TY - JOUR
AU - Lin, Le
AB - Abstract Drawing on ABCKID, one of the world’s largest online education platforms, and its six competitors, this paper rethinks control and consent of work in the platform economy. Falling into a new category of home-based virtual platforms, ABCKID has mobilized a group of American contractors (e.g. stay-at-home moms) who were previously marginalized in the US labor market. Using interviews, surveys and online data, I found that most ABCKID contractors show high levels of job satisfaction due to their relative gains. Furthermore, although ABCKID imposes strict and direct control over contractors’ work content and schedules, it opens new room for contractors to adapt to control. Contractors’ adaptive practices have also made the platform’s efforts to organize control as games less effective. I argue that consent of work on these platforms is not only a product of relative gains, but also a result of interactions between contractors’ adaptive practices and platforms’ direct controls. 1. Introduction Online platforms, such as Uber and DoorDash, are not only changing our daily lives but also redefining the way we understand labor and employment issues. Although some studies characterize working on platforms as flexible and sharing-oriented (e.g. Fitzmaurice et al., 2020), others warn against romanticizing the employment relationship on platforms (Kenney and Zysman, 2016; Ravenelle, 2017). In fact, Uber and other platforms may have exacerbated the already precarious lives of workers in both developed and developing countries (Scholz, 2016; Graham et al., 2017). This is not surprising given the worldwide lack of employee status and social rights of independent contractors working on platforms (hereafter ‘contractors’) (Prassl and Risak, 2016). The extant literature has documented platforms’ power and control over contractors as significant contributors to the latter’s precarity (e.g. Kenney and Zysman, 2016; Rosenblat and Stark, 2016; Scholz, 2016). However, existing studies often focus on a single form of control, such as control over contractors’ schedules (e.g. Lehdonvirta, 2018). There is a dearth of research on multiple forms of control and the combination of these forms on any given platform. Neither have existing studies explored the possibility, as shown in Burawoy (1979), of whether and how control is associated with contractor consent. Consent is defined as what is meaningful in a job and/or a sense of achievement, belonging or community that motivate workers to align with employer interests and even exceed employer goals (Burawoy, 1979; Hodson, 1999; Padavic, 2005; Kojima, 2015). This paper uncovers the dialectical relationship between control and consent of work on platforms.1 This paper rethinks control and consent of work in the platform economy by introducing transnational online education platforms. I focus on ABCKID (pseudonym) and its six competitors that are based in China and specialize in online English tutoring. ABCKID, the primary case, is one of the world’s largest online education platforms and the most sought-after education brand among global investors (Business Wire, 2018, 2019). Using peer-to-peer matching and one-on-one live video technologies, ABCKID connects over 700 000 Chinese children with over 100 000 American contractors who live all over the world. Behind these large numbers are China’s rising urban middle-class families who are eager to embrace American culture on one end, and an educated US labor force that has been struggling with bleak job prospects, wage stagnation and student loans on the other end. ABCKID is also a perfect case for understanding control and consent of work on platforms. This platform is famous for its extensive control over teachers on what and how to teach. Despite multiple forms of platform control, contractors’ positive reviews help this platform rank at the top on multiple job review websites. Some contractors even write comments in online forums about their sense that working with ABCKID is a meaningful job. ABCKID and its competitors fall into a new category of platform that has escaped from researchers’ radar. Unlike Uber and DoorDash jobs that involve working out of home, require contractors’ physical mobility or a certain degree of face-to-face interactions with consumers, and are primarily strangers-serving, ABCKID and its competitors are home-based, virtual and primarily familiar-customer-facing. Hereafter, I call this new category home-based virtual platforms. Empowered by new technologies and applied to a transnational context, these platforms have mobilized a distinct group of American contractors: ABCKID contractors are predominantly white, female and well educated, and many ABCKID contractors are stay-at-home moms, freelance artists and military wives (hereafter the group of moms, artists and military wives, or MAMs). This group of MAMs used to be marginalized in the US labor market in the sense that they were previously denied access to standard and platform jobs. Drawing on interviews, surveys and online data, I found that these teachers show high levels of job satisfaction even though many of them rely on this ABCKID job as their primary source of income,2 elevating the overall job satisfaction rate among teachers. The majority of ABCKID teachers are satisfied with this job due to the relative gains compared with their disappointing previous or current full-time employment situations. I also found that ABCKID combines multiple forms of control, especially fines and other punitive control that are enforced through camera monitoring and customer feedback, on contractors’ work content and schedules. A home-based virtual platform, however, ABCKID opens new room for contractors to cope with platform control with adaptive practices,3 such as making use of the familiarity with students to circumvent platform requirements on standard work content and teaching styles. When exercised on this new category of platforms, control interacts with adaptive practices to generate consent of contractors through three mechanisms: (a) adaptive practices of contractors in response to control reduce contractor dissatisfaction over control, elevating their job satisfaction rates and paving the way for the production of consent;4 (b) online coordination among contractors helps them adapt to control and build a sense of community and (c) being home-based virtual platforms allows a wide range of contractors to adapt to control with buffers and backup plans, cultivating their sense of belonging and fitness for the job. Contractors’ adaptive practices have also made the platform’s efforts to organize control as games less effective. I argue that consent on this new category of platform is not only a product of relative gains, but also a result of the interactions between contractors’ adaptive practices and platforms’ direct controls. My paper makes several contributions. First, it revives the tradition of synthesizing control and consent, a tradition that seems to have lost steam in recent literature. My findings not only corroborate Burawoy (1979)’s thesis that control can generate worker consent, but also contribute specific mechanisms to this work process. Second, this paper illustrates the intersection of multiple forms of control beyond any single form on which current literature tends to focus. It depicts not only how a new category of platforms expands control on contractors, but also how these platforms open new room for contractors to maneuver and circumvent control. The possibility for contractors to adapt to and get around control is not primarily due to the capabilities of the contractors. Rather, it is conditioned upon the structural features of a platform—the extent to which a platform is home-based, virtual, familiar-customer-facing and transnational. Third, this paper introduces a new category of home-based virtual platforms by using less explored dimensions to differentiate platforms. At the end of the paper, the theoretical and policy implications of the findings are discussed. 2. Theoretical framework 2.1 Control and consent of work in the platform economy The platform economy is not only gaining popularity in our daily life but is also a booming area in labor and employment studies. The platform economy entails nonstandard and contingent work which is often associated with precarity and worker isolation (e.g. Kalleberg, 2018). Some platforms, such as Uber, are notorious for poor work conditions and contractor insecurity (e.g. Scholz, 2016; Graham et al., 2017). A major source of precarity comes from the platform workforce’s weak labor status and lack of social rights. Across the world, thriving platforms categorize platform workers as independent contractors rather than employees. Such a categorization means that a series of social rights often given to regular employees—minimum wages, health and safety regulations, as well as unfair dismissal protection—have been denied by platforms (Prassl and Risak, 2016). Stripped of employee status, contractors face unemployment and other risks by themselves. This makes their poor work conditions less visible to the public (Moore, 2019). Platform contractors’ insecurity and precarity are partly rooted in platforms’ formidable power over contractors. The power asymmetry of platforms over contractors is manifest in multiple aspects: (a) platforms enjoy advantages over contractors in data and information due to the empowerment of algorithms and other technologies; (b) contractors lack organized and collective bargaining due to the absence of unions and (c) oligopoly arises because only a few platforms dominate the market, leaving contractors with few job choices (Rosenblat and Stark, 2016; MacDonald and Giazitzoglu, 2019; Prassl and Risak, 2016). Kenney and Zysman (2016, p. 62) even warn that online platforms ‘are seemingly developing power that may be even more formidable than was that of the factory owners in the early industrial revolution.’ The power asymmetry of platforms over contractors enables the former to control the latter in forms that have rarely been seen. On the one hand, platforms allow work content to be more individualized, thereby reducing both alienation and centralized control commonly found in the regular workplace (e.g. Fitzmaurice et al., 2020). On the other hand, many studies point out how platforms impose control on contractor schedules (e.g. Scholz, 2016; Rosenblat and Stark, 2016). Some new forms of control, empowered by recent technological advancements, enable platforms to monitor contractors’ schedules and locations with censors (Moore, 2019). Even platforms’ control over worker performance differs from that seen in the regular workplace: platforms rarely enforce performance control by themselves but instead involve customers to provide reputational evaluations (Gandini, 2019; Prassl and Risak, 2016). Furthermore, platforms invoke discourse and ideologies to exercise seemingly softer and less visible forms of control over contractors. Power does not necessarily present itself in terms of making decisions or setting agendas (Lukes, 1974). Rather, it can operate through discourse and ideology so that the dominated may support and legitimize the dominators’ control (ibid). For example, platforms and government agencies often promote neoliberal discourses, such as ‘enterprise culture’ and ‘resilience of workers’, so that contractors perceive themselves as entrepreneurs (Cicmil et al., 2016; MacDonald and Giazitzoglu, 2019). Consequently, contractors might seem to be more autonomous and free than was previously the case. Nonetheless, such autonomy and freedom are often ‘fictitious’ given platforms’ extended control over contractors (Shibata, 2019). Despite their attention to control, existing studies often focus on a single aspect of control, such as control over schedules (e.g. Lehdonvirta, 2018). In doing so, these existing studies of platforms lag behind previous studies on nonstandard and contingent work in the pre-platform era. Among these pre-platform era studies, Sallaz (2015) has already shown how outsourcing firms combine direct and indirect control over contractors. He depicts direct control as a despotic intervention that often entails punishment and coercion, whereas indirect control takes the form of reward games and learning games (Sallaz, 2015).5 Existing studies of platform control have not adequately discussed the combination of different control forms. Nor have existing studies explored if and how platforms can combine these already-known forms with new forms of control enabled by technologies, such as video monitoring, location sensors and algorithm-based consumer evaluations. A greater gap in the extant literature lies in the dearth of research on the relationship between control and consent of work on platforms. Consent is associated with meaning in the work, a sense of achievement or a sense of belonging (Kojima, 2015; Mears, 2015). Workers with consent will discipline and motivate themselves to engage in self-exploitation—aligning themselves with employer interests and even exceeding employer goals (Padavic, 2005). Pre-platform era studies have observed the simultaneous occurrence of control and consent, and more importantly, have inquired how control manufactures consent (e.g. Burawoy, 1979; Sallaz, 2015). For example, employers’ indirect control in the form of game manufactures consent through workers competing for prestige and acquiring a sense of achievement and belonging (Burawoy, 1979). For a variety of contingent workers in the pre-platform era, control generates consent through such an indirect and game-like process (e.g. Padavic, 2005; Mears, 2015). When it comes to the platform era, some studies with an interest in consent link it with control and collective bargaining (e.g. Johnston, 2020). However, some other existing studies only look at how consent and a sense of solidarity emerge in a seemingly control-free contractor-organized occupation community (e.g. Schwartz, 2018). Conversely, many existing studies of platform control tend to only look for contractor resistance and alienation, neglecting whether and how control is associated with contractor consent (e.g. Rosenblat and Stark, 2016; Moore, 2019). Another missing piece in existing studies is the role of technology in the control–consent relationship. According to adaptive structuration theory, technology embodies the power and resource imbalance of the social world and, more importantly, social actors often adapt to and cope with technology to alter its original goal (e.g. Markus and Robey, 1988; DeSanctis and Poole, 1994; Orlikowski, 2000). To be sure, some studies of platforms have documented how platform contractors develop daily routines and other informal practices to adapt to control (Lehdonvirta, 2018; Petriglieri et al., 2019). However, these studies have glossed over the diversity of adaptive practices on different categories of platforms. Neither have these studies linked adaptive practices to the production of consent. In sum, although a few strands of literature touch upon the issue of control or consent on the platform economy, these existing studies are inadequate for answering the research question regarding the high level of job satisfaction and consent among ABCKID teachers despite multiple forms of control adopted by the platform. 2.2 A new category of platforms This paper will draw on a new category of platforms to elucidate the relationship between control and consent. Discovering new platforms requires using less explored dimensions of structural features for differentiation. Existing research differentiates platforms along some critical dimensions, such as for-profit versus nonprofit platforms and a high versus low level of contractors’ economic dependence on platforms (Kuhn and Maleki, 2017; Schor and Attwood-Charles, 2017). However, there are three vital dimensions that have been overlooked in existing studies. Table 1 introduces these three dimensions. The first dimension, as shown in the first row, is whether the work of contractors can be done at home. For example, ABCKID allows contractors to stay at home, while working with Uber requires contractors to leave home. Being able to work from home is a significant determinant in the quality of work life, especially for women (e.g. Shamir and Salomon, 1985). There are studies that touch upon home-based platforms, but they do not single out being home based as a distinct dimension (Schor et al., 2016; Lehdonvirta, 2018). Table 1 Three new dimensions Three new dimensions . Examples . Home-based versus non-home-based [Home-based]: ABCKID Amazon Mechanical Turk [Non-home-based]: Uber DoorDash TaskRabbit Postmates Virtual versus face-to-face- [Virtual]: ABCKID Amazon Mechanical Turk [Face-to-face]: Uber DoorDash TaskRabbit Postmates Familiar-customer-facing versus unfamiliar-customer-facing [Familiar-customer-facing]: ABCKID [Unfamiliar-customer-facing]: Uber DoorDash Three new dimensions . Examples . Home-based versus non-home-based [Home-based]: ABCKID Amazon Mechanical Turk [Non-home-based]: Uber DoorDash TaskRabbit Postmates Virtual versus face-to-face- [Virtual]: ABCKID Amazon Mechanical Turk [Face-to-face]: Uber DoorDash TaskRabbit Postmates Familiar-customer-facing versus unfamiliar-customer-facing [Familiar-customer-facing]: ABCKID [Unfamiliar-customer-facing]: Uber DoorDash Open in new tab Table 1 Three new dimensions Three new dimensions . Examples . Home-based versus non-home-based [Home-based]: ABCKID Amazon Mechanical Turk [Non-home-based]: Uber DoorDash TaskRabbit Postmates Virtual versus face-to-face- [Virtual]: ABCKID Amazon Mechanical Turk [Face-to-face]: Uber DoorDash TaskRabbit Postmates Familiar-customer-facing versus unfamiliar-customer-facing [Familiar-customer-facing]: ABCKID [Unfamiliar-customer-facing]: Uber DoorDash Three new dimensions . Examples . Home-based versus non-home-based [Home-based]: ABCKID Amazon Mechanical Turk [Non-home-based]: Uber DoorDash TaskRabbit Postmates Virtual versus face-to-face- [Virtual]: ABCKID Amazon Mechanical Turk [Face-to-face]: Uber DoorDash TaskRabbit Postmates Familiar-customer-facing versus unfamiliar-customer-facing [Familiar-customer-facing]: ABCKID [Unfamiliar-customer-facing]: Uber DoorDash Open in new tab The second dimension, shown in the second row, is whether a platform allows contractors to interact with customers virtually or requires face-to-face interactions. On this dimension, ABCKID again differs from many other platforms because ABCKID teachers are entirely freed from face-to-face interaction. Pure virtual interaction means there is no concern for safety, a crucial indicator for job precarity (Kuhn and Maleki, 2017). Platform jobs that only require virtual interactions are also more portable, allowing contractors to move more freely and conduct work transnationally. The third dimension distinguishes a new platform category that is primarily familiar-customer-facing from those that are not, as seen in the third row in Table 1. Being primarily familiar-customer-facing means the majority of contractors spend most of their work time interacting with customers they already know through the platform. ABCKID encourages teachers to build a group of familiar student followers, and familiar students can even request slots of a teacher’s schedule much earlier than can new students. In contrast, most of the time Uber drivers do not ride with people they know. An Uber driver does not even have the chance to know whether the next passenger is a stranger before the platform accomplishes the passenger assignment. This distinction is important because social interactions among familiar social actors are different from those among strangers. 3. Methods This paper draws on ABCKID as the primary case. This case is ideal for understanding platform control. It video-records every single class and employs a special team in China to monitor teaching. Teachers are also subject to multiple forms of punishment, such as fines for canceling classes. Like many other platforms, ABCKID adopts a customer evaluation system that encompasses a numerical scale of one to five apples and written comments. The platform can punish teachers by allowing negative customer evaluations and comments to stay on the front page of teachers’ profiles, thus hurting teachers’ subsequent bookings. ABCKID’s control over the workforce starts at the recruitment and screening periods: it only hires teachers who can show an American or a Canadian passport, speak with North American accents and have obtained bachelor’s degrees or above. During the teaching process, ABCKID also encourages student parents to enforce rules and punishment by reporting any teacher violation to the headquarters. Despite being subject to extensive control, ABCKID teachers expressed high levels of job satisfaction. In a survey conducted in 2017, teachers reported ‘an average rating hovered around a perfect 5 stars’ (Elstrom and Ramli, 2017, p. 20). In 2018, ABCKID was rated by Flexjobs and Forbes as No. 1 for work-from-home jobs. In 2019, Glassdoor picked ABCKID as among the Top 100 ‘Best Places to Work’ based on employee reviews. I also include six competitors of ABCKID, such as ELIKID and Mango Lingo (pseudonyms), as cases. Inquiring about these competing platforms and including their information in the empirical analysis is helpful for reaching more generalized conclusions about online education platforms as well as their shared features and practices. Since many teachers who teach on ABCKID also teach with its competitors, the data on these competing platforms come from my ABCKID informants and secondary sources, such as newspaper articles. To study these online education platforms, I collected interviews, surveys and online data. The analyses and findings primarily draw on in-depth interviews with and surveys of 37 current ABCKID teachers, as well as online data from YouTube and Glassdoor, a top job review website. I interviewed 37 ABCKID teachers via telephone or Skype between September 2018 and February 2019. I recruited ABCKID teachers primarily through their Reddit community, one of the largest ABCKID teacher forums and the only forum open to non-ABCKID visitors. After I described my project on Reddit, 45 teachers contacted me. Of those, I selected 37 who showed sufficient proof of employment (e.g. job ID and screen shots of personal portal). Each interview was semi-structured. Some of my interview questions asked for quantitative information such as ‘from a scale of 1–5, how would you describe your satisfaction of working with ABCKID’. Other questions, such as ‘please describe a case where you dealt with ABCKID’s cancelation policy’, focus on documenting each informant’s story. On average each interview lasted for approximately an hour. To protect the identities of my informants, I use pseudonyms for all people and organizations throughout the paper.6 I include six additional interviews to increase the generalizability of the findings. I interviewed two senior managers of ABCKID’s competitors and two investors who specialized in online education platforms. I contacted ABCKID’s founder, but she declined the request to interview her and her managers.7 I also interviewed two parents of ABCKID students. I selected these additional interviewees through the introduction of a few managers I met when conducting fieldwork in China’s for-profit education industry for a larger project. Since the main focus of this paper is on teachers, the number of these additional interviewees is sufficient to provide non-teacher perspectives. Furthermore, I use the information provided by these additional interviewees to cross-check the validity of ABCKID teachers’ stories. Immediately after the interviews, I collected survey answers through emails from 36 informants. One interviewee did not respond to the email. This wave of survey focused on the demographic background and financial information. I conducted a second wave of survey in September 2019 to confirm if the pattern found in my sample is generalizable to the ABCKID teacher population. I complement interviews and surveys with online data. The online data include 1164 reviews of ABCKID by teachers on Glassdoor and multiple YouTube videos posted by ABCKID teachers and viewed more than 10 000 times. I further add secondary data, such as scholarly studies, reports, newspapers and websites. Secondary data are especially helpful for illuminating the pattern of ABCKID competitors whose primary data are limited. I analyzed survey results of the 37 teachers and reported their demographic information by age, sex, race and education level. I then coded interview data and categorized interviewees into different groups based on their experience in the traditional job market. For example, I categorized stay-at-home moms and military wives into the MAM group because they both were marginalized from the traditional job market. I calculated mean job satisfaction scores, compared these scores across different groups and conducted a statistical significance test. 4. Control and consent of work ABCKID and its competitors are all for-profit platforms requiring high levels of cultural capital but low levels of assets from contractors. For example, all of these platforms require applicants to have a bachelor’s degree or higher, but the only hardware a teacher needs for this job is a computer. More importantly, ABCKID and its competitors are home-based, virtual and primarily familiar-customer-facing. 4.1 A unique workforce and its relative gains Being home-based, virtual and primarily familiar-customer-facing, ABCKID has a population of contractors that is predominantly American, White, female and well-educated (with bachelor’s or above).8 This demographic profile of contractors is distinct to the extent that many existing studies find platform contractors, such as Uber drivers, to be predominantly non-white and male with a considerable proportion who did not finish college education (Schor and Attwood-Charles, 2017). Teachers in my sample are also predominantly American, White, female and well-educated. The average age of the 37 informants is 31.5 years. Among them, 36 are American citizens and 1 is Canadian. Thirty-two of the 37 are White and 29 are female. The rest of the sample is comprised of one African American, two Hispanics, one Asian-White and one Hispanic-White. All my informants have bachelor’s degrees or above (including 18 with master’s degrees and four with PhDs or JDs). The interviewees give lessons from all over the world: currently 26 of them live in the USA, 1 in Canada, 3 in Central and South America and the rest are in Europe, Asia and Africa. A closer look into the profiles of my informants reveals another pattern: there are three stay-at-home moms, three military wives and three freelance artists.9 This group of MAMs used to be marginalized in the US job market since they were often denied access to standard jobs. To date, few studies have mentioned the participation of MAMs in the platform economy. On ABCKID, MAMs are an active group and some evidence suggest that the percentage of ABCKID teachers from the group of MAMs is even higher than what the number in my sample reveals. For example, there are multiple Facebook groups for stay-at-home moms as ABCKID teachers, and one of these groups is one of the largest and most active ABCKID teacher communities on Facebook.10 The fact that four of these nine teachers are also teaching or have taught on ABCKID’s direct competitors indicates that these online English tutoring platforms have attracted teachers with similar profiles. This pattern about the six competitors is corroborated by online job review websites and other secondary source data.11 Working with ABCKID has become increasingly popular among stay-at-home moms in America. Glassdoor asks each reviewer to list the ‘Pros’ and ‘Cons’ for a job. The most frequently mentioned ‘Pro’ for working with ABCKID that appears in 254 reviews is ‘work from home or while traveling’. Another ‘Pro’ that appears in 32 reviews explicitly says this is a great job for ‘stay-at-home moms’. A YouTube video titled ‘Mom Boss’ that answers questions regarding how to work with ABCKI'D as a stay-at-home mom had been viewed over 10 000 times by January 1, 2019. In some areas, such as Utah, where there is a culture of mothers staying at home, there could be five or six stay-at-home moms in the same neighborhood working with ABCKID (Informant No. 17). ABCKID is among the very few jobs at which military wives can keep working, given that military wives are constantly on the move with their husbands (Informant No. 8, No. 23 and No. 31). The third group is comprised of freelance artists. Among my informants, one teacher auditions for a dancing company for free, another does handicrafts and paintings at home and the third is a freelance musician. They face highly inconsistent income and an unstable career path for their art-related jobs. Among the nine teachers who belong to this group of MAMs, six consider the ABCKID job their primary source of income (over 70% of personal income from ABCKID). Other groups of ABCKID teachers, such as graduate students and educators, are not commonly seen on other platforms. They boast high levels of cultural capital but do not necessarily possess high-value properties. These people, however, are less likely to draw on a ABCKID job as their primary source of income than the group of MAMs mentioned earlier. The prevalence of MAMs on ABCKID should encourage a reconsideration of economic dependency and job satisfaction. Conventional wisdom suggests there is a reverse relationship between the two: within the same platform, those who work on the job for supplemental income are more likely to be satisfied than others who are economically dependent upon the platform job (Schor et al., 2020). My findings, however, point to a different conclusion. In Table 2, I compare mean job satisfaction rates between informants who use an ABCKID job as the primary, partial or supplemental source of income. The job satisfaction rates across the three dependency levels are all above 4.3 and are not significantly different from each other (P > 0.05). Table 2 Average job satisfaction across groups . Number of informants . Average job satisfaction score (1 = extremely unsatisfied and 5 = extremely satisfied) . Average job satisfaction score Within the group that uses ABCKID as a primary source of income . Primary source (>70% income from ABCKID) 17 4.35† Six contractors belonging to the group of MAMs = 4.83‡ Other 11 contractors = 4.09‡ Partial source (30–69% income from ABCKID)  8  4.31† Supplement source (<30% income from ABCKID)  11  4.32† . Number of informants . Average job satisfaction score (1 = extremely unsatisfied and 5 = extremely satisfied) . Average job satisfaction score Within the group that uses ABCKID as a primary source of income . Primary source (>70% income from ABCKID) 17 4.35† Six contractors belonging to the group of MAMs = 4.83‡ Other 11 contractors = 4.09‡ Partial source (30–69% income from ABCKID)  8  4.31† Supplement source (<30% income from ABCKID)  11  4.32† Note: There are only 36 respondents because there is one respondent with missing data. † The result of ANOVA test is not statistically significant (P > 0.05). ‡ The result of t-test is statistically significantly (P < 0.05). Open in new tab Table 2 Average job satisfaction across groups . Number of informants . Average job satisfaction score (1 = extremely unsatisfied and 5 = extremely satisfied) . Average job satisfaction score Within the group that uses ABCKID as a primary source of income . Primary source (>70% income from ABCKID) 17 4.35† Six contractors belonging to the group of MAMs = 4.83‡ Other 11 contractors = 4.09‡ Partial source (30–69% income from ABCKID)  8  4.31† Supplement source (<30% income from ABCKID)  11  4.32† . Number of informants . Average job satisfaction score (1 = extremely unsatisfied and 5 = extremely satisfied) . Average job satisfaction score Within the group that uses ABCKID as a primary source of income . Primary source (>70% income from ABCKID) 17 4.35† Six contractors belonging to the group of MAMs = 4.83‡ Other 11 contractors = 4.09‡ Partial source (30–69% income from ABCKID)  8  4.31† Supplement source (<30% income from ABCKID)  11  4.32† Note: There are only 36 respondents because there is one respondent with missing data. † The result of ANOVA test is not statistically significant (P > 0.05). ‡ The result of t-test is statistically significantly (P < 0.05). Open in new tab Closer scrutiny reveals the role of the MAMs in job satisfaction. Among the nine contractors in the sample who belong to the group of MAMs, six of them are economically dependent on ABCKID. In the third column of Table 2, the economically dependent contractors are divided into two sub-groups—those who belong to the group of MAMs and others. The mean satisfaction rate for the former group is 4.83, while the rate for the later is 4.09. The mean satisfaction rates between the two sub-groups are significantly different from each other (P < 0.05). In other words, the reverse relationship may hold true for contractors surveyed in previous studies. Including people from the group of MAMs, however, disrupts the reverse relationship and elevates the job satisfaction for the economically dependent. To understand why contractors of the MAM group are satisfied with their ABCKID job despite this platform job being their primary source of income, we need to differentiate household economic dependence from individual economic dependence. Of the six contractors who belong to the MAM group and who use this platform as their individual primary source of income, ABCKID income constitutes less than half of the household income for five of them. For these five contractors, the major responsibility to financially support the family and provide healthcare insurance falls on the shoulders of their husbands. The reason the majority of ABCKID teachers are satisfied with this job is related to the relative gains of teachers. By relative gains, I do not mean that ABCKID teachers are highly satisfied because their current working conditions are desirable. Rather, their previous situation—either being denied access to other jobs as in the case of those in the MAM group, or being stuck in work conditions with low income as in the case of many others—has lowered their expectations, making them more likely to be satisfied now. To begin with, the financial gains for most ABCKID teachers are substantial. ABCKID pays an hourly wage ranging from $14 to $22, and the average hourly pay among teachers I interviewed is $20. To be sure, this pay level is not high and teachers rarely experience any pay raise. However, teachers appreciate the fact that ABCKID income is steady and is significantly higher than federal and various state minimum wages. Since ABCKID has not yet produced any profit, this pay level is sustained by the financial support from global investors to prioritize market share over profit. Furthermore, 14 of the 19 teachers who do not use this job as their personal primary source of income are working other full-time jobs. They are school teachers, graduate students and legal staff who do not earn a high salary from their full-time jobs. They were only expecting to earn a little extra income from the ABCKID job; therefore, their expectations were easily met. The relative gains also concern opportunities, especially for those in the MAM group. Previously, many people in the MAM group had no opportunity to work in standard 9–5 jobs. Now, stay-at-home moms no longer have to face the dilemma of no-job-while-staying-home versus not-taking-care-of-children-while-working-outside. Grace, a 36-year-old female teacher and the mother of a 2-year-old girl, said, ‘I couldn’t stand being away, you know, work for ten hours a day, missing everything, missing her growing up, basically. That’s why I have been looking for something remote for so long and I haven’t been able to find anything. So it’s definitely a motivator (for a stay-at-home mom to work for ABCKID)’ (Informant No. 14). Even when compared with working on other platforms, the opportunity gains from working on ABCKID are substantial. Non-home-based platforms do not fit stay-at-home moms and non-portable platforms do not serve military wives’ needs. There is also a safety concern present for women who choose to work for non-virtual platforms. Even among home-based and virtual platforms, very few prove to be sustainable over the long run. Platforms that were used by the informants include those for exchanging courses and editing services, as well as others that connect clothes-making contractors with their customers. Informants stopped working on these platforms because these platforms could provide neither steady bookings nor a genuine sense of social interactions. The relative gains can also be psychological, especially for those in the MAM group. First, people like stay-at-home moms who used to lack novelty in their social lives now have ‘something new to talk about’ with fellow stay-at-home mom teachers (Informant No. 17). Second, people who spend the entire day at home feel they have ‘something refreshing’ in the morning and ‘something else to break up the day’ (Informant No. 17 and No. 29). Third, freelance artists who used to struggle with the financial prospects of creating artwork can now enjoy art as a pure interest (e.g. Informant No. 13). Fourth, teachers are now empowered to do things they previously could not have imagined to be possible. For example, Brian, a 27-year-old male teacher who graduated from a liberal arts college with student loans, became a digital nomad12 in Europe when he could not find a satisfactory job and began teaching at ABCKID (Informant No. 4). Three other ABCKID teachers have also expressed their interest in becoming future digital nomads. Due to these psychological gains, it is no wonder people describe their ABCKID jobs as a ‘joy’, ‘fulfilling’ and ‘incredibly rewarding’ (Informant No. 13, No. 14 and No. 17). Relative gains can be highly visible for contractors who used to work on other full-time jobs but their previous jobs were such disappointments in terms of monetary and psychological payoffs. This situation applies to contractors in the MAM group as well as the rest in my sample. For example, Clara is a 27-year-old military wife. She was also a former public school teacher. She compares her experience in ABCKID with that of her previous teaching job: I am a teacher by trade. I taught high school English in the public school system for two years. And that was incredibly draining, exhausting and personally costly … In this job (ABCKID), there are so many teachers because we can work fantastically well to support students. I connect with them. That is amazing … I am really able to build that connection with students while still being myself … (But in American high schools), it sucks everything, not to mention working insane hours, not to mention the emphasis on everything about test scores. I literally had a student dying. But you want to talk more about test scores than the student who died. So teaching in the brick-and-mortar setting, I felt I was dying. I felt the part of me that matters was dying. (Informant No. 8) Clara is far from the only ABCKID teacher who has experienced substantial gains compared with a previous teaching position in the brick-and-mortar education system in the US. Julia, another 27-year-old female ABCKID teacher, does not belong to the MAM group. She gives a 4.5 score when asked about her job satisfaction, even though this job is her individual primary source of income. She earned her bachelor’s and master’s degrees in education and dreamed of becoming an elementary school teacher. However, she was caught in ‘overwhelming student loans’ and ‘could not afford to be a teacher’ (Informant No. 18). She added that ‘I don’t know what I would have done if I haven’t met ABCKID’. Another ABCKID teacher whose son is a high school teacher said ‘parents (in the USA) always blame teachers, but Chinese parents blame their kids’ (Informant No. 13). Apparently, teachers feel that teaching jobs in the USA are underpaid and underappreciated but there has been a leap forward in terms of job experience when they start working with ABCKID. 4.2 Control, adaptive practices and the generation of consent Being satisfied with one’s job is a necessary but not sufficient component for the generation of his or her consent.13 I asked all my informants ‘do you consider teaching with ABCKID a meaningful job’ and 18 of them said yes. I further asked ‘which parts of the job do you find meaningful’ and I allowed them to provide multiple answers. Among the 18 contractors who found the job meaningful, 8 associate the meaning of the job with the fact that the platform is home-based, virtual and primarily familiar-customer-facing but they do not explicitly mention any use of adaptive strategies. Seven of them replied that they found it meaningful to bond with familiar students, to watch these students grow and ‘push the limit of what they can talk about’ (Informant No. 16). Two contractors indicated that they found the job meaningful in connecting with a foreign culture. For example, a 33-year-old female teacher said ‘I do enjoy kind of the cultural outreach that’s happening as a result. I do feel that I have a vested interest now in what happens in China’ (Informant No. 30). Three contractors broadly mentioned they had experienced a sense of achievement or a sense of solidarity in the teachers’ community. There are 10 others among the 18 contractors who found meaning in the job. These 10 teachers used adaptive strategies as responses to the extensive control imposed by the platform. I organize this section around multiple forms of direct and indirect control imposed by the platform. I further specify how control by the platform and adaptations by contractors interact with each other to generate consent through the three mechanisms mentioned earlier. 4.2.1 Standardization of work content Unlike some other platforms that strive to offer personalized services (Fitzmaurice et al., 2020), ABCKID standardizes what to teach. To begin with, all teachers are required to use a centrally prepared and American common-core-based curriculum. The course materials and interface are also standardized. For ABCKID’s mainstream English courses, there are six levels, each with 12 units and each unit contains 12 lessons. With a similar structure (e.g. songs at the beginning followed by grammar and math), each lesson is 25-min long and comes with 25 prepared slides. When a teacher and a student are in a class, the slides always appear on the left side of the screen and the teacher and student share the right side. Furthermore, ABCKID enforces control on what and how to teach with the help of camera monitoring, customer ratings and the involvement of student parents. First, there are a series of ‘Dos’ teachers must follow. Teachers must finish all 25 slides in any class and are not allowed to skip any slide. Teachers are required to give students ‘stars’ on the screen as a primary reward, use other incentives as a secondary reward and use props (e.g. toys) as often as possible. ABCKID enforces these rules by encouraging parents to participate in class or watch recorded videos, and by designing two features in the customer-rating system. For the first feature, parents are provided with bubble tags to review a teacher (e.g. ‘this teacher did not finish all slides’). Negative bubble tags will appear under the profile of a teacher and decrease the chance for this teacher to be booked by other parents. The other feature is that, although ABCKID allows teachers to dispute ratings that are below ‘Four Apples’, teachers who do not obey the rules are not allowed to dispute. Second, there is a long list of ‘Don’ts’. Teachers are not allowed to discuss politically sensitive topics, such as the 3Ts (Taiwan, Tibet and Tiananmen). Neither are teachers allowed to be seen yawning, sipping coffee or having knees over the desk. To enforce these rules, ABCKID records all classes and employs specialists to regularly check recorded videos. For example, ABCKID asked a teacher to stop using a map that does not include the South China Sea after discovering this map through watching the video (Informant No. 20). Standardization of work content has intended and unintended benefits for the platform. It greatly reduces the instability and inconsistency that are common for online education. As one parent says, no matter who the teacher is or how fast the teacher turnover rate is, teachers will deliver predictable results (Informant No. 38). In case there is a change of teachers, the new teacher can easily pick up what has been left by the previous one. Standardization also makes teachers more replaceable. According to an investor, making teachers replaceable is critical for reducing teachers’ bargaining power for a pay raise (Informant No. 42). Moreover, standardization makes it easier for the platform to recruit teachers because teachers do not have to spend much time on preparation. Despite the abovementioned benefits, 13 teachers in my sample have expressed dissatisfaction over standardization. Although teachers enjoy not having to prepare curriculum and materials, many are annoyed by the fact that the platform has been ‘micromanaging’ how to teach (Informant No. 26 and No. 37). According to a number of teachers, the centrally prepared teaching materials are often ‘redundant’ (Informant No. 9), ‘inconsistent’ (Informant No. 16 and No. 32) and ‘inflexible’ (Informant No. 26). Some teachers mentioned that they had been reluctant to adopt the recommended ‘theatrical’, ‘overly entertaining’ and ‘clown-like’ style of teaching (Informant No. 25 and No. 33). Others could not understand why teachers were not allowed to possess teaching materials and why these materials were not released to teachers until a few days before the class (Informant No. 7). Despite the dissatisfaction of these 13 teachers, 6 of them acquired consent entirely or partly through mechanism (a): Adaptive practices of contractors help reduce their dissatisfaction over platform control and elevate their job satisfaction, an essential component of consent. As previously mentioned, most teachers primarily teach familiar students. With new students and their parents, a teacher’s yawning can often lead to a bad rating because these unfamiliar parents tend to follow the platform’s rules (Informant No. 22). When it comes to familiar students, parents tend to follow what the teacher says, instead of what the platform wants. Bella, a 27-year-old female teacher, gives a full score for her job satisfaction rating even though she dislikes the rule about having to finish all slides. She talked about how she skipped slides with her ‘regulars’—students she has known for a while and taught regularly: Especially when I was first starting working here, I tried to really stay in contact with parents through feedbacks and ask them, is this what you want me to focus on? Do you really want me to focus on free talk? If that is your priority, is it OK for me to skip these math slides? It is not helpful for students who are already beyond those math slides. I can skip this, happy to go do other stuff … So I definitely think the regulars that I have are really comfortable … That comfort, that knowing them, is really giving me the confidence to do that. For a new student, I almost never skip a slide. Very rare … Because I don’t quite know yet what that parent wants from me. (Informant No. 3) Skipping slides with regular students helps reduce Bella’s dissatisfaction over control, which in turn leads to her high satisfaction rate. The high satisfaction rate paves the way for her comment on teaching with ABCKID as ‘super meaningful’ (Informant No. 3). There are more cases where teaching regular students allows disgruntled teachers to shrug off punishments. Once a teacher yawned and, more unfortunately, the parent was in a singularly bad mood that day, so the parent left a bad rating. ABCKID decided to let the bad rating stay on the front page of the teacher’s profile because the yawning was captured on camera. Despite all the mishaps, the teacher and the parent soon reconciled because they had known each other for a long time. Eventually, the parent gave the teacher 15 ‘Five Apples’ in subsequent classes to offset the previous bad rating (Informant No. 22). Furthermore, mechanism (b) is at play to produce contractor consent: there has been frequent use of online coordination by contractors to adapt to control, building the sense of community among contractors. Given the fact that the work content is exactly the same for all contractors and that contractors can easily connect with each other virtually, most contractors in my sample actively participate in online communities to compare, discuss and imitate each other’s adaptive practices. Two of them even became leaders in these communities and grew their sense of achievement by developing innovative adaptive practices. Dustin, a 30-year-old male teacher living in the US South, is one of the leaders. He relies on ABCKID as a primary source of income but has experienced a strong sense of community. Drawing on the standardized work content and multiple online forums such as Facebook and Reddit, Dustin developed a Google-document-based template for writing feedback for students and shared the template with other teachers in those online forums. With the assistance of the template, teachers only need to fill in a few blanks, such as student name and sex. The template can then automatically generate feedback to students. Using this template, according to Dustin, can reduce four-fifths of the time previously needed to write up the feedback. Dustin felt ‘so proud’ of being part of a community after his version became one of the ‘most imported’ templates in online forums (Informant No. 12). Dustin is far from the only one who communicates and coordinates with other teachers online to adapt to standardization. There is a Facebook group where ABCKID teachers share and improve their Google slides. With over 30 000 members, this group is one of the largest among various ABCKID groups on Facebook. 4.2.2 Control over schedule Like many other platforms, there is considerable flexibility in working on ABCKID. Teachers decide what slots they will open 1 or 2 weeks in advance. However, there is a variety of constraints and control over teachers’ schedules. One constraint of this transnational platform is the time zone difference between teachers and students. To fit students’ peak time for after-school learning which is 7–9 pm Beijing time, teachers in the USA and Canada need to adjust their working hours to 6–8 am on the East Coast or 3–5 am on the West Coast. Eight teachers interviewed who live in PST and MST zones have to wake up before 5 am to teach. Some teachers even have to pull all-nighters in summer and winter vacations when Chinese students enjoy abundant learning time at homes. Mia, a 25-year-old female teacher living in the Midwest, spent the summer of 2018 working from 8 pm to 8 am, 7 days a week. She said, ‘when I was doing the overnight, it felt like my days are longer than 24 h, because, I mean, I would work 12 h and I wouldn’t want to sleep all during the day because I wanted to feel that I had a life. So I was sleeping very minimally during the day, then staying up all night long and trying to be awake. I mean, it was awful. Awful’ (Informant No. 26). Besides these constraints, ABCKID directly intervenes in teachers’ schedules with fines, firing and other forms of punishment. What makes teachers most nervous and afraid is the platform’s strict policies on cancelations and no-shows. At the beginning, the platform stipulated that a teacher could only have six cancelations within 6 months, otherwise the contract renewal would be jeopardized. Although six cancelations seem lenient, a teacher who is sick can easily use up the quota within 2 days (Informant No. 33). A 2017 Bloomberg Businessweek article criticizes ABCKID’s harsh cancelation policies with the following title: ‘Migraine? Appendectomy? Giving Birth? That’s No Reason to Cancel Classes’ (Elstrom and Ramli, 2017). Some tragic events described in the article, such as one teacher who was fired after canceling her classes to mourn the loss of her child, led to outcries on social media and a drop in job satisfaction ratings on Glassdoor. After these tragic events, ABCKID modified its cancelation policy into one with hard, medium and soft cancelations. However, many punitive and unfair elements persist. A teacher can only get a soft cancelation with documents showing the teacher’s admittance in hospital, death certificates of immediate family members or with proof of absolutely uncontrollable issues. Using hospitalization as a condition means a teacher is unlikely to get a soft cancelation even if she has a miscarriage but does not stay in hospital (Informant No. 26). These conditions for soft cancelations are also considered unfair because teachers, as contractors who might not have health insurance, could face difficulties accessing hospitals and doctors (Informant No. 31 and No. 37). Helen, a 27-year-old female teacher, illustrated how hard it was to obtain a soft cancelation. Even after she had obtained a document from the power company showing an outage in her neighborhood, ABCKID asked her to show the exact size of the affected area to determine if she could have had the chance to go to a Starbucks in an adjacent town to use wifi and continue teaching (Informant No. 16). Fines and other punitive means are commonly used to enforce cancelation rules.14 For any medium or hard cancelation, not only is the cancelation counted against the contract renewal, but teachers also have to pay fines. Mia mentioned that even though she had a stomach flu and sent ABCKID a picture proving her ‘sitting next to the toilet throwing up’ (Informant No. 26), the platform only agreed to not count this case toward her contract and still fined her. She was furious because ABCKID ‘charged her for being sick’. Despite the presence of extensive control on teacher schedules, the platform being home-based and virtual facilitates mechanism (c): a wide range of contractors can adapt to control with buffers and backup plans, cultivating their sense of belonging and fitness with the job. Contractors who live in the USA and do not travel while teaching only need in-home backup equipment, such as a wifi hotspot or power generator. These contractors consider this job easy and a good fit because ‘all it takes is to change your lifestyle a little bit’ (Informant No. 19). Those who consider themselves organized or self-disciplined are most likely to acquire a sense of belonging with the job. Clara puts it this way, ‘I am a very organized person. This job is like playing my strings. I have four alarms go off every morning. I never miss waking up because I never take chances’ (Informant No. 9). Contractors who travel internationally and frequently can also experience a sense of fitness or belonging. For frequent travelers and ‘digital nomads’, the fact that this platform is virtual, portable and transnational means that there are a variety of buffers and backup plans available for travelers, such as using the first one or two classes to test wifi in a new place before opening more slots. For those who perform teaching in Europe and Asia, time zone differences make it relatively easy to prepare buffers and backup plans. They do not have to worry about oversleeping in the early morning and can readily prepare alternative plans for emergencies, such as power outage, because alternative plans can be easily carried out during daytime or early evenings. Kristin, a female teacher currently living in Japan, appreciates the fact that she has been teaching on ABCKID in Japan rather than the USA. When her home wifi stopped working before a class in the early evening, all she needed to do was to bike to her nearby daytime workplace, a place that was still open, bright and equipped with high-speed wifi. Such a readily available backup plan is a luxury for those who live in the USA and conduct teaching before sunrise. Even those who do not teach many regular students can benefit from the fact that the platform’s resources tilt toward serving regular students while underserving new students. To better serve new students, the platform offers bonuses to teachers who accept trial classes and short-notice bookings. A few teachers who lack regular students consider short-notice bookings an opportunity to earn extra money and buffer themselves from cancelation punishment. Their strategy entails avoiding opening slots in advance. There can be no cancelation if there is no booking ahead of time. Julia describes how she implements this method: 80 or maybe 90% of my classes are short-noticed, which means they are booked within 24 hours …. The reason I do that is because of the flexibility in scheduling and the pay incentives to do so … I never open any slots ahead of time … For example on Monday, during classes I got this stomach bug which was terrible. But then I was like, OK, I am done with Tuesday. Tuesday is not going to happen. I am not teaching Tuesday morning. So I was able to save myself in the sense that I wasn’t getting cancelled or, you know, having to call someone for permission. (Informant No. 18) 4.3 Ineffective games Although ABCKID adopts both reward and learning games, teachers rarely play these games. These games are therefore ineffective and irrelevant for producing consent on this platform. 4.3.1 Reward games ABCKID does use a wide range of reward games. Some of the bounties are monetary. There are bonuses for converting trial students into customers, for accepting short-notice bookings, for completing classes on time and for referring other teachers. Some other bounties are non-material. For example, teachers are encouraged to earn certificates which open the door to different kinds of specialized courses, such as grammar and pronunciation. ABCKID also encourages teachers to compare themselves with each other and compete for social prestige. For example, it periodically places updates on social media about participating in contests among teachers to win free trips to Beijing. Despite the abundance of reward games, few teachers interviewed are interested in these games. They show little interest in competing for social prestige on this platform. In fact, most teachers claim that they do not find the idea of hanging out with other teachers appealing, except for helping each other on practical troubleshooting on Reddit and Facebook. Some are busy taking care of children and spending quality time with their partners. Others need to work on their nine-to-five jobs and coursework. A few teachers even consider reward games artificial and ill-devised (e.g. Informant No. 26). Teachers experience little difficulty in circumventing these games because the platform is home-based, virtual and thus does not provide a shared working space. In such a situation, the pressure to compare themselves with each other and compete for social prestige subsides. There is also little pressure to supervise others or to be supervised. The majority of teachers are happy about the fact that they do not have to please bosses for this job. Indeed, there are teachers outside the sample who play reward games. A few stay-at-home mom teachers on YouTube are actively promoting their referral links. Since ABCKID provides lucrative bonuses for successful referrals, some teachers prioritize referrals over teaching. Their high energy and dedication to these games, however, keep other teachers from joining certain communities. When asked about those who share teaching styles and promote referral links on YouTube, my informants recruited through Reddit describe them as ‘bubbly’, ‘cheesy’, ‘over the top’, ‘crazy’ and ‘too much’ (Informant No. 1, No. 14, No. 28 and No. 37). Some ABCKID teachers on Reddit explicitly mentioned that they would like to stay away from YouTube and Facebook where active game-players abound. The possibility to avoid these communities makes this type of game-like control ineffective and irrelevant. 4.3.2 Learning games Learning games are not effective on ABCKID either. Like any other jobs, of course, the early phase of a teacher’s experience on ABCKID is associated with uncertainty and stress. Overall, however, ABCKID is not interested in adding uncertainty or stress with learning games. Only seven interviewees said they had difficulty or stress during the initial hiring, preparation and booking periods. Preparation is easy thanks to centrally prepared work content. Various online forums also provide clues to technological challenges. Even for those who experienced difficulties, the most common cause for their difficulties was that they started working during Chinese New Year when many Chinese students were traveling and the number of bookings declined. It is not the intention of the platform to make the initial phase of the job challenging. On the contrary, to make the initial phase easy is crucial for any platform to survive and thrive. Contractors will only find the job worth doing if a platform can attract abundant customers, and customers will only want to use a platform if there are readily available contractors. As the two managers from competing companies and an investor agreed, the key for the success of a platform is to increase the number of customers and contractors rapidly and simultaneously (Informant No. 39, No. 40 and No. 42). Financial support from global investors makes this feasible. By using cash from global investors to make courses cheaper for consumers, ABCKID has been able to build a gigantic client base in a short period of time. To grow the number of teachers rapidly, ABCKID is eager to hire people with minimal requirements. Teachers mention that ABCKID promotes the idea during the recruitment and hiring processes that this job is so easy that ‘you don’t have to take it seriously’ (Informant No. 26). Some teachers mentioned that when they first saw ABCKID’s recruitment advertisement, they thought this company was a scam or multi-level marketing scheme because everything seemed too good to be true (Informant No. 4, No. 14, No. 26 and No. 33). Without any requirement for a license or education-related major, this platform has considerably lowered the bar for teaching. Some teachers even complain that ABCKID ‘is attracting people who shouldn’t be teachers’ (Informant No. 26). In a word, the platform wants to make the initial recruitment easier, not harder. For example, ABCKID wants to let teachers know they should not worry about technology. Simon, a 36-year-old male teacher, shares his observation: I think one of the appealing things about ABCKID is that they don’t care that much about the quality of the Internet… I think the ABCKID’s strategy is they want to attract people. They want to attract a lot of teachers obviously. And one of the ways they do that is they kind of make it seem that ‘you can travel the world while teaching.’ And sometimes it doesn’t matter if the Internet is not that good. (Informant No. 35) 5. Discussion This study has important theoretical and policy implications. Theoretically, this paper confirms the vitality and relevance of the control–consent framework in the connected age of digital platforms.15 On the one hand, consent indeed emerges as employers wield control, which supports the conclusions by Burawoy (1979). On the other hand, the ways in which control is associated with consent on platforms are different from what Burawoy describes. In this study, consent is not manufactured through indirect game-like control. Rather, contractors find meaning in the job despite extensive and punitive control imposed by platforms. Direct control is not being replaced by indirect control in the digital platform age, as Shibata (2019) claims. Rather, direct control is likely to gain new momentum since consent of work emerges in the interactions between contractors’ adaptive practices and direct control. Given the significance of direct control, future research will benefit not only from asking how consent is generated in spite of direct control, but also from probing how consent is manufactured because of direct control. A key takeaway from the findings is that new technologies and the distinct structural features these technologies have enabled—the extent to which a platform is familiar-customer-facing, home-based, virtual and transnational—are double-edged swords: On the one hand, these technologies and structural features have enriched platforms’ repertoires of control and further expanded platform power over contractors; on the other hand, these technologies and structural features have also empowered contractors with new adaptive practices to circumvent control. This implication echoes the conclusion about the complex role of technologies in the adaptive structuration theory. The findings emphasize that the empowering effects of technologies and the effectiveness of adaptive strategies on a particular platform are largely conditioned upon the extent to which the platform is home-based, virtual and familiar-customer-facing. Moreover, my findings suggest that the successful implementation of adaptive strategies by workers does not necessarily lead to their greater autonomy from management control. Indeed, being able to circumvent some forms of control helps contractors lift heavy burdens and acquire satisfaction and meaning in the work. Nevertheless, the production of more satisfied contractors who find meaning in platform jobs is likely to lead to less resistance to control, to the extent that consent often legitimizes control (Mears, 2015). Another thematic implication concerns the generation of consent among workers previously excluded or marginalized in the labor market, such as those teachers who belong to the MAM group. After being exposed to the neoliberal culture, being denied job market opportunities and having lowered job expectations, these contractors do not necessarily find punitive control intrusive. It is most likely that they are more willing than others to internalize platform control and value self-discipline (Cicmil et al., 2016). In the words of the debate about intrinsic versus extrinsic motivations, these previously excluded and self-disciplined contractors are so intrinsically motivated that they no longer need extrinsic motivations, such as various forms of incentives. The presence and proliferation of intrinsically motivated contractors can further inhibit resistance against control. For now, as revealed in one of the mechanisms, occupational communities serve as a venue where contractors experience a sense of community and solidarity. However, such solidarity does not necessarily translate into a coherent collective identity for platform contractors. The fragmentation of occupational communities is one barrier to collective identity. For example, some teachers only hang out on Facebook or YouTube, while others only frequent Reddit. As previously mentioned, many ABCKID teachers on Reddit frown upon the exaggerative teaching style and the deep involvement in reward games of other ABCKID teachers on Facebook and YouTube. The fact that teachers from different online forums dislike each other casts doubt on the rosy view of one seamlessly integrated community as described by Schwartz (2018). This paper also illuminates the current legal and policy debate with regard to the labor status of contractors. To understand this debate, we need to move beyond the binary classification of platform workers as either employees or contractors (Prassl and Risak, 2016). We should shift our focus to platforms and understand platforms as an entity with multiple functions, rather than a single counterparty to the contract of employment. These multiple functions, ranging from receiving labor and its fruits to providing work and pay, qualify platforms as employers because all platforms perform these multiple functions (ibid). This paper lends credence to this argument not only because these multiple functions are enforced through platform control over all factors of production, but also because platform capacity to combine multiple forms of control undergirds the multiple functions of platforms. Furthermore, the control–consent connection illustrated in this paper suggests the limitation of perceiving the labor status issue from a purely contractual perspective. Previous studies emphasizing the non-employee status of contractors start with the premise that contractors spontaneously enter and reproduce the voluntary contractual relationship.16 This line of research echoes the idea that consent emerges in the absence of power or control (Schwartz, 2018). As my findings show, however, power and control are deeply involved in the production of consent. This is why we need to look beyond employment on platforms as a voluntary contractual relationship and look at control more closely as foundations for platforms’ multiple functions. Footnotes 1 By dialectical relationship, I mean how control imposed on contractors generates their consent and how consent in turn shapes control. This paper will put more focus on the control–consent direction, although my findings also have rich implications for the impact of consent on control. 2 Primary source of income is defined here as over 70% of income coming from a single employer. Using a platform job as primary source of income, therefore, is associated with a high level of economic dependence on the platform. 3 Adaptive practices can be defined as the informal actions of social actors to circumvent the inscribed use of a technology or to bypass control of other actors (e.g. Orlikowski, 2000). This paper uses ‘adapting’, ‘circumventing’, ‘get around’ and ‘bypass’ interchangeably. 4 Hodson (1999) shows that a high level of job satisfaction constitutes an essential element of worker consent. I will demonstrate in the empirical section that high job satisfaction rates are necessary for the production of worker consent. 5 In this Burawoyian framework, reward games refer to employer-organized competition among workers for tangible bounty, such as wage increases. Learning games involve employer efforts to make workers’ learning stage stressful and difficult, so workers can gain a sense of competence after the learning process ends. 6 Since this study is citing newspaper articles and reports that use real names, it seems impossible to keep organizations completely anonymous. Using pseudonyms for organizations in this paper, however, creates an additional buffer for the informants. I also revised some titles in the references when real organizations names are used in the original titles. 7 The reason, according to an investor who specializes in online education, is that ABCKID does not want any coverage of their tight financial situation in order to better secure their next-round investment. 8 In the second-wave survey sent in September 2019, I asked the 37 contractor informants whether they agreed with this statement. Twenty-eight of them responded and twenty-six agreed. They relied on teacher profiles they saw on Facebook and ABCKID’s official forum, Hutong. The two parents interviewed also agreed with this statement (Informant No. 38 and No. 43). This paper focuses on control and consent, and inequality among education platform contractors based on sex, class, nationality and race is discussed more systematically in a separate paper. 9 Freelance artists are those who systematically work on art-related projects but are not formally employed. There are also ABCKID teachers who are formally employed on a long- or short-term basis. 10 This Facebook has 7600 members and an average of 70 posts per day. Of course, there are a few ABCKID groups on Facebook, such as ‘ABCKID Google Slide Group’ and ‘ABCKID Props and Rewards’, that have more members. But those are general groups that serve all ABCKID teachers. Among Facebook groups that are based on a particular social groups, this group for stay-at-home moms is the largest known to this researcher. 11 See, for example, Glassdoor’s review of ABCKID’s competitors. 12 Digital nomads are those who use telecommunication and other technologies to earn a living. They often conduct their jobs while traveling. 13 In the sample, for example, 17 of the 18 teachers who found meaning in their work gave a job satisfaction rate of 4 or above. On the other hand, a number of teachers who rated their job satisfaction with 4 or above had not found the job meaningful. 14 Some might suggest that ABCKID’s standardization of work content means this platform could use substitute teaching and therefore does not need such stringent and punitive rules on course cancelation. Although standardization and substitute teaching can smoothen the impact of cancelation, ABCKID still prefers stringent cancelation rules to deter frequent changes and avoid chaos associated with last-minute cancelations. An informant also mentions that such strict and punitive rules are understandable since a large number of ABCKID teachers are traveling and some are far from being qualified (Informant No. 26). Another informant adds that ABCKID Chinese managers’ inflexibility with sickness might be partly due to a cultural misunderstanding: Chinese people think it is heroic and admirable to continue working while being sick, and ‘going to hospital’ in China does not necessarily mean a serious situation; in the USA, however, it is taken for granted that a sick person should rest, and ‘going to hospital’ is different from visiting a clinic and usually indicates a more serious situation (Informant No. 13). 15 ‘The connected age’ has been used to indicate the present time when a large number of people work on the interconnected Internet. For example, see, accessed at https://gigaom.com/2007/10/06/from-the-information-age-to-the-connected-age/ on September 20, 2020. 16 See Prassl and Risak (2016) for some studies that carry this line of logic. Acknowledgements I am grateful to my anonymous informants for sharing their stories. I thank Aya Kimura, Hagen Koo, Yan Long, Pat Steinhoff, Yuki Asahina and Rumika Suzuki for their insightful comments and suggestions. Special thanks to the five anonymous reviewers and editor William Attwood-Charles for their constructive feedback. Previous versions of this work benefited from participants’ comments at the Pacific and Asia Workshop at Department of Sociology, University of Hawaii. Conflict of interest I hereby declare that there is no financial interest or connection, direct or indirect, or other situations, that may raise the question of bias in the work reported or the conclusions, implications or opinions stated. References Burawoy M. ( 1979 ) Manufacturing Consent: Changes in the Labor Process under Monopoly Capitalism , Chicago , University of Chicago Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Business Wire. ( 2018 ) ‘ABCKID (pseudonym chosen by the author for protecting informant rights) Raises $500 Million in Series D+ Financing, led by Coatue Management, Tencent, Sequoia Capital and Yunfeng Capital’, June 21, 2018. Business Wire. ( 2019 ) ‘ABCKID (pseudonym chosen by the author for protecting informant rights) Honored by Glassdoor as One of the Best Places to Work in 2020 for Second Year in a Row, Rising in Rank’. December 11, 2019. Cicmil S. , Lindgren M., Packendorff J. ( 2016 ) ‘ The Project (Management) Discourse and Its Consequences: On Vulnerability and Unsustainability in Project-Based Work ’, New Technology, Work & Employment , 31 , 58 – 76 . Google Scholar Crossref Search ADS WorldCat DeSanctis G. , Poole M. S. ( 1994 ) ‘ Capturing the Complexity in Advanced Technology Use: Adaptive Structuration Theory ’, Organization Science , 5 , 121 – 147 . Google Scholar Crossref Search ADS WorldCat Elstrom P. , Ramli D. ( 2017 ) ‘ Migraine? Appendectomy? Giving Birth? That’s No Reason to Cancel Classes ’, Bloomberg Businessweek , November 27, 2017. Google Scholar OpenURL Placeholder Text WorldCat Fitzmaurice C. J. , Ladegaard I., Attwood-Charles W., Cansoy M., Carfagna L. B., Schor J. B., Wengronowitz R. ( 2020 ) ‘ Domesticating the Market: Moral Exchange and the Sharing Economy ’, Socio-Economic Review , 18 , 81 – 102 . Google Scholar Crossref Search ADS WorldCat Gandini A. ( 2019 ) ‘ Labour Process Theory and the Gig Economy ’, Human Relations , 72 , 1039 –10 56 . Google Scholar Crossref Search ADS WorldCat Graham M. , Hjorth I., Lehdonvirta V. ( 2017 ) ‘ Digital Labour and Development: Impacts of Global Digital Labour Platforms and the Gig Economy on Worker Livelihoods ’, Transfer: European Review of Labour and Research , 23 , 135 – 162 . Google Scholar Crossref Search ADS WorldCat Hodson R. ( 1999 ) ‘ Organizational Anomie and Worker Consent ’, Work and Occupations , 26 , 292 – 323 . Google Scholar Crossref Search ADS WorldCat Johnston H. ( 2020 ) ‘ Labour Geographies of the Platform Economy: Understanding Collective Organizing Strategies in the Context of Digitally Mediated Work ’, International Labour Review , 159 , 25 – 45 . Google Scholar Crossref Search ADS WorldCat Kalleberg A. ( 2018 ) Precarious Lives: Job Insecurity and Well-Being in Rich Democracies , Cambridge , Polity Press . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kenney M. , Zysman J. ( 2016 ) ‘ The Rise of the Platform Economy ’, Issues in Science & Technology , 32 , 61 – 69 . Google Scholar OpenURL Placeholder Text WorldCat Kojima S. ( 2015 ) ‘ Why Do Temp Workers Work as Hard as They Do?: The Commitment and Suffering of Factory Temp Workers in Japan ’, The Sociological Quarterly , 56 , 355 – 385 . Google Scholar Crossref Search ADS WorldCat Kuhn K. , Maleki A. ( 2017 ) ‘ Micro-Entrepreneurs, Dependent Contractors, and Instaserfs: Understanding Online Labor Platform Workforces ’, Academy of Management Perspectives , 31 , 183 – 200 . Google Scholar Crossref Search ADS WorldCat Lehdonvirta V. ( 2018 ) ‘ Flexibility in the Gig Economy: Managing Time on Three Online Piecework Platforms ’, New Technology, Work and Employment , 33 , 13 – 29 . Google Scholar Crossref Search ADS WorldCat Lukes S. ( 1974 ) Power: A Radical Review , London; New York: McMillan . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC MacDonald R. , Giazitzoglu A. ( 2019 ) ‘ Youth, Enterprise and Precarity: Or, What Is, and What Is Wrong with, the ‘Gig Economy ’, Journal of Sociology , 55 , 724 – 740 . Google Scholar Crossref Search ADS WorldCat Markus M. L. , Robey D. ( 1988 ) ‘ Information Technology and Organizational Change: Causal Structure in Theory and Research ’, Management Science , 34 , 583 – 598 . Google Scholar Crossref Search ADS WorldCat Mears A. ( 2015 ) ‘ Working for Free in the VIP: Relational Work and the Production of Consent ’, American Sociological Review , 80 , 1099 – 1122 . Google Scholar Crossref Search ADS WorldCat Moore P. V. ( 2019 ) ‘E(a)ffective Precarity, Control and Resistance in the Digitalised Work-Place’, In Chandler D., Fuchs C. (eds) Digital Objects, Digital Subjects: Interdisciplinary Perspectives on Capitalism, Labour and Politics in the Age of Big Data . London : University of Westminster Press . pp. 125–144. Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Orlikowski W. ( 2000 ) ‘ Using Technology and Constituting Structures: A Practice Lens for Studying Technology in Organizations ’, Organization Science , 11 , 404 – 428 . Google Scholar Crossref Search ADS WorldCat Padavic I. ( 2005 ) ‘ Laboring under Uncertainty: Identity Renegotiation among Contingent Workers ’, Symbolic Interaction , 28 , 111 – 134 . Google Scholar Crossref Search ADS WorldCat Petriglieri G. , Ashford S., Wrzesniewski A. ( 2019 ) ‘ Agony and Ecstasy in the Gig Economy: Cultivating Holding Environments for Precarious and Personalized Work Identities ’, Administrative Science Quarterly , 64 , 124 – 170 . Google Scholar Crossref Search ADS WorldCat Prassl J. , Risak M. (2016) ‘ Uber, TaskRabbit, and Co.: Platforms as Employers? Rethinking the Legal Analysis of Crowdwork ’, Comparative Labor Law and Policy Journal , 37 , 1 – 30 . OpenURL Placeholder Text WorldCat Ravenelle A. ( 2017 ) ‘ Sharing Economy Workers: Selling, Not Sharing ’, Cambridge Journal of Regions, Economy and Society , 10 , 281 – 295 . Google Scholar Crossref Search ADS WorldCat Rosenblat A. , Stark L. ( 2016 ) ‘ Algorithmic Labor and Information Asymmetries: A Case Study of Uber’s Drivers ’, International Journal of Communication , 10 , 3758 – 3784 . Google Scholar OpenURL Placeholder Text WorldCat Sallaz J. ( 2015 ) ‘ Permanent Pedagogy: How Post-Fordist Firms Generate Effort but Not Consent ’, Work and Occupations , 42 , 3 – 34 . Google Scholar Crossref Search ADS WorldCat Scholz T. ( 2016 ) Uberworked and Underpaid: How Workers Are Disrupting the Digital Economy , Cambridge, MA , Polity . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Schor J. B. , Fitzmaurice C., Carfagna L. B., Attwood-Charles W., Poteat E. D. ( 2016 ) ‘ Paradoxes of Openness and Distinction in the Sharing Economy ’, Poetics , 54 , 66 – 81 . Google Scholar Crossref Search ADS WorldCat Schor J. , Attwood-Charles W. ( 2017 ) ‘ The “Sharing” Economy: Labor, Inequality, and Social Connection on for-Profit Platforms ’, Sociology Compass , 11 , e12493 – 16 . Google Scholar Crossref Search ADS WorldCat Schor J. B. , Attwood-Charles W., Cansoy M., Ladegaard I., Wengronowitz R. ( 2020 ) ‘ Dependence and Precarity in the Platform Economy ’, Theory and Society , 49 , 833 –8 61 . Google Scholar Crossref Search ADS WorldCat Schwartz D. ( 2018 ) ‘ Embedded in the Crowd: Creative Freelancers, Crowdsourced Work, and Occupational Community ’, Work and Occupations , 45 , 247 –2 82 . Google Scholar Crossref Search ADS WorldCat Shamir B. , Salomon I. ( 1985 ) ‘ Work-at-Home and the Quality of Working Life ’, Academy of Management Review , 10 , 455 – 464 . Google Scholar Crossref Search ADS WorldCat Shibata S. ( 2019 ) ‘ Gig Work and the Discourse of Autonomy: Fictitious Freedom in Japan’s Digital Economy ’, New Political Economy , May , 1 – 17 . Google Scholar OpenURL Placeholder Text WorldCat © The Author(s) 2021. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Control and consent in the connected age: the work of contractors on transnational online education platforms
JF - Socio-Economic Review
DO - 10.1093/ser/mwab007
DA - 2021-11-03
UR - https://www.deepdyve.com/lp/oxford-university-press/control-and-consent-in-the-connected-age-the-work-of-contractors-on-5eskX9X0M9
SP - 1291
EP - 1313
VL - 19
IS - 4
DP - DeepDyve
ER -