Abstract
Few texts have done more to shape how sociologists and social theorists think about workplace politics, labour agency, the everyday production of cooperation and even the sociology of work than Michael Burawoy’s Manufacturing Consent. As a result, discussing it poses a two-fold challenge – of scope and of depth. The first requires a profound understanding of the theoretical foundation of industrial sociology on which Burawoy built. The second, in-depth expertise of the organisation-level issues Burawoy puts forward. Paul Thompson, Emeritus Professor of Employment Studies at the University of Stirling, is a scholar equal to this challenge. Paul’s influential work has been central to shaping labour process analysis and debates, which Manufacturing Consent opened. Having extensively engaged with Burawoy’s work, Paul is uniquely placed to comment on the book’s legacy, over 40 years on. In the conversation which follows he reflects on how Burawoy is being taken up today, how Manufacturing Consent and its successor, The Politics of Production, changed the discipline, which of its insights have endured, and what still requires reflection and critical conversation.
Keywords
Few texts have done more to shape how sociologists and social theorists think about workplace politics, labour agency, the everyday production of cooperation and even the sociology of work than Michael Burawoy’s Manufacturing Consent. As a result, discussing it poses a two-fold challenge of scope and of depth. Tackling the first challenge requires a profound understanding of the theoretical foundation of industrial sociology on which Burawoy built; the second is the in-depth expertise of the organisation-level issues Burawoy puts forward. Paul Thompson, Emeritus Professor of Employment Studies at the University of Stirling, is a scholar equal to this challenge. Paul’s influential work has been central to shaping labour process analysis and the debates that Manufacturing Consent opened. Having engaged with Burawoy’s work extensively, Paul is uniquely placed to comment on the book’s legacy over 45 years on. In the conversation below, Paul reflects on how Burawoy is being taken up today, how Manufacturing Consent and its successor, The Politics of Production, changed the discipline, which of its insights have endured, and what still requires reflection and critical conversation.
To start as we mean to go on, Paul, what’s your headline assessment of Burawoy’s legacy, 45 years after the publication of Manufacturing Consent?
The interesting thing about Burawoy’s legacy is the kind of research people are doing now under his influence. One can treat that in a kind of objective way. In other words, it exists, and it’s obviously doing important things. But I have this argument that some people may find curious, which is that some of this work is being done in spite of what Burawoy did.
That sounds fantastic and I’m anticipating some of that critical engagement! Let me enquire about your initial impression or your first encounter with Michael Burawoy’s work. Can I take you back to that moment, at the end of the 1970s? Was there anything that stood out, or anything to which you had a strong reaction?
Yeah, I mean, I do oddly enough. And there’s no doubt that reading Manufacturing Consent for me was as important as [Harry Braverman’s] Labour and Monopoly Capital (1974). A little anecdote here. In 1983 I wrote The Nature of Work (Thompson, 1983), which became for many people, many academics, young academics, their introduction to the labour process debate; it did have, you know, a kind of orienting effect. In it, I really foregrounded Burawoy’s work, which was sent out to reviewers by what was then Macmillan Press. One of the reviewers, a very prominent scholar, accused me of being obsessed with Burawoy. So, that shows you how important an effect it had. . .
Did this ‘obsession’ influence the questions you addressed when writing The Nature of Work?
The question, which I think I opened with from the first page of the first chapter in The Nature of Work, about workers’ own agency. At the time, the Left was considering ruling class ideologies and the general effects of the class structure and so on. He’s [Burawoy] encouraging us to think of the everyday dynamics in which workers are bound up. Alex Wood [economic sociologist at the University of Cambridge] uses the term ‘micro foundations’ to discuss this and I think that’s a good term (Wood, 2026). Of course, this doesn’t require us to stop talking about managerial controls, because – don’t forget – labour process theory at the time was emerging as this, if you like, control and resistance model. Richard Edwards [American economist] is famous for his Contested Terrain (1974/1979) book. His argument is that management builds up these systems of control, which generate their own contradictions. Workers build up their own resistance or counter measures to this. This is kind of true, but there is this other dimension of worker agency around their own day-to-day practices.
So I was always very clear. Labour process theory had to be a control, resistance and consent model. So, to circle back to your original question, my ‘younger self’ criticism of Burawoy was actually not about consent, which I thought was a really important breakthrough. It was about his regime constructs.
Have your views changed over time?
I am now much more critical of how pessimistic the consent idea is, for example, as expressed in Burawoy’s The Politics of Production (1985). The text is almost a compendium of different things he’d written. So, a lot of it was already there. If you think about Manufacturing Consent, it’s partly a general conceptual and empirical argument about worker agency and consent, at a fairly general level.
But it also contains a regime argument – about how that agency interacts with the bigger picture. At the time, of course, this was the notion of hegemonic regimes: basically, a kind of social settlement inside the workplace, which Burawoy called the ‘internal state’.
Later he began to call these things ideological and political apparatuses. But the argument was, in a sense, that consent is ultimately based on everyday practices, reinforced by production apparatuses that are linked to wider processes to do with the state, labour markets, and so on.
And already then I was identifying a problem of pessimism in his work. Burawoy argues that you can’t play a game and question the rules at the same time. This is pretty bleak. It’s combined with this idea that hegemony enables workers to buy into the capitalist system, or at least capitalist relations of production. Then there’s the claim that this hegemonic regime had been present for about four decades. I never bought that. I thought it was US-centric, and far too universalised across time and space.
It also didn’t make a massive amount of sense in the UK context at the time, where we were in the middle of the Winter of Discontent. Industrial action had already brought down one government in 1974. It was just about to bring down the Labour government and usher in Thatcherism. So, the whole ‘hegemonic regime’ thing always felt overblown to me.
It was part of a wider debate in labour process research: people reacting against overarching, linear regime constructs, like Edwards’s (1974/1979) movement from simple control to technical control to bureaucratic control. There was a reaction against that, and I think a lot of labour process scholars, particularly in the UK, never really bought into the regime construct – not the idea that there are regimes, by the way, or that you can give them names, but the idea of overarching regimes across massive time and space, covering the whole of capitalism.
When he got to The Politics of Production, he replaced the hegemonic regime with hegemonic despotism – a new, all-pervasive, all-consuming regime. So even when I was celebrating Burawoy’s impact, I never bought into those aspects of the regime stuff.
Thank you. I’d like to go back to a point you just made and ask you to say a bit more about the, perhaps problematic, certainly complex, relationship between consent and coercion. What do you think Burawoy got right about labour agency, and what do you think is missing from his account?
I think there are two things going on here that can be distinguished. One is the consent–coercion binary, which has provided the conceptual language for a lot of research. It’s clearly present in Burawoy’s own work: his regime constructs are based on consent and coercion (if not a strict binary, then at least a spectrum) where regimes are characterised by the extent to which managerial practices, or state practices, are organised around consent or coercion. And if you look at the language people use at labour process conferences now, we still get a lot of papers naming regimes, for example, as varieties of despotism. Burawoy provides a conceptual vocabulary around coercion and consent that allows people to be flexible and creative. And it’s important to say that people who apply Burawoyan categories aren’t just blindly following the original conceptual architecture.
I remember going to a labour process conference in the States – 1998, I think – and there were still a lot of his PhD students there giving papers where the punchline was always ‘hegemonic despotism’. You’d be led through all the detail and then, at the end: yes, and this shows it’s a regime of hegemonic despotism. And I think that’s a classic example of conceptual overreach. If you compare that to now, most labour process research doesn’t buy into that kind of extension of regimes across time and space.
You are suggesting that the consent–coercion binary, as you put it, is too abstract?
People tend to be much more specific now. That is, regimes that might be dominant in particular sectors, or at particular times and places, in particular countries. So they develop this language – flexible despotism, this kind of despotism, that kind of despotism. And they could, if you like, go to the other end of the spectrum and name consent categories too. But obviously times have got darker. Since The Politics of Production, we’ve had the financial crash, financialisation, austerity, precarity, so you see that reflected in the regime categories people use. We’re getting more ‘despotism’ – more coercion than consent – in how people describe things. Though nobody is really arguing that consent isn’t there, or isn’t an issue, or that there are no mechanisms producing it.
I mean, I’ve gone through my long career doing workplace regime research without ever using ‘consent’ and ‘coercion’, in a Burawoyan sense, as labels for regimes. And neither do most labour process researchers. But to be honest, I don’t really care. I’m not critical of people who do use it, if what they’re doing has explanatory power. It’s part of the conceptual architecture of labour process theory. I’ve just never felt the need to and partly that’s because I think the consent–coercion binary tends to diminish the complex dynamics of which workplace regimes consist. For example, when I look at workplace regimes, I also want to look at the mechanisms of value extraction; the mechanisms through which people are placed in jobs, and the links to race, gender, and so on; work intensification; work organisation. In other words, you can’t filter everything through consent and coercion – and I do get a bit cheesed off at times when people try, when that becomes the whole picture.
In other words, it is not a case of the consent–coercion dynamics being absent, but there being more nuance behind the terminology?
When you look at Burawoy’s work, I think you can distinguish an argument about consent itself from the consent–coercion regime construct stuff. There’s no doubt that when you read Manufacturing Consent, Burawoy is very aware of the whole legacy of British and American industrial sociology – effort bargains, games, relative satisfactions and compensations, and so on. It’s less obvious on the page, perhaps, but it’s there to an extent. He was also strongly influenced by Adam Przeworski and rational-choice Marxism – this underlying notion about games in general, and the idea that workers can be treated, in part, as rational game-players. Przeworski’s work, Capitalism and Social Democracy (1985), say, has a parallel to Burawoy’s argument about why workers buy into capitalist relations of production. Przeworski’s argument is that workers buy into reformism because reformism is a game that produces outcomes that benefit people. They don’t buy into it because they’ve been misled by bad leaders, or false consciousness, or some other nonsense. So you get this parallel: workers as game-players.
And by the way, there’s another aspect that very few people ever pick up on, but it’s there in Manufacturing Consent: Burawoy was also playing around with ideas of Marxist psychology. He talks about this, although he doesn’t ultimately take it anywhere. But he’s trying to think about workers’ predilection for games and there’s something there where he links it to human nature in some way. So, all of this can be distinguished from the regime constructs as such.
And look, I think there’s a lot to be said for the consent argument. I thought it was flawed at the time, but going back and looking at it more, I can see more of where it comes from. Still, it’s an extraordinarily strong claim when you think about it: that when you play a game, you can’t question the rules. That’s an incredibly absolutist argument with an absolutist outcome. Game playing leads to consent and, in turn, to the obscuring and securing of capitalist relations of production. They’re not the same thing. Clearly, obscuring means workers don’t see their own exploitation, in effect. Securing means it enables capitalism to go on reproducing itself. Now, the point is: as a general argument, it’s astonishingly pessimistic because it doesn’t really leave us anywhere to go.
This is interesting! You regard consent as absolute and therefore devoid of hope – for the worker?
In this framing, consent becomes a mechanism for the endless buying into and reproduction of capitalism. And where I’m going with this – although it wasn’t obvious at the time – is that a few years after The Politics of Production Burawoy lost interest in the workplace and production. And not only lost interest in it in the way we all sometimes lose interest in spheres we’ve been researching: he began to argue that exploitation and production were no longer a source of resistance, or class struggle, or social change and he basically moved to market exchange and civil society. This wasn’t a minor argument. This was his fundamental direction in what he then called sociological Marxism. He no longer wrote about the workplace and production. So all these people who are now applying Burawoyan categories are, in many ways. . .
Behind the curve a little bit?
Well, they’re kind of engaging with the ghost of Burawoy. Not in terms of his passing, but in the sense that, for the last 20, 25, 30 years, it’s not been his concern. He’s had no interest in work and production.
There are a few problems with his argument about securing and obscuring – this ‘obscuring’ stuff – and the claim that you can’t play a game and question the rules. And I think there are two sides to that coin.
One goes back to agency. I learned this predominantly from Alex Wood: Burawoy was really only interested in actions and outcomes. He didn’t show much interest in the meanings that workers brought to their work activities. He even has this line, buried in a fairly obscure article, where he says that Braverman was right not to look at workers’ consciousness and resistance; he was just wrong in the reasons he gave for doing so.
But for whatever reason, Burawoy wasn’t interested in workers’ subjective motivations. He simply said: if you play this game, your actions produce securing and obscuring. But when you dig into the research, into any real workplace situation, you find that workers are often highly insightful about relationships of exploitation. They know what the game is. It’s absurd to say the game is somehow behind people’s backs.
That’s not to say they understand the origins and consequences of exploitation for capitalism, but they know it’s a game in which they’re active agents. And of course, he’s right that a lot of what we do can generate consent in different ways; but it doesn’t necessarily mean it automatically generates consent, or that people can’t be critical of exploitative relationships and act on those criticisms at the same time as they’re reproducing capitalist relations of production.
I’m currently writing a paper with Sarah Nies (Professor of Sociology, University of Göttingen, Germany) on what she calls people’s ‘subjective interests at work’ – which often lead them to be critical of the games they’re embedded in (Thompson & Nies, forthcoming).
And there’s a second point that’s actually consistent with Burawoy’s conceptual architecture, but he doesn’t develop it properly, which is: how many games are going on at the same time? So you’ve got, if you like, the labour process game (which is another version of the effort bargain) but you’ve also got other games. And I’m using ‘game’ loosely here, because I don’t necessarily want to call everything a game; I’m just playing along with his conceptual architecture.
You’ll know, because you’re in these kinds of departments [management; business school], that you get concepts like the psychological contract. It’s not a term I use, but if you ignore the label, it’s about the broader employment relationship in the firm – what people expect from employment, from the contract, from the relationship. . .
And since these expectations and the contexts in which these games unfold are dynamic, the outcomes cannot be expected to generate consent from the outset?
To me, if you treat that, roughly, as a game, with its own rules, institutions, mechanisms, perceptions, then that game often undermines the consent that might be generated in the labour process game. That was, in effect, my argument in the ‘disconnected capitalism’ thesis: workers may want to engage in more innovative work practices – teamworking and so on – because they want work to be more interesting, they want it to have use-value, and so on. But then they run up against work intensification, employment insecurity, and so forth – precarity and insecurity undermining whatever consent is generated in the labour process.
This is actually rooted in Burawoy’s own conceptual architecture, where he distinguishes between relations in and of production – the relations in production as the labour process, and the relations of production as exploitation. In practice, those things can’t be distinguished. You can distinguish them abstractly – Marx has his definition of the labour process and its different components – but in practice, when people do their work in particular contexts of exploitation and exchange, those things are always connected.
The point is: there’s always more than one game going on. And I think the problem with Burawoy’s notion of games is its tendency to universalise the process; separate it from other processes. You know, ‘you can’t play the game and question the rules’. Well, what if those rules run up against other rule-based processes and mechanisms?
And my point would be that when you look at contemporary research on games, particularly in digital labour, a lot of this is going on. Looking at the literature on digital platforms, scholars do start from games, consent and worker agency, but in practice they almost always reach the conclusion that consent is conditional, partial, contested, or all three. I can’t find many papers where someone actually follows the argument that you can’t play the game and question the rules at the same time.
Now, that’s a really powerful point. Let me try to summarise it and link it back to your recent paper. On the one hand, there’s Burawoy’s tendency to generalise – to present ‘the game’ and its rules as not only widely applicable, but also singular: as if there is one game and one set of rules. As you’ve just suggested, what if there are multiple games and multiple rule-sets operating at the same time? In your paper, you argue that these kinds of generalisations are problematic not only for understanding subjective work experience, but also because they can obscure – or fail to recognise – forms of resistance that are displaced, either geographically (to the Global South) or temporally (into the future).
So let me make a slightly controversial point, and ask you a question. Has this tendency to generalise – whether through the consent–coercion spectrum or through the ‘game’ metaphor – ended up leaving the person out of the process? Has it become more about the game than the people who play it? And if so, in the context of newer research on gamification and algorithmic management, how do we rectify that – so that Burawoy remains relevant going forward, rather than becoming obsolete?
I think, to start with, the point in your last question, about Burawoy’s pessimism over agency. I haven’t really spoken about this yet, but when I first read The Politics of Production, I was incredibly excited by the argument about a ‘politics of production’. That argument wasn’t present in Manufacturing Consent. And ‘politics of production’ is a term that has really stuck in labour process research. There’s no doubt about it. Labour process research focuses on workplace regimes in the broadest sense – some use Burawoyan categories, some don’t – and within that, it focuses on the workplace politics that arise from the contestation between capital and labour in the workplace, rather than in the broader society. That’s its domain specialism, if you like.
And Burawoy has this fantastic argument in the introduction to The Politics of Production that speaks directly to your point about whether we’re forgetting real agency. He argues that we need to focus on what workers actually do at work, rather than on their imputed interests – generated from what he describes as a kind of messianic Marxism, where workers are attributed the interests of being the universal class that will liberate humanity.
He says: let’s forget that, and focus on workers as real agents at work, and look at the politics of production. And I thought: this is great. It gives us a conceptual language. I was also very happy that he distinguished this from other domains – state politics, gender politics, consumption politics – because we need, on the left, a pluralist notion of politics that doesn’t read everything through some universal lens of class struggle.
My point, looking back on it now, is that even though he developed a very clear, potentially usable concept of ‘politics of production’, it bumps up against the rest of his conceptual architecture because you can’t really have a politics of production if everything that workers do secures and obscures capitalist relations. Those things just can’t coexist.
How did Burawoy resolve this seeming impasse?
Two things happened. First, the ‘politics’ part of ‘politics of production’ became something that was mainly about external political and ideological apparatuses – so in other words, where the state impacts the workplace. And you still get this with some people who apply Burawoy today; that the politics of production is not actually about production but about how the state interacts with production. And then even that diminishes to the point where production and exploitation is no consideration. They’re no longer where the action is, or where the potential for political change lies. In a recent paper in Work in the Global Economy, Alex Wood argues that a lot of this was to do with Burawoy’s disillusionment about Eastern Europe – his work there, and his belief that a form of democratic socialism would emerge out of the collapse of communism. Alex knows this far better than I do and it is not something that I can make any great claims about.
Second, Burawoy lost any interest in the agency of workers at work. With the sociological Marxism stuff, he switched, particularly, to his work with Erik Olin Wright on real utopias (see Wright, 2010) which was much more about decommodifying social relations and experimenting with alternative forms within civil society. And I’m not going to knock those arguments for themselves. But I will knock the either/or argument; the claim that we should abandon the workplace as a source of change and contestation.
So that’s a slightly long-winded way of coming back to your point about worker agency. I think we have to go back and ask questions that are often missing. One of the problems, even now, when you read a lot of work by people who are writing in a Burawoyan tradition is that they’re not asking deep enough questions about what consent is, and what workers are actually consenting to, as if consent is one thing. And that’s how you end up with this complicated picture of work where workers are consenting to some things but not others.
So, in order to dig deeper into this, we have to focus more on workers’ subjective interests at work on the meaning end of the agency spectrum, rather than only the action end of the spectrum.
When we talk about agency, there’s a meaning dimension and an action dimension, and both are really important. People can act against their employers at work without necessarily being fully conscious of the underlying social relations. But Sarah and I are very clear that there remain sources of critique of capitalist social relations that arise out of people’s subjective interests at work – what they want to do at work, how they think about their jobs, how they think about their identities at work. And these create multi-level and multidimensional notions of consent.
Thank you for such a comprehensive and engaging overview. So, this will be my last question, and I think it’s an important one, because it aligns with your forthcoming paper with Sarah Nies, and it also feels like we’re beginning to set an agenda for scholars who want to follow and develop Burawoy’s work.
You mentioned earlier that, very early on, Burawoy flirted with a kind of Marxian psychology – trying to get at workers’ motivations – but he never really developed it or at least didn’t return to it. And you’ve also talked about how, in your own work, you’ve come to see the need to keep the worker and worker subjectivity front and centre.
Part of the issue, as you’ve put it, is that resistance can sometimes be an action without meaning – without sense-making on the part of the worker – and the challenge is to connect those two pieces: action and meaning.
So, let me ask a potentially unfair, or provocative, question: is Burawoy’s conceptual architecture sufficient to achieve that? And if not, what would you supplement it with in order to keep subjectivity front and centre?
It’s a very fair question. And in fact, I’d criticise my own work with Stephen Ackroyd) on misbehaviour at work for much the same problem (Ackroyd & Thompson, 2022). That’s not to say I don’t think it’s valuable: the misbehaviour argument was about expanding our conceptual vocabulary for worker agency; about showing that everything wasn’t necessarily capital-R Resistance.
And don’t forget the context: that work was developed in a period when both HRM and post-structuralism were, in different ways, pretty much writing off agency at work. So, we did something – in a very Burawoyan spirit – we expanded the conceptual vocabulary. But it was really about action. It had those four domains – appropriation of work, of product, time and identity and it generated a massive amount of research. But we didn’t really develop a parallel language of meaning-making.
I think meaning was there in the sense that, whenever we discussed what people’s behaviour or misbehaviour consisted of, you would refer to it. But it didn’t have that parallel conceptual vocabulary. In fact, the argument we tended to make was that these forms of contestation were predominantly driven by the drive for autonomy. Now, of course, autonomy itself has to be linked to people’s perceptions of themselves – as workers, as professionals, whatever it may be.
I’ve resisted going too far into a conceptual vocabulary of meaning-making from a materialist point of view. But in the paper I’ve been working on with Sarah, partly influenced by her work on subjective interests at work, we’re definitely moving in that direction. One of the things we’re trying to do is to be much clearer about the different dimensions of consent and worker subjectivity, and how they relate.
In my defence, I have tried to develop a notion of workers’ activities at work that rests on the interplay of interests and identities. There’s a chapter in the labour process theory book (Thompson & Smith, 2010), that I wrote with Abigail Marks, called ‘Beyond the Blank Slate’. And that has an interesting connection back to Burawoy’s discussion of human nature, because it treats identities in terms of scarce symbolic resources, and contestation around those resources.
So, it’s not that I don’t know what the problem is. But I think it’s a reasonable argument that I haven’t developed the conceptual vocabulary to the extent needed to meet it. But hey – maybe I’ll get there before my time is up. Who knows?
There’s plenty of time! Paul, this has been absolutely incredible. It’s taken me into realms I hadn’t even considered. Thank you for your generosity, your time, and your scholarship.
Thank you. We often neglect conversations as a source of developing ideas, but they’re really important. A lot of the conversations we have are about developing our own work, but sometimes conversations also generate ideas, or help you clarify your own ideas – in a spirit of open-ended, open-minded discussion. So thank you for your questions. They were classically tough, but they’ve helped me think things through as well.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.