Lives in the Balance: Asylum Adjudication by the Department of Homeland Security

10. The Asylum Officers

>> 177 10 The Asylum Officers In chapter 9, we saw that there was great variability among the grant rates of asylum officers deciding similar cases within the same regional office. What might account for these striking disparities? One asylum officer with whom we spoke closed our interview by saying “there are so many factors that play into the data—the age of the asylum officer, whether they have an old or a new supervisor, whether they are close to retirement, their background, whether they had breakfast that morning.”1 To the extent possible, we used the data that DHS supplied to us to identify relationships between personal characteristics of the asylum officers and grant rates. DHS did not record which officers ate breakfast on which days, but the agency did keep records on many attributes of officers who were trained during a particular five-year period. This chapter studies the 31,635 decisions made by the 221 asylum officers who attended Asylum Officer Basic Training (AOBTC) classes at the Federal Law Enforcement Training Center (FLETC) between July 2003 and July 2008.2 Of those 221 officers, 196 were new hires trained shortly after beginning service, while the other 25 officers had joined DHS before July 2003. Most of the cases in that database (a subset of the larger database used for the analysis in chapters 6 through 9) were therefore decided by officers who joined the asylum officer corps between 2003 and 2008. AtFLETC,theofficerscompletedaquestionnairethatcollectedcertainbiographical characteristics, including their educational degrees, ethnicity, gender , and prior work experience.3 We examined the relationship between these variables and the officers’ asylum grant rates. To confirm the statistical significanceoftheserelationships ,werantwologisticregressions,includingonewith standard errors clustered by asylum officer. We also ran a hierarchical logistic regression. The three regressions contained the same independent variables.4 Unless otherwise noted, the relationships between these variables and the dependent variable of grant rate were statistically significant5 and confirmed thedirectionoftherelationshipspresentedinthecross-tabulationanalyses.6 We also constructed a variable from our primary database of 329,336 decisions on the merits made by 1,232 asylum officers between October 1995 and June 2009. This variable measured, for each case, the number of asylum 178 > 179 rates of male officers for male applicants (44 percent); in other words, female officers granted asylum to male applicants at a rate 7 percent greater than their male counterparts’ rate.9 These rather small differences put in perspective some of the much larger differences reported in this chapter and elsewhere in this book. Prior Government Service Some asylum officers we spoke with speculated that disparities between individual adjudicators within an office were caused by the prior work experiences of the officers. One officer observed that officers who have relatively lower grant rates are those who came into the system earlier, having been INS inspectors or border patrol agents. They never made the transition to asylum. . . . Some mind-sets are not capable of ever letting go. There is always a tangential document in the case that bothers them. Of course there are also people who don’t pay attention, and grant everything. That’s just as bad. Fig. 10-1. Grant Rates by Gender of Officer and Applicant 180 > 181 The grant rates of the Asian or Pacific Islander and White officers were in between these two figures.15 The regression analyses, however, showed that factors other than race contributedtothesedifferences .16 Infact,holdingallothervariablesinthemodelconstant, two of the regressions failed to confirm a statistically significant relationship with grant rates for either the Black or Hispanic categories.17 This lack of significance probably resulted from the fact that the grant rates for these ethnicities (once all othervariablesareheldconstant)wereindistinguishablefromthoseofWhiteofficers (the control category). But in all three regressions, with all other variables in themodelheldconstant,AsianorPacificIslanderethnicityoftheofficercorrelated withgrantrateslowerthanthoseoftheotherthreeself-reportedethnicgroups. Officer’s Ethnicity and Applicant’s Status at Entry The ethnicity variable played out in interesting ways when combined with other variables.18 Black and Asian or Pacific Islander asylum officers granted Fig. 10-3. Grant Rates for Inspected and Uninspected Applicants, by Ethnicity of Asylum Officer 182 > 183 that this smaller database focuses on decisions made toward the end of the time frame of the study. Asian or Pacific Islander, Black, and Hispanic officers granted asylum more often to unrepresented applicants (figure 10-4).20 Black officers granted 24 percent more often to unrepresented applicants. In contrast, White officers granted 9 percent more often to represented applicants. Because there were moreWhiteofficersthannonwhiteofficers,theoverallgrantratewasnearlythe sameforrepresented(46percent...