Data Feminism

Overview

Data Feminism (2020) by Catherine D'Ignazio and Lauren Klein is a manifesto and a method for rethinking data science from a feminist perspective. Its founding claim is that "data feminism is about power": every dataset, every visualization, every algorithm is a product of choices about what to count, who to include, and whose interests to serve — and those choices are made by people with power, usually to the benefit of people with power. The book offers seven principles for doing data work otherwise: examine power, challenge power, elevate emotion and embodiment, rethink binaries and hierarchies, embrace pluralism, consider context, and make labor visible.

The book is written for practitioners as well as critics: it shows, with dozens of concrete examples, how data feminism can be put into practice in data collection, analysis, visualization, and communication. It draws on intersectional feminism, critical race theory, disability studies, and the long history of feminist science studies, and it insists that the "objectivity" of data is a myth that serves the status quo. Published in open access by MIT Press, the book has become the standard introduction to the field of feminist data studies.

Core Ideas

Data Feminism Is About Power

The book's first and governing principle is that data feminism is about power. Every data project begins with decisions that are political: what is counted, what is not, who is included in the dataset, who is left out, who owns the data, who benefits from the analysis. The claim to neutrality — "the data speak for themselves" — is, in D'Ignazio and Klein's analysis, a claim made by those whose power is served by the status quo. The task of data feminism is to "examine power" and "challenge power": to ask who profits from a dataset, and to build data practices that redistribute rather than entrench advantage.

Rethink Binaries and Hierarchies

Feminist thought has long criticized the binary oppositions that structure Western knowledge — man/woman, nature/culture, mind/body, objective/subjective — and data feminism applies this critique to the binary thinking embedded in data science: male/female checkboxes, black/white categories, included/excluded thresholds. These binaries are not neutral descriptions; they erase, flatten, and rank. The principle to "rethink binaries and hierarchies" asks data workers to resist the default categories, to make room for multiplicity and intersection, and to notice when a "simple" variable is doing complex political work. Data feminism refuses the false choice between "objective" data and "subjective" experience.

Elevate Emotion and Embodiment

Against the ideal of the detached, disembodied analyst, data feminism insists that "emotion and embodiment" are sources of knowledge, not noise. The data workers most likely to notice harm are those closest to it — the communities being measured, the patients being scored, the workers being optimized. Elevating emotion and embodiment means taking seriously the felt experience of data subjects, building analysis from the standpoint of those affected, and refusing the fantasy of the view from nowhere. This principle draws directly on the tradition of feminist standpoint epistemology: the marginalized are not biased observers but privileged ones, because they see what the powerful cannot.

Embrace Pluralism and Consider Context

Data feminism is pluralist: it insists that there is no single correct answer or universal method, and that the best analyses are built from many perspectives in coalition. It is also contextual: it demands that data be understood in the full situation of its production and use, including the histories of colonialism, racism, and sexism that shape who is counted and how. The principles of pluralism and context are the book's answer to the failures of "all-powerful" algorithms — models that ignore the local, the historical, and the human. The final principles, to "consider context" and "make labor visible," complete the program: data work is work, done by people, embedded in institutions, and its true costs — human and environmental — must be counted too.

Counterdata and Data Justice

The book's practical program centers on counterdata: the deliberate production of data that challenges official accounts. When the state fails to count, activists count for themselves; when institutions erase, communities document. Counterdata is the data-feminist answer to the weaponization of statistics: the tools of measurement, turned against the measurers. The book's examples range from the grassroots documentation of police violence to participatory mapping of environmental harm, and its overall argument is that data justice is not a technical fix but a democratic practice — the right of communities to shape the data that shapes them.

Make Labor Visible and the Work of Data

The seventh principle of data feminism is to "make labor visible." Data work — collection, cleaning, coding, analysis — is vast, skilled, and almost entirely invisible: it is the hidden labor of data janitors, content moderators, and crowd workers, disproportionately performed by women and people of color in the global South. The principle demands that this labor be counted, credited, and compensated, and that the products of data science not be presented as if they emerged from nowhere. Making labor visible is also a method of critique: when we see who builds the systems, we see whose interests they serve. The principle connects the book to the broader movement for data justice, which insists that the workers of the data economy are as much its subjects as its beneficiaries.

Historical Context

Data Feminism appeared in 2020, at the height of the public reckoning with algorithmic harm: the #MeToo movement, Black Lives Matter, and the COVID-19 pandemic had all made visible the political stakes of data and counting. The book belongs to a longer tradition of feminist science studies — from Donna Haraway's situated knowledges to Sandra Harding's standpoint theory — and to the emerging field of critical data studies. Its open-access publication reflected its own politics: the authors practiced what they preached, making the book available to the communities it was written for.

Legacy

Data Feminism has become the standard text of feminist data studies and a touchstone for the broader movement for data justice. Its seven principles are used in classrooms, research labs, and policy discussions as a framework for evaluating data work, and its insistence that data science is political has moved from the margins to the mainstream of the field. The book connects feminist epistemology to information ethics in a way that is both rigorous and practical.

Critics have asked whether a "feminist" framing is the most capacious umbrella for all the injustices data can serve, and whether the principles can be operationalized without losing their critical edge. The authors' answer is that the principles are a beginning, not a checklist: data feminism names a tradition and a commitment, and it invites — indeed requires — ongoing critique. As automated systems extend into every domain of life, Data Feminism remains the essential argument that the data revolution must be a justice revolution, or it will not be a revolution at all.

The book's principles have become a framework taught in classrooms and applied in research labs, and its insistence that data feminism is about power has moved from provocation to premise. Its deepest contribution is to have shown that data justice is not a technical fix but a democratic practice — and that the work of making it real is never finished.

Sources

• D'Ignazio, Catherine, and Lauren Klein. Data Feminism. Cambridge, MA: MIT Press, 2020.
• D'Ignazio, Catherine, and Lauren Klein. Data Feminism. Open access edition. Cambridge, MA: MIT Press, 2020. https://data-feminism.mitpress.mit.edu/.
• Benjamin, Ruha. Race After Technology: Abolitionist Tools for the New Jim Code. Cambridge: Polity Press, 2019.