Abstract
I am grateful to Critical Sociology for hosting this review symposium and to Benassi, Ikeler and Wood for their deep engagement with my book. Management Divided (hereafter, MD) presents a synthetic theory – which I call organisational political economy – integrating key concepts from organisation theory into a historical materialist foundation. MD is thick and dense, so it would be difficult to address all the arguments and evidence in a short book review. It is thus unsurprising that there is substantial variation in what the three reviewers have focused on.
Further, as Benassi perceptively notes: ‘Vidal’s book contains provocative and original ideas, which will have a large following but will also raise some scepticism among scholars close to a more “orthodox” Marxist tradition’. Indeed. I have been fortunate to receive many of reviews of MD. A pattern has emerged that Marxists from the Bravermanian school misinterpret or mischaracterise my basic argument regarding lean production while dismissing my evidence (Carter, 2023; Rosenfeld, 2022). Ikeler can now be added to this list. These scholars have an unshakable commitment to Braverman’s (1974) thesis that all managers everywhere are concerned with controlling and disempowering labour, and therefore lean production can only be a more refined system for managerial control and deskilling.
Bravermanian theory is neo-Marxist insofar as it rejects the relative autonomy of the forces of production, which means rejecting what Marx saw as a primary source of dynamism within and across human societies: the contradictory relation between the forces and relations of production. For Braverman (1974), capitalist production relations fundamentally shape the trajectory of technical change: technology from machines to management systems is allegedly designed in service of the capitalist interest in controlling and disempowering labour. Despite being neo-Marxist in this sense – hence distinct from Kautskyite orthodox Marxism (a technologically determinist theory of the inevitability of socialism) – Bravermanian theory has indeed become an orthodoxy of the radical wing of the American labour movement, specifically the Labor Notes network (Parker and Slaughter, 1988; Parker, 2017), and of the British Socialist Workers’ Party (Carter, 2023).
I work within the tradition of classical Marxism and MD is an attempt to flesh out a classical Marxist alternative to Bravermanian neo-Marxism (and to Kautskyite orthodox Marxism), based on an interpretation of Marx as arguing that the forces of production have a relative autonomy, meaning that technical change develops relatively independently from the concerns of capitalists. Elsewhere (Vidal, 2022a) I provide extended defence of this interpretation with extensive textual support. There, I write:
Marx distinguished technical development in general from specifically capitalist motivations for developing technology and specifically capitalist uses for technology. Against specifically capitalist motivations, Marx (1965 [1867]) counterposes science, suggesting it is a relatively independent domain that is exploited by capital: ‘Science, generally speaking, costs the capitalist nothing, a fact that by no means hinders him from exploiting it. The science of others is as much annexed by capital as the labour of others’. (p. 333) Critically, ‘we must distinguish between the increased productivity which is due to the development of the social process of production, and that which is due to the exploitation by the capitalists of that development’ (Marx, 1990 [1867]: 547).
Being a committed Bravermanian (Ikeler, 2015; Ikeler and Limonic, 2018), Ikeler attempts to refute my finding that some US factory managers adopt an upskilling and empowerment approach. Unfortunately, Ikeler fundamentally misunderstands the basic theory articulated in MD, attributing to me several arguments which are gross distortions of what I argue, while ignoring the robust evidence I present in support of my arguments. Fortunately, Benassi and Wood clearly and accurately summarise (different parts of) my argument while both accept my evidence on upskilling and empowerment, which suggests the source of Ikeler’s many misunderstandings is not unclear or ambiguous writing on my part.
Benassi summarises my theory of the labour process and of lean production:
Vidal argues that in post-Fordism lean is the best practice of work organisation, and that lean with substantive workers empowerment represents the most efficient way of organizing work. Yet, in the implementation of lean systems the management experience the contradiction between coordination and discipline, that distinguishes all management systems, in the form of empowerment vs deskilling (p. 80), and they often satisfice rather than maximizing efficiency. Managers settle for ‘limited forms of empowerment that produce satisfactory increases in efficiency’ (p. 85) based on their aspiration levels and cognitive orientations. They might satisfice because the former logic of Taylorism still resonates with them; or because they are happy to settle for good enough between deskilling and empowerment instead of going all the way to substantive empowerment; or they might satisfice in the routine politics of production, when they engage with workers who defend their existing routines and resist change. (p. 95)
Wood complements this with a summary of my theory of routine politics of production:
Faced with competing demands and workers’ resistance most managers with moderate aspiration levels aim for ‘good enough’ which invariably means the securing of sufficient output rather than trying to achieve best practice, that is, empowering workers – a far more challenging proposition. One reason that achieving empowerment is challenging is that workers tend to defend existing routines and ways of doing things. Vidal, in contrast to Burawoy, terms this workplace conflict the ‘routine politics of production’ and convincingly argues that it develops, not so much out of the ‘effort-bargain’ but from contrasting views among workers and managers regarding the best means to achieve efficiency and quality as well contrasting views on the proper division of labour between workers and managers.
Wood also clearly summarises my conceptualisation of growth stages as settled organisational fields:
Organisational fields are . . . conceived as becoming settled around particular notions of best practice (but also retaining influence from past dominant logics). This framework enables Vidal to articulate a Marxist theory of growth stages in which stages are constituted by unique formations of production, competition, and state institutional settlements and the emergence of a web of settled organisational fields . . . Providing a conceptual framework that enables the identification of Fordism and post-Fordism as distinct stages of capitalism is just one of the major contributions of the book.
While Benassi and Wood broadly accept my findings, Ikeler concludes that
Empowerment . . . is assumed rather proven. Does it consist in better information about customer wishes and plant processes? Perhaps cross-training, rotation, and direction to produce smaller batches of output? These are what Vidal’s informants offer. Yet they offer unconvincing evidence of rising leverage or skill, much less durable upgrading and a refutation of Braverman’s thesis . . . We are thus left with little evidence in support of the book’s central claim that lean empowers workers.
This statement manages to both misunderstand my basic argument and to ignore the evidence I present. I begin by setting the record straight on my argument and evidence regarding upskilling and empowerment. I then turn in sequence to several issues raised by Benassi and Wood: politics of production, control and consent, power resources, the generalisability of my findings, and the possibility of balancing contradictory pressures. The final section turns to again to correct gross misinterpretations by Ikeler on my analysis of deskilling – which, as I clearly argue in MD, remains the dominant tendency across a broad range of occupations – and my suggestions for worker/union strategy regarding lean production.
Upskilling and Empowerment
It is puzzling how Ikeler concludes that one of ‘book’s larger-order conclusions’ is ‘that lean methods “substantively empower” workers’. He attributes to me the argument that: ‘lean . . . demands reintegration of manual and mental labour’. But I explicitly criticise this position along with the opposite view that deskilling and intensification are inherent to lean. Immediately after criticising both positions, I write: ‘my research demonstrates that lean is a management model which can be implemented in distinct ways, with distinct emphases, depending on the orientation of management and the power and orientation of labour’ (pp. 7–8).
This argument is reflected in the title of the book:
Capitalist management is divided between those who pursue the more dynamically efficient, but also more difficult, strategy of substantive empowerment and those who settle for good enough by focusing on standardization, limited training, and consultative worker participation (or no employee involvement). (p. 39)
One of the basic findings of MD is that all of the factories I studied had implemented lean yet the major majority (12 out of 22) did not adopt an upskilling/empowerment strategy. I note that my ‘sample is likely biased toward [high-involvement] factories insofar as factories taking a low-road approach to labour management may be less likely to agree to allow a researcher inside their organization to examine their operations’ (p. 319). My expectation is that under capitalist management, upskilling/empowerment will remain relatively rare and my theory of management decision making in MD seeks to explain why this is the case.
The theory proposes that two-processes drive decision making: a top–down process in which institutional logics shape how individuals understand and define their situation, and a bottom–up, feedback-driven process in which individuals evaluate existing performance for satisfactoriness according to their own aspiration levels. As Chapter 4 demonstrates, in the US manufacturing field, the dominant institutional logics of production include lean production and worker empowerment. Yet, the formerly dominant logics of Fordism and Taylorism continue to exert influence on many managers.
An upskilling/empowerment strategy will be adopted only where a manager has (1) embraced the logic of empowerment and (2) has high aspiration levels (or where a powerful union forces managers to adopt the strategy). I wrote:
To the extent that managers are facing more intense contradictions, and . . . that best practice [a complementary package of lean practices, including substantive worker empowerment] is difficult and requires high aspiration levels, deviation from best practice has likely become more common in the postfordist era. (p. 18)
In my analysis, a traditional Taylorist deskilling approach is a deviation from best practice in post-Fordist manufacturing.
My findings suggest that many workers also continue to be influenced by the logics of Fordism and Taylorism. In attempt to make their work meaningful and to ensure the success of their company, some workers think in terms of the Fordist notion of efficiency, which leads them to resist or be reticent towards lean practices they see as inefficient. Benassi and Wood both note this argument without criticising it.
By contrast, Ikeler completely misunderstands the point of my analysis, writing that I interpret ‘worker resistance to lean as various forms of irrationality’. But the point of institutional logics theory is that each logic is a form of rationality. Neither the term ‘irrational’ nor any such idea appears in MD. The point of my analysis is that in the factories I studied, worker resistance was primarily driven by workers adopting the rationality of Fordism and Taylorism, according to which scale economies dictate maximising output on individual machines or tasks – with narrow job classifications enshrined for decades in collective bargaining contracts – and the ‘proper’ division of labour between managers and workers entails the former doing the brain work and the latter doing the physical labour. Some workers thus see lean practices as inefficient and resist them because they think it will make their employer uncompetitive.
Examples include workers expressing frustration with smaller baches and more changeovers (which are elements of just-in-time production focused on economies of scope and flexibility) – ‘I’m putting this same die in every other week. It doesn’t make sense’ (p. 229); skilled machinists resisting cross-training on multiple machines, which contrast with some union workers taking the view that cross-training ‘makes us more overall machinists rather than machine operators . . . training everybody that they can go on a situation and perform the job’ (p. 210); seeing teamwork as inefficient – ‘they’ve got four people over there doing what I used to do by myself, you know. And it’s like, I just don’t see how that’s benefiting anybody or anything’ (p. 224); and workers not wanting to be given tasks, such as participation in problem solving and decision making, which they see as the responsibility of management, in the words of two different workers: ‘they have too many committees’, ‘too many little committees set up to fix a problem’ (pp. 211, 225).
On the specific question of skill, Ikeler adopts a classic definition based on a romanticised notion of pre-industrial craft work: task complexity and autonomy. For Ikeler, skill consists of ‘independent mastery, both conceptual and executive, over a process’ (p. 2). So what is the problem here?
Conceptually, Ikeler makes obligatory reference to the social construction of skill but follows this by asserting a singular, objectivist definition of skill: independent mastery of a process, measured in terms of task complexity and autonomy. He cites Attewell (1990) but fails to heed the latter’s warning that skill is ‘a complex and ambiguous idea’ (pp. 422, 443, 445). Attewell concludes that, if the comparison is with pre-capitalist craft work, then ‘most all occupations within capitalism will appear unskilled, by definition’. Yet, this is precisely what Ikeler does.
Empirically, though he cites Attewell’s (1987) classic study of skill trends, Ikeler fails to acknowledge the study’s headline finding: ‘Empirical evidence was reviewed on the relative size of occupations over time, and on changes in economywide skill levels. Neither supported Braverman’s contention of pervasive deskilling within the twentieth-century U.S. economy’ (p. 341). More recently, Autor et al. (2006) found that most job growth in the 1980s was in high-skill occupations and job growth in the 1990s was characterised by polarised growth with ‘the most rapid increases in high-skill jobs’ (p. 191) (for a similar finding with different data, see Wright and Dwyer, 2003).
The three studies just cited all find a broad increase in the share of high-skill work in the United States. Among the fastest growing occupations in 2022, according to the Bureau of Labor Statistics (BLS), are wind turbine service technicians, nurse practitioners, data scientists, statisticians, epidemiologists, software developers and computer scientists (in addition to a host of low-wage jobs). 1 All of these occupations entail being managed within a capitalist labour process. Bravermanian theory cannot explain this extensive growth of high-skill occupations. It has nothing to offer here. Bravermanians argue that managers and engineers prioritise disempowering labour over all other concerns, even efficiency (Gordon, 1976; Noble, 1986).
By contrast – and against the received wisdom – Marx’s theory provides an explanation for upskilling. Here I draw on the work of Adler (1990, 2007), who, as I explain in MD, offers an interpretation:
that alongside shorter-term processes of degradation and immiseration, Marx theorized a longer-term tendency toward a more progressive form of socialization. This notion of socialization (Vergesellschaftung) refers to the transformation of people and means of production from individual/private into collective/social. Marx’s use of the concept of socialization, what I refer to as productive socialization, is expansive. Building on Adler, I read Marx as theorizing a process of upgrading the capabilities of humans and technology over time based on the progressive accumulation of scientific and technical knowledge. (p. 78)
Much of this productive socialisation happens via popular education, which for Marx (1990 [1867]: 613, 617–618; 1981 [1894]: 414–415) was a fundamental force increasing the technical sophistication – hence managerial capabilities – of workers. Chapter 3 of MD has extensive textual reference to Marx. Here I provide a single quote:
Modern industry never views or treats the existing form of a production process as the definitive one. Its technical basis is therefore revolutionary . . . [I]t is continually transforming not only the technical basis of production, but also in the functions of the worker and the social combinations of the labour process. Marx (1990 [1867]: 617, quoted in MD pp. 78–79)
On skill, Adler (2007: 1319) argues,
viewed theoretically (and indeed politically), autonomy is, I submit, the wrong yardstick. It is backward-looking, reflecting nostalgic regret for the passing of the autonomous craftsman (or alternatively, reflecting the ideal of alienated, self-sufficient individualism that is the spontaneous ideology of market society). This yardstick may allow us to measure what often has been lost in the development of the capitalism; but to formulate an assessment, we surely also need a way to understand what has replaced that lost autonomy . . . Autonomy is merely the converse of interdependence, and if we want to understand the changing nature of work relations and of skill more generally, we need a way to understand the changing forms of interdependence.
Adler concludes, correctly, that: ‘interdependence can take either coercive or collaborative forms, and we therefore need a theory of skill that can help us understand these forms and the relations between them’. Lean production is a case-in-point: there is a decrease in individual autonomy because there is an increase in interdependency. This neo-Taylorist system maintains task fragmentation, but managers with high aspiration levels and the right orientation will enlarge the task bundle and enrich it with responsibilities for quality control, problem solving and continuous improvement, which entail giving workers a better understanding of the overall process via both job rotation with cross-training and worker participation in value-stream mapping.
I found evidence of substantial cross-training and substantive worker empowerment in 7 of the 22 factories; the latter generally took the form of workers participating in problem decision making in teams and in particular in kaizen (i.e. continuous improvement) events. According to one manager of a union job shop with around 90 shopfloor workers making hydraulic presses:
We focus in from value-stream mappings on where we need to target to get the biggest bang for the buck for the kaizen activities . . . I’m in the process of challenging each cell leader [an hourly wage position] to perform X amount of kaizen events in their cell, but that’s just really getting rolling at this point. We’re still in the process of educating people on the different lean tools and how to implement them.
His goal was to ‘move the people to where the work is, have them as highly trained and able to run different machines, perform different operations, as possible’. He was training them in ‘how to do brainstorming, problem-solving analysis, root-cause analysis, those sorts of things, team building’, with the goal of having the workforce drive continuous improvement, including ‘having team meetings that just involve the people out there on the factory floor’.
Another case is a non-union, precision casting operation with around 500 shopfloor workers. According to the vice president: ‘The president and I get up in front of all the employees on a team-by-team basis, and we say, ‘Here’s where we fell short last quarter. Here’s what needs to be improved. Go back and work on it’’. The personnel manager described how workers are allowed to run continuous improvement projects on their own initiative:
You can have many going on at one time, and I wouldn’t even be aware of them. Some of the employees in Module 1 at one point in time had a real interest in the grind area and knew how they could streamline it and make it better. They talked to the plant manager. ‘Fine, go ahead. Start working on it’. They’d come in and work on things here and there.
They’d just come in and do it on their own?
Mm-hmm. And then they came up with some programs to really help the process, implement it, reduced times drastically, you know, in setups and things like that to standardize stuff.
What did the workers say? According to one machinist at the union job shop making hydraulic presses: ‘The overall input, as far as how the machines go together in the particular cell to produce the product, that was basically our input’. Another machinist from the same factory said, in contrast to the old days: ‘we’ve got teams now. We’ve got cell lead [hourly] people that are involved, that has just taken off. We’ve got more committees than we had before. We have union involvement, more so than we had before’.
A third worker at this factory explained how they are being cross-trained, increasing their skillset:
It makes us more overall machinists rather than machine operators . . . [in] the old apprenticeship system where you learned in a work circle, and you learned how to be a machinist on a drill press, on a lathe, and a grinder and other stuff. I think we’re moving back toward that now where we’re training everybody that they can go on a situation and perform the job.
Another worker at this factory had a similar view, noting that being broadly cross-trained gives them greater flexibility:
Say you’re down one man in that particular cell. Knowing how to do the job, one man conceivably – or two people – could do the job of three, or just picking up the slack of one . . . But having the ability to perform all of those jobs – at least most of the customers are going to get their product.
I did not interview workers in the precision casting operation, so I present quotes from another non-union shop, which had around 40 workers making high-end pumps for the food and beverage industry. An assembler explained how he has more input under lean production: ‘We’ll get together and brainstorm, you know, ‘What do you feel about this, or what’s your idea on it?’ He said that workers are allowed to experiment with changes, ‘which is nice, they’ll give us the option to be able to do that, you know’.
A senior assembler described being empowered through participation in team-based continuous improvement activities and how, being more responsible for quality and aware of the entire value-stream, he is able to approach workers in other areas:
We were using old methods of putting bearings on shafts . . . So that saved us a lot of time . . . through the mapping of the processes, and through the lean manufacturing that we had done. We found out where our time – we did a lot of time studies. We did some videotaping of some of the activities to find out where the wastes were . . . I feel like I’ve got a better handle on my job, and I have a little bit more control of saying . . .
Mm-hmm, so you do feel like you have more control?
Yeah, I do . . . Through the value stream, you’re finding out where the processes start out, you know, in doing the maps and finding out what is an outside vendor issue and what isn’t. I’ve been able to approach . . . purchasing and ask them about when something’s going to come in and not feel like, “Oh, I’m stepping out of my bounds.
He concluded that under lean, every day he feels like: ‘I’ve gotten a lot more accomplished than I did before, and a lot less stressed out as far as feeling like I went to work and I actually accomplished something’. A polisher in this factory confirmed that they were being upskilled via cross-training: ‘They’re starting to cross-train everybody, so everybody gets a break and gets to jump back, and other people get a feel of knowing more pumps’. A third assembler noted how his job had been enlarged:
When I first started, it was just building pumps. Now it’s went to pulling your own parts, refilling bins, when we went to kanban. I mean, we’ve eliminated other jobs that people have done for us, but we’re building just as fast. So that’s a plus that you’ve seen to the lean.
This is just a sampling of the extensive evidence presented over five empirical chapters in MD (plus a sixth empirical chapter on the manufacturing field).
Politics of Production
My concept of routine politics of production highlights situations in which:
workplace politics largely revolve around managerial attempts to change routines and workers’ defense of their existing routines, based on competing visions of efficiency, concerns for product quality, and understandings of the proper division of labor between management and workers. (p. 89)
More precisely, routine politics obtain where managers attempt to upgrade routines and workers dispute these changes based on competing visions of efficiency. Benassi suggests – correctly – that ‘Unions’ commitment to Fordist ideas and resistance to change is typical of Anglo-Saxon (manufacturing) unions, which have little control over changes in the work organisation due to their poorly institutionalized voice rights’, whereas in countries like Germany, where unions ‘strong voice rights and built on the tradition of humanisation of work’, the politics of production are different.
I fully agree. Routine politics are a particular outcome produced by the institutional context and the theory suggests variations in production politics may be generated by differences at the organisation level in (1) managerial strategy and/or (2) local workplace culture, both of which are shaped but not determined by differences in the broader organisational field. The latter incorporates both aspects of national variation that Benassi highlights: formal employment relations institutions and distinct union traditions and orientations.
Wood suggests I make too much of the difference between how I use the concept of production politics and how Burawoy used it. I wrote:
The term politics of production is associated with Michael Burawoy (1982, 1987), but my usage is more general and in some ways opposed to his. For Burawoy, production politics refer to the ideological effects of the workplace . . . I use the term production politics more broadly to refer to workplace conflict and culture within the workplace. (p. 35)
Wood is surely correct here: I tried too hard to differentiate my approach from Burawoy’s (for reasons I discuss in the following section). Wood is also correct that Burawoy’s argument that ‘conflict can be contained, dispersed and transformed by workplace institutions into lateral competition between workers’ is important and can complement routine politics of production.
Control and Consent
Wood is a leading sociologist working in the Burawoyian tradition. He has made important contributions to Marxist labour process theory, including, among others, the concept of ‘flexible discipline’, in which precarious scheduling gives managers extensive power over job quality and a despotic mechanism of control (Wood, 2020), and the concept of ‘schedule gifts’, which allow managers to gain ‘gratitude and moral obligation’ from workers in regimes flexible discipline (Wood, 2017).
I accept Wood’s argument that while Burawoy saw consent as inevitable in the Fordist firm, increasing marketisation has subsequently eroded key elements of the Fordist firm (particularly internal labour markets) and increased the incidence of despotic managerial regimes. I also concede that my ‘account of alienation is actually highly amenable to Burawoy’s view that workers invest in work games so as to make their work meaningful and avoid the tedium of capitalist work organisation’, a point which I acknowledged in the book (p. 73).
But I continue to find Burawoy’s concept of consent to be problematic. The problem here, as I write in MD is an underling rational choice analysis:
one vein of labour process theory postulates that because workers have ‘been forced to sell their labour power to another’ (Braverman, 1974: 39), they ‘have an interest in expending as little effort as possible’ (see also Edwards, 1979: 12; Lazonick, 1990: 14). In this rational choice analysis – a crude model with no textual basis in Marx – workers are a homogeneous category with uniform interests in minimizing effort due to the fact that they have alienated their labor to the authority of capitalist management (p. 202).
In MD, I placed Burawoy in the sociological camp as opposed to the foregoing rational choice analyses. But upon further reflection, it seems clear that Burawoy also adopts this rational choice framing, attributing to workers an interest in minimising effort. As I have noted (Vidal, 2022b), Burawoy asserts, without any textual evidence:
‘Marx’s analysis of the labour process in general’ rests ‘on the assumption that the expenditure of effort is decided by coercion’, and that Marx ‘had no place in his theory of the labour process for the organization of consent, for the necessity or willingness to cooperate in the translation of labour power into labour’. (Burawoy, 1982: 27, emphasis mine)
Burawoy provides no textual references to Marx discussing the question of effort. The only place I am aware of is a sentence in Chapter 13 of Capital Vol 1 – a chapter titled ‘Co-operation’, which begins Marx’s (1990 [1867]) historical discussion of the labour process – where Marx makes a point very similar to Burawoy’s theory of game playing at work: ‘mere social contact begets in most industries an emulation and a stimulation of the animal spirits that heighten the efficiency of each individual workman’ (pp. 443, 477). Chapter 14 is entitled ‘Division of Labour and Manufacture’. These chapters discuss authority extensively, and indeed contrast the ‘anarchy’ of the market with managerial ‘despotism’ in the workshop, but they do not discuss coercion and they do not imply workers will expend as little effort as possible.
Marx’s (1978 [1844]: 75–76) discussion of alienation, on the contrary, is premised on the idea that labour is the ‘life-activity’ of humans, that is, workers realise their species-being by creatively expressing themselves through work. In other words, workers have an independent interest in working hard: to realise purpose and meaning through work. Nowhere does Marx argue that alienation extinguishes this human drive to find meaning and realise purpose through work. Nowhere does he argue, as Braverman does, that workers ‘have an interest in expending as little effort as possible’.
MD documents workers being sceptical towards management but committed to their work, an outcome I have subsequently labelled alienated commitment (Vidal, 2022b). The concept of consent cannot explain this contradictory outcome, which is surely widely experienced within the workforce. The workers I observed were committed to being high-quality, efficient workers, something that was a rooted deep desire to have meaningful work, not something which emerged from game playing at work; yet many of them were highly critical of their managers.
On control, Wood notes that I prefer the term ‘discipline’. He invokes a classic distinction Edwards (1986) made between, in Wood’s words: ‘Detailed Control’ (i.e. control over details of work tasks) vs “General Control” (accommodation of workers to the overall aims of the enterprise)’. This distinction is useful. General control is what I mean by discipline. As discussed elsewhere (Vidal, 2022a), I dislike the term control because it is so often confused not simply with detailed control but with control as an end in itself. Braverman, Gordon (1976) and Noble (1986) all argue that managers prioritise control over efficiency. Yates (1999) explicitly asserts that control of labour is an end in itself: ‘the essence of capitalist management is control: control over the labor process and therefore control over the worker’.
I interviewed 47 managers for this project and not one appeared to exhibit any concern with having absolute control over the labour process or with disempowering labour, let alone being singularly obsessed with it. Nor did any of the 59 workers I interviewed complain of too much managerial control or of being deskilled. To be sure, ‘Employer anti-unionism is a central feature of the history of capitalism, and union-busting tactics are core tools in the capitalist struggle against labor’ (Vidal, 2022a: 36). But managers, especially middle managers like those running individual factories, must be concerned with quality control, flexibility, problem solving, continuous improvement, workflow, the scale and scope of production, production control and so on. Survival in the market entails attending to all of these concerns and a manager who single-mindedly prioritises absolute control over the labour process will drive the company out of business.
Wood mentions many strategies and practices managers can use to secure commitment and/or effort from their workers. I agree on the examples he gives (loyalty systems, responsible autonomy, etc.) but simply do not think it is accurate or useful to describe these mechanisms as securing consent, because the concept assumes that wage labourers have an interest in expending as little effort as possible, ignoring the interest humans have in making their work meaningful.
Power Resources
Benassi makes the important point that ‘The power resources available to labour also vary across institutional contexts’. She notes I acknowledge ‘that organized labour, adversarial tactics and militancy play an essential and indispensable role (pp. 38, 322) but such a role is not integrated in the theory when discussing the mechanisms leading to high-involvement lean regimes’ (pp. 153–155). She concludes: ‘I wonder therefore whether the US context is partly responsible for Vidal’s claim that managers with the highest aspiration level are primarily responsible for the adoption of lean with substantive empowerment’.
I agree fully that power resources are critical and also with her point that ‘that strong labour can serve as “beneficial constraint” on management because it prevents the compression of labour costs and forces management to invest in their workforce and in workplace innovation to improve productivity’ (Sorge and Streeck, 1988). She is correct that the institutional context of my study is critical here: in my study of the United States, managerial orientation and aspiration level were the most important variables explanation the variation I found. I do cite the comparative employment relations literature and note that:
Consistent with this [literature], my research found that where a local union is focused on increasing worker empowerment rather than resisting managerial restructuring altogether, union power is an important factor that can increase the likelihood, extent, depth, and stability of worker empowerment. But my findings also demonstrate the critical role of low or moderate managerial aspiration levels in limiting the diffusion of best practice, including the substantive empowerment of labour, in the face of the management and workforce contradictions. (p. 38)
I do concede that union power is underemphasised in MD, which is because I was attempting to highlight factors that were more important in my case and, indeed, are underemphasised in the sociology of work literature. Within debates on lean, as far as I am aware, managerial orientation has been emphasised only by a couple of major employment relations contributions (Appelbaum and Batt, 1994; Kochan et al., 1997) and here the point is mentioned but not fleshed out. Managerial orientation has not been a variable for labour process theory in general, and with regard to lean specifically, the existing labour process literature nearly universally assumes lean is designed to deskill and intensify labour (important exceptions include Delbridge, 2003; Elger and Smith, 2005). By contrast, my theory makes variations in managerial orientation a central variable and offers a theory to explain these variations. The satisficing aspect of my model is even more original; as far as I am aware this concept has not been deployed or developed within the sociology of work, employment relations or indeed even within institutional theory in organisational sociology (including institutional logics theory).
The Generalisability of the Theory
Benassi, a leading scholar in comparative employment relations (see, e.g. Benassi et al., 2016, 2022; Doellgast et al., 2016), writes that ‘a more detailed upfront discussion of the influence of the (institutional) context on the theoretical framework might have clarified how the latter could be applied to a broader range of cases beyond the US’. I do note in MD that ‘Although my case study is on contemporary American manufacturing, the framework, which I call organizational political economy, is explicitly developed for comparative and historical research on organizations across sectors and countries’ (p. viii). That said, I concede that there are only a few scattered comments about this in the book.
The theory conceptualises growth stages as settled organisational fields, which are constituted by dominant institutional logics across key domains: the production model, the form of competition and the form of state. In MD, I conceived the production model broadly to include the employment relations system and finance, but elsewhere I have conceptualised these as distinct domains and specified the Fordist and post-Fordist growth regimes in the United States, United Kingdom and Germany, including variations in power resources and union orientation (Vidal, 2019).
Wood also questions the generalisability of the argument, suggesting:
it could be argued instead that the tension between empowerment and control is specific to American lean production as in such work organisation attempts at normative control are not concretely underpinned by hegemonic institutions (indeed this is what I found in my study of US lean retail (Wood, 2020)) and thus, potentially, not a general tendency of late capitalism.
In my analysis, this tension is a deep one that is inherent to the manager–employee relation and does not depend on external institutional supports, though I agree with Wood that the latter can moderate the tension (see the following section). As I write in MD, the underlying contradiction is between:
ensuring worker discipline and harnessing worker creativity . . . This contradiction is manifest in conflicting pressures to use workers for routine manual versus abstract cognitive labour power; to provide minimal training versus substantial training (e.g. training in a single skill versus multiple skills or training in narrow skills versus broad skills); or to standardize work versus allowing discretion and autonomy. (pp. 4, 6)
The argument is that all managers everywhere face these contradictory pressures. Even traditional Taylorism, which tried to resolve the tension by turning workers into robots – all discipline, no creativity – has always been and remains problematic because workers are not robots and they resist being treated as such (Friedman, 1977). Purely Taylorist systems are only able to work because workers use their brains to make them work, engaging in quality control against the managerial emphasis on maximising output, which often undermines quality (Forrant, 2000). One form the contradiction between ensuring discipline and harnessing creativity takes is conflicting pressures between ensuring discipline and empowering workers. Another form it takes is between standardisation and flexibility (Korczynski, 2009) or between quantity and quality (Taylor and Bain, 1999).
Balancing the Tensions
Wood suggests that these tensions can be balanced and, indeed may
not necessarily exist in opposition nor necessarily independently of each other (a necessary precondition for contradictions). For example, increasing empowerment might lessen the need for discipline (direct control) as doing so increases the potential for effective normative control, legitimation, and responsible autonomy and thus achieves a higher level of general control.
This is a perceptive point and it overlaps with an argument Benassi (2023) has made in her own work on lean. I do write in MD that ‘It is possible – if a manager or management team is sufficiently competent and has sufficiently high aspiration levels – to forge a configuration of routines and practices that establishes a balance’ (p. 3). As noted, I understand discipline to mean general control, not detailed or direct control. That being said, my argument clearly implies – and my findings show – that high levels of empowerment may allow a strategy of responsible autonomy with little or no need for direct control. My argument is not that the tension is unresolvable but that managers across all sectors and institutional contexts experience conflicting pressures and that most will opt for discipline over empowerment due to a combination of intense competitive pressure, managerial satisficing and the continued influence of the logic of traditional Taylorism. Firms must establish minimum levels of efficiency, quality and flexibility to survive, though given that most manufacturing firms are middling performers (evidence reviewed on pp. 92, 116), ‘the small number of top performers cannot drive out the large number of middling performers’ (p. 315).
Finally, Wood questions my claim that, in his words, ‘Even if one accepts that the tension between empowerment and control constitutes a contradiction between capitalist forces and relations of production it is not evident prima facie that it constitutes ‘one of the most fundamental contradictions’’ (my emphasis). I argue that
One of the most fundamental contradictions of human society is between the forces and relations of production. Within the workplace, the contradiction between the forces and relations of production operates via conflicting pressures associated with the labour process and the valorization process . . . Managers coordinate the division of labour to increase efficiency but they also must discipline the workforce to ensure workers follow instructions and exert sufficient levels of effort. (pp. 75, 77, emphasis added)
Wood notes several additional contradictions of capitalism. I accept these and do not think it is particularly important to rank their relative importance. The reason I suggest the contradiction between the forces and relations is fundamental is because it is transhistorical, existing across modes of production and acting as a central dynamic shaping class relations within societies and institutional change across societies and modes of production. Indeed, the argument is that this contradiction drives skill trends in capitalism, as I now turn to discuss.
Deskilling, Upskilling and Union Strategy
Ikeler writes that one of ‘the book’s larger-order conclusions’ is ‘that deskilling is not a fundamental tendency of capitalist production’. This interpretation is puzzling, to say the least. I consistently and unambiguously argue that deskilling has been and remains a dominant tendency within the capitalist employment relation (emphasis added in all cases):
If mainstream management scholars tend to emphasize celebrity cases, sociologists and critical management scholars tend to focus on more typical cases, often finding deskilling and work intensification. (p. 4) Managers experience conflicting pressures for standardization versus discretion, deskilling versus upskilling, and routine manual versus abstract cognitive labor in a wide range of occupations. (p. 6) Braverman’s contribution was monumental, inaugurating an important new research program, which has indeed documented widespread deskilling and intensification under capitalist management, past and present. (p. 11) Marx’s analysis was prophetic insofar as he highlighted deskilling as a major tendency, forecasting the coming dominance of scientific management. However, Marx was emphasizing the dominant tendency across particular growth stages and not positing a singular, timeless imperative à la Braverman. (p. 19) In line with his theory of technical change as a contradictory process, Marx (1990 [1867]: 617–618, emphasis added) envisioned a progressive tendency conflicting with the regressive tendency of deskilling. (p. 75) To be sure, deskilling remains widespread in the twenty-first century. (p. 76) Postfordist capitalism has thus initiated an intensification of the contradiction between coordination and discipline, which takes the form of empowerment versus deskilling. (p. 80) The contradictions of the labor process, along with competitive pressures, encourage continued deskilling and disempowerment despite decades of advocacy of the high-involvement model of human resource management within and beyond business schools. (p. 322)
My argument is that deskilling remains the dominant tendency within the capitalist labour process but there is a weaker – though fundamentally important – tendency towards upskilling and empowerment of labour.
What does it mean that deskilling is a strong tendency and upskilling/empowerment is a weak one? As explained in MD:
Lean reflects and intensifies a material contradiction that is inherent to the capitalist employment relation: the need to ensure discipline versus the need to harness the creativity of workers. The core argument of this book is that in facing the conflicting pressures generated by this contradiction, managers often satisfice – settle for good enough – rather than maximize profit, efficiency, labour control, or anything else. (p. 9)
As argued in MD, deskilling was the only logic of management from the advent of the factory through the 1950s, when job enlargement and enrichment were introduced in the United States and autonomous teamwork in the United Kingdom. But deskilling remains a strong tendency today because most managers are satisficers and hybrid approaches, such as what I call lean enough – which achieve sufficient degrees of productivity, quality and flexibility by using lean process control tools and deploying standardisation, rotation among deskilled tasks and consultative employee involvement – are sufficient to survive and profit in the market.
Unfortunately, Ikeler’s fundamental misunderstandings of MD do not end there. He writes that I criticise the view that the Toyota Production System is ‘an enhanced or more reflective form of Taylorism’? In fact, I argue repeatedly throughout the book that lean is neo-Taylorist. Here is the definition I give:
lean production includes (1) demand-driven, flow production, (2) economies of flexibility and continuous improvement (with economies of scale), and (3) neotaylorism with worker empowerment, consisting of (3a) teamwork and cross training on fragmented tasks and (3b) employee involvement. (pp. 32–33, emphasis added)
Finally, Ikeler grossly misrepresents my argument when he asserts: ‘Vidal ultimately calls for workers and unions to jettison opposition and fully participate in lean because it represents “real movement[. . .] toward the empowerment of labor” (p. 321)’. This suggests that Ikeler did not read the Conclusion, which is the only place where I discuss my views on worker and union strategy, where my position is clear (emphasis added in all cases):
Militant opposition is an important, often-necessary, and appropriate tool for organized labor. (p. 320) If there are movements toward empowerment and democracy within capitalist firms, unions and workers can fight to increase and expand those, rather than resist them. Indeed, unions can push for more direct and institutionalized forms of power, such as the formal Labor-Management Partnership of companies like Second Tier Specialist, driven and facilitated by the International Association of Machinists (IAM) union. This explicitly does not entail foregoing militancy. (p. 321) Organized labor, adversarial tactics, and militancy play an essential and indispensable role, for worker dignity and the extension of democracy within the workplace. (p. 322) Union demands should go beyond what most lean advocates suggest: in the US context, this entails pushing and organizing for labor-management partnerships with institutionalized decision-making rights over business strategy and other high-level decisions . . . This requires a strong union ready and willing to engage in militant tactics. (p. 322)
My proposition is that the toolbox of unions should include militancy and conditional cooperation, the latter being pursued – in certain circumstances – to increase workplace democracy and build union power and union capacity, with the goal of worker control. I have elaborated on this position elsewhere (Vidal, 2022c: 150). The broader implication of MD is that workers are capable of running firms more efficiently than capitalist managers; workers and their unions should make the inefficiency of capitalist management a key point in their efforts to organise the working class and establish worker control.
