One of the ways that platforms such as Uber and Lyft ensure good customer service is by allowing riders to rate drivers on an app. If drivers want to keep working, they’ve got to keep those ratings up. But anyone who has been in a service job knows that customers can be difficult and sometimes even dangerous. So, how can workers in the gig economy exert any kind of control over their jobs when they’re at the mercy of digital bosses?
That’s the question that Wharton management professor Lindsey Cameron studied in her latest paper, “Expanding the Locus of Resistance: Understanding the Co-Constitution of Control and Resistance in the Gig Economy.” It was co-authored with Hatim Rahman, a management and organization professor at Northwestern University’s Kellogg School of Management, and published in Organization Science.
The study is one of six that Cameron is conducting on gig work. A previous paper looked at how independent workers create meaning to find satisfaction in their temporary jobs, while another looked at strategies gig workers have used to navigate risk during the pandemic.
Cameron joined Knowledge@Wharton to talk about the study. Listen to the podcast above or read an edited transcript of the conversation below.
Knowledge@Wharton: A Pew Research Center survey found that 16% of Americans have earned money on a gig platform, which speaks to the scope of the gig economy. What do you define as worker resistance, and what made you want to study it?
Lindsey Cameron: The gig economy, I love to say, punches above its weight class in that 90% of Americans have heard of it. More than three-quarters have been in one of these vehicles before. A lot of individuals earn money on these platforms. It’s amazing how, in 10 years, it has become almost ubiquitous or embedded not only in American life but across the world.
I became interested in the ride-hailing industry because I was watching a lot of people, including family members, who had lost their jobs during the Great Recession, and they were doing gig or independent work as a way to close the gap. This study is a comparison between two different companies, one that we call RideHail and the other we call FindWork. In this comparison, we’re looking at how workers resist on these two different types of platforms.
RideHail is a closed labor market. You turn on the app and you drive, and there are ratings that are evaluating performance. On a platform like FindWork, it’s an open labor market, so you list what your skills are online, and customers reach out to you and say, “Hey, I’d like you to edit this website or write this source code for me.”
“What’s really interesting in the gig economy is there’s not a human boss.”
My co-author and I got together because we wanted to do a comparison between the two different types of work, one more simple, one more complex. And how we came to this study around resistance is there has been a fair amount of research or even just public attention to overt acts of resistance. That’s what we think of as strikes, as organizing, as work stoppages — the mainstays of the labor unions. We wanted to explore covert resistance, the type that is less visible to the public, organizations, and even customers.
Knowledge@Wharton: You mention in your paper some familiar examples, like the waitress spitting in your hamburger or the cashier walking away from a customer who’s rude. But worker resistance is a little different when you’re talking about gig platform work.
Cameron: Right. Those examples are not overt resistance, not the strikes. It’s covert resistance. What do people do behind the scenes? And what’s really interesting in the gig economy is there’s not a human boss. Typically, it’s the customer who’s the boss. So, what is the equivalent of spitting in someone’s food or drink in the gig economy, when there’s not even a manager looking over your shoulder?
This is sort of the comparison and the question we’re looking at in the study. I had firsthand experience in this because I actually worked on and off as a driver for three years in the ride-hailing industry. I wasn’t full-time, but I did get experience about what it was like.
Knowledge@Wharton: You’re literally going to put us in the driver’s seat of this research. What are the key findings in this study?
Cameron: In a nutshell, we’re talking about workers trying to resist when customers have the control — the digital boss. Typically, when you think of a customer service encounter you’re going up to a cashier, paying for your item, and then you’re gone. But one of the things that’s so interesting about gig work is that it extends this service encounter. You have the period of time where you’ve just requested the ride. Then you have the period of time when the person is actually in the car with you when you’re driving. Then you have all this time afterwards, where you’ve been given the rating. Are you going to push back against that rating? As you can see, the length of the customer service encounter gets elongated. There are different parts of the service encounter where there’s more or less control being had by the platform and the customer, and there’s more or less latitude for the worker’s resistance. That’s what this study looks at. It looks at the elongation of the service encounter, and when workers have more or less latitude or ability to resist.
Knowledge@Wharton: This study brings something new to the table: It highlights how gig workers manipulate that algorithm so that they can get back some control from those customers. Can you talk about how they do that, and what is the lesson for companies like Uber and Lyft and other platforms that rely on digital bosses?
Cameron: What they can do to resist depends on what part of the now elongated the service encounter I was telling you about. If it’s at the very beginning, before they’ve even picked up the customer or they started the project, they can vet the customer. “Do you seem like you’re going to be difficult? I’m going to decide not to take your ride. I’m not going to take the assignment. You might mess up my rating.”
While we’re in the middle of the work, where the customer actually has the most control, the gig worker has to use more finesse to resist and not cause trouble. There might be times where they’ll just be like, “Oh, I need to end the assignment early,” if they feel like the customer is being difficult. That way, they can preserve their rating. Sometimes they’ll even pay back the money to make sure the customer is happy, so they avoid getting a low rating. But there are actually times where they’ll try to trick you or dupe the customer. Or they’ll hold the work hostage and won’t give you the work until you promise a good rating. So, the middle is where workers have less latitude.
“Workers are always going to find ways to have resistance, to have agency.”
And when the work is complete, there is much less latitude for resistance. One is called mediated retaliation: “I’m going to give you a low rating if I think you’re going to give me a low rating.” Regardless of how the customer behaved during the work. Or they can file what we call Hail Mary disputes. The purpose of these disputes is more symbolic. They’re just more to let the platform know, “Hey, there was a problem,” not that they think it will actually fix anything with the customer, but it will be less likely they’ll be kicked off the platform because they alerted them that there was a problem with the customer.
Knowledge@Wharton: Do you think the platforms are aware of these tactics? What does it mean for them?
Cameron: While technically the platforms could be aware of these, we don’t actually think they are actively monitoring these discussions. There are all these back-and-forth dialogues [between workers and customers] to ensure a good rating. Such as a worker asking “Can you promise a good rating?” before starting a project. Or asking a customer: “Can you break up this project into small projects?” so they can inflate their ratings. We don’t think the companies are monitoring this behavior in the chat, and they’re sort of letting it go.
Workers are always going to find ways to have resistance, to have agency. And the more workers understand how work is being managed by the algorithm and by the customers, the more they can understand how to maneuver around these digital bosses.
Knowledge@Wharton: As you point out in the paper, they don’t even have a manager that they can complain to or get support from.
Cameron: Exactly. So, that ties into some of the implications I see for workers. I think what the implications are for companies that rely on customers as the new boss is that there are limitations because the customer doesn’t care as much [as a typical human manager would]. They just want to get the service encounter done with. They want the ride done, the coding done. That’s opposed to a manager, where you develop this deeper human relationship.
If a company has really strict deactivation policies — like being kicked off if your rating falls below [a certain level] and there’s no way to get back into the company’s good graces — it might kick out people who are actually really good workers because they just had a customer who doesn’t care and who gives a low rating just because they’re going to give a low rating regardless.
I think what the study shows is the flexibility that these companies need to have around using these [customer] evaluations as the final arbiter of performance. I just got back from Africa, where I interviewed people who are on these platforms, and I was pleased to see that the algorithmic control isn’t as tight there as we see in the U.S. because there are more cultural contexts that need to be taken into account. The customer is really just not putting a lot of thought into their rating.
“What the study shows is the flexibility that these companies need to have around using these [customer] evaluations as the final arbiter of performance.”
Knowledge@Wharton: Is there an implicit takeaway in your study for consumers?
Cameron: To be honest, I just don’t know if consumers actually have enough of the political will or the impetus to do anything.
Knowledge@Wharton: The customer is still king.
Cameron: Right, and particularly when there are so many other alternative options out there. They think, “I’m not getting what I want. Let me just go to this different platform.” The best I can say is to know that ratings matter. But even more than the ratings, it’s showing care and concern in these one-off interactions. For us as consumers, [the concern is] did we get our groceries delivered? But for this other person, they’re fighting traffic. They’re waiting in line. Remember what it was like at the beginning of the pandemic, and there were hour-long lines to get into grocery stores? All of the invisible work just to make a dollar. How can we as consumers make the transaction as simple and as straightforward for the worker as possible? Like don’t add five things to your Instacart cart once the shopper starts. And tip well. I’m glad that many of these platforms now have automated tipping because those tips matter a lot in boosting up the wages to a more livable amount.
Knowledge@Wharton: The very last statement in your paper is: “As the nature of technology and work changes, we anticipate the relationship between control and resistance will continue to evolve in ways that will require innovative data collection and theory-building opportunities.” That feels foreboding. What do you see as you continue to study this topic?
Cameron: One of the reasons I find this gig economy work so fascinating is just that it’s a major disruptor. It’s a disruption in different industries. It’s a disruptor in how we as consumers live our lives, and that’s why it’s going to continue to change and evolve.
I’m now looking at overt resistance, particularly the Decline Now movement in DoorDash, where you find workers voicing complaints on these forums and then taking action saying, “No, we’re going to decline these food delivery rides until we can try to drive up the prices so we can earn a higher wage.”
This study is about control, and resistance is very much part of the discipline of labor studies. But I’m also interested in how the workers find choice and agency in this work, and not just schedule flexibility. I do think there’s something fundamentally different about how this work is structured. That was one of the reasons why I was recently in West Africa collecting data — to see how this work is experienced differently, even though the [user interface] platform is mostly the same across all these countries.
