In April 2018, the American Immigration Council (the Council), alongside the National Immigrant Justice Center, the Women’s Refugee Commission, Kids in Need of Defense, and the Florence Immigrant and Refugee Rights Project, filed requests under the Freedom of Information Act (FOIA) with U.S. Customs and Border Protection (CBP), the U.S. Department of Homeland Security (DHS), the Department of Health and Human Services (HHS), and the Department of Justice (DOJ). The FOIA requests sought records relating to the separation of adult family members from minor children; the process for detaining the adult family members and minor children in different DHS, HHS, or DOJ facilities; and the criminal prosecution of adult family members. As a result of litigation, the Council received approximately 3,500 responsive records totaling tens of thousands of pages, including emails, policy documents and reports, and data files. Most of the records received were created between 2016 and 2019.
The FOIA request spurred the government production of records that informed the Council’s prior reporting on the harm caused by family separation. This tranche of records complements the archive published by The Atlantic in December 2022 and records published by other organizations. The Council’s work here presents an analysis of FOIA records and data interactives it developed from never-before-seen government datasets on family separations.
The analysis uses a mixed-methods design to examine the role government agencies played in facilitating the separation of families at the U.S.-Mexico border. The research integrates quantitative data from internal spreadsheets with qualitative insights from internal email correspondence and policy documents. Both types of data were obtained through FOIA and analyzed independently before being merged for interpretation.
Qualitative Data
The qualitative aspect of the research involved thematic analysis to group core documents into key themes. The Council reviewed the 3,500 documents produced in their entirety and found that approximately 150 documents could be classified under three general themes:
- Records showing how the administration sought to use the media to amplify its message of deterrence then had to control its messaging when public attention shifted to the horrors families faced when separated.
- Internal government communications demonstrating the lack of systems in place to collect information about family separations and challenges in tracking and reunification.
- Records documenting the role oversight agencies and Congress played in exposing and ending the zero-tolerance policy.
Quantitative Data
The quantitative research involved an analysis of spreadsheets detailing U.S. Border Patrol entries of apprehended and separated families. The records productions included ten spreadsheets with data on Border Patrol apprehensions during different time periods between 2017 and 2018. Ultimately, from the ten spreadsheets, we selected a 2,867-page spreadsheet provided by the government in response to the FOIA request. The Council received the spreadsheet in PDF format on all Border Patrol nationwide apprehensions by demographic from October 1, 2017, to October 27, 2018, and it was selected because this dataset had apprehension information of the longest time period of each dataset.
Variables in the spreadsheet obtained included border sector, apprehension date, age, gender, whether the adult was referred for prosecution, citizenship, and separation reason. This dataset involved the largest time frame of all documents received and had a significant number of variables that allowed us to demonstrate the impact of the zero-tolerance policy. Our review of documents also showed that Border Patrol data on separation was more comprehensive than other agencies’ data because Border Patrol initiated most family separations immediately upon encountering individuals crossing the U.S.-Mexico border. By contrast, Office of Refugee Resettlement (ORR) data lagged, and only included individuals who were referred to them.
We scraped the data from the PDF file to an Excel spreadsheet through Python’s pdfplumber library. We then filtered the data to show only entries of individuals with a family separation code listed, as we believe that these were the families who CBP separated at the border. After scraping data from the PDF, we fact-checked the data through public records. Finally, we used descriptive statistics to summarize trends in apprehensions, separations, and adult prosecutions. We subsequently created data interactives through tools such as DataWrapper, Python, and D3.js to visualize the results.
Time Frames: We split the data into four time frames to note any changes in separations affected by the policy.
- Before Zero Tolerance Policy (ZTP)—The period between October 1, 2017 and April 5, 2018, before the ZTP was officially announced or implemented.
- ZTP Announced—The period between April 6, 2018 and May 6, 2018, when the ZTP was announced by the Department of Justice but had yet to be officially implemented.
- ZTP Active—The period between May 7, 2018 and June 20, 2018, when ZTP was active.
- ZTP Ends—The period between June 21, 2018, and October 27, 2018, after the ZTP had officially ended.
Family Separation Reason: The records contained ten listed reasons for family separation:
- Immigration Violation (Prosecuted)
- Other (Prosecuted)
- Criminal History (Prosecuted)
- Fraudulent Claim: No Family Relationship (Prosecuted)
- Fraudulent Claim: No Family Relationship (No Prosecution)
- Fraudulent Claim: Child Determined to be Over 18
- Family Member Hospitalized
- Gang Affiliation
- Extraditable Warrant
- Redacted
In the very small number of cases where two different separation reasons were listed, we grouped them into “Multiple Reasons.” For example, a person whose separation code was listed as “Criminal History (Prosecuted)” twice was kept as “Criminal History (Prosecuted).” But an individual whose separation code listed two different separation codes such as “Criminal History (Prosecuted) and Other (Prosecuted)” was labeled “Multiple Reasons.”
Prosecution Rates: We calculated the adult prosecution rate by filtering the data for adults (18+) with a family separation code. This is because only adults were referred for prosecution. We then divided the number of adults separated who had been referred for prosecution by the total number of separated adults to obtain a percentage.
Family Status: The data showed an inconsistency with an individual’s family status (labeled “Current Demo”). Some adults (18+) who had been labeled as being separated from a family unit were also labeled as single adults, and some children (<18) who had been labeled as being separated from a family unit were labeled as unaccompanied minors. Evidence from other records we received (more in discussion) as well as outside sources demonstrate that Border Patrol deleted family information and classified parents as single adults and children traveling with parents as unaccompanied children in order to facilitate family separation. Thus, rather than focusing on the “family status” variable collected by Border Patrol, we isolated those who had a “family separation” code and described separations based on how many children (<18) were separated from adults (18+).
Research Integration
After conducting separate qualitative and quantitative analyses, the Council merged the two lines of analysis, which enabled a more comprehensive understanding of family separation and the ZTP. The qualitative research pointed to key themes arising from government documents, including responses to oversight from Congress, responses to journalists and advocates, and challenges compiling data and identifying parent-child relationships. The quantitative analysis verified findings from the records reviewed, especially the challenges in compiling data.
