Social data governance: From reflective practices to comparative synthesis

Abstract

This special theme brings together reflections and deliberations regarding the design, implementation, and development of data governance. By addressing “social data governance” as the keyword of the special theme, we aim to further the discussion on a contextual understanding of both the governing foundations and effects of data, dataism, and datafication in different societies. Such a discussion reminds us to pay particular attention to—and thus account for—the social dynamics that underpin and contextualize the design, operation, and promotion of quantified governing mechanisms in which information on social behaviors is collected, datafied, manipulated, and represented. Essentially, the social dynamics of data governance have existed for a long time and in many forms, ranging from credit bureaus’ scrutiny, evaluation, and labeling of their customers to internet-enabled massive data collection and scoring systems used by governments, and to automated contact tracing techniques as a centerpiece of dataveillance and infection control amid the COVID-19 pandemic. Nevertheless, scholarly work from a wide range of disciplines like law, mathematics, and business and with diverse geographical foci has not yet been comparatively and reflectively articulated. Being rich and diverse, the special theme advances such a requisite understanding of the status and relevance of social dynamics of data governance mechanisms based on a wide range of empirical cases around the globe. To scrutinize the social dynamics helps illuminate and contrast divergent manifestations of data governance and their underlying mechanisms.

This article is a part of special theme on Social Data Governance. To see a full list of all articles in this special theme, please click here: https://journals.sagepub.com/page/bds/collections/socialdatagovernance

While datafication (Mayer-Schönberger and Cukier, 2013: 198) has come to be seen as an inherently normative mechanism in various social institutions in contemporary societies (e.g. Jarke and Breiter, 2019; Ruckenstein and Schüll, 2017; Sadowski, 2019), the process—and, in a broader sense, the governance—of this mechanism remains contextually situated and geographically dispersed. In practice, driven by increasing concerns over surveillance, privacy, security, and law enforcement, governments are erecting borders in data governance, such as efforts regarding data sovereignty (e.g. Hummel et al., 2021; Vatanparast, 2020) and data localization (e.g. Selby, 2017; Taylor, 2020) that address the maintenance, process, and control of datafication within the boundaries of its state of origin. In scholarship, some challenge the assumed “digital universalism” (Loukissas, 2019) or “data universalism” (Chan, 2021)—as Milan and Treré elaborate, the tendency to “assimilate the heterogeneity of diverse contexts and to gloss over differences and cultural specificities” (Milan and Treré, 2019: 324) in (the governing of) datafication. Instead, “different geographical, political, social, organizational, and jurisdictional contexts … affect roles and power in the (data) governance discourse” (Micheli et al., 2020: 4). Others uncover distinct sociotechnical imaginaries (Jasanoff and Kim, 2015) for the governance of data and datafication in terms of, for instance, diverse stakeholders (Micheli et al., 2020) or institutional and legal contexts (Guay and Birch, 2022). All of these speak to the necessity and urgency to integrate contextually specific concerns and considerations when engaging with the relationship between data(fication) and governance.

Seeking the social in data governance

This special theme joins the existing efforts against a one-size-fits-all elaboration or solution regarding the design, implementation, and development of data governance (e.g. Delacroix and Lawrence, 2019; Guay and Birch, 2022; Micheli et al., 2020). By addressing “social data governance” as the keyword of the special theme, we aim, more specifically, to advance the discussion on a contextual understanding of both the governing foundations and effects of data, dataism, and datafication (van Dijck, 2014) in different societies. While an explication of the keyword is detailed further in Liu's (2022) discussion of social data governance, here we suggest that this emphasis reminds us to pay particular attention to—and thus account for—the social dynamics that underpin and contextualize the design, operation, and promotion of quantified governing mechanisms in which information on social behaviors is collected, datafied, manipulated, and represented. To do so, we seek to avoid the pitfalls of the “asocial and ahistorical”—that is, unpacking data governance as “something operating outside of history and of specific sociopolitical, cultural, and economic contexts” (Milan and Treré, 2019: 325).

Essentially, the social dynamics of data governance have existed for a long time and in many forms, ranging from credit bureaus’ scrutiny, evaluation, and labeling of their customers (Lauer, 2017; Pasquale, 2015) to internet-enabled massive data collection and scoring systems used by governments—for example, the social credit systems in China (e.g. Liang et al., 2018) and Aadhaar in India (e.g. Singh, 2021); those used by tech companies to rate or score citizens at an increasing number of occasions—for example, “The Scored Society” (Citron and Pasquale, 2014); automated contact tracing techniques as a centerpiece of dataveillance—that is, surveillance using digital methods (Clarke, 1988; van Dijck, 2014); and infection control amid the COVID-19 pandemic (e.g. Sweeney, 2020; Whitelaw et al., 2020). While digitalization may make datafication an intense and prominent phenomenon nowadays, as Guay and Birch point out, an understanding of the governing of datafication necessitates an interrogation of “the pre-histories of contemporary data governance” (Guay and Birch, 2022: 10, emphasis in original). In other words, just like Iliadis and Russo indicate, “[d]ata, along with its sciences and infrastructures, are informed by specific histories, ideologies, and philosophies that tend to remain hidden” (Iliadis and Russo, 2016: 2). These histories, ideologies, and philosophies not only underpin specific data governance practices but also sustain and legitimize the consequence of the existing governing mechanism. As elaborated further by scholarship, these entrenched histories, ideologies, and philosophies may exacerbate already existing deep-rooted vulnerabilities and inequalities (Arora, 2019; Milan and Treré, 2019; O’Neil, 2016). A comprehensive picture of the social dynamics of data governance must thus avoid rendering either data or governance into value-freed and value-neutral phenomena out of the deeper meanings and the complexity of its context.

Beyond the scrutiny of social dynamics in data governance, this special theme brings together reflections and deliberations that allow badly needed comparative discussion in the study of data and datafication governance. While this field is growing exponentially, scholarly work from a wide range of disciplines like law, mathematics, and business (e.g. Ladley, 2019; O’Neil, 2016; Zuboff, 2019) and with diverse geographical foci (e.g. Khatri and Brown, 2010; Singh, 2021; Tupasela et al., 2020; van Dijck et al., 2018), has not yet been comparatively and reflectively articulated. Flensburg and Lomborg (2021: 15) stipulated that a comparative focus should include studies of datafication across the globe. To scrutinize the social dynamics thereby helps illuminate and contrast divergent manifestations of data governance and their underlying mechanisms. For instance, who are the stakeholders when defining what kind of behavior-associated data should be governed and how? Who holds the decision rights and accountability, and in what sense, regarding behavior-related data assets, and who has influence in the decision-making process of data-related policy and practices? How are the answers to the above questions engendered by specific political systems and social settings? Is universal “data justice” (Dencik et al., 2016; Taylor, 2017), or a contingency model in data governance that entails relevant contextual considerations, pursuable (Arora, 2019; Weber et al., 2009)? Could disparate types of data governance (Arora and Stevens, 2019; Redden et al., 2020) across geographical borders better understand one another, thereby enabling mutual support, conversation, or collaboration? How can one overcome data universalism (Milan and Treré, 2019) and reduce the big data divide (Andrejevic, 2014) between the paradigmatic binary of the global North and South (e.g. Arora, 2016)? Being rich and diverse, the special theme advances such a requisite understanding of the status and relevance of social dynamics of data governance mechanisms based on a wide range of empirical cases around the globe.

Grounding frameworks for comparison and generalization

This special theme includes four articles that examine the social dynamics of data governance from different perspectives. These perspectives, as Miller and Mansilla (2004: 4) contended, essentialize different ways of “seeing and thinking that…[are]…based on commitment to a system of theories, a body of professional knowledge, a discipline, or a discourse community.” The richness and diversity of the articles thus exemplify the heart of our understanding of interdisciplinarity that does not simply juxtapose but deeply intertwines disciplines. In such a way, each article contributes to the deliberation on governing mechanisms in specific social contexts. Simultaneously, the four research and commentary pieces together demonstrate a set of frameworks for comparative studies of the social dynamics of data governance. This framework further shows how these discussions could be generalized for theoretical or practical applications (see Table 1 for a summary of the comparative framework).

Table 1.

Framework for comparative studies on social dynamics of data governance.

Perspectives for comparison	Variables
Political/Cultural ideologies	- Authoritarianism- Libertarianism- Communitarianism- Individualism- Combination of any of the above
Communicative approaches	- Transactional approach to the public- Dialogical approach to the public- Open systems approach to data work
Governing rationales	- Democratic values- Market capitalism- Privacy protection- Efficiency- Pro-innovation
Governing actors	- Governmental actors- Non-governmental actors- Corporate actors

Liu's “Social data governance: Towards a definition and model” draws attention to the role of socio-political cultures in shaping the modes of data governance from a comparative perspective. As summarized in “Political/Cultural ideologies” in Table 1, the two pairs of contradictions—authoritarianism versus libertarianism and communitarianism versus individualism—form a four-quadrant model which allows Liu to categorize data governance arrangements in different countries along the four dimensions. This model sheds light on the ideological and historical grounding of data governance. It also encourages future research on data-based public management to examine the social dynamics of data governance, which is often beyond data practices and overrides privacy protection. Such driving forces are often not determined by the governing organizations alone but also by the interactions among the government, the public, and governing executives in historically, socially, and politically specific ways. Based on the proposed model, Liu further exemplifies how Hungary, Singapore, China, Japan, South Korea, and Norway govern digital contact tracing practices during the COVID-19 pandemic. Theoretically, this article reminds data governance researchers to welcome the conversation between practice-oriented fields and theory-based disciplines. As an emerging and practical topic, data governance needs to return to and be nurtured by pertinent political science and public management theories. In turn, data governance research also has considerable potential to contribute to political and management theories.

Tan and Lim's commentary on Singapore's COVID-19 contact-tracing initiative resonates with the authoritarian/communitarianism quadrant in Liu's model. It zooms in on the case of the TraceTogether app promoted by the Singapore government. As the authors point out, to get the TraceTogether app adopted and diffused at a societal level, the Singapore government took a transactional approach to communicating with the public. Information about the app was sent to the public from the top down, and the government tended to protect app users’ privacy only through legislative means. While the government's choice encountered resistance from the public, this commentary suggests instead a dialogical approach to informing the public, particularly to enhance the state-public conversation on social media (see the “Communicative approaches” summary in Table 1). Tan and Lim argue that the technocratic bent of the Singapore government has enabled the governor's transactional approach. This comment reinforces Liu's positioning of Singapore in the four-quadrant model, but it is also a critical reflection of Liu's argument. If Liu stresses more that socio-political culture (not data performances and practices) shapes the mode of data governance, Tan and Lim's analysis points out that data practices also constitute and reinforce technocratic cultures endorsed by the Singapore government.

Similarly, Cheong and Nyaupane tackle the communicative issues surrounding the app-mediated interaction between the governing and the governed. The authors choose a more specific context, an American research university that values innovation, to explore how the Internet of Things, in the form of a Smart Campus, was promoted. This specific research unit represents a sector of governing institutions in the US that tries to integrate democratic value with market capitalism and accommodate collisions between digital innovation and resistance to privacy violations (see the summary of “Governing rationales” in Table 1). Based on focus group interviews with the students, this article recommends that stakeholders in data governance practices should preserve space for human communicative opportunities, and networked computing program and sensor regulations should value the voice and experience of the users. Furthermore, data governance is not only about privacy protection. It is also a matter of digital participation and social inclusion. As Cheong and Nyaupane show, the governing actors need to ensure that everyone in the university community has equal access to the informational services provided in the hyper-connected environment.

While these three studies, broadly speaking, explore data assemblages—“the technological, political, social and economic apparatuses and elements that constitute and frame the generation, circulation and deployment of data” (Kitchin and Lauriault, 2014: 1)—Wang draws attention to the social consequences of the new way of corporate management and governance shaped by extensive data practices. Based on semi-structured interviews and participatory observation in business meetings at three fintech companies in China, Wang asserts that data governance is widely considered an advanced version of data management work in the big data era. In this context, to turn data into the company's assets, an open-system approach to management (in Table 1, a new communicative approach between the data-related stakeholders) and the affordance of big data are considered the dynamics of innovation, the key to break through economic stagnation and deal with economic crises of all kinds. Nevertheless, by closely examining how this data-based ideology of innovation (summarized in “Governing rationales” in Table 1) has been realized in daily work, Wang points out that such innovations are somewhat performative. At the executive level, data governance is an indispensable strategy for governing the risks brought by business competition: if your competitors are building a data governance infrastructure (the so-called middle ground) and you are not, you are already at risk. Yet, at the staff level, the essential tasks of data governance fulfilled by programmers and product managers aim more at building a clean database and expansive digital infrastructure than at generating new values for users or society at large in the specific Chinese context.

All four articles, in explicit and implicit ways, consider non-governmental actors (see the summary of “Governing actors” in Table 1), including citizens, students, administrative officers, programmers, and business executives, as relevant stakeholders in the governance of big data practices. While Liu (2022) and Tan and Lim (2022) remain largely focused on the governing agencies, what is at stake is the social dynamics, namely the political cultures and governing ideologies shaped by the long-term interactions between the government and the public. The four pieces also share an understanding of “the social” as an ensemble of actors and their interactions, app-mediated communications involved in the governing process, as opposed to the government, legislators, and regulators in sole charge of data governance work. In accordance, the coverage of data governance work has been redefined by these social actors.

Despite our navigation toward a comparative synthesis of the social dynamics of data governance, several aspects of the topic remain unchecked. All four contributions take a grounded approach and use qualitative methods to identify the understudied aspects of data governance. While they reveal or point out new research directions, this interrogation would benefit from more nuanced data and longer-term observation of the social dynamics of data governance. For now, at least two types of work are needed to remedy the flaws. One is empirical case studies examining the data-based governance of social behaviors. The key question is: how should the governing stakeholders deal with data that are highly unstructured, such as information about social behavior, images, audio, and video data? The other type of work is a historical analysis of how non-governmental stakeholders join or are assigned as the governing subjects in the transition from the pre-digital age to the big data era.

In addition to these future agendas, we expect this special theme will lead to rethinking the fundamental concepts and interconnections between data, datafication, digitalization, and governance. In other words, as data governance has become a social project, the questions of what should be considered as data and thus subject to what sense of governance and regulation, who has the right to govern data work, and for whose interest, surely need more sophisticated answers.