Synopsis
Data justice is the political movement that names as decisions the things that look like facts: what to measure, whose experience generates the data, whose categories organise it, whose life is rendered legible by the resulting numbers. Participatory statistics is the practice of making those decisions differently. The Detroit Digital Justice Coalition’s 2009 broadband survey found 40% effective access in communities that federal ISP-reported maps showed as 90% served. The gap was not an accident of measurement. It was the direct consequence of using ISP-reported data rather than resident-reported data, and it determined which communities received infrastructure investment. In Kerala in 1996, the Communist Party of India (Marxist) government devolved 40% of the state’s development budget to local bodies who conducted participatory resource mapping using community-developed survey instruments. The data collected was generated by the community it described. In 2018, Joy Buolamwini’s “Gender Shades” study demonstrated that commercial facial recognition systems misclassified darker-skinned women at rates up to 34 percentage points higher than lighter-skinned men — not because of technical failure but because the training data over-represented the faces of those who built the systems. The First Nations Information Governance Centre’s OCAP principles (Ownership, Control, Access, Possession) and the Maori Data Sovereignty Network’s CARE principles (Collective Benefit, Authority to Control, Responsibility, Ethics) are not aspirational but operational: concrete governance frameworks asserting that data about a community belongs to that community. Participatory statistics is not a soft alternative to rigorous statistics. It is more rigorous, because it cannot afford to be wrong about whose data it is.
This chapter’s core argument is that data justice is not a critique of technical practice but a claim about political architecture — and that realising it requires both political transformation and technical innovation.
In This Chapter
- How the data justice movement’s founding texts and institutional expressions — from Linnet Taylor’s early work on the political economy of data to the African feminist data movement and Indigenous data sovereignty frameworks — constitute a coherent political programme with specific implications for how poverty should be measured
- How D’Ignazio and Klein’s feminist data science methodology provides practical tools for auditing poverty datasets — examining missing data patterns, proxy variables, measurement definitions and their exclusions — in ways that connect technical critique to structural analysis
- How community-controlled measurement projects — Detroit’s neighbourhood data initiative, the CARE principles for Indigenous data, participatory budgeting data systems — demonstrate what measurement accountability looks like in practice, and what their experiences suggest about the technical and political conditions that make it possible
- How the compatibility between participatory data collection and topological analysis opens a pathway toward a poverty measurement practice that is both mathematically rigorous and politically accountable — the technical and political arguments converging on the same conclusion
Connection Forward
Chapter 16 examines one specific case of activist measurement revision — the sustained campaign by poverty researchers and community organisations to modernise and replace the Orshansky line — as a concrete illustration of what politically accountable poverty measurement looks like when pursued through existing institutions.