Chapter 39 Further Reading: Race, Representation, and Data Justice

Foundational Works in Critical Data Studies

Benjamin, Ruha. Race After Technology: Abolitionist Tools for the New Jim Code. Polity Press, 2019. The central text for understanding how algorithmic systems reproduce racial hierarchy through design. Benjamin's concept of the "New Jim Code" — the double entendre of Jim Crow racial hierarchy and computer code — is the most productive analytical frame for understanding racially biased political targeting models. Essential reading.

Noble, Safiya Umoja. Algorithms of Oppression: How Search Engines Reinforce Racism. NYU Press, 2018. Noble's analysis of how Google's search algorithms produce racially harmful results focuses on commercial search but the analytic framework transfers directly to political analytics. Her concept of representational harm is particularly relevant to how political data systems can damage communities not just through inaccurate counting but through reductive or stereotyping representation.

Buolamwini, Joy, and Timnit Gebru. "Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification." Proceedings of the 1st Conference on Fairness, Accountability and Transparency (2018): 77–91. The landmark study documenting that commercial face classification systems perform dramatically worse on darker-skinned women. The Gender Shades methodology — disaggregated performance evaluation across demographic groups — is a direct model for algorithm auditing in political analytics.

D'Ignazio, Catherine, and Lauren F. Klein. Data Feminism. MIT Press, 2020. An accessible and rigorous examination of how power shapes data — who collects it, who analyzes it, who benefits, and who is harmed. Chapter 4, "What Gets Counted Counts," is directly relevant to the Census undercount discussion; Chapter 5, "Unicorns, Janitors, Ninjas, Wizards, and Rock Stars," examines diversity and power in data science workplaces.

The Census Undercount

U.S. Census Bureau. 2020 Census Post-Enumeration Survey. U.S. Department of Commerce, 2022. The primary government source on 2020 Census undercount rates by demographic group. The technical reports include methodology documentation for the coverage measurement surveys and estimates of net undercount by race, ethnicity, and geography. Available at census.gov.

Citro, Constance F., Daniel L. Cork, and Janet L. Norwood, eds. The 2000 Census: Counting Under Adversity. National Academies Press, 2004. Though focused on the 2000 Census, this comprehensive review provides essential historical context for understanding the persistent differential undercount problem, its mechanisms, and the Bureau's past efforts to address it. The mechanisms documented here remain operative.

Rodriguez, Clara E. Changing Race: Latinos, the Census, and the History of Ethnicity in the United States. NYU Press, 2000. Examines how Census racial classification categories have historically failed to adequately capture Latino identity, with consequences for both representation and research. Essential context for understanding why the Census undercount of Hispanic communities is not simply a technical problem but reflects a deeper conceptual mismatch.

Polling and Minority Communities

Kuo, Alex, Neil Malhotra, and Cecilia Hyunjung Mo. "Why Do Asian Americans Identify as Democrats? Testing Theories of Social Exclusion and Intergroup Solidarity." Journal of Politics 79, no. 3 (2017): 1020–1036. While focused on Asian American political identification, this study illuminates the challenges of capturing minority political opinion with standard survey methods, including the role of interviewer effects and sample selection in producing systematic measurement error.

Griffin, John D., and Brian Newman. "Are Voters Better Represented?" Journal of Politics 67, no. 4 (2005): 1206–1227. An empirical examination of the gap between minority constituent preferences and congressional roll call voting, which is partly attributable to measurement failures in capturing minority opinion — the methodological concern that drives the affirmative data practices discussion.

Lax, Jeffrey R., Justin H. Phillips, and Alissa F. Stollwerk. "Are Survey Respondents Lying About Their Support for Same-Sex Marriage? Lessons for Analyzing Sensitive Topics." Political Analysis 24, no. 2 (2016): 153–171. Though focused on social desirability bias on LGBTQ topics, this study's methodology for identifying systematic measurement error in sensitive contexts is directly applicable to the problems of measuring minority political opinion with standard polling instruments.

Algorithmic Bias and Political Data

Eubanks, Virginia. Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor. St. Martin's Press, 2018. Eubanks documents three cases of algorithmic decision-making in public services — welfare administration, child protective services, criminal risk assessment — where automated systems reproduce and amplify inequality. The mechanisms she documents (training on biased historical data, proxy variables for race, lack of accountability for algorithmic outputs) are directly applicable to political analytics.

Obermeyer, Ziad, Brian Powers, Christine Vogeli, and Sendhil Mullainathan. "Dissecting Racial Bias in an Algorithm Used to Manage the Health of Populations." Science 366, no. 6464 (2019): 447–453. A landmark study documenting that a widely used commercial healthcare algorithm systematically underestimated the health needs of Black patients relative to white patients — not due to explicit racial variables but because a racially biased proxy (healthcare cost) was used as a training outcome. The mechanism is directly analogous to the political targeting bias described in this chapter.

Voting Rights and Data

Keele, Luke J., et al. "An Overview of the Methodological Literature on Race and Redistricting." Political Analysis (forthcoming). A review of quantitative methods for analyzing racial effects in redistricting, including the use of Census data, ecological inference methods, and algorithmic fairness approaches to evaluating district plans.

McDonald, Michael P. United States Elections Project. The primary web resource for voter turnout data, including turnout by race and ethnicity where available. Essential for anyone working on empirical questions about differential participation. Available at electproject.org.

Data Justice: Practice and Advocacy

Taylor, Linnet. "What Is Data Justice? The Case for Connecting Digital Rights and Freedoms Globally." Big Data & Society 4, no. 2 (2017). The foundational academic article articulating the data justice framework as a global concept. Taylor distinguishes data justice from privacy rights frameworks and conventional accuracy-and-validity frameworks, arguing for a conception centered on equity, power, and democratic accountability.

Data for Black Lives. Data for Black Lives Principles. 2017. The founding document of the Data for Black Lives movement — a network of activists and researchers committed to using data to serve Black communities. The principles provide a practical articulation of the data justice framework from an advocacy perspective. Available at d4bl.org.

Crenshaw, Kimberlé, ed. Mapping the Margins: Intersectionality, Identity Politics, and Violence Against Women of Color. Stanford Law Review, 1991. Though predating the contemporary data justice literature, Crenshaw's framework of intersectionality — the interaction of race, gender, and other identity dimensions in creating distinctive patterns of disadvantage — is essential for understanding why race alone is insufficient as an analytical category in political data work.