share this article

Listen to this episode and subscribe to the FA podcast on Apple PodcastsSpotifyGoogle Podcasts, Overcast, or wherever else you usually get your podcasts.

When the gig economy hit cities across the world in the early 2010s, gig companies promised flexible working hours to their “contractors” and on-demand ease to their customers. In reality, the companies and their algorithms have induced a monumental change in patterns of work and consumption, recomposing commercial districts in pursuit of more efficient last-mile delivery and invisibilising deeply exploitative and often criminally underpaid labour practices.

In a bid to understand this new form of “algorithmic management”, researcher Callum Cant took a job riding for Deliveroo, a food delivery startup that was founded in the UK in 2013 by Will Shu and Greg Orlowski, which has since become one of the most prominent new gig companies. Riding for Deliveroo is also the title of a book that Callum wrote about his experiences and the subject of this two-part Breezeblock interview.

This first conversation involves a discussion of the “workers’ inquiry” methodology and class composition analysis that inspired Callum’s research, both of which draw on a remarkable history stretching all the way back to an obscure 1880 essay written by an equally obscure German philosopher. But first, it begins with Callum offering a stirring description of what it’s actually like to ride for Deliveroo, a job that forces the workers into a relationship with the city that is, by equal turns, both dangerous and exhilarating.

Listen to the second part here.

Below is a transcript of the conversation.


Riding for Deliveroo w/Callum Cant (pt.1)

Deliveroo strike

Duncan Cumming/Flickr

Charlie Jay Clemoes (CJC): It might be nice to talk what these apps are doing to people’s bodies. There was this part where you were talking about it in very specific detail about the way that the delivery process occurs, at least in the hands of the rider, where you find out very late on in the actual process. The app tells you OK, you need to go here now so there’s a certain unpredictability about whether or. You want to take a job and then also there’s, as you keep doing it, a repetitive process of learning the city and learning its mysteries, in a sense, it’s shortcuts, its obstacles as well, and I don’t know. I just thought it was one of my favourite bits of the book where you talked about it in terms of the video game Crazy Taxi. 

Callum Cant (CC): Yeah, it is crazy. Taxi. Yeah. 

CJC: Yeah, yeah. So, do you want to just talk through the process? 

CC: Yeah, for sure. Well, I mean. So, in a shift you’d go out of an evening and you’d probably have a number in your head. “I want to do this many deliveries and make this kind of money. I’ve got these hours to work”. When I was working, you’d go out and be freezing cold and you’d want to start working straight away. But usually, because you’re heading out at 5pm or 6pm, it’s a little bit before demand has really picked up, you don’t want to miss any of the initial surge, right? So you head out a little early and it’s freezing cold and you start to get cold because you’re just sat around and you sit at the bench, you chat with people, you hope that you’ve got enough battery in your lights and your phone because, in my case, I’ve been working all day and I’ve only just been back at the flat for 20 minutes or something and come out again. You’re probably quite hungry. And then suddenly you start getting notifications, right? And you have the app up and you just swipe, swipe, swipe, swipe, swipe, swipe. You stick on your bike and then bang, you’re like, “OK, I’m off to this restaurant”. Once you were at the restaurant (at the time, this has now changed). It was only after I clicked the order I’d be told where I was going and the whole time you’re trying to go as fast as you can. You’re trying to save as much time as possible, but you’re also trying not to exhaust yourself because you need to keep working for the dinner shift or like work three or four hours, right? So if I sprint everywhere, I’m going to be too knackered to work for the whole time, so I need to hit the maximum consistent intensity I’m going to be able to maintain for that whole period. And there are these periods of very frustrating waiting where you suddenly go from being very, very fast to the restaurant and then bang, you’re waiting inside, and you’ve got 10 minutes because the burger isn’t ready and you just sit there for 10 minutes unpaid. So it’s just constant acceleration, deceleration, the constant search for spare seconds that you can pick up because this is a major impact of the piece wage. Right, it’s that you are doing the work, intensification yourself, you are forcing yourself to go fast. Then there’s these wild spikes and dips and adrenaline as you come very close to buses and taxis and you’re flying up a hill to flying down a hill, someone tries to start a fight with you. You get very angry because you’re not making any money or you’re like, “Oh my God, I’m making really good money and need to make up for that terrible day last week and I need to do as many orders as possible”. So it’s quite a roller coaster. The guys who do it all the time must find a way to calm down, but I find myself very emotionally stressed by the whole labour process. And there were these moments of intense enjoyment like I loved learning the city. That’s one of the best parts of it is to start being on your bike. I love riding my bike, but it was getting to know where I am at what time and what happens and seeing all the little back alleys and wanting to get across the city. I can do it this way and be so much faster than Google Maps or whatever right? Beating Google Maps became quite an enjoyable thing. So, in summary, the labour process is underinformed, overstressed, very fast, then sometimes very slow, then very fast again, very variable and, like I say in the book, there are a couple of occasions where I got in full-on screaming matches with people. I’m quite a chilled person and there’s this occasion, the guy who came along with a bottle of cider, who tried to start a fight with me from a van, which is, I swear to God, 100% true that he just, like, went out of a van and screamed at me in the face. “Your mum’s a cunt,” and then I was like, “No, your mum’s a cunt” and it’s like bizarre. But that was the kind of emotional state you were in. You were so amped. And I think that, yeah, that whole process, the whole experience of work, you’re right on the edge. Because there’s no security right, it’s precarious in an employment sense but it’s also emotionally precarious because you don’t know if your time is going to be wasted. You don’t know how valued you are. You don’t know how much you’re going to make, so that double use of precarity is, I think, quite emotionally true. It’s shit. When you have a bad shift, you feel awful when there are no orders, you’re just: “Why the hell am I here? What am I doing?” I remember I did one Uber Eats shift where I made £2.00 an hour and I tried to work for four hours on a Sunday and I did three or four deliveries, and it was just pointless. It’s like what, what, “what am I doing? Why am I doing this?” And it can have a really intense impact. So, I think we underestimate, people who have worked most of their lives in relatively stable salary jobs, how time can feel when your time can either be valuable or not, and there is this constant variability in what your wage rate is. And I suspect that a lot of riders get very desensitised to it over time. Certainly, that would be my impression from working with people and talking to people. But it’s easy to overlook the fact that that’s actually a deeply unpleasant aspect of the work. Particularly the health and safety stuff, I mean. I think it should be illegal for anyone to pay piece rates to people who work on the road, because fundamentally the faster they go, the more money they make. You’re already paying them, probably below minimum wage after costs, so they need to go really fast to try and make decent money. So basically, you’re incentivising people to helter-skelter around the city, taking all sorts of risks, mixing it up with taxis and buses, all the rest of it, just so that people get their dinner faster. I mean there are many elements of the Deliveroo model that are morally reprehensible, right? But I really do wonder how Will Shu the Deliveroo boss sleeps at night knowing that that payment system, that’s what it does, that’s the point of it, that’s the point is to make you go fast so that you work really, really hard and you deliver food quickly, but the cost is huge. 

CJC: The thing you said about the excitement, but also the danger of it all and talking about here as well, the fact that like it does encourage risk taking. To what extent do you think it does turn you into, in a sense, an appendage of the algorithm, or an appendage of the app? I don’t know if I got that right… 

CC: No, no, no. An appendage of the machine is what Marx says in Capital you’ve got it perfect. Yeah. It’s really, I mean I guess the Marx reference gives it away, but I do think it’s a structural relationship where living labour becomes just the moving bit on the end of dead labour, right? Or like the feeling bit on the end of dead labour. There are these huge constructions of capital that are basically funded by previous generations of labour that then just flap us around like the moving bits on the end. And this is the nature of alienation in the fundamental sense. Right? You are turned from someone who has goals and objectives and feelings into, for them, just a piece of Unit 5, right? A bit of a commodity that’s been bought and that now has to be controlled in such a way that it produces value. 

CJC: I wanted to say about that as well, like the description at the beginning of the book this manager, who’s being surrounded by, is it Will Shu… 

CC: Oh, it’s Dan Warne. 

CJC: That’s it. It is funny that it’s a very striking moment of suddenly being faced with the human consequences of seemingly seamless cloud-like platform apps, which sort of present themselves as this smooth surface. And then there are these moments where suddenly the bodies are piling up in front of them… 

CC: And this happens. All the time with platforms they run away from… So at Stewart, the longest-running strike in the gig economy has been going on for ages and they set up a feedback meeting where they wanted to meet with individual workers in Sheffield. Instead, a load of them turned up with their union representatives and said “We want to talk to you with our union. We want to talk together. To you.” And Stewart’s representatives ran away. There is this thing where the actual existence of these workers is always quite an unpleasant shock to certain kinds of platform managers, they want them to remain dots on the map. They don’t want them to have physical representation. And this is one of the interesting things about you could say one of the more subjective elements of platform labour protest. Is that it is hyper-visible. The thing it does is people get together in huge convoys and roll through the city. Like in Sao Paulo there was one of these that was 1000 workers strong that went across a major central bridge. And you see pictures of it, it’s incredible. They do the same thing in Athens and they’re not platform workers per se, but the Athens couriers, right? These huge protests. And it reminds me a bit of like the demand that comes up again and again in UVW organising for justice or for dignity, there is an extent to which platform worker protest is about representing that we are real. We exist, we’re not dots. We refuse to be that. Together we’re powerful. There’s a reclaiming of that experience of being on the street and travelling together as a powerful experience and it becomes a festival procession rather than a boring strike rally and people are yelling and lighting flares and all the rest of it. So, I think there is that kind of carnivalesque element of the protest which is a direct reaction to the fact that that’s really what they don’t want you to do, right? You are manifesting yourself in a way. It reminds me of the French word “manifestation” for “demonstration”. It’s like you’re showing that you are real and in person in a way that the app seems to deny.  

CJC: Just moving onto something else, the book starts with an outline of the Workers’ Inquiry method you used in your research, could you explain what this method involves and a little bit about its history? 

CC: We actually take the term workers inquiry from Marx directly. In 1880, he wrote a 101-question-long survey for French workers. Incredibly lengthy, I don’t think we’d ever get away with sending that out now as academics, and it covered everything from “What is the price of the goods you need to survive in your town?” To “How big is the factory you work at?”, “How many workers are employed there?”, “What kind of machinery is used?”, “Is there a history of strike action?”  

The survey basically has kind of two goals within it. First of all, Marx says in the absence of factory inspectors’ reports, which is what he drew on for the writing of Capital, in order to understand the state of industry in France, he needed workers to self-report on what was going on in their location. So there was a kind of a fundamental scientific need for more information to describe the state of French industry and to understand what conditions workers were facing and so he wanted an auto description, a direct self-investigation of what was going on and he wanted workers to reach out and tell him what was happening.  

At the same time, a lot of these questions are written in a way that provokes increasingly more militant responses as the inquiry goes on. There’s almost a didactic purpose as well as an informational purpose, whereby people’s narratives about their own lives and their own work are one of the most effective ways of communicating the reality of capitalism. If you want people to understand what is life like under the system, what’s life like for other people, to what degree emotions which I perceive to be purely personal, are actually collective class expressions, all these kinds of issues, then narratives are a really, really effective way of doing this.  

The workers’ inquiry tradition kind of starts with Marx in 1880 and then continues variously through other currents. You have the Johnson-Forest Tendency in the US, a group of dissident Trotskyists who kind of give up on being Trotskyists, move to Detroit and start organising the car workers there. CLR James was part of that tendency and they do really incredible work around the car plants, writing narratives of what it’s like to work in a factory. And there’s one of those that’s called The American Worker, which becomes incredibly influential in France and Italy, where actually in France it’s first translated by a group of people who go on to found something called Socialisme ou Barbarie, which is kind of a dissident Marxist group and they do the same kind of thing at Renault plants, looking at conditions of production there and then also in Italy it becomes very influential on a group we call the Workerists or the Operaists, who are essentially a group of Italian Marxists who perceive that there is a new industrial reality that workers in, for instance, the Fiat Mirafiori Plant in Turin, these workers are somehow more militant than previously but in perhaps in a different way, right? So, when a deal gets announced between the unions and management that they don’t like they riot, and they attack the union’s offices. There’s this disconnect that they’re very interested in and they want to know what’s producing this and they come up with a whole series of things around migration from the South and the creation of new industrial processes, and so on and so forth, but essentially that tradition carries on up to Notes from Below and ourselves presently and organisations like Viewpoint in the US and others. As the Notes from Below collective, which I helped found a few years ago, now our real intuition was that work has continued to change very rapidly. Particularly, we’ve gone through this huge wave of deindustrialization that smashed the British workers movement, and then there was the creation of progressively more and more new kinds of workplaces. And myself, Jamie Woodcock and a few others were very interested in the new reality facing the British working class, particularly after 2008, because with 2008 you saw the acceleration of a whole series of long-standing economic trends that really intensified and there was a kind of a renewed capitalist assault on the working class that made, for instance, things like precarity fundamentally the norm of the British economy. Non-union service workers are a norm of the British economy, a lot of those trends took off in the period directly after 2008 and you know the repression of real wages, all that kind of stuff, that’s really that initial period and it goes alongside the remodelling of the state through austerity.  

All of this is to say that the Workers’ Inquiry is fundamentally a method that’s interested in investigating work from the worker’s point of view, that’s fundamentally partisan, that says, “We’re not just investigating this because we’re abstractly interested in the sociological reality of modern society”. Instead, it’s fundamentally a form of militant inquiry, a way of understanding the world so as to change it. It often uses methods which are very sensitive. It’s basically a lot of sociological methods slightly turned on their heads. So, for instance, Riding for Deliveroo is really an ethnography, Jamie Woodcock’s, Working the Phones, is also, likewise, an ethnography. But in my PhD, we also did semi-structured interviews. We did some group interviews that turned into a Marxist reading group with workers at a Wetherspoons pub. There are all kinds of different ways you can do this. The ultimate goal is always to solicit information about the balance of forces in the workplace that can be useful to workers and even when, like nowadays, I’m an academic, I conduct all my Workers’ Inquiry stuff from the outside, per se, there’s this attempt to give ownership and control to workers themselves and Notes and Below we still, by and large, publish more workers than we do academics, which I think is one of its great advantages.  

I also think there is a degree to which this was a submerged history of what working-class movements have always done. Having to understand the workplace, whether it’s done under the name of workers’ inquiry or not Mao is doing stuff like this where he goes and investigates the peasant movement in Hunan and starts travelling around. And that’s no particular endorsement of Mao in any direct way, but just to say that if one wants to act on the world, one needs to understand it and workers‘ inquiry is this ongoing current. So, workers inquiry, by participating in it, we’re recovering its sub-current or sub-history of what it meant to be a militant in the 19th and 20th centuries. Because often this stuff was the bread and butter of how you went and organised. Organisation requires knowledge, which requires research.  

CJC: So you’ve established workers’ inquiry at this point, and I suppose, yeah, the next question I had was about this approach you used to interpret your findings, class composition analysis, which is very geographical, very spatial… maybe you can explain what that us by way of the specific, what is the class composition of Deliveroo? But maybe it’s worth, yeah, explaining class composition by way of the example of Deliveroo. 

CC: Yeah, that sounds. Good. So, class composition is in essence the object of investigation when you’re doing workers inquiry, what you’re trying to find out is how is the class composed here. So, the first thing to say is that class composition is a set of possible parameters. Fundamentally, under this mode of production, there are certain classes that exist in certain relationships, right? There are bosses, there are workers. There’s the capital relation. There’s all this stuff that Marx sets out in Capital Vol.1. The point of class composition isn’t really to change any of that, or it’s certainly not to change any of that. It’s rather to interpret what that means in a specific circumstance. So, like, how do these big-picture forces and social relations become fundamentally embodied in one particular workplace context? And what implications does that have for how we organise and how we act in that context? So, we talk about three parts, the technical composition of the class is the first one, so this would be in production. What machinery is being used? What technology is being used? How does the labour process work? Is it cooperative or is it siloed? Are workers formally employed or are they self-employed? All the kinds of questions are actually profoundly similar to what Marx was asking back in his questionnaire in 1880: a lot of detail about how production is organised and how a group of workers is turned into a working class involved in the labour process that produces value for capitalists.  

Then we also talk about social composition. This is one of the moves that Notes from Below made, to increase the emphasis placed on social composition, which is to say that workers don’t just appear at the gates of the workplace each day and then work and then disappear again. There is a whole life that exists outside of that which can have profound implications for the balance of forces between classes. So, for instance, when a worker leaves work, are they chased down by immigration enforcement at their home address? Or is there a network of food cooperatives that allows them to live below subsistence wages? One of my favourite examples of this is South Africa. The gold and diamond mines used to be basically worked on by people who came from the interior of Africa, who migrated specifically to work in those mines, and they were able to reproduce their labour power at costs far below the cost of subsistence in South Africa. So basically, they could get paid incredibly low wages, then go home and those wages would be worth a lot more. So that created, over time, the specific relationship between races and workers and bosses that led to the apartheid system. This idea that you had had an external cheap source of labour that couldn’t be fully integrated.  

Then, political composition kind of emerges out of the other two. Given that the labour process is set up this way, given that society is set up in this way, it explores the questions: How are workers reacting and what are they doing? So, we look for instance at how forms of informal, localised small-scale resistance tip over into collective organisation, that’s what we’re really interested in. So, this is both the formal stuff of trade unions and political parties. But it’s also about whatever form of self-organisation emerges. You know, like if 10 workers refused to do a job to the point that that job stops being done and instead that problem is resolved in a different way, that would be a form of organisation that we’re interested in, but fundamentally the object of investigation here is working class self-organisation. It’s not official politics, right? So yeah, the, the, the P in political composition is very much a small P.  

So, what does this mean to Deliveroo? Well, I mean, in the first instance, when you’re looking at the technical composition of delivery, you see all kinds of factors jump out at you. One of the major ones is algorithmic management right now. Algorithmic management is a term we use to describe, more or less, the management of workers by relatively complex computer programs that can do things you would normally expect a human manager to do. So, for instance, distributing work tasks or supervising task completion or monitoring how fast you’re working, all these kinds of things. So, algorithm management isn’t just part of the platform economy, but it is a big part of what makes that possible. It’s kind of the automation of supervision in a way. Historically delivery work could have a human dispatcher on the phone being like, “OK, you’ve just done this delivery, now you need to go to this restaurant or do this delivery and they would literally be telling you job by job what you should be going and doing. Instead, that’s then replaced by this automated supervisor who does the job instead, so that’s the key factor of the technical composition.  

In terms of social composition, you might well talk about migration. What’s often really fascinating is the fact that a lot of the bulk of work done in these platforms is done by people who are part of the, for want of a better term, the urban surplus population: these people have been excluded from the labour market because of irregular migration status, racial discrimination, child-rearing responsibilities, a whole number of things that push them out of the standard labour market, and they’re absorbed by the gig economy, by platforms like delivery and they form the core of that workforce that does the vast majority of work, because this is often the only way they have access to a wage. So, in particular here we might think about the role of the hostile environment. A lot of people who work for Deliveroo are renting accounts of people who have the right to work but do not themselves have the right to work. So, if you’re a regular migrant and you’re renting an account, you say you’re getting 6 pounds £8.00 an hour, you’re going to be shelling out a reasonable amount of that to the person who actually owns the account. And you’ll take the rest home yourself. But here you can see how the border regime, which makes migrants vulnerable if they don’t have a particular status, drives down wages. Because it drives down what people are willing to accept. It makes them vulnerable to hyper-exploitation. There’s a French theorist called Emmanuel Terray who talks very persuasively about how this kind of stuff is outsourcing to the Global South, but without actually geographically moving production. You just bring the Global South to you. You bring the externality inside. It goes back to that apartheid example a bit, I think.  

And finally, in terms of political composition, well, I mean we see that over and over again, workers kick off in this context. There’s a Leeds Index of Platform Labour protest, which is really, really interesting, a global tracker of protests and you basically see overtime this huge ramping up from about 2017 of platform worker protest that comes basically out of nowhere. So really, really pronounced phenomena that we’ve now seen. If you’re to track this globally there were strikes in the UAE the other day, right, like there aren’t wildcat strikes in the UAE, but delivery workers and platform workers have been striking there. I remember back in 2017, I had just finished my master’s and I was sitting there in the most boring office job in the world, like one of the real brain-dead jobs. You can feel your brain leaking out your ears, that kind of job and whilst I was sitting there in 2017 the gig economy is obviously spreading. I’m interested in work. I’m interested in strike action. I’m interested in all these kinds of things and to that point, we haven’t really seen large-scale action and then bang out of nowhere in London we see a really big well-organised strike by both Uber Eats and Deliveroo workers. And I said at the start of the book I rang up a trade unionist and I was like, “What’s going on?” He was on the ground, and I thought this looked crazy. I’m seeing these videos and stuff like “What’s happening?” And he just said to me, “I don’t really know like this is quite confusing to me. I’m not really sure.” And to me, that was like, wow, I mean, we thought these workers were disempowered. We thought they didn’t have much capacity to self-organise. We didn’t think they were going to, you know, fight for better wages. And yet here they are. There are 500 workers. They’ve surrounded their boss, and they’re screaming at him to demand better wages. You know that’s not how industrial relations happen in the UK. Most of the time these are big, effective wildcat strikes, clearly led by migrant workers. Clearly, very militant. They’re exerting a loss of workplace leverage and have come out of absolutely nowhere. And no one understands how that happened. Right. So, from that point onwards, I was like, “Well, I really, really want to understand how that process happened”. Because it’s potentially remarkable and I think there is significance for it as we now talk about probably 4.4 million gig workers in the UK, platform workers in the UK. So, a lot of people who work part- or full-time on platforms, that’s about 15% of the labour market, for context. So, this is an expanding sector that is going to net in more and more people and we need to understand it. But also, we’re talking about algorithmic management. That is categorically spreading and it’s spreading, really, really fast now. I’ve been saying this for years. Back in 2019, when we first launched the book, I was like “Algorithm management, it’s spreading” and people were like “Uuh, is it?” and then now after the pandemic, it’s profoundly clear. Zoom, for instance, has a facility built into it to monitor workers’ Zoom calls and rate their effectiveness on Zoom calls. If they’re doing pitches for sales right, like this kind of stuff is coming not only to Amazon warehouses but also to white collar. So, I was really interested in seeing this management technology in one of its earlier iterations and starting to think through what possibilities offered workers and how we could fight back against some of the more negative implications.