Proceedings of the 50th Hawaii International Conference on System Sciences | 2017

A Conceptual Framework for Investigating Organizational Control and Resistance in Crowd-Based Platforms

David A. Askay California Polytechnic State University [email protected]

Abstract Crowd-based platforms coordinate action through This paper presents a research agenda for crowd decomposing tasks and encouraging individuals to behavior research by drawing from the participate by providing intrinsic (e.g., fun, organizational control literature. It addresses the enjoyment) and/or extrinsic (e.g., status, money, need for research into the organizational and social social interaction, etc.) motivators [3]. Three recent structures that guide user behavior and contributions Information Systems (IS) reviews of crowdsourcing in crowd-based platforms. Crowd behavior is research [6, 4, 7] emphasize the importance of situated within a conceptual framework of designing effective incentive systems. However, organizational control. This framework helps research on motivation is somewhat disparate, with scholars more fully articulate the full range of various categorizations and often inconsistent control mechanisms operating in crowd-based findings as to which incentives are the most effective platforms, contextualizes these mechanisms into the [4]. Moreover, this approach can be limited by its context of crowd-based platforms, challenges existing often deterministic and rational assumptions of user rational assumptions about incentive systems, and behavior and motivation, which overlooks normative clarifies theoretical constructs of organizational and social aspects of human behavior [8]. To more control to foster stronger integration between fully understand the dynamics of crowd behavior and information systems research and organizational and governance of crowd-based projects, IS researchers management science. need a sophisticated way to conceptualize the complex social and organizational processes that guide the behavior and actions of the crowd. To fill this conceptual gap in the crowd behavior 1. Introduction literature, this paper draws from organizational control research. Organizational control refers to the Organizations are increasingly interested in strategies used by managers to align worker harnessing the collective actions of crowds to meet behaviors to organizational objectives [9, 10]. While business goals [1]. This has resulted in the IS scholars have occasionally focused on proliferation of crowd-based platforms supporting organizational control mechanisms [11, 12, 13], an collective intelligence and crowdsourcing efforts that explicit consideration of control mechanisms is coordinate the contribution of user-generated particularly limited in the crowdsourcing and content—product reviews, encyclopedia articles, and collective intelligence literature. current traffic conditions. Consequently, the crowd This paper makes four contributions to the has become a powerful source of knowledge that literature. First, I present a framework that more fully guides not only the decisions of the public, but also articulates the range of control mechanisms that decisions made by organizations [2]. As guide the behavior of the crowd. Second, I organizations like Amazon, Waze, and Wikipedia contextualize this framework through providing turn to the crowd to produce the content that their examples of organizational control research in business models depend on, the challenge of how to crowd-based platforms. Third, I identify promising motivate and govern the crowd has become the topic new directions for crowd behavior research by of much research [1, 3]. While there is scholarly bringing in the ontologically and methodologically interest in the “actions and policies employed to diverse organizational control literature [14]. Fourth, effectively manage the crowd and steer them” [4], this paper clarifies theoretical constructs of there remains limited research about effective organizational control in order to foster integrative mechanisms for governing the crowd [5].

URI: http://hdl.handle.net/10125/41364 ISBN: 978-0-9981331-0-2 CC-BY-NC-ND 1727 research drawing from both IS and organizational and-effect relationship between incentives and action science literature. [22]. By contrast, informal control strategies (also 2. Organizational Control and Crowds called normative or social control) seek to guide behaviors through fostering collective norms, identity, culture, and values in members. A This section defines organizational control, theoretically and ontologically diverse range of summarizes an input-process-output model of scholars have considered these kinds of informal organizational control, and describes resistance to control strategies [23, 21, 20, 9, 24, 14, 25]. In organizational control. general, socio-ideological control occurs when individuals are socialized into an organization’s 2.1. What is Organizational Control? system of beliefs, values, and norms (i.e., corporate culture), which provides a particular organizational While IS researchers often note the ‘loss of reality that acts as a sense-making device for member control’ associated with crowdsourcing [5], control is behaviors. When these values and norms are a fundamental aspect of all organizations and is internalized, members can monitor their own needed to coordinate goal-oriented behaviors [15]. As behaviors to ensure that they align with will be described below, crowd-based platforms use a organizational goals. Informal control mechanisms diverse range of control mechanisms. Organizational are particularly useful for knowledge work, where control includes the mechanisms used by managers to required creativity, ambiguity, and flexibility are not direct, motivate, and encourage worker behaviors to conducive to bureaucratic control strategies that align with organizational objectives [9, 10]. The need dictate a rule for every contingency [26]. Still, for organizational control stems from the competing control strategies often do not fit firmly in one of interests that exist between individuals and these two categories, and a single mechanism may organizations, which can be resolved though control possess attributes of both formal and informal mechanisms (e.g., rules, rewards, punishments, strategies [10, 21]. supervision, etc.) that subordinate the individual’s goals to those of the organization. Control strategies 2.2. Framework of Organizational Control include the incentives that motivate some behaviors (e.g., incentive systems) and also the mechanisms A recent synthesis of organizational control that constrain other behaviors. research uses an input, process, output (IPO) model Various control mechanisms are used by to produce a useful model for conceptualizing the organizations, including policies [16], bureaucratic various mechanisms in a control system [10]. This rules [17], culture [18], incentive systems and framework classifies control systems on the degree to supervision [19], identity [20], and technological which formal and informal control mechanisms target design [9]. Scholars have offered several approaches the input, behavior, and output of the system. It to categorizing these control strategies [9, 10, 21, 22], presents a control system as comprised of several though most distinguish between formal and informal control mechanisms (e.g., bureaucratic, control strategies [22, 11]. technological, direct, socio-ideological, etc.) that Formal control strategies (also called rational each have a control target (e.g., input, behavior, control) seek to influence behavior through the output) in the organization. Input targets include explicit and codified structuring of inputs, tasks, mechanisms that control how inputs (e.g., who is processes, behaviors, and/or outcomes—in essence allowed to participate) to the system are qualified, providing incentives in return for completing well- chosen, and prepared, such as through the selection, designed tasks within certain parameters or punishing training, and socialization of individuals. Behavior those that deviate from this organizationally designed targets focus on controlling the manner in which process. This forms the basis for the incentives individuals perform actions or complete tasks. systems currently used by many crowd-based Finally, output targets include mechanisms that platforms. Additionally, behavior can be directed control qualitative and quantitative outcomes, such as through bureaucratic rules, direct supervision, and profits, quality standards, production volume, and technological design. These control strategies are speed. This framework draws attention to the various premised on the assumptions that individuals are control mechanisms that collectively comprise a primarily rational and view work as an economic control system. exchange, and that there is a generally linear cause- Most IS research on crowd behavior has focused on the study of formal control mechanisms (i.e.,

1728 incentive systems) that target the outputs produced by neighborhood, local residents began posting false the crowd [4]. By contrast, little scholarly attention reports during rush hour in effort to redirect traffic has been directed towards control mechanisms that away [33]. However, these acts of resistance were target inputs (e.g., who is included and excluded from suppressed through two organizational control participating, socialization and training, etc.) and mechanisms. First, other Waze users were able to behaviors (e.g., how and when contributions are invalidate the false reports through submitting their produced). Additionally, informal control own report. Second, an algorithm used data mechanisms have been understudied in the context of transmitted by the smartphone (e.g., speed, direction, crowd behavior, if mentioned at all. Finally, little etc.) to suspend the user accounts of those suspected research has addressed how several control strategies of submitting false reports. Together, these control may work in tandem to control crowd behavior, an mechanisms were sufficient to overcome the approach that is advocated for by organizational resistance efforts of these stakeholders and scholars [21, 26, 27]. marginalize their interests. Given the scale at which crowd-based platforms operate, the question of how 2.3. Resistance to Organizational Control control systems produce winners and losers as stakeholders struggle over divergent goals becomes an important area of study. Control is often implicitly considered a straightforward—even deterministic—process by scholars, whereby the application of proper control 3. Organizational Control Mechanisms mechanisms produce desired user behavior. This is reflected in IS studies suggesting that “task designers A summary of formal and informal control must guess at the right combination of incentives and mechanisms is presented. Examples of crowd-based iterate until success” [3] or simply identify the “right platforms are provided, drawing from studies that incentive mechanism” [4] for encouraging explicitly or implicitly embody these control participation. However, other scholars challenge this mechanisms. This reveals a variety of yet rational deterministic view of control by questioning unconsidered control mechanisms currently used in the assumption that “control is achieved by designing crowd-based platforms. and applying appropriate structures, procedures, measures and targets; and, relatedly, that resistance to 3.1. Bureaucratic Control these mechanisms is symptomatic of `poor design’ or `poor management’ that can be rectified by restructuring” [20]. Bureaucratic control relies on impersonal and Critical approaches to organizational control, by formal system of rules, procedures, and roles to guide contrast, view organizations as socially constructed worker behavior [17], although some rules may also and as sites of power and resistance [14], where be informal. Such systems can, for example, dictate divergent interests among stakeholders often the kinds of people selected for a task, the manner by provokes subversion, protest, sabotage, and other which tasks are supposed to be completed, the kinds deviant behaviors [28, 29, 30, 31, 32]. Indeed, of behaviors that are acceptable, and how conflict and struggle may be as commonplace in performance is evaluated, rewarded, and punished. crowds as compliance and consent. Yet resistance as Often organizational control is achieved by explicitly core aspect of organizational life is overlooked in linking worker behaviors to sanctions or rewards, crowd behavior research, where participation is often targeting extrinsic and intrinsic motivators of these conceived of as binary—individuals participate or individuals [34]. Extrinsic motivation induces they do not. By focusing scholarly attention to the cooperation through the promise of instrumental struggles experienced by various stakeholders, we benefits, such as money, status, or reputation. can better understand the full range of stakeholder Conversely, intrinsic motivation results from experiences in these platforms and design more individuals cooperating due to the task being humane and democratic platforms. personally rewarding, such as enjoyment, fun, and A recent example underscores the less visible altruism. actions of control and resistance that occur in crowd- The use of bureaucratic control is identifiable in based platforms. Waze is a GPS-based navigation many crowd-based platforms, as rule-based application for smartphones that incorporates real- contributions are used to coordinate the collective time user-reported accidents, traffic, and police cars action of unconnected individuals. Other rules may to optimize routes. When a detour consistently motivate participation by providing clear guidelines rerouted traffic through a previously quiet or criteria for achieving certain symbols of status

1729 (e.g., badges or titles) or earning rewards (e.g., but that they were also sufficiently internalized to the money). Alternatively, individuals may participate point that they would self-police content despite no more when they know that their contributions are clear incentive. A deeply unsatisfying explanation of quantified and made visible on their profile. A recent this behavior is that some members just happen to literature review concluded that studies of incentive take on extra tasks for no apparent reason. These systems and motivation are among the most studied members are likely driven by commitment to the aspects of crowd behavior in the IS literature [4]. Yet, organization and its goals through socio-ideological these findings are often inconsistent, with it control (described below) or perhaps the kind of remaining unclear which, if any, of these intrinsic concertive control found in self-managed teams [24]. and extrinsic incentives are the most important for Here, the alignment of two different control recruiting and fostering participation. mechanisms work together to produce a control Two important observations about crowd system that discourages and eliminates certain kinds behavior research arise from these findings. First, of contributions to Yelp. However, this may also be a extant IS research focuses heavily on incentives that cautionary lesson as organizational research reveals target outputs where participants are rewarded for that rules deemed ‘bad’ by members can also become their contributions with status, recognition, badges, or dysfunctional [37, 38, 39], leading individuals to money. Less often examined are formal rules that resist, engage in mischief, or leave the organization. target inputs (e.g., who gets to participate) or In other words, some formal rules may act as a behaviors (e.g., how people perform tasks) in the disincentive for participation. system. Second, most research examines rules that motivate certain actions, while overlooking formal 3.2. Technological Control rules act to constrain other actions. As organizations seek to adequately govern and control the crowd, Technological control is exercised through more research is needed to understand the extent to organizational technologies that substitute for the which formal rules reward and punish, not just presence of direct supervision and constrains the outputs, but also inputs and behaviors. amount of discretion available to workers [40]. For

example, assembly lines can control the pace of work 3.1.1. Example of Bureaucratic Control. A in factories, while soda fountains at McDonald’s study of reviewers on Yelp reveals how bureaucratic dispense precisely the right amount of soda into cups rules, in the form of the terms of service, impacts to eliminate waste [30]. Such mechanisms severely crowd behavior. While Yelp promotes empowerment limit the ability of workers to deviate from the and transparency through its online review system, organizationally planned process, highlighting their the terms of service disallow users from including use in targeting the behavior of participants— civic or political discourses in their reviews [35]. constraining how and when they work. Other Kuehn describes how some participants on Yelp technologies can provide the constant threat of “monitored and flagged political reviews that did not surveillance, acting as an invisible supervisor that can fit Yelp’s terms of service” [36], which limited the notify managers or take action when a worker discursive agency available to users in their deviates from a task. evaluations on Yelp. This bureaucratic rule, enforced There are several ways in which technological through self-policing, suppressed and filtered reviews control can be exerted in crowd-based platforms. deemed inappropriate by Yelp. These rules construct Contractors on the digital labor market Upwork, for Yelp as ‘‘not the place’’ for consumer politics despite example, can be surveilled by their employers promoting itself as a platform for empowerment and through recording their desktop screens while they transparency. More importantly, users are work. More recently, the use of algorithms has also empowered to produce only the type of content that explored as a means of control [41, 42]. For example, aligns the objectives of the greater organization—the the Yelp filter automatically removes roughly 20% of crowd is not simply encouraged to contribute, but to reviews submitted to the website that are suspicious contribute the ‘right’ kind of content. [43]. While this is done to remove fake or paid-for This study reveals an important, yet understudied, reviews, it also may upset individuals that provided way in which formal rules are used to constrain the honest opinions only to find them removed for an kind of content that is produced in a crowd-based unknown reason. platform. Perhaps most interesting is that some users Technological control can also be linked to the chose to self-police the reviews to ensure that these concept of affordances, which has been increasingly rules were being followed by others. This reveals that used in the IS field [44, 45]. Affordances of a not only did these members know about these rules, technological artifact are the potentials for action

1730 perceived by an individual [45], which can be as needed. This definition suggests that it is unlikely mediated by group norms [46]. As a result, that direct supervision exists in crowdsourcing. affordances of technology are not viewed Indeed, scholars assert that crowdsourcing “implies deterministic, but rather accept that technologies can voluntary participation of individuals, with no be adopted in unintended ways by users. This has hierarchy or contract related constraint … been applied to crowdsourcing by IS researchers who [c]oordination by hierarchy does not take place” [49]. recognize that “design elements of a variety of Other conceptualizations of crowdsourcing and architectures that constrain and afford purposeful collective intelligence emphasize that lack of human actions” [47]. In this sense, the affordances hierarchy is a condition of crowdsourcing [1] and that designed into crowdsourcing platforms might act to crowds lack direct supervision [3]. shape crowd behavior. Still, while affordances are often conceived of as enabling and constraining 3.3.1. Example of Direct Control. Emerging action, the explicit study of affordances (or lack research, however, challenges the assumed lack of thereof) being strategically designed (or omitted) direct supervision of the crowd. Several studies have from crowd platforms to further organizational goals identified people holding the position of a community emerges as understudied. manager operating in crowd-based platforms. In one study, the community manager was found to play an 3.2.1. Example of Technological Control. One integral role in managing the production of user- study investigating a peer feedback system on a generated content. As Kerr and Kelleher explain, popular hospitality exchange website provides “[w]hen share value depends on market share, unique insights to how technological control through users, and clicks, companies are turning to a range of affordances may function in crowd-based platforms techniques to “engage,” retain, and convert user [48]. This study found that the contribution of activity into revenue [including] the employment of negative feedback was inhibited by the affordances of community managers” [50]. These community the system. Specifically, users of this website managers were found to organize offline events, reported that their negative experiences could not be manage problematic user behaviors, and deal with the adequately explained within the 1,000 character limit emotions of users. Other scholars found that available for references and that the required labeling community managers monitor and control crowd of feedback as negative, neutral, or positive did not activity and “promote the participation and align with their subjective experience. As a result, collaboration of stakeholders in order to improve many users did not feel empowered to contribute some ‘crowdsourcing’ processes” [51]. Still, their negative experiences. This contributed to the relatively little is known about these community overall positive bias of feedback on the system and managers and the extent to which they are effective. may have (intentionally or unintentionally) benefited They reveal, however, that as harnessing crowds has the organization because new members (and hence become more central to organizations, more formal income) were recruited through its advertised high and rational forms of direct control appear to be level of positive feedback. Interestingly, though, implemented to guide crowd behaviors. It also these users resisted this technological control by illustrates the importance considering how instead expressing negative experiences in the unexpected forms of control may operate in crowd- discussion forums—a channel with different based platforms. technological affordances that did not restrict the length of comments nor require labels. This shows 3.4. Socio-Ideological Control that users were willing to contribute the content of their negative experiences, but that the technological While previous control mechanisms are founded affordances of this platform restricted the kinds of on a rational approach to human behavior, socio- feedback that users were able to produce. ideological control mechanisms stem from a normative approach to control, in which the 3.3. Direct Control organization can foster the development of beliefs and values that workers are expected to internalize Direct control is generally enacted through a and use as a guide for their behaviors [21]. Scholars formalized hierarchical supervisor-subordinate have noted a lack of conceptual consensus of relationship in which the supervisor has the informal approaches [10], contributing to legitimate authority to instruct workers on what to do, development of several overlapping theories of socio- monitor their behavior, and reward and punish them ideological control, including corporate culture,

1731 cultural control [52, 53], clan mechanisms [9], donating and suppresses the idea that they might be disciplinary control [54, 55], and identity-based deserving of payment. control [20, 23]. These theories are likewise Despite the popularity of investigating socio- differentiated by their intellectual traditions and ideological control mechanisms in the organizational ontological and epistemological orientations, and managerial sciences, this is understudied in including structural functionalism, post-structuralism, studies of crowd behavior. Yet, IS scholars have social constructivist, post-modernism, critical theory, called for research to investigate the social processes positivism, and interpretism. operating in crowd-based projects [8]. Moreover, Informal control mechanisms generally involve “a research demonstrating frequent communication process of reality construction that allows people to among crowd workers [72] and the employment of see and understand particular events, actions, objects, community managers [51] suggests that utterances, or situations in distinctive ways” [56]. crowdsourcing platforms may be more social and Through discourse that shapes the organizational have greater organizational contact than sometimes reality of members, organizations can influence believed. behavior by guiding worker values, beliefs, norms, taken-for-granted assumptions, concepts of right and 3.4.1. Example of Socio-Ideological Control. The wrong, and good and bad. Ultimately, this serves to findings of a recent ethnography of the invite-only eliminate or minimize the incongruence that exists Yelp Elite Squad implicitly speaks to how socio- between individual and organizational goals. ideological control functions in a crowd-based Socio-ideological control mechanisms are less platform [73]. The Yelp Elite Squad exists in cities obtrusive than rational control systems as they do not around the world, members of which spend rely on external structures of the organizations significant time visiting business, producing reviews, (supervisors, technology, or rules). Rather, through a and participating in online forum discussions. In strong culture or identity fostered through return, they are invited to free offline member-only organizational discourses, workers are encouraged to events each month. They are incidentally coordinated monitor their own behavior to ensure that they make by a community manager (CM), who is a trained and decisions and actions that align with organizational paid employee of Yelp. The study describes how goals. This places emphasis on understanding how discursive resources on the Yelp website and language and organizational discourse (e.g., written communication by the community manager foster a documents, speech, pictures, symbols, etc.) is drawn specific organizational reality. Elite Squad members upon by members to constitute their social reality and are constructed as community members through organizational life [57]. Discourse can be thought of discourse defining them as “the true heart of the Yelp as “a way of reasoning…anchored in a particular community” and “a local authority and role model for vocabulary that constitutes a particular version of the the Yelp community” that has “got a lot of sway in social world” [58], based in the assertion that the community” [74]. Coinciding with this, members “language constructs organizational reality, rather reported that they strive to be Yelpy and Yelp-like, than simply reflects it” [59]. This emphasizes the which includes behaviors like representing Yelp well analysis of verbal and written communication to businesses, not demanding special treatment, including organization stories [60, 61], rituals [62, recruiting new members, producing high quality 63], narratives [64], and metaphors [65] used by the reviews, and producing reviews often. Furthermore, organization and shared among members and used as the construction of their participation as community a sense-making device [66]. Several recent studies members was sufficiently powerful that when asked offer detailed accounts of socio-ideological control in about their role in Yelp’s business, members organizational contexts [67, 68, 69, 21]. responded saying that they simply had not considered One example of discourse constructing a certain Yelp as a business or their labor as producing organizational reality can be found in the concept of financial value. One member said that, “I just feel playbour (a play on the words play and labour) [70]. like, oh it’s just a fun community and you know, get This occurs when organizations encourage free dinners and it’s just like friends are in it . . . I see participants to understand their unpaid contributions Yelp as a community more than I do a business. Like, (such as unpaid video game modding) as ‘fun’ or a I never really thought of it being a business” [73]. ‘hobby’, while the organization extracts value from Several implications arise from considering the their efforts. Their labor is paradoxically “voluntarily Elite Squad from the perspective of socio-ideological given and unwaged, enjoyed and exploited” [71]. control. First, the desire to be ‘Yelpy’ suggests that This framing obscures other constructions of their these individuals strongly identity with the participation—such as volunteering, work, or organization and have internalized a set of values and

1732 beliefs that guide their behavior, which leads to them the IS literature has revealed much about incentive to self-monitor their actions even when they are not systems [4] that target outputs, other types of formal supervised (e.g., not demanding special treatment). and informal control mechanisms have been How Yelp was able to achieve this becomes an understudied. Yet these are critical aspects to interesting scholarly question. Second, these investigate as scholars and designers seek to members internalized Yelp’s version of understand the best ways to govern the crowd. organizational reality, in which their participation is To address this need, this paper encourages IS understood through a civic-minded framing of scholars to consider the various formal and informal ‘community’, rather than as a volunteer or as rational control mechanisms operating at the input, behavior, exchange with a for-profit business that financially and output targets of crowd-based platforms [10]. benefits from their unpaid labor. The discursive Through applying these concepts to studies of various resources that constitute the Elite Squad as a crowd-based platforms, some key insights emerge. community obscure how their labor financially While bureaucratic incentives systems that benefits Yelp—it was not until confronted with this encourage certain outputs emerged as the most idea by the interviewer that members were able to commonly studied type of control, far less examined make sense of their participation differently. This are the rules and policies that act to constrain crowd suggests that hegemony—a process of gaining the behavior. Moreover, the study of self-policing at active consent of a dominated group [75, 76]—and Yelp [36] suggests that some control mechanisms related considerations of asymmetrical power work in concert with each other (i.e., bureaucratic differences is a fruitful area of consideration for and socio-ideological) to encourage some users to further research. Third, it highlights that some crowd- flag content that deviates from what Yelp considers based platforms require individuals to operate in appropriate. Effectively, a control system was offline settings (e.g., visiting businesses to review) devised that encourages some users to flag and where they are not as easily survailled by the remove content that deviates from organizationally organization. Yet, the organization still has an appropriate standards. This type of crowd behavior is interest in controlling the behavior of participants in both incredibly interesting and understudied. Future these offline spaces. Here, disciplinary control, in research may draw from concertive control literature which members self-monitor their behavior [55, 54, [24] and studies that examine the interplay of several 23], emerges a powerful mechanism of organizational control mechanisms [21, 26] to further examine control. Fourth, it demonstrates that not all similar phenomena. individuals in the crowd are subject to the same In regard to technological control, a study of the control mechanisms. Rather, there appears to be be feedback system on a hospitality website revealed less visible sub-groups [77] operating in the crowd that the design and absence of some affordances subject to different mechanisms of organizational acted to inhibit the contribution of negative reviews control. Overall, this study brings attention to the [48]. This suggests that a promising vein of crowd importance of considering discourse in the behavior research could more explicitly integrate the investigation of informal control strategies in crowd- technological affordances and materiality literature based platforms. [44, 46] in studies of organizational control. Similarly, the use of algorithms to automate control is 4. A Research Agenda for Crowd Behavior likewise an emerging area of relevance to the Based on Organizational Control Theory functioning of crowd-based platforms [41, 42]. Emerging research on community managers suggests that some crowd-based platforms are using There is an assumption that crowd-based direct control of crowd members through local paid platforms imply the loss of control by organizations employees, who monitor and manage members of the [2, 5]. However, by clarifying mechanisms of crowd [73, 50, 51]. Further research is needed to organizational control and situating these within clarity the role of these community managers and empirical studies of crowd-based platforms, this how the crowd responds to and resists being managed paper demonstrates that a diverse array of formal and and managing each other. Still, this brings to focus informal organizational control mechanisms do the need to better understand how control systems indeed operate in this setting. Organizational control evolve over time [10] on crowd-based platforms. scholars note that researchers are often too singular in Informal control mechanisms emerged as their conceptualizations and empirical focus when important and understudied in crowd-based studying control mechanisms, frequently choosing to platforms. A study of the Elite Squad reveals strong examine one type of control over others [10]. While commitment and identification to Yelp [73].

1733 Moreover, how these members made sense of their kind of content produced by the crowd and who organizational reality was shown to be influenced defines this? To what extent can control be strongly by organizational discourse and discursive resisted? How do control mechanisms produce resources. Just as an incentive system can encourage winners and losers?

certain behaviors through motivation, organizational • What are the similarities and differences between discourse can encourage certain behaviors through control systems in traditional organizations and constructing organizational realities and identities of crowd-based platforms? crowd members. Future research could adopt • What emergent or organizationally constructed discourse analysis [64, 57] to investigate how socio-ideological norms guide crowd behavior? organizational language constructs certain meanings How are identities of crowd members socially for participants and influences overall crowd constructed through organizational discourses? behavior.

Finally, this review of the literature reveals that In conclusion, many studies of crowd behavior the power struggles inherent in organizational life focus on a limited range of incentive systems. Less that manifest in resistance, hegemony, and frequent are investigations of the organizational and domination [29, 30, 14, 75] have largely been social aspects that influence how and what content is unexamined in the crowd behavior literature. Yet, produced. This review presents the ontologically, examples presented in this paper illustrate the theoretically, and methodological diverse literature of struggles emerge over defining what kind of content organizational control as a promising framework for is appropriate [36, 35] or resisting the material investigating these aspects of crowd behavior. This impacts of crowdsourcing technologies [33]. review also encourages a more careful consideration Research adopting a critical perspective [14] to of the term crowd and its popularization as an organizational control in crowd-based platforms alternative to organizations. While crowd highlights could provide valuable insights into these issues. many of the novel and different aspects of crowd- Limitations are present in this review. While an based platforms, it may also act to obscure from effort was made to draw from literature investigating scholars the more traditional organizational structures a variety of crowd-based platforms (e.g., Amazon, that operate in crowd-based platforms. Scholars can Reddit, etc.), research on Yelp is over-represented. build a more complete understanding of the complex This is because studies that explicitly or implicitly sociotechnical processes influencing crowd behavior addressed issues relating to organizational control by remaining open to discovering unexpected control were scarce in other platforms. Rather, studies of mechanisms operating in this context. While much these platforms would often examine large datasets of can be learned about crowds by examining how they what is produced by the crowd (e.g., reviews.). Far are different from organizations, there is also value in fewer studies have investigated how this content is examining how they are similar. produced by the crowd (e.g., interviews, ethnography, etc.). The scarcity of these kinds of studies underscores the need for further research 5. References analyzing the important social and organizational [1] T. W. Malone, R. Laubacher and C. Dellarocas, "The processes influencing crowd behavior. Ultimately, Collective Intelligence Genome", IEEE Engineering this review leads to a range of interesting research Management Review 38(3), 2010, pp. 38-52. questions: [2] E. Bonabeau, "Decisions 2.0: The Power of Collective Intelligence", MIT Sloan Management Review 50(2), 2009, • What is the full array of control mechanisms pp. 45-52. used in crowd-based platforms and how do they [3] A. Kittur, J. V. Nickerson, M. Bernstein, E. Gerber, A. influence participation? How do control Shaw, J. Zimmerman, M. Lease and J. Horton, "The Future mechanisms enable and constrain the content of Crowd Work", CSCW 2013 Proceedings, ACM, New produced by the crowd? How do various control York, NY, 2013, pp. 1301-1318. [4] J. Pedersen, D. Kocsis, A. Tripathi, A. Tarrell, A. mechanisms reinforce or contradict each other Weerakoon, N. Tahmasbi, J. Xiong, W. Deng, O. Oh and and how are these resolved? How do control G.-J. De Vreede, "Conceptual Foundations of mechanisms evolve over time and how does the Crowdsourcing: A Review of Is Research", 46th Hawaii crowd respond to changes? International Conference on System Sciences, IEEE,

• Who are the stakeholders in crowd-based Wailea, Maui, HI, 2013, pp. 579-588. platforms and what are their goals? What [5] R. Jain, "Investigation of Governance Mechanisms for Crowdsourcing Initiatives", AMCIS 2010 Proceedings, tensions arise from divergent stakeholder goals Lima, Peru, 2010, pp. 557. and how are they resolved? What is the ‘right’

1734 [6] R. Buettner, "A Systematic Literature Review of [23] H. C. White, Identity and Control: A Structural Theory Crowdsourcing Research from a Human Resource of Social Action, Princeton University Press, 1992. Management Perspective", 48th Hawaii International [24] J. R. Barker, "Tightening the Iron Cage: Concertive Conference on System Sciences, IEEE, 2015, pp. 4609- Control in Self-Managing Teams", Administrative Science 4618. Quarterly, 1993, pp. 408-437. [7] Y. Zhao and Q. Zhu, "Evaluation on Crowdsourcing [25] P. K. Tompkins and G. Cheney, "Communication and Research: Current Status and Future Direction", Unobtrusive Control in Contemporary Organizations", in Information Systems Frontiers 16(3), 2014, pp. 417-434. R. D. McPhee and P. K. Tompkins, eds., Organizational [8] P. Shachaf and H. Rosenbaum, "Online Social Communication: Traditional Themes and New Directions, Reference: A Research Agenda through a STIN SAGE, Newbury Park, CA, 1985, pp. 179-210. Framework", iConference ’09, ACM, Chapel Hill, NC, [26] D. Kärreman and M. Alvesson, "Cages in Tandem: 2009. Management Control, Social Identity, and Identification in [9] W. G. Ouchi, "A Conceptual Framework for the Design a Knowledge-Intensive Firm", Organization 11(1), 2004, of Organizational Control Mechanisms", Management pp. 149-175. Science 25(9), 1979, pp. 833-848. [27] M. O’Toole and C. Grey, "‘We Can Tell Them to Get [10] L. Cardinal, S. Sitkin and C. Long, "A Configurational Lost, but We Won’t Do That’: Cultural Control and Theory of Control", in S. Sitkin, L. Cardinal and K. Resistance in Voluntary Work", Organization Studies Bijlsma-Frankema, eds., Organizational Control, 37(1), 2016, pp. 55-75. Cambridge University Press, New York, NY, 2010. [28] T. Lawrence and S. Robinson, "Ain't Misbehavin: [11] L. J. Kirsch, "The Management of Complex Tasks in Workplace Deviance as Organizational Resistance", Organizations: Controlling the Systems Development Journal of Management 33(3), 2007, pp. 378-394. Process", Organization Science 7(1), 1996, pp. 1-21. [29] S. R. Clegg, D. Courpasson and N. Phillips, Power and [12] M. L. Harris, R. W. Collins and A. R. Hevner, Organizations, SAGE, Thousand Oaks, CA, 2006. "Control of Flexible Software Development under [30] J. M. Jermier, D. E. Knights and W. R. Nord, eds., Uncertainty", Information Systems Research 20(3), 2009, "Resistance and Power in Organizations", Routledge, pp. 400-419. London, UK, 1994. [13] M. T. Rao, C. Brown and W. Perkins, "Host Country [31] P. Fleming, "Metaphors of Resistance", Management Resource Availability and Information System Control Communication Quarterly 19(1), 2005, pp. 45-66. Mechanisms in Multinational Corporations: An Empirical [32] M. Parker, Against Management: Organization in the Test of Resource Dependence Theory", Journal of Age of Managerialism, Polity, Cambridge, UK, 2002. Management Information Systems 23(4), 2007, pp. 11-28. [33] S. Hendrix, "Traffic-Weary Homeowners and Waze [14] R. Delbridge, S. Sitkin, L. Cardinal and K. Bijlsma- Are at War, Again. Guess Who’s Winning?", The Frankema, "Critical Perspectives on Organizational Washington Post, 2016. Control: Reflections and Prospects", in S. Sitkin, L. [34] A. Weibel, "Managerial Objectives of Formal Control: Cardinal and K. Bijlsma-Frankema, eds., Organizational High Motivation Control Mechanisms", in S. Sitkin, L. B. Control, 2010, pp. 80-108. Cardinal and K. Bijlsma-Frankema, eds., Organizational [15] A. S. Tannenbaum, "Control in Organizations: Control, 2010, pp. 434-462. Individual Adjustment and Organizational Performance", [35] K. M. Kuehn, ""There’s Got to Be a Review Administrative Science Quarterly, 1962, pp. 236-257. Democracy": Communicative Capitalism, Neoliberal [16] C. I. Barnard, The Functions of the Executive, Harvard Citizenship and the Politics of Participation on the University Press, 1968. Consumer Evaluation Website Yelp.com", International [17] M. Weber, Economy and Society: An Outline of Journal of Communication 7, 2013, pp. 19. Interpretive Sociology, Univ of California Press, 1978. [36] K. M. Kuehn, "Brand Local: Consumer Evaluations as [18] E. H. Schein, "Organizational Culture", American Commodity Activism on Yelp.com", Journal of Consumer Psychologist 45(2), 1990, pp. 109-119. Culture, 2015. [19] H. Mintzberg, The Structuring of Organizations: A [37] B. Bozeman, Bureaucracy and Red Tape, Prentice Synthesis of the Research, Prentice-Hall, Englewood Cliffs, Hall, Upper Saddle River, NJ, 2000. NJ, 1979. [38] P. S. Adler and B. Borys, "Two Types of Bureaucracy: [20] M. Alvesson and H. Willmott, "Identity Regulation as Enabling and Coercive", Administrative Science Quarterly Organizational Control: Producing the Appropriate 41(1), 1996, pp. 61-89. Individual", Journal of Management Studies 39(5), 2002, [39] C. Perrow, Complex Organizations, Random House, pp. 619-644. New York, NY, 1986. [21] M. Alvesson and D. Kärreman, "Interfaces of Control. [40] L. Gossett, "Organization Control Theory", in S. Technocratic and Socio-Ideological Control in a Global Littlejohn and K. Foss, eds., Encyclopedia of Management Consultancy Firm", Accounting, Communication Theory, SAGE, Thousand Oaks, CA, Organizations and Society 29(3), 2004, pp. 423-444. 2009, pp. 706-710. [22] S. R. Barley and G. Kunda, "Design and Devotion: [41] J. Cheney-Lippold, "A New Algorithmic Identity Soft Surges of Rational and Normative Ideologies of Control in Biopolitics and the Modulation of Control", Theory, Managerial Discourse", Administrative Science Quarterly, Culture & Society 28(6), 2011, pp. 164-181. 1992, pp. 363-399. [42] T. Gillespie, "The Relevance of Algorithms", in T. Gillespie, P. J. Boczkowski and K. A. Foot, eds., Media

1735 Technologies: Essays on Communication, Materiality, and [59] C. Hardy, T. B. Lawrence and D. Grant, "Discourse Society, MIT Press, London, UK, 2014, pp. 167. and Collaboration: The Role of Conversations and [43] M. Luca and G. Zervas, "Fake It Till You Make It: Collective Identity", Academy of Management Review Reputation, Competition, and Yelp Review Fraud", 30(1), 2005, pp. 58-77. Harvard Business School NOM Unit Working Paper No. [60] D. M. Boje, "The Storytelling Organization: A Study 14-006, 2015. of Story Performance in an Office-Supply Firm", [44] S. Faraj and B. Azad, "The Materiality of Technology: Administrative Science Quarterly, 1991, pp. 106-126. An Affordance Perspective", in P. M. Leonardi, B. Nardi [61] Y. Gabriel, Storytelling in Organizations: Facts, and K. J., eds., Materiality and Organizing: Social Fictions, and Fantasies, Oxford University Press, Oxford, Interaction in a Technological World, Oxford University UK, 2000. Press, Oxford, UK, 2013, pp. 237-258. [62] H. M. Trice and J. M. Beyer, "Studying Organizational [45] R. F. Zammuto, T. L. Griffith, A. Majchrzak, D. J. Cultures through Rites and Ceremonials", Academy of Dougherty and S. Faraj, "Information Technology and the management review 9(4), 1984, pp. 653-669. Changing Fabric of Organization", Organization Science [63] H. M. Trice and J. M. Beyer, The Cultures of Work 18(5), 2007, pp. 749-762. Organizations, Prentice Hall, Englewood Cliffs, NJ, 1993. [46] V. Kaptelinin and B. Nardi, "Affordances in Hci: [64] D. Grant, C. Hardy, C. Oswick and L. Putnam, eds., Toward a Mediated Action Perspective", Proceedings of "The Sage Handbook of Organizational Discourse", SAGE, the SIGCHI Conference on Human Factors in Computing Thousand Oaks, CA, 2004. Systems, ACM, 2012, pp. 967-976. [65] S. Koch and S. Deetz, "Metaphor Analysis of Social [47] A. Majchrzak and A. Malhotra, "Towards an Reality in Organizations", Journal of Applied Information Systems Perspective and Research Agenda on Communication Research 9(1), 1981, pp. 1-15. Crowdsourcing for Innovation", The Journal of Strategic [66] S. G. Harris, "Organizational Culture and Individual Information Systems 22(4), 2013, pp. 257-268. Sensemaking: A Schema-Based Perspective", Organization [48] D. A. Askay, "Silence in the Crowd: The Spiral of Science 5(3), 1994, pp. 309-321. Silence Contributing to the Positive Bias of Opinions in an [67] M. Ezzamel and H. Willmott, "Strategy as Discourse Online Review System", New Media & Society 17(11), in a Global Retailer: A Supplement to Rationalist and 2014, pp. 1811-1829. Interpretive Accounts", Organization Studies 29(2), 2008, [49] E. Schenk and C. Guittard, "Towards a pp. 191-217. Characterization of Crowdsourcing Practices", Journal of [68] H. M. Zoller, "Health on the Line: Identity and Innovation Economics & Management (1), 2011, pp. 93- Disciplinary Control in Employee Occupational Health and 107. Safety Discourse", Journal of Applied Communication [50] A. Kerr and J. Kelleher, "The Recruitment of Passion Research 31(2), 2003, pp. 118-139. and Community in the Service of Capital: Community [69] S. Mantere and E. Vaara, "On the Problem of Managers in the Digital Games Industry", Critical Studies Participation in Strategy: A Critical Discursive in Media Communication 32(3), 2015, pp. 177-192. Perspective", Organization Science 19(2), 2008, pp. 341- [51] D. Ribeiro Soriano, F. J. Garrigos-Simon, R. Lapiedra 358. Alcamí and T. Barberá Ribera, "Social Networks and Web [70] J. Kücklich, "Precarious Playbour: Modders and the 3.0: Their Impact on the Management and Marketing of Digital Games Industry", Fibreculture 5, 2005. Organizations", Management Decision 50(10), 2012, pp. [71] T. Terranova, "Free Labor: Producing Culture for the 1880-1890. Digital Economy", Social Text 18(2), 2000, pp. 33-58. [52] G. Kunda, Engineering Culture: Control and [72] M. L. Gray, S. Suri, S. S. Ali and D. Kulkarni, "The Commitment in a High-Tech Corporation, Temple Crowd Is a Collaborative Network", CSCW 2016 University Press, Philadelphia, PA, 2006. Proceedings, 2016. [53] K. A. Merchant, Control in Business Organization, [73] D. A. Askay and L. Gossett, "Concealing Pitman Publishing, Marshfield, MA, 1985. Communities within the Crowd Hiding Organizational [54] M. Foucault, Discipline and Punish: The Birth of the Identities and Brokering Member Identifications of the Prison, Vintage, New York, NY, 1979. Yelp Elite Squad", Management Communication Quarterly [55] J. R. Barker and G. Cheney, "The Concept and the 29(4), 2015, pp. 616-641. Practices of Discipline in Contemporary Organizational [74] "Yelp Elite Squad", www.yelp.com/elite, Accessed Life", Communications Monographs 61(1), 1994, pp. 19- 9/20/2015 43. [75] M. Humphreys and A. D. Brown, "Narratives of [56] G. Morgan, Images of Organization, SAGE, Thousand Organizational Identity and Identification: A Case Study of Oaks, CA, 2006. Hegemony and Resistance", Organization Studies 23(3), [57] N. Phillips and C. Oswick, "Organizational Discourse: 2002, pp. 421-447. Domains, Debates, and Directions", The Academy of [76] S. R. Clegg, Frameworks of Power, SAGE, Thousand Management Annals 6(1), 2012, pp. 435-481. Oaks, CA, 1989. [58] S. Sveningsson and M. Alvesson, "Managing [77] C. R. Scott, Anonymous Agencies, Backstreet Managerial Identities: Organizational Fragmentation, Businesses, and Covert Collectives: Rethinking Discourse and Identity Struggle", Human Relations 56(10), Organizations in the 21st Century, Stanford University 2003, pp. 1163-1193. Press, Stanford, CA, 2013.

1736