Audit Objectives

Our audit objectives include determining whether the USMS: (1) adequately designed, tested, and implemented a workforce management plan that sufficiently assesses its human resources and capacity requirements based on current and expected work loads by function; (2) evaluates, monitors, and corrects, if necessary, its personnel utilization to ensure it directs appropriate resources to its highest priorities and achieves its organizational objectives; (3) has sufficiently addressed pay compensation, including job-grade and career progression; and (4) has provided adequate and appropriate training to its operational employees.

Scope and Methodology

We performed our audit in accordance with the Government Auditing Standards and included such tests of the records and procedures that we considered necessary to accomplish the audit objectives. We performed testing of the USMS’s compliance with certain internal controls in the accomplishment of our audit objectives and include discussion on such in Chapters 2 through 5 of our report. The objectives of our audit did not require that we perform testing of the USMS’s compliance with laws and regulations.

To accomplish our objectives, we interviewed officials at USMS headquarters and district offices, and reviewed empirical resource utilization and workload data.

Interviews

Much of our work centered on interviews with USMS officials at various levels within the organization, which were conducted at the headquarters and district office levels. Additionally, we interviewed DOJ Justice Management Division (JMD) officials involved with budget and human resource matters. These interviews, as well as documentation obtained during the interviews, provided perspective on the resource-related issues covered by our audit objectives. In total, we interviewed 186 USMS and DOJ officials.

Of these interviews, 50 were conducted with USMS personnel at its headquarters in Arlington, Virginia, and the Federal Law Enforcement Training Center in Glynco, Georgia. Specifically, we spoke with the USMS Director and Deputy Director, USMS Training Academy officials, as well as Assistant Directors and other staff in several headquarters divisions, including the Investigative Services Division, Judicial Security Division, and Witness Security and Prisoner Operations Division. We also met with the Acting Assistant Attorney General for Administration, as well as JMD budget and human resources representatives.

Further, we conducted fieldwork at seven USMS district offices to obtain the district-level perspective on various issues pertaining to USMS human resources. During fieldwork, we interviewed 119 district representatives, including U.S. Marshals, Chief Deputy U.S. Marshals, deputies, and administrative staff. Following is a list of the USMS district offices visited:

• Central District of California
• Western District of Texas
• Northern District of Illinois
• District of Rhode Island
• Southern District of New York
• D.C. Superior Court
• Southern District of Florida

During our fieldwork at the seven district offices, we met with 13 USMS representatives on the Regional Fugitive Task Forces, which were located in the same jurisdictional area. Specifically, we interviewed USMS officials from the Capital Area Regional Fugitive Task Force, Great Lakes Regional Fugitive Task Force, New York / New Jersey Regional Fugitive Task Force, and Pacific Southwest Regional Fugitive Task Force.

Utilization and Workload Data

To assist in accomplishing the audit’s objectives, we analyzed data provided by the USMS. Specifically, we conducted analyses of USMS data on its resource allocation, resource utilization, and casework. We requested data from the Administrative Office of the U.S. Courts (AOUSC) to provide more insight on the USMS’s court-related workload. However, the AOUSC was unwilling to provide us with this information.

To examine the USMS’s human resource utilization, we examined data from the USMS’s USM-7 system – a module of its payroll system – for the period October 1, 1999, through September 30, 2005. The USM-7 contains work-hour data for all USMS employees, as well as for personal services contract guards. To examine the types and quantity of work addressed by the USMS during this same time period, we analyzed data from the USMS’s Warrant Information Network (WIN) and Prisoner Tracking System (PTS).

Data Analysis

We performed analyses of USMS resource allocation, resource utilization, and casework data to identify trends and note significant changes in the USMS’s operations from FY 2000 to FY 2005. In total, this data amounted to 12,884,978 records. As noted in the body of the report, we identified several concerns with the accuracy and reliability of the data contained within the USMS’s automated systems. While we believe that these issues affect the analyses we performed, we believe that the overall results presented have utility for looking at the USMS as a whole to get a basic understanding of the agency’s resource utilization and workload. We did not perform an independent, overall assessment of the reliability of the data in the USMS’s automated systems.

USMS Human Resources

We conducted analyses of USMS positions (authorized or allocated levels) and full-time equivalent (resource utilization) data.

Allocated Resource Levels

We used the USMS’s position data established by the Management and Budget Division to analyze the allocation of USMS resources. We obtained operational and administrative allocations for each fiscal year, for both USMS district offices and headquarters divisions for FYs 2000 through 2005. The USMS does not allocate district resources to specific mission activities, such as fugitive apprehension or judicial security. We reviewed these resource allocation levels, focusing on changes occurring between FYs 2000 and 2005. The total position data amounted to 2,034 records.

Resource Utilization Level

USMS employees and personal services contract guards (also known as District Security Officers) record their time on the USM-7 every 2 weeks. When completing the USM 7, USMS personnel record the number of hours worked (tracked to 0.25 hours) to project codes that are attributed to the type of function or duty being performed. For example, for any 8-hour day, a Deputy might record that he worked 4 hours apprehending fugitives, 2 hours producing prisoners for court, and 2 hours in annual leave status. The USM-7 data is only as valid as the information reported by USMS personnel. However, the USMS considers the USM-7 system’s data to be the best way to assess the actual amount worked by USMS employees and independent contractors in specific mission areas.

The USM-7 data run was provided in a text file, which we imported into a database file. The data run contained 7,981,715 records, each containing information within 19 different fields. Following is a listing of the fields used as part of our analyses:

We elected to analyze resource utilization data by fiscal year. To do this, we totaled the hours for all pay periods within each fiscal year for each program and project code. Next, we divided this total by 2,087 hours to obtain the average number of personnel working on a particular program or project code in a given fiscal year. This computation provided us with the reported number of full-time equivalents (FTE) involved in the USMS’s various mission activities. For example, if three deputies within a particular district office each spent one-third of their time (approximately 696 hours) on fugitive apprehension within a given fiscal year, the resource utilization for that district office (for fugitive apprehension during that fiscal year) would be equal to one operational FTE.

USMS Casework

The USMS uses many systems for tracking or measuring its casework. In particular, we focused on analyzing data contained in the USMS’s Prisoner Tracking System (PTS) and Warrant Information Network (WIN).

Fugitive Warrants

For our analyses of the USMS’s fugitive warrant casework, we received a data file from the WIN system, and focused on warrants received from FYs 2000 through 2005. We also reviewed warrants closed during this same time period. The data run was provided in a text file and imported into a database file containing 843,944 records, separated into the following fields used for our analyses:

We confined our analysis to the data we obtained from the WIN system, and did not review individual case files to examine the actual level of effort expended on a single warrant. Thus, if a case was open during a particular timeframe, we considered it to be worked during that period.

Case Openings – The number of warrants received in a given period of time demonstrates the fugitive warrant workload that the USMS was handling. In order to conduct such an evaluation, we first organized the cases according to the fiscal years in which they were received by all USMS district offices. Then, we analyzed the difference in warrants received between FYs 2000 and 2005 by warrant type and warrant category. This analysis afforded perspective on the changes in the USMS’s level of investigative effort among various types and categories of warrants. However, it should be noted that the USMS has very little control over the type or category of warrants it is given. Instead, the USMS’s fugitive workload heavily depends on the warrants issued by the federal courts.

Case Closings – Similar to our case openings analysis, the number of warrants cleared in a given period of time demonstrates the types of cases handled or completed by the USMS. For this analysis, we first organized the cases according to the fiscal years in which they were closed by each district office. Then, we analyzed the difference in warrants cleared between FYs 2000 and 2005 by warrant type, warrant category, and reason for closure.

Prisoner Transportation

For our analyses of the USMS’s prisoner-related casework, we received a data file from the USMS’s PTS, and focused on prisoner movements conducted during FYs 2000 through 2005. Additionally, we reviewed the level of prisoners moved by the USMS during this same period of time. We confined our casework analysis to the data we obtained from the PTS, and did not review individual case files to determine the actual level of effort expended on a single prisoner or prisoner movement. The data run was provided in a text file and imported into a database file containing 4,047,450 records, separated into the following fields:

Prisoner Movements – The number of prisoner movements conducted in a given period of time demonstrates one facet of the USMS’s prisoner-related workload. In order to conduct such an evaluation, we first organized the prisoner movements according to the fiscal years in which they were conducted for each USMS district office. Then, we analyzed the difference in prisoner movements conducted between FYs 2000 and 2005 by the USMS’s production codes, i.e., the reason for transporting a prisoner. This analysis afforded perspective on the changes in the USMS’s level of prisoner-related casework. However, it should be noted that there are other variables that influence the number of prisoner movements, including the federal judicial system’s court proceedings and the investigative efforts of other law enforcement agencies.

Prisoners Handled – Besides determining changes in the number of prisoner movements, we also reviewed the changes in the number of prisoners handled by the USMS. Specifically, we focused on prisoners whom the USMS would handle on a given day involving the transportation of a prisoner from one location to another. For example, it could involve transporting a prisoner from a correctional facility to the courthouse for an initial appearance, or transporting a prisoner from a correctional facility to a hospital for medical care.

We focused on identifying the number of unique prisoners handled during any particular day. In other words, the same prisoner could be transported multiple times during the same day, such as appearing for a bond hearing at a courthouse and then going to a hospital for medical treatment. In these instances, we only wanted to count that the USMS handled that prisoner once during that particular day. To perform this analysis, we had to create a unique field. Specifically, we combined the USMS’s prisoner identification field with the prisoner production (or transportation) date. An example of the resulting value would be 101 10/11/2002, which means that the prisoner with an identification number of 101 was transported on October 11, 2002. From this resulting value, we were able to identify those instances in which the same prisoners were transported multiple times during the same day. We excluded any such records from our analyses.

Moreover, there were several instances in which the USMS used a prisoner identification value of “0.” According to USMS officials, this value indicates that there was not a USMS number for these prisoners. Further, this value may refer to the same prisoner, or it could be associated with different prisoners. Since there was no readily available way to determine if these records applied to the same prisoner, we excluded all records with such a prisoner identification value.

In order to complete our evaluation, we organized the values within our created field according to the fiscal years in which the prisoners were handled by the USMS within each district office. Then, we analyzed the difference in the number of prisoners handled between FYs 2000 and 2005. This analysis afforded another perspective on the changes in the USMS’s level of prisoner-related casework. Again, it should be noted that there are other variables that influence the number of prisoners handled or transported by the USMS.

Potential Threats

For our analyses of the USMS’s threat-related casework, we received a data file from the USMS’s WIN system, and focused on potential threats tracked during FYs 2000 through 2005. We performed our casework analysis on the data we obtained from the WIN system, but did not review individual case files to determine the actual level of effort expended on a single potential threat. The data run was provided in a text file and imported into a database file containing 9,835 records, separated into the following fields that were used for our analyses:

In conducting our evaluation of the USMS’s casework with potential threats, we first organized the records according to the fiscal years in which the USMS was notified of the potential threat. Then, we analyzed the difference in potential threats received between FYs 2000 and 2005 by method of delivery, residency information, and the perceived threat level.

FY 2006 Update of Threat-Related Data

Subsequent to our analyses of FYs 2000 through 2005 threat-related data, we obtained and performed similar analyses on resource utilization and casework data from FY 2006. For our resource utilization analysis, we received a USM-7 data file, which contained 32,229 records associated with protective investigations. For our analysis of the USMS’s threat-related casework, we received a data file from the USMS’s WIN system containing 11,230 records and focused on those records reported to USMS district offices during FY 2006.