‘Are they out to get us?’ Power and the ‘recognition’ of the subject through a ‘lean’ work regime

Critical studies of ‘lean’ work regimes have tended to focus on the factory shop floor or public and healthcare sectors, despite its recent revival and wider deployment in neoliberal service economies. This paper investigates the politics of the workplace in a United Kingdom automotive dealership group subject to an intervention inspired by lean methods. We develop Foucauldian studies of governmentality by addressing lean as a technology of power deployed to act on the conduct of workers, examining how they debunk, distance themselves from and enact its imperatives. Our findings support critiques of lean work regimes that raise concerns about work intensification and poor worker health. Discourses of professional autonomy allow workers to distance themselves from lean prescriptions, yet they are reaffirmed in their actions. More significantly, we illustrate the exercise of a more encompassing form of power, showing how lean harnesses the inherently exploitable desire for recognition among hitherto marginalised workers, and its role as a form of ‘human capital’. The paper contributes to critical studies of lean by illustrating its subtle, deleterious and persistent effects within the analytical frame of neoliberal governmentality. We also demonstrate how studies of governmentality can be advanced through the analysis of contested social relations on the ground, highlighting the ethico-political potential of Foucauldian work.


Introduction
The automotive industry has historically been at the forefront of efficiency drives to enhance productivity and capital accumulation, from the mass production assembly line of Ford (Wilson & McKinlay, 2010), to the 'lean production' regime of the Toyota production system (Ezzamel, Willmott, & Worthington, 2001;Krafcik, 1988). Lean 'philosophies' and 'tools' are designed to match production and service provision with the market, to eliminate 'waste', costly stockpiles and shortages, thus creating a symmetry between demand and supply for the reduction of costs and the enhancement of profit. Critical analysis of lean regimes, both in the automotive industry and through its diffusion into the civil service, health and social care, demonstrates concerns about work intensification, deskilling and poor worker health (Carter et al., 2011(Carter et al., , 2013(Carter et al., , 2014Charlesworth, Baines, & Cunningham, 2015;Stewart, 2013;Stewart, Mrozowicki, Danford, & Murphy, 2016;Stewart et al., 2009). Calls have thus been made for a critical analysis and understanding of the persistence of lean and its socio-cultural implications at work today (Rees & Gauld, 2017;Stewart et al., 2016).
Critical scrutiny of lean in the automotive industry has focused on the production line and the factory shop floor (Ezzamel et al., 2001;Zanoni, 2011). Servicing and aftersales, as in our case, remains a neglected area of study. Nevertheless, it is intimately bound up in the networks of power produced by major transnational automotive manufacturers at the heart of fossil-fuel driven economies (Fleming & Spicer, 2007;Rhodes, 2016), and controlled via quasi-judicial franchise contracts granted to dealership groups (Arruñada, Garicano, & Vázquez, 2001). This paper investigates an organisational intervention inspired by lean methods in a major United Kingdom dealership over a two-year period. We ask, how does lean, as governmental technology of power, act upon the conduct and subjectivity of workers in the 'witches' brew' of everyday practice? More specifically, we ask: How are workers 'made up' through lean and its associated rationalities, and how might they resist or distance themselves from it? What kind of constraining and 'enabling' effects does lean have for workers operating in the context of aftersales and servicing? How do workers resist or consume lean amid the politics of organisational transformation?
Our analysis illustrates how workers live through neoliberal discourses that sustain the legitimacy of lean, yet they are not fully determined by them. Critiques of surveillance and workflow standardisation call upon a discourse of professional autonomy that emphasises the tacit selfknowledge of workers beyond lean prescriptions. Yet, despite these dis-identifications, lean discourses are nonetheless reaffirmed in the actions of workers (Fleming & Spicer, 2003). Moreover, we illustrate how among increasingly demanding organisational circumstances lean discourses create common modes of perception (Rose, 1999) that serve to harness the inherently exploitable desire for recognition and respect among hitherto marginalised workers. This, coupled with the appeal of lean as an individualised and reproducible form of 'human capital' (Weiskopf & Munro, 2012), demonstrates the encompassing effects of lean as a technology of 'government' power within and beyond the workplace (Fleming, 2014;Foucault, 2008). This paper makes three original contributions to Foucauldian studies of work and organisation and lean work regimes. First, we demonstrate how studies of governmentality may be advanced through an examination of contested social relations in the everyday (McKinlay, Carter, & Pezet, 2012). Moving beyond the 'programmer's perspective', we argue, allows for the workplace to be conceptualised as a site of political struggle between the ethical demands of others, and what we may wish for in actualising our freedom (Foucault, 1996). Second, we identify the harnessing of the inherently exploitable desire for 'recognition' and 'respect' as a technique of subjectivisation among previously marginalised workers. Third, the paper contributes to critical studies of lean (Stewart et al., 2016) by extending critique to account for its more subtle and deleterious power effects within the analytical frame of neoliberal governmentality.
The paper is organised as follows. First, we outline the field of Foucauldian organisation studies and the importance of the perspective of governmentality. Second, we discuss lean in public and automotive sectors, detailing its persistence as a technology of 'government' power. We then discuss the research methods and data analysis undertaken. The paper then turns to the organisational intervention in question before analysing its reception among workers 'on the ground'. We begin, however, by discussing Foucauldian studies of work and organisation and their importance for this research.

An Argument for the Perspective of Governmentality
Foucault's writings shifted over time from the archaeology of knowledge and discourse to the genealogy of knowledge and power. Yet, there is neither a pre-nor post-archaeology or genealogy in Foucault, but rather clear changes in emphasis (Dreyfus & Rabinow, 1982, pp. 104-8). Foucault's work began to encompass not solely autonomous discourses as regulative 'discursive formations' (Foucault, 1972), but rather the hazardous realities and relations of power which frame 'discursive regimes' (Foucault, 1980). Foucault's genealogical work adopted a more general interpretive analytics of what forms, restricts and institutionalises discursive regimes according to specific historical power/knowledge configurations. Our thoughts and actions, the games of truth or rationalities which we play out, have a history and are a product of particular struggles and contingencies (Brown, 1998). Foucault argued that relations of power significantly curtail the degree to which the human subject can fashion their own existence. Nevertheless, his later work on the art of governing, ethics and care of the self signalled a transformative agenda and a concern for what autonomy could look like for our present (Barratt, 2008;Foucault, 1996).
Foucault's oeuvre has been influential in the field of organisation studies for over three decades. Four interweaving waves of influence, drawing on (i) discipline and disciplinary power, (ii) discourse, (iii) governmentality (iv) and subjectivity and care of the self, demonstrate a wide-ranging 'Foucault effect' (Raffnsoe, Mennicken, & Miller, 2019). Inspiration from Foucault's 'middle' genealogical period, drawing from Discipline and Punish and The History of Sexuality, volume one (Foucault, 1977(Foucault, , 1978, brought a welcome yet partial initial reading to the field, arguably creating a misrepresentation of his contribution (Barratt, 2008;McKinlay & Taylor, 1998;Raffnsoe et al., 2019). A focus on disciplinary power within prisons, schools and factories (Foucault, 1977) legitimised commonsense notions of concrete organisations with clear-cut boundaries (Knights, 2002). Studies of the workplace, including that of lean regimes (Barker, 1993;Sewell, 1998), emphasised omniscient surveillance through electronic, human resource and peer review systems. The constitutive nature of power became encapsulated in 'self-discipline' and in the delegation of responsibilities for performance to teams through both 'vertical' and 'horizontal' surveillance (Sewell, 1998;Sewell & Wilkinson, 1992a). Nevertheless, by depicting obedient and normalised subjects, there was little or no illustration of agency, subversion or contestation in these studies (Newton, 1998). As already 'disciplined' and docile subjects (see Barker, 1993;Sewell, 1998), workers were depicted as complicit in their own subjugation, overlooking the manner in which power and knowledge may play out (McKinlay & Taylor, 1998;Raffnsoe et al., 2019). Power became synonymous with repression, coercion and limitation, erroneously creating a dichotomous interplay between passive recipients and opponents of repressive structures. Foucault's approach to power, then, is best understood not as a theory, but as a tautology, implying that the dynamic of power/knowledge and resistance is a matter for empirical investigation.
Readings of the 'later' Foucault have opened up new possibilities for Foucauldian scholarship in the field (Barratt, 2008;Fleming, 2014;Munro, 2012). The interrelated concepts of biopower (Foucault, 2008) and governmentality (Foucault, 1982), although peripheral in organisation studies, offer critical viewpoints upon work in neoliberal societies through which individuals are not only targets of power but active in its operation (Fleming, 2014;Munro, 2012). Ambitions and desires are mobilised and hastened rather than coercively shaped through rigorous procedures alone. The instrumentalisation of a population's intuition, sociality and desire, what Fleming (2014) elucidates as 'biocracy', means that life itself, in activities, work, joys and miseries, can become politically and managerially useful (Foucault, 2008).
Nevertheless, it is perhaps the concept of governmentality that has been adopted most enthusiastically from Foucault's experimental toolkit (Barratt, 2008;McKinlay & Pezet, 2017). In contrast to discipline, 'government' is predicated on the premise that the governed will continually disrupt, adjust, resist and distance themselves from the practices of governing (Foucault, 1982). 'Government', recalling sixteenth-century connotations, forms a practical and strategic relation between the governors and the governed, the 'modes of action, more or less considered or calculated, which were destined to act upon the possibilities of action of other people. . .to structure the possible field of action of others' (Foucault, 1982, p. 790). 'Governmentality' is a link between techniques of domination and techniques of the self, implying forms of agency as power/knowledge relations structure possibilities not only in ostensibly repressive ways, but also in seemingly seductive and appealing ways. In neoliberalism delimited 'freedom' is the architect of control, as subjects are required to recognise themselves, readily or not, as entrepreneurs of themselves (Foucault, 2008). Through notions such as learning, competency, employability and career, personal choices are delimited according to a logic of self-fulfilment and through the accumulation of 'human capital' (Weiskopf & Munro, 2012). Individuals and collectives are then 'offered' to participate in action to resolve matters previously in the hands of their superiors. This can be understood as a kind of 'responsibilisation', corresponding with ways in which 'the governed are encouraged, freely and rationally, to conduct themselves' (Burchell, 1996, p. 29).
Inspired by Foucault's genealogical methods, the 'London governmentalists' (Miller & Rose, 2008;Rose, 1996Rose, , 1999 have charted modern power in 'advanced liberalism' as the effect of diverse calculative techniques, forms of expertise and professional vocabularies designed to reconfigure identities at work. Power is not exercised by the powerful but instead through 'grey sciences' (Rose, 1996, p. 54) of efficiency and productivity, those that govern by creating calculable spaces in which workers calculate for themselves, and begin to know themselves accordingly, 'to seek to maximise productivity for a given income, to cut out waste, to restructure activities that were not cost effective' (Rose, 1999, p. 153). Managerial expertise plays the role of relay between the aspirations of corporate authorities and the ambitions of individuals and groups. Forms of 'translation' in codified knowledge and practice, such as lean, produce loosely affiliated networks and attempt to construct common modes of perception. When 'translation' is achieved between the values of others into one's own terms, judgements and conduct, then rule is established 'at a distance' (Rose & Miller, 2010).
Nevertheless, there is more to governmentality than an individual's assimilation into managerial techniques and corporate networks. The discursive and 'technical' means of influence on which 'governmentality' depends necessarily must align with an individual's or group's delimited 'freedoms' for its legitimation and neutrality (Foucault, 1982). McKinlay and Pezet (2017) suggest that the 'London governmentalists' have under-theorised resistance as always inherent in power relations, instead relying on the 'programmer's perspective' (see Rose, 1999) as a textual approach to historical writing (McKinlay et al., 2012). Moreover, the governmental rationality of 'enterprise' (du Gay, 1996) has been portrayed deterministically, often without reference to individual or collective agents (Fournier & Grey, 1999). Calls have thus been made for empirical research that explores how discourses are received, modified and resisted in everyday practice (Fleming, 2014;McKinlay et al., 2012). As McKinlay and Pezet argue, 'every governmentalist technology is depicted as always producing a full-blown neoliberal subjectivity, irrespective of how promising the ground is' (McKinlay & Pezet, 2017, p. 18).
Governmentality studies of work have typically 'borrowed' from the genealogies of (neo)liberal governmentality (Foucault, 2008;Miller & Rose, 2008;Rose, 1999) and applied them to the politics of the workplace. They contrast with the discipline power/knowledge couple by placing an emphasis on the mundane and ostensibly 'liberating' aspects of modern power. Studies, for example, of teamworking (Knights & McCabe, 2003) and project management (Clegg, Pitsis, Rura-Polley, & Marosszeky, 2002) show how liberal technologies render subjects ever more calculable according to economic criteria. By querying the idea of power as repressive, insight is gained into the production of new identities, forms of agency and practices of resistance amid complex organisational circumstances.
Nevertheless, for us 'governmentality' is less a coherent theoretical perspective on 'what works' organisationally (Clegg et al., 2002), and more a comparatively open-ended approach to the politics of situated research (McKinlay et al., 2012). Equally for us it is not that 'governmentality matters' (Clegg et al., 2002), but rather that governmentalities matter (Barratt, 2008). That is to say, the perspective of governmentality does not depend on formal texts and official programmes of neoliberal rationalities alone, but also on heterogeneous social relations and governmentalities that take shape 'on the ground'. We argue that the relationship between individuals and regulative governmental technologies such as 'lean' are best examined by taking account of 'what matters to them'. This permits a non-deterministic analysis of power relations beyond the received dichotomies of power and freedom, compliance and resistance (Raffnsoe, Gudmand-Hoyer, & Thaning, 2016). By acknowledging both the productive and limiting aspects of modern power, we are sensitive to the ethico-political ambitions of Foucauldian work and the possibilities of fashioning alternative ways of being (Barratt, 2008;Munro, 2014;Weiskopf & Willmott, 2013). With this in mind, below we outline 'lean' and its associated techniques as a governmental technology of power, before turning to the case in hand.

Governing Through Lean
Lean production, commonly known simply as 'lean', has been progressively implemented in workplace organisation in industrialised economies over the last three decades. Its 'philosophy' and techniques originated in the global automotive industry, and as some argue (Bhasin, 2015;Womack, 1990), more specifically the Toyota production system (Ohno, 1988). Following Japan's apparent immunity to the late 1970s financial crisis, principles and techniques under the banner of 'lean' were imported into the US and latterly British manufacturing and service sectors (Stewart, 2013). Although lean emerged in the highly regulated economy of post-Second World War Japan, it maintains significant appeal in neoliberal capitalist economies in which market rationalities advocate consumer authority and 'efficiency' to maintain profit, compete, or survive in recessionary conditions (Stewart et al., 2016(Stewart et al., , 2009. Lean entails the adoption of so-called 'neo-Taylorist' (Crowley, Tope, Chamberlain, & Hodson, 2010) techniques for fragmenting tasks, standardising operating procedures, performance monitoring, and eradicating any work that will not produce a profit or 'customer value'. Advocates profess that it removes 'obstacles' in the flow of production and quality through 'continuous improvement' (kaizen), and through teamworking arrangements stated to 'build up a system that will allow the workers to display their full capabilities by themselves' (Sugimori, Kusunoki, & Uchikawa, 1977, p. 554). Prescriptive texts claim that lean eradicates wasted time (muda) and overburden (muri) through 'just-in-time' (JIT) delivery and inventory systems (kanban) (Bhasin, 2015). As a form of workplace control, it is deployed to streamline work processes and heighten productivity while lowering costs, often involving a reduction in the labour force (Carter et al., 2013;Smith, 2000). Advocates argue that lean is participatory and democratic by 'offering' workers the means to be more involved in response to market pressures (see Bhasin, 2015;Womack, 1990). Workers, it is claimed, have enhanced 'freedom' to control their work through decision-making authority and skills enhancement. The imperative of 'working smarter not harder' proposes that lean is not only less wasteful of resources, but supposedly less burdensome (muri) by involving employees as 'agents of change'.
Nevertheless, research in both manufacturing and service sectors demonstrates that lean involves more intensive and centralised control, deskilling, work intensification, low staff morale, poor worker health, and a strengthening of the functions of capital through performance targets, surveillance and teamworking from above (Carter et al., 2014(Carter et al., , 2017Ezzamel et al., 2001;Stewart et al., 2016;Stewart et al., 2009). Delbridge (1995) plainly noted that the adoption of JIT and total quality management (TQM) in automotive manufacturing implied that workers were being coerced into 'surviving rather than resisting their exploitation' (Delbridge, 1995: 814). Lean regimes in the automotive industry have involved the speeding up of production and service provision with fewer workers, thus increasing the risk of work overload and poor worker health (Graham, 1995;Stewart et al., 2009Stewart et al., , 2016. Claims of a more democratic style of management following the decline of 'Fordism' (Crowley et al., 2010), whereby responsibility is devolved to workers in the efficient design and execution of work (see Sugimori et al., 1977), are discredited in empirical accounts where, for example, kaizen represents the imperative to 'produce more or risk the sack' (Ezzamel et al., 2001(Ezzamel et al., , p. 1072. The claim that lean 'empowers' workers through teamwork are invalidated in the analysis of systems such as JIT and TQM (Sewell & Wilkinson, 1992b), which instil higher degrees of surveillance upon the workforce. Lean, then, is representative of a growth in bureaucratic and technical control (Carter et al., 2014) and a reduction in the discretion of middle managers and supervisors through expanding hierarchies (Stewart, 2013).
The last two decades have witnessed efforts to transfer lean from manufacturing to service and administrative sectors, and particularly to the British civil service, health and social care sectors amid the politics of 'austerity' (Carter et al., 2011(Carter et al., , 2013Charlesworth et al., 2015;Radnor & Osborne, 2013;Rees & Gauld, 2017). The 'new public management' regime, with its doctrine of reduced public funding, commercialism, value-for-money and management-by-measurement control (Lapsley, 2008), aligns neatly with the 'better for less' principles of lean (Radnor & Osborne, 2013). Yet, in public service and clerical work, employees have eschewed claims of 'empowerment' by pointing to a reduction in decision-making, increased workload expectations and a narrowing of tasks (Carter et al., 2011(Carter et al., , 2013. In social care settings lean interventions have entailed overworking, low pay and stress related to overstretched budgets and resources (Charlesworth et al., 2015). Teamworking is shown not to be the harnessing of employee 'voice', multiskilling or job rotation, but rather a top-down scheme that places more demanding responsibilities and targets upon middle managers and non-managers, diffusing stress throughout the workforce (Carter et al., 2017;Procter & Radnor, 2014).
In healthcare, the deployment of lean has prioritised 'tool-based' applications as stand-alone interventions, such as value stream mapping exercises for identifying and eliminating 'wasteful' activity, and rapid improvement events where staff are required to meet, evaluate and streamline processes (Radnor, Holweg, & Waring, 2012, p. 370). 'Tool-based' approaches are thought to undermine the 'true' value of lean thinking as a broader 'model of cultural change' (Radnor & Osborne, 2013, p. 283), where kaizen denotes a broader programme of professionalisation (Radnor et al., 2012). Such perspectives demonstrate an academic 'cottage industry' (McCann, Hassard, Granter, & Hyde, 2015: 1559 on the subject of lean adoption in healthcare, where problems are not correlated with lean itself, but rather its incorrect implementation. While lean may improve processes such as patient turnaround time, information duplication and lengths of stay (Radnor et al., 2012), hopes of resolving overdue workplace frustrations are met with heavy workloads, insufficient resources and the view that lean is ill-suited to complex healthcare settings, where the importance of professional judgement outweighs the importance of process improvement (McCann et al., 2015). Here, a familiar critique is advanced, outlining lean as managerial rhetoric designed to obscure the reality of 'little more than an extension of Taylorism' (McCann et al., 2015(McCann et al., , p. 1560, and where 'continuous improvement' is ascribed to almost any kind of commonsense organisational improvement as an example of leaning. Nevertheless, while the efficacy of lean interventions in themselves is dubious, its 'philosophy' and techniques persist and have a long and embedded history, particularly in the automotive industry (Stewart, 2013;Stewart et al., 2009Stewart et al., , 2016. The 'continuous rationalization' of work (Stewart et al., 2016) involves not just the control and coordination of labour, but also the inculcation of discourses related to competency and career. The persistence of lean, as Stewart (2013) note, is in response to the continuing crisis of twentieth-and twentyfirst-century capitalism and recurring attempts to implicate workers in the 'strategic struggle' of production. It is therefore a mistake to write off lean as a collection of ill-placed techniques and managerial fantasies, given that its rationalities of efficiency, cost reduction and productivity lie at the heart of efforts to responsibilise workers for the risks and costs of modern capitalism (Stewart et al., 2016).
As a technology of power, it is precisely lean's apparent superficiality and mundanity that warrants further investigation of its effects (Foucault, 1982). Its extension as a 'grey science' (Rose, 1996, p. 54) of efficiency into automotive aftersales and servicing involves the neoliberal imperative that workers should make corporate ambitions their own (Kiff, 2000). Crucially, lean's regime of truth produces effects by encouraging individuals and groups to know themselves differently, not simply through prescriptive interventions, but also in relation to professional expertise, competency and career, extending the reach of capital into life itself (Fleming, 2014), and through the circulation of 'human capital' (Weiskopf & Munro, 2012). Lean production, then, is much more than a series of prescriptive tools and professional know-hows with rules to enforce. As a governmental technology it relies on harnessing the self-governing capacities of workers, willingly or not, for its effect. It therefore constrains subjectivities while producing others among the politics of everyday working life.

The Study and Research Method
This paper draws from a larger study conducted at an independent family-owned business, which we label VehicleCo, employing 1500 people. It is one of the largest private companies in the motor retailing business in the UK, with a chain of 34 franchised dealerships. The study focused, first, on how the company organised and governed its franchised dealerships, management and workers. Second, on how systems and processes influenced by major car manufacturers (under the banner of 'lean') were used to govern VehicleCo, its management and workers. Under the authority of major automotive manufacturers, margins in distribution, aftersales and servicing are small and downturns can have major financial implications for dealerships.
The study investigated a knowledge transfer programme (KTP) set up to implement transformational change. The objective was to improve 'efficiency', which, senior management hoped, would increase the 'capacity' of the company's resources and increase output. The KTP would instil kaizen through process improvements. It also proposed a comprehensive programme to develop the skills and knowledge of workers in line with lean thinking. The training of employees, it was argued, would allow for new skillsets that would aid recruitment practices and align performance criteria with those delimited by major car manufacturers.
A KTP associate was active in the intervention and worked across a variety of dealership sites evaluating existing processes using lean 'tools'. For example, the technique of value stream mapping (Womack & Jones, 2003) was deployed to eradicate task duplication, standardise operating procedures and reduce 'waste' in terms of labour time. The aim was to produce streamlined, convenient (and thus more profitable) customer appointments and services. The KTP associate was regularly interviewed by the researchers on the intervention's reception. We use a pseudonym, 'Work Savvy', to refer to the intervention under investigation.
The empirical research comprised interviews, observations, weekly meetings with VehicleCo's Strategic Director, attendance at management meetings, and analyses of company audit and job specification documents. Data collection occurred between January 2014 and January 2016. Audio-recorded interviews were conducted across four different company franchises. In two franchises, selected for longitudinal access, interviews were conducted in three stages: before, during and after the intervention. Data were transcribed from 74 interviews, five strategic management meetings, weekly team meetings and observations of work practices. Interviews were conducted using semi-structured questions accompanied by follow-up questions to explore in more depth the experiences of Work Savvy. As the researchers progressed, questions were refined to probe more deeply into emerging discourses. The first round of interviews typically lasted one hour, and in stages 2 and 3 between 30 and 45 minutes. For quotations, we use pseudonyms and refer to the participant's job title and stage of the research process.

Data analysis
Data analysis followed a Foucauldian approach to discourse analysis, remaining sensitive to particular 'regimes of truth' (Foucault, 2008) in official company texts, patterns of talk, assumptions and rationalities in the discursive strategies of participants (Alvesson & Karreman, 2011). The methodological commitments of Foucauldian discourse analysis propose the study of relations of power in the historical constitution of individuals and groups as subjects (Foucault, 1980(Foucault, , 1982. The constitution of subjectivity is dependent on norms which are facilitated through structures of recognition. Yet these norms are not deterministic and emerge and fade depending on the operation of power in specific contexts. Our analysis adopted an iterative approach between empirical material and theory to investigate how participants related to their work, others and the self (Alvesson & Karreman, 2011). We address workplace 'culture' as that which arises from particular power/ knowledge configurations and problematisations. Such problematisations were evident not only in official statements, but also in the self-governing discourses of managers and workers 'on the ground' (McKinlay et al., 2012). By paying attention to discursive rules, as 'what counts as what', we remained sensitive to linguistic practices as they emerged. Selves are situated in discourse, and what becomes pertinent is not the interpretation but the discursive rule as to which it serves (Potter & Wetherall, 1987).
The two-year study enabled the research team to document changes to processes and working practices, and the responses of workers. Two researchers independently coded data using NVivo software. Each researcher reviewed interviews at stage 1 to identify the discursive strategies that employees drew upon. Four interviews were independently coded by the two researchers to develop participant-derived, first-order codes (Miles, Huberman, & Saldana, 2014). The researchers then compared the similarities and differences between coding to agree consistency (Silverman, 2012). Interviews were then independently coded, and meetings were held to discuss themes. The same procedure was followed for stages 2 and 3. Thereafter, the two researchers developed an iterative dialogue between data and theory to identify relationships between first-order codes and group them into second-order codes. The latter were refined and grouped into more robust categories, and theoretical themes were established. The final phase of coding identified discourses that highlighted distinct relations of power in the context of the study (Foucault, 1991). This analytical process resulted in the identification of three primary categories of discourse outlined in the 'Analysis' section below. First, however, the paper turns to a summary of the 'Work Savvy' intervention.

'Work Savvy' -The governmental intervention
VehicleCo's business strategy was 'to be a "world class" retailer delivering the best customer experience in a fun and expert way' (Work Savvy proposal document, August 2013). The aftersales division, dealing in maintenance and repairs, was problematised by senior management as failing to generate income. Aftersales, it was argued, could provide a greater source of profit that would insulate the company against seasonal variations. The strategic director identified 'lean' methods as a favourable approach. By assessing and standardising work in terms of its predictability, 'wasted' time would be eradicated and job turnaround time would be lessened, allowing for the sale of maintenance and repairs to more customers and thereby increasing profits. The expertise to deliver this change was thought not to exist within the company itself, and previous attempts to use consultants had failed (notes from meeting, July 2013).
The intention was to undertake transformational change throughout the whole dealership network. Lean processes were to become 'embedded' (project proposal document, August 2013) among the workforce. The project was branded as 'Work Savvy', officially commencing in January 2014. VehicleCo's senior management sanctioned a review to assess current knowledge of lean knowledge and practice, and 'readiness' for change, in order to develop and test Work Savvy. Obtaining 'buy-in' from employees was considered necessary before inculcating lean thinking. After the review, the company sought to redesign the flow work and remove 'wasted' activity, thereby heightening customer turnaround and satisfaction. The aim was to 'involve a wide spread of staff across the project' with a view to 'standardising work and improving efficiency' (Work Savvy briefing document, 2013). Time and motion studies of technicians, including process maps (or 'spaghetti maps'), were conducted, videotaped and reviewed by management and Work Savvy teams. The aim was to identify and eliminate 'waste' and produce standard operating procedures for predictable work. The Work Savvy programme aligned with audits from car assemblers which stipulated times and processes. Teams included those who were 'championing' Work Savvy, typically made up of volunteers. Senior management monitored the progress of changes through the 'continual improvement' plan, do, check, act (PDCA) cycle (see Deming, 1950).
Senior management agreed (project proposal document, August 2013) that the first year of Work Savvy would be spent on 'groundwork' for transformational change. Weekly meetings were held at head office where the KTP associate's activities and outcomes were communicated. In May 2014, the strategic director 'pushed' for the use of a 'pilot site' as a way to 'promote' Work Savvy and demonstrate its benefits across the company (notes from weekly meeting, May 2014). In June 2014, an induction event took place at the pilot site introducing Work Savvy for aftersales. In late 2014, two more pilot sites were identified by the strategic director (notes from weekly meeting, December 2014). Weekly meetings focused on 'efficiency' in the aftersales function, and how effective attempts to achieve employee 'buy in' had been. Additionally, 'Masterclass' training sessions (see Pullin, 1998) were introduced focusing on the structuring of processes for technicians and customer service staff. Furthermore, a project team made up of employees engaged with external lean consultants who put on training sessions inclusive of 'exemplary' lean practices from different industries. In May 2015, drawing on observed 'best practice', the intervention incorporated the rationalisation of work into standard operating procedures, grouped as follows: (i) predictable work (the green lane), (ii) may cause unanticipated issues (the amber lane) and (iii) could become complex (the red lane). In the section below, we discuss three prominent categories of discourse that emerged in our analysis.

Staff shortages and overworking
A consistent discourse running throughout the interviews we conducted at multiple dealership sites centred on staff shortages. Both managers and workers bemoaned that staff shortages not only were producing stress through overworking, but that such conditions were significantly curtailing time available to contribute to Work Savvy. Jamie, a dealership manager, for example, commented that 'resource has been tough, particularly since the 2008 crash'. The industry had become 'scared of having too much resource, and so we find where we might think two administrators would be okay, instead of working one really hard, we're not, we're working one very hard' (Jamie, business centre manager, stage 1, 28/06/2015). Working during days off and during holiday time was considered commonplace, with one account manager, Steve, stating that: If you look downstairs at the moment, there's a chap who's on holiday. He's in sorting cars out. . . You quite often come in on your day off to sort stuff out. . .it is quite stressful, because you have to do everything. (Steve, account manager, stage 1, 22/6/15) As noted elsewhere (Stewart et al., 2016), workplaces subject to lean interventions are characterised by a form of responsibilisation through which workers are made accountable, directly or indirectly, for labour utility problems. Claims of 'empowerment' and participation (Bhasin, 2015;Womack, 1990) are therefore misleading given that, when workers assume more decision-making responsibility, they do so in combination with a heightened responsibility for the effectiveness of their labour and that of their colleagues. As the above accounts show, this can give rise to overworking and work-related stress (Carter et al., 2014;Stewart et al., 2016).
Participants also emphasised long working hours, which for some, were indicative of a 'sink or swim' working environment, denoting the sense that 'all you're doing is standing there with a hosepipe' (Simon, service advisor, stage 2, 22/12/2015). As Simon, a service advisor, elaborated: Well, again, to get support, you need people, and there aren't any people. Within this context, Work Savvy was frequently understood to be exacerbating, rather than alleviating, stressful working conditions. As Paul, another service advisor, commented, 'Well, to be honest with you, at first, I think that [Work Savvy] was quite frowned upon. . . It is a lot of stress and strains when you take people out of a dealership for any period of time' (Paul, service advisor, stage 2, 29/12/2015). The issue was that 'it's difficult to put something like this in place when there's simply not enough staff. . . to do something smartly, you need time, and nobody has any time at the minute' (Simon, service advisor, stage 2, 22/12/2015). One's obligation to become involved in Work Savvy was, for some, undermining the more immediate responsibilities of fulfilling day-to-day tasks. Working 'smartly', then, seemed beyond reach for some of these workers. Such accounts support claims that lean interventions, by attempting to involve employees in the 'strategic struggle' of production, deepen experiences of work intensification and work-related stress (Carter et al., 2013;Stewart, 2013).

Surveillance, professional autonomy and distancing
Against the background of intensive working conditions participants regularly commented on the nature of surveillance and its effects. Eric, a dealership manager, worried about workforce morale, outlined his concerns in regard to new time and motion studies and the monitoring and accounting of performance to reduce 'wasted' activity: Some of these guys in the aftersales might have been there twenty years and then [for] somebody to say, 'Right we're monitoring on this, this, this and this, it's just changing, you have to.' This account taking is not well received, and they [workers] view it like a negative. (Eric, general manager, stage 1, 24/6/2015) For those subject to monitoring, discursive strategies regularly called upon a discourse of professional autonomy to counter surveillance and the rationalisation of tasks. As Brad, an experienced technician, commented when discussing the initial stages of Work Savvy: It's just. . . I think at first when people were in watching you, I think that's a bit nerve-racking for everybody. . . And everyone was a little bit under pressure. There's nothing worse than being watched by someone, like how you're doing something. In Brad's statement his professional experience ('it's something you've done for years') is destabilised in the process of being observed by others in the early stages of Work Savvy ('there's nothing worse than being watched by someone. . . especially when it's something you've done for years'). Brad, then, is struggling with practices designed to assess and standardise his performance, those which are thought to detract from his autonomy as an established and experienced professional. His comments are not directed at a particular group of professionals or colleagues, but instead the techniques of power by which his performance is to be measured, inscribed and transformed (Rose, 1996). Such accounts are indicative of a lack of professional agency in shaping the manner in which standard operating procedures are to take shape, and where 'teamworking' equates to 'claustrophobic monitoring' (Carter et al., 2017, p. 463).
Elucidations on the theme of professional autonomy were commonplace among technicians, recurring during discussions about Work Savvy. As Marcus, another experienced technician, remarked: Back in the day. . . you know, if you messed up there's no one breathing down your neck. Your manager would take you in the office, explain to you, you'd get shouted at and then you'd come out. But at the moment. . . everyone's got that feeling now of like, 'Am I being watched? Is someone watching us on the CCTV? Is someone checking all my paperwork? Are they out to get us?' (Marcus, technician, stage 2, 4/1/2016) In this account Marcus describes the power effects of a 'vertical' system of surveillance analogous to Sewell and Wilkinson's (1992b) analysis of the JIT/TQM labour process. Nevertheless, rather than a 'superstructure of control' to which docile bodies do not respond other than to submit (Foucault, 1977;Newton, 1998), Marcus is critiquing the effects of disciplinary power, illustrating how the governed continually distance themselves from the practices of governing (Foucault, 1982). The ethical demands of others, in this case, do not align with what Marcus wishes for in actualising his own professional freedom (Weiskopf & Willmott, 2013). Marcus is responding as an inventive and thinking worker with capacities of his own (Barratt, 2003(Barratt, , 2008McKinlay et al., 2012), and calls upon historical ideals of clear hierarchical relationships ('Your manager would take you in the office, explain to you, you'd get shouted at and then you'd come out') to counter the moral efficacy of workplace surveillance deployed to increase his productivity.
By standardising operational procedures, Work Savvy was, for some, discouraging established professional practices that had been developed over years. Dave, a technician, stated that 'changing the way you've worked for years [is the biggest challenge]. . . If you're in a routine, it's hard to alter that routine without overlooking something. . . Everybody does every job differently' (Dave, stage 2, 29/7/16). Within the discourse of professional autonomy discursive strategies were not only set in opposition to the standardisation of procedures, but also emphasised concerns about deskilling. In these incidences, the effects of workflow standardisation were thought to place limitations on the possibilities for both job satisfaction and professional development. Tom, a technician, stated that: One of the lads had had enough of the green lane and asked to be moved, because it was pretty much servicing all day, every day. So it can get quite mundane. You're repeating stuff over and over and over again. (Tom, technician 2, stage 3, 20/12/2015) For Dave, the monotony of tasks also posed a threat to his ability to learn a sufficiently broad range of skills, given that 'you get people set in a routine because they're doing the same job all the time', and therefore, '[you're] not getting the experience of our trade' (Dave, technician, stage 2, 29/7/16). One's potential standing in the labour market, and therefore one's 'human capital' (Weiskopf & Munro, 2012), were considered to be under threat for these workers. Such accounts contest claims that teamworking through lean provides employees with control over their work, and instead emphasises vulnerability, ambivalence and a lack of agency over the rationalisation of tasks. Indeed, as another technician, Luke, aptly summed up, 'I don't have any control, you get what you're given' (Luke, technician, stage 1, 27/06/2015).
Aside from the notable discourse of professional autonomy, participants also found ways in which to express cynicism in order to distance themselves (Fleming & Spicer, 2003) from prescribed subject positions delineated through Work Savvy. Discursive strategies questioned aspects of the new governmental regime which participants felt detracted from, or ignored, who they were as experienced and knowledgeable workers. In these incidences, participants articulated ambivalence concerning their own role as a vehicle for power through lean discourses (Foucault, 1982). As Dave, a technician, argued: I don't know, they seem to think they want you 100% efficient and they think you've got nothing better to do than fill out these silly little bloody 'be Work Savvy' things. But it's like being back at school. We don't want that. As long as we do the job and we do it correctly and on time. . . they [senior management] think you've nothing better [to do], they think you've got loads of time on your hands. (Dave, technician, stage 1, 27/06/2015) The obligation to take part in practices intended to shape appropriate conduct detracts from Dave's and his colleagues' understanding of organisational matters, and of themselves. 'Investing' in Work Savvy is not considered a means to achieve meaningful engagement with work. Such a perspective not only calls upon a discourse of professional autonomy, but also serves as a critique of the 'conduct of conduct' (Dean, 1999), illustrating that governmentality is determined only insofar as what is wanted from workers matches what they want for themselves. The requirement to inscribe a 'version' of oneself in Work Savvy training materials detracts from being able to recognise oneself as a dependable worker ('As long as we do the job and we do it correctly and on time'). 'Resistance' thus manifests as a kind of 'irresponsibility', expressed in a reluctance to fully recognise oneself as being involved in the 'strategic struggle' of Work Savvy. Being responsible for one's performance is therefore not synonymous with the rationalities and principles of lean (McCann et al., 2015). Rather, lean is understood to undermine the day-to-day operation of running a productive dealership, reducing the time needed to perform well.

'Recognition' and the production of new identities through Work Savvy
In this section we examine the more 'productive' aspects of governmental power exercised through Work Savvy. Here, we ask what kind of subjectivities and experiences it served to 'produce' in this context. We now have a better understanding of the discursive strategies deployed by participants emphasising a 'distancing' (Fleming & Spicer, 2003) from Work Savvy, often construed as a time-consuming encumbrance that overlooked who they were as experienced and autonomous professionals. Nevertheless, in conditions in which the value of one's labour was at stake, Work Savvy was at times both seductive and seemingly necessary for managers and workers. The following section illustrates that the negotiation of subject positions among the artificial 'freedoms' of this governmental intervention are both complex and contradictory; power both produces and constrains (Foucault, 1982). John, for example, discussed the personal benefits of partaking in Work Savvy: I'm respected a lot more by the management. . .I seem to get approached a lot, so I think it has been positive, being part of the Work Savvy team. And that's one of the reasons I did stick by it. . . it could lead to better things. (John,technician,stage 2,4/1/16) In John's comment subjectivity is a process of becoming as he recognises himself in light of the power relations in which he is involved. Being a part of the Work Savvy team is directly related to a heightened sense of self-respect. The appeal of Work Savvy as an ostensibly 'liberating' technology of governmental power is emphasised, as John addressed it as a means to enhance his individual 'freedom' in relation to his career and the wider labour market ('it could lead to better things'). Rather than discussing the benefits for the workforce, John is turning in on himself as an individual in recognition of his own 'human capital' (Weiskopf & Munro, 2012). Problematising one's identity as a worthy contributor in the eyes of one-s superiors gives rise to a responsibility to oneself, and not to one's colleagues. This would also suggest that under conditions of stress and overworking, a preoccupation with identity concerns may support a more concrete sense of one's own significance (Knights & McCabe, 2003 ). Yet, at the same time Work Savvy provides 'liberating' potential by highlighting Mark's and his colleagues' worth to others. As Mark commented further, 'Work Savvy seems to be pushing the case of people going, 'Right, we'll stop and listen to the technician because he knows how to do his job and he's been doing this a lot longer than most people' (Mark, technician, stage 2, 4/1/16). Rather than addressing his new-found visibility in terms of surveillance and discipline, the Work Savvy intervention appears as a way to demonstrate a particular set of concerns about the value of one's contribution. At the same time, however, this form of 'empowerment', through the production of new 'lean' identities, aligns these workers with the governmental schemes that define them; as those who must produce more 'output' as responsibilised managerial subjects. Recognising the self as a 'lean' subject, then, produces a more concrete sense of self as one's ambitions become more intimately and subtly aligned with corporate objectives (Rose, 1999). As another technician commented: By encouraging contributions from workers, the governmental rationality of lean, in this case, hails a new form of managerial agency for these technicians. By actively disclosing one's abilities and capabilities to superiors, one becomes more amenable to intervention and evaluation, while simultaneously assuming more responsibility for productive activities. These workers are thus assuming ownership of responsibilities that were previously in the hands of their superiors. Yet at the same time, as a technology of agency and performance (Dean, 1999), lean serves to animate these workers to act on themselves as they become more 'aware' of their new identities within a network of recognition. Discipline and surveillance do not adequately explain these power effects, insofar as work intensification is intimately tied up with a heightened sense of self-control and self-government. Surveillance is thus 'designed in to flows of everyday existence' (Rose, 1999, p. 234) and it does not take shape as exhaustive regulation (see Sewell & Wilkinson, 1992b). Instead, a devolved governmental framework is deployed to frame the ways in which choices and decisions are to be made in the interests of 'efficiency' and in relation to one's involvement in the 'strategic struggle' of production. Productivity and efficiency become personal matters, so much so that these workers actively construct the means by which productivity is to be increased.

Discussion and Conclusion
Our findings make a novel contribution to critical studies of lean regimes (Carter et al., 2013(Carter et al., , 2017Stewart et al., 2016) by illustrating its deleterious and subtle power effects within the analytical frame of governmentality studies. Overworking, stress and the detrimental effects upon worker health were clearly evident within what was dubbed a 'sink or swim' working environment. In addition to a strain on labour, articulated most expressively in relation to staff shortages and a lack of available time, participants also challenged claims of working 'leaner' and 'smarter' by drawing upon a discourse of professional autonomy to emphasise their tacit knowledge and experience. This, in combination with a critique of enhanced surveillance and workflow standardisation, put to question the moral efficacy of this lean intervention, given that for some it detracted from, or ignored, who they were as experienced and knowledgeable professionals. Concerns about deskilling were articulated in relation to standardised operating procedures that were thought not to provide a sufficient range of tasks or skills necessary to maintain a trade or produce job satisfaction. Such findings add a new viewpoint upon critical studies of lean, where overstretched resources produce stress (Charlesworth et al., 2015), professional self-knowledge is seen to be antagonistic to narrowly defined lean process improvements (McCann et al., 2015) and teamworking is often indicative of a lack of professional agency over how work will be organised and executed (Carter et al., 2017). Expressions of ambiguity in our analysis related to struggles over which fields of judgement were in play, be that professional or managerial, illustrating that this governmental intervention did not cultivate its subjects exhaustively (McKinlay et al., 2012). For some, 'investing' in Work Savvy was not considered to be a meaningful way in which to engage with work, demonstrating that lean must align with an individual's or group's delimited 'freedoms' for its legitimation and neutrality (McKinlay & Pezet, 2017). Yet, responses to lean principles of employee involvement and 'respect' (Bhasin, 2015;Sugimori et al., 1977) did not always constitute explicit acceptance or rejection. Rather, employees dis-identified with managerial prescriptions and subject positions and developed a kind of professional 'irresponsibility' towards the 'strategic struggle' of production. Although lean discourses were shunned through these practices of 'dis-identification', they were nonetheless reaffirmed, often begrudgingly, in the actions of these workers (Fleming & Spicer, 2003). The possibility for and indeed the very notion of practical refusal was noticeably absent in the discourse of these participants.
Without downplaying the undoubtedly harmful effects of lean, we suggest that its persistence in work and organisation can be better understood through a more encompassing analysis of its effects. The perspective of governmentality demonstrates how in increasingly demanding circumstances (Stewart et al., 2016), and despite the discontent of many workers, the boundaries between work and life can become less distinct (Fleming, 2014;Munro, 2012). Subtler power effects were evident in the harnessing of the inherently exploitable desire for respect and recognition among workers. Within this frame, acknowledgement of one's labour was facilitated through a common network of perception, where lean principles of efficiency and productivity began to translate into one's own terms, judgements and conduct (Rose & Miller, 2010). For some, then, the effect of the Work Savvy intervention produced 'self-respect', where the issue of work intensification was diluted through networks of mutual recognition (Knights & McCabe, 2003). Lean was surprisingly well received among technicians after it ostensibly enabled them to 'think for themselves' (Knights & McCabe, 2003, p. 1613 as 'worthy' decision-making subjects. This process of identity affirmation gave rise to enthusiasm, not explicitly for lean itself, but instead as a way out of historical identity-related problems as estranged workers. The role of 'recognition' and 'respect' through lean as a form of 'translation' (Rose, 1999), then, points to a subtle technique of subjectivisation in this context.
Moreover, lean was addressed as a form of professional competency that could potentially assist one's prospects in the eyes of superiors and in the wider labour market, possibly leading to 'better things'. The reframing of labour as individualised 'human capital' illustrates an important effect of lean within the broader frame of neoliberal governmentality (Weiskopf & Munro, 2012). Not only were workers targets of power as disenfranchised professionals, they were also active in its operation through neoliberal modes of self-government (McKinlay & Pezet, 2017). Implicitly, such examples show how economic security and risk are at play in the constitution of the 'lean subject', where an individual cost-benefit analysis of gains and losses can permeate into a broader life project (Fleming, 2014) amid increasingly competitive and precarious circumstances (Stewart et al., 2016).
The perspective of governmentality shifts beyond 'discipline' not only by querying how lean places restraint on the freedom of skilled workers (Carter et al., 2011;Sewell & Wilkinson, 1992a), but also by querying the conditions by which specific forms of agency are animated, and at what cost (McKinlay et al., 2012). Our analysis has sought to go beyond reductive dichotomies such as power and freedom, compliance and resistance (Raffnsoe et al., 2016), pointing to a more intimate form of power than the genealogy of discipline can explain. Power did not explicitly dominate subjects, but instead presupposed the autonomy of agents in attempting to align them with centralised corporate objectives. A contribution of this paper has thus been to show that the perspective of governmentality can be adapted to address strategic interventions on the one hand, and the divergent responses of 'real' actors on the other; those who are resourceful subjects with histories and capacities of their own (Barratt, 2003).
Lean, as a technology of governmental power, is one way for capital to endeavour to increase its hold on labour through a more encompassing and competitive framework of reasoning (Dean, 1999;du Gay, 1996;Rose, 1999). Nevertheless, as our analysis of an automotive dealership demonstrates, the virtuous claims of upskilling, job rotation and employee voice (Womack, 1990) are met with accounts of work intensification (Stewart, 2013), claustrophobic monitoring (Carter et al., 2017) and discourses of professional autonomy deployed to critique a lack of agency over the pace and standardisation of work. However, we suggest that the study of responses to prevailing economic rationalities in particular socio-historical contexts can provide a more nuanced perspective upon their effects and costs (McKinlay et al., 2012). In doing so we have highlighted some of the more subtle power effects of lean at work today, remaining alert to post-disciplinary forms of power in a neoliberal society (Fleming, 2014).
Although the incidences of identity affirmation we observed may be understood and felt individually as 'empowering', the costs for collective labour and the quality of work are unmistakably deleterious. Security, in this case, comes as a matter of aligning with representative truths, those that solidify individualised material and symbolic possibilities in increasingly competitive and precarious circumstances (Foucault, 2007). If what is understood as intrusive surveillance, for some, becomes identity-giving for others, then there is a pressing need for a more radical critique of the persistence and politics of lean interventions, and post-disciplinary work, in work and life today. One way in which this project could take shape is by exploring how discourses of professional autonomy may evolve into a broader transformative politics of the workplace. As Foucault (1996, p. 448) noted, 'the concept of governmentality makes it possible to bring out the freedom of the subject and its relationship to others, which constitutes the very stuff of ethics'. Governmentality on the ground, as we have observed, implies that social relations are a site of struggle between the ethical demands of others, and who we might wish to become in actualising our freedom (Weiskopf & Willmott, 2013). Following Foucault, we suggest that this freedom is fundamentally political. To that end, it must involve a care of the self that goes beyond prescriptive neoliberal imperatives, and where 'being free means not being a slave to oneself' (Foucault, 1996, p. 437).
Chris Hicks is Professor of Operations Management at Newcastle University Business School in the UK. Chris has a strong interest in lean manufacturing and has participated in European Regions for Innovative Manufacturing, an EU-funded project that has helped transfer lean expertise into small companies throughout the North Sea region of Europe. He also undertook a project that evaluated transformational change in the National Health Service in the northeast of England.
Tracy Scurry is a Senior Lecturer in human resource management at Newcastle University Business School in the UK. Her research focuses on careers from the perspectives of the individual and the organisation. She has worked with a number of organisations, in the public and private sectors, evaluating the processes and impact of organisational change. Work to date has adopted a multi-stakeholder perspective, examining both the organisational and individual implications of the practices for all those involved. Current research interests include graduate careers, underemployment, career resilience, dual careers, mobility and organisational change.