NEXT STEPS: MUCH NEEDS TO BE DONE TO ENSURE IDENT'S SUCCESS

INS has succeeded in demonstrating the potential value of using fingerprints to identify aliens apprehended at the border. In San Diego, where the United States Attorney for the Southern District of California is using IDENT to focus prosecutorial resources on criminal aliens, IDENT is proving to be a useful tool. While its deterrent effect cannot be clearly demonstrated, the system has been used successfully in Southern California to identify and prosecute aggravated felons, and drug and alien smugglers.

However, much needs to be done if IDENT is to be used consistently along the Southwest border to identify the most dangerous criminal aliens, and help make the most effective use of prosecutorial resources.

The Border Patrol must develop and implement strategies in each sector to enroll all apprehended aliens in IDENT.

INS must ensure that all criminal and previously deported aliens are placed in the IDENT lookout database.

The United States Attorneys need to better understand how IDENT works, and begin to make consistent use of it as part of a broader border enforcement strategy.

Use of IDENT by other enforcement and benefits programs will pose even greater challenges for INS. The integration of IDENT and biometrics into INS programs such as benefits processing, inspections, detention and deportation, and investigations has yet to be accomplished. Linking IDENT with other Federal and state law enforcement and benefit systems will also require considerable INS resources. Although we are limiting our recommendations to INS' current use of IDENT as a border enforcement tool, we strongly urge INS to proceed with ongoing efforts to expand its use of fingerprints in other aspects of its work.


RECOMMENDATIONS

The Inspections Division recommends that the Commissioner, Immigration and Naturalization Service:

1. Amend the existing INS Servicewide Fingerprint Policy to establish explicit procedures and criteria for placing individuals in the IDENT lookout database.

2. Develop and implement a plan to ensure that the fingerprints of criminal and deported aliens are consistently entered into the IDENT lookout database.

3. Develop and implement a plan to ensure that all IDENT lookout records contain aliens' photographs to assist arresting agents in identifying them.

4. Implement input controls for critical data entered into the IDENT lookout database to ensure the accuracy of that data.

5. Identify, and to the extent possible, correct invalid data in the IDENT lookout database.

6. Develop and implement a strategy for consolidating multiple FINs that exist for individual aliens.

7. Develop and implement a strategy for sufficiently training all INS personnel using IDENT (e.g., making IDENT a part of an INS data systems training module at the INS training academies).

8. Direct that Chief Border Patrol Agents in each sector work with the United States Attorneys to develop and implement a plan to use IDENT to identify and refer the most egregious offenders for prosecution.

 


Appendix I

INSPECTION METHODOLOGY

Our inspection included interviews with officials from various divisions of the INS Office of Information Resources Management, including the IDENT Project Manager, as well as headquarters and field personnel from investigations, examinations, inspections, and detention and deportation. We solicited opinions and information from INS contractors, Biometrics Users Group members, and INS Statistics Division personnel.

In order to observe how IDENT was being used, we visited four Border Patrol Sectors along the Southwest border: San Diego, Tucson, El Paso and McAllen. In addition to the Sector headquarters offices, we visited 27 Border Patrol stations and 10 checkpoints, or about 70 percent of all Border Patrol locations within those sectors. In each Sector, we toured the border with agents, observed alien processing during the day and at night, and discussed IDENT with Border Patrol agents, supervisors, and managers.

To discuss future and planned uses of IDENT, we visited four district offices in the Southwest region: San Diego, Phoenix, El Paso and Harlingen, as well as the Western Regional Office. At the District Offices we met the District Directors, when available, and interviewed the Assistant District Directors for Examinations, Investigations, Deportation, and Inspections. In these districts, we visited 11 ports of entry and 4 detention facilities.

In addition to visiting INS offices, we used financial and IDENT deployment data provided by INS to assess INS' progress in implementing IDENT. We also performed analyses of IDENT enrollments, the IDENT lookout database, and IDENT lookout matches.

In assessing INS' use of IDENT, we obtained data from INS on apprehensions and IDENT enrollments and performed analysis at both the INS sector level and calculated total enrollment statistics for the nine Southwest border sectors.

To determine whether INS was placing criminal and previously deported aliens in the IDENT lookout database, we obtained from INS the entire lookout database from its inception. We then obtained DACS records for all aliens deported or excluded during FY 1996 and matched these records against the lookout database. We also selected a sample of 498 FY 1996 DACS records out of a population of 66,744 deported or excluded aliens listed. We used this sample to determine whether fingerprint cards for these removals were sent by INS field offices to the INS Fingerprint Storage Facility (FSF), and whether the FSF was entering these alien fingerprint records into the lookout database.

We analyzed the entire IDENT lookout database to assess the integrity of the lookout data, and analyzed compliance by INS field offices and Regions with the requirement that fingerprint cards be submitted as required by the INS Fingerprint Policy.

We analyzed a random sample of 495 aliens out of a population of 8,025 matched by IDENT to individuals previously entered in the lookout database. We subsequently determined that only 371 aliens in our sample of 495 had in fact been matched to the IDENT lookout database. The remaining 124 aliens in our sample, due to a programming error by INS contractor personnel, were erroneously included in the population from which we drew our sample. We used this sample of 371 aliens to generate descriptive statistics about the composition of the database, and determine the locations at which these matches occurred.

We used prosecution data obtained from the United States Attorney for the Southern District of California to assess whether IDENT lookout matches were leading to prosecutions. We found that 46 (or 9 percent) of the aliens in our sample of 495 had been prosecuted by the United States Attorney for the Southern District of California. Based upon this sample result, we projected that a total of 745 (or 9 percent) out of our population of 8,025 aliens were prosecuted by this United States Attorney.

We spoke with Assistant United States Attorneys in San Diego and El Paso to determine whether they were familiar with IDENT, and if so, what role it played in prosecutions of aliens. By telephone, we contacted Assistant United States Attorneys in Tucson, Arizona; McAllen, Texas; and Brownsville, Texas.

We also conducted fingerprint tests of IDENT at most of the locations we visited. The 5 inspectors on our team were enrolled in IDENT using their own fingerprints and a mock photograph at a total of 28 sites. This included 147 IDENT enrollments. We used this test to assess the reliability with which IDENT matches individuals to the IDENT databases.

 


Appendix II

PERCENT OF APPREHENSIONS ENROLLED IN IDENT

Attached are charts that show IDENT enrollments as a percentage of apprehensions for each of the nine Southwest Border Patrol sectors from September 1996 through July 1997. The data for each Sector includes all stations where IDENT has been installed.

Enrollment calculations are based upon data provided by the INS Statistics Division. We did not independently validate this data.

 

SAN DIEGO

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


YUMA

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


TUCSON

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


LAREDO

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


DEL RIO

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


MARFA

Values of over 100% for April and June were apparently the result of enrollments credited erroneously to the Marfa Sector. The INS Statistics Division is working with IRM and Field Operations to solve the problem.

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


EL CENTRO

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


EL PASO

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


MCALLEN

SOURCE:  IMMIGRATION AND NATURALIZATION SERVICE, STATISTICS DIVISION

 


 

Appendix III


U.S. Department of Justice
Immigration and Naturalization Service
Office of the Commissioner

425 I Street NW
Washington, DC 20536


MAR 23 1998

MEMORANDUM FOR  MARY W. DEMORY
                                        ASSISTANT INSPECTOR GENERAL FOR INSPECTIONS
                                        DEPARTMENT OF JUSTICE

FROM:          Doris Meissner
                      Commissioner
                      Immigration and Naturalization Service

SUBJECT:     Draft Inspection Report: INS' Automated Biometric Identification
                      System (IDENT); Assignment Number A-97-04

We appreciate your providing us the opportunity to comment on the subject report. Our request that your office review IDENT was based on our need to have an independent assessment of this critical operations subsystem. We agree with your recommendations and overall conclusions about what we need to do if IDENT is to be used consistently along the Southwest border. Your report has provided us with information that will help us in our efforts to continuously improve operations. I have attached comments from the Associate Commissioner for Programs and the Executive Associate Commissioners for Management and Field Operations that address each of the eight recommendations in the report.

The report raised some valid concerns about IDENT that we, too, shared as the review was ongoing. You reported that INS did not achieve its FY 1997 goal to process 85 percent of apprehended aliens along the Southwest border through IDENT. By the fourth quarter FY 1997, we had processed about 57 percent of the aliens apprehended through IDENT. Factors contributing to not meeting the goal included hardware problems and staffing constraints at some of our Sectors that faced an overwhelming increase in apprehensions. Our goal for FY 1998 is 88 percent for all sites that have IDENT. For the first quarter FY 1998, the usage rate had climbed to almost 75 percent for apprehended aliens, three Sectors having exceeded 85 percent. The January 1998 data shows the usage rate has increased to almost 80 percent. Processing of apprehended aliens through IDENT was a Commissioner's Priority in FY 1997, and it continues to be a priority in FY 1998. Both Headquarters and the field share these goals, because IDENT has consistently demonstrated to our agents how useful it can be in their operations.

Our Border Patrol Agents are becoming more comfortable with using IDENT in day-to-day operations. The last Southwest Border Patrol Station received IDENT in February 1998. Our response to recommendation 7 concerning a training strategy provides our plan to incorporate systems into the course work at the academies. Follow-on training will be achieved through our "distance learning" efforts. However, I want to reemphasize that our deployment strategy provides training using the "train-the-trainer" concept. We train at each site upon initial deployment and thereafter with each system upgrade. The "train-the-trainer" concept facilitates "on-the-job" training, which we believe is the most critical aspect of our training strategy. Finally, we have developed a Web Site on the INS Intranet that provides training Information-nation on IDENT.

The problems you have identified with data entry into the IDENT lookout database have also frustrated us. When we established this database, our source documents were old ten-print cards. The quality of these source documents was poor, both in their physical condition and in the information provided. However, at the time, we felt that including what data we had was more important than not including it because of missing photographs or biographic data. Future data collected from IDENT bookings will include photographs, where they are available. As discussed in our comments to recommendations 4 and 5, we have installed 49 of 52 required live-scan, ten-print devices at facilities that take ten prints for inclusion in the IDENT lookout database. We installed the live-scan devices at our detention facilities and other INS sites that do volume bookings that require ten prints as part of the booking process or because the site acts as a transit point for collecting persons bussed somewhere else. These live-scan, ten-print devices include editing software to both make sure that data entered meet minimum edit standards and to make certain fields mandatory. This should help improve data integrity within the IDENT lookout database. In addition, your recommendations to develop better criteria and procedures for placing fingerprints in the IDENT lookout database are being addressed, and we hope to have them completed and to the field by June 1998. Our plan for implementing the procedures will include monitoring compliance.

We also want to reduce the number of times the same individual appears as a separate record in IDENT. Our response to recommendation 6 provides you with our strategy for consolidating multiple records for the same individual. We agree that multiple records for the same individual is not optimum, and the attached strategy will help address this situation. However, IDENT provides users will all possible records, which probably includes the multiple records for individuals where they exist.

Although your report did not mention this issue, we plan to improve our process for evaluating systems once they are in the field. Our Office of Information and Resources Management staff recently shared with the Office of Inspector General staff a post-implementation review process, proposed to evaluate the effectiveness of system deployment after a new system or significant system revision has been operational for at least 6 months. The review provides insight in areas such as lessons learned, opportunities for improvement, and issues that require special attention.

We appreciate your providing us with the opportunity to respond to the draft report. If you have any questions, please contact Kathleen Stanley, Audit Liaison, at (202) 514-8800.

Attachments

cc: Vickie L. Sloan, DOJ Audit Liaison

 

 


 

Memorandum

Subject   DRAFT INSPECTION REPORT
Review of the INS' Automated Biometric
Identification System (IDENT), #A-97-04
Date    MAR  3 1998
To  Doris Meissner
      Commissioner
From   Office of Programs


RECOMMENDATION 1:
Amend the existing INS Servicewide Fingerprint Policy to establish explicit procedures for criteria for placing individuals in the IDENT lookout database.

INS RESPONSE: Concur. The Office of Programs will convene the INS Fingerprint Working Group, including all major program users of the IDENT system, to develop uniform criteria for placing individuals in the IDENT database. The Office of Programs will prepare an amendment to the existing INS Servicewide Fingerprint Policy to include these criteria by June 30, 1998. The Office of Field Operations will be responsible for communicating the criteria to INS field organizations and coordinating any training activities that may be required.

RECOMMENDATION 2: Develop and implement a plan to ensure that the fingerprints of criminals and deported aliens are consistently entered into the IDENT lookout database.

INS RESPONSE: Concur. A plan will be developed and implemented to ensure that the fingerprints of criminals and deported aliens are consistently entered into the IDENT database. The Office of Programs will facilitate the development of this Ian with the Office of Information Resources Management and the Office of Field Operations. The Office of Field Operations will be responsible for communicating the procedures to the INS field units and monitoring compliance. The development of the plan will be completed by June 30, 1998.

RECOMMENDATION 3: Develop and implement a plan to ensure that all IDENT lookout records contain aliens' photographs to assist arresting agents in identifying them.

INS RESPONSE: Concur. The Office of Programs will work with the Office of Field Operations to develop a plan to ensure that all IDENT records contain aliens' photographs to assist arresting agents. The Office of Field Operations will be responsible for implementing the plan through field units, and for monitoring compliance. The development of the plan will be completed by June 30, 1998.

If you have any questions, please contact Ms. Janet Keys, Director, Office of Strategic Information and Technology Development, Office of Programs, at 202-514-8223.

Michael D. Cronin
Acting Associate Commissioner

 

 


 

U.S. Department of Justice
Immigration and Naturalization Service

Office of the Executive Associate Commissioner

425 I Street NW
Washington, DC 20536

MAR  6 1998

MEMORANDUM FOR  KATHLEEN M. STANLEY
                                        ASSISTANT DIRECTOR
                                        INTERNAL REVIEW BRANCH

FROM:     Michael A. Pearson
                 Executive Associate Commissioner
                 Field Operations

SUBJECT:  Draft Inspection Report: Review of IDENT, #A-97-04

In accordance with your request, following is the Field Operations response to recommendation 8 in the draft report by the Office of the Inspector General on its review of IDENT.

RECOMMENDATION 8: Direct that Chief Border Patrol Agents in each sector work with the U.S. Attorneys to develop and implement a plan to use IDENT to identify and refer the most egregious offenders for prosecution.

INS RESPONSE: CONCUR. Through the Priorities Management System, the INS directs the Chief Patrol Agents to maximize the use of IDENT and identify and remove criminal aliens. To the extent possible, all sectors coordinate with their local U.S. Attorneys to accept the maximum number of prosecutions, including those who previously could not be identified but for whom a complete record of illegal entry attempts is now available through IDENT. The Regional Directors also direct the field offices to use IDENT to identify and refer offenders for prosecution under local guidelines made jointly with the Office of General Counsel and the U.S. Attorneys. As IDENT is implemented at more sites the value of expanding the number of prosecutions to accept IDENT recidivists and potential smugglers will again depend on the local negotiations. All criminal aliens identified through IDENT are already considered or referred for prosecution.

If you have any questions, please contact Karen Hess, Headquarters Border Patrol, at 202-514-2698.

 


 

Memorandum

Subject   DRAFT INSPECTION REPORT
Review of the INS' Automated Biometric
Identification System (IDENT), #A-97-04
Date    FEB  13 1998
To  Doris Meissner
      Commissioner
From   Office of Management

 

The following are comments provided by the Office of Management in response to the recommendation made in the Draft IG Report #A-97-04.

Recommendation 4: Implement input controls for critical data entered into IDENT lookout database to ensure the accuracy of that data

INS Response: Concur

We have installed over 50 ten print live scan devices, which have a build in data capture screen that validates the incoming data, making selected fields mandatory, thus ensuring that errors and omissions are caught.

Recommendation 5: Identify, and to the extent possible, correct invalid data in the IDENT lookout database

INS Response: Concur in Part

We have researched this problem by reviewing the original source documents and have concluded that all of the data that was available has been accurately and completely entered. Where data was missing from the source document, we have concluded that there is no way to go back and collect it, (e.g. person who was booked and ten printed is now gone).

To address this in the future, we have installed over 50 ten print live scan devices, which have a build in data capture screen that validates the incoming data, making selected fields mandatory, thus ensuring that errors and omissions are caught.

Recommendation 6: Develop and implement a strategy for consolidating multiple FINS that exist for individual aliens

INS Response: Concur

Current process works in three parts. Part one involves the booking agent who is asked to perform a manual verification when a close match occurs, the second part involves an automated consolidation of records when the score is sufficiently high enough to permit a safe consolidation, and the third part involves possible matches that have a score too low for the computer to make an automated match and were not matched by the agent. In part three, the possible matches are reviewed by finger print classification experts who review the records and perform a consolidation of those that are the same person. We will further prepare guidance to the field instructing them to make review of close matches part of their mandatory booking process, thus ensuring that the transaction is closed along with the booking. For the future, we are looking at facial recognition as a backup, which will allow for more automated consolidation.

Recommendation 7: Develop and implement a strategy for sufficiently training all INS personnel using IDENT (e.g., making IDENT a part of an INS data systems training module at the INS training academies)

INS Response: Concur in Part

We concur that more training is needed; however, we have previously looked at adding system specific training to the curriculum at the academies and found the adding of more days or more training within existing days to be problematic. The courses are already very long and the feasibility of adding more does not look practical. The solution that we are pursuing is to integrate the systems into the course work so that the trainees become familiar with the systems as they learn their future job. We plan to make the systems available for use during practice exercises. We are addressing follow-on training once the employee/officer has been on the job (e.g., through our distance learning program/efforts). We are also attempting to build systems that require minimum training and systems that utilize automated edits and validation logic to help shift more of the burden to the computer software to reduce the system specific training need.

For further information, please do not hesitate to call Ronald Collison, Associate Commissioner, Office of Information Resources Management, on 514-2547.

George H. Bohlinger,
Executive Associate Commissioner

 

 


 

Appendix IV


OFFICE OF THE INSPECTOR GENERAL'S ANALYSIS
OF MANAGEMENT'S RESPONSE

On January 16, 1998, the Inspections Division sent copies of the draft report to INS with a request for written comments. INS responded by memorandum from the Commissioner dated March 23, 1998 (APPENDIX III).

INS concurred or concurred in part with all eight recommendations and proposed corrective actions for each of them. However, many of these actions lack the clarity and level of detail necessary for a full assessment of their effectiveness. In addition, several of the proposals fail to address the full scope of the problems the recommendations raise. We must receive more explicit, detailed information to fully assess whether the actions proposed, when implemented, will correct the problems raised in the report and reflected in the recommendations. We will unresolve any recommendation for which we are unable to make such an assessment.

Recommendation Number:

1. Resolved-Open. INS concurred with our recommendation to amend the existing INS Fingerprint Policy to establish explicit procedures and criteria for placing individuals in the IDENT lookout database. INS' response is that the Office of Programs will prepare an amendment to the existing INS Servicewide Fingerprint Policy to include these procedures and criteria by June 30, 1998. Office of Field Operations will be responsible for communicating the procedures and criteria to the field and coordinating any training required.

Please provide us by July 15, 1998, the procedures and criteria for placing individuals in the IDENT lookout database. Please also provide us a copy of the memorandum transmitting this information to the field, including references to any required training activities.

2. Resolved-Open. INS concurred with our recommendation to develop and implement a plan to ensure that the fingerprints of criminals and deported aliens are consistently entered into the IDENT lookout database. INS' response is that the Office of Programs, Office of Information Resources Management and Office of Field Operations will develop a plan by June 30, 1998. The Office of Field Operations will be responsible for communicating the procedures to the field and monitoring compliance.

Please provide us by July 15, 1998, the plan that will ensure the entry of fingerprints of criminals and deported aliens into the IDENT lookout database. Please also provide us a copy of the memorandum transmitting the procedures to the field and describe the means by which compliance will be monitored.

3. Resolved-Open. INS concurred with our recommendation to develop and implement a plan to ensure that all IDENT lookout records contain aliens' photographs to assist arresting INS agents in identifying them. INS' response is that the Office of Programs and Office of Field Operations will develop a plan by June 30, 1998. The Office of Field Operations will be responsible for implementing the plan through field units and for monitoring compliance.

Please provide us by July 15, 1998, the plan that ensures that all IDENT lookout records contain aliens' photographs, whether the records are electronically submitted or received by mail. Please also provide us with evidence that the plan has been implemented in the field and tell us the means by which compliance will be monitored.

4. Resolved-Open. INS concurred with our recommendation to implement controls for critical data entered into the IDENT lookout database to ensure the accuracy of that data. INS' response is that over 50 ten-print live scan devices have been installed, which will automatically validate incoming data, making selected fields mandatory, thus ensuring that errors and omissions are caught.

This response only partially addresses the data integrity issues raised in the report. It does address the data integrity of entries made directly to the lookout database through use of the 10-print live scan devices. It fails to address controls necessary to ensure the accuracy of the data input from the fingerprint cards that are electronically transmitted or received by mail and then manually entered by INS staff. Finally, the response does not identify the mandatory fields or edit checks being used to ensure data accuracy.

Please provide us by July 15, 1998, the details of how data integrity will be ensured for fingerprint cards whose data will not be directly transmitted to the lookout database by the 10-print live scan devices. Also please identify the mandatory fields and edit checks used to ensure data accuracy for fingerprint cards generated using the 10-print live scan devices.

5. Resolved-Open. INS concurred in part with our recommendation to identify, and to the extent possible, correct invalid data in the IDENT lookout database. INS' response is that it determined the scope of the problem by reviewing original source documents. INS personnel concluded that all available data has been accurately and completely entered. Where data was missing from the source document, INS concluded there was no way to go back and collect it because the person who was booked and fingerprinted is no longer in custody.

The response refers to a review of original source documents but it is not clear how this review was done. INS then reiterates that the installation of 10-print live scan devices and their associated data validity checks will be the primary means by which errors and omissions will be caught in the future.

Please provide us by July 15, 1998, details of how the review of original source documents was performed, and the specific results of that review.

6. Resolved-Open. INS concurred with our recommendation to develop and implement a strategy for consolidating multiple FINs that exist for individual aliens. The INS response is that the current process works in three different ways: 1) an agent performs a manual verification when a close match occurs; 2) the system automatically consolidates records when the match score is sufficiently high to permit a safe consolidation; and 3) a fingerprint specialist reviews possible matches that have scores too low for the computer to make an automated match and where these possible matches were not confirmed by the agent. The fingerprint specialist consolidates those records that he determines come from the same person.

The INS response states that guidance will be prepared, instructing INS personnel to include the manual confirmation of close matches as a mandatory part of the booking process. While, if followed, these procedures could help ensure that each of the multiple FINs generated for an individual is confirmed during the booking process, the INS response does not explain how such confirmations will be used by the system to consolidate multiple FINs that are identified. The response also does not address how multiple FINs will be consolidated when confirmation of close matches cannot be completed by the agents due to technical or communications problems.

Please provide by July 15, 1998, a more detailed explanation of how multiple FINs consolidation will be accomplished. This should include procedures to address those instances in which match confirmations cannot be completed due to technical or communications problems. Please also provide us a copy of the guidance, instructing INS personnel to include the manual confirmation of close matches as a mandatory part of the booking process.

7. Resolved-Open. INS concurred in part with our recommendation to develop and implement a strategy for sufficiently training all INS personnel using IDENT (e.g., making IDENT a part of an INS data systems training module at the INS training academies).

INS' response agrees more training is needed but states adding separate system specific training to the curriculum at the academies is problematic. INS proposes instead to integrate systems training into course work so that trainees become familiar with the IDENT system as they learn their job. INS plans to make systems available during practice exercises. Follow-on training once the employee is on the job will be addressed through the distance learning programs, e.g., a Web Site on the INS Intranet providing training information on IDENT or perhaps the use of a CD-ROM training disk.

INS' approach proposed for IDENT training seems reasonable but is incomplete. The response does not address how INS will evaluate the effectiveness of IDENT training provided for INS agents in the field. The "train the trainer" concept used up to now, has had limited success because not all the agents were trained, and those that were never had to demonstrate that they had mastered the basic skills. A method needs to be identified to ensure that INS officers successfully learn how to properly use IDENT.

Please provide us by July 15, 1998, a more specific post-Academy IDENT-related training plan. The plan should include how INS will evaluate the effectiveness of IDENT training provided for INS agents in the field.

8. Resolved-Open. INS concurred with our recommendation to direct the Chief Border Patrol Agent in each sector to work with the United States Attorneys to develop and implement a plan to use IDENT to identify and refer the most egregious offenders for prosecution. The INS' response says that the Border Patrol Chiefs are already directed to maximize use of IDENT to identify and remove criminal aliens. It states that "to the extent possible" all sectors coordinate with their local United States Attorneys to accept the maximum number of prosecutions including those who previously could not be identified, but for whom a complete record of illegal entry attempts is now available through IDENT. The Regional Directors also direct the field offices to use IDENT to identify and refer offenders for prosecution under "local guidelines" made jointly with the Office of General Counsel and the United States Attorneys. The response points out that as IDENT is implemented at more sites, the value of expanding the number of prosecutions to accept IDENT recidivists and potential smugglers will again "depend on the local negotiations." The response states that all criminal aliens identified through IDENT are already considered or referred for prosecution.

While INS concurs with our recommendation, the response highlights instructions or agreements already in place that have not led to a systematic use of IDENT-supported prosecutions along the border. INS needs to initiate systematic discussions with the United States Attorneys that will lead to strategies that make more effective use of IDENT to prosecute border crimes. It would be in the context of such strategies that local guidelines should be applied and negotiations occur.

Please inform us by July 15, 1998, of the actions taken to complete this recommendation.

#####