This academic post analyses the role played by public reasoning and power in Taylor’s (2017) data justice framework in light of the main discussions I have presented until now in the blog. Firstly, the concept of data justice is explained and complemented by arguments on justice and the contentious politics of data, which highlight public deliberation and universalism as core elements of justice. Secondly, the international dynamics of data power relations is explored based on the definitions of informational state and data colonialism. Thirdly, the study-case of the Brazilian Parliament’s Inquiry Commission on Fake News (from now on, CPMI) is analysed in light of the concepts laid out above. Finally, the conclusion proposes a decolonizing approach to public deliberation and power concerning the data justice framework.
Data justice and the contentious politics of data
Taylor (2017) proposes a framework for data justice that articulates a methodological engagement based on three pillars: visibility, digital (dis)engagement and countering data-driven discrimination. Visibility “deals both with privacy and representation”, relating to issues such as privacy at the social margins and data as a public good; engagement concerns “the freedom to control the terms of one’s engagement with data markets”, for instance in how data processing establishes hierarchical relations; nondiscrimination is composed by “the power to identify and challenge bias in data use, and the freedom not to be discriminated against”, underscoring increasingly complexity of data production and the need for the governance of algorithmic processes and the enforcement of rules by the government against data bias (p.9).
Based on those pillars, Taylor devises a theory of data justice that is informed by public reasoning and operates on a global scale. Firstly, the author argues for a conceptual framework that asks “what kind of organising principles for justice can address the marginalised and vulnerable to the same extent as everyone else” (p.10). Instead of determining such principles, he affirms that public reasoning (“debate, activism and regulation”) should identify the basic needs of people in relation to data technologies and convert those into functionings (‘doings’ or ‘beings’), such as data privacy, digital (dis)engagement and living free from data discrimination (p.10). Then, conversion factors (political, legal and educational support) can translate the functionings into capabilities, like “participation in data value chains” and “access to data affecting oneself” (p.10). Both the transformation of people’s basic data needs into functionings and the conversion of these into capabilities reinforces Sen’s argument that public reasoning plays a central role in the understanding of justice, which in turn leads us to a connection between justice and democracy, since the idea of democracy as government by discussion is widely accepted (2011, p.358). In other words, Sen explains that, if justice demands can only be assessed with the support of public reasoning, which is constitutively related to the idea of democracy, it follows that “there is an intimate connection between justice and democracy, which share discursive characteristics” (2011, p.360).
Secondly, Taylor proposes “a global ecosystemic approach that could look across borders”, since operationalizing data justice would only be possible with the connection between different theories (such as Critical Data Studies and digital geographies) to “policy and law, and particularly to the transnational political responsibilities of online service providers” (2017, p.11). The author advances the idea that these “translations and negotiations” should take place in multiple locations, turning once again to “the domain of public reasoning”, instead of a particular “high-income, high-technology location” (2017, p.11). The outcome of this framework would be a paradigmatic change from an individualistic take on data use responsibility to national and international authorities accountability (2017, p.12). Such a task may look impossible at first, but the history of human rights shows that however imperfect, such efforts may achieve concrete and widespread results in the long term. In fact, Taylor’s proposition on enforcing data justice through global-scale public reasoning echoes Sen’s argument that the defense of human rights is based on the assumption that their underlying ethics is irrefutable and therefore would survive open and well-informed inquiry that considers multiple viewpoints, including national and international experiences (Sen, 2011, p.420).
The relevance in scope and depth of Taylor’s proposition notwithstanding, the inescapable power relations entailed in such a public reasoning process concerning the governance of data justice must be further assessed. By this token, it is important to mention here the concept of data politics, as defined by Ruppert, Isin & Bigo, which “is concerned with not only political struggles around data collection and its deployments, but how data is generative of new forms of power relations and politics at different and interconnected scales” (2017, p. 2). Building on the definition of data politics, Beraldo & Milan devised a complementary concept: the contentious politics of data, or “the multiplicity of bottom-up, transformative initiatives interfering with and/or hijacking dominant, top-down processes of datafication” (2019, p.2). They argue that data activism can be “mapped along two analytical dimensions”: “data-as-stakes (that is, as issues/objects of political struggle in their own right) and data-as-repertoires (in other words, as modular tools for political struggle)” (2019, p.1).
Taylor’s framework entails both analytical dimensions. The public reasoning processes necessary for the process of converting theory into the assessment of basic data needs of people, functionings and then capabilities are basically data-as-stakes debates (for instance, the discussion over a particular norm concerning data privacy), which may also be engaged using data-as-repertoire (for example, by the deployment of social media bots to create a false sense of majority and exert pressure on legislators). In this regard, one should be wary of the mounting power inequalities inside and outside national borders concerning data and what that means for Taylor’s data justice proposition.
The informational state and data colonialism
The idea of informational state is useful for one to grasp how data inequalities translate into power inequalities. When many of the established or trending technologies of today were still only dots in the horizon, Braman already envisioned informational power as the one that “dominates power in other forms, changes how they are exercised, and alters the nature of their effects” (2006, p.26). This kind of power is at the core of what she defines as the informational state, which “uses control over information to produce and reproduce loci of power and to carve out areas of autonomous influence within the network environment”.
Braman (2006) lists 18 trends concerning the social impact of information policy, many of which may hinder open public deliberations on data justice: “the informational state knows more and more about individuals, while individuals know less and less about the state” (p.314), “the subjects of surveillance never know when, how or why they might become visible on the panspectral screen (p.315), “the use of digital technologies may actually decrease, rather than increase, the possibilities of meaningful participatory democracy” (p.315); among other trends. If the informational state tends to show little respect for their own citizens’ information rights, what about the population of technologically vulnerable, poor nations? If history is any guide, no one should expect the informational state to hold back its available resources based on ethical considerations or public reasoning.
In this sense, Braman’s analysis can be complemented by Couldry and Mejia’s concept of data colonialism, “which combines the predatory extractive practices of historical colonialism with the abstract quantification methods of computing” (2018, p.1). The authors identify at least two poles of colonial power, the USA and China. I would name them informational metropoles, since they combine a high level of informational power with practices of data colonialism. Like the chartered companies of yore (for example, Dutch East India Company), “the elites of data colonialism (think of Facebook) benefit from colonization in both dimensions, and North-South, East-West divisions no longer matter in the same way” (p.2). It is noteworthy that the informational state has deep connections to the elites of data colonialism, since many of their technologies are of dual use and the informational state allowance is essential for these companies to profit.
In data colonialism, Couldry and Mejia argue, “life needs to be configured so as to generate” data (2018, p.3). Hence the platform “produces the social for capital, that is, a form of ‘social’ that is ready for appropriation and exploitation for value as data, when combined with other data similarly appropriated” (2018, p.4). In the same way the terra nullius dogma justified colonialism – despite local populations that occupied many of the colonized lands for many generations -, data colonialism is based on the assumption that personal data is freely available for appropriation, like “a natural resource, a resource that is just there” (p.5).
The concepts of informational state and data colonialism demonstrate that power relations pose a major challenge to Taylor’s reliance on public reasoning to advance the basic data needs on a global scale. There is an increasing information gap between the conjunction of the informational states and information elites vis-à-vis the general public. Moreover, the digital platforms in which such debates increasingly occur are themselves devised and regulated by information elites based in and closely related to the informational states. To further illustrate this point, the recent debates over disinformation in Brazil provide evidence on how difficult it may be to promote an open debate over data justice.
The Parliamentary Inquiry Commission on Fake News
A week prior to the final vote on the 2018 Presidential Election in Brazil, an illegal scheme to disseminate “hundreds of million” of Whatsapp messages in favour of Jair Bolsonaro was being prepared. This accusation was the headline of Folha de S. Paulo, the most important newspaper in Brazil in October, 18 of that year (The Guardian, 2018). No relevant measures were undertaken by electoral authorities to investigate the matter and Bolsonaro was elected president. In September, 2019, a Parliamentary Inquiry Commission on Fake News was created by the National Congress to prove what has really happened in the 2018 elections and whether the federal government is behind a series of disinformation / hate speech episodes favouring president Bolsonaro since then. The five-month investigation has held 23 public hearings and gathered over a hundred documents so far, achieving mixed results.
During the CPMI hearings, a witness falsely claimed that the Folha de S. Paulo’s reporter who wrote the original story on the Bolsonaro’s illegal Whatsapp campaign had offered him sex in exchange for information. Immediately, coordinated attacks against her on digital platforms ensued, including not only a virtual army of bots and trolls but one of the president’s sons (CPJ, 2020). No measures were undertaken by digital platforms to stop the harassment and the spread of fake news about the episode from happening. This is a prime example of data-as-stakes debate in which data-as-repertoire is used in an unethical way to derail public reasoning.
Former allies of president Bolsonaro have denounced to the CPMI that there is a “digital militia” formed by parliamentary assistants and government officials connected to the Bolsonaro family, but it has been hard to prove. It should not be. There is mounting suspicion over a number of social media accounts, whose record retention was demanded by the CPMI, including copies of conversations and erased content. However, Facebook and Twitter have denied such measures (DOC 59 and DOC 57 at Senado Federal, 2020), arguing that since the data is property of a US American company, it could only be demanded using the Treaty on Mutual Legal Assistance in Criminal Matters (MALT) between the USA and Brazil. If such a process will be effective remains unclear, since the executive branch of the Brazilian government has no interest in helping the investigations. In addition, WhatsApp has denied access to data history on groups suspected to disseminate fake news by arguing that the company does not store this information (DOC 58 at Senado Federal, 2020). Consequently, the location of social media infrastructure (in the USA) influenced the prerogatives of the Brazilian Congress during the CPMI.
On the other hand, the collaboration of social media companies with the investigations, albeit limited, has finally paid off when a Facebook’s answer to a formal request of the CPMI proved the connection between president Bolsonaro’s son, the federal deputy Eduardo Bolsonaro, to a fake news / hate speech profile (UOL, 2020). Facebook informed that the e-mail account used to register the page was the same address used by Eduardo Bolsonaro’s staff. The conclusion is appalling: fake news and hate speech has been disseminated from inside the Brazilian Parliament, by the president’s son, using public resources to attack fellow congress men and women, the Supreme Court and other democratic institutions. If the discovery shows that it is possible for democratic institutions to assess online misbehaviour, it is also a prime example of how difficult and long such a process may be when the interested party is a developing country.
Conclusion: decolonizing public deliberations in digital platforms
In order to advance the data justice agenda, data colonialism must be addressed accordingly both in public reasoning and (global) power relations. In regard to public reasoning – and in line with Taylor’s approach – “the underlying rationality that enables continuous appropriation to seem natural, necessary and somehow an enhancement of, not a violence to, human development” must be attacked, as proposed by Couldry and Mejia (2018, p.16). Furthermore, Milan and Trere warn that “we continue to frame key debates on democracy, surveillance, and the recent automation turn by means of ‘Western’ concerns, contexts, user behavior patterns, and conceptual frameworks” (2019, p.320). In effect, it is necessary to further investigate “the impact of datafication on the disadvantaged, the silenced, and the invisible” (2019, p.321).
The global ecosystem proposed by Taylor also provides some important insights. Sen advocates that “capability is a kind of power, and it would be a mistake to see capability only as a concept of human advantage, not also as a central concept in human obligation” (2008, p.336). In effect, the responsibility of effective power entails that informational metropoles must consider data inequalities – which may seem naive but is a strong political argument – or that social and political movements in those countries can play a central role in decolonizing data, since they have capabilities toward the informational metropoles that the peoples of informational colonies have not. Finally, information policy implemented by the Global South(s) must advance the interests of its people, not the ones of informational metropoles and the elites of data colonialism. In spite of the importance of the role played by international institutions and NGOs, this challenging task demands a number of interventions by the three branches of the state in developing nations, as demonstrated in the study-case. In summary, a change of rationale is needed so as not to repeat past colonial experiences. It is time to imagine the Global South(s) not only as a locus of information rights or basic needs, but one of informational power.