Not long ago, labor scholars and activists fastened on Walmart’s labor practices as providing the most influential template for the “low road” approach toward employment generally. Since then, Amazon has in many ways surpassed Walmart, overtaking it in many retail markets, and bringing into play a whole new set of labor practices, many of which are equipped with powerful digital surveillance tools. This raises the question: What, precisely, do we know about the labor control mechanisms that workers encounter in Amazon’s warehouses? Despite journalistic forays and scattered but growing academic research, we have only a faint and tenuous outline of the company’s managerial regime, and of the workers’ responses to it.[1]
My colleagues and I pursued this question for two interrelated reasons. First, especially with the rise of on-line commerce, warehouse workers hold a centrally important position at the very heart of consumer capitalism. Second, and for this reason, Amazon poses a key strategic challenge for the workers’ movement generally. There will be no upsurge of worker power unless Amazon can be brought to heel.
Given our critical approach, Amazon would hardly have allowed us entrée. Using the Schneider-Harkett (2019) method of sampling, however, we found it possible to conduct a mixed methods study of Amazon’s warehouse workers, eventuating in a survey of ~700 respondents, coupled with 46 semi-structured follow up interviews with hourly employees. Below we provide a summary of the insights we gleaned from these data, as reported in two recent journal articles, available here and here. (A third paper, available here, addressed Amazon Flex.) The results were in some respects surprising.
Labor process theory (LPT) suggests that managerial regimes have assumed an increasingly despotic form (Greisbach et al 2019; Hatton 2020; Wood 2020). The reasons are complex, rooted in cutbacks in the welfare state, the ever-present threat of capital mobility, and the availability of digital technologies that can enforce a relentless pace of work. Though our research found substantial truth in this view, we also found complexities that labor process theory has yet to acknowledge. Indeed, Amazon’s power rests in its ability to invoke on a broad array of labor control mechanisms, many of which rely more heavily on the generation of consent than on despotism or coercion.
In our 2022 paper, we used qualitative data to tease out the labor control mechanisms that were apparent in the company’s operations. We did find substantial evidence of highly coercive controls, especially among Amazon workers performing “direct functions” –those picking, sorting, or packing customer orders, who commonly use hand-held scanners to perform their tasks. Crucially, these scanners track not only the movement of the customers’ packages but also the movements of the workers’ bodies. Workers’ movements are measured in a two dimensional space. One dimension involves the speed of the workers’ movements, which tracked in terms of the Units Per Hour that workers post in real time. The second dimension captures the continuity of the workers’ performance, using their “Time Off Task” record (the number of minutes during which workers have been at rest). Algorithms rank workers individually, and those who fall behind are subject to “coaching” –company speak for warnings, formal discipline, or termination. We found that the effectiveness of this surveillance system depended to a significant degree on the workers’ economic vulnerability: The more precarious their economic position, the greater the pressure they felt to “make rate” –a pattern than led us to speak of techno-economic despotism as a powerful lever of labor control.
Yet such despotic or coercive controls hardly exhausted the company’s organizational repertoire. We were surprised to find a substantial proportion of workers who spoke in highly appreciative terms regarding Amazon’s treatment of them, expressing their allegiance with the firm and even defending it against its critics. Probing into the basis of such consent, we identified three distinct mechanisms that seemed to account for such consent. One involved normative control, in which the company adopted job structures that rewarded compliant workers, moving them into positions (e.g., as “learning ambassadors” or “problem solvers”) which carried a measure of symbolic authority –but no extra pay. A second mechanism hinged on relational control, in which workers –especially those with blemished work records— developed a sense of gratitude toward the company for hiring them when no one else would. Partly an effect of the dismal state of the labor market for workers with only a high school degree, this second mechanism encouraged workers to feel obliged toward the company, as if the employment relation were akin to a gift (a point that echoes findings in Wood 2020). A third mechanism involved a set of formal provisions and rules governing attendance, promotions, and scheduling. Following Foucault, we dubbed these governmental control, since they helped Amazon interpellate the employment relation, imbuing it with the appearance of agency despite the presence of structural constraints.
What our first paper revealed, in effect, was that Amazon’s ability to dominate its workforce stems from its use of a broad array of labor controls –a form of organizational bricolage. While it relies on despotism to a substantial extent, establishing a largely punitive regime based on fear and coercion, it also wields mechanisms that are more affirmative, eliciting a sense of pride, gratitude and even a semblance of autonomy, available to workers who comply with the company’s demands. These factors leave the workforce fractured and disempowered, helping explain the low levels of resistance our interviews unearthed.
Our second paper developed a mixed methods analysis that extended our qualitative analysis and introduced new empirical insights. Using structural equation modeling we again found much evidence of algorithmic despotism. Subjection to digital surveillance tools left workers feeling highly coerced and more highly alienated from their jobs, in turn generation sharp resentment toward the firm. On the other hand, the greater the hardship workers faced in the broader economy, the more likely they were to view their jobs as a sanctuary from the precarity they had previously faced. The more precariously situated the worker, the more tolerant they were of highly alienating jobs. In effect, our second paper again points toward the duality of Amazon’s managerial regime: a punitive system of digital controls exists and clearly generates the elements of an oppositional consciousness. But this system is overlaid with a second layer of controls that enable the company to offer a level of employment stability that few workers have previously known. Here again we are reminded that coercion and consent operate hand in glove.
One final point that emerged in our second paper relates to the role of education. We found a significant relation between education and class consciousness that was both positive and pronounced: More highly educated workers (those with some college, or beyond) are much more likely to view the employment relation through an exploitative lens. What our data cannot address, however, is the nature of this relationship. Does the education effect stem from the material resources (a stronger position in the labor market) that education confers? Or does its effect stem from the symbolic resources that education provides (e.g., the ability to articulate a discourse of workers’ rights)? Arguably, both explanations warrant support, but more research is needed on this question of the education/class consciousness link, so evident in the polity more broadly.
References
Alimahomed-Wilson, J., Allison, J., & Reese, E. (2020). “Introduction: Amazon capitalism.” Pp. 1-20 in J. Alimahomed-Wilson & E. Reese (Eds.), The cost of free shipping: Amazon in the global economy. Pluto Press. https://doi.org/https://doi.org/10.2307/j.ctv16zjhcj.
Allison, Juliann E and Ellen Reese. 2023. Unsustainable: Amazon, Warehousing, and the Politics of Exploitation. Berkeley: University of California Press.
Delfanti, Alessandro. 2021. The Warehouse: Workers and Robots at Amazon. London: Pluto.
Griesbach, Kathleen, Adam Reich, Luke Elliott-Negri, and Ruth Milkman. 2019. “Algorithmic control in platform food delivery work.” Socius 5: 2378023119870041.
Hatton, E., 2020. Coerced: Work under threat of punishment. Univ of California Press.
Wood, A.J., 2020. Despotism on demand: How power operates in the flexible workplace. Cornell University Press.
Read More
Vallas, S. P., Johnston, H., & Mommadova, Y. (2022). Prime Suspect: Mechanisms of Labor Control at Amazon’s Warehouses. Work and Occupations, 49(4), 421-456. https://doi.org/10.1177/07308884221106922
Vallas, S. P., & Kronberg, A.-K. (2023). Coercion, Consent, and Class Consciousness: How Workers Respond to Amazon’s Production Regime. Socius, 9. https://doi.org/10.1177/23780231231216286
Mommadova, Y. (2024). A tale of two platforms: Habitus as the structuring force of gig workers’ experience. New Technology, Work and Employment. https://doi.org/10.1111/ntwe.12304
Image: Aaron Hughes, via justseeds.org
[1] For recent studies of Amazon workers in the United States, see Alimahomed-Wilson, Allison and Reese (2020) and Allison and Reese (2023). For a provocative study of Amazon in Italy, see Delfanti (2021).