Return to the USDOJ/OIG Home Page
Return to the Table of Contents

Follow-up Inspection of the United States Marshals Service's
Fugitive Apprehension Program

Report Number I-2000-02
January 2000


The Office of the Inspector General (OIG), Inspections Division, initiated a follow-up inspection to review the United States Marshals Service (USMS) efforts to improve the fugitive apprehension program — the program responsible for executing federal fugitive warrants — in light of recommendations made in a 1995 Inspections Division report (I-94-04). That report recommended establishing national goals and priorities for the fugitive apprehension program, developing performance measures to assess these goals and priorities, increasing investigative support to field personnel working on fugitive investigations, and enhancing the Warrant Information Network (WIN) system. The WIN system is a USMS database containing fugitive information.

In response to our recommendations, the USMS established national goals to reduce the backlog (defined as warrants older than one year) by 15 percent and to close 80 percent of new warrants within one year. In addition, the USMS established the Quality Point Index (QPI) system as a management tool intended to assess the effectiveness of the fugitive apprehension program by measuring whether the USMS is closing warrants consistent with its national goals and priorities. The USMS also created the Analytical Support Unit (ASU) to provide additional investigative support to its field personnel. The USMS agreed to give each district access to the WIN system and agreed to provide more training to field personnel using the system.

This follow-up report analyzes the actions taken by the USMS since the 1995 report. We focused most of our efforts on assessing the establishment of national goals and priorities and the usefulness and accuracy of the QPI system.


Responsibilities of the USMS

The USMS performs a variety of law enforcement and civil service functions. The primary responsibility of the USMS is to protect the Federal courts. The USMS is also responsible for executing court orders, providing secure prisoner confinement and transportation for judicial proceedings, collecting seized assets, and ensuring the long-term safety of protected government witnesses.

In July 1979, the USMS assumed responsibility for executing federal fugitive warrants. The USMS investigates warrants issued for probation violators, parole violators, bond defaulters, and escapees, as well as subjects of Drug Enforcement Administration (DEA) warrants, fugitives wanted by agencies without arrest authority, and bench warrants issued by federal judges.1 Collectively, these activities make up the fugitive program.

Organizational Structure of the Fugitive Program

The fugitive program is administered by the Investigative Services Division (ISD) at USMS headquarters. The ISD oversees the enforcement of court orders, fugitive investigations, and the execution of federal warrants. In addition, ISD maintains the WIN system, provides electronic surveillance, and manages the international extradition program. Within the ISD, the domestic fugitive program is primarily administered by three units:

Domestic Investigations Unit (DIU): The DIU provides investigative advice and analysis to USMS district offices (herein referred to as districts); coordinates the USMS's high priority cases; oversees districts' task force participation; reviews districts' compliance with headquarters performance plans; creates policies and procedures for fugitive investigations; and authorizes informant payments.

Analytical Support Unit (ASU): The ASU provides tactical and strategic information to the districts for older cases whose leads have been exhausted; maintains the WIN system; and analyzes threats made against the judiciary.

Electronic Surveillance Unit (ESU): The ESU provides electronic surveillance such as cellular telephone and pager intercepts; advises districts about appropriate surveillance techniques; assists districts in preparing court orders requesting electronic surveillance; and analyzes information obtained through electronic surveillance.

While the USMS headquarters administers the fugitive program, the districts conduct the actual investigations and execute warrants. There are 94 USMS districts, each headed by a Presidentially appointed U.S. Marshal. Because resource allocation and internal structure of the districts are determined by each U.S. Marshal, the districts vary somewhat in structure.

This variation is also present in the districts' fugitive apprehension efforts. Some districts have warrant squads staffed with USMS Deputy Marshals who are dedicated full time to conducting fugitive investigations, while other districts assign fugitive investigations to deputies as a collateral duty. These deputies conduct fugitive investigations when they are not transporting prisoners or protecting the courts. Lastly, some districts supplement warrant squads staffed by deputies who only conduct fugitive investigations with additional deputies who conduct investigations as a collateral duty.

Fugitive Investigations

A USMS district initiates a fugitive investigation upon receipt of a warrant. A supervisor assigns the warrant to a deputy, who collects as much information on the fugitive as possible by using a variety of investigative techniques. Not only do the deputies have access to databases maintained by the USMS, but they also have access to systems maintained by other federal and local law enforcement agencies. Deputies interview parole and probation officers if a fugitive has already been through the penal system, speak with family members and associates, and research employment records in order to generate leads to determine a fugitive's location. Deputies also may request electronic surveillance support provided by the ESU for high priority investigations.

If the fugitive does not already have a WIN record, the deputy enters all background and criminal history information into the WIN system. The system contains records of fugitives, including biographical information, the fugitive's criminal history, and all warrants issued in their name. If the fugitive has an existing WIN record, the deputy updates the record with the new information contained in the warrant. Once the new warrant is entered into the WIN system, the deputy then creates an entry in the National Crime Information Center (NCIC) database, alerting all law enforcement officers to contact the USMS if they encounter the fugitive.

The USMS closes a fugitive investigation when the warrant is executed. Warrants are executed when a fugitive is physically arrested, or surrenders, or when the warrant is dismissed (e.g., if the fugitive is deceased). After the deputy executes the warrant, he closes the warrant in the WIN system by entering the date the warrant was executed.

The USMS can place warrants in a suspended category if the immediate apprehension of the fugitive is not possible even though the USMS has located him (e.g., if a fugitive has fled to a country that does not have an extradition treaty with the United States). Suspended warrants are reviewed less frequently than active warrants to determine whether the status of the fugitive has changed. If the fugitive is already incarcerated (e.g., for a state criminal violation), the USMS and the law enforcement agency holding the fugitive place a detainer in the fugitive's file. Typically, the law enforcement agency informs the USMS of the fugitive's scheduled release. At the time of the release, the USMS will take the fugitive into custody and execute the warrant.

The Quality Point Index System

The QPI system is a management tool intended to assess the effectiveness of the fugitive apprehension program by measuring whether the USMS is closing warrants consistent with its national goals and priorities. In addition, it can serve as a tool for headquarters and districts to manage district priorities, monitor district performance, and prioritize warrant investigations by concentrating efforts primarily on wanted violent felons.2

Prior to the development of the QPI system, the USMS measured the effectiveness of the program solely on the basis of one factor — the length of time taken to close a warrant. In 1995, the USMS determined this singular measure was inadequate to measure the effectiveness of the program. The QPI system was designed to take into account additional factors involved in closing warrants that are consistent with USMS national goals and priorities.

Currently, the USMS national goals and priorities reflected in the QPI system are apprehending violent fugitives to prevent future acts of crime, reducing the warrant backlog (defined as warrants over one year old), and apprehending fugitives who are wanted on federal charges as well as state or local charges.3

Assignment of Points: The QPI system assigns points to indicate the relative investigative priority of a warrant. Once the USMS receives a warrant and enters the warrant into the WIN system, the system assigns the number of points that the district will receive when the warrant is closed by the USMS. The WIN system calculates the points based on the offender category, the age of the case, and the existence of state and local warrants. Warrants receive a specific number of points as shown in Table 1.

Table 1: Distribution of QPI Points4

Top 15 Most Wanted 11 3 2
Major 10 3 2
1 6 3 2
2 3 3 1
3 2 3 1
Source: USMS

The USMS created five offender categories to classify different types of warrants based on the criminal history of the fugitive. Almost all fugitives labeled as Top 15 Most Wanted, Major, or Category 1 are considered violent.5 A description of each offender category follows.

Top 15 Most Wanted: These fugitives are selected by headquarters as the USMS's highest priority fugitives to investigate and apprehend. Although these cases are chosen based on subjective criteria, almost all of the fugitives have extensive criminal histories involving violent crimes. These cases receive national publicity and a list of the 15 fugitives is distributed throughout the country.

Major: These fugitives are selected by headquarters as high priority fugitives to investigate and apprehend. Although these cases are chosen based on subjective criteria, almost all of the fugitives have extensive criminal histories involving violent crimes, but are of lower priority than the Top 15 Most Wanted cases.

Category 1: These fugitives have committed at least one homicide, rape, robbery, assault, or kidnapping; are wanted by the DEA for serious drug charges or association with a gang; or have been arrested on charges of weapons or firearms possession.

Category 2: These fugitives are designated as drug offenders (who do not fall into category 1) or habitual non-violent offenders who have been arrested at least six times for immigration violations, forgery, burglary, narcotics, extortion, and other miscellaneous crimes.

Category 3: These fugitives have non-violent felony charges over which the USMS has primary investigative responsibility and who do not fall into the previous categories.

The WIN system first assigns the appropriate number of points to the warrant based on its offender category. For example, an individual who is identified as a category 1 fugitive receives six points, while an individual classified as a category 2 fugitive receives three points.

Second, the WIN system determines whether the warrant is older than one year by calculating the number of days between the date the warrant is issued and the date the warrant is closed. Finally, the system assigns extra points if the fugitive has a state or local warrant in addition to a federal warrant.

The WIN system tracks and displays the potential point value for each warrant when the warrant is entered into the system, thus making the relative investigative priority known to the district. The higher the warrant's point value, the more important it is for districts to give priority to the investigation. For example, a warrant with 16 points should be investigated more aggressively than a warrant with 6 points. In addition, warrants with an equal number of points should receive a similar investigative priority.

Once a warrant is closed, points are awarded to the district maintaining the warrant. Districts receive no points until a warrant is closed. The WIN system continuously updates the record when new information is entered into the system. If, in the course of the investigation, the case becomes older than one year or the presence of a state and local warrant associated with the fugitive is entered into the WIN system, the warrant's points increase to reflect these changes.

The QPI System as a Management Tool: The QPI score, part of the QPI system, was designed as a tool to monitor district performance. This score provides management with a tool to assess district offices' effectiveness and may indicate a need for either headquarters or district management to make improvements to achieve national goals and priorities.

The district's QPI score is calculated based on the total points assigned to fugitives apprehended by the district, divided by the number of fugitives apprehended. For example, if District A apprehends a Top 15 Most Wanted fugitive whose warrant is three years old, the warrant's points would equal 14 points — 11 points for the warrant closure and 3 points because the warrant was older than one year. If District A then apprehends a Category 3 fugitive whose warrant is less than one year old and has an additional state warrant, the district would receive 3 points — 2 for the warrant closure, 0 for the age of the warrant and 1 for the state warrant. The district's QPI score would equal 8.5 — the total points for both fugitives (17) divided by the number of fugitives apprehended (2).

Headquarters sets a yearly QPI target by calculating the average of all the districts' yearly QPI scores. For fiscal years (FY) 1998 and 1999, the QPI target set for all districts was 4.8. On a monthly basis, headquarters generates a report comparing the performance of all districts with the target. In addition, headquarters produces a yearly report of each district's cumulative QPI score.

Scope and Methodology

We interviewed 50 USMS managers and staff from 12 district offices either during on-site visits or by telephone. We visited five USMS districts and one sub-office representing small, medium, and large districts. During our site visits, we conducted interviews with chief deputies, warrant squad supervisors, deputies, and investigative research analysts. We also conducted telephone interviews with warrant squad supervisors from seven additional districts. We interviewed and obtained relevant documents from headquarters officials from the DIU, the ASU, the ESU, and the Information Technology Division (ITD).

We received data from the WIN system on all 82,682 warrants closed by the USMS from FY 1994 through FY 1998. With the information that headquarters provided, we created a database from which we conducted the majority of our data analysis. From this information, we were able to generate statistics regarding the number of warrants closed, the types of warrants executed, and the number of violent fugitives apprehended.

Generally, we did not examine open warrants; however, we did gather statistics on open warrants to calculate the percentage of warrants closed within one year and to calculate warrant backlog reduction. We did not analyze warrants categorized as Class II — warrants issued for misdemeanors and traffic violations occurring on federal property.6 Refer to Appendix 1 for a full discussion of our scope and methodology.



In our 1995 report, we found disparity among districts in their commitment to investigate warrants and allocate resources to the fugitive apprehension program. Consequently, we recommended that the USMS establish national goals and priorities to improve the effectiveness of the fugitive program. In response, the USMS established national goals to (1) reduce its warrant backlog and (2) close warrants more quickly.

Our recent inspection found that the USMS met its backlog reduction goal for FY 1999. The USMS has exceeded it goal to close warrants under one year old, but achievement of this goal does not represent an improvement in performance.

The USMS Met Its Backlog Reduction Goal For FY 1999

The USMS established backlog reduction as one of its goals in FY 1996; however, only limited progress had been made to reduce the backlog until FY 1998. At the direction of the Attorney General, the USMS established a warrant backlog reduction goal of 20 percent in FY 1999. The USMS has significantly reduced its FY 1998 backlog and met its warrant backlog reduction goal of 20 percent in FY 1999.7

Before the warrant backlog reduction goal was set at 20 percent, the USMS assessed the composition of the warrant backlog. This assessment resulted in the USMS redefining what types of warrants are included in the warrant backlog. Previously, the USMS defined its warrant backlog as all warrants where the USMS had investigative responsibility that were over one year old. This broad definition included many warrants where there was no realistic opportunity to execute the warrant in the foreseeable future. To better define the warrant backlog, the USMS excluded from definition all warrants for (1) incarcerated fugitives where the USMS has lodged a detainer to take custody of the prisoner once he is released and (2) fugitives located in foreign countries with no United States extradition treaty. While the USMS monitors the status of these warrants, substantial investigative effort will not necessarily result in closure of these warrants. When the USMS removed these warrants from the backlog, it established the outstanding warrant backlog at 10,600 as of October 1, 1998. The USMS reduced this backlog to 8,480 warrants to meet its 20 percent goal by September 30, 1999.

To accomplish this goal, the USMS created a backlog reduction plan. The backlog reduction plan outlined procedures for updating the WIN system, reviewing fugitive investigation staffing, and establishing Cold Case Consultant Teams (CCCT) to visit the 12 districts with the highest number of backlog warrants.8 These teams reviewed district policies and identified district problems to help develop the backlog reduction plan.9 In addition, the plan called for improving cooperation with DEA by emphasizing the importance of promptly handing over warrants to the USMS within the required seven days as outlined in the DEA-USMS Memorandum of Understanding (MOU). The plan also incorporated fugitive apprehension activities in district performance reviews.

The backlog reduction plan primarily focuses on closing warrants through administrative and policy means. For example, the CCCT recommended working with the U.S. Attorneys' Offices to dismiss older cases, and working with the DEA to permit the USMS to return warrants missing vital suspect information back to the DEA.10 These actions will increase the number of warrants closed; however, these closures are primarily due to administrative actions and will not necessarily increase the number of fugitives arrested.

Although there is not yet enough information available on the overall impact of the backlog reduction plan to fully assess its long term success, the USMS has met its targeted goal in FY 1999. To analyze the backlog from FY 1994 through FY 1999, we applied the current backlog definition to the entire time period. In FY 1999, the USMS made substantial progress to reduce the backlog to 8,186 warrants, exceeding its goal by 3 percent (see Chart 1).

Chart 1: USMS Warrants Over One Year Old (Backlog)

Source: USMS WIN System

The USMS Has Exceeded Its Goal To Close Warrants Under One Year Old, But Achievement Of This Goal Does Not Represent An Improvement In Performance

The USMS consistently exceeded its goal of closing 80 percent of new warrants within one year (see Chart 2).

Chart 2: Percentage of Warrants Closed By USMS Within One Year11

Source: USMS WIN System

In FY 1994 and FY 1995, before the USMS established the 80 percent goal, the USMS was already closing 85 to 86 percent of warrants within one year. The USMS established the 80 percent goal in FY 1996; however, the average rate of warrant closure remained the same 86 percent from FY 1996 through FY 1999. We found similar patterns when we examined warrant closure rates at the district level. In FY 1994 and FY 1995, an average of 86 percent of the districts closed over 80 percent of new warrants. After the goal was established, districts meeting the goal remained at an average of 86 percent from FY 1996 through FY 1999.

The establishment of the 80 percent goal has produced no substantive changes in the percentage of warrants closed at either the district or the national level. In fact, the USMS's performance could drop 6 percentage points and the USMS could still claim achievement by meeting its goal.

An additional assessment of the USMS's achievement regarding its management of warrants would include consideration of the work years and funding dollars that are expended to close warrants. However, the USMS does not relate these variables to its closure rate. We understand from discussions with USMS personnel that some doubt exists with the reliability of the work year statistics devoted to the fugitive program. Specifically, the work years attributed to the fugitive program are not all directly related to federal fugitive apprehensions (e.g., time spent investigating state and local warrants, time spent on processing threats against judges) and variations exist in how districts collect work year data.


In our 1995 report, we recommended that the USMS establish national goals and priorities to improve the effectiveness of the fugitive program. Subsequent to that report, the USMS made the apprehension of violent fugitives an explicit national priority.

In our current inspection, we found that the USMS has not established a quantifiable goal for apprehending violent fugitives.

The Attorney General has stressed the importance of reducing crime and removing violent offenders from the street. Accordingly, the DOJ strategic plan established a number of goals for other DOJ components to target violent criminals. To parallel these efforts, the USMS is also emphasizing the apprehension of violent fugitives. However, unlike its goal for reducing backlogged cases and closing warrants less than one year old, the USMS has not set a quantifiable goal for targeting violent fugitives. Therefore, the USMS is not systematically tracking and assessing its efforts to apprehend violent fugitives.

We analyzed WIN data from FY 1994 through FY 1998 to determine how many violent fugitives the USMS has apprehended during this period. We examined the criminal history of all fugitives classified as Top 15 Most Wanted, Major, and category 1 fugitives — all defined as violent offenders.12 These categories include fugitives arrested for homicide, rape, kidnapping, weapons, assault, robbery, a serious drug violation or association with a gang. We found that the percentage of violent fugitives apprehended out of all warrants closed increased by 27 percent from FY 1994 through FY 1998.13 In FY 1994, 41 percent of the warrants closed involved violent fugitives; in FY 1998, 52 percent of the warrants involved violent fugitives (see Chart 3).

Chart 3: Percentage of Warrants Closed for Violent Fugitives14

>Source: USMS WIN System

The USMS currently has the data to track the percentage of violent fugitives apprehended using its violent offender categories but has not done so. Given the importance of apprehending violent fugitives to the USMS, establishment of a goal tracking the number of violent fugitives apprehended would provide a method to gauge USMS performance, target investigative resources, and serve as an incentive to increase the number of violent warrants closed. In addition, establishing and tracking of a quantifiable goal would allow trend analysis to assess program improvements over time.


Based on the recommendation to establish national goals and priorities in our 1995 report, we also recommended that the USMS develop and implement performance measures to assess the effectiveness of the fugitive apprehension program. In response, the USMS created the QPI system.

In our current inspection, we found that the QPI system is not an effective method to prioritize violent fugitive investigations. Supervisors do not use the QPI to manage their district's priorities or to measure their district's effectiveness in apprehending violent fugitives. Deputies do not use the QPI to prioritize warrant investigations and there are no consequences for districts that do not meet the QPI target.

Due To Design Flaws, The QPI System Is Not An Effective Method To Prioritize Violent Fugitive Investigations

The QPI system is a flawed management tool. We identified problems with the QPI point structure that diminishes its usefulness as a tool to prioritize fugitive investigations on violent fugitives. If the QPI system was a useful tool, higher points should result in greater investigative priority in apprehending violent fugitives. However, we found that this is not always the case. Based on the current point structure, non-violent fugitives can receive the same or a greater number of points than a violent fugitive would; therefore, the number of points does not necessarily distinguish violent fugitives.

The QPI system does not easily identify violent fugitives because points are awarded for variables unrelated to violence. Points based on other variables, such as the age of the warrant or the existence of a state or local warrant, can result in a non-violent fugitive receiving higher points than a violent fugitive. Neither of these variables are directly related to violence. Some supervisors and deputies stated that they have encountered cases in which non-violent fugitives received the same number of points as violent fugitives. In addition, some deputies we interviewed questioned why the warrant received extra points because their experience was that the presence of additional warrants does not necessarily mean that a fugitive is more violent.

We analyzed FY 1998 closed warrants to determine the type and percentage of violent fugitives apprehended for given point totals. For every point total, we compared the percentage that were violent and non-violent. We found instances where fugitives classified as non-violent received the same number of points as fugitives classified as violent. Based on our analysis, 93 percent of the warrants closed in FY 1998 receiving six points were for violent fugitives, while 7 percent were for non-violent fugitives. If districts prioritized cases based on the current point structure, non-violent and violent fugitives could receive equal investigative priority. We found, in FY 1997, a district received six points for arresting a fugitive on a warrant for homicide who had an extensive criminal history involving three prior arrests for homicide. Another district also received six points for arresting a fugitive on a warrant for marijuana possession who had no known prior criminal history (see case summaries).

Case Summary 1
  • The fugitive's criminal history involved three prior arrests for homicide and narcotics. In addition, he was convicted once on narcotic charges.
  • The fugitive was categorized as a violent fugitive (offender category 1).
  • His warrant was issued for a homicide charge.
  • He was physically arrested by the USMS within one year of the issuance of the warrant.


Case Summary 2
  • No known prior criminal history.
  • The fugitive was categorized as a non-violent fugitive (offender category 2).
  • His warrant was issued for possession of marijuana.
  • He was physically arrested by the USMS more than one year after the issuance of the warrant.


    Source: USMS WIN System

Finally, if the QPI system truly measured violence, higher point values should have been assigned to violent fugitives than to non-violent fugitives. While violent fugitives generally have higher QPI points than non-violent, this is not always the case. We found that 93 percent of the warrants closed in FY 1998 receiving six points were for violent fugitives, yet all the warrants receiving seven points were for non-violent fugitives.

Supervisors Do Not Use The QPI To Manage District's Priorities Or To Measure District's Effectiveness

None of the warrant supervisors in the 12 districts whom we interviewed use the QPI system to manage their deputies' priorities, nor do they use the QPI as an indication of how their district office is performing.15 The supervisors reported that they do not use the QPI as a measure of effectiveness or as a method for selecting the next warrant to be investigated. Specifically, district officials reported that they lack the time, resources, and staffing to review and fully understand the QPI numbers; the QPI does not reflect the level of effort required to investigate a fugitive case; the QPI conflicts with other priorities that the district office is expected to fulfill (e.g., closing warrants quickly); the small size of their districts (and thus the smaller caseload) makes it difficult, if not impossible, to meet the QPI target number; and the QPI can be higher (or lower) depending on the abundance (or dearth) of fugitive information available to calculate the offender categories in the WIN system.

Headquarters distributes monthly reports listing each district's QPI. Although the supervisors confirmed the receipt of these reports, they told us that they did not use them as a management tool.

Deputies Do Not Use The QPI To Prioritize Warrant Investigations

None of the deputies with whom we spoke used the QPI system to prioritize the order in which they investigate their warrant caseloads. Instead, the deputies work their caseloads by the type of leads they receive — time-sensitive leads receive first priority for investigation. Because time-sensitive leads need to be worked immediately, the deputies explained that putting the warrants in priority order is not usually an option. Deputies exhaust all avenues until no new information is available on the warrant. In general, deputies do not believe the QPI system can replace their investigative judgement.

Furthermore, out of all our 50 respondents, only one supervisor expressed confidence in his understanding of the QPI system. Everyone else was either "not very" familiar or were only "vaguely familiar" with the QPI's structure and purpose. In addition, only 4 of the 12 districts had at least one supervisor or deputy who was aware of the FY 1999 4.8 QPI target established by headquarters.

Districts Do Not Suffer Consequences For Failing To Meet The QPI Target

Our respondents report that there are no consequences for failure to meet headquarters' QPI target. As of September 1998, 37 out of 94 districts (39 percent) were not meeting the FY 1998 national 4.8 QPI target. Some supervisors and deputies felt that this lack of involvement by headquarters sends the message that the QPI is not important and that the QPI system is not supported by USMS as an important tool to prioritize investigations.


In our 1995 report, we found that the WIN system could be used more effectively as a warrant investigative tool. We also found that the districts could benefit from more comprehensive investigative information. Based on these findings, we recommended that the USMS enhance the WIN system and provide additional investigative support to districts.

In our current inspection, we found limited progress in enhancing the WIN system and providing investigative support to the districts.

No Enhancements Have Been Made To The WIN System's Application Software

The WIN system is an investigative tool that can provide vital information necessary for apprehending fugitives. During the previous review, the team learned that not all districts had access to the WIN system. Deputies working in districts without access sent all fugitive information to headquarters for entry into the system. Because this was an inefficient and unreliable process, we recommended enhancements to the WIN system. The USMS agreed to ensure that all districts had direct access to the system. Headquarters also decided to offer more training to those using the WIN system, and dedicated one ITD programmer to making system enhancements requested by the ISD.16 In addition, headquarters stated that it would make plans for on-going maintenance of the system.

During the course of our field work, we found that while every district now can access the WIN system, many were unsatisfied with their level of training and their ability to make maximum use of system capabilities. Supervisors and deputies expressed dissatisfaction with WIN's format. Many respondents suggested improvements, including: downloading photographs and fingerprint records directly into WIN, enlarging the comment field, and creating a more user-friendly system. Based on these suggestions and headquarters requirements, headquarters identified a series of WIN enhancements to begin in October 1998.17

The ITD has not implemented any of these improvements to the WIN system's application software.18 For the past several years, the ITD focused efforts on upgrading the infrastructure of the WIN system by replacing older computer servers with newer, more advanced ones. This project has taken nearly three years to complete.

No formal process is followed for the assignment and prioritization of projects requested by the DIU and the ASU. Headquarters cited severe budget constraints as a key factor in not making WIN enhancements. In addition to budget constraints, only 1 person out of approximately 30 is permanently assigned to handle WIN issues in the ITD.19 The ITD staff told us that until the unit receives more funding and staffing for the fugitive program, it does not plan to make enhancements until infrastructure improvements are completed.20

We were informed by field personnel that limited resources were available for WIN system training. Some field personnel with whom we spoke stated that they had not received adequate guidance for using the system when they attended USMS basic deputy training at the Federal Law Enforcement Training Center in Glynco, Georgia. Headquarters has sent some of its staff to train district personnel, but no consistent plan exists for providing WIN-related information and training to the districts.21

Districts Report That Additional Resources Devoted To The ESU Are Helpful To District Fugitive Apprehension Efforts

Deputies rely on investigative assistance from the ESU. To assist in the apprehension of fugitives, the unit provides electronic surveillance assistance to headquarters and districts in the form of vehicle tracking, audio and video surveillance, pager intercepts, and cellular telephone tracking. ESU staff also advise deputies on procedures for requesting necessary court orders to use these electronic surveillance methods. Even though we did not specifically recommend expansion of the ESU in our 1995 inspection report, the ESU has grown from 5 deputies in 1995 to 22 deputies in 1999 located throughout the United States to support the 94 USMS districts.

Almost all deputies and supervisors we contacted have requested ESU assistance. In a 1999 survey of district offices conducted by the USMS, the ESU received, along with Legal Affairs, the highest rating on responsiveness, timeliness, and guidance. Those responding to the survey were highly complimentary of the ESU's technical ability and workmanship quality. However, they also noted that ESU does not have enough staff to meet the demand for their services.

An ESU official reported that the office receives approximately 150 requests from different districts each week (the ESU responds to requests regarding fugitives designated as Top 15 Most Wanted, Major, and category 1 — fugitives who are typically violent). Almost 25 percent of these requests are not acted on because the ESU said that they do not have the staffing to meet the demand.

Districts And Headquarters Report That The ASU Has Not Provided The Level Of Investigative Assistance To The Fugitive Apprehension Program As Intended

Our original report recommended that the USMS provide additional investigative support to the districts. As a result, the USMS created the ASU and staffed it with a team of analysts who assist deputies on high priority investigations, research information for older cases whose leads have been exhausted, investigate judicial threats and inappropriate communications to the court, and maintain the WIN system.22

Despite its mission to assist with fugitive investigations, ASU analysts spend the majority of their time working on judicial threats. Headquarters reported fugitive investigations are not a priority for the ASU. When we questioned supervisors and deputies in the field about investigative assistance provided by the ASU, the responses were mixed. Some deputies used the ASU and were satisfied with the help they received. More typically, the deputies felt they rarely needed to contact the ASU for assistance because the ASU primarily provides judicial threat assessments. A few deputies were not even aware of the ASU's existence.


Overall, the USMS has made progress since our 1995 report by establishing and implementing its national goal to reduce the warrant backlog. Before the warrant backlog goal was set at 20 percent, the USMS assessed the composition of the backlog to determine whether any adjustments were necessary. As a result of this assessment, the USMS established a more accurate definition of warrant backlog. In addition, the USMS backlog reduction plan provided practical steps for districts to reduce the warrant backlog. These actions are positive steps that will help alleviate the longstanding problem of the warrant backlog.

Progress on implementing a meaningful goal to close cases quickly has not been as successful as the backlog reduction effort because the USMS has not increased the percentage of warrants closed within one year since FY 1995. The achievement of this goal has not encouraged substantive changes in the percentage of warrants closed because the USMS has exceeded this goal by 6 percentage points in each of the last four years. The USMS needs to examine its goals, priorities, and resources to reassess and perhaps raise the 80 percent goal.

The USMS has no quantifiable goal for targeting violent fugitives, even though targeting violence is an integral part of the USMS fugitive program. Given the importance of targeting investigative resources on violent fugitives, the USMS needs a goal and an appropriate performance measure to ensure that violent fugitives receive investigative priority. Even though the QPI system was developed as a method to set investigative priorities, currently it is not an effective tool to set priorities nor does it provide a direct method for the USMS to monitor progress in targeting violent fugitives. The USMS needs an accurate, more direct method to track its progress in targeting violent fugitives.

Finally, we found the ASU has provided only a limited amount of additional investigative support to the field. While all districts now have direct access to the WIN system, the system's capabilities have not been enhanced. In addition, headquarters plan to use the ASU to support the fugitive program has not materialized because the ASU focuses almost exclusively on investigating judicial threats and maintaining the WIN system.

The OIG, Inspections Division recommends that the Director, USMS:

  1. Reevaluate the goal to close 80 percent of warrants within one year, given that it has little meaning, and set a more appropriate goal.
  2. Establish a quantifiable goal for apprehending violent fugitives.
  3. Either abolish the QPI system or correct the system to make it an effective management tool.

  1. The USMS fugitive program accounted for 14 percent of FY 1998 USMS budget expenditures.
  2. FY 1997 Annual Report of the United States Marshals Service, page 7.
  3. The QPI does not take into account the USMS national goal of closing 80 percent of warrants within one year of the warrant date.
  4. All points shown in the table are given to the district maintaining the warrant. Not shown in the table are additional points a district not maintaining the warrant receives if it arrests a fugitive on a collateral lead. Collateral leads are requests for assistance from one district when a fugitive has fled to another district. Currently, the same number of points are given regardless of how the warrant is executed (e.g., physical apprehension, surrender).
  5. Exceptions include fugitives that the USMS or the Department of Justice (DOJ) designate as Top 15 or Major because the fugitive has committed a non-violent crime that has received national attention. In FY 1998, the USMS closed 10,046 warrants categorized as Top 15, Major, and Category 1. Only 19 (0.2 percent) of those warrants closed in FY 1998 were categorized as non-violent Major cases. (All Top 15 and Category 1 fugitives for FY 1998 were violent fugitives).
  6. The USMS spends most of its time investigating Class I fugitives wanted on felony warrants. However, the USMS also has investigative responsibility for misdemeanors and traffic warrants occurring on federal property.
  7. The USMS met its backlog reduction goal by reducing the warrant backlog carried over from FY 1998 and also closing any warrants added to the backlog during FY 1999.
  8. The CCCT's visited the following 12 largest warrant backlogged districts: Northern Illinois, Central California, Southern California, Northern California, Southern Texas, Western Texas, Northern Texas, Southern Florida, Middle Florida, Eastern New York, Southern New York, and Arizona.
  9. Major problems outlined in the CCCT's January 25, 1999, report included: the U.S. Attorneys' Offices refused to dismiss warrants; duplicate fugitive warrants were issued for the same offense; the USMS definitions for out-of-country cases were inconsistent; and some districts were unsatisfied with the amount of WIN training and administrative help.
  10. The USMS is responsible for entering warrant information it receives from DEA into NCIC. However, during a recent review conducted by the USMS, warrants were found that lacked vital suspect information (e.g., race, sex, height, weight, name), making it impossible for the USMS to enter the information into NCIC. Without this information in NCIC, the chances of locating the fugitive will decrease.
  11. The USMS calculates the percentage of warrants closed within one year of the warrant date by dividing all warrants closed within one year by all warrants under a year old. The following number of warrants were closed within one year by the USMS: 15,658 in FY 1994; 16,248 in FY 1995; 16,130 in FY 1996; 18,846 in FY 1997; 21,351 in FY 1998; and 22,150 in FY 1999.
  12. To determine whether the offender categories are a valid measure of violence, we compared the violent offender categories with the FBI definition of a violent criminal and found that they are similar.
  13. We recognize that this increase may be explained by a number of factors, including the USMS emphasis on apprehending violent fugitives and the completeness of the criminal history field in the WIN system. We do not believe the increase can be explained by an increase of violent fugitives in the general population because nationally the number of violent crimes fell 7 percent in 1998, creating the largest annual decline since crime began to decrease in 1992.
  14. To calculate the percentage of warrants closed for violent fugitives, we divided the number of closed warrants for violent fugitives (fugitives categorized as Top 15 Most Wanted, Major or category 1) by all USMS warrants closed. The USMS closed the following number of warrants for violent fugitives: 6,052 in FY 1994; 6,615 in FY 1995; 7,192 in FY 1996; 9,078 in FY 1997; and 10,046 in FY 1998.
  15. However, one warrant squad supervisor stated that he did use the QPI as one criteria for choosing deputies who investigate warrants as a collateral duty when selecting new members of the permanent warrant squad.
  16. The ISD was formerly known as the Enforcement Division.
  17. As outlined in a September 22, 1998, memorandum to the ITD, the improvements included efforts to: increase the districts' ability to generate new reports and queries in order to assist districts to meet the Attorney General's backlog reduction goal; create new reports to record accomplishments of task force operations; develop additional data fields to monitor work being done on misdemeanor and traffic warrants; upgrade the system to better record tracking, sending, and receiving of collateral leads; and upgrade the system to allow districts to enter fugitive photographs.
  18. Even though WIN has not been enhanced to store photographs, the districts recently received the capability to send, store, and receive photographs through their automated prisoner booking stations.
  19. When time permits, another employee can spend up to 50 percent of his time assisting with WIN issues.
  20. As of July 1999, all infrastructure changes have been completed. If additional funding becomes available, the ITD plans to enhance the WIN system by incorporating it into the Justice Detainee Information System (JDIS). The ITD intends to combine the WIN system, the Prisoner Tracking System, the Automated Prisoner Scheduling System, and automated booking stations into the JDIS system.
  21. Staff turnover due to deputies rotating among districts is another problem with keeping WIN training current.
  22. During the previous inspection, the DIU provided investigative support for Top 15 and Major cases. The DIU continues to provide investigative support to districts and has grown from four deputies in 1995 to eight deputies 1999.