Immigration Court's Data on Minors Facing Deportation is Too Faulty to Be Trusted
After careful analysis and consideration, TRAC is forced to suspend its publication of data on juveniles facing deportation in Immigration Court due to serious, unresolved deficiencies in the EOIR's data. TRAC's analyses indicate that the data used by the Immigration Court for tracking and reporting on juveniles who are facing deportation appear to be seriously flawed to the point that we question whether the agency has the ability to meaningfully and reliably report on juveniles in its caseload.
We wrote to EOIR's Acting Director Jean King on September 22, 2021 to share TRAC's findings and request "any feedback or insight the agency would like to share on our preliminary findings before we move forward to finalize and publish them." We offered to "share additional details with the agency should [EOIR] find that helpful." TRAC did not receive any response to that letter.
We wrote to the EOIR again on October 15, 2021, this time to Director David Neal who had subsequently been appointed as EOIR's permanent director by Attorney General Merrick Garland. We reiterated our concerns and noted:
TRAC did not hear a response to that letter either. EOIR subsequently published its juvenile tables without commenting on these substantial data quality concerns, and without correcting its methodology.
TRAC is now regretfully withdrawing its own Juvenile App. Because these significant data problems arose only at the time EOIR implemented a series of changes in the latter part of 2017 impacting how unaccompanied juveniles were tracked, the results compiled before these changes occurred will be retained online for use in historical research. However, until EOIR corrects their currently flawed data tracking of juveniles, TRAC will be unable to reinstate updating EOIR's handling of juveniles, whether unaccompanied or accompanied as part of a family group.
TRAC has concluded that these flaws, as detailed below, are so serious that the resulting statistics based on these data are not an accurate or reliable indicator of the quantity or characteristics of juvenile cases currently being handled by the Immigration Court.
Methodology TRAC Used
Under the Freedom of Information Act (FOIA), TRAC previously requested "access to whatever source or sources EOIR is currently using to identify juveniles, as well as to identify that segment who are unaccompanied children (UACs), as each of their cases reach the Immigration Court." The EOIR responded by advising us that it relied solely on the case-by-case records from its "juvenile history" file.
TRAC had submitted this 2021 FOIA request given issues we were increasingly seeing in the juvenile history file and wanting to be sure that the agency hadn't implemented supplemental tracking systems that TRAC wasn't receiving. Given the EOIR's confirmation that the agency relied solely on the juvenile table for its own reporting, TRAC undertook a detailed study analyzing the quality of this data. The study was aided because information recorded by EOIR that would indicate a juvenile status had more recently become available. We used the following three points of comparison in our analyses:
TRAC's Findings Explained
We summarize our findings from each of the comparisons described above.
A. Juvenile History File v. Age at Time of NTA
First, we compiled the reported age at the time of the NTA for all pending cases. The month and year of birth was recorded for most (77%) of these and resulted in a file of 466,397 pending cases where we could determine that they were children 0-17 years of age at the time their NTA was issued. We then compared these cases with all records in the juvenile history file, along with how that file classified them.
These pending juvenile cases based on age, showed little agreement with the juvenile history file. Errors were of two different types: errors of omission and errors of over inclusion. Errors of omission occur when an individual clearly identified as a juvenile in terms of age was not included in the juvenile history table. Errors of inclusion occur when an individual clearly identified as outside the juvenile age range was included in the juvenile history file.
B. Juvenile History File v. Juvenile Court Assignment
We then compared all cases assigned to hearing locations that are specifically set up to handle cases for unaccompanied minors to the coverage in the juvenile file.
Nearly three out of ten (29%) of the individuals whose cases were pending at hearing locations that have been set up to handle cases for unaccompanied children were not included at all in the juvenile history file or were misclassified (not classified as unaccompanied children).
C. Juvenile History File v. Dedicated Docket
We compared cases assigned to the Dedicated Docket with the juvenile history file. We found that nearly nine out of ten (88%) children 0-17 years of age in families seeking asylum who are part of the administration's recent Dedicated Docket initiative were not included in the juvenile history file.
TRAC Calls on Biden Administration to Address Significant Data Quality Concerns
Government agencies should be transparent and accountable to the American public—but this is difficult if they don't collect reliable information on what they are actually doing. The Immigration Court's failure to respond to or address TRAC's findings of significant data quality issues regarding minors is particularly concerning given the highly sensitive nature of children facing deportation.
This data quality problem on tracking juvenile cases adds to EOIR's earlier refusal to address data quality issues regarding asylum cases that were disappearing from the agency's master database which it relies on to manage its workload. A number of these asylum cases continue to disappear each month.
And the problems don't end there. TRAC recently uncovered additional issues when developing a new public web query tool to track the asylum backlog. TRAC found that the EOIR had apparently lost track of about 50,000 additional pending asylum applications. This has led it to falsely report that its asylum backlog had been reduced this past year when in fact it had markedly grown.
Taken individually, each specific issue is significant and noteworthy in its own right. But taken together, these now multiple unresolved data quality issues are compounding upon each other. TRAC has repeatedly offered to work with the EOIR to aid the agency as it seeks an understanding of the problem and a meaningful solution—yet thus far EOIR has continued to ignore its growing data management problems.
The public should be increasingly troubled by the indifference that the Immigration Courts have shown to these issues and should push for improved transparency and accountability.
 New codes were added to the juvenile history file beginning in 2017 that substantially changed how juveniles were being tracked. Up until that time the focus had been on unaccompanied juveniles. Then the coverage was extended to also cover cases for accompanied juveniles - that is, children who were part of a family group. In addition, a new code "not applicable" was added. Because the file is designed to record changes in status, a child can "age out" of a juvenile status. However, even those cases that represented the only record for that individual rather than a change in status soon became the vast majority of all records in the juvenile history file after this new "not applicable" code was added. For example, at the time of our study during FY 2021, we found 79 percent of all records newly added during FY 2021 to the juvenile history file had assigned the "not applicable" code. Further, 40 percent reporting birthdate with this code, were adults at the time their NTA was issued. Why then were these cases on adults so recently added to a file designed to track "juvenile history" included in the first place?
 Some of these were coded "not applicable" in the juvenile history file, but many adults were coded as unaccompanied or accompanied juveniles.
 Recorded month and year of birth with rare exception agreed (>98%) entirely with being a juvenile assigned to the juvenile docket.