The Bureau of Alcohol, Tobacco, Firearms and Explosives’
National Integrated Ballistic Information Network Program

Audit Report 05-30
June 2005
Office of the Inspector General


Findings and Recommendations


    1. NIBIN CAPABILITY TO COMPARE BALLISTIC IMAGES ON A NATIONAL LEVEL

    We determined that the ATF completed deployment of the IBIS equipment during FY 2003 to all 231 sites contained in its deployment plan. However, some of the equipment was not sent to agencies that could best utilize it. Our analyses found that 37 agencies that did not receive IBIS equipment had submitted more firearms data for entry into NIBIN than some agencies that received the IBIS equipment. This analysis indicates a need for the ATF to determine whether additional IBIS equipment should be purchased and deployed to these high-usage agencies, or whether IBIS equipment should be redistributed from low- to high-usage agencies. We also determined that NIBIN is capable of comparing ballistic images on a national level and that most NIBIN users are aware of this capability. We identified seven agencies that could benefit when performing nationwide searches with additional guidance, training, or assistance from the ATF.

Deploying IBIS Equipment

To evaluate the ATF’s deployment of IBIS equipment to participating law enforcement agencies, we: (1) interviewed NIBIN program officials at the ATF, (2) obtained and reviewed the ATF’s deployment plan for NIBIN, (3) reviewed the methodology used by the ATF to select participants for the program, (4) compared the current status of deployment to the deployment schedule, and (5) obtained and analyzed firearms data entered into NIBIN for both partner agencies and non-partner agencies.

In a June 7, 2000, memorandum, the NIBIN Executive Board established a deployment plan with a tentative 24-month schedule for delivering the IBIS equipment during FYs 2001 and 2002. NIBIN sites were selected to receive equipment based on such factors as population, rate of violent crime, and demonstration of commitment to ballistic technology through the past use of IBIS or DRUGFIRE equipment. Through this preliminary deployment plan, site surveys were conducted at each agency scheduled to receive equipment to ensure that the type of equipment sent matched the needs and capabilities of the receiving agency. The ATF’s NIBIN contractor conducted site visits and met with upper management from the partner agencies. The contractor used a site survey to obtain information to deliver and install the IBIS equipment. The contractor also discussed the responsibilities of each agency and provided a copy of the MOU that was to be executed between the ATF and each partner agency. The contractor also provided each partner agency with the technical requirements that the facility needed to meet before the IBIS equipment could be provided. The contractor coordinated with local ATF personnel and each partner agency’s staff on the details of deploying the equipment and coordinating the necessary training. Finally, NIBIN officials told us that they committed to placing at least one site in each state to ensure nationwide coverage.

We determined that much of the equipment was delivered before the deployment plan, methodology, and budget were established. Specifically, according to the ATF, the IBIS equipment was sent to 65 of the 231 NIBIN sites before FY 2001.30 Candidates were evaluated on a case-by-case basis and, as funding became available, the equipment was delivered.

In addition, the IBIS equipment was sent to 19 of the 231 NIBIN sites after FY 2002. An ATF official explained that most of the equipment was sent prior to development of the deployment plan because the ATF’s focus was on moving the equipment as quickly as possible. The official attributed the delay in completing full deployment by the end of FY 2002 to: (1) delays in receiving the congressional appropriation for FY 2002, (2) problems during equipment installation, and (3) the transition of the FBI’s responsibilities for the network to the ATF. Despite the delays, to date the ATF has deployed the IBIS equipment to all agencies for which delivery was planned.

While the IBIS equipment has been fully deployed, we determined that it was not sent to agencies that could best utilize it. We obtained a copy of the NIBIN database as of October 22, 2004, and found that 196 partner agencies had entered 888,447 pieces of firearms data into the system.31 We also determined that at least 7,653 law enforcement agencies that had received Originating Agency Reporting Identifier (ORI) numbers from the FBI contributed the 888,447 data items.32 We analyzed the entries and found that many of the partner agencies that received the IBIS equipment had entered only minimal firearms data into NIBIN, whereas some non-partner agencies that did not receive the IBIS equipment had submitted much more firearms data for entry into the system. Our analysis showed that:

  • 196 partner agencies had entered 888,447 records of firearms data into NIBIN;
  • the top 30 (15 percent) partner agencies entered 608,280 (68 percent) records of firearms data;
  • 71 (36 percent) partner agencies entered fewer than 1,000 total records of firearms data, and 4 of those 71 partner agencies entered fewer than 100 total records of firearms data;
  • 37 non-partner agencies submitted a substantial number of firearms records, ranging from 1,491 to 39,200; and
  • 7 of the top 20 contributors of data were non-partner agencies.

As the data indicates, some partner agencies that received IBIS equipment contributed very little evidence to NIBIN. Conversely, some non-partner agencies that did not have the IBIS equipment submitted considerable evidence through a partner agency. Accordingly, the ATF needs to determine whether additional IBIS equipment should be purchased and deployed to these high-usage agencies, or whether IBIS equipment should be redistributed from low-usage agencies to high-usage agencies. The NIBIN Program Director told us that the ATF’s initial focus was to deploy the equipment as quickly as possible and that not enough attention was directed towards ensuring the equipment was sent to sites that could best utilize it. Around 1999, the ATF began monitoring the partner agencies, but it was a simplistic measurement of the monthly usage of each partner. The NIBIN Program Director told us that no in-depth analyses or studies were done to compare participation of partner agencies to non-partner agencies. The official also stated that the ATF is considering redistributing equipment from low-usage agencies to high-usage agencies because they can better utilize the equipment.

Comparing Images

We interviewed ATF officials to see if IBIS was capable of performing nationwide comparisons of ballistic images associated with crime firearms. Nationwide comparisons can help law enforcement officials link crimes across jurisdictions when the same firearm is used to commit crimes in multiple jurisdictions. We also visited 22 of the 196 NIBIN partner agencies that contributed evidence to NIBIN as of October 22, 2004, and sent survey questionnaires to the remaining 174 partner agencies that contributed evidence to NIBIN to understand how each uses the nationwide comparison.

We determined that prior to November 2003, NIBIN tracked and compared ballistic images either locally or regionally. The NIBIN system is made up of servers located in Ammendale, MD; Atlanta, GA; and Walnut Creek, CA, connected to the 12 NIBIN regions as shown in Appendix VI. Most of the regions are further divided into sub-regions covering NIBIN partner agencies within the same region as shown in Appendix VII . For example, the Walnut Creek, CA, server is connected to four regions, one of which is Region 1a for Southern California . The Southern California region is partitioned into 3 sections covering the southern, central, and northern sections of Region 1a. The southern partition serves 2 partner agencies; the northern partition serves 3 partner agencies; and the central partition serves 10 partner agencies.

NIBIN automatically performs a local search as data is entered into the system. A local search is one where the data is compared against other data in the same partition within the region. For example, in Region 1a when a partner agency in the southern partition enters evidence into NIBIN, the evidence is automatically compared against other evidence entered by that partner agency as well as evidence entered by the other partner agency within the southern region partition of Region 1a.

Before November 2003, a regional search could also be performed by the agency entering the data, but the search was not automatic and had to be manually selected. The agency performing the search could select the entire region, or could select which agencies within the region to include and not include in the search. Nationwide searches were not possible.

In November 2003, NIBIN was enhanced to allow nationwide searches against all data in the system. The nationwide search is similar to the regional search, except that the search must be made against one server at a time, and the searching agency must first select which server to search against. The server then can choose to search against either all regions and agencies, or specific regions or agencies within the server.

According to NIBIN officials, the ATF delegates to users the authority and control over how to search the database to compare nationwide firearms evidence. Officials believe that the users are the best judges in conducting searches within the system because they know best whether the crime evidence collected might be linked to other crimes. Further, the users are more familiar with the cases and crimes in their respective areas. Officials added that, with a handful of exceptions, almost all firearms crimes occur in the vicinity where the firearms are found. Generally, users make nationwide database searches only on high-profile cases that are either national in nature or where the user has specific leads in the case.

Through site visits to 22 NIBIN partner agencies and survey questionnaires sent to the remaining 174 NIBIN partners that received RDAS equipment, we asked each agency how it performs searches. We found that the nationwide search capability was rarely used because the participating agencies rarely had a need to perform such searches or rarely received requests for such searches from agencies that submitted evidence to them for entry into NIBIN. Examples of typical responses we received from the 22 NIBIN partner agencies we visited related to how they use the search capabilities are included in Appendix XII.

We received responses from 160 of the 174 NIBIN partner agencies that were mailed surveys. For the 160 surveys we received back, 153 answered our question related to regional searches and 152 answered our question related to national searches. As shown in the following table, the survey results showed that national and regional searches are not used often by some partner agencies.

Responses from Surveyed Partner Agencies
Regarding their Use of NIBIN Searches
  Always Seldom Rarely Never Upon
Request
Regional searches 104 23 5 4 17
National searches 2 37 25 15 73
 Source: Survey Questionnaires from NIBIN Partner Agencies

The explanations regarding the regional searches for the agencies that responded “rarely” or “never” are contained in Appendix XIII. The explanations regarding the national searches for the agencies that responded “rarely” or “never” are contained in Appendix XIV.

The site visit and survey results showed that most of the agencies were aware of the regional and nationwide search capabilities but did not perform them, because the agencies either did not see a need for searching beyond their local area or had received no requests from the submitting non-partner agencies. However, we identified seven partner agencies that indicated they did not perform the regional or national searches because of:

  • unfamiliarity with, or lack of training on, how to use the system to perform the searches (Omaha Police Department Crime Laboratory; Prince George’s County (MD) Police Department; and the Idaho State Police); and
  • the time, manpower, or difficulty to perform the searches (West Virginia State Police – Charleston; Kansas Bureau of Investigation – Topeka; Massachusetts State Police – Sturbridge; and the Lake County (IN) Crime Laboratory).33

As a result, if a situation arises where a regional or national search might be warranted, these agencies may not be able to perform the searches and may have to seek assistance from the ATF, which could delay the results and affect the agencies’ investigations. The ATF needs to ensure that these agencies receive the training, guidance or assistance to perform regional or national searches from one of the ATF laboratories or the NIBIN contractor.

Conclusion

The ATF did not fully deploy the NIBIN program equipment as planned within the two-year schedule. While the ATF planned to complete deployment by the end of FY 2002, eight percent of the NIBIN partner agencies received their equipment in FY 2003. Although the system was not deployed on schedule, it was fully deployed by the time of our audit. However, the ATF was more concerned with distributing the equipment as quickly as possible instead of ensuring that it was sent to sites that could best utilize it. As a result, we found that many non-partner agencies that did not receive IBIS equipment submitted more evidence into NIBIN than partner agencies that received the IBIS equipment. Consequently, the ATF needs to determine whether additional equipment should be purchased and sent to the high-usage non-partner agencies, or whether it should be redistributed from the low- to the high-usage non-partner agencies.

We also determined that: (1) NIBIN does have the capability to perform comparisons on a nationwide basis, and (2) most of the NIBIN partner agencies were aware of the nationwide search capabilities of NIBIN. However, the nationwide search feature of NIBIN is rarely used because most participating partner agencies rarely need to perform such searches. Also, those agencies rarely receive requests for nationwide searches from non-partner agencies submitting evidence to them. We did identify a small number of partner agencies that might benefit when performing nationwide searches with additional guidance, training, or assistance from the ATF.

Recommendations:

We recommend that the ATF:

  1. Determine whether additional IBIS equipment should be purchased and deployed to high-usage non-partner agencies, or whether equipment should be redistributed from the low-usage partner agencies to high-usage non-partner agencies.
  2. Provide additional guidance, training, or assistance to the partner agencies that indicated they did not perform regional or nationwide searches because they either lacked an understanding of the process or lacked manpower to perform such searches.


    2. ENTERING BALLISTIC IMAGES INTO NIBIN

    The ATF needs to take steps to increase the number of ballistic images entered into NIBIN. Only about 20 percent of the law enforcement agencies with potential to participate in NIBIN actually contributed evidence into the system. Participation in NIBIN could be improved if the ATF: (1) better promotes the use and benefits of NIBIN to law enforcement agencies, and (2) involves the partner agencies more in promoting the NIBIN program to other law enforcement agencies in their area. We also found that participating agencies were submitting a much smaller proportion of bullets into NIBIN than cartridge casings because NIBIN did not produce good quality images of bullets. Most of the agencies stated that it was not worth the resources needed to enter bullets in the system because the poor quality images resulted in few hits. In addition, we found potential matches in NIBIN had not been reviewed by one high-volume partner agency since January 2002, and therefore potential matches were not identified and pursued. The ATF also needs to provide guidance to NIBIN partners by stressing the importance of viewing potential matches (correlations) in a timely manner and by ensuring that NIBIN accurately reflects the status of the viewed correlations.

Data Entered Into NIBIN

We requested from the ATF a copy of its NIBIN database showing all the records entered regarding discharged bullets, cartridge casings, related information records, associated case-information records, and firearms information records. The ATF provided the requested data as of October 22, 2004, through its contractor (Forensic Technology, Inc.) located in Montreal, Canada . The data was presented in multiple tables in a relational database. For more details about the tables in the NIBIN database, see the Data Analysis section of Appendix I. The database contained the following records for 196 NIBIN partner agencies:

  • 888,447 records of firearms evidence (bullets and cartridge casings) with each record containing 11 fields of information,
  • 514,731 records of cases with each record containing 8 fields of information, and
  • 254,187 records of firearms with each record containing 15 fields of information.

The law enforcement agency that entered the evidence, or submitted the evidence for entry, is identified in the database cases table by its Originating Agency Reporting Identifier (ORI) number. To determine the agencies (based on ORI number) that submitted evidence into NIBIN, we linked the NIBIN cases table to the NIBIN evidence table. However, we were unable to link the ORI numbers in the cases table for 55,193 of the 514,731 cases to evidence in the evidence table because of the following omissions or errors in the cases table.

  • 52,392 records in the cases table had “unknown” entered in the ORI field.
  • Two NIBIN partner agencies (Colorado Bureau of Investigation – Montrose and Rhode Island State Crime Laboratory) entered duplicate case ID numbers for its own cases. As a consequence, it was impossible to link the cases table to the evidence table for these agencies. A total of 2,801 records in the cases table for these two agencies contained duplicate case ID numbers. Of the total, 478 records were from the Colorado Bureau of Investigation – Montrose and 2,323 records were from the Rhode Island State Crime Laboratory.

As a result of these omissions and errors, we were unable to link the ORI numbers from the 55,193 cases in the cases table to 115,092 evidence items in the evidence table. Consequently, we were only able to link the remaining 459,538 cases in the cases table to 773,355 records of evidence in the evidence table from 194 NIBIN partner agencies. Thus, our analyses of the NIBIN data in this report are limited to the 773,355 records of evidence contributed by 194 NIBIN partner agencies.

Law Enforcement Participation in NIBIN

One factor essential to the success of the NIBIN program is the wide participation of law enforcement agencies. The ATF identified 38,717 law enforcement agencies or divisions of law enforcement agencies that received an ORI number from the FBI and that were potential candidates for participating in NIBIN. We analyzed the 773,355 evidence records that were entered into, or submitted for entry into, NIBIN as of October 22, 2004 . Of that total, we determined that only 7,653 (20 percent) of the 38,717 law enforcement agencies that had received ORI numbers had actually contributed evidence to NIBIN. These 7,653 agencies included partner and non-partner agencies. A closer examination of the evidence records revealed that the top 20 percent of law enforcement agencies (1,520) submitted 96 percent (739,959) of the 773,355 evidence records in NIBIN. The remaining agencies submitted only 4 percent (33,396) of the evidence records in NIBIN. Since the non-partner agencies submit their evidence to partner agencies for entry into NIBIN, we further analyzed the 773,355 evidence records to determine if a similar trend existed based on the partner agencies that contributed the evidence data. This analysis identified a similar trend as shown in the following table.

Contributed Records Into NIBIN
  Top 20 Percent
of Partner Agencies
Next 25 Percent
of Partner Agencies
Bottom 55 Percent
of Partner Agencies
Cases Data 75% 15% 10%
Evidence Data 75% 16% 9%
Firearms Data34 73% 17% 10%
 Source: OIG Analysis of NIBIN Database Records

The distribution of data contributed by the partner agencies is found to be highly skewed. More descriptive results of our analyses, including the number of cases, evidence, and firearms contributed, are presented in the following charts, which show columns for the amount of data entered into NIBIN by each group of partner agencies, and also show a line above the columns to represent the cumulative amount of data entered.


Distribution of NIBIN Case Records
Among Contributing Partner Agencies

Total cases entered/percent of cases entered: 20% of Partners - 386,026/75%;  25% of Partners - 78,838/15%;  55% of Partners - 49,867/10%.


Source: OIG Analysis of NIBIN
Database Cases Records

 
Distribution of NIBIN Evidence Records
Among Contributing Partner Agencies

Total evidence entered/percent of evidence entered: 20% of Partners - 666,322/75%;  25% of Partners - 138,325/15%;  55% of Partners - 83,800/10%.


Source: OIG Analysis of NIBIN
Database Evidence Records

 
Distribution of NIBIN Firearms Records
Among Contributing Partner Agencies

Total firearms entered/percent of firearms entered: 20% of Partners - 185,892/75%;  25% of Partners - 42,596/16%;  55% of Partners - 25,699/10%.


Source: OIG Analysis of NIBIN
Database Firearms Records

We sought to evaluate the extent of evidence contributed to NIBIN in comparison to the extent of potential evidence based on the FBI’s Uniform Crime Reports (UCR).35 Through the UCR, law enforcement agencies report to the FBI data regarding a variety of crimes, including firearms-related crimes of robbery and aggravated assault. Other reported crimes may involve firearms but are not reported as a firearms-specific crime. We obtained the FBI’s UCR data for 2000 through 2003, the most recent data available as of our audit. From the data, we extracted the number of firearms crimes reported each year and compared this data to the amount of evidence entered into NIBIN during the same time period.

We originally attempted to compare, for NIBIN participating agencies, the crime rate according to the UCR to the level of evidence contributed to NIBIN. However, meaningful comparisons were not possible based on the available data because of variables such as population size, population density, geographic location, and other demographic factors. Instead, to construct a rough estimate of the extent to which NIBIN contains a reasonable proportion of potential firearms evidence, we compared the nationally reported number of reported firearms crimes to the amount of evidence entered into NIBIN for calendar years 2000 through 2003. The results of our comparison are illustrated in the following chart.

Crimes Involving Firearms Reported in UCR
and Firearms Evidence Entered into NIBIN


[Not Available Electronically]


Source: FBI UCR and NIBIN Database Records

The chart shows that the gap between the firearms crimes reported and the evidence entered into NIBIN has narrowed from about 169,000 in 2000 to about 118,000 in 2003. This result was to be expected since the ATF was deploying much of the IBIS equipment during this time period. However, the chart also shows that there is potential for significant amounts of additional evidence to be entered into NIBIN each year.

As also would be expected, we found that the number of hits identified for a partner agency is closely tied to the number of cases and evidence contributed by that partner agency, as shown in the following chart.

Distribution of Hits Compared to
Cases and Evidence Entered

Percent of Hits/Cases/Evidences: 20% of Partners - 72%/75%/75%; 25% of Partners - 19%/15%/16%; 55% of Partners - 9%/10%/9%.
Source: OIG Analysis of NIBIN
Database Cases and Hits Data

The data indicates that as more available data is entered into NIBIN, the greater the chance of getting hits. Since a hit can provide a link to another crime, the hit could result in additional leads to investigators in solving a crime involving the use of a firearm. Therefore, to maximize the success of the NIBIN program, it is essential that the ATF and the partner agencies fully promote participation by more law enforcement agencies. The ATF’s efforts to promote the program are discussed in the next section of this finding.

While the aggregate of the NIBIN data shows that more hits result as more data is entered, we found that some partner agencies had very high hit rates even though the data they entered was comparatively low as illustrated in the following table.

Analysis of Hits Compared to Cases Entered
Partner Agency State Total
Cases
Entered
Hits Hit
Rate
Johnson Co. Sheriff’s Office
KS
13
5
38.5%
Allegheny Co. Coroner’s Office
PA
5,030
813
16.2%
Kansas Bureau of Investigation – Topeka
KS
308
40
13%
Monroe Co. Department of Public Safety
NY
1,045
134
12.8%
Maryland State Police
MD
1,535
193
12.6%
Massachusetts State Police – Danvers
MA
166
20
12%
Minneapolis Police Department
MN
1,331
160
12%
Long Beach Police Department
CA
1,072
110
10.3%
Santa Ana Police Department
CA
1,976
183
9.3%
West chester Co. Department of Public Safety
NY
640
54
8.4%
San Diego Police Department
CA
531
43
8.1%
Boston Police Department
MA
5,741
431
7.5%
 Source: OIG Analysis of NIBIN Database Cases and Hits Data

We contacted the 12 agencies and asked what they did to get such a high hit rate. Eight of the 11 agencies that responded told us the primary reason for the high hit rate was because they enter almost all of the recovered bullet and cartridge casing evidence into NIBIN. The remaining three agencies did not believe their hit rates were as high as the ATF’s NIBIN data indicated, but provided no documentation to dispute the ATF’s NIBIN data. The ATF should further research why these agencies are so successful and should apply the results of such research to the majority of partner agencies that are not as successful.

Program Promotion

Given the low participation by most law enforcement agencies, we discussed with ATF officials the process of promoting the NIBIN program. ATF officials recognized that for the program to be most effective, state and local law enforcement agencies must be well informed of the existence of the NIBIN program and how it can benefit them. We found that while the ATF has taken steps to promote the program, its promotion has been mostly effective for the partner agencies, and less effective for the other agencies that submit data through the partner agencies.

We found that the ATF took the following steps to promote the program.

  • The ATF staff address law enforcement groups and appear at law enforcement conferences, including the National Association of Chiefs of Police conference, of which the NIBIN program is affiliated .
  • The ATF also conducts regional conferences for NIBIN partner agencies that provide briefings on developments in the system; answer questions; and allow partner agencies to make connections and share experiences.
  • The ATF publishes program information on the Internet at the NIBIN program website (www.nibin.gov).
  • The ATF also publishes limited program information on the agency’s main website (www.atf.gov).
  • The ATF develops program publications that it makes available to interested agencies.

We asked both participating and non-participating law enforcement agencies to evaluate the ATF’s promotion of the NIBIN program and its effectiveness. Specifically, we asked 16 NIBIN partner agencies we visited about the ATF’s promotion efforts. We also sent survey questionnaires to: (1) 174 partner agencies and 411 participating non-partner agencies that had contributed evidence to NIBIN, and (2) 85 non-participating non-partner agencies.

Of the 16 partner agencies we visited, we found that 11 had mostly positive comments about the ATF’s promotion of the program while 5 had mostly negative comments. Positive comments include the following.

  • The ATF has gone to various precincts to promote the program. (Detroit, Michigan, Police Department)
  • The local ATF District Office’s contractor holds regional meetings annually where presentations are made to promote the benefits of participating in the program. Also, the ATF contractor holds the same types of meetings locally. (Boston, Massachusetts, Police Department and Rhode Island State Crime Laboratory)
  • The ATF provides program brochures to regional and state laboratories. (Houston, Texas, Police Department)
  • The ATF provides assistance or guidance in promoting the program, conducts presentations, and provides brochures and training. (Tulsa, Oklahoma, Police Department)
  • The ATF sent a spokesman to a conference, and also presented a slide show at a District Attorney conference. (Mississippi State Crime Laboratory)
  • The ATF made a video about the NIBIN program and the IBIS equipment and sent a copy to every agency here. (Indianapolis/Marion County, Indianapolis, Forensic Laboratory)

Negative comments include the following.

  • Limited assistance is provided by the ATF, and so the department is not promoting the program. (Minneapolis, Minnesota, Police Department)
  • The department does not solicit other law enforcement agencies to submit evidence, so the ATF’s promotional guidance is not used. (Denver, Colorado, Police Department)
  • The ATF does not provide assistance or guidance in promoting the program in our area because of a state mandate that requires local law enforcement agencies to work and train with the state and local law enforcement agencies in the use of the laboratories. (Tacoma, Washington, State Patrol Crime Laboratory)
  • The ATF does not provide any specific assistance or guidance in promoting the program. However, the ATF does attend and sponsor law enforcement meetings in the area. (Los Angeles, California, Police Department)

Of the 174 partner agencies we surveyed, we obtained responses from 148 regarding the ATF’s promotion of NIBIN. While 70 percent (103) of the responding agencies indicated that the ATF had provided assistance or guidance in promoting the NIBIN program, the remaining 30 percent (45) indicated that the ATF did not provide such assistance or guidance.

Of the 411 participating non-partner agencies we surveyed, we obtained responses from 228 agencies. From the responses, it appears that the ATF’s promotion of the NIBIN program has not been effective. The following table illustrates that the majority of these agencies were not aware of the following ATF initiatives to promote the NIBIN program.

Analysis of Participating Non-Partner Agency Responses
Related to the ATF’s Promotion of the NIBIN Program
ATF Initiative Aware Not Aware
Number Percent Number Percent
NIBIN Publications
59
26%
169
74%
NIBIN Pamphlets
57
25%
171
75%
Presentations at law
enforcement conferences
53
23%
175
77%
Visual aids at law
enforcement conferences
47
21%
181
79%
Communication with NIBIN users
56
25%
172
75%
Communication with ATF local
and regional representatives
90
39%
138
61%
 Source: Survey Questionnaires Returned by Participating Non-Partner Agencies

We also asked the participating non-partner agencies about their level of understanding of the NIBIN program as another way to measure the ATF’s effectiveness in promoting it. We found that about half the agencies had a high or medium level of understanding of certain aspects of the NIBIN program, while the other half had a low or no understanding, or did not answer the question, as shown in the following table.

Analysis of Participating Non-Partner Agencies’
Understanding of the NIBIN Program
NIBIN Program Issues High Medium Percent
High/
Medium
Low None No Answer Percent
Low/None/
No Answer
Ballistic imaging 39 64 45% 68 20 37 55%
Usefulness of the entry of all
projectiles, discharged cartridge
casings, and test-fired evidence
from firearms collected
46 70 51% 63 20 29 49%
Role of state and local law
enforcement agencies in the
NIBIN program
43 71 50% 64 21 29 50%
Definition of a “hit” in NIBIN 56 61 51% 58 22 31 49%
 Source: Survey Questionnaires Returned by Participating Non-Partner Agencies

For the 85 non-participating non-partner agencies surveyed, we obtained responses from 28 agencies. From the responses, it appears that the ATF’s promotion of the NIBIN program has been even less effective. The table below shows that the majority of these agencies were not aware of the following ATF initiatives to promote the NIBIN program.

Analysis of Non-Participating Agency Responses
Related to the ATF’s Promotion of the NIBIN Program
ATF Initiative Aware Not Aware
Number Percent Number Percent
NIBIN Publications
6
21%
22
79%
NIBIN Pamphlets
2
7%
26
93%
Presentations at law enforcement conferences
3
11%
25
89%
Visual aids at law enforcement conferences
3
11%
25
89%
Communication with NIBIN users
0
0%
28
100%
Communication with ATF local and
regional representatives
10
36%
18
64%
 Source: Survey Questionnaires Returned by Non-Participating Agencies

We also asked the non-participating non-partner agencies about their level of understanding of the NIBIN program as another way to measure the ATF’s effectiveness of promoting the program. We found that few agencies had a high or medium level of understanding of certain aspects of the NIBIN program, while most had a low or no understanding of the program, or did not answer the question, as shown in the table below.

Analysis of Non-Participating Non-Partner
Agencies’ Understanding of the NIBIN Program
NIBIN Program Issues High Medium Percent
High/
Medium
Low None No Answer Percent
Low/None/
No Answer
Ballistic imaging
0
3
11%
6
7
12
89%
Usefulness of the entry of all projectiles, discharged cartridge casings, and test-fired evidence from firearms collected
0
4
14%
6
6
12
86%
Role of state and local law enforcement agencies in the NIBIN program
0
2
7%
6
8
12
93%
Definition of a “hit” in NIBIN
1
2
11%
6
7
12
89%
 Source: Survey Questionnaires Returned by Non-Participating Non-Partner Agencies

In addition to the ATF’s promotion of the NIBIN program, the partner agencies also promote the program to law enforcement agencies within their area to increase participation by non-partner agencies. During our visits to the NIBIN partner agencies, we determined that while some of them encourage other law enforcement agencies to participate, many do not encourage such participation. Examples of responses from the partner agencies that encourage participation included:

  • At least three times a year, representatives attend the police chief’s monthly meetings and discuss the NIBIN program. (Rhode Island State Crime Laboratory)
  • Letters are sent to other law enforcement agencies promoting the use of the NIBIN program. (Omaha, Nebraska, Police Department Crime Laboratory)
  • Brochures and e-mails are sent to other law enforcement agencies promoting the NIBIN program. Contacts with other agencies are made in person and by phone. The IBIS equipment is co-located with the training academy, and as a result, one of the courses for the recruits is related to NIBIN. (New Mexico Department of Public Safety – Santa Fe)
  • Routine visits are conducted at rural police departments to inform them of the importance of submitting their evidence. Also, demonstrations are done on how to test-fire firearms and how to submit firearms evidence. In addition, monthly presentations are made to police departments and police chiefs about the laboratory operations, which include sections on the NIBIN program. (Tacoma, Washington, State Patrol Crime Laboratory)
  • Presentations, which include sections about the NIBIN program, are made to local police departments and police chiefs. (Erie County, New York, Forensic Laboratory)

Examples of responses from the partner agencies that did not encourage participation included:

  • We do not solicit other law enforcement agencies because the agency is too busy to take on more work, and because the county and the state already have their own equipment. (Minneapolis, Minnesota, Police Department)
  • Law enforcement agencies are not solicited to submit firearms evidence. Most of the department’s 11 police districts are serviced and very few requests are from outside of the department. (Boston, Massachusetts, Police Department)
  • Because of the caseload and because the state bureau of investigations performs firearms evidence entries for other law enforcement agencies, specific solicitations to law enforcement agencies are not performed. Entries are performed for federal agencies, however. (Denver, Colorado, Police Department)
  • As a crime laboratory, the agency does not solicit law enforcement agencies within their service area to submit firearms evidence, because it creates the feeling of animosity by encroaching on someone else’s area of responsibility. (Los Angeles, California, Police Department)

We also sent survey questionnaires to the remaining 174 partner agencies that contributed evidence to NIBIN to determine whether they solicit other law enforcement agencies to participate in the NIBIN program. Of the 160 partner agencies that returned the questionnaires, we determined that:

  • 77 percent (123) stated that they do solicit other law enforcement agencies within their service area to submit firearms evidence,
  • 21 percent (33) of respondents indicated they do not encourage law enforcement agencies within their service area to submit firearms evidence, and
  • 2 percent (4) did not respond.

We assessed the level of outreach of the NIBIN partner agencies from the data in the NIBIN database by analyzing the number of law enforcement agencies that had submitted evidence to each partner agency. The analysis showed that while some partner agencies received evidence from other law enforcement agencies for entry into NIBIN, many partner agencies received very little evidence from other law enforcement agencies. For the 196 partner agencies that had contributed data to NIBIN, the average number of law enforcement agencies that submitted data to a partner agency was 48 and the median was 32. One partner agency (Pennsylvania State Police – Harrisburg) contributed data to NIBIN that was submitted by 478 other law enforcement agencies, while 13 partner agencies only contributed their own data. We found that 26 percent (51) of the 196 partner agencies either had not received evidence from any other law enforcement agencies or had received evidence from 10 or fewer other law enforcement agencies.

Our analyses of data in the NIBIN database; interviews of partner agency officials during site visits; and analyses of responses obtained from partner agencies, participating non-partner agencies, and non-participating non-partner agencies led us to conclude that the NIBIN program has not been fully promoted to law enforcement agencies by the ATF or by the NIBIN partner agencies. In our judgment, this lack of promotion contributes significantly to the low participation in the program by many partner and non-partner agencies. The effectiveness of the NIBIN program could be improved if the ATF more aggressively promotes NIBIN’s benefits in helping to solve crimes, and encourages the partner agencies to promote the NIBIN program to other law enforcement agencies in their area.

Collecting and Submitting Firearms Evidence

Another factor essential to the success of the NIBIN program is for the participating agencies to enter the maximum amount of evidence into the system. As shown earlier, the more evidence entered into the system, the greater the chance that hits will be identified. ATF officials told us it is not only important to enter as much evidence as possible, but also to enter both bullets and cartridge casings to take full advantage of the system’s capabilities. To determine if the participating partners and non-partners were entering the maximum evidence into the system, we determined the type of data being entered by: (1) analyzing the evidence data entered into NIBIN for each partner agency, (2) visiting 22 partner agencies and discussing with agency officials the type of evidence entered into the system, (3) sending survey questionnaires to the remaining 174 partners that contributed evidence to NIBIN and to a sample of 411 participating non-partner agencies asking similar questions, and (4) contacting via telephone all the agencies that received the RBI units for imaging cartridge casings.

Our analysis of the evidence entered into NIBIN showed that 72 percent (640,652) of the evidence entered was cartridge casings and only 28 percent (247,702) of the evidence was bullets. We analyzed this data further and found that the top 10 percent of partner agencies that contributed the most balanced mix of bullets and cartridge casings contributed from 55 percent bullets/45 percent cartridge casings to 42 percent bullets/58 percent cartridge casings as shown in the following table.

Analysis of Evidence Entered into NIBIN by Partner Agencies
Contributing a Balanced Mix of Bullets and Cartridge Casings
Partner Agency Total
Evidence
Entered
Bullets
Entered
% Bullets
To Total
Evidence
Cartridge
Casings
Entered
% Cartridge
Casings
To Total Evidence
ATF Laboratory – Walnut Creek (CA)
12,141
6,638
55%
5,503
45%
Florida Department of Law Enforcement – Pensacola
1,575
817
52%
758
48%
North Carolina State Bureau of Investigation Crime Lab
14,624
7,350
50%
7,274
50%
Baltimore Co. PD
1,770
863
49%
907
51%
Florida Department of Law Enforcement – Tampa
7,307
3,521
48%
3,786
52%
Erie Co. (NY) Forensic Lab
2,411
1,131
47%
1,280
53%
Georgia Bureau of Investigation – Decatur
33,175
15,209
46%
17,966
54%
Charlotte PD
9,339
4,322
46%
5,017
54%
Knoxville PD
235
106
45%
129
55%
Greenville (SC) PD
1,745
761
44%
984
56%
Georgia Bureau of Investigation – Savannah
6,276
2,738
44%
3,538
56%
ATF Laboratory – Atlanta
19,744
8,649
44%
11,095
56%
Hickory (NC) PD
1,372
604
44%
768
56%
Bergen Co. (NJ) Sheriff’s Office
1,243
551
44%
692
56%
North Dakota Dept. Of Health
9
4
44%
5
56%
State of Connecticut Department of Public Safety
5,663
2,418
43%
3,245
57%
San Mateo Co. Sheriff's Office
814
349
43%
465
57%
Alabama Department of Forensic Sciences – Huntsville
1,052
446
42%
606
58%
Pasadena (TX) PD
1,566
664
42%
902
58%
Kansas City Regional Crime Lab
903
383
42%
520
58%
 Source: OIG Analysis of NIBIN Database Evidence Data

By comparison, the bottom 12 percent of partner agencies that entered the fewest bullets as compared to cartridge casings contributed from 0 (zero) percent bullets/100 percent cartridge casings to 1 percent bullets/99 percent cartridge casings as shown in the following table.

Analysis of Evidence Entered into NIBIN by Partner Agencies
Contributing Mostly Cartridge Casings and Few or No Bullets
Partner Agency Total
Evidence
Entered
Bullets
Entered
% Bullets
To Total
Evidence
Cartridge
Casings
Entered
% Cartridge
Casings To
Total Evidence
Newark (NJ) PD
896
-
0%
896
100%
Massachusetts State Police – Sturbridge
344
-
0%
344
100%
Washington State Police – Spokane
587
-
0%
587
100%
Northern Utah Laboratory – Ogden
370
-
0%
370
100%
Maryland State Police
1,924
-
0%
1,924
100%
Illinois State Police – Fairview Heights
1,387
-
0%
1,387
100%
Los Angeles Co. Sheriff's Office
3,544
-
0%
3,544
100%
Illinois State Police – Springfield
481
-
0%
481
100%
San Bernardino Co. Sheriff’s Office
1,634
1
0%
1,633
100%
New York State Police – Albany
955
1
0%
954
100%
Washington State Patrol – Seattle
5,659
7
0%
5,652
100%
Las Vegas Metro PD
703
1
0%
702
100%
California Department of Justice – Riverside
462
1
0%
461
100%
Wisconsin State Patrol – Milwaukee
2,370
6
0%
2,364
100%
Massachusetts State Patrol – Sudbury
1,564
7
0%
1,557
100%
Massachusetts State Patrol – Danvers
192
1
1%
191
99%
Miami–Dade PD
7,597
43
1%
7,554
99%
Minneapolis PD
1,713
10
1%
1,703
99%
Tucson PD
1,284
9
1%
1,275
99%
Southwestern Institute of Forensic Sciences – Dallas
1,378
12
1%
1,366
99%
Bexar Co. (TX) Laboratory
525
5
1%
520
99%
Rhode Island State Crime Laboratory
7,988
82
1%
7,906
99%
Illinois State Police – Morton
485
5
1%
480
99%
 Source: OIG Analysis of NIBIN Database Evidence Data

In our survey questionnaires and during our site visits to the partner agencies, we asked officials to explain why they were not entering bullets into NIBIN. Examples of responses from the partner agencies included:

  • Bullets are not entered due to the shortage of manpower. There are priority cases that have to be worked in addition to NIBIN. More hits are from cartridge casings. (Prince George ’s County, Maryland, Police Department)
  • The entry of bullets is extremely time-consuming and the system was not well-designed for bullets. (Minneapolis, Minnesota, Police Department)
  • Some bullets from evidence are entered, but not from test-fires. The process takes considerably more time than entering cartridge casings and the hit rate on bullets is much lower than on cartridge casings. (Omaha, Nebraska, Police Department Crime Laboratory)
  • Bullets are not entered due to the backlog of evidence awaiting entry. (Allegheny County, Pennsylvania, Coroner’s Office Forensic Laboratory Division)
  • Bullets are not entered because: (1) the quality of the correlation result is poor, (2) minimal returns on results are obtained, (3) enough manpower is not available to enter both bullets and cartridge casings, and (4) the IBIS equipment does not allow for easy imaging of bullets. (Los Angeles, California, Police Department)
  • Bullet evidence is not entered due to the low probability of success. (Illinois State Police – Joliet)

We obtained similar responses from the participating non-partner agencies that responded to our survey, 25 percent of whom indicated that they submit few to no bullets for entry into NIBIN.

As the responses show, the consistent theme from both the partner agencies and the participating non-partner agencies is that the: (1) process of entering bullets is difficult and time consuming, and (2) quality of bullet images produced by the IBIS equipment is not adequate to produce enough hits for the agencies to spend the time and resources necessary to enter bullet evidence into NIBIN. The NIBIN Program Director agreed that some partner agencies are not entering bullets into NIBIN. Consequently, the ATF needs to determine whether new technology exists that will improve the image quality of bullets enough to make it worthwhile for the participating agencies to spend resources to enter the bullet data into NIBIN.

In our interviews with the agencies that received RBI units for imaging cartridge casings, we determined that most of the agencies were not satisfied with the operation of the RBI units. Consequently, some of the agencies returned the RBI units to the ATF and others were no longer using them to enter cartridge casing evidence into NIBIN. We contacted 28 agencies that had received the RBI units to inquire about their use of the RBI units. We obtained responses from 25 of the 28 agencies regarding their satisfaction and use of the RBI units. The discussions we had with 20 of the 25 agencies indicated they were not satisfied with the operation of the RBI units. A sample of the comments we received from these agencies included:

  • The RBI unit is very unreliable, the modem fails, and the agency has trouble connecting to its RDAS partner. (Aurora, Colorado, Police Department)
  • The RBI unit is no good, the lighting is bad, the images are not clear, and the download is slow. (Bridgeport, Connecticut, Police Department)
  • The equipment breaks down often. (North Louisiana Crime Laboratory – Alexandria)
  • All RBI units should be replaced with RDAS units. The lines should be upgraded to download faster. Also, 3-D imaging technology should be incorporated into the system. (Rockland County, New York, Sheriff’s Department)
  • The system failed on several occasions and was down at least four months this past year. (Youngstown, Ohio, Police Department)

From our discussions, we determined that 6 of the 25 agencies had returned the RBI units to the ATF. In addition, 4 of the remaining 19 agencies indicated that they had either stopped using the RBI units to enter cartridge casing evidence into NIBIN or used the units sparingly. The NIBIN Program Director told us that the equipment manufacturer had developed a new unit that is much better at imaging cartridge casings than the RBI units, but the new unit is much more costly. We later were told by the NIBIN program staff that the RBI unit is considered to be obsolete and is no longer in production by the manufacturer. Given the dissatisfaction among users of the RBI units, the ATF needs to perform an analysis of the current RBI users, and any other potential users, to determine if they would use the new unit enough to warrant the additional cost.

Images Correlation Process

Another key factor that contributes to the success of the NIBIN program is measuring similarities between ammunition components by comparing their images entered into NIBIN using the IBIS equipment. This process is referred to as the “correlation” process and resulting measurements are called “correlations.” The process works as follows:

  • When images are captured by an RDAS unit, the data is sent from the unit to the regional server for storage and to generate correlations against other images in the regional database. Images that are captured by an RBI unit are sent by telephone line to the assigned RDAS unit. After a quality review, the images are transmitted to the regional server for comparison against images of other items in the regional database.
  • The correlations of images are computed by algorithm electronically after examining the similarities between two images. If the images are very similar, they likely represent images of ammunition components fired from the same firearm.
  • After the algorithm on the server compares the images, it ranks the list of images based on their correlation results. Since this process is automated by comparing images algorithmically, potential matches can be found in a large volume of images much faster than a firearms technician or examiner can view them manually.
  • The correlation results for images with very high correlation scores are electronically sent back to the RDAS unit for evaluation.
  • If the correlation results originated from an RBI unit, the agency possessing the RBI is telephoned with the results.

After the correlation process is completed, the IBIS technician at the submitting agency reviews the correlation results, selecting the top matches and identifying these matches in the IBIS system as high-confidence candidates. The high-confidence candidates must then be reviewed and examined by a licensed firearms examiner who determines whether they are actual matches resulting in a hit. To make this determination, the firearms examiner obtains the original evidence to compare with the high-confidence candidates. If the comparison results in a hit, the firearms examiner notifies the IBIS technician, who then identifies the hit in the system.

During our initial visits to NIBIN partner agencies, we found a very high-volume NIBIN partner (Georgia Bureau of Investigation – Decatur) that did not perform reviews of high-confidence candidates. At the time of our audit, the Georgia agency had not reviewed and examined high-confidence candidates since January 2002 – about 3,350 high-confidence candidates. An official from the Georgia Bureau of Investigation stated that high-confidence candidates were not being reviewed because: (1) funding for overtime for firearms examiners was not available, and (2) firearms examiners had not had the opportunity to work with the IBIS equipment. The agency official also told us that using the IBIS equipment was considered a duty outside of the firearm examiner’s daily laboratory activities.

When partner agencies do not review the hits on high-confidence candidates, it is a serious failure in utilizing the program’s capability. Further, if high-confidence candidates are not reviewed, hits from the data entered into NIBIN cannot be identified and leads that might help find and convict a person involved in a firearms-related crime may not be pursued. In addition, the significant amount of resources used to collect, transport, store, protect, and enter the evidence into IBIS could be wasted. This situation was magnified because the Georgia Bureau of Investigation was a state crime laboratory that participating non-partner agencies relied on by submitting evidence for entry into IBIS. However, because the state crime laboratory did not review high-confidence candidates, the submitting non-partner agencies did not receive any benefits from the submitted data.

After identifying this problem, we asked the ATF if it had a routine report showing the non-viewed correlations for each NIBIN partner agency to determine whether other partner agencies were not reviewing the correlations. The ATF responded that it did not have a standard NIBIN report that showed this data, but stated that it was able to query NIBIN and produce a report (NIBIN Non-Viewed Correlation Requests Report) showing the non-viewed correlations as of September 30, 2004. The report identified 100 partner agencies that had a total of 4,024 bullet evidence correlations recorded in NIBIN as non-viewed and 155 partner agencies that had a total of 18,379 cartridge case evidence correlations recorded in NIBIN as non-viewed.

During our subsequent visits, we asked 15 NIBIN partner agencies to verify the non-viewed correlation numbers shown on the NIBIN Non-Viewed Correlation Requests Report. The partner agencies confirmed that the report numbers agreed with the data in NIBIN. However, the partner agencies told us that while the correlations were shown in NIBIN as non-viewed, they actually had been reviewed, but NIBIN had not been updated to show this. The partner agencies gave the following reasons for why the correlations had not been updated in NIBIN:

  • The NIBIN partners did not know how to use the “correlation viewed” feature within the system. As a result, the correlations were shown as not being viewed on the ATF’s report.
  • The number of non-viewed correlations was representative of the volume of evidence that is put into the IBIS system on a daily basis. It takes about four hours from the time evidence is entered until the time any correlations are ready for viewing. Consequently, the amount of non-viewed correlations was not really a backlog.
  • The number of non-viewed correlations represented evidence that was downloaded from an RBI unit to another agency’s RDAS unit and the correlations were marked as viewed on the other agency’s RDAS unit. Later this agency received a RDAS unit and the NIBIN contractor downloaded the previous correlations to the new RDAS unit. Consequently, all the correlations appeared erroneously as non-viewed.
  • A software upgrade caused previously viewed items to be reported as non-viewed.
  • Correlations on the system were generated as experimental entries.
  • The correlations were primarily from test-fired firearms. Personnel that were responsible for the entries for two years were in training to become a full-time firearms examiner and were not able to keep up with the correlations.

Given the problems associated with the Georgia Bureau of Investigation and the inaccurate data contained in the NIBIN Non-Viewed Correlation Requests Report, the ATF needs to: (1) provide guidance to partner agencies on the necessity to view correlations in a timely manner and to ensure that they are properly marked as viewed in NIBIN, and (2) monitor the non-viewed correlations by partner agencies, and take corrective actions when a backlog of correlations is identified.

Backlogged Evidence

The success of NIBIN depends on firearms evidence being entered into the system as soon as possible. Otherwise, leads generated from the system and from subsequent firearm examiners’ analyses of the evidence may not be identified and followed-up on quickly enough to prevent additional crimes from occurring. To determine whether participating agencies were timely entering firearms evidence into NIBIN, we visited 22 NIBIN partner agencies and sent survey questionnaires to the remaining 174 partner agencies that had contributed evidence to NIBIN and to 411 participating non-partner agencies.

During our visits to the 22 partner agencies, we found that many of the agencies had significant backlogs of firearms evidence that had not been entered into NIBIN. Below are some of the results:

  • 1,000 or more bullets and cartridge casings, and 269 test-fired bullets, cartridge casings, and firearms (Prince George ’s County, Maryland, Police Department)36
  • 3,079 test-fired bullets and 5,202 test-fired cartridge casings (Georgia Bureau of Investigation – Decatur)
  • 500 bullets and cartridge casings, and 6 firearms awaiting test-fire (Hickory, North Carolina, Police Department)
  • 331 bullets and 940 cartridge casings (New Orleans, Louisiana, Police Department)
  • 100 test-fired bullets and cartridge casings and 300 cartridge casings (Rhode Island State Crime Laboratory)
  • 716 cartridge casings and bullets, and 1,730 test-fired cartridge casings and bullets (Los Angeles, California, Police Department)

For the 174 partner agencies surveyed, 160 returned the survey questionnaires to us. From the questionnaires returned, we determined that many of the partner agencies had significant backlogs of firearms evidence that had not been entered into NIBIN as shown in the following table.

Summary of Partner Agency Survey Responses Indicating
a Backlog of Evidence Awaiting Entry into NIBIN
Type of Backlog Number
of
Agencies
Amount
Awaiting
Entry37
Bullets collected 86 4,905
Cartridge cases collected 107 10,796
Bullets from test-fired firearms collected 55 10,911
Cartridge casings from test-fired firearms collected 70 5,289
Firearms awaiting test-fire 119 9,738
 Source: Survey Questionnaires from Partner Agencies

The primary reasons that the partner agencies provided for backlogged evidence were staffing shortages, other priorities, and the large volume of evidence submitted.

For the 411 participating non-partner agencies surveyed, 228 returned the survey questionnaires to us. From the questionnaires returned, we determined that the participating non-partner agencies did not have a significant amount of firearms evidence that had not been entered into NIBIN.

Any delays in entering firearms evidence into NIBIN prevents matching the evidence to other evidence in the system, which could lengthen the time it takes to identify and apprehend a criminal, or could result in additional crimes being committed before the criminal is caught. Therefore, the ATF needs to develop strategies to help the partner agencies eliminate the current backlog of firearms evidence awaiting entry into NIBIN. The research should consider whether: (1) the partner agencies can send their backlogged evidence to the ATF laboratories or to other partner agencies for entry into NIBIN, and (2) improvements to the efficiency of NIBIN would facilitate more rapid and easy entry of evidence.

Federal Agencies

In January 2001, the Secretary of Treasury and the Attorney General issued memoranda requiring that all law enforcement agencies within those two departments should trace every recovered crime firearm through the ATF’s National Tracing Center and enter bullets and cartridge casings found at crime scenes into NIBIN (See Appendix X).

To determine whether the Department of Justice agencies were complying with the January 2001 directive, we sent survey questionnaires to the ATF, FBI, Drug Enforcement Administration (DEA), Federal Bureau of Prisons (BOP), and United States Marshals Service (USMS) to ask whether or not they collect firearms evidence. We found that all of the agencies except the BOP collect such evidence. The BOP indicated that it does not collect firearms evidence because this type of enforcement falls within the jurisdiction of the FBI. Four agencies – the ATF, FBI, DEA, and USMS – also indicated that they submit firearms evidence for entry into NIBIN. The DEA said that the only locations that participate in the NIBIN program are the Los Angeles, Chicago, and New York division offices. The DEA also explained that it is currently working to revise the guidance given to division offices on the ballistic testing of seized firearms, and that it will be working with the ATF on specific requirements for firearms submissions to the NIBIN program. We reviewed the firearms evidence data in NIBIN as of October 22, 2004, and found that while the ATF and FBI had many offices contributing data to NIBIN, the DEA and USMS had much lower participation as shown below.

Participation in NIBIN by Department
of Justice Law Enforcement Agencies
Agency Offices
Contributing
Data
Amount
of Data
Contributed
ATF 114 (Gun Center). The pilot project required the ATF to hire an IBIS technician on a contract basis to work at the Gun Center and enter images of the ATF’s recovered firearms evidence, as well as images of all firearms evidence recovered by other federal agencies in Southern California .

One of the purposes of the pilot program was to determine the feasibility and resources for imaging 100 percent of the recovered firearms by federal agencies. Before the pilot program ended, the ATF and its NIBIN contractor (Forensic Technology, Inc.) developed protocols ce: NIBIN database

To ensure the ATF and other federal agencies adhere to the directive established by the Attorney General and the Secretary of the Treasury, the ATF developed a pilot program in August 2003 that placed IBIS equipment for at least one year in the ATF’s newly formed Southern California Regional Crime Gun Center, known as the Los Angeles Gun Center (Gun Center). The pilot project required the ATF to hire an IBIS technician on a contract basis to work at the Gun Center and enter images of the ATF’s recovered firearms evidence, as well as images of all firearms evidence recovered by other federal agencies in Southern California .

One of the purposes of the pilot program was to determine the feasibility and resources for imaging 100 percent of the recovered firearms by federal agencies. Before the pilot program ended, the ATF and its NIBIN contractor (Forensic Technology, Inc.) developed protocols and procedures for test-firing and imaging all firearms recovered in the Southern California area. Protocols were also developed for maintaining and archiving test-fired evidence and associated documentation, so that the firearms evidence could be destroyed. The NIBIN Program Director told us, however, that funding was not available for the ATF to test-fire all federally seized firearms and enter all firearms evidence collected by federal agencies into NIBIN. Given the funding situation, the ATF needs to coordinate with the other Department of Justice law enforcement agencies that seize firearms and firearms evidence to help them establish a process for entering the seized evidence into NIBIN.

Conclusion

We found that only about 20 percent of the law enforcement agencies nationwide participate in NIBIN. We believe the ATF could improve participation by: (1) better promoting the use and benefits of NIBIN to law enforcement agencies, and (2) involving the partner agencies more in promoting the NIBIN program to other law enforcement agencies in their area. We also found that many participating agencies: (1) were not satisfied with the quality of images produced by NIBIN from bullet evidence, and (2) believed that it was not cost-effective to enter bullet evidence into NIBIN because very few hits result from the poor-quality bullet images.

The ATF needs to determine if new technology exists that will improve the image quality of bullets enough to make it worthwhile for the participating agencies to spend limited resources to enter the bullet data into NIBIN. In addition, we found the potential matches in NIBIN had not been reviewed by one high-volume partner agency since January 2002. Consequently, potential leads that might result from reviewing potential matches were not identified and pursued. Finally, the ATF needs to provide guidance to NIBIN partners stressing the importance of viewing correlations in a timely manner and for ensuring that NIBIN accurately reflects the status of those viewed correlations.

Recommendations:

We recommend that the ATF:

  1. Ensure that NIBIN partner agencies enter the ORI number of the contributing agency for all evidence entered into NIBIN.
  2. Resolve the duplicate case ID number issue in the NIBIN database for the Colorado Bureau of Investigation – Montrose; and the Rhode Island State Crime Laboratory.
  3. Research the reasons why 12 agencies have achieved high hit rates with relatively low number of cases entered into NIBIN and share the results of such research with the remaining partner agencies.
  4. Establish a plan to enhance promotion of NIBIN to law enforcement agencies nationwide to help increase participation in the program. The plan should address steps to: (1) increase the partner agencies’ use of the system, (2) increase the non-partner agencies’ awareness and use of the system, and (3) encourage the partner agencies to promote the NIBIN program to other law enforcement agencies in their area.
  5. Determine whether new technology exists that will improve the image quality of bullets enough to make it worthwhile for the participating agencies to spend valuable resources to enter the bullet data into NIBIN, and deploy the technology if it is cost-effective.
  6. Perform an analysis of the current RBI users, and any other potential users, to determine if they would use an improved system enough to warrant the additional cost. If the analysis concludes that another system would be cost-effective, then the ATF should pursue funding to obtain the system.
  7. Provide guidance to partner agencies on the necessity to view correlations in a timely manner and to ensure that correlations viewed in NIBIN are properly marked.
  8. Monitor the non-viewed correlations of partner agencies and take corrective actions when a backlog is identified.
  9. Research ways to help the partner agencies eliminate the current backlog of firearms evidence awaiting entry into NIBIN. The research should consider whether the partner agencies can send their backlogged evidence to the ATF Laboratories or to other partner agencies for entry into NIBIN, and whether improvements to the efficiency of NIBIN would facilitate more rapid and easy entry of evidence.
  10. Coordinate with Department of Justice law enforcement agencies that seize firearms and firearms evidence to help them establish a process for entering the seized evidence into NIBIN.
    3. PREVENTING ENTRY OF NEW FIREARMS DATA INTO NIBIN

    Federal law prohibits the entry of ballistic images from newly manufactured, imported, or sold firearms into NIBIN. To ensure that this prohibition is followed, the ATF established a control that consists of NIBIN users signing an MOU agreeing not to enter such data. We tested the effectiveness of this control system and found no evidence that prohibited data had been entered into NIBIN.

NIBIN Data Prohibitions

The Firearms Owners’ Protection Act of 1986 (the Act) prohibits the establishment of any system of registration of firearms, firearms owners, or firearms transactions or dispositions. This provision prohibits directly linking ballistic images through a centralized computer database to both the firearms themselves (which would constitute a firearms registry), and the identities of the private citizens who possess image firearms (which would constitute a firearms owners’ registry). The Act states that:

No such rule or regulation prescribed after the date of the enactment of the Firearms Owners’ Protection Act may require that records required to be maintained under this chapter or any portion of the contents of such records, be recorded at or transferred to a facility owned, managed, or controlled by the United States or any state or any political subdivision thereof, nor that any system of registration of firearms, firearms owners, or firearms transactions or dispositions be established.

Language included in the ATF’s annual appropriations bill also requires that ballistic images of bullets and cartridge casings from newly manufactured, imported, or sold firearms are not to be available in or connected to NIBIN. Therefore, the IBIS equipment deployed to law enforcement agencies participating in the NIBIN program should not be used to capture or store such ballistic images.

Controls Established by the ATF

We interviewed ATF officials, observed the entry of data into NIBIN by users, and obtained and reviewed documentation provided by ATF officials. The ATF officials explained that it requires NIBIN users to sign an MOU that forbids them from entering the prohibited data. The MOU states:

Ballistic imaging systems provided and deployed by ATF to other federal, state, or local law authorities may be used only for imaging operations associated with criminal law enforcement functions. Systems deployed to other federal, state, or local authorities shall not be used to capture, directly search, or store ballistic images acquired at the point of manufacture, importation, or sale.

Since NIBIN does not contain controls to prevent the entry of prohibited data, we asked an ATF official how such entry can be prevented. She told us that the ATF does not monitor the data entered by NIBIN users, but that it is unlikely that they are entering prohibited ballistic images into the system because they are fully aware of the MOU prohibition.

To evaluate whether NIBIN users were aware of and complying with the MOU prohibition, we visited 22 NIBIN partner agencies during the audit and sent survey questionnaires to the other 174 NIBIN partner agencies that operated NIBIN RDAS units. We determined that all the NIBIN partner agencies were aware of the MOU prohibition and we found no indication of non-compliance. We also sent survey questionnaires to 411 participating non-partner agencies in which we asked if the agencies submitted prohibited data for entry into NIBIN. None of the participating non-partner agencies that returned the questionnaires indicated that they submitted such data.

States That Maintain Their Own Databases

An ATF official told us that two NIBIN partner agencies, the Maryland State Police (MSP) Forensic Sciences Division and the New York State Police (NYSP), also maintain state databases with ballistic images from newly manufactured firearms. The ATF official said that it is unlikely these NIBIN users are entering prohibited ballistic images into NIBIN because the agencies are also fully aware of the MOU prohibition. We confirmed the ATF official’s statement that the States of Maryland and New York had established database systems for entering ballistic images from new firearms.38 Since these two agencies operate both NIBIN and a separate state system for new firearms images, we evaluated their compliance with the MOU prohibition by performing a comparison between the firearms data in NIBIN and the firearms data in each agency’s system.

Maryland State Forensic Sciences Division: The MSP system that contains new firearms data is the Maryland-Integrated Ballistic Identification System (MD- IBIS). MSP officials informed us that the State of Maryland purchased its own IBIS equipment in September 2000 under the assumption that they eventually would be able to connect it to the ATF's NIBIN and share data. However, the MSP was told by the ATF that the data could not be shared between the two systems because it was illegal to do so. We observed that at the sole location where they co-exist, the NIBIN and MD- IBIS systems were kept in separate rooms and each was accessible to authorized staff only. In addition, according to MSP officials, the two systems were not connected.

The MSP officials also told us that they are aware of the federal law prohibiting entry of new firearms information into NIBIN. Further, officials believe their controls are strong enough to prevent any new firearms entry into NIBIN, because the systems are kept separate and require checks and double-checks before entry. We reviewed the MSP’s controls and did not identify any weaknesses.

To determine the MSP’s compliance with the prohibition against entering firearms data from newly manufactured, imported, or sold firearms, we obtained an electronic copy of the firearms data contained in the MSP’s MD- IBIS system and the firearms data contained in the ATF’s NIBIN. The MSP’s MD- IBIS system contained 47,798 firearms records and NIBIN contained 254,187 firearms records. We performed a computerized comparison between the data in both systems. The comparison identified only 42 firearms entries from new firearms in the MD- IBIS system where the same firearms were also entered in NIBIN. Of the 42 matching entries in NIBIN, 32 were entered into NIBIN by the MSP, 8 were entered by the Washington, D.C., Metropolitan Police Department, 1 was entered by the Colorado Division of Investigation – Pueblo, and 1 was entered by the Los Angeles, California, Police Department. However, all four agencies provided documentation to show that the firearms were each entered into NIBIN as a result of a crime and not as new firearms. Therefore, we concluded that the MSP was complying with the prohibition against entering firearms data from newly manufactured, imported, or sold firearms into NIBIN.

New York State Police: The NYSP officials explained that they have two state systems for entering firearms data. One system is used for the entry of new firearms information and the other system is used for the entry of crime-related firearms information. The two systems together are known as the Combined Ballistic Identification System (COBIS). We will refer to the COBIS system for new firearms data as COBIS 1 and the COBIS system for crime-related firearms data as COBIS 2. The NIBIN and both the COBIS systems were located only at the NYSP headquarters in Albany, New York. The NIBIN and the COBIS 2 system were located together in one room and the COBIS 1 system was located in a separate room across the hall.

The NYSP officials told us they were aware that the MOU with the ATF prohibits the entry of newly manufactured, imported, or sold firearms into NIBIN. Officials also said that they have not violated this MOU condition and that since the State of New York requires the tracking of newly manufactured, imported, or sold firearms, the state set up the separate system (COBIS 1) with controls to identify and track such firearms. We reviewed the NYSP’s controls and did not identify any weaknesses.

To determine the NYSP’s compliance with the prohibition against entering firearms data from newly manufactured, imported, or sold firearms, we obtained an electronic copy of the firearms data contained in the NYSP’s COBIS 1 system and the firearms data contained in the ATF’s NIBIN.39 The NYSP’s COBIS 1 system contained 90,063 firearms records and the NIBIN contained 254,187 firearms records. We performed a computerized comparison between the data in both systems. The comparison identified only 14 firearms entries from new firearms in the COBIS 1 system where the same firearms were also entered into NIBIN. Of the 14 matching entries in NIBIN, 5 were entered into NIBIN by the NYSP; 3 were entered by the Erie County, New York, Forensic Laboratory; 2 were entered by the Monroe County, New York, Public Safety Laboratory; 2 were entered by the Westchester County, New York, Police Department; 1 was entered by the Massachusetts State Police; and 1 was entered by the Alabama Division of Forensic Sciences laboratory. However, all six agencies provided explanations to show that the firearms were each entered into NIBIN as a result of a crime and not as new firearms. Therefore, we concluded that the NYSP was complying with the prohibition against entering firearms data from newly manufactured, imported, or sold firearms into NIBIN.



Footnotes

  1. According to the ATF, full deployment of the IBIS equipment was completed in May 2003, at which time it had been sent to 227 NIBIN partner agencies. Subsequent to May 2003, IBIS equipment was delivered to 13 additional agencies that requested it and taken back from 9 sites that no longer needed it. Therefore, as of January 2005, the IBIS equipment was in use at 231 NIBIN partner agencies.

  2. The remaining partner agencies received only RBI units that submit data through an RDAS unit. The NIBIN tracks data by RDAS and not by RBI.

  3. Every agency that reports data for inclusion in the FBI’s Uniform Crime Reports (UCR) is assigned a unique ORI number by the FBI. For the most part, only police agencies have an ORI and report crimes, but other agencies such as fire marshals, alcoholic beverage control agencies, regional and special-purpose task forces, federal agencies, and private colleges also have law enforcement responsibilities and are assigned ORI numbers. Some agencies also have separate divisions for reporting purposes and each reporting division has a separate ORI number. We could not match the data entered to the ORI numbers contributing the data for 115,092 of the 888,447 data items entered because the ORI numbers were not entered into the system.

  4. Two of the seven partner agencies (Omaha Police Department Crime Laboratory and Prince George’s County (MD) Police Department) were part of the 22 partner agencies we visited. These agencies’ responses regarding searches performed are included in the table at Appendix XII. The other five partner agencies (West Virginia State Police – Charleston; Idaho State Police; Kansas Bureau of Investigation – Topeka; Massachusetts State Police – Sturbridge; and the Lake County (IN) Crime Laboratory) were part of the 174 partner agencies surveyed. These agencies’ responses regarding searches performed are included in the tables at Appendices XIII and XIV.

  5. The 254,187 records in the NIBIN firearms table were contributed by 184 partner agencies.

  6. The UCR program encompasses approximately 17,000 law enforcement agencies nationwide that voluntarily contribute crime statistics. The program seeks to generate a reliable set of crime statistics for use in law enforcement administration, operation, and management. Agencies voluntarily provide summarized reports on offenses, persons arrested, and law enforcement officers killed and assaulted. For the most part, agencies submit monthly crime reports to a centralized crime-records facility within their state. The state UCR program then forwards the data, using uniform-offense definitions, to the FBI’s national UCR program. Agencies in states that do not have a state program submit their statistics directly to the national program. The FBI compiles, publishes, and distributes the data to participating agencies, state UCR programs, and others interested in crime data.

  7. An official at this partner agency told us that the backlog of bullets numbered in the thousands. Since an exact number was not known, we used a conservative estimate of 1,000 or more.

  8. The amount of evidence awaiting entry into NIBIN is a conservative estimate because some agencies did not provide numbers and other agencies provided only a range of numbers. In all cases, we used the lower estimate.

  9. In April 2000, the State of Maryland adopted the first ballistic imaging law that required the establishment of state ballistic imaging systems that directly link the images of newly manufactured or sold firearms to firearms ownership. In August 2000, the State of New York enacted a similar law that required the ballistic imaging of all handguns shipped into the state after March 1, 2001.

  10. The NYSP provided us a copy of the data in their COBIS 1 system while we were on-site and required that we return the data copy to them before we left the premises. Since we were not able to keep a copy of the NYSP’s data, we had a NYSP official sign a statement confirming the matches we found between the NYSP’s COBIS 1 data and the NIBIN data.



Previous Page Back to Table of Contents Next Page