What ethnography can teach the gig economy
A conversation with Niels van Doorn
Happy Sunday!
This week I spoke with Niels van Doorn, principal investigator of the Platform Labor research project and assistant professor of New Media and Digital Culture in the Department of Media Studies at the University of Amsterdam. Van Doorn recently published an article, Platform capitalism’s social contract, in Internet Policy Review, a journal on internet regulation (check it out, it’s free!). In it, he considers the social contract behind platform capitalism through three different case studies. We chatted about how he got interested in the gig economy, the gendered dynamics of platform labor and hosting on Airbnb, and what ethnography can contribute to our understanding of the gig economy.
This interview has been condensed and edited, with links added for clarity.
Oversharing: So to start I wanted to ask, what got you interested in researching the gig economy?
Van Doorn: Well, unless I'm very mistaken, in terms of timing, I think you've been at it longer than I have. So I think you might be one inspiration.
Oh, thank you.
I read some of your articles, along of course with articles and a book by Sarah Kessler. So to be honest, it was a combination of two things. It was reading really good and super interesting journalism—I would call it investigative journalism—and thinking, wow, there's something really interesting going on here. But at the same time, since 2012, I've been an assistant professor here of New Media and Digital Culture. In 2014-15, I was mainly working on more conceptions, practices, and value in social media environments. And the notion of platform started to appear but more in a social media platform, like Web 2.0, social media, and platformization of the web.
I didn't want to really go do that. I wanted to do something else with platforms. And that was around the time when, again, I saw these things about the so-called sharing economy and collaborative consumption, remember that term? And the gig economy came up and I was like, yeah, that is super interesting, because it’s not just something that happens where so-called sharing is online and where algorithms and datafication play an increasingly big role, but it actually translates into our everyday lives. So that was for me the moment where things clicked.
It's interesting because like you say, a lot of the early work on the gig economy did come from journalism. I'm curious how you think about what can be contributed instead from perhaps a more rigorous academic perspective.
That's a really good question. Because sometimes my initial response is not much! But I think the answer to your question is, well, we generally get more time institutionally and qua funding. I'm lucky enough to have that funding for the Platform Labor research project from the European Research Council for five years, to really sit down to read and then to engage with these very situated and kind of fast-moving, fast-paced dynamics and phenomena and really kind of zoom out and place that in a particular context. And again, some of the investigative journalism definitely touches on that. But I guess as academics, we're just privileged enough to sit with the material a little longer and provide more philosophical, theoretical, conceptual—whatever term you prefer—to add that layer of knowledge-making.
Your research is mostly focused on three cities: Amsterdam, Berlin, and New York. How did you choose those as a group?
So this was back in 2016-2017. I wanted to look at cities that in their respective area were quite pioneering and that were also tech and startup hubs. Amsterdam, increasingly Berlin the last five to six years, and of course, New York as well. You can compare them because they're all in what we now call the Global North. But individually, if you look at how they're embedded in national regulatory and policy settings and also economic markets, they’re different enough where you could have an interesting comparative angle.
One of the case studies in your paper is home cleaning service Helpling and its “selective formalisation of domestic work.” Can you explain what you mean by that?
Well, so instead of just making everything formal about the labor arrangements for these cleaners, what they do is they formalize some elements of the labor process that are congenial to their business model. So you know, electronic payments, monitoring through reviews and ratings, particular forms of communication. They formalize a number of elements, but they also strategically keep a number of dimensions of informal work. Lack of professional advancement is one obvious way, but also just insecurity, information asymmetries, lack of redress or accountability. And by the way, I'm not equating informal labor markets with ‘bad’ work necessarily. But one thing I've learned is that a lot of people in the gig economy would actually prefer going back to the formal labor market because it actually offers them more and not less freedom because of this platform and how it mediates.
It sounds like part of your argument is that this ‘selective formalization’ widens the gap between the people in power, which would be the companies, and the people who are sort of disempowered in the system, which would be the workers.
Yes, because they get the worst of both worlds, in some ways. Because again, let's take Helpling, they definitely monitor [their workers], they make sure that there can be no communication with the client off the platform, etc. But this is my point. They don't only mediate or govern the labor process. They also engineer through software labor markets that are designed very much in favor of the customer and the platform. The customer is being given this interface with which they can compare and contrast super easily. And then it's just this abundance of labor, which is made to look comparable, look fungible, and you can just pick and choose the cheapest one because they all have a high rating, or most of them do. So it becomes very difficult for these [workers], even when they can set their own rates, to set a higher rate because you price yourself out of a very artificially inflated and governed market.
So cleaners on Helpling can set their own rates?
Yes, sorry. In Helpling for a while now, also after a lawsuit, they can set their own rates. But in my interviews, what comes up a lot is that ‘yeah, I mean, of course I could and sometimes I try, but I get less work because others like me also have good ratings and reviews and they have lower rates.’
Sort of a race-to-the-bottom dynamic?
Even if there's not a race to the bottom directly, there is definitely an artificial price ceiling. It's just a ceiling of how the market works. You can’t go much higher because then you're going to have less work. However, this remains relatively low-wage labor.
There is a big difference between people that are new on the platform and are still trying things out and trying to get a clientele. Once you get regular customers, there's less stress. You know what you get paid, they trust you, it's easier to negotiate higher wages. So it does get better. However, of course, who needs the platform when you have regulars?
Yeah, it’s interesting how a lot of these companies that effectively function as lead-generation services, because they allow workers to find clients, then invest heavily in making it impossible to actually separate from the platform.
Exactly. I love that term by Benjamin Bratton about how platforms operate. He calls it ‘generative entrenchment.’ So it decreases the cost of staying and it increases the cost of leaving the platform over time. And that is really what these platforms are so good at.
Have you spoken to anyone who has successfully shifted their work off Helpling?
Yes. I found people that strangely do not completely leave the platform behind but increasingly shift more and more clients offline. But interestingly enough, they never fully leave the platform. And when I asked like, ‘but you don't like the platform, why do you stay with the platform?’ they always say, ‘that's a backup in your pocket.’ Suppose other things fail. Fairly quickly, you can have that generate leads for you again and you can have some fairly quick income, and that's what it's all about.
The other case study in your article that I wanted to talk about is Airbnb. You draw an interesting distinction between the capital and labor sides of Airbnb, in that you need capital to own properties that can be rented, but the act of hosting and managing involves labor that as Airbnb has increasingly professionalized is often outsourced, your paper argues to migrants and to women.
In my PhD’s research, what we see is that the owners of these properties are usually men or companies headed by men. Because it's very idealized, this home-sharing idea. You are the owner of the home and you host and you do that hosting labor, including some cleaning, and that's it. But that's not ultimately what Airbnb is most interested in, as we know. Airbnb also needs to scale, needs to professionalize. There's a paper soon coming out by the both of us that talks about labor-based professionalization and asset-based professionalization.
The reality of a large share of Airbnb is this cottage industry that’s built on Airbnb as kind of its operating system. The platform is an operating system. There is still the host who owns, but what we see more of is either an institutional investor or a rich, wealthy family with a bunch of homes. The work of hosting is done by these management companies and the management companies, they subcontract and it goes further downstream. And the further downstream you go, the more migrants and women we usually see. And the higher upstream you go, the more men and they usually are also white.
And then finally, what my PhD student finds in his research is that it's mostly women who can do the labor-based professionalization. So being a superhost or co-hosting and then expanding that portfolio. But more often than men, they don't have the capital needed to do the asset-based professionalization, you know, actually owning your home or they have a home but there's all kinds of restrictions in Berlin.
One challenge of reporting on the gig economy is its scale. I could never interview enough people to know if something is true on a statistical level, and we also generally don’t have access to large-scale company data. But you take an ethnographic approach. So I'm wondering if you can talk a bit about why that's useful for understanding the dynamics of these platforms, even if it's not necessarily ‘statistically representative.’
There was a long time when policymakers were just interested in numbers and facts. And ethnography and sociology more generally were frowned upon. But that has really luckily changed over the last five years or so. And I think that policymakers and regulators are starting to realize that besides the fact that we need data—which we cannot get, so we need new regulations to get that data, and Europe is working on it—we also need to have an in-depth knowledge of how these technologies are actually impacting people in their particular situations.
The gig economy workforce is incredibly heterogeneous if you compare it to other kinds of industries, because it actually integrates all kinds of sectors and industries, all kinds of people. And before you can regulate properly, accurately—really for the common good—you need to know who is part of that common good and what are the frictions within that common good. Because a lot of quantitative-based policymaking says like ‘best practices, one size fits all.’ And I hope that ethnography or qualitative social research can contribute to a bit more nuance, a bit more critical thinking, and a bit of second-guessing before you actually do this blanket regulation. For instance, reclassification. I am not against reclassification, but for some groups, particularly migrants in the Netherlands, it does not always work. And I want to get that out there. I want to be able to tell people, not based on numbers, but based on experiences, what they struggle with and why we should have another thought about maybe there’s other ways to go at this.