CoChat
Sign Up
Cited by 512Digital Economy and Work TransformationLegal and Labor StudiesJob Satisfaction and Organizational Behavior

Algorithms as work designers: How algorithmic management influences the design of jobs

Open access
Parent-Rocheleau, Xavier, Parker, Sharon K.·2022·Human Resource Management Review
DOI 10.1016/j.hrmr.2021.100838
Open this paper in CoChat — free accountOpen PDF

Abstract

Abstract We review the literature on algorithmic management (AM) to bridge the gap between this emerging research area and the well-established theory and research on work design. First, we identify six management functions that algorithms are currently able to perform: monitoring, goal setting, performance management, scheduling, compensation, and job termination. Second, we show how each AM function affects key job resources (e.g., job autonomy, job complexity) and key job demands (e.g., workload, physical demands); with each of these resources and demands being important drivers of worker motivation and their well-being. Third, rejecting a deterministic perspective and drawing on sociotechnical systems theory, we outline key categories of variables that moderate the link between AM on work design, namely transparency, fairness and human influence (e.g., whether workers can control the system). We summarize our review in the form of a model to help guide research on AM, and to support practitioners and designers in the creation and maintenance of meaningful jobs in the era of algorithms.

Integrity signals

Registry

Crossref · 10.1016/j.hrmr.2021.100838

Cites

183 of 183 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

69 of 269 cited sources checked · 14 supported · 1 flagged · accuracy not yet measured

  • Findingcitation

    In gig work, the algorithmic task allocation process is highly contingent on demand or customer satisfaction scores.

    “In gig work, the algorithmic task allocation process is highly contingent on demand or customer satisfaction scores (Gerber & Krzywdzinski, 2019;Gregory, 2020;Griesbach et al., 2019).”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Gerber & Krzywdzinski, 2019 source

      The Microtask platforms section (pp. 21–22) says algorithmically organised rankings help allocate tasks and draw on ratings of completed jobs, including client ratings. The paper does not establish that allocation is contingent on demand, and it describes other factors besides customer ratings, including experience and activity (p. 21; Discussion and Conclusions, p. 33). Its findings concern crowdwork platforms, not gig work generally. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Gregory, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Griesbach et al., 2019 source
  • Interpretationcitation

    Descriptive algorithms are likely to multiply the metrics available to quantify employees' actions, emotions, performance, behaviors, attitudes, physical state, and customer-related information such as satisfaction and traffic.

    “Descriptive algorithms are likely to reinvent the world of performance management (PM) in multiplying the metrics available to precisely quantify various facets of employees' actions, emotions, performance, behaviors, attitudes and physical state (Angrave et al., 2016;Tambe et al., 2019), but also customer-related information such as satisfaction and traffic (Evans & Kitchin, 2018;Levy & Barocas,”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites Angrave et al., 2016 source

      In ‘Defining big data and HR analytics’ (pp. 5–6), the paper describes existing employee performance and attitude data and the possibility of using digital records to infer mood, morale and activity. It does not establish that descriptive algorithms will precisely quantify employees’ physical state or reinvent performance management; it also says the means of integrating and analysing the larger data sources are not yet well established. The customer-related clause is attached to other citations, not Angrave et al. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Tambe et al., 2019 source
    • cites Evans & Kitchin, 2018 source

      The Conclusion (page unavailable) says customer satisfaction is delegated to big-data and algorithmic processing. The section “Non-coded activities and symbolic labour” (page unavailable) describes a running count of customers passing through the self-scan area, supporting the traffic example. The paper does not establish that descriptive algorithms will multiply metrics that precisely quantify customer satisfaction and traffic. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
  • Premisecitation

    Job resources and job demands are important drivers of workers' motivation, well-being, and performance.

    “with each of these resources and demands being important drivers of worker motivation and their well-being.”
    • Also stated in: § THE EFFECT OF ALGORITHMIC MANAGEMENT ON WORK DESIGN · boosted
      “Considerable evidence links job resources and job demands with workers' motivation, well-being and performance (for meta-analyses or reviews, see: Humphrey et al., 2007;Knight & Parker, 2021;Parker & Ohly, 2008).”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Humphrey et al., 2007 source
    • cites Knight & Parker, 2021 source

      In “Work design theories and performance” (page unavailable), the paper says job characteristics affect motivation and performance. Its discussion of the Job-Demands Resources model (page unavailable) links resources and demands to strain, well-being, and subsequent performance. It does not establish the citing sentence’s more specific implication that demands, as well as resources, drive worker motivation. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites Parker & Ohly, 2008 source
  • Findingcitation

    The impact of using algorithmic management to set goals on the reduction of workers' autonomy has been outlined by many scholars, namely in gig work where autonomy in task selection is often viewed by externals as a major advantage of platform-based work.

    “The impact of the use of AM to set goals on the reduction of workers' autonomy has been outlined by many scholars, namely in gig-work where autonomy in task selection is often viewed by externals as a major advantage of platform-based work (Curchod, Patriotta, Cohen, & Neysen, 2019;Gerber & Krzywdzinski, 2019;Griesbach et al., 2019;Rosenblat, 2018;Shapiro, 2018).”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites Curchod et al., 2019 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites Gerber & Krzywdzinski, 2019 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites Griesbach et al., 2019 source
  • Findingcitation

    AI-driven algorithmic monitoring allows the collection and recording of a vast array of new information and metrics about workers.

    “AI-driven algorithmic monitoring thus allows to collect and record a vast array of new information and metrics, such as emotions (Chan, 2018;Pinheiro, Ramos, Donizete, Picanço, & De Oliveira, 2017), movements, sleep time, physical and health condition, social media activity, internet browser history (Angrave, Charlwood, Kirkpatrick, Lawrence, & Stuart, 2016;Leicht-Deobald et al., 2019)”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites Chan, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Angrave et al., 2016 source

      In “Defining big data and HR analytics” (pp. 5–6), the paper describes worker data already held in HR systems and the potential to combine it with location, communications and internet-browsing data. This supports the broad point that analytics can draw on many worker metrics, including movement and browsing history. It does not identify sleep time or physical health as monitored metrics, and presents integration of the newer data sources as a possibility rather than an established capability. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Leicht-Deobald et al., 2019 source

      In “A Brief History of Algorithm-Based HR Decision-Making” (page unavailable), the paper describes monitoring location, health and fitness data, Facebook activity, and browser histories, supporting the claim that algorithmic tools can collect varied new information about workers. It does not identify sleep time as a monitored metric, so that item in the citing sentence is unsupported by our copy. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Classified as probably not a substantial claim. 54
  • May combine two claims. 45

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic scheduling is negatively associated with social support.

    “Social support (-)”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Uhde et al. (2020) source

      In “Data souvereignity and transparency” (page unavailable), the paper says that algorithmic scheduling systems, if designed correctly, can facilitate pro-social behavior and may benefit team climate. It does not report a negative association with social support; its discussion points in the opposite direction. (LLM, conf=0.78)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingcitation

    Narrow algorithmic monitoring systems lead to a greater emphasis on quantified work objectives at the expense of other and potentially meaningful aspects of one's job, reducing autonomy in the choice of work methods.

    “A similar consequence of narrow algorithmic monitoring systems is a greater emphasis on quantified work objectives at the expense of other and potentially meaningful aspects of one's job (Schafheitle et al., 2020), reducing the autonomy of workers in the choice of their work methods (Day, Paquet, Scott, & Hambley, 2012;Wang, Liu, & Parker, 2020)”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites Schafheitle et al., 2020 source

      In “Framework on Datafication Technology Control Configurations” (page not provided), the paper describes technology that tracks employee outputs against performance targets. It does not say that narrow monitoring increases emphasis on quantified objectives at the expense of other meaningful aspects of work. The autonomy clause is cited to different sources, so it is not judged here. (LLM, conf=0.82)

      Experimental — accuracy not yet measured
    • cites Day et al., 2012 source
    • cites Wang et al., 2020 source

  • May combine two claims.

    Findingcitation

    Algorithmic task allocation is largely found in the gig economy and platform work, and in traditional work sectors like telecommunication, electronics, public transports, and parcel delivery.

    “Algorithmic task allocation is largely found in the gig economy and platform work (Gerber & Krzywdzinski, 2019;Lee et al., 2015;Rosenblat, 2018), and in traditional work sectors like telecommunication (Leclercq-Vandelannoitte, 2017), electronics (Gershgorn, 2015;Hitachi, 2015), public transports (Hodson, 2014), and parcel delivery (Stevens, 2015).”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Gerber & Krzywdzinski, 2019 source
    • cites Lee et al., 2015 source
    • cites Rosenblat, 2018

  • May combine two claims.

    Premisecitation

    Algorithmic management is a technological phenomenon that is not contingent on the type of organization and can be found in a large array of work contexts.

    “However, recent reviews and conceptualizations (Kellogg et al., 2019;Mateescu & Nguyen, 2019a;Schafheitle et al., 2020) consider algorithmic management as a technological phenomenon that is not contingent on the type of organization and can be found in a large array of work contexts.”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Kellogg et al., 2019 source
    • cites Mateescu & Nguyen, 2019a

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Schafheitle et al., 2020 source

      The Introduction (page unavailable in our copy) presents datafication technologies as applicable to both easily automated tasks and “thinking work,” supporting the claim that technology-based control spans varied work contexts. The paper discusses algorithmic decision-making in organizational control, but the passages do not establish that algorithmic management is independent of organization type. (LLM, conf=0.78)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingcitation

    Because data collected through monitoring devices is often used to assess performance, workers sometimes tend to work for data, focusing more on monitored tasks and eluding unmonitored ones, which can hamper task variety.

    “Because data collected through monitoring devices is often used to assess their performance, workers sometimes tend to”
    Quote in the paperSubstantial claim2 of 3 sources checked
    • cites Evans & Kitchin, 2018 source

      The ‘Working for data’ section says PDA-directed work is measured through system data and leaves little room for customer interaction. ‘Non-coded activities and symbolic labour’ says customer service is not captured in performance metrics, but it does not show workers choosing to evade unmonitored tasks; it says pickers must assist customers when asked. The claimed reduction in task variety is therefore supported only indirectly. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites Schafheitle et al., 2020 source

      The Framework on Datafication Technology Control Configurations section describes sensor and GPS data linked to individual performance management at UPS; the ORION discussion also considers how those data might be used to rank drivers. Neither passage says workers focus on monitored tasks, evade unmonitored ones, or experience reduced task variety. Page numbers are unavailable in our copy. (LLM, conf=0.83)

      Experimental — accuracy not yet measured
    • cites Tomczak et al., 2018 source

  • May combine two claims.

    Premisecitation

    AI algorithms differ from previous automated decision-making devices in their high level of autonomy, self-learning capacities, potential for interconnectedness, and ability to handle massive and heterogeneous data.

    “In sum, AI algorithms, largely powered by machine learning, differ from previous automated decision-making devices in their high level of autonomy (requiring little human input), their self-learning capacities, their potential for interconnectedness, and their ability to handle massive and heterogeneous data in order to perform description, prediction, and prescription tasks (Makarius et al., 2020”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites Makarius et al., 2020 source

      Section 2 contrasts AI with earlier task automation and identifies learning and decision-making as distinguishing features. Section 3 describes autonomous systems that can learn from data without human input, while Section 5 discusses big-data inputs and data sharing. The copy does not establish the claimed ability to handle heterogeneous data or the full description–prediction–prescription task range; page numbers were not available. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Murray et al., 2020 source

      The ‘Conjoined Agency with Automating Technologies’ section (page unavailable) says these technologies process large amounts of unstructured data, learn continuously, and select actions without human intervention, supporting most of the comparison. The cited paper does not establish interconnectedness as a distinguishing feature or present description, prediction, and prescription as a set of tasks. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Limitation of workers' voice in scheduling can occur through direct scheduling decisions or through scheduling nudging.

    “This can occur through direct scheduling decisions (Uhde, Schlicker, Wallach, & Hassenzahl, 2020), or through scheduling nudging, particularly common in gig work (Gal et al., 2020;Mateescu & Nguyen, 2019a).”
    Quote in the paperSubstantial claim1 of 3 sources checked
    • cites Uhde, Schlicker, Wallach, & Hassenzahl, 2020 source
    • cites Gal et al., 2020 source
    • cites Mateescu & Nguyen, 2019a

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation

    Customer satisfaction ratings are questioned for their volatility, reliability, and validity.

    “The strong impact of these ratings, largely questioned about their volatility, the reliability and validity (Gerber & Krzywdzinski, 2019;Greenwood, Adjerid, & Angst, 2019;Luca & Zervas, 2016;Orlikowski & Scott, 2013;Rosenblat, Levy, Barocas, & Hwang, 2017), has been associated with workers' perception of job insecurity (Curchod et al., 2019;Veen et al., 2019).”
    Quote in the paperSubstantial claim3 of 3 sources checked
    • cites Gerber & Krzywdzinski, 2019 source
    • cites Greenwood, Adjerid, & Angst, 2019 source

      The Related Literature section (page unavailable) discusses selection and exaggeration biases in online ratings, and the Results section finds that female drivers can receive a larger rating penalty than male drivers for a low-quality experience. These findings support questioning whether ratings accurately reflect performance, but the paper does not establish that customer satisfaction ratings are volatile or generally unreliable. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Luca & Zervas, 2016 source

      Section 2.2, “Fake and Filtered Reviews” (page unavailable), says fraudulent and selectively solicited Yelp reviews can produce a positively biased review sample, supporting concern about the reliability and validity of those ratings. The paper does not establish that customer satisfaction ratings generally are volatile, nor does it study ratings of workers. (LLM, conf=0.86)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Sociotechnical capital can only be optimized when AI technologies and employees act as a tightly coupled system exhibiting increased responsiveness.

    “These authors claim that sociotechnical capital can only be optimized when AI technologies and employees act as a tightly coupled system exhibiting increased responsiveness.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Makarius et al., 2020 source

      Section 3 says sociotechnical capital is likely to result when AI and employees are successfully integrated as a tightly coupled, responsive system. But it also says loosely coupled AI and employees can produce moderate sociotechnical capital. The source therefore does not support the claim that tight coupling is the only way to optimize it. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Premisecitation

    Digitalization and automation can contribute to stimulating and healthy jobs if their implementation is designed to increase resources and reduce demands, and if people are in control and craft their use of the system.

    “Consistent with this idea, one of the authors of the JD-R model stated that”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Demerouti, 2020 source

      The “(Negative) Consequences and How to Avoid Them” section (page not supplied) says digitalization and automation can contribute to stimulating, healthy jobs when people can control and craft their use and job resources are maximized while demands remain affordable. The claim omits two further conditions the paper includes: sharing economic growth with employees and protecting employees and employment conditions. (LLM, conf=0.83)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Both labor needs and supply in algorithmic scheduling rely on algorithmic estimation.

    “Both needs and supply rely on algorithmic estimation, with the precise prediction of needs in workforce based on information such as expected or real-time customer traffic, deadlines, real-time monitoring of fluctuating demand, weather forecasts, or on previous occupancy at the same date (Cheng, Rao, Jiang, & Zhou, 2015;Hoshino, Slobodin, & Bernoudy, 2018).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Cheng, Rao, Jiang, & Zhou, 2015 source
    • cites Hoshino, Slobodin, & Bernoudy, 2018 source

      The Mathematical Model (p. 3) uses employee availability and minimum and maximum staffing requirements as inputs to an algorithm that optimizes schedules. The Next Steps section (p. 6) says managers gather employee preferences and enter them into the program. The paper therefore supports algorithmic scheduling using information about needs and supply, but not the claim that the algorithm estimates both. (LLM, conf=0.79)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The fast pace of recent technological progress has created a disequilibrium between social and technical systems, such that technical systems and humans are often not jointly optimized.

    “Some scholars have begun to apply STS theory to AI and the digitalization of work (Makarius et al., 2020;Parker & Grote, 2020;Shin & Park, 2019;Winby & Mohrman, 2018), stating that the fast pace of the recent technological progress has created an disequilibrium between the two components, such that technical systems and humans are often not jointly optimized.”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Makarius et al., 2020 source

      The Introduction (page unavailable) says STS theory focuses on jointly optimizing an organization’s human and technological dimensions. Section 2.2 (page unavailable) says few organizations have achieved their AI aspirations at scale. The paper does not establish that rapid technological progress has caused a general disequilibrium between social and technical systems, or that they are often not jointly optimized. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Parker & Grote, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Shin & Park, 2019 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Providing transparency in the data (data visibility) could conflict with the need to preserve the privacy of the data.

    “One of the advocated solutions to this issue is to provide transparency in the data (i.e., data visibility), which could however conflict with the need to preserve the privacy of the data (Stohl et al., 2016;Zuboff, 2015).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Stohl et al., 2016

      The Relationship Between Information Visibility and Transparency section links information visibility to transparency, and Legal obligation describes proprietary data being made public. But the passages do not discuss a conflict between data visibility and preserving data privacy; proprietary information is not necessarily private data in the sense claimed. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Zuboff, 2015 source

  • May combine two claims.

    Premisecitation

    Transparency in the decisions of very complex machine learning algorithms with limited room for explainability is very difficult (and perhaps impossible) to achieve, and if achieved could be very demanding in terms of cognitive load for workers.

    “transparency in the decisions of very complex machine learning algorithms with limited room for explainability is very difficult (and perhaps impossible) to achieve, and, if achieved, could be very demanding in terms of cognitive load for workers (Gal et al., 2020;Stohl, Stohl, & Leonardi, 2016).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Gal et al., 2020 source
    • cites Stohl, Stohl, & Leonardi, 2016

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Instead of hampering social support between nurses, the easiness to use the system to swap shifts in order to accommodate coworkers triggered a positive team climate and a reduction of interpersonal conflicts, increasing social support.

    “Their findings indicate that, instead of hampering social support between nurses, the easiness to use the system to swap shifts in order to accommodate coworkers triggered a positive team climate and a reduction of interpersonal conflicts, increasing social support.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites Uhde et al., 2020 source

      In “Data souvereignity and transparency” (page not provided), the paper suggests that a scheduling system could facilitate pro-social behavior and may benefit team climate. It does not report that an easy-to-use system actually reduced interpersonal conflicts or increased social support. “Cultivate a positive team spirit” instead reports that existing respect and empathy increased willingness to swap shifts. (LLM, conf=0.88)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic scheduling is positively associated with job insecurity.

    “Job insecurity (+)”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Van Oort (2019) source
    • cites Moore and Hayes (2017) source

      The paper’s discussion of electronic monitoring and zero-hours contracts describes work-on-demand scheduling and uncertainty about paid time. Its ‘Electronic Monitoring – rationale and reality’ section describes a system that checks visits against a roster, but does not report an association between algorithmic scheduling and job insecurity. I searched for job insecurity, employment security, and scheduling in our copy; no passage establishes the claimed relationship. (LLM, conf=0.75)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic goal setting is positively associated with job insecurity.

    “Job insecurity (+)”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Griesbach et al. (2019) source
    • cites Lehdonvirta (2018) source
    • cites Wood et al. (2019) source

      In Findings and discussion, the paper says algorithmic ratings and rankings can filter work away from lower-rated workers, and Worker power and competition reports that workers lacked formal job security and feared replacement. It does not identify algorithmic goal setting as a practice or report a positive association between goal setting and job insecurity. (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic compensation is negatively associated with job autonomy.

    “Autonomy (-)”
    Quote in the paperNot a substantial claim1 of 3 sources checked
    • cites Moore and Hayes (2017, 2018)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Griesbach et al. (2019) source
    • cites Wood et al. (2019) source

      The Findings and discussion section describes algorithmic management as affording remote gig workers significant autonomy and discretion. A search for compensation, pay, and wages found discussion of earnings and platform ratings, but no assessment of whether algorithmic compensation is negatively associated with job autonomy. (LLM, conf=0.76)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management is positively associated with workload.

    “Workload (+)”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Levy (2015) source
    • cites Bakewell et al. (2018) source

  • Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

    Premisecitation

    Most authors consider that only the advent of strong artificial intelligence (general AI), involving fully agentic and autonomous AI algorithms with potentially greater capabilities than human beings, could reasonably lead one to adopt a strict determinism perspective.

    “Most authors consider that only the advent of strong artificial intelligence (also called general AI), involving fully agentic and autonomous AI algorithms with potentially greater capabilities than human beings, could reasonably lead one to adopt a strict determinism perspective(Braga & Logan, 2017;Clifton, Glasmeier, & Gray, 2020).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Braga & Logan, 2017 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites Clifton, Glasmeier, & Gray, 2020 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management is positively associated with job insecurity.

    “Job insecurity (+)”
    Quote in the paperNot a substantial claim2 of 3 sources checked
    • cites Veen et al. (2019) source
    • cites Wood et al. (2019) source

      In ‘Autonomy in the shadow of algorithmic management’ (page unavailable), the paper says algorithmic ratings can filter work away from lower-rated workers, making a living on the platform less viable. ‘Worker power and competition’ (page unavailable) reports job insecurity, but also attributes it to competition and the absence of formal job security. The paper does not report a measured positive association between algorithmic performance management and job insecurity. (LLM, conf=0.84)

      Experimental — accuracy not yet measured
    • cites Evans and Kitchin (2018) source

      The Introduction (page unavailable) links algorithm-governed work to precarious employment, which supports a general connection with insecurity. It does not establish a positive association specifically between algorithmic performance management and job insecurity; its discussion of performance monitoring addresses control of work rather than that outcome. (LLM, conf=0.78)

      Experimental — accuracy not yet measured

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been associated with a reduction of workers' autonomy.

    “In particular, AM has been associated with a reduction of workers' autonomy (Jabagi, Croteau, Audebrand, & Marsan, 2019;Mohlmann & Zalmanson, 2017;Toyoda, Lucas, & Gratch, 2020)”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Jabagi et al., 2019 source
    • cites Mohlmann & Zalmanson, 2017
    • cites Toyoda et al., 2020 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in low trust.

    “These consequences potentially explain why AM has also been shown to result in in negative emotions, unfairness perceptions, low trust (Lee, 2018;Zarsky, 2016)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Lee, 2018 source
    • cites Zarsky, 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in negative emotions.

    “These consequences potentially explain why AM has also been shown to result in in negative emotions, unfairness perceptions, low trust (Lee, 2018;Zarsky, 2016)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Lee, 2018 source
    • cites Zarsky, 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in unfairness perceptions.

    “These consequences potentially explain why AM has also been shown to result in in negative emotions, unfairness perceptions, low trust (Lee, 2018;Zarsky, 2016)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Lee, 2018 source
    • cites Zarsky, 2016 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been associated with the creation of power asymmetry in information availability.

    “and the creation of power asymmetry in information availability (Calo & Rosenblat, 2017;Jarrahi & Sutherland, 2018;Rosenblat & Stark, 2016;Shapiro, 2018).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Calo & Rosenblat, 2017

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Jarrahi & Sutherland, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Rosenblat & Stark, 2016

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    In our review, we identified literature about algorithmic management both in the gig economy and in other sectors.

    “In our review, we similarly identified literature about algorithmic management both in the gig economy and in other sectors.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Findingnone given

    The emerging literature on algorithmic management is spread across different streams of research and conceptual approaches, and has mainly examined one algorithmic function at a time, lacking theoretical integration.

    “First, the emerging literature on the topic is spread across different streams of research and conceptual approaches, and has mainly examined one algorithmic function at the time, lacking theoretical integration.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in reduced engagement in work.

    “reduced engagement in work (Bucher, Fieseler, & Lutz, 2019)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bucher et al., 2019 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in psychological contract breach.

    “psychological contract breach (Tomprou & Lee, 2019)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Tomprou & Lee, 2019

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Findingcitation

    Reyes (2018) reports that an algorithmic system used to dispatch room cleaning assignments based on instant clients' arrival time has taken away housekeepers' autonomy to organize their day and made their jobs more physically demanding.

    “Reyes (2018) namely reports how an algorithmic system used to dispatch room cleaning assignments based on instant clients' arrival time has taken away housekeepers' autonomy to organize their day and made their jobs more physically demanding”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Reyes, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in lower life satisfaction.

    “lower life satisfaction (Keith, Harms, & Tay, 2019)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Keith et al., 2019 source

  • Classified as probably not a substantial claim. May combine two claims.

    Findingcitation

    Monitoring of numerous features of employees' attitudes, behaviors, performance, and emotions has largely been viewed as a pervasive form of surveillance, focused on controlling employees, which can correspond to a stress-related emotional demand.

    “Such monitoring of numerous features of employees' attitudes, behaviors, performance, and emotions has largely been viewed as a pervasive form of surveillance, focused on controlling employees (Mateescu & Nguyen, 2019b), which can correspond to a stress-related emotional demand.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mateescu & Nguyen, 2019b

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingcitation

    Algorithmic management has been shown to result in low job satisfaction.

    “low job satisfaction (Brawley & Pury, 2016;Griesbach, Reich, Elliott-Negri, & Milkman, 2019)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Brawley & Pury, 2016 source
    • cites Griesbach et al., 2019 source

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Findingnone given

    Existing research often seems to be based on deterministic assumptions about the consequences of technology, implying pre-determined outcomes.

    “Second, existing research often seems to be based on deterministic assumptions about the consequences of technology, implying pre-determined outcomes.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Findingnone given

    Research on the role, operationalization, and interconnectedness of the proposed constructs is nascent and warrants further development.

    “In sum, research of the role and on the operationalization and interconnectedness of these constructs is nascent and warrants further development.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    The impact of algorithmic management on workers' autonomy is consistently negative across the six functions, while the effect of each AM function on other job resources and demands is highly variable.

    “while the impact on workers' autonomy is consistently negative across the six functions, the effect of each AM function on other job resources and demands is highly variable, suggesting the importance of considering functions separately.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Algorithmic management has six actual functions that can be examined as distinct components.

  • Classified as probably not a substantial claim.

    Findingreasoning

    There is a lack of intervention research that examines whether proactively making changes to real algorithmic management systems positively impacts work design and outcomes.

    “there is a lack of intervention research that actually examines whether making these sorts of changes to real systems in a proactive way indeed positively impacts work design and outcomes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingreasoning

    The evidence reviewed highlights the influence of algorithmic management on work design characteristics, with most of this influence appearing to be negative.

    “the evidence reviewed clearly highlights the influence of AM on work design characteristics, with most of this influence appearing to be negative.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Algorithmic management has six actual functions that can be examined as distinct components.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingreasoning

    Several further outcomes of work design could be included in the model, but the evidence base currently does not exist for them.

    “several further outcomes of work design could be included in the model, yet we have not done so as the evidence base currently does not exist.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    Algorithmic performance management and scheduling have more salient consequences on job demands than job resources, whereas for the other functions there is more evidence of effects on job resources.

    “algorithmic performance management (PM) and scheduling seem to have more salient consequences on job demands than job resources, whereas for the other functions, there is more evidence of effects on job resources.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Algorithmic management has six actual functions that can be examined as distinct components.

  • Classified as probably not a substantial claim. May combine two claims.

    Findingreasoning

    To date, the evidence for the effect of the proposed socio-technical moderators on the link between algorithmic management and work design is limited, and the paper's propositions are largely based on extrapolating from other fields of research.

    “To date, the evidence for the effect of these elements is limited, and our propositions are largely based on extrapolating from other fields of research (computing and data science, namely).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingnone given

    Algorithmic performance management is the function with the most evidence for a bright side, with a potential improvement of role clarity and feedback from the job.

    “PM is also the function for which there is the most evidence for a bright side, with a potential improvement of role clarity and feedback from the job.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Algorithmic management has six actual functions that can be examined as distinct components.

  • May combine two claims.

    Interpretationcitation

    Monitoring systems used to monitor and optimize warehouse workers' hands and arms movements or track parcel delivery features are likely to simplify workers' jobs or solve problems for workers, hampering job complexity and problem-solving opportunities.

    “Such systems used, for instance, to monitor and optimize warehouse workers hands and arms movements (Engler, 2018;Yeginsu, 2018), or to track different features of parcel delivery process such as the geolocation, the cleanliness of the trucks (Woyke, 2018) and the weight and size of each package (Rosenbush, 2018) are likely to simplify workers jobs or solving problems for workers”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Engler, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Yeginsu, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Woyke, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim.

    Interpretationcitation

    Existing research on algorithmic management tends to suggest that it generates more negative than positive outcomes for workers.

    “Existing research on algorithmic management tends to suggest that it generates more negative than positive outcomes for workers.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    Three critical and potentially inter-connected parameters function as socio-technical moderators of the link between algorithmic management and its impact on work design: system transparency, system fairness, and human influence.

    “More specifically, as depicted in our model, we identified three critical and potentially inter-connected such parameters, which function as socio-technical moderators of the link between AM and its impact on work design: System transparency, system fairness, and human influence.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    Algorithmic scheduling may facilitate the implementation of an organizational strategy towards zero-hours contracts, resulting in lower job stability.

    “Similarly, algorithmic scheduling may facilitate the implementation of an organizational strategy towards zero-hours contract (Fleming, 2017;Moore & Hayes, 2017, 2018), resulting in lower job stability.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Fleming, 2017 source
    • cites Moore & Hayes, 2017, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationcitation

    Negative reactions to algorithmic feedback could yield confusion about employer or platform expectations and reduce the quality of feedback and role clarity.

    “which could, rather yield confusion about the employer's or platform's expectations towards employees, reduce the quality of feedback and role clarity.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationnone given

    The ability of algorithms to decide, implement, or facilitate job termination makes it easier for employers to proceed to job termination, which could potentially influence job insecurity perceptions.

    “Overall, the ability of algorithms to decide, implement, or facilitate job termination makes it easier for employers to proceed to job termination, which could potentially influence job insecurity perceptions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Performance-based compensation lowers the perception of autonomy and is more likely to generate a controlled and extrinsic type of motivation.

    “In orienting efforts towards activities that will be financially rewarded, performance-based compensation lowers the perception of autonomy and is more likely to generate a controlled and extrinsic type of motivation (Gagné & Forest, 2008).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Gagné & Forest, 2008 source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    Companies' objectives in delegating scheduling to algorithms is generally to create on-demand and just-in-time schedules quickly and automatically.

    “The companies' objectives in delegating scheduling to algorithms is generally to create on-demand and justin-time schedules quickly and automatically.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The use of algorithmic management has consequences on workers, but these effects can be influenced and managed by decisions of stakeholders in organizations or institutions, as well as to some extent by actions of individuals themselves.

    “Our moderate voluntaristic point of view leads us to postulate that the use of algorithmic management has consequences on workers, but that these effects can be influenced and managed by the decisions of stakeholders in organizations or institutions, as well as (to some extent) by the actions of individuals themselves.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Greater use of customer monitoring and data in worker management requires workers to display exaggeratedly and sustained positive emotions in online interactions with customers, increasing the risk of emotional demands.

    “Literature also suggest that, like the call centers' emotional state monitoring described above and involving continuous requests of rapid emotional shifts, the greater use of customer monitoring and data in workers management require them to display exaggeratedly and sustained positive emotions in their online interactions with customers, increasing the risk of emotional demands (Van Oort, 2019).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Van Oort, 2019 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Organizations perform better when their design jointly optimizes the social systems and the technical system.

    “Our research was guided by sociotechnical systems theory (STS : Cherns, 1976;Emery & Trist, 1978), which proposes that organizations perform better when their design jointly optimizes the social systems (i.e., the humans within and around the organization) and the technical system (i.e., technologies and mechanics supporting work processes) (Cherns, 1976;Emery & Trist, 1978).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Cherns, 1976 source
    • cites Emery & Trist, 1978

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Premisenone given

    An emerging base for bonus or pay raise allocation is the algorithmic prediction of hypothetical future performance.

    “Furthermore, an emerging base for bonus or pay raise allocation is the algorithmic prediction of hypothetical future performance.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Gig food deliverers feel obligated to follow scheduling nudges displayed by the app and to sometimes work in adverse weather or very difficult conditions because they don't know to what extent compliance is reflected in performance scores or future assignments.

    “In the same vein, Gregory (2020) showed that gig food deliverers feel obligated to follow the scheduling nudges displayed by the app and to sometimes work in adverse weather or very difficult conditions (physical demands), mainly because they don't know to what extent compliance to these incentives is reflected in performance scores or in future assignments.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Gregory, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Premisecitation

    Algorithmic compensation encourages workers to internalize a logic of efficiency and productivity, infusing an ideal of hyper-meritocratic justice, which could result in the embracement of higher workloads.

    “Another study observed that algorithmic compensation encourages the workers”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Galière, 2020 source

  • May combine two claims.

    Interpretationreasoning

    The above discussion suggests that outcomes for work design are principally negative, with algorithmic management tending to lower job resources and increase job demands.

    “The above discussion of AM functions and their consequences suggests that the outcomes for work design are principally negative, with AM tending to lower job resources and to increase job demands.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Algorithm-based pay decisions may be perceived as reductionistic, and rewarding quantified, unsignificant and decontextualized aspects might alter task significance.

    “The study by Newman et al. (2020) also showed that algorithm-based pay decisions made be perceived as reductionistic, and that such rewarding of quantified, unsignificant an decontextualized aspect might alter task significance.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Newman et al., 2020 source

  • May combine two claims.

    Interpretationreasoning

    Algorithmic performance management is likely to augment the amount of feedback from the job and align it more precisely on employees' actual performance.

    “In terms of work design, algorithmic PM is, on the one hand, likely to augment the”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    Algorithms can make the decision to terminate employment and automatically notify the employee.

    “First, algorithms can make the decision to terminate the employment of a worker and automatically notify the employee.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    Because algorithmic metrics are generally easy to understand and straightforward, they may increase clarity of expectations and foster role clarity.

    “Because these metrics are generally easy to understand and straightforward (see Rosenblat & Stark, 2016, pp. 73-74 for examples of platform feedback notification), they also may increase the clarity of the expectations toward them, fostering role clarity.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenblat & Stark, 2016

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning

    Algorithmic management could either enhance or hamper the achievement of quality work design, and hence impact individual outcomes and firm productivity, depending on the interplay of technical characteristics and human workers.

    “Following this statement, and in accordance with the postulates of STS theory, the logic of our model is that algorithmic management could either enhance or hamper the achievement of quality work design, and hence impact individual outcomes and indeed firm productivity (or sociotechnical capital), depending on the interplay of technical characteristics and human workers.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationcitation

    Algorithmic compensation might contribute to reducing workers' autonomy by constraining them to work longer hours and decreasing their control over rewards.

    “In terms of work design, algorithmic compensation might contribute to reduce the autonomy of the workers in constraining them to work longer hours and decrease the control they have on the rewards they gain from work (Goods, Veen, & Barratt, 2019;Griesbach et al., 2019;Moore & Hayes, 2017, 2018;Rani & Furrer, 2020;Wood et al., 2019).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Goods, Veen, & Barratt, 2019 source
    • cites Griesbach et al., 2019 source
    • cites Moore & Hayes, 2017, 2018

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    The most common and accepted components of algorithmic fairness are the absence or minimization of bias and discrimination, the accuracy and appropriateness of decisions, the relevance, reasonableness or legitimacy of the criteria or information used, and the privacy of data and decisions.

    “Although the conceptualizations of algorithmic fairness differ among authors, its most common and accepted components are the absence (or minimization) of bias and discrimination (Choudhury, Starr, & Agarwal, 2020;Cowgill & Tucker, 2020;Jean, 2020;Lepri et al., 2018;Robert et al., 2020;Shin & Park, 2019;Wong, 2019), the accuracy and appropriateness of decisions (Shin, 2021;Shin & Park, 2019), the”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Choudhury, Starr, & Agarwal, 2020 source
    • cites Cowgill & Tucker, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Jean, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The three proposed categories of moderators present significant practical challenges.

    “the three proposed categories of moderators present significant practical challenges.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    There is likely to be a moderating role of individual differences and of wider contextual elements such as organizational culture and labor market variables.

    “we recognise there is likely to be a moderating role of individual differences and of wider contextual elements such as organizational culture and labor market variables.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    Keeping humans in the decision loop and preserving human voice is important, but too much voice could impair the benefits of automated decision-making and lead to a heavy and unfruitful decision-making process.

    “Also, keeping humans in the decision loop and preserving human voice is certainly important, but too much voice could impair at some point the benefits of automated decision-making and lead to a heavy and unfruitful decision-making process.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    There are potentially important moderating factors related to how algorithmic management is designed and implemented that could mitigate its negative effects and enhance its positive effects.

    “there are potentially important moderating factors to do with the way that AM is designed and implemented that could mitigate AM's negative effects and enhance its positive effects.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic goal setting is positively associated with physical demands.

    “Physical demands (+)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Reyes (2018)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Gregory (2020)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management is negatively associated with social support.

    “Social support (-)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Leclercq-Vandelannoitte (2017) source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic goal setting is negatively associated with job complexity.

    “Job complexity (-)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Brione (2017, p. 12)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    The extent to which schedules or goal-setting decisions are based on previous performance could shape the burdens associated with those decisions in terms of workload or physical demands.

    “Similarly, the extent to which schedules or goal-setting decisions are based on previous performance could shape the burdens associated with those decisions in terms of workload or physical demands.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic scheduling is positively associated with workload.

    “Workload (+)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Lehdonvirta (2018) source
    • cites Levy and Barocas (2018)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Most experts concur that strong AI is either not likely to happen in the near future, or is unlikely to happen at all.

    “most experts concur that strong AI is either not likely to happen in the near future, or is unlikely to happen at all(Hagendorff & Wezel, 2019;Wang & Siau, 2019).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Hagendorff & Wezel, 2019 source
    • cites Wang & Siau, 2019 source

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

    Premisecitation

    The provision of fair algorithms, especially machine learning algorithms, is a complex and challenging issue for computer scientists and managers, and choices and decisions may depend on ethical orientations.

    “The provision of fair algorithms, especially machine learning algorithms, is also a complex and challenging issue for computer scientists and managers (Jean, 2020;Lepri et al., 2018), and choices and decisions may depend on ethical orientations (Gal et al., 2020;Gibert, 2020;Martin, 2019a;Mittelstadt, Allo, Taddeo, Wachter, & Floridi, 2016).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Jean, 2020

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Lepri et al., 2018 source
    • cites Gal et al., 2020 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management has a mixed (positive and negative) association with feedback from the job.

    “Feedback from the job (+/-)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Rosenblat and Stark (2016)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management is negatively associated with task significance.

    “Task significance (-)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Moore (2016, 2017)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Newman et al. (2020) source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic performance management has a mixed (positive and negative) association with role clarity.

    “Role clarity (+ / -)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Rosenblat and Stark (2016)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Bernstein and Li (2017) source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationreasoning

    The ability to contest prejudicial or erroneous algorithmic compensation or termination decisions may protect workers' autonomy and lower the probability of stressful job demands such as job insecurity.

    “The ability to contest prejudicial or erroneous algorithmic compensation or termination decisions might also help to protect workers' autonomy, as well as lower the probability of stressful job demands such as job insecurity.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The reasonableness (level of intrusiveness, range of information collected) of the system is likely to attenuate the impact of monitoring on task variety and significance and to preserve the ability of workers to organize their tasks and methods.

    “In that sense, the reasonableness (level of intrusiveness, range of information collected) of the system is likely to attenuate the impact of monitoring on task variety and significance and to preserve the ability of workers to organize their tasks and methods.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationnone given

    Fostering transparency about the use and content of algorithmic systems may contribute to lower biases, and allowing more human influence in the system is likely to trigger more relevant and legitimate automated decisions.

    “fostering transparency about the use and content of algorithmic systems may naturally contribute to lower biases, and allowing more human influence in the system is likely to trigger more relevant and legitimate automated decisions.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The ability to opt out of the automated system or of its decision will increase job autonomy and maintain problem-solving opportunities.

    “We expect that this ability to opt out of the automated system or of its decision will increase job autonomy and maintain problemsolving opportunities.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

    Premisecitation

    Algorithmic compensation is negatively associated with task significance.

    “Task significance (-)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Newman et al. (2020) source

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Premisereasoning

    Bus drivers have access to a detailed and confidential breakdown of their areas of possible improvement, which seems to be perceived by drivers as feedback from the job and role clarity improvement.

    “Bus drivers have access to a detailed and confidential breakdown of their areas of possible improvement, which seem to be perceived by drivers as feedback from the job and role clarity improvement.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Algorithmic goal setting is positively associated with workload.

    “Workload (+)”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Liao (2018)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Burin (2019)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Yeginsu (2018)

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

86 of 86

  • The paper relies on this without support.

    Algorithmic management has six actual functions that can be examined as distinct components.

    “in unpacking the phenomenon of algorithmic management and its six (actual) functions, we develop a more holistic approach of algorithmic management, allowing its examination across the distinctly operating components rather than as a monolithic piece.”
1 of 1

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Managers and organizations might reduce workers' autonomy by closely monitoring them, such as checking computer or phone usage.

    “However, managers and organizations might rob workers of this autonomy by closely monitoring them, for example by checking their computer or phone usage123.”

    From Understanding and shaping the future of work with self-determination theory (2022) · cites this paper as [123]

    The 'Algorithmic monitoring' section says monitoring can reduce workers' autonomy in choosing their work methods and lists internet browser history among the information systems can collect. This supports the general claim and a computer-use example, but the passages do not identify checking phone usage as an example. (LLM, conf=0.82)

    Experimental — accuracy not yet measured

  • Whether algorithmic management systems consider human motivational factors influences whether workers are given enough autonomy, skills usage, task variety, social contact, role clarity, and manageable workload.

    “Specifically, whether algorithmic management systems consider human motivational factors in their design influences whether workers are given enough autonomy, skills usage, task variety, social contact, role clarity (including knowing the impact of one’s work) and a manageable workload123.”

    From Understanding and shaping the future of work with self-determination theory (2022) · cites this paper as [123]

    The introduction and the section on job resources and demands describe how algorithmic-management design can shape autonomy, task variety, social support, role clarity and workload. The paper discusses job complexity and problem-solving opportunities, but does not specifically establish skills usage as an outcome. It also treats knowing the impact of one’s work as task significance, not as part of role clarity. Page numbers were not available in our copy. (LLM, conf=0.83)

    Experimental — accuracy not yet measured

  • Algorithmic management is often associated with simplified tasks, lower problem-solving opportunities, and job variety.

    “In addition, algorithmic management is often associated with simplified tasks, and with lower problem-solving opportunities and job variety123.”

    From Understanding and shaping the future of work with self-determination theory (2022) · cites this paper as [123]

  • Because quantified metrics are often linked to performance scores, pay incentives, and future allocation of tasks or schedules, algorithmic management reduces workers' freedom in decision-making, which can significantly reduce self-determined motivation.

    “Because these types of quantified metric are often directly linked to performance scores, pay incentives and future allocation of tasks or schedules (that is, getting future work), algorithmic management reduces workers’ freedom in decision-making related to their work, which can significantly reduce their self-determined motivation123.”

    From Understanding and shaping the future of work with self-determination theory (2022) · cites this paper as [123]

Cites

Flagged first, then by citations

2 of the 185 references couldn’t be matched to a known paper.

Research works printed but not linked: 15

  1. De La Garza, 2019, This AI software is “Coaching” customer service workers. Soon It Could Be Bossing You Around, Too, Times, 7/8/2019
  2. Eveleth, 2019, Your employer may be spying on you—And wasting its time, Scientific American, 321
  3. Gershgorn, 2015, Hitachi hires artificially intelligent bosses for their warehouses, Popular Science
  4. Gibert, 2020, Vol. 17
  5. Hernandez, 2018, How computers could make your customer-service calls more human, The Wall Street Journal (online)
  6. Hodson, 2014, The AI boss that deploys Hong Kong's subway engineers, New Scientist
  7. Levy, 2018, Refractive surveillance: Monitoring customers to manage workers, International Journal of Communication, 12, 1166
  8. O'Connor, 2016, When your boss is an algorithm, Financial Times.
  9. O'Donovan, 2018, An Invisible Rating System At Your Favorite Chain Restaurant Is Costing Your Server, BuzzFeed News
  10. Rosenbush, 2018, The Morning download: UPS expands role of predictive analytics, The Wall Street Journal.
  11. de Ruyter, 2019, Gig work and the Foruth industrial revolution, Journal of International Affairs, 72, 37
  12. Stevens, 2015, At UPS, the algorithm is the driver, The Wall Street Journal
  13. Waterson, 2016, Daily telegraph installs workplace monitors on journalists' desks, BuzzFeed News
  14. Williams, 2017, How the gap used an app to give workers more control over their schedules, Harvard Business Review
  15. Yeginsu, 2018, If workers slack off, the wristband will know. (and Amazon has a patent for it.), The New York Times

Cited by

Sorted by citations

Related

Authors
Parent-Rocheleau, X., & Parker, S. K.
Affiliations
Curtin University (AU)
ORCID
Parker, Sharon K.
Venue
Human Resource Management Review
Publisher
Elsevier BV
Published
2022-09
Volume
32
Issue
3
Pages
100838
ISSN
1053-4822 (ISSN-L)
DOI
10.1016/j.hrmr.2021.100838
OpenAlex
W3161181201
Semantic Scholar
7f867c7f4b67961c21507e959d13ecd5af949179
Cited by
512
References
185
Citations by year
2026: 199 · 2025: 137 · 2024: 93 · 2023: 43 · 2022: 37 · 2021: 3
FWCI
60.93
Type
journal article
Language
en
Open access
Licence
Locations
  • Human Resource Management Review (acceptedVersion)
  • eSpace (Curtin University) (submittedVersion, cc-by-nc-nd)
Funders
  • Social Sciences and Humanities Research Council of Canada (756-2019-0094)
  • Australian Research Council (FL160100033)
Topics
  • Social Sciences › Sociology and Political Science › Digital Economy and Work Transformation
  • Social Sciences › Business, Management and Accounting › Industrial relations › Legal and Labor Studies
  • Social Sciences › Business, Management and Accounting › Organizational Behavior and Human Resource Management › Job Satisfaction and Organizational Behavior
Keywords
algorithmic management, job design, employee autonomy, job complexity, workload, performance management, goal setting, job scheduling, compensation, dismissal, employee motivation, employee well-being, transparency, fairness, human influence, socio-technical systems theory, meaningful work

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-30.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-30.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review