Strangers to the Algorithm: the Effects of Algorithmic Decision-Making in Removal Proceedings on Due Process Rights

This paper investigates the impact of algorithmic decision-making tools on due process rights, particularly within immigration removal proceedings. Deviating from prior studies, which primarily focused on various measures of fairness or the long-term racial stratification as a consequence of these tools, this research emphasizes the implications for due process—a fundamental legal principle enshrined in the U.S. Constitution. By examining ICE's use of these tools between July 2012 (when they were adopted) and October 2016, the study reveals the ways algorithmic input may inadvertently erode principles of procedural justice, especially adjudicatorial independence, in a setting where the rights of non-citizens are already under significant strain. Through an analysis of FOIA-sourced ICE risk assessment data and detention rates, the study looks at the discretionary behavior of ICE field officers—as a proxy for independent decision-making—and how this behavior is affected by the adoption of algorithmic recommendations and subsequent updates to the algorithmic regime. The results indicate that updates to the algorithmic regimes employed by ICE Officers produced pronounced and sustained reductions in discretionary behavior, providing troubling evidence that the integration of these tools may unduly constrain officer judgment and erode the foundational principle of adjudicative independence within immigration proceedings.