Background Investigations Conducted by the United States Marshals Service

E&I Report No. I-2005-002
February 2005
Office of the Inspector General

SCOPE We reviewed the process for all background investigations and reinvestigations of employees, CSOs, and other contractors whose field investigations are adjudicated by USMS headquarters personnel in either the Human Resources Division or the Judicial Security Division. We did not review low-risk contractors whose field investigations and adjudications are conducted at the district office level. METHODOLOGY Specifically, we reviewed policies and procedures, interviewed officials involved in the process, and analyzed selected files to determine whether the USMS provides timely and thorough background investigations and reinvestigations that comply with federal regulations and Department policy for its employees and contractors. Interviews We conducted 21 interviews, listed below, at USMS headquarters in Arlington, Virginia, and at SEPS’s offices in Washington, D.C. USMS Headquarters Human Resources Division Chief of Human Resources Services Chief of the Operational Security Branch, Human Resources Services Chief of the Personnel Security Branch, Human Resources Services Acting Assistant Director, Management Services Branch Chief of Adjudications, Human Resources Services Two Personnel Security Specialists (adjudicators), Human Resources Services Operations Support Division Chief Inspector for Internal Affairs Judicial Security Division Chief of the Judicial Protective Service Chief Inspector of the Judicial Security Division Senior Inspector Information Technology Administrator SEPS Personnel Security Group Assistant Director Chief of Operations Security Personnel Security Specialist, Information & Technical Security Group Personnel Security Specialists, Policy Section Compliance Review Group Senior Security Specialist Policy Analyst Office of the Inspector General Two OIG Supervisory Investigative Agents, OIG Investigations Division Office of Human Resources Specialist Other Government Officials OPM Training Officer OPM Customer Service Group Chief National Finance Center Data Analyst Data and Sample File Review The USMS Human Resources Division, the Judicial Security Division, SEPS, and the OIG Investigations Division provided us with data from five databases, described below. Because of the poor quality of data in the two USMS Human Resources Division databases for employees and contractors, projections cannot be made from the statistics generated from these databases. We used these databases for four purposes: 1) to evaluate the quality and completeness of data; 2) to generate random and judgmental sample files for review; 3) to determine the adjudications workload in recent years; and 4) to determine the databases’ usefulness for case tracking and program management. USMS Employees The Human Resources Division provided us with a copy of its Microsoft Access employee database, as of March 2004, that the Human Resources Division uses to track background investigations and reinvestigations of current employees. The database contains 41 fields to record biographical data, position and clearance data, key event dates for processing investigations, the type of investigation requested, and its cost. We started with the full database of 7,309 entries and refined it to include only entries with an entered-on-duty date, or 4,424 entries.25 We chose the entered-on-duty date because we were told that data field would identify active employees. Therefore, the database statistics on USMS employees provided in this review were derived from the universe of the 4,424 database entries with entered-on-duty dates on the date the database was obtained (March 2004). In March 2004, SEPS provided us with specified fields for the 127 USMS employees it manages in its database, TRAQ. We used this database to generate a random sample of 26 files to review the timeliness and thoroughness of investigations and adjudications for USMS political appointees, attorneys, and other designated employees. To conduct our sample file review, we generated a random sample of 66 cases from the universe of 4,424 employees.26 We prepared a checklist for our review of each case file and used the information on the checklist to assess the thoroughness of OPM investigations and USMS adjudications and to determine compliance with USMS policy. USMS Contractors (Excluding CSOs) The Human Resources Division provided us with a copy of its Microsoft Access contractor database, as of March 2004, that the Human Resources Division used to track background investigations and reinvestigations of current contractors. The database fields are identical to those in the employee database. We started with the full database of 8,844 entries and refined it to include only entries with an entered-on-duty date, or 5,335 entries.27 We used these 5,335 entries to select a random sample of 33 contract guards and a random sample of 25 other contractors, for a total of 58 review files.28 Of the 58 contractor files we requested for our sample review, the USMS was able to locate 25 of the files at headquarters. We used those files to evaluate compliance with federal and Department regulations and the thoroughness of investigations and adjudications. The remaining 33 contractors had been hired at the local level without adjudication at headquarters, so their files were stored in various districts or could not be located. We did not review the files stored at the districts. Court Security Officers The Judicial Security Division provided a list (and later a copy of the Judicial Security Division Oracle database) of the 4,323 contract CSOs as of May 2004. We used these cases to generate a random sample of 33 background investigation files to evaluate the thoroughness of CSO investigations and adjudications.29 Misconduct Cases The OIG Investigations Division provided a copy of the Investigations Data Management System (IDMS) database, which contained all USMS misconduct allegations made from fiscal year (FY) 2000 through April 2004, against USMS employees and contractors. We used the database to create a judgmental sample of 42 cases for the file review. These cases fell into three categories: Cases specifically mentioned during interviews with USMS staff; Cases that resulted in a criminal sentence, termination, or retirement, regardless of the nature of the misconduct; and All cases in which there was an allegation of violence, a threat of violence, or harassment involving domestic relationships, colleagues, the public, or prisoners.30 We requested from the USMS Office of Internal Affairs the case dispositions of the 42 cases that had met at least one of our criteria, and we determined that 29 (28 employees and 1 contractor) had misconduct allegations that were sustained. There were no political appointees, attorneys, or CSOs who met our criteria, so the sample did not include anyone from these categories. Statistics are based on these 29 cases or a subset of these cases. These 29 sample misconduct files were used for three purposes: 1) to review the thoroughness of the background investigations and whether reinvestigations were initiated within five years; 2) to evaluate whether there was any relationship between the thoroughness of the background investigations and reinvestigations and subsequent sustained misconduct allegations; and 3) to determine whether reinvestigations of these individuals at the five-year deadline might have identified a problem before it led to a misconduct investigation. Other Sources We reviewed federal regulations, Central Intelligence Agency directives, OPM and SEPS guidance, OPM and USMS policy and procedures manuals, training materials, checklists, and the most recent USMS Office of Internal Affairs report (FY 2001). We also obtained relevant portions of a standard CSO contract from the Judicial Security Division. Footnotes The remaining 2,885 cases did not have entered-on-duty dates because they were in various stages in the investigation and hiring process. Over 75 percent of USMS employees occupy law enforcement positions. To ensure that we reviewed a sufficient number of employees who were not in law enforcement positions, including a range of positions and clearance levels, we separated the two caseloads and generated separate random samples of 33 cases each. The remaining 3,509 cases did not have entered-on-duty dates and were not included in the random sample because there was insufficient information to determine their current status. There were 5,233 contract guards with entered-on-duty dates and only 102 other contractors. To ensure that we reviewed a sufficient number of contractors who were not contract guards, including some who had received national security clearances, we separated the two caseloads and generated separate random samples of 33 contract guards and 25 other contractors. Because all CSOs nationwide have the same responsibilities and are investigated and adjudicated to the same standard, we did not need to create a separate random sample within this caseload. Thirty-three files were necessary for a review of 4,323 cases at the 90 percent confidence rate. We chose these files based on the misconduct codes used by the OIG Investigations Division.