About | HeinOnline Law Journal Library | HeinOnline Law Journal Library | HeinOnline

128 Yale L. J. 2218 (2018-2019)
Bias in, Bias out

handle is hein.journals/ylr128 and id is 2309 raw text is: SANDRA G. MAYSON
Bias In, Bias Out
A B S T R A C T. Police, prosecutors, judges, and other criminal justice actors increasingly use al-
gorithmic risk assessment to estimate the likelihood that a person will commit future crime. As
many scholars have noted, these algorithms tend to have disparate racial impacts. In response,
critics advocate three strategies of resistance: (1) the exclusion of input factors that correlate closely
with race; (2) adjustments to algorithmic design to equalize predictions across racial lines; and (3)
rejection of algorithmic methods altogether.
This Article's central claim is that these strategies are at best superficial and at worst counter-
productive because the source of racial inequality in risk assessment lies neither in the input data,
nor in a particular algorithm, nor in algorithmic methodology per se. The deep problem is the
nature of prediction itself. All prediction looks to the past to make guesses about future events. In
a racially stratified world, any method of prediction will project the inequalities of the past into the
future. This is as true of the subjective prediction that has long pervaded criminal justice as it is of
the algorithmic tools now replacing it. Algorithmic risk assessment has revealed the inequality
inherent in all prediction, forcing us to confront a problem much larger than the challenges of a
new technology. Algorithms, in short, shed new light on an old problem.
Ultimately, the Article contends, redressing racial disparity in prediction will require more
fundamental changes in the way the criminal justice system conceives of and responds to risk. The
Article argues that criminal law and policy should, first, more clearly delineate the risks that matter
and, second, acknowledge that some kinds of risk may be beyond our ability to measure without
racial distortion -in which case they cannot justify state coercion. Further, to the extent that we
can reliably assess risk, criminal system actors should strive whenever possible to respond to risk
with support rather than restraint. Counterintuitively, algorithmic risk assessment could be a val-
uable tool in a system that supports the risky.
A U T H O R. Assistant Professor of Law, University of Georgia School of Law. I am grateful for
extremely helpful input from David Ball, Mehrsa Baradaran, Solon Barocas, Richard Berk, Steph-
anie Bornstein, Kiel Brennan-Marquez, Bennett Capers, Nathan Chapman, Andrea Dennis, Sue
Ferrere, Melissa Hamilton, Deborah Hellman, Sean Hill, Mark Houldin, Aziz Huq, Gerry Leon-
ard, Kay Levine, Truman Morrison, Anna Roberts, Bo Rutledge, Hannah Sassaman, Tim
Schnacke, Andrew Selbst, Megan Stevenson, Lauren Sudeall, and Stephanie Wykstra; for
thoughtful comments from fellow participants in the 2017 Southeastern Junior/Senior Faculty
Workshop, CrimFest 2017 and 2018, and the 2017 and 2018 UGA-Emory Faculty Workshops; for
invaluable research support from T.J. Striepe, Associate Director for Research Services at UGA
Law; and for extraordinary editorial assistance by the members of the Yale Law Journal, especially
Yasin Hegazy and Luis Calder6n G6mez. Title credit to Maron Deering, way back in 2016.


What Is HeinOnline?

HeinOnline is a subscription-based resource containing nearly 3,000 academic and legal journals from inception; complete coverage of government documents such as U.S. Statutes at Large, U.S. Code, Federal Register, Code of Federal Regulations, U.S. Reports, and much more. Documents are image-based, fully searchable PDFs with the authority of print combined with the accessibility of a user-friendly and powerful database. For more information, request a quote or trial for your organization below.

Short-term subscription options include 24 hours, 48 hours, or 1 week to HeinOnline with pricing starting as low as $29.95

Contact us for annual subscription options:

Already a HeinOnline Subscriber?

profiles profiles most