Asylum Discord: Disparities in Persecution Assessments
60 Pages Posted: 31 May 2015
Date Written: May 29, 2015
Abstract
Asylum claims often hinge on applicants’ ability to prove that they suffered or fear harm that rises to the level of persecution. Nevertheless, scholars have not comprehensively surveyed how courts measure harm and to what extent courts’ harm assessments are inconsistent. This study undertakes the task. It reviews approximately 900 published asylum cases decided by the federal appellate courts from 1996 through 2013, since the appellate courts have served as the final arbiter for when harm is sufficiently severe. From these 900 cases, the study isolates the decisions where an appellate court specifically determined whether a set of harms is necessarily sufficient to establish an applicant was persecuted. After isolating these decisions, the study then categorizes them based on the type, frequency, and duration of harms sustained by the applicants.
The results demonstrate incredible divergences in most of the segmented categories, from a single instance of abuse and detention to prolonged psychological suffering. The data reveals that asylum applicants’ ability to avoid deportation may depend on the appellate court jurisdiction they happen to fall under or even the particular judges within a given circuit that happen to be assigned to their cases. Disparities in how courts assess and measure harm also contribute to the inconsistent persecution outcomes. The article recommends heightened scrutiny of whether being persecuted requires systematic abuse, increased standardization of harm severity thresholds, greater attention to the details of harm inquiries, and more precise opinion drafting.
The Supplemental Tables for this paper are available at the following URL: http://ssrn.com/abstract=2612282
Keywords: asylum, refugee, judicial process, persecution, persecute, immigration, study, empirical, court of appeals, appellate review, substantial evidence, harm
JEL Classification: K1, K10, K19, K23, K39, K40, K41
Suggested Citation: Suggested Citation