A study by the Government Accountability Office (GAO) has found that U.S. Border Patrol agents inconsistently documented checkpoint data, including how many smuggled people are apprehended and how many drug seizures were made using canines.
Border Patrol operates immigration checkpoints at more than 110 locations on U.S. highways and secondary roads, generally 25 to 100 miles inland from the southwest and northern borders. According to its own data, from fiscal years 2016 through 2020, Border Patrol apprehended about 35,700 potentially removable people in about 17,500 events at checkpoints. During the same period, Border Patrol seized drugs in about 17,970 events at checkpoints. About 70 percent of Border Patrol checkpoint apprehension events took place in three southwest border sectors (Big Bend, Laredo, and Rio Grande Valley), with fewer than one percent on the northern border. GAO’s study found that most drug seizure events involved only U.S. citizens (91 percent), of which 75 percent involved the seizure of marijuana and no other drugs.
The watchdog determined that while Border Patrol data on apprehensions and drug seizures were generally reliable, certain other checkpoint activity data, including on apprehensions of smuggled people and canine assists with drug seizures, were unreliable.
GAO’s analysis of Border Patrol checkpoint activity data showed that checkpoint agents did not consistently and appropriately document human smuggling events in the agency’s e3 data system (Border Patrol’s application that captures data related to checkpoint activity). According to Border Patrol documentation and leadership, the primary purpose of checkpoint operations is to interdict human smuggling events and identify recent entrants to the U.S. who are potentially removable. According to Border Patrol’s e3 guidance, agents are to document people involved in human smuggling events by, among other things, designating the principal smuggler and each smuggled person in e3. GAO found that Border Patrol agents documented about 600 people as “smuggled” between fiscal years 2016 and 2020 while simultaneously documenting them as lawfully present in the U.S. Of these, 485 people were documented as both U.S. citizens and smuggled people. GAO reports that 88 percent of these records were from two checkpoints in the Rio Grande Valley sector. Rio Grande Valley sector data integrity officials told the watchdog that the incorrect smuggled person designations were likely due to user data entry error or system error. GAO said about 85 percent of the incorrect designations occurred in 2019. Rio Grande Valley data integrity officials noted that in 2019 Border Patrol transitioned to a new version of the e3 application.
In addition, although sector officials said canines were integral to checkpoint drug seizures, there was wide variation across sectors in how often agents documented canine assists with such events. In February 2018, Border Patrol issued a memorandum reminding sectors of the requirement to record asset assists with apprehensions or seizures, including canine assists. The four sectors with the highest number of drug seizure events from fiscal years 2016 through 2020 were Yuma, Big Bend, Tucson, and Rio Grande Valley. However, Border Patrol data showed substantial variation in the frequency of canine assists with drug seizures in these sectors. For example, data showed canines assisted with a small proportion of drug seizure events in Yuma (2 percent) and Big Bend (4 percent) and a large proportion of such events in Tucson (77 percent) and Rio Grande Valley (88 percent). Data integrity officials from the Yuma and Big Bend sectors offered contrasting explanations regarding the relatively low number of canine assists their agents documented in e3. Yuma sector officials said that it did not seem reasonable that canines assisted with only 2 percent of drug seizure events at checkpoints in their sector. These officials acknowledged that they did not have a good oversight mechanism in place at the sector level to ensure that agents appropriately documented canine assists in e3. In contrast, Big Bend sector data integrity officials said that it seemed reasonable that canines assisted with 4 percent of drug seizure events at sector checkpoints. However, officials from a selected checkpoint in Big Bend sector told GAO that canines were critical to their mission and that the sector regularly assigned canine teams to work at the checkpoint.
The government watchdog also reported concerns over Border Patrol’s use of the Border Enforcement Secondary Tool (BEST), which was developed in 2019 to document information about secondary inspections at checkpoints, among other functions. GAO found that Border Patrol has not required checkpoints to use BEST, only about half of checkpoints with the capability are using BEST as intended, and Border Patrol has not reviewed or analyzed BEST data and does not have plans to do so. Officials in sectors with checkpoints that were not using the system or not using it consistently attributed this to various factors. These factors included that their use of BEST is not required; some checkpoints do not have the necessary technological capabilities, such as adequate bandwidth and computer terminals, to use the system; and documenting secondary inspections in the system is time-consuming. Border Patrol headquarters officials told GAO that they have not required checkpoints to use BEST and that the agency did not handle its implementation as well as it could have.
Border Patrol established the Checkpoint Program Management Office (CPMO) in 2013 to oversee checkpoint operations. However, Border Patrol has not demonstrated a sustained commitment to ensuring that CPMO carries out its checkpoint oversight activities or held CPMO accountable for implementing these activities. For example, CPMO was not reviewing checkpoint resources and technology—an activity assigned to it by Border Patrol. In January 2021, in response to GAO’s request for a current list of checkpoints and their features, CPMO provided the watchdog with a list of checkpoint locations and resources from 2019 that included duplicative checkpoint locations and had outdated information on checkpoint features, including inaccurate information on the types of technologies in use at checkpoints. Further, GAO found that Border Patrol has not established clear roles and responsibilities for CPMO, consistently and adequately staffed it, or ensured that newly assigned staff have the information they need to carry out CPMO activities. For example, documentation for new staff did not include details on tasks or the past activities of the office. In June 2021, the newly assigned Associate Chief responsible for CPMO acknowledged that the agency had not provided sustained and effective oversight of CPMO, nor held CPMO accountable for its activities. A senior Border Patrol official responsible for CPMO staffing also told GAO that Border Patrol had historically not provided the office with the continuity and appropriate levels of staff necessary to fulfill its checkpoint oversight mission. In December 2021, officials assigned to CPMO in August 2021 said that they were working to establish relationships with sector and headquarters officials involved in checkpoint operations.
While finding fault with data and oversight procedures, GAO was generally satisfied with Border Patrol’s training and guidance. Border Patrol agents receive classroom and on-the-job training in conducting traffic stops and in search and seizure law. For example, during Border Patrol’s basic training academy, new agents are to receive classroom training and complete practical exercises in performing checkpoint operations, searching vehicles, and searching suspects. This training includes such topics as court cases relevant to checkpoints and their importance in checkpoint operations; identifying indicators of concealed smuggling loads; and inspecting a noncompliant motorist. Sectors manage on-the-job training in search and seizure activities, and may provide or require other training, including training related to checkpoint legal authorities.
To help Border Patrol meet its checkpoint mission, GAO has made seven recommendations:
- Ensure that sectors and the Checkpoint Program Management Office (CPMO) are overseeing that checkpoint activity data are consistently and accurately recorded. This should include (but not be limited to) overseeing data on apprehensions of smuggled people, canine assists with drug seizures, seizures of trace amounts of marijuana, non-drug property seizures, and attempted checkpoint circumventions.
- Ensure that Border Patrol uses its data system to document information about all people for whom a checkpoint encounter leads to a subsequent enforcement action.
- Require checkpoints with license plate readers to document secondary inspections by using BEST as intended.
- Provide sustained oversight of CPMO and hold it accountable for implementing its checkpoint oversight activities.
- Ensure that CPMO is consistently and adequately staffed.
- Update CPMO’s roles and responsibilities and communicate them across Border Patrol, including to sectors.
- Develop policies and procedures for CPMO’s checkpoint oversight activities.
The Department of Homeland Security concurred with all of the recommendations. It stated that CPMO plans to coordinate with Border Patrol and CBP offices and divisions, as appropriate, to review checkpoint activity data and guidance related to checkpoint processing and to regularly review data related to checkpoint operations. In addition, CPMO plans to hold regular meetings with Border Patrol headquarters and sector officials to define organizational structures, headquarters and sector responsibilities, and standard operating procedures.