X. Chapter Ten: Integration of IDENT with FBI Automated Fingerprint Identification Systems

This chapter describes the efforts by the INS and the FBI to integrate their automated fingerprint identification systems. The INS and the FBI have periodically discussed integrating their systems since they began developing them around 1990. However, the INS had differing needs and developed its automated fingerprint identification system faster than the FBI developed its own systems. The INS therefore deployed IDENT in 1994, well before the FBI’s systems – the Integrated Automated Fingerprint Identification System (IAFIS) or NCIC 2000 – were deployed. Since then the INS and the FBI continued to discuss integration of their systems, but no agreements were reached on such integration. As a result, when IAFIS and NCIC 2000 were deployed in the summer of 1999, they were not linked with IDENT.

It is important to keep in mind that IAFIS and NCIC 2000 were not yet operational when the problems of the Resendez case unfolded. But the problems in the Resendez case vividly illustrate the need for such integration. This case has spurred the FBI and the INS, at the direction of Congress and under the guidance of the DOJ Justice Management Division (JMD), to develop a plan for integration. Unfortunately, this plan could take five years to implement fully. Based on our review, we believe that integration of IDENT and IAFIS is a critical step that should be accomplished as expeditiously as possible to avoid a repetition of a tragedy similar to the Resendez case. This integration needs to be aggressively pursued even after the publicity over the Resendez case recedes.

A. History of Efforts to Integrate INS and FBI Automated Fingerprint Identification Systems

1. Early Discussions About Integration

Since the 1920s, the FBI’s Identification Division has maintained a central repository of ten-fingerprint cards of criminal offenders. In 1967, the National Crime Information Center (NCIC) was created to provide a national database of information on wanted individuals and stolen articles, vehicles, guns, and license plates submitted by participating federal, state and local law enforcement agencies. The Interstate Identification Index (III), which contains information on arrests and dispositions, was added to NCIC in 1983.

The Advisory Policy Board (APB), composed of approximately 30 representatives from the federal, state and local criminal justice community, was established in 1967 to establish policy and procedures for the management and use of the NCIC system and make recommendations to the FBI Director. In June 1989, the FBI Director asked the APB to expand its role and provide advice and guidance on fingerprint identification issues as well. In February 1990, the APB recommended that the FBI overhaul its paper-based fingerprint identification system and create a new system, IAFIS, which would allow electronic searches for fingerprint matches.

In 1990 and 1991, the INS and the FBI met regularly to discuss their planned automated fingerprint identification systems. Memoranda summarizing some of these meetings stated that the INS and the FBI were coordinating the development of an integrated automated fingerprint identification system to determine if an integrated system could satisfy the INS’s needs. The memoranda included preliminary diagrams and narratives of how FBI’s IAFIS and NCIC 2000 and the INS’s system (not yet officially referred to as IDENT) could be linked in an overall automated fingerprint identification system (AFIS) network. There was also discussion of how to ensure common high-quality fingerprint image and electronic transmission standards for fingerprints and identification data so that they could be transmitted among different AFIS’s.

During these planning meetings, the INS and the FBI specifically discussed whether the FBI would be willing to store fingerprint records of criminal and non-criminal aliens or whether it would endorse the INS establishing a separate system if the FBI’s system could not support all of the INS’s requirements. In these early discussions, both the INS and the FBI thought that the INS would not need to check all apprehended aliens against the FBI’s databases. Rather, they believed that the INS would only check those aliens it identified as potential criminals. They assumed that the FBI’s IAFIS could handle a small volume of INS requests with a quick response time. The INS emphasized that a fast response time was critical because the INS could not detain large numbers of apprehended aliens for long periods of time while waiting for responses to criminal checks.

An early difference between the INS and the FBI was whether the INS should take two fingerprints or ten fingerprints from the aliens it apprehended. The FBI, along with state and local law enforcement agencies, believed that the INS should take ten fingerprints so that they could be matched against ten fingerprint records in the law enforcement agencies’ databases or any latent fingerprints obtained at crime scenes. The INS believed that taking ten-prints of all apprehended aliens would take too long and adversely affect its ability to carry out its mission, because large numbers of aliens would have to be detained for increased lengths of time while waiting for fingerprint checks. This also would require increased detention space at Border Patrol stations. The INS Associate Commissioner for OIRM also told the OIG that the INS had no intention of taking ten-prints because the INS intended to use the FBI’s IAFIS, when it became operational, to meet any INS need for checking ten-prints. He said the INS could not justify spending money to duplicate IAFIS’ ten-print capability.

2. Deployment of IDENT

Between 1991 and 1993, the INS operated a pilot AFIS project in the San Diego Border Patrol Sector. In 1994, Congress allocated funding for IDENT.

The INS discussed its plans to deploy IDENT at a meeting between the INS and the FBI in April 1994. At the meeting, the INS pointed out its need to handle a high volume of fingerprints (for approximately 1.5 million aliens apprehended annually) and its need for quick response times (in two minutes or less) for each encounter. The FBI responded that the INS requirements could not be handled by the FBI’s planned IAFIS without additional development time and money, which the FBI did not have. The FBI also stated that it had a wider constituency that required the ten-fingerprint system necessary for criminal investigations.

The INS asked whether the FBI would consider searching and matching the two fingerprints captured by IDENT against the FBI’s planned IAFIS database of ten-prints. The FBI believed that this was problematic and would require much greater computer power than it had in order to provide the response time INS needed. The FBI suggested that the INS should make funding available to the FBI to pay for any changes to IAFIS that would have to be made to meet INS requirements. The INS believed that the FBI wanted too much money to modify IAFIS to search the INS two-prints against the IAFIS ten-print database.

An FBI official involved with these integration issues told the OIG that the FBI was frustrated in its efforts to get from the INS all its requirements for its automated identification system. He also said the FBI was never comfortable attempting to match the INS two-prints against the IAFIS ten-print database. In addition to the volume and response time requirements, accuracy of results was a concern to the FBI. In addition, the FBI did not have the funding to provide a specialized AFIS system for each law enforcement user. FBI officials said they expected the INS to pay for any modifications in IAFIS to accommodate the INS’s needs.

In 1994, the development of IDENT by the INS was ahead of the development of the FBI’s IAFIS, which was behind schedule and not likely to be implemented any time soon. INS officials were not certain that IAFIS was going to meet the INS needs. The INS therefore decided to move forward with implementation of IDENT in 1994, first in the San Diego Border Patrol Sector and later along the Southwest Border. The INS’s project plan for IDENT described the reasons that the INS decided to implement IDENT independently from the FBI’s IAFIS. It stated that the IAFIS design

has not addressed the INS’ functional requirements to conduct quick lookouts for 1.5 million arrestees per year against a database of about 450,000 lookouts . . . The FBI IAFIS would need substantial re-engineering to accommodate the INS requirements and still achieve its original system performance objectives . . . For most encounters, the INS requires a fast and simple fingerprint collection technique, such as pressed index fingerprints not 10 rolled prints, for quick lookups. Finally, IAFIS is not yet available and the INS needs an identification solution now.

3. JMD’s 1995 Report

Around this time, the Department of Justice’s JMD reviewed the plans for the various AFIS fingerprint systems, including IDENT, the FBI’s IAFIS, and NCIC 2000. A February 1995 paper prepared by JMD’s Systems Policy Staff summarized the capabilities of these three fingerprint identification initiatives. The paper identified the different operational requirements that led the INS to favor the two-fingerprint system and the FBI to favor a ten-print system. It stated that the INS system had to have databases of sufficient capacity (450,000 for the lookout database, 1.5 million for the recidivist database) and had to be able to handle a high volume of checks (up to 8,000 a day) with a two-minute response time for checking both databases.

The JMD report concluded that the three systems complemented each other and were not duplicative. It also stated that the INS would collect and store all fingerprint records in an IAFIS-compatible format and resolution, and that when IAFIS became operational the INS should be able to electronically submit fingerprint records to IAFIS. The report suggested that neither IAFIS nor NCIC 2000 could meet the INS’s operational requirements, nor could the FBI afford the cost to reconfigure the systems to integrate INS records.

However, INS officials told us that after the INS had deployed IDENT, it was convinced that the FBI was no longer interested in working out a means of matching the two fingerprints taken by IDENT against the FBI’s planned IAFIS. By contrast, FBI officials told us they were convinced that INS did not want to integrate IDENT with IAFIS. They suggested that the INS was fearful that integration would result in the FBI taking over IDENT and not being responsive to INS needs. As a result, further efforts to integrate IDENT with IAFIS made no progress.

Stephen Colgate, the Assistant Attorney General for Administration, described for the OIG three main obstacles to integration of IDENT and IAFIS. First, the INS and the FBI, the second and third largest components in the Department, each had its own parochial agenda and made little effort to understand the operational requirements of the other. Second, there was insufficient money for the development of the different AFIS projects, and when money was forthcoming each agency focused on meeting its own requirements. Third, the further each agency progressed toward the respective goals of their own automated fingerprint identification systems, the harder it was to back up to accommodate the needs of integration.

Colgate also told the OIG that he believed that the FBI’s delay and over-promising on IAFIS made the INS rightfully skeptical about whether the FBI could support INS fingerprint needs.64 On the other hand, Colgate believed the INS may have over-promised Congress about IDENT’s integration with the FBI’s systems.65

In 1997, controversy arose regarding the way fingerprint checks were handled for aliens applying for citizenship under the INS’s Citizenship USA (CUSA) program. During CUSA, the INS was sending ten-print fingerprint cards of citizenship applicants to the FBI to be checked for evidence of criminal histories in the FBI Criminal Master File. If the INS did not receive a response after a certain period of time (60 days), the INS assumed that the alien had no criminal record. This practice led to many instances in which aliens who had a disqualifying criminal record nevertheless received citizenship.

In 1997, the Attorney General established the Fingerprint Coordination Working Group, chaired by the Assistant Attorney General for Administration. The Working Group, which included representatives from the INS and the FBI, was charged with resolving problems that were identified in fingerprint procedures between INS and the FBI. As part of this effort, JMD reviewed issues of integration of fingerprint systems and attempted to obtain better cooperation between the INS and the FBI.

4. JMD’s 1998 Report

On February 2, 1998, Congressman Alan Mollohan, the ranking member of the House Appropriations Subcommittee for Commerce, Justice, State, and Judiciary, sent a letter to the Attorney General questioning whether IAFIS and IDENT were redundant. The letter stated that “major duplication” exists between the two systems and that consideration should be given to integrating IDENT and IAFIS.66

In response to Mollohan’s letter, JMD’s Management and Planning Staff (MPS) was directed to review IDENT and IAFIS and report on the feasibility of converting IDENT to a ten-fingerprint system. After reviewing IDENT operations and interviewing INS and FBI employees, JMD issued a report on May 28, 1998, which concluded that IDENT and IAFIS were not redundant and that IDENT satisfied an important INS need. The report stated that converting IDENT to a ten-print system was “untenable in the current operating environment” because it would add significantly to the time it took to process aliens and would require additional INS equipment, personnel, and space.

The report proposed three options. The first option was to retain IDENT as a two-print system but improve its policies and procedures, including the enrollment of aliens and lookouts in IDENT. The second option was to adopt ten-printing as a long-term policy goal and retain IDENT as a system designed to meet INS’s internal requirements. Under this option, IDENT and IAFIS would both operate in tandem, but strategies for the INS taking ten-prints would be explored “when it makes economic and operational sense to do so.” The third option was to terminate IDENT and move to a full ten-print system that would be part of IAFIS.

The JMD report recommended the second option. It stated that although the second option was the costliest of the three, it

is the unanimous choice of those who have been involved in the IDENT/IAFIS fingerprint issue, i.e., JMD, the FBI, the Border Patrol, and the INS. Properly funded, the option will permit the Border Patrol to maintain its current processing times while providing other law enforcement agencies with a voluminous fingerprint database that can be searched to solve crimes committed in the communities. At the same time, by retaining IDENT, the Border Patrol is able to capitalize on the benefits that system has to provide as an intelligence-gathering and investigative tool. Moreover, the rest of the INS will be able to continue its plans to integrate IDENT with other internal functions unique to the Service.

As part of the second option, the JMD report recommended a nationwide 12-month pilot study, to begin in the fall of 1999, of a ten-print system in selected Border Patrol stations to determine the impact of taking ten-prints on Border Patrol operations, to evaluate IAFIS’s ability to respond in a timely manner to fingerprint check requests made by the Border Patrol, and to study recidivist populations along the border to find out to what extent it included aliens engaged in criminal activity. The estimated cost of the pilot study was approximately $18 million, $15 million of which would be for system changes to IAFIS necessary to ensure a two-minute response time. The report estimated that it would take five years to implement the second option. On June 25, 1998, the Attorney General approved the second option.

However, Congress did not approve the funding for this option in the fiscal year 1999 budget. As a result, in April 1999 JMD proposed a more limited pilot study that would cost $500,000 and would be implemented in a smaller number of Border Patrol stations. No action was taken on this proposal.

5. Current Status of Integration

The FBI’s NCIC 2000 became operational on July 11, 1999, and IAFIS became operational on July 28, 1999. IDENT is not linked to either system.

IAFIS is an automated ten-fingerprint matching system. It contains in its Criminal Master File over 40 million ten-print fingerprint records. IAFIS is connected electronically with all 50 states and some federal agencies, including the INS. The IAFIS records can be electronically compared against submitted fingerprints and can provide a response within two hours for a limited number of high-priority requests.

NCIC 2000 allows law enforcement agents across the United States to quickly search 20 NCIC databases containing over 40 million records when investigating crimes or questioning suspects. The databases include wanted persons, stolen vehicles, deported felons, gang and terrorist members, and missing persons. In addition, in a matter of minutes, NCIC 2000 allows agents in the field in a properly equipped police vehicle to transmit a single fingerprint that is searched against the NCIC wanted, missing, and unidentified person files. NCIC 2000 also lets officers view mug shots to help confirm a suspect’s identity.

The controversy over the Resendez case made clear the need for the INS to be able to check apprehended aliens against FBI records and receive quick responses, as well as for the FBI state and local law enforcement agencies to obtain access to alien fingerprints collected by INS.

In July 1999, a House Committee on Appropriations report stated, “[T]he Committee continues to be concerned about the inadequacies of this system [IDENT], specifically with regard to its ability to identify wanted criminals who may be apprehended by INS Border Patrol agents and inspectors along the border . . . the Committee repeatedly raised concerns that the IDENT database was not integrated with FBI’s IAFIS database.” The House report also expressed the belief that federal, state, and local law enforcement should have access to INS fingerprint information and that the INS should have the full benefit of FBI criminal history records. The House report directed that the INS suspend further deployment of IDENT until the Department of Justice submitted to the Committee a plan for integration of IDENT and IAFIS. The conference report for the Department of Justice’s fiscal year 2000 appropriations included this provision.

In response to the congressional directive, JMD convened a “Fingerprint Summit” meeting on August 12, 1999, attended by FBI, INS, and JMD representatives, to discuss a plan for integrating IDENT with IAFIS. JMD’s Management Planning Staff (MPS) is coordinating the efforts to develop the integration plan. After meetings and discussions among senior managers from the various components and technical specialists, MPS issued a report on March 1, 2000, entitled; “Implementation Plan for Integrating the INS’s IDENT and the FBI’s IAFIS Fingerprint Data.”

According to the March 1 report, “The INS and the FBI have acknowledged that integrating IDENT with IAFIS would greatly benefit both agencies, as well as Federal, state and local law enforcement.” The report further states that such integration

has the potential to: reduce the likelihood that a wanted individual would be released from the INS’s custody; provide Federal, state, and local law enforcement an integrated picture of the criminal activity known by agencies in DOJ, including the INS histories of illegal border crossers; and enable Federal, state and local law enforcement to search latent prints against additional illegal border crossers, especially if ten-prints are taken.

The March 1 report therefore discussed various options for integration. In considering these options, the report raised four main questions: 1) how many fingers would be printed and used in the INS search; 2) would a separate database parallel with IAFIS be established to handle the heavy INS workload or would the current IAFIS be enhanced to accommodate the increase in workload; 3) how quickly would the new system be deployed; and 4) which databases would the new system search?

Each of the options considered assumed the following conditions were required: a quick response time to INS fingerprint searches, similar to the response time in IDENT; responses would be a computer-generated hit/no hit response; the INS would submit a maximum of 1,050 searches per hour with an average of 18,000 a day; the INS would remain responsible for obtaining fingerprints from aliens and maintaining its communication links; and IDENT would continue operating until an integrated system was implemented.

The March 1 report recommended one of the options as a “conceptual model” for integrating IDENT and IAFIS over several years. The report stated that the option chosen was less expensive, offered earlier operational capability, and early identification of potential problems. The report also stated that considerable work, including addressing technical and operational complexities, still needed to be done before a final integration proposal could be presented. According to the report, the proposed conceptual plan could take up to five years and cost more than $200 million to implement fully.

A key element of the proposed conceptual model was that IAFIS would eventually be supplemented by adding fingerprint files maintained by IDENT, including its lookout and recidivist databases. Also, Federal, state, and local law enforcement agencies would be able to access the fingerprint files through IAFIS, and the INS would be able to check fingerprints of apprehended aliens against all files in IAFIS.

Before finalizing any recommendation for the integration of IDENT and IAFIS, however, the report states that in fiscal year 2000 the INS, FBI, and JMD will conduct the following three studies:

A study to assess criminality, which involves matching a sample of current INS recidivist records against the FBI’s Criminal Master File in IAFIS. The matching will help estimate the percentage of apprehended illegal border crossers that have been charged with more serious crimes. It will also help quantify the risks of the current system and the benefits from improvement.

An engineering study that will develop integrated systems requirements, define the system architecture, and produce a cost analysis for the development and implementation of an integrated system. The study will include consideration of the effects of alternative query response times.

An operational impact study to assess the effect of the new system on the operations and procedures at the border, focusing primarily on the impact of taking ten prints of individuals at these locations, as well as the effects of alternative query response times.

If, as a result of these studies, the conceptual model proposed by the March 1 report is approved, the integration of IDENT and IAFIS would occur in stages. The first stage would result in the creation and installation of an interim system that would allow the INS to search aliens’ two fingerprints against a new, two million-subject FBI Federal Lookout File (FLF), which would be a subset of the FBI’s much larger Criminal Master File. The FBI would create the FLF file and maintain it in Clarksburg, West Virginia, where IAFIS is located. The FLF would include individuals from the Criminal Master File who have outstanding warrants or have “an INS arrest cycle.”

The second stage would integrate the interim FLF and the INS’s recidivist database into IAFIS. IDENT recidivist files would be entered into IAFIS and searches would be made against the recidivist files and the FLF, all of which would be maintained in Clarksburg, West Virginia.

According to the report, the third stage would be the full integration of IDENT with IAFIS. The integrated system would permit the search of two fingerprints of aliens against the recidivist files and the full IAFIS Criminal Master File.

According to the conceptual model, the Department might also conclude as a result of its studies that the INS should ten-print all apprehended aliens. The report states that the proposed new system would have the flexibility to allow the INS to take and submit ten-prints for all apprehended aliens, if warranted, and if sufficient INS personnel, resources, and space are available at border stations and ports of entry.

The Department forwarded the March 1 report to Congress on Monday, March 6, 2000.

B. Conclusions

We agree with the assessment that it is critical for IDENT and IAFIS to be integrated. We also believe that it was reasonable for the INS in 1994 to move forward with the IDENT as a separate system. The INS had a congressional mandate and the funding to implement an automated fingerprint identification system to identify and track recidivist and criminal aliens. IDENT served an important and useful function for the INS. At the time, the FBI could not meet the INS requirements with the available funding, and the FBI was behind schedule getting IAFIS operational.

Our interviews and review of documents revealed that although there were periodic meetings about integrating IDENT and IAFIS, there was never a sustained effort to achieve that goal. Although controversies such as CUSA generated flurries of activity, the plans for integration were not achieved. As a result, even though IAFIS is now operational, the integration of IDENT and IAFIS is still years away.

The Resendez case clearly demonstrates the pressing need for such integration. A detailed evaluation of the merits of JMD’s proposed “conceptual model,” which was just issued this month, or whether changes should be made to it, is beyond the scope of this review. What is clear, however, is that the plan is still in the developmental stages and will take years to accomplish. Our review of the Resendez case also makes clear that it is critical that integration of IDENT and IAFIS be pursued in an expeditious and sustained fashion, even after the controversy about the Resendez case recedes. We believe that Department of Justice managers must continue to provide leadership on this issue to ensure that any differences between the FBI and the INS about integration of their automated fingerprint systems are overcome.


XI. Chapter Eleven: OIG Recommendations

In this chapter, we provide our recommendations to address the systemic problems we found as the result of this review. The recommendations fall into four categories: 1) the operation of IDENT; 2) increased training and education about IDENT; 3) integration of IDENT with other criminal and INS databases; and 4) other recommendations.

A. The Operation of IDENT

1. The INS needs to improve its procedures to ensure that all aliens who fit within the criteria of the IDENT lookout policy are in fact entered into the lookout database.

2. We found that some INS offices sent in few or no records to the BSC for entry into the lookout database and others sent in many records that did not qualify for entry into the lookout database. The INS should develop an oversight mechanism to ensure that records for aliens who fit the lookout criteria are submitted to the BSC for entry into the lookout database, and records that do not fit those criteria are not submitted.

3. The INS should review the current records in the IDENT lookout database and eliminate those that do not belong there. Many aliens who do not fit the criteria of the INS’s lookout policy have nevertheless been entered into the lookout database.

4. The processing of aliens in IDENT by INS agents needs to be improved. For example, we found that Border Patrol agents made various mistakes in how they enrolled Resendez in IDENT, including allowing IDENT to auto verify previous apprehensions; denying valid matches because the agents were overly cautious or afraid to mistakenly confirm a match; failing to view all previous apprehensions that IDENT provided on Resendez; and enrolling Resendez in IDENT under other agents’ passwords. These types of deficiencies should be corrected.

5. The INS should include compliance with IDENT policies as part of the Office of Internal Audit’s INSpect reviews, including its reviews of INS program offices outside the Border Patrol.

6. The INS should assign one person or entity within INS the central responsibility for IDENT policies and the responsibility to resolve IDENT issues that affect more than one INS component. IDENT affects many INS offices that report to different managers. While OIRM has technical expertise in the operation of IDENT, it is not equipped to address policy issues relating to it. The INS should assign one central authority, outside OIRM, with the responsibility for ensuring that the appropriate components within INS are using IDENT and that the INS achieves the maximum benefits from this important system.

B. Training and Education on IDENT

7. IDENT training for INS employees needs to be improved. Contrary to the INS’s claims in response to our 1998 Inspection report, we found a widespread lack of knowledge throughout the INS about IDENT and its operation. This is particularly true in program offices outside the Border Patrol. Moreover, the “train the trainer” approach used by the INS has not been effective. High turnover of agents has resulted in few current agents having received direct IDENT training. We believe the INS should implement better methods of training for IDENT. For example, the INS should consider requiring INS employees to complete computerized tutorial training on IDENT tailored to each relevant INS component. As part of this tutorial training, the INS should include a test to determine whether the agent understands IDENT’s important concepts.

8. The INS should emphasize in training what lookouts and alerts are, how to enter them, and what agents should do if they see a lookout or alert when processing an alien.

9. The INS should provide training in IDENT to Intelligence personnel (in the Intelligence, Investigations, Border Patrol, or Inspections offices) who may have a role in analyzing data or assisting with the apprehension of criminal aliens.

10. The INS should tailor its IDENT training to the needs of different components. For example, training in Investigations, Detention and Deportation, and Intelligence should emphasize the standards and procedures for entering lookouts in IDENT and stress how IDENT can enable each component to perform its tasks more effectively.

C. Integration of IDENT with Other Criminal and INS Databases

11. We believe that IDENT and the FBI’s IAFIS should be integrated as soon as possible. IDENT serves an important function in identifying and tracking recidivist illegal aliens who are repeatedly apprehended in the United States. But the INS’s lookout database of criminal aliens is incomplete. Although the recommendations we make in this chapter may correct some of the problems we found, the IDENT lookout database, which relies upon INS employees to populate it with selected aliens, will inevitably contain gaps. Technology should be used to effectively provide what the IDENT lookout database has not provided: a check of all apprehended aliens to determine if they have serious criminal records, prior orders of deportation, or any outstanding arrest warrants. We believe that the INS and the FBI should expeditiously and aggressively pursue integration of IDENT with the FBI’s IAFIS.

12. In a report issued on March 1, 2000, the Department’s Justice Management Division describes the plans it is developing for integrating IDENT with IAFIS. A detailed analysis of the merits of the specific proposals in this plan is beyond the scope of this review. However, we believe it critical for the Department of Justice to pursue plans for integration of IDENT and IAFIS as rapidly as feasible and to seek adequate funding to implement a comprehensive plan for integration.

13. The Department of Justice should designate one entity as ultimately responsible for the integration of IDENT with IAFIS. Without such a designated entity, we believe that the necessary direction and oversight to ensure timely completion of the project will be lacking. A likely candidate for this assignment is the Department’s Justice Management Division. Whatever entity is chosen, it should convene regular meetings, require status reports, and frequently consult with the various entities involved to ensure that any obstacles to integration are overcome.

14. The INS should study the feasibility and costs and benefits of linking the information in IDENT to state automated fingerprint identification systems.

15. The INS should study the feasibility and costs and benefits of linking the information in IDENT to other INS text-based systems.

D. Other Recommendations

16. We found wide disparities in the “threshold” number of apprehensions required before an alien is prosecuted for entry without inspection. In some jurisdictions, the threshold is very high (such as Redacted in Santa Teresa, where Resendez was apprehended on June 1, 1999). Also, because IDENT does not always recognize previous apprehensions and not all apprehensions are enrolled in IDENT, the actual number of apprehensions before an alien is prosecuted may be even higher. We believe that the INS and the U.S. Attorneys’ Offices should consider lowering high thresholds that exist in certain jurisdictions. While that action may involve additional costs, it could increase the deterrent effect on some aliens who repeatedly attempt to enter the United States.

17. The INS and United States Attorneys’ Offices should consider modifying their policies and submit aliens for prosecution for entry without inspection if the alien is apprehended repeatedly within a short time frame, even if the alien has not yet reached the normal prosecution threshold.

18. We believe that the wide disparities in threshold levels should also be reviewed. It seems inequitable that U.S. Attorneys’ Offices treat similar conduct so differently. We also found wide differences in the numbers of misdemeanor and felony immigration cases that are brought by U.S. Attorneys’ Offices along the Southwest Border. Some of these differences may be attributable to varying levels of resources devoted to the prosecution of immigration-related offenses. We recommend that the Department of Justice, including the INS and the Executive Office for United States Attorneys, review these differences and determine whether more uniform policies on the prosecution of immigration offenses should be encouraged.

19. We found that some Border Patrol agents did not know the applicable prosecution thresholds. The INS should educate its agents about the prosecution thresholds and direct that they be followed.

20. The INS and U.S. Attorneys’ Offices should attempt to implement the practice used in the Del Rio Sector where Border Patrol Prosecution Officers represent the government in misdemeanor entry without inspection cases and perform critical tasks necessary for preparing felony immigration cases.

21. The INS should improve and standardize how intelligence officers communicate information and wanted posters to other INS entities regarding dangerous or wanted aliens such as Resendez. The INS should establish standard procedures for intelligence officers to use when they need to notify other INS employees, such as Border Patrol agents, of a dangerous or wanted alien.

22. During our review, we had difficulty obtaining timely and complete responses about IDENT from the INS’s OIRM. OIRM and contractor records regarding IDENT appeared scattered, disorganized, and incomplete. We recommend that OIRM make better efforts to organize the records regarding IDENT, particularly records kept by contractors, so that information related to IDENT can be better accessed. We also recommend that when information is sought from OIRM in the future, it designate one person as responsible for full and timely responses, rather than allowing that responsibility to be passed from person to person.

23. ENFORCE is used regularly in some Border Patrol sectors but only sporadically or not at all in other sectors where it also has been deployed. The INS should require consistent use of ENFORCE once it has been deployed.

24. The INS should make a concerted effort to notify other federal, state, and local law enforcement agencies of the existence and possible uses of IDENT to aid in the apprehension of dangerous and wanted criminal aliens. Virtually every law enforcement officer to whom we spoke knew nothing about IDENT or its existence prior to the Resendez case. All said IDENT would have helped in their investigation of Resendez. Although the Resendez case publicized IDENT widely, the INS should take affirmative steps to let other agencies know about IDENT and how they can use it to alert the INS about aliens who are wanted for serious criminal offenses. The INS notes that if IDENT is more widely used by these agencies, there may be additional costs to the INS, such as those costs associated with entering more records in the IDENT lookout database and the additional costs involved with detaining aliens wanted for state and federal criminal offenses. We believe the benefits of detaining serious criminal offenders outweigh these added costs.

25. The INS and the FBI should work together to revise the FBI’s lookout forms for wanted aliens. Currently, the FBI’s only lookout form is for NAILS, a database primarily used at ports of entry to check travelers seeking admission to the United States. The FBI’s form is sent to INS Headquarters Inspections rather than directly to the Inspections Lookout Unit, which enters lookouts for aliens in NAILS. The INS should work with the FBI to create another FBI form for IDENT lookouts. Moreover, the INS should request that the FBI revise its NAILS lookout form so that it is sent to the Inspections Lookout Unit directly.


XII. Chapter Eleven: OIG Conclusions

Despite Resendez’s extensive criminal record and history of prior deportations, he was voluntarily returned to Mexico by the Border Patrol seven times in 1998 and again on June 1, 1999, even after federal and state warrants for his arrest had been issued. After June 1, he allegedly re-entered the United States and committed four more murders. When this shocking chain of events was discovered, it generated serious questions about the actions of the INS employees who were involved in the case as well as the operation of IDENT, the INS database system that was supposed to identify recidivist and criminal aliens.

Because of the importance of this matter, the OIG expended significant resources conducting a detailed review of the actions of INS employees in connection with the Resendez case. We also reviewed the operation of IDENT to ascertain how the problems in the Resendez case occurred.

Our review determined that several INS investigators were contacted early in 1999 by law enforcement officers who were seeking Resendez in connection with the brutal murder of Dr. Benton on December 16, 1998. These INS investigators were asked for help placing a border lookout for Resendez. One investigator was specifically asked to place a “hold” on Resendez in the INS’s “internal system.” Yet, none of these investigators – or any of their supervisors or co-workers to whom they spoke about the case – thought to place a lookout for Resendez in IDENT. Two of the INS investigators instead suggested to law enforcement officers seeking Resendez that a lookout be placed for him in a database maintained by the Customs Service. But no one even mentioned IDENT to the officers seeking Resendez. Also, when several INS intelligence officers learned about the case, they suggested sending a wanted poster to INS employees along the border, but made no mention of placing a lookout in IDENT (and they failed to follow through with the idea to send the wanted poster).

Because no lookout was entered in IDENT for Resendez, the Border Patrol agents who apprehended him on June 1, 1999, did not know that he was wanted in connection with the murders. They therefore voluntarily returned him to Mexico. We concluded that the Border Patrol agents who returned Resendez to Mexico on June 1, 1999, as well as the agents who returned him seven times in 1998, did so in accord with standard Border Patrol procedures. Although these Border Patrol agents made some mistakes in how they processed Resendez in IDENT, their mistakes did not affect the decision to return Resendez to Mexico. Rather, the most egregious deficiency we found in this matter was the failure of the INS agents who were contacted by the law enforcement officers seeking to arrest Resendez to consider IDENT as an option.

Yet, through our interviews, we found that their failure to think of IDENT was largely caused by broader problems in the way that the INS implemented and oversaw this database. Although the INS proclaimed IDENT’s benefits in identifying criminal aliens, and identified IDENT to Congress as a system that would be used for that purpose, the INS took insufficient steps to ensure that IDENT was understood and used effectively to identify criminal aliens.

In 1998, the OIG had reviewed the operation of IDENT and made recommendations regarding problems we found in its implementation and operation. During this review of the Resendez case, we found that some problems we identified in the 1998 report were not sufficiently addressed. We saw that IDENT was still predominantly viewed by INS employees as a Border Patrol tool to identify and count recidivists, rather than a tool to be used throughout the INS. We found a surprising and disturbing lack of knowledge about IDENT lookouts or how to enter them. We found that training on IDENT was ineffective. Particularly outside the Border Patrol, very few INS employees to whom we spoke understood IDENT or the lookout policy. Despite periodic directives from INS Headquarters, INS supervisors failed to ensure that lookouts for criminal or deported aliens were consistently entered in IDENT. And INS Headquarters did not follow up on these directives and ensure that the lookout policy was understood and followed.

We were told by many INS employees, and we agree, that IDENT is a valuable system used to identify and track recidivist border crossers. Before IDENT, it was impossible for the INS to ascertain the true identity of the many aliens it apprehended each day or for the INS to determine how often these aliens had been previously detained. However, many INS employees also told us that it is critical that IDENT be linked to FBI criminal databases so that criminal aliens can be identified. We also agree. We think the Department should aggressively and expeditiously pursue integration of IDENT with the FBI’s IAFIS. In the interim, the INS should take specific measures, such as the ones we recommended in the previous chapter, to improve the operation and usefulness of IDENT. While these measures will no doubt be expensive, they are critical steps that could help prevent tragedies similar to the Resendez case.


March 20, 2000 ___________________
Robert L. Ashbaugh
Acting Inspector General


Glenn A. Fine
Special Investigations and
  Review Unit


Principal Contributors
Office of the Inspector General


Stephen H. Fallowfield
Adrienne M. Flave
Diana L. Gordon
Peter M. Martz
Eugenio Pedraza
Gregory L. Tyler


64 In 1993, IAFIS was originally budgeted to cost $520 million but, according to the FBI, ultimately cost $640 million to develop.

65 For example, in a May 8, 1996, hearing of the House Appropriations Subcommittee on Commerce, Justice and State, the Judiciary, and Related Agencies, Congressman Alan Mollohan asked INS Associate Commissioner for OIRM Ronald Collison if the INS “database will be able to be shared and it will be integrated with the FBI’s.” Collison replied, “It is sharable. The FBI and INS systems are designed so that a transaction coming into their system, if permissible legally, could check our database and vice versa. So, there will be full compatibility to move the actual fingerprint images between the three systems.” In testimony before the same committee on the INS’s fiscal year 1998 and 1999 budget requests, INS Commissioner Meissner stated that IDENT served INS mission purposes well, noted that the FBI automated fingerprint identification systems were not yet operational, and insisted that INS was working toward ensuring that IDENT would be compatible with them.

66 These issues were also raised in hearings before the House Appropriations Subcommittee from 1996 to 1999.