By Author:
Lindsey D. Cameron – Wharton School, University of Pennsylvania
Interviewers:
Xian Zhu – McGill University
Krishna Dermawan – University of Queensland
Article link: https://doi.org/10.1177/00018392241236163
- What brought you to this intersection between algorithms at work and gig work? How did this paper start and develop?
Taking inspiration from Foucault, I start with my genealogy to understand how I position myself as a management scholar and what led me to study the gig economy. My mom lost her job in middle management during the height of the Great Recession and, sadly, found out that age discrimination is real. I watched her do a lot of what we call the O.G. or “old school” gig work. She sold samples in grocery stores, opened her own small business selling purses at the World Trade Center, and did the overnight shift at a warehouse. She was truly trying to maintain her middle-class status even though, at times, she earned less than my graduate school stipend. Watching her perseverance led to my research interest of downward social mobility and how people try to stop it. There is a fetishization in our field and American society at large about upward social mobility, leadership, and “higher-skilled” work. You probably saw Damon Phillips’ [Research in Organizational Behavior] article which discusses how, even in our field, we privilege the study of “elite” entrepreneurship even though that’s not the type of entrepreneurship that most people participate in. What most people call entrepreneurship is include starting a mom and pop business, selling items for Mary Kay or Amazon, or becoming an Uber driver. Now, we can go back and forth about if we want to label those activities entrepreneurship, self-employment, or gig work, but often what these people are doing, especially when under financial duress, is trying to stop downward social mobility. That’s exactly what my mom was doing.
During my third year in the doctoral program, I shifted from my prior research topics—two different streams on mindfulness and ethnic entrepreneurship (both were related to my time living in the Middle East, but for completely different reasons)—to study downward social mobility. I was reading works by Katherine Newman and Kathy Edin, and doing pilot interviews with people who had been recently laid off when my advisor – who I often jokingly call “The Great Jerry Davis” – made a seemingly small suggestion that shaped the next ten years of my life: “Why don’t you bound this topic by an organization?” Looking through my data, I saw many people who were recently laid off or encountering financial hardship were working at gig companies like Uber, DoorDash, and TaskRabbit. More people were ride-hailing drivers for Uber and Lyft, so that’s where I started. From there, I moved from a purely sociological area of study to one that also had a strong organizational component. The eventual focus on algorithmic management came later as I analyzed the data, though I would say that my interest in technology is long-standing. I studied engineering and computer science as an undergrad at Harvard and in my prior career, I was a technical analyst at the National Security Agency.
Oh, and it ended all well for my mom. After about five years, she was hired in a more traditional job. She’s now happily retired and spoiling her six grandkids.
To your question of the origin of this specific article: My dissertation committee had originally suggested a “book” dissertation with one main empirical chapter, but I decided to push myself and write two empirical chapters. The first empirical chapter of my dissertation eventually became my 2022 Organization Science. By far, this was the easiest paper to write. The paper started with the basic idea that some people like ride-hailing work [while] some people don’t and I found a theoretical frame, Burawoy’s workplace games, to help makes sense of it. I was a visiting student at the University of Pennsylvania and I remember my garret apartment in West Philadelphia being filled with color-coded sticky notes on the floor and walls and writing endless stream-of-consciousness memos. As I was coding and getting to know –eat, sleep, and breathe – my data, I realized there was more to the story. Drivers would say things like the “Uber made me do it” which, at face value, doesn’t make sense. “The Uber” doesn’t make anyone do anything. What I eventually came to understand was that “The Uber” was the algorithmic management system that underpinned the digital platform. And that the algorithmic management system was a series of interlocking levers. This insight became the first section of my findings in this article. Piecing together this story took many years and reminds me of the story where ten blind people are touching parts of an elephant and each describing something different. I wanted to understand the whole elephant!
A wise person once told me that you never get as much time as when you are PhD student, and you never get as much feedback as when you are on the job market, so I soaked it up. Even though the idea for the second dissertation paper was less developed that the first one, I thought it had the potential for more impact, so I took a risk and made it my job market paper to get more feedback. I remember walking around AOM and other conferences with a handout of the model for feedback and even asked Wanda Orlikowski for advice at a bus stop at Stanford. (Thanks, Wanda!) And here we are eight years later.
- Did you collect all of the data during your PhD?
Not all; in total there is more than seven years of data collection in that one article! My data collection strategy was based on advice from my advisors: collect enough data that you can “eat” on it for your first three years as faculty and you don’t need to run out and collect more data. I half-way listened to that advice since I’ve never really stopped collecting data, but what I collected my last three years in the PhD program is the core of my data that I draw on for this article. The other data I collected after my dissertation expanded to look at boundary conditions and flesh out the impact of major events (e.g., COVID). I’ve looked at other industries within the gig economy (e.g., food delivery) and studied ride-hailing in other countries. So, I’ve never really stopped thinking about the gig economy.
- Are there other things you learnt during the PhD that was useful in writing this paper?
Umm, everything? Michigan really is wonderful. Full stop. I can think of no better place to be a student and develop as a scholar. Full stop. We had these mini-classes – I think they were called Research 2.0 – in which faculty shared their “super-powers” with us. Jane Dutton’s was developing relationships. Jeffrey Sanchez-Burks’ and Leigh Tost’s was experiments. Bob Quinn and Jim Walsh helped us think through our research identity. David Mayer’s taught us how to deconstruct a paper for mechanisms; a technique I still use every time I read a paper. Sue Ashford taught us, essentially, how to theorize by asking ourselves thought-provoking questions. Her notecards hang on my office wall. Ross was such a great learning incubator, along with all the other courses I took in Michigan sociology, psychology, and anthropology.
And I learned some skills in unexpected places, such as how to manage realtionships with companies from Gretchen Spreitzer, create a really good PowerPoint deck from Tawanna Dillahunt and Alex Liu in the Information School, different genres of literature reviews from Christine Feak in the English Language Institute, creative non-fiction writing from Ruth Behar in Anthropology, and different ways to brainstorm from the professional development coach Melanie Sobocinski. (Yes, Michigan offers coaching to students!) And everyone needs a good non-academic hobby and mine was gymnastics/acrobatics. I am thankful for Hartland Gymnastics opening their doors to a non-traditional athlete and helping me get new tumbling skills! 🤸🏽 🤸🏽 🤸🏽
More specifically regarding skills, I started writing groups every year from my second year onwards. Every year, I’d start a new writing group with a different theme. One year it was folks at Michigan, another year all qualitative researchers on any topic, another year was “Micro meets Macro” and included students at other schools (I was the most micro!), and another year it was all students studying technology. These groups helped me brainstorm ideas, get feedback early, and iterate, which meant I was better prepared when I met with my advisors. Overall, my writing groups helped me develop my network of potential collaborators, methods experts, and even friends.
My committee and, in particular, Jane Dutton and Beth Bechky, emphasized slow science. High-quality research isn’t produced in a rush and it’s okay to spend a significant amount of time to understand a phenomenon deeply. Everyone on my committee (and Michigan, tbh) poured into me, encouraging me that I was onto a “big idea.” That’s not to say I didn’t feel pressure to publish in the PhD program, but more that I never felt it was “bad” that it took so long – about eighteen months – to analyze my data. The idea that big ideas take time was normalized.
So, why did it take so long? I had a fair amount of personal upheaval with all the moving and whatnot, but qualitative analysis takes time. I coded data by hand and had colored index cards all around my apartment. I would print out transcripts, circle important quotes, and tack them up on the walls. I would listen to my interviews and fieldnotes, and record voice memos to myself. My data is in me: if you give me any quote, I can exactly tell you who said it, the context, and where, both in the transcript and where I was at in my life. Because I knew my data so well, I could more easily try on different theories to fit my data. And by understanding the data and my phenomenon so well, I was able to easily make connections to theories in other projects.
While your question was about what I learned during my PhD program I’d like to emphasize how much I keep learning! Someone once told me that a PhD just teaches you how not to get desk rejected and the rest of your career is spent learning how to respond to revisions. I typically read one to three books a week. I like to stay in the mix: I stay in interdisciplinary conversations, listen to a lot of podcasts, watch documentaries, and go to conferences. Lawyers, restaurant managers, union organizers, your mother – they all have an opinion of the gig economy. I listen to them all so I can understand as many perspectives of the gig economy as possible, especially the ones I don’t agree with.
- When did this theoretical framing came out? What’s the Aha moment that you see these different tactics can lead to the workplace consent, which gives you the punchline of this article?
Throughout the dissertation, I was obsessed with the interplay of autonomy and control, you can see in the dissertation’s title: The Rise of Algorithmic Work: Implications for Managerial Control and Worker Autonomy. In my first round R&R (revise and resubmit), I was told that the bar to contribute to autonomy and control was high and I needed to rethink my framing, i.e., bring it down a level of analysis. This is where my friends from writing groups especially came in to help: Nick Occhiuto was generous enough to hop on a call with me. And when I told him what my core finding was—that algorithmic management enables a small amount of autonomy—he said, “I think you’re talking about consent,” and directed me to a specific page of Burawoy’s Manufacturing Consent. While I had read the book multiple times, I now read it with a different eye. It took me three years to really understand that paragraph, the book, and the concept of consent. I remain deeply influenced by Burawoy and I’ve written a series of articles grappling with different themes in his work. He was a brilliant scholar who left us all too soon.
Much of the fleshing out of the tactics I describe in this article came through giving talks. For instance, the difference between simple and complex tactics came from my talk at UC Santa Barbara, specifically the helpful feedback of Steve Barley, Matt Beane, Paul Leonardi, and others. The section on withdrawing consent came from a talk at MIT and comments from Ezra Zuckerman, Emilio Castillo, and Wanda Orlikowski. I can point to other footnotes that came from conversations with Peter Cappelli, Beth Bechky, Seth Carnahan, Mauro Guillen, Katherine Klein, and Kate Kellogg. And that’s just a few of the folks who helped shaped my ideas. All this to say that knowledge is socially constructed, and I am forever grateful to all the individuals who helped me develop as a scholar.
- In terms of the inputs with the ASQ editors, how did you navigate between the feedback from workshopping your research in all of these different forums, with the reviewers’ feedback?
I’ve had amazing review experiences at several management journals, and what happened to ASQ was unparalleled. The ASQ reviewers could have been the same person because their feedback was so similar! They truly understood what I was trying to say and helped me to articulate it. Let me give you an example: In the first round, they told me, “You’re using the word ‘autonomy’ sometimes from a psychological perspective and other times from a sociological perspective. You can’t have both ways in the paper, so pick one. We think it is sociology.” They didn’t tell me what to do, but how to think, and for that I am deeply grateful. Importantly, I did not feel like I had to check boxes or like there was a way my research ‘had to be.’ (I was especially glad the reviewers did not tell me about their last Uber ride and how it was alike/different from my findings – a particular sore spot of anyone studying a popular topic!) And Forrest was a magnificent editor. Of course, he found the right reviewers for the paper, but he also handled the manuscript thoughtfully in each revision. In one review, he said something along the lines of “I will not be the fourth reviewer. The reviewers are consistent. Please respond.” I realize this is not typically what an editor does, but in this case it felt appropriate because the reviewers were so synchronized. It takes a considerable amount of self-knowledge and skill as an editor to know if, when, and how to offer comments. In another review, he directed me to legal research on consent, which I hadn’t thought about before, and ultimately helped me flush out the discussion. I have the utmost respect for Forrest, the reviewers, and ASQ as a journal. And I am happy the paper has been received so well in the world, both in terms of academic citations and being drawn upon in several legal cases.
- We think one of the reasons why the paper was so impactful was that you jumped into the world of gig-work yourself by becoming a driver in the ride-hailing industry, and then there is a lot of other data collected. How did you manage the collection and systematic analysis of all this data? Are there any fun stories, surprising, or anecdotes from your experience as a driver?
I collected data through what I call “punctuated immersion.” I would spend two weeks in the field – driving myself, interviewing other drivers, and scraping online forums — and then spent about another ten days analyzing the data, followed by three to four days “off” working on other projects. I must have done this over a period of more than a year. But let me say: collecting data was easy, it was the analysis that was hard. For that, I leaned into my writing groups, presenting at many conferences, and consulting my committee members.
Fun experiences driving? Ha! I seriously dislike driving in general – I don’t even own a car anymore – which makes my choice of a research site ironic. Honestly, what I remember most is the frustration, emotional labor—or worse being ignored—and, at times, feeling scared.
There are definitely some interesting — if not fun — stories. I lost my very first ride. My phone overheated and it fell out of my hand, because it buzzes a lot when you receive a ride request, and then it went underneath the car seat. It was August in DC, which was like 105 degrees, and I didn’t know that I was supposed to buy a mount to hold my phone in front of the AC vent to keep it cool, so I was holding it in my lap. Not smart. Later in that day, my phone overheated again while I was driving someplace, so I had to rely on my own memory to figure out how to take that person where they needed to be. I’m not sure if I got paid properly for that ride. I remember taking a woman to her son’s chemotherapy treatment and comforting her on the way. Someone once howled at the moon. (Luckily, I don’t believe in werewolves.) I remember having three very large men in the car, one right next to me, as I’m driving against a setting sun into the countryside—I was really scared. Another time, I remember hearing gunshots and seeing blue [police] lights flashing and thinking, “Oh **** what am I going to do?! Am I going to go and cancel the ride? Am I going to stay and do this pick-up?” Fortunately, my prior government training kicked in and I “Got off the ‘X.’ (i.e., hot spot).” Most vividly, though, I remember the monotony. Every day, if you went out at the same time, you’d be going to the exact same places. North Capitol Street to downtown DC in the mornings, Georgetown or Adams Morgan on weekends, and so on. I felt like I was trapped in Pleasantville — except it wasn’t so pleasant.
I will say, too, that while I dislike driving, in general, many drivers didn’t and had a completely different experience of the work. I was awed when they shared stories with me about going to concerts with customers or helping them with their romantic problems. I go into this in-depth in my Organization Science article.
- In the paper’s Discussion section you paraphrased Jerry Davis, “Our grandparents had careers, our parents had jobs, and today we have tasks.” What are our concerns about gig work, based on your rich observation, for the workers and for the societies?
There are so, so many concerns. In 2023, the World Economic Forum declared we are in a polycrisis, where rapidly converging environmental, technology and social changes at the global level interact with one another to form a cluster of compounding risks. The rise of gig work and the subsequent concentrated power in digital platform companies is just one factor in today’s polycrisis. In a paper with Moira Weigel, we describe a series of promises that platforms make to individuals, communities, and nation states that they simply cannot keep and cannot be held accountable for. Given how these platforms often attract and recruit the most marginalized portions of the labor force – in a process Tressie McMillan Cottom calls exclusion by inclusion – this deeply concerns me.
While the taskification of work, both in the gig economy and in traditional companies, is on the rise, I also want to mention that my colleague, Matthew Bidwell, has a forthcoming book called “Insider Advantage” that points out that rates of employment tenure haven’t really changed since the year 2000, nor have the rates of independent contracting. There’s still a lot of regular employment around.
- Way towards just general, academic advice. You had a very, very impressive career before you did your PhD. You worked in the diplomatic community, and I think it shows why you’ve been so successful in terms of channelling your research towards policy outcomes. Our question is, what do you think academic research gives you that, being in the field, in the diplomatic community, or people even in businesses, may not give you? What do you think is the niche key contribution for academics really to impact the world?
The world of business moves fast – but the government and academia, not so much. But, as I mentioned earlier when describing my research process, there is real value in being slow and deliberate. Going slow means you can be fast and precise when needed. In many of the legal cases I’ve worked on, there has been tremendous time pressure to analyze large volumes of data (I’m talking multiple file cabinets of paper) and produce a report. While demanding, the years I’ve devoted to understanding the gig economy helps me parse the data, evaluate multiple perspectives, and build an argument easily. I enjoy becoming laser focused on a complex topic, like a surgeon with a scalpel. It feels like intellectual jiujutsu. You’ve probably seen something like this in a really good seminar talk when the speaker and the audience can volley with one another in lighting speed. It takes real depth and mastery in one’s research area to do that and it’s one of the things I love about academia. So my advice: Go slow, so you can be fast when needed.
Interviewer bios:
Xian Zhu is a PhD candidate at Desautels Faculty of Management, McGill University. His research explores cross-disciplinary collaboration in extreme contexts. He has conducted qualitative fieldwork in hospital emergency departments, trauma centers, and emergency call centers.
Krishna Dermawan is a PhD student at the University of Queensland Business School.