How Uber and Airbnb Created a Parasite Economy
An examination of more than a dozen companies reveals the nuances between gig platforms and their workers
Uber and Lyft’s high-profile battle with California over its AB5 law, which requires them to classify their drivers as employees, has driven the debate over gig workers and their platforms to a fever pitch. Marker sought out the perspective of Juliet B. Schor, a professor of sociology at Boston College and author of After the Gig: How the Sharing Economy Got Hijacked and How to Win It Back, to shine some light on how these companies really treat their workers, and how these platforms might be fixed.
Schor and her team collected data on 13 different gig work platforms including Airbnb, TaskRabbit, Postmates, Uber, and Lyft over seven years, and interviewed 278 different gig workers about their experiences working on these platforms in the process of writing the book. We spoke with Schor to discuss whether these platforms expand opportunities for workers or exploit them, and if the gig economy can disrupt the discriminatory barriers of the traditional employment economy.
This interview has been edited and condensed for clarity.
Marker: What do you find most accounts of the gig economy or the platform-sharing economy get wrong?
Juliet Schor: Well, they tend to go wrong in one of two ways. On the one hand, you’ve got the boosters who just think everything about it is wonderful, and it’s all about how the technology is going to allow workers so much freedom, and so forth. Economists and the business literature have really tended to see just the sunny side of the street.
On the other hand, you’ve got the overly dystopian point of view about it, in the sense that the platforms have all the power, workers are getting exploited, and conditions of work are being degraded.
I would say of the two views, the latter has been more accurate. We have seen that borne out in our study of ride-hail and food delivery platforms, and to a certain extent, on apps like Instacart. But there’s two things about that view that my book and my research counters. One is, we have found that for people who are using these platforms to supplement their income, rather than depending on them for an income, the experiences are much better. They get better wages. They retain more flexibility and control. It’s less risky, and overall, they’re just a lot happier with the experience.
The second thing is, the negative accounts are very critical of the technology as enabling surveillance and algorithmic control, and creating precarious conditions. But the potential of the technology to create efficiencies and flexibility is really pretty fantastic.
What do these gig platforms have in common, and what are the differences between them?
First, they use some kind of matching algorithm to connect buyers and sellers. Second, they have a lot of automated quality control and HR. They crowdsource ratings and reputational information to try and ensure quality.
Third, very few platforms hire workers as employees. They almost all hire them as independent contractors. This is really important since, in comparison to a conventional employer, like a hotel, or a taxi company, the platforms cede much more control to the earners around level of work, effort, and scheduling, and that they talk about that in terms of flexibility.
This retreat from control is an important step the platforms have taken, giving much more control to the workers. The flip side of that is that the workers bear much more risk. They have to bring their own capital, and the companies contribute nothing to security. No benefits, no insurance, none of that. That’s pretty much common across all of the platforms that we’ve studied.
As a consequence of this, in comparison to conventional employment, the platforms have a much more heterogeneous labor force, in terms of the number of hours they work, their orientations to the work, the kind of earners that they are.
What have you found that gig workers most like — and dislike — about their work on these platforms?
It’s pretty consistent that they love having flexibility and not having a boss, even on something like ride-hail apps like Uber or Lyft, which have the most intrusive algorithmic control, compared to say, Airbnb or TaskRabbit.
We found that it’s really hard for people to make a living on these apps.
I would say the biggest problems are the wages and the availability of work. In the beginning, the wages were really quite good. But now, you have multiple problems on the lower-paid apps. Either wages are not enough, or there’s not enough work. You have too many people chasing too little work. This seems to have intensified since the pandemic. I’ve been part of a research team that’s interviewing shoppers and delivery workers, and getting tasks or getting shifts has become really hard.
There’s been a shift over time, particularly on the ride-hail, delivery, and shopper platforms, toward more full-time workers. We found that it’s really hard for people to make a living on these apps. There’s not enough work, and they don’t pay enough.
What Went Wrong at Uber
Super Pumped author Mike Isaac and Gigged author Sarah Kessler discuss the cautionary tale of Uber
So would you say that the fundamental divide in the experience that workers have on these platforms is whether or not they are dependent on platform work as a primary source of income?
Yes. We have a paper out on that called “Dependence and Precarity in the Platform Economy,” and that’s exactly the argument we make.
There’s also what we’ve called the platform hierarchy. Some of the platforms just pay more than others. If you’re higher up on that hierarchy, you’ll do a lot better. To some extent, it’s correlated with the amount of capital you can bring into it, so on a platform like Airbnb, you can actually earn much better. We also didn’t have any workers in our sample who were fully dependent on Airbnb for a living, unlike platforms like Uber. We had some who were partially dependent, a middle category.
Do the platforms themselves have any kind of preference or bias for whether their workers are dependent on them?
I haven’t done many interviews with people from companies, so I don’t have a sense from them directly. But I would say it’s pretty clear that the platforms try to get people to work longer and do more. We’ve especially seen that with ride-hail’s use of gamification and bonus structures. You’ve got it in delivery, now, too.
It’s complicated, because on the one hand, we’ve found in our research that the dependent worker is a more controllable worker. Workers who are “supplementals” won’t necessarily do what they’re expected to do. They’re not as bound by the ratings. They do things more the way they want to rather than what the platform policies say.
In that sense, you could say, the platforms should want a dependent worker. On the other hand, the supplemental workers may be more desirable workers as they are often the workers with higher levels of education and tend to be more satisfied with the work since they aren’t doing it out of desperation.
When I started studying TaskRabbit, one of the things I observed (though this wasn’t something we studied) was that there were a lot of highly educated people on TaskRabbit. College graduates and people with graduate degrees, too. I thought, because the customers tend to be pretty well-off and highly educated, maybe they prefer people of similar social class and education.
You have a really interesting section in the book where you describe the gig work platforms as parasites. What do you mean by that?
I say they’re parasites because in order to have a satisfactory experience on a gig work platform, you need another employer or another source of income. If you’re a dependent worker, it doesn’t really work. It’s hard to even earn up to poverty level.
Those conventional employers are paying the benefits. They’re paying the salaries and so forth. The gig platforms are living off that. They’re parasites in the literal sense that they need that host to be able to keep those workers coming. In other words, they’re free riders.
Uber and Lyft’s Business Model May Be Dead. Good.
The biggest startups in modern history were built on old-fashioned worker exploitation. Time for an upgrade.
In your research, have you found that these platforms break down or reinforce social inequalities?
The platforms claim that they’re going to break down social inequalities. They’re very easy to access, and they should help people who are otherwise discriminated against, because they don’t have the same institutional structures of discrimination that exist in the conventional economy.
Why would you expect that just because it’s a new kind of business, inequality would suddenly melt away?
We studied this in the context of Airbnb. The idea was that, if you’re a Black renter or homeowner, we know you’re getting discriminated against in the “legacy economy.” Airbnb comes along to open up a new frontier with a ratings and reputation system that reveals true information about you and is supposed to make it fairer.
But inequality reinscribes itself even in new places. Why would you expect that just because it’s a new kind of business, inequality would suddenly melt away?
On Airbnb, we found that in non-white areas, non-white neighborhoods, people did list homes at a higher rate, which supported the idea that people of color would be more eager to try these new opportunities. But they were still getting fewer bookings, lower ratings, less revenue, although not by huge amounts.
The audit studies that have been done on Airbnb do show discrimination, with hosts rejecting guests of color, and our work and others have shown that hosts also receive less if they’re in neighborhoods with more people of color. There’s one study that coded hosts by race using their pictures, and showed that they earn less.
Hotels are governed by laws against discrimination, whereas it’s legal to discriminate on Airbnb. If you showed up on my phone with an Indian name, and I decide I don’t like Indians, I don’t have to let you stay in my Airbnb, and that’s perfectly legal. It’s not legal if I’m running a real hotel.
In ride-hail, meanwhile, it’s been much better for the passengers, because taxis were much more discriminatory. If you’re a Black person, it’s a lot easier to get a ride-hail than it was to get a taxi. Ride-hail drivers are more willing to go into neighborhoods of color than taxis, but it’s still not perfect. Studies are showing more cancellations if you’re Black and longer wait times. If you’re a woman, there was one study showing you get taken on a much longer ride, and you pay much more. There are a couple of other studies of other platforms showing reluctance around delivery and tasks in Blacker neighborhoods.
So the platforms haven’t eliminated discrimination, but I do think that, outside of accommodation, they’re somewhat better than the legacy systems. There are more women ride-hail drivers than there are taxi drivers, for example. Another good thing about the platforms has been that a lot of people with disabilities use them, because of the ability to work whenever you want. There are a lot of disabilities where people don’t know, day-to-day, or even hour-to-hour, whether they’re going to be able to work. They’ve been good for that.
Do you think it’s viable for these platforms to reclassify all their workers as employees?
It’s viable in one sense, which is that they could classify their workers as employees. It would be more costly for them, because they’d have to pay more, and pay into the various benefits funds, unemployment insurance, worker’s comp, and so on. There are going to be more issues around the hours worked, because some of these drivers work very long hours, because the wages are so low.
What’s not viable for the platforms is to charge what they’re charging customers and pay more to the drivers. That’s where it gets hard, because they’re already losing a lot of money. That’s why they don’t want to do it.
With colleagues at Northeastern, I have been studying a delivery platform that did convert its independent contractors to employees in California. The data we’ve been getting are showing that the shift produced a lot of benefits around productivity and reliability of the service. There’s no question they can do it. It’s just a question of the price point, I think.
The companies argue that the workers will lose flexibility if they do this. But in the company we studied, the workers still retained a fair amount of flexibility. They didn’t have as much flexibility as in Uber and Lyft, where drivers can shut the app on and off at will. These workers would have to sign up for shifts more, but you can still retain a fair amount of flexibility. Just not as much.
The company would have to do more to get that elastic labor supply that it’s relying on. They would probably have to pull back from dynamic pricing. They wouldn’t want to give employees the same amount of flexibility, because once a firm has per-person costs that don’t vary with hours, like unemployment insurance, worker’s comp, and so on, which are capped, they just want the workers to work really long hours, because those hours are then cheaper for the firm. The solution here is to figure out a cost structure that will really not put in a big disincentive for flexibility.
The other thing they’re going to have to do is put in a vehicle cap. This has already happened in New York. They’re talking about it in California, too, because a firm is not going to be able to give employee status to everybody, so a cap would probably be the way to make it work. You’ve already got too many drivers there and too much idle time — dead miles, they call it. Right now, the estimates are looking like about a third of the time, drivers are sitting without people in the car.
My Startup Could’ve Exploited Gig Economy Workers — Here’s Why We Didn’t
By making every one of our employees a W-2 worker with full benefits, we’ve helped them succeed and done right by our…
Is there a business model that you see as better than what these platforms currently offer?
There are some platforms that take a small-ish fee for arranging transactions between individuals. That model works. That’s Airbnb’s model. Etsy uses that model. Ride-hail by contrast is a labor control system. It specifies the routes, the drivers don’t pick the passengers, or determine the rates, and so on, but these others really are more like matching exchanges.
Where there’s a reasonable fee for arranging those transactions, and maybe some insurance, which is part of what Airbnb is offering, that seems like a good model to me. They’re not setting prices, for one thing.
Uber’s low prices were a way to achieve market domination. They didn’t have to price so low. Uber could have still had a business, but it was attempting to wipe out all the competitors, including public transportation, which it admitted in its IPO documents, although it later took that back.
In ride-hail, the disruption came about because you had a regulated industry where a lot of rents were being extracted by the owners of medallions. Then, ride-hail came in, and basically illegally began offering cheap rides with massive subsidies by investors, so it was very easy to knock off that industry. Taxi was the industry everybody loved to hate.
What Exactly Is Uber’s End Game?
Food delivery is a losing game — so why is the most aggressive unicorn on the planet betting everything on it?
In accommodations, what companies like Airbnb did was also illegal, because a lot of cities had laws against people renting out for these short periods of time. The authorities just looked the other way. If you ask how these platforms were able to disrupt so much, the mainstream answer would be the new technology, convenience, and cost. The critic’s answer is, they disobeyed the law, and so they had unfair advantages.
I think cooperatives are the best model out there. Those are platforms that are owned and governed by the workers. The technology, like the matching algorithms, and the rating and reputations systems, are now pretty replicable, and they allow companies to get rid of huge amounts of human management. That means that a worker-owned firm with the technology is just so much more efficient, because there’s just the workers and the technology. Why do we need these investor-owned firms? If the workers could get some access to capital, they could run these technologies themselves.
Are there examples of that happening already in the world?
Yes, there are. I did what I think is the first academic study of one of them, a stock photography platform called Stocksy. It’s got, at this point, over 1,000 members, and has been successful. It was started by people with a lot of experience in the business, and they were able to attract really great artists because of the co-op feature. They pay a lot more than Getty, which is the dominant platform in the industry.
You have co-ops in house cleaning, ride-hail, and bicycle delivery too. There’s something called Fairbnb, which is an alternative to Airbnb that puts money back into the local community. There’s a really big freelancers’ cooperative called Smart, which operates in more than 40 cities in Europe.
These are models that are working, today, and there’s actually a lot of interest in them. There are a number of different groups that are doing incubators for platform co-ops. I think you’ll see a lot more of them in coming years.