Justice in Numbers: Statistics and the Transformation of Civil Rights in Modern America

Abstract

This dissertation examines the convergence of the struggle for civil rights led by Black Americans and the precipitous rise of government quantification in the twentieth-century United States. Focusing on litigation campaigns waged by the NAACP Legal Defense Fund against employment discrimination and the death penalty, it charts the trajectory of strategies to prove racial discrimination with statistics. It argues that technical standards and social scientific theories shaped legal ideas about fairness, and that the retrenchment of antidiscrimination law was as much a struggle over knowledge as politics, doctrine, or cultural memory. To make these claims, Justice in Numbers traces the “disparate impact” theory of discrimination from its emergence in the 1960s to its codification in the 1970s through the backlash against it in the 1980s. At the peak of the Civil Rights Movement, concerns about racial statistics divided activists and administrators who embraced data to hone litigation as a weapon against structural inequality. Lawyers collaborated with social scientists to mount a constitutional challenge to the death penalty rooted in racial disparities research. Their efforts turned debates on capital punishment into proxy battles over criminal justice expertise amid a rising tide of law-and-order politics. Meanwhile, statistics secured key victories in employment lawsuits. But what social science giveth, it taketh away, as disciplinary dissensus threatened to undermine disparate impact theory. By the Reagan presidency, opponents of civil rights no longer rejected quantification. Rather, they excelled at it, moving law’s benchmark for discrimination even further from the structural and historical concerns that inspired the adoption of statistics in the first place. To tell the crucial story of how quantification shaped legal strategy and colorblind politics, Justice in Numbers uses methods from the history of science and Black studies to analyze milestone cases, providing a new perspective on computing, race, and social power in modern America. Past efforts to mobilize computational modeling and scientific anti-racism in the courtroom have shored up pernicious standards that occlude the historical processes of racism. Scholars confronting the challenges presented by AI and machine learning to civil rights today should take heed of this history, lest it be repeated.