Sage Journals: Discover world-class research

Abstract

Although the “logistics revolution” has powerfully transformed contemporary capitalism, little is known about how logistics workers view such changes. To address this gap, the authors explore how Amazon’s labor practices shape workers’ orientations toward their jobs and class positions. Using a mixture of in-depth interviews and survey data, the authors find that Amazon’s production regime has contradictory effects on workers’ orientations. Digital surveillance augments managerial coercion, fostering an alienated relation toward work and a heightened sense of exploitation. Yet the company’s ability to hire precariously situated workers supplies it with workers who clearly consent to such labor practices. These results underscore the duality of managerial regimes among warehouse workers and begin to explain the ordeal of union formation, as workers adopt sharply distinct views of the working conditions they confront.

Keywords

logistics work digital surveillance labor process Amazon warehouses class consciousness

In recent years scholars have spoken with growing frequency about the logistics revolution, viewing it as inaugurating seismic shifts in the political economy of contemporary capitalism (Appelbaum and Lichtenstein 2006; Bonacich and Wilson 2008; Cowen 2014). In this view, the market power of massive retailers like Walmart and Amazon stems largely from their deployment of sophisticated logistics systems, which enable them to undercut the position that domestic manufacturing firms previously enjoyed. At the same time, scholars argue that logistics knowledge and expertise leave many firms vulnerable to disruption, as job actions at critical “choke points” can bring the accumulation process to a screeching halt (Bonacich and Wilson 2008; Scheiber 2023; Silver 2003). Missing in much of this literature, however, has been any sustained effort to explore the meanings and beliefs that logistics workers bring to bear on their jobs. Because scholars in this field have mainly focused on the political economy of logistics work, they have neglected the moral economy that emerges among workers’ ranks (Dörflinger, Pulignano, and Vallas 2021). As a result, we have little detailed knowledge regarding the subjective orientations that logistics workers hold with respect to their jobs and their positions as wage laborers. This gap is especially glaring in the postpandemic United States, where labor struggles are becoming increasingly evident in the transportation, travel, distribution, and warehousing sectors, among others.

In this article, we examine how logistics jobs impinge on warehouse employees’ work and class orientations. We focus on Amazon, the corporate behemoth that has gained a massive presence in multiple fields and whose e-commerce operations have become especially prevalent in everyday life. The subjective orientations of Amazon workers are critical in light of the company’s influence as an exemplar of digital capitalism and its strategic importance as a pivotal arena of labor struggle in the contemporary context (Rosenberg 2023).

There has been no shortage of commentary on Amazon’s internal operations, including on issues such as digital surveillance technologies, the onerous demands management places on both its workers, the unusually high rates of injury that occur in the company’s warehouses, its repressive, antiunion work culture, its failure to adopt protective measures against COVID-19, and its apparent tolerance of a racialized division of labor (Alimahomed-Wilson and Reese 2021; Green and Alcantara 2021; Gurley 2020; Gutelius and Pinto 2023; Kantor, Wiese, and Ashford 2021; Roosevelt 2021). Although this literature has prompted important debates over workers’ rights, the great bulk of it has been limited to journalistic inquiry, with social scientific contributions only beginning to emerge (see Allison and Reese 2023; Delfanti 2021; Henaway 2023; Vallas, Johnston, and Mommadova 2022). Likewise, discussions seldom focus on how workers view the company’s operations and their positions within the firm.

In this article, we overcome these limitations by collecting unique mixed-methods data on Amazon workers. Building on the sampling strategy pioneered by Schneider and Harknett (2019), we used targeted Facebook ads to generate a national sample of 700 Amazon employees. We then interviewed 46 Amazon workers, allowing us to explore in depth how workers make sense of their working conditions, management’s labor practices, the company’s use of digital surveillance technology, and workers’ outlooks toward management.

As outlined later, two distinct but overlapping theoretical approaches guide our analysis: labor process analysis, especially as it bears on what scholars have termed “digital Taylorism” (Delfanti 2021; Schaupp 2022), and studies of labor market precarity and the dampening effect it may have on working-class resistance. The results support both lines of analysis, indicating that Amazon’s production regime functions in a dual way, reflecting the company’s reliance not only on heightened coercion (as critics often stress) but also on consent (generated by workers’ vulnerable position in the labor market). We do indeed find that the company’s reliance on digital surveillance enables it to impose a highly coercive regime, policing worker productivity in ways that workers find highly alienating and that provoke an oppositional consciousness. Yet the company’s ability to recruit workers experiencing economic uncertainty has the opposite effect, since such workers seem significantly more tolerant of the harsh and demanding conditions the company imposes on them. Thus, whereas some workers experience Amazon’s production regime as a crucible of class consciousness, others view it as a sanctuary from precarity. This duality begins to explain the ordeal of union formation at Amazon, as workers exhibit divergent views of their interests and relation toward the firm. Although our analysis is subject to important qualifications, it has implications for the sociology of logistics labor and studies of worker consciousness.

Approaching Amazon

Founded as an online bookseller in 1994, Amazon has achieved a massive presence in myriad economic domains, often using its market strength in one sector to gain advantages in others (Kenney, Bearson, and Zysman 2021). Responding to the public’s reliance on e-commerce during the pandemic, the company leveraged its position as a logistics powerhouse and hired more than 500,000 new employees during the second half of 2020. Although its growth and profits have slowed since then, the company remains the second largest private employer in the United States (behind only Walmart). The company’s massive influence reaches across multiple sectors of the U.S. retail and service economy (Kenney et al. 2021), lending it enormous influence over the conventional and digital economies.

The firm’s labor practices have garnered much critical commentary over the last several years. In 2018, when journalists reported that a substantial share of Amazon workers relied on food stamps and Medicaid, the company sought to address its growing legitimacy problem by raising its hourly wage from $11 to $15 an hour and offering other benefits. Yet criticism of Amazon’s practices continued to threaten the company on at least two distinct fronts. First, Amazon’s warehouse workers have formed associations to challenge the company’s power, such as Amazonians United (a worker support group taking root in several U.S. cities) and the Amazon Labor Union, which won a union election at JFK8 in Staten Island, New York, and has sought to expand its role nationally (see Allison and Reese 2023). The Teamsters have joined the fray, moving to organize workers employed by Amazon’s Delivery Service Partners, a strategic effort the union hopes will broadly inspire worker mobilization.

Second, public criticism has given rise to direct government intervention. An example is California Assembly Bill 701, which bans injurious production quotas by warehouse companies and expands the rights of warehouse workers throughout that state (Hussain 2021; Roosevelt 2021). New York and Minnesota have followed suit, heightening regulatory threats to the company’s operations. Given the ongoing debate over Amazon’s labor practices, analysis of its internal operations and especially of its workers’ responses to them warrant careful empirical research.

The Work/Consciousness Nexus: Digital Coercion and Worker Precarity

Labor process theory has drawn on Marxist and left-Weberian views toward work and class relations, inaugurating an important line of analysis of work organizations under contemporary capitalism. Rather than asking, “Why do workers resist management’s authority?”—a problematic that takes managerial prerogatives for granted—labor process scholars have inverted these assumptions, seeking to explain why workers so often comply with managerial directives (Burawoy 1979, 1985; Edwards 1979; Mears 2015). In this view, purchasing and selling labor power is a most peculiar transaction. Although capitalists gain legal ownership over the labor power they buy, workers remain in possession of this commodity and can therefore limit efforts to generate surplus value. This confronts employers with two interrelated problems: indeterminacy regarding the realization of profits and the ever-present possibility of worker resistance to management directives. As a result, capital is compelled to develop systems of control and surveillance that involve far more than the exercise of unity and coordination (Edwards 1986). The question labor process theory pursues is how management ensures its ability to exploit the full value of the labor power it has bought.

Efforts to address this question have generated an ongoing contest between competing typologies and historical interpretations (Burawoy 1985; Edwards 1979). During the 1980s and 1990s, theorists stressed the capacity of managerial regimes to elicit workers’ consent to employer demands (Burawoy 1979; Kunda 1992; Mears 2015; Wood 2020). Given more recent shifts in political and economic conditions, however, scholars have concluded that capital no longer needs to foster consent, instead increasingly relying on production regimes steeped in highly coercive practices (Burawoy 1985; Chun 2001). This is especially true where digital technologies can be used to police the behavior of individual employees, resulting in what scholars often dub “algorithmic despotism” (Griesbach et al 2019; Kristal 2013; Taylor and Bain 1999; Wood 2020) or “digital Taylorism” (Delfanti 2021; Henaway 2023). The argument is that digital technologies have augmented capital’s power over labor, allowing employers to direct, evaluate and discipline workers with much greater granularity, exposing workers to increasingly demanding and highly individualized forms of control. Such claims have particular resonance in the case of Amazon, which has developed powerful systems that track workers’ bodily movements, their efforts to aggregate into informal groups, and even their social media posts (Wiggin 2023).

A second, related strand of thinking stems from research on the proliferation of precarious forms of employment (Kalleberg 2011, 2018; Pugh 2015; Sharone 2013; Vallas and Prener 2012; Wood 2020). Research has documented an erosion of the standard employment relation, which now encompasses a smaller proportion of the contemporary workforce than before, routinely exposing workers to risks that firms and governments had previously assumed (Hacker 2006). Although this literature is complex, one of its more provocative claims involves the assertion that precarity is not simply an externality or side effect; instead, it has come to furnish “a mode of domination of a new kind” (Bourdieu 1998:85) or “an instrument of governing” labor in its own right (Lorey 2015:1). The notion here is that employers have implicitly deployed risk, using it to disable workers’ capacity to resist capital’s demands, and even to alter their expectations regarding the employment relation itself. Pugh (2015) provided a prominent example and finds that workers experiencing highly precarious labor market conditions have come to accept what she termed the “one-way honor code,” in which workers feel obliged to work hard but with little expectation that management will reciprocate (see also Lane 2011; Sharone 2013).

Research based on these two approaches has contributed much to our understanding of the work situations that predominate under contemporary capitalism. Yet a key gap in both genres concerns the nexus between the production regimes they identify and workers’ subjective orientations and dispositions in response. Despite early arguments in the labor process genre calling for greater attention to worker agency and resistance (Knights and McCabe 2000; Thompson and Ackroyd 1995), this literature has focused on managerial practices, with notably less emphasis on worker consciousness. We argue that labor process theory has continued to emphasize the deterministic stand in Marx’s thinking, neglecting his recognition that the capitalist labor process often serves to foster an oppositional consciousness, furnishing cultural and organizational resources that workers can use to transform the production regime itself.

A similar limitation characterizes precarity research, which has largely focused on the organizational arrangements that impose risk on workers, whether through the outsourcing of workers’ jobs, the use of contingent or gig work arrangements, or the manipulation of work schedules (Kalleberg 2018; Schneider and Harknett 2019; Vallas 2019; Weil 2014). Although studies such as those by Lane (2011), Pugh (2015), and Sharone (2013) suggest that precarity has reduced workers’ willingness to contest managerial dominance, other scholars find that the use of lean staffing, outsourcing, or downsizing leads workers to perceive a growing divergence between their interests and those which their employers seek to advance (Crowley, Payne, and Kennedy 2020; Reisel et al. 2010). Precisely how precarity impinges on worker consciousness has thus remained unclear (Pedulla 2013).

In the following sections, we draw on labor process theory and precarity research to fill these gaps concerning the work-consciousness nexus. We introduce hypotheses predicting how specific facets of workers’ jobs impinge on their orientations toward work and class situations. The key is how exposure to coercive forms of managerial control, digital surveillance, and economic precarity bear on two outcome variables: workers’ experience of alienation from work and their manifestation of an oppositional consciousness.

Labor Process Theory: Digital Surveillance, Managerial Coercion, and Oppositional Orientations

The literature indicates that Amazon’s establishments frequently use rituals and arrangements designed to “naturalize managers’ power as an objective constraint and to foster workers’ involvement” (Massimo 2020:133). Walls are decorated with slogans enjoining workers to embrace the firm’s mission to “Work Hard, Have Fun, Make History.” Management often seeks to infuse work with gamelike features, offering prizes to workers who outperform others (Delfanti 2021). It also provides various extrinsic rewards to cultivate its workforce’s allegiances. Yet existing literature also suggests that these efforts to establish normative control are powerfully undermined by the material circumstances of workers’ jobs, in particular the extreme intensity of labor that management demands, its use of digital technology to surveil worker productivity, and its tendency to favor customer interests over worker well-being, which seem to jeopardize workers’ safety and health (Henaway 2023; Struna and Reese 2020). In keeping with theory that stresses the salience of such material conditions (Fantasia 1988), which often invite workers to adopt a critical view toward managerial “common sense” (Femia 1975; Vallas 1993), we formulate three hypotheses regarding job conditions and worker consciousness.

The first concerns workers’ exposure to information technologies that enable firms to automate the labor of supervision, a phenomenon that theorists have dubbed “digital Taylorism,” “algorithmic despotism,” and “algocracy” (Aneesh 2009; Delfanti 2021; Griesbach et al. 2019; Henaway 2023). This use of digital technology has received particular attention from students of labor platforms, who view the phenomenon as representing a new stage in the subordination of labor (Kenney and Zysman 2016; Rosenblat and Stark 2016; Vallas and Schor 2020). Yet this use of information technology has equally important implications when applied to the conventional economy, as it enables firms not only to monitor the bodily exertions of individual workers in real time, but also to automate decisions about the imposition of discipline or even termination. Amazon has acknowledged this use of technology, notifying the National Labor Relations Board in 2018 that its “production system generates all production related warning and termination notices automatically with no input from supervisors” (in Lecher 2019). In this context, we expect that the availability of such algorithmic technologies will enable management to impose operational demands with impunity. Thus, we predict that the greater the worker’s exposure to digital surveillance, the higher the level of coercion workers will report (hypothesis 1).

Our second hypothesis is rooted in literature addressing the consequences of managerial coercion for workers’ orientations toward their work and class situations. Thus, qualitative studies have long established that production regimes that violate workers’ expectations regarding dignity, autonomy, or respect provoke pronounced feelings of resentment and even resistance among their ranks (Gouldner 1954; Hatton 2020; Hodson 2001). Although acknowledging that workers are not blank slates when they arrive in the labor process (a point discussed later), we contend that exposure to highly coercive forms of management will have an “educative” function, shaping workers’ views of the wage-labor relation as such. We reason that workers who encounter punitive and unilateral treatment during their everyday working lives will experience resentment toward corporate employers. Specifically, we predict that exposure to managerial coercion will generate an oppositional class orientation, in which corporations are viewed as taking unfair advantage of the workers they employ (hypothesis 2).

A third hypothesis draws on recent literature on the centrality of worker alienation in accounting for workers’ views toward their employers. Although the concept of alienation languished for an extended time (Vallas 1993), recent studies have shown revived interest in the concept, with scholars seeking to clarify its meaning and to adopt novel strategies regarding its measurement (Nair and Vohra 2009; Hardering 2020). Studies in this vein report that manifestations of worker alienation have independent effects on such outcome variables as worker performance, the likelihood of engaging in illicit activity, and workers’ views regarding the social value of their own jobs (Schantz et al. 2015; Soffia et al. 2022). In keeping with this literature, we reason that worker’s feelings regarding their jobs will play a central role in shaping their views of the employer. We therefore predict that the relation between exposure to managerial coercion and an oppositional class orientation will be mediated by the experience of alienation in one’s work (hypothesis 3). We therefore predict that the covariation between coercion and consciousness will weaken or disappear when controlling for work alienation.

Thus, our use of labor process theory leads us to envision a straightforward link among four variables in which digital surveillance heightens managerial coercion, engendering greater alienation from work and a growing view of the management-labor relation in antagonistic terms. Figure 1 illustrates our conceptual map. Specifically, we indicate hypotheses related to labor process theory with solid black arrows.

Figure 1.

Conceptual map.

Sanctuary Thesis: Financial Strain and Oppositional Class Orientation

The labor process does not operate in a vacuum, and among other factors the labor market situation of workers very likely affects their response to the working conditions they confront. That is, the proliferation of economic insecurity is likely to have a sobering effect on workers’ orientations, fostering a more acquiescent or deferential view of capital’s labor practices than would otherwise be the case (Bourdieu 1998; Pugh 2015). Although much of this literature has focused on “contingent” workers or independent contractors (Lane 2011; Padavic 2005; Pedulla 2013), workers in the standard employment relation can also face uncertain, unstable, and insecure conditions (Kalleberg 2011, 2018). This indeed was the thrust of Burawoy’s (1985) argument concerning “hegemonic despotism,” in which the hypermobility of capital threatens workers’ livelihoods, compelling them to defer to their employers. Although Amazon’s warehouses cannot easily be relocated, given their need for proximity to consumer markets and the logistics infrastructure they require, its workers do face precarious conditions in several respects. One is the possibility of injury to their bodies. Another is the threat of termination should they fail to satisfy the company’s production demands. But precarity also stems from the low-wage labor markets in which Amazon typically recruits its workforce, conditions that have required a significant proportion of its workforce to rely on food stamps (Bhattarai 2018), which is also true of employees at Walmart, Amazon’s competitor. The point is that Amazon recruits its workforce within low-wage labor markets, where many workers are likely to encounter severe forms of financial hardship before being hired at the company’s warehouses.

We therefore expect that workers who experience financial strain will look to Amazon as providing a sanctuary from financial strain and uncertainty (a prediction portrayed by the gray dotted lines in Figure 1). Put differently, we predict that the association between work conditions and class orientation will be significantly weaker among workers facing economic precarity than less precarious workers (hypothesis 4). Because of limited empirical evidence on how exactly precarity limits class orientation, we do not have specific expectations. We use our mixed-methods data to explore whether precariousness mutes the effect of digital surveillance, managerial coercion, alienation, or all.

Data and Methods

Data Collection and Sampling

We use data collected over 14 months beginning in June 2020. As part of a larger mixed-method project on Amazon’s logistics workforce, we conducted a national survey of warehouse workers and in-depth semistructured interviews with warehouse workers. We recruited respondents by posting targeted ads on Facebook, utilizing the sampling method developed by Schneider and Harknett (2019, 2022; see also Griesbach et al. 2019). Our ads used images of Amazon warehouses as backgrounds with text in both English and Spanish that invited workers to participate in an independent, university-affiliated study of Amazon’s working conditions. Ads were aimed at workers whose profiles indicated that they worked for Amazon and resided in or near ZIP codes adjacent to Amazon warehouses. We entered workers into a raffle with a $250 prize distributed to a winner chosen randomly. Our quantitative survey captured variables including the nature of workers’ tasks, levels, and forms of supervision, the discipline they encountered as well as their employment histories, exposure to financial insecurity, health, and well-being, their feelings about their jobs, and the fairness of the company’s labor practices. In total, we collected 669 survey responses, and after listwise deletion of missing responses on key variables, our quantitative analysis includes responses from 558 Amazon workers.

The survey also asked respondents to indicate whether they were interested in being interviewed, allowing us to follow-up via telephone. Additionally, we recruited interviewees directly, running English and Spanish language ads in Facebook groups. We offered workers compensation of $50 to $75, depending on when we recruited them. Interviews lasted roughly an hour, were conducted via telephone in English or Spanish, recorded, and transcribed. Topics in our interviews ran parallel to those included in the survey instrument. They also explored workers’ class identification and perceptions of the episodic labor protests at Amazon warehouses. Because of time constraints, we gathered our qualitative data simultaneously with the survey results. All 46 interviews were transcribed and coded using NVivo 12.

It is important to acknowledge several limitations. First, our reliance on cross-sectional data introduces ambiguity in the causal ordering of the variables in our analysis. Our inferences are vulnerable to rival interpretations; for instance, workers harboring critical attitudes toward management may perceive their jobs in more negative terms, confounding the hypothesized relations we have advanced. We address this issue in the latter stages of our analysis, offering robustness checks and other considerations.

A second limitation concerns our sample’s underrepresentation of workers of color. Amazon’s equal employment opportunity (EEO) data indicate that the company’s blue-collar workforce is predominantly Black or Hispanic. However, workers in these groups constituted only a third (15 of 46) of our interviews and survey respondents. An underrepresentation of minority respondents stems partly from our Facebook sampling strategy (see Schneider and Harknett 2022). Additionally, Facebook comments on our ads suggest that some Amazon workers were afraid and distrusted us. If so, our data will likely be skewed toward more favorable accounts of Amazon. To address underrepresentation, we developed survey weights on the basis of the “laborer and helper” occupation in Amazon’s 2020 EEO-1 report. We divided our data into 12 groups (2 gender × 6 ethnoracial groups) and generated survey weights so that the relative size of these groups mirrors the EEO-1 report.

A third limitation is our sample’s lack of organizational data, which forces us to rely on individual reports of job conditions and limits our grasp of micro-structural processes. Here too, we acknowledge the need for studies that allow scholars to capture the role of workplace conditions at various levels of analysis.

Independent Variables

Four key explanatory variables interest us: Digital surveillance, managerial coercion, alienation from work, and economic precarity. Several of our variables are latent constructs, and we provide the list of survey items measuring each construct, item means, factor loadings, and reliability coefficients in Table 1. In support of our measurement model, Table 1 shows that our factor loadings are all within the acceptable range of greater than 0.5. Each factor’s Cronbach’s α coefficient meets or exceeds the threshold of 0.7 and is in the acceptable range. Moreover, Table S1 in the Online Supplement assesses the overall measurement model, confirming our choice of four separate latent variables.

Table 1.

Means, Factor Loadings, and Cronbach’s α Coefficients.

	Mean	SD	Factor Loading	Cronbach’s α
Digital surveillance
Digital technology is used to determine where I am at any given moment	3.79	.10	.68	.71
I feel as if someone is watching me and monitoring my performance	3.73	.09	.64
The company uses tools like scanners, computer monitors, and other technology to closely track the speed of my work	4.53	.06	.59
Managerial Coercion
I feel so exhausted at the end of the working day that it’s hard to do anything else	3.44	.10	.67	.78
I can be penalized if I don’t meet the company’s performance goals	3.15	.11	.60
I am required to work overtime even if this is inconvenient for me	2.82	.11	.63
My work schedule makes it difficult to meet my personal, family or social commitments	2.53	.10	.67
The people I work with are afraid to talk about certain things because of what the company might do	2.30	.11	.60
Alienation
The only thing I look forward to at my job is getting paid	3.39	.11	.72	.80
When I am working I feel like I become just another part of the machinery	3.19	.11	.73
I really have to force myself to go into work	2.70	.11	.72
Class orientation
Workplace decisions would be better if they were made by the workers themselves, rather than by supervisors	3.36	.09	.59	.72
If given the choice, I would join an organization that would support workers and advocate for their rights	3.75	.08	.59
Workers receive less income than they deserve	3.88	.08	.62
Corporations benefit owners and managers at the expense of workers and their families	3.66	.08	.58

Note: We used sampling weights when calculating means and standard deviations.

Table 2 shows the means, standard deviations, and ranges for all variables in the model (we factorize latent variables). Similarly, Table S2 in the Online Supplement provides the bivariate correlations between our analysis variables. In the following, we define key concepts and outline our operationalization of them.

Table 2.

Means, Standard Deviations, and Ranges.

	Mean or %	SD	Minimum	Maximum
Latent variables
Digital surveillance	.11	.0502	−2.1	.8
Managerial coercion	−.04	.07	−1.6	1.7
Alienation	.04	.06	−1.4	1.4
Class orientation	−.02	.07	−2.5	1.4
Observed job characteristics
Respondent financially strained	23%	—	0	1
Permanent contract^a	90%	—	0	1
Facility type: fulfillment center	84%	—	0	1
Observed employee characteristics
Woman (reference: man)^a	52%	—	0	1
Asian^a	10%	—	0	1
Black^a	34%	—	0	1
White^a	36%	—	0	1
Other + multiple racial groups^a	20%	—	0	1
Hispanic^a	23%	—	0	1
Age in years	35.6	1.03	18	71
Highest degree: high school or less	34%	—	0	1
Highest degree: some college or associate’s degree	50%	—	0	1
Highest degree: bachelor’s or higher	15%	—	0	1
Firm tenure: >6 mo	30%	—	0	1
Firm tenure: 6–24 mo	41%	—	0	1
Firm tenure: ≥24 mo	29%	—	0	1
n (respondents)	558

Note: We used sampling weights when calculating means and standard deviations.

Variable included for descriptive purposes only (i.e., not included in our multivariate structural equation model).

Digital Surveillance

Company documents and existing accounts (Lecher 2019; Struna and Reese 2020) suggest that Amazon deploys a twofold system of digital surveillance. One facet of this system measures units per hour, typically representing the number of items that workers scan. These data are tabulated in real time, enabling supervisors and workers to determine whether they are meeting the production quota. A second facet captures time off task (ToT), a measure of the accumulated time during which workers failed to register productive work. The company has adopted rules, encoded in its software, that automatically flag problematic ToT readings and specify any requisite discipline (e.g., ranging from warnings to immediate termination).

Not all workers are subjected to such systems, however. Those performing “direct” functions, who use scanners to pick, stow, or pack customer orders, are closely surveilled; those performing “indirect” functions (those loading or unloading trucks or working as “problem solvers”) enjoy greater latitude. Mindful of these points, we used three survey items aimed at capturing workers’ exposure to technology that tracks their bodily movements and the feeling of being monitored (see Table 1 for exact item wording).

Managerial Coercion

Debates in labor process theory often hinge on the particular mix of coercion and consent that employers are said to employ. Labor processes that rely primarily on coercion essentially involve “low trust” regimes, or what Friedman (1977) called “direct control,” in which management strictly defines the pace and method of production and enforces compliance using punitive sanctions rather than more affirmative means (Gouldner 1954). Our measure of this concept thus captures the degree to which management seeks to impose its will on workers regardless of their preferences. Our survey instrument includes five items such as penalties for missing goals, feeling exhausted, being required to work overtime regardless of one’s personal obligations, and being afraid to criticize management for fear of retaliation (see Table 1).¹

Alienation from Work

In keeping with recent efforts to revive this concept, we define alienation as a cognitive and affective condition in which the worker feels objectified in the job, expresses feelings of repulsion toward it, and views the work in purely monetary terms. Note that this usage hews more closely to classical Marxist thinking than the later, Seeman-Kohn variation (see Sayers 2011; Vallas 1993:130–31; Hardering 2020). As listed in Table 1, we measured alienation using a one-factor measure composed of three survey items that capture the worker’s feeling of being treated like a part of the machinery, that the job provides only monetary benefits, and that one has to force oneself to go into work. Our use of a one-factor measure resembles that used by Nair and Vohra (2009).

Precarity

The concept of precarity generally centers on exposure to economic conditions involving risk, uncertainty, or instability (Kalleberg 2011, 2018). It has been operationalized in different ways. One approach has focused on the distinction between “standard” and “nonstandard” work arrangements (Vallas and Prener 2012). A second approach has dealt with the job or labor market uncertainty that workers confront (i.e., the likelihood of job loss or the difficulty workers would face in finding comparable work). A third approach fastens on economic insecurity, or exposure to unpredictable reductions in one’s annual income (Hacker et al. 2014). Although our survey sought to capture all elements, our measures encountered certain limitations. First, perhaps because high consumer demand during the pandemic led the company to rely largely on permanent workers, more than 90 percent of our respondents were permanent employees, thus limiting our ability to explore precarity on the basis of employment status. Perhaps also reflecting pandemic era labor market conditions, standard measures of job- and labor market uncertainty seemed too weakly intercorrelated to be used with much confidence. In testing our sanctuary hypothesis (hypothesis 4), we have therefore constructed a measure of precarity that focused on the degree of financial hardship or strain that workers reportedly endure. Three items were especially helpful here. We count respondents as financially strained (a condition affecting 23 percent of our respondents) when they said yes to each of the following items: “I have no emergency savings,” “missed some bills in the past month,” and “omitted at least one medical service (e.g., dentist, specialist, or prescription drugs) because I could not afford it.”

Dependent Variables

Our outcome of interest here is working-class consciousness. We acknowledge the overly deductive and even mechanistic assumptions that have often plagued discussions of this dimension (Fantasia 1988; Marshall 1997). For this reason, we prefer to use the term class orientation, meaning workers’ conceptions of the class system and their position within it. We see their orientations as lying on a continuum that has deferential or consensual views at one end and more conflictual or oppositional views at the other. We define workers as holding an oppositional orientation when they view corporations as exploiting workers or profiting at their expense while also expressing their support for alternative arrangements that would better advance workers’ rights (see Vallas 1993; Wright 1985). Toward this end, the survey instrument includes four items that capture workers’ agreement regarding corporations only benefiting owners, workers receiving less than they deserve, preference for worker-based decisions, and willingness to organize (see Table 1).

Control Variables

We control several other worker characteristics that affect perceptions of surveillance, managerial coercion, and workers’ class orientation. Employees’ firm tenure measures how many months employees have worked for Amazon at the time of the survey. We group employees into three categories: 6 months or less, 6 to 24 months, and more than 24 months. To account for employees’ education, we group responses into three groups: high school degree or less, some college or an associate’s degree, and bachelor’s degree or more. Accounting for employees’ age, we include a linear and quadratic term for employees’ age in years.

In additional models (not shown), we also accounted for gender and race. We found these variables were uncorrelated with our dependent and independent variables and had no significant effect in the multivariate model. We removed these variables for parsimony, which did not affect substantive results.

Analytic Strategy

We coded the qualitative interview data using an iterative coding process (Tracy 2019; Vallas et al. 2022). The coding scheme began by identifying parent codes suggested by extant theory and journalistic reports of Amazon’s working conditions. In the second stage of coding, we identified emergent themes and subcodes. Members of the research team met regularly to identify and resolve instances that seemed complex, revising the coding scheme as needed.

We use structural equation modeling (SEM) to analyze the survey data. SEM allows us to examine the latent factors of class orientation, alienation, surveillance, and managerial coercion. We use a multigroup analysis to examine the moderated mediation proposed in hypothesis 4. This analysis is most appropriate when the moderating variable is a single observed binary indicator, which is the case for financial strain (Cheung et al. 2021; Pieters, Pieters, and Lemmens 2022). Findings from our interviews are woven into the quantitative analysis, lending greater depth and meaning to the associations our survey results provide.

Results

How Does the Labor Process Shape Class Orientation?

In the following, we examine the multivariate SEM analyses to test hypotheses 1 to 3, derived from the labor process approach. In Table 3, we examine oppositional class orientation and add the predictors of digital surveillance, managerial coercion, and alienation step by step while controlling for workers’ education, tenure, and age in all models. Model 1 reveals that greater exposure to digital surveillance is indeed associated with a more oppositional class orientation (b = 0.446, p = .000). Model 2 reveals that the link between digital surveillance and an oppositional consciousness is not due simply to the automation of management functions; indeed, the surveillance coefficient loses significance when the managerial coercion variable is introduced into the model. Instead, digital surveillance augments managerial coercion (b = 1.097, p = .000), suggesting that it enables management to tighten its control over its employees, which in turn fosters an oppositional orientation among the workforce (b = 0.366, p = .000). These results lend support to hypotheses 1 and 2.

Table 3.

Structural Equation Model: Effect of Digital Surveillance, Managerial Coercion, and Alienation on Class Orientation.

	Model 1: Just Surveillance		Model 2: Adding Coercion		Model 3: Adding Alienation
	b	p	b	p	b	p
Managerial coercion
Digital surveillance	—		1.097	.000***	1.103	.000***
Alienation
Digital surveillance	—		—		−.124	.178
Managerial coercion	—		—		.820	.000***
Class orientation
Digital surveillance	.446	.000***	.138	.123	.210	.066
Managerial coercion	—		.366	.000***	−.185	.382
Alienation	—		—		.690	.004**
Controls
Education: high school or less	−.313	.001***	−.288	.000***	−.185	.028*
Education: Some college/associate’s degree	−.211	.008**	−.285	.000***	−.200	.004**
Tenure: <6 mo	−.689	.000***	−.465	.000***	−.407	.000***
Tenure: 6–24 mo	−.297	.000***	−.193	.008**	−.158	.030*
Age	.033	.013*	.025	.041*	.031	.011*
Age²	−.001	.000***	.000	.004**	.000	.002**
χ²	171		523		766
df	49		117		168
CFI	.978		.944		.933
RMSEA	.067		.079		.080
SRMR	.061		.084		.086

Note: Estimates were weighted with survey weights. CFI = confirmatory fit index; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.

p < .05. **p < .01. ***p < .001.

The data presented in model 3 provide a more fully specified analysis of the relations among these variables, suggesting that alienation from work plays a key role in the coercion-consciousness link. When alienation from work is included, the effect of managerial coercion on class orientation becomes nonsignificant while that for alienation is highly significant. The overall pattern is much as predicted; the company uses digital surveillance to undergird its coercive practices, generating greater alienation, which in turn translates into a more oppositional view of management. The estimated coefficients for this chain of influences are all relatively strong (1.103, 0.820, and 0.690, respectively), suggesting that the company’s deployment of technology and coercive practices reverberate strongly through the chain, predicting substantial increments in worker opposition.

One final point in Table 3 concerns our control of education and tenure. Educational level predicts class orientation in highly suggestive ways. Workers with high school educations or less are much less critically disposed toward management than are workers with a college degree. Tenure, too, is associated with class orientations as relative newcomers to Amazon are less critical of management than those who have been on the company payroll for three years or more. This suggests a learning effect, with more experienced workers becoming increasingly critical of management over time.

The interview data provide a deeper understanding of the patterns in Table 3. Workers do indeed perceive the company’s digital surveillance systems as greatly heightening the production pressures they experience in their everyday working lives. References to the intense pressure workers feel to “make rate” (i.e., to comply with the digitally enforced quota their departments have set) are strewn throughout our interviews, but mainly among workers performing direct functions. Such workers spoke with pronounced resentment at the intensity of the company’s demands, knowing that failure to comply exposed them to punitive treatment. Rick, in his late 50s and working in Pennsylvania, told us how the company posted worker units-per-hour scores in highly visible locations continuously throughout the day. This meant that if your scores were lagging, you had to “really bust your ass for the rest of the night, so that you don’t fall into that category of being [deficient].” He continued,

if you have more than say 15 minutes or 20 minutes of ToT, you’ll be approached by a manager, saying that, “You were time off task. Can you explain to me what you’re doing during that time?” Oh, you might’ve had a piece of equipment that broke down and you had to go switch a piece of equipment and get one that was actually operating properly and unless you keep track of that, then it just automatically goes towards what they call “accumulated time off task.” It’s a rigorous environment, believe me.

A second point concerns the apparent consequences of such coercive treatment for workers’ views of their work situations and management’s treatment of them. Feelings reflecting alienation were common, with many feeling treated like machines. Sara, 42 and working in South Carolina, commented that “we don’t have robots at our center. So we are the robots.” Workers employed as sorters or pickers often reported dissociating themselves from the labor they had to perform. Felix, a Wisconsin man in his early 30s, told us that “you just ‘clock out’ [stop thinking] when you’re clocking in and then vice versa when you’re leaving the building.” Helena, in her early 30s and working in southern California, said that she tends to “leave my body on autopilot . . . just not to be there and help make time pass.” Rick, quoted previously, noted that “people are praying for stuff to beak down so they can do nothing, because that’s how much you’re moving here.” Many workers spoke appreciatively of the company’s provisions for voluntary time off (VTO), which enabled them to leave their shifts early if they desired, but without any pay for those working hours. Sara, just quoted, made the following observation, indicative of the repulsion many workers feel toward their jobs.

[When] they throw out VTO, or voluntary time off, people will seize, like they lose themselves, they’re like, “Oh my God, I got to go!” People just, they want out. You don’t get paid to take VTO, you’re shortchanging your paycheck, it’s just so miserable people are just like, “I want to go!” They’re happy when they get it and they run out the door.

References to the coercion and production pressures workers faced were commonly interlaced with expressions of resentment toward the firm for such treatment. LaTonya, a Black woman in her 40s, recalled that while employed as a packer,

I’d be sweating so hard I felt like I was gonna die. A lot of people were on 12-hour energy drinks, drinking Red Bulls to keep up with the pace. I would never work as a packer like that again.

Stephanie, a mother of four in her late 20s, was hesitant to confide in us at first, but eventually shared her view about Amazon as an employer, saying simply that “it’s a slave trade . . . or a sweatshop actually.” Kelley, in her 40s and working in Kentucky, told us simply that “They want it every second from you. They own you every minute.” She continued, making references to Jeff Bezos, then Amazon’s CEO, who embodied the stark unfairness many workers perceived. Her hope, that research on Amazon might be of material benefit to workers generally, is worth quoting in full:

I don’t want to see more work places go to this sort of way of doing business because it’s not good for us as employees, for anybody. It’s horrible. It’s good for Jeff Bezos, but not for us. I think most people need to know that, and I think that more people need to understand what they’re supporting when they order something from Amazon. So I want this out there.

This comment captures an essential feature of oppositional consciousness –its focus on the structured antagonism between the corporation and the workers it employs.

One last point concerns workers’ views regarding the concrete forms of resistance that might be expected to accompany opposition to the company’s labor practices, especially in the light of its extraordinary profits. Clearly, resistance has occurred at many Amazon establishments (Allison and Reese 2023). Predictably, our interviews did unearth instances of overt job actions. In one case, a young woman ripped down instruction sheets management had posted that sought to intensify the work effort. Department team meetings at the outset of workers’ shifts at times provided occasions for workers to challenge managerial practices (prompting some managers to discontinue such meetings, citing the need for social distancing). We heard one instance in which a worker deliberately angled large packages in such a way as to bring the line down, thus gaining unauthorized rest. We also heard frequent support for the work stoppages and protests that broke out during the worst of the pandemic.

Yet such expressions of overt resistance were rare even among the company’s fiercest critics. Our interview data suggested two interconnected reasons: the fear of managerial retaliation and economic vulnerability. Fear was evident in numerous comments about the company’s “ruthless” behavior. Robert, a 41-year-old man in Wisconsin, said, “Amazon had put fear, instilled fear into them . . . so people won’t stand up for themselves.” Felix, 30, a Latinx man in Colorado, argued that “you even breathe the word, start the word with the letter ‘u,’ they’re going to shut you down.” He continued, “a lot of people are low income . . . and this is their livelihood. This is their means of eating, paying rent, getting gas in their car.” In the absence of any protective force, the risks are simply too great to bear.

Does Economic Precarity Mute the Effect of Job Conditions on Class Orientation?

These qualitative observations lead us to examine the validity of the sanctuary thesis (hypothesis 4). Table 4 shows the results of multigroup analyses comparing our core model for respondents who differ in terms of their exposure to precarity or financial strain (see Lin et al. 2010; Schoemann and Jorgensen 2021).

Table 4.

Multigroup Analysis, by Financial Strain.

	Financially Very Strained		Financially Not or Somewhat Strained		Difference (Very Strained − Somewhat)	p
	b	p	b	p	Difference (Very Strained − Somewhat)	p
Managerial coercion
Digital surveillance	1.361	.000	.952	.001	−.409	.238
Alienation
Digital surveillance	−.092	.559	−.092	.559
Managerial coercion	.726	.000	.869	.000	.143	.168
Class orientation
Digital surveillance	.221	.146	.221	.146
Managerial coercion	−.273	.397	−.273	.397
Alienation	.533	.169	.836	.046	.303	.040*
Controls
Education: high school or less	−.224	.068	−.224	.068
Education: some college/associate’s degree	−.235	.049	−.235	.049
Tenure: <6 mo	−.403	.002	−.403	.002
Tenure: 6–24 mo	−.154	.110	−.154	.110
Age	.027	.141	.027	.141
Age²	.000	.070	.000	.055
	χ²	Dif χ²	Dif df	p
Model: invariant	1,330.4
Model: moderated mediation	1,309.9	20.5	3	.024**
Model: full moderation	1,303.3	6.6	3	.170

Note: We used sampling weights when calculating means and standard deviation. The table shows a group-specific model in which the effect of surveillance, managerial coercion, and alienation vary by financial strain (i.e., moderated mediation). Moderated mediation fits the data significantly better (p = .003) than an invariant model that forces coefficients to be the same for all employees. A fully moderated model, in which direct effects on class orientation also vary among groups, did not improve model fit (p = .128).

p < .05. **p < .01.

Our results align with hypothesis 4, suggesting that the relation between working conditions and class orientation grows weaker under conditions of economic precarity. More specifically, precarity does not alter the link between surveillance and coercion or coercion and alienation, suggestive that all employees experience management as more coercive when their workplaces use more digital surveillance. Likewise, all employees feel more alienated when exposed to coercive management. But precarity does inhibit alienated workers from adopting an oppositional orientation, as the effect of alienation on class orientation is significantly weaker (p = .040) among strained respondents (b = 0.533) than for individuals facing less financial strain (b = 0.836). In short, the need for economic security among workers facing hardship makes them more tolerant of alienated labor than workers in stronger economic positionss.²

The interview data reinforce this conclusion. Although we have focused on workers manifesting an oppositional consciousness, a substantial proportion of the sample defended Amazon, expressing their gratitude for the company’s treatment of them. In some cases, this deferential view was linked to a broader investment in a “diligent worker” identity that was partly rooted in the wider culture but reinforced by Amazon’s job structures (Pugh 2015; Sharone 2013; Vallas et al. 2022). Yet expressions of deference to the company often seemed to stem from workers’ appreciation of the financial stability it provided, which was a sharp departure from what they had known before. Chaz, a 46-year-old man in New Jersey, allowed that

it used to be basically the only time I ever paid my bills was when they [creditors] were threatening to shut stuff off. I’m actually able to pay my bills now, so I have honestly nothing but good things to say about working for Amazon.

Kathy, a southern California woman in her late 50s, was behind in her property taxes and carried substantial credit card debt. Her situation, combined with her blemished work record, left her feeling thankful to the company for hiring her even though she has suffered two serious injuries while on the job. When we asked Bernard, a 28-year-old Connecticut man how he viewed the future, he said,

I don’t know what to expect. That’s kind of how my life can be, kind of chaotic. I’m hoping that I can keep working here for a while and just kind of keep everything afloat, how it’s going right now, because everything’s going all right.

Such workers have known hardship and uncertainty in the recent past and have no desire to risk returning to such chaos. Their image of the future has receded from their view, much as Bourdieu (1998) suggests.

Robustness Checks

Our analysis has made important assumptions concerning causal ordering. Yet arguably, reverse causation may operate, where workers’ class orientations and job attitudes shape their reports of management practices. That is, workers predisposed to negative views of Amazon may be more likely to perceive their work situations as coercive, thus confounding the analysis. To address this issue, we conducted a two-stage least squares regression using factor scores for our latent variables (managerial coercion, alienation, class orientation) and facility type as instrumental variable. Table S3 in the Online Supplement shows the results. In summary, even when regressing facility type on managerial coercion, our core model holds. Although we cannot definitively rule out reverse causality without experimental or longitudinal designs, the present two-stage least squares regression analysis suggests some significant causality flows from experiences of managerial coercion to alienation and class orientation.

Conclusions

Although there is widespread recognition that logistics workers hold a strategic role in the political economy of contemporary capitalism, there has been little systematic research on the normative outlooks that emerge among this workforce. Aiming to fill this gap, we have examined the work situations and worldviews of warehouse workers at Amazon, one of the most powerful corporations in the United States, and beyond. We have paid particular attention to the work-consciousness nexus, exploring the factors that affect workers’ responses to the production regime that Amazon has deployed. Our analytic framework has been informed by two theoretical problematics: one rooted in Marxist theories of the labor process and the other drawing on recent theories of precarity. Although our findings should be read with important cautions in mind, they support predictions offered by both approaches, suggesting that managerial practices at Amazon have contradictory effects.

On the one hand, characteristics of the labor process promote an oppositional consciousness among Amazon employees, much as classical Marxist analysis expects. Here we find a set of punitive and coercive practices accentuated by digital technologies (as predicted in hypothesis 1), enabling the firm to police worker productivity with a granularity that would otherwise be unattainable. Moreover, the ideological consequences of such a heightened level of discipline are highly significant (as hypothesis 2 predicts) and largely due to the sense of alienation workers feel concerning their labor (hypothesis 3). Thus, the experience of the labor process itself is highly consequential for workers’ class orientations. The operational strategies on which Amazon relies not only accelerate the movement of goods through its distribution system; they also generate normative support for the very forms of resistance the company has struggled to contain. Our quantitative and qualitative findings align closely with accounts by journalists and activists, who regularly (and rightly) decry the abusive conditions that Amazon deploys. Although some theorists have declared the concept of class to be a “zombie” category, as if class identities have largely waned (Beck 2007; Eidlin 2014), our data suggest that the operational pressures built into logistics systems cut in the opposite direction, kindling the basis on which worker resistance might grow.

Yet this is not the only pattern evident in the experience of Amazon employees, many of whom express their appreciation for the job rewards that Amazon provides. Such appreciation seemed especially common among workers who held vulnerable positions in the labor market, and indeed, our survey results reveal that financial strain weakens workers’ ability to embrace an oppositional consciousness in response to the deprivations their jobs impose. This suggests that the company’s ability to maintain control over its workforce depends partly on its access to precariously situated workers willing to tolerate the company’s regime, many of whom view the job as a sanctuary from the economic hardship they would otherwise confront. Exposure to high levels of precarity—having no savings, being unable to pay one’s bills, and having to forego medical care because of its cost—does not affect workers’ feelings about their jobs. Poor work conditions result in alienation among all workers. But being financially strained does weaken the worker’s ability to link such alienating conditions to the corporate practices that produce them. Precarity seems to weaken workers’ ability to articulate a discourse of workers’ rights, perhaps reflecting their single-minded focus on meeting their subsistence needs. In this view, harboring critical thoughts toward management becomes a luxury workers cannot afford. Better by far to develop the ability to withstand hardship, to do what one must, and even to be thankful for what one has.

These findings suggest that the ordeal of labor struggle at Amazon may stem in part from the simultaneous coexistence of two distinct currents among the company’s workers, dividing workers between those who see labor struggle as threatening their sole source of security and those who feel driven to challenge the company’s control over their working lives. The implication here is not simply that coercion and consent often coexist, a point that has been a subtext of labor process theory for several decades (Burawoy 1979, 1985; Wood 2020). Rather, distinct categories of workers experience the “same” production regime in sharply different ways, depending on their positions in the labor market. Analysis of production regimes and of capital’s ability to limit worker resistance must pay greater attention to the different cultural logics that arise as workers make sense of their work and class situations (Vidal 2022).

Of particular interest are the effects of education on workers’ responses to the production regime on which Amazon has relied. Our analysis finds that workers with the least education (high school or less) hold a less oppositional view of management than do workers with postsecondary schooling. What remains unclear however, is why this pattern obtains. Does it stem from the lower value of their labor power (i.e., their weaker position in the labor market)? Or is this difference rooted in the greater cultural resources that educated workers can invoke, enabling them to articulate a set of claims regarding the rights that employers owe their workers? Our qualitative data suggest that both interpretations are plausible, prompting the need for further research.

Our analysis suggests that Amazon’s ability to maintain its production regime depends partly on its ability to recruit a financially strained workforce, providing a brake on the frequency and intensity of worker resistance. Consequently, any improvements in the labor market conditions for low-wage workers will likely alter the coalitional dynamics within the company’s establishments. The effects here may themselves be contradictory. On the one hand, if an improving economy provides alternative employment opportunities for low-wage employees, the sanctuary effect may wane, empowering workers to a greater extent. On the other hand, an improving economy may drain the warehouse workforce of its better educated workers, whose exit depletes the workers’ movement of its most critically minded constituency. Counteracting this latter trend, however, is the erosion of many career pathways that educated workers previously enjoyed, forcing some to resort to jobs (such as Amazon provides) they might previously have refused. How these various trends impinge on the composition of Amazon’s workforce may play an important role affecting the course of worker mobilization on this terrain.

Another point warranting attention concerns the timing of our study. We collected our data during the onset of the pandemic (the summer of 2020), when employment was in free fall, when protests had arisen concerning the risks to which many workers were exposed, and when reliance on e-commerce contributed to an explosion of hiring on Amazon’s part. Arguably, the company’s surge in hiring led it to recruit a more heterogenous workforce than previously, as many white-collar and service workers resorted to warehouse work out of sheer desperation. This raises questions about our findings’ enduring validity and applicability to the postpandemic time. How, for example, has the rise of labor activism affected workers’ views of the wage-labor relation itself? As labor market conditions have improved, has the sanctuary effect tended to wane? These are matters for future research on Amazon, other major employers in the logistics sector, and beyond.

One last point concerns the methods that such research might best use. Some scholars hold that survey methodology typically fixates on individual attitudes rather than organizational entities and social movements (Fantasia 1988; Marshall 1997), thus obscuring the emergent effects of workplace culture as such. We disagree and argue that survey research can capture the presence of oppositional sentiments—James Scott’s “infrapolitics” (Scott 1990)—that provide the raw materials with which organizations and movements can build. We very much acknowledge the difficulties that inhere in cross-sectional survey designs such as ours. Although we have sought, through statistical maneuvers, to address the assumptions on which our analysis relies, longitudinal designs are indispensable if we are to sort out the work-consciousness link. Such research seems especially needed when scholars debate the sources of the new labor activism now on the political horizon and the role of the labor process in fueling such events.

Supplemental Material

sj-docx-1-srd-10.1177_23780231231216286 – Supplemental material for Coercion, Consent, and Class Consciousness: How Workers Respond to Amazon’s Production Regime

Supplemental material, sj-docx-1-srd-10.1177_23780231231216286 for Coercion, Consent, and Class Consciousness: How Workers Respond to Amazon’s Production Regime by Steven P. Vallas and Anne-Kathrin Kronberg in Socius

Footnotes

Acknowledgements

The authors wish to thank the following people for their generous participation in this research: Yana Mommadova, Hannah Johnston, Chris Prener, and Raimy Jaramillo. We are also grateful to the two anonymous reviewers and the editors for their insightful comments. The authors are grateful to Yana Mommadova, Hannah Johnston, and Chris Prener for their tireless work on this project.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: We gratefully acknowledge material support from the Office of the Provost, Northeastern University.

ORCID iD

Steven P. Vallas

Supplemental Material

Supplemental material for this article is available online.

1

One reviewer raised questions about our delineation of predictors, arguing that digital surveillance is itself a form of managerial coercion. Our approach is based on the long-standing tradition in organization research in which technology and work organization are considered conceptually distinct. shows that a three-factor model combining digital surveillance and managerial coercion fits noticeably worse than a four-factor model separating surveillance and coercion.

2

Only respondents experiencing the most extreme form of financial precarity (no savings, missing some bills, and omitting at least one medical service) exhibit the sanctuary effect. Looking at our financial indicators individually yielded nonsignificant group differences. Perhaps because economic uncertainty is a common experience among many Amazon employees, some workers have grown inured to its effects. These results imply that our sample provides a conservative estimate of precarity and work conditions.

Author Biographies

Steven P. Vallas is professor emeritus at Northeastern University. His current research centers on two overlapping projects: one on the nature of platform work and a second on work and authority relations in the logistics sector. He has written or edited numerous studies on work, technology and labor relations in various branches of the economy, including telecommunications, manufacturing, biotechnology, and social media.

Anne-Kathrin Kronberg is an assistant professor of sociology at the University of North Carolina at Charlotte. She has been trained in both the United States and Germany, studying how organizations shape careers and career outcomes. Her research has been published in Social Forces, Work and Occupations, and Mobilization, and her work has been funded by the German Research Foundation. Her current research explores how the structure of Twitch, the social media platform, affects gender and racial inequality among influencers.

References

Alimahomed-Wilson

Jake

Reese

Ellen

. 2021. “Surveilling Amazon’s Warehouse Workers: Racism, Retaliation, and Worker Resistance Amid the Pandemic.” Work in the Global Economy 1(1–2):55–73.

Allison

Juliann E

Reese

Ellen

. 2023. Unsustainable: Amazon, Warehousing, and the Politics of Exploitation. Berkeley: University of California Press.

Aneesh

. 2009. “Global Labor: Algocratic Modes of Organization.” Sociological Theory 27(4):347–70.

Appelbaum

Richard

Lichtenstein

Nelson

. 2006. “A New World of Retail Supremacy: Supply Chains and Workers’ Chains in the Age of Wal-Mart.” International Labor and Working-Class History 70(1):106–25.

Beck

Ulrich

. 2007. “Beyond Class and Nation: Reframing Social Inequalities in a Globalizing World.” British Journal of Sociology 58(4):679–705.

Bhattarai

Abha

. 2018. “Thousands of Amazon Workers Receive Food Stamps.” The Washington Post, August 23. Retrieved November 24, 2024. https://www.washingtonpost.com/business/2018/08/24/thousands-amazon-workers-receive-food-stamps-now-bernie-sanders-wants-amazon-pay-up/.

Bonacich

Edna

Wilson

Jake B.

2008. Getting the Goods: Ports, Labor, and the Logistics Revolution. Ithaca, NY: Cornell University Press.

Bourdieu

Pierre

. 1998. Acts of Resistance: Against the Tyranny of the Market. Trans. Nice

. Cambridge, UK: Polity.

Burawoy

Michael

. 1979. Manufacturing Consent: Changes in the Labor Process Under Monopoly Capitalism. Chicago: University of Chicago Press.

10.

Burawoy

Michael

. 1985. The Politics of Production: Factory Regimes under Capitalism and Socialism. London: Verso.

11.

Cheung

Gordon W.

Cooper-Thomas

Helena D.

Lau

Rebecca S.

Wang

Linda C.

2021. “Testing Moderation in Business and Psychological Studies with Latent Moderated Structural Equations.” Journal of Business and Psychology 36:1009–1033.

12.

Chun

Jennifer

. 2001. “Flexible Despotism: The Intensification of Insecurity and Uncertainty in the Lives of Silicon Valley’s High-Tech Assembly Workers.” Pp. 127–154 in The Critical Study of Work: Labor, Technology, and Global Production, edited by Baldoz

Kroeber

Kraft

Philadelphia: Temple University Press.

13.

Cowen

Deborah

. 2014. The Deadly Life of Logistics: Mapping Violence in Global Trade. Minneapolis: University of Minnesota Press.

14.

Crowley

Martha

Payne

Julianne

Kennedy

Earl

. 2020. “Manufacturing Discontent: The Labor Process, Job Insecurity and the Making of ‘Good’ and ‘Bad’ Workers.” Pp. 221–47 in Professional Work: Knowledge, Power and Social Inequalities, edited by Gorman

E. H.

Vallas

S. P.

Bingley, UK: Emerald.

15.

Delfanti

Alessandro

. 2021. The Warehouse: Workers and Robots at Amazon. London: Pluto.

16.

Dörflinger

Nadja

Pulignano

Valeria

Vallas

Steven P.

2021. “Production Regimes and Class Compromise Among European Warehouse Workers.” Work and Occupations 48(2):111–45.

17.

Edwards

Paul K.

1986. Conflict at Work: A Materialist Analysis of Workplace Relations. New York: Blackwell.

18.

Edwards

Richard

. 1979. Contested Terrain. New York: Basic Books.

19.

Eidlin

Barry

. 2014. “Class Formation and Class Identity: Birth, Death, and Possibilities for Renewal.” Sociology Compass 8(8):1045–62.

20.

Fantasia

Rick

. 1988. Cultures of Solidarity: Consciousness, Action, and Contemporary American Workers. Berkeley: University of California Press.

21.

Femia

Joseph

. 1975. “Hegemony and Consciousness in the Thought of Antonio Gramsci.” Political Studies 23(1):29–48.

22.

Friedman

Andrew

. 1977. Industry and Labour: Class Struggle at Work and Monopoly Capitalism. London: Macmillan.

23.

Gouldner

Alvin

. 1954. Patterns of Industrial Bureaucracy. New York: Free Press.

24.

Green

Jay

Alcantara

Chris

. 2021. “Amazon Warehouse Workers Suffer Serious Injuries at Higher Rates Than Other Firms.” The Washington Post, June 1. Retrieved November 24, 2023. https://www.washingtonpost.com/technology/2021/06/01/amazon-osha-injury-rate/.

25.

Griesbach

Kathleen

Reich

Adam

Elliott-Negri

Luke

Milkman

Ruth

. 2019. “Algorithmic Control in Platform Food Delivery Work.” Socius. doi:10.1177/2378023119870041.

26.

Gurley

Lauren Kaori

. 2020. “Secret Amazon Reports Expose the Company’s Surveillance of Labor and Environmental Groups.” Vice, November 23. Retrieved November 24, 2023. https://www.vice.com/en/article/5dp3yn/amazon-leaked-reports-expose-spying-warehouse-workerslabor-union-environmental-groups-social-movements.

27.

Gutelius

Beth

Pinto

Sanjay

. 2023. “Painpoints: Data on Work Intensity, Monitoring and Health at Amazon Warehouses.” Chicago: Center for Urban and Economic Development, University of Illinois Chicago. https://cued.uic.edu/wp-content/uploads/sites/219/2023/10/Pain-Points_Final_Oct2023.pdf

28.

Hacker

Jacob

. 2006. The Great Risk Shift. Oxford, UK: Oxford University Press.

29.

Hacker

Jacob

Huber

Gregory

Nichols

Austin

Rehm

Philipp

Schlesinger

Mark

Valletta

Rob

Craig

Stuart

. 2014. “The Economic Security Index: A New Measure for Research and Policy Analysis.” Review of Income and Wealth 60(Suppl.):S5–S32.

30.

Hardering

Fredericke

. 2020. “The Critique of Alienation and the Integration of the Actor’s Perspective: Understanding Alienation in Biographical Identity Work Processes.” Distinktion: Journal of Social Theory 21(1):46–62.

31.

Hatton

Erin

. 2020. Coerced: Work under Threat of Punishment. Oakland: University of California Press.

32.

Henaway

Mostafa

. 2023. “Amazon’s Distribution Space: Constructing a ‘Labour Fix’ through Digital Taylorism and Corporate Keynesianism.” ZFW–Advances in Economic Geography. doi:10.1515/zfw-2022-0017.

33.

Hodson

Randy

. 2001. Dignity at Work. Cambridge, UK: Cambridge University Press.

34.

Hussain

Suhauna

. 2021. “Newsom Signs Bill Taking Aim at Labor Practices at Amazon Warehouses.” The Los Angeles Times, September 22. Retrieved November 24, 2023. https://www.latimes.com/business/technology/story/2021-09-22/newsom-signs-ab-701-amazon-warehouse-algorithm-labor-practices.

35.

Kalleberg

Arne L.

2011. Good Jobs, Bad Jobs: The Rise of Polarized and Precarious Employment Systems in the United States, 1970s to 2000s. New York: Russell Sage.

36.

Kalleberg

Arne L.

2018. Precarious Lives: Job Insecurity and Well-Being in Rich Democracies. New York: John Wiley.

37.

Kantor

Jodi

Wiese

Karen

Ashford

Grace

. 2021. “Inside Amazon’s Employment Machine.” The New York Times, June 15. Retrieved November 24, 2023. https://www.nytimes.com/interactive/2021/06/15/us/amazon-workers.html.

38.

Kenney

Martin

Bearson

Dafna

Zysman

John

. 2021. “The Platform Economy Matures: Measuring Pervasiveness and Exploring Power.” Socio-economic Review 19(4):1451–83.

39.

Kenney

Martin

Zysman

John

. 2016. “The Rise of the Platform Economy.” Issues in Science and Technology 32(3):61.

40.

Knights

David

McCabe

Darren

. 2000. “‘Ain’t Misbehaving’? Opportunities for Resistance under New Forms of ‘Quality’ Management.” Sociology 34(3):421–36.

41.

Kristal

Tali

. 2013. “The Capitalist Machine: Computerization, Workers’ Power, and the Decline in Labor’s Share within US Industries.” American Sociological Review 78(3):361–89.

42.

Kunda

Gideon

. 1992. Engineering Culture. Cambridge, MA: MIT Press.

43.

Lane

Carrie M.

2011. A Company of One: Insecurity, Independence, and the New World of White-Collar Unemployment. Ithaca, NY: Cornell University Press.

44.

Lecher

Colin

. 2019. “How Amazon Automatically Tracks and Fires Warehouse Workers for ‘Productivity.’” The Verge, April 25. Retrieved November 24, 2023. https://www.theverge.com/2019/4/25/18516004/amazon-warehouse-fulfillment-centers-productivity-firingterminations.

45.

Lin

Guan-Chyun

Wen

Zhonglin

Marsh

Herbert W.

Lin

Huey-Shyan

. 2010. “Structural Equation Models of Latent Interactions: Clarification of Orthogonalizing and Double-Mean-Centering Strategies.” Structural Equation Modeling: A Multidisciplinary Journal 17(3):374–91.

46.

Lorey

Isabell

. 2015. State of Insecurity. London: Verso.

47.

Marshall

Gordon

. 1997. Repositioning Class: Social Inequality in Industrial Societies. London: Sage Ltd.

48.

Massimo

Francesco

. 2020. “A Struggle for Bodies and Souls: Amazon Management and Union Strategies in France and Italy.” Pp. 129–45 in The Cost of Free Shipping: Amazon in the Global Economy, edited by Alimahomed-Wilson

Reese

London: Pluto.

49.

Mears

Ashley

. 2015. “Working for Free in the VIP: Relational Work and the Production of Consent.” American Sociological Review 80(6):1099–1122.

50.

Nair

Nisha

Vohra

Neharika

. 2009. “Developing a New Measure of Work Alienation.” Journal of Workplace Rights 14(3):293–309.

51.

Padavic

Irene

. 2005. “Laboring under Uncertainty: Identity Renegotiation among Contingent Workers.” Symbolic Interaction 28(1):111–34.

52.

Pedulla

David S.

2013. “The Hidden Costs of Contingency: Employers’ Use of Contingent Workers and Standard Employees’ Outcomes.” Social Forces 92(2):691–722.

53.

Pieters

Constant

Pieters

Rik

Lemmens

Aurélie

. 2022. “Six Methods for Latent Moderation Analysis in Marketing Research: A Comparison and Guidelines.” Journal of Marketing Research 59(5):941–62.

54.

Pugh

Allison

. 2015. The Tumbleweed Society. New York: Oxford University Press.

55.

Reisel

William D.

Probst

Tahira M.

Chia

Swee-Lim

Maloles

Cesar M.

König

Cornelius J.

2010. “The Effects of Job Insecurity on Job Satisfaction, Organizational Citizenship Behavior, Deviant Behavior, and Negative Emotions of Employees.” International Studies of Management & Organization 40(1):74–91.

56.

Roosevelt

Margot

. 2021. “‘The Algorithm Fired Me’: California Bill Takes on Amazon’s Notorious Work Culture.” The Los Angeles Times, August 31. Retrieved from https://www.latimes.com/business/story/2021-08-31/la-fi-amazon-warehouse-injuries-ab701-bill-calosha.

57.

Rosenberg

Jonathan

. 2023. “There Is No Future for a Labor Movement That Fails to Organize at Amazon.” The Nation, April 24. Retrieved November 24, 2023. https://www.thenation.com/article/activism/labor-union-amazon-organizing/.

58.

Rosenblat

Alex

Stark

Luke

. 2016. “Algorithmic Labor and Information Asymmetries: A Case Study of Uber’s Drivers.” International Journal of Communication 10:27.

59.

Sayers

Sean

. 2011. Marx and Alienation: Essays on Hegelian Themes. New York: Palgrave Macmillan.

60.

Schaupp

Simon

. 2022. “Algorithmic Integration and Precarious (Dis)obedience: On the Co-constitution of Migration Regime and Workplace Regime in Digitalised Manufacturing and Logistics.” Work, Employment and Society 36(2):310–27.

61.

Scheiber

Noam

. 2023. “Amazon Is Everywhere. That’s What Makes It so Vulnerable.” The New York Times, May 19. Retrieved November 24, 2023. https://www.nytimes.com/2023/05/19/business/amazon-union-choke-points.html.

62.

Schneider

Daniel

Harknett

Kristen

. 2019. “Consequences of Routine Work-Schedule Instability for Worker Health and Well-Being.” American Sociological Review 84(1):82–114.

63.

Schneider

Daniel

Harknett

Kristen

. 2022. “What’s to Like? Facebook as a Tool for Survey Data Collection.” Sociological Methods & Research 51(1):108–140.

64.

Schoemann

Alexander M.

Jorgensen

Terrence D.

2021. “Testing and Interpreting Latent Variable Interactions Using the semTools Package.” Psych 3(3):322–35.

65.

Scott

James C.

1990. Domination and the Arts of Resistance: Hidden Transcripts. New Haven, CT: Yale University Press.

66.

Schantz

Amanda

Alfes

Kerstin

Bailey

Catherine

Soane

Emma

. 2015. “Drivers and Outcomes of Work Alienation: Reviving a Concept.” Journal of Management Inquiry 24(4):382–393.

67.

Sharone

Ofer

. 2013. Flawed System/Flawed Self: Job Searching and Unemployment Experiences. Chicago: University of Chicago Press.

68.

Silver

Beverly

. 2003. Forces of Labor: Workers’ Movement and Globalization since 1870. Cambridge, UK: Cambridge University Press.

69.

Soffia

Magdalena

Wood

Alex

Burchell

Brendan

. 2022. “Alienation is Not ‘Bullshit’: An Empirical Critique of Graeber’s Theory of BS Jobs.” Work, Employment and Society 36(5):816–40.

70.

Struna

Jason

Reese

Ellen

. 2020. “Automation and the Surveillance Driven Warehouse in Inland Southern California.” Pp. 86–101 in The Cost of Free Shipping: Amazon in the Global Economy, edited by Alimahomed-Wilson

Reese

London: Pluto.

71.

Taylor

Phil

Bain

Peter

. 1999. “‘An Assembly Line in the Head’: Work and Employee Relations in the Call Centre.” Industrial Relations Journal 30(2):101–17.

72.

Thompson

Paul

Ackroyd

Stephen

. 1995. “All Quiet on the Workplace Front? A Critique of Recent Trends in British Industrial Sociology.” Sociology 29(4):615–33.

73.

Tracy

Sarah J.

2019. Qualitative Research Methods: Collecting Evidence, Crafting Analysis, Communicating Impact. 2nd ed. Hoboken, NJ: John Wiley.

74.

Vallas

Steven P.

1993. Power in the Workplace: The Politics of Production at AT&T. Albany: State University of New York Press.

75.

Vallas

Steven P.

2019. “Platform Capitalism: What’s at Stake for Workers?” New Labor Forum 28(1):48–59.

76.

Vallas

Steven P.

Johnston

Hannah

Mommadova

Yana

. 2022. “Prime Suspect: Mechanisms of Labor Control at Amazon’s Warehouses.” Work and Occupations 49(4):421–56.

77.

Vallas

Steven P.

Prener

Christopher

. 2012. “Dualism, Job Polarization, and the Social Construction of Precarious Work.” Work and Occupations 39(4):331–53.

78.

Vallas

Steven P.

Schor

Juliet B.

2020. “What Do Platforms Do? Understanding the Gig Economy.” Annual Review of Sociology 46:273–94.

79.

Vidal

Matt

. 2022. Management Divided: Contradictions of Labor Management. Oxford, UK: Oxford University Press.

80.

Weil

David

. 2014. The Fissured Workplace. Cambridge, MA: Harvard University Press.

81.

Wiggin

Teke

. 2023. “Weaponizing the Workplace: Algorithmic Management, ‘Worker Captivity’ and Amazon’s Anti-union Campaign in Bessemer, Alabama.” Paper presented at the Annual Meetings of the Labor and Employment Relations Association, Detroit.

82.

Wood

Alex

. 2020. Despotism on Demand. Ithaca, NY: Cornell University Press.

83.

Wright

Erik O.

1985. Classes. London: Verso.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

Coercion,Consent,and Class Consciousness: How Workers Respond to Amazon’s Production Regime

Abstract

Keywords

Approaching Amazon

The Work/Consciousness Nexus: Digital Coercion and Worker Precarity

Labor Process Theory: Digital Surveillance, Managerial Coercion, and Oppositional Orientations

Sanctuary Thesis: Financial Strain and Oppositional Class Orientation

Data and Methods

Data Collection and Sampling

Independent Variables

Digital Surveillance

Managerial Coercion

Alienation from Work

Precarity

Dependent Variables

Control Variables

Analytic Strategy

Results

How Does the Labor Process Shape Class Orientation?

Does Economic Precarity Mute the Effect of Job Conditions on Class Orientation?

Robustness Checks

Conclusions

Supplemental Material

sj-docx-1-srd-10.1177_23780231231216286 – Supplemental material for Coercion, Consent, and Class Consciousness: How Workers Respond to Amazon’s Production Regime

Footnotes

Acknowledgements

Funding

ORCID iD

Supplemental Material

1

2

Author Biographies

References

Supplementary Material