The Human Factor: Algorithms, Dissenters, and Detention in Immigration Enforcement

iLCSS Working Paper | no. 1

13 Pages Posted: 15 Apr 2019

See all articles by Robert Koulish

Robert Koulish

University of Maryland

Ernesto F. Calvo

University of Maryland

Date Written: March 16, 2019

Abstract

This article examines changes in the punitive bias of Immigration and Customs Enforcement (ICE) officers in the 2012 through 2016 period. We define punitive biases as the officers’ higher likelihood of expressing a written dissent with Low Risk Classifications cases when compared to high risk classification ones. We provide evidence that contextual factors enter into the immigration decision process in two different ways: one, by shifting the officers’ frame of references based on the number of cases, the compounded risk portfolio available in processing centers (e.g. the caseloads flow), and in response to external shocks, such as elections. Second, by allowing policy preferences to edit the risk algorithm (i.e. inserting, deleting, and reweighting the importance given to different types of violations). The analysis draws data from 1.4 million immigration detention risk classification assessment cases between 2012 and 2016 received pursuant to the Freedom of Information Act (FOIA).

Suggested Citation

Koulish, Robert and Calvo, Ernesto F., The Human Factor: Algorithms, Dissenters, and Detention in Immigration Enforcement (March 16, 2019). iLCSS Working Paper | no. 1. Available at SSRN: https://ssrn.com/abstract=3355663 or http://dx.doi.org/10.2139/ssrn.3355663

Robert Koulish (Contact Author)

University of Maryland ( email )

College Park, MD 20742
United States
301-405-3175 (Phone)

Ernesto F. Calvo

University of Maryland ( email )

Register to save articles to
your library

Register

Paper statistics

Downloads
6
Abstract Views
117
PlumX Metrics