Review of Shooting Incidents in the Department of Justice

E & I Report I-2004-010
September 2004

We found that the ATF, the DEA, the FBI, and the USMS have different policies and practices for reporting, investigating, and reviewing shooting incidents involving law enforcement officers (LEOs). Although we found that LEOs reported most shooting incidents, the field offices did not always report shooting incidents to headquarters within the time required and did not consistently report shooting incidents to the Office of the Inspector General or the Civil Rights Division. The components also conducted shooting incident investigations differently. The ATF and the USMS conducted a headquarters investigation in every case, while the DEA and the FBI delegated some investigations to field offices. The FBI conducted its own criminal investigations in every case, while the ATF, the DEA, and the USMS requested that local law enforcement agencies conduct the criminal investigations of their shooting incidents. The components used different standards for obtaining declinations of prosecution and for compelling statements during administrative investigations. In addition, the components' Review Boards applied the standard for the reasonable use of deadly force differently and referred cases for discipline in different ways. The components did not systematically share the lessons learned from shooting incident reports, and the Department did not aggregate shooting incident data to identify needed improvements to training.

Reporting to Supervisors, Component Headquarters, the CRD, and the OIG.

Reports of shooting incidents to supervisors. Every component requires LEOs to report all firearm discharges (other than for training or recreation) to their supervisors immediately after they occur because an immediate report by the LEO is essential to begin the investigative process. In the 103 incidents we reviewed, the LEOs immediately reported 100 of the shooting incidents. However, in three cases, LEOs failed to report shooting incidents as required. In those three incidents:

• A DEA Special Agent failed to report that he discharged his weapon during an enforcement operation. After a month-long investigation into the unexplained gunfire heard by other LEOs at the scene, the Special Agent admitted to investigators that he had fired his weapon. The Special Agent received a 30-day suspension for failing to report the firearm discharge immediately.

• A DEA Special Agent failed to report that he discharged his weapon during an enforcement operation. Six months later, while testifying at the suspect's trial, the Special Agent admitted to discharging his weapon. The Special Agent received a 35 day suspension for failing to immediately report the firearm discharge.

• An FBI Special Agent waited one day to report that he had unintentionally discharged his weapon outside his residence while chasing a burglar. The Special Agent received a 5 day suspension for failing to exercise appropriate firearms safety and failing to report the incident immediately.

Two of these three incidents were not investigated fully because a shooting incident investigation was no longer practicable when the shooting incident was discovered. According to the DEA Chief of Inspections, failure to report the incidents immediately constituted misconduct. Therefore, the Office of Inspections transferred these two investigations to the DEA's OPR. The OPR conducted misconduct investigations and imposed discipline.

Field office reporting of shooting incidents to headquarters. The components require timely written reports so that senior managers can oversee the investigative process. Following the immediate reporting of a shooting incident by an LEO, each component requires the LEO's supervisor to submit an initial written report on the shooting incident to headquarters within one day.³¹ Of the 103 incidents we reviewed, we were able to analyze 97 for timeliness. We did not consider in our analysis the above three incidents that were not reported as required and an additional three from the FBI because a SAC may classify a shooting incident as unintentional and, if no one is injured or killed, an immediate report is not required.³² We reviewed the files of the remaining 97 incidents for copies of the written report to headquarters. Of the 97 incident files, 11 did not contain the required written report, and in 1 file the report was undated. Although these 12 files contained evidence of telephone calls and e-mails informing headquarters of the shooting incidents, these informal communications did not include all the information that the components require in the formal written report (Figure 7).

Figure 7: Missing or Undated Written Shooting
Incident Reports, FY 2000 - FY 2003

Source: OIG analysis of components' shooting incident files.

For the 85 incident files that contained dated initial written reports, we examined how long the supervisors took to submit the reports to headquarters. Of the 85 incidents, a formal written report was submitted within one day in 54 of the incidents (64 percent). On average, the ATF and the FBI met the 1-day reporting requirement, while DEA and USMS supervisors took two to three days on average to submit the reports (Figure 8).

Figure 8: Average Number of Days to Submit Initial Written
Report to Headquarters, FY 2000 - FY 2003

Source: OIG analysis of components' shooting incident files. Note: Reporting time is measured in complete days following the day of the incident. Reports submitted on the day of the incident were counted as zero days to report, reports submitted on the following day were counted as one day to report, etc.

We discussed the reporting delays with each of the components. The USMS staff stated that the USMS organizational structure was one factor that delayed the reporting process, because the USMS has 94 decentralized district offices, each of which rarely has a shooting incident. The DEA policies, provided by the DEA shooting incident investigators, require at least three reports following a shooting incident:

• Based on an initial telephonic report, DEA headquarters may designate a shooting incident as a "critical incident." If this happens, the SAC is required to send a "preliminary shooting incident report" to headquarters within six hours. We found these critical incident reports in 26 of the 35 DEA incident files. A critical incident report may not have been required in the remaining nine cases, but we did not find any documentation of the critical incident decisions.

• In every shooting incident, the SAC must send an electronic communication describing the shooting incident to headquarters within 24 hours. We found an electronic communication from the SAC in 33 of 35 incident files and used this initial written report for our timeliness assessment.

• Finally, the Shooting Incident Investigator must submit a Report of Shooting Incident after completing the investigation. We found a Report of Shooting Incident in 30 of the 35 incident files.

Overall, we identified an initial written report in 85 of the 97 incidents. We found no dated, initial written report in 12 cases, and in 31 of the 85 incidents the initial reports did not meet the 1-day deadline. As a result, in 43 of 97 cases (44 percent) we could not validate that senior managers received the written information in time to use it for oversight of the investigation.

Reporting shooting incidents to the CRD. The Department's CRD has the authority to review and prosecute complaints of violations of federal civil rights statutes, including complaints related to the shooting incidents we reviewed. In 2000, the CRD established written agreements with the DEA and the FBI that require them to report to the CRD shooting incidents resulting in injury or death.³³ The CRD does not have similar written agreements with the ATF or the USMS.

We examined the DEA's and the FBI's compliance with the agreed upon reporting requirements. Since the FBI's agreement went into effect, 15 FBI shooting incidents have resulted in injury or death. The FBI reported 14 of these 15 shooting incidents to the CRD. The FBI could not explain why one of its cases was not reported. Since the DEA's agreement went into effect, 11 DEA shooting incidents resulted in injury or death. The DEA reported 8 of these 11 shooting incidents to the CRD. The DEA did not report three incidents to the CRD. Under the DEA's procedures, its OPR is responsible for reporting potential civil rights violations to the CRD. However, DEA's shooting incident investigators do not forward to the OPR those cases that they determine do not involve misconduct. The investigators never forwarded the three incidents to the OPR because, though they involved injury or death, the investigators did not identify any potential misconduct. Under the procedures in effect at the time, the OPR did not receive separate notice of shooting incidents.

The ATF and the USMS have no written agreements with the CRD to report shooting incidents. However, the USMS Chief Inspector told us that the USMS would report any shooting incidents involving a potential violation of civil rights. Also, the ATF Special Agent in Charge, Investigations Division, Office of Inspection, told us that the ATF would report any shooting incident involving the allegation or suggestion of a civil rights violation to the appropriate United States Attorney's Office. We found no evidence (e.g., declinations of prosecution, memoranda, telephone records, e mail) in the ATF and USMS incident files of any reports provided to the CRD. We also found that the CRD considered two USMS cases we reviewed, but not because they were reported by the USMS. In one USMS shooting incident, the CRD received an allegation of a civil rights violation from the suspect's family, and in another, a United States Attorney consulted with the CRD before deciding not to prosecute the Deputy Marshal. The ATF case files contained no evidence that any of the 23 ATF cases resulted in a civil rights complaint against an ATF Special Agent.

In addition to the inconsistency among the components in reporting to the CRD, we found that the DEA's and the FBI's agreements with the CRD do not ensure that all incidents with the potential to result in civil rights complaints were reported to the CRD. As noted, the DEA and the FBI agreed to report to the CRD all shooting incidents that resulted in injury or death. However, incidents without injury or death that are not reported to the CRD may still result in allegations of civil rights violations. In one incident we reviewed, a suspect, who was not injured, made a civil rights complaint after local authorities arrested him. The suspect alleged that an FBI Special Agent unnecessarily shot at him as he eluded arrest by the FBI Special Agent. As this case illustrates, firearm discharges that do not cause injury or death may result in complaints of civil rights violations by LEOs that the CRD should review.

Reporting shooting incidents to the OIG. Attorney General Order 2492-2001 (Order 2492) requires components to report to the OIG "evidence and non-frivolous allegations of criminal wrongdoing and serious administrative misconduct by Department employees" so that the OIG may decide whether to exercise its statutory authority to investigate the matter or to delegate the investigation to the component. All shooting incidents have the potential to involve criminal wrongdoing or serious administrative misconduct. Of the 103 incidents we reviewed, 46 were not reportable to the OIG because they occurred before this OIG had the authority to investigate misconduct involving FBI and DEA employees. ³⁴

Of the incidents we reviewed, 57 were reportable to the OIG. We found documentation in the files that the components submitted formal reports in 35 of the 57 reportable incidents (61 percent). The USMS and the ATF reported every incident, but we found no documentation for more than half of the DEA and FBI incidents (Figure 9).

Figure 9: Shooting Incidents Reported to the OIG

Source: OIG analysis of components' shooting incident data.

The average time that each component took to formally report shooting incidents to the OIG ranged from 7 to 205 days. The ATF, the DEA, and the USMS reported most shooting incidents to the OIG within a few days, but the FBI did not formally report its incidents to the OIG until after the report of the administrative investigation of the shooting incident was submitted to the FBI Review Board (Figure 10).³⁵

Figure 10: Average Time to Report Shooting Incidents to the OIG

Source: OIG analysis components' shooting incident files and IDMS data.

The USMS reported all shooting incidents as required, and after reviewing the reports, the OIG delegated all of the investigations back to the USMS. Although the ATF, the DEA, and the FBI did not consistently provide timely written reports, the OIG ASACs who are the primary liaisons to each component's internal affairs unit told us that they visit those components' headquarters regularly and review the components' internal shooting incident reports. The ASACs said that they base their decisions to delegate shooting incident investigations on their review of these reports, as well as other interactions with component staff, and annotate the components' files to indicate the OIG decision.

We also found that the components had different interpretations of what constituted a shooting incident reportable to the OIG. The ATF and the USMS reported all shooting incidents to the OIG. In contrast, when the FBI and the DEA determined initially that a shooting incident was unintentional, they did not report it to the OIG. This practice can lead to failing to report an incident that should be reported. In one case we reviewed, which occurred prior to the OIG obtaining jurisdiction to investigate FBI and DEA employees, a DEA Special Agent at the shooting scene stated that he unintentionally discharged his weapon. This case would not have been reported to the OIG under current DEA procedures. However, in a later statement to the local authorities, the Special Agent gave a different version of events and stated that he intentionally discharged his firearm because he feared for his life. Intentional discharges of this kind are now reportable under DEA procedures. Overall, we found that the components did not report to the OIG 22 of the 57 reportable shooting incidents. Moreover, the FBI did not formally report any of its shooting incidents to the OIG until after the FBI completed its investigation.

All firearms discharges, except those in connection with training or off duty sports and hobbies, must be reported to the OIG. To ensure that shooting incidents are properly reported to the OIG, on July 27, 2004, the Assistant Inspector General of the Investigations Division sent a memorandum to the ATF, the DEA, and the FBI reiterating the existing reporting requirements in order to eliminate any confusion over the requirement to report firearms discharges.

The Components' Shooting Investigations

Resolution 13 does not specify how shooting investigations should be conducted, but requires that all investigations be objective, thorough, and timely. Specifically, Resolution 13 states that investigations must:

• Avoid any appearance of conflicts of interest or impropriety,

• Present all the information necessary for analysis, and

• Be conducted expeditiously.

Resolution 13 also requires that an investigation be "appropriate for the type of incident involved," but authorizes the components to decide who will conduct the investigation:

The decision whether a shooting inquiry will be conducted by investigators assigned to the field office where the incident occurred or by investigators assigned to a component Headquarters Office of Inspection or other Headquarters element, will be made by designated component Headquarters Senior Management following consultation with field office Senior Management.

The components call the decision to have a field office conduct the investigation "delegating the investigation."

Delegation of the shooting incident investigation. The criteria for delegating shooting incident investigations vary by component. We found that the ATF has the strictest limitation on delegating shooting incidents. It requires that all intentional and unintentional firearms discharges by ATF employees be investigated by inspectors from the Office of Inspections. In all 23 ATF cases in our review, ATF headquarters dispatched an Office of Inspections team to conduct the shooting incident investigation.

The DEA generally delegated shooting incidents that did not result in "significant or life threatening injuries, deaths, or other significant liabilities" to the SAC of the DEA field office to which the Special Agent was assigned. We found that the DEA delegated 22 (59 percent) of its 37 shooting cases to the SAC for investigation. Of these 22 cases, 7 resulted in injury or death and 15 did not (9 were intentional discharges and 6 were unintentional discharges during enforcement operations).

According to the FBI Manual of Investigative Operations and Guidelines, the decision to delegate an investigation is not based on the seriousness of the incident but rather "on the extent of the SAC or ASAC participation in the planning and operational events of the incident." Of the 39 FBI cases we reviewed, 16 (41 percent) were delegated to the SAC. In two cases (5 percent), a team was dispatched from headquarters to investigate, and in 21 cases (54 percent), headquarters assigned an IIP to oversee the investigation. The FBI policy distinguishes investigations overseen by an IIP from investigations delegated to the SAC; however, in 6 of the 21 investigations we reviewed, we found that the IIPs directed to investigate were assigned to the same field offices as the Special Agents involved in the shooting incident.

The USMS Office of Internal Affairs (OIA) forms a team to conduct the investigation of each shooting incident involving a Deputy Marshal. USMS case files did not record whether investigators were dispatched from headquarters, assigned from another district, or assigned from within the local district office, but every investigator reported directly to the USMS OIA.

Criminal investigations of shooting incidents. In reviewing 124 individual cases arising from 103 shooting incidents, we found that local law enforcement agencies conducted the criminal investigations of most ATF, DEA, and USMS shooting incidents, while the FBI conducted the criminal investigations of all of its shooting incidents (Table 3).

Table 3: Criminal Investigations of ATF, DEA, FBI, and USMS Shooting Cases by Federal and Local Law Enforcement
Investigated by:
	State/Local	Component	Both	Cases
ATF	21	2	0	23
DEA	27	13	0	40
USMS	14	8	0	22
FBI	12	39	12	39
				124

Source: OIG analysis of components' shooting cases.

The ATF, the DEA, and the USMS told us that they always requested that state or local law enforcement agencies conduct the criminal investigations and that they only conducted a criminal investigation themselves if the state and local agencies declined the request. That was reflected in our review of the case files. We found that state and local agencies conducted the criminal investigations in 62 of the 85 shooting cases involving LEOs of the ATF, the DEA, and the USMS. The components conducted the criminal investigations in the remaining 23 cases after local law enforcement declined to investigate.³⁶

The ATF, the DEA, and the USMS shooting incident investigators gave several reasons for having state or local agencies conduct the criminal investigations. First, because state and local agencies have no vested interest in the outcome of an investigation, having them conduct the criminal investigation of the LEOs' actions avoids any appearance of a conflict of interest. The ATF, DEA, and USMS investigators also told us that, in their opinion, local authorities are better able to secure the scene, preserve physical evidence, and interview witnesses. One ATF Shooting Incident Investigator noted that local law enforcement authorities routinely investigate firearms discharges in their jurisdictions. Local authorities are also better able to address any investigative issues that arise away from the scene of the incident or after the initial shooting response ends. This was demonstrated in two of the cases we reviewed.

• A Deputy Marshal observed a man threatening two women and intervened. The man lunged at him and the Deputy fired at the man in self-defense, but did not know if he hit him. The man and the two women fled the scene. The man later sought treatment for his gunshot wound, and the hospital staff called the local police. When the police arrived at the hospital, the man complained to them that the Deputy Marshal had unfairly shot him.

• A DEA Special Agent was involved in a shooting incident, but thought he had missed the suspect. A week later, local law enforcement officers arrested the suspect and found a bullet wound in the suspect's arm. The suspect attributed the wound to the earlier shooting incident.

The states have the primary authority and responsibility to investigate and prosecute assaults, homicides, or other felonies occurring in their jurisdictions and, in fact, investigated the vast majority of firearms discharges, including most of the 9,369 homicides committed with a firearm in 2002.³⁷ Allowing local authorities to carry out their duties may eliminate the need for an extensive federal investigation. Relying on local authorities to conduct criminal investigations also reduces the possibility that separate local and criminal investigations will lead to conflicting conclusions.

In contrast to the ATF, the DEA, and the USMS, the FBI always conducted a criminal investigation on any shooting incident involving its Special Agents. The FBI conducted criminal investigations on all 39 cases we reviewed. However, we found that state or local agencies also conducted an investigation in 12 of those 39 cases.³⁸

We asked the Deputy Assistant Director why the FBI conducted a federal criminal investigation in every case, even if it duplicated a local criminal investigation. He told us that for cases involving injury or death, the FBI's reporting agreement with the CRD requires that the FBI investigate any potential violation of civil rights.³⁹ He said that policy has been extended, by practice, to every shooting incident. He also told us that the FBI would not discourage any local investigation and would cooperate with state or local investigators if they conducted an investigation, but that they would not request an investigation by a local law enforcement agency because the FBI had the resources to conduct investigations in all cases.

The DEA has a similar requirement in its reporting agreement with the CRD but takes a different approach to the criminal investigation of shooting incidents. The DEA meets its requirement by asking local law enforcement authorities to investigate the shooting and submitting the local criminal investigations to the CRD.

We discussed with CRD attorneys the fact that the DEA and the FBI have similar agreements with the CRD but have different approaches to conducting criminal investigations into shooting incidents. The CRD attorneys said that they used the local criminal investigations of the DEA shooting incidents to determine whether a civil rights violation had occurred. The CRD attorneys also said that local authorities questioned the objectivity of the FBI's criminal investigations of shooting incidents involving its Special Agents in at least one case. In that case, FBI headquarters admonished the FBI SAC for turning over the scene of an FBI shooting incident to local and state criminal investigators who insisted on conducting a criminal investigation of the incident.

The rights of the LEOs involved in shooting incidents. Both Resolution 13 and the components' policies direct that the investigation balance the importance of conducting an objective, thorough, and timely criminal investigation with protecting the rights of the LEOs for whom shooting incidents are traumatic events. LEOs may exercise their constitutional right to remain silent during the criminal investigation or can make voluntary statements, with the advice of an attorney if they choose. In 101 of the 124 shooting cases we reviewed, the LEOs who fired their weapons made voluntary statements, and in 30 of the 101 cases, an attorney representing the LEO was present during the statement to advise the LEO (Table 4).

Table 4: Voluntary Statements With Attorneys Present
Component	Cases	Cases With Voluntary Statements	Attorney Present at Voluntary Statement	Percentage
ATF	23	21	6	29
DEA	40	40	11	28
FBI	39	25	5*	20
USMS	22	15	8	53
TOTAL	124	101	30	30

Source: OIG analysis of components' shooting cases. * Three additional FBI Special Agents were represented by counsel, but their attorneys were not present when the voluntary statements were given.

Declinations of prosecution. After the criminal investigation is complete, each component requires investigators to obtain a declination of prosecution from state or federal prosecutors or otherwise ensure that there is no criminal action pending before completing the administrative investigation of a shooting incident. Because both state and federal declinations of prosecution are recorded in the case files, we were able to evaluate their use by the components. When they sought a declination, we noted that only the FBI followed the practice of obtaining declinations from both federal and state prosecutors (Table 5).

Table 5: Declinations of Prosecution Obtained by Components
Component	Cases	Federal Declinations	State Declinations	Federal and State Declinations
ATF	23	1	18	0
DEA	40	9	21	4
FBI	39	23	22	21
USMS	22	2	3	0
TOTAL	124	35	64	25

Source: OIG analysis of components' shooting cases.

We also examined components' compliance with their individual policies regarding obtaining declinations of prosecution. The DEA and the FBI reported to us that they do not routinely obtain declinations of prosecution in cases that do not involve injury or death. However, we found that the DEA and the FBI did not obtain a declination of prosecution in one case involving death or injury each. The ATF and USMS are required by their policies to obtain a declination of prosecution in every case. Yet, the ATF did not obtain a declination of prosecution in 4 cases and the USMS did not obtain a declination of prosecution in 17 cases (Table 6).

Table 6: Cases in Which No Declination of Prosecution Was Obtained
Component	Cases	No Declination	Death or Injury	No Death or Injury
ATF	23	4	0	4
DEA	40	14	1	13
FBI	39	15	1	14
USMS	22	17	5	12
TOTAL	124	50	7	43

Source: OIG analysis of components' shooting cases.

The ATF investigators acknowledged that they had not obtained the required declinations in four cases due to an administrative oversight; these four ATF cases involved shooting incidents in which the suspect was neither injured nor killed. The DEA investigators said that a declination of prosecution was not obtained in one case because it involved a minor wound to a suspect who eluded arrest. The FBI could not explain why one of its cases was not reported to the CRD for the required declination of prosecution. The USMS told us that its policy of obtaining declinations of prosecution in every case "needed to be more actively enforced."

Compelled statements in the administrative investigation. As noted above, during criminal investigations, LEOs have a constitutional right under the Fifth Amendment to remain silent. But once the criminal investigation is complete, the government can compel LEOs to provide statements for the administrative investigation and can discipline them if they refuse to comply. If a government agency compels an employee to provide a statement, it may use information in that statement to take administrative action against the employee, but it may not use the information against the individual in a criminal prosecution.⁴⁰ In the 124 cases we reviewed, the components compelled 30 administrative statements (Table 7).

Table 7: Statements by LEOs Involved in Shooting Incidents
Component	Cases	Voluntary Statement	Compelled Statement	Both Voluntary and Compelled Statements	No Statement
ATF	23	21	2	0	0
DEA	40	40	0	0	0
FBI	39	25	14	1	1*
USMS	22	15	14	7	0
TOTAL	124	101	30	8	1

Source: OIG analysis of components' shooting cases. * One FBI Special Agent retired before the administrative investigation was completed.

We discussed the practice of compelling statements with investigators from each of the components. The ATF and DEA shooting incident investigators we interviewed said that it is their components' policy not to compel statements unless they are needed to complete the administrative investigation. The ATF shooting investigator said this was the ATF's reason for compelling statements in two ATF cases. The FBI Deputy Assistant Director we interviewed confirmed that it is FBI practice to compel an administrative statement in all administrative cases unless the LEO has already made a voluntary statement in the criminal one. The FBI Deputy Assistant Director also stated that the administrative warnings and criminal immunity associated with compelled administrative statements were necessary to protect the rights of the LEOs. The Deputy Marshal now responsible for investigating shooting incidents told us that, in the past, the USMS's practice was to compel an administrative statement in every case, but that the USMS discontinued the practice. Currently, USMS policy authorizes compelled administrative statements only when necessary to complete the administrative investigation.

Our review of the case files also found that the components followed substantially different practices for compelling administrative statements in similar situations. In four FBI shooting cases, investigators compelled administrative statements (without declinations of prosecution) from the Special Agents involved in shooting incidents for use in the criminal cases against the suspects involved in the shooting incidents. In similar DEA shooting incidents, the DEA did not compel the Special Agents involved to provide administrative statements to support the prosecution of the suspects. Instead, the DEA consulted with the Assistant United States Attorney, and the DEA Special Agents testified before the federal grand jury considering the charges against the suspects.

In the DEA and FBI cases, the suspects were charged under 18 U.S.C. § 115 with assaulting the Special Agents. Although the assaults may be related to shooting incidents, both the FBI and the DEA policies require that the charges of assaulting the LEOs must be investigated separately, not as a part of the shooting incident administrative investigation. The four FBI case files included no explanation for the shooting incident investigators' decision not to conduct a separate criminal investigation and to rely instead on compelled administrative statements.

Timeliness of shooting incident investigations. All of the components have established timeliness standards for their shooting incident investigations. We found that all components began their administrative investigations while the criminal investigations were in progress, and all components required that the criminal investigations be completed before the administrative investigations were closed. The FBI requires that an administrative investigation be completed within two weeks of the incident, and the ATF, the DEA, and the USMS require completion within 30 days. All of the components allow for extensions of the time required to complete the administrative investigation.

During our review, we found that the components documented the "completion" of administrative investigations differently. The ATF, the DEA, and the USMS Inspections Divisions prepared formal memoranda to document the completion of the administrative investigation and to forward the case to the Review Board. We used the dates of those memoranda as the dates that the investigations were completed. The FBI completed a Form FD-263, Report of Administrative Investigation, but did not submit the administrative investigation to the FBI Review Board at that time. We found no documentation in the FBI's files as to when the administrative investigations were actually submitted to the FBI Review Board. The case files we reviewed did not contain sufficient documentation for us to determine when, or for what reasons, extensions were granted.

We determined that the completion dates recorded on the FBI's Reports of Administrative Investigation could not have been the actual dates that the FBI completed the administrative investigations. On the date these forms were completed, most criminal investigations had not been completed and the CRD declinations of prosecution had not been obtained, both of which are required before an administrative investigation can be closed. We examined the time taken by the FBI to obtain declinations of prosecution from the CRD and found that it took the FBI an average of 117 days to obtain the declinations required to close the criminal investigations. Therefore, the completion dates recorded on the FBI's Reports of Administrative Investigation appeared to represent only the completion of the on-site investigations.

We compared the time required by all the components to obtain declinations of prosecution to the time required to complete an administrative investigation (Figure 11).

Figure 11: Comparison of Average Time to Obtain
Declination of Prosecution to Average Time to Complete
and Submit Administrative Investigation

Source: OIG analysis of components' shooting incident data. Note: The chart shows the time required by the ATF, the DEA, and the USMS to obtain either a local or federal declination of prosecution and the time required by the FBI to obtain a declination of prosecution from the CRD.

As Figure 11 shows, the ATF, the DEA, and the USMS took longer to close their investigations than they took to obtain declinations of prosecution, which was the expected result. However, the FBI records indicated that the FBI completed a Form FD-263, Report of Administrative Investigation, in far less time than it took to obtain declinations of prosecution. Specifically, it took the FBI an average of 117 days to obtain the declination of prosecution required to close a criminal investigation, but only 25 days, on average, to record the administrative investigation as complete. Because our review of the FBI's case files found no other documentation that recorded the completion of the overall investigation, we used the FBI's average time to obtain declinations of prosecution as the average time in which FBI administrative investigations were completed and ready for Review Board consideration.

Overall, we found that all of the components exceeded their established time frames for completing administrative investigations. We recognize that the amount of time taken to complete any particular investigation may have been reasonable based on the complexity of the incident, pending criminal charges, or other factors outside the control of the components. However, the files we reviewed generally did not contain sufficient documentation to enable us to conduct a detailed analysis of these factors. Moreover, for the FBI's cases, the files did not contain sufficient documentation to enable us to accurately determine when the administrative investigations were actually completed and ready for review.

The Components' Review Boards

Resolution 13 requires the Review Boards to "determine the reasonableness of the application of deadly force in accordance with the Department's Deadly Force Policy and the law" based on the shooting investigation and to provide senior management with analyses, observations, and recommendations concerning operational training and discipline. Resolution 13 also requires that Review Boards be independent and objective, and their decisions and recommendations free of the control or direction of component management.

Although each of the components had established a Review Board to consider shooting incident investigations, we found significant differences among the Boards, including the composition of the Boards, the length of time the Boards took to complete the review process, the decisions made and the process used to reach them, the extent to which decisions were documented, and whether and how the Boards referred cases for discipline.

The composition of Review Boards. Resolution 13 requires that "the investigation and review process must be overseen to ensure that ... potential conflicts of interest are avoided, including even the appearance of conflict of interest or impropriety." The composition of the components' Review Boards varied significantly, including differences in the number of Board members, the grade level and position of the members, and whether members from outside the component were included.⁴¹ Specifically:

• The seven-person ATF Review Board included five senior ATF managers and two senior managers from other law enforcement components of the Department.⁴²

• The DEA Review Board consisted of three of the highest-ranking officials in the agency - the Chief Inspector, Inspections Division (the Chairperson); the Chief of Operations (Vice-Chairperson); and the SAC of the Office of Training.

• The composition of the FBI's Review Board varied. During the period of our review it included 14 members representing various divisions and positions within the FBI (e.g., Criminal Investigative Division, National Security Division, Training Division) as well as one Department attorney from the CRD.

• The USMS Review Board had seven members, each an LEO with at least four years of law enforcement experience. Five of the members were senior management staff, one was a USMS Instructor from FLETC, and one was a nonsupervisory Deputy Marshal. The USMS Board also included an advisory representative from the USMS Office of the General Counsel. The USMS placed restrictions on members that other components did not, including:

  • Prohibiting membership to anyone disciplined for an incident involving the use of a weapon or under investigation for a disciplinary matter; and

  • Prohibiting any Review Board member who investigated a shooting incident from participating in the final decision.

In examining the composition of the components' Review Boards, we were told by Board members that including outside members could improve the independence and objectivity of Review Boards. Two former ATF Review Board members, one from the Department of the Treasury and one from the Department of Justice, believed that including representatives from peer organizations helped ensure the independence and objectivity of the Boards in which they participated.

We also concluded that the composition of the DEA's Review Board might not be consistent with the Resolution 13 requirement for an independent review. The DEA's Review Board is composed of three of the highest-ranking individuals in the agency's management structure. Their only supervisors are the DEA Administrator and Deputy Administrator. Because the members of the Review Board are at the highest levels of DEA management, their decisions are not only subject to the control or direction of component management, but may, in fact, be the decisions of the component's management. Although the DEA Administrator and Deputy Administrator can provide some independent oversight, limiting Review Board membership only to individuals who bear significant responsibility for the operations being reviewed can create an inherent conflict with the need for the Board to independently examine and potentially criticize those operations. In response to our observation, DEA Review Board members asserted that the more extensive training and experience of their Review Board outweighed any theoretical reduction in the Review Board's independence.

Timeliness of reviews by Review Boards. Resolution 13 states that prompt reporting, investigation, and review of shooting incidents are important, although it does not establish time standards for Review Boards' consideration of cases. To evaluate the timeliness of reviews, we examined the average time it took for each component's Board to meet after the completion of the investigation. The average time for Boards to meet to consider completed investigations ranged from 39 days at the ATF to 226 days at the DEA (Figure 12).

Figure 12: Average Number of Days to Convene
the Review Board After Completing the
Administrative Invesigation

Source: OIG analysis of components' shooting incident data.

In examining the reasons for the differences in the time that the components' Review Boards took to convene after investigations were completed, we found that the ATF required its Review Board to meet within a specified time (60 days after the completion of a shooting incident investigation). The other components did not require their Boards to meet within a specified time. During our review, the USMS Board decided to begin meeting on a quarterly basis beginning on March 1, 2004.⁴³ DEA Inspections Division staff told us that the lengthy delays in convening the DEA Review Board were due to difficulties in coordinating the full schedules of its three high-ranking members.

Lengthy review delays had two negative effects. First, recommendations to senior management regarding operational, training, and safety issues were delayed, which hindered management's ability to make prompt corrections. Second, lengthy delays increased the time that employees remained under investigation. For example, the USMS Chief Inspector told us that the delays had a negative effect on the careers of Deputy Marshals involved in shooting incidents, even those ultimately cleared of any misconduct, because their promotions and transfers were delayed until the reviews were complete.

During our examination of Review Board actions, we also found one DEA incident that was never considered by the DEA Review Board. During a multijurisdictional operation, two LEOs (a DEA Special Agent and a local detective) fatally shot two suspects because, according to the LEOs, they believed that the suspects were accelerating toward them in a car. However, evidence from the scene conflicted with the LEOs' account. The car was in reverse, had backed up against another vehicle, and could not have been accelerating toward the LEOs. The DEA OPR and the CRD investigated this case extensively, and the DEA Administrator directed a special review of the tactics used in this and similar cases. The DEA OPR investigation cleared the DEA Special Agent of misconduct, but the case was never forwarded to the DEA Review Board for consideration as a shooting incident.

Components' Review Boards apply the standard for the reasonable use of deadly force differently. Resolution 13 requires components' Review Boards to determine whether the application of deadly force was reasonable.⁴⁴ Of the 121 shooting incidents that were considered by Review Boards, 14 were determined to be unreasonable uses of force (Table 8).⁴⁵ Of those 14 determinations, 11 were made in cases in which the discharge was unintentional. In only three cases, all relating to USMS Deputy Marshals, did a Review Board find that an intentional discharge was unreasonable.

Table 8: Review Board Decisions
Component	Unreasonable	Reasonable	Total
ATF	0	23	23
DEA	7	30	37
FBI	4	35	39
USMS	3	19	22
TOTAL	14	107	121

Source: OIG analysis of components' shooting incidents.

We found that when considering whether the actions of LEOs were reasonable, the ATF, the DEA, and the FBI Review Boards focused on the moment the LEOs decided to discharge their firearms. In contrast, the USMS Review Board took into account the total circumstances preceding the incident. These different approaches can lead to different conclusions about similar sets of facts. For example, the ATF, the DEA, and the FBI approach would find "reasonable" the actions of an LEO who failed to properly identify a suspect and consequently shot an innocent civilian if, at the moment the LEO fired, he or she believed that the civilian was the suspect and was acting in a way that the LEO believed was threatening. The USMS's approach could find the same actions to be "unreasonable" because the LEO had not taken steps to properly identify the individual. Although no incidents are exactly alike, several incidents we reviewed demonstrated that the components were applying the standard for the use of deadly force differently. In each incident, although we found no evidence that the findings we reviewed were inconsistent with the review procedure established by each component, the Review Boards reached different conclusions on the reasonableness of LEOs' actions. For example:

Incidents Involving Fleeing Suspects

A Deputy Marshal fired into a fleeing fugitive's car as the fugitive drove away from the Deputy Marshal after a failed arrest attempt. The Review Board found that the Deputy Marshal's use of deadly force was unreasonable, and the Deputy Marshal was suspended for five days.

A DEA Special Agent fired into a pickup truck driving away from him after shots were fired in the general direction of DEA Special Agents during a surveillance operation. The Review Board found that the use of deadly force was reasonable, and no discipline was imposed.

Incidents of Mistaken Identity

A Deputy Marshal shot and wounded a civilian mistakenly identified as a suspect. The civilian was in a vehicle that matched the description provided by an informant and arrived at the location and time that the informant had arranged. The Deputy Marshal and local police on the scene believed that the person was the suspect. The Deputy Marshal stated that, as the vehicle door opened, he saw the individual reach across the steering wheel with a gun, and he fired because he perceived an imminent threat to his safety. The Deputy Marshal did not properly identify the individual in the vehicle and rushed the vehicle rather than wait for assistance. The USMS Review Board determined that the Deputy Marshal did not have a reasonable belief that the individual posed an imminent danger of death or serious physical injury.

A FBI Special Agent shot and wounded a civilian mistakenly identified as a suspect. The civilian was in a vehicle that matched the description provided by an informant and arrived at the location and time that the informant had arranged. The Special Agent believed that the person was the suspect, and other Special Agents at the scene were communicating over the radio that they shared a similar belief. The Special Agent stated that, upon approaching the vehicle, he believed that he saw the individual reaching into his waistband for a weapon, and he fired his weapon because he perceived an imminent threat to his safety. Although the administrative investigation stated that no exigency was identified justifying the rush to physically extract the occupants, the FBI Review Board determined that the Special Agent's use of deadly force was reasonable.

Documentation of findings and recommendations. Resolution 13 requires Review Boards to provide component management with their "analyses, observations, and recommendations concerning operational, training, and other relevant issues." All of the Review Boards prepared memoranda for senior management summarizing their findings on each case. In almost all of the cases that we reviewed, the memoranda prepared by the FBI and the USMS Review Boards were thorough and provided useful recommendations. For instance:

• FBI Board memoranda for some cases noted that Special Agents did not wear body armor, created dangerous one-on-one arrest scenarios, had not attended deadly force instruction training sessions, had not qualified with Bureau-issued weapons, placed a loaded firearm near a handcuffed prisoner, did not display clear law enforcement identifiers, or did not use shields.

• In November 2002, the USMS Board observed that personnel frequently attempted to arrest fugitives while the fugitives were in their vehicles. Attempts to block the vehicles were often unsuccessful and could lead to the discharge of firearms. The Review Board recommended that the USMS conduct research to identify devices that could be used to immobilize vehicles, preventing the fugitive from fleeing and possibly preventing serious injury to law enforcement personnel and others.

In contrast, the ATF Review Board's memoranda provided a written summary of the shooting incident and Review Board discussion, but almost no analysis of the incident and few training recommendations for senior management. Two of the 13 ATF Review Board memoranda included recommendations for training improvements.⁴⁶

The DEA Review Board memoranda contained standard language describing their findings that the facts and circumstances were accurately and completely reported in the investigation, that the employee was acting within the scope of employment and authority, that the action taken by the employee was in compliance with component policies and procedures, that there was no evidence of employee misconduct or malfeasance, that the use of force was justified, and that the use of force did or did not violate the law. The DEA required Board members to complete and sign a two-page worksheet indicating that their case review covered all the above factors. We found that 32 of 37 memoranda consisted only of the standard language contained in the worksheet and a brief summary of the incident.

Referral of cases for discipline. Resolution 13 requires Review Boards to make "appropriate, timely recommendations to senior management … including, if necessary, referral to appropriate entities for disciplinary review." The ATF Review Board did not refer any of the cases that we reviewed for discipline because it did not find any shooting cases unreasonable. The DEA Board referred eight cases, the FBI Board referred four cases, and the USMS referred three cases (Table 9).

Table 9: Review Board Discipline Referrals
Component	Reason for Review Board Referral
ATF	No referrals.
DEA	Firing into a vehicle containing a passenger and in the proximity of law enforcement. Change in statement regarding the shooting Use of an unauthorized weapon. Five unintentional discharges during enforcement operations.
FBI	Four unintentional discharges during enforcement operations.
USMS	Firing at a vehicle's tires. Shooting the wrong suspect. No reasonable belief of imminent danger. Firing at a vehicle that was driving away. No reasonable belief of imminent danger.
Total Referrals:	15

Source: OIG analysis of components' shooting incidents.

The referral for discipline was not always clearly stated in the memorandum of the Review Board's finding. For example, a Deputy Marshal fired one round at a suspect's car to disable it and, when this failed, fired again at the driver. The Review Board found that the round intended to disable the vehicle was unauthorized, but that the rounds fired at the driver were authorized. The Chairman of the Board summarized these findings in a memorandum to the Deputy Director dated June 18, 2003. In a letter dated August 12, 2003, the Chairman of the Board informed the Deputy Marshal that the Review Board recommended that he "review the USMS shooting policy … and undergo 16 hours of remedial [firearms] training." It was unclear from the files we reviewed whether the case was referred for discipline, and the Deputy Marshal's personnel record did not contain any disciplinary action.

When a case is referred for discipline, the components follow their disciplinary adjudication processes. For the 15 cases in which the Review Boards made a referral for discipline, the components took the following actions:

• DEA. The DEA imposed no discipline in any of the eight referred cases. The DEA cleared three of the Special Agents and issued Letters of Caution to the other five Special Agents.⁴⁷

• FBI. The FBI followed the Board's recommendation and suspended two Special Agents for three days without pay, but increased the suspensions it imposed on the two other Special Agents from the recommended three days to five days.⁴⁸

• USMS. The USMS imposed discipline in two of the three cases. In the first, a Deputy Marshal was demoted and suspended without pay for 30 days. The individual resigned before the penalty was imposed. The second Deputy Marshal was suspended for five days without pay. The USMS was the only component to impose discipline on an employee for using deadly force against an innocent person or under other unreasonable circumstances.

We also found that the process by which cases were referred for discipline affected the actual discipline imposed by the components. For example, because of its policy of recommending a minimum 3-day suspension for all unintentional discharges, the FBI Review Board recommended a 3-day suspension (increased to a 5-day suspension by the deciding official) for a Special Agent who unintentionally discharged one round into a dying suspect. The other components' Review Boards are not required to recommend a minimum disciplinary action. Consequently, in a similar case in which a DEA Special Agent unintentionally discharged her firearm into a suspect and another DEA Special Agent, wounding both, the Review Board referred the shooting incident to the Board of Professional Conduct without a specific recommendation and the Special Agent received a Letter of Caution.

Overall timeliness. Under Resolution 13, Review Boards are supposed to provide a check and balance on the reporting and investigative process and ensure that the process is completed expeditiously. We reviewed the files on 100 incidents to determine how long the shooting review process took. We used the date of the shooting incident as the start date and the date that a signed letter was sent to the subject LEO with the Review Board's decision as the date the review was closed. We found that the ATF and the FBI took the least amount of time on average to complete their reviews, while the DEA took the longest (Figure 13).

Figure 13: Average Number of Days to Close Shooting Review Process

Source: OIG analysis of components' shooting incidents.

Conclusion. We concluded that the components' differences in implementation of the Review Board process could lead to inconsistent determinations and different disciplinary actions for similar shooting incidents. That outcome undermines Resolution 13's objective of "appropriate, consistent operational guidelines for the criminal investigative agencies of the Department." Although the weaknesses we identified need to be corrected to ensure the effectiveness of the Review Board process, each component's system had strengths that the other Review Boards could consider as benchmarks for improvement and that the Department could use to improve the shooting review process overall.

The most significant strength we noted was the inclusion of LEOs from other components as members of the Review Boards. Including representatives from peer organizations broadens the knowledge and experience applied to any shooting incident review and can help to ensure the independence and objectivity of the Review Boards.

The most significant weakness we noted was the components' failure to uniformly apply the standard for determining the reasonableness of the use of deadly force. The ATF, the DEA, and the FBI looked at the "reasonableness" of the LEO's belief that the suspect posed an imminent threat at the moment deadly force was used. The USMS Review Board considered the reasonableness of the Deputy Marshal's actions as a whole, including the actions that created the necessity for deadly force. A main purpose for reviewing shooting incidents is to ensure that LEOs use deadly force only when necessary to protect themselves and the public. Based on Resolution 13, we believe that Review Boards should uniformly apply the Departmentwide standard to determine the reasonableness of the use of deadly force.

We also concluded that uniform requirements are needed to ensure that Review Boards meet regularly to consider shooting incident investigations promptly, that they fully document their decisions, and that they provide the expected recommendations for improving training and operational procedures.

Sharing of Lessons Learned

Resolution 13 requires that "operational, safety, training or other relevant issues disclosed during the investigative or review process should be promptly communicated to component employees, and must be incorporated in policy manuals and training curriculae, as appropriate." Resolution 13 also requires that components conduct meaningful shooting data and trend analyses.

During site visits and interviews, we asked senior component officials how they communicate operational, safety, and training issues to LEOs. We found that the ATF prepared semiannual summaries of its shooting incidents that it made available to all employees on the ATF intranet. The DEA prepared annual summaries of its shooting incidents that were available to employees on the DEA intranet. We found that the FBI had not prepared summaries of its shooting incidents but began to prepare them during the course of our review, an effort that was ongoing when we completed our fieldwork. The USMS did not prepare any shooting incident summaries before we began our review. In March 2004, the USMS Review Board suggested to the Director that the USMS prepare shooting incident summaries, stating:

We further suggest, in the same spirit of making information available to the Service that at the end of each Board session, a summary of all cases could be created, removing any specific individual names or districts. These could then be emailed or otherwise sent to all USMS operational personnel advising them of the facts and findings of each of the cases by the Training Academy representative. This summary could be used for educational, training and compliance purposes.

The USMS liaison told us that the USMS is implementing the Review Board's recommendation.

Incorporation of shooting incidents in training curricula. Although all of the components' training directors and supervisory firearms instructors we interviewed said that their headquarters would notify them of any safety issues disclosed during the investigative or review process and that they would promptly incorporate safety issues in policy manuals and training curricula, we found that only the DEA sent shooting incident files from its Review Board directly to its training academy for operational and training analysis. After completing the analysis, the DEA incorporated its findings directly into the training curriculum for new Special Agents. In contrast:

• The ATF training staff at FLETC did not receive the shooting incident files for review and, during our interview, was unaware that shooting incident summaries were available on the ATF's intranet;

• The FBI training academy only retained preliminary reports of shooting incidents; and

• The USMS supervisory firearms instructor at FLETC, who was a member of the Review Board, was prohibited from discussing shooting incident reviews with anyone not a member of the Review Board.

Consequently, the ATF, the FBI, and the USMS did not incorporate lessons learned from shooting incidents in their training curricula. According to one USMS trainer, the lack of a mechanism to incorporate lessons learned from shooting incidents into the training provided at the Academy meant that Deputy Marshals must rely on "word of mouth" for lessons learned from shooting incidents and that important details may be lost or inaccurately conveyed. Resolution 13 requires components to use statistical techniques to describe, summarize, and compare shooting incident data and to identify patterns and changes that have occurred over time in order to minimize risks to LEOs and public safety. However, the Resolution does not specifically require components to share information with one another, and we found no evidence that the components shared shooting incident data among themselves.

We believe that components could benefit from a review of other components' shooting incidents. Only a small number of shooting incidents occurred in each component yearly. During the period we reviewed, the annual number of shooting cases involving the use of deadly force never exceeded 16 in any component. However, across the Department, the number of shooting incidents was higher. Aggregating the data on all the Department's shooting incidents would provide enough data to allow the components to use statistical techniques to describe, summarize, and compare shooting incident data and to identify long-term patterns and changes that have occurred over time.

Sharing information among components could enable them to identify significant trends earlier. For instance, we compiled the shooting incident data reported for all components over the last four years and found that more than half involved vehicles (Table 10).

Table 10: Shooting Cases Involving Vehicles
Component	Total Cases	Involved Suspect in Vehicle	Percentage
ATF	23	9	39%
DEA	40	25	63%
FBI	39	20	51%
USMS	22	15	68%
TOTAL	124	69	56%

Source: OIG analysis of components' shooting incidents.

Because they did not share data with each other, the components learned each lesson independently. For example, in November 2002 the USMS Review Board observed that Deputy Marshals were frequently involved in shooting incidents while trying to arrest fugitives in stopped vehicles and recommended that the USMS research "devices that could be used to immobilize vehicles." During the same time, the DEA identified vehicles as a principal factor in shooting incidents and conducted research on vehicle containment. Had the components shared lessons learned, the DEA could have informed the USMS that its research on vehicle containment concluded "that tire spike strips and an intentional puncture of the subject vehicle's tire(s) for the purpose of preventing a subject from fleeing proved ineffective and unreliable."⁴⁹ The DEA could also have shared the "active vehicle containment technique" that the DEA developed in response to shooting incidents to make it more difficult for suspects to use their vehicles as weapons against LEOs. We found there was no Departmentwide system to disseminate information on shooting incidents. Sharing information on shooting incidents among the components as soon as they are reported would allow the components to improve their operational planning and tactics immediately and avoid mistakes made by other components. All the individual LEOs we interviewed told us that knowing about shooting incidents involving other LEOs could improve the planning of enforcement operations.

CONCLUSIONS

We noted several areas in the components' shooting review processes that differed significantly. We believe that the results of our review should be examined carefully by the components to identify areas that could be improved and to ensure that shooting incidents are reported promptly, investigated thoroughly, and reviewed by an objective and independent Review Board.

All of the components require a written report within one day so that senior management can make investigative decisions, but, on average, only the ATF and the FBI consistently met the requirement. Further, the FBI and the DEA are required to report shooting incidents involving injury and death to the CRD, and all the components are required to report shooting incidents to the OIG, but neither the CRD nor the OIG were informed of all reportable incidents.

Three of the components - the ATF, the DEA, and the USMS - rely on local law enforcement to conduct the criminal investigations of shooting incidents, but the FBI conducts all its own criminal investigations. Investigators assigned by the components' headquarters conducted the administrative investigation of every ATF and USMS shooting incident, but the DEA and the FBI delegated the administrative investigation of some shooting incidents to the field office to which the LEO involved was assigned.

While each of the components' Review Boards prepares a memorandum for every shooting incident reviewed, we found that only those prepared by the FBI and USMS Review Boards consistently included analysis and recommendations specific to the incident being reviewed. We also found that each component has different Review Board membership requirements, ranging from only senior-level managers to outside law enforcement to nonsupervisory personnel. Outside representation on Review Boards can improve objectivity and independence, and reduce inconsistencies among the components. The ATF approach of including experienced LEOs from peer law enforcement agencies appeared to enhance the independence and objectivity of its Review Board.

The most important difference in the components' review of shooting incidents was the lack of uniform application of the standard for determining the reasonableness of the use of deadly force. The ATF, the DEA, and the FBI Review Boards looked at the reasonableness of the LEO's belief that the suspect posed an imminent threat at the moment deadly force was used. The USMS Review Board considered the reasonableness of a Deputy Marshal's actions as a whole, including the actions that created the necessity for deadly force. As a result, Review Boards made different decisions regarding the reasonableness of the use of deadly force for similar shooting incidents. We believe that Review Boards should apply a uniform Departmentwide standard for determining the reasonableness of the use of deadly force.

Overall, we found that the components varied substantially in the time they took to complete the administrative investigation of shooting incidents. The ATF averaged 176 days; the DEA, 440; the FBI, 184; and the USMS, 262. Much of this difference appeared to be due to the time it took for each component's Review Board to meet after the completion of the investigations. The average times ranged from 39 days at the ATF to 226 days at the DEA. Only the ATF required its Review Board to meet within a specified time period (60 days after the completion of a shooting incident investigation). The other components did not require their Boards to meet within a specified time.

Although the areas we identified need to be corrected to ensure the effectiveness of the shooting review process, each component's system had strengths that the other components and the Department could use as benchmarks to improve the shooting review process. Moreover, better sharing and analysis of information on shooting incidents could identify improvements to operational procedures and training.

RECOMMENDATIONS

To better ensure timely, thorough, and objective reporting, investigation, and review of shooting incidents, we recommend that the Department:

1. Establish a working group to consider uniform Department standards for the components' shooting incident reviews. The standards should include requirements regarding:

   1. Content and timeliness of preliminary shooting incident reports and investigations.

   2. Delegation of investigative responsibility, involvement of local authorities in criminal investigations of shooting incidents, and requirements for declination of prosecution.

   3. Composition of the Review Boards, application of the standard for determining the reasonableness of deadly force, and documentation of Review Board decisions.

   4. Aggregation within the Department of shooting incident data and lessons learned, including the preparation and distribution of shooting incident summaries.

   With regard to specific component practices, we recommend that the components take the following actions to correct the specific weaknesses identified in this report.

2. ATF Recommendations - We recommend that the ATF:

   1. Establish a formal reporting relationship with the CRD.

   2. Improve the documentation of Review Board findings and recommendations.

3. DEA Recommendations - We recommend that the DEA:

   1. Improve the consistency and timeliness of shooting incident reports.

   2. Ensure compliance with its CRD reporting agreements.

   3. Consider changing the composition of its Review Board and including outside members on its Review Board.

   4. Improve documentation of Review Board findings and recommendations.

   5. Improve the timeliness of review of shooting incidents.

4. FBI Recommendations - We recommend that the FBI:

   1. Ensure compliance with its CRD reporting agreements.

   2. Establish specific criteria for when to delegate shooting incident investigations to field offices.

   3. Consider requesting local criminal investigation of shooting incidents and avoid duplication of local criminal investigations.

5. USMS Recommendations - We recommend that the USMS:

   1. Streamline its shooting incident reporting system to improve consistency and timeliness in reporting shooting incidents.

   2. Establish a formal reporting relationship with the CRD.

   3. Consider including outside members on its Review Board.

31. The ATF and the DEA require a written report within 12 hours of the incident. The USMS requires a written report within 24 hours. The FBI requires an immediate written report by electronic communication.

32. The field office has ten business days to report incidents classified as unintentional if no injury or death is involved. In the three cases subject to the 10-day rule, there was no initial report in one case, one report was timely, and one was late.

33. See DEA - Cover letter from Section Chief Criminal Section, CRD, to Deputy Chief Inspector, DEA OPR, September 12, 2000, and attachment: Procedures for referral and investigation of DEA incidents suggesting possible criminal civil rights violations. The DEA agreement also lists incidents other than shootings that must be reported. See also FBI - Letter from Section Chief Criminal Section, CRD, to Deputy Assistant Director of Inspections Division, FBI, June 26, 2000.

34. Order 2492 gave the OIG the authority to investigate criminal wrongdoing and serious administrative misconduct in the DEA and the FBI on July 11, 2001. The ATF has been required to report shooting incidents to the OIG since it was transferred to the Department on January 24, 2003. The USMS has been required to report shooting incidents to the OIG since July 1, 1998.

35. The USMS included copies of formal reports and recorded the delegation of investigative authority received from the OIG in its investigative files. Because the USMS recorded the date it reported to the OIG, we were able to verify that the report date listed by the USMS was within one business day of the date the OIG reported being informed of the case. The ATF's Inspections Division established the practice of reporting all shooting incidents to the Treasury OIG and continued that practice when the ATF moved to the Department on January 24, 2003.

36. We asked Shooting Incident Investigators from the ATF, the DEA, and the USMS why these 23 shooting incidents were not investigated by state or local law enforcement as required by their components' policies. They explained that state or local law enforcement authorities sometimes declined to investigate shooting incidents that did not involve injury or death, or that occurred on federal property.

37. FBI, Crime in the United States, Summary of Uniform Crime Reports, 2002.

38. Local law enforcement officers may be present at the scene of a shooting incident as part of the task force or ad hoc team assembled to conduct the operation or because they were the first responders to the scene. We found indications that local authorities were present in 29 of the 39 FBI cases we reviewed.

39. Civil rights violations are federal offenses punishable under 18 U.S.C. § 242.

40. However, an employee who makes a false statement in the compelled interview may be criminally prosecuted for the false statement.

41. Resolution 13 does not require outside membership on the Review Board.

42. The ATF began including outside representatives while it was part of the Department of the Treasury and continued this practice after transferring to the Department.

43. Prior to our review, the USMS Review Board met only when the Chairperson believed there were a sufficient number of cases to review.

44. The Review Boards applied the standard for reasonableness in the use of deadly force established in the Department's Resolution 14, approved by the Attorney General on October 17, 1995, which states that an LEO "may use deadly force only when necessary, that is, when the officer has a reasonable belief that the subject of such force poses an imminent danger of death or serious physical injury to the officer or to another person." The Department issued a revised deadly force policy on July 1, 2004, but did not change the standard for the use of deadly force.

45. Resolution 13 is specific in requiring Review Boards to decide whether a use of force was reasonable, but none of the Review Boards used that terminology. The ATF, the DEA, and the FBI Review Boards used "justified" and "not justified" to describe their determinations. The USMS Review Board used "authorized" and "not authorized." For consistency, we use the terms "reasonable" and "unreasonable" to describe Review Board determinations.

46. ATF Order 0 8200.3 in effect from April 22, 1996, until November 3, 2000, and superseding ATF Order 0 8200.3A, which took effect on November 3, 2000, and is still in effect requires that the Review Board's report include findings "relative to the following issues: use of force, compliance with [ATF] policies and directives, and determinations of wrongful or inappropriate actions of those involved in the shooting incident. The [Review Board] will also make recommendations relative to needed changes to policy, directives, equipment, training, supervision, and safety issues identified in the [report]."

47. A Letter of Caution is not a formal disciplinary action and does not become part of the employee's personnel record.

48. The FBI Review Board recommendations were based on guidance outlined in an August 27, 1997, memorandum to all SACs, "Accidental Discharges of Weapons," in which the Director stated that accidental discharges of a firearm as a result of disregarding established safety procedures would result in a minimum penalty of a 3-calendar-day suspension, absent definitive mitigation. The FBI Review Board recommended 3-day suspensions for unintentional discharges, regardless of whether the discharge occurred at the range, in the office, or during enforcement operations.

49. DEA Deputy Administrator Memorandum, "Vehicle Containment and Vehicle-Involved Arrest Procedures," dated August 22, 2002.