Race After Technology

Ruha Benjamin’s Race After Technology: Abolitionist Tools for the New Jim Code asks why systems presented as objective, efficient, or progressive can reproduce racial hierarchy—and why the language of innovation can make that continuity harder to see. “New Jim Code” is her name for discriminatory designs that encode inequity while appearing more neutral than the systems they replace. It is an analogy about how racial control is reorganized, not a claim that contemporary software is literally the same thing as Jim Crow law. The book’s feminist value is its insistence that technical choices are social choices: they distribute visibility, risk, labor, and authority.

The book organizes this diagnosis through four linked mechanisms. “Engineered inequity” asks how hierarchy can be built into a device or design; “default discrimination” asks what happens when developers treat a socially unequal world as though it were neutral; “coded exposure” examines the terms on which people become visible or watched; and “technological benevolence” questions products that claim to correct bias while leaving the underlying distribution of power intact. Together these ideas move the conversation beyond whether a particular algorithm contains a measurable error. They ask what history the system inherits, what its designers count as a problem, and who can contest its decisions.

Hannah Robbins Hopkins’s full review in the University of Texas at Austin’s E3W Review of Books describes Benjamin’s first section through robots marketed for policing, military use, sex, and household labor. That range matters for feminist reading: the imagined “helpful” domestic machine and the coercive machine of the carceral state may look like different products, yet both can carry assumptions about who serves, who is served, and whose safety is the priority. The point is not that all robots do the same harm. It is that a technical artifact can make a social arrangement appear natural, inevitable, or beyond public accountability.

The chapter on default discrimination reframes the “glitch.” When an automated system excludes or misclassifies people, the mistake may reveal the design’s ordinary assumptions rather than an isolated malfunction. Hopkins discusses Benjamin’s treatment of predictive policing and recidivism calculations: predictions can acquire an aura of accuracy even as they direct intensified scrutiny toward the people already marked as risky. This creates a feedback loop—surveillance produces records, records train future predictions, and those predictions can justify more surveillance. Calling the outcome a glitch can make a durable political arrangement sound temporary and technical.

“Coded exposure” adds a tension that simple calls for inclusion miss. Some people are rendered invisible when a system cannot recognize or serve them; others become hypervisible as targets of monitoring. Visibility is therefore not automatically recognition or safety. Hopkins notes that Benjamin traces this question through racial representation, eugenic histories, and contemporary technologies. The feminist question is not only who appears in a dataset or image, but who controls the terms of appearance, whose account is believed, and what institutions do with the resulting information.

The section on technological benevolence turns the critique toward well-intentioned fixes. A product may advertise fairness, inclusion, or neutral service yet define the problem so narrowly that it leaves the relevant hierarchy untouched—or makes it more efficient. For Benjamin, a more diverse engineering team or a corrected dataset cannot by itself answer who set the system’s goals, who benefits from automation, or who bears its costs. Her abolitionist horizon asks readers to imagine forms of social life that do not depend on automated policing, exclusion, or extraction, and to treat collective practices as sources of invention rather than obstacles to innovation. That is a more demanding proposal than “make the model less biased”: it asks whether the system should exist in that form at all.

The framework is useful beyond the book’s examples, but it should not be stretched into a universal account of every racial order. Its genealogy is anchored in U.S. histories of racialization, and international examples do not make it a comparative survey of all contexts. Readers applying it elsewhere should ask how local histories, categories, and institutions differ rather than simply importing the Jim Code analogy. For two complementary angles, read Safiya Umoja Noble’s Algorithms of Oppression for search and representation, and Virginia Eubanks’s Automating Inequality for automated welfare and eligibility systems. The chapter examples above are attributed to Hopkins’s review. See the author’s book page, the IUCAT bibliographic record, and the full review.