producers and contract labour, with implications for the composition of the producers, contractors and their workforces (see also Bair and Werner (2015) on a relational concept of labour in uneven development). This example highlights how the labour process retains a central role in the restructuring of value chains.

This example also underlines our third point, which relates to the central role of segmentation processes in the articulation of chain- and firm-level circuits of capital. Insofar as a specific ‘fix’ is necessarily asymmetric and contested, it fuels continuous changes in capital–capital and capital–labour relations (Grimshaw and Rubery, 2005). Barrientos and Kritzinger’s (2004) research also highlights how the re-segmentation of capital and labour mediates the articulation of circuits of capital across different strands and tiers of the value chain. It is argued here that an emphasis on segmentation helps in conceptualising a disconnected global political economy, going beyond rigidly defined forms of governance, the confines of the workplace, and too unilinear and unidirectional relations between them (Thompson et al., 2015). Labour process and GVC analyses both work with an implicit understanding of segmentation. Both, for example, recognise that market mechanisms are insufficient to, respectively, ensure the extraction of labour from labour power or to coordinate geographically dispersed production processes. Fundamental antagonisms in the labour process as well as across the horizontal and vertical divisions of labour ensure that accumulation dynamics produce segmentation as much as they are the result of continuous re-segmentation processes (e.g. Fine, 1998; Peck, 1996; Rubery, 2007; also, for feminist approaches to global commodity chains, Dunaway, 2014). Segmentation dynamics shape how particular firms and workplaces can compete for positions in the division of labour (Hammer and Riisgaard, 2015). Yet, insofar as they are key in translating GVCs’ quality and performance requirements, they also underpin any disconnections in the functional division of labour.

As will be shown below, for example, the small, frequent but intermittent nature of orders within fast fashion has made it extremely difficult for manufacturers to produce according to the ‘ethical’ standards prescribed by lead firms. In other words, pressures that originate within buyer-lead firms’ intensified circuit of capital are ‘solved’ in the manufacturers’ circuit of capital, yet, in ways that disconnect them from the ethics of the former and threaten the sustainability of the entire value chain. The problem, thus, is not only in articulating the elements within one circuit of capital (which is what labour process analysis has pointed to) but how different circuits of capital can be coordinated in a temporally stable fashion (an implication of GVC analysis). The case discussed below will demonstrate the emerging tensions stemming from fast fashion’s mode of governance as well as how re-segmentation constitutes a solution and problem at the same time.

Methods

Two key challenges within value chain research stem from epistemological and methodological problems. First, there is a question of how to establish a causality of impacts between actors that are not connected through contractual relations and are often separated by numerous stages in the spatial and functional division of labour, part of which is located in the informal economy. The challenge, thus, consists in isolating and analysing the mechanisms through which value chain governance is effected. This has put the emphasis on formal and informal private ‘standards’ that link otherwise fragmented production networks (Ponte and Gibbon, 2005). Furthermore, while value chain drivenness shapes key parameters of the division of labour, it needs to be analysed in its embeddedness in macro dynamics of accumulation as well as in local struggles around the structured antagonism between capital and labour (Edwards, 2005; Thompson et al., 2015).

The present research has tried to tackle this challenge through a dual route, a mixed-methods design (Morse, 2010) that complements research on the supply chain with a sector focus, both focused on the East Midlands, which accounts for one-third of UK apparel production. On the one hand, research was conducted on the concrete supply chains of 10 UK high-street retailers. As they were prepared to share details of their supply chain for this research, it was possible to trace the links between specific buyer-lead firm strategies through to manufacturers. This emphasis on concrete supply chains was complemented, on the other hand, through a focus on the sector in a wider sense, which allowed the lead firm sample to be contextualised within broader product and labour market dynamics. Such a triangulation of supply chain and sector perspectives was crucial in delineating the role of fast fashion in contemporary apparel manufacturing.

The second, methodological challenge arises from the need to analyse and evaluate data from a wide range of firm and labour actors, who all have antagonistic interests and are embedded in different product and labour market contexts. The research gathered data in four interlinked stages (for more detail, see Hammer et al., 2015). First, a database on subcontracting networks and work and employment conditions was developed from publicly available data as well as confidential information from lead firms. Second, these data were evaluated further in 30 semi-structured interviews with major stakeholders along the supply chains of the mentioned 10 high-street retailers, from ethical trade managers and buyers in lead firms to trade unions, industry organisations, and social auditors to suppliers, manufacturers (cut-make-trim (CMT) production, printing, labelling and packaging), workers and their community organisations. Third, information on purchasing practices and production for fast fashion were also gained through a forum that gathered 30 suppliers and first and second tier manufacturers without the presence of lead firms. Fourth, a face-to-face survey was carried out with 30 workers, two-thirds of which rarely or not at all produced for the 10 lead firms in our sample. These interviews were carried out by a community organisation, principally by a female interviewer with a South Asian background, thereby matching key social characteristics of the respondents.

This design allowed to develop a robust empirical basis, one that reached across supply chain functions and appreciated the institutional and social embeddedness of different actors. In particular, triangulating data from respondents at different positions in the division of labour proved invaluable insofar as the engagement and complicity in extensive smoke and mirror games is often crucial for survival, particularly at the boundary of the informal economy. This also has consequences for the way the data are reported below: as respondents routinely concealed strategies that could incriminate them, the empirical basis often relies on material that has already been triangulated and counterfactually analysed.

The emergence of fast fashion in the UK

Apparel manufacturing in the UK responded to international competition through a combination of spatial and functional strategies. The outsourcing of production to rural areas within the UK in the 1970s (Phizacklea, 1990) soon gave way to international sourcing (Pickles and Smith, 2011; Warren, 2005). China quickly became the main importer in the early 2000s, accounting for more than a third of UK clothing imports by 2010. In the same period, Eastern Europe and North Africa emerged as key sourcing destinations, with Turkey being responsible for around 10% of UK clothing imports in the 2000s. Significantly, from 2007 onwards, UK wearing apparel manufacturing reversed its decline and grew its turnover by 110% (2008–2016): turnover increased by almost £2 billion, whereas the UK’s trade deficit in clothing shrunk by more than US$5 billion. These statistics, however, must not hide the existence of different strands within the apparel value chain, subject to different forms of governance. Below we focus on the fast fashion value chain for which speed-to-market is particularly crucial.

At the downstream end, the rise of buyer-lead firms (clothing retailers that neither owned nor directly contracted to factories) was accentuated at the same time as competition increased greatly through global market pressures, the increasing financialisation of lead firms, the intensification of buyer logics, as well as new business models. After branded retailers had to face competition through the market entry of general retailers (e.g. most supermarkets now have their own clothing brands), the latter now have to respond to the challenge of online platforms and online-only brands. The pressures resulting from these developments are evidenced in an enormous price squeeze: consumer prices for clothing in the UK, for example, declined at an annual average rate of 3.4% between 1997 and 2016.

Part of these economic pressures stems from forms of value chain governance that emerged in the late 1990s (Winterton and Winterton, 1997) and was further entrenched through fast fashion (Hammer and Plugor, 2016; Taplin, 2014). Fast fashion lead firms harnessed a host of new marketing channels (linked with social media), organisational changes in lean retailing, supply chain management and quick response manufacturing, and built a business model that challenges more established lead firms with a high-street presence. Insofar as online brands’ only ‘shop’ consists in their online portal, they rely on small batch orders that can be repeated as and when real-time sales data indicate. Fast fashion, therefore, ‘pulls’ production, rather than manufacturers ‘pushing’ goods into sales channels (Taplin, 2014). Sourcing near consumer markets, as opposed to South East Asia, allowed lead firms to quickly respond to market trends (within a week rather than months) and reduced the need for complex logistics, stocks and inventories.

At the upstream end, fast fashion entailed a different mix of quality/price/speed-to-market requirements. It focused on simple products, at very low price, ready for sale within days from the point of order. Importantly, the parameters of fast fashion have implications for the temporal and spatial dimensions of competition: on the one hand, small repeat orders put manufacturers in competition with each other every time such an order is placed (at least implicitly); on the other hand, the quality/price/speed-to-market mix of fast fashion put UK manufacturers in competition with producers from Eastern Europe and the North African rim. While only Turkish apparel imports are significant compared to total UK imports (at 11.2%, 2015), Romania, Morocco and Bulgaria, for example, are important competitors in the fast fashion value chain even if their combined imports only constitute about 13% of the overall turnover in UK apparel manufacturing (which, in 2015, stood at £4.1 billion). ²

Thus, while the rise of fast fashion pulled growth (real turnover in East Midlands apparel manufacturing grew by 98% in 2008–2016), its benefits were not shared equally across the value chain. Manufacturers found it increasingly difficult to meet the temporal, price and quality demands of fast fashion in the face of lead firms’ pricing power. The structure of orders – large orders at no margin or very small batches (trials and repeats) – further entrenched intense product market competition: the overwhelming majority of manufacturers had to chase order volumes below 1000 units, which made it virtually impossible to achieve continuous and economically viable production at full capacity.

Fast fashion was also made possible through the re-segmentation of manufacturers that entailed significant downsizing and downgrading in the context of heightened horizontal competition. The decline in the number of large UK-based manufacturers was accompanied by a parallel growth in small factories, often staffed, managed and owned by minority ethnic workers who lost their jobs in the decline of automotive manufacturing (Warren, 2005). These factories specialised in technical textiles or low-value areas such as socks, knitwear and the CMT of casual wear, tended to be undercapitalised and therefore had little scope to develop more vertically integrated businesses (Winterton and Winterton, 1997). Average employment fell from 19.7 to 7.7 per enterprise (1998–2015), with only 17% employing 10 and more workers in 2015. The integration into the fast fashion value chain saw the East Midlands rise to the largest sourcing hub, accounting for a third of UK apparel production at its height in 2014. Industry statistics point to faster growth in the East Midlands compared to the UK (2008–2016): turnover saw an annual average growth of 9.6%, aGVA grew by 7.6% on average (3% and 4.1% above the respective UK values).

In response to the emerging fast fashion value chain governance and structural changes in the product market, manufacturers resorted to subcontracting (without the knowledge of lead firms, in factories that were not audited by lead firms, and at times in unregistered factories) and the extensive reliance on informal employment. The development of further tiers within the manufacturer base, the fluid combination of authorised and unauthorised production in the same factory in order to maximise capacity, and the simultaneous quotation of ‘compliant’ and ‘non-compliant’ prices as reported by a lead firm buyer, were consistent consequences of the asymmetries in the value chain. Increasingly, the polarisation between monopsonistic buyer-lead firms and manufacturers subject to cut-throat competition could not be reconciled anymore through product and process standards (price, quality, speed, social standards). The next sections detail the consequences of value chain and product market restructuring for the workplace, focusing in turn on the labour process, wage formation and workforce segmentation.

Market volatility and the labour process

The organisation of the UK apparel industry – with low entry barriers, labour intensive work, high levels of competition, global sourcing and extended subcontracting – has subjected it firmly to the vagaries of the market, resulting in power asymmetries in the value chain that were accentuated further through the requirements of fast fashion (Taplin, 2014). These requirements meant that a key function of the wage-effort bargain was to balance order fluctuations with labour supply and the intensity of work. The centrality of the link between labour power and time is exemplified by historical and comparative research. Jenkins and Blyton’s (2017) account of the ‘comp-off’ system in Bangalore garment factories, for example, is reminiscent of Cunnison’s (1966) analysis of the ‘dead-horse’ system at the ‘Dee’ factory. Both were designed to even out fluctuations in the order book whereby wages were advanced in exchange for time debt: power relations ensured, however, that this relationship was never reversed to increase workers’ time autonomy.

Compared to Cunnison’s case study, key parameters of the fast fashion labour process have changed: increases in task division and subcontracting, as well as the shift from piece rates to hourly wages (Ram et al., 2003: 76; Warren, 2005), have altered the way workers manoeuvre to obtain ‘good’ work and maximise their income. In our research on fast fashion manufacturers in the East Midlands, control over the scale and intensity of the labour process was achieved through an extensive informalisation of the employment relationship. This allowed to eliminate unproductive gaps in the order book and production process, and resulted in a more unmediated management authority.

Of the 30 workers surveyed, only four held employment contracts. Informal employment represented an industry norm: 75% stated that people in their workplace normally do not hold employment contracts:

‘The job is as and when required.’ ‘There is no such thing [an employment contract] in the garment sector. This is very unusual.’ ‘We never asked and were never given one.’ ‘I have been working for a long time and have never been given a contract. I thought this is normal.’

Two-thirds of those workers said that their employer did not want to issue a contract, with 40% adding that they themselves did not want a contract either. Part of the latter group, when asked for a reason, pointed to visa restrictions. Interestingly, regardless of their informal contractual situation and the actual flexibility of their employment, just over half of the respondents considered themselves in permanent employment, while a quarter thought of their job as an occasional one, and the remaining 17% as currently out of a job.

An important consequence of informal employment was that the link between market volatility, management strategy and employment became more immediate and direct. The integration into extremely responsive supply chain practices was accompanied by strategies to further increase the extraction of absolute surplus in the workplace. This was achieved, first, through increasing the flexibility in the length and scheduling of work. For example, two-thirds of the survey respondents reported considerable variations in their working time, with regard to starting and finishing times as well as daily and weekly working hours. Asking workers to return home (without pay) whenever the order book or production process demanded, was standard practice. Reasons given were:

‘No new orders.’ ‘Not enough work.’ ‘Machinery and equipment maintenance.’ ‘Some machines were not working, they required repairs.’ ‘They don’t have enough work, that’s why most of us are not working now.’

Second, work in apparel manufacturing is very intense: 80% of the survey respondents were never or only rarely able to take a break when they wished and worked over three hours before they took a break. Stress in the production process was ubiquitous: 43% experienced stress at work always or most of the time, while another 50% were sometimes under stress.

Direct supervision by the supervisor played a central role in securing control over the labour process and tended to exploit worker characteristics. Apart from the taxing pace of production, workers pointed to verbal abuse, threats and humiliating behaviour, as well as bullying and harassment as constituent features of workplace management. These practices were used to secure a certain pace and quality in the CMT process, but also – by discriminating against the mostly skilled female workers based on their language skills and ethnicity – to reinforce the segmentation of the workforce and achieve discipline through ‘divide-and-rule’ practices. For example, workers were abused and humiliated in front of fellow workers for mistakes and missed performance targets (see also Jenkins and Blyton, 2017: 99), and had wages deducted even if the underlying problem was not in their control. Another important disciplinary practice was to regulate the continuity of employment. It was generally accepted that daily employment depended on the state of the order book. However, workers who challenged management were told that there was no work for the time being and were sent home while the rest of the factory seemingly worked at capacity. One respondent told of a practice to make new employees sign a blank sheet of paper which could then be used to ‘dismiss’ workers in the form of ‘resignations’ if market fluctuations or labour-management problems so required.

Findings on health and safety were somewhat ambiguous, possibly reflecting its centrality in lead firms’ or social auditors’ factory inspections. Thus, the most serious health and safety problems might already have been dealt with; equally, workers might have been aware of the sensitivity of health and safety and under-reported incidents and concerns. Almost two-thirds of the survey respondents reported their factory’s machinery, equipment and the factory environment overall as safe and healthy, while the rest considered their workplace unsafe. However, those respondents who did speak of accidents, reported 10 and 14 accidents (within the last year), respectively.

Historically, it was homework that allowed to eliminate overheads and unproductive gaps in the production process, while piece rates provided a productivity incentive for workers. The speed and coordination demands of contemporary fast fashion, however, are met more seamlessly in a factory environment, while informal employment – particularly when pay is based on hourly wages – establishes a supervision incentive for managers, keeps labour costs low and strictly ties labour costs to market volatilities. Still, this system is not free of tensions: in the same way as Cunnison’s workers could not achieve their target earnings when there was less work ‘going around’, today’s hourly paid workers often find it difficult to earn above the poverty line if they are told to stay at home whenever their factory is low on orders. Rather than through a debt relation, as in Jenkins and Blyton’s (2017) and Cunnison’s (1966) cases, the extraction of labour in fast fashion is based on the link between precarious employment (mixing formal and informal employment) and welfare payments. This gives rise to complex ‘games’ around wage formation, welfare and taxation on the part of both employers and workers.

Wage formation and welfare

Workplace studies have not only shown how horizontal competition impacts on the wage-effort bargain (e.g. Cunnison, 1966), but also how the functional division of labour shapes working conditions (James and Lloyd, 2008; Newsome et al., 2015; Ram et al., 2011). Cunnison’s study, written in the tradition of the human relations school, illustrates how workers reconsider the wage-effort bargain for different products and tasks, as well as how those strategies are likely to position them with regard to market fluctuations and the likelihood of continuous employment. A skilled worker in Cunnison’s (1966) ‘Dee’ factory, for example, might have taken on a ‘bad, more complex, job’, which reduced her overall earnings but, by demonstrating ‘good will’, put her into the core segment of the workforce. While subcontracting and the under-payment of wages have long been key features of the industry (see, for example, Phizacklea, 1990), this dynamic of the wage-effort bargain has changed considerably in the context of fast fashion with regard to its level, composition and gender implications.

First, there were unabated downward pressures on wages. Insofar as wages combined formal and informal elements (see, for example, Williams (2014) on ‘envelope wages’), they were difficult to trace in official statistics. However, an analysis of hourly wages in Leicester between 1990 and 1997 shows that some workers started at rates of 17–33% below those set in national bargaining agreements in the garment industry (Leicester City Council/Low Pay Campaign, various years; see also Warren, 2005). These data were confirmed by our findings on pay practices in Leicester’s fast fashion manufacturers in 2014. Wages were paid by the hour and cash-in-hand. While wage slips recorded a certain number of hours (normally around 16–20 hours per week) paid at National Minimum Wage (NMW) levels, actual working hours tended to be far higher. Overall, these practices resulted in a virtually uniformly confirmed industry norm of £3 per hour. This wage norm was differentiated further according to job (packers or less experienced workers, for example, tended to be on lower rates) or residency and right-to-work status (undocumented workers could earn as little as £1 per hour). Deviations from quality standards were subject to wage deductions or extended working hours in order to make up for losses on quality or quantity. The growth of the industry in general, of fast fashion in particular, also accentuated longstanding gender wage inequalities (e.g. Phizacklea, 1990). This is supported by UK wide data on median wages in wearing apparel (from the Annual Survey of Hours and Earnings), which show a gender gap of 19.6% for the period 1999–2007, which rose to 23.4% in the boom period 2008–2016.

Second, problems resulting from low wages were further compounded by practices that considered wages as ‘composite wages’, consisting of one part paid by the employer, which was ‘topped up’ through welfare benefits such as working tax credits, child tax credits and housing benefits. In fact, half the workers interviewed received one or more of those benefits. Views on whether it was employers or workers who ‘drove’ such practices of composite wages varied and were by necessity blurred as they required workers’ agency and complicity. Composite wages entailed a continuing practice of re-negotiation between worker and employer, in addition to the wage-effort bargain, as they also encompassed a social wage-effort bargain with considerable stakes: recorded wages that were too low or too high undermined workers’ eligibility for certain benefits or rights, which gave rise to complicated negotiations. Community organisations reported how their clients often fell foul of regulations (e.g. when they received irregular wage slips) and had to repay working tax credits. Other cases have been reported where, in order to qualify for leave-to-remain status, workers received wage slips recording the required higher number of hours and wages but had to repay their employer what was in excess of the industry norm. Thus, while workers used their agency to make the best of a very difficult situation, the advantages clearly lay with the employers. Not only did these practices reduce their wage liabilities significantly (in 2014, the defrauded weekly wage sum in the East Midlands alone was estimated to around £1 million per week), they also reduced their Pay-As-You-Earn and National Insurance liabilities.

Third, this conjuncture of market fluctuations and informal employment and pay practices reinforced gendered wage and poverty dynamics. For example, considering the social wage as part of the firm-level ‘wage package’ is likely to reinforce gendered patterns of wage discrimination insofar as women are more likely perceived to make up their low wages through welfare benefits. Furthermore, the unpredictability of work – 40% of survey respondents thought of themselves as permanently on-call and were sent home between three and four times per month due to a lack of orders – had serious implications for their income continuity (our respondents’ average wage in July 2014 was £584). It is fitting that, according to minority ethnic community organisations, workers sought support for poverty-related problems such as arrears in rent, utility bills or Council Tax payments in the first instance; it was only in the process that underlying problems of employment and wage theft came to light.

The informalisation of employment came with a shift from a wage-effort bargain to an income-effort bargain with a specific focus on the social wage. Paradoxically, spaces of de-commodification (access to certain public welfare benefits) have made it easier for employers to increase the commodification of labour; that is, by intensifying the labour process and shifting the employment relationship outside the regulatory purview. While workers were able to exercise agency, making the best out of precarious and insecure employment, this agency did not result in NMW level incomes and only rarely offered an escape route out of the industry. From a (fast fashion) value chain perspective, this suggests that the circuit of capital at the upstream end of the value chain could only be closed through a colonisation of the sphere of reproduction.

Workforce segmentation

Management practices of coordination and control, wage and income formation reinforced each other in establishing and reproducing a segmented workforce. These practices of segmentation were certainly not homogenous across the industry as they were contingent on the detailed product market, business strategies and management styles (which create different capacities to respond to price and quality parameters; see also Ram et al., 2011). Nonetheless, regarding the articulation of the labour process into the fast fashion value chain, two key functions of workforce segmentation need to be highlighted. First, and in line with labour process analysis, workforce segmentation served a control function that was expressed in daily negotiations over – actual and recorded – hours, hourly rates, the type of tasks and the intensity of work. Second, workforce segmentation also had a strong competitive function insofar as it was adapted in order to satisfy the varying price, quality and social compliance requirements of lead firms as well as to maintain profit levels in a very volatile market. Unauthorised subcontracting is driven by a lack in order continuity but also constitutes a strategy to increase the level of absolute surplus value. Thus, unauthorised subcontracting in fast fashion was driven as much by a coordination imperative, to deal with order (dis-)continuities, as by a profit imperative, to maintain levels of absolute surplus value as first tier manufacturers were required to adhere to lead firms’ quality and process standards.

Broadly, four segments could be identified in the workforce. The top segment was composed of managers and supervisors who were sometimes family members but certainly had the trust of the owner. They were formally employed on contracts and received firm-level benefits. At times, a small number of senior and skilled workers also formed part of this segment and would be paid at NMW rates.

A second segment was formed by naturalised, mostly female workers of (mostly) South Asian background who, on paper, were employed at NMW rates but earned around £3 per hour when their actual working hours were taken into account. This group constituted the upper tier of CMT workers and was forced to ‘maximise’ incomes from underpaid factory wages and welfare benefits, thereby exercising a modicum of agency at the same time as subsidising the employer. Workers in this segment often encountered debt and in-work poverty and were unlikely to find employment in other sectors as their language and socio-cultural capabilities were limited.

The third segment of the workforce was made up of migrants who had a right to work in the UK but only a limited right to remain and thereby no recourse to public funds (at the time of the research, this included workers on a student visa, a route that has since been closed). Thus, they were unable to make up their factory wages and tended to suffer worse terms and working conditions than the second segment.

The lowest segment was composed by workers who, under current regulations, were not allowed to work and/or stay in the UK. This constituted the most vulnerable group and consisted of under- or undocumented migrants as well as workers who over-stayed their (student or visitor) visa. Such workers were forced to work for even lower or no wages, work night shifts, and were dismissed at will. Segmentation played a crucial role in fulfilling orders in a flexible way, according to the ebbs and flows of the market, the varying product and process standards of lead firms, as well as changing pressures on profits. Respondents, for example, repeatedly emphasised how workers in the second segment who, officially, were on NMW rates, were substituted by workers from the third and fourth segments, invariably at lower rates, when the factory encountered difficulties.

However, workforce segmentation cannot be explained solely with reference to productive requirements (Peck, 1996; Rubery, 2007). While it is important to be aware of the social, economic and compliance function of workforce segmentation, it is equally crucial to see how this factory regime was driven by complex socio-economic processes on the supply side, providing a captive and partly under- or undocumented workforce. The segmentation of (in-)formal firms and workforces was given further space through inappropriate and inefficient forms of enforcement, of tax, as well as work and employment regulations. The segmentation of firms and workforces proved central in two respects: it enabled the reproduction of the formal chain-level capital circuit with its economic and social standards, and it enabled the reproduction of a largely informal firm-level capital circuit, one that was only viable at the price of a fundamental disconnection between the labour process and minimum work and employment regulations.

Conclusions

This article demonstrated how fast fashion accentuated the economic and operational demands in the supply chain and intensified competition amongst manufacturers to secure orders. Many firms could only absorb these pressures by adopting informal employment practices and/or by subcontracting to other firms that operated in the informal economy. For the core workforce, this resulted in a form of permanent on-call informal employment at a poverty wage rate of £3 per hour, although workers in the second, ‘core’, segment could be replaced by more vulnerable, and cheaper, segments if profits were threatened. The article highlighted how competition and segmentation at different levels – in the functional division of labour, in product markets, as well as the workforce – are at the centre of translating the market power of lead firms into the sphere of production. Regarding the article’s main contribution, the interlinkages between the labour process and value chains, we emphasise two issues pertaining to the role of labour in value chains.

First, linking labour process and value chain analyses through circuits of capital established a two-way interlinkage, treating labour as a central value producer rather than confining it to an organised collective actor or rule taker of market and regulatory dynamics (Bair and Werner, 2015). Labour might well take a lot of rules, yet as the argument about disconnections has shown, it also establishes the terrain on which those rules prove viable (or not) and to what extent specific labour processes can support production for fast fashion. A distinction needs to be made here between value distribution (labour’s share from the capital-labour bargain) and value creation (labour’s structural position in the production process) (Thompson et al., 2015). While the labour process might be disconnected in the sense that lead firms cannot keep their side of the ‘ethical fashion’ bargain, this by no means implies that the labour process does not have any impact on the accumulation dynamics across different tiers of the value chain. The UK fast fashion chain certainly seems very dysfunctional, cutting the majority of the workforce out of lead firms’ ‘codes of ethics’ and disconnecting labour (and a good number of the manufacturers) by shifting it into the informal economy.

Second, irrespective of this central role of labour, the analysis of competition and segmentation across a series of capital circuits has demonstrated the central role of domains other than the labour process. In particular, regarding the functional division of labour, the argument emphasised how the re-segmentation of firms and the workforce was as much a driver as an outcome of the emerging form of fast fashion governance. The same was argued with regard to the sphere of reproduction and the way welfare benefits were appropriated to establish an income-effort bargain (see Dunaway, 2014; Pickles and Smith, 2011; Thompson, 2013, respectively, on households, capital markets and the state). Thus, the phenomenon of fast fashion, its place in changing sourcing portfolios, needs to be understood not only as a form of governance that coordinates disaggregated production processes. Rather, it constitutes an articulation of capital circuits that are based on specific political settlements across different domains and thereby reconfigure the actors in the value chain (Bair and Werner, 2015). This analysis underlines the political character of GVCs, not only of prices and wages, but also of the rules and institutions that underpin the various bargains across value chains.