Abstract
Platform work is one prominent type of independent contracting in the United States. Yet the independent status of platform workers is contested. Some scholars call platform workers dependent contractors. Others are measuring workers’ economic dependence to support better classification. Given platform workers’ heterogeneity, current efforts to classify workers’ dependence might be missing different kinds of dependencies. This article asks the following: What are the dimensions of economic dependence that platform workers experience? I interviewed 47 individuals working on the microwork platform Prolific and analyzed three dimensions that were salient in workers’ accounts: “episodic,” “discretionary,” and “projected” dependencies. These dimensions can help us to measure platform dependence. Furthermore, this article theorizes how each form of dependence might reinforce economic precarity. This study calls for further connections between platform studies and literature on household finance, consumption, and culture.
Independent contracting has increased in the United States over the last several decades, reflecting a broader shift toward a new set of employment relations (Abraham and Houseman 2019; Kalleberg 2009; Katz and Krueger 2019). Independent contractors include self-employed workers in jobs ranging from cleaners to caretakers, tradespersons, skilled technicians, and financial consultants. The platform economy is one growing site for independent contracting jobs (Dubal 2017a; Spreitzer et al. 2017:485; Vallas and Schor 2020). In 2021, 16 percent of Americans reported having done platform-mediated work, with 9 percent having done so in the last year (Anderson et al. 2021). Among both platform companies and platform workers, independence, flexibility, and being your own boss are widely expressed as job benefits (Anderson et al. 2021; Lehdonvirta 2018; Smith 2016). Yet platform employers have shifted many risks—financial, safety, and others—on to their workers (Hacker 2008; Kalleberg 2009; Vallas and Schor 2020).
As employers become more dependent on independent contracting, the classification of what is or is not independent work has been increasingly subject to scrutiny and contestation. Misclassification occurs when both the employer and the worker are highly dependent on each other but the employer withholds employee-related benefits (Dubal 2017b). In the platform economy, there have been prominent legal challenges to the classification of platform workers as independent contractors (Cherry and Aloisi 2016; Dubal 2020). The mixed outcomes of these legal challenges illustrate the contested meanings of dependence, independence, and interdependence. These classification issues have real consequences for millions of independent workers and their communities and the economy as a whole.
Scholarly research has already contributed to a better understanding of worker classification in the platform economy. In particular, platform studies have focused on defining and measuring concepts such as control and precarity (for a review, see Vallas and Schor 2020). However, platform studies have avoided directly theorizing dependence. The exception to this lacuna is the work of Schor et al. (2020). They argue that platform dependence can be measured for individual workers, who can then be classified based on their degree of dependence on this source of income.
Furthermore, Schor et al. (2020) argue for a co-constitutive relationship between dependence and precarity. For workers who are the most dependent, platform work—as it is currently structured, like independent contracting—generally reinforces precarity and dependence. On the other hand, “independent” workers earning supplemental income pull benefits from the system, including better wages and higher satisfaction. The unequal distribution of rewards toward the more independent group is theorized as a source of class inequality (Schor 2017).
Schor’s dependence framework is highly cited in the field of platform studies. Yet few scholars have attempted to extend and critique her notion of dependence (cf. Maffie 2022). This article takes Schor’s definition as a starting point, and asks the following: What dimensions of dependence are salient among a heterogeneous group of platform workers? I use interviews with 47 individuals who work on the platform Prolific (for details, see case background) to explore a variety of platform dependencies.
I identify and describe three dimensions—or types—of dependence that broaden and complicate Schor’s core definition. First, “episodic dependence” underlines the role of episodes of precarity (i.e., instability, volatility) in shaping platform dependence. Second, “discretionary dependence” underlines the importance of consumption for the purposes of maintaining and performing: status, identity, and social relations. Third, “projected dependence” emerges from the future-oriented narratives that motivate platform workers even when money is not being spent right away on necessities (Beckert 2013). In addition to describing these three dimensions, I explore how they interact with precarity; this connection between precarity and dependence contributes new evidence to Schor’s theory about the reproduction of precarity on platforms (Schor et al. 2020). We would benefit from future work that extends and refines the dependence framework and work that tests related concepts empirically on a range of platforms (see Glavin and Schieman 2022).
What Do We Know about Precarity and Dependence on Platforms?
Precarity is one of the dominant perspectives in platform studies (Vallas and Schor 2020). Scholarship in this field argues that workers experience precarious working conditions and financial outcomes through a number of mechanisms. Platform workers can be laid off or fired by their employer without notice or explanation, including through automated and opaque algorithmic systems (Gray and Suri 2019; Griesbach et al. 2019; Rosenblat 2018). Platform workers also experience income insecurity, or the lack of an adequate, stable income while working on the platform (Griesbach et al. 2019; Rosenblat 2018; Standing 1997). Many studies of major platforms, including Uber and Amazon Mechanical Turk, conclude that platform pay is not tied to minimum wage laws and is not amenable to fairness (Dubal 2017a; Hara et al. 2018). Integrating these different facets of precarity, scholars have underlined the multidimensional and nested insecurities that workers face (Duffy et al. 2021). According to the precarity perspective, workers are experiencing insecurity in work conditions that hinge in large part on their legal status as independent contractors; this implies that moving platform workers toward employee status would ensure more protections and fairer conditions (Dubal 2020).
In contrast to this first perspective, other studies highlight that platform jobs can act as protection against precarity, as a type of private safety net. Although this perspective recognizes the precarity inherent in platform jobs, it focuses on the broader conditions of precarity in the labor market (Barrios, Hochberg, and Yi 2022). First, platform work can be critical to smooth income following a job loss or period of unemployment (Fos et al. 2019; Huang et al. 2020). During the COVID-19 pandemic, platform work was essential for certain workers experiencing hardship (Ravenelle 2023). Furthermore, a 2016 study using data from over 2 million households with bank accounts at JP Morgan Chase concluded that “[i]ndividuals can, and do, generate additional income on labor platforms in a timely fashion when they experience a dip in regular earnings” (Farrell and Greig 2016:29). These studies suggest that platform jobs, although precarious, can also become a solution to financial insecurity faced by certain households.
The work of Schor et al. (2020; Schor 2021) has been critical in synthesizing these two perspectives: platforms as a source of and solution to precarity. In analyzing interview data with 112 workers from several platforms, Schor introduced the concept of dependence to capture the heterogeneous experiences in her sample. First, dependence is operationalized as the relative percentage of overall income earned on platforms. Second, dependence is qualitatively assessed by whether workers report spending their platform income on basic expenses (i.e., essentials).
Schor outlined three categories of worker dependence on platform income. One in four individuals in the sample was classified as “dependent” because they made most or all of their income on platforms and spent that money on essentials. This category included drivers who earned their income solely from Uber and other ride-hailing applications. Second, “supplemental earners” were 40 percent of the sample; these individuals had other jobs and did not rely on platform income for basic expenses (Schor et al. 2020:835). Instead, supplemental earners were described as using their income on discretionary spending, such as vacations, or saving the money. Finally, one in three workers was “partially dependent,” working part-time on platforms and having another major source of income but using some or all their earnings for essentials.
Critically, Schor theorized that workers’ category of dependence correlated with their experiences of precarity on the same platform. Because supplemental earners tended to have higher incomes, and therefore relied less on platform income, they had more autonomy to choose higher paying jobs, and they earned more per hour than dependent workers. The experiences of independent workers were positive. In contrast, the dependent category experienced precarity on the platform, including lower wages, higher control, and lower satisfaction. This dynamic was akin to a cycle; precarity off the platform—due to lack of alternatives or insufficient income—reinforced precarious conditions on the platform.
Overall, Schor’s large, multiplatform interview study induced three categories of economic dependence and offered a preliminary definition that could be operationalized, measured, and tested. However, I hypothesize that Schor’s definition is limited and therefore misses several subtypes of dependence. In particular, a close reading of Schor’s account suggests that supplemental earners’ experiences are varied and that their motivations are connected to precarity. For example, Bobby supplemented his full-time job as a teacher by driving Uber for 6 to 10 hours each week “depending on where [he was] at financially” (Schor et al. 2020:846). Although a supplemental earner, he reported “running out of money” on a Friday and using Uber to earn enough money for activities taking place that very weekend. When someone is on the verge of illiquidity, platform income is not only about buying essentials, it is essential; this lack of savings is a kind of precarity that needs to be considered as part of the picture. Another of Schor’s respondents, Lenny, was labeled as a supplemental earner; as a graduate student in social work, he described his $750 in monthly income from TaskRabbit as being put toward savings, as a kind of “safety net” (Schor et al. 2020:845). Saving for a safety net indicates that Lenny might need the money for basic needs, although not immediately. From these two examples, we see that there is an opportunity to build conceptually on Schor and her collaborators’ contribution to the classification of dependence. This article asks, What are the dimensions of economic dependence that platform workers experience? I use interviews with workers on one microwork platform, Prolific, to explore this theoretically motivated question.
Case Background
Labor platforms are organizations that mediate relationships between clients and contractors, allowing clients to find and hire someone to work on tasks or projects and enforcing contracts and negotiating resolutions when issues arise (Kenney and Zysman 2019). For example, Uber matches riders with drivers, and in exchange for these services, it takes a commission. Platform jobs are typically a hybrid of contract work and independent contracting (Barley, Bechky, and Milliken 2017).
Platforms come in many forms. Uber and other geographically tethered platforms facilitate work in real life. Such in-person services are different from remote work, where digitization (e.g., digital tools such as Zoom, Google Docs, and platform interfaces) allows work to be completed away from a standard workplace—including at home (Vallas and Schor 2020). For example, freelancing sites allow workers from all over the world to sell their skills, such as design, copywriting, or computer programming.
One growing form of remote platform labor is microwork (also called crowdwork), which minimizes the skills needed to complete tasks by making them small and routinized—like cleaning and annotating data (Woodcock and Graham 2019). Microwork platforms are being used by those developing machine learning algorithms; human feedback improves the algorithm’s outputs (Casilli 2019). Importantly, micro workers in the United States and Europe are typically classified as independent contractors (Berg et al. 2018; Hara et al. 2018).
Previous work on microwork sites, notably Amazon Mechanical Turk, has noted the diverse range of motivations for doing this kind of work (Gray and Suri 2019; Irani 2015). In addition to offering spatial flexibility, microwork sites tend to have high temporal flexibility (Woodcock and Graham 2019). As Vallas and Schor (2020:283) note, platforms with lower barriers to entry, like microwork, “foster greater variation in the work orientations, labor market positions, and sociodemographic composition of their workers.” Microwork platforms are a “strategic research site” from which to observe a range of platform dependencies (Merton 1987:1).
I investigate the dynamics of dependence using one microwork platform, Prolific. Prolific originated in 2015 when two graduate students in the United Kingdom who had a hard time finding participants for their research solved the problem by building a digital platform that connected researchers from academia with research participants. More recently, Prolific has updated its offerings to attract organizations building machine learning algorithms, with clients including companies like Google and Meta. This article is one of the first to explore the dynamics of work on Prolific (Muldoon and Apostolidis 2023).
When people looking for work sign into their Prolific accounts, they see a list of possible tasks—such as taking a survey or participating in an interview—that they are already eligible to complete based on the demographic information that they give to the platform (including age, gender, race, and income level). This is closer to an open labor market because it allows for the worker’s discretion in selecting jobs instead of relying entirely on automatic matching by an algorithm (Cameron and Rahman 2021:41). Before starting a job, people receive information about the approximate time to complete the task and the pay.
I estimate that my interviewees earned on average $100 a month, with some earning as little as $20 a month and others earning upward of $300 a month. There also appeared to be an upper limit on how much workers could earn from Prolific, which I estimate to be $600 a month. Therefore, workers on Prolific could earn enough to qualify as dependent but were more heavily concentrated in the partially dependent and supplemental categories (Schor et al. 2020). 1 Prolific is different from a platform like Uber, which has a higher concentration of highly dependent earners; each labor platform has its own unique mix of dependence categories. My findings speak to conditions on Prolific, but they are meant to be useful for theoretically motivated comparisons of dependence on other platforms.
Methods
Starting in August 2022, I posted a work task directly on Prolific that introduced my project. Interviewees were compensated $15 for approximately an hour of their time, and their names and certain details have been changed to ensure anonymity. My interview guide asked each respondent about how they came to work on Prolific, the experiences of doing work on Prolific on a daily basis and over time, and the relationship between Prolific and the individual’s other work and obligations. Interviews enabled me to gather accounts from platform workers including the vocabularies of motive that platform workers used to make sense of their work and the meaning they derive from it (Mills 1940; Scott and Lyman 1968).
My approach to sampling through Prolific was geared toward achieving variation on variables such as age, gender, race, employment status, and income (Lofland et al. 2006). Overall, my sample mirrors many of the demographic characteristics found in the population of Prolific. As Table 1 shows, my sample of 47 interviewees skews older. My sample skews more female, which matches the pattern on the platform. My sample is also slightly more educated and whiter than the population of the platform.
Demographics for Study Interviewees (N = 47) and Active American Participants on Prolific.
Source. Statistics for Prolific were created using data from Prolific’s website, the “find your participants” tool on May 23 and May 24, 2023. Note that not all participants opt to fill in demographic characteristics. Some percentages have been rounded and will not add up to exactly 100%.
I infer that my sample might differ from the population in other ways. Almost all of my interviewees reported earning at least $20 a month on the platform. This is significantly higher than the average user of Prolific given that most micro workers make close to nothing on these websites (Berg et al. 2018; Muldoon and Apostolidis 2023; Piasna, Zwysen, and Drahokoupil 2022). Overall, it is likely that I was speaking to the 20 percent of Prolific workers who were doing a majority of the work on Prolific (Gray and Suri 2019).
Analyses for this article relied primarily on a set of questions that asked interviewees to describe how they thought about the money that they earned from Prolific. This included specific instances of income that had been used over time and plans for how it might be used in the future. My coding of the data focused on identifying the different practices that my interviewees described. I started to classify and compare individual cases and write analytic memos exploring types of dependence. Overall, my findings underline similarities and differences across cases (Small 2009).
Dimensions of Dependence: Episodic, Discretionary and Projected
“Three Days Straight”: Financial Dynamics of Precarity and Episodic Dependence
Many of my interviewees experienced moments of intense dependence on Prolific income. These episodes of precarity focus our attention on a form of dependence that we might otherwise miss using Schor’s analysis. When I met Fallon over a video call one evening, she was in the midst of an episode of dependence. She sat on the floor surrounded by moving boxes. A couple of months before, she had landed a new job at a design consultancy. Fallon’s paycheck from her new, salaried job was coming in eight days. But moving costs had wiped out her savings. She was running low on food and gas for her car and had a $65 internet bill to pay. To avoid getting her internet turned off, she was planning on working on Prolific—filling out as many surveys as she could regardless of the hourly pay. This was not the first time that Fallon was in a position like this. About four years before, when she was still working in food and beverage, a trusted friend told Fallon about Prolific. That same weekend she worked “three days straight” and made $150, enough to keep her utilities from being shut off.
Fallon did not feel like she had other options except working on Prolific. She had a credit card at one point in her early 20s, but she had not paid off her debt. Now her credit score was low. Furthermore, Fallon told me that it did not feel good to be in debt to friends. “I hate asking for money. It’s a point of pride.” Overall, she explained “I’ve always made things work” by working multiple jobs and extra hours.
Sam was 28 years old when I spoke to her, and she echoed many elements of Fallon’s narrative. Sam lived with her two children and boyfriend and worked in customer support from home for two companies, both part-time. During the holiday season, her hours had been cut unexpectedly when business slowed down. Her son needed diapers. Her rent was due. She needed to buy gifts for her family and friends, too. As a result, Sam turned to Prolific: I was checking like, all day, I was just checking throughout the day, because you know, times have been hard. If I’m like “I really need to get this,” I’ll be really dedicated to doing Prolific. Prolific has really helped me out in those times where I was struggling financially.
She described making $500 in one month. But it came at a cost—of constantly checking the website and doing many jobs that were paying less than she would have normally agreed to. When I interviewed her, Sam and her partner were earning enough, so she was saving Prolific income as an emergency fund or for little extras. She was also checking Prolific less frequently.
When I asked Sam what might have happened without Prolific, she explained that she could have asked family or friends for the money. “But I hate it . . . I don’t want to owe people. They can use it against me,” she explained. Sam preferred working—instead of borrowing— because it preserved her independence from her network. Furthermore, Sam described getting into credit card debt at the beginning of the pandemic. “I wish I never had,” she continued, because she was still working to pay off the debt.
Fallon and Sam’s movements in and out of precarity and dependence complicate the application of Schor’s three categories. We can better capture these financial dynamics using concepts that relate to core economic problems in households. The economist Jonathan Morduch, in his review of key concepts in microfinance, offers three financial problems that households experience: insufficiency, instability, and illiquidity (Morduch 2023; see also Collins et al. 2009). These three problems overlap, co-constitute, and in some cases, compound experiences of financial insecurity. And these financial dynamics motivate relations of dependence with employers, family and friends, the government, and credit card companies.
Fallon and Sam both described living in relatively expensive urban areas and working in service jobs and therefore had exposure to insufficiency of income. Fallon, for example, had worked two jobs to earn enough income to get by. Yet even still, episodes of instability—whether rapid changes in income or spikes in consumption—required financial problem-solving. For both Fallon and Sam, these moments of instability led them to start working on platforms and periodically return when instability was high. In certain cases, like Fallon’s cross-country move, a bit of instability and bad timing could lead to illiquidity due to a lack of savings. Because both Fallon and Sam did not feel like they could rely on their network or other financial tools, such as credit cards, platforms became a safety net. Both described high dependence on the platform for short bursts, like a weekend or one month: working longer hours and with lower pay. Yet as their financial situation stabilized, their dependence shifted back toward something like supplemental earning (Schor et al. 2020). By paying attention to platform workers’ experiences over time, we are more likely to observe insufficiency, instability and illiquidity, which might differentially motivate episodes of platform dependence.
“A Little Extra”: Discretionary Dependence and Social Meanings of Money
Schor defines dependence in part by what workers buy with their platform money. For the supplemental earner, platform income is assumed to be extra and is labeled “discretionary.” On the other hand, for those who are dependent and experiencing precarity, platform income is assumed to be spent on basic necessities. My data, however, suggest that both dependent and independent workers could spend some or all of their Prolific money on nonessentials. I argue that people were dependent on this discretion. The meanings of a little extra Prolific income could be highly motivating because it was a means of expressing identity and social status (Sykes et al. 2015; Zelizer 1994). By integrating findings from platform studies with the literature on consumption, we can move toward a theory of platform work that integrates the social meanings of money.
When I spoke with Gary (47 years old), his wife had recently fallen ill with cancer, and he and his son were caring for her at the family’s home. “We don’t crave huge amounts of cash,” he mused, “but we do need money.” Especially with his wife’s medical care, bills were piling up, and they had limited savings for retirement. Gary had transitioned from being a company man to an entrepreneur years ago. He took pride in being his own boss, managing and designing websites for local businesses. But his business income could be volatile, especially during the pandemic. His wife, a copartner in the business, was not currently able to work.
Prolific was interspersed throughout Gary’s days. He described his morning ritual: drinking coffee while checking his work email, reading the news, and watching TV; he could intersperse these activities with a few Prolific surveys. “I do like doing it. But I don’t know how much I would like doing it if it didn’t pay me,” he explained. He described Prolific as “free money,” and when I asked what he meant, he compared the ease and simplicity of Prolific with his web design business, which could involve demanding customers and complex task management. Furthermore, he cared that Prolific had a minimum wage, in contrast with Amazon Mechanical Turk, which he described as “basically just slave labor.” Prolific money piled up slowly, and “One day you turn around and you say ‘Oh my goodness, I got $80 worth of pennies. What the hell do I do with it?’”
Gary described spending Prolific money in various ways. One category was gifts for his family. He had bought his wife a Star Trek flip phone as a silly joke, which she loved. Excited, he bought one for his son and finally one for himself. He planned to delight his wife with a big feather pen, which she could use when she did the family’s accounting in an old-fashioned ledger. He stated openly that this money was going more toward “luxury items that we don’t necessarily need.” If Gary lost income from Prolific, he would not have been able to afford the little extras that brought moments of joy to his family and himself during a difficult period.
Camden (36 years old)—also married and living with his wife and stepdaughter—was upfront with me: that he was using his Prolific earnings to fuel his personal addiction to online poker. He took whatever he made and funneled it into his gambling accounts. “When I need money, I often . . . there’s Prolific. Just for something stupid. Like, I’m broke. If I’m out of money, I’ll just go hit up Prolific and after a while I’ve got 20 bucks.” He claimed to have made over £1,200 since joining Prolific. Camden was one of five of my interviewees who reported using some or all of their Prolific earnings on gambling.
Camden graduated high school and worked a series of highly demanding physical jobs, which took a toll on his body, and he transitioned to working as a customer support agent remotely in the last few years. During the pandemic, he started trying many small platform jobs, including survey sites and games where you get paid to play and watch ads. Camden spent time on Prolific throughout the week. He would check it throughout the day. “If I see something that pays over like a buck or two, and it’s 5 to 10 minutes, I’ll do it on my [work] break.” He would also look to Prolific when he got off work and on his days off.
Camden compared Prolific to other sites that “aren’t very good,” with Prolific being “hands down the best. I mean, it’s the highest paying that I’ve found, and you’re never rejected.” He described going through Prolific trying to maximize his earnings. “I like when I scroll through Prolific to pick [a survey]. I like when I see a good one. It gives me a little bit of a high. Like ‘Oh, it’s 13 bucks. And it only takes like fifteen minutes. That’s cool.’ [chuckles]” But he admitted that some of these high-paying surveys were tedious and boring.
Nancy did platform work in her spare time and spent the money that she earned in ways that she called “free and guilt-free.” She and her husband earned a combined $300,000 a year in their stable, technical careers. Their salaried income was tied up in the budget, including funding their retirement and savings accounts. But the money earned from Prolific and other platforms fell outside of the budget and took on special properties. Nancy explained: “It’s just fun money. It’s nice, it goes in my PayPal account and then if I want to go buy a purse or a new sweater or a pair of shoes? Like I don’t have to feel guilty about it. Do that! Use that money!” Over the years, Nancy had made over $3,000 on Prolific. She had spent the money spontaneously on $300 tickets to a concert with her husband to see one of their favorite rock bands. She also recalled sending $50 over GoFundMe to support a fundraiser that her daughter was involved in.
Based on Schor’s degrees of dependence, we might consider these three individuals—and their households—to be supplemental earners. They worked a little extra on Prolific and spent this extra money on nonessentials like gifts, gambling, and concert tickets. Yet these discretionary choices were deeply meaningful and therefore motivating (Zelizer 1994). The purposes of money were related to expressing their identity, maintaining their social relations, and preserving their social status.
Cultural practices can constitute their own dependencies. In particular, in the American context, individuals are exposed to cultural practices of consumption that encourage more spending and therefore more working (Schor 1998). A needs-centered definition of platform dependence (Schor et al. 2020) conceives precarity as a reinforcing dynamic based mainly on subsistence. By considering Schor’s (1991) seminal work on consumption, notably the “cycle of work-and-spend,” we can elaborate another pattern of platform dependence that integrates cultural practices and social relations.
“My Curiosity Is Just Waiting”: Projected Dependence and Imagined Alternatives to Precarious Futures
Platform jobs have been theorized as a safety net, either for those who are already experiencing financial trouble or for those who are preparing for instability (Ravenelle 2023). Schor chooses to categorize those who are able to save their platform income as independent, supplemental earners. My data complicates this account. Prolific income could be consciously set aside for future needs connected to uncertainty and precarity. Furthermore, Prolific workers who were living in precarious situations—and therefore appeared dependent on platform income—made strategic choices to save, often by reducing spending. Based on these observations, Schor and her coauthors might be missing the breadth of platform dependence among supplemental earners. Three vignettes from my interviews illustrate these dynamics.
Stacy joined Prolific and other platforms after her husband lost his job early in the COVID-19 pandemic. In addition to her full-time job as an administrative assistant, she spent many moments throughout her workdays doing surveys and logging in after hours. For a time, this income was spent on bills. When her husband got a new job, the money started going toward paying off the unpaid debt that the couple had accumulated early in their marriage.
Gradually, Stacy and her husband made the decision to not simply pay off the debts but also set apart about half of their “little side hustles”—including Prolific—for savings. Stacy explained that she and her husband communicated a lot about what the money should be spent on. “If we want it, you know, we think about it, and see if we need it now or want it now. Or if we can save for it, and then get it later, we see it as a reward.” She continued: You don’t get a lot of money from [work on Prolific and other platforms]. And you do put time and effort into it. Fitting it in between a work schedule. So if we were to just be spending it all over, all that time and effort that I put into this . . . that doesn’t give me a lot. It’s like getting something and then just throwing it away. It wouldn’t make sense for me.
Stacy saw Prolific income as hard-earned and therefore more meaningful when it was saved. In particular, Stacy and her husband liked to imagine how their savings might be spent in the longer term, in 3 or 10 years. “My curiosity is just waiting, waiting more and seeing what happens,” Stacy told me with visible excitement. This future-oriented thinking was made more realistic by the relatively stable financial situation that Stacy and her husband were currently experiencing after a period of instability during the pandemic.
Lea was also imagining how she might spend the money that she was able to save from Prolific. She was 54 years old when I interviewed her and received a fixed income from the government for her disability. Living on $1,000 a month was difficult. Saving was only possible through meticulous cuts in her consumption. She described cutting back on food, including subsisting on peanut butter sandwiches: “I’ve got myself eating between $70 and $100 a month worth of food. . . . If I were to buy the food like I was buying before inflation hit, I’d be spending like $600 a month.”
To gain some financial independence, Lea worked on various online businesses that fit her lifestyle and made her additional income. She had recently come across difficulties with selling jewelry and other products on Etsy and shifted her attention to small jobs on platforms, including Prolific. She explained to me matter-of-factly that these platforms were multiple streams of income, “and who doesn’t want more money coming in? [chuckles].” She wanted to earn extra in order to save and invest it: I took like two months educating myself on the stock market before I even got involved. And so now I am diversified. I have money in stocks and bonds and ETFs. I’ve got physical silver, physical gold, I’ve got my real estate [through an app called Fundrise], I’ve got my crypto.
Lea described her project: to hold on to her various investments for retirement. But she also had shorter-term goals, like replacing her car, which was not working. A working car would allow her to get out of her neighborhood, which she was afraid to walk in, and take trips to fish on the coast. Yet despite Lea’s best efforts, these longer-term goals were at risk because of the volatility of living paycheck to paycheck. She described recently cashing out almost half of her Prolific savings to pay out-of-pocket expenses for her dog’s medical emergency.
Evan, like Lea, was investing his Prolific savings, but unlike Lea, he was younger (35 years old) and was able to work full-time in skilled jobs. Saving up money felt important to Evan, who had seen friends and family struggle to live comfortably on Social Security income. He saw his savings as contributing to a retirement that would be a “little more comfortable, and [I’m] not going to be as stressed about a fixed income.”
Evan was exploring how to save money for his future. He described becoming more frugal over time, spending less of his money on “things that aren’t necessary, and finding ways to reduce the expense of the necessities.” This included avoiding name-brand foods, buying rice and beans, and less eating out at restaurants. To increase his rate of savings, Evan turned increasingly toward investments. Like many Americans during the COVID-19 pandemic, he researched more about the stock market. “And I was like, ‘Okay, there’s actually feasible, reasonable strategies.’ You’re not going to get rich quick, but you’re eventually going to be able to live off the income.” He started to “squirrel away little bits here and there” and put it into the stock market slowly. He saw Prolific as another source of extra income that he could funnel reliably month to month into his investment portfolio.
In contrast to individuals who experienced discretionary dependence, this group of interviewees tended to identify themselves as savers who were actively engaged in strategies to reduce their spending and change their lifestyle choices. We can categorize these savers as “downshifters,” people who take an active approach to reign in their spending (Schor 1998:111–42). For downshifters, working a little extra translated into extra savings and security—and less work—in the future. Downshifting is a future-oriented project that taps into imagination and cultural practices. Stacy, Lea, and Evan were all projecting themselves into the future, imagining what uncertainties might lay ahead while facing facets of precarity in their daily lives (Beckert 2013). Working for extra money on Prolific and the savings that they could gather from it were critical to feeling more independent and in control of their future. These future goals reinforced dependence on Prolific income in the present.
Discussion
Scholarly literature on platforms warns against treating platforms in isolation. Individual workers are embedded in relationships with nonplatform employers, financial institutions, or families and communities (Schor, Tirrell, and Vallas 2023; Vallas and Schor 2020:286; Wood et al. 2019; Wood and Lehdonvirta 2022). To understand the value of platform work, we need to consider the broader picture of the dependencies of individuals who work on platforms. This article argues that we can expand Schor et al.’s (2020) definition of platform dependence and in the process, capture more of the workers’ heterogeneous attachments, including but not limited to platform employers. I identify and analyze three dimensions of platform dependence—episodic, discretionary, and projected—and contextualize these using concepts from household finance, consumption, and cultural studies.
First, my interviewees described dependence based on episodes of financial insecurity. They turned to Prolific as a safety net. Although they were often grateful for Prolific, they also described enduring more difficult working conditions and lower pay. They also mentioned a lack of alternatives, including savings, credit, and networks. Yet these situations lasted for a relatively short period of time—a weekend or a month—and their degree of dependence shifted as their situation stabilized.
The financial insecurity that my interviewees described echoes the experiences of many Americans. Since 1970, Americans are experiencing higher levels of income insecurity, including income volatility (Desmond 2023:16). New high-frequency data illustrate how even weekly and monthly volatility in both income and consumption can create financial difficulties for low- and middle-income households (Farrell and Greig 2016; Morduch and Schneider 2017). Many families experience episodic poverty that annual statistics do not capture (Morduch and Siwicki 2017). These dynamics have encouraged scholars in finance to be more precise about identifying and measuring the problems that insecure households face, notably, insufficiency, instability, and illiquidity (Morduch 2023).
We can identify each of these three financial problems in the existing literature on platform studies. The dominant, precarity perspective of platform work is that individuals are mainly motivated by insufficiency of their existing income sources (Vallas and Schor 2020). On the other hand, those who focus on platform work as a safety net tend to understand gig work in relation to instability, including job loss and volatility (Barrios et al. 2022; Ravenelle 2023; Rosenblat 2018). The third financial motivation for doing this work, managing liquidity, has received less attention (cf. Koustas 2018). My data—including Fallon and Sam’s stories—suggest that illiquidity can create particularly strong conditions for dependence and ultimately, exploitation. In situations of illiquidity, platform jobs can start to look like other highly exploitative financial services, such as payday lenders (Morse 2011). Zooming out, platform studies can benefit from adopting more conceptual tools from household finance (Tufano 2009).
The “discretionary” dimension of dependence emphasizes the cultural practices and social relations of platform workers. Like in other platform studies, many of my interviewees described this source of income as supplemental or extra, as “fun” money. However, instead of categorizing discretionary money as a source of independence, I explored the uses and meanings of money, including status and belonging (Sykes et al. 2015; Zelizer 1994).
The earlier work of Schor (1991, 1998) can help us theorize how patterns of consumption are the basis for platform dependence. Schor observed that Americans have worked more and more over the last few decades. In particular, she speculated that households use their extra income from these long hours to spend more; as money is spent, new needs arise for more money, which sends the individual back to work. Schor (1991:107–38) calls this the “work-spend cycle.” High consumption can be financially damaging, leading to a deficit of savings. Although Schor’s work on consumption practices is well cited, its implications have not been fully applied to analyzing the dynamics in the platform economy. Therefore, this article brings together multiple strands from Schor’s (1991, 1998, 2017; Schor et al. 2020) work to motivate an exploration of the nexus of peoples’ work and consumption habits on platforms.
Although the work-spend cycle appeared salient for many, some of my interviewees managed to save some or all of their Prolific earnings. Research over the last several decades has shown that Americans have incredible difficulties saving money (Schneider and Tufano 2009). About 40 percent of Americans do not have enough cash on hand for a $400 emergency (Federal Reserve System and U.S. Board of Governors 2021). So it was noteworthy that among the Prolific workers that I interviewed, so many were claiming to save platform income.
One of the mechanisms that I identified as important in saving was narratives about the future. These goals, or projects, were often connected to the stated desire for independence. Yet these narratives of independence typically were anchored in real and perceived threats of dependence on employers, families, and the government. Projected dependence connects back to the lived experiences of financial uncertainty in the United States, which impacts those in the lower and middle classes (Hacker 2008). It also is resonant with American cultural norms such as self-reliance (Fridman 2020; Lamont 2019). Prolific workers who save are part of a larger cultural resistance to both work and consumption that Schor (1998) calls “downshifting.”
Other platform studies have underlined the importance of imagined futures in motivating platform work. Uber drivers made sense of their activity as “temporary or marginal” to preserve their sense of dignity (Josserand and Kaine 2019:566). Platform workers also fantasize about the potential of their work to create future opportunities as a way to defend against doubts and insecurity in the present (Anicich 2022). My work adds to this strand of research by focusing not only on the career ambitions of the worker but also the projected uses and meanings of the money earned in platform jobs.
I find differences in the ways that more dependent and less dependent workers described their relations to discretionary and projected dependencies. Although discretionary dependence was more common among supplemental earners, many dependent workers also spent Prolific income on nonessentials. When highly dependent individuals spent their money on nonessentials, they missed an opportunity to save and build a safety net for instability and illiquidity. Yet if they could not spend on extras, they missed opportunities to meet other important needs, including status and social connection. Furthermore, because Prolific income could be volatile, workers’ ability to make informed choices about spending was diminished. Choosing between subsistence and status was less of a concern among nondependent individuals, who—like Nancy—described “guilt-free” spending with Prolific income.
For some, platform work was motivated by projected dependence, which connected their earnings to longer-term goals. The dependent category of workers described saving, often with great trouble, including strict discipline in spending. As we saw in the case of Lea, these savings were threatened by financial problems. Data from the financial diaries of Americans find something similar; lower-income families save money but dip into savings more frequently to smooth volatility (Morduch and Schneider 2017). Those in the relatively independent class had a higher chance of holding onto their savings over longer periods; this also made investing in long-term, low-risk assets possible.
My data on Prolific link platform work to larger dynamics of inequality. Schor et al. (2020) claim that the experiences of the dependent group of platform workers can be self-reinforcing: Individuals can be trapped in a cycle of precarity. My data support Schor’s conclusion by looking at how different socioeconomic positions thought about and used their platform income, with evidence suggesting reinforcing dynamics of precarity for the dependent workers.
In her 2017 article, “Does the Sharing Economy Increase Inequality within the Eighty Percent?,” Schor speculates that those who use platforms to earn extra income are increasingly competing with and taking opportunities from those in the working class. Platforms have made low-wage, low-status jobs—including microwork sites like Prolific—more accessible and less stigmatized for educated and middle-class workers (Schor 2017). My study of Prolific suggests that a group of middle-class workers might also be using their platform income differently than those who are dependent, including investing in financial products and services. This finding can motivate a larger study of how working and middle classes not only compete on platforms but also use income differently from these supplemental jobs and the consequences of these twin dynamics for inequality.
Conclusion
The increasing precarity of work is a longer-standing concern in the United States (Hacker 2008; Kalleberg 2009). Recently, the classification of platform workers as independent contractors has raised questions about how to best regulate employment relations in the digital age (Dubal 2017b). Schor et al. (2020) offer a compelling framework that contrasts the experiences of two broad categories—dependent and independent workers. They argue that dependent workers struggle with economic subsistence and therefore need better conditions at work. This theory of dependence is widely cited, and scholars have started to measure and test dependence (Glavin and Schieman 2022) and expand on the concept of dependence (Maffie 2022).
Given the heterogeneous workforce employed by labor platforms, this article seeks to broaden the definition of dependence. It uses in-depth interviews with 47 platform workers to gather information about the various ways that they think about and use platform income. I argue that Schor’s definition misses forms of dependence that drive participation in the platform economy. A multidimensional framework can help us to better measure dependence. Furthermore, I speculate about how dependence might reinforce precarity and vice versa. My contribution pulls from the literature on household finance, consumption, and culture to contextualize my findings.
This study has several limitations. It is focused on one platform, Prolific. My sample, although heterogeneous on a number of factors, including degrees of dependence, was based on a convenience sample of active workers. Although my in-depth interviews allowed me to gather accounts of people’s time on platforms, this limits my ability to generalize to the population. However, future work can build on the theory of dependence through comparisons with other platforms and test these three dimensions with representative samples using survey methods.
In conclusion, it is important to note that this article has used the term “dependence” practically but also uncritically. The negative connotations associated with dependence in everyday language risk reifying the neoliberal embrace of independence (Fraser and Gordon 1994). Durkheim highlighted this mythology around independence in his conclusion to The Division of Labor in Society: “We see how inexact it is to define [the division of labor], as is often done, through liberty. It rather consists in a state of dependence” (Durkheim and Simpson 1960:398). Following Durkheim, scholars can reinvigorate the sociological appreciation of concepts like dependence and solidarity. Relations between workers and employers might be more fairly regulated if our language more fully embraces the social reality of interdependence.
Footnotes
Acknowledgements
I am grateful for constructive feedback and support offered by Matthew Baggetta, Lindsey Cameron, Katherine Chen, Neil Fligstein, Marion Fourcade, Thomas Gepts, Neil Goteiner, Heather Haveman, Nadine Joseph, Maya Joseph-Goteiner, Aslı Kimya, Nicole Lopez, Jonathan Morduch, Calvin Morrill, Trond Petersen, Julian Posada, Alexandre Rigal, Daniel Schneider, Ben Shestakofsky, and Nicholas Theis, Kadir Uysal, SASE’s Early Career Workshop, the Problem Solving Sociology Coffee Hour, the Laboratory for Studies in Economic Sociology at HSE, and Boston University’s Precarity Lab. I also thank Socius editors and anonymous reviewers for their valuable feedback.
