Data Science as Political Action: Grounding Data Science in a Politics of Justice

Abstract

In response to public scrutiny of data-driven algorithms, the field of data science has adopted ethics training and principles. Although ethics can help data scientists reflect on certain normative aspects of their work, such efforts are ill-equipped to generate a data science that avoids social harms and promotes social justice. In this article, I argue that data science must embrace a political orientation. Data scientists must recognize themselves as political actors engaged in normative constructions of society and evaluate their work according to its downstream impacts on people's lives. I first articulate why data scientists must recognize themselves as political actors. In this section, I respond to three arguments that data scientists commonly invoke when challenged to take political positions regarding their work. In confronting these arguments, I describe why attempting to remain apolitical is itself a political stance-a fundamentally conservative one-and why data science's attempts to promote "social good" dangerously rely on unarticulated and incrementalist political assumptions. I then propose a framework for how data science can evolve toward a deliberative and rigorous politics of social justice. I conceptualize the process of developing a politically engaged data science as a sequence of four stages. Pursuing these new approaches will empower data scientists with new methods for thoughtfully and rigorously contributing to social justice.

From Ethics Washing to Ethics Bashing: A Moral Philosophy View on Tech Ethics

“…[49] Ethnographers speak of "ordinary ethics" as the descriptive way ethics and morality structure routine social interaction. [13] Zigon however emphasizes the importance of distinguishing routine and unconscious moral claims from conscious ethical claims that arise during "breakdown" moments and are aimed at changing a culture and at "returning to the unreflective mode of everyday moral dispositions". [50] While Zigon's anthropological perspective on morality and ethics captures the pivotal role played by moments of breakdown and moral dilemma, he still sees morality and ethics as fundamentally about the need to return to unreflected normality, to revise beliefs so they can be fixed, routinized, and remain unchallenged once again.…”

“…Their efforts to promote and arguably build more trustworthy and ethical AI indicate a calculative stance, a method for preempting financial and reputational risk, more than a recognition of the political nature of AI and its implications. [13,14,16] Even though it might be argued that the intentions behind these initiatives are good, the practices themselves are too limited and opportunistic to be in line with a conception of morality and ethics as reflexive capacious exercises that can foster disinterested selfless change. Overall, speaking of AI "ethics" instead of AI "politics" can be seen as a way to depoliticize and normalize the impacts of company efforts in this space, [14] allowing companies to "ethics wash" their reputations and to narrow the space for real debate and change in AI.…”

“…They have been criticized for bringing about a performative "transformation of ethics and design into discourses about ethics and design", [11] a routinized checklist approach to ethics that is powered by capitalist logics and a technosolutionist mindset. [13] Companies are "learning to speak and perform ethics rather than make the structural changes necessary to achieve the social values underpinning the ethical fault lines that exist". [13] For Greene, Hoffmann, and Stark, these practices are both too focused on technical tweaks, blinded by technical concerns about how to embed fairness and accountability within machine learning systems and neglectful of structural injustice, and are universalist projects "justified by reference to a hazy biological essentialism".…”

“…This is a tendency, common amongst non-philosophers, to simplify the issues around tech "ethics" and "moral philosophy" either by drawing a sharp distinction between ethics and law and defining ethics as that which operates in the absence of law [12] or by conflating all forms of moral inquiry with routine politics, for instance by merging or drawing artificial separations between the frameworks of "ethics", "justice", and "political action". [13,14] Distinguishing between "law" and "ethics" is a common legal positivist move, configurable within a long philosophical tradition that sees the practice of making, interpreting, and applying law as processes whose existence and relevance are distinct and separable from their moral and societal implications. [15] The relation between "ethics", "justice", and "political action" instead is complex.…”

“…As described by Jacob Metcalf, Emanuel Moss, and Danah Boyd, the distinction between narrow "ethics" and capacious "justice" became a central focus of discussions during the 2019 ACM Conference on Fairness, Accountability and Transparency. [13] Equating serious engagement in moral argument with the social and political dynamics within ethics boards or understanding ethics as a methodological stance that is antithetic to-instead of complementary to and inherent in-serious engagement in law-making and democratic decision-making, is a frequent and dangerous fallacy. The misunderstandings underlying the broad trend of ethics bashing are at least three-fold: (1) philosophy is either confused with "self-interested politics" or understood in opposition to law, justice, political representation, and social organizing; (2) philosophy and "ethics" are seen as a formalistic methodology, vulnerable to instrumentalization and abuse, and thus ontologically flawed; and (3) engagement in moral philosophy is downplayed and portrayed as mere "ivory tower" intellectualization of complex problems that need to be dealt with through alternative and more practical methodologies.…”